Impacts of Lining Materials on Water Quality

Transcription

1 Impacts of Lining Materials on Water Quality Subject Area: Infrastructure

2

3 Impacts of Lining Materials on Water Quality

4 About the Water Research Foundation The Water Research Foundation (formerly Awwa Research Foundation or AwwaRF) is a member-supported, international, 501(c)3 nonprofit organization that sponsors research to enable water utilities, public health agencies, and other professionals to provide safe and affordable drinking water to consumers. The Foundation s mission is to advance the science of water to improve the quality of life. To achieve this mission, the Foundation sponsors studies on all aspects of drinking water, including resources, treatment, distribution, and health effects. Funding for research is provided primarily by subscription payments from close to 1,000 water utilities, consulting firms, and manufacturers in North America and abroad. Additional funding comes from collaborative partnerships with other national and international organizations and the U.S. federal government, allowing for resources to be leveraged, expertise to be shared, and broad-based knowledge to be developed and disseminated. From its headquarters in Denver, Colorado, the Foundation s staff directs and supports the efforts of more than 800 volunteers who serve on the board of trustees and various committees. These volunteers represent many facets of the water industry, and contribute their expertise to select and monitor research studies that benefit the entire drinking water community. The results of research are disseminated through a number of channels, including reports, the Web site, Webcasts, conferences, and periodicals. For its subscribers, the Foundation serves as a cooperative program in which water suppliers unite to pool their resources. By applying Foundation research findings, these water suppliers can save substantial costs and stay on the leading edge of drinking water science and technology. Since its inception, the Foundation has supplied the water community with more than $460 million in applied research value. More information about the Foundation and how to become a subscriber is available on the Web at

5 Impacts of Lining Materials on Water Quality Prepared by: Arun Deb, Sandra B. McCammon, and Jerry Snyder Weston Solutions, Inc Weston Way, West Chester, PA and Andrea Dietrich Virginia Polytechnic Institute, Department of Civil Engineering 413 Durham Hall, Blacksburg, VA Sponsored by: Water Research Foundation 6666 West Quincy Avenue, Denver, CO U.S. Environmental Protection Agency Washington, D.C. and Drinking Water Inspectorate 55 Whitehall, London SW1A 2EY, England Published by:

6 DISCLAIMER This study was jointly funded by the Water Research Foundation (Foundation) and the U.S. Environmental Protection Agency (USEPA) under Cooperative Agreement No. X , and by the Drinking Water Inspectorate (DWI). The Foundation, USEPA, and DWI assume no responsibility for the content of the research study reported in this publication or for the opinions or statements of fact expressed in the report. The mention of trade names for commercial products does not represent or imply the approval or endorsement of the Foundation, USEPA, or DWI. This report is presented solely for informational purposes. Copyright 2010 by Water Research Foundation and Drinking Water Inspectorate ALL RIGHTS RESERVED. No part of this publication may be copied, reproduced or otherwise utilized without permission. ISBN Printed in the U.S.A.

7 CONTENTS LIST OF TABLES... xi LIST OF FIGURES... xiii FOREWORD... xxi ACKNOWLEDGMENTS... xxiii EXECUTIVE SUMMARY...xxv CHAPTER 1 INTRODUCTION...1 Background and Introduction...1 Project Objectives...2 Project Approach...2 Organization of the Report...3 CHAPTER 2 LITERATURE REVIEW...5 Introduction...5 Cement-Mortar Lining...5 Impact on Water Quality...6 Microbial Growth and Bio-film Formation...7 Disinfectant Decay...7 Utility Survey...8 Field Testing...8 Epoxy-Resin Lining...9 Evolution of Epoxy-Resin Lining...10 Quality Assurance and Quality Control (QA/QC)...10 Utility Experience...11 Water Quality Impacts...11 Polyurethane Lining...13 Evolution of Polyurethane Lining...13 Quality Assurance and Quality Control (QA/QC)...14 Utility Case Studies of Polyurethane Linings...14 Impact on Water Quality...15 CHAPTER 3 SURVEY OF CURRENT PRACTICES AND WATER QUALITY...17 Questionnaire...17 Questionnaire Responses and Results...17 CHAPTER 4 OVERALL EXPERIMENTAL DESIGN OF BENCH-SCALE LEACHING STUDIES...21 Objectives...21 Cement-Mortar Coupon Fabrication...21 v

8 vi Impacts of Lining Materials on Water Quality Cement-Mortar Composition...22 Mold Design...22 Cement Coupon Production Process...23 Polyurethane Coupon Preparation...23 Epoxy Coupon Preparation...23 Mature Pipe Sample...24 Pretreatment of Test Coupons...24 Test Vessels Leaching of Lining Materials...24 Test Waters for Leaching of Lining Materials...25 Protocol for Water Changes and Analyses...26 Statistical Analysis...29 CHAPTER 5 TESTING OF CEMENT-MORTAR LINING...31 Test Results...31 Effect of Cement-Mortar Lining on Inorganic Water Quality Parameters...31 ph...31 Alkalinity...32 Hardness...32 Elemental Analysis...33 Total Solids...36 Disinfectant Residual...37 Ammonia...40 Total Organic Carbon (TOC)...40 Semivolatile Organic Compounds (SVOCs)...42 Trihalomethanes (THMs)...42 THM Formation Potential and Sorption to Cement-Mortar Lining...42 Summary of Impacts to Water Quality from New Cement-Mortar Lining...48 Major Impacts...48 Minor Impacts...48 Testing of Cement-Mortar Lining with Corrosion Prevention Additive (CPA) Waters...49 Objective...49 Test Results...49 Summary of Test Results with CPA...57 CHAPTER 6 TESTING OF POLYURETHANE LINING...59 Results...59 Effect of Polyurethane on Inorganic Water Quality Parameters...59 ph...59 Total Alkalinity...59 Hardness...59 Elemental Analyses...61 Total and Dissolved Solids...62 Disinfectant Residual...62 Ammonia...62 Effect of Polyurethane on Organic Water Quality Parameters...65 Total Organic Carbon (TOC)...65

9 Contents vii Trihalomethanes (THMs)...67 Haloacetic Acids (HAA5)...67 Semivolatile Organic Compounds (SVOCs)...69 Odor...70 Physical Changes to Polyurethane Coupons...72 Summary of Impacts to Water Quality from Polyurethane...73 CHAPTER 7 TESTING OF EPOXY LINING...75 Results for Epoxy Lining...75 Effect of Epoxy on Inorganic Water Quality Parameters...75 ph...75 Alkalinity...75 Hardness...78 Elemental Analyses...79 Total and Dissolved Solids...79 Disinfectant Residual...79 Ammonia...82 Organic Water Quality Parameters...83 Total Organic Carbon (TOC)...83 Trihalomethanes (THMs)...84 THM Formation/Sorption...87 Haloacetic Acids (HAA5)...88 Semivolatile Organic Compounds (SVOCs)...90 Odor...91 Microbiological Analysis of Mature Epoxy Lining...94 Summary of Impacts to Water Quality from New Epoxy and Mature Epoxy...94 New Epoxy...94 Mature Epoxy...95 CHAPTER 8 EVALUATION OF WATER QUALITY IMPACTS...97 Introduction...97 Lining Materials Evaluated...98 Cement Mortar...98 Epoxy...99 Polyurethane...99 Construction Parameters...99 Cement Mortar...99 Epoxy...99 Polyurethane Background Water Quality Water Quality Impacts Cement-Mortar Lining Epoxy Lining Polyurethane (PU) Lining Health Impacts Cement Mortar...106

10 viii Impacts of Lining Materials on Water Quality Epoxy Polyurethane Regulatory Impacts Cement Mortar Epoxy Polyurethane (PU) Potential Regulatory Measures CHAPTER 9 DEVELOPMENT OF A METHODOLOGY FOR SELECTION OF A LINING MATERIAL Introduction Methodology for Analysis of Water Quality Impacts Development of Methodology for Mass-Based Parameters Development of Methodology for ph Estimation of Useful Life of Liners Cement Mortar Epoxy Polyurethane (PU) Economic Analysis of Liners Summary CHAPTER 10 TEST CASE ANALYSIS Case Study South Central Connecticut Regional Water Authority Woodward Avenue Renewal Impact from Cement Mortar Impact from Epoxy Lining Impact from Polyurethane Lining Cost Impacts Sensitivity Analysis Conclusions CHAPTER 11 RESEARCH GAPS Cement-Mortar Linings Polyurethane Lining Epoxy Lining General CHAPTER 12 FINDINGS AND RECOMMENDATIONS Findings Recommendations APPENDIX A: WATER UTILITY QUESTIONNAIRE APPENDIX B: METHODS OF CHEMICAL MEASUREMENTS APPENDIX C: MASS RATE RELEASE CURVES...153

11 Contents ix APPENDIX D: USER S MANUAL REFERENCES ABBREVIATIONS...179

12

13 TABLES 3.1 Lining material summary Reported impacts from lining materials Cement-mortar recipe Reference water composition Test water recipes for CPA impact testing Typical frequency and quality control for chemical water quality parameters for benchscale testing of lining materials Measurement method and sampling procedures Calcium and magnesium concentrations (as CaCO 3 ) for cement mortar in contact with no disinfectant water and data for the corresponding control water Calcium and magnesium concentrations (as CaCO 3 ) for cement mortar in contact with ph 6.5, chlorine water and data for the corresponding control water Calcium and magnesium concentrations (as CaCO 3 ) for cement mortar in contact with ph 8, chlorine water and data for the corresponding control water Calcium and magnesium concentrations (as CaCO 3 ) for cement mortar in contact with ph 8, chloraminated water and data for the corresponding control water Total solids in water in contact with cement-mortar coupons THM formation test matrix THM sorption test matrix Summary of CPA testing results Sorption of THMs from aqueous solution and into polyurethane after a 72-hour contact time; initial THM concentration was 100 µg/l Summary of organic chemicals detected in leachate water with polyurethane coupons but not detected in the control water Properties of organic compound detected in leachate from polyurethane...70 xi

14 xii Impacts of Lining Materials on Water Quality 7.1 Results of microbiological analysis of 2 cm 2 of mature epoxy pipe surface Composition of four test waters used in leaching tests Knowledge matrix leaching from lining materials and its effect on water quality Highest concentrations measured during tests of cement mortar Constituents monitored during laboratory leaching tests Drinking water quality limits for constituents of interest Required removal of TOC from source water Summary of regression equations for mass loss curves Summary of impact module predictions for Woodward Avenue test case Sensitivity of target ph and flushing volume A.1a Inventory of lining projects (Part 1) A.1b Inventory of lining projects (Part 2) A.2a Water quality measurements made before lining A.2b Water quality measurements made after lining...148

15 FIGURES 4.1 Photo of cement-mortar coupon mold Immersion test apparatus with coupons Corrosion preventative experimental testing setup ph as a function of time Alkalinity as a function of time for all four test waters and the controls in contact with cement mortar Calcium and magnesium hardness as a function of time for ph 8, 2 mg/l Cl 2 water (water 3) and its control in contact with cement mortar Aluminum and chromium as a function of time for ph 8, 2 mg/l C1 2 water (water 3) and its control in contact with cement mortar Test vessel containing cement-mortar coupons and precipitate that was typical of all test conditions up through day Chlorine decay, showing residual disinfectant and consumed disinfectant as a function of time, corrected for decay in the controls (water: ph 6.5, 2 mg/l Cl 2 ) in contact with cement mortar Chlorine decay, showing residual disinfectant and consumed disinfectant as a function of time, corrected for decay in the controls (water: ph 8, 2 mg/l Cl 2 ) in contact with cement mortar Monochloramine decay, showing residual disinfectant and consumed disinfectant as a function of time, corrected for decay in the controls (water: ph 8, 4 to 5 mg/l NH 2 Cl) in contact with cement mortar Disinfectant decay rate based on contact time of cement mortar with disinfectant Ammonia as a function of time for ph 8 water dosed with monochloramine in contact with cement mortar TOC as a function of time for water type in contact with cement mortar TOC leached per unit time and surface area when in contact with cement mortar Chloroform concentration and chlorine residual as a function of time after chlorination of sample waters in contact with cement mortar...44 xiii

16 xiv Impacts of Lining Materials on Water Quality 5.14 Loss in chloroform concentration as a function of time after exposure to cement-mortar coupon segments Concentrations of HAA5 present in water in contact with cement mortar (top) and production of HAA5 as a function of contact time (bottom) Cement odor intensity as a function of time and water type in contact with cement mortar ph as a function of exposure time for three CPA types and control water in contact with cement mortar Alkalinity as a function of exposure time for three CPA types and control water in contact with cement mortar Total hardness as a function of exposure time for three CPA types and control water in contact with cement mortar Calcium concentrations as a function of exposure time for three CPA types and control water in contact with cement mortar Magnesium concentrations as a function of exposure time for three CPA types and control water in contact with cement mortar Aluminum concentrations as a function of exposure time for three CPA types and control water in contact with cement mortar Phosphorous concentrations as a function of exposure time for three CPA types and control water in contact with cement mortar Zinc concentrations as a function of exposure time for three CPA types and control water in contact with cement mortar Chromium concentrations as a function of exposure time for three CPA types and control water in contact with cement mortar ph as a function of time as measured by test day for three test waters in contact with polyurethane and water change Alkalinity as a function of time for the three test waters in contact with polyurethane and corresponding control waters without polyurethane Concentrations of selected elements in no disinfectant water in contact with polyurethane coupons...61

17 Figures xv 6.4 Concentrations of selected elements in 2 mg/l Cl 2 water in contact with polyurethane coupons Concentrations of selected elements in 5 to 6 mg/l NH 2 Cl water in contact with polyurethane coupons Chlorine decay, showing residual disinfectant, and consumed disinfectant, as a function of time Monochloramine decay, showing residual disinfectant and consumed disinfectant, as a function of time Rate of chlorine and monochloramine decay; data were corrected for loss of residual in the controls (mg/l/day) Total organic carbon (TOC, mg/l) as a function of time for all three waters in contact with polyurethane with controls (error bars represent standard deviation) Rate of TOC leachate (TOC, mg/cm 2 /day) as a function of time for all three waters in contact with polyurethane and corrected for TOC measured in controls HAA5 concentrations as a function of exposure time for all three water types (corrected for controls) in contact with polyurethane HAA5 rate of leaching over time, as μg/cm3/day, as a function of time for all three waters in contact with polyurethane Odor intensity, divided among the categorized odors, as a function of time for no disinfectant water exposed to polyurethane Odor intensity, divided among the categorized odors, as a function of time for chlorinated water exposed to polyurethane Odor intensity, divided among the categorized odors, as a function of time for chloraminated water exposed to polyurethane Polyurethane coupons after extended exposure to: (1) no disinfectant, (2) monochloramine, and (3) chlorine ph as a function of time in all three waters and their controls in contact with epoxy (error bars indicate standard deviations) ph as a function of time in chlorinated matured epoxy-lined pipe and chlorinated control...76

18 xvi Impacts of Lining Materials on Water Quality 7.3 Alkalinity (mg/l CaCO 3 ) as a function of time for water types exposed to epoxy and their controls (error bars indicate standard deviations) Alkalinity (mg/l CaCO 3 ) as a function of time in mature epoxy-lined pipe sample exposed to chlorinated water and corresponding chlorinated control Calcium and magnesium hardness (mg/l as CaCO 3 ) concentrations for test waters exposed to epoxy as a function of time Calcium and magnesium hardness (mg/l as CaCO 3 ) concentrations for control waters without epoxy as a function of time Chlorine decay, showing residual disinfectant and consumed disinfectant as a function of time, corrected for decay in the controls (water: ph 8, 2 mg/l Cl 2 ) Monochloramine decay, showing residual disinfectant and consumed disinfectant as a function of time, corrected for decay in the controls (water: ph 8, 4.5 to 5.8 mg/l NH 2 Cl) Chlorine and monochloramine residuals (mg/l) with controls as a function of time Chlorine and monochloramine consumption rate (mg/l/day), corrected for decay in the controls, as a function of time Chlorine decay, showing residual disinfectant and consumed disinfectant as a function of time, corrected for natural decay in the controls for matured epoxy-lined sample (water: ph 8, 2 mg/l Cl 2 ) Ammonia (mg/l as N) in chloraminated sample and control waters (4.5 to 5.8 mg/l NH 2 Cl) as a function of time (error bars indicate standard deviations) TOC (mg/l) as a function of time for all three water types, corrected for controls (error bars indicate standard deviations) TOC (mg/cm 2 ) as a function of time for all three water types and mature pipe sample, corrected for controls (error bars indicate standard deviations) Rate of TOC formation (mg/cm 2 /day) in all three water treatments as a function of time, corrected for controls (error bars indicate standard deviations) CHCl3 concentration (µg/l) as a function of time in chlorinated test waters, corrected for controls CHCl3 concentration (µg/cm 2 ) as a function of time in chlorinated test waters, corrected for controls...86

19 Figures xvii 7.18 CHCl3 formation rate (µg/cm 2 /day) in chlorinated test waters, corrected for controls Concentrations of disinfectant residual (mg/l), TOC (mg/l), and CHCl 3 (µg/l) in THM formation study as a function of time HAA5 concentrations (µg/l) corrected for controls as a function of time for all three water types HAA5 concentrations (µg/cm 2 ) corrected for controls as a function of time for all three water types and mature pipe sample HAA5 rate of formation (µg/cm 2 /day) corrected for controls as a function of time for no disinfectant, chlorinated, and chloraminated test waters Bis-phenol A concentration (µg/l) in all three waters in contact with epoxy as a function of time, corrected for controls Odor intensity, divided between categorized odors, as a function of time (water: ph 8, no disinfectant) (error bars indicate standard deviations) Odor intensity, divided between categorized odors, as a function of time (water: ph 8, 2 mg/l Cl 2 ) (error bars indicate standard deviations) Odor intensity, divided between categorized odors, as a function of time (water: ph 8, 4.5 to 5.8 mg/l NH 2 Cl) Schematic diagram for evaluation of water quality impact of lining Calcium concentrations in water in contact with cement mortar Rate of alkalinity increase versus age of cement-mortar lining Alkalinity mass loss rate versus age of cement-mortar lining from laboratory tests Comparison of alkalinity mass loss rates between laboratory and field tests Alkalinity increases in water versus age of cement-mortar 6,000 gpd Effect of pipe diameter on increased alkalinity under uniform 1-day hydraulic residence time Impact of flushing on alkalinity concentration on a 6-in. pipe Variation of ph with time for daily and weekly flushing frequencies ph increase versus age of cement mortar for four HRTs...121

20 xviii Impacts of Lining Materials on Water Quality 9.9 ph increase versus age of cement mortar for three HRTs ph increase versus HRT for six ages of cement mortar Measured ph before and after cleaning and lining water main on Woodward Avenue Woodward Avenue ph after cleaning and lining Alkalinity increase in drinking water versus age of cement mortar for the Woodward Avenue example Calcium increase in drinking water versus age of cement mortar for the Woodward Avenue example Aluminum increase in drinking water versus age of cement mortar for the Woodward Avenue example TDS increase in drinking water versus age of cement mortar for the Woodward Avenue example TOC increase in drinking water versus age of epoxy for the Woodward Avenue example Disinfectant demand in drinking water versus age of epoxy for the Woodward Avenue example Disinfectant demand in drinking water versus age of polyurethane for the Woodward Avenue example HAA5 increase in drinking water versus age of polyurethane for the Woodward Avenue example C.1 Impact of cement mortar on calcium concentration rate of increase over time C.2 Calcium mass loss rate versus time C.3 Calcium mass loss rate versus time C.4 Aluminum concentration increase over time for cement mortar C.5 Aluminum mass loss rate µg/cm 2 /day with age C.6 Total dissolved solids (TDS) concentration increase over time for cement mortar C.7 TDS mass release rate from cement mortar with time...157

21 Figures xix C.8 TOC concentration rate of change for epoxy material C.9 TOC mass release rate from epoxy with time C.10 Disinfectant demand concentration rate over time for epoxy material C.11 Disinfectant demand mass loss rate for epoxy with time C.12 Disinfectant demand concentration rate over time for water in contact with polyurethane C.13 Disinfectant demand mass rate over time for water in contact with polyurethane C.14 HAA5 concentration increase rate over time for water in contact with polyurethane C.15 HAA5 mass formation rate for water in contact with polyurethane D.1 Data input screen for the Impact Module D.2 Database of mass release rates D.3 Impact Calculations worksheet ph calculation D.4 Impact Calculations worksheet alkalinity calculations D.5 Impact Calculations worksheet calcium calculations D.6 Calculated alkalinity concentration increase due to cement mortar D.7 Calculated calcium concentration increase due to cement mortar D.8 Additional volume of water reduced calcium concentration D.9 Select Goal Seek under the Tools Menu D.10 Goal Seek data entry window D.11 Goal Seek status window D.12 Result after Goal Seek D.13 Flushing Cost Module input...174

22

23 FOREWORD The Water Research Foundation (Foundation) is a nonprofit corporation that is dedicated to the implementation of a research effort to help utilities respond to regulatory requirements and traditional high-priority concerns of the industry. The research agenda is developed through a process of consultation with subscribers and drinking water professionals. Under the umbrella of a Strategic Research Plan, the Research Advisory Council prioritizes the suggested projects based upon current and future needs, applicability, and past work; the recommendations are forwarded to the Board of Trustees for final selection. The Foundation also sponsors research projects through the unsolicited proposal process; the Collaborative Research, Research Applications, and Tailored Collaboration programs; and various joint research efforts with organizations such as the U.S. Environmental Protection Agency, the U.S. Bureau of Reclamation, and the Association of California Water Agencies. This publication is a result of one of these sponsored studies, and it is hoped that its findings will be applied in communities throughout the world. The following report serves not only as a means of communicating the results of the water industry's centralized research program but also as a tool to enlist the further support of the nonmember utilities and individuals. Projects are managed closely from their inception to the final report by the Foundation's staff and large cadre of volunteers who willingly contribute their time and expertise. The Foundation serves a planning and management function and awards contracts to other institutions such as water utilities, universities, and engineering firms. The funding for this research effort comes primarily from the Subscription Program, through which water utilities subscribe to the research program and make an annual payment proportionate to the volume of water they deliver and consultants and manufacturers subscribe based on their annual billings. The program offers a cost-effective and fair method for funding research in the public interest. A broad spectrum of water supply issues is addressed by the Foundation's research agenda: resources, treatment and operations, distribution and storage, water quality and analysis, toxicology, economics, and management. The ultimate purpose of the coordinated effort is to assist water suppliers to provide the highest possible quality of water economically and reliably. The true benefits are realized when the results are implemented at the utility level. The Foundation's trustees are pleased to offer this publication as a contribution toward that end. Roy L. Wolfe, Ph.D. Chair, Board of Trustees Water Research Foundation Robert C. Renner, P.E. Executive Director Water Research Foundation xxi

24

25 ACKNOWLEDGMENTS The authors of this report thank the Water Research Foundation (Foundation) for its support of this study, and the Foundation project manager, Dr. Jian Zhang, for his help and guidance in the execution of the project. The authors acknowledge the input provided by the members of the Project Advisory Committee (PAC), namely: U.S. Environmental Protection Agency, Mr. Michael R. Schock Charlotte Mecklenburg Utilities, Mr. Rusty Campbell Drinking Water Inspectorate, U.K. (DWI), Joanna Hunt The participating utilities include: Bucks County Water and Sewer Authority, Warrington, Pa., Mr. John Butler Yorkshire Water, Yorkshire, United Kingdom, Mr. David Brown Heitkamp, Inc., Watertown, Conn., Mr. Leonard Assard Cohesant Materials, Inc., Ms. Joanne Hughes Denver Water, Denver, Colo., Mr. John H. Bambei, Jr. Halifax Regional Water Commission, Halifax, Canada, Mr. Jamie Hannam Madison Chemical Industries, Inc., Ont., Canada, Ms. Marilou Soria City of Cleveland, Cleveland, Ohio, Mr. J. Christopher Nielson South Central Connecticut Regional Water Authority, Conn., Mr. Ted Norris Mr. Ted Norris of South Central Connecticut Regional Water Authority graciously allowed the use of the system for the test case data. The authors would also like to thank the workshop participants for their input. xxiii

26

27 EXECUTIVE SUMMARY BACKGROUND AND INTRODUCTION As water mains become old and reach the end of their useful lives, their performance diminishes gradually, resulting in high maintenance costs, deterioration of water quality, loss of hydraulic capacity, and a significant increase in customer complaints. Cleaning and lining of a poorly performing water main will decrease customer complaints and will improve water quality, flow, and pressure significantly. This will improve the long-term performance of the pipe and improve the reliability of the system. Rehabilitation of water mains using cleaning and lining results in significant cost savings compared to replacement. In North America, most of the water main rehabilitation is conducted using cementmortar lining. Alternatives to cement-mortar lining of water mains are epoxy lining and polyurethane lining. In the United Kingdom (U.K.), epoxy has been used as the preferred lining material in rehabilitation of water pipelines. In the last few years, however, polyurethane has become the most widely used lining material in the U.K. One of the important purposes of pipe lining is to prevent leaching of metals from the wall of water mains and thus, to improve water quality. However, there is concern about potential leaching of metals and organic chemicals from these lining materials. Water quality problems as a result of leaching chemicals from these linings are of great concern in the United States. A few years ago the National Sanitation Foundation (NSF) provided the approval certification (NSF-61) for use of epoxy and polyurethane as viable alternatives to cement-mortar lining for drinking water mains. Water quality impact cost may be significant but, in general, is not considered in selection of lining materials. When comparing the cost of alternative lining materials available for water main rehabilitation, it is necessary to consider the water quality impact cost while estimating the total cost of the lining. Water quality impacts of lining materials may be an important decision criterion in selecting a lining material or rehabilitation technique. In this study a methodology for a decision support system for selection of an appropriate lining by water utilities that considers potential water quality impacts has been developed. OBJECTIVES OF THE RESEARCH The objectives of this study were the following: Gather, compile, and synthesize detailed information on water quality impacts associated with cement mortar, epoxy, and polyurethane water main lining materials and conduct bench-scale testing of water quality impacts of newly laid cement mortar, epoxy, and polyurethane lining materials under various exposure times and water characteristics. Develop a methodology to help water utilities to consider water quality impacts in order to select the most suitable lining rehabilitation method for long-term performance. xxv

28 xxvi Impacts of Lining Materials on Water Quality Identify research needs and knowledge gaps on water quality impacts from lining materials. RESEARCH APPROACH The research approach to accomplish the project objectives and goals consisted of the following tasks: A comprehensive review of literature and available information was conducted to study literature on water quality impacts from cement-mortar, epoxy, and polyurethane lining rehabilitation methods. A group of selected water utilities with water main lining experience were contacted to conduct a questionnaire survey of historical lining rehabilitation projects and their impacts on water quality. Data were compiled and evaluated to identify principal parameters impacting water quality. A bench-scale leaching study was formulated and conducted with cement-mortar, epoxy, and polyurethane lining materials, and water quality impacts were evaluated. In addition, leaching tests with one mature sample of epoxy lining was conducted to evaluate impacts of age of epoxy lining on water quality. Information gathered was evaluated and a methodology was developed to identify water quality impacts due to various lining materials, and to estimate the cost of such impacts for selection of cost-effective lining rehabilitation technologies. The methodology has been applied to one test case utility to demonstrate its suitability and applicability. Knowledge gaps were identified, and findings and recommendations were made. SURVEY OF CURRENT PRACTICES A questionnaire survey of water utilities of North America and the U.K. was conducted. North American water utilities installed epoxy lining in about 100,000 feet (ft) of water mains, whereas, the U.K. utilities constructed about 50,000,000 ft of water mains. Utilities surveyed used cement-mortar, epoxy, and polyurethane linings. The need to improve water quality was found to be the primary reason for lining water mains. Significant improvement in water quality and hydraulic performance was observed by water utilities after installation of liners. All surveyed utilities reported to have a good written quality assurance/quality control (QA/QC) procedure and conduct a closed-circuit television (CCTV) inspection of water mains before and after the lining. DESIGN OF BENCH-SCALE STUDY Bench-scale testing was undertaken that simulated the water volume to surface area ratio of a 4-inch (in.) diameter pipe. Coupons of the lining materials were used in these tests. The bench-scale testing used low alkalinity/low hardness water with either no disinfectant, free chlorine, or chloramines. Four water types were tested for cement-mortar lining and three water types were tested for epoxy and polyurethane linings. Complete water changes occurred on days 1, 2, 4, 9, 11, 14, 15, and 19; the tests continued up to 30 days with occasional water changes

29 Executive Summary xxvii from day 19 to day 30. The contact times between the water and lining materials represented worst-case conditions similar to those in low flow and dead-end water mains. TESTING OF CEMENT-MORTAR LINING The following are the water types used in testing cement-mortar lining: ph = 8.0, low alkalinity/hardness, no disinfectant ph = 6.5, low alkalinity/hardness, 2.0 mg/l chlorine ph = 8.0, low alkalinity/hardness, 2.0 mg/l chlorine ph = 8.0, low alkalinity/hardness, 4.0 mg/l monochloramine The main water quality impact of cement-mortar linings is to increase ph, alkalinity, calcium, aluminum, and total dissolved solids (TDS). The duration of the increases in alkalinity, calcium, aluminum, and TDS lasts from 2 to 4 weeks. The impact on ph is more persistent, lasting more than 30 days. The cement-mortar material created a significant increase in the alkalinity of all four test waters. The impact on alkalinity in general terms was the same for all four test waters increasing it to 600 milligrams per liter (mg/l) as CaCO 3 and then stabilizing at 100 mg/l after day 19 of the test. The background alkalinity of the test waters was 35 mg/l. Cement mortar increased the ph with nearly identical values for all four test waters. The starting ph of test waters did not impact the resulting ph after contact with the cement mortar. The ph after the first day in contact with the cement mortar was After 30 days in contact with the cement mortar and 9 changes of water, the ph was Cement-mortar increased the calcium concentrations in the test waters. Increased concentrations of calcium were measured over a 14-day period, after which calcium stabilized and remained at 7 to 9 mg/l as calcium, slightly below the background concentration of 11.5 mg/l as calcium. The aluminum concentrations increased in all four test waters after being in contact with the cement-mortar coupons, showing the same trend in the increase. The aluminum concentrations increased from day 1 through day 9, then decreased significantly at day 11 of the test, probably due to a lower contact time of 2 days. The concentrations of aluminum exceeded the drinking water secondary maximum contaminant level (SMCL) of 200 micrograms per liter (µg/l) for the first 9 days of the test. The chromium concentrations increased to 0.07 mg/l in 24 hours, but decreased significantly on day 9. Cement-mortar lining increased consumption of chlorine. Chloramines were more stable than chlorine in the presence of cement mortar. The total solids (TS) concentrations increased up to 1,500 mg/l on the first day and declined substantially after day 9. Cement-mortar lining was also tested with Corrosion Prevention Additive (CPA) waters to find its impact on water quality parameters of ph, alkalinity, and metals from newly installed cement-mortar lining. The three different CPA chemical regimes used are: Orthophosphate (OP) at 1.0 mg/l as P Polyphosphate (PP) at 1.0 mg/l as P Orthophosphate and zinc (OPZn) at 1.0 mg/l as P and 0.3 mg/l as Zn

30 xxviii Impacts of Lining Materials on Water Quality Results show that the ph of all three tests increased to 12 and then dropped in the range of to The alkalinity of all samples increased to 800 mg/l as CaCO 3 and dropped to about 100 mg/l as CaCO 3. Similarly the hardness of all samples increased to 600 mg/l as CaCO 3 and finally dropped to less than 100 mg/l as CaCO 3. TESTING OF POLYURETHANE LINING Three standard water compositions were tested before and after contact with polyurethane-coated coupons. The three water types were: ph = 8.0, low alkalinity/hardness, no disinfectant ph = 8.0, low alkalinity/hardness, 2.0 mg/l chlorine ph = 8.0, low alkalinity/hardness, 4.0 to 6.0 mg/l monochloramine In the presence of polyurethane, the ph was reduced from ph 8 to about ph 6. The ph drop was observed within 24 hours and persisted for 30 days. Polyurethane consumed chlorine and chloramine disinfectant although chlorine was consumed at a greater rate than chloramine. The consumption rate decreased over time, but still persisted at the end of 30 days of testing. Organic carbon was leached from polyurethane, with a greater amount leached in the presence of chlorine than in its absence. Leached total organic carbon (TOC) reacted with free chlorine to form up to 30 µg/l of five regulated haloacetic acids (HAA5), but no trihalomethanes (THMs) were detected. The low ph of 6 would favor HAA5 formation over THM formation. None of the THM or HAA5 concentrations exceeded the drinking water standard of 80 µg/l and 60 µg/l, respectively. Weak to moderate odor intensities were released from the polyurethane and it persisted for the 30 days of this study. TESTING OF EPOXY LINING Three standard water compositions that were tested with epoxy-coated coupons were the same as those tested with polyurethane lining. The greatest impact from epoxy was on TOC and disinfectant residual concentrations. Epoxy exposed to each of the three water types produced significant concentrations of TOC (3.5 to 6.3 mg/l) during the first 24 hours of exposure to water. By the second 24-hour exposure period the TOC decreased substantially. By the end of the 30 days, each of the water types exposed to epoxy had TOC present in concentrations between 0.5 and 1.7 mg/l, with chlorinated water having the highest TOC concentration. Epoxy reduced the concentrations of both chlorine and chloramine disinfectants. Free chlorine was consumed at a greater rate than chloramine. Mature pipe samples lined with epoxy also reduced the disinfectant residual concentrations. Water exposed to epoxy showed some increase in the THM and HAA5 concentrations, but none of these increases exceeded the drinking water standards. The increases in THM and HAA5 concentrations were greatest in the chlorinated water.

31 Executive Summary xxix METHODOLOGY A simplified procedure that provides guidance to water utilities in the selection of lining materials for pipe-lining projects was developed. The methodology incorporates water quality impacts, useful life of alternative lining materials, and comparative costs of water quality impacts. The methodology includes a procedure for estimating the water quality impacts from lining materials using the following three procedural steps: Convert concentration increases to daily rates Calculate mass leaching rates in milligrams per square centimeter per day (mg/cm 2 /day) Apply mass rates to distribution-system-scale conditions In this methodology, bench-scale study data of water quality impacts were first converted into daily rates. These data were then used to develop mass leaching rates in mg/cm 2 /day for all important parameters and lining materials. This procedure has been applied to each major waterquality-impacted parameter identified in the bench-scale tests for each of the three lining materials. Calculated mass rates were then applied to estimate water quality in distribution system water mains. For each water main, pipe size and average flow data are required to analyze turnover time or contact time. This information provides the user the duration of a particular parameter concentration above the accepted limit prescribed by the utility. Analyzing all the major water quality parameters for cement-mortar, epoxy, and polyurethane linings, a worst-case scenario for each lining material requiring maximum flushing was identified. The procedure includes the ability to calculate the volume of additional flushing required to keep the water quality to a desired level. Once a water quality impact mitigation plan is developed, the costs of any additional required flushing are estimated. The whole procedure was developed in two easy-to-use spreadsheets labeled as Impact Module and Cost Module to help water utilities to analyze the suitability of alternative lining materials and the relative water quality impact cost of each option in order for them to select a cost-effective lining material. TEST CASE STUDY The Impact Module software developed in this study for evaluation of cost effectiveness and suitability of a lining material was tested with South Central Connecticut Regional Water Authority s (SCCRWA) data for a lining project to determine the applicability of the software in determining a suitable lining material. An SCCRWA lining project was selected and required data on pipe size, length, and average flow. Background water quality parameters were collected and applied to the software to determine the water quality impacts of each liner and the relative costs of mitigation in order to identify the most suitable lining material for the project. The software developed was able to calculate water quality impacts, and the relative costs for each lining material and helped to select the most suitable liner. The test case results indicate that the water quality impacts from the three lining materials were minimal.

32 xxx Impacts of Lining Materials on Water Quality FINDINGS The following are the important findings and recommendations of this study: A survey of North American and European water utilities conducted for this study indicates that North American water utilities have experience primarily in using cement-mortar lining, whereas in Europe most of the water utilities use epoxy and polyurethane linings. The most frequently reported water quality impact from cement- mortar lining was an increase in the ph of the water. Cement-mortar lining: Impacts to water quality from cement mortar were most severe up to 9 days. After day 9, a significant decrease in most water quality parameter release rates was observed. The ph of water increased drastically to a value of 12.5 with 24 hours of contact with cement mortar and maintained values of ph 10.5 to 11.5 throughout the 30- day test period. Likewise, the alkalinity increased from 35 to 600 mg/l (as CaCO 3 ) within the first 24 hours of contact time with cement mortar. After 9 days of contact, the alkalinity declined to about 100 mg/l as CaCO 3. The total solids content of the water increased to up to a maximum of 1,760 mg/l in the presence of cement-mortar coupons. Cement mortar significantly increased the calcium, aluminum, and chromium concentrations in the water. The aluminum concentrations exceeded the U.S. Environmental Protection Agency (EPA) SMCL, while the chromium levels remained below the EPA maximum contaminant level (MCL). After day 9, release of aluminum and chromium to the water decreased. Cement mortar created a substantial chlorine demand, but the demand for chloramine was much less and ceased after a few days of contact. Cement-mortar lining with corrosion prevention additives (CPA) in water: The findings of differences in the effects produced by any of three CPAs types when compared to a control are: Polyphosphate (PP) reduced ph increases more substantially than orthophosphate (OP) or zinc orthophosphate (OPZn) after day 9. There are no significant differences in increase of alkalinity, hardness, and calcium concentrations among various CPAs. The magnesium (Mg) concentration is highest and aluminum (Al) is lowest for the PP additive. Polyurethane: In the presence of polyurethane, the ph was reduced from ph 8 to about ph 6. The ph drop was observed within 24 hours and persisted for 30 days. Free chlorine was consumed in the presence of polyurethane. The rate of chlorine decay was greater during days 1 through 4 than in later exposure times. TOC was leached from polyurethane, with a greater amount leached in the presence of chlorine than in its absence. Leached TOC reacted with free chlorine to form up to 30 µg/l HAA5, but no THMs were detected.

33 Executive Summary xxxi Weak to moderate odor intensities were released from the polyurethane and it persisted for the 30 days of this study. New epoxy: Epoxy exposed to each of the three water types produced significant concentrations of TOC (3.5 to 6.3 mg/l) during the first 24 hours of exposure to water and then the TOC decreased substantially. The epoxy reacted readily with both chlorine and chloramines during the first 24 hours of exposure. The disinfectant consumption rate decreased over the 30 days. Disinfectant byproducts (DBPs) were present in most samples, with the highest concentrations detected in chlorinated water. Bis-phenol A (BPA) was detected in concentrations of 22 to 33 µg/l during the first 24 hours in all three waters exposed to epoxy. Concentrations decreased substantially by test day 2. Weak to moderate odor intensities were released from epoxy exposed, which persisted all 30 days. Mature epoxy: Free chlorine exposed to the mature pipe was almost completely consumed by each test day. In contrast to the new epoxy, the pipe did not show a decrease in disinfectant demand over the 30 days. TOC leached from the mature pipe sample ranged in concentration from 3.5 to 1.6 mg/l. Test case study: Hydraulic residence time (HRT) is a key parameter in controlling water quality impacts from lining materials. RECOMMENDATIONS The following general recommendations are made as a result of this study: The methodology developed in this study should be used to analyze water quality impacts of the alternative lining materials cement-mortar, epoxy, and polyurethane, and to rationally select a lining material. Cement-mortar lining, in general, should be avoided in low water circulating areas such as dead ends. In such areas, water should be flushed regularly. In other areas, water should be flushed initially. The frequency and duration of flushing can be determined by using the methodology developed in this study. Epoxy lining and polyurethane lining for water mains have been found to be good alternatives to cement-mortar lining and all water utilities should consider these technologies as an alternative to cement-mortar lining.

34 xxxii Impacts of Lining Materials on Water Quality

35 CHAPTER 1 INTRODUCTION BACKGROUND AND INTRODUCTION As water mains age and reach the end of their useful lives, their performance diminishes, resulting in high maintenance costs, deterioration of water quality, loss of hydraulic capacity, and a significant increase in customer complaints. Rehabilitating a poorly performing main via cleaning and lining decreases maintenance costs and customer complaints, while significantly improving flow, pressure, and water quality. About 50% of water mains in North American water systems are composed of cast iron; of these, most installed prior to the 1950s are unlined. Although structurally sound, they show severe tuberculation. The results are a reduction in hydraulic capacity, water quality problems, and high energy costs. If the water main is structurally sound, rehabilitation through cleaning and lining is the most economical option. If the main is structurally unsound, however, either due to graphitization or external corrosion, replacement of the pipe is the desired option. Many different pipe cleaning and lining techniques exist. In North America, more than 90% of the in situ pipe lining conducted used cement-mortar or epoxy lining, with use of epoxy lining gaining prominence during the last decade. In the United Kingdom (U.K.), epoxy has been used as the preferred lining material in rehabilitation of water pipelines. In the last few years, however, polyurethane has become the most widely used lining material in the U.K. One of the important purposes of pipe lining is to prevent leaching of metals from the wall of water mains and thus, to improve water quality. However, there is concern about leaching of metals and organic chemicals from these lining materials. The effectiveness and the water quality impacts of these linings depend on the properties of the lining material itself, installation procedures, water characteristics, exposure time, and surface area to volume ratio. Rehabilitation of water mains using cleaning and lining results in significant cost savings compared to replacement. However, if the proper lining is not used, the water quality impact may be severe and the life of the rehabilitated pipe may also be reduced. The first test case of epoxy lining of water mains in the United States was conducted by the American Water Works Association Research Foundation (AwwaRF) in 1993 (Conroy et al. 1993). Water quality problems as a result of leaching chemicals from the epoxy lining were of great concern in the United States at that time. The AwwaRF study showed that leaching from epoxy lining is minimal and satisfies all U.S. Environmental Protection Agency (EPA) primary drinking water regulations. However, without the approval of the National Sanitation Foundation (NSF), water utilities were very hesitant to accept epoxy lining as an alternative to cement lining. The quality of epoxy and related installation procedures have significantly improved during the last decade. In 2004, NSF provided the approval certification (NSF 61) for use of epoxy as a viable alternative for cement-mortar lining for drinking water mains. As a result, there is an impetus for North American water utilities to consider the epoxy lining technique as an alternative to cement lining. Water quality impacts from lining leachate are still a concern. When comparing the cost of alternative lining materials available for water main rehabilitation, it is necessary to consider the total cost of the lining material, including the water quality impact cost. The potential for water quality impact of each lining material may vary depending on the characteristics of the 1

36 2 Impacts of Lining Materials on Water Quality lining material, the water, and the effectiveness of lining construction. Many studies have been performed to identify water quality impacts from water pipe linings. These study results indicate that water quality impacts of lining materials may be an important decision criterion in selecting a lining material or rehabilitation technique. While water quality impact is considered in selection of a lining material, it is hardly considered or integrated into the cost analysis and selection process for rehabilitation decisions. In this study a methodology for a decision support system for selection of an appropriate lining by water utilities that considers potential water quality impacts has been developed. This system considered different water characteristics, different hydraulic capacities (exposure times), and different treatment processes to guide water utilities in selecting the most suitable lining based on structural and water quality considerations. PROJECT OBJECTIVES The objectives of this study were the following: Gather, compile, and synthesize detailed information on water quality impacts associated with different water main lining materials from past and ongoing research, published literature, water system compilations, and other relevant sources of information from different research organizations and water utilities. Conduct bench-scale testing of various lining materials in use in water main lining under various exposure times and water characteristics, and document water quality impacts. Develop criteria to control or minimize leaching. Conduct tests on matured water main liners to determine impacts on the condition of the lining and water quality. Develop a methodology and protocols to help water utilities to consider water quality impacts in order to select the most suitable lining rehabilitation method for long-term performance. Identify research needs and knowledge gaps on water quality impacts from lining materials. PROJECT APPROACH The technical approach to accomplish the project objectives and goals consisted of the following tasks: A comprehensive review of literature and available information was conducted to study literature on water quality impacts and regulatory compliance from various lining rehabilitation methods and to identify utilities with extensive experience with lining rehabilitation. Selected water utilities with lining experience were contacted to conduct a questionnaire survey of historical lining rehabilitation projects and their impacts on water quality. Data were compiled, synthesized, and evaluated to identify principal parameters impacting water quality. A bench-scale leaching study was conducted with various combinations of lining materials and water qualities to evaluate impacts of lining materials and background water chemistry on water quality.

37 Chapter 1: Introduction 3 Leaching tests with one matured sample of epoxy lining were conducted to evaluate the impacts of age of epoxy lining on water quality. Information developed was evaluated to develop a methodology to identify water quality impacts due to various lining materials, and to estimate the cost of options to mitigate water quality impacts. The methodology will be suitable for selection of cost-effective lining rehabilitation technologies by water utilities. Knowledge gaps were identified to formulate future research needs. An industry workshop was conducted on September 23, 2008 to review and validate the decision methodology and study findings. The methodology was applied to one test case water utility to verify the applicability of the system developed. Findings and conclusions were drawn, and recommendations are made. Based on these tasks, the final report was prepared. ORGANIZATION OF THE REPORT This report is organized as follows: Chapter 1 introduces the problems of water quality impacts as a result of water main lining by cement mortars, epoxy, and polyurethane. Chapter 2 summarizes the literature of bench-scale, field, and other studies on water quality impacts due to cement-mortar, epoxy, and polyurethane lining of water mains. Chapter 3 presents a survey of water utilities describing their experiences with cement mortar lining, epoxy lining, and polyurethane lining, and their impacts on water quality. Chapters 4, 5, 6, and 7 describe bench-scale laboratory tests of new water main lining sample studies considering various inlet water qualities, and flushing times using cement-mortar, epoxy, and polyurethane linings and one matured sample of epoxy lining. Data on water quality impacts were collected and analyzed. Chapter 8 provides an evaluation of water quality impacts. Chapter 9 analyzes the data collected in Chapters 4, 5, 6, and 7 and develops a methodology for water utilities to use to select proper lining materials for their specific utilities with due consideration of water quality impacts. This chapter estimates the life of lining materials and develops a procedure to estimate the cost of water quality impacts of various alternative linings in order to select an efficient and cost-effective lining. Chapter 10 describes the ease of use of the methodology developed in Chapter 9 by practical application in a test case utility. All problems of usage are identified and resolved for easy use. Chapter 11 identifies research gaps for future research. Chapter 12 summarizes the findings of the project and recommends measures for proper selection of water main lining for rehabilitation of mains.

38 4 Impacts of Lining Materials on Water Quality

39 CHAPTER 2 LITERATURE REVIEW INTRODUCTION A literature review was conducted to establish the materials, practices, applications, and impacts on water quality as a result of using cement-mortar, epoxy, and polyurethane resin lining in water main rehabilitation processes. This literature review is part of an effort to develop a protocol and procedure for laboratory testing that will help water utilities assess the potential water quality impacts of in situ applied liners used in water main rehabilitation. Cement-mortar, polyurethane, and epoxy-resin linings are all materials used in the nonstructural rehabilitation of potable water mains. The potential impacts on water quality of the different pipe lining technologies vary, depending on the lining material, initial water quality, adherence to application procedures, and quality assurance/quality control (QA/QC). Linings for potable water pipes must be certified according to American National Standards Institute/National Sanitation Foundation (ANSI/NSF) Standard 61 (NSF 61) (AWWA 2002) in the U.S., and Drinking Water Inspectorate Regulation 31 in the U.K. prior to use. These regulations test for aesthetic impacts, microbial growth, presence of substances of possible concern to public health, bacterial count, and heavy metals (Kut 1993). Both regulations ensure acceptable water quality and negligible human health impacts for water that comes in contact with the material being tested. ANSI/NSF 61 covers indirect additives, products, and materials, which must be tested for effects on water quality and their fitness for purpose (WRc-NSF 2004). CEMENT-MORTAR LINING In North America, cement mortar lining is the conventional choice to renovate unlined cast iron mains of 4 inches (100 millimeters [mm]) and larger in diameter. Studies and field experience have shown that, in soft waters, the cement-mortar lining may corrode and adversely affect water quality. In the U.K., cement-mortar lining was used from the 1930s through the 1990s, but was found to be ineffective as a long-term solution when water is relatively soft and aggressive (WRc 1997). Cement is composed of numerous compounds, principally silicates (Ca 3 SiO 5, Ca 2 SiO 4 ) and aluminates (Ca 3 Al 2 O 6 ). When these compounds come into contact with water and hydrate, calcium hydroxide (Ca(OH) 2 ) forms and then dissolves into Ca 2+ and OH -. This leads to a rise in ph and the potential for calcium precipitation (Slaats et al. 2004). The first cement-lined iron pipe was installed in Charleston, S.C. in The decades following have seen numerous improvements in manufacturing techniques and QC, as well as technical understanding of the interaction between water and the cement itself. In 1929, a tentative standard for cement-mortar lining was established by the American Standards Association; this was later adopted and published by the American Water Works Association (AWWA) in 1932 as Specifications for Cement-Mortar Lining for Cast Iron Pipe and Fittings (Bonds 2005). During the 1940s and early 1950s, much research regarding types of cement, manufacturing techniques, and curing was conducted. The centrifugal process for pipe lining and the use of asphaltic sealing materials were also developed during this time. Later revisions to the 5

40 6 Impacts of Lining Materials on Water Quality document included a reduction in the minimum permissible thickness in 1964, and the addition of a standard test for toxicity from sealant materials in 1980 (Bonds 2005). Impact on Water Quality Soft waters have a low concentration of carbonates and bicarbonates and may also be acidic due to the presence of dissolved CO 2. Soft waters are aggressive to the calcium hydroxide (Ca(OH) 2 ) in cement, and may result in leaching of this compound. Soft water will also reduce the mechanical strength of pipe surfaces by hydrolyzing calcium silicate hydrates into soft silica gels (Bonds 2005). Waters with low ion content are aggressive to calcium hydroxide contained in the cement mortar lining (Leroy, Schock, Wagner, and Holtschulte 1996) resulting in leaching of calcium hydroxide. These waters will also react with calcium silicate hydrates forming silica gels, resulting in a soft surface with reduced mechanical strength. The amount by which the ph is increased depends on the alkalinity and buffer capacity of the water in contact with the cement. Aggressive waters may also impact water quality by leaching metals from cement-mortar pipe walls and linings. Aluminum has been known to leach in significant quantities and pose a health threat to dialysis patients (Berend and Trouwborst 1999). Over a course of 2 months, aluminum levels were found to increase from 5 micrograms per liter (µg/l) to 690 µg/l. Chromium, barium, and cadmium are also of concern. Leaching of metals typically occurs under low flow conditions and can leach from newly installed materials for as long as 4 years (Berend and Trouwborst 1999). The leachability of regulated metals from cement-mortar linings has also been potentially linked to the manufacture of dry cement mixture in kilns burning hazardous wastes (Guo et al. 1998). In a laboratory investigation, sections of test pipe were shown to leach significant levels of barium, cadmium, and chromium after being exposed to water for 14 days. Concrete-mortar linings can impact water quality due to the chemical reactions that take place between the mortar and the water the pipe carries. Significant deterioration of water quality can occur from corrosion of cement mortar, alkalization of water, consumption of chlorine (Schwenk 1990), and by bacterial growth on the lining itself (Niquette et al. 2000; Momba and Makala 2004). It has even been suggested that a link could exist between corrosion of cementmortar pipes and acid rain (Anderson 1994). EPA has reported that lakes fed by acid rain experience a decrease in alkalinity. Water with weak buffering, low ph, and low alkalinity will corrode cement-mortar by leaching calcium in an attempt to restore equilibrium (Anderson 1994). Focus is given to the water quality impacts from in situ applied cement-lined pipe (Douglas and Merrill 1993). Chung et al. (Chung et al. 2004) attempted to outline a method to predict corrosion in concrete-lined water pipes. The interaction between cement and water occurs in pores in the cement. There are two types of chemical interaction. The primary interaction is the rapid degradation of cement via the calcium carbonate saturation and the carbonate species of water. The secondary interaction involves lime leaching. Calcium aluminates or calcium silicates leached into the water can raise ph, affecting water quality. The buffer capacity is related to the alkalinity of the water and thus, increased alkalinity is desired to reduce corrosion. Corrosion rate decreases with increased free CO 2 or bicarbonate (Chung et al. 2004). A laboratory investigation conducted in Germany tested six different water types to determine which water parameters most heavily influenced mortar corrosion and water quality impact (Schwenk 1990).

41 Chapter 2: Literature Review 7 The effects of mortar variables were also investigated (Schwenk 1990). The sand-tocement ratio had no effect on corrosion. Lining technology did appear to impact corrosion rates. The amount of corrosion correlates with the dissolved CO 2 content of the water, with the threshold concentration needed to initiate mortar wear equaling 7 milligrams per liter (mg/l). Water quality parameters such as CO 2, calcium concentration, and physical parameters such as volume to surface area ratio, had a significant influence on corrosion and water quality (Schwenk 1990). An AwwaRF study (Douglas and Merrill 1991) attempted to identify factors leading to the corrosion of in situ applied cement-mortar lining and to study the effects of water quality and cement matrix variables on rates of corrosivity. Sets of 10 lined pipe sections mounted vertically on a support rig were monitored for corrosion and effect on water quality. Results showed that the most important impacts on water quality were ph, alkalinity, and calcium concentration. Extended exposure to water under low-flow conditions results in increased leaching of ph, alkalinity, and calcium. Contact under high-flow conditions seems to accelerate these effects. Waters low in ph, alkalinity, and calcium are the most aggressive to cement linings. Corrosion rates are also influenced by cement-mortar composition, method of lining, and use of a protective coating. It is recommended that utilities routinely flush low-flow sections of pipe to avoid adverse effects on water quality (Douglas and Merrill 1991). Microbial Growth and Bio-film Formation The impact of concrete liners on bacterial growth as compared to plastic pipes was examined during an experiment in which asbestos-cement and cemented cast iron, as well as polyvinyl chloride (PVC) and polyethylene (PE) pipes were exposed to different source waters, after which bacterial biomass was measured (Niquette et al. 2000). Results showed that steel and iron pipes had the greatest biomass, PE and PVC pipes the least, and asbestos-cement and cemented cast iron fell in the middle range. This indicates pipe material can impact biomass growth, and therefore, water quality. This is an example of an indirect water quality impact from a pipe material. Biofilm growth is based on disinfectant, temperature, and biodegradable organic carbon (BDOC). This suggests that surface roughness and leaching of organic compounds from a pipe liner could affect water quality and should be examined in testing (Niquette et al. 2000). Cement-mortar based pipes also fell between plastic and metal pipes in a study that examined the effects of distribution system materials on bacterial regrowth (Camper et al. 2003). This study compared coupons lined with test materials exposed to water in both a laboratory and field investigation. A similar investigation in South Africa compared the effect of various pipe materials on biofilm formation (Momba and Makala 2004). Water from a surface source was exposed to a variety of plastic-based pipes as well as cement and asbestos cement. The colonization and growth of microbes on the pipe surface was then studied by examining attached coliforms and heterotrophic plate count bacteria. No difference was detected between cement and asbestos-cement. Disinfectant Decay Disinfectant concentrations in the distribution system can cause water quality issues by either being higher than intended and imparting a chlorinous odor to water, or by not being adequate to stop bacterial growth, which can lead to organic off-odors. The consumption of chlorine in concrete pipes is attributed to the oxidation of inorganic compounds in the cement, with chlorine consumption affected by the same parameters that increase alkalinity (Schwenk

42 8 Impacts of Lining Materials on Water Quality 1990). The impact of cement-mortar pipe and pipe liners on chlorine decay when compared to other types of pipe was also examined. Plastic pipes were compared to cement-lined iron and analyzed for both the rate and mechanism of chlorine decay. Results showed that cement-mortarlined iron pipes have a slightly higher reactivity than PVC or PE pipes, and that the decay was controlled by the reactivity of the pipe. Utility Survey In 1991, an AwwaRF study (Douglas and Merrill 1991) conducted a survey of utilities with low alkalinity aggressive water in the United States. All of the utilities were located in areas with low alkalinity water, generally obtained from rainfall or snowmelt as opposed to groundwater. Nine of the utilities questioned reported problems with their mortar liners, and seven of these reported water quality impacts from ph increase. Problems typically occurred in low-flow areas in newly lined pipes. Four utilities (South Central Connecticut Regional Water Authority [SCCRWA], Marin Metropolitan Water Department [MMWD], Washington Suburban Sanitary Commission [WSSC], and East Bay Municipal Utility District [EBMUD]) identified significant water quality impacts from cement-mortar linings. Water ph increased significantly, but no structural deteriorations were observed. SCCRWA water was found to be the most aggressive with a Langlier index of To solve the problems of water quality, all utilities adopted a measure to flush the water. Field Testing Field tests of newly installed cement-mortar-lined pipes were conducted with waters of SCCRWA and MMWD for 13 weeks (Douglas and Merrill 1991). The pipes were flushed regularly. In SCCRWA water under a continuous flow regime, no water quality impacts in terms of ph, alkalinity, and calcium were observed. However, water in the test pipe with weekly flushing, registered an increased ph of 12.5, alkalinity of 760 mg/l CaCO 3, and calcium of 135 mg/l CaCO 3 after 1 week of operation. Water quality impacts increased from the continuous flow system to the daily flushing system, to the weekly flushing system. Alkalinity and calcium values stabilized after 7 to 9 weeks of operation, whereas ph values did not stabilize even after 12.5 weeks of operation. Field tests at locations in New Haven, Conn. and Corte Madera, Calif. demonstrated that leaching of lime from cement-mortar corrosion can significantly impair water quality by raising ph, alkalinity, and calcium concentrations (Douglas and Merrill 1996). Utilities at both locations examined the effects of flow rate, lining composition, lining technique, seal coating, and water quality. Both sites showed that stagnant water, which remained in newly lined sections of pipe for a week, experienced ph increases up to 12. In Johnson County, Kan. and Howard County, Md., water main rehabilitation using in situ cement-mortar lining was conducted after customer complains about rusty water at the tap. Cleaning and cement-mortar lining eliminated consumer complaints (Hall 1999; Mitchell and Dumas 1992). An incidence of metals leaching from a factory cement-lined newly installed iron pipe leading to health effects occurred in Cuaçao, the Dutch Antilles (Berend and Trouwborsk 1999). After installation of the water main, patients in a nearby dialysis center became sickened with acute aluminum intoxication. While it is known that metals, including barium, cadmium, and

43 Chapter 2: Literature Review 9 chromium, can leach from cement-mortar, the high concentrations of aluminum leached in this case were attributed to new pipe, corrosive water, high ph, water temperature, and long residence time, as well as higher-than-usual concentrations of aluminum in the cement itself. EPOXY-RESIN LINING Epoxy resin has replaced cement mortar in many areas in the U.K. as a material for nonstructural pipe rehabilitation. Epoxies consist of two components that react to form a hard inert material. The first component consists of an epoxy resin, a low-molecular-weight polymer with epoxy groups at each end. The second component contains diamines, and is known as the curing agent, or hardener (Dekker 1988). Employing different types and amounts of hardener controls the cross-link density between the resin and hardener, varying the structure of the material (Bhangale 2003). Different structures yield different material characteristics, such as strength or hardness. It is important to choose epoxy resin materials carefully and ensure that the mix ratio between the resin and hardener components is correct and carefully monitored during the lining processes (WRc 1997). The correct mix ratio of resin to hardener varies depending on the epoxy used. Epoxy resin manufacturers specify the optimum mix ratio required to achieve a lining that exhibits good adhesion to the interior surface of the pipe, resistance to moisture, minimum leaching potential, and adequate hardness (WRc 1997). Attention to curing time is also critical to ensuring that water quality is not impacted (DWI 1990). All epoxy resin material used for in situ lining of potable water mains must be solventand benzyl-alcohol-free, slump-resistant, rapid-curing, and moisture-tolerant (AwwaRF 2003). The basic formulations are all very similar because of the limitation put on the formulators when developing materials that will cure within 16 hours at temperatures as low as 3 Celsius ( C), while providing acceptable water quality (DWI 2005). General information on the formulations is in the public domain, while the specific individual formulations are proprietary. The basic principles of epoxy resin formulation are well documented. Resin material types are, in general, classified as being Type 1 or Type 2 materials. Type 1 materials are typified by the early formulations and are generally low viscosity materials that can be pumped by ordinary nonumbilical hose lining equipment and have a curing agent based on trimethyl hexamethylene diamine (Deb et al. 2006). Type 2 resins are more recent and are higher viscosity materials requiring heated umbilical lining hoses and rigs for application and have curing agents based on more complicated alternative amine systems. Type 2 formulations are very similar to the above for the resin component, but in the hardener, the amine curing agents are more complex (Warren 1983). Epoxy-resin lining provides a corrosion-resistant lining system in pipes that improves flow capacity and does not affect water quality. Application of epoxy-resin lining is more efficient and less expensive than structural pipe rehabilitation or replacement. To successfully use epoxy lining, the pipe section needing repair must be cleaned according to recommended methods and the carefully chosen epoxy material must be applied as per published guidelines (WRc 1997). In addition, testing and monitoring should be implemented before, during, and after application processes as part of QA/QC. Epoxy-resin-lining techniques have been used successfully in the U.K.

44 10 Impacts of Lining Materials on Water Quality Evolution of Epoxy-Resin Lining Epoxy-resin-lining techniques were first introduced in 1979 in the U.K., and were widely used and accepted for lining potable water pipelines. Research has been conducted to improve materials for lining applications and to develop operational practices designed to aid in using epoxy-lining techniques as a method of nonstructural pipe rehabilitation. These operational guidelines set the standard that contractors in the U.K. must achieve in order to participate in epoxy-lining procedures (WRc 1997). More recently, use of epoxy-resin lining has been discontinued throughout the U.K. and the majority of water utilities have chosen to use polyurethanes instead of epoxy resin. Although polyurethane linings have only been used in recent history, operational standards and guidelines have been published and the use of polyurethane linings has been widely accepted throughout the U.K. The benefits of epoxy linings over concrete mortar are numerous, mostly relating to the smoother surface epoxy coatings have when compared with cement (Kut 1993). Improved flow characteristics due to greater C-values can increase system capacity. Smoother surfaces result in reduced tuberculation and biofilm growth, and easier pigging and maintenance. Other benefits of epoxy over cement mortar are increased durability in soft water, and reduced impact on ph. Following adequate results of laboratory-based accelerated testing, in-service assessments ensued, which involved lining pipe samples with different epoxy resins and installing inspection pits in various distribution systems covering a range of pipe diameters and water qualities. Over a 4-year study period, no evidence of deterioration was observed and testing showed that there was no short-term breakdown of the linings (Warren 1983). Water quality assessments were also conducted and showed that epoxy-resin linings satisfied all test criteria (Warren 1983). Due to its success in rehabilitating water mains with minimal disruption and restoring water quality, the use of epoxy resin is widespread, especially in the U.K. Quality Assurance and Quality Control (QA/QC) QA/QC procedures are outlined in the Operational Guidance and Code of Practice (OG/CP) for practices in the U.K. (WRc 2001). In developing similar standards, recommendations, and guidelines in the U.S., the following QA/QC procedures were considered to ensure linings are applied correctly to the interior surfaces of potable water mains (WRc 2001): Routine weight checks Use of metering devices for mix ratio control Use of spin-up time before lining Use of approved lining rigs, static mixers, and lining application heads Guaranteed control over minimum cure times Pre- and post-lining closed-circuit television (CCTV) surveys Limitation of lining in cold weather In addition to the AWWA standard, the epoxy-resin materials used in lining processes must be certified under NSF 61. NSF 61 addresses indirect additives, products, and materials that come in contact with drinking water. The standard evaluates materials and products for

45 Chapter 2: Literature Review 11 contaminant leaching and certifies that the leached contaminant levels are within acceptable limits. The standard also evaluates a material s potential for microbial growth. Utility Experience Between 1982 and 1988 approximately 247 kilometers (km) of potable water pipes were lined with epoxy resin in the U.K. within the Western Division of Yorkshire Water (Warren Associates 1997). According to the report (Warren Associates 1997), the quality of epoxy-resin linings in potable water pipes varied drastically during this time, which can be largely attributed to the fact that this method of rehabilitation was in its infancy and many of the mains lined were part of major trials. Published operational guidelines had not yet been released and QC methods were not as stringent as those implemented in later years. Investigations of pipes lined with epoxy during ensued by gathering information, sampling pipes at various locations, and analyzing samples to determine the quality of epoxy lining. For each sample, the mix ratio of resin to hardener was determined and in many cases found to be incorrect, which led to the finding that approximately 50% of the linings were in need of remedial work. In many cases, the epoxy was found to be inadequately cured, potentially impacting water quality. There are many reasons that may have caused an incorrect mix ratio. In North America, epoxy-resin lining did not make an appearance until the mid-1990s. The first trial in the U.S. of in situ epoxy-resin lining was in Chester, Pa. Chester Water Authority (CWA) was chosen by AwwaRF as a demonstration site for the epoxy-lining process (Conroy et al. 1993). The test main was a 6-in., 80-year old, unlined dead-end cast-iron water main. A 30-day water quality monitoring program concluded that leaching epoxy was well within water quality guidelines. The hydraulic performance gained with the epoxy-lining process was confirmed with flow test results showing an increase in the Hazen-Williams coefficient of roughness from 22 to 114 (Conroy et al. 1993). The first projects involving the application of NSF 61-certified epoxy-resin lining were carried out in the City of Ottawa, Canada (AWWA 2002). Epoxy-resin lining was selected because previous regional experience with this technology yielded an innovative, cost-effective solution to rehabilitating existing, structurally sound pipes (Marcuccio and Salvo 2000). In addition, several states and other Canadian provinces were involved with epoxy-lining projects during 1998 (Salvo 1998). The states of Colorado, Indiana, Kansas, Minnesota, New York, and Pennsylvania undertook projects. In Pennsylvania, Bucks County Water and Sewer Authority (BCWSA) selected an area to apply epoxy-resin-lining techniques (Salvo 1998). Water Quality Impacts Laboratory testing of epoxy-resin lining is essential in determining the impacts on water quality, which may occur following installation. Two main mechanisms exist by which epoxy lining may impact water quality. The first is the direct leaching of organic compounds from the lining itself. The second is the growth of microbes that may utilize these compounds as a substrate and whose presence may degrade water quality (Khiari et al. 1999). Additionally, the interaction between types of disinfectant used (e.g., chlorine and chloramines) with each other, pipe surfaces, microbial growth, and other chemical constituents in the distribution system should also be examined. Several laboratory studies were reviewed for this literature review. Generally, these articles discuss testing methods and ways to determine how epoxy lining will affect water quality and aesthetics.

46 12 Impacts of Lining Materials on Water Quality During the approval process for epoxy lining in the U.K., efforts were made by both the National Water Commission (NWC) and Department of the Environment (U.K.) Chemicals, Construction Products and Materials (DoE-CCM) to examine the potential impacts of epoxy lining on water quality. The NWC study examined the effects of epoxy on the following criteria prior to approval, and determined that all had been adequately satisfied (Warren 1983): Ability to support microbial growth Potential to cause organoleptic and physical deterioration of the water Potential to release toxic metals into the water Potential to release cytotoxic compounds into the water The DoE-CCM testing focused on assessing health implications of epoxy-lining materials by examining in detail the material specification, composition, properties, use, and potential toxicity (Warren 1983). The DoE-CCM determined that epoxy lining has no risk of acute or subacute toxicity, but did not examine longer term effects. Although epoxy lining met with NWC and DoE-CCM standards, more comprehensive testing reported by Warren revealed an increase in microbial growth following epoxy rehabilitation. A general increase in nonpathogenic microbes was observed following epoxy lining of distribution systems, with levels increasing to several orders of magnitude above background before eventually leveling off (Warren 1983). Further testing also led to the identification of a number of organic compounds with the potential for organoleptic impacts that leached from epoxy-lining materials. The impact of epoxy linings on water quality was examined by exposing epoxy-coated steel panels, along with four other lining materials, to test water over a period of 30 days (Alben et al. 1989). Results showed that epoxy leached the greatest amount of total organic carbon (TOC), with rates of 40 to 187 μg/l-day. Compounds identified included methyl isobutyl ketone (MIBK), and xylenes. Field studies by the same author examining recently epoxy-coated water storage tanks yielded the same compounds. Five NSF- and AWWA-approved epoxy resins were evaluated by the city of Calgary, Alberta, Canada, by exposing coated coupons to test water for 72 hours. The water was then analyzed for TOC and volatile organic compounds (VOCs). Results showed that all five epoxies leached significant quantities of both. TOC ranged between 34 and 345 mg/l and benzene, toluene, ethylbenzene, xylene (BTEX) from 0.2 to 48 mg/l for all samples (Satchwill 2002). Specific research into the aesthetic impacts on water quality has repeatedly shown the ability of epoxy to contribute off flavor and odors to drinking water. This was evidenced by two studies that used the Utility Quick Test (UQT), a leaching and migration protocol developed in the U.S., and human panelists to evaluate drinking water odors. In the first, pipe materials were subjected to longer-term analysis in a series of three plumbing rigs, which cycled water with chlorine, chloramine, and no disinfectant for over a year (Durand and Dietrich 2005). Samples were taken regularly and analyzed for water quality parameters including odor descriptors and intensity. Data from the rigs initially indicated odor descriptors of a plastic odor for epoxycoated copper pipe. Odors decreased with the passage of time, however, and after 1 month, the odors had been reduced to threshold. Levels of TOC leaching from the pipe also tended to drop significantly during this initial 1-month period. A second study, involving shorter exposure times, indicated that the epoxy liner leached compounds that contributed a plastic/adhesive/putty odor to drinking water. The odor was

47 Chapter 2: Literature Review 13 present in water exposed to the pipe material, and its intensity did not diminish during subsequent flushing and stagnation periods within the time frame of the UQT. While significant odor occurred after 72 hours of leaching, short-term laboratory tests demonstrated that the odor imparted was proportional to stagnation time and was not detectable after only 1 hour of leaching. Field tests with flowing water after full-scale installation of the epoxy liner in an apartment building showed no detectable odor, indicating that water usage determined the extent of the odor. The presence of chlorine or chloramines did not change either the intensity or the descriptors for the plastic/adhesive/putty odor (Heim and Dietrich 2005). These results demonstrate the ability of epoxy resin liners to significantly impact water quality and aesthetics, especially immediately following installation. The curing process has been determined to be the key factor in the leaching of organics. POLYURETHANE LINING Polyurethane is a polymer used in nonstructural pipe rehabilitation. The compound was first discovered by Otto Bayer in 1937, and was developed during World War II as a substitute for rubber. Although polyurethane has been used extensively as an industrial coating material for some time, only recently has approval by regulatory agencies and advances in application technology spurred its increased use in water main rehabilitation (Warren Associates 2000). The manufacture of polyurethane involves the reaction of a polyol with a diisocyanate or polyisocyanate. Manufacturers generally designate the isocyanate component A and the polyol component B. Pigments are usually added to the polyol component to determine color (Guan 1995). Various additives and catalysts may be used depending on the specific characteristics of the desired product. The polyurethane used for rehabilitating water mains is strong, durable, and water-resistant. In water main relining, this allows rapid, same day return to service. Recently, 100% solids, 2-component polyisocyanate-cured materials have been developed with extremely short curing times and no negative effect on water quality (Warren Associates 2000). Their highly durable nature and suitability to rapid application have made polyurethane linings poised to become the next generation of in situ relining products, already replacing epoxy in many parts of the U.K. Although polymeric linings are not currently used in North America, the U.K. has already generated operational guidelines for the application of polymeric pipe linings (Warren Associates 2000). The application of polyurethane linings is a complex process, and rigorous QA measures must be undertaken to ensure success (Guan 1995). Evolution of Polyurethane Lining The recent increase in the utilization of rapid-setting polyurethane polymers for water main rehabilitation has progressed following the success for epoxy-resin-lining systems for the same purpose. Epoxy-lining products have been proven to successfully and cost effectively restore water mains and extend their service life by 30 to 50 years (Salvo 1998) without compromising water quality. Rapid-setting polymer linings are the next evolutionary step in in situ relining products, offering the benefits of fast curing times and rapid, same-day return to service (Warren Associates 2000). Although epoxy-resin products have been used extensively and successfully in the U.K. since the 1970s, they were limited by a mandatory 16-hour cure period. This necessitated a 36- hour return to service time before water was available to consumers. After polyurethane liners

48 14 Impacts of Lining Materials on Water Quality with a cure time of 30 minutes were introduced in 1999, service in the U.K. could be restored the same day with greatly reduced disruption (Metcalfe 2005). Polyurethane lining products are generally a solvent-free (100% solids) isocyanate-cured material based on Interpenetrating Polymer Network (IPN) technology. They are highly tolerant of moisture and cure sufficiently within 30 minutes of application to allow disinfection and flushing procedures to occur. The resulting product has a hard glossy surface, which provides high C-values (Metcalfe 2005). In a comparative study among 100% solids epoxies, 100% solids elastomeric polyurethane, and 100% solids rigid polyurethane linings (Walker and Guan 1997), it was determined that the 100% solids rigid polyurethane system was the best performing liner. Rigid polyurethane was determined to be the best system due to its high tensile adhesion, low undercutting, and excellent abrasion and impact resistance. These attributes make 100% rigid polyurethane a high performance corrosion protection system that can be used in cases of aggressive environments, repeated hydration and dehydration, high flow velocities, and possible pipe strains (Walker and Guan 1997). The most recent advancements involving polyurethane coatings incorporate antimicrobial additives (Guan 2001). This could prove highly beneficial to both the water and wastewater industry, offering substrate protection by hindering biofilms and microbial growth, while at the same time protecting the pipe from long-term corrosion (Guan 2001). Quality Assurance and Quality Control (QA/QC) QA/QC procedures are outlined in Operational Guidelines and Code of Practice (Warren Associates 2000) and Polyurethane Coatings for the Interior and Exterior of Steel Water Pipe and Fittings (AWWA 2000), as well as Assuring Quality When Applying 100 Percent Solids Polyurethanes developed by Madison Chemical Industries (Guan 1995). QA measures should include: Routine weight checks Metering devices for mix ratio control Use of spin-up time before lining Approved lining rigs, mixers, and application heads Pre- and post-lining CCTV surveys Limitation of lining in cold weather Utility Case Studies of Polyurethane Linings One of the first trial projects involving the use of a rapid-setting polymeric lining to rehabilitate water mains occurred in Plymree, U.K. and was conducted by South West Water. The utility sought to restore 2.5 km of water pipe near the village of Plymree, and wished to avoid the 36-hour interruption of service that the use of epoxy would bring. The polymer selected was Copon Hycote 169, applied by Pipeline Polymers. The project was conducted successfully, and service was restored to customers within the same day. After its initial success, South West Water, U.K., continued to increase its use of polyurethane liners, using it to line 29 km of pipe in 2000 compared to only 4 km with the older epoxy method. The utility has now adopted rapid-setting polyurethane as its principal in situ lining material for pipes ranging from 3 to 36 inches in diameter (Metcalfe 2005).

49 Chapter 2: Literature Review 15 Yorkshire Water, U.K., utilized eight specially designed plumbing rigs to spray-apply rapid-setting polyurethane lining to rehabilitate water mains in situ (Parker Hannifin 2002). In 2001, the Sandwich Water District, Mass., conducted a successful lining trial using rapid-setting polyurethane as a lining material for a 8,500-foot vinyl-lined asbestos-cement pipe to prevent leaching of tetrachloroethylene (PCE) from the vinyl lining. This project included 1,400 feet of water mains lined and put back into service within a 12-hour period (Gove, Oram, and Mahoney 2003). Impact on Water Quality Rapid-setting, 100%-solids polyurethane linings have a negligible impact on water quality due to both lack of solvent and rapid cure time. The absence of the partially soluble constituents and nonreactive accelerators present in epoxy resins means minimal potential for aesthetic complaints from consumers related to taste and odors. Additionally, the rapid curing time, when compared with epoxy resins, reduces the likelihood that a rehabilitated pipe will be returned to service before being allowed to adequately cure, which could result in the migration of volatile compounds, which negatively impact water quality (Metcalfe 2005).

50

51 CHAPTER 3 SURVEY OF CURRENT PRACTICES AND WATER QUALITY QUESTIONNAIRE A survey questionnaire was prepared to collect existing information on the experiences of water utilities on water main lining with cement-mortar, epoxy, and polyurethane materials, and their impacts on water quality. The questionnaire included questions summarized in the following sections: Contact information Historical use of lining materials Reasons for lining pipes Water quality samples and results before and after the liner was placed Water quality changes, good or bad, due to the liner material and changes with time Reports of the lining installation projects QC procedures used to install or apply the liner Expectation of the useful life of the lining Time duration between lining and putting the pipe back into service Inventory of lining projects and their descriptions Water quality measurements made before and after lining A copy of the questionnaire is included in Appendix A. QUESTIONNAIRE RESPONSES AND RESULTS Questionnaires were sent to 48 water utilities in North America, the U.K., Sweden, Japan, Norway, and Australia. After a second request, replies were received from 22 water utilities. Of those responding, one declined to participate citing time and workload constraints, 2 had not done any lining projects and had no first-hand experience to offer, 15 completed the questionnaire, and 4 said that they had someone working on the questionnaire. Of the 15 utilities that completed questionnaires 3 were from Canada, 1 was from Sweden, 2 were from the U.K., 1 was from Norway, and 8 were from the U.S. The respondents said that they had used cement-mortar lining (CML), epoxy, polyurethane, and cured-in-place pipe (CIPP) to solve various hydraulic and water quality problems. Tables 3.1 and 3.2 summarize the lining materials and impacts of lining pipes, respectively, as reported by the respondent water utilities. The most frequent reason for using a lining product was hydraulic capacity, followed by tuberculation and red water. Structural integrity and chlorine residual were also mentioned as reasons for lining water mains. Both CML and epoxy lining were effective at eliminating red water and rust particles from delivered water. Iron and manganese concentrations were effectively reduced to within applicable drinking water standards using CML, epoxy, and polyurethane. CIPP and replacement polyethylene (PE) pipe were reported to correct structural problems without any adverse water quality impacts. 17

52 18 Impacts of Lining Materials on Water Quality Respondent Table 3.1 Lining material summary Cementmortar lining Epoxy Polyurethane Polyethylene CIPP Other 1 Unspecified solventbased coating 2 X 3 X 4 X 5 X 6 X 7 X 8 X 9 X 10 X X 11 X 12 X X 13 X X 14 X 15 X X X Table 3.2 Reported impacts from lining materials Respondent Tuberculation Hydraulic capacity Red water Chlorine residual Structural Positive impacts Negative impacts 1 X Coating did not fix leaking 8-million gallon storage tank 2 X X Lining eliminated red water problem 3 X X Fire flow improved 128% 4 X X X Red water and rust particles were eliminated; C-factor improved 5 X ph increased after lining, especially in low demand areas ph improved with time and with increased demand 6 No negative impact from one epoxy product "Off-gassing" from another epoxy product (continued)

53 Chapter 3: Survey of Current Practices and Water Quality 19 Table 3.2 (Continued) Respondent Tuberculation Hydraulic capacity Red water Chlorine residual Structural Positive impacts Negative impacts 7 Low-level VOCs in 10-million gallon storage tank 8 X No water quality impacts from PE material 9 X X X Red water eliminated 10 X X X Eliminated red water 11 X X X Chlorine residual improved in summer conditions 12 X Iron and manganese reduced in concentration to within standard 13 Study provided a physical condition assessment of CML, epoxy, and polyurethane linings 14 X Improved C-factor and hydraulic capacity Reduced number of customer complaints 15 X X X Iron and manganese concentrations reduced Increased ph with CML High ph required months of flushing No significant reduction in iron or other water quality improvements Customers complained about traffic, pedestrian disruption, and wasting water No negative impacts The most frequently reported water quality impact from CML was an increase in ph. In one case the elevated ph required months of flushing before the newly lined main could be put back into service. In other cases the elevated ph occurred immediately after lining and was greatest in areas with low water demand. After time and increased water demand, the ph impact subsided. It is possible that calcium and aluminum were not noted in the questionnaire because they are not routinely monitored. One utility reported that CML improved C-factor but did not reduce the iron concentrations. The lining project did generate complaints of traffic and pedestrian disruptions. Some customers complained that the flushing required after CML was wasteful of water during a period when the utility had imposed water use restrictions on its customers. In most cases where epoxy lining had been used it was reported that no negative impacts were observed. In a few cases it was reported that VOCs and off-gassing from the coating material had occurred.

54

55 CHAPTER 4 OVERALL EXPERIMENTAL DESIGN OF BENCH-SCALE LEACHING STUDIES OBJECTIVES The objectives of this research are to measure the initial impact of cement mortar, polyurethane, and epoxy lining materials on drinking water quality parameters including ph, ammonia, alkalinity, total organic carbon (TOC), disinfectant residual, hardness, total and dissolved solids, trihalomethanes (THMs), five regulated haloacetic acids (HAA5), semivolatile organic compounds (SVOCs), metals and nonmetals, and aesthetic quality as measured by odor. The lining materials tested were approved for potable water through certification by NSF under protocol ANSI/NSF-61. The experimental procedure was designed to compare water quality changes for selected materials and water types with and without lining material coupons present. This approach was necessary because selected water quality parameters, such as decrease in disinfectant residual, would happen whether lining materials were present or not. In such a case, the rate of change in comparison to the control is necessary to evaluate the impact of the lining material. The leaching protocol was adapted from ANSI/NSF-61. Water changes were performed periodically with variable days of contact time between water changes. This represented water that undergoes varying times of stagnation in pipes and allows the assessment of contact time as a variable. In addition, testing was conducted to determine if industry-standard pipe corrosion preventative additives (CPAs) had any impact on the water quality effects of newly installed cement mortar lining materials. This testing, which used materials and methods similar to those reported cement mortar leaching tests, evaluated three different corrosion preventative chemical regimes: Orthophosphate (OP) at 1.0 mg/l as P Polyphosphate (PP) at 1.0 mg/l as P OP with zinc (OPZn) at 1.0 mg/l as P and 0.3 mg/l as Zn Samples with no CPA added were also tested and served as a comparative baseline. The tests performed to evaluate these potential impacts were as follows: ph Metals including total hardness (calcium [Ca] and magnesium [Mg]) Alkalinity These parameters were evaluated because they produced the most significant impacts during the earlier testing and they have a substantial influence on drinking water quality. CEMENT-MORTAR COUPON FABRICATION The fabrication of the cement-mortar coupons was performed in two phases, prototype and final production. The prototype phase was an opportunity to test out the cement-mortar recipe, mold design, mold-filling process, and determine if appropriately sized coupons would result. There 21

56 22 Impacts of Lining Materials on Water Quality were two iterations of mold design, fabrication, and filling completed during this phase. The original mortar recipe, meeting the requirements of AWWA C602, Cement Mortar Lining of Water Pipelines in Place 4 In. and Larger, and recommended by a mortar-lining contractor (Heitkamp, Inc.), performed successfully in all cases. The coupon size also proved to be appropriate. The final production recipe, design, and process that resulted are described in more detail below. Table 4.1 Cement-mortar recipe Ingredient Amount Sand (medium grade, washed) 1:1 ratio with cement Cement (Type II Portland cement) As needed Water (tap) 0.43 weight fraction of cement weight CEMENT-MORTAR COMPOSITION The cement-mortar composition was determined by following the guidelines listed in the AWWA publication listed above and conversing with a representative from Heitkamp, Inc. who had extensive experience in the cement-mortar pipe-lining industry. The composition shown in Table 4.1 was used successfully throughout the prototype and production phases. MOLD DESIGN The two production molds consisted of a plywood base that was covered with a thin Teflon sheet upon which a 1-centimeter (cm) high grid was affixed. Each wooden member of the grid was wrapped in Teflon film. Each mold grid produced 50 of the 8 cm x 8 cm x 1 cm coupons. The design allowed for easy filling and mortar leveling during the fabrication process. A photograph of the mold is provided in Figure 4.1. Figure 4.1 Photo of cement-mortar coupon mold

57 Chapter 4: Overall Experimental Design of Bench-Scale Leaching Studies 23 CEMENT COUPON PRODUCTION PROCESS The coupon production process consisted of the following steps: 1. Mortar mixing the mortar was mixed in an electric drum mixer using the composition shown in Table 4.1. Dry ingredients were mixed initially, followed by the slow addition of water. The mortar was mixed for approximately 15 minutes. Two batches of mortar mix were produced, one for each mold assembly, over a 90-minute time-frame. 2. Mortar placement the mortar was transferred to each grid segment using aluminum scoops. The mold was placed on a vibrating table to aid in distributing the mortar. The mortar was initially leveled using a wooden screed. Low spots were filled with fresh mortar and final leveling was accomplished using steel cement finishing trowels. 3. Coupon curing the filled molds were placed in a humid room and covered with one layer of polyethylene sheet to preserve moisture during the curing process. The coupons were cured for 24 hours. 4. Coupon removal the molds were removed from the humid atmosphere and examined for anomalies. Each grid member was then removed, thus allowing the coupon to release. The coupons were easily removed due to the Teflon surfaces. All coupons were found to be acceptable and were transported back to the testing laboratory so that they could be immersed in the test waters. Fabrication of cement-mortar coupons used in CPA testing was also prepared using the procedures outlined above. POLYURETHANE COUPON PREPARATION A single polyurethane was tested. Sand-blasted and roughened 8-cm x 8-cm glass coupons were prepared. Coating of the glass coupons with polyurethane was achieved with the cooperation and coordination of Madison Chemical Industries, Inc., Ont., Canada. Clean sandblasted 8-cm x 8-cm glass coupons were coated on all sides with polyurethane in accordance with procedures recommended by the manufacturer. This polyurethane consists of two components in a 1:1 mixture that was applied using heated airless plural spray equipment. After the recommended 48-hour curing time, the leaching tests for the polyurethane-coated coupons were begun. EPOXY COUPON PREPARATION Coating of the coupons with epoxy was achieved with the cooperation of Cohesant Technologies and RLS Solutions. Sand-blasted and roughened 8.1-cm x 8.7-cm x 0.52-cm glass coupons were shipped to Cohesant Technologies. The coupons were encapsulated with epoxy and shipped back to the Virginia Tech laboratory. Coupons were cured for 60 hours before leaching tests for the epoxycoated coupons began.

58 24 Impacts of Lining Materials on Water Quality MATURE PIPE SAMPLE For mature pipe leaching tests, a 5-year old epoxy-lined copper pipe was obtained. The mature epoxy-lined copper pipe sample (26.70 cm in length, 3.63 cm interior diameter) was cut from an active service line in British Columbia and was obtained from RLS Solutions as a comparison. The pipe was in service from April 2003 to May 2007 and carried chlorinated water at 140 C. The epoxy layer was a yellow-green color, smooth in finish, and with no visible tears or holes. No scale or biofilm could be seen on the surface of the lining. For this experiment the pipe was filled with the same chlorinated reference water as prepared for the other samples. For the leaching study, the pipe sample was sealed on one side with a Teflon-coated rubber stopper and filled with the same ph chlorinated water prepared for the coupon leaching study. The pipe was filled headspace-free, capped with another Teflon-coated rubber stopper, and stored in the dark adjacent to the test vessels. The pipe was approximately 1.5 inches (in.) in diameter, less than half the diameter of the pipe the epoxy coupons were meant to simulate, so it had a much greater surface area to volume ratio. When making comparisons between new and mature epoxy, the concentrations were normalized for surface area. The total volume of the pipe was cubic centimeters (cm 3 ) and the surface area was square centimeters (cm 2 ). PRETREATMENT OF TEST COUPONS The cement, polyurethane-coated, or epoxy-coated coupons were washed and prepared as specified in ANSI/NSF-61. The coupons was exposed to a 200-mg/L Cl 2 solution for 30 minutes and rinsed. TEST VESSELS LEACHING OF LINING MATERIALS The prepared coupons of lining materials were placed in a fully enclosed container to simulate contact with water in an enclosed pipe (Figure 4.2). Triplicate all-glass immersion test vessels of approximately 3-liter (L) volume and containing coupons were prepared for each water type. The ratio of volume of water and lining material surface area mimics a 4-inch diameter pipe. Coupons were held upright by perfluoroethylene support racks. A control for each water type, that contained no coupons, was also prepared. All vessels were maintained headspace-free throughout the experiment. The temperature was between 21 and 23 C for cement and polyurethane, and between 19 and 23 C for epoxy. During the leaching procedure, all test vessels were stored in the dark. Figure 4.2 Immersion test apparatus with coupons

59 Chapter 4: Overall Experimental Design of Bench-Scale Leaching Studies 25 The test vessels used for the CPA evaluation were different from those used during the leaching tests. They were square, plastic, 1-L containers with open tops. The use of plastic containers was acceptable because there were no organic chemical tests planned as part of the evaluation. The vessels were filled with 410 milliliters (ml) of prepared sample water and one coupon was immersed in the water. The coupon was supported at an angle by a small glass cylinder. This ensured that all coupon surfaces were in contact with the water. The surface area to volume ratio created by this procedure was 0.39, which is the same ratio created by the original setup using six coupons and nearly 3 L of water. A photograph of the test setup is provided in Figure 4.3. There were duplicate test vessels used for each water type (control, OP, PP, and OPZn), resulting in eight test vessels being used. The vessels were stored in the dark except when being sampled and having the water changed. Each was covered with plastic wrap to minimize evaporation. TEST WATERS FOR LEACHING OF LINING MATERIALS Four water types were tested for cement-mortar lining and three water types were tested for epoxy and polyurethane lining so as to investigate the different disinfectants that are typical of U.S. drinking waters. A total of 11 L of each water sample were prepared for each water change/sampling date. Additional tests were conducted for cement-mortar lining with corrosionresistance additives. The following are the water types used in testing cement-mortar lining. 1. ph = 8.0, low alkalinity/hardness, no disinfectant 2. ph = 6.5, low alkalinity/hardness, 2.0 mg/l chlorine 3. ph = 8.0, low alkalinity/hardness, 2.0 mg/l chlorine 4. ph = 8.0, low alkalinity/hardness, 4.0 mg/l monochloramine The following three types of water were used for testing of epoxy and polyurethane lining: 1. ph = 8.0, low alkalinity/hardness, no disinfectant 2. ph = 8.0, low alkalinity/hardness, 2.0 mg/l chlorine 3. ph = 8.0, low alkalinity/hardness, 4.0 to 6.0 mg/l monochloramine Figure 4.3 Corrosion preventative experimental testing setup

60 26 Impacts of Lining Materials on Water Quality The composition of the ph 8, low alkalinity/low hardness reference water is presented in Table 4.2. Reference water was prepared by adding salts to Nanopure reagent water. The same low alkalinity, low hardness water reference or base water recipe was used for the CPA experiments. The reference water was augmented with CPA chemical and adjusted to a ph 8.0 for CPA samples or only ph-adjusted when used as a control within the testing regime. There were three CPA water types created for use in the test and their recipes are provided in Table 4.3. The CPA dosage levels are representative of corrosion prevention programs used throughout the potable water treatment industry. A total of 11 L of sample was prepared for each water change/sampling date. Chlorinated test water was prepared by adding NaOCl. Chloraminated test water was prepared by first adding NH 4 OH while being continuously stirred. After 5 minutes of mixing, NaOCl was quickly added and the water was shaken vigorously for 30 seconds. ph was adjusted as necessary by adding hydrochloric acid. The chloraminated water ph fluctuated between 8 and 9, depending on the amount of ph adjustment after addition of ammonium hydroxide. PROTOCOL FOR WATER CHANGES AND ANALYSES The leaching protocol was adapted from ANSI/NSF-61, which evaluates the leaching of organic and inorganic components from materials in contact with drinking water. The leaching test was performed for up to 30 days and the frequency for measuring different water quality parameters is presented in Table 4.4. Water was changed on the days indicated and subsamples of the water were obtained to measure selected analytes as indicated in the table. The sample water was tested for ph, TOC concentration, disinfectant residual, ammonia concentration as N-NH 3, hardness as combined calcium and magnesium concentrations, alkalinity (as CaCO 3 ), temperature, and ph on days when the sample water was changed. Total solids (TS), odor, THMs, HAA5, and SVOCs were determined on days 1, 4, 9, and 14. Table 4.2 Reference water composition Chemical Molarity Concentration of chemical (mg/l) MgSO 4 * NaHCO CaSO 4 * CaC1 2 * Na 2 SiO 3 * x KNO x HCl *Added as hydrates

61 Chapter 4: Overall Experimental Design of Bench-Scale Leaching Studies 27 Table 4.3 Test water recipes for CPA impact testing Water type CPA added CPA addition form Resulting CPA concentrations Base (control) None NA NA Orthophosphate (OP) Na 3 PO 4 Stock aqueous solution 1.0 mg/l as P Polyphosphate (PP) Na 8 P 6 O 19 Stock aqueous solution 1.0 mg/l as P OP and zinc (OPZn) Na 3 PO 4 and ZnCl 2 Both stock aqueous solutions 1.0 mg/l as P 0.3 mg/l as Zn Table 4.4 Typical frequency and quality control for chemical water quality parameters for benchscale testing of lining materials Parameter Measurement frequency (day) TOC 1, 2, 4, 9, 11, 14, 15, 19 Disinfectant residual 1, 2, 4, 9, 11, 14, 15, 19, 21, 30 Ammonia 1, 2, 4, 9, 11, 14, 15 Hardness 1, 2, 4, 9, 11, 14, 15, 19, 21, 30 Alkalinity 1, 2, 4, 9, 11, 14, 15, 19, 21, 30 ph 1, 2, 4, 9, 11, 14, 15, 19, 21, 30 Temperature 1, 2, 4, 9, 11, 14, 15, 19, 21, 30 Total solids 1, 4, 9, 14 Odor 1, 4, 9, 14 THMs 1, 4, 9, 14 HAA5* 1, 4, 9, 14 SVOCs* 1, 4, 9, 14 Elements 1, 2, 4, 9, 14, 15 * A single sample was analyzed that is a composite from the triplicate sample for each pipe lining/water combination.

62 28 Impacts of Lining Materials on Water Quality The protocol for water changes and chemical analysis for CPA testing was the same as the original testing except that the test period lasted 19 days instead of 30 days. The water changes occurred on days 1, 2, 4, 9, 11, 14, 15, and 19 with samples taken for each of the tests listed above at each water change. The water was agitated before sampling to improve the homogeneity of the obtained sample. The test methods used were identical to those used in the leaching testing. The shorter testing period was selected because the previous results indicated, for the three tests being performed, that little change occurred after day 19. The specific sampling protocols and measurement methods used to determine the water quality parameters are presented in Table 4.5. Details of methods and chemical measurements of all analyses are given in Appendix B. Parameter Table 4.5 Measurement method and sampling procedures Standard method* Instrumentation Sample container/ replicates Preservation/ storage TOC SM 5310c TOC analyzer Glass/3 In dark, 4 C ± 2 C Disinfectant residual Maximum holding time 28 days SM 2350 HACH colorimeter Glass/3 4 C ± 2 C <24 hours Ammonia SM 4500 HACH colorimeter Glass/3 H 2 SO 4 to ph <2, 7 days 4 C ± 2 C Hardness SM 2340b ICP-MS Glass/3 HCl or H 2 SO 4 to 6 months ph <2, 4 C ± 2 C Alkalinity SM 2320 HACH colorimeter Glass/3 4 C ± 2 C 14 days ph SM 2310 ph meter Glass/3 N/A Immediately Temperature SM 2550 Thermometer Glass/3 N/A Immediately Total solids SM 2540 Evaporation Glass/3 4 C ± 2 C 7 days Odor SM 2170 FPA Odor-free 4 C ± 2 C >2 days glass/1 THMs SM 6232D Purge and trap Glass/3 Ascorbic acid 7 days (20 mg/40 ml) 4 C ± 2 C HAA5 SM 6251 Extraction Glass/1 Dechlorinate, 9 days 4 C ± 2 C SVOCs SM 6410 GC/MS Glass/1 Dechlorinate, 4 C ± 2 C 7 days until extraction Metals/elements SM 3125 ICP-MS Glass/3 HNO 3 to ph <2, 4 C ± 2 C 6 months * SM = Standard Methods for the Examination of Water and Wastewater Quantity of samples is number of samples per triplicate lining material and water types; 3 = each replicate will be measured; 1 = composite of triplicates Hardness = sum of calcium and magnesium concentrations

63 Chapter 4: Overall Experimental Design of Bench-Scale Leaching Studies 29 STATISTICAL ANALYSIS Statistical analysis of data was performed using both a linear and logarithmic regression analysis and analysis of variance (ANOVA) using the statistical program R version 2.6.2, with an α = A p value was generated where if the p value <0.05 the data are statistically significant, and if the p value >0.05 the data are not statistically significant. Logarithmic analysis could not always be used due to the negative values, which cannot be analyzed logarithmically, and therefore, decreased the data points available. Paired t tests were most commonly used as a comparison between the samples and the corresponding controls. Where triplicate samples were collected, the mean value of the samples was used for the t test. An alpha value of 0.05 was used for statistical significance. Where appropriate, error bars are drawn on the graphs to indicate the standard deviation between the triplicate samples. No bars are drawn for controls or pipe samples because only one sample was taken.

64

65 CHAPTER 5 TESTING OF CEMENT-MORTAR LINING TEST RESULTS Leaching of cement-mortar lining was tested using four types of water as outlined in the methodology discussed in Chapter 4. Test results of water quality impacts are discussed in this chapter. EFFECT OF CEMENT-MORTAR LINING ON INORGANIC WATER QUALITY PARAMETERS ph The initial water quality condition, i.e., either ph 6.5 or ph 8 with no disinfectant, chlorine, or monochloramine, did not affect the final ph after the cement-mortar coupons were placed in the test vessel (Figure 5.1). The control waters, which were test vessels without cement-mortar coupons, were able to maintain their target ph values of 8 or 6.5. This is similar to results reported by Douglas et al. in As shown in Figure 5.1, the ph for all four test water conditions increased from to during the leaching tests. ph was measured on days that the water in the vessels was changed; thus, the length of time that the coupons were in contact with the water is reflected by the days between water changes. For sampling days 1, 2, 4, and 9, the ph was consistently about 12. The water change on day 11 (48-hour contact time) and 4 previous water changes resulted in a ph drop to about After a contact time of 72 hours, when the water was changed on day 14, the ph was still about When the water was changed 24 hours later on day 15, it declined sharply to about ph 10. This is likely due to a combination of the short contact time, the previous leaching, and curing of the cement. Continued contact of cement and water resulted in an elevated ph between 11.4 and 11.6 for the 30-day duration of the test. Figure 5.1 ph as a function of time 31

66 32 Impacts of Lining Materials on Water Quality Alkalinity The target concentration for the reference water was 35 mg/l and the average for all the controls for all sampling days was 35.5 mg/l, indicating excellent agreement between prepared and measured concentrations. Similar to the ph data, there were no impacts from the initial water quality condition, i.e., either ph 6.5 or 8 with no disinfectant, chlorine, or monochloramine, on the amount of alkalinity released from the cement mortar. All water conditions performed similarly as shown in Figure 5.2. Also similar to the ph data, there was a substantial drop in alkalinity between the water changes on days 9 and 11. After this time, the alkalinity leveled off and remained about 100 mg/l for all 4 test waters for days 19, 24, and 30. Also, after day 9, precipitates no longer appeared. The high alkalinity and ph is likely due to the dissolution of Ca(OH) 2, which is a byproduct of the cure reactions of cement. Hardness The following data are typical of those observed for the metals released from new cement-mortar coupons for all ph and disinfectant combinations of the experiment (Figure 5.3). As was seen in previous reports, concrete mortar contributes a significant amount of calcium and magnesium to the water. The calcium concentration is much greater than the magnesium concentration and in the range of hundreds of mg/l for the first few days. However, after the concrete cured substantially (around day 9), the concrete actually began to sorb calcium, as the values were below that of the controls. The magnesium behaved differently, and is consistently sorbed by the concrete before or after substantial curing. The controls contained calcium and magnesium because these cations were added in the reference water at 11.5 mg/l and 8 mg/l, respectively. The high calcium levels for the early days of cement mortar in contact with water are likely due to the dissolution of Ca(OH) 2, which is a byproduct of the cure reactions of cement. Figure 5.2 Alkalinity as a function of time for all four test waters and the controls in contact with cement mortar

67 Chapter 5: Testing of Cement-Mortar Lining Concentration (mg/l) Calcium, ph 8, 2 mg chlorine Calcium Control Magnesium, ph 8, 2 mg chlorine Magnesium Control Test Day Figure 5.3 Calcium and magnesium hardness as a function of time for ph 8, 2 mg/l Cl 2 water (water 3) and its control in contact with cement mortar The magnesium and calcium hardness data are presented in Tables 5.1 through 5.4. Release and sorption of these metals followed a similar pattern for all four test waters. For days 1, 2, 4, and 9 there was a great release of calcium with simultaneous sorption of magnesium. Sorption of magnesium continued throughout the 30 days of the test. Sorption of calcium began on day 15 and continued to the end of the test. Elemental Analysis Of the metals and nonmetals tested, the leachate concentrations of chromium and aluminum were consistently higher than their controls (Figure 5.4) for all ph and disinfectant combinations. 450 Concentration (µg/l) Aluminum, ph 8, 2 mg chlorine Aluminum Control Chromium, ph 8, 2 mg chlorine Chromium Control Test Day Figure 5.4 Aluminum and chromium as a function of time for ph 8, 2 mg/l Cl 2 water (water 3) and its control in contact with cement mortar

68 34 Impacts of Lining Materials on Water Quality Table 5.1 Calcium and magnesium concentrations (as CaCO 3 ) for cement mortar in contact with no disinfectant water and data for the corresponding control water (The delta metal concentration is for sample control) Calcium and magnesium hardness, mg/l as CaCO 3 Day Ca in cementmortar water Ca control Δ Ca Mg in cementmortar water Mg control Δ Mg Table 5.2 Calcium and magnesium concentrations (as CaCO 3 ) for cement mortar in contact with ph 6.5, chlorine water and data for the corresponding control water (The delta metal concentration is for sample control) Calcium and magnesium hardness, mg/l as CaCO 3 Day Ca in cementmortar water Ca control Δ Ca Mg in cementmortar water Mg control Δ Mg

69 Chapter 5: Testing of Cement-Mortar Lining 35 Table 5.3 Calcium and magnesium concentrations (as CaCO 3 ) for cement mortar in contact with ph 8, chlorine water and data for the corresponding control water (The delta metal concentration is for sample control) Calcium and magnesium hardness, mg/l as CaCO 3 Day Ca in cement-mortar water Ca control Δ Ca Mg in cementmortar water Mg control Δ Mg Table 5.4 Calcium and magnesium concentrations (as CaCO 3 ) for cement mortar in contact with ph 8, chloraminated water and data for the corresponding control water (The delta metal concentration is for sample control) Calcium and magnesium hardness, mg/l as CaCO 3 Day Ca in cement-mortar water Ca control Δ Ca Mg in cementmortar water Mg control Δ Mg

70 36 Impacts of Lining Materials on Water Quality The release of aluminum is from the available aluminum silicates in uncured cement; aluminum becomes bound in the mortar matrix as curing proceeds. Aluminum in the initial period exceeded the EPA secondary standard of 0.2 mg/l (EPA 2003). The chromium levels did not exceed the EPA maximum contaminant level (MCL) of 0.1 mg/l. As with hardness, there is a significant changeover when the concrete cured substantially (about day 9), after which the quantities of aluminum and chromium became negligible and within EPA limits (although still slightly higher than the corresponding controls). The concentrations of other elements monitored, including sodium, phosphorus, sulfur, chlorine, iron, manganese, nickel, zinc, potassium, and copper, were similar in the test water and the controls indicating that they were not leached from the cement mortar. Total Solids The total solids data are present in Table 5.5. These data represent the sum of dissolved and suspended solids from dissolution of solids from the cement coupons and also the approximately 160 mg/l dissolved solids from the reference water. Measurement of dissolved and suspended solids in selected samples indicated that the majority of the solids were dissolved with samples containing from zero to only few percent suspended solids. The photograph in Figure 5.5 suggests that suspended solids were present through day 9. The apparent solids in all test vessels were gel-like in nature and dissipated on physical contact. The dissolved solids concentrations in the day 1, 4, and 9 samples were in the range of 1,000 to 1,700 mg/l. Similar to data trends for ph, alkalinity, and hardness, the total solids concentrations declined substantially after day 9 when the cement cured. Although the later concentrations of total solids declined to 400 to 700 mg/l, these values likely exceeded the EPA secondary maximum contaminant level (SMCL) for total dissolved solids (TDS) of 500 mg/l. Table 5.5 Total solids in water in contact with cement-mortar coupons Total solids (mg/l) ph 8 ph 6.5 ph 8 ph 8 Day No disinfectant chlorine chlorine monochloramine 1 1,500 1, , ,740 1,410 1,760 1, ,210 1, ,

71 Chapter 5: Testing of Cement-Mortar Lining 37 Figure 5.5 Test vessel containing cement-mortar coupons and precipitate that was typical of all test conditions up through day 9 Disinfectant Residual Figures 5.6 through 5.8 show the disinfectant loss in mg/l as Cl 2 for the chlorinated waters at ph 6.5 and 8 and also the chloraminated water. The water without disinfectant treatment is not shown because there was no Cl 2 in the initial test water and the concentration did not change over the course of the study. The residuals in the chlorinated samples were consistently lower than the demand in both chlorinated waters indicating that the cement mortar caused increased consumption of chlorine. The amount of chlorine demand in the presence of cement mortar decreased over time, which implies that the chlorine demand should eventually level off, but this did not happen within the 30-day test period. Chloramines were more stable than free chlorine in the presence of cement mortar (Figure 5.8). Chloramines exposed to cement-mortar coupons had a similar moderate rate of consumption on days 1, 2, 4, and 9, after which chloramines were stable in the presence of cement mortar. Figure 5.9 shows the disinfectant decay rate in milligrams per liter per day (mg/l/day) for ph 8, chlorine water, and ph 8, monochloramine water. The ph 6.5 chlorine water demonstrated a trend similar to the ph 8 chlorine water, and is not shown. Data for the controls are shown as both chlorine and monochloramine decay spontaneously even without the presence of cement mortar. The decay rate for both chlorine and monochloramine was enhanced by the presence of cement mortar. This higher rate of decay was maintained until test day 9, after which the cement mortar performed more similarly to the controls. This pattern is similar to that observed for ph and alkalinity.

72 38 Impacts of Lining Materials on Water Quality Disinfectant Residual (as mg/l Cl 2) Residual Demand Test Day Figure 5.6 Chlorine decay, showing residual disinfectant and consumed disinfectant as a function of time, corrected for decay in the controls (water: ph 6.5, 2 mg/l Cl 2 ) in contact with cement mortar Disinfectant Residual (as mg/l Cl 2) Residual Demand Test Day Figure 5.7 Chlorine decay, showing residual disinfectant and consumed disinfectant as a function of time, corrected for decay in the controls (water: ph 8, 2 mg/l Cl 2 ) in contact with cement mortar

73 Chapter 5: Testing of Cement-Mortar Lining 39 Disinfectant Residual (as mg/l Cl 2) Residual Demand Test Day Figure 5.8 Monochloramine decay, showing residual disinfectant and consumed disinfectant as a function of time, corrected for decay in the controls (water: ph 8, 4 to 5 mg/l NH 2 Cl) in contact with cement mortar 2.0 Disinfectant Decay Rate (mg/l/day) ph 8, 2 mg/l Chlorine ph 8, 4-5 mg/l Chloramine ph 8, 2 mg/l Chlorine, Control ph 8, 4-5 mg/l Chloramine, Control Test Day Figure 5.9 Disinfectant decay rate based on contact time of cement mortar with disinfectant

74 40 Impacts of Lining Materials on Water Quality Ammonia Data for ph 8 monochloramine water are shown in Figure Ammonia is a decay product of monochloramine and thus, only this water type was expected to have increased levels of ammonia. The other waters and controls, with chlorine or no disinfectant, contained lower and similar concentrations of ammonia. As shown in Figure 5.10, water with ph 8 and monochloramine, with the cement mortar, had similar concentrations of ammonia as its control and thus, cement mortar had little effect on monochloramine decay. Total Organic Carbon (TOC) Data for TOC are shown in Figures 5.11 and The background level of TOC fluctuated between 0.2 and 0.4 mg/l with a mean of 0.26 mg/l. These concentrations are typical of reagent water produced for laboratory deionization/carbon filtration systems used to produce purified water. The cement water generally contributed <0.5 mg/l TOC to the test waters through day 9 of the experiment, and then leveled off to background levels similar to the controls. There appears to be variation in TOC leaching and this may be due to the different water quality conditions or to variability in the materials used to make the coupons. The highest TOC levels were observed in the first days of the leaching experiment. These can also be seen in later discussions that THM and HAA5 concentrations are also highest in the first days of the leaching experiment. Ammonia 1.80 Ammonia (mg/l as N) Test Day Water 4 Control 4 Figure 5.10 Ammonia as a function of time for ph 8 water dosed with monochloramine in contact with cement mortar

75 Chapter 5: Testing of Cement-Mortar Lining 41 TOC (mg/l) ph 8, no disinfectant ph 6.5, chlorine ph 8, chlorine ph 8, monochloramine Controls (average) Test Day Figure 5.11 TOC as a function of time for water type in contact with cement mortar TOC (mg/cm 2 /day) 8.00E E E E E E E E E+00 ph 8, no disinfectant ph 8, no disinfectant, control ph 6.5, chlorine ph 6.5, chlorine, control ph 8, chlorine ph 8, chlorine, control ph 8, monochloramine ph 8, monochloramine, control Test Day Figure 5.12 TOC leached per unit time and surface area when in contact with cement mortar

76 42 Impacts of Lining Materials on Water Quality Figure 5.12 represents the amount of TOC leached per unit time and surface area of cement mortar. While the trend is similar to that of TOC concentration vs. time for the initial days of contact of cement mortar with water, there is a distinct difference. The rate of TOC release per unit time indicates that the amount of TOC leached in a short contact period is substantial (i.e., between day 9 and 11 and between day 14 and day 15). Semivolatile Organic Compounds (SVOCs) No specific SVOCs were extracted and identified by gas chromatography/mass spectrometry (GC/MS) in the cement-mortar leachates, which generally contained <0.5 mg/l TOC. The likely explanation for the TOC is that it is comprised of humic and fulvic acids present in the sand. Trihalomethanes (THMs) The data indicate that THMs did not form within the first 14 days of the leaching test when cement-mortar coupons were in contact with chlorinated water (data not shown). No THMs were detected in the no disinfectant water or the chloraminated water. THM Formation Potential and Sorption to Cement-Mortar Lining The objective of this testing was to evaluate the THM formation potential and the THM sorption properties of the cement-mortar lining material. Materials and Methods Cement-mortar coupons were fabricated as previously described and cured for 24 hours before being immersed in water. The reference water used in this experiment was water 1 (ph 8, low alkalinity and no disinfectant) as described earlier in this chapter. THM Formation In order to create water with TOC resulting from contact with the cement mortar only, coupons were placed in immersion test vessels (ITVs) used for the leaching experiments. The ph-adjusted base water was used to fill the ITV until it was headspace-free. The surface area to volume ratio of 0.39 created was identical to that used during the leaching test. The coupons were in contact with the base water for 48 hours after which the water was removed and placed in 500-ml glass amber bottles. No headspace was allowed in the bottles and they were stored at 4 C until tested for TOC or used within the test itself. The TOC was determined to be mg/l TOC. This water was diluted by a factor of four with base water to a resulting TOC level of 3.34 mg/l. This TOC level is higher than what would normally be found in a distribution system but ensured there was ample TOC available for the THM formation reactions. This was then chlorinated using sodium hypochlorite to a free chlorine content of mg/l. This again is higher than that which would normally be evident in a distribution system, but it ensured ample chlorine would be available for the THM reactions. The high TOC and high chlorine conditions might also be present in recently lined water mains that are being disinfected before being placed back into service. The chlorinated water was quickly poured into the 20-ml amber glass sample vials. The samples supported the test matrix shown in Table 5.6. The samples were stored at room temperature during the THM formation period.

77 Chapter 5: Testing of Cement-Mortar Lining 43 Table 5.6 THM formation test matrix Number of samples Hours after chlorination Test 3 0 Chloroform concentration 3 24 Chloroform concentration 3 48 Chloroform concentration 3 72 Chloroform concentration 4 0, 24,48, 72 Free chlorine (1 test for each sampling time) At each of the times shown in Table 5.6, the sample was injected with 100 microliters (μl) of concentrated sodium thiosulfate, which stopped the THM formation reaction, and was then stored at 4 C until analyzed. Additionally, one sample vial was tested for free chlorine at each sampling time to track the chlorine residual. The THM formation samples were then analyzed using the same methods outlined in the leaching test procedures (purge and trap, GC). THM Sorption The two sectioned coupons were separated into 12 segments that measured 1.0 cm thick x 1.25 cm wide x 8.0 cm long. The variability among segments was 5% for thickness and width and near 0% for length. The segments were then placed in a square, plastic jar test container and covered with base water. Because the purpose of the immersion was only to condition the coupons for use in sorption tests, the amount of water was not critical. The vessel was covered with plastic wrap to prevent evaporation and placed in a dark cabinet for 24 days. This longer immersion time ensured that all entrapped air was released from within the cement segment and would not alter the THM sorption results. The segments were rinsed with reagent water before being immersed in the chloroform-spiked water. A 1-L batch of base water was obtained within a glass bottle. A standard aqueous solution of chloroform was added to the base water resulting in a chloroform concentration of micrograms per liter (μg/l). The 12 cement-mortar segments were each placed into an amber 40-ml glass vial. The chloroform-spiked water was added to each of the vials such that no headspace was present. The vials were capped and stored at room temperature. The samples were prepared to support the test matrix shown in Table 5.7. Table 5.7 THM sorption test matrix Number of samples Hours after exposure to mortar segment Test 3 8 Chloroform concentration 3 24 Chloroform concentration 3 48 Chloroform concentration 3 72 Chloroform concentration 3 (Controls) 24,48, 72 Chloroform (1 test for each sampling time)

78 44 Impacts of Lining Materials on Water Quality The water samples and controls were tested for chloroform concentration using the same THM analysis methods previously described. Results: THM Formation Potential. The THM formation results are presented graphically in Figure The data presented are the average of the triplicate samples analyzed. Figure 5.13 shows that the TOC produced by the immersed cement-mortar coupons does have the potential to form chloroform, a common THM. At 72 hours after chlorination, approximately 34 μg/l of chloroform had formed. Based on the curve shown, the chloroform concentration may have reached between 35 and 40 μg/l given additional time. The base water at t = 0 had a chloroform concentration of approximately 3.5 μg/l, and this value should be subtracted from the total chloroform concentration. The chlorine residual declined as expected as the chloroform reaction progressed, dropping from 10.3 mg/l to 7.8 mg/l. Although the test conditions were a worst-case scenario, the results indicate that the cement-mortar lining material can generate TOC that has a THM formation potential of at least 30 μg/l. Results: THM Sorption. The THM sorption results are provided in Figure 5.14 and show that the cement mortar does adsorb chloroform. Figure 5.14 shows loss of chloroform with duration of exposure to the cement mortar. The 40 ml of vial volume is reduced to 30 ml due to the addition of the 10-cm 3 coupon segment. The segment has a surface area of approximately 38 cm 2 giving a surface area to volume ratio of This ratio is 3.25 times higher than the 0.39 ratio used for other cement-lining testing, which was meant to simulate the ratio for a 4-inch diameter pipe. The adsorption rate shown in Figure 5.14 is adjusted for a ratio of Because the rate is linear during the entire test period, CHCl 3 Concentration (µg/l) Chloroform Formed Chlorine Residual Cement mortar exposed water with an average TOC = 3.44 mg/l was used for this test Chlorine Concentration (mg/l) Hours after Chlorination 0.00 Figure 5.13 Chloroform concentration and chlorine residual as a function of time after chlorination of sample waters in contact with cement mortar

79 Chapter 5: Testing of Cement-Mortar Lining 45 Loss in CHCl 3 Concentration (µg/l) Initial CHCl 3 concentration was µg/l Chloroform Adsorbed Adjusted Chloroform Adsorbed Hours after Exposure to Cement Mortar Figure 5.14 Loss in chloroform concentration as a function of time after exposure to cement-mortar coupon segments a maximum sorption level cannot be ascertained. At 72 hours the adjusted adsorbed amount results in a chloroform concentration loss of 1.34 μg/l. Based on the surface area of the coupon segment, this equates to approximately μg/l/cm 2. To summarize, a determination of the THM formation potential and THM sorption rate was made using chloroform as the THM of interest. Cement-mortar coupons were immersed in base water to either create TOC-laden test water or to condition cement coupon segments for the sorption test. Analytical methods identical to those used in previous leaching tests were used to determine chloroform concentrations within samples taken at predetermined time intervals. The THM formation results indicated that the cement-induced TOC does have a THM formation potential of approximately 30 μg/l (chloroform). The THM sorption results, when corrected for the higher surface area to volume ratio, indicate that the chloroform concentration would be reduced 1.34 μg/l due to sorption on the coupon surface after 72 hours of exposure. Haloacetic Acids (HAA5) Figure 5.15 indicates that HAA5 formation was highest in the first days of leaching of the cement-mortar coupons when the TOC concentration was also highest. The lowest concentrations, 0 to 4 μg/l, formed in the no disinfectant control. All conditions with disinfectant produced more HAA5 than their controls, with the chlorinated waters producing more HAA5 than the ph 8, chloraminated water. Concentrations formed were in the range of 10 μg/l, which is substantial considering that the EPA limit is 60 μg/l. When the data are plotted to account for contact time, the production of HAA5 declined over time and by day 8 there was no

80 46 Impacts of Lining Materials on Water Quality Total HAA5 Concentration HAA5 Concentration (µg/l) Water #1 (No Disinfectant) Water #1 (Control) Water #2 (Chlorine, ph 6.5) Water #2 (Control) Water #3 (Chlorine) Water #3 (Control) Water #4 (Chloramines) Water #4 (Control) Test Day Daily Change in HAA5 Concentration (µg/l) Total HAA5 Concentration Change Per Day Water #1 (No Disinfectant) Water #1 (Control) Water #2 (Chlorine, ph 6.5) Water #2 (Control) Water #3 (Chlorine) Water #3 (Control) Water #4 (Chloramines) Water #4 (Control) Test Day Figure 5.15 Concentrations of HAA5 present in water in contact with cement mortar (top) and production of HAA5 as a function of contact time (bottom)

81 Chapter 5: Testing of Cement-Mortar Lining 47 difference between controls and test waters. Thus, production of HAA5 occurred with newly formed cement-mortar lining material when TOC was leached and then declined rapidly once the TOC was depleted. Odor Water samples were evaluated for odor by three to four members of a human panel trained in Flavor Profile Analysis (FPA) (Standard Method 2170). A noticeable cement odor was imparted to the water by the cement-mortar coupons (Figure 5.16). The intensity of this cement odor was in the 3.5 to 4.5 range on the FPA scale of 0 to 12. This is a weak intensity (defined as FPA intensity = 4) and corresponds to the sweetness of canned fruit (for comparison, FPA intensity = 8 = moderate and corresponds to canned soda and FPA = 12 = strong and corresponds to syrup or jelly). The odor intensity did not diminish over the 14-day period that data were collected. The cement odor intensity or description was not impacted by the presence of chlorine or monochloramine. On day 1, nearly all the residual disinfectant had been consumed in the samples. The sensory panelists detected a chlorine odor (data not shown) on day 14 in the ph 8 chlorinated and chloraminated waters. This is consistent with the chlorine residual loss data, which show that the chlorine and monochloramine were being consumed by the cement mortar early in the test but had leveled off by day 14. Figure 5.16 Cement odor intensity as a function of time and water type in contact with cement mortar

82 48 Impacts of Lining Materials on Water Quality SUMMARY OF IMPACTS TO WATER QUALITY FROM NEW CEMENT-MORTAR LINING Major Impacts 1. All ph and disinfectant combinations resulted in similar water quality effects. Thus, the initial water quality (ph 6.5 or 8; no disinfectant, chlorine, or monochloramine) had no impact on the release of chemicals from the cement. Impacts to water quality from cement mortar were most severe on days 1, 2, 4, and 9. Day 9 was a critical point after which a significant decrease in most water quality parameter release rates and curing of the cement occurred. 2. The ph increased drastically to ph 12.5 with 24 hours of contact with cement mortar and maintained values of ph 10.5 to 12.5 throughout the 30-day test period. Likewise, the alkalinity was increased from 35 to 600 mg/l (as CaCO 3 ) within the first 24 hours of contact with cement mortar. After 9 days of contact, the alkalinity declined to about 100 mg/l as (as CaCO 3 ). 3. The total solids content of the water increased to up to 1,700 mg/l in the presence of cement-mortar coupons. 4. High ph (up to 12.5 ph units), alkalinity (up to 600 mg/l), calcium concentrations (up to 260 mg/l), and total solids (up to 1,700 mg/l) are driven by dissolution of Ca(OH) 2, which is a byproduct of the curing of cement. 5. Cement mortar significantly increased the calcium, aluminum, and chromium concentrations in the water. The aluminum concentrations exceeded the EPA SMCL, while the chromium levels remained below the EPA MCL. After day 9, release of aluminum and chromium to the water decreased. After day 15 for calcium and from day 1 for magnesium, these metals sorbed to the cement from the water. 6. Cement mortar created a substantial chlorine demand but the demand for chloramine was much less and ceased after a few days of contact. 7. All water types containing cement-mortar coupons had an intense cement odor, which remained moderately high for first 14 days of testing. The presence of chlorine or chloramines did not mask this cement odor. This odor would be readily detectable by consumers. Minor Impacts 1. The presence of cement mortar had no impact on the ammonia concentration in no disinfectant, chlorinated, or chloraminated water. 2. TOC release was generally <0.5 mg/l. In chlorinated water, this TOC reacted to produce <10 μg/l HAA5; HAA5 formation decreased rapidly after 2 days. 3. Although THMs were not measured in the leachate water during the ANSI/NSF-61 procedure, the TOC leached from the cement-mortar coupons was demonstrated to form chloroform when reacted with chlorine without the presence of coupons. The cement-mortar coupons were shown to reduce the chloroform concentration through either sorption or reaction. Therefore, the potential formation of THMs exists when new cement mortar that is leaching TOC is contacted with free chlorine.

83 Chapter 5: Testing of Cement-Mortar Lining 49 TESTING OF CEMENT-MORTAR LINING WITH CORROSION PREVENTION ADDITIVE (CPA) WATERS Objective The objective of this testing was to determine if industry-standard pipe corrosion CPAs had any impact on the water quality effects of newly installed cement-mortar lining materials. This testing, which used materials and methods similar to those reported cement-mortar leaching tests, evaluated three different corrosion preventative chemical regimes: OP at 1.0 mg/l as P PP at 1.0 mg/l as P OPZn at 1.0 mg/l as P and 0.3 mg/l as Zn Samples with no CPA added were also tested and served as a comparative baseline. The tests performed to evaluate these potential impacts were as follows: ph Metals including total hardness (calcium and magnesium) Alkalinity These parameters were evaluated because they produced the most significant impacts during the earlier testing and they have a substantial influence on drinking water quality. The CPA results are presented graphically with the three CPA types and the control water shown on each figure. The data presented are the average of the duplicates used in the CPA testing. (Note: A very slow leak formed in one of the two control water immersion vessels, which resulted in some water loss between days 4 and 9. Approximately 30% of this water leaked over the 5-day period. The impact of this issue is unknown because the time at which the leak occurred during the period is uncertain. However, the data associated with day 9 control water may be skewed slightly upward.) Test Results ph The ph results are shown in Figure 5.17 and show that for all of the samples the ph rises rapidly, from ph 8.0, after exposure to the uncured cement mortar. The three CPA and control water types behaved similarly until day 11 when the polyphosphate (PP) water began to indicate a lower ph. This trend continued throughout the remaining test period with the PP additive showing the largest drop in ph to 10.65, departing from the two other CPA types and the control. The orthophosphate (OP) additive and CPA control behaved very similarly dropping to ph 11.2 to 11.3 and leveling off. The zincorthophosphate (OPZn) additive lowered the ph to 11.0, a value which fell in between the PP and OP values. The CPA results showed greater differences between types as the test period progressed.

84 50 Impacts of Lining Materials on Water Quality ph Base Water (Control) OP PP OPZn Initial ph for all water types Test Day Figure 5.17 ph as a function of exposure time for three CPA types and control water in contact with cement mortar The rise and overall trend in ph values during CPA testing can be attributed to the same causes that were outlined earlier in this chapter. The exposure time between water changes and the cement s cure progression affect the amount of calcium hydroxide leached into the sample water. Longer exposure times increase hydroxide concentration while cement cure progression dampens the increases in hydroxide concentration due to the more complete hydration of the cement s calcium silicate component. The PP additive s ability to limit the ph increase may be due to its enhanced ability to form a protective layer on the concrete surface, thus, slowing the movement of calcium hydroxide into the sample water. Alkalinity The alkalinity results are provided in Figure The alkalinity values for all samples are very high on day 1 due to a high rate of calcium hydroxide leaching. The alkalinity for all samples then drops rapidly on day 2. Subsequently, the three CPA types and control follow similar trends, only departing from one another on day 9. The data converge for the remainder of the test period with very similar results on day 19. These data suggest that these CPA types at the dosages tested do not have a substantial impact on changes in alkalinity.

85 Chapter 5: Testing of Cement-Mortar Lining Alkalinity (mg/l CaCO 3 ) Base Water (Control) OP PP OPZn Test Day Figure 5.18 Alkalinity as a function of exposure time for three CPA types and control water in contact with cement mortar Metals Samples were tested for the content of 17 different metals. This section will discuss the following metals testing results in detail: Calcium and magnesium because they are the major constituents of hardness. Aluminum because it is a major component of cement in the form of aluminum silicates. Phosphorous because the phosphate ion is associated with the CPAs that were added to the various test waters. Zinc because it was added as a component in one corrosion preventative type and the results showed some significant concentrations in all samples. Chromium because of its potential health effects and it was found in significant enough quantities to warrant discussion in the CPA tests and the previous leaching tests. The other metals and organics that were part of the analysis (sodium [Na], silicon [Si], sulfate [SO 4 ], chlorine [Cl], potassium [K], vanadium [V], iron [Fe], manganese [Mn], cobalt [Co], nickel [Ni], and copper [Cu]) were either present in very small amounts or did not show any significant variation from the control data or among the CPA types. All metal samples were filtered through a 0.45-micrometer (μm) glass filter before being tested so as to remove any precipitate. Hardness (Ca and Mg). The hardness results are provided in Figure 5.19 and show that the hardness is very high at day 1 and then falls asymptotically toward a value of approximately 50 mg/l as CaCO 3. This hardness is attributed almost entirely to calcium, the concentration of which is plotted in Figure 5.20, until day 14 for PP samples and day 15 for OP, OPZn, and control. After day 14 or 15, Mg begins to make a contribution (Figure 5.21). The Mg contribution remains small, especially for the CPA types other than PP. The high calcium values

86 52 Impacts of Lining Materials on Water Quality early in the test period are generated by the same reaction mechanisms that increase ph and alkalinity. As the cement cures, calcium hydroxide is produced and dissolves until super saturation conditions are met after which precipitation of calcium hydroxide and calcium carbonate occurs. The calcium leaching can be further augmented by simple dissolution of free lime from the mortar to the water. Figure 5.19 Total hardness as a function of exposure time for three CPA types and control water in contact with cement mortar Figure 5.20 Calcium concentrations as a function of exposure time for three CPA types and control water in contact with cement mortar

87 Chapter 5: Testing of Cement-Mortar Lining Average Mg Concentration in Base, OP, PP and OPZn Blank Waters [Mg] (mg/l) Base Water (Control) OP PP OPZn Test Day Figure 5.21 Magnesium concentrations as a function of exposure time for three CPA types and control water in contact with cement mortar Figure 5.21 indicates that the dissolved Mg in all of the water samples is 9.5 mg/l as Mg before placed in contact with the cement mortar. During the period before day 14 for PP and day 15 for the other samples, nearly all of this Mg falls out of solution as a precipitate or as a coating on the cement surface. This occurs due to the high ph seen during this period at which the solubility of magnesium salts is very low. It is not until the ph drops below 11.3 that appreciable amounts of Mg are found in solution. The higher dissolved Mg concentration found in the PP sample is the result of the lower ph values evident in the same samples. It was also noted that the PP water contained much less precipitate than the other samples during this period of lower ph. There were no readily observed hardness differences in the results among the CPA samples. There was an increase in the range of hardness values observed on day 9 but the range decreases significantly at day 11 and by day 19 the hardness values among the CPA types are very nearly the same (49 to 57 mg/l CaCO 3 ). The differences seen in Mg concentration may have substantial impacts at lower hardness levels. Aluminum. The results for Al concentration are provided in Figure 5.22 and indicate that this metal does dissolve into the water during the cement mortar s curing period. The data also suggest that there are differences among the various CPA types as to the magnitude of the dissolution. The OP and OPZn additives behaved similarly throughout the test period with both limiting aluminum dissolution better than the control on most days. The PP limited the Al concentration most effectively until day 15 and performed much better at days 9, 11, and 14. At day 19 the three CPA types all limit the Al concentration to substantially lower values than the control.

88 54 Impacts of Lining Materials on Water Quality Base Water (Control) OP PP OPZn [Al] (µg/l) SMCL for Al Average Al Concentration in Base, OP, PP and OPZn Blank Water Test Day Figure 5.22 Aluminum concentrations as a function of exposure time for three CPA types and control water in contact with cement mortar Phosphorous. The phosphorous concentration results are provided in Figure 5.23 and show that the P concentration drops to virtually zero for all CPA types until day 14. The PP water sample shows some P on day 14, most likely in the form of phosphate, while the others remain at near zero. On day 15, the PP water shows a rapid increase in phosphate content to slightly more than 800 μg/l as P while the OP and OPZn show an increase to nearly 60 μg/l as P. These results suggest that at the higher ph values prevalent during the earlier portion of the test period the phosphate was precipitating out of solution or forming a coating on the coupon surface. The phosphorous curve is very similar to the magnesium curve and may suggest the formation of magnesium phosphate precipitate or coating. This indicates that the lower ph values produced by the PP additive allowed a greater fraction of the added phosphate to dissolve back into the water (approximately two-thirds of the available phosphate). The dashed line represents the average phosphorous content of the CPA types in each of the conditioned waters before contact with the coupons (1283, 1261, and 1253 μg/l as P for PP, OP, and OPZn, respectively). The control water, which had no phosphate added, never shows any phosphate content indicating that the mortar does not contribute phosphate to the water. In summary, the PP additive allowed more phosphate to dissolve after day 14 due to its ability to limit the ph increases during the same period.

89 Chapter 5: Testing of Cement-Mortar Lining Average P Concentration in OP, PP and OPZn Blank Water [P] (µg/l) Base Water (Control) OP PP OPZn Test Day Figure 5.23 Phosphorous concentrations as a function of exposure time for three CPA types and control water in contact with cement mortar Zinc. The Zn concentration results are provided in Figure 5.24 and show that the CPA type had little impact on the Zn concentrations found in the waters over the entire test period. Even the OPZn water, which had an average of 360 μg/l as Zn before contact with the coupons (upper dashed line), had Zn concentration values that were similar to the other two CPA types and the control (these samples had nearly 0 μg/l as Zn before contact with coupons as shown by the lower dashed line). This suggests that two mechanisms are at work with respect to Zn concentration. First, the cement mortar is contributing Zn to the water sample in all cases. Second, the ph conditions are then controlling the concentration of Zn that remains in solution. The shape of the Zn concentration closely resembles the trend shown in the ph graph and suggests that as the ph is lowered, less of the Zn is staying in solution. The Zn is either precipitating or forming a surface coating at the lower ph values seen later in the test period. Overall, there is little difference in the behavior of the three CPA types and the control with respect to Zn concentration. Chromium. The Cr concentration results are provided in Figure 5.25 and show that the cement mortar increases the Cr concentration substantially on day 1 for all CPA water samples and the control. The Cr concentration then declines rapidly for all samples as the cement cures. The PP additive does drive the Cr concentration to a slightly lower value (to 1 μg/l) than the other CPA types and the control (2, 3, and 4 μg/l for OPZn, OP, and the control, respectively). The Cr concentration never approaches the EPA MCL of 100 μg/l. Overall, the CPAs and the control samples behaved similarly with respect to Cr concentration.

90 56 Impacts of Lining Materials on Water Quality Average Zn Concentration in OPZn Blank Water Base Water (Control) 300 OP PP [Zn] (µg/l) Average Zn Concentration in Base, OP and PP Blank Waters OPZn Test Day Figure 5.24 Zinc concentrations as a function of exposure time for three CPA types and control water in contact with cement mortar Note: MCL for Cr is 100 µg/l [Cr] (µg/l) Base Water (Control) OP PP OPZn 5 0 Average Cr Concentration in Base, OP, PP and OPZn Blank Waters Test Day Figure 5.25 Chromium concentrations as a function of exposure time for three CPA types and control water in contact with cement mortar

91 Chapter 5: Testing of Cement-Mortar Lining 57 SUMMARY OF TEST RESULTS WITH CPA An evaluation of the effects of CPA types was conducted using materials and methods similar to those used in evaluating the effects of cement-mortar lining material alone. The objective of the testing was to determine if there were differences in the effects produced by any of three CPA types when compared to a control. Table 5.8 summarizes the results. The key difference between the CPA types is the PP additives ability to limit the increase in ph that is being driven by the leaching of calcium hydroxide from cement mortar, particularly after day 11. It is this factor that drives most of the other differences seen between the CPA types and the control. ph Table 5.8 Summary of CPA Testing Results Parameter Differences between CPA types Differences versus control PP reduced ph increases more substantially than OP or OPZn after day 9 All CPA types reduced ph increases more than the control, OP only slightly Alkalinity Minimal difference Minimal difference, particularly at end of 19-day test period Hardness (Ca and Mg) Aluminum Phosphorous (phosphate) Minimal difference in hardness and Ca concentration; Mg concentration highest for PP at end of 19-day test period due to lower ph PP limited Al concentration to lower values on most days, similar values for all types at test period end All CPA types performed similarly until day 11 (zero P concentration), PP allowed for higher P content after day 11 due to lower ph and increased phosphate solubility. Minimal difference in hardness and Ca; all CPA types had higher Mg concentration at test end due to lower ph values All CPA types limited Al concentration to lower values on most test sampling days, all lower than control at test period end PP allowed for much higher P content after day 11 due to lower ph and increased P solubility; OP and OPZn only slightly higher than control value of zero Zinc Minimal difference Minimal difference Chromium Minimal difference Minimal difference

92

93 CHAPTER 6 TESTING OF POLYURETHANE LINING RESULTS Leaching of polyurethane lining was tested using three types of water as outlined in the methodology discussed in Chapter 4. In this chapter results of leaching tests are discussed. EFFECT OF POLYURETHANE ON INORGANIC WATER QUALITY PARAMETERS ph Regardless of the disinfectant type, the presence of polyurethane caused the leachate ph to decrease 2 to 3 ph units to approximately ph 6. Because ph is a log scale, a 2- to 3-unit change in ph corresponds to a 100- to 1,000-fold change in [H + ]. Values near ph 6 persisted throughout the 30-day exposure period to polyurethane with very small standard deviations in ph measurement (Figure 6.1). All of the initial water conditions began with a ph of 8 for no disinfectant and chlorine or ph 9 for monochloramine. The controls with no polyurethane were able to maintain their target ph values throughout the 30-day test period (Figure 6.1). The waters were changed and ph measured on days 1, 2, 4, 9, 11, 14, 15, 19, 21, and 30. The longer the length of exposure times to polyurethane between the water changes, the greater the decrease in ph. When the exposure time exceeded 1 day, seen on days 4, 9, 14, 19, and 30, the ph dropped by up to 0.7 ph units compared to the previous measurement. This translates to approximately a 3- to 5-fold change in the concentration of hydronium ion for a ph change of 0.2 to 0.7. The initial ph for monochloramine-containing water was 9, which dropped to about ph 6 in the presence of polyurethane, making the polyurethane-induced change in ph more dramatic for monochloramine than for chlorine or no disinfectant. Statistical analysis was performed using linear regression where only exposure time statistically impacted the change in ph (p = 0.013). The type of disinfectant used, chlorine or monochloramine, had no impact on the variance of ph (p = and , respectively). Total Alkalinity The total alkalinity was not significantly different in the presence of polyurethane compared to the corresponding control water with no polyurethane (Figure 6.2). No trends were seen in either increasing or decreasing alkalinities in the presence of polyurethanes for the no disinfectant-, chlorine-, or monochloramine-containing waters. The measured alkalinity is slightly higher for the monochloraminated test water and its control due to the addition of ammonium hydroxide. Hardness Similar to alkalinity, the presence of polyurethane did not affect water hardness (data not shown). The concentrations of calcium and magnesium, which constitute hardness, in leachate water were not statistically different from their corresponding control waters without 59

94 60 Impacts of Lining Materials on Water Quality ph ph 8, no disinfectant ph 8, chlorine ph 8, monochloramine ph 8, no disinfectant, control ph 8, 2 mg/l, control ph 8, 4 mg/l monochloramine, control Test Day Figure 6.1 ph as a function of time as measured by test day for three test waters in contact with polyurethane and water change (Control waters did not contain polyurethane coupons) Alkalnity (as mg/l CaCO 3 ) ph 8 ph 8, chlorine ph 8, monochloramine ph 8, control ph 8, chlorine, control ph 8, monochloramine, control Test Day Figure 6.2 Alkalinity as a function of time for the three test waters in contact with polyurethane and corresponding control waters without polyurethane

95 Chapter 6: Testing of Polyurethane Lining 61 polyurethane for no disinfectant, chlorinated, or monochloraminated water types. The hardness did not change over the 30-day exposure time (p = 0.264, 0.721, and 0.329, respectively, based on linear regression analysis for no disinfectant, chlorinated, or monochloraminated water types). Elemental Analyses Metals and nonmetals were measured by inductively coupled plasma (ICP)-MS. There was no substantial change in the concentration of barium (Ba), tin (Sn), Na, Si, K, cadmium (Cd), V, Co, arsenic (As), and molybdenum (Mo) in any test water containing polyurethane coupons compared to its corresponding control water without polyurethane. Selected trace metals and nonmetals, in the µg/l range, either adsorbed onto the polyurethane or leached out of the polymer when the concentrations in the test water were compared to those of the respective control. An element was considered to leach from the polymer if its concentration was higher than the control and adsorbed onto the polymer if the concentration was lower than the control. Five elements, two nonmetals (sulfur [S] and phosphorous [P]) and three metals (aluminum, iron [Fe], and zinc), were different in all three disinfectant treatments of the polyurethane and will be used for comparison. The graphs represent the metals that changed (>10% wet weight [wt/wt]) in the test waters compared to the controls and eliminated samples with values below detection limit readings (Figure 6.3). The data indicate that the three water types had different trends in the decrease or increase of selected elements. Concentrations of Al, Fe, P, and Zn consistently increased in the range of 1 to 40 µg/l concentrations in the test water with polyurethane coupons and no disinfectant (Figure 6.3) although Zn and P sorbed on day 14. Sulfur was consistently sorbed into the polymer and its concentration decreased in the water in comparison to the control. The no disinfectant water had the greatest change in the elemental concentrations. Other elements that were leached from the polymer, above 10% wt/wt change, in this water were Cl, Cr, Mn, Ni, Pb, and Cu (data not shown). Figure 6.3 Concentrations of selected elements in no disinfectant water in contact with polyurethane coupons (The element concentrations in the control water were subtracted from the test data)

96 62 Impacts of Lining Materials on Water Quality The chlorinated water both increased and decreased element concentrations with respect to the controls throughout the testing period. The elements did not have general trends but the elements that were both leached and sorbed from the polymer, above 10% wt/wt change, were Al, Cr, Fe, Ni, P, S, Cu, and Zn (Figure 6.4). The water with monochloramine both increased and decreased elemental concentrations but to a lesser extent than the chlorinated water (Figure 6.5). Again, there was not a trend in all of the elements, which were both sorbed and leached, with a greater than 10% wt/wt change, and include P, Cr, Fe, Ni, S, Al, and Zn. Of the metals measured, several had EPA regulatory limits, as MCLs or Action Levels, or guidelines through SMCLs. No measured values for any metal in any test water with polyurethane exceeded EPA regulated or guideline limits. Total and Dissolved Solids The presence of polyurethane did not add or remove measurable amounts of solids (data not shown). The values for total solids and dissolved solids were similar indicating that only dissolved solids were present in the control waters or waters containing polyurethane coupons. Disinfectant Residual Data for chlorine and monochloramine disinfectant loss in mg/l as Cl 2 are shown in Figures 6.6 and 6.7, respectively. The water with no disinfectant is not shown because there was no change in the disinfectant concentration. The data are shown as disinfectant residual as mg/l Cl 2 for both chlorine and monochloramine disinfectant. The residuals in the controls were higher than the test water residuals, indicating that the polyurethane is causing the disinfectant to degrade more quickly than it would otherwise. As expected, free chlorine decayed to a greater extent than monochloramine. When the disinfectant consumed per day exposed (mg as Cl 2 /L/exposure time) is graphed, the reaction rate of the chlorine with polyurethane decreases with time (Figure 6.8). For chlorine, the rate of decay is variable but overall shows a decreasing trend over time (R 2 = ). The decay rate was the highest in the first few days, then leveled off by day 9 (Figure 6.8). Therefore, within 2 weeks after lining, the chlorine consumption by the polyurethane will lessen. With monochloramine there is no real trend with time (R 2 = ). Ammonia Only the chloraminated water and its control contained measurable ammonia, which was formed by the decay of monochloramine. The test and control waters had similar concentrations of ammonia and thus, the presence of polyurethane had no effect on ammonia production. These data are consistent with those in Figure 6.8, which indicate that monochloramine was fairly stable.

97 Chapter 6: Testing of Polyurethane Lining 63 Element Concentration (µg/l) Test Day Figure 6.4 Concentrations of selected elements in 2 mg/l Cl 2 water in contact with polyurethane coupons (The element concentrations in the control water were subtracted from the test data) Al Fe P S Zn Element Concentration (μg/l) Al Fe P S Zn Test Day Figure 6.5 Concentrations of selected elements in 5 to 6 mg/l NH 2 Cl water in contact with polyurethane coupons (The element concentrations in the control water were subtracted from the test data)

98 64 Impacts of Lining Materials on Water Quality 2.5 Residual (mg/l as Cl 2 ) Consumed Residual Test Day Figure 6.6 Chlorine decay, showing residual disinfectant, and consumed disinfectant, as a function of time Figure 6.7 Monochloramine decay, showing residual disinfectant and consumed disinfectant, as a function of time

99 Chapter 6: Testing of Polyurethane Lining 65 Figure 6.8 Rate of chlorine and monochloramine decay; data were corrected for loss of residual in the controls (mg/l/day) EFFECT OF POLYURETHANE ON ORGANIC WATER QUALITY PARAMETERS Total Organic Carbon (TOC) Data for TOC are shown in Figure 6.9. The test waters are all higher than their controls, especially after polyurethane exposure times exceeding 1 day. It is important to note that the TOC in the controls decreases from ~0.5 mg/l to ~0.2 mg/l over 15 days, likely due to changes in TOC in Nanopure water used to prepare reference water. Comparison to the controls illustrates that the test water TOC notably increases when exposed to the polyurethane coupons for longer time periods. Over 15 days and 6 water changes the TOC levels in waters with polyurethane coupons decreased and reached approximately the concentrations found in the controls (Figure 6.9). The convergence of TOC leaching over 15 days is significant using regression on a logarithmic scale (p = 0.027). A trend in the leaching of TOC with disinfectant type was noted with the most TOC leached in the presence of chlorine, then in the presence of monochloramine, and the least TOC leached in the presence of no disinfectant. Although this relative amount of TOC leaching is apparent in Figure 6.9, it is not statistically significant due to the moderate variability between replicates. The convergence of the TOC leached to the concentration found in the controls shows that the decrease in leached TOC was not likely due to the decay of the polymer from disinfectant but to leachable TOC in the polymer. When plotted as a rate of TOC leaching, the data indicate that the rate of TOC leaching during the initial 4 days of contact with water is higher than the rate after 4 days (Figure 6.10).

100 66 Impacts of Lining Materials on Water Quality Total Organic Carbon (mg/l) ph 8, no disinfectant ph 8, chlorine ph 8, monochloramine ph 8, no disinfectant, control ph 8, chlorine, control ph 8, monochloramine, control Test Day Figure 6.9 Total organic carbon (TOC, mg/l) as a function of time for all three waters in contact with polyurethane with controls (error bars represent standard deviation) Total Organic Carbon (mg/cm 2 /day) ph 8, no disinfectant ph 8, chlorine ph 8, monochloramine Test Day Figure 6.10 Rate of TOC leachate (TOC, mg/cm 2 /day) as a function of time for all three waters in contact with polyurethane and corrected for TOC measured in controls

101 Chapter 6: Testing of Polyurethane Lining 67 Trihalomethanes (THMs) THM formation did not occur within the first 14 days of the polyurethane leaching test for any of the test water conditions. As expected, THMs were not detected in the water that had no disinfectant or the water that contained monochloramine. The chlorinated waters contained only <2 μg/l chloroform in both the chlorine test water and its control. As was stated previously, about 0.2 mg/l TOC was in the reference water and therefore, some formation of THMs could occur in the control. Additional experiments were performed to determine if THMs could be formed from the leached TOC and possibly sorb into the polyurethane. TOC was extracted from the polyurethane coupons with disinfectant-free water, then chlorine was added to this water in glass volatile organic analysis (VOA) vials and THM formation monitored. Although the TOC was about 1 mg/l and the free chlorine was not detectable after 24 hours, no THMs were detected. Thus, THM formation does not occur for TOC leached from this polyurethane lining. The absorption of THMs into polyurethane was also investigated. Reference water without disinfectant and containing 100-μg/L THMs was exposed to the polyurethane coupons for 3 days, after which THMs were measured in triplicate samples. The THMs decreased by 21.3, 34.8, 38.6, and 40.2% for CHCl 3, CHCl 2 Br, CHClBr 2, and CHBr 3, respectively (Table 6.1). Interestingly, as the number of bromines in the THM molecule increased, so did the amount of absorption of that THM into the polyurethane. Haloacetic Acids (HAA5) The HAA5 are comprised of monochloroacetic acid (MCAA), dichloroacetic acid (DCAA), trichloroacetic acid (TCAA), monobromoacetic acid (MBAA), and dibromoacetic acid (DBAA). EPA regulates total HAA5 species concentration <60 μg/l. The major HAA5 species that formed was MCAA, with lesser amounts of DCAA and TCAA. There was no bromide added to the water, therefore, it is logical that none of the species containing bromine were detected. The amount of HAA5 was minimal and the same for both the no disinfectant test water with polyurethane coupons and its control; thus, polyurethane does not leach HAA5. The monochloraminated test water formed approximately 5 μg/l more HAA5 than its corresponding control; this would be due to the presence of free chlorine in equilibrium with the ammonia and monochloramine. Limited HAA5 formed in the disinfected control waters due to the background level of TOC in the reference water. Table 6.1 Sorption of THMs from aqueous solution and into polyurethane after a 72-hour contact time; initial THM concentration was 100 µg/l Aqueous Concentration (µg/l) Replicate CHCl 3 CHCl 2 Br CHClBr 2 CHBr Avg

102 68 Impacts of Lining Materials on Water Quality Substantial amounts of HAA5 formed (approximately 30 μg/l) in the chlorinated water in the presence of the polyurethane coupons when compared to the controls (Figure 6.11). When the data were plotted as a rate of HAA5 formation in terms of μg/cm 3 /day HAA5 (Figure 6.12), the chlorinated water had the highest rate between days 1 and 4 and then decreased. The rate of formation of HAA5 in the chlorinated water in contact with polyurethane coupons was consistently greater than that for the corresponding chlorinated control, indicating that the TOC leached from the coupons was responsible for increased HAA5 formation. This is consistent with the TOC data that indicated that the rate of TOC leaching was the greatest in the initial contact of polyurethane with test water. It is also consistent with the higher rate of chlorine decay in the first days of the test. MCAA comprised most of the HAA5 (85%) that were present, which is to be expected in chlorinated waters if HAA5 are formed. Day 9 also formed large amounts of HAA5 compounds (approximately 30 μg/l), which is most likely due to the high TOC leached on this day as well as the long contact time (5 days). Figure 6.12 also indicates that the rate of formation of HAA5 in the chloraminated water was substantially less than in the chlorinated water, which is to be expected because the free chlorine level is very low when in equilibrium with chloramines. HAA5 Concentration (mg/l) ph 8, no disinfectant ph 8, chlorine ph 8, monochloramine Test Day Figure 6.11 HAA5 concentrations as a function of exposure time for all three water types (corrected for controls) in contact with polyurethane

103 Chapter 6: Testing of Polyurethane Lining 69 HAA5 Concentration (μg/cm 3 /day) ph 8, no disinfectant ph 8, no disinfectant, control ph 8, chlorine ph 8, chlorine, control ph 8, monochloramine ph 8, monochloramine, control Test Day Figure 6.12 HAA5 rate of leaching over time, as μg/cm 3 /day, as a function of time for all three waters in contact with polyurethane Semivolatile Organic Compounds (SVOCs) SVOCs are a specific group of compounds that are extractable by methylene chloride liquid-liquid and detectable by GC/MS. Only nontargeted organic compounds were detected in the test waters in contact with polyurethane and these compounds were not detected in the corresponding controls. One organic compound was detected in the test water leachates with the National Institute of Standards and Technology (NIST) MS spectral database (Table 6.2). Analysis of the SVOC data shows that chlorophenyl isocyanate was detected in the chlorinated water samples and one of the chloraminated water samples but not in the water without disinfectant. This is supported by previous data that demonstrated more TOC leached from the polyurethane in contact with chlorinated water. Chlorophenyl isocyanate is possibly related to the isocyanate compound used to manufacture the polyurethane. Although the concentration was not measured, it is estimated to be present in only a few µg/l based on its area counts for the GC/MS chromatogram. The toxicological properties of this compound are presented in Table 6.3. Table 6.2 Summary of organic chemicals detected in leachate water with polyurethane coupons but not detected in the control water Sample ID (CAS #) No disinfectant Cl 2 NH 2 Cl Sampling day 1 d* 4 d 9 d 14 d 1 d 4 d 9 d 14 d 1 d 4 d 9 d 14 d Chlorophenyl isocyanate ( )1 * Tentative identification based on >90% library match with NIST library mass spectrum

104 70 Impacts of Lining Materials on Water Quality Table 6.3 Properties of organic compound detected in leachate from polyurethane Acute oral LD 50 (mg/kg) Acute dermal LD 50 (mg/kg) Compound name (CAS #) Odor descriptor Color Irritant? Chlorophenyl isocyanate Slight corn Pale ( )* oil yellow 4,977 + >2,000 No * Identification tentative based on match of mass spectrum of sample and NIST library + Rat Rabbit Odor Water samples from all treatments were evaluated for odor by three to six members of a human panel trained in Flavor Profile Analysis (FPA) (Standard Method 2170). On the FPA intensity scale 0 is odor free (OF), 4 is weak, and 8 is moderate. Many drinking water customers begin to complain when a nonchlorinous odor intensity reaches 2 or above. There were distinct odors associated with water in contact with polyurethane coupons. To some of the panelists there was a putrid smell, which was also described as a smelly locker room, wet socks, and organic. Another set of panelists described the odor as pleasant or sweet chemical and vegetative. These descriptors were found in all of the samples, however, the prominence of the odors varied in each water. The chlorine smell was also seen in both the water with chlorine and chloramines. In fact, several of the samples, most notably day 1, 9, and 30, had concentrations of chlorine >0.10 mg/l Cl 2 and therefore, were not likely to be the source of such a strong odor. The odors did not exceed an odor intensity = 8 in the odor panel or a moderate intensity. In Figures 6.13 through 6.15, the presence of the pleasant smell was what separated the water odors most appreciably. With no disinfectant present the pleasant odor was detected on the first day, yet, with chlorine it is not detected until day 4 and not again until day 9 in the presence of monochloramine. The pleasant odor was also the most difficult to detect and was described as: organic, vegetative, pleasant, sweet, sweet chemical, beets, and burning meat (still described as a pleasant odor). The panelists often took several sniffs to be able to describe the odor accurately. Over the course of the tests there was no decrease in the average intensity of the waters, therefore, it is not possible to determine how long these odors will be present in the water. The odors detected in the FPA test cannot be directly related to the SVOC identified. Additionally, perception of odor is related to the composite of all compounds present and many odorants are detectable by the human nose as nanograms per liter (ng/l) concentration, which is lower than the detection limit of the SVOC method utilized. The extraction process used to measure SVOCs would not have extracted VOCs that could be responsible for these odors; more tests should be performed to confirm this.

105 Chapter 6: Testing of Polyurethane Lining 71 Odor Intensity Pleasant Putrid Chlorine Test Day Figure 6.13 Odor intensity, divided among the categorized odors, as a function of time for no disinfectant water exposed to polyurethane Pleasant Putrid Chlorine 8 6 Odor Intensity Test Day Figure 6.14 Odor intensity, divided among the categorized odors, as a function of time for chlorinated water exposed to polyurethane

106 72 Impacts of Lining Materials on Water Quality 12 Odor Intensity Pleasant Putrid Chlorine Test Day Figure 6.15 Odor intensity, divided among the categorized odors, as a function of time for chloraminated water exposed to polyurethane PHYSICAL CHANGES TO POLYURETHANE COUPONS A color change was noted after the polyurethane was exposed to the test waters (Figure 6.16). The chlorine-exposed polyurethane coupons had the most noticeable change followed by those exposed to monochloramine (which is always in equilibrium with free chlorine). The change was a graying of the purplish color of the polyurethane. Although no physical or chemical analyses have been performed on the polyurethane material, the color change strongly supports the idea that chlorine reacted with the polyurethane. This could be the source of an increased leaching of TOC and specific SVOCs and odors Figure 6.16 Polyurethane coupons after extended exposure to: (1) no disinfectant, (2) monochloramine, and (3) chlorine

107 Chapter 6: Testing of Polyurethane Lining 73 SUMMARY OF IMPACTS TO WATER QUALITY FROM POLYURETHANE Major impacts observed under the conditions of this study are as follows: 1. In the presence of polyurethane, the ph was reduced from ph 8 to about ph 6. The ph drop was observed within 24 hours and persisted for 30 days. The source of the ph drop is not known. Changes in concentrations of metals, nonmetals, or SVOCs cannot account for the 2-pH unit decrease. Analytes that were not measured in this study, such as carbon dioxide, strong acid anions, or volatile organic acids, should be investigated. 2. Free chlorine was consumed in the presence of polyurethane and its consumption was greater than that for monochloramine. The rate of chlorine decay was greater during days 1 through 4 than in later exposure times. 3. Organic carbon was leached from polyurethane, with a greater amount leached in the presence of chlorine than in its absence. Approximately 1 mg/l was leached in the first 24-hour contact period; afterward the rate of TOC leaching declined over time for all disinfectant treatment types. Leached TOC reacted with free chlorine to form up to 30-µg/L HAA5 but no THMs were detected. The low ph of 6 would favor HAA5 formation over THM formation. 4. Weak to moderate odor intensities were released from the polyurethane and it persisted for the 30 days of this study. These odor intensities would be objectionable to consumers. The odor was described as pleasant or putrid by different people on the odor panel. While the presence of a disinfectant caused a chlorinous odor, the disinfectant did not substantially alter the intensity or descriptors of the polyurethane odors. The identity of the odorants was not determined. Minor impacts observed under the conditions of this study are as follows: 1. Variable concentrations of metals and nonmetals were detected in the leachate water during the course of this study, indicating that elements both leached from and sorbed into the polyurethane. 2. The concentration differences were less that 40 µg/l and, in particular, Al, Fe, P, S, and Zn concentrations changed. 3. No observed metal concentrations exceeded EPA limits. 4. Leaching of elements was most substantial for the no disinfectant water, which could mean that the presence of an oxidant changed the surface chemistry of the polyurethane. 5. Surface chemistry changes are supported by the color change observed for polyurethane exposed to disinfectants.

108

109 CHAPTER 7 TESTING OF EPOXY LINING RESULTS FOR EPOXY LINING Leaching of epoxy lining was tested using the three test waters as outlined in the methodology discussed in Chapter 4. In this chapter results of leaching tests are discussed. EFFECT OF EPOXY ON INORGANIC WATER QUALITY PARAMETERS ph No difference in ph was seen between any of the three water types and their controls and no change in ph over time was found. All of the initial water conditions began with a ph of 8 with varied disinfectant types and concentrations (no disinfectant, 2 mg/l Cl 2, and 5 to 6 mg/l NH 2 Cl). The control waters were the reference waters (salts, ph 8, and disinfectants) placed in vessels with no coupons. Waters were adjusted to ph 8 using either dilute HCl or dilute NaOH. The addition of chloramines to the test water was found to increase the ph to 9 in the chloraminated water. Water from days 1 and 2 shows a ph of 9 in both chloraminated samples and controls, which is not due to interaction with the epoxy but rather the tendency of the ph to resist adjustment. After day 2 the ph was adjusted more often to ensure it remained at ph 8. Paired t tests were used for statistical analysis. Results showed no statistical difference between the test waters and the controls (p = 0.102, 0.180, and for no disinfectant, chlorinated, and chloraminated waters, respectively). The data indicate the epoxy has no effect on ph in any of the three water treatments. Figure 7.1 shows the ph change over the course of the study for the three water types. The mature pipe sample, in contrast, did show a tendency to decrease ph by about 0.3 ph units in chlorinated test waters, on average. Statistical analysis comparing the aged epoxy to the chlorinated control found the ph to be statistically lower in the aged epoxy (p < 0.001, one-sided paired t test). It is uncertain whether or not aging of the epoxy lining or the specific epoxy compound is responsible for the small, but detectable, change in ph. Figure 7.2 shows the ph in the mature pipe sample and in the chlorinated control. Alkalinity No change in alkalinity was detected in any of the three disinfectant treatments with respect to their controls (Figure 7.3). Target alkalinity for the reference waters was prepared as 35 mg/l as CaCO 3. Statistical analysis using paired t tests showed no difference in alkalinity between waters with coupons and their controls (p = 0.939, 0.995, and for no disinfectant, chlorinated, and chloraminated waters, respectively). The higher alkalinity readings on days 1 and 2 in both the chloraminated test waters and their corresponding controls can be attributed to the higher ph readings of ph 9 on both days. (Variability in chloraminated water triplicate samples was extremely low for all days, thus, error bars are not visible in Figure 7.3.) Analysis performed to compare alkalinity in the mature pipe sample to the chlorinated control also showed no difference between alkalinity in the chlorinated mature pipe sample and in the control (p = 0.121, paired t test). Figure 7.4 shows alkalinity variation with time in the matured 75

110 76 Impacts of Lining Materials on Water Quality sample. Both analyses indicate that neither the new epoxy coupons nor the mature pipe sample had any effect on alkalinity ph ph 8, no disinfectant ph 8, chlorine ph 8, monochloramine ph 8, no disinfectant, control ph 8, chlorine, control 4 ph 8, monochloramine, control Test Day Figure 7.1 ph as a function of time in all three waters and their controls in contact with epoxy (error bars indicate standard deviations) ph ph 8, chlorine, control ph 8, chlorine, mature pipe Test Day Figure 7.2 ph as a function of time in chlorinated matured epoxy-lined pipe and chlorinated control

111 Chapter 7: Testing of Epoxy Lining 77 Alkalinity as mg/l CaCO ph 8, no disinfectant ph 8, chlorine ph 8, monochloramine ph 8, no disinfectant, control ph 8, chlorine, control ph 8, monochloramine, control Test Day Figure 7.3 Alkalinity (mg/l CaCO 3 ) as a function of time for water types exposed to epoxy and their controls (error bars indicate standard deviations) 50 ph 8, chlorine, control 45 ph 8, chlorine, mature pipe Alkalinity as mg/l CaCO Test Day Figure 7.4 Alkalinity (mg/l CaCO 3 ) as a function of time in mature epoxy-lined pipe sample exposed to chlorinated water and corresponding chlorinated control

112 78 Impacts of Lining Materials on Water Quality Hardness No difference in hardness was detected in the waters in contact with epoxy coupons compared with the controls. The two primary constituents of hardness in the waters are calcium and magnesium, both added as salts to the test waters at a concentration of 11.5 mg/l Ca and 8 mg/l Mg and both recovered at the same concentrations. The results (Figures 7.5 and 7.6) show the same total hardness (as mg/l CaCO 3 ) in both sample and control waters, and analysis using paired t tests shows no difference between test waters and controls (p = 0.900, 0.605, and for no disinfectant, chlorinated, and chloraminated waters, respectively). Statistical analysis again using a paired t test shows no significant difference in total hardness between the mature pipe sample and the chlorinated controls (p = 0.439). Data show neither new nor mature epoxy has an impact on total hardness. 75 Ca and Mg Hardness (mg/l as CaCO 3 ) ph 8, no disinfectant ph 8, chlorine ph 8, monochloramine Test Day Figure 7.5 Calcium and magnesium hardness (mg/l as CaCO 3 ) concentrations for test waters exposed to epoxy as a function of time 75 Ca and Mg Hardness (mg/l as CaCO 3 ) ph 8, no disinfectant, control ph 8, chlorine, control ph 8, monochloramine, control Test Day Figure 7.6 Calcium and magnesium hardness (mg/l as CaCO 3 ) concentrations for control waters without epoxy as a function of time

113 Chapter 7: Testing of Epoxy Lining 79 Elemental Analyses No trends concerning selected trace metals and nonmetals were found in any of the three water types exposed to epoxy (data not shown). ICP-MS was used to quantify specific trace metals (sodium, aluminum, potassium, vanadium, chromium, iron, manganese, cobalt, nickel, copper, zinc, arsenic, molybdenum, cadmium, tin, barium, and lead) and nonmetals (sulfur, phosphorus, chlorine, and silicon). Analysis using paired t tests shows no statistical difference between the concentration of Na, Al, Cr, Fe, Mn, Ni, Cu, Zn, lead (Pb), sulfur (S), phosphorus (P), or Si between the test waters and the controls (p > 0.05 for all treatments). Concentrations of V, Co, As, Mb, Cd, and Sn were below detection limits on most test days and thus, could not be statistically analyzed. Analysis comparing metals and nonmetals concentrations in the aged pipe sample to its chlorinated control also failed to detect a statistical difference between any of the elemental concentrations (p >0.05 for all tests) (data not shown). Again, both epoxy linings had no discernible effect on any of the selected metals or nonmetals sampled in the study. Total and Dissolved Solids The total solids concentrations (soluble solids + suspended solids) of the three sample waters and controls were not found to be different than the controls or follow a trend over time (data not shown). Sample water was clear of particulate solids and total solids were recovered in approximately the same concentrations in both test waters and controls. Paired t tests were used to compare total solids concentrations in the test waters with the controls and confirmed that the concentrations of solids in all three treatments were the same as the controls (p = 0.887, 0.915, and for no disinfectant, chlorinated, and chloraminated waters, respectively). These data indicate epoxy does not contribute to the presence of solids in the waters. Disinfectant Residual Figures 7.7 through 7.10 show the disinfectant loss in mg/l as Cl 2 for both chlorinated and chloraminated waters. Water with no disinfectant is not shown because there was no Cl 2 in the initial test water and the concentration did not change over the course of the study. The residuals in the control samples were consistently higher than in the test waters in both chlorinated and chloraminated waters (p <0.001, p = for chlorinated and chloraminated waters, respectively), indicating epoxy is causing increased consumption of both chlorine and monochloramine. As consistent with earlier data from studies with cement-mortar and polyurethane-lining materials, the rate of disinfectant consumption shows a trend of high initial consumption rate, which decreases over the study. Longer exposure times correspond to greater disinfectant consumption but the rate appears to follow an exponential form of decay, with the highest rate of decay during the first 24 hours and a slower decay rate on subsequent test days. Chloraminated water exposed to epoxy coupons had the highest rate of consumption during the first 24 hours but consumption rates slowed significantly after this initial exposure. By contrast, free chlorine was still being consumed by the epoxy at a higher and more consistent rate even after 30 days of exposure. Although Figure 7.9, showing the disinfectant rate of consumption, suggests fairly similar rates between the two disinfectants, it is important to note that the proportion of chlorine consumed is much greater. Free chlorine residual in the water exposed to

114 80 Impacts of Lining Materials on Water Quality 2.5 Demand Residual Disinfectant Residual (mg/l Cl2) Test Day Figure 7.7 Chlorine decay, showing residual disinfectant and consumed disinfectant as a function of time, corrected for decay in the controls (water: ph 8, 2 mg/l Cl 2 ) Disinfectant Residual (mg/l Cl2) Demand Residual Test Day Figure 7.8 Monochloramine decay, showing residual disinfectant and consumed disinfectant as a function of time, corrected for decay in the controls (water: ph 8, 4.5 to 5.8 mg/l NH 2 Cl)

115 Chapter 7: Testing of Epoxy Lining 81 Disinfectant Residual (mg/l Cl 2 ) ph 8,chlorine ph 8, monochloramine ph 8, chlorine, control ph 8, monochloramine, control Test Day Figure 7.9 Chlorine and monochloramine residuals (mg/l) with controls as a function of time 2.0 Disinfectant Residual Consumption Rate (mg/l/day) ph 8,chlorine ph 8, monochloramine Linear (ph 8, chlorine) Linear (ph 8, monochloramine) Test Day Figure 7.10 Chlorine and monochloramine consumption rate (mg/l/day), corrected for decay in the controls, as a function of time

116 82 Impacts of Lining Materials on Water Quality epoxy decreased on average by 80% compared with the chlorine concentration in the control samples. In contrast, chloramines residual decreased on average by only 20% compared with the controls. The epoxy lining reacts considerably with chlorine in the water to decrease it or, at longer exposure times, completely consume it. After the initial 24 hours, epoxy reacts with chloramines to a lesser extent and shows a decreasing ability to consume chloramines over time. Interestingly, the mature pipe sample showed a much higher tendency to consume disinfectant over the course of the study. Whereas the new epoxy showed a decreasing tendency to react with and consume chlorine over time, the mature pipe sample maintained a disinfectant residual that averaged only 0.05 mg/l over the study. Even when normalized for surface area, the disinfectant was still significantly lower than the new epoxy exposed to chlorinated water (p = 0.006, one-sided paired t test). The disinfectant residual for the mature pipe is shown in Figure These findings show that maintaining a disinfectant residual in both new and mature epoxy-lined pipes could be problematic in waters treated with chlorine. Ammonia Data for ammonia, shown in Figure 7.12, is only for chloraminated water. Neither the no disinfectant water nor the chlorinated water had ammonia levels above background levels of between 0 and 0.05 mg/l as N. Figure 7.12 shows similar concentrations of ammonia in both the sample and control waters, with the notable exception of day 1. The higher ammonia concentration in the samples on day 1 can likely be attributed to the high rate of monochloramine consumption by the epoxy, releasing free ammonia into the water. Other than the high initial rate of monochloramine consumption, epoxy has no discernible effect on the concentration of Disinfectant Residual (mg/l Cl2) Chlorine Residual Chlorine Demand Test Day Figure 7.11 Chlorine decay, showing residual disinfectant and consumed disinfectant as a function of time, corrected for natural decay in the controls for matured epoxy-lined sample (water: ph 8, 2 mg/l Cl 2 )

117 Chapter 7: Testing of Epoxy Lining Ammonia (mg/l as N) ph 8, monochloramine ph 8, monochloramine, control Test Day Figure 7.12 Ammonia (mg/l as N) in chloraminated sample and control waters (4.5 to 5.8 mg/l NH 2 Cl) as a function of time (error bars indicate standard deviations) ammonia in the water. Statistical analysis using paired t tests showed no difference between the ammonia as nitrogen (N-NH 3 ) concentration in the chloraminated test waters and controls (p = 0.522). Also of note is the trend seen in the N-NH 3 concentration in the mature pipe sample. The concentration of N-NH 3, while still low, was consistently and significantly higher than the concentration in the chlorinated control waters (p <0.001, paired t test). The slightly higher concentration of N-NH 3 found in the pipe samples, which averaged 0.06 mg/l over the course of the study (compared with 0.00 in the control), could be due to the presence of bacteria or accumulations found on the surface of the epoxy. ORGANIC WATER QUALITY PARAMETERS Total Organic Carbon (TOC) The test waters all showed significantly higher TOC concentrations than the controls (p = for all three waters). The highest TOC leached in the samples was after the first 24 hours of contact between epoxy coupons and water, when concentrations in no disinfectant and chlorinated waters reached 6.3-mg/L and 5.5-mg/L TOC, respectively (Figure 7.13). TOC in chloraminated waters was noticeably lower, with only 3.4 mg/l detected. By the next 24-hour contact period, concentrations were shown to have decreased significantly. All treatments shared this trend of high initial leached TOC for the first 24 hours of exposure, with significantly reduced TOC leaching after the initial 24-hour period. Interestingly, the rate of TOC leaching from the epoxy coupons remained more constant in the chlorinated and chloraminated water samples, with TOC leaching at a rate of 0.30 mg/l/day by day 4 and decreasing to 0.19 mg/l/day by day 30 in chlorinated waters. In chloraminated waters the rate decreased from 0.33

118 84 Impacts of Lining Materials on Water Quality ph 8, no disinfectant ph 8, chlorine ph 8, monochloramine TOC (mg/l) Test Day Figure 7.13 TOC (mg/l) as a function of time for all three water types, corrected for controls (error bars indicate standard deviations) mg/l/day on day 4 to 0.12 mg/l/day by day 30. The no disinfectant water showed a rate decrease from 0.30 mg/l/day to 0.07 over the same time period. Figures 7.14 and 7.15 show the TOC leaching rate, normalized for epoxy surface area (mg/cm 2 /day) to show a comparison between the new epoxy coupons and the mature pipe. The concentration of TOC in the mature pipe sample exposed to chlorinated water (shown in Figure 7.14 as mg/cm 2 surface area) remained fairly stable over the study. TOC ranged from 1.6 to 3.5 mg/l over the 30 days. When normalized for surface area, TOC leached from the pipe during the first 24 hours was initially much lower than all other water types in contact with new epoxy coupons. TOC on subsequent days, however, was at comparable levels. The data show that mature epoxy still leaches organic carbon, particularly under long exposure times, even after a substantial amount of time has passed. Trihalomethanes (THMs) THMs were not detected in either the no disinfectant or the chloraminated waters and only chloroform (CHCl 3 ) was detected in the chlorinated waters (Figures 7.16 and 7.17). With the exception of day 4, the rate of THM formation remained fairly constant. The concentration of THMs in the chlorinated test waters was low, ranging from 0.3 to 11.8 µg/l above the background concentration in the controls, much lower than the MCL of 80 µg/l. Figure 7.17 shows the concentration (µg/cm 2 of epoxy surface area) as a function of time for chlorinated test waters. Figure 7.18 shows the CHCl 3 formation rate (µg/cm 2 /day) as a function of time. Overall concentrations increased with exposure time and THMs were still being formed at the end of the study.

119 Chapter 7: Testing of Epoxy Lining 85 TOC (mg/cm 2 ) ph 8, no disinfectant ph 8, chlorine ph 8, monochloramine ph 8, chlorine, mature pipe Test Day Figure 7.14 TOC (mg/cm 2 ) as a function of time for all three water types and mature pipe sample, corrected for controls (error bars indicate standard deviations) TOC (mg/cm 2 /day) ph 8, no disinfectant ph 8, chlorine ph 8, monochloramine ph 8, chlorine, mature pipe Test Day Figure 7.15 Rate of TOC formation (mg/cm 2 /day) in all three water treatments as a function of time, corrected for controls (error bars indicate standard deviations)

120 86 Impacts of Lining Materials on Water Quality 14 CHCl 3 Concentration (µg/l) Test Day Figure 7.16 CHCl 3 concentration (µg/l) as a function of time in chlorinated test waters, corrected for controls CHCl 3 Concentration (µg/cm 2 ) Test Day Figure 7.17 CHCl 3 concentration (µg/cm 2 ) as a function of time in chlorinated test waters, corrected for controls

121 Chapter 7: Testing of Epoxy Lining CHCl3 Concentration (µg/cm 2 /day) Test Day Figure 7.18 CHCl 3 formation rate (µg/cm 2 /day) in chlorinated test waters, corrected for controls Negligible concentrations of THMs were formed in chlorinated water exposed to the mature pipe sample. THM Formation/Sorption A separate two-part study was conducted using the same test vessels and new epoxycoated coupons to measure THM formation and sorption potential in waters exposed to epoxy. The first portion of the experiment involved exposing 10 coupons to no disinfectant reference water for 72 hours, under the same experimental conditions as the other portion of this study. After the initial exposure time, triplicate water samples were collected and spiked with a high concentration of Cl 2 (6.7 mg/l) and allowed to react for 1, 24, or 72 hours before quenching with sodium thiosulfate. Samples were then analyzed for disinfectant residual, TOC, and THM concentration. Controls were also collected of the initial contact water before Cl 2 was added. The total surface area of the coupons was 1,573.8 cm 2 and the total volume was 2,970 cm 3, giving a surface area to volume ratio of The data confirm the potential for THMs to form in the presence of Cl 2 and epoxy. Only chloroform (CHCl 3 ) was formed, and at a concentration of approximately 10 µg/l. When normalized for surface area, the amount of chloroform produced averaged µg/cm 2, approximately the same levels as detected in earlier chlorinated samples. Figure 7.19 shows that even a very high initial disinfectant concentration is mostly consumed within the first hour and completely consumed within 24 hours. THMs were still forming after the first hour but increased only slightly afterwards because all the chlorine was consumed. The initial TOC concentration in the water was 7.5 mg/l, which decreased 2 mg/l over the next 3 days. Normalized for surface area, the initial TOC was mg/cm 2. This concentration was slightly lower than the

122 88 Impacts of Lining Materials on Water Quality Concentration TOC (mg/l) chloroform (µg/l) chlorine (mg/l) Test Day Figure 7.19 Concentrations of disinfectant residual (mg/l), TOC (mg/l), and CHCl 3 (µg/l) in THM formation study as a function of time mg/cm 2 initial TOC leached from new coupons during the bench-scale study. Interestingly, the additional chlorine consumed did not substantially increase the THMs formed in the water. A THM sorption experiment was also conducted to determine whether or not a portion of the chloroform that was being formed from leached TOC was absorbing back into the epoxy lining. In this study, performed under the same test conditions and again using new epoxy-coated coupons, 10 coupons were exposed to no disinfectant reference water spiked with approximately 14.5 µg/l of CHCl 3. Samples were collected from the headspace-free vessels after 8, 24, 48, and 72 hours to determine CHCl 3 concentration in the samples. No statistically significant change in CHCl 3 was observed over the 72 hours (data not shown), showing epoxy does not absorb CHCl 3. Haloacetic Acids (HAA5) Results showed low concentrations of HAA5 in the test waters (less than 12 µg/l) when adjusted for the background concentrations in the control waters (Figure 7.20). HAA5 present in the water consisted primarily of DCAA and TCAA, with mostly negligible concentrations of MCAA being detected. No brominated HAA5 were found. With the exception of day 1, there were consistently higher concentrations of HAA5 produced in the chlorinated waters. The concentration showed an increase over time across the test days. Figure 7.21 shows the difference in concentration (µg/cm 2 of epoxy surface area) between the HAA5 in no disinfectant, chlorinated, and chloraminated test waters, corrected for the controls. Figure 7.22 shows the formation rate of HAA5 (µg/cm 2 /day) in the same treatments, also corrected for controls. The third figure shows the concentration of HAA5 (µg/l) in chlorinated water exposed to the aged epoxy pipe sample and epoxy coupons, corrected for controls. Analysis of the data using one-sided paired t tests confirmed a statistically higher

123 Chapter 7: Testing of Epoxy Lining 89 HAA5 Concentration (µg/l) ph 8, no disinfectant ph 8, chlorine ph 8, monochloramine Test Day Figure 7.20 HAA5 concentrations (µg/l) corrected for controls as a function of time for all three water types HAA5 Concentration (µg/cm 2 ) ph 8, no disinfectant ph 8, chlorine ph 8, monochloramine ph 8, chlorine, mature pipe Test Day Figure 7.21 HAA5 concentrations (µg/cm 2 ) corrected for controls as a function of time for all three water types and mature pipe sample

124 90 Impacts of Lining Materials on Water Quality HAA5 Concentration (µg/cm 2 /day) ph 8, no disinfectant ph 8, chlorine ph 8, monochloramine ph 8, chlorine, mature pipe Test Day Figure 7.22 HAA5 rate of formation (µg/cm 2 /day) corrected for controls as a function of time for no disinfectant, chlorinated, and chloraminated test waters concentration of CHCl 3 in the chlorinated test waters compared with the no disinfectant test waters but no difference compared with the chloraminated waters (p = and no disinfectant and chloraminated waters, respectively). No difference was found in the HAA5 concentration between chlorinated test waters exposed to the new epoxy and the chlorinated water exposed to the mature coupons (p = 0.798). It is uncertain why the mature epoxy sample had the tendency to react with disinfectant to produce HAA5 but not THMs, while the new epoxy produced both disinfection byproducts (DBPs). Overall the data indicate that both new and mature epoxy does react with disinfectant in the water to produce small, but measurable, concentrations of HAA5. However, it is not clear whether the mature epoxy or the deposits on the epoxy are responsible for DBP formation. Semivolatile Organic Compounds (SVOCs) None of the targeted SVOCs were detected in the water samples in contact with epoxy coupons. Samples were also analyzed for the presence of bis-phenol A (BPA), an endocrinedisrupting compound commonly found in epoxy resins, and any other SVOCs using a GC/MS library search. Bis-phenol A was detected in substantial concentrations (22 to 33 µg/l) in all three water types exposed to epoxy (Figure 7.23). Concentrations were greatest during the initial 24 hours and were highest in both chlorinated and chloraminated waters. Bis-phenol A decreased significantly by day 4, but trace concentrations were still detected on both day 9 and 14. No other SVOCs were detected in the samples.

125 Chapter 7: Testing of Epoxy Lining 91 Bis-Phenol A Concentration (µg/l) ph 8, no disinfectant ph 8, chlorine ph 8, monochloramine Test Day Figure 7.23 Bis-phenol A concentration (µg/l) in all three waters in contact with epoxy as a function of time, corrected for controls Odor Water samples were evaluated for odor by three to five members of a panel trained in FPA (Standard Method 2170). The FPA intensity scale: is 0 = odor free, 1 = threshold, 2 = very weak, 4 = weak, 8 = moderate, 10 = strong, and 12 = very strong. A weak intensity of 4 corresponds to the sweetness of canned fruit while a moderate intensity of 8 corresponds to the sweetness of canned soda. An intensity of 12, very strong, corresponds to the sweetness of syrup or jelly. There were distinct odors associated with the three water treatments. The most commonly used descriptors were chlorinous, sweet, plastic, and chemical. For the purpose of quantifying the odor intensity of the waters according to descriptors, two categories were used: chlorinous and sweet/plastic/chemical. Panelists commonly used sweet, plastic, and chemical interchangeably to describe the same odor. Of the test waters, the chlorinated samples had the highest odor intensity, followed by chloraminated samples and then no disinfectant samples. None of the odor intensities of the test waters changed statistically over time (linear regression, p > 0.05). The no disinfectant test water had an average intensity of 3 over the 7 test days (Figure 7.24). Panelists described the no disinfectant samples as an overall pleasant odor and used the labels sweet, chemical, plastic, and waxy to describe it. By comparison, the no disinfectant control had an average intensity of 0.7 over the test days and the odor was typically described as either undetectable or an unidentifiable threshold odor. Panelists reviewing the chlorinated test waters could detect both a chlorinated odor and a plastic/chemical odor on most test days. They commonly described the chlorinous odor as

126 92 Impacts of Lining Materials on Water Quality Chlorinous Plastic/Chemical/Sweet Odor Intensity Test Day Figure 7.24 Odor intensity, divided between categorized odors, as a function of time (water: ph 8, no disinfectant) (error bars indicate standard deviations) burning and the plastic/chemical odor as burning chemical or burnt rubber. The combination of chlorine with epoxy appeared to have a synergistic effect on the plastic/chemical odor, with panelists no longer describing the odor as sweet and pleasant with a very weak to weak intensity but rather an offensive chemical odor with a moderate intensity. Over the 7 test days the average intensity of the chlorinous odor was 7 and the intensity of the plastic/chemical odor was 7.6 (Figure 7.25). Intensity was highest on days 4, 9, 14, 19, and 21, with lower intensities found on both the first and last test days. Notably, on day 30, after an exposure period of 9 days and no chlorine disinfectant residual remaining in the test water, panelists could no longer detect any chlorinous odor and described the plastic/chemical odor for the first time as sweet and having a weak intensity. The chlorinated control water maintained a more constant chlorinous odor over the test days, with an average intensity of 7.1 (Figure 7.25). The control still had a moderate chlorinous odor on day 30, also consistent with the disinfectant residual concentration in the control. Panelists reported the chloraminated test waters also had both a chlorinous and plastic/chemical odor, but with weaker intensities of both chlorine and plastic (Figure 7.26). The chlorinous odor appeared to be masked by the plastic/chemical odor on day 1 and had a much lower intensity than in chlorinated test waters (average intensity: 2.7). The plastic/chemical odor was also not as intense as in chlorinated test waters (average intensity: 4.4) and panelists frequently described the odor as sweet and pleasant. The sweet descriptor was also commonly used to describe the chloraminated control water, most likely due to chloramines having a characteristically sweeter odor.

127 Chapter 7: Testing of Epoxy Lining Odor Intensity Chlorinous Plastic/Chemical/Sweet Test Day Figure 7.25 Odor intensity, divided between categorized odors, as a function of time (water: ph 8, 2 mg/l Cl 2 ) (error bars indicate standard deviations) Chlorinous Plastic/Chemical/Sweet Odor Intensity Test Day Figure 7.26 Odor intensity, divided between categorized odors, as a function of time (water: ph 8, 4.5 to 5.8 mg/l NH 2 Cl)

128 94 Impacts of Lining Materials on Water Quality Table 7.1 Results of microbiological analysis of 2 cm 2 of mature epoxy pipe surface Plate Plate counts (CFU*/cm 2 ) Morphological description R2A Original sample 103 Diverse circular colonies of different colors: pink, yellow, and white R2A Diluted x NA PCA Original sample 4 Circular colonies, one pink one yellow, and two white PCA Diluted x NA McConkey Original sample 0 NA McConkey Diluted x NA * CFU = Colony-forming unit Microbiological Analysis of Mature Epoxy Lining A microbiological analysis was performed on the mature pipe sample to determine whether a biofilm was present on the epoxy lining. To perform the analysis, a 2-cm 2 area was scraped using a spatula, swabbed for bacteria, and added to a sterile tap water solution for culturing. Detection of heterotrophic organisms was performed using both R2A and PCA agar. MacConkey (lactose) agar was used for detection of fecal coliforms. The results of the analysis are shown in Table 7.1. A visible biofilm was not present on the pipe surface, but data do show the presence of bacterial colonies in low densities. SUMMARY OF IMPACTS TO WATER QUALITY FROM NEW EPOXY AND MATURE EPOXY New Epoxy The following are the major impacts observed under the conditions of the study: 1. Epoxy exposed to each of the three water types produced significant concentrations of TOC (3.5 to 6.3 mg/l) during the first 24 hours of exposure to water. By the second 24-hour exposure period the TOC decreased substantially. Epoxy coupons in both chlorinated and no disinfectant water leached the highest concentrations of TOC. By the end of the 30 days, each of the water types exposed to epoxy had TOC present in concentrations between 0.5 and 1.7 mg/l, with chlorinated water having the highest TOC concentration. 2. The epoxy reacted readily with both chlorine and chloramines during the first 24 hours of exposure. After the initial 24 hours of exposure, free chlorine was consumed at a much greater rate than chloramines. The disinfectant consumption rate decreased over the 30 days for both disinfectants but longer exposure times corresponded with higher consumption. Although some DBPs formed, the high consumption of disinfectant is not accounted for by the low formation of HAA5 and THMs. In fact, the THM formation data show a significantly higher amount of chlorine was consumed but significantly higher concentrations of DBPs were not formed.

129 Chapter 7: Testing of Epoxy Lining DBPs were present in most samples, with the highest concentrations detected in chlorinated water. THMs were present only in the chlorinated water exposed to the epoxy and at relatively low levels (<12 µg/l). Chloroform was the only THM detected. HAA5 were present in all three water types and the highest concentrations were found in chlorinated waters. Concentrations in waters exposed to epoxy were never greater than 8 µg/l above the controls. Both THM and HAA5 concentrations increased with exposure time. 4. BPA was detected in concentrations of 22 to 33 µg/l during the first 24 hours in all three waters exposed to epoxy. Concentrations decreased substantially by the second test day, but trace amounts of BPA were still detected by the last test day. No other targeted SVOCs were detected in the samples and it is uncertain what organic compounds contributed to the high TOC levels. 5. Weak to moderate odor intensities were released from water exposed to epoxy, which persisted all 30 days. The odor was strongest in chlorinated waters, with panelists detecting both a chlorinous and a burnt plastic odor. In chloraminated and no disinfectant waters panelists detected a sweet and pleasant odor, in addition to a less intense chlorinous odor in chloraminated waters. The intensities from any of the samples would be objectionable to customers. Minor impacts observed under the conditions of the study are as follows: 1. Ammonia was present in high concentrations in water exposed to epoxy only during the first 24 hours. This can be attributed to the high consumption of chloramines during the first 24 hours, which released nitrogen into the water. Mature Epoxy Major impacts observed under the conditions of the study are as follows: 1. ph in the mature pipe sample was found to decrease slightly when compared with the chlorinated control. The ph consistently dropped approximately 0.3 ph units on each test day. 2. Free chlorine exposed to the mature pipe was almost completely consumed by each test day. In contrast to the new epoxy, the pipe did not show a decrease in disinfectant demand over the 30 days. 3. TOC leached from the mature pipe sample ranged in concentration from 3.5 to 1.6 mg/l. There was a slightly higher TOC concentration leached during the first 24 hours, but overall the leaching rate remained constant over the experiment. 4. DBPs were formed in chlorinated water exposed to the pipe. Only HAA5 were detected and, when normalized for surface area, these were formed at the same concentration as in the chlorinated samples exposed to new epoxy. Minor impacts observed under the conditions of the study are as follows: 1. Ammonia was statistically higher in the pipe sample than in controls, but only by 0.6 mg/l. The slightly higher ammonia concentration in the pipe sample could be due to the presence of bacteria or buildup on the surface of the pipe.

130

131 CHAPTER 8 EVALUATION OF WATER QUALITY IMPACTS INTRODUCTION This chapter synthesizes the laboratory data and identifies and quantifies the background water quality parameters that impact leaching from lining materials. These background water quality parameters are then correlated with drinking water impacts and used in the development of a methodology for selection of a liner. The impacts of the tested lining materials on the quality of the water in contact with the liners were measured during laboratory-scale batch tests. During the batch tests coupons of three different materials (cement mortar, epoxy, and polyurethane), currently in use for lining water mains, were immersed in water. Four different waters were used to represent different finished water conditions. The water in contact with the coupons was analyzed after different contact times. The water was replaced with fresh water after each contact period. The replacement of the water after each contact period can be considered as the periodic flushing of a newly lined water main. The batch contact represents conditions that more closely resemble the conditions in a dead-end water main. The volume to surface area ratio used in the batch tests represented the ratio present in a 4-inch diameter pipe. Therefore, the test conditions for the laboratory tests are representative of conditions that may be encountered in small diameter pipes in a water distribution system. Therefore, the test conditions represent conditions in the field that result in maximum impact from the lining material on water quality. The following sections describe the changes in water quality that occurred in test waters in contact with three lining materials. These changes are correlated with the quality of the test waters. Compatibilities and incompatibilities between the characteristics of the test waters and the lining materials and their implications for water quality impacts in distribution systems are described. A general schematic diagram for evaluating the potential water quality impacts from lining materials used in public water supply systems is shown in Figure 8.1. The methodology considers the lining materials, source water quality, and regulatory limits for drinking water. Figure 8.1 describes a uniform approach for evaluating three lining materials. Construction parameters include those factors that can vary during application of the lining materials and affect their durability and water quality impacts. Background water quality includes water chemistry parameters that may influence the leaching of materials from the liners. Drinking water impacts include the concentrations of all constituents leached from the lining materials and consideration of the interaction of background water quality with the lining materials and the duration of water quality impacts. The duration of the water quality impacts is a function of two factors, the age of the pipe lining material, and the contact time between the water and the pipe liner. Health impacts are qualitative descriptions of the health impacts of the constituents leached from the lining materials. Regulatory impacts refer to a comparison of the concentrations of compounds and elements released from the lining materials against current primary and secondary drinking water standards. Drinking water standards anticipated in the future are also considered. 97

132 98 Impacts of Lining Materials on Water Quality Lining Material: Cement mortar Epoxy Polyurethane Construction Parameters Technology Used Mix Ratio Cure Time Background Water Quality Parameters ph, alkalinity, hardness, TOC, disinfectant type and residual, antiscalant type and concentration. Drinking Water Quality Impacts of Various Contact Times Leachate composition, ph, alkalinity, metals concentration, taste and odor, DBPs, VOCs, SVOCs impacts on disinfection residual. Health Impacts Figure 8.1 Schematic diagram for evaluation of water quality impact of lining LINING MATERIALS EVALUATED The methodology outlined in Figure 8.1 is applied to the three new lining materials that were tested; cement mortar, epoxy, and polyurethane. In addition, one mature epoxy lining was also tested. Cement Mortar Regulatory Impacts Testing was performed on water in contact with Type II Portland cement. The experimental protocol is described in the Quality Assurance Project Plan (QAPP) and is intended to represent conditions in a 4-in. diameter pipe that has been newly lined with cement mortar. The cement-mortar coupons were placed in contact with batches of water for various lengths of time. The water was periodically removed and tested for water quality while the testing vessels were filled with fresh water. This represents periodic flushing of newly lined mains.

133 Chapter 8: Evaluation of Water Quality Impacts 99 Epoxy The epoxy material tested was manufactured by Cohesant Materials of Broken Arrow, Okla. It is certified by NSF for potable water applications. It is a solvent-free, 100% solids epoxy. The mature samples of lined pipes had been lined with another epoxy manufactured by Cohesant Materials of Broken Arrow, Okla. and certified by NSF for potable water applications. Polyurethane The polyurethane material tested was manufactured by Madison Chemical Industries of Milton, Ont., Canada. This product has NSF-61 approval for use in potable water pipes. CONSTRUCTION PARAMETERS It is known that several factors can influence the integrity of the applied liner. These include the methods used to clean and prepare the interior pipe surface, the mixing and preparation of the material used to line the pipe, and the curing time before the liner is placed in contact with the water. The materials, mixing ratios, mixing time, and curing time used in the preparation of the cement-mortar coupons are described below. The coating materials, mixing ratios, mixing time, and curing times used in the preparation of the coupons were similar to those used in actual pipe-lining projects following manufacturers recommendations and AWWA manuals of practice and standards. Cement Mortar The construction parameters for the cement-mortar coupons tested are described as follows: 1. Technology cast in mold coupons 8 cm x 8 cm x 1cm 2. Material Type II Portland cement 3. Cement ratio - 1:1 sand:cement weight-to-weight (w/w) 4. Water ratio 0.43 water:cement w/w 5. Curing time: a. 15 minutes mixing time b. 90 minutes working time c. 24-hour curing time 6. Surface preparation The cement-mortar coupons were not adhered to any other materials, thus, no surface preparation was necessary. The cement-mortar coupons were placed directly in contact with the water and were not molded against any other surfaces. Therefore this experiment does not simulate or test the impact of surface preparation methods as a variable. Epoxy The construction parameters for the epoxy coupons tested are described as follows: 1. Technology sprayed onto glass blanks (8.1 cm x 8.7 cm x 0.52 cm)

134 100 Impacts of Lining Materials on Water Quality 2. Material Manufacturer - Cohesant Materials 3. Application ratio 3:1 resin (Part A) to hardener (Part B) 4. Curing time: a. 5-hour curing time at 72 ºF b. Manufacturer specifies that after cure a 15-minute flush is required prior to being placed into service c. Testing was performed within 48 hours after coating 5. Application surface 8.1-cm x 8.7-cm x 0.52-cm glass coupons 6. Film thickness 30 mils (manufacturer recommends 10 to 1,000 mils minimum and maximum coating thickness) 7. Surface preparation The glass coupons were sand-blasted and roughened before the epoxy material was applied. 8. Cleaning All coupons were disinfected prior to beginning the study by soaking the coupons in a 200 mg/l Cl 2 solution for 30 minutes. Polyurethane The polyurethane is a two-component product that requires a Graco model EX-7 plural component spray gun to apply. The construction parameters for the polyurethane lining are as follows: 1. Technology sprayed onto glass blanks, (8.1 cm x 8.7 cm x 0.52 cm) 2. Material Manufacturer - Madison Chemical Industries 3. Application ratio 1:1 4. Curing time: a. 3 minutes setting time b. 8 minutes curing time before handling c. 24-hour curing time before immersion (manufacturer s recommendation) d. The coupons were tested within 48 hours after coating 5. Surface preparation The glass coupons were sand-blasted and roughened before the polyurethane material was applied BACKGROUND WATER QUALITY The background water quality used in the laboratory tests represented four different combinations of ph and disinfectant, which represented finished water qualities that can be measured in public water supply systems. Test waters 1, 3, and 4 had a ph of 8, while test water 2 had a ph of 6.5. Test water 1 had no disinfectant. Test waters 2 and 3 had 2-mg/L chlorine disinfectant. Test water 4 had 4 mg/l of monochloroamine. The four test waters are summarized in Table 8.1. Cement-mortar coupons were tested with all four waters listed in Table 8.1. Coupons cured for 60 hours before leaching tests for the epoxy-coated coupons began. The epoxy coupons were tested with test waters 1, 3, and 4 listed in Table 8.1. The epoxy coupons were not tested with the low ph water 2. The polyurethane coupons were tested with test waters 1, 3, and 4 listed in Table 8.1. The polyurethane coupons were not tested with the low ph water 2.

135 Chapter 8: Evaluation of Water Quality Impacts 101 Table 8.1 Composition of four test waters used in leaching tests Water ID ph Alkalinity Hardness Disinfectant Disinfectant s.u. mg/l mg/l as CaCO 3 mg/l Name N/A Chlorine Chlorine Monochloramine WATER QUALITY IMPACTS The impacts on the quality of the waters in contact with three lining materials are summarized in Table 8.2. The impacts are graded qualitatively as low (L), medium (M), or high (H). Cement-Mortar Lining The greatest impact of cement-mortar lining on water quality is to increase the ph, alkalinity, calcium, aluminum, and total dissolved solids (TDS) in the water in contact with the lining. The increase in the calcium concentration in water was the cause of elevated hardness of the water. In the laboratory leaching tests, cement-mortar lining increased the aluminum concentrations in the test waters. The cement-mortar material created a significant increase in the alkalinity of all four test waters. The impact on alkalinity in general terms was the same for all four test waters. The alkalinity stabilized at 100 mg/l as CaCO 3 after day 19 of the test. The background alkalinity of the test waters was 35 mg/l. Sudden drops in alkalinity between days 9 and 11, and between days 14 and 15, are due to changing of water in 2 days, and 1 day, respectively. These are equivalent to 2-day and 1-day interval flushing. Douglas and Merrill (1991) reported similar impacts of cement-mortar lining on water alkalinity in South Central Connecticut Regional Water Authority (SCCRWA) water, which also has low alkalinity. They also tested waters with continuous flushing, 1-day flushing, and weekly flushing and found impacts increased with flushing interval. Continuous flushing had minimum impact whereas weekly flushing had maximum impact. Alkalinity stabilized at 40 mg/l as CaCO 3 after 3 weeks under the weekly flushing condition, and 70 mg/l as CaCO 3 after 9 weeks under the weekly flushing condition. Cement mortar increased the ph of the test waters. The resulting ph values of the test waters in contact with the cement mortar were nearly identical for all four test waters. The starting ph of test waters did not impact the resulting ph after contact with the cement mortar. The ph after the first day in contact with the cement mortar was After 30 days in contact with the cement mortar and 9 changes of water, the ph was Therefore, cement-mortar s impact on ph was persistent during the 30-day period of the test. The ph of the test waters exceeded the drinking water secondary maximum contaminant level (SMCL) of 8.5 standard units (s.u).

136 Lining material Cement mortar Epoxy Polyurethane Mature epoxy Table 8.2 Knowledge matrix leaching from lining materials and its effect on water quality TOC M H M H ph H L H L Alkalinity H L L L Hardness Disinfectant residual THMs HAA5 Note: Odor was not tested on mature epoxy samples. * THMs were only significant in the chlorinated test waters (2 and 3). Aluminum concentrations increased significantly. Concentrations of other metals increased only slightly. Decay rate decreases with exposure time. Monochloramine has low (L) impact while chlorine has high (H) impact. H L L L L H H H M* M L M L L * L Metals H L L L Odor L H H N/A 102 Impacts of Lining Materials on Water Quality

137 Chapter 8: Evaluation of Water Quality Impacts 103 The increased ph and alkalinity that resulted from cement-mortar lining have the potential to cause some water utilities to be outside the ranges of ph and alkalinity established for them under the lead and copper rule. Douglas and Merrill (1991) also found an increase of ph value from 7.2 to about 12 after 1 week and to about 10.5 to 11 after 3 weeks of testing, which is similar to the findings of this study. For a continuous flow condition, however, there was no significant change in ph. Cement mortar increased the calcium concentrations in the test waters. The measured calcium concentrations in the test waters over time are shown in Figure 8.2. Cement mortar increased the calcium concentrations in the test waters. Increased concentrations of calcium were measured over a 14-day period, after which calcium stabilized and remained at 7 to 9 mg/l as calcium, slightly below the background concentration of 11.5 mg/l as calcium. Except for the initial calcium concentrations measured on day 1, the calcium concentrations measured in all four test waters followed the same trend and were quite close in value. The differences in initial water composition did not result in different calcium leaching rates from the cement mortar. Douglas and Merrill (1991) in their tests with SCCRWA low alkalinity water also found that calcium was stabilized at about 12 to 14 mg/l as calcium after 3 to 7 weeks, depending on the frequency of flushing. The aluminum concentrations increased in all four test waters after being in contact with the cement-mortar coupons. All four test waters showed the same trend in aluminum concentrations. The aluminum concentrations increased from day 1 through day 9, then decreased dramatically at day 11 of the test. The concentrations of aluminum exceeded the drinking water SMCL of 200 µg/l for the first 9 days of the test. Figure 8.2 Calcium concentrations in water in contact with cement mortar

138 104 Impacts of Lining Materials on Water Quality Construction Methods Douglas and Merrill (1991), in their study of water quality impacts due to cement-mortar lining, also tested the impacts of construction methods on water quality. They used two different cement-mortar lining construction methods, drag and trowel method, and centrifugal method of lining. They found that after the first few weeks, the centrifugally lined section performed better than the drag and trowel section. However, the performances of the two sections became similar after the ninth week, and the water quality differences were minimal. After 13 weeks of tests, the water qualities of both sections had a ph of 11.1 to 11.3, alkalinity of 80 to 110 mg/l as CaCO 3, and calcium of 80 to 85 mg/l as CaCO 3. Summary for Cement-Mortar Lining The highest concentrations of measured constituents during the cement-mortar leaching tests are summarized in Table 8.3. Table 8.3 Highest concentrations measured during tests of cement mortar Parameter Highest concentration measured Control sample concentration Units Alkalinity mg/l ph s.u. Hardness mg/l as CaCO 3 Calcium mg/l Magnesium mg/l Aluminum mg/l Chromium mg/l Ammonia mg/l as N Total dissolved solids 1, mg/l Total organic carbon mg/l Total trihalomethanes mg/l Total haloacetic acids mg/l The concentrations of these parameters were measured in batches of water exposed to the cement-mortar coupons in closed vessels. The test conditions are most similar to dead-end, 4-in. diameter, water mains. The impacts and concentrations are expected to be lower in flowingwater distribution pipes. During the laboratory tests, the highest concentrations of most constituents were measured after the initial exposure periods and generally decreased with time. This is believed to be due to curing of the cement-mortar test coupons and gradual exhaustion of the store of leachable constituents from the cement-mortar matrix. An exception to this was ammonia, which was a byproduct of monochloramine decay. For samples tested with monochloramine disinfectant, ammonia was measured in control and test waters at similar concentrations. The age of the cement mortar did not affect the ammonia concentrations. The main water quality impact of cement-mortar linings is to increase ph, alkalinity, calcium, and aluminum. The duration of the increases in alkalinity, calcium, and aluminum are

139 Chapter 8: Evaluation of Water Quality Impacts 105 finite and lasts from 2 to 4 weeks. The impact on ph is more persistent. Flushing of water mains after cement-mortar lining will reduce the water quality impact. There is no evidence in the literature that indicated that the loss of calcium from the cement-mortar matrix weakened the structural integrity of the cement-mortar lining. The ph, alkalinity, and hardness of the water in contact with cement-mortar linings did not appear to significantly shorten the useful life of the liner. There were reports that cement-mortar linings could be eroded by sand and grit in the water and by water flowing at high velocities. Large diameter pipes have a greater volume to surface area ratio than smaller pipes, thus, the impact of cement-mortar linings on water quality is less for larger diameter pipes than for smaller diameter pipes. The laboratory tests for this study were performed at a surface area to volume ratio equivalent to a 4-in. diameter pipe. An 8-in. pipe would provide four times the dilution volume of a 4-in. pipe. Strategies for minimizing the water quality impact from cementmortar linings include the following: Monitor the water quality in the newly lined mains. Maintain the water quality in the main by flushing the newly lined main as needed. Consider lining small, dead-end mains with other lining materials, such as epoxy or polyurethane. Epoxy Lining The greatest impact from epoxy was on TOC and disinfectant residual concentrations. Epoxy lining produced significant increases in TOC concentrations for all three test waters. Although the concentrations of TOC decreased with time, there were still measurable increases in TOC concentrations at the end of the 30-day exposure period. The chlorinated water produced the highest TOC concentrations of the three test waters. There is no drinking water standard for TOC. There is a treatment technology rule that prescribes the percent removal of TOC as a function of source water TOC concentration and alkalinity (see Table 8.3). From Table 8.3, the allowable TOC concentration for low alkalinity water is between 1.2 and 4 mg/l, after applying the required removal rate. Epoxy reduced the concentrations of both chlorine and chloramine disinfectants. Free chlorine was consumed at a greater rate than chloramine. This effect continued throughout the 30-day test period. Mature pipe samples lined with epoxy reduced the disinfectant residual concentrations. Water exposed to epoxy showed some increase in the THM and HAA5 concentrations, but none of these increases exceeded the drinking water standards. The increases in THM and HAA5 concentrations were greatest in the chlorinated water. Leaching of bis-phenol A (BPA), an unregulated compound, was detected in water samples in contact with epoxy. Polyurethane (PU) Lining Three standard water compositions were tested before and after contact with polyurethane-coated coupons. The three water types were: ph = 8.0, low alkalinity/hardness, no disinfectant ph = 8.0, low alkalinity/hardness, 2.0 mg/l chlorine ph = 8.0, low alkalinity/hardness, 4.0 to 6.0 mg/l monochloramine

140 106 Impacts of Lining Materials on Water Quality The greatest impact of polyurethane on water quality was to decrease the ph by 2 to 3 s.u. The magnitude of the drop in ph was directly proportional to the duration of the contact between the polyurethane-coated coupons with the water. Polyurethane consumed chlorine and chloramine disinfectant although chlorine was consumed at a greater rate than chloramine. The consumption rate decreased over time but still persisted at the end of 30 days of testing. Polyurethane promoted the creation of HAA5 (five regulated HAA5 compounds). None of the HAA5 concentrations exceeded the drinking water standard of 60 µg/l. HEALTH IMPACTS It is important to know which compounds may be leached from lining materials and under what conditions leaching is likely to occur. Impacts from leached compounds in drinking water can be health impacts and aesthetic impacts. EPA has established primary and secondary standards for drinking water. Table 8.4 contains a list of constituents monitored during laboratory-scale leaching tests and their potential impacts on drinking water quality. The list of constituents in Table 8.4 is not exhaustive but is limited to the constituents monitored during the leaching tests. Cement Mortar In the laboratory-scale testing performed as part of this study cement mortar was associated with elevated concentrations of alkalinity, calcium (hardness), aluminum, TDS, and increased ph. These increases have the potential to exceed secondary drinking water standards. They do not represent a health threat but may generate customer complaints. Epoxy Epoxy was associated with a significant increase in the TOC concentrations and odor. There was a significant decrease in the disinfectant residual (both chlorine and chloramine) in water in contact with epoxy. Epoxy material has the potential to reduce the disinfectant concentration below the minimum drinking water standards or disinfectant residual. Epoxy has the potential to increase TOC concentrations above the guidelines prescribed by the treatment technology (TT) removal requirements for TOC. TOC does not represent a health threat by itself but could be converted to HAA5 in the presence of chlorine, which does represent a potential health threat. Although there is no numerical standard for odor, odors may generate customer complaints and loss of customer goodwill.

141 Chapter 8: Evaluation of Water Quality Impacts 107 Table 8.4 Constituents monitored during laboratory leaching tests Secondary Parameter Primary health impacts impacts Other impacts Alkalinity N/A Aesthetic Low ph: bitter taste High ph: slippery feel and soda taste ph N/A Hardness N/A Increases soap usage, precipitates Calcium N/A Same as hardness Magnesium N/A Same as hardness Aluminum N/A Aesthetic colored water Chromium Allergic dermatitis Ammonia N/A Aesthetic bitter or salty taste, colored water Can form chloramines; could possibly contribute to nitrification in distribution system TDS N/A Technical scaling TOC THMs HAA5 Bis-phenol A Odor Cancer risk, reproductive, liver, kidney, and nervous system impacts Cancer risk Can create THMs Potential developmental and reproduction effects; undesirable to customers

142 108 Impacts of Lining Materials on Water Quality Polyurethane Polyurethane was associated with decreases in ph, loss of disinfectant residual, and odors. The decrease in ph associated with freshly applied polyurethane material has the potential to decrease the ph below the secondary standard of 6.5 s.u. The loss of disinfectant residual was significant for both chlorine and chloramine. Although there is no numerical standard for odor, odors may generate customer complaints and loss of customer goodwill. REGULATORY IMPACTS The numerical regulatory limits established by EPA for drinking water are listed in Table 8.5. Cement Mortar None of the primary drinking water limits or action levels were exceeded by any measured concentrations during the leaching tests for cement mortar. The secondary standards for aluminum, ph, TDS, and odor were exceeded. The elevated ph and alkalinity could impact limits on ph and alkalinity established for some water utilities under the lead and copper rule. The treatment technology rule for TOC would not strictly apply to elevated TOC concentrations originating from newly placed cement-mortar liners. However, the removal efficiencies for TOC listed in Table 8.6 could be achieved by flushing the newly lined water mains, either periodically or continuously. Epoxy None of the primary drinking water limits or action levels were exceeded in any of the test waters in contact with epoxy. The secondary standard for odor was exceeded. Polyurethane (PU) None of the primary drinking water limits or action levels were exceeded in any of the test waters in contact with polyurethane. The secondary standard for odor was exceeded. Polyurethane has the potential to lower the ph below the secondary limit of 6.5 if the source water ph is below 8.5 s.u. This may have impacts on the lead and copper rule requirements.

143 Chapter 8: Evaluation of Water Quality Impacts 109 Table 8.5 Drinking water quality limits for constituents of interest Regulatory Parameter limit Units Alkalinity N/A mg/l as CaCO 3 ph s.u. Hardness N/A mg/l as CaCO 3 Calcium N/A mg/l Magnesium N/A mg/l Aluminum mg/l Chromium 0.1 mg/l Ammonia N/A mg/l as N TDS 500 mg/l TOC TT mg/l THMs mg/l HAA mg/l Odor 3 Threshold odor number (TON) TT = Limited by treatment technique Table 8.6 Required removal of TOC from source water* Source water TOC (mg/l) Source water alkalinity (mg/l as CaCO 3 ) 0-60 > >120 > % 25.0% 15.0% > % 35.0% 25.0% > % 40.0% 30.0% * Systems meeting at least one of the alternative compliance criteria in the rule are not required to meet the removals in this table. Systems practicing softening must meet the TOC removal requirements in the last column to the right. POTENTIAL REGULATORY MEASURES There were no observed water quality impacts from cement-mortar, epoxy, or polyurethane lining that are subject to anticipated future regulation.

144

145 CHAPTER 9 DEVELOPMENT OF A METHODOLOGY FOR SELECTION OF A LINING MATERIAL INTRODUCTION In this chapter a simplified procedure is presented that provides guidance to water utilities in the selection of lining materials for pipe-lining projects. The methodology will incorporate the useful life of alternative lining materials, comparative costs, and compatibilities and incompatibilities between water quality and liner materials in the analysis. METHODOLOGY FOR ANALYSIS OF WATER QUALITY IMPACTS A standardized methodology for analysis of water quality impacts by synthesizing the leaching test data was developed. The methodology includes procedures for evaluating the data measured during the bench-scale tests described in Chapters 5 through 7 in this report. The data measured during the bench-scale tests were used in the methodology to estimate the water quality impacts from lining materials on distribution mains in the field. The methodology involves three procedural steps. These are: Step 1. Convert concentration increases to daily rates Step 2. Calculate mass loss rates Step 3. Apply mass loss rates to field-scale conditions Development of Methodology for Mass-Based Parameters The development and application of these steps for mass-based parameters are described in the following subsections. Step 1. Convert Concentration Increases to Daily Rates The laboratory bench-scale leaching tests involved taking samples of water in contact with the test coupons at irregular intervals. The ratio of the surface area of the lining material to the volume of water in the test vessel was the same as a 4-in. diameter pipe. The water in the vessels containing the test coupons was replaced with fresh water after each interval of contact time. Using alkalinity concentration data measured during the cement-mortar tests as an example, the alkalinity concentrations were converted to a daily rate by dividing the increase in alkalinity above background by the number of days in the contact interval. The rate of alkalinity concentration increase, in milligrams per liter per day (mg/l/day) was greater when the cement mortar was fresh and decreased as the cement mortar aged. Figure 9.1 shows the alkalinity concentration rate versus age of the cement-mortar coupons. The calculated alkalinity rates are plotted at the midpoint of each exposure interval. The alkalinity concentration increases are not significantly affected by the test water compositions. All four test waters show the same pattern of decreasing impact with time. The 111

146 112 Impacts of Lining Materials on Water Quality Figure 9.1 Rate of alkalinity increase versus age of cement-mortar lining plots of the four test waters are nearly identical. A 1-day alkalinity measurement was only reported for test water 2. The initial increase in alkalinity is nearly 600 mg/l/day on the first day of exposure. This rate drops off rapidly and falls below 100 mg/l/day after day 5. This suggests that a 4-in. diameter low flow or dead-end pipe should be flushed on a daily basis to keep the alkalinity below 100 mg/l at the customer s tap. Step 2. Calculate Mass Loss Rates In order to apply these results to pipes of any size, and any flow rate the alkalinity concentration rates (mg/l/day) were converted to a specific mass-leaching rate (milligrams per square centimeter per day [mg/cm 2 /day]) by multiplying by the net volume of the test vessel (2.3 L) and dividing by the surface area of cement mortar exposed to the water (912 cm 2 ). The alkalinity specific mass-leaching rate for the cement-mortar coupons in contact with four test waters is shown in Figure 9.2. From Figure 9.2 it is clear that the mass loss of alkalinity from cement mortar is not constant but decreases with time. A regression curve was developed from the data so that the mass loss of alkalinity could be described analytically. The mathematical mass loss function can be applied to pipes of any diameter and length under any flow condition to calculate the resulting alkalinity concentration in water. The regression equation and coefficient of determination (r 2 ) are shown in Figure 9.2. The impact of source water alkalinity was evaluated by comparing the laboratory results from this study, which used source water with an alkalinity of 35 mg/l (as CaCO 3 ), with the results of a previous AwwaRF #415 study (Douglas and Merrill 1991), which had a source water alkalinity of 13 mg/l (as CaCO 3 ). The loss rates for alkalinity, normalized by surface area, were developed for both sets of data. The alkalinity loss rates (in mg/cm 2 /day) and best fit regression curves for

147 Chapter 9: Development of a Methodology for Selection of a Lining Material 113 both studies are shown in Figure 9.3. The 35-mg/L alkalinity water has an initial alkalinity loss rate of 1.5 mg/cm 2 /day. This rate rapidly decreases to 0.25 mg/cm 2 /day at day 4, then to less than 0.05 mg/cm 2 /day by day 27. The mass loss rate for the low alkalinity water is initially less than 0.4 g/cm 2 /day. This rate decreased slowly and is still above 0.1 mg/cm 2 /day at day 30. The loss of alkalinity persisted longer for the lower alkalinity water. The data shown in Figure 9.2 are limited to two studies. They suggest that loss of alkalinity from cement mortar persists longer for low alkalinity (13 mg/l as CaCO 3 ) water than from higher alkalinity (35 mg/l as CaCO 3 ) water. Mass loss rate curves were developed for all parameters that were considered to be high impact, except for odor and ph, which are not amenable for evaluating as a mass rate Alkalinity Mass Loss Rate, mg/cm 2 /day y = x r 2 = Water #1 Water #2 Water #3 Water #4 Regression (Water #2) Age of Cement Mortar, days Figure 9.2 Alkalinity mass loss rate versus age of cement-mortar lining from laboratory tests

148 114 Impacts of Lining Materials on Water Quality Figure 9.3 Comparison of alkalinity mass loss rates between laboratory and field tests Step 3. Apply Mass Loss Rates to Field-Scale Conditions An impact module was developed in an Excel spreadsheet. The function of the impact module is to calculate the increase in water concentrations due to the mass loss of constituents from lining materials. The impact module uses the regression equation developed from the mass loss curves described in Step 2. This function, along with pipe diameter, length, and daily flow through the pipe, are used to calculate the change in alkalinity concentration in water passing through the pipe at any time after liner installation. The mass is assumed to be released from the lining material in accordance with the regression equation and is uniformly distributed over the volume of water passing through the pipe section each day. The calculations take into account the interior surface area of the pipe (the source), the volume of the pipe section, and the total volume of water flowing through the pipe each day. Using the normalized mass curve for alkalinity (Figure 9.2), alkalinity concentration increase in three 500-ft long sections of 4-in., 6-in., and 8-in. diameter pipes with 6,000 gallons per day (gpd) usage is calculated using the impact module and plotted in Figure 9.4. For a constant flow it is apparent that water stagnates more in a larger diameter pipe, resulting in more water quality impact than a smaller diameter pipe. Next, a constant 1-day hydraulic residence time (HRT) was considered for all size pipes, and applied to each pipe. Each pipe is 500 ft long. Figure 9.5 represents the increased alkalinity concentrations for the first month after applying cement-mortar lining under daily flushing conditions. Under identical HRT conditions, larger diameter pipes have a smaller increase in alkalinity because the volume of water contained in the pipe increases in proportion to the square of the diameter while the interior surface area of the pipe increases in direct proportion to the

149 Chapter 9: Development of a Methodology for Selection of a Lining Material 115 diameter. Therefore, the ratio of water volume to pipe surface area is greater in large diameter pipes, resulting in less impact from alkalinity leached from the cement-mortar liner. A 1-day HRT is equivalent to one pipe volume exchange of water flowing through a pipe each day. As the number of exchange volumes flowing through a pipe each day increases, either from customer usage or flushing activities, the impact from the lining material decreases. Figure 9.6 illustrates this for a 6-in. diameter pipe 500 ft in length. The volume of water in 500 ft of 6-in. diameter water main is 734 gallons. One exchange volume is 734 gallons. Figure 9.6 illustrates that the increased alkalinity concentration is lower when the number of pipe volume exchanges is greater Alkalinity Concentration Increase, mg/l inch pipe 6-inch pipe 8-inch pipe Age of cement mortar, days Figure 9.4 Alkalinity increases in water versus age of cement-mortar 6,000 gpd

150 116 Impacts of Lining Materials on Water Quality Alkalinity Concentration Increase, mg/l inch 6-inch 8-inch 10-inch 12-inch 18-inch Age of Cement Mortar, days Figure 9.5 Effect of pipe diameter on increased alkalinity under uniform 1-day hydraulic residence time Figure 9.6 illustrates that the initial water quality impacts from cement-mortar lining are less when there is more flow through the hypothetical water main. The exchange volumes can be due to customer usage, flushing, or a combination of the two. Let us assume that the water utility has established a policy that the allowable increase in alkalinity after a cement-mortar lining project is 100 mg/l above background. If the water usage from this water main is 3 exchange volumes (2,202 gallons) or greater, the alkalinity increase will be less than the target of 100 mg/l and no additional flushing will be needed. If the water demand on the main is 734 gpd (one exchange volume), Figure 9.6 illustrates that two additional pipe volumes of flow on day 1 will keep the alkalinity increase below 100 mg/l. On day 2, only one additional pipe volume of water flushed through the pipe will maintain the alkalinity increase below the target value of 100 mg/l. If the main has no water demand, the required flushing to maintain the target water quality would be 3 exchange volumes flushed on day 1, 2 volumes flushed on day 2, and 1 volume flushed on day 3. No additional flushing would be needed after day 3. Using the same procedure, the user can calculate the water quality impacts of other parameters using the mass rate curves developed for all high impact parameters for all three lining materials considered in this study. The following mass loss rate curves were developed and are shown in Appendix C: Alkalinity from cement-mortar lining Calcium from cement-mortar lining

151 Chapter 9: Development of a Methodology for Selection of a Lining Material 117 Aluminum from cement-mortar lining Total dissolved solids (TDS) from cement-mortar lining Total organic carbon (TOC) from epoxy lining p 300 Alkalinity Concentration Increase, mg/l exchanges/day 2 exchanges/day 3 exchanges/day 4 exchanges/day 5 exchanges/day 6 exchanges/day Age of Cement Mortar, days Figure 9.6 Impact of flushing on alkalinity concentration on a 6-in. pipe Disinfectant demand from epoxy lining Disinfectant demand from polyurethane lining HAA5 formation rate from polyurethane lining The power and exponential regression equations developed to the mass release curves are shown as follows: Exponential equation: y = a e bx Power equation: y = a x b Where: y = constituent of interest; x = age of lining in days; and a and b are the coefficient and exponent of the equations. The values of a and b coefficients and exponents of all the equations are summarized in Table 9.1. The impact module has been simulated in a spreadsheet. It has a user input sheet where the user can specify the lining materials to be analyzed, the compounds of interest, and mass release functions. The length, diameter, and daily water usage on the water main are also specified on the input screen. For constituents other than ph, the impact module calculates the

152 118 Impacts of Lining Materials on Water Quality concentration increase of these constituents above background. For ph, the impact module adds the increase in ph to the background ph and calculates the resulting ph. The impact module calculates the increase in concentration for each constituent and lining material combination similar to the example in Figure 9.4. The user can determine the critical or limiting constituent and the number of additional flushing operations, if any, needed to maintain the concentrations below target levels. Development of Methodology for ph The impact of cement-mortar material on the ph of water is significant and persistent. The ph is elevated significantly in water that is in contact with fresh cement-mortar material. This was measured in the laboratory tests performed as part of this study, as well as in a previous AwwaRF #415 study (Douglas and Merrill 1991). Figure 9.7 is a plot of measured ph values from this study and the previous study (Douglas and Merrill 1991) showing ph measured for water in contact with 4-in. diameter pipes, 3 ft long, lined with cement mortar. The set of points near ph 7 represent ph measurements of water continuously flowing at 10 gallons per minute (gpm) through the pipe, which is equivalent to five volume exchanges per minute in the test pipes. The ph values measured in water flowing continuously through the pipes were very close to the source water ph value. The lower regression line is fitted to data from tests where the water was changed on a daily basis (i.e., daily flushing). The upper regression curve is fitted to data that approximate weekly flushing. It is apparent that the frequency of flushing can mitigate the impact of cement mortar on ph. The impact of cement mortar on ph persists beyond the 90- day test period. The ph values measured during the laboratory testing with cement mortar performed in this study fall between the daily and weekly flushing regression lines. This is consistent with the water change-out frequencies used in the laboratory tests, which were between 1 and 6 days.

153 Material Disinfectant Table 9.1 Summary of regression equations for mass loss curves Constituent of interest, y Units Regression form Coefficient a Exponent b Coefficient of determination, r 2 CM Waters 1-4 Chromium µg/cm 2 /day Power y = x CM Water 3 Aluminum µg/cm 2 /day Exponential y = e x CM Water 2 Alkalinity mg/cm 2 /day Power y = x CM Water 1 Calcium mg/cm 2 /day Power y = x CM Water 2 HAA5 µg/cm 2 /day Exponential y = e x CM Water 1 TDS mg/cm 2 /day Power y = x CM Water 2 TOC mg/cm 2 /day Exponential y = e x Epoxy Chlorine Disinfectant loss mg/cm 2 /day Power y = x Epoxy Monochloramine Disinfectant loss mg/cm 2 /day Power y = x Mature epoxy Chlorine Disinfectant loss mg/cm 2 /day Power y = x Epoxy No disinfectant TOC mg/cm 2 /day Power y = x Epoxy Chlorine TOC mg/cm 2 /day Power y = x Epoxy Monochloramine TOC mg/cm 2 /day Power y = x Mature epoxy Chlorine TOC mg/cm 2 /day Power y = x PU Chlorine Disinfectant loss mg/cm 2 /day Power y = x PU Chlorine HAA5 µg/cm 2 /day Power y = 0.028x CM Chlorine Calcium * mg/cm 2 /day Power y = x CM Chlorine Alkalinity mg/cm 2 /day Power y = x Note: CM = Cement mortar PU = Polyurethane * Daily calcium data from AwwaRF Project #415 (Douglas and Merrill 1991) Daily alkalinity data from AwwaRF Project #415 (Douglas and Merrill 1991) Equation Chapter 9: Development of a Methodology for Selection of a Lining Material 119

154 120 Impacts of Lining Materials on Water Quality Figure 9.7 Variation of ph with time for daily and weekly flushing frequencies The HRT represents the contact time between the water and the cement-mortar material and is calculated as the inverse of the flushing frequency. The ph data presented in Figure 9.7 represent HRTs of 7 days, 1 day, and 0.2 minute (0.001 day). A relationship between an increment of ph and age of cement-mortar lining for various HRT values was developed. The ph increase versus age of cement mortar is shown in Figure 9.8 for weekly, daily, and continuous flushing along with the data from a test case of South Central Connecticut Regional Water Authority (SCCRWA) Woodward Avenue lining project, which represented 0.32 day HRT. The continuous flushing with very low HRT (0.001 day) showed no increase in ph above background. Best fit lines were fitted through three data sets representing 0.32-day, 1-day, and 7- days HRT. These are shown in Figure 9.9. Data for Figure 9.9 were used to develop a ph increase versus HRT relationship for six different cement mortar ages (10, 20, 30, 60, 90, and 100 days) and are shown in Figure The methodology for evaluating ph impacts from cement mortar requires that a target maximum ph value be selected so that a maximum ph increase can be calculated. The maximum increase in ph is calculated as the difference between the background ph and the maximum target ph. The required, or target, HRT is estimated from Figure 9.10 and the appropriate curve representing the age of the cement mortar. The existing and required HRTs are compared. If the required HRT is greater than the existing HRT, then no supplemental flushing is required to maintain the ph within acceptable

155 Chapter 9: Development of a Methodology for Selection of a Lining Material 121 limits. If the required HRT is less than the actual HRT, then additional flushing or other water usage is needed to maintain the ph within acceptable limits. Figure 9.8 ph increase versus age of cement mortar for four HRTs Figure 9.9 ph increase versus age of cement mortar for three HRTs

156 122 Impacts of Lining Materials on Water Quality ph Increase, s.u days old 20 days old 30 days old 60 days old 90 days old 100 days old Hydraulic Residence Time, Days Figure 9.10 ph increase versus HRT for six ages of cement mortar The number of pipe volumes of water flowing through the pipe is the inverse of the HRT. If additional flushing is needed to control the ph, the procedure for calculating the additional volume of water needed is to take the inverse of the existing and target HRT values, calculate the number of additional pipe volumes of water as the difference, then multiply by the number of gallons in one pipe volume. Considering an example of a pipe with an HRT of 0.4 day, ph of water of 6.5 with an allowable ph of 8.5, the allowable rise of ph will be 2.0. In Figure 9.10, for a ph rise of 2.0, the required HRT is equal to 0.41, which is greater than the existing HRT of 0.4, and therefore, would not require any additional flushing. However, if the HRT of the pipe is 0.5 day, it will be necessary to flush the pipe for about 70 days, with a daily flushing of about 0.5 [(1/0.41) - (1/0.5)] pipe volume. This methodology has been implemented in an Excel spreadsheet titled Impact Module. The Impact Module is attached to this report and is documented in Appendix D. ESTIMATION OF USEFUL LIFE OF LINERS Cement Mortar From utility experiences of North America, it has been found that the performance of cement-mortar lining is very good even after 60 years in operation under normal water conditions. The Los Angeles Department of Water and Power (LADWP) began their cementmortar lining operation in 1944 (Deb et al. 1990). Therefore, from North American water utility experience, the useful life of cement-mortar liners can be conservatively estimated to be 50 to 60 years for regular water. Cement-mortar liners installed in large diameter water mains by

157 Chapter 9: Development of a Methodology for Selection of a Lining Material 123 SCCRWA in 1958 were still in service in 1991 and in good condition (Douglas and Merrill 1991). The water supplied by SCCRWA has low alkalinity. The low alkalinity water had considerable leaching of calcium from cement-mortar lining during the initial few weeks of installation, but no significant evidence has been found to support the opinion that low alkalinity water has significantly reduced the structural integrity or service life of cement-mortar liners. However, for our analysis, it would be safe to assume that for low alkalinity water and for small diameter water mains, the service life of cement-mortar lining can be estimated, at a minimum, to be 40 to 50 years. Epoxy In a previous AwwaRF study (Deb et al. 2006) it was found that the useful service life of epoxy liners was conservatively estimated to be between 50 and 60 years. Epoxy-coated plates were exposed to an environment to induce accelerated aging of the coating. The barrier properties of the epoxy coatings were tested using electrochemical impedance spectroscopy (EIS). Specimens of pipes that had been lined with the same epoxy material and placed back into service for periods of 4 to 13 years were also tested with EIS. Using regression analysis, the useful life of the epoxy material was calculated to be in excess of 120 years. After applying a safety factor to account for imperfections in surface preparation and coating application, it was concluded that the useful life of a water main can be extended by 50 to 60 years by rehabilitating the main using epoxy lining. This assumes that the residual strength of the host pipe is sufficient to provide a remaining life equal to or greater than the lining material. Because epoxy lining is a nonstructural lining, it is not proper to place a liner in a host pipe that has a shorter remaining life. This logic is applicable to any nonstructural lining material. Polyurethane (PU) The useful life of polyurethane lining materials is expected to be at least as long as epoxy materials. In a previous AwwaRF study (Deb et al. 2006) it was found that the polyurethane material tested had excellent barrier properties, as measured with EIS. Those barrier properties remained stable and did not decrease during long-term accelerated aging tests. Therefore, it is expected that polyurethane will have a useful service life at least as long as epoxy, possibly longer. No pipe samples historically lined with polyurethane were tested in the AwwaRF study (Deb et al. 2006). ECONOMIC ANALYSIS OF LINERS The cost of a lining project can vary greatly if the project includes the restoration of shutoff valves, curb stops, valve replacement, and appurtenances such as fire hydrants. Cementmortar lining projects require more material than epoxy- or polyurethane-lining projects because a thicker liner is used for cement mortar than for epoxy or polyurethane. A cost module was developed earlier in another AwwaRF project (#2519) (Deb et al. 2002) to calculate the cost of cleaning and lining water mains, but has not been used in this project. The cost to mitigate potential water quality impacts from lining materials is due to flushing of water mains for a limited period of time after the new lining is installed. Because these costs will occur shortly after liner installation, they can be considered part of the construction cost. The cost of mitigating the water quality impacts for the lining with three

158 124 Impacts of Lining Materials on Water Quality materials of interest (cement mortar, epoxy, and polyurethane) can be compared directly and considered when evaluating the merits of alternative lining materials. Water quality impacts from newly lined mains may require daily flushing until the impacts are within acceptable levels. This will vary between lining materials and constituents of interest. The specific flushing requirements are discussed in Chapter 10. A standalone cost module for flushing is developed and implemented in an Excel spreadsheet. The cost module can calculate additional flushing costs to meet water quality requirements for each of three lining materials. As an alternative, flushing costs may be estimated from the utility s experience with previous flushing projects. SUMMARY While selecting the best suited water main lining material a water utility may use the impact module spreadsheet to analyze mass rate curves to evaluate the impact of water quality and cost on any pipe in their distribution system. The following steps can be used by water utilities to identify the impacts of pipe lining on water quality, to recommend a flushing frequency that will maintain acceptable water quality, and to calculate the costs of additional flushing: 1. Identify worst case areas in the distribution system where waters are relatively stagnant (areas such as low residual chlorine, dead ends, etc.). The water quality impacts from lining materials will be greatest in these areas. The recommended flushing frequencies calculated using the methodology will be adequate for any pipes in the distribution system. 2. Select pipes from the area that the utility wants to analyze (size and length of the pipe). For this example a 4-in. diameter pipe 500 ft in length has been selected. 3. Identify the number of customers or houses connected with the pipe. Assuming 10 customers on each side of the pipe, the pipe in this example serves 20 customers. 4. Estimate average water flow in the pipe based on average customer water usage rates and down-flow usage from billing records when available. For example, if there are 20 houses with an average of 3 persons per household and a usage rate of 100 gallons per capita per day (gpcd), the water usage from that pipe would be a minimum of 20*3*100 = 6,000 gpd. This flow value will be larger if there are further water demands in the downstream side of the pipe. If necessary this flow can also be estimated using a hydraulic simulation module for an average day demand. 5. Estimate turnover of water per day. The turnover is the number of pipe volumes that flow through the pipe each day. The volume of a 4-in. pipe, 500 ft long is 326 gallons. For a pipe with 20 customers connected, the turnover is 6,000/326 = 18.4 turnovers or pipe volume exchanges per day. 6. Identify background water quality of the area (alkalinity, ph, calcium, hardness, metals, TOC). We assume a background alkalinity of 35 mg/l, similar to the alkalinity of the test waters used in this study. 7. Decide tolerance of customers to high alkalinity of water and select maximum value of alkalinity that will be acceptable. Let us assume a value of 100 mg/l as CaCO 3 as the maximum acceptable alkalinity due to lining with cement-mortar lining. 8. Using proper mass curves (mg/cm 2 /day) that represent the mass transfer of alkalinity into the water, calculate the resulting alkalinity concentration in the pipe. The mass

159 Chapter 9: Development of a Methodology for Selection of a Lining Material 125 curve for alkalinity is shown as Figure 9.2. The Impact Module spreadsheet was used to calculate the alkalinity concentrations in a hypothetical 500-ft long pipe, nominally 1 block long. A water usage rate of 6,000 gpd was used to represent the demand from 20 residential customers, assuming 3 persons per household and 100 gpcd average water use. The calculated alkalinity concentrations are shown in Figure 9.4. The calculated alkalinity concentrations for 6- and 8-in. pipes are also shown for comparison. The alkalinity concentration increased more in the 8-in. diameter pipe because the constant water demand represented fewer exchanges of water through the 8-in. pipe compared to the 4- and 6-in. diameter pipes. 9. Determine the number of days when water alkalinity will be expected to exceed 100 mg/l as CaCO 3. In the example presented in this chapter, the alkalinity does not exceed the acceptable threshold of 100 mg/l, thus, it is concluded that no water main flushing is needed to mitigate the water quality impact in this case. If the water demand on the pipe had been less, the calculated alkalinity concentration would be greater. For example, if it is only the first 5 days that alkalinity will exceed the limit, then the area distribution system needs to be flushed for 5 days to maintain the alkalinity concentration below the target concentration of 100 mg/l. The required frequency and duration of flushing can be decided on the basis of needed turnover rate and the maximum acceptable alkalinity concentration in the delivered water. 10. Estimate the cost of flushing and add this cost to the total cost of cement-mortar lining while evaluating alternative linings. A cost-estimating spreadsheet has been developed in a previous study (Deb et al. 2002) to calculate the cost of lining. The cost of lining can also be estimated using bids from contractors. The water quality impacts and the cost of mitigating the impacts of the water quality from the lining material can be evaluated on the spreadsheet shown in Appendix D and compared between different lining materials. This impact assessment procedure may also be used to analyze the impact of other constituents released from cement mortar as well as impacts from epoxy and polyurethane. The potential impacts from all three lining materials should be evaluated and compared in order to select the right lining material for the utility. This methodology is illustrated in Chapter 10 by applying the methodology on a test case lining project of a water utility.

160

161 CHAPTER 10 TEST CASE ANALYSIS In this chapter the methodology for evaluating the water quality impacts of lining materials presented in Chapter 9 is applied to a case study. The case study is a cleaning and lining project performed by the South Central Connecticut Regional Water Authority (SCCRWA) in 2006 using cement-mortar lining. In this test case study epoxy and polyurethane linings were also considered for comparison for selection of a lining. CASE STUDY SOUTH CENTRAL CONNECTICUT REGIONAL WATER AUTHORITY WOODWARD AVENUE RENEWAL SCCRWA provides an average of 50 million gallons per day (mgd) to 100,000 customers through 1,700 miles of water mains. They have a program of renewing 8 to 10 miles of water mains per year. Their preferred renewal method is cleaning and lining with cement mortar. A typical 8-mile renewal program impacts 1,500 customers and takes 7 months to complete. Customers are on bypass water service for 3 weeks. The average diameter of a water main is 8 in. The average residential customer uses 230 gpd (2,800 cubic feet per quarter). It is the policy of SCCRWA to flush the water mains for several hours after the main has been lined with cement mortar before the main is restored to service. The mains are also disinfected in accordance with the requirements of AWWA Standard C (Standard for Disinfection of Water Mains). SCCRWA has an unwritten policy to flush cement-mortar-lined water mains when the ph remains above 8.5 (Norris 2008). In July 2006, SCCRWA cleaned and lined 503 ft of 8-in. diameter water main on Woodward Avenue in New Haven, Conn. with cement mortar. The main serves 14 customers, and the average water usage is 4,159 gpd. This represents over 3 pipe volume exchanges per day and an HRT of 0.32 day. The customers were on bypass water supply between May 25, 2006 and June 13, The water main was cleaned and lined between June 1, 2006 and June 9, The water main was restored to service on June 13, 2006 when the liner was 4 days old. ph measurements were taken from the Woodward Avenue water main before and after lining with cement mortar. Alkalinity and hardness were not measured. Therefore, the impact methodology will be demonstrated for ph only. The first ph measurement was taken on June 26, 2006 when the liner was 17 days old. Impact from Cement Mortar SCCRWA measured ph before and after cleaning and lining. These measurements are shown in Figure The measured ph values after the main was restored to service are shown in Figure 10.2, along with a best curve fitted to the data. The 10-, 20-, 30-, 60-, 90-, and 100-day ph increase values were calculated from the data shown in Figure 9.9 and are plotted as points in Figure 9.10, along with the predictive ph increase curves. The predictive curves were developed from the test data of this project and data collected in an AwwaRF study (Douglas and Merrill 1991) in which test lining specimens were not disinfected with high doses of chlorine prior to the start of the testing. 127

162 128 Impacts of Lining Materials on Water Quality ph after Cleaning and Lining ph, s.u. 7 ph before Cleaning and Lining 6 Woodward Avenue ph Data before Cleaning and Lining Woodward Avenue ph after Cleaning and Lining /14/2005 2/2/2006 3/24/2006 5/13/2006 7/2/2006 8/21/ /10/ /29/2006 1/18/2007 Date Figure 10.1 Measured ph before and after cleaning and lining water main on Woodward Avenue ph, s.u Woodward Ave. ph Power Best Fit (Woodward Ave. ph) Age of Cement Mortar, days Figure 10.2 Woodward Avenue ph after cleaning and lining

163 Chapter10: Test Case Analysis 129 Given a background ph at Woodward Avenue of 7.3 and a maximum ph target of 8.5, the maximum ph increase is 1.2 standard units (s.u.). The HRT is 0.32 day. One pipe volume for 503 ft of 8-in. diameter pipe is 1,313 gallons. For the first 10 days after cement-mortar lining an HRT of 0.24 day is needed to keep the ph increase below 1.2 s.u. (Figure 9.10). Because the HRT from water demand alone is 0.32 day, the flow of water through the newly lined pipe needs to be increased in order to decrease the HRT from 0.32 day to 0.24 day. The number of pipe volumes of water flowing through the pipe is the inverse of the HRT. The procedure for calculating the additional volume of water required to limit the ph increase below the target value is to take the inverse of the existing and target HRT values, calculate the number of additional pipe volumes of water as the difference, then multiply by the number of gallons in one pipe volume. These calculations are shown for days 1 through 10 below. Actual Needed Difference HRT, days /HRT, pipe volumes/day Additional pipe volumes needed: Pipe volume = 1,313 gallons Additional flow needed: 1,368 gallons per day for the first 10 days The additional flow could be in the form of flushing, bleeding, or other form of additional water demand. Additional water demand could be in the form of a beneficial use of the water such as storage and irrigation, street sweeping, or dust control. The procedure is repeated for other ages of cement-mortar lining until the required HRT becomes greater than the existing HRT of 0.32 day. In this case, the HRT required to maintain a ph increase less than 1.2 s.u. exceeds 0.32 day after 90 days. Therefore, additional water will have to be withdrawn for the first 90 days after lining to keep the ph increase below 1.2 s.u. In this example, the total additional volume of water needed to keep the ph below 8.5 over the first 90 days after lining is 94,000 gallons. The project parameters (length and size of pipe and flow) for the Woodward Avenue project were entered into the Impact Module spreadsheet to evaluate the relative impacts of cement mortar and alternative lining materials (epoxy and polyurethane) on water quality. Table 10.1 summarizes the results from the Impact Module. The concentrations of alkalinity, calcium, aluminum, and TDS calculated for the Woodward Avenue project by the Impact Module are shown in Figures 10.3 through This result should be compared with the user s input target values for each constituent. Target values should be selected by the utility except aluminum, TDS, and HAA5, which have either primary or secondary standards established (EPA 2003). For this test case the assumed target concentrations of constituents are shown in Table In addition to ph, only one constituent, calcium, exceeded the target concentration set for this example, which was 50 mg/l (Figure 10.5). The calculated calcium concentration was 95 mg/l on day 1 and 25 mg/l on day 2. The Impact Module calculated that the calcium concentration could be decreased to 50 mg/l by flushing 3,700 gallons on day 1. All other constituent values were within the targets and therefore, no further actions were required.

164 130 Impacts of Lining Materials on Water Quality Table 10.1 Summary of impact module predictions for Woodward Avenue test case Duration when target concentration was exceeded (days) Material Constituent Predicted concentration increase Units Target concentration increase Units Cement mortar ph 1.6 s.u. 1.2 s.u. 90 Cement mortar Alkalinity (as 62 mg/l 100 mg/l 0 CaCO 3 ) Cement mortar Calcium (as 95 mg/l 50 mg/l 1 CaCO 3 ) Cement mortar Aluminum 50 µg/l 200 µg/l 0 Cement mortar TDS 138 mg/l 500 mg/l 0 Epoxy TOC 0.2 mg/l 5 mg/l 0 Epoxy Disinfectant 0.21 mg/l 0.5 mg/l 0 loss Polyurethane Disinfectant 0.24 mg/l 0.5 mg/l 0 loss Polyurethane HAA5 1.7 µg/l 60 µg/l 0 Polyurethane ph N/A s.u. 10 s.u. N/A 70 Alkalinity Concentration Increase, mg/l as CaCO Age of Cement Mortar, days Figure 10.3 Alkalinity increase in drinking water versus age of cement mortar for the Woodward Avenue example

165 Chapter10: Test Case Analysis Calcium Concentration Increase, mg/l mg/l Calcium Limit Established for this Example Case Study Age of Cement Mortar, days Figure 10.4 Calcium increase in drinking water versus age of cement mortar for the Woodward Avenue example 60 Aluminum Concentration Increase, µg/l Age of Cement Mortar, days Figure 10.5 Aluminum increase in drinking water versus age of cement mortar for the Woodward Avenue example

166 132 Impacts of Lining Materials on Water Quality TDS Concentration Increase, mg/l Age of Cement Mortar, days Figure 10.6 TDS increase in drinking water versus age of cement mortar for the Woodward Avenue example Impact from Epoxy Lining In Chapter 7, it was determined that in epoxy lining TOC and chlorine demand are the major impacted constituents. For comparison, the Impact Module calculated the concentrations of TOC and chlorine demand from epoxy material that would occur if the water main on Woodward Avenue had been lined with epoxy. None of the concentrations of TOC or chlorine demand exceeded the target concentrations listed in Table The calculated concentrations of TOC and disinfectant demand for epoxy lining, if that had been used, were below their target concentrations of 5 mg/l and 0.5 mg/l, respectively, at all times. This indicates that no supplemental flushing is required. The calculated concentrations of TOC and chlorine demand for epoxy material are shown in Figures 10.7 and 10.8, respectively. No additional flushing would be required to maintain water quality goals if the epoxy lining material had been used. Impact from Polyurethane Lining In Chapter 6, it was determined that the major impacted constituents are ph, chlorine demand, and HAA5. In this study it has been shown that ph value drops by 2 units consistently throughout the 30-day test period. The Impact Module does not have the capability to predict reduction of ph values. The Impact Module calculated the concentrations of chlorine demand and the concentrations of regulated HAA5 due to the use of polyurethane. The chlorine demand and HAA5 concentrations for polyurethane were both below their target concentrations of 0.5 mg/l and 60 µg/l, respectively, at all times (Figures 10.9 and 10.10), therefore, no supplemental flushing was required.

167 Chapter10: Test Case Analysis TOC Concentration Increase, mg/l Age of Epoxy, days Figure 10.7 TOC increase in drinking water versus age of epoxy for the Woodward Avenue example Disinfectant Demand Concentration Increase, mg/l Age of Epoxy, days Figure 10.8 Disinfectant demand in drinking water versus age of epoxy for the Woodward Avenue example

168 134 Impacts of Lining Materials on Water Quality 0.30 Disinfectant Demand Concentration Increase, mg/l Age of Polyurethane, days Figure 10.9 Disinfectant demand in drinking water versus age of polyurethane for the Woodward Avenue example Figure HAA5 increase in drinking water versus age of polyurethane for the Woodward Avenue example

169 Chapter10: Test Case Analysis 135 Cost Impacts The cost to mitigate water quality impacts from cement-mortar lining was the cost of flushing an additional 3,700 gallons of water to maintain the calcium concentration on the first day after the main was put back into service. The Cost Module calculated the cost, which includes equipment, vehicles, labor, and the cost of water, as $247. However, to mitigate ph to keep its value less than 8.5 s.u. in cement-mortar lining will require 90 days of flushing, which will result in a cost of $3,554. There were no additional costs for epoxy or polyurethane because the Impact Module calculated that the concentrations of all high impact constituents (identified in Chapters 7 and 8) were below the target concentrations established for these constituents. SENSITIVITY ANALYSIS A sensitivity analysis was performed on the target ph value used in the Impact Module. Target ph values from 8.4 to 9.0 in increments of 0.1 s.u. were used as input to the Impact Module. The calculated volume of water needed for flushing the water main was quite sensitive to target ph value. A target ph value of 8.4 required nearly 150,000 gallons of water to maintain the ph below that level while a target ph value of 8.9 did not require any additional water for flushing to keep the ph below that level. Table 10.2 summarizes the results of the sensitivity analysis for ph. Table 10.2 Sensitivity of target ph and flushing volume Target maximum Volume required Duration of impact ph ,800 gallons 100 days ,000 gallons 90 days ,700 gallons 60 days ,400 gallons 60 days 8.8 7,100 gallons 60 days gallons 0 days gallons 0 days It is apparent that the selection of the target ph can significantly affect the volumes of water required to control the ph impacts from cement-mortar lining as well as the duration of the impacts.

170 136 Impacts of Lining Materials on Water Quality CONCLUSIONS Two significant impacts were predicted by the Impact Module. These were an increased calcium concentration and an increase in ph. Both of these impacts were from cement mortar. The 50-mg/L target concentration for calcium was only exceeded for one day at 95 mg/l. The hydraulic retention time of 0.32 day provided sufficient flow to keep the concentrations, and water quality impact, low. For epoxy and polyurethane materials, none of the calculated concentrations exceeded their target values. However, for polyurethane there will be a reduction of ph, which should be considered in the selection of the lining. In this example, polyurethane and epoxy did not have any water quality impacts that required flushing to mitigate. Cement mortar was associated with elevated calcium concentrations and ph levels. The impact from ph was the most significant and persistent, requiring 94,000 gallons of additional water over 90 days to maintain the ph below the target value of 8.5. By comparison, the calcium concentration exceeded the target value of 50 mg/l for only 1 day and required only 3,700 gallons of additional water to mitigate. The additional cost to mitigate the increased calcium from cement mortar was $247. The cost to mitigate the ph impact from cement mortar was $3,534. There were no additional flushing costs for epoxy or polyurethane. In this example case study, there are no water quality or cost factors that would favor selection of one material over another. The selection of a lining material should be based on other factors. These factors may include: Cost of the project bid with alternative lining materials Availability of contractors Experience with a particular lining material ph is a complex phenomenon. The methodology for estimating ph impacts from cement mortar and the example calculations presented should be used with caution and modified with the user s own data. ph values should be monitored after cement-lining projects and plotted as shown in Figure The user should calculate the HRT in the pipe being evaluated. The HRT in the case study was estimated from the volume of the rehabilitated pipe on Woodward Avenue and the number of customers and their average quarterly water usage as reported by SCCRWA. This method likely over-estimates the HRT because it does not include any additional water passing through the pipe to downstream users. A more precise estimate of HRT can be obtained from computerized hydraulic models of the distribution system. The required flushing volumes and durations will vary with the acceptable target ph value selected by the water utility. This target value may depend on the utility s lead and copper rule compliance requirements and on the method of corrosion control practiced. If the impact and cost procedures had been applied to a lining project in an area with water restrictions or rationing, the cost of water used for flushing would have a greater cost and social impact and could affect the cost comparison of lining material alternatives. Water restrictions would also reduce the volumes of water flowing through the distribution mains, increase the HRT, and result in higher concentrations of constituents in the water.

171 CHAPTER 11 RESEARCH GAPS This study identified the following research gaps that should be studied in the future. CEMENT-MORTAR LININGS The study identified the following research gaps for cement-mortar lining: Study should be conducted using various types of cement mortars to identify cement mortars that will reduce lime leaching. Study should be conducted to estimate residual strengths of cement-mortar linings with time in contact with aggressive waters. Study should be conducted to determine the relationship of residual strengths of cement mortars with calcium loss. Long-term water quality and residual strengths should be studied. Impact of cure time on leaching from cement-mortar lining should be studied. Study should be conducted to compare leaching characteristics and mortar strengths of rapid-curing and regular cement. Impact of additives in cement on water quality should be studied. Long-time bench-scale leaching tests, along with corresponding field tests, would be useful. Study should be conducted to find techniques for maintaining lower ph after a cement-mortar lining has been installed. The impact of various doses of corrosion inhibitors and their interaction with lime leaching from fresh cement mortar should be investigated, with special emphasis on implication for lead and copper rule requirements. POLYURETHANE LINING The study identified the following research gaps for polyurethane lining: The ph drop was observed within 24 hours in leaching tests with polyurethane and persisted for 30 days. The source of the ph drop is not known. Changes in concentrations of metals, nonmetals, or semivolatile organic compounds (SVOCs) cannot account for the 2 ph unit decrease. Analytes that were not measured in this study, such as carbon dioxide, strong acid anions, or volatile organic acids should be investigated. Although the setting time for polyurethane is in minutes, to be on the safe side a 48- hour setting time was used in this study. It would be prudent to know the impacts of setting time on water quality so that water utilities can decide on the minimum setting time they can use without sacrificing water quality. Conduct a comparison of water quality impacts of a factory polyurethane-lined pipe with an in situ polyurethane-lined pipe. 137

172 138 Impacts of Lining Materials on Water Quality EPOXY LINING GENERAL The study identified the following research gaps for epoxy lining: Although the setting time for epoxy is only a few hours, to be on the safe side a 48- hour setting time was used in this study. It would be prudent to know the impacts of setting time on water quality, particularly for TOC leaching, so that water utilities can decide on the minimum setting time they can use without sacrificing water quality. Bis-phenol A (BPA) is a component in the epoxy resin and was detected in the water for a short duration. This impact should be studied further. A comparison should be made of water quality impacts of a factory epoxy-lined pipe with an in situ epoxy-lined pipe. Reasons for increased chlorine decay and increased THM formation should be studied. General research gaps that should be studied are the following: A comparison of leaching contaminants from various lined pipes, along with polyvinyl chloride (PVC), high density polyethylene (HDPE), and other unlined pipes, should be conducted. Institutional drivers in Europe and in North America that create preferences for use of different lining materials should be studied.

173 CHAPTER 12 FINDINGS AND RECOMMENDATIONS Cleaning and lining is one of the most economical options for rehabilitation of old water mains with good structural condition. One of the important purposes of pipe lining is to prevent leaching of metals from the wall of water mains and thus, to improve water quality. However, there is concern about leaching of metals and organic chemicals from these lining materials. The effectiveness and the water quality impacts of these linings depend on the properties of the lining material itself, installation procedures, water characteristics, exposure time, and surface area to volume ratio. When comparing the cost of alternative lining materials available for water main rehabilitation, it is necessary to consider the total cost of the lining material, including the water quality impact cost. In North America, most of the water main rehabilitation is conducted using cementmortar lining. As an alternative to cement-mortar lining, epoxy and polyurethane linings have been found to have high durability and higher hydraulic capacity. This project developed water quality impacts of cement-mortar, epoxy, and polyurethane liners under various background water quality conditions. The key findings and recommendations of this project are provided in the following subsections. FINDINGS The key findings of this study are as follows: Questionnaire survey: A survey of North American and European water utilities conducted for this study indicates that North American water utilities have experience primarily of using cement-mortar lining, whereas in Europe most of the water utilities use epoxy and polyurethane linings. Red water and reduction of hydraulic capacity are the main concerns that appear to be the primary reasons for lining water mains. The utilities that responded, in general, noticed significant improvements in hydraulic capacity and in elimination of red water after lining. The most frequently reported water quality impact from cement-mortar lining was an increase in the ph of water. After time and increased water demand, the ph impact subsided. Use of proper quality assurance/quality control (QA/QC) procedures for the construction of linings is common, and significantly improved the performance of linings. Bench-scale testing was undertaken that simulated the water volume to surface area ratio of a 4-in. pipe. Coupons of the lining materials were used in these tests. The bench-scale testing used a low alkalinity/low hardness water with either no disinfectant, free chlorine, or chloramines. Water changes occurred on days 1, 2, 4, 9, 11, 14, 15, and 19; the tests continued up to 30 days with occasional water changes from day 19 to day 30. Cement-mortar lining: Under the conditions used for the bench-scale testing of cement-mortar coupons, impacts to water quality from cement mortar were most severe up to 9 days. After 139

174 140 Impacts of Lining Materials on Water Quality day 9, a significant decrease in most water quality parameter release rates was observed. The ph of water increased significantly to a value of 12.5 with 24 hours of contact with cement mortar and maintained values of ph 10.5 to 12.5 throughout the 30-day test period. Likewise, the alkalinity was increased from 35 to 600 mg/l (as CaCO 3 ) within the first 24 hours of contact time with cement mortar. After 9 days of contact, the alkalinity declined to about 100 mg/l as (as CaCO 3 ). The total solids content of the water increased to up to 1,700 mg/l in the presence of cement-mortar coupons. High ph (up to 12.5 ph units), alkalinity (up to 600 mg/l), calcium concentrations (up to 260 mg/l), and total solids (up to 1,700 mg/l) are driven by dissolution of Ca(OH) 2, which is a byproduct of the curing of cement. This may result in a temporary elevation of turbidity in the water in the distribution system. Cement mortar significantly increased the calcium, aluminum, and chromium concentrations in the water. The aluminum concentrations exceeded the EPA SMCL, while the chromium levels remained below the EPA MCL. After day 9, release of aluminum and chromium to the water decreased. Cement mortar created a substantial chlorine demand, but the demand for chloramine was much less and ceased after a few days of contact. All water types containing cement-mortar coupons had an intense cement odor, which remained moderately high for first 14 days of testing. The presence of cement mortar had no impact on the ammonia concentration in water samples. TOC release was generally < 0.5 mg/l. In chlorinated water, this TOC reacted to produce < 10 µg/l HAA5; HAA5 formation decreased rapidly after 2 days. Potential formation of THMs exists when new cement mortar that is leaching TOC is contacted with free chlorine. Cement-mortar lining with corrosion prevention additives (CPAs) in water: An evaluation of the effects of CPA types on water quality was conducted. The findings are: Polyphosphate (PP) reduced ph increases more substantially than orthophosphate (OP) or zinc orthophosphate (OPZn) after day 9. There are no significant differences in increase of alkalinity, hardness, and calcium concentrations among various CPAs. Magnesium concentration is highest and aluminum is lowest for PP additive. PP allowed for higher P content after day 11 due to lower ph and increased P solubility. Polyurethane: In the presence of polyurethane, the ph was reduced from ph 8 to about ph 6. The ph drop was observed within 24 hours and persisted for 30 days. Free chlorine was consumed in the presence of polyurethane and its consumption was greater than that for monochloramine. The rate of chlorine decay was greater during days 1 through 4 than in later exposure times.

175 Chapter12: Findings and Recommendations 141 TOC was leached from polyurethane, with a greater amount leached in the presence of chlorine than in its absence. Leached TOC reacted with free chlorine to form up to 30 µg/l HAA5 but no THMs were detected. Weak to moderate odor intensities were released from the polyurethane and persisted for the 30 days of this study. New epoxy: Epoxy exposed to each of the three disinfectant water types produced significant concentrations of TOC (3.5 to 6.3 mg/l) during the first 24 hours of exposure to water. By the second 24-hour exposure period the TOC decreased substantially. By the end of the 30 days, TOC decreased to concentrations between 0.5 and 1.7 mg/l. The epoxy reacted readily with both chlorine and chloramines during the first 24 hours of exposure. After the initial 24-hours exposure, free chlorine was consumed at a much greater rate than chloramines. The disinfectant consumption rate decreased over the 30 days. Disinfectant byproducts (DBPs) were present in most samples, with the highest concentrations detected in chlorinated water. THMs were present only in the chlorinated water exposed to the epoxy and at relatively low levels (< 12 µg/l). HAA5 (<8 µg/l) were present in all three water types and the highest concentrations were found in chlorinated waters. Bis-phenol A (BPA) was detected in concentrations of 22 to 33 µg/l during the first 24 hours in all three waters exposed to epoxy. Concentrations decreased substantially by the second test day, but trace amounts of BPA were still detected by the last test day. Weak to moderate odor intensities were released from epoxy, which persisted for 30 days. The odor was strongest in chlorinated waters. Ammonia was present in high concentrations in chloraminated water exposed to epoxy only during the first 24 hours. Mature epoxy: ph in the mature-epoxy-lined pipe sample was found to decrease slightly. Free chlorine exposed to the mature pipe was almost completely consumed within each test period. In contrast to the new epoxy, the pipe did not show a decrease in disinfectant demand over the 30 days. TOC leached from the mature pipe sample ranged in concentration from 3.5 to 1.6 mg/l. DBPs were formed in chlorinated water exposed to the pipe. Only HAA5 were detected, and, when normalized for surface area, these were formed at the same concentration as in the chlorinated samples exposed to new epoxy.

176 142 Impacts of Lining Materials on Water Quality RECOMMENDATIONS The following general recommendations are made as a result of this study: Cement-mortar lining has a severe initial water quality impact of raising ph, alkalinity, and calcium. For very low alkalinity water, rise of ph may continue for a long period of time. High water flows reduce this impact significantly. Therefore, cement-mortar lining, in general, should be avoided in low water circulating areas such as dead ends. In such areas, water should be flushed regularly. In other areas, water should be flushed initially for a required period. The frequency and duration of flushing can be determined by using the methodology developed in this study. Epoxy lining and polyurethane lining for water mains have been found to be a good alternative to cement-mortar lining and all water utilities should consider this technology as an alternative to cement-mortar lining. Water quality impacts for epoxy and polyurethane linings should be analyzed using the methodology developed in this study before making the final selection. In this study a software has been developed to analyze water quality impacts of the alternative lining materials of cement mortar, epoxy, and polyurethane; identify the mitigation required; and estimate the mitigation cost to select the best lining material for a particular site. Before selecting the lining material for a lining project water utilities should consider using the software developed in this study for a detailed analysis of water quality impacts, their mitigation, and the cost effectiveness of alternative linings before making a final decision on a lining material.

177 1. Contact Information APPENDIX A WATER UTILITY QUESTIONNAIRE IMPACTS OF LINING MATERIALS ON WATER QUALITY Name of Organization: Address: Contact Name/Title: Phone: Fax: 2. Historical use of lining materials. Provide list of lining projects on Table 1 (attached). 3. Why did your utility line the pipes? Tuberculation Hydraulic capacity (C-factor) Red water Other (specify) 4a. Did you collect water quality samples before the liner was placed? yes no 4b. If yes, list tests and results on Table 2a (attached). 5a. Did you collect water quality samples after the liner was placed? yes no 5b. If yes, list tests and results on Table 2b (attached). 143

178 144 Impacts of Lining Materials on Water Quality 6a. Have you experienced water quality changes, good or bad, due to the liner material? yes no 6b. If yes, describe the water quality changes. 6c. Do the water quality changes from the liner diminish or increase with the age of the liner? yes no 6d. If yes, describe (list by project using project numbers on Table 1). 7. Are reports of the lining installation project available? yes no If yes, please provide a copy. 8. What QC procedures were used to install or apply the liner? Preparation/cleaning Mix ratio Temperature Moisture Other (explain) 9. Are written QC procedures for installation or application of the liner available? yes no If yes, please provide a copy.

179 Appendix A: Water Utility Questionnaire What QC procedures were used to evaluate the quality of the installed liner? 11. Are written QC procedures for post-installation evaluation available? yes no If yes, please provide a copy. 12a. Do you have an expectation of the useful life of the lining? yes no 12b. If so, what is your expectation for the useful life of the epoxy lining? (years) 13. What is the time duration between lining and putting the pipe back into service? (hours).

180 Project number Year of lining (year) Lining material Project name/location Table A.1a Inventory of lining projects Part 1 Length (feet) Pipe diameter (inches) Pipe material (attach separate pages as necessary to describe lining projects) Year pipe installed (year) Lining material manufacturer 146 Impacts of Lining Materials on Water Quality

181 Table A.1b Inventory of lining projects Part 2 Project number Contractor name Cleaning method used Problems encountered during installation (description) (attach separate pages as necessary to describe lining projects) Target liner thickness (mils) Time to full cure (hr:min) Appendix A: Water Utility Questionnaire 147

182 148 Impacts of Lining Materials on Water Quality Project number (from Table 1) Table A.2a Water quality measurements made before lining Test (analyte/compound) Result Units Test standard number/id Table A.2b Water quality measurements made after lining Project number (from Table 1) Test (analyte/compound) Result Units Test standard number/id Time after installation of liner (years/months/days)

183 APPENDIX B METHODS OF CHEMICAL MEASUREMENTS The following methodologies are used in chemical analysis of parameters: ph and Temperature Temperature and ph were measured using Standard Method 4500-H + using a Corning 315 ph/ion probe that was calibrated using standard ph solutions 4, 7, and 11. Total Alkalinity Alkalinity was measured using Standard Method H 2 SO 4 (CAS # ) with concentration N was titrated into 100-mL samples to ph 4.5. Disinfectant Residual Disinfectant residual was measured using Standard Method 2350 with a HACH kit (HACH, free chlorine CAT # , total chlorine CAT # ) measuring free and total chlorine. Free chlorine was used to measure the chlorinated sample water. Monochloramine was calculated using free and total chlorine (total free = chloramine). Nitrogen-Ammonia (N-NH 3 ) Nitrogen-ammonia (N-NH 3 ) was measured using Standard Method 4500 with a HACH kit using salicylate (HACH, CAT # ) and cyanurate reagents (HACH, CAT # ). Solids Total solids (TS) was measured using Standard Method 2540B. Selected samples were filtered through a 0.5-μm filter to measure dissolved solids and particulate solids. Elemental Analysis for Metals, Nonmetals, and Hardness Selected trace metals (sodium [Na], magnesium [Mg], aluminum [Al], potassium [K], vanadium [V], chromium [Cr], iron [Fe], manganese [Mn], cobalt [Co], nickel [Ni], copper [Cu], zinc [Zn], arsenic [As], molybdenum [Mo], cadmium [Cd], tin [Sn], barium [Ba], and lead [Pb]), and nonmetals (sulfur [S], phosphorous [P], chlorine [Cl], and silicon [Si], a semi-metal) were measured. Standard Method 3125 was used and the samples were analyzed on a ThermoElectron Corporation inductively coupled plasma mass spectrometer (ICP-MS) X-Series. Hardness, as the sum of calcium and magnesium concentrations, was measured using the elemental analysis results. Total Organic Carbon (TOC) TOC was measured using Standard Method 5310C using a Shimadzu TOC-V TOC analyzer. The samples were acidified to < ph 2 using nitric acid (CAS # ). Two measurements were taken from each replicate. If the samples were not within 20%, a third sample was taken and an average was calculated to give the TOC concentration for that sample. 149

184 150 Impacts of Lining Materials on Water Quality Trihalomethanes (THMs) THMs were analyzed using Standard Method 6232D using a Tre Metrics 9001 gas chromatograph, Tracor 1000 Hall detector, Tekmar 2016 purge and trap autosampler, and Tekmar 3000 purge and trap concentrator. The column used was a DB-624 with helium as the carrier gas with the initial oven temperature at 45 ºC and held for 3 minutes, increasing at 11 ºC/min. with a maximum oven temperature of 200 ºC. Standard curve was performed for each run unless run sequentially (THM Ultra 5000 μg/ml standard). Haloacetic Acids (HAA5) The HAA5 extractions were performed following EPA Method using liquid-liquid extraction, derivitization, and GC with electron capture detection (ECD). The standard (Methylated Haloacetic Acid Standard, Chem Service, West Chester, Pa.) contained concentrations of HAA5 ranging from 20 μg/l to 60 μg/l. The GC column used was a DB-1701 (30 m length, 0.25 diam., 0.25 μm film) with helium as the carrier gas and nitrogen as the makeup gas. The initial oven temperature was 35 C, which was held for 10 min. and increased at 5.7 C/min. to 75 C and held for 5 min. Temperature was again increased at 5 C/min. to 100 C and held for 5 min., then increased at 20 C/min. to 140 C and held for 5 min. After being removed and equilibrating to room temperature, the samples were identified and quantified using procedural standard calibration. Semivolatile Organic Compounds (SVOCs) SVOCs were measured using Standard Method 6410B, which is a liquid-liquid extraction using methylene chloride and then concentrated using a Kaderna-Danish apparatus at 75 C water bath. The samples were analyzed using a GC/MS using a DB-5 column (30.0 m x 250 µm x 0.30 µm). The initial oven temperature was 40 ºC and was held for 3 min., then increased by 8 ºC/min. to 200 ºC and held for 1 min. The column was then heated by 10 ºC/min. to 300 ºC and held for 11 min. The targeted compounds were identified using standard NIST elution times as well as mass spectra. Nontargeted compounds were tentatively identified based on library matching of mass spectra. For selected compounds, chemical standards were purchased and retention times and mass spectra compared to confirm their qualitative identity and measure quantitative amounts. Odor Analysis Odor analysis was performed using the Flavor Profile Analysis (FPA) using Standard Method 2170 where FPA participants underwent a 1-day training session to learn the FPA method. The study protocol was approved by the Institutional Review Board for Research Involving Human Subjects at Virginia Tech. The FPA intensity scale is 0: odor free (OF), 1: threshold, 2: very weak, 4: weak, 8: moderate, 10: strong, and 12: very strong. The weak intensity corresponds to the sweetness of canned fruit (for comparison, an FPA taste intensity of 8 is moderate and corresponds to canned soda, while an FPA intensity of 12 is strong and corresponds to syrup or jelly).

185 Appendix B: Methods of Chemical Measurements 151 THM Sorption/Formation Experiment An issue for leaching tests, including those described in NSF-61, is that DBPs can form in the water during leaching and then be re-sorbed into the lining material and consequently, not measured. This phenomenon was investigated for selected THMs. The formation of THMs from the reaction of TOC leached from the lining material and added chlorine used for disinfection was investigated by placing coated coupons in reference water with no disinfectant and allowing leaching to occur over a 72- to 96-hour period under headspace-free conditions at room temperature. The TOC was then measured from aliquots of the leached water that were placed in headspace-free VOA vials with the addition of free chlorine. At 0-, 24-, and 72-hour time intervals, N sodium thiosulfate solution was added to quench the THM formation reaction. The samples were then analyzed for disinfectant residual and THMs. To investigate if THMs could sorb into the lining material, coated coupons were immersed headspace-free for up to 72 hours in reference water containing 15 to 100 µg/l of THMs and no free chlorine. Lining material coupons were placed in the test apparatus with the standard water without disinfectant for 72 to 96 hours under headspace-free conditions. The sorption of THMs into the lining was tested by determining the concentration of THMs in the water over time.

186

187 APPENDIX C MASS RATE CURVES In this appendix the mass rate curves developed from the laboratory data are presented. CEMENT MORTAR (CM) Calcium Cement mortar increased the calcium in water. Each mg/l of calcium represents 2.5 mg/l of hardness as CaCO 3. Although there is no numerical standard for hardness, the increased calcium concentrations could be noticeable to customers. The impact of cement mortar on calcium concentration is of short duration. The increased calcium concentration beyond 15 days is not significant. Figure C.1 shows the rate of increase in calcium concentration over the 30-day test period. The normalized calcium loss rate (in mg/cm 2 /day) is shown in Figure C.2. The normalized loss rate is a useful tool that can be used to calculate the impact on water quality for pipes of any diameter. Impact of Source Water Alkalinity on Calcium Loss The impact of source water alkalinity was evaluated by comparing the laboratory results from this study, in which alkalinity was 35 mg/l, with the results of a previous study (Douglas and Merrill 1991), which had a source water alkalinity of 13 mg/l. Normalized loss rates for calcium are shown in Figure C.3 for both data sets. The calcium release rate from the low alkalinity water was initially low (around 0.15 mg/cm 2 /day) and maintained a persistent loss through day 90 of the test. The calcium release rate from the higher alkalinity water was initially around 0.6 mg/cm 2 /day but decreased rapidly to insignificant levels by day 12 of the test. These data suggest that low alkalinity may cause a more extended release of calcium than higher alkalinity water. Aluminum Cement mortar was associated with an increase in aluminum concentration in water. The aluminum concentration increased above the secondary drinking water limit of 200 µg/l. The rate of increase in aluminum concentrations measured over the 30-day test period is shown in Figure C.4. The rate of increase was greatest at the end of day 4, represented by the data plotted at day 1.5 (the center point between day 1 and day 4). The impact of cement mortar on aluminum concentration was minimal after day 10. The normalized aluminum mass loss rate over time is shown in Figure C.5. Total Dissolved Solids (TDS) The impact of cement mortar on TDS is significant but is of short duration. The rate of increase in TDS concentration over time is shown in Figure C.6 The mass loss rate for TDS is shown in Figure C

188 154 Impacts of Lining Materials on Water Quality 300 Calcium Concentration Rate of Increase, mg/l/day Water 1 Water 2 Water 3 Water Time, days Figure C.1 Impact of cement mortar on calcium concentration rate of increase over time Calcium Mass Loss Rate, mg/cm 2 /day y = x r 2 = Series1 Series2 Series3 Series4 Regression (Series 1) Age of cement mortar, days Figure C.2 Calcium mass loss rate versus time

189 Appendix C: Mass Rate Curves Calcium Mass Loss Rate, mg/cm 2 /day y = x r 2 = y = x Series1 Series2 Series3 Series4 #415 data Regression (Series 1) Regression (#415 data) r 2 = Age of cement mortar, days Figure C.3 Calcium mass loss rate versus time 450 Aluminum Concentration Rate of Increase, µg/l/day Water #1 Water #2 Water #3 Water # Time, Days Figure C.4 Aluminum concentration rate over time for cement mortar

190 156 Impacts of Lining Materials on Water Quality 0.8 Aluminum Mass Loss Rate, µg/cm 2 /day y = e x r 2 = Water #1, No Disinfectant Water #2, Chlorine, ph = 6 Water #3, Chlorine, ph = 8 Water #4, Chloramine Regression (Water #3, Chlorine, ph = 8) Age of cement mortar, days Figure C.5 Aluminum mass loss rate µg/cm 2 /day with age TDS Concentration Rate, mg/l/day Series1 Series2 Series3 Series Time, Days Figure C.6 Total dissolved solids (TDS) concentration increase over time for cement mortar

191 Appendix C: Mass Rate Curves TDS Mass Release Rate, mg/cm 2 /day y = x r 2 = Water #1, No Disinfectant Water #2, Chlorine, ph = 6 Water #3, Chlorine, ph = 8 Water #4, Chloramine Regression (Water #1, No Disinfectant) Time, Days Figure C.7 TDS mass release rate from cement mortar with time EPOXY The parameters having the most significance impact on water concentrations with respect to epoxy material were total organic carbon (TOC) and disinfectant demand. The TOC concentration increases with time for water exposed to epoxy material, however, the rate of increase decreases with time and is shown in Figure C.8. The TOC concentration rate in water in contact with fresh epoxy material had a similar pattern for chlorinated water, chloraminated water, and water without disinfectant. The TOC concentration rate of increase for water in contact with mature epoxy is also shown in Figure C.8. The mass release rate of TOC from epoxy material for all test conditions is shown in Figure C.9. The disinfectant demand from fresh and mature epoxies was significant. The rates of disinfectant demand are shown in Figure C.10. Although the disinfectant demand for chlorine was greater than for monochloramine, the demand rates were measurable and significant for both disinfectants. The disinfectant mass loss rate curves are shown in Figure C.11.

192 158 Impacts of Lining Materials on Water Quality 7 TOC Concentration Rate of Increase, mg/l/day No Disinfectant Chlorine Chloramine Mature Pipe with Chlorine Time, Day Figure C.8 TOC concentration rate of change for epoxy material y = x r 2 = (No disinfectant) y = x r 2 = (Chlorine) y = x r 2 = (Monochloramine) y = x r 2 = 0.84 (Mature epoxy and chlorine) TOC Mass Release Rate, mg/cm 2 /day No Disinfectant Chlorine Chloramine Mature Pipe Regression (No Disinfectant) Regression (Chlorine) Regression Chloramine) Regression (Mature Pipe) Time, Days Figure C.9 TOC mass release rate from epoxy with time

193 Appendix C: Mass Rate Curves Disinfectant Demand Concentration Rate, mg/l/day Chlorine Chloramine Mature Pipe with Chlorine Time, Days Figure C.10 Disinfectant demand concentration rate over time for epoxy material Disinfectant Demand Mass Loss Rate, mg/cm 2 /day y = x r 2 = (Chlorine) y = x r 2 = (Monochloramine) y = x r 2 = (Mature epoxy and chlorine) Chlorine Chloramine Mature Epoxy with Chlorine Regression (Chlorine) Regression (Chloramine) Regression (Mature Epoxy with Chlorine) Time, Days Figure C.11 Disinfectant demand mass loss rate for epoxy with time

194 160 Impacts of Lining Materials on Water Quality POLYURETHANE (PU) The water chemistry parameters of significance for PU material were disinfectant demand rate and formation of five regulated haloacetic acids (HAA5). The disinfectant loss rates for chlorinated water are shown in Figure C.12. The mass loss rates of chlorine disinfectants in contact with PU material are shown in Figure C.13. Only the regression curve for chlorine is shown because the mass loss rates for monochloramine were widely scattered and did not show a clear pattern. The rates of formation of HAA5 in waters in contact with PU material are shown in Figure C.14. The mass formation rates of HAA5 formation are shown in Figure C.15. Only chlorinated water showed a clear pattern of HAA5 formation. 2.0 Disinfectant Demand Concentration Rate, mg/l/day Chlorine Time, Days Figure C.12 Disinfectant demand concentration rate over time for water in contact with polyurethane

195 Appendix C: Mass Rate Curves Disinfectant Demand Mass Rate, mg/cm 2 /day y = x r 2 = Chlorine Regression (Chlorine) Time, Days Figure C.13 Disinfectant demand mass rate over time for water in contact with polyurethane HAA5 Concentration Increase Rate, µg/l/day No Disinfectant Chlorine Chloramine Time, Days Figure C.14 HAA5 concentration increase rate over time for water in contact with polyurethane

196 162 Impacts of Lining Materials on Water Quality HAA5 Mass Formation Rate, µg/cm 2 /day y = x r 2 = No Disinfectant Chlorine Chloramine Power (Chlorine) Time, Days Figure C.15 HAA5 mass formation rate for water in contact with polyurethane

197 APPENDIX D USER S GUIDE FOR THE IMPACT MODULE INTRODUCTION The Impact Module calculates the concentration of seven constituents in water from three common materials used to line water mains. The Impact Module is implemented in a spreadsheet computing environment. This User s Guide explains the data requirements, calculation procedures, output, and interpretation of the results. The spreadsheet contains 13 worksheets. These include a data input sheet, a database of mass loss functions, an output worksheet, and 10 charts displaying the calculated concentrations of various constituents over time. DATA REQUIREMENTS The data required to run the model are: The diameter, in inches, of the water main that will be cleaned and lined. The length, in feet, of the water main to be cleaned and lined. The daily water usage or flow, in gallons per day, through the water main. The disinfectant used in the water flowing through the main. The maximum increase established by the water utility for constituents of interest. INPUT SCREEN The data input screen is shown in Figure D.1. The user types the diameter, length, and daily water use in cells B6, B7, and B8, respectively. Cell B9 contains a drop-down menu allowing the user to select chlorine, chloramine, or none as the disinfectant. Cells C13 through C21 of the input screen contain target concentrations, which are maximum concentration increases allowed in the water for the purpose of determining whether additional flushing will be required to meet the target concentrations. In the case of ph, the Impact Module calculates the maximum allowable increase in ph as the difference between the background ph (cell B10) and target ph (cell C13). These can be changed by the user. Note that the limits for aluminum and TDS are secondary drinking water MCLs. The limit for HAA5 is a primary drinking water MCL. The target values shown for the other constituents are provided as examples only. The length of time when the concentration of any constituent exceeds the target concentration is automatically displayed by the program. The volume of additional water required to meet the target concentrations is read from the Impact Calculations worksheet, which requires some entries by the user. Those will be described in the Impact Calculations section. DATABASE OF MASS RELEASE RATES A database contains mass release rates for 18 combinations of lining materials, disinfectants, and water quality constituents. The mass release rates were developed from the bench-scale testing performed as part of this study. Figure D.2 shows the database. The coefficients and exponents 163

198 164 Impacts of Lining Materials on Water Quality could be modified by the user if they had data to support the mass release rate curves. The mass release rates are in units of mg/cm 2 /day or µg/cm 2 /day, depending on the constituent. Figure D.1 Data input screen for the Impact Module NUMERICAL OUTPUT The numerical output is displayed in the Impact Calculations worksheet and summarized on the data input screen (columns E and H). The ph calculations are found in columns A, B, and C from the Impact Calculations worksheet and are shown in Figure D.3. The Impact Module calculates the maximum allowable ph increase as the difference between the background ph and the target ph. The required hydraulic residence time (HRT), flushing volume, and duration of the ph exceedance are calculated and displayed in the calculations worksheet. The flushing volume and duration of exceedance are also displayed on the Input worksheet. An example calculation for alkalinity, columns D, E, and F from the Impact Calculations worksheet, is displayed in Figure D.4. The upper section of Figure D.4 shows the calculations performed based on the input provided in the Input worksheet. The total volume of water, interior surface area, and HRT are calculated from the data supplied in the Input worksheet. The program obtains the correct coefficients and exponents describing the mass loss rates based on the material, constituent of interest, and disinfectant, if the loss rate is affected by the disinfectant from the Mass Loss Functions worksheet.

199 Appendix D: User s Guide for the Impact Module 165 Material Disinfectant Constituent of Interest Units Regression Form Coefficient Exponent Coefficient of Determination, r 2 Equation CM Waters 1-4 Chromium ug/cm 2 /d Power y = x CM Water 3 Aluminum ug/cm 2 /d Exponential y = e x CM Waters 2 Alkalinity mg/cm 2 /d Power y = x CM Water 1 Calcium mg/cm 2 /d Power y = x CM Water 2 HAA-5 ug/cm 2 /d Exponential y = e x CM Water 1 TDS mg/cm 2 /d Power y = x CM Water 2 TOC mg/cm 2 /d Exponential y = e x Epoxy Chlorine Disinfectant Loss mg/cm 2 /d Power y = x Epoxy Monochloramine Disinfectant Loss mg/cm 2 /d Power y = x Mature Epoxy Chlorine Disinfectant Loss mg/cm 2 /d Power y = x Epoxy No disinfectant TOC mg/cm 2 /d Power y = x Epoxy Chlorine TOC mg/cm 2 /d Power y = x Epoxy Monochloramine TOC mg/cm 2 /d Power y = x Mature Epoxy Chlorine TOC mg/cm 2 /d Power y = x PU Chlorine Disinfectant Loss mg/cm 2 /d Power y = x PU Chlorine HAA-5 ug/cm 2 /d Power y = 0.028x CM Chlorine Calcium 1 mg/cm 2 /d Power y = x CM Chlorine Alkalinity 2 mg/cm 2 /d Power y = x Daily calcium data from AwwaRF Project #415 2 Daily alkalinity data from AwwaRF Project #415 CM, cement mortar PU, polyurethane Water 1, ph 8, no disinfectant Water 2, ph 6.5, chlorine 2 mg/l Water 3, ph 8, chlorine 2 mg/l Water 4, ph 8, monochloramine 4 mg/l HAA-5, haloacetic acid, 5 regulated compounds TDS, total dissolved solids TOC, total organic carbon Figure D.2 Database of mass release rates

200 166 Impacts of Lining Materials on Water Quality Cement Mortar ph Diameter 8 in. Length 503 ft. end area 0.35 s.f. circumference 2.09 ft. surface area 1,053 s.f. surface area 978,716 cm2 volume 176 c.f. volume 1,313 gallons daily water use 4,159 gallons/day pipe volumes volume exchanges/day velocity fps hydraulic residence time 0.32 days Maximum ph 8.5 s.u Maximum DpH 1.2 s.u. Time Required HRT Additional Flow Required Days Days gpd Target = 8.5 s.u. Figure D.3 Impact Calculations worksheet ph calculation

201 Appendix D: User s Guide for the Impact Module 167 Cement Mortar Alkalinity Diameter 8 in. Length 503 ft. end area 0.35 s.f. Circumference 2.09 ft. surface area 1,053 s.f. surface area 978,716 cm2 Volume 176 c.f. Volume 1,313 gallons daily water use 4,159 gallons/day pipe volumes volume exchanges/day Velocity fps hydraulic residence time 0.32 days Coefficient Exponent Time Alkalinity release rate Alkalinity Concentration Days mg/cm2/d mg/l Target = 100 mg/l Figure D.4 Impact Calculations worksheet alkalinity calculations

202 168 Impacts of Lining Materials on Water Quality The lower section of Figure D.4 displays the number of days the liner is in contact with water, the alkalinity release rate, and the calculated alkalinity concentration. Note that in this case, all of the calculated alkalinity concentration increases are below the target concentration of 100 mg/l as CaCO 3. The calculated calcium concentrations are shown in Figure D.5. Note that the calcium concentration on day 1 exceeds the target concentration increase of 50 mg/l. In another section of the worksheet, the user can specify an additional volume of water to be flowed through the main in order to reduce the calculated concentration. This procedure will be explained in the Calculating Additional Flushing Volumes section. GRAPHICAL OUTPUT For each constituent and material combination, the Impact Module creates a graph of the calculated concentration increases over time. An example of the graphs created for alkalinity and calcium are displayed in Figures D.6 and D.7, respectively. CALCULATING ADDITIONAL FLUSHING VOLUMES In the example for calcium, the calculated concentration on day 1 was 95 mg/l, which exceeded the target concentration (not a standard) of 50 mg/l. The Impact Calculations worksheet has an area where the user can enter a volume of water, in gallons, which will be added to the specified daily flow volume. This will lower the calculated concentration. The user can use a trial and error method of entering a flow volume, compare the calculated and target concentrations, then enter another flow volume until the target and calculated concentrations are acceptably close. Figure D.8 shows the calcium portion of the Impact Calculations worksheet (columns G, H, and I) with a value of 2,000 gallons entered in cell H37 as an additional water volume. The calcium concentration increase is 64 mg/l, which is still above the target concentration increase of 50 mg/l. The flow value should be increased further until the target concentration is reached. As an alternative, the user can use the Goal Seek function of Excel to calculate the flow volume required to reach the target concentration. The Goal Seek function is accessed through the Tools Menu as shown in Figure D.9. After the the Goal Seek window appears, specify the cell where the target value is needed (F21 contains the calculated calcium concentration for day 1), specify the target value (50), and the cell where the independent variable is entered (H37). These are shown in Figure D.10. The Goal Seek function performs the calculations, then displays the results as shown in Figure D.11. After selecting OK, the cell H37 of the worksheet contains a water volume of 3,734 gallons and the calculated calcium concentration is 50 mg/l as shown in Figure D.12. A flow value of 3,734 gallons has been inserted in cell H37 and the calcium increase on day 1 is now 50 mg/l. INTERPRETATION OF RESULTS In this example, ph and calcium exceeded the target concentration increases. The exceedance for ph lasted 90 days and required 94,000 gallons of water to keep the ph increases less than the target value of 1.2 s.u., which, in this example, is equivalent to ph 8.5. The exceedance for calcium lasted for 1 day.

203 Appendix D: User s Guide for the Impact Module 169 Cement Mortar Calcium Diameter 8 In. Length 503 ft. end area 0.35 s.f. circumference 2.09 ft. surface area 1,053 s.f. surface area 978,716 cm2 volume 176 c.f. volume 1,313 gallons daily water use 4,159 gallons/day pipe volumes volume exchanges/day velocity fps hydraulic residence time 0.32 days Coefficient Exponent Time Calcium release rate Calcium Concentration Days mg/cm2/d mg/l Target = 50 mg/l Figure D.5 Impact Calculations worksheet calcium calculations

204 170 Impacts of Lining Materials on Water Quality Figure D.6 Calculated alkalinity concentration increase due to cement mortar Figure D.7 Calculated calcium concentration increase due to cement mortar

205 Appendix D: User s Guide for the Impact Module 171 Cement Mortar Calcium Diameter 8 in. Length 503 ft. end area 0.35 s.f. circumference 2.09 ft. surface area 1,053 s.f. surface area 978,716 cm2 volume 176 c.f. volume 1,313 gallons daily water use 4,159 gallons/day pipe volumes volume exchanges/day velocity fps hydraulic residence time 0.32 days Coefficient Exponent Time Calcium release rate Calcium Concentration Days mg/cm2/d mg/l Target = 50 mg/l Day 1 Flow 2,000 gallons Figure D.8 Additional volume of water reduced calcium concentration

206 172 Impacts of Lining Materials on Water Quality Figure D.9 Select Goal Seek under the Tools Menu Figure D.10 Goal Seek data entry window Figure D.11 Goal Seek status window

207 Appendix D: User s Guide for the Impact Module 173 Figure D.12 Result after Goal Seek The calculated flushing volume is quite sensitive to the target ph value selected. A target ph of 8.4 requires 150,000 gallons of water, while a target ph of 8.9 requires no additional water to achieve this value (see Table 10.2). COST COMPARISON The installation costs and life expectancies for cement mortar, epoxy, and polyurethane liners are comparable. One factor that can distinguish the cost of these alternative flushing materials is the cost of additional flushing needed to mitigate the water quality impacts from newly placed liners. The Impact Module calculates the additional flushing volumes needed to meet user-specified concentrations. A cost module to calculate the additional costs of flushing to mitigate water quality impacts from various liners has been developed and is included on the attached CD. The Cost Module input screen is shown in Figure D.13. The Cost Module requires as input the volume and unit cost of water, personnel costs, and equipment costs. If available, unit costs for flushing, based on the company s experience, may be used to calculate the cost of water main flushing.