Effect of Nitrification
on Corrosion in the
Distribution System
Subject Area: Water Quality
Effect of Nitrification
on Corrosion in the
Distribution System
©2010 Water Research Foundation. ALL RIGHTS RESERVED
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
www.WaterResearchFoundation.org.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Effect of Nitrification
on Corrosion in the
Distribution System
Prepared by:
Yan Zhang, Marc Edwards, Ameet Pinto, and Nancy Love
Virginia Tech
418 Durham Hall, Blacksburg, VA 24061
Anne Camper and Mohammad Rahman
Montana State University, Civil Engineering Department
205 Cobleigh Hall, Bozeman, MT 59717
and
Helene Baribeau
Carollo Engineers
199 Los Robles Avenue, Suite 530, Pasadena, California 91101
Jointly sponsored by:
Water Research Foundation
6666 West Quincy Avenue, Denver, CO 80235-3098
and
U.S. Environmental Protection Agency
Washington, D.C.
Published by:
©2010 Water Research Foundation. ALL RIGHTS RESERVED
DISCLAIMER
This study was funded by the Water Research Foundation (Foundation) and the U.S. Environmental
Protection Agency (USEPA) under Cooperative Agreement No. X-83294801. The Foundation and
USEPA 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 or
USEPA. This report is presented solely for informational purposes.
Copyright © 2010
by Water Research Foundation
ALL RIGHTS RESERVED.
No part of this publication may be copied, reproduced
or otherwise utilized without permission.
ISBN 978-1-60573-047-9
Printed in the U.S.A.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
CONTENTS
LIST OF TABLES ......................................................................................................................... ix LIST OF FIGURES ....................................................................................................................... xi FOREWORD .............................................................................................................................. xvii ACKNOWLEDGMENTS ........................................................................................................... xix EXECUTIVE SUMMARY ......................................................................................................... xxi CHAPTER 1: INTRODUCTION .................................................................................................. 1 Background ......................................................................................................................... 1 Overview of Nitrification Effects on Corrosion ................................................................. 1
Factors Affecting Nitrification Occurrence: Interaction of Materials and Water
Chemistry ...................................................................................................................... 2
Nutrient Effects on Nitrification ............................................................................. 2
Effect of Pipe Materials and Corrosion Control Strategies on Nitrification
Occurrence ........................................................................................................ 2
Special Focus on the Interaction of Phosphorus, pH and Copper Pipe .................. 3
Granular Activated Carbon (GAC) Filters in Homes ............................................. 4
Interaction of Nitrifiers and Heterotrophic Bacteria and Heterotrophic
Nitrification in Drinking Water Systems .......................................................... 4
Research Objectives ............................................................................................................ 4
Figures and Tables .............................................................................................................. 5
CHAPTER 2: EFFECT OF AMBIENT NUTRIENTS AND METALS ON NITRIFICATION
IN PREMISE PLUMBING....................................................................................................... 9
Introduction ..........................................................................................................................9
Materials and Methods .......................................................................................................11
Experimental Setup ................................................................................................11
Analytical Methods ................................................................................................12
Results and Discussion ......................................................................................................13
Effect of Nutrient Level on Nitrifier Growth and Activity ....................................13
Direct and Indirect Effects of Alkalinity ...............................................................15
Nutrient Release and Ammonia Cycling from Different Pipe Materials ...............16
Effect of Organic Carbon .......................................................................................17
Conclusions ........................................................................................................................17
Figures and Tables .............................................................................................................18
CHAPTER 3: EFFECT OF ORGANIC AND AMMONIA LEVELS ON NITRIFICATION
AND COPPER RELEASE ......................................................................................................23
Introduction ........................................................................................................................23
Nitrification and its Effects on pH and Alkalinity ................................................ 23
v
©2010 Water Research Foundation. ALL RIGHTS RESERVED
vi | Effect of Nitrification on Corrosion in the Distribution System
Materials and Methods .......................................................................................................24
Reactor Setup .........................................................................................................24
Water Preparation ..................................................................................................25
Experimental Approach .........................................................................................25
Analytical Methods ................................................................................................26
Results and Discussion ......................................................................................................28
Ammonia Utilization in PVC and Copper Reactors ..............................................28
Effect of Ammonia and TOC Levels on Nitrification and Copper ........................28
Microbial Community Composition ......................................................................29
Conclusions ........................................................................................................................29
Figures and Tables .............................................................................................................30
CHAPTER 4: NITRIFICATION IN PREMISE PLUMBING: ROLE OF PHOSPHATE,
PH AND PIPE CORROSION .................................................................................................37
CHAPTER 5: LEAD CONTAMINATION OF POTABLE WATER DUE TO
NITRIFICATION ....................................................................................................................39
CHAPTER 6: ACCELERATED CHLORAMINE DECAY AND MICROBIAL GROWTH
RESULTING FROM NITRIFICATION IN PREMISE PLUMBING....................................41
Introduction ........................................................................................................................41
Materials and Methods.......................................................................................................43
Pipe Rig Setup........................................................................................................43
Water Chemistry and Nitrification Inoculation .....................................................43
Analytical Methods ................................................................................................44
Results and Discussions .....................................................................................................45
Disinfectant Decay .................................................................................................45
Heterotrophic Bacterial Growth .............................................................................46
Field Data ...............................................................................................................47
Conclusions ........................................................................................................................48
Figures and Tables .............................................................................................................48
Supporting Information......................................................................................................54
CHAPTER 7: NITRIFICATION EFFECT ON CORROSION OF GALVANIZED IRON,
COPPER AND CONCRETE...................................................................................................55
Introduction ........................................................................................................................55
Materials and Methods.......................................................................................................56
Large Scale Pipe Rig Testing .................................................................................56
Mechanism Investigation of Galvanized Iron Corrosion .......................................56
Nitrification Effect on Concrete Corrosion Test ....................................................57
Analytical Methods ................................................................................................57
Results and Discussions .....................................................................................................58
Nitrification Activity in Different Pipe Materials (Phase II) .................................58
Nitrification Effect on Water Chemistry................................................................58
Effect of Nitrification on Pipe Corrosion...............................................................59
Nitrification Effect on Concrete Corrosion............................................................61
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Contents | vii
Conclusions ........................................................................................................................61
Figures and Tables .............................................................................................................62
CHAPTER 8: EFFECT OF NITRIFICATION AND GAC FILTRATION ON COPPER
AND LEAD LEACHING IN HOME PLUMBING SYSTEMS .............................................69
Introduction ........................................................................................................................69
Effect of Disinfectant .............................................................................................69
Effect of Nitrification.............................................................................................70
Effect of Granular Activated Carbon (GAC) Filters..............................................71
Materials and Methods.......................................................................................................71
Pipe Rig Setup........................................................................................................71
Water Chemistry ....................................................................................................72
Experimental Monitoring and Analysis .................................................................73
Results and Discussions .....................................................................................................73
Disinfectant Effect in Blacksburg Water Without GAC-Phase I...........................73
Overall Impacts of GAC Treatment on Water Chemistry-Phase II and
Phase III ...........................................................................................................75
Nitrification Effects-Phase II .................................................................................77
Field Studies...........................................................................................................79
Summary and Conclusions ................................................................................................79
Figures and Tables .............................................................................................................80
CHAPTER 9: UTILITY INTERVIEW AND CASE STUDIES ..................................................87
Methodology ......................................................................................................................87
Utility Interviews ...................................................................................................87
Review of Existing Water Quality Data ................................................................87
Distribution System Sampling ...............................................................................87
Results and Discussion ......................................................................................................88
Utility Interviews ...................................................................................................88
Review of Existing Water Quality Data ................................................................90
Distribution System Sampling ...............................................................................92
Conclusions ........................................................................................................................96
Utility Interviews ...................................................................................................96
Review of Existing Water Quality Data ................................................................96
Distribution System Sampling ...............................................................................97
Figures and Tables .............................................................................................................97
APPENDIX A: NITRIFICATION IN DRINKING WATER SYSTEMS .................................115
APPENDIX B: NITRIFICATION IN PREMISE PLUMBING: ROLE OF PHOSPHATE
PH, AND PIPE CORROSION ..............................................................................................115
APPENDIX C: CULTURE INDEPENDENT QUANTIFICATION OF AMMONIA
OXIDIZING BACTERIAL IN DRINKING WATER SYSTEMS .......................................151
©2010 Water Research Foundation. ALL RIGHTS RESERVED
viii | Effect of Nitrification on Corrosion in the Distribution System
APPENDIX D: LEAD CONTAMINATION OF POTABLE WATER DUE TO
NITRIFICATION ..................................................................................................................117
APPENDIX E: UTILITY INTERVIEW QUESTIONNAIRE ...................................................191
REFERENCES ............................................................................................................................205
ABBREVIATIONS .....................................................................................................................223
©2010 Water Research Foundation. ALL RIGHTS RESERVED
TABLES
1.1
Key effects of nitrification on water quality and corrosion .................................................7
1.2
Factors affecting nitrification in drinking water systems ....................................................8
1.3
Effect of pipe materials and GAC on nitrification occurrence ............................................8
2.1
Nutrient levels investigated................................................................................................21
2.2
Illustrative prediction of free Cu2+ at two different alkalinity levels for a range of pHs ...21
2.3
Change in nutrient level by leaching from conrete materials to water ..............................22
2.4
Predicted and observed effects of different pipe materials on nitrification .......................22
3.1
Influent water quality for different reactors .......................................................................35
6.1
Field study results ..............................................................................................................53
7.1
Nitrification in different pipe materials .............................................................................66
7.2
Weight loss for different pipe materials.............................................................................66
7.3
Predicted and confirmed effect of nitrification on corrosion .............................................67
8.1
Water quality comparison before and after filtration through GAC ..................................85
8.2
Average ammonia consumption % after 24 hours stagnation in the indicated
pipe material...........................................................................................................86
8.3
Final pH after 24 hours stagnation in the indicated pipe material .....................................86
8.4
Field studies at three utilities .............................................................................................86
9.1
Utilities that participated in the distribution system samplings .......................................112
9.2
Nutrient levels at different sampling locations ................................................................112
9.3
Metal Release at IRWD utility .........................................................................................113
9.4
Metal Release in Hampden Water District ......................................................................113
A.1
Important reactions influencing nitrogen in the distribution system ...............................142
ix
©2010 Water Research Foundation. ALL RIGHTS RESERVED
x | Effect of Nitrification on Corrosion in the Distribution System
A.2
pH-dependent growth range and optimal pH for nitrifying bacteria ...............................143
A.3
Limiting and inhibiting range of various nutrients and their occurrence range in US
raw drinking water ...............................................................................................144
A.4
Key effects of nitrification on water quality and corrosion .............................................146
A.5
Nitrification indicator and rationale for its use ................................................................147
A.6
Situations where nitrification indicators do not always work ..........................................147
A.7
Nitrification control methods: effectiveness and rationale .............................................148
B.1
Metal release in brass pipes .............................................................................................162
B.2
Logarithm of nitrifier MPN at different phosphorus and pH levels in different
pipe materials .......................................................................................................162
C.1
Primers used for P. aeruginosa quantification .................................................................175
C.2
Primers used for N. Europaea amplicon generation and total AOB quantification .........175
C.3
Ammonia oxidizing bacteria concentrations for laboratory scale set-ups .......................175
C.4
AOB concentration for large scale rigs ............................................................................176
C.5
DAPI staining for water samples for five different PVC plumbing systems ...................176
D.1
Water utility studies results ..............................................................................................187
D.2
Montana bench test results ...............................................................................................188
©2010 Water Research Foundation. ALL RIGHTS RESERVED
FIGURES
1.1
General effect of a given nutrient concentration on nitrifier activity ..................................5
1.2
Free Cu concentration in relation to pH and phosphorus levels ..........................................6
2.1
Average ammonia loss % and nitrifier MPN at different copper concentrations ..............18
2.2
Average ammonia loss % at different phosphorus concentrations ....................................18
2.3
Average ammonia loss % and nitrifier MPN at different zinc concentrations ..................19
2.4
Average ammonia loss % and MPN values at different alkalinity added..........................19
2.5
Effect of copper and zinc concentrations on ammonia loss at two different alkalinities .....20
2.6
Average ammonia loss % at different alkalinity and phosphate levels in lead pipes ........20
2.7
Nitrifier MPN at 5 ppb phosphorus ...................................................................................21
3.1
Modified CDC reactor .......................................................................................................30
3.2
Typical setup for a reactor .................................................................................................30
3.3
Bulk water NH3 concentration in different reactors during the eight hours stagnation ....31
3.4
Biofilm cell density in different reactors ...........................................................................31
3.5
Average bulk water MPN for AOB and NOB from different reactors ..............................32
3.6
Average (n=12) HPC value for different reactors..............................................................32
3.7
Average (n=12) pH of bulk water after eight hours of stagnation .....................................33
3.8
Average (n=12) alkalinity used/mg of ammonia nitrogen oxidized in different reactors .... 33
3.9
Aferage effluent total and dissolved copper concentrations in different copper reactors ....34
3.10
DGGE profile of biofilm sampled from different reactors ................................................34
6.1
Rough conceptualization of relative secondary disinfectant advantages ...........................48
6.2
Pipe rig setup......................................................................................................................49
6.3
Total chlorine residual decay in different pipe materials...................................................49
xi
©2010 Water Research Foundation. ALL RIGHTS RESERVED
xii | Effect of Nitrification on Corrosion in the Distribution System
6.4
Chlorine and chloramine decay after 6 hours stagnation (Phase III) as total chlorine
doses were ramped up ............................................................................................50
6.5
Ammonia loss % and MPN with PVC, iron and new lead pipes .......................................51
6.6
Bulk water HPC in PVC pipes ...........................................................................................51
6.7
Biofilm HPC (Phase II)......................................................................................................52
6.8
Bulk water HPC in galvanized pipe (Phase I-no nitrification) ..........................................52
6.9
Biofilm HPC with and without nitrification ......................................................................53
7.1
Nitrification effect on water chemistry ..............................................................................62
7.2
Total zinc release versus time in galvanized pipes ............................................................63
7.3
Water sample from galvanized pipes .................................................................................63
7.4
Zinc release in galvanized pipe ..........................................................................................64
7.5
Total copper release versus time in copper pipes ..............................................................64
7.6
Predicted soluble copper at different pH and alkalinities ..................................................65
7.7
pH change and calcium leaching from concrete w/ and w/out nitrification at different
intitial pH and alkalinities ......................................................................................65
8.1
Pipe rig setup......................................................................................................................65
8.2
Water preparation for Phase II and Phase III study ...........................................................80
8.3
Average lead releae after a stagnation period of 3.5 days in the pipes ..............................81
8.4
Average copper release after a stagnation period of 3.5 days in the pipes ........................81
8.5
pH values after water passing through GAC filtration or after mixing fresh water with
aged water ..............................................................................................................82
8.6
Nitrification in PVC pipes with chloramine disinfectant ...................................................82
8.7
Representative nitrogen balance 24 hours after water was fed into PVC pipes ................83
8.8
Effect of GAC treatment on lead release (Phase II) ..........................................................83
8.9
Effect of GAC treatment on copper release (Phase II) ......................................................84
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Figures | xiii
8.10
Final pH and ammonia consumption correlation in PVC pipes after 24 hours stagnation .. 84
8.11
Lead release in copper-lead rig, non-GAC ........................................................................84
8.12
Coppper release in copper-lead rig, non-GAC ..................................................................85
9.1
Geographic location of the respondents .............................................................................97
9.2
Number of persons served by each respondent..................................................................98
9.3
Total storage capacity in the distribution system...............................................................98
9.4
Proportion of pipe material for specific pipe diameter categories in the distribution
system ....................................................................................................................99
9.5
Percentage of homes built before 1986 and served by the respondents ..........................100
9.6
Number of years of experience with chloramination .......................................................100
9.7
Frequency of nitrification occurrences in the respondents’ distribution system .............100
9.8
Methods used to monitor for nitrification ........................................................................101
9.9
Frequency of monitoring..................................................................................................101
9.10
Methods used to prevent for nitrification.........................................................................102
9.11
Methods used to correct nitrification ...............................................................................102
9.12
90th percentile of lead concentrations measured in Bangor’s distribution system ...........103
9.13
Lead concentrations measured in Bangor’s distribution system in June 2007 ................103
9.14
90th percentile of copper concentration measured in Bangor’s distribution system ........104
9.15
Copper concentrations measured in Bangor’s distribution system in June 2007 ............104
9.16
Lead and copper levels at the Pinellas’ Keller 1 POE and corresponding distribution
system sampling locations ...................................................................................105
9.17
Nitrification activity in first draw samples in St. Paul, MN ............................................105
9.18
Water quality in first draw samples in St. Paul, MN .......................................................106
9.19
Lead levels in St. Paul, MN .............................................................................................106
©2010 Water Research Foundation. ALL RIGHTS RESERVED
xiv | Effect of Nitrification on Corrosion in the Distribution System
9.20
Nitrification activity in the anonymous utility .................................................................106
9.21
Water quality in first draw samples in the anonymous utilty ..........................................106
9.22
Lead, copper and zinc levels in the anonymous utility ....................................................107
9.23
Nitrification activity at IRWD utility...............................................................................108
9.24
Chlorine decay at IRWD utility .......................................................................................108
9.25
Nitrification activity in first draw samples at Portalnd utility .........................................108
9.26
pH and HPC at Portland utility ........................................................................................109
9.27
Lead and copper levels at Portland utility........................................................................109
9.28
Nitrification activity in the first draw samples in Bangor utility .....................................110
9.29
pH and total chlorine in the first draw samples in Bangor utility ....................................110
9.30
Lead and copper levels in Bangor utility .........................................................................110
9.31
pH and total chlorine in the first draw samples in Hampden utility ................................111
9.32
Water quality change in Hampden Wate District ............................................................111
A.1
Conceptual model illustrating coexistence of nitrifiers and heterotrophs .......................139
A.2
Four states of nitrifier activity changing with metal concentration .................................140
A.3
Free Cu (II) concentrations in relation to pH and phosphorus levels ..............................140
A.4
Disinfectant, nitrifiers interaction and nutrient release on iron pipe surface ...................141
B.1
Ammonia loss % after 3.5 day stagnation and total Cu release from brass pipes ...........159
B.2
Ammonia loss % after 3.5 day stagnation and total Cu release from brass pipe .............159
B.3
TOC levels after 3.5 day stagnation .................................................................................160
B.4
Copper release in copper pipes ........................................................................................160
B.5
Average ammonia loss % versus pH at different phosphorus levels in brass pipes ........161
B.6
Correlation of ammonia % vs. soluble Cu in copper and brass pipes..............................161
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Figures | xv
C.1
Standard curve for primer set PAER-f/PAER-r used for quantification of
P. aeruginosa: example 1 ....................................................................................171
C.2
Standard curve for primer set PAER-f/PAER-r used for quantification of
P. aeruginosa: example 2 ....................................................................................172
C.3
Standard curve for primer set CTO189f/RT1r used for quantification of AOB ..............172
C.4
DNA extraction efficiencies for laboratory scale plumbing systems ..............................173
C.5
DNA extraction efficiencies for large scale rig samples .................................................173
C.6
DAPI staining results for a 400x concentrated sample ....................................................174
C.7
DAPI staining results for a N. europaea cells at 108 cells/ml .........................................174
D.1
Ammonia loss % and final pH, versus time when initial alkalinity was reduced in stages
from 100 to 0 mg/L ..............................................................................................184
D.2
Total and soluble lead release versus time in lead pipes at 1 mg/L as P .........................185
D.3
Average soluble and particulate lead release at different alkalinity levels and
1 mg/L as PO4-P ...................................................................................................185
D.4
Predicted % soluble lead increase due to pH decrease (assuming 2000 ppb lead) ..........186
D.5
Actual soluble lead vs. predicted soluble lead in lead pipes at the control condition......186
D.6
Resulting pH drop based on ammonia oxidized at different initial alkalinity and pH ....187
©2010 Water Research Foundation. ALL RIGHTS RESERVED
©2010 Water Research Foundation. ALL RIGHTS RESERVED
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.
David E. Rager
Chair, Board of Trustees
Water Research Foundation
Robert C. Renner, P.E.
Executive Director
Water Research Foundation
xvii
©2010 Water Research Foundation. ALL RIGHTS RESERVED
©2010 Water Research Foundation. ALL RIGHTS RESERVED
ACKNOWLEDGMENTS
The authors of this report greatly appreciate the following water utilities for their
cooperation and participation in the project: Dallas Water Utilities (TX), Irvine Ranch Water
District (CA), Saint Paul Regional Water Services (MN), Greenville Utilities (NC), Portland
Water Bureau (OR), Bangor Water District (ME), Hampden Water District (ME), East Bay
Municipal Water District (CA), City of Ottawa (Ontario, Canada), Philadelphia Water
Department (PE), City of Raleigh (NC), Portland Water District (ME), Pinellas County Utilities
(FL) and Copper Development Association for their support.
We also appreciate the input of the Project Advisory Committee (PAC) members:
Franklyn Smith, Regional Municipality of Waterloo; Daniel Noguera, University of Wisconsin at
Madison and Philippe Piriou, CIRSEE, Suez Environment and the help of the Water Research
Foundation project officer Djanette Khiari. The authors would also like to thank Paul Blouch,
Ben Custalow, Jeff Coyne, Allian Griffin, Rachel Methvin, Anusha Kashyap, Meridith Raetz and
Paolo Scardina for help in maintaining experiments.
xix
©2010 Water Research Foundation. ALL RIGHTS RESERVED
xx | Effect of Nitrification on Corrosion in the Distribution System
©2010 Water Research Foundation. ALL RIGHTS RESERVED
EXECUTIVE SUMMARY
OBJECTIVES
The objective of this research was to comprehensively study nitrification in drinking
water systems targeting two aspects: the impact of water quality and pipe corrosion on
nitrification occurrence and the impact of nitrification on water quality and pipe corrosion.
BACKGROUND
In the United States, utilities are increasingly using chloramine to comply with
regulations for disinfection by-products. The ammonia formed via chloramine decay can support
autotrophic microbial nitrification. Nitrification can create levels of nitrite that exceed the
maximum contaminant level (MCL), stimulate growth of heterotrophic bacteria, contribute to
loss of disinfectant, and also create problems with lead and copper contamination from corrosion
of premise plumbing systems. Previous research indicates that nitrification was detected in about
two thirds of medium and large U.S. systems using chloramines. Given the high costs of
corrosion to utilities and consumers, and other possible health concerns related to nitrification, it
is critical to better understand nitrification.
APPROACH
The study used a systematic approach to study nitrification in drinking water systems,
including a comprehensive literature review, bench and then large scale studies, followed by
utility sampling. The specific research conducted in this study included the following:
•
•
•
•
•
Investigated effects of significant trace micro and macro nutrients on nitrification
activity, and studied the role of different pipe materials and pipe corrosion in
modifying nutrient levels and nitrification activity
Examined the interplay between nitrifiers and heterotrophic bacteria
Verified anecdotal links established between nitrification and pipe corrosion in
both well-controlled laboratory and simulated home plumbing systems, and
conducted a mechanistic basis for these links
Examined the role of nitrification in modifying disinfectant efficiency
Conducted utility sampling to document nitrification effects on corrosion and the
effects of corrosion on nitrification
RESULTS/CONCLUSIONS
Plumbing materials had profound impacts on the incidence of nitrification in homes.
Effects were due to toxicity (i.e., release of Cu+2), recycling of nitrate back to ammonia substrate
by reaction (zerovalent iron, lead, or zinc materials), or release of nutrients that are essential to
nitrification by leaching from concrete or other materials.
xxi
©2010 Water Research Foundation. ALL RIGHTS RESERVED
xxii | Effect of Nitrification on Corrosion in the Distribution System
Phosphate plays a key role in determining where, when, and if problems with nitrification
will occur in a given water distribution system. High levels of phosphate inhibitor can provide
nutrients to nitrification, reduce the concentration of Cu+2 ions, and make nitrification more
likely, but phosphate can also sometimes lower corrosion rates and increase the stability of
disinfectant and its efficacy on controlling nitrifiers.
Dependent on circumstances, nitrification had an increased, decreased, or no effect on
aspects of materials corrosion. For example, nitrification markedly increased lead contamination
of low alkalinity potable water by reducing the pH, but dramatically decreased leaching of zinc
to potable water from galvanized iron. Nitrification did not affect copper solubility in low
alkalinity water, but is expected to increase copper solubility in higher alkalinity waters.
Experiments also verified that nitrification could affect the relative efficacy of chlorine
versus chloramines in controlling heterotrophic bacteria in premise plumbing.
APPLICATIONS/RECOMMENDATIONS
Many of the key findings of the research are captured in the following decision tree
(Figure ES.1) that can be used by utilities, regulators, and scientists. If there is nitrification in
the main water distribution system or in the premise plumbing sampled for lead and copper
under the USEPA Lead and Copper Rule (LCR), under certain circumstances significant effects
on lead corrosion (i.e., low alkalinity water) and copper corrosion (i.e., high alkalinity) can be
anticipated. It may be necessary to implement corrosion control or nitrification control to stop
these problems.
Even when problems are not apparent through sampling in the pre-1986 homes in the
LCR sampling pool, “hotspots” can occur in other “worst case” consumers’ homes or in schools.
Specifically, buildings with galvanized iron, iron, or plastic plumbing materials can sometimes
develop rampant nitrification problems that can be self-perpetuating due to rapid loss of
disinfectant. Other worst case situations can occur for homes near the end of the distribution
system or which utilize GAC treatment devices. These hotspots could be sites where secondary
disinfection is ineffective, or where lead, copper, and other corrosion problems are exacerbated.
Depending on the seriousness of the problems, corrosion or nitrification control could be
implemented to try and improve water quality in these situations.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Executive Summary | xxiii
Figure ES.1 Decision Tree
RESEARCH PARTNER
USEPA
PARTICIPANTS
Thirteen utilities from the United States and Canada participated in this project.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
xxiv | Effect of Nitrification on Corrosion in the Distribution System
©2010 Water Research Foundation. ALL RIGHTS RESERVED
CHAPTER 1
INTRODUCTION
BACKGROUND
The use of chloramines is often the lowest cost means of complying with the United States
Environmental Protection Agency (USEPA) Stage 1 and Stage 2 Disinfectants and Disinfection
By-Products Rule (D/DBPR). About 30% of surface water treatment plants currently use
chloramines versus 20% in 1990, and the percentage of surface water treatment plants using
chloramines might rise to as high as 40-65% in the near future (Seidel et al. 2005; USEPA 2005;
Zhang et al. 2009a). Free ammonia present in the finished drinking water or formed during
chloramine decay can trigger nitrification, which is the conversion of ammonia to nitrite and then
nitrate by autotrophic ammonia oxidizing bacteria (AOB) and nitrite oxidizing bacteria (NOB),
respectively. Nitrification can have significant effect on water quality and pipe corrosion (Table
1.1). See Appendix A for a comprehensive review of nitrification in drinking water systems.
Past research on nitrification has been focused on water utilities and main distribution
systems emphasizing consequences from lower pHs and loss of disinfectant residual. But recent
experience demonstrates that nitrification sometimes causes serious corrosion problems in premise
(building) plumbing systems, as manifested by increased lead and copper leaching to water, blue
water, copper pinhole leaks and even brass failures. It is also clear that nitrification not only
influences corrosion, but that corrosion of iron, copper and concrete can influence nitrification in
ways that are sometimes synergistic and other times antagonistic.
OVERVIEW OF NITRIFICATION EFFECTS ON CORROSION
Nitrification occurs via the following two step microbiological process:
NH4+ + 1.38 O2 + 0.0172 HCO3- + 0.069 CO2Æ 0.0172 C5H7O2N + 0.983 NO2- + 0.966
H2O + 1.97 H+
(1.1)
NO2- + .00875 NH4+ + 0.035 H2CO3 + .00875 HCO3- + 0.456 O2 + 0.00875 H2O Æ
0.00875 C5H7O2N + 1.0 NO3(1.2)
These reactions are mediated by autotrophic ammonia oxidizing bacteria (AOB) and nitrite
oxidizing bacteria (NOB), respectively. The overall acid producing reaction yields 2 molecules of
H+ for every mole of ammonia-N converted to nitrite. Dissolved inorganic carbon (DIC) in the
water is used primarily to neutralize the acid generated by the AOBs, but some is removed and
converted to organic carbon in the form of biomass (Zhang et al. 2009a).
Although the profound adverse impacts of nitrification on corrosion of lead pipe in
drinking water were first noted more than 100 years ago (Garret 1891) and the issue of nitrification
in chloraminated water supplies was reasonably described 70 years ago (Feben 1935; Hulbert
1933; Larson 1939), there have been no modern studies explicitly examining the
inter-relationships between nitrification and corrosion problems for a wide range of distribution
system materials and water qualities. Excellent work has been done in Pinellas County, Florida
(Powell 2004) that highlighted some concerns related to iron corrosion control and red water.
1
©2010 Water Research Foundation. ALL RIGHTS RESERVED
2 | Effect of Nitrification on Corrosion in the Distribution System
Furthermore, elevated copper levels at the tap were clearly tied to activity of nitrifying bacteria in
Willmar homes (Murphy et al. 1997a; Murphy et al. 1997b), and nitrification was implicated in
higher lead leaching in Ottawa (Douglas et al. 2004) and to some extent in Washington D.C.
(Edwards and Dudi 2004).
FACTORS AFFECTING NITRIFICATION OCCURRENCE: INTERACTION OF
MATERIALS AND WATER CHEMISTRY
Many factors (Table 1.2) are well known to influence nitrification occurrence in drinking
water systems. Besides these well known factors, the authors also identified several other
important factors in drinking water systems include nutrients, pipe materials and household
treatment methods (Zhang et al. 2009a).
Nutrient Effects on Nitrification
Nitrifier growth depends on the availability of nutrient(s), as illustrated by the following
overly simplistic equation:
DIC + N + P + Trace Nutrients (K, Cu, Fe, etc.) + O2 → Nitrification
(1.3)
Each nutrient has four possible metabolic impacts on nitrification, depending on its
concentration (Figure 1.1). Deficiencies of nutrients can decrease growth or cause bacterial death
(Reeves et al., 1981; Fransolet et al., 1988). As the nutrient concentration increases, bacterial
activity may be restored, and reaches an optimum depending on the circumstance of growth. At
excess concentrations, however, detrimental effects may result.
The role of nutrients in nitrification has been studied since the 1960’s, but only in fairly
well mixed pure lab culture studies or activated sludge. For water chemistries found in drinking
water and during unmixed conditions occurring during stagnation, there is no research on the
potential role of trace nutrients in nitrification occurrence and control. This is an important gap
given the extreme variability of key nutrients for nitrification illustrated by analysis of survey data
from 330 U.S. raw drinking waters— nutrients influencing nitrification were found to occur at
levels expected to limit nitrifier growth in some water systems while inhibiting growth in others
(AWWARF, 2004, Zhang and Edwards, 2005). Our bench scale studies also demonstrated effects
of trace nutrients under typical drinking water conditions in home plumbing (Zhang and Edwards,
2005). Specifically, the concentration of calcium, copper, potassium, zinc and phosphorus affected
nitrification when varied over the concentration range encountered in drinking water. The different
nutrient levels might help explain variations observed from system to system and houses within a
system. For nutrients/toxins that arise from corrosion of distribution system materials, such as
Cu+2 from copper pipe, effects on nitrification rate can even help explain variability from one line
of plumbing in a home to another.
Effect of Pipe Materials and Corrosion Control Strategies on Nitrification Occurrence
The preceding examples illustrate that distribution system materials and corrosion control
strategies can be expected to influence the rate and extent of nitrification. Thus, nitrification not
only influences corrosion, but corrosion can also influence or even control nitrification. Possible
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 1: Introduction | 3
mechanisms regarding the effects of different pipe materials and home treatment device are
summarized (Table 1.3).
This issue has already been examined extensively for heterotrophic bacteria. In general,
unlined iron is more supportive of heterotrophic growth relative to PVC, cement and epoxy
materials (Camper et al. 2003), due to removal of the disinfectant and release of iron and other
trace nutrients such as phosphorus from the corroding metal (Morton et al., 2005).
It has been speculated that iron tubercles in distribution pipes are also conducive to nitrifier
growth (Odell et al., 1996), a hypothesis supported by extensive utility experience and recent
experimental work of others (Karim and LeChevallier, 2006). Iron pipe can accelerate chloramine
decay and release ammonia for nitrifying bacteria growth (Woolschlager et al., 2001):
1/2NH2Cl + H+ + Fe2+→ Fe3+ + 1/2NH4+ +1/2 Cl-
(1.4)
Piping made of concrete materials are also of special interest, since they can leach lime and
therefore have a higher surface pH (> pH 12 for fresh concrete) than bulk distributed water
(AWWARF and DVGW-TZW 1996). Leaching of lime and other trace nutrients from concrete is
very dependent on the water chemistry (Guo et al., 1998). Perhaps not surprisingly, concrete lined
pipes had the lowest AOB and HPC in a study at two California utilities using chloramine
(Steward and Lieu 1997), yet cement lined iron pipe had a higher heterotrophic biomass than did
unlined iron in Pinellas county, Florida (LePuil et al., 2003). It is speculated that high pH on fresh
concrete surface can act to prevent nitrification, whereas on older concrete slightly higher pH,
leached nutrients and free ammonia formed by chloramine auto-decomposition can enhance
nitrifier growth (Woolschlager and Soucie, 2003).
For lead pipes, leaded solder and leaded brass, two contrary effects on nitrifying bacterial
growth could exist. On one hand, corrosion reactions between lead and nitrate could recycle
ammonia from nitrate and support a large nitrifier population even when there are very low levels
of ammonia (Uchida and Okuwaki, 1998; Edwards & Dudi, 2004). On the other hand, soluble lead
leached during corrosion might be able to inhibit nitrification, although a reported level of 0.005 to
0.5 ppm lead had no effect on Nitrosomonas in a pure culture suspended growth experiment
(Loveless and Painter, 1968).
Special Focus on the Interaction of Phosphorus, pH and Copper Pipe
Copper pipe is of special interest since it is the dominant plumbing material in buildings,
and it is also in electrical contact with lead solder and leaded brass resulting in galvanic corrosion.
Free copper released from copper pipe is a key parameter in controlling nitrification (Figure 1.2),
especially given that the biofilm attached to the pipe can locally depress pH.
Phosphate and pH interact to control free and total copper concentrations (Edwards et al.
2002; Schock et al. 1995). Fifty six percent of water utilities add phosphorus inhibitor to water at a
level of 0.2-3 mg/L PO4-P. It is believed that a low solubility Cu3PO4 or similar scale forms,
whereas in situations without phosphate other minerals such as Cu(OH)2 scale, CuO, malachite or
other layers are present (McNeill and Edwards, 2002). When Cu(OH)2 is present as in the case of
new pipe, free Cu 2+ released to the water is much higher than when phosphorus is added,
especially at lower pHs (Figure 1.2). Consequently, nitrification may be inhibited by toxic levels
of free copper at lower pH, and might be accelerated by trace levels of free copper at higher pH.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
4 | Effect of Nitrification on Corrosion in the Distribution System
Assuming a threshold of 0.1 ppm Cu+2 based on the completely mix suspended growth
tests by Skinner & Walker, 1961, toxicity to nitrifiers is expected at surface pHs of about 7.3 or
lower in the absence of phosphate, but pH as low as 6.7 would not be toxic in the presence of high
phosphorus (Figure 1.2). Thus, use of phosphate inhibitor might be expected to reduce copper
toxicity to nitrifiers in copper tube when biofilm pH is below 7. Use of phosphorus inhibitor could
also eliminate potential P-limitations to nitrifier growth.
Granular Activated Carbon (GAC) Filters in Homes
Some consumers use whole house or point of use filters for removal of organic matter, and
control of taste and odor problems. As early as 1935, Feben recognized that filter beds receiving
water with ammonia can offer a nearly ideal environment for nitrifying bacteria. Granular
activated carbon (GAC) has high porosity and can increase nutrient adsorption and bacterial
attachment (Rollinger and Dott, 1987). Nitrification was reported to occur after installing GAC in
Ann Arbor, Michigan (Skadsen, 1993) and one distribution system in Finland.
GAC filters convert chloramines to free ammonia and therefore enable nitrifiers to
proliferate (Tokuno 1997; Vahala 2002). Our field observations indicate that if GAC filters are
installed for whole house or point of use filters in chloraminated systems, nitrification effects can
be dramatic as measured by corrosion impacts on copper and lead.
Interaction of Nitrifiers and Heterotrophic Bacteria and Heterotrophic Nitrification in
Drinking Water Systems
Increased heterotrophic bacteria are always found in association with nitrifying bacteria
when nitrification problems occur (Wolfe et al., 1990). Nitrifiers can increase heterotrophic
growth by producing soluble organic (Rittmann et al., 1994). Heterotrophs can be beneficial to
nitrification by producing stimulating organics for nitrifiers (Hockenbury et al. 1977; Pan and
Umbreit, 1972b) and protecting nitrifiers from detachment (Rittmann and Manem, 1992; Furumai
and Rittmann, 1994; Rittmann et al., 1994). In other cases, heterotrophs can be detrimental to
nitrifiers since they compete for surfaces, dissolved oxygen, and ammonia (Rittmann and Manem,
1992). Nitrifiers and heterotrophs could be detrimental or beneficial to corrosion depending on
circumstance.
Many heterotrophic bacteria have also been found to contribute to nitrification (Verstraete
and Alexander, 1973 and 1986; Focht and Verstraete, 1977; Watson et al., 1989; Killham, 1986;
Bock et al., 1992), although with a slower rate and different mechanisms than autotrophic
nitrifiers. This is of special interest in drinking water systems, because unlike autotrophic
nitrification, heterotrophic bacteria do not consume dissolved inorganic carbon, and net changes to
water chemistry from heterotrophic nitrification will differ from autotrophic nitrification.
RESEARCH OBJECTIVES
The objective of this research is to study nitrification effects on corrosion, and corrosion
effects on nitrification, in order to answer many of the key questions confronting consumers,
regulators and the water industry. Three phases of research were performed.
For Phase 1, a comprehensive literature review was completed, and data was collected
from participating utilities to document effects of nitrification on corrosion, and effects of
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 1: Introduction | 5
corrosion on nitrification in drinking water distribution systems and premise (buildings/homes)
plumbing.
For Phase 2, bench scale studies were conducted to evaluate 1) impact of nutrients on
nitrification occurrence 2) interplay between pH, certain premise plumbing materials (copper,
brass, PVC, lead) and phosphate in controlling nitrification and metals leaching, and 3) effects of
organic carbon on nitrification and corrosivity.
For Phase 3, large scale pipe rigs were used to examine 1) role of different pipe materials
on nitrification occurrence, 2) effect of nitrification on the corrosion of 10 materials (new lead
pipe, old lead pipe, PVC, PEX, Stainless Steel, Galvanized Iron, Copper, Epoxy lined copper, new
cast iron, old cast iron), and 3) other nitrification effects including disinfectant decay and bacterial
growth. Another set of experiments in Phase 3 investigated the effect of nitrification and Granular
Activated Carbon (GAC) on corrosion of galvanically connected copper pipe/lead solder, copper
pipe/leaded brass, and copper pipe/lead pipe.
FIGURES AND TABLES
Optimal
Inhibiting
Toxic
Nitrifier Nutrient
Figure 1.1 General effect of a given nutrient concentration on nitrifier activity
©2010 Water Research Foundation. ALL RIGHTS RESERVED
6 | Effect of Nitrification on Corrosion in the Distribution System
Figure 1.2 Free Cu concentrations in relation to pH and phosphorus levels
Source: (Zhang et al. 2009a)
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 1: Introduction | 7
Table 1.1
Key effects of nitrification on water quality and corrosion
(Zhang et al. 2009a)
Change due
to
nitrification
Possible
direct/indirect effect
on water quality
Possible effect on
corrosion-materials
degradation
Nitrite MCL,
increased lead from
brass, disinfectant
loss
Samples for nitrite MCL not
Increased microbial corrosion, collected in premise plumbing,
nitrite catalyzed stress
so potential problem can be
corrosion failures and attack
missed.
on grain boundaries (brass)*
T&O from corrosion,
red and blue water
complaints.
Divergence from targeted
optimal corrosion control
relative to finished water
Lead and Copper Rule, Toxicity
from Blue Water typically
occurs in new homes not tested
in LCR
Concern over
pathogen re-growth
and loss of
disinfection
More microbial corrosion,
likely link to some cases of
pinhole leaks in copper tube
Total coliform rule, HPC action
levels, Legionella and
Mycobacterium as emerging
issues
Failure to maintain
Rapid decay
residual at distant
of chloramine parts of distribution
system
Effect on corrosion rates
dependent on relative
corrosivity of chloramine vs.
decay products
Nitrite
production
Lower pH,
alkalinity and
DIC
Higher HPC
Decreased
DO
Low redox in iron
pipe associated with
more red water
Highly corrosive sulfate
reducing bacteria are
anaerobic
Other Concerns
Chlorine residual not routinely
monitored in premise plumbing,
where it controls opportunistic
pathogens
Lead and Copper Rule, color,
taste and odor complaints
©2010 Water Research Foundation. ALL RIGHTS RESERVED
8 | Effect of Nitrification on Corrosion in the Distribution System
Table 1.2
Factors affecting nitrification in drinking water systems
Requred Range
Ammonia
0.05-3000 mg/L-N (Baribeau, 2006)
Dissolved oxygen required, but can survive
oxygen
at low DO
Temperature
25-30 °C
light
sensitvie to light
4.6-11.2 (Wolfe et al., 2001; Prakasam and
pH
Loehr, 1972)
Optimal
14-350 mg/L-N (Wolfe and Lieu,
Drinking water condition
2001)
0.05-1 mg/L-N
O2 except for dead ends and
within pipe scale
varies by time and location
absent in distribution system,
present in open resevoir
> 0.5 to 4 mg/L
4-60 °C (Wolfe et al., 2001)
7--8
neutral to alkaline, optimal
for nitrifier growth
sufficient alkalinity to support
nitrification
(USEPA, 2005)
In wastewater 14 mg/L as CaCO3
Alkalinity
required/mg NH4+-N
Table 1.3
Effect of pipe materials and GAC on nitrification occurrence
Postive impact on nitrification
Negative impact on nitrification
iron
Source of Fe, P, other nutrients, allows ammonia cycling;
destroys chloramine and creates free ammonia
None known.
concrete
nutrients released, destroys chloramine and creates free
ammonia, higher pH and alkalinity
pH can be too high on new
concrete
copper
lead
GAC
serve as source of Cu, destroys chloramine and creates
free ammonia
toxic Cu level
convert nitrate to ammonia
None known.in water environment
destroy disinfectant and create free ammonia, good
surface for attachment
©2010 Water Research Foundation. ALL RIGHTS RESERVED
CHAPTER 2
EFFECT OF AMBIENT NUTRIENTS AND METALS ON NITRIFICATION
IN PREMISE PLUMBING
Yan Zhang and Marc Edwards
This work screened for effects of possible trace micro and macro nutrients on nitrification
rates. Different pipe materials were then examined from the perspective of their ability to modify
nutrient levels in nitrifier biofilms. Finally, the role of organic carbon was examined, from the
perspective of the possible interplay between nitrifiers and heterotrophic bacteria.
INTRODUCTION
Nitrification is an increasing concern in U.S. drinking water systems, since utilities are
more frequently using chloramines to comply with regulations for disinfection byproducts (Zhang
et al. 2009a). The ammonia formed via chloramine decay can support autotrophic microbial
nitrification. A survey conducted in 1991 indicates that 63% of the utilities using chloramines
experienced nitrification problems (Wilczak et al. 1996). Nitrification can stimulate growth of
heterotrophic bacteria, contribute to loss of disinfectant, and also create problems with lead and
copper contamination from corrosion of premise plumbing systems.
Nitrification in drinking water can be affected by the availability of nutrients and potential
microbial toxins in a potable water supply. This dependency has been explicitly addressed for
chlorine disinfectants and free ammonia substrates in work by other researchers (Fleming et al.
2005; Fleming et al. 2008). For example, a nitrification potential curve is proposed based on
equation (Fleming et al. 2005; Fleming et al. 2008):
[Total chlorine] =
Rgi [ free ammonia ]
(2.1)
Ks + [ free ammonia ]
Where [Total chlorine] = the sum of free chlorine, monochloramine and dichloramine
concentrations, mg/L-Cl2
Rgi: the minimum total chlorine concentrations needed to prevent nitrification for any free
ammonia concentration, mg/L-Cl2
Ks: half saturation constant for Ammonia oxidizing bacteria, mg/L-N
[Free ammonia]: the sum of ammonia (NH3) and ammonium (NH4+) concentrations,
mg/L-N
In essence, this equation illustrates that if the death rate of nitrifying bacteria via
disinfection exceeds the growth rate from ammonia consumption, then nitrification has difficultly
becoming established. In contrast, if the nitrifier growth rate exceeds the death rate, then serious
nitrification problems can occur even under continuous flow conditions present in water
distribution systems (Fleming et al. 2005; Fleming et al. 2008).
While this earlier work (Fleming et al. 2005; Fleming et al. 2008) is an important first step,
there are many other nutrients and toxins present in a water distribution system that also can play
9
©2010 Water Research Foundation. ALL RIGHTS RESERVED
10 | Effect of Nitrification on Corrosion in the Distribution System
controlling roles in nitrification. For example, like all microbial growth, nitrifiers need certain
nutrients in order to grow as represented with an overall equation (Zhang et al. 2009a):
Carbon + Nitrogen + Phosphorus + Other Nutrients (K, Mg, Ca, Co, Mo) + O2 →
nitrification
(2.2)
In this work, inorganic carbon, nitrogen, phosphorus and oxygen are termed macro
nutrients, because they need to be present in significant quantities, whereas other nutrients are only
required at trace levels. Each nutrient has four possible metabolic impacts on nitrification,
depending on its concentration. If the nutrient concentration is zero or too low, the deficiency can
decrease growth rate or even cause bacterial death (Reeves et al., 1981; Fransolet et al., 1988). As
the nutrient concentration increases, bacterial activity may be restored, and can reach an optimum
concentration depending on the circumstance of growth. At excess concentrations, for metals,
detrimental effects may result (Zhang et al. 2009a).
In a recent survey of 330 drinking waters in the U.S. (Parks et al. 2004), many elements
important to nitrifier growth were found to have a wide range of occurrence in the source water
before treatment (Zhang et al. 2009a). For example, Loveless and Painter 1968 demonstrated that
addition of 12.5 to 50 ppm magnesium stimulates the growth of nitrifying bacteria, whereas levels
above 50 ppm inhibit growth (Loveless and Painter 1968). If these thresholds are applied to
potable water systems, it is estimated that in about 40 % of US waters magnesium is below 12.5
ppm, whereas about 10 % waters have magnesium levels that may inhibit growth of nitrifying
bacteria.
In addition to the nutrients coming from the raw water, pipe materials can also release or
modify the nutrient levels available to nitrifiers (Zhang et al. 2009a). For example, earlier studies
demonstrated that iron pipe can release bioavailable phosphorus for bacterial growth (Morton et al.
2005), and similar reactions might be important for nitrifiers. Iron pipe may also form ammonia
from nitrate (Huang and Zhang 2005; McIntyre and Mercer 1993; Zhang and Edwards 2007),
thereby allowing a relatively small amount of free ammonia to sustain large populations of
nitrifying bacteria. Lead pipe is also capable of recycling ammonia from nitrite and nitrate
(Edwards and Dudi 2004; Zhang et al. 2009a).
The interplay between organic carbon, and the potential competition between
heterotrophic bacteria and nitrifying bacteria also is worthy of consideration. Research in
wastewater systems has demonstrated that nitrifiers and heterotrophs compete for surfaces,
dissolved oxygen, ammonium and other nutrients (Rittmann and Manem 1992). Nitrification is an
energy intensive process and nitrifiers have lower growth rates compared to heterotrophs
(Rittmann and Manem 1992). So if there is higher organic carbon or if ammonia is limited,
nitrifiers’ numbers may be reduced by heterotrophs due to competition for ammonia (Ohashi et al.
1995; Verhagen and Laanbroek 1991). On the other hand, when no dissolved organic carbon is
present in the water, heterotrophs can be completely dependent on the lysis and extracellular
products of nitrifiers as their source of organic carbon (Rittmann et al. 1994).
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 2: Effect of Ambient Nutrients and Metals on Nitrification in Premise Plumbing | 11
MATERIALS AND METHODS
Experimental Setup
Nutrient Effect Study in PVC Pipe
PVC pipes were used as a control material for this study since it is relatively inert to
nitrifiers. Nitrifiers were established in PVC pipes (30 cm length × 1.9 cm diameter) using
dechlorinated Blacksburg tap water. The dechlorinated tap water was adjusted for easy
nitrification establishment by adding 2 ppm-N (NH4)2SO4 and pH to 8. The specifics of the
approach were described elsewhere (Zhang et al. 2008a). Water in the pipes was changed twice a
week using a “dump and fill” protocol to simulate infrequent water use in buildings and to
replenish nutrients for microbial nitrifier growth. These pipes were maintained at room
temperature and covered with black plastic throughout the whole experiment to prevent possible
light inhibition to nitrification (Wolfe and Lieu 2001). Complete nitrification (100% ammonia
loss) was confirmed to occur in PVC pipes within 24 hours after 3 months of inoculation.
The water fed into the pipes was then changed to synthesized potable water. This
synthesized water contained (NH4)2SO4 (1 ppm-N), MgSO4 (0.5 ppm-Mg), CaCl2 (1 ppm-Ca),
KCl (10 ppm-K), Na2HPO4 (10 ppb-P), NaHCO3 (500 ppm) and other trace nutrients (5 ppb
Cu2+, 1.7 ppb Mo6+, 0.1 ppb Co2+, 5.6 ppb Mn2+, 2.6 ppb Zn2+ and 0.1 ppm Fe2+). The pH was
adjusted to 8 before filling up the pipes. This water simulates potable water conditions while
providing essential nutrients for nitrifying bacteria. Complete nitrification continued in all PVC
pipes. After two months of conditioning time, then the control water was modified to contain
different nutrient levels (Table 2.1). During the testing, the alkalinity of all water was also reduced
down to 10 mg/L to compare the effect of nutrient variation at different alkalinities. The role of
trace nutrients was also studied by eliminating a certain trace nutrient (no Mo6+, no Co2+, no
Mn2+, no Zn2+ or no Fe2+).
Effect of Inorganic Carbon and Phosphate Nutrients in Lead Pipes
Nitrification was established in lead pipes at 5, 60 and 1000 ppb orthophosphate-P (4 pipes
at each phosphate level). The alkalinity of the synthesized feeding water was dropped stepwise
from 100 ppm down to 30, 15 and then 0 ppm alkalinity, with sufficient time maintained at each
alkalinity for relatively stable levels of nitrification, pH and lead leaching to be achieved. Water in
the pipes was changed twice a week and nitrification activity was tracked at these different
phosphate and alkalinity conditions. The specifics of this test were described elsewhere (Zhang et
al. 2009b).
Nutrient Leaching from Copper, Concrete and PVC pipes
Some pipes (new PVC, soft and hard copper pipes) were purchased from the hardware
store to test the nutrient leaching from these materials. Pipes are of 30 cm length × 1.9 cm
diameter. For these pipes, the control water (without phosphate added) from the above nutrient
effect study in PVC pipes was filled into the pipes and kept stagnant for 96 hours, after which
samples of water were collected for analysis. A small piece of soft and hard copper was also
©2010 Water Research Foundation. ALL RIGHTS RESERVED
12 | Effect of Nitrification on Corrosion in the Distribution System
dissolved in 5% nitric acid and tested for composition to confirm the source if any nutrient was
released under potable water conditions
To test leaching of nutrients from concrete materials, glass pipes (30 cm length × 1.9 cm
diameter) containing small mounted coupons of aged concrete were used (Concrete SA:Volume =
0.6 cm-1). Parallel tests were conducted using glass pipes without concrete. Nitrifier growth was
established in both sets (w/ or w/o concrete) by inoculating pure culture Nitrosomonas europaea,
but no effort was made to sustain the pure culture over a period of three months. After stable
nitrification was established in both sets of glass pipes with or without concrete, inoculation was
stopped and water with different nutrient levels was introduced to the PVC pipes (Table 2.1).
Water was held stagnant for 96 hours between water changes and tested for nutrient changes and
nitrification activity.
Recycling of Ammonia from nitrite and nitrate by iron, copper, lead and zinc
To test the recycling of ammonia from nitrite and nitrate by iron, copper, lead and zinc,
amber glass TOC test vials (25 mm × 95 mm) with Mininert valve that allows for headspace
sampling were used (Lee 2004). Pure metal wires purchased from sigma Aldrich were used. Metal
wires (20 mm length × 1 mm diameter) were placed into the vials with 30 ml solution. There were
3 waters (control, control + 10 mg/L NO2-, control + 10 mg/L NO3-) × 3 metals (iron, lead and
zinc) × 2 oxygen conditions (aerobic and anaerobic-created by purging solution with N2 gas) × 2
pHs (6 and 8) × 3 duplicates = 108 vials. The control water was simulated Potomac water without
any nitrogen species added. Water in the vials was changed twice a week and tested for the change
of ammonia, nitrite and nitrate.
Effect of Organic Carbon Study in PVC Pipes
PVC pipes of 30 cm length × 1.9 cm diameter were used. The PVC pipes were exposed to
a synthesized water for one year without nitrification, and then exposed to water with ammonia
(and resulting nitrification) for 15 months as described elsewhere (Zhang et al. 2008a). No
disinfectant was ever added to these pipes. Thirty pipes were exposed at 5, 60 and 1000 ppb
orthophosphate-P (10 at each phosphate level). The ten replicate pipes were separated into three
groups. The first group was a control which was unchanged (4 pipes). The second group was
modified by incrementally adding organic carbon (3 pipes) at 20, 100, 500 and 1000 ppb TOC.
The organic used initially was ozonated natural organic matter (preparation method, see (Zhang
and Edwards 2007)). Thereafter, a more bioavailable organic carbon (glucose) which had been
used in other studies (Lechevallier et al. 1990) was used. In parallel, of the third group of pipes was
tested with the organic carbon, but without ammonia (3 pipes). So, overall, three types of water
were tested including 1) with ammonia and without added TOC, 2) with ammonia and with added
TOC, and 3) without added TOC or ammonia.
Analytical Methods
Nutrient Effect Study in PVC and Lead Pipes
Combined samples from the five duplicate pipes for each nutrient condition were collected
after 7 hours stagnation for quantification of nitrification activity. Nitrifier activity was mainly
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 2: Effect of Ambient Nutrients and Metals on Nitrification in Premise Plumbing | 13
tracked by measuring loss of ammonia, but production of nitrite and nitrate was also analyzed to
confirm the ammonia loss was due to nitrification. NH4-N was measured with salicylate method
using a HACH DR/2400 spectrophotometer, according to standard method 4500 NH3 (Clesceri et
al. 1998). NO2-N and NO3-N were measured using DIONEX, DX-120 ion chromatography (IC),
according to standard method 4110 (Clesceri et al. 1998). At the end of the experiment, nitrifier
density was also monitored by a five-tube Most Probable Number (MPN) procedure (Wolfe et al.
1990), the specifics of the method was described elsewhere (Zhang et al. 2008a).
Nutrient Release from Copper, Concrete and PVC pipes
Total metal and phosphorus release was quantified by digesting samples with 2% nitric
acid for 24 hours in a 90 °C oven. Metal concentrations were quantified using an Inductively
Coupled Plasma Mass Spectrophotometer (ICP-MS) according to Standard Method 3125-B
(Clesceri et al. 1998). Nitrifier activity was also tracked by ammonia loss and nitrite and nitrate
production.
Recycling of Ammonia from nitrite and nitrate by iron, copper, lead and zinc
Initial water and final water (combined sample from triplicate samples) for each condition
was measured at each water change for pH, NH4+ NO2- and NO3-.
Effect of Organic Carbon Study in PVC Pipes
Nitrifier activity was monitored by ammonia loss. Nitrifier density was monitored by the
five-tube Most Probable Number (MPN) procedure (Wolfe et al. 1990; Zhang et al. 2008a).
Heterotrophic bacteria were monitored with Heterotrophic Plate Count (HPC) according to
Standard Method 9215 (Clesceri et al. 1998) using the spread plate method with R2A medium.
RESULTS AND DISCUSSION
Results are organized in sections that examine 1) broad survey of nutrient effects on
nitrifier activity, 2) nutrient release and ammonia cycling by different pipe materials, and 3)
interaction of nitrifiers and heterotrophs.
Effect of Nutrient Level on Nitrifier Growth and Activity
Copper
Copper (II) is a key component of the ammonia monooxygenase (AMO) enzyme, which is
essential for ammonia oxidation and nitrifier growth (Ensign et al. 1993; Richardson and
Watmough 1999). However, excess copper is known to be toxic to nitrifiers (Braam and Klapwijk
1981). In this study, at alkalinity of 10 mg/L as CaCO3, the percentage ammonia oxidized
increased from 50 % to 70 % when the copper level was increased from 0 to 20 ppb. Increases of
the copper concentration to 100 ppb decreased ammonia oxidation down to 28 %, and further
increases in copper eliminated nitrifier activity (Figure 2.1). The nitrifier MPN increased by an
order of magnitude when 5 ppb copper was added versus no copper, and when copper
©2010 Water Research Foundation. ALL RIGHTS RESERVED
14 | Effect of Nitrification on Corrosion in the Distribution System
concentration was increased to above 100 ppb nitrifier MPN decreased by two orders of magnitude
comparing MPN value at 20 ppb copper. Hence, copper exhibits the complete range of metabolic
impacts on nitrifier activity. These results also provide an important practical confirmation
regarding the role of copper in preventing colonization by nitrifiers in copper containing materials
including brass and copper pipesversus PVC (Zhang et al. 2008a).
Phosphorus
Phosphorus is an essential nutrient for all microbial growth including nitrifying bacteria
(Zhang et al. 2009a). Previous studies have reported that at least 3-20 ppb phosphate is necessary
for nitrification in drinking water treatment (Van der Aa et al. 2002; Van Droogenbroeck and
Laudelout 1967). However, in this study, for the first two weeks, ammonia oxidation (30-60%)
was observed even in the pipes with no phosphate added (Figure 2.2), possibly because 10 ppb
phosphorus was previously dosed in order to establish nitrification and this phosphate might be
stored in biofilm or pipe surface deposits. This stored phosphorus can serve as a reservoir for
phosphorus after the phosphate dosing was stopped (Van Droogenbroeck and Laudelout 1967).
After two weeks, the effect of phosphorus on nitrification activity became very clear
(Figure 2.2). Specifically, only about 10% of the ammonia was oxidized when there was no
phosphorus dosed. As phosphorus concentration was increased to 5 and 10 ppb, ammonia
increased to 40% and 50%, respectively (Figure 2.2). The highest percentage of ammonia
oxidiation was observed when there was 1000 ppb phosphate. Consistent with trends in ammonia
loss, nitrifier BART testing indicated that nitrifiers at 1000 ppb-P (100000 cfu/ml) were two orders
of magnitude higher than that at no and 5 ppb phosphate (1000 cfu/ml).
Zinc
Zinc at high levels can cause toxicity to nitrifiers, possibly by precipitating phosphorus
nutrients (Harper et al. 1996). In a study with a full scale drinking water treatment plant, no
nitrification was expected when 0.5 ppm zinc were added (Bott 2005). In this study, while the
effect was weak, in PVC pipes dosing of 1000 ppb zinc significantly decreased ammonia loss %
and nitrifier MPN (Figure 2.3).
Other Trace Nutrients Tested
For all the other elements investigated in this study, including magnesium, potassium and
calcium, no significant ammonia loss and nitrifier MPN difference was observed at different
levels, although high levels of magnesium (50 ppm) seem to slightly decrease nitrification
compared to lower levels (p = 0.006). Eliminating these trace nutrients from the synthesized water
also had no effect on ammonia loss in eight week experiments.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 2: Effect of Ambient Nutrients and Metals on Nitrification in Premise Plumbing | 15
Direct and Indirect Effects of Alkalinity
Direct Effect
A large amount of alkalinity is consumed during the oxidation of ammonia to nitrate.
Specifically, 8.62 mg HCO3- (14 ppm as CaCO3) is consumed for every mg NH4+-N oxidized
(Grady et al. 1999). If alkalinity is consumed in wastewater applications, which are open systems,
nitrification activity can be severely impacted. It was previously believed that this limitation arose
due to acidic pH (pH < 6) inhibiting nitrification, but it was recently suggested that the loss of
alkalinity corresponded with a virtual complete loss of inorganic carbon in the open systems under
investigation. That is, nitrifiers were demonstrated to remain active even at pH 3.2 in a closed
system (Tarre and Green 2004).
In this study, the role of alkalinity was examined in PVC pipes that were closed to the
atmosphere. Higher alkalinity was clearly beneficial to nitrifier growth, as indicated by higher
MPN and ammonia loss (Figure 2.4). But nitrification activity was not completely stopped when
there was no alkalinity added, even if the only source was a small 2-3 ppm as CaCO3 attributable to
CO2 contamination from NaOH used to adjust pH. However, in later experiments with lead pipes,
in which fresh caustic without any significant CO2 contamination was tested, nitrification activity
was immediately stopped when inorganic carbon was reduced to 0 ppm (Figure 2.6). The key point
is that nitrifiers in closed potable water systems only require very low levels of alkalinity (< 5
mg/L as CaCO3) to fulfill their need for inorganic carbon.
Indirect Effect of Alkalinity on Copper and Zinc Toxicity
Low alkalinity did have an important indirect effect on the levels of copper and zinc that
where toxic to nitrifiers. For example, at 10 ppm alkalinity, ammonia loss was significantly lower
when copper levels were >100 ppb, but no such inhibiting effect was observed for the same level
of copper at 500 ppm alkalinity (Figure 2.5). This might be expected given that the level of free
Cu2+ can be reduced by complexation or precipitation with carbonate species (Braam and
Klapwijk 1981). Indeed, solubility models (AWWARF and DVGW-TZW 1996; Edwards et al.
1996) predict much less free Cu2+ at 500 mg/L alkalinity than with 10 ppm alkalinity (Table 2.2). It
is quite possible, then, that the high alkalinity (400-450 mg/L) in Willmar, MN, was a key reason
that nitrification was believed to occur in copper pipes, whereas in our research at 100 ppm
alkalinity or lower, no nitrification activity could be established in copper pipe even after 2.5
years. Similar impacts and mechanisms are likely to explain the mitigating effect of alkalinity on
zinc toxicity to nitrifiers.
Indirect Effect of Alkalinity on Phosphate Limitations
Alkalinity had no effect on phosphorus limitations in PVC pipes. That is, phosphate
limitation was observed at both low (10 ppm) and high (500 ppm) alkalinity. But in lead pipes, at
100 ppm alkalinity, complete nitrification (100% ammonia loss) occurred at all three phosphate
levels (5, 60 and 1000 ppb-Figure 2.5), indicating that nitrification was not limited by the low
levels of phosphate. When alkalinity was reduced to 30 and 15 ppm, ammonia loss in the pipes
with 5 and 60 ppb phosphate was reduced down to < 60%, while 100% ammonia loss was still
occurring in the pipes with 1000 ppb-P.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
16 | Effect of Nitrification on Corrosion in the Distribution System
There are two possible explanations for this effect. First, it is important to note that there
was substantial uptake (> 90%) of the phosphorus by the lead material and the higher alkalinity
might tend to reduce the uptake of phosphate from the water by lead corrosion. Hence, while
phosphate limitation is obvious in lead pipes at 5 and 60 ppb-P (Figure 2.6), the concentrations will
not directly apply to other systems such as PVC, for which the pipe does not uptake phosphate
from the water and decrease its bioavailability. Indeed, a parallel test with PVC showed no
phosphate limitation even with 5 ppb phosphate dosed. Second, lower alkalinity and lower
phosphate might induce a “dual limitation” to growth rates. Further research is needed to examine
these hypotheses explicitly.
Nutrient Release and Ammonia Cycling from Different Pipe Materials
Pipe materials can directly release or modify the nutrient levels available to nitrifiers
(Zhang et al. 2009a). For example, earlier studies have demonstrated that P-contaminants in iron
metal can release to water, and increase bioavailable phosphorus for bacterial growth (Morton et
al. 2005). In this study, soft copper pipe was confirmed to contain 0.05% (weight %) phosphorus,
and 20-40 ppb phosphorus was detected in water held in soft copper pipe after 96 hours stagnation.
Phosphorus released from these pipes can support nitrifier growth if no phosphorus is available in
bulk water. In contrast, no phosphorus is released from hard copper and PVC pipes, although
earlier studies have indicated that reduced phosphorus (i.e., phosphite) can be released from
polyethylene pipes (Lehtola et al. 2004).
Concrete materials can also leach essential nutrients like phosphorus, magnesium, calcium
and potassium (Table 2.3). But concrete can also reduce the levels of toxic elements like copper
and zinc (Table 2.3). However, due to the leaching of lime from concrete and resulted high pH
(initially pH = 11), nitrifier growth in the pipes with concrete did not start until the pH was dropped
down to below 9.3. However after lower pHs and stable nitrification was established, the leaching
of nutrients for glass pipes with added concrete, could completely eliminate nutrient limitations
observed in parallel tests with glass pipe alone. These observations can help explain why in some
studies lower nitrifiers and HPC were associated with concrete pipe materials (Steward and Lieu
1997), while in other studies, more heterotrophic bacterial counts are observed with concrete lined
pipes (LePuil et al. 2003). It’s very likely that higher pHs inhibited nitrifier growth in the former
study, whereas availability of nutrients from concrete caused higher growth in the latter study.
Iron and lead pipes are predicted to abiotically convert nitrate to ammonia (Edwards and
Dudi 2004; McIntyre and Mercer 1993; Zhang and Edwards 2007; Zhang et al. 2009a). In this
study, at all pH and oxygen conditions, iron consistently converted 0.5-1.2 mg/L ammonia from
nitrite and nitrate. The amount of ammonia produced was relatively small compared to the nitrite
and nitrate added (10 ppm), but this was not surprising considering both the LeChetalier's principle
and the small surface area to volume ratio (1: 2000) used. A larger surface area to volume ratio
would increase the extent of the reaction (Kielemoes et al. 2000). For lead and zinc, ammonia was
only observed at pH 6, aerobic condition: up to 1.3 mg/L ammonia was converted by lead and up to
0.5 mg/L ammonia was converted by zinc. Furthermore, ammonia was not observed after four
weeks’ testing with lead and zinc, possibly due to the formation of oxide layer on the metal surface
which can prevent further reaction between nitrite, nitrate and lead and zinc. This mechanism
might eventually stop the ammonia production from iron if the test were conducted long enough.
Nevertherless, the overall results indicate that under some circumstances, metal (iron, lead and
zinc) can clearly generate ammonia from nitrite and nitrate at drinking water conditions. If these
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 2: Effect of Ambient Nutrients and Metals on Nitrification in Premise Plumbing | 17
reactions were coupled with nitrification, very significant nitrification and pipe corrosion
problems might occur even with low levels of free ammonia (Edwards and Dudi 2004).
Overall, iron and lead pipes are beneficial to nitrification versus inert materials, whereas
copper pipe can completely stop nitrification due to disinfection properties of copper. But in other
situations soft copper (not hard) could supply a key trace nutrient (phosphorus) (Table 2.4).
Concrete materials can be either beneficial or detrimental to nitrification occurrence depending on
the circumstances (Table 2.4).
Effect of Organic Carbon
Previous modeling efforts (Zhang et al. 2009a) at 2 mg/L ammonia, determined that
organic carbon levels need to be very high (> 10 mg/L) in order to significantly reduce nitrification
by competition with nitrifiers. The expectation that organic carbon levels typically present in
drinking water (0-1 mg/L) would not affect nitrification, was tested in experiments with 60 and
1000 ppb-P. The nitrifier MPN values with or without organic carbon were similar (p ≥ 0.23); but
at 5 ppb-P, nitrifier MPN values in water with added TOC were slightly lower (Figure 2.7) than
without TOC (p = 0.03). This suggests that while heterotrophs are unlikely to outcompete nitrifiers
for ammonia, they might be out-competed by heterotrophs for phosphorus. As expected, at all
three phosphate levels, nitrifier MPN values without added ammonia were significantly lower than
that with ammonia (Figure 2.7).
In terms of heterotrophic bacterial growth, on average, there were 105 cfu/ml heterotrophic
bacterial counts even without added dissolved organic carbon to the water (organic free water was
used in the study). So it is clear that heterotrophs can grow satisfactorily on organic products of
autotrophic nitrification, consistent with expectations based on earlier studies in wastewater
systems (Rittmann et al. 1994). Compared to the pipes without added organic carbon, the addition
of organic (as glucose) up to 1 mg/L either slightly increased HPC or had no effect.
CONCLUSIONS
Nitrifying Bacterial growth was demonstrated to be affected by different nutrient levels,
specifically:
Copper > 100 ppb and zinc at 1000 ppb significantly reduced nitrifier activity, but this
effect was much more significant at low alkalinity (10 mg/L) and was not observed at higher
alkalinity (500 mg/L)
Nitrification activity was limited when no phosphorus was added in PVC pipes. Higher
levels of phosphorus increased nitrifier activity. The effect of phosphorus was not affected by
alkalinity levels in PVC pipes, but in lead pipes, phosphate limitation was only observed at low
alkalinity levels.
Pipe materials can release necessary nutrients (copper, phosphorus, magnesium and
potassium from copper and concrete pipe), reduce toxicity (i.e., concrete removed Zn2+ and Cu2+
from the water), and regenerate ammonia for nitrification (iron, lead and galvanized iron).
Adding organic carbon (up to 1 mg/L) did not significantly reduce nitrifier MPN values.
The exception was at a level of 5 ppb phosphate, which indicates that nitrifiers might be
out-competed by heterotrophs for low levels of phosphorus. Very high levels of heterotrophic
bacterial growth can be completely sustained by the organic carbon produced during nitrification.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
18 | Effect of Nitrification on Corrosion in the Distribution System
FIGURES AND TABLES
Figure 2.1 Average ammonia loss % and nitrifier MPN at different copper concentrations
ammonia loss %
Note: Alkalinity-10 mg/L. For Ammonia loss %, data reported were the average of 8
measurements; each measurement was taken 7 hours after water change. MPN measurement was
done at the end of the test. Error bars indicate 95% confidence interval.
0
5
10
50
1000
90
80
70
60
50
40
30
20
10
0
8/20/08 8/30/08 9/9/08 9/19/08 9/29/08 10/9/08
Time, day
Figure 2.2 Average ammonia loss % at different phosphorus concentrations
Note: Alkalinity-10 mg/L. Data reported were the average of five duplicates measured
after 7 hours, and error bars are 95% confidence levels.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 2: Effect of Ambient Nutrients and Metals on Nitrification in Premise Plumbing | 19
Figure 2.3 Average ammonia loss % and nitrifier MPN at different zinc concentrations
Note: Alkalinity-10 mg/L. For Ammonia loss %, data reported were the average of 8
measurements; each measurement was taken 4 hours after water change. MPN measurement was
done at the end of the test. Error bars indicate 95% confidence interval.
Figure 2.4 Average ammonia loss % and MPN values at different alkalinity added
Note: For Ammonia loss %, data reported were the average of 9 measurements; each
measurement was taken 4 hours after water change. MPN measurement was done at the end of the
test. Error bars indicate 95% confidence interval. “0” added alkalinity actually had a trace 2-3
mg/L inorganic carbon from contamination of NaOH used to adjust pH.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
20 | Effect of Nitrification on Corrosion in the Distribution System
Figure 2.5 Effect of copper and zinc concentrations on ammonia loss at two different
alkalinities
Note: Data reported for each alkalinity is the average of at least four measurements. At
very high alkalinity zinc and copper toxicity was eliminated.
Figure 2.6 Average ammonia loss % at different alkalinity and phosphate levels in lead pipes
Note: data reported were the average of three replicates; each measurement was taken 3.5
days after water change.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 2: Effect of Ambient Nutrients and Metals on Nitrification in Premise Plumbing | 21
T
Figure 2.7 Nitrifier MPN at 5 ppb phosphorus
Table 2.1
Nutrient levels investigated
Copper Zinc Phosphorus Magnesiu Calcium Potassium Alkalinity
ppb
ppb
ppb
m ppm
ppb
ppb
ppm
0
0
0
0
0
0
0
5
2
5
0.05
50
50
10
20
20
10
0.5
300
1000
50
100
100
50
5
1000
10000
100
200
500
100
25
4900
500
500 1000
1000
50
Note: for each nutrient level tested, there are 5 duplicate pipes, so there are 42 ×5 = 210 pipes.
Table 2.2
Illustrative prediction of free Cu2+ at two different alkalinity levels for a range of pHs
pH
Alkalinity, mg/L
5
Total
20
Cu
100
added,
200
ppb
500
6
500
0.9
3.6
18.1
36.2
90.9
7
10
4.4
17.6
88.3
176.0
440.3
500
0.10
0.38
1.9
3.8
6.0
8
9
10
500
10
500
10
1.3 0.008 0.023 0.000 0.000
5.1 0.026 0.092 0.001 0.001
25.5 0.13 0.46 0.003 0.005
50.9 0.17 0.92 0.006 0.010
60.0 0.17 1.73 0.006 0.024 ©2010 Water Research Foundation. ALL RIGHTS RESERVED
22 | Effect of Nitrification on Corrosion in the Distribution System
Table 2.3
Change in nutrient level by leaching from concrete materials to water
Concentration Actual concentration
with concrete, ppb
added, ppb
P
0
233
Mg
0
1910
Ca
0
9845
K
0
2588
100
6.5
Cu
200
14.2
500
66.4
100
9.1
Zn
500
52.2
Concrete added essential trace nutrients such as P, Mg, Ca and K, and removed toxic
constituents including zinc and copper.
Table 2.4
Predicted and observed effects of different pipe materials on nitrification
Predicted
Water Quality Change Impact on Nitrification
Iron
Serve as source of Fe,
P; convert nitrate to
ammonia; destruct
disinfectatn; good
surface for attachment
Concrete
Nutrients released,
destroys chloramine
and creates free
ammonia, higher pH
and alkalinity
Copper
Lead
Beneficial to
nitrification
Destruct disinfectant
fast, convert nitrite
and nitrate to
ammonia
Very high rate of
nitrification observed
High pH inhibit
nitrification
High pH inhibited
Release necessary
nitrification
nutrient (P, Mg, Ca
and K), absorb toxic
Release nutrient and nutrient (Cu and Zn), Modification of nutrient
destroy chloramine,
stopped nutrient
increase pH (up to
beneficial to nitrification 11) and alkalinity
limitation and toxicity
Release copper
copper can be toxic to
nitrifiers
Convert nitrate to
ammonia, release Pb
If convert nitrate to
ammoia, beneficial to
nitrification; but lead
might be toxic to
nitrifiers
1: (Zhang et al. 2008a)
Confirmed in this Project
Water Quality
Impact on Nitrification
Release copper
Nitrification can not be
(>200 ppb) and
established on copper
phosphorus (20-40
pipe1
ppb)
Convert nitrate to
ammonia under some
up to 8000 lead not toxic
circumstances, up to
to nitrifiers1
8000 ppb lead
released
©2010 Water Research Foundation. ALL RIGHTS RESERVED
CHAPTER 3
EFFECT OF ORGANIC AND AMMONIA LEVELS ON NITRIFICATION
AND COPPER RELEASE
Mohammad Rahman, Ann Camper
This chapter discusses the impacts of different organic and ammonia levels on nitrification
and copper release. The bench-scale study was conducted using nitrifiers naturally occurring in
Bozeman tap water.
INTRODUCTION
Nitrification and Its Effect on pH and Alkalinity
Nitrification is the oxidation of ammonia to nitrite and nitrate. Nitrification is primarily
accomplished by autotrophic nitrifying bacteria. The first step is the oxidation of ammonia to
nitrite by ammonia oxidizing bacteria (AOB) according to equation 3.1. The second step is the
oxidation of nitrite to nitrate by nitrite oxidizing bacteria (NOB) (equation 3.2). Overall the
nitrification process can be represented chemically according to equation 3.3.
NH3 + O2
NO2- + H2O
NH3 + O2 + H2O
NO2- + 3H+ + 2eNO3- + 2H+ + 2eNO3- + 5H+ + 4e-
(3.1)
(3.2)
(3.3)
Based on the above equation, it is expected that nitrification can decrease pH and
alkalinity. Theoretically 7.14 mg/L of alkalinity as CaCO3 is consumed for the oxidation of 1 mg/L
of NH3-N to NO3-N (equation 3.3). Earlier studies have reported alkalinity consumption in the
range of 6.2 to 7.4 mg/L as CaCO3 (Parker 1975) which is less than theoretical.
Nitrification can impact corrosion by decreasing pH, alkalinity and dissolved oxygen (Zhang
et al. 2009a). It is expected that lower pH can increase copper release to the water (Schock et al.
1995), while lower alkalinity might reduce copper corrosion (Edwards et al. 1996).
Effect of Organic Carbon and Ammonia Levels on Microbial Populations and the Interaction
between Autotrophic and Heterotrophic Populations
Carbon and nitrogen are the major nutrients necessary for microbial growth. Autotrophic
bacteria like nitrifiers use inorganic carbon, while heterotrophic bacteria use organic carbon as their
carbon source. However, besides the typical autotrophic nitrifiers, heterotrophic bacteria and fungi
can also carry out nitrification, although these organisms nitrify at a lower rate (Watson et al. 1989).
Organic carbon is generally considered to have a negative effect on nitrification (Zhang et al.
2009a), although there are many studies reporting nitrifiers incorporate selected organic compounds
(Clark and Schmidt 1966; Clark and Schmidt 1967; Hommes et al. 2003; Pan and Umbreit 1972).
Elevated organic carbon may also stimulate heterotrophs which might out-compete nitrifiers for
surfaces, dissolved oxygen, ammonia and other nutrients. On the other hand, when no dissolved
organic carbon is present in the water, heterotrophs are completely dependent on the lysis and
23
©2010 Water Research Foundation. ALL RIGHTS RESERVED
24 | Effect of Nitrification on Corrosion in the Distribution System
extracellular products of nitrifiers as a source of organic carbon (Rittmann et al. 1994). Wolfe et al.
(1990) reported a strong correlation between heterotrophic and nitrifying populations in water
distribution systems.
Different nitrogen compounds (NH3 or NO3-) or natural organic matter (NOM) in the water
can also directly affect corrosion. Both ammonia and NOM might increase copper solubility by
complexing with cupric ions (Korshin et al. 1996; Schock et al. 1995). However, NOM can also
reduce copper corrosion by fueling microbial removal of oxygen, causing re-deposition of Cu(I) to
the pipe wall, and sorption of soluble organic matter onto pipe surfaces during stagnation, which
decreases the solution's complexation capacity for copper (Edwards and Boulay 2001). Campbell
(1950) also reported that NOM can inhibit pitting corrosion and the presence of NOM can produce
a thick protective cuprous oxide layer (Campbell 1950). Nitrate (NO3-) might serve as electron
acceptor and increase copper corrosion when oxygen is not available. Low concentration of NO3can decrease pitting corrosion by modifying the scale characteristics (Edwards et al. 1994), but
higher concentrations (>40 mg/L) increases pitting (AWWARF and DVGW-TZW 1996).
In this study, the effect of different organic carbon and ammonia levels on nitrification was
studied in simulated household plumbing systems with two types of materials (i.e. copper and
PVC). Experiments were conducted to understand the effect of different levels of TOC and NH3 on
autotrophic nitrifiers (AOB/NOB) and heterotrophic populations in simulated household systems.
The pH and alkalinity change due to nitrification was also examined at different carbon and
nitrogen concentrations in these two different pipe materials.
MATERIALS AND METHODS
Reactor Setup
To simulate a domestic plumbing system the commonly used CDC (Goeres et al. 2005)
reactor was modified (Figure 3.1). These modified reactor’s coupons, bottom plate and stirring
blades have the same surface to volume ratio as that of a six foot long ¾” diameter domestic
copper plumbing pipe. The rotational speed of the blade inside these reactors was 300 rpm. This
speed was chosen as it creates 3fps velocity in bulk water which can be found in domestic water
lines. Volume of the reactors is 120 ml. Two types of materials were tested, copper and PVC. All
coupons were washed with 0.1N NaOH three times to remove any biological materials from their
surface prior to use.
Operation Scheme
To simulate periods of stagnation in home plumbing the reactors were flushed with
peristaltic pumps for five minutes and then the water inside the reactors remained stagnant for
eight hours. The feed pumps and stirplates were on two different timers, which controlled the
power supply. The timers were offset from each other by one minute with the stirplate starting
before the pumps. At the end of five minutes the stirplates stopped, followed by pumps. This cycle
was repeated three times a day. As a result, fresh influent mixed with the stagnant bulk water and
excess water spilled through the effluent port. Because there is mixing, the effluent water is always
diluted with fresh influent feed, which prevents the effluent NH3 concentration from becoming
zero, even though the bulk NH3 concentration (the concentration in the reactor) is zero due to
nitrification. Because sampling occurred in both the bulk and the effluent, a mass balance of the
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 3: Effect of Organic and Ammonia Levels on Nitrification and Copper Release | 25
reactor was conducted to determine the relation between effluent NH3 concentration and NH3
utilization in these reactors. This approach was used to calculate the % NH3 utilization and the %
of ammonia present as NO2, and NO3. The approach is described in detail elsewhere (Rahman
2008).
Water Preparation
Typical setup of these reactors is shown in Figure 3.2. The ratio of flow is RO: humics:
BAC= 50:5:1, so these reactors are mostly supplied with RO water. Soil derived humics
substances (Elliot Silt Loam) was used as carbon source and (NH4)2SO4 was used as a nitrogen
source. Humics were chosen because they occur naturally in water and previous studies showed
that biofilm bacteria can use and grow using humic materials (Camper 2004; Volk et al. 1997).
Initially influent carbon concentration as humics was 4 ppm for all reactors. Lab grade chemicals
were added to reverse osmosis (RO) water to give it target alkalinity of 35 mg/L as CaCO3. The
final concentrations of these chemicals are: MgSO4 (39.6 mg/L), CaSO4 •2H2O (25 mg/L),
Al2(SO4)3•18H2O (0.62 mg/L), CaCl2 •2H2O (20.8 mg/L), Na2SiO3 •9H2O (26 mg/L).
Bozeman tap water was flowed through a biologically activated carbon (BAC) column to
remove any residual chlorine. It should be noted that Bozeman water comes from surface water
source, does not contain ammonia, and chlorine is used as the final disinfectant. BAC water was
pumped to the reactors to ensure a supply of indigenous bacteria. These organisms were the only
inoculum supplied to the reactors. Humics were supplied using a separate pump.
Humics preparation
Fifty grams of Elliot silt loam soil (International Humic Substances Society) was added to
500 ml of 0.1 N NaOH and mixed for 48 hours. This solution was then centrifuged at 10,000 X g
for 20 minutes. After centrifugation the supernatant was collected in carbon free glassware (made
by baking at 3900 C for five hours) and stored at 40 C in the dark. Total organic carbon content of
the humics was measured using a Dohrman DC-80® and subsequently diluted to the appropriate
concentration using the RO water feeding the reactors.
Experimental Approach
Modified CDC reactors equipped with PVC and copper coupons were used. There are
three duplicate reactors for both PVC and copper. All of these reactors had been running for almost
two years and consistently converted all the added ammonia to nitrite and nitrate. Initially all
nitrifying reactors were operated at 0.71 ppm NH3-N and 4 ppm TOC (humics) in the influent.
Influent water pH and alkalinity were always kept at 8.15 and 35 mg/L as CaCO3 respectively.
NH3-N and TOC content of the influent of two reactors from two sets was subsequently raised as
shown in Table 3.1. When the reactors adjusted to the changed conditions as evidenced by stable
nitrification, a biofilm sample was collected for analysis. Regular measurement of bulk water
alkalinity and pH was done at the end of stagnation period for the entire period. When stable
nitrification was observed, regular measurement of bulk water NH3 concentration was done to
estimate the NH3 utilization pattern in those reactors.
Effluent water was collected and tested three times (Monday, Wednesday and Friday)
every week. A bulk water sample from the reactors was taken once every week for similar
©2010 Water Research Foundation. ALL RIGHTS RESERVED
26 | Effect of Nitrification on Corrosion in the Distribution System
measurements. Effluent water concentrations were converted to bulk values using the model
(Rahman 2008).
Analytical Methods
Chemical Analysis
Both total and dissolved copper were measured according to Standard Methods (Clesceri et
al. 1998). Dissolved copper is operationally defined as the portion that passes through a 0.45µm
pore size syringe filter. Measurements were done using a portable HACH 2000
spectrophotometer.
Ammonia and nitrite was measured using a HACH 2000 spectrophotometer according to
standard method 4500 NH3- and 4500 NO2-. Nitrate was measured using a Dionex® ion
chromatography system with a CD20 conductivity detector and GP40 gradient pump unit. Water
was filtered through a sterilized 0.2 µm polycarbonate filter and 5ml BD® syringe to remove any
bacteria or suspended particles. The filtered sample was collected in a sterilized 15 ml Falcon®
tube and stored in the refrigerator at 40 C. Stored samples were measured within two weeks of
collection. pH was measured using an Accumet AP-63P pH probe according to Standard Method
4500 H+ B. Alkalinity was measured according to Standard Methods 2320 (Clesceri et al. 1998).
Total Organic Carbon (TOC) was measured using a Dohrman DC-80® carbon analyzer, with
potassium hydrogen phthalate as standard and potassium persulfate as the oxidizing agent. Weight
loss was calculated for copper reactors from the difference of initial dry weight and final weight
after removing the scale.
Microbiological Analysis
Heterotrophic plate counts (HPC) of the water samples and the biofilms were done
according to Standard Methods (Clesceri et al. 1998) 9215A using R2A agar plates.
Biofilm Analysis
Biofilm was collected at different time points in each experiment. Autoclaved reverse
osmosis (RO) water was filtered through a sterilized 0.2µm polycarbonate filter to remove any
foreign DNA. Filtered DNA free water was placed in a DNA free glass tray (baked at 3900C for 5
hours). One coupon was removed from the reactor and placed in the glass tray containing the
water. The coupon was then scraped using an autoclaved rubber policeman inside a laminar flow
hood. After scraping, the biomass with water was poured in a sterilized 50 ml Falcon® tube, which
was then homogenized with a homogenizer (Biohomogenizer® Model M133/12810, ESGE®) for
30 sec. From the homogenized biomass, samples were taken for MPN and HPC analysis. The
remaining biomass was used for DNA extraction and community analysis.
DNA extraction
Homogenized biomass was collected on a 0.2 µm polycarbonate filter using a three
channel manifold (Pall® Life Science) with filter funnels. DNA extraction of the collected biofilm
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 3: Effect of Organic and Ammonia Levels on Nitrification and Copper Release | 27
sample was done using a Fast DNA® SPIN Kit for soil (Q-BIOgene catalog #6560-200).
Collected DNA was stored in a -300 C freezer.
PCR (Polymerase Chain Reaction)
GoTag Green® master mix from Promega Inc. was used for amplifying the extracted DNA
through PCR using an Eppendorf Mastercycler®. Due to the inhibitory effect of humics present
and the low quantity of biomass, a two stage PCR was performed. In the first stage a 50µl reaction
volume was used which contains 25µl master mix, 10 pM of the universal primers 1070F and
1392R, 1µl DNA suspension and 20 µl of water. The amplification process involved initial
denaturation at 940C followed by 15 cycles of 30 second denaturation at 940C, 45 second
annealing at 520 C and 2 minutes of extension at 720 C with a final 5 minutes extension at 720C. In
the second stage, 5 µl of the product from the first stage was used as template. In the second stage
a similar reaction composition was used except 1392+GC was used instead of the 1392 primer and
20 amplification cycles were used. PCR products were evaluated by an agarose gel
electrophoresis. Negative controls without template addition were treated identically through the
PCR and evaluated by agarose gels to confirm the absence of contaminant DNA.
DGGE (Denaturing Gradient Gel Elecrophoresis)
DGGE was performed at 600 C with a D-Code Universal Mutation Detection System
(Bio-Rad Laboratories). Eight and twelve percent (w/v) acrylamide gels with denaturant gradients
of 40 to 70% were used for analyzing fragments amplified using 1070 and 1392+GC. A 25 ml
volume denaturing gel was poured and allowed to polymerize prior to pouring of a zero percent
denaturant stacking gel for the loading wells. Sixteen hours of electrophoresis were performed for
the gels at 60 V. After electrophoresis the gels were subsequently stained with SYBR Green I
(Cambrex Bio Science). Images of the gel were obtained using a FluorChem 8800® Imaging
system and AlphaEase FC® software (Alpha Innotech). Composition of all the reagents and
conditions used for DGGE are presented elsewhere (Rahman 2008).
MPN (Most Probable Number
AOB and NOB populations were enumerated using the most probable number (MPN)
technique (Lipponen et al. 2002) using Costar® Clear-Bottom 96 well microtiter plates. The
mineral medium used for AOB contained per liter: (NH4)2SO4, 330 mg; KH2PO4, 100 mg; MgSO4
7H2O 40 mg; CaCl2, 15 mg and 1 ml of a trace-element solution. The trace element solution
contained per liter: Na2EDTA, 4292 mg; FeCl2 4H2O, 1988 mg; MnCl2 H2O, 99 mg; NiCl2
6H2O,24 mg; CoCl2 6H2O, 24 mg; CuCl2 2H2O, 17 mg; ZnCl2, 68 mg; Na2MoO4 2H2O, 24 mg and
H3BO3, 62 mg. Bromothymol blue (5ml/ l of 0.04% solution in water) was added as a pH indicator.
The pH was adjusted to 8 using 1M NaOH before autoclaving at 1100 C for 15 min. The NOB
medium has the same composition except that it did not contain (NH4)SO4 and bromothymol blue,
and was supplemented with 34.5 mg/ l of NaNO2. The pH was adjusted to 6.5 with 1M NaOH
before autoclaving at 1100 C for 15 min. 175µl of media was poured into each of the 96 wells of the
plate. An equal amount of sample was then inoculated into the first column of wells in the plate
using a multi channel pipette. The content of the first column of wells was carefully mixed, and
175 µl of inoculated media was transferred to the next column of wells, and this process continued.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
28 | Effect of Nitrification on Corrosion in the Distribution System
After inoculation the microtiter plates were sealed with polyester tape to prevent evaporation and
incubated for 9 weeks at 200C in the dark. After the incubation period AOB presence was
determined by detecting the presence of nitrite or nitrate in the medium by adding 40 µl of 0.2%
diphenylamine in H2SO4 in the well. This reagent reacts with nitrite or nitrate and forms a blue
color. The absorbance of the blue color was measured at 630 nm with a microplate reader (EL808
ultra microplate reader BioTek instruments®). The blue color indicated nitrite or nitrate had
formed and the well was scored as positive.
Griess Ilosvay reagent was used to detect NOB activity in the samples. This reagent was
made by mixing three separate solutions. In the first solution exactly 0.6 gm of sulfanilic acid was
dissolved in 70 ml of hot distilled water. After cooling to room temperature, 20 ml of concentrated
HCl and 10 ml of distilled water was added. The second solution was made by dissolving 0.6 gm of
α-napthylamine in 20 ml of distilled water containing 1 ml of concentrated HCl and then diluting it
to 100 ml with H2O. The third solution is 16.4% (w/V) of CH3COONa. 3H2O in water. All the
solutions were made separately in dark bottles and kept in the refrigerator. Equal volumes of these
three solutions were mixed and 40 µl of the mixture was added to each well after the incubation
period. In the presence of nitrite, the Griess Ilosvay reagent produced a red color within five
minutes. The absorbance of the well was measured after five minutes at 540 nm with the
microplate reader. If nitrite in the well was detected it was scored as negative. The MPNs were
calculated according to Rowe et al. (1977) and finally expressed as cells/ml of sample (Rowe et al.
1977).
RESULTS AND DISCUSSION
Ammonia Utilization in PVC and Copper Reactors
Initially, at 0.71 ppm NH3-N and 4 ppm TOC, 100% ammonia utilization was occurring in
both PVC and copper reactors for over two years (Rahman 2008). Then for the reactors with
increased ammonia (2.13 ppm), complete ammonia utilization (> 95%) occurred in PVC reactors
two weeks after the increase of NH3, while it did not occur in copper reactors until two months
after the change. The ammonia conversion rate was also slower in copper reactors than that in PVC
reactors over 8 hour stagnation time (Figure 3.3). However, the nitrifying biofilm MPN cell
density was not significantly different between the copper and PVC reactors (Figure 3.4), possibly
because the underestimation of nitrifying community by this method (Both et al. 1990; Fairey et al.
2004; Rennie and Schmidt 1977).
The slightly lower ammonia utilization and longer adaptation time in copper reactors might
be due to the toxicity of copper to nitrifiers. Lower level of nitrifier activity in copper pipe and
thinner biofilms on copper surfaces than plastic have been reported (Zhang et al. 2009a; Schwartz
et al. 1998).
Effect of ammonia and TOC levels on Nitrification and Copper
At all ammonia and TOC levels, 85~100% ammonia was utilized in PVC reactors and
60-90% was utilized in copper reactors, indicating that ammonia and TOC levels did not affect
nitrification efficiency. For bulk water nitrifier MPN, increase of ammonia increased AOB and
NOB MPN by two orders of magnitude in PVC reactors and one order of magnitude in copper
reactors (Figure 3.5). This is not surprising, considering that higher ammonia would support more
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 3: Effect of Organic and Ammonia Levels on Nitrification and Copper Release | 29
nitrifier growth. Increase of TOC did not affect nitrifier MPN (Figure 3.5), consistent with the
findings of Chapter 2. For bulk water HPC, although higher ammonia supported more nitrifier
growth, heterotrophic bacterial growth was not affected (Figure 3.6), indicating that the organic
produced from nitrifiers was not the major carbon source for heterotrophs. Consistent with
expectation, increase of TOC slightly increased HPC by an order of magnitude (Figure 3.6).
Biofilm HPC and nitrifier MPN was not affected by ammonia and TOC levels (Figure 3.4).
For all reactors the final bulk water pH is significantly lower (around 7~7.25) than the
influent pH (8.15) (Figure 3.7). A greater pH drop was observed with higher level of ammonia
(Figure 3.7), which is consistent with the prediction from reaction stoichiometry (Equation 3).
Increase of TOC did not affect the final pH (Figure 3.7). The majority of the alkalinity used during
nitrification is for pH adjustment (Grady et al. 1999). In this study, alkalinity destroyed in PVC
reactors was close to theoretical values, while in copper reactors, much less alkalinity was
destroyed (Figure 3.8). This is possibly because copper corrosion reaction can raise pH so that less
alkalinity is needed for pH adjustment than PVC reactors. In this study, lower alkalinity destruction
was observed with the increase of TOC and decrease of ammonia (Figure 3.8). This is consistent
with previous researchers who reported that higher alkalinity destruction was associated with lower
C/N ratio (Benninger and Sherrard 1978; Sherrard et al. 1980).
Increase of ammonia by three times significantly increased total and soluble copper by
40-50% (Figure 3.9), while increase of TOC by two times slightly decreased total and soluble
copper by 20% (Figure 3.9). This observation was consistent with earlier studies that ammonia can
increase copper level due to the strong complexion (Schock et al. 1995) and NOM can decrease
copper release (Edwards and Boulay 2001; Campbell 1950). But copper weight loss was not
affected by ammonia or TOC increase.
Microbial Community Composition
Molecular techniques (DGGE) also indicated that community composition was not
significantly different at different ammonia and TOC levels except for a few species
(Figure 3.10-marked by white circles). Different microbial species was also observed in PVC and
copper reactors, as indicated by molecular techniques-DGGE (Figure 3.10-bright white band in
copper reactors was absent in PVC reactors).
CONCLUSIONS
•
•
•
•
•
Higher nitrification rate and faster adaptation time for nitrifiers were observed with
PVC than copper
Increase of ammonia levels slightly increases the AOB/NOB population (1~2 logs)
and decreased pH more significantly, but it had no impact on HPC
Increase of TOC slightly increased the suspended HPC level but it had no impact
on AOB/NOB population
Using molecular techniques, different bacterial species were observed in PVC and
copper surfaces.
Alkalinity and pH decrease was decreased due to nitrification, but less alkalinity
was destroyed in copper than in PVC, possibly due to copper corrosion reaction can
also adjust pH; less alkalinity was destroyed when there is more ammonia and TOC
©2010 Water Research Foundation. ALL RIGHTS RESERVED
30 | Effect of Nitrification on Corrosion in the Distribution System
•
Increase of ammonia increased total and soluble copper, while increase of TOC
decreased total and soluble copper
FIGURES AND TABLES
Figure 3.1 Modified CDC reactor
Tap Water Humics RO
WATER
pH ~8.15
Pump Alk 35
mg/L
Pump Pump Figure 3.2 Typical setup for a reactor
Note: This figure illustrates the typical set-up for either PVC or copper CDC reactors. The
reactors are fed with reverse osmosis water adjusted to pH 8.15, a separate feed of humics to
provide organic carbon, and Bozeman tap water that runs through a biologically active carbon
filter to remove background organics and provide a consistent microbial load to the system. The
reactors are held on a stir plate and operated in a mode of 8 hr. stagnation and 5 min. of flow.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter
C
3: Efffect of Organiic and Ammon
nia Levels on Nitrification
N
a Copper Release
and
R
| 31
Figure 3.3
3 Bulk watter NH3 con
ncentration in differentt reactors during the eiight hours
stagnatioon
2XC-3XN: influent NH3N 2.13ppm
m and
TOC=8ppm
m;
2
AOB/NOB and HPC cell density(cells/cm )
1.0E+07
1.0E+06
1XC-3XN: influent NH3N 2.13ppm
m and
TOC=4ppm
m;
AO
OB-PVC
AO
OB-Copper
NO
OB-PVC
NO
OB-Copper
1.0E+05
1XC-1XN--control:
influent NH
H3-N
0.71ppm an
nd
TOC=4ppm
m
HP
PC-PVC
HP
PC-Copper
1.0E+04
1.0E+03
1.0E+02
2X C-3XN
1XC-3XN
N
1XC-1XN co
ontrol
Reactor ty
ype
Figure 3.4
3 Biofilm cell
c density in
i different reactors
©2010 Water Research Foundation. ALL RIGHTS RESERVED
32 | Effect of Nitrification on Corrosion in the Distribution System
10000
AOB-PVC
AOB-Copper
NOB-PVC
NOB-Copper
AOB/NOB(cells/ml)
1000
100
10
1
2X C-3XN
1XC-3XN
1XC-1XN-CONTROL
Figure 3.5 Average bulk water MPN for AOB and NOB from different reactors
1.0E+07
PVC
Copper
HPC(CFU/ml)
1.0E+06
1.0E+05
1.0E+04
1.0E+03
2X C-3XN
1XC-3XN
Figure 3.6 Average (n=12) HPC value for different reactors
©2010 Water Research Foundation. ALL RIGHTS RESERVED
1XC-1XN-CONTROL
Chapter 3: Effect of Organic and Ammonia Levels on Nitrification and Copper Release | 33
8.00
1XC-1X NH3
CONTROL
2XC- 3X NH3
2XC-3XN: influent NH3-N 2.13ppm and TOC=8ppm;
1XC-3XN: influent NH3-N 2.13ppm and TOC=4ppm;
1XC-1XN-control: influent NH3-N 0.71ppm and TOC=4ppm
1XC- 3X NH3
pH
7.50
7.00
6.50
6.00
PVC
Copper
Coupon/Reactror type
Figure 3.7 Average (n=12) pH of bulk water after eight hours of stagnation
Alkalinity(mg/L as CaCO3 ) used/NH3-N(mg/L)
oxidized
10.00
9.00
Theoretically 7.14
mg as CaCO3 will
be used for 1mg/L
NH3-N oxidation.
1XC-1X NH3 CONTROL
2XC- 3X NH3
1XC- 3X NH3
8.00
7.00
2XC-3XN: influent NH3-N
2.13ppm and TOC=8ppm;
6.00
1XC-3XN: influent NH3-N
2.13ppm and TOC=4ppm;
5.00
1XC-1XN-control: influent NH3
N 0.71ppm and TOC=4ppm
4.00
3.00
2.00
1.00
0.00
PVC
Copper
Coupon/Reactror type
Figure 3.8 Average (n=12) alkalinity used /mg of ammonia nitrogen oxidized
©2010 Water Research Foundation. ALL RIGHTS RESERVED
34 | Effect of Nitrification on Corrosion in the Distribution System
1.20
Effluent Total and Dissolved Copper(mg/L)
Total Copper
Dissolved Copper
1.00
0.80
0.60
0.40
0.20
0.00
2X C-3XN
1XC-3XN
1XC-1XN-CONTROL
Reactor type
Figure 3.9 Average effluent total and dissolved copper concentrations in different copper
reactors
Note: error bar showing 95% confidence interval; for student paired t-tests (n =12), p ≤10-5
PVC
2XC3XN
Copper
1XC3XN 1XC-1XN-control 2XC3XN
1XC3XN
1XC-1XN-control
Figure 3.10 DGGE profile of biofilm sampled from different reactors
Note: one band represent one species in theory
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 3: Effect of Organic and Ammonia Levels on Nitrification and Copper Release | 35
Table 3.1
Influent water quality for different reactors
Coupon type
PVC
Copper
Reactor
NH3-N(ppm)
1XC-1XN1 Control
0.71
4
2XC-3XN2
2.13
8
1XC-3XN3
2.13
4
1XC-1XN1 Control
0.71
4
2XC-3XN2
2.13
8
1XC-3XN3
2.13
4
©2010 Water Research Foundation. ALL RIGHTS RESERVED
TOC(ppm)
36 | Effect of Nitrification on Corrosion in the Distribution System
©2010 Water Research Foundation. ALL RIGHTS RESERVED
CHAPTER 4
NITRIFICATION IN PREMISE PLUMBING: ROLE OF PHOSPHATE, PH
AND PIPE CORROSION
Yan Zhang, Allian Griffin and Marc Edwards
This study focused on the interaction of phosphorus, pH, and pipe corrosion. The
hypothesis for this test was: the corrosion of copper based materials can prevent nitrification in
premise plumbing and reducing copper levels by increasing pH and using phosphate corrosion
inhibitor might help nitrification. On the other hand, lead materials are capable of converting
products of nitrification (nitrite and nitrate) back to ammonia, so that nitrification is enhanced.
Four different pipe materials were investigated for their ability to support nitrification:
PVC, lead, copper and brass. Complete nitrification was established in PVC and lead pipes within
two weeks, with lead pipes slightly enhanced nitrification relative to PVC. For brass pipes,
nitrification was not established until three months after the inoculation, while for copper pipes
(<1.5 years old), no significant nitrification activity could be established. Nitrification in PVC and
lead pipes was a weak function of pH over the range 6.5–8.5 and was insensitive to phosphate
concentrations of 5–1000 ppb. However, in brass pipes, higher nitrification activity and more
nitrifier growth were observed at higher pH and phosphate levels, indicating that phosphate and
pH can affect nitrification by affecting copper level and thus controlling its toxicity to nitrifiers.
This study was published in Environmental Science and Technology, 2008, 42 (12), 4280-4284
and permission is granted by ACS (American Chemical Society) to be included in the report. The
paper and its supporting materials are included in the Appendix B.
In the published paper, nitrifiers were quantified using traditional culturing technique-most
probable number (MPN), but these techniques have the limitations of low recovery efficiency and
long incubation time. So effort was made to develop protocols to monitor nitrifiers with molecular
techniques. However, due to the low nitrifier quantities in our samples (105-106 cfu/ml) and the
limitation of sample amount (< 300 ml for each condition), the protocol initially designated was
not able to detect DNA signal from these samples. So the main goal of the sample collection at the
time was to maximize nitrifier quantity, and less care was carried in sterile sample handling. After
this study was finished, the team successfully developed a reliable protocol for DNA extraction
and qPCR analyses for the nitrifier quantification for low nitrifier quantity samples, the specifics
of the protocol are presented in Appendix C. However, the protocol was not developed in time to
collect new samples, so old samples with a very high probability of cross contamination were used
for the analysis (Appendix C). And not surprisingly, the PCR results (Appendix C) were quite
contradictory to MPN and ammonia loss results (Appendix C). Also, the primers used for
quantification of ammonia oxidizers can also present false positives, by amplifying non-AOB
targets. It is impossible to separate false positives from the specific targets through qPCR analyses.
Though the protocol developed showed good efficiency of DNA extraction, it may be appropriate
to perform additional techniques such as denaturing gradient gel electrophoresis to ascertain the
identity of amplified targets through amplicon sequencing. Nonetheless, the developed protocol
provides a good starting point for future researchers interested in the quantification/identification
of microbial communities colonizing drinking water systems.
37
©2010 Water Research Foundation. ALL RIGHTS RESERVED
©2010 Water Research Foundation. ALL RIGHTS RESERVED
CHAPTER 5
LEAD CONTAMINATION OF POTABLE WATER DUE TO
NITRIFICATION
Yan Zhang, Allian Griffin1, Mohammad Rahman, Ann Camper, Helene Baribeau and Marc
Edwards
This study focused on the effect of nitrification on lead release. Nitrification has been
associated with corrosion problems including elevated lead contamination, but there has been no
research that directly confirmed this effect and its mechanism. So the goal of this study was to
understand nitrification effect on corrosion and metal release in typical premise plumbing
condition through a well-controlled laboratory study.
Results confirmed that nitrification can increase soluble lead release by reducing pH. The
magnitude of the pH effect depends on the initial alkalinity, extent of nitrification and associated
acid production. At 100 mg/L alkalinity as CaCO3, complete nitrification did not significantly
decrease pH (pH stayed > 7.5) or increase lead contamination from lead pipe, but at 15 mg/L
alkalinity, nitrification decreased the pH by 1.5 units (pH reduced to < 6.5) and increased soluble
lead contamination by 65 times. Lower pH values from nitrification also leached 45% more lead
and 81% more zinc from leaded brass connected to PVC pipes versus copper pipes. Particulate
lead leaching was important, but did not seem to be impacted by nitrification. Production of nitrite
and nitrate, or reductions in inorganic carbon and or dissolved oxygen via nitrification, did not
significantly influence lead leaching. This study was published in Environmental Science and
Technology, 2009, 46 (3), 1890-1895 and permission is granted by ACS (American Chemical
Society) to be included in the report. The paper and its supporting materials are included in the
Appendix D.
39
©2010 Water Research Foundation. ALL RIGHTS RESERVED
©2010 Water Research Foundation. ALL RIGHTS RESERVED
CHAPTER 6
ACCELERATED CHLORAMINE DECAY AND MICROBIAL GROWTH
RESULTING FROM NITRIFICATION IN PREMISE PLUMBING
Yan Zhang and Marc Edwards
The objective of this study is to rigorously examine the conceptual and anecdotal evidence
that chloramine is sometimes less persistent than free chlorine in certain situations encountered in
premise plumbing. This is important, if only to provide an understanding of possible “exceptions
to the rule” regarding enhanced control of pathogens that might result from chloramine use. It also
sheds light on potential indirect implications of nitrification as a public health concern.
INTRODUCTION
In the United States, utilities are increasingly using combined chlorine (chloramines) to
comply with regulations for disinfection by-products (Seidel et al. 2005). Accepted advantages of
chloramine include reduced formation of many regulated Disinfectant By-Products (DBP) and
greater persistence relative to free chlorine (Neden et al. 1992a; Norton and LeChevallier 1997).
Chloramines are often more effective than chlorine in reducing heterotrophic plate counts (HPC)
(Neden et al. 1992a).
In comparing the relative effectiveness of chlorine and chloramine for disinfection of
biofilms on pipe walls under continuous flow conditions found in distribution systems,
LeChevallier et., al. 1990 determined that both free chlorine and monochloramine were effective
for PVC, copper, and galvanized iron pipes (Lechevallier et al. 1990). However, biofilm growth on
iron pipe was better controlled by monochloramine than the same dose of free chlorine
(Lechevallier et al. 1990). This was attributed to a very high reaction rate between free chlorine
and iron corrosion products on the pipe surface, essentially rendering the disinfectant unavailable
for controlling biofilm bacteria. In contrast, monochloramine is less affected by reactions with iron
scale, and is therefore said to have greater “penetrating power” for microbial control (Lechevallier
et al. 1990; Lechevallier et al. 1993). In essence, the greater persistence of chloramine can
outweigh its reduced efficacy as a disinfectant versus the same dose (as Cl2) of free chlorine
(Wolfe et al. 1984).
If situations existed in which monochloramine was consumed at a faster rate than chlorine,
chlorine can be expected to outperform chloramine in control of biofilms and bacteria in the bulk
water. Anecdotally, at least one utility that converted to chloramine believed that chlorine had been
more persistent as a disinfectant than chloramine in their main distribution system in warm summer
months (Powell 2004). Nitrification of free ammonia, which occurs from chloramine decay, might be
expected to contribute to this type of situation (Norton and LeChevallier 1997; Wilczak et al. 1996;
Wolfe and Lieu 2001; Zhang et al. 2009a), since the nitrite produced from nitrification accelerates
decay of chloramine residuals (Harrington 2002; Valentine 1984). Because free chlorine cannot
co-exist with ammonia, nitrification is unlikely to occur if free chlorine were used as a secondary
disinfectant, providing a possible relative advantage for free chlorine. Indeed, a loss of chloramine
disinfectant during nitrification has frequently been associated with relatively high levels of HPC’s in
main water distribution systems (Powell 2004; Skadsen 1993; Wilczak et al. 1996; Wolfe and Lieu
2001; Zhang et al. 2009a).
41
©2010 Water Research Foundation. ALL RIGHTS RESERVED
42 | Effect of Nitrification on Corrosion in the Distribution System
To oversimplify the situation on iron pipe surfaces with nitrification, chlorine decay rates
are controlled mainly by the reactions with bulk water species and iron corrosion products
(equation 6.1), whereas chloramine decay is controlled by reactions with bulk water and nitrite
produced by nitrifying biofilms (equation 6.2).
(dCl2/dt) = (dCl2/dt) bulk + (dFe/dt) iron corrosion
dNH2Cl/dt =(dNH3/dt)autodecomposition + bulk + (dNH3/dt)nitrification
(6.1)
(6.2)
Using the above equations to consider extremes in relative rates of chloramine (as Cl2) and
chlorine (as Cl2) residual decay is instructive. In a pipe without significant nitrification,
chloramine is always expected to be more persistent than chlorine as a disinfectant, because
autodecomposition and reaction rates in bulk water for chloramine are relatively low compared to
reactions for Cl2 (Neden et al. 1992a; Norton and LeChevallier 1997; Vikesland et al. 2001).
Chloramine will have even greater advantages relative to free chlorine if iron corrosion is
significant due to the higher reactivity between chlorine and iron scales.
Conversely, if significant nitrification occurs and controls the chloramine decay rate,
chloramine might be less persistent than chlorine when iron corrosion rates are low. For example,
assuming initial total chlorine levels of 2 mg/L and a typical first order decay constant 1.16 d-1
(Vasconcelos et al. 1997), only 0.6 mg/L chlorine (as Cl2) chlorine would disappear after an 8 hour
stagnation time without considering demand from iron corrosion. However, microbial conversion of
1 mg/L NH3 to nitrite over an 8 hour time period would be expected to cause complete consumption
of a 4 mg/L chloramine residual based on the stoichiometry and decay models presented elsewhere
(Margerum et al. 1994; Vikesland et al. 2001).
In situations where chlorine decay is controlled by reactions with iron corrosion products
and chloramine decay is controlled by reactions with nitrite, the relative persistence of the two
disinfectants will be a function of the relative rates of nitrification and iron corrosion. Isopleths of
equal persistence could be roughly defined as:
(dFe/dt)for free chlorine = (dNH3/dt)nitrification for chloramine
(6.3)
Assuming that all the iron corrosion reacts to form Fe+2 which destroys free chlorine (See
details in Support Information); an iron corrosion rate of 0.8 uA/cm2 in an 3/4 inch pipe would
produce equal persistence (for free chlorine) as a nitrification rate of 0.2 mg NH3/L-h (for
chloramine). At lower iron corrosion rates or higher rates of nitrification, chlorine would be more
persistent than chloramine (Figure 6.1). If only 10% of the iron corrosion led to formation of scales
or corrosion products that reacted with free chlorine (i.e., 90% was non-reactive), then an iron
corrosion rate of 8 uA/cm2 would produce equal persistence (for free chlorine) as a nitrification
rate of 0.2 mg NH3/L-h (for chloramine).
Nitrification, chloramine decay rates and increased HPC are expected to be especially
significant under the frequent stagnation conditions found in premise plumbing (Edwards et al. 2005;
Zhang et al. 2009a). Once established, nitrification in such situations might not be controlled by even 8
mg/L chloramine-Cl2 (Fairey et al. 2004; Skadsen 1993). The loss of chloramine is a possible concern
in premise plumbing, since it might allow for re-growth of Legionella (and other opportunistic
pathogens) in water tanks, which has recently been identified as a major cause of waterborne disease
outbreaks (CDC 2008 ; Edwards et al. 2005; Strickhouser 2007). Limited data collected to date has
strongly indicated that chloramine is usually far superior to free chlorine in terms of Legionella control
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 6: Accelerated Chloramine Decay and Microbial Growth | 43
due to greater persistence of the chloramine disinfectant without nitrification (DeGraca 2005; Flannery
et al. 2006; Kool et al. 1999a; Kool et al. 1999b; Strickhouser 2007) and conversion to chloramine is
considered a key community-based intervention associated with reduced risk of Legionnaire’s disease
(Flannery et al. 2006).
MATERIALS AND METHODS
Pipe Rig Setup
A large scale pipe rig setup was designed to roughly simulate operating conditions in a
premise plumbing system (Figure 6.2a). The rig utilized a large reservoir to feed 8 parallel pipe
sections, each plumbed with a different type of material commonly used in home plumbing. Three
rigs were used to explore extremes of disinfection encountered throughout water distribution
systems (Figure 6.2a). Water flowed through the pipe loops on a regular schedule of two minutes
hydraulic flushing after every 8 hours’ stagnation.
Initially, the 8 pipe materials used in this study were: glass tubes covered with foil as a
control, cross linked polyethylene (PEX), polyethylene (PE), chlorinated polyvinyl chloride
(CPVC), copper, copper with epoxy coating (CuE), galvanized iron and stainless steel. CPVC,
PEX, copper, PE, galvanized iron and stainless steel pipes were purchased from a local hardware
store. Copper pipes lined with epoxy were acquired from ACE Duraflow®.
Later in the testing, PEX, glass and PE pipes were replaced with iron, new lead and old lead
pipes. Since unlined iron pipe could not be obtained, cylindrical cast iron samples of size 1 inch
(diameter) × 3/4 inch (length) coupons (Grammer,Dempsey& Hudson, Inc,Newark, NJ) were
placed inside Tygon tubing (Figure 6.2c). Sixteen new iron and sixteen old iron coupons derived
from (Zhang and Edwards 2007) were used. The iron coupons were coated with epoxy on all
except one side. New pure lead pipes were specially fabricated, and “old” lead pipes were created
by subjecting new lead pipes to accelerated aging via continuous exposure to 10 mg/L free
chlorine for 3 months. The latter condition reproducibly creates a thick PbO2 scale, thought to be
analogous to that produced naturally in many distribution systems such as in Washington D.C.
This approach was selected over use of harvested lead pipe in this work, because the method is
scientifically reproducible and also has less variability than is observed for harvested lead pipe.
All pipe materials except for the iron coupons are tubular with a ¾ inch internal diameter.
Each material had a continuous pipe section 5 ft long, as well as thirty-two 2 inch long coupons
connected by Teflon/Tygon tube, which allow occasional sampling for biofilm and weight loss
(Figure 6.2b).
Water Chemistry and Nitrification Inoculation
The test water used for this study utilized a recipe for synthesized Potomac River water
containing MgSO4 (39.3 mg/L), CaCl2•2H2O (20.5 mg/L), KNO3 (9.7 mg/L), Na2HPO4 (0.046
mg/L), NaHCO3 (56.8 mg/L), NaSiO3 (26.3 mg/L), CaSO4•2H2O, Al2(SO4)3•18H2O and ozonated
Natural Organic Matter (NOM) (0.18 mg/L TOC) with an initial pH of 8.
Three phases of work were performed on this test rig. During the first phase, the three
different disinfectant regimes investigated included: no disinfectant as a control, chlorine and
chloramine. The initial target disinfectant dose for chlorine and chloramine pipe loops were 2 and
4 mg/L (Cl2) respectively for 2.5 months; and then the doses were dropped to 0.5 and 1 mg/L (Cl2)
©2010 Water Research Foundation. ALL RIGHTS RESERVED
44 | Effect of Nitrification on Corrosion in the Distribution System
respectively for 15 months. The 200% higher dose of chloramine versus chlorine residual was
maintained in all experiments, as is consistent with established U.S. residual disinfectant practices,
and it is not an attempt to bias results in favor of chloramine. No detectable signs of nitrification or
nitrite production occurred in any of the rigs during this first phase of testing.
A second phase of testing established nitrification in the system that was dosed with
chloramine. The disinfectant doses in both chlorine and chloramine pipe loops were stopped and
0.5% dechlorinated Blacksburg, VA tap water containing nitrifying bacteria (Zhang et al. 2008a)
and 2 mg/L-N (NH4)2SO4 was introduced into the pipe loop that was chloraminated during phase
one. Trace nutrients (Zhang et al. 2008a) were also added and the level of ozonated NOM was
reduced to 0.02 mg/L TOC. The inoculation was continued for a year. During this time iron
coupons and lead pipes were placed into the rig. The pipe loops were then operated for another 4.5
months with stable nitrification occurring in all pipe materials except for copper pipes as
confirmed by ammonia conversion to nitrite and nitrate.
For the third phase of work, the disinfectant doses for the chlorine and chloramine rigs
were incrementally increased (i.e. for chlorinated pipe, to 0.125, 0.25, 0.5, 1, and 2 mg/L-Cl2 and
for chloraminated pipe, to 0.25, 0.5, 1, 2 and 4 mg/L-Cl2), with four weeks’ exposure at each
disinfectant level. The goal was to examine the possible inactivation of nitrifiers by secondary
disinfectants once they were established in the chloramine rig, and to also quantify the relative
rates of chloramine and chorine decay.
Analytical Methods
Changes in water quality were monitored by collection of water samples before and after 6
hours stagnation in the pipes. Nitrifier activity was tracked by measuring loss of ammonia,
production of nitrite and nitrate and reduction of pH. pH was monitored by using a pH electrode
according to Standard Method 4500-H+ B (Clesceri et al. 1998). NH4+-N was measured with
salicylate method using a HACH DR/2400 spectrophotometer, according to Standard Method
4500-NH3 (Clesceri et al. 1998). NO2--N and NO3--N were measured using DIONEX, DX-120 ion
chromatography according to Standard Method 4110 (Clesceri et al. 1998).
Chlorine and chloramine levels were measured with a DPD colorimetric method using the
HACH DR/2400 spectrophotometer, according to Standard Method 4500-Cl
(Clesceri et al. 1998). Bulk water and biofilm heterotrophic bacteria were monitored with
Heterotrophic Plate Count (HPC) according to Standard Method 9215 using the spread plate
method with R2A medium. Biofilm analysis was performed on the two inch coupon at the end of
each phase and before the replacing of pipe materials. The procedure of the biofilm collection was
modified from work of others (Silhan et al. 2006). The inner surface of the coupons was swabbed
with two sterile cotton sticks, which were then transferred to 20mL of sterilized 25 mM phosphate
solution (pH = 7.2) and vortexed vigorously for 1 min to release the bacteria from the sticks. Two
coupons were sampled for each pipe. The biofilm solutions from the two coupons were then
combined together and vortexed for 20 s. The combined biofilm solution from each pipe was
analyzed by HPC similar to bulk water HPC. Semi-quantitatively standardized colorimetric kits
called BART tests were also used in examining bacterial type and concentration for instant
measurement (Droycon Inc.).
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 6: Accelerated Chloramine Decay and Microbial Growth | 45
RESULTS AND DISCUSSIONS
Disinfectant Decay
Free chlorine Decay in Different Pipe Materials
The rate of free chlorine decay was strongly affected by the type of pipe material. During
both phase I and phase III, a high free chlorine decay rate was observed in highly corroded metal
pipes, including cast iron, galvanized, copper and old lead pipes (Figure 6.3). This is attributed to
consumption of the free chlorine by corrosion reactions. Over 80% of the chlorine disappeared in
these metal pipes within 6 hours of stagnation (Figure 6.3). In contrast, only 10 – 60% of the
chlorine decayed in plastic (PVC, PEX, PE, CuE), glass and stainless steel pipes that had very little
corrosion (Figure 6.3). The average chlorine decay rate in new lead pipes was less than the other
metal pipes, with about 50% of free chlorine decay after 6 hours (Figure 6.3). The overall
dependency of free chlorine decay rates on corrosion is consistent with other research which
determined that pipes like iron exerted a high chlorine demand whereas plastic pipes like PVC
were classified as unreactive (Hallama et al. 2002).
Chloramine (combined chlorine) decay in different Pipe Materials
Before nitrification was established (Phase I), after 6 hours stagnation time less than 30%
of the chloramine residual decayed in the pipes. The only exception was copper pipes, in which
about 60% of the chloramine decayed in 6 hours. The higher chloramine demand in copper pipes is
consistent with earlier findings attributed to an abiotic corrosion reaction (Nguyen 2005).
The net result is that before nitrification was established in these systems, free chlorine
decay rates were always higher than chloramine decay rates in all materials (p ≤ 0.006), consistent
with the conventional wisdom and equation 1 and 2 (Neden et al. 1992a; Norton and LeChevallier
1997).
During the third phase of the study when nitrification had been established in the pipe rig
with chloramine, the situation was markedly changed. In relatively inert pipe materials like PVC,
CuE, new lead and stainless steel pipes, much more chloramine (as Cl2) decayed than free chlorine
(p ≤ 0.0006) (Figure 6.4). For example, in PVC pipes, at all levels of applied disinfectant, at least
40% more chloramine disappeared during the 6 hour stagnation than was the case with free
chlorine disinfectant (Figure 6.4). A lower final chloramine residual (in mg/L) was even observed,
despite the fact that the initial chloramine dose (as Cl2) was twice as high as the chlorine dose for
conditions in this work.
In situations where free chlorine decay is controlled by corrosion reactions with pipe
materials like iron, copper, galvanized and old lead pipes, and chloramine decay is controlled by
nitrification, roughly similar decay rates were observed for chlorine and chloramine (Figure 6.4).
Even so, greater total chlorine residual was always present when free chlorine was used, compared
to the same situation with chloramine (p ≤ 0.03). The only exception was copper pipes, where
residual chlorine and chloramine were similar (p = 0.41), perhaps due to the fact that very little
nitrification occurred in copper pipes (ammonia loss < 40%) (Zhang et al. 2008a).
These results confirmed our key hypothesis regarding the relative stability of chloramine
and chlorine in premise plumbing systems.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
46 | Effect of Nitrification on Corrosion in the Distribution System
Comparing the situations with and without nitrification (Phase I vs. Phase III), the
chloramine decay rate (% decayed) with nitrification was 4-40 times faster than without
nitrification in the same pipe materials (Figure 6.3). This higher rate might be due to production of
nitrite and/or reducing pH. That is, nitrite reacts rapidly with chloramine and 1 mg/L nitrite can
consume 5 mg/L chloramine stoichiometrically (Margerum et al. 1994; Valentine 1984; Vikesland
et al. 2001). In this study, when 0-2 mg/L chloramine was dosed, in all pipe materials except for
copper pipes, on average, pH was decreased by 0.6-1.1 unit and 0.8-1.7 mg/L nitrite was detected.
Assuming that the chloramine decay rates only double for each 0.7 unit decrease in pH (Thomas
1987), it is most likely that the dominant factor for accelerated chloramine decay was nitrite
production.
Inactivation of Nitrification
Earlier studies (Wolfe et al. 1990) indicated that chloramine levels typically used for
potable water disinfection (1 to 2 mg/L) should be sufficient to eliminate nitrifiers in water main
distribution systems, but there are also cases where nitrification was not controlled by up to 8 mg/L
chloramine (Fairey et al. 2004; Skadsen 1993). In this study using long stagnation times found in
many premise plumbing systems, for most pipe materials nitrification activity did not decrease
until the chloramine level was increased to 4 mg/L and maintained at that level for several weeks
(Figure 6.5). The result indicated that 4 mg/L is the critical disinfectant level that nitrifier death
rate finally exceeded the growth rate (Fleming et al. 2005). The exceptions were PVC pipe, in
which nitrification activity started to decrease at 2 mg/L chloramine residual, and in the rig with
iron coupons for which nitrification activity was not reduced even at 4 mg/L chloramine (Figure
6.5). In both PVC pipe and iron coupons, nitrifier MPN only decreased markedly when chloramine
was increased to 4 mg/L (Figure 6.5).
As expected, the rate of chloramine decay was correlated to changes in nitrifier activity as
measured by ammonia loss %. For example, in PVC pipes, at 2 mg/L chloramine, ammonia loss
decreased (Figure 6.5) and the chloramine decay rate was also decreased from 100% to 70%
(Figure 6.4). With iron coupons, even at 4 mg/L chloramine, ammonia loss was not decreased
(Figure 6.5), and the chloramine decay rate stayed above 90% (Figure 6.4). The quantity of
nitrifying bacteria (MPN) in the bulk water did not directly impact chloramine decay rate (Figure
6.5).
Heterotrophic Bacterial Growth
Effect of Nitrification on Heterotrophic Bacterial growth
During the second phase of the study, when there was no disinfectant applied, nitrification
did not increase bulk water HPC (Figure 6.6-Phase II). This was contradictory to expectations
based on previous research (Powell 2004; Skadsen 1993; Wilczak et al. 1996; Wolfe et al. 1990).
However, biofilm HPC densities in most of the pipe materials were slightly higher during
nitrification when compared to control without nitrification in the same material (Figure 6.7). It is
suspected that the deionized water used for the makeup water had significant assimilable organic
carbon, and this might account for the relatively high HPCs both with and without nitrification.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 6: Accelerated Chloramine Decay and Microbial Growth | 47
Effectiveness of chlorine vs. chloramine in reducing HPC before and after nitrification
During the first phase of the study, chloramine was more persistent than chlorine without
nitrification (Figure 6.3), but chloramine was not statistically better in controlling bulk water HPC
(p > 0.07) (Figure 6.6). However, in galvanized iron pipe, chloramine consistently reduced HPC to
a greater extent than did chlorine, confirming the results of (Lechevallier et al. 1990) (Figure 6.8).
During the third phase of the study, with nitrification, chloramine is less persistent than
chlorine (Figure 6.3), and less effective in reducing bulk water HPC (Figure 6.6). Compared to the
control, bulk water HPC was decreased by ten to a hundredfold with free chlorine while it was not
significantly affected by chloramine (Figure 6.6).
Similar trends were observed for measurements of biofilm HPC’s, in that there was little
difference in situations without nitrification, but more biofilm bacteria for chloramine systems
when nitrification caused rapid loss of the disinfectant residual (Figure 6.9). For iron coupons,
neither chlorine nor chloramine reduced biofilm HPC even at chloramine dose up to 4 mg/L when
nitrification was occurring (Figure 6.9).
One exception to the trends, occurred in galvanized pipe, in which chloramine decayed
faster than chlorine (Figure 6.4) but lower HPCs were still observed with chloramine (Figure 6.9).
It is possible that this is due to the much lower presence of dissolved oxygen in the galvanized
pipes with nitrification. Specifically, bulk water DO was decreased to 2.8 mg/L in these pipes
versus the 5.7-6.2 mg/L DO observed in the same pipes using no disinfectant or chlorine. The
lower DO favors the growth of anaerobic bacteria like sulfate reducing bacteria (SRB),
methanogens or iron reducing bacteria (IRB) versus HPCs (Lovley and Phillips 1987). Indeed, in
this work, a low level of SRB (200 cfu/ml) and a very high level of IRB was detected (> 140000
cfu/ml) in the chloraminated galvanized pipe. In the pipes with no disinfectant and chlorine, lower
levels of SRB (< 200 cfu/ml) and IRB (< 9000 cfu/ml) were detected. The conclusion is that HPC
counts do not directly translate to lower levels of total bacteria in the system, and if total
quantifiable bacteria are considered (HPCs, IRB, SRB), more microbes were present in the rig
with chloramine than in the rig with chlorine.
Field Data
In parallel field work, samples were collected from homes at chloraminated utilities and
analyzed for possible water changes due to nitrification. In many of the utilities, nitrification was
occurring to markedly different degrees in premise plumbing in the same distribution system
(Table 6.1). This might be expected given differences in water usage rates, flow patterns,
plumbing materials and influent disinfectant residuals from home to home. For example, at
anonymous utility A, compared to site #2 and #3, site #1 had a larger ammonia decrease (up to
20% higher) and nitrite and nitrate increase, indicating a higher nitrification activity.
Correspondingly, this house had the largest pH decrease, chloramine decay and HPC increase (up
to 0.7 pH decrease, 95% chlorine decay and 105 cfu/ml were observed) (Table 6.1). The trend of
higher nitrification rates associated with higher chloramine decay rates and HPC counts were
observed at all utilities studied (Table 6.1).
©2010 Water Research Foundation. ALL RIGHTS RESERVED
48 | Effect of Nitrification on Corrosion in the Distribution System
CONCLUSIONS
The relative efficacy of chlorine versus chloramine in terms of controlling HPC in premise
plumbing is a complex function of autodecomposition, reactions with metals scale produced via
corrosion, nitrification rates, and the pipe materials themselves:
• Without nitrification, chloramine was always more persistent than chlorine.
• With rapid nitrification, in relatively inert pipe materials such as PVC, CuE, Stainless
Steel pipes, chloramine was less persistent than chlorine and less effective in reducing
bulk and biofilm HPC.
• In materials that are reactive to both chloramine and chlorine such as cast iron,
galvanized iron and old lead pipes, the relative decay rate and efficiency of controlling
bacterial growth is likely to depend on the relative rate of corrosion and rate of
nitrification.
FIGURES AND TABLES
Figure 6.1 Rough conceptualization of relative secondary disinfectant advantages
Note: Assume chlorine decay is controlled by reactions with iron corrosion products and
chloramine decay is controlled by reaction with nitrite from nitrification. Calculations assume a
pipe size of 3/4’’.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 6: Accelerated Chloramine Decay and Microbial Growth | 49
Figure 6.2 Pipe Rig Setup
Note: there are three water scenarios × eight pipe materials
Figure 6.3 Total chlorine decay in different pipe materials
Note: For phase I, data reported are the typical decay rate at 5 months pipe age, with initial
chlorine of 0.6 mg/L and chloramine of 1.1 mg/L. For Phase III, data reported are the average of
three weeks’ measurements. Error bars indicate a 95 % confidence interval on replicate samples.
Initial chlorine was 1.1 ±0.2 mg/L and chloramine was 2.0 ±0.2 mg/L.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
50 | Effect of Nitrification on Corrosion in the Distribution System
Figure 6.4 Chlorine and chloramine decay after 6 hours stagnation (Phase III) as total
chlorine doses were ramped up
Note: the initial chlorine dose is half of the indicated chloramine dose at any time.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 6: Accelerated Chloramine Decay and Microbial Growth | 51
Figure 6.5 Ammonia loss % and MPN with PVC, iron and new lead pipes
Note: the trend of ammonia loss with CuE, old lead, galvanized and stainless steel was
similar to that observed for new lead pipe. Ammonia loss % with Cu was below 40% throughout
the test.
control
chlorine
chloramine
HPC, cfu/ml
1000000
100000
10000
1000
100
10
1
8/1/04
Phase I
2/17/05
Phase II
9/5/05
3/24/06 10/10/06
Date
Phase III
4/28/07
11/14/07
6/1/08
Figure 6.6 Bulk water HPC in PVC pipes
Note: chlorine and chloramine doses were stopped during Phase II. Nitrification was
occurring in the chloraminated rigs during Phase II and Phase III.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
52 | Effect of Nitrification on Corrosion in the Distribution System
Figure 6.7 Biofilm HPC (Phase II)
Note: For PVC, Cu, CuE, Galvanized and Stainless Steel, data reported were the average of
two measurements and the upper and lower limits of the error bars are the data of the two
measurements. For Fe, new lead and old lead, the data reported were the result of one time
measurement.
Figure 6.8 Bulk water HPC in galvanized pipe (Phase I-no nitrification)
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 6: Accelerated Chloramine Decay and Microbial Growth | 53
Figure 6.9 Biofilm HPC with and without nitrification
Note: Data reported for phase I was the average of five measurements over the first year;
Data reported for phase III was the measurement at the end after applying the highest disinfectant
levels for a month (No data available for copper pipes with chloramine).
Table 6.1
Field study results
#1
#2
#3
#1
#2
#3
#1
#2
#1
#2
#3
Nitrite +
Ammonia
pH
Nitrate
decrease,
increase, decrease
mg/L
mg/L
Anonymous
0.41
0.54
0.7
0.34
0.17
0.4
0.38
0.15
0.1
St Paul. MN
0.8
0.36
0.72
0.4
0.35
0.12
0.18
0.02
-0.18
Bangor, ME
0.6
0.28
0.71
0.62
0.68
0.58
Hampden, ME
0.42
0.55
1.8
0.44
0.53
1.99
-0.39
0.02
0.15
Cl2 decay %
HPC,
cfu/ml
95
66
41
109000
172000
90
94
30
9
140
98
87
7300
16500
99
98
27
263000
138000
21000
40
Note: Data reported are for the 1st draw samples. Ammonia, nitrite, nitrate and actual pH
drop were calculated against the flush sample at the entry point of the distribution system.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
54 | Effect of Nitrification on Corrosion in the Distribution System
SUPPORTING INFORMATION
Modeling Approach for Figure 6.1:
Chlorine decay with iron corrosion products:
−
columb 6.24 × 1018 e −
1mol − e −
μA
−11 mol − e
1 2 = 1 × 10 −6
×
×
=
1
.
04
×
10
cm
s • cm 2
columb
6.02 × 10 23 e −
s • cm 2
Iron Corrosion: Fe →Fe2+ + 2eChlorine reacts with Fe2+: 2Fe2+ + Cl2→2Fe3+ + 2ClCl
1
so −2 =
e
4
mg − Cl2
μA
mol − e − Cl2 71000mg 3600 s
1 2 = 1.04 × 10 −11
× −×
×
= 6.62 × 10 −4
2
4e 1mol − Cl2
1h
cm
s • cm
cm 2 • h
Chloramine decay with nitrite:
Nitrification: 2NH4+ + 3O2→2H2O + 2 NO2- + 4H+
Chloramine reacting with nitrite: NH2Cl + NO2- + H2O →NH3 + NO3- + HCl
NH 2Cl 1mol 71mg − Cl2
=
=
So
+
1mol 14mg − N
NH 4
If ammonia consumption rate is 1
mg − N
, then chloramine decay rate would be
L−h
mg − Cl 2
, in a 3/4 inch diameter size pipe, this would translate
L−h
mg − Cl2
mg − Cl2
L
π / 4 × 1.9 2 × length cm 3
× 3 3×
= 2.4 × 10 −3
5
2
L•h
π × 1.9 × length cm
cm 2 • h
10 cm
So at equal decay rate of chlorine and chloramine,
mg − Cl2 1mg − N
μA
cm 2 • h
mg − N
1 2 = 6.62 × 10 −4
×
×
= 0.276
−3
2
cm
L•h
2.4 × 10 mg − Cl2
L•h
cm • h
5
©2010 Water Research Foundation. ALL RIGHTS RESERVED
to
CHAPTER 7
NITRIFICATION EFFECT ON CORROSION OF GALVANIZED IRON,
COPPER AND CONCRETE
Yan Zhang, Allian Griffin, Marc Edwards
The objective of this study was to examine nitrification in a wide range of pipe materials,
with particular attention focused on establishing nitrification and corrosion impacts.
INTRODUCTION
As more utilities in the United States switch to chloramine in order to comply with
Disinfectants and Disinfection By-Products Rule (D/DBPR) (Seidel et al. 2005; Wilczak et al. 1996),
there is increasing concern about potential costs and possible health implications of nitrification
(Wilczak et al. 1996; Zhang et al. 2009a). Nitrification is the conversion of ammonia to nitrite (NO2-)
and then nitrate (NO3-) by nitrifying bacteria. A 1991 survey indicated that two thirds of the medium
and large utilities that use chloramine report nitrification problems in water mains (Wilczak et al.
1996), and it is very likely that even a greater percentage have nitrification issues if premise plumbing
is considered (Zhang et al. 2008a; Zhang et al. 2009a) .
Nitrification is expected to be strongly influenced by different pipe materials since the pipe
can serve as a source of trace nutrients, toxic metals, attached growth and disinfectant destruction.
Earlier bench scale studies demonstrated that nitrification established readily in PVC and lead
pipes, and much less readily in copper or brass pipes due to copper toxicity (Zhang et al. 2008a).
Likewise, it was speculated that iron tubercles in distribution pipes may exert chlorine demand and
facilitate nitrifier growth (Odell et al. 1996). Iron and lead materials are also predicted to facilitate
nitrifier growth by recycling ammonia from nitrification end product nitrite and nitrate (Edwards
and Dudi 2004; Uchida and Okuwaki 1998; Zhang et al. 2009a). For concrete materials,
contradictory results have been reported: at two California utilities using chloramines, concrete
lined pipes had the lowest nitrifiers and HPC, possibly because of the high pH due to concrete
leaching (Steward and Lieu 1997); while in Pinellas county, Florida, cement lined ductile iron pipe
supported a higher heterotrophic biomass than did unlined iron (LePuil et al. 2003), possibly
because concrete accelerated chloramine decay.
Many studies have reported possible links of corrosion problems to nitrification (Douglas
et al. 2004; Edwards and Dudi 2004; Edwards and Triantafyllidou 2007; Powell 2004;
Triantafyllidou et al. 2007). The reduction of pH due to nitrification has been suspected (Douglas
et al. 2004; Odell et al. 1996) and confirmed to be the major mechanisms to increase lead release
(Zhang et al. 2009b). However, besides pH decrease, other changes by nitrification might also
impact corrosion. For example, accelerated disinfectant decay and increase of heterotrophic
bacterial growth might stimulate microbial induced corrosion (Cantor et al. 2006). The production
of soluble microbial product (SMP) might increase metal release by complexing with metal ions
(AWWARF and DVGW-TZW 1996). The production of nitrite and nitrate might also increase the
corrosion of iron and lead by chemical reactions (Kielemoes et al. 2000). The decrease of
dissolved oxygen might also affect corrosion since oxygen is the oxidizing agent for corrosion
reaction, and removing dissolved oxygen might decrease corrosion or increase corrosion by
inducing “concentration cell corrosion.”
55
©2010 Water Research Foundation. ALL RIGHTS RESERVED
56 | Effect of Nitrification on Corrosion in the Distribution System
MATERIALS AND METHODS
Large Scale Pipe Rig Testing
A large scale pipe rig setup was designed to roughly simulate operating conditions in a
premise plumbing system (Figure 6.2). The rig utilized a large reservoir to feed 8 parallel pipe
sections, each plumbed with a different type of material commonly used in home plumbing. The
eight pipe materials used were: glass tubes covered with foil as a control, cross linked polyethylene
(PEX), polyethylene (PE), chlorinated polyvinyl chloride (CPVC), copper, copper with epoxy
coating (CuE), galvanized iron and stainless steel. Later in the test, PEX, glass and PE pipes were
replaced with iron, new lead and old lead pipes. More specifics of the large scale pipe rig testing
was described in Chapter 6.
The test water used for this study utilized a recipe for synthesized Potomac River water as
described in chapter 6. Three rigs were used to explore extremes of disinfection encountered
throughout water distribution systems. Three phases of work were performed on this test rig.
During the first phase, the three different disinfectant regimes investigated included: no
disinfectant as a control, chlorine and chloramine. No detectable signs of nitrification or nitrite
production occurred during this first phase of testing. Based on the nitrification potential curve
proposed by other researchers (Fleming et al. 2005; Fleming et al. 2008), in order for nitrification
to occur, low chlorine and high ammonia are necessary, which was not the case during this phase.
Besides, nitrifiers were never introduced to the rig. A second phase of testing was to purposefully
establish nitrification in the system that was dosed with chloramine. The disinfectant doses in both
chlorine and chloramine pipe loops were stopped and 0.5% dechlorinated Blacksburg, VA tap
water containing nitrifying bacteria (Zhang et al. 2008a) and 2 mg/L-N (NH4)2SO4 was introduced
into the pipe loop that had been chloraminated. The inoculation was continued for a year. During
this time iron coupons (replacing PEX) and lead pipes (replacing glass and PE pipes) were placed
into the rig. The pipe loops were then operated for another 4.5 months with stable nitrification
occurring in all pipe materials except for copper pipes as confirmed by ammonia conversion to
nitrite and nitrate. For the third phase of work, the disinfectant doses for the chlorine and
chloramine rigs were incrementally increased (i.e. for chlorinated pipe, to 0.125, 0.25, 0.5, 1, and 2
mg/L-Cl2 and for chloraminated pipe, to 0.25, 0.5, 1, 2 and 4 mg/L-Cl2), with four weeks’ exposure
at each disinfectant level.
Mechanism Investigation of Galvanized Iron Corrosion
Galvanized pipes were newly purchased with a size of 30 cm (length) × 1.9 cm (diameter).
There are three replicate pipes for each condition. Pipes were kept stagnant and water in the pipes
was changed every 3.5 days (twice a week) using a “dump and fill” protocol. The control water
was the same water used in the large scale test without any ammonia or nitrate added and with an
initial pH of 7. Then the effect of different pH (pH 6 or 8), anaerobic condition and the addition of
different nitrogen species (2 mg/L ammonia, 2 mg/L nitrite or 4 mg/L nitrate) on galvanized iron
corrosion was studied by modifying the control water. Zinc and iron release was measured at each
water change.
The effect of different pH (pH 6 or 8), oxygen condition (aerobic and anaerobic) and
different nitrogen species (5 mg/L ammonia, 10 mg/L nitrite or 10 mg/L nitrate) was also
investigated with pure iron and zinc metals, which are the major components of galvanized iron.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 7: Nitrification Effect on Corrosion of Galvanized Iron, Copper and Concrete | 57
Pure iron and zinc wires (20 mm length × 1 mm diameter) were placed inside amber glass TOC test
vials (25 mm × 95 mm) with 30 ml solution. For each metal, there are fifteen conditions tested: 3
waters (control, control + 10 mg/L nitrite, control + 10 mg/L nitrate) × 2 oxygen conditions
(aerobic and anaerobic-achieved by purging with N2) × 2 pHs (6 and 8) + pH 8, aerobic, 5 mg/L
ammonia = 15 × 3 duplicates = 45 vials. The control water was simulated Potomac water without
any nitrogen species added. Water in the vials was changed twice a week and metal release was
tested each week.
Nitrification Effect on Concrete Corrosion Test
Reactors used were 25 mm × 95 mm Amber Glass TOC Test Vials (Fisher Brand) with 30
ml solution. Four pieces of Portland cement (37 mm × 5 mm × 4 mm) were added into each
reactor. The test water used was synthesized water contained MgSO4 (1 ppm-Mg), CaCl2 (4.9
ppm-Ca), KCl (10 ppm-K), Na2HPO4 (1 ppm-P), NaHCO3 (10 mM), and other trace nutrients (5
ppb Cu2+, 1.7 ppb Mo6+, 0.1 ppb Co2+, 5.6 ppb Mn2+, 2.6 ppb Zn2+, and 0.1 ppm Fe2+) and 2
mg/L-N ammonia. There were twelve reactors totally, each inoculated with Blacksburg, VA tap
water containing nitrifiers for two weeks. Then, in order to demonstrate the effect of nitrification,
for six of the reactors, nitrification was stopped by adding chlorine (up to 1 mg/L), and then these
reactors were used as control without nitrification. The initial pH was adjusted to 8 or 7 at different
time of the experiment, and initial alkalinity (added as NaHCO3) used was 10 and 75 mg/L at
different time of the experiment. Water was changed every weekday using a dump and fill
protocol.
Analytical Methods
Water samples were collected before and after introduction to the pipe/reactors, so that
nitrification activity and associated water quality changes could be quantified. Nitrifier activity
was tracked by measuring loss of ammonia, production of nitrite and nitrate, and reduction of pH.
pH was monitored by using pH electrode according to Standard Method 4500 H+ B. NH4-N was
measured with salicylate method using a HACH DR/2400 spectrophotometer, according to
Standard Method 4500NH3. NO2-N and NO3-N were measured using DIONEX, DX-120 ion
chromatography (IC), according to Standard Method 4110. Nitrifier most probable number (MPN)
was evaluated using the same approach as described in the bench scale study (Zhang et al. 2008a).
Total organic carbon (TOC) was analyzed using a SIEVERS 800 Total Organic Analyzer
according to Standard Method 5310C. Dissolved oxygen was quantified according to Standard
Method 4500 O G using a dissolved oxygen meter YSI Model 58. Soluble and total metal release
was also quantified. Soluble metal concentration was operationally defined by filtration though a
0.45 μm pore size syringe filter. Total metal release was digested with 2% nitric acid in a 80 °C
oven. Metal concentrations (Fe, Cu, Zn, Pb) were quantified using an inductively coupled plasma
mass spectrophotometer (ICP-MS) according to Standard Method 3125-B. Weight loss was
determined at the end of the test after removing loose scale from the coupons with a Dremel tool.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
58 | Effect of Nitrification on Corrosion in the Distribution System
RESULTS AND DISCUSSIONS
Nitrification Activity in Different Pipe Materials (Phase II)
Earlier bench scale studies under mostly stagnant conditions demonstrated that nitrification
established readily in PVC and lead pipes, and much less readily in copper or brass pipes due to
copper toxicity (Zhang et al. 2008a). In this larger-scale test using a wider range of materials and
more realistic flow regimes (water changes were three times per day), nitrification activity was
rapidly established in iron, plastic and lead materials (Table 7.1). In these materials, significant
nitrification was established almost immediately after inoculation, and when the inoculation was
stopped, about 70% ammonia loss was consistently observed after 6 hours stagnation (Table 7.1).
Although less ammonia loss (more ammonia residual) was observed with in the line with
cast iron coupons (Table 7.1), bulk water nitrifier MPN with iron coupons was similar or even
higher than PVC (chapter 6). In fact, the low surface area/volume for iron relative to the other
materials (Table 7.1) indicates extremely high rates of nitrification. It is also possible that the
ability of cast iron to reduce nitrate/nitrite back to ammonia can cause less ammonia loss for a
given amount of nitrification (Westerhoff and James 2003; Zhang and Edwards 2007).
In stainless steel and galvanized pipes, nitrification did not start until one month after
inoculation. Lower ammonia loss was also observed in galvanized pipes, possibly due to the
toxicity from Zn2+, since high Zn2+ is expected to inhibit nitrification under at least some
circumstances (Bott 2005; Zhang and Edwards 2005). It is also possible that nitrate is converted to
ammonia by galvanized pipes, since nitrate has been reported to completely convert back to
ammonia after 24 hours in galvanized pipes (Wagner 1993).
In copper pipe, average ammonia loss was about 40% (Table 7.1), but this ammonia loss
was believed not caused by nitrifier biofilm in copper pipes, but rather, the nitrifiers flowing from
the reservoir tank (Figure 7.1). The interior of the reservoir tank was made of glass, so similar to
other pipe materials; nitrification was also established in the tank during the inoculation. After the
inoculation was stopped, stable ammonia loss (76 ± 5 % after 6 hour stagnation) was observed in
the tank, which in fact continues to serve as an inoculation for the downstream pipes including
copper pipes. The lower ammonia loss (38%) in copper pipes compared to that in the tank (76%)
confirmed the toxicity of copper pipes. Additionally, compared to the nitrifier MPN in the tank
(160000 MPN/100 ml), nitrifier MPN decreased in copper pipes (500000 MPN/100 ml), but
stayed the same or increased in other pipe materials. The work conducted in Phase III further
confirmed that the limited nitrifier activity observed in copper pipe was coming from the tank: as
chloramine levels were increased to 1 mg/L, ammonia loss in the tank and copper pipes were
simultaneously decreased down to 6-30% while it stayed > 50% in the other pipes.
Nitrification Effect on Water Chemistry
Nitrification significantly decreased pH and Dissolved Oxygen (DO) during stagnation in
the pipe (Figure 7.1). After 6 hour stagnation time, in the pipes with nitrification, pH was
decreased by 0.6 to 1 unit in all pipe materials and DO was also decreased down to < 3 mg/L
(Figure 7.1). Nitrification stoichiometry (Zhang et al. 2009a) predicts a pH drop of up to 0.9 unit
and DO drop up to 10 mg/L if complete nitrification (100% ammonia conversion to nitrate) occurs,
so the observed pH and DO drop was in the range of expectation considering nitrification alone. In
the pipes without nitrification, pH and DO was not significantly changed during stagnation, except
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 7: Nitrification Effect on Corrosion of Galvanized Iron, Copper and Concrete | 59
that with iron coupons, pH and DO were also decreased (Figure 7.1), possibly due to iron
corrosion. In copper pipes, average ammonia loss of 40% (Table 7.1) can only account for a pH
drop of 0.5 units. The higher than expected pH and DO drop (Figure 7.1) might be due to reactions
like corrosion (Zhang et al. 2008b) and the growth of other bacteria. Total Organic Carbon (TOC)
was not significantly increased by nitrification in most pipe materials, but with cast iron coupons,
very high TOC was occasionally observed (Figure 7.1), which is correlated with the higher level of
nitrification and possibly other autotrophic bacterial growth (Morton and Edwards 2005) and
biofilm sloughing.
Effect of Nitrification on Pipe Corrosion
Galvanized Pipe
Before nitrification occurred in the galvanized iron pipe (Phase I), total zinc release from
galvanized pipes was similar at all three conditions (p ≥ 0.13) (Figure 7.2). After nitrification was
established (Phase II), zinc release from the nitrifying rig was significantly lower compared to the
rigs without nitrification (6 times lower on average, p ≤ 0.002) (Figure 7.2). The difference in zinc
levels with/without nitrification had very dramatic visual impacts (Figure 7.3). Water samples
from the pipes without nitrification had an obvious white turbidity; while in contrast, water
samples from pipes undergoing nitrification were very clear (Figure 7.3). This clearly
demonstrates beneficial effect from nitrification on water.
This effect was further confirmed at phase III, when disinfectants were re-dosed to the
chlorine and chloramine rigs. As chloramine levels gradually increased and nitrification activity
decreased, zinc release in the nitrifying pipe gradually increased to the point where it was not
statistically different from the pipes without ammonia (p ≥ 0.08 at 4 mg/L chloramine) (Figure 7.2).
Iron release from the nitrifying rig was also lower compared to the rig without nitrification
(p = 0.002), although the trend was not as dramatic as zinc release. Consistent with the lower zinc
and iron release, galvanized pipe with nitrification had a yellowish, dense scale, while the pipes
without nitrification had very thick whitish loose scale.
Mechanistic Investigation of Galvanized Iron Corrosion
The reduction of metal release from galvanized pipe by nitrification was initially deemed
contrary to the general expectation that nitrification would invariably increase corrosion and metal
release (Zhang et al. 2009a). Since nitrification converted ammonia to nitrite/nitrate and
significantly reduced pH and DO in the pipes (Figure 7.1), a mechanistic investigation was
designed to evaluate the effects of lower DO, lower pH, and addition of ammonia, nitrite and
nitrate on zinc release from galvanized pipes.
Among the conditions investigated, lower dissolved oxygen (DO) was proved to reduce
zinc release from galvanized pipe most significantly (p = 0.003); average zinc release from
anaerobic conditions was only 20% of that from aerobic conditions (Figure 7.4). The addition of
ammonia also reduced average zinc release by 40% (p = 0.03). Average zinc release at pH 6 and 7
was 60% of that at pH 8, although this was not statistically significant at > 95% confidence (p =
0.08 comparing pH 6 and 8, and p= 0.1 comparing pH 7 and 8). Adding nitrite or nitrate had no
effect on zinc release (p = 0.16 comparing control and adding nitrite, and p = 0.06 comparing
control and adding nitrate) (Figure 7.4).
©2010 Water Research Foundation. ALL RIGHTS RESERVED
60 | Effect of Nitrification on Corrosion in the Distribution System
So, overall, amongst the factors tested, lower DO, free ammonia, and lower pH might all
contribute to the reduced zinc leaching from the galvanized pipe with nitrification. And lower DO
is the most significant factor, which is not surprising since dissolved oxygen can accelerate the
corrosion from Zn metal to Zn2+ (Jones 1996).
For the test with pure iron and zinc metal, anaerobic conditions also reduced metal release
most significantly. Specifically, at pH 8, over the first four weeks of testing, average iron released
without O2 was 23% of the iron released with O2 (p = 0.002 ), and average zinc released without O2
was 14% of the zinc released with O2 (p = 0.07). Lower pH did not have a consistent effect on metal
release with pure iron and zinc. For nitrogen species, over the two months of testing, addition of
ammonia and nitrate did not affect the metal release from iron or zinc, but 10 mg/L nitrite reduced
iron release by 50% (p = 5 × 10-7) and zinc release by 30% ( p = 0.015). The beneficial effect of
nitrite on iron and zinc release was contradictory to the expectation that chemical reaction between
nitrite and iron/zinc would increase corrosion. Although earlier studies have concluded that nitrite
inhibit corrosion at high concentrations, but the concentrations were a lot higher (> 100 mg/L-N)
(Hubert et al. 2005; Kielemoes et al. 2000) than the level used in this study.
Copper Pipe
In copper pipes, the reduction of pH and DO did not significantly affect copper release
(Figure 7.5). This observation was consistent with solubility modeling (AWWARF and
DVGW-TZW 1996; Zhang et al. 2009b). At low alkalinity (30 mg/L): when copper pipe is old and
the controlling solid is CuO (tenorite), copper solubility is not a function of pH (Figure 7.6). In
contrast, when copper pipe is relatively new and the controlling solid is Cu(OH)2, soluble copper
increases with decreased pH (Figure 7.6), as observed in a previous study (Zhang et al. 2008b).
At high alkalinity, copper solubility is strongly affected by pH, consistent with earlier
studies (Dodrill and Edwards 1995; Edwards et al. 1996), particularly when the controlling solid is
Cu(OH)2 (Figure 7.6). A small pH drop of 0.2 unit would increase soluble copper by 30%, and pH
drop of 0.5 unit would double the soluble copper levels (Figure 7.6). This prediction can help
explain why in Willmar, Minnesota, a relatively small pH drop (0.4-0.9) caused by nitrification
significantly increased copper levels due to the high alkalinity (400-450 mg/L) (Murphy et al.
1997a; Murphy et al. 1997b). That is, in terms of copper release, the otherwise negligible pH drop
caused by nitrification (0.2 unit) at low alkalinity might be very significant at high alkalinity.
Metal Release from other pipe materials and weight loss
In new lead pipes, total lead release was generally increased by nitrification (p ≤ 0.02),
consistent with bench scale results at low alkalinities (Zhang et al. 2009b). However, no significant
metal release differences were observed in old lead coated with Pb (IV) scale by pre-exposing to
high chlorine levels, cast iron and stainless steel pipes between the rig with nitrification and the
rigs without nitrification (p ≥ 0.09).
Weight loss is generally considered to have less error in evaluating corrosion rate
compared to electrochemical measurements (AWWARF and DVGW-TZW 1996). In this study,
for all the pipe materials, no significant weight loss differences were observed among the three
disinfectant conditions (Table 7.2).
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 7: Nitrification Effect on Corrosion of Galvanized Iron, Copper and Concrete | 61
Nitrification Effect on Concrete Corrosion
Concrete materials can leach lime and when in contact with water, the bulk water pH is
usually increased (AWWARF and DVGW-TZW 1996). The effect of nitrification on the bulk
water pH was affected by the initial pH (Figure 7.7). Specifically, at an initial pH 8, nitrification
had no effect on pH, but when initial pH was reduced to 7, lower final pH was observed with
nitrification (Figure 7.7). Initial alkalinity had no effect on bulk water pH change (Figure 7.7).
Although nitrification lowered the final pH by 0.3-1.1 unit, this pH reduction was not
significant enough to increase concrete corrosion (Figure 7.7). As indicated by calcium leaching,
similar calcium leaching was observed with or without nitrification (Figure 7.7).
So overall, nitrification can have various impacts on the corrosion of different pipe
materials (Table 7.3). The reduction of pH due to nitrification has been confirmed to increase lead
release in earlier study (Zhang et al. 2009b). However, decrease of pH had no impact on concrete
corrosion and actually decreased zinc release from galvanized pipe. Copper release was also not
affected by pH decrease at low alkalinity waters. Besides pH reduction, reduction of dissolved
oxygen also decreased metal release from galvanized pipe and pure iron and zinc meals. The
production of nitrite and nitrate from nitrification is predicted to increase the corrosion of iron and
lead (Kielemoes et al. 2000), but no effect was observed except that high nitrite level decreased
iron and zinc release from pure metals (Table 7.3).
CONCLUSION
Nitrification was studied in a wide range of different plumbing materials in a large scale
pipe rig system:
• Nitrification activity was rapidly established in iron, plastic,lead and old concrete
materials, and less readily in galvanized, stainless steel, new concrete and copper pipes.
• Nitrification decreased pH by 0.6 to 1 unit, and reduced DO down to 2.4 mg/L; high
TOC was also observed with cast iron coupons.
• Nitrification decreased zinc release from galvanized pipe by 6 times due to the
reduction of DO; total lead release was generally increased by nitrification in new lead
pipes; nitrification did not impact corrosion or metal release in copper, old lead, cast
iron, concrete and stainless steel pipes.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
62 | Effect of Nitrification on Corrosion in the Distribution System
FIGURES AND TABLES
Figure 7.1 Nitrification effect on water chemistry
Note: Data reported were the average of four measurements for TOC, and two
measurements for pH and DO, ±indicate standard deviation. w/o nitrification data is the rig with no
disinfectant added throughout the test. Data for other pipe materials are similar to PVC but not
presented.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 7: Nitrification Effect on Corrosion of Galvanized Iron, Copper and Concrete | 63
Figure 7.2 Total zinc release versus time in galvanized pipes
Note: chlorine and chloramine doses were stopped during phase II, nitrification was
occurring in chloraminated pipe in Phase II and Phase III.
Figure 7.3 Water sample from galvanized pipes
©2010 Water Research Foundation. ALL RIGHTS RESERVED
64 | Effect of Nitrification on Corrosion in the Distribution System
Figure 7.4 Zn release in galvanized pipe
Note: Data reported are the averages of three week testing. The error bars represent the
95% confidence.
Figure 7.5 Total copper release versus time in copper pipes
Note: chlorine and chloramine doses were stopped during phase II, nitrification was
occurring in chloraminated pipe in Phase II and Phase III.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 7: Nitrification Effect on Corrosion of Galvanized Iron, Copper and Concrete | 65
Figure 7.6 Predicted soluble copper at different pH and alkalinities
Note: Assume 50000 ppb total copper.
Figure 7.7 pH change and calcium leaching from concrete w/ and w/o nitrification at
different initial pH and alkalinities
©2010 Water Research Foundation. ALL RIGHTS RESERVED
66 | Effect of Nitrification on Corrosion in the Distribution System
Table 7.1
Nitrification in different pipe materials
Surface
Ammonia
Time when
Rank
Pipe Materials
nitrification loss % after 6 Area/Volume,
hours
started
cm-1
0.12
Cast Iron
57 ± 3
PVC
72 ± 21
Polyethylene
70 ± 12
Cross linked Polyethylene
73 ± 5
< 1 week of
2.1
1
Epoxy Coated Copper
68 ± 5
inoculation
Old Lead
76 ± 16
New Lead
71 ± 7
Glass
52 ± 6
1.3
Old Concrete 1
100 2
2
Stainless Steel
Galvanized
< 1 month of
inoculation
67 ± 14
2.1
57 ± 11
0.6
3
3 months
New Concrete 1
60 3
2.1
4
Copper
N/A
38 ± 9
Note: Reported ammonia loss was the average of at least three measurements. ±indicate
95% confidence interval. 1: concrete was tested in glass pipes with small concrete pieces (30 cm ×
2.5 cm) placed inside; pure culture Nitrosmonas europaea was used, but effort was not made to
maintain the purity. Old concrete pieces have been exposed to other water for a year (Parks et al.
2008). 2: Ammonia loss tested after 72 hours; 3: Ammonia loss tested after 24 hours.
Table 7.2
Weight loss for different pipe materials
cast iron
Copper
New lead
Old lead
Galvanized
Stainless Steel
Weight loss, mg
no disinfectant chlorine chloramine + nitrification
35 ± 14
41 ± 16
41 ± 15
25 ± 13
13 ± 10
21 ± 6.5
55 ± 8
54 ± 11
60 ± 8.7
97 ± 55
85 ± 41
91 ± 19
94 ± 23
117 ± 30
127 ± 17
-11 ± 7
-3.4 ± 1.6
-0.3 ± 5.8
Note: ± indicate 95% confidence error. Note: the exposure time is four years for copper and
stainless steel pipes and 11 months for cast iron, new and old lead pipes.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 7: Nitrification Effect on Corrosion of Galvanized Iron, Copper and Concrete | 67
Table 7.3
Predicted and confirmed effect of nitrification on corrosion
Nitrification Effect on
Water Quality
Predicted Effect on
Corrosion
Confirmed Effect in this Project
Increase nitrite and
nitrate
Increase corrosion by
chemical reaction with metal
High level nitrite (10 mg/L) reduce iron and zinc
release from pure iron and zinc, no effect
observed for other metals
Increase lead release from lead pipe and brass1
Decreased pH and
alkalinity
Increase in HPC
Rapid decay of
chloramine
Decreased dissolved
oxygen
Decrease zinc release from galvanized pipe
Increase corrosion of copper
and lead, have varying effect
No effect on concrete
on iron corrosion
No effect on copper release with low alkalinity,
but increase copper release with high alkalinity
Increase microbiologically
N/A
influenced corrosion
Increase/decrease corrosion
N/A
depends on specific system
Change redox potential of
Decrease zinc release from galvanized pipe, iron
pipe surface, might increase
and zinc release from pure iron and zinc
or decrease corrosion
1: (Zhang et al. 2009b)
©2010 Water Research Foundation. ALL RIGHTS RESERVED
68 | Effect of Nitrification on Corrosion in the Distribution System
©2010 Water Research Foundation. ALL RIGHTS RESERVED
CHAPTER 8
EFFECT OF NITRIFICATION AND GAC FILTRATION ON COPPER AND
LEAD LEACHING IN HOME PLUMBING SYSTEMS
Yan Zhang; Simoni Triantafyllidou; and Marc Edwards
Reprinted with permission from ASCE Journal of Environmental Engineering 134 (7): 521-530.
Copyright 2008 ASCE
This study investigated nitrification and metal release under a simulated complex home
plumbing system: four different disinfectant scenarios, the use of Granular Activated Carbon
(GAC) and three types of leaded materials galvanically connected to copper pipes were
investigated. Nitrification and GAC filtration impact leaching of lead/copper to potable water
under typical home plumbing configurations. GAC filters removed the disinfectant and caused
rapid establishment of nitrification in chloraminated systems. The potential adverse consequences
of whole house GAC filters deserve increased scrutiny in chloraminated systems. The lower pH
values from nitrification and other microbes during overnight stagnation in pipes can markedly
increase (up to 800%) lead and copper contamination of water.
INTRODUCTION
Elevated lead and copper levels in drinking water have recently received increased scrutiny
due to cases of childhood lead poisoning tied to potable water in Durham, NC; Greenville, NC; and
Washington DC (Edwards and Triantafyllidou 2007; Triantafyllidou et al. 2007). Potable water
leaving treatment plants almost invariably has low copper and lead levels, but lead and copper can
increase in water as it sits stagnant within home plumbing (Federal-Register 1991; Schock 1989).
Copper pipe is the dominant material used to plumb buildings; lead pipe and leaded solder could
be used until officially banned in about 1986, and leaded brass (up to 8% lead by weight) are still
routinely installed in homes. The most common use of lead pipe is the service line connection
between the water main and home plumbing.
Lead can leach from brass, solder or lead pipe by itself, or lead corrosion can be driven by
direct galvanic connections with copper (Dudi 2004; Reiber 1991). In such situations, the lead
bearing material generally serves as the anode and is sacrificed by the copper:
(8.1)
Pb Æ Pb+2 + 2eand the cathodic reaction (e.g., O2 reduction) occurs on the copper surface:
O2 + 4e- + 2H2O Æ 4OH(8.2)
The rate of attack on the lead material alone or via galvanic corrosion is influenced by
numerous factors including water chemistry, disinfectant types (Dudi 2004) and bacterial growth
(Douglas et al. 2004; Reiber and Edwards 1997).
Effect of Disinfectant
Copper and lead release in water can be strongly influenced by disinfectant types
(AWWARF and DVGW-TZW 1996; Edwards and Dudi 2004) Theoretically, chlorine can
69
©2010 Water Research Foundation. ALL RIGHTS RESERVED
70 | Effect of Nitrification on Corrosion in the Distribution System
increase lead levels in drinking water by accelerating the corrosion of metallic lead to Pb (II) or Pb
(IV), but it was also predicted (AWWARF and DVGW-TZW 1996; Schock et al. 2001) to greatly
reduce lead leaching by converting more soluble Pb (II) to less soluble Pb (IV). In contrast, due to
its lower oxidation-reduction potential, chloramine does not have this ability to maintain the less
soluble Pb (IV) oxides (AWWARF and DVGW-TZW 1996), so it can sometimes worsen lead
levels in drinking water compared to chlorine. The loss of chlorine’s inhibiting effect on lead
leaching from service line pipes is now understood to be a major factor in a recent serious problem
with lead in Washington D.C. lead problem (Edwards and Dudi 2004; Lytle and Schock 2005),
although lead hazards also arose from corrosion of lead solder and brass due to reactions with
chloramine. Similar trends have been observed in other field and laboratory studies (Cantor et al.
2003; Vasquez et al. 2006).
The effect of chlorine on copper release varies dependent on circumstance. For example,
chlorine has been observed to increase copper leaching at lower pH values (pH < 7) (Atlas et al.
1982; Boulay and Edwards 2001; Hong and MacCauley 1998) but decrease copper corrosion and
copper release when there are problems with copper particulates at higher pH (e.g., “blue water”)
(Boulay and Edwards 2001; Edwards and Ferguson 1993). Some practical experience has
indicated that at some utilities copper solubility increased during periods of chloramination and
decreased with breakpoint chlorination (AWWARF and DVGW-TZW 1996). However, other
studies did not detect differences in copper corrosion rates or copper release when free chlorine or
combined chlorine was used (AWWARF and DVGW-TZW 1996).
Another underappreciated difference between the two disinfectants is the final decay
product: both produce chloride during disinfectant decay but chloramine also forms free ammonia.
In dead ends or in home plumbing systems with reactive materials, the free ammonia from
chloramine might have different impacts on corrosion and water quality relative to experience with
chlorine, decayed chlorine or even chloramine. It is therefore desirable to study the possible effects
of chlorine and chloramine on lead and copper corrosion which includes conditions with free
ammonia and no disinfectant.
Effect of Nitrification
On the basis of the above discussion, nitrification is another possible explanation for
divergent impacts of chlorine versus chloramine on lead and copper corrosion. Free ammonia
remaining after chloramine decay can be converted to nitrite and then nitrate by nitrifying bacteria
(USEPA 2005) with production of acid (Grady et al. 1999):
NH4+ + 1.9 O2 + 0.069 CO2 + 0.0172 HCO3- Æ 0.0172 C5H7O2N + 0.983 NO2- + 0.966
H2O + 1.97 H+
(8.3)
+
NO2 + .00875 NH4 + .035CO2 + .00875 HCO3 + 0.456 O2 + 0.00875 H2O Æ 0.00875
(8.4)
C5H7O2N + 1.0 NO3Nitrification can decrease pH and alkalinity (Equation 3 and 4) and produce nitrite at levels
of concern relative to drinking water guidelines, and create organic carbon that can serve as a
substrate for other bacteria.
This increasing concern over lead leaching due to nitrification arose from cases of lead
poisoning from water that occurred in chloraminated systems (Edwards and Triantafyllidou 2007;
Triantafyllidou et al. 2007), although it remains uncertain whether nitrification was a major
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 8: Effect of Nitrification and GAC Filtration on Copper and Lead Leaching | 71
contributor to the problem. The adverse effects of nitrification on lead leaching to water were first
noted more than 100 years ago (Garret 1891) and virtually no research on the subject has been
conducted since then. However, one study tied elevated copper levels at the tap to activity of
nitrifying bacteria in homes of Willmar, Minnesota (Murphy et al. 1997a).
Corrosion and metal leaching could be affected by nitrification through reduced pH and
alkalinity, and perhaps, through other indirect effects on bacterial growth and microbially induced
corrosion (McNeill and Edwards 2001; Schock 1989). In some cases microbially influenced
corrosion (e.g. blue water) and associated metal leaching is readily controlled by increased
disinfection (Cantor et al. 2006).
Effect of Granular Activated Carbon (GAC) filters
Home treatment devices that filter water at the Point-of-Use and Point-of-Entry were used
in about 37% of consumer homes in 2003 (USEPA 2003a). These devices are installed by
consumers to improve taste, color, and odor problems (McFeters 1990), as well as to possibly
increase the safety of their water. Devices employing activated carbon are widely used due to their
low cost and ability to remove chlorine. Activated carbon can also remove a wide range of
contaminants including organic compounds that are suspected carcinogens and that cause taste and
odor problems (Lykins et al. 1992). Elevated total bacterial growth and opportunistic pathogens
have sometimes been associated with Granular Activated Carbon (GAC) filters in homes (Lykins
et al. 1992), possibly due to the removal of disinfectant residual and the large surface area for
bacterial attachment in GAC. In chloraminated water distribution systems, installation of GAC
filters has been associated with nitrification problems in the U.S. (Feben 1935; Skadsen 1993) and
Europe (Vahala 2002; Vahala and Laukkanen 1998).
If GAC filters were installed in a home served chloraminated water, and nitrifier growth
were to occur and lower pH in the premise plumbing system, corrosion of copper and lead pipes
downstream of the filter could be adversely impacted. GAC filters also remove chlorine and
chloramine disinfectant residuals and significantly reduce organic levels, which can also affect
corrosion of copper and lead (Boulay and Edwards 2001; Korshin et al. 1994; Rehring and
Edwards 1996).
In light of the above discussion, this study is aimed at a broad preliminary investigation
into effects of disinfectant type, GAC filters and nitrification on leaching of metals from home
plumbing.
MATERIALS AND METHODS
Pipe Rig Setup
The basic experimental rig simulates a connection between copper pipe and a lead bearing
material such as brass, lead solder or lead pipe (Figure 8.1). All pipe materials were new. Copper
pipes and brass fixtures were purchased new from local hardware store; pure lead pipes were
specially ordered; and solder pipes were made by covering new copper pipes with melted solder
(50:50 lead: tin). The copper pipe is a 91.4 cm (3 ft - length) × 1.9 cm (¾ in. – diameter) section
that is electrically connected to lead-based material (lead, tin solder or brass fixture) via an
external grounding wire. The surface area ratio of the copper pipe to the leaded material was
approximately 12:1. A plastic ball valve is also placed between the two materials. By closing the
©2010 Water Research Foundation. ALL RIGHTS RESERVED
72 | Effect of Nitrification on Corrosion in the Distribution System
ball valve before collecting a sample, it is possible to collect samples of the water from the lead
anode and copper cathode sides of the pipe rig, which is advantageous in determining differences
in pH and nitrification in each segment. As a control and basis for comparison, pipe rigs were also
constructed using PVC pipes.
Water Chemistry
Phase I-Investigation of Disinfectant Effects on Different Pipe Materials
Blacksburg, VA tap water was used as the base water for this study. The chloramine
residual in this tap water was first removed by adding extra chlorine to achieve breakpoint
chlorination (Snoeyink and Jenkins 1980). The dose of Cl2 required to achieve the breakpoint was
estimated based on measurement of total chlorine (in chloramine) and ammonia in the water:
Cl2 Dose (mg/L) = 0.5 × Measured Cl2 (mg/L) + NH3 (mg/L) × 3.6
(8.5)
After allowing the water to sit for three days to allow disappearance of residual chlorine, it
was then modified to achieve four extremes of disinfection that might be encountered in potable
water systems including: no disinfectant, free chlorine, chloramine and free ammonia. Water with
free chlorine and chloramine represent water quality in homes located nearer the treatment plant
where disinfectant decay is negligible. The conditions with no disinfectant and free ammonia
represent conditions that might be encountered at distal portions of the distributions system.
Free chlorine and chloramine were targeted at 1.5 mg/L and 4 mg/L, respectively.
Chloramine was prepared with 4:1 chlorine to ammonia mass ratio. One mg/L NH3-N was added
to simulate a condition when chloramine was completely decayed. The final pH of all test waters
was adjusted to 7.8 ± 0.1, which is the target pH of finished water for Blacksburg water. Each
water quality was tested in duplicate, so there were 4 pipe materials (copper-lead, copper-solder,
copper-brass, and PVC) × 4 disinfectant types × 2 duplicates = 32 tests.
Water was changed in the pipes every 3.5 days (twice a week) using a “dump and fill”
protocol for 11 months duration. This was done in order to replenish nutrients and disinfectant
residual. The experiment results therefore simulate corrosion that might occur in sections of
infrequently used home plumbing.
Phase II and Phase III: Investigation of GAC and Nitrification Effect
After 11 months baseline behavior was established, duplicates for each condition were
treated differently, in that one pipe was fed with water pre-treated by a GAC filter whereas the
second was fed the same water not treated by GAC (Figure 8.2). The home GAC filters
impregnated with carbon used in this study were purchased off the shelf from a local hardware
store. The filters were conditioned with Blacksburg tap water with indications of nitrifying activity
(up to 0.03 mg/L-N NO2- detected) for two weeks prior to their use. By adding 1 mg/L-N NH3 into
the conditioning water, Nitrifiers rapidly colonized these filters.
During these two phases of the study, the water conditions tested included 4 pipe materials
(copper-lead, copper-solder, copper-brass, and PVC) × 4 disinfectant conditions × 2 treatment
conditions (GAC treated and non-GAC treated) = 32 tests. Water changes were made every day
during the work week (Monday-Friday) for better nutrient replenishment.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 8: Effect of Nitrification and GAC Filtration on Copper and Lead Leaching | 73
In order to determine the effect of GAC treatment on corrosion, each of the four waters (4
disinfectant types) were divided into two portions (Figure 8.2). One portion was passed through a
GAC filter (Figure 8.2), during which time the new water mixed with 500 mL old water retained in
the filter from the last water change. To simulate the same effect in water not treated by GAC, 500
ml water from the previous water change was left in a plastic container and mixed with freshly
prepared water before changing water in the rigs (Figure 8.2).
Two phases of experiments (Phase II and Phase III) were designed to examine effects of
GAC treatment on metals leaching. In Phase II, there was no pH adjustment before water was
filled into the pipes, as would occur in normal installation. But in phase III, to isolate the effect of
pH changes occurred due to GAC filtration/mixing, the pH was adjusted back to the target value of
7.8 before water was filled into pipes (Figure 8.2). To investigate the effect of nitrification
simultaneously, chronically, six months of Phase III work was conducted before the three day
duration study of Phase II.
Experimental Monitoring and Analysis
The extent of lead, copper leaching was quantified by measuring metal concentrations in
the water after sitting stagnant inside the pipes. Lead, copper, phosphorus, iron, aluminum were
measured using Inductively Coupled Plasma Mass Spectrophotometer (ICP-MS) according to
standard method 3125-B (Clesceri et al. 1998)
Nitrification and nitrifier bacteria activity was tracked by measuring loss of ammonia,
production of nitrite and nitrate and reduction of pH. NH4-N was measured with a colorimetric
method using a HACH DR/2400 spectrophotometer. NO3-N was measured using DIONEX,
DX-120 ion chromatography (IC). NO2-N was measured by both Hach and IC methods. Alkalinity
was monitored using HACH alkalinity test kit (model # AL-DT). Total Organic Carbon (TOC)
was monitored by SIEVERS 800 Total Organic Analyzer according to standard method 5310C
(Clesceri et al. 1998). In addition, the absorbance at wavelength UV254, which is often associated
with humic substances (or natural organic matter-NOM), was monitored on BECKMAN DU640
Spectrophotomer. Heterotrophic bacteria growth was monitored during the experiment by
Heterotrophic Plate Counting (HPC) method according to Standard Method 9215 C (Clesceri et al.
1998). R2A agar was used as the media and plates were incubated in a 20 °C constant temperature
room for 7 days prior to counting. The final HPC values were derived from the mean of two
duplicate plates at the appropriate dilution.
RESULTS AND DISCUSSION
Experimental findings are organized into four sections including 1) Effect of Disinfectants,
2) Effect of GAC, 3) Nitrification impacts, and 4) Field Studies.
Disinfectant Effect in Blacksburg Water without GAC-Phase I
For the first 11 months the study examined the effects of the four different disinfectant
conditions on copper and lead leaching. Nitrification did not occur in any of the rigs to a large
extent during this phase of testing as indicated by measurements of ammonia (< 20% loss), low
nitrite production (< 0.03 mg/L-N) and lack of a pH change in PVC pipes. This might indicate that
a long time is required to establish nitrification, or because the makeup water was completely
©2010 Water Research Foundation. ALL RIGHTS RESERVED
74 | Effect of Nitrification on Corrosion in the Distribution System
changed on a regular basis. Later experiments (Figure 8.2) retained some “old” water during water
changes which seemed to markedly enhance nitrifier activity.
Lead Release
Under these conditions without significant detectable nitrification, for a given type of lead
bearing material, average lead release was not significantly affected by disinfectant types
(Figure 8.3). Lead release from copper-lead and copper-solder rig was above the 15 ppb lead
action limit over the whole exposure time. For copper-lead rig, average lead release increased:
monochloramine < no disinfect < chlorine < ammonia. And for copper-solder rig, average lead
release follow the order of monochloramine < ammonia < chlorine < no disinfectant. Because total
lead release was gradually decreasing from 1000 ppb to about 100 ppb in copper-Lead and
copper-solder rigs, no confidence interval analysis was conducted for the average; rather, a paired
t-test was conducted to compare the head to head results of lead release with time. The difference
was rarely significant at 95% confidence (p > 0.05). The only exception for copper-lead rig was
that lead release with ammonia was 60% higher than that with monochloramine at 99% confidence
(p = 0.01). The exception for copper-solder rig was: lead release with monochloramine is lower
than that with chlorine and no disinfectant significantly, with p values of 0.03 and 0.01
respectively. Levels of lead leached from the brass containing rig and PVC were well below the 15
ppb lead action limit in this water.
Copper Release
Copper release was much lower than the LCR Action Level in all waters tested, even
during the first month of exposure. For copper-lead rig, average copper release follows the order:
ammonia < no disinfectant < chlorine < monochloramine. For copper-solder rig and copper brass
rig, average copper release follows the order of no disinfectant < ammonia < chlorine <
monochloramine. Because total copper release didn’t change as significant as lead release over
time, 95% confidence interval was applied to the pooled data (Figure 8.4). Copper release with
monochloramine was always higher than with other disinfectants, but only in copper-brass rigs
was this trend greater than 95% confidence.
Paired t-tests were also conducted to compare the head to head results of copper release. In
copper-lead rigs, the copper release with free ammonia was lower than with other disinfectants >
99% confidence (p < 0.01). In copper-solder rigs, copper release with ammonia was lower than
with monochloramine at 99% confidence (p = 0.01) and copper release with no disinfectant was
smaller than with monochloramine and chlorine at 99.4% (p = 0.006) and 98.5% (p = 0.015)
confidence. For copper-brass rig, the trend of no disinfectant < ammonia < chlorine <
monochloramine was confident at > 99%. As would be expected, there was no effect of
disinfectant type on copper release from PVC.
HPC
The HPC levels in water collected from the pipes after stagnation were ranked
monochloramine < chlorine < no disinfectant < ammonia. This result confirms the effectiveness of
the disinfectants, and the detriments of ammonia as a nutrient for re-growth in home plumbing, as
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 8: Effect of Nitrification and GAC Filtration on Copper and Lead Leaching | 75
described by (Edwards et al. 2005). But the HPC levels were not correlated to the extent of lead or
copper leaching in this water.
Overall Impacts of GAC Treatment on Water Chemistry-Phase II and Phase III
Impact of GAC on Water Chemistry
Treatment of water by GAC produced a number of changes in water chemistry that can
directly and indirectly affect corrosivity (Table 8.1). pH was decreased by about 1 pH unit after the
water passed through the GAC filter (Figure 8.5). While microorganisms can produce acidic
metabolites and decrease the pH (Bremer et al. 2001; Davidson 1996), since alkalinity did not
change significantly, the reduced pH is attributed mostly to increased CO2 arising from microbe
respiration. Parameters associated with humic substances including Total Organic Carbon (TOC)
and UV254 absorbance were decreased by 80-95%.
Heterotrophic bacterial counts were also increased by GAC treatment, especially when free
chlorine was the disinfectant where a 40X increase was observed relative to the same water
without a filter. This is most likely due to the complete removal of chlorine residual by GAC
filtration (Table 8.1). Although increased heterotrophic bacterial growth on GAC filters has shown
no direct adverse health effects (Wqa 2006), it can still indirectly affect water quality because
increased bacterial growth might increase pipe corrosion and metals leaching (McNeill and
Edwards 2001). Chloramine residual was also removed by GAC, and this doubled HPC counts
(Table 8.1). GAC treatment reduced HPC counts in water without disinfectant, perhaps due to the
effective removal of organic carbon (Edwards et al. 2005).
Of these changes lower pH is generally expected to increase corrosivity and metals
leaching, but removal of organic matter can also be highly influential. Organic matter has been
reported to increase copper release in some cases (Korshin et al. 1994; Kristiansen 1982; Rehring
and Edwards 1996) but decrease copper release in others (Rehring and Edwards 1996). Edwards
and Boulay demonstrated organic matter can increase copper release by complexation or colloid
mobilization/dispersion, but it can also decrease copper release by fueling microbial removal of
oxygen or via adsorption to the pipe surface and inhibition of corrosion (Edwards and Boulay
2001). Other researchers also suggest that if the organic matter is readily assimilable, then
microbially influenced corrosion might occur (Cantor et al. 2006).
GAC effect on nitrification occurrence
When chloramine disinfectant was used, GAC treatment dramatically increased nitrifying
bacterial growth. Specifically, in the water without GAC treatment with a high chloramine
residual, ammonia loss attributed to nitrification was only 19 % on average, whereas in the water
with GAC treatment and where chloramine had been removed, ammonia loss was 90 % on average
(Figure 8.6).
In waters with free ammonia, GAC treatment produced waters in which ammonia was
converted completely to nitrate. In contrast, higher levels of nitrite and some nitrate were present
in water not treated by GAC (Figure 8.7). We speculate that the incomplete conversion of nitrite to
nitrate in non-GAC treated water versus the complete conversion to nitrate in GAC treated water
was because GAC filtration removed the inhibition of chlorate on nitrite oxidizing bacteria (NOB)
activity. Chlorate has been noted to partially inhibit the activity of nitrite oxidizing bacteria (NOB)
©2010 Water Research Foundation. ALL RIGHTS RESERVED
76 | Effect of Nitrification on Corrosion in the Distribution System
at levels below 1 mg/L (Belser and Mays 1980). Chlorate dropped from 0.43 ppm to 0.12 ppm
after GAC treatment. This result illustrates that nitrification can be occurring without detectable
nitrite, thus hinder the detection of nitrification in home plumbing systems since nitrite is often
used as an indicator of nitrification in drinking water systems (Wilczak et al. 1996).
GAC effect on copper and lead release-Phase II
The relative effect of GAC treatment was calculated using a ratio of metal leaching (e.g.,
lead) as follows:
Ratio = Pb concentration (ppb) using GAC treated water/ Pb concentration (ppb) using
non-GAC treated water
(8.6)
If the ratio < 1, GAC treatment reduced lead leaching whereas a ratio > 1 indicates that
GAC treatment increased lead leaching.
During Phase II work, no pH adjustment was made after GAC treatment in homes to
simulate actual installed conditions. The pH depression by GAC treatment would be expected to
adversely affect lead corrosion in downstream pipes. However, the qualitative effects of GAC
treatment for lead corrosion in Blacksburg water dependent on the plumbing material examined,
suggesting that pH depression was not always the dominant factor controlling corrosion.
GAC treatment did increase lead release from copper-lead pipe rigs when the water
contained monochloramine, chlorine and no disinfectant (Figure 8.8). In the case without
disinfectant lead leaching was increased by 316% after GAC treatment. The lack of effect in water
with free ammonia is to be expected, given the similarity in pH value of the water with and without
GAC treatment in this case (Figure 8.5).
The qualitative effect of GAC treatment for copper-solder rigs was the same as for
copper-lead pipe rigs, as GAC treatment had no effect on lead release with ammonia and increased
lead release when there is no disinfectant. But when monochloramine and chlorine were used, lead
release was decreased by GAC treatment in copper-solder rigs. It is possible that the benefits of
organic matter removal by GAC filtration overcame the detriments of the lower pH values for
copper-solder rigs. Earlier research demonstrated that higher organic matter increased lead
leaching from lead-tin solder (Korshin et al. 2005).
For copper-brass rigs, the final average lead concentration was always reduced by GAC
pre-treatment (from 4.7 ppb to 1.4 ppb on average). Part of the reduced lead in GAC treated water
was due to removal of trace lead from the water even before contacting the pipes (initial water lead
decreased from 2.6 ppb in non-GAC treated water to 0.8 ppb in GAC treated water), consistent
with the general observation that GAC can remove heavy metals including copper, lead, cadmium,
etc (Reed et al. 1993; Scholz and Xu 2002). Another part of the reduced lead in GAC water was
due to lower lead release from brass side. It is possible that the benefits of NOM removal
outweighed the detriments of lower pH values after GAC treatment, since removal of NOM would
be expected to decrease lead leaching from brass (Korshin et al. 2000),
GAC treatment increased copper leaching for all conditions. The largest increase was
approximately 800% in copper-brass rigs with chlorine disinfectant. The increase in copper
leaching was not as significant when ammonia was used compared to other disinfectants
(Figure 8.9). Clearly, copper release was more consistently impacted by the lower pH values after
GAC treatment (Figure 8.5).
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 8: Effect of Nitrification and GAC Filtration on Copper and Lead Leaching | 77
GAC effect on copper and lead release-Phase III
The relative role of pH depression versus other impacts of GAC treatment was obviously
complex and dependent on the plumbing material. The Phase III experiments attempted to
examine this in greater detail by re-adjusting the pH of the water back to its initial value (pH 7.8)
before filling the pipes.
In the copper-lead rigs, lead leaching was no longer worsened by GAC treatment if pH
depression was countered. In fact, when free ammonia was present, lead release was decreased by
GAC treatment, perhaps due to slightly lower pH in non-GAC treated water (pH = 7.39) than GAC
treated water (pH = 7.57) after stagnation (Table 8.3).
In copper-solder rigs, even without the lower pH values, GAC treatment still increased lead
release by three times and 50% when no disinfectant and ammonia were used. However, GAC
treatment didn’t have an effect when chlorine or monochloramine disinfectants were used (data
not shown). In copper-brass rigs, similar to phase II work, the final average lead concentration was
reduced by 34-89% after GAC treatment.
In copper-lead and copper-solder rigs, GAC treatment increased copper release only when
using chlorine as the disinfectant (data not shown). This might be because chlorine helps to control
copper release either by control microbially influenced corrosion (Cantor et al. 2006) or by
decreasing corrosion rate directly (Edwards and Ferguson 1993) and GAC removed the chlorine.
However, in copper-brass rigs, GAC treatment still increased copper release for all conditions (up
to 500% when chlorine was used). In the copper brass rigs it was consistently observed that the
copper side of the galvanic connection had higher pH values than the brass side in non-GAC
treated water. We define this pH difference (pH of copper side – pH of brass side) as ΔpH. The
ΔpH of GAC treated water is much smaller than for non-GAC treated water. Galvanic current and
voltage were plotted against ΔpH and no correlation was found, indicating that change in galvanic
corrosion was not the major mechanism affecting the difference in ΔpH and copper release
between non-GAC treated and GAC treated waters. (Rehring and Edwards 1996) also found
increases in copper leaching and corrosion rate after waters were treated by GAC.
Nitrification Effects-Phase III
Nitrification occurrence
Factors that trigger nitrification in potable water systems are poorly understood. A number
of observations were made during the course of this study which gave insight to factors which
might be important contributors to nitrification in premise plumbing systems. During the first 11
months of the study when no GAC treatment was applied, nitrification was insignificant as
indicated by low ammonia consumption (< 20%), low nitrite production (< 0.03 mg/L-N) and no
pH change even in PVC pipes. Three months after starting to use GAC, nitrification was detected
in the water with free ammonia in both GAC treated and non-GAC treated waters. For the
non-GAC water, to isolate the effect of GAC treatment, 500 mL of water sat stagnant between
water changes, and was mixed with new water each time (Figure 8.2). This mixing of aged water
with freshly prepared water, instead of using 100% freshly prepared water, is believed responsible
for initiation of nitrification that had not been observed at all during 11 months of previous study.
The stagnant water used for mixing initially had 1 mg/L NH3-N present. This ammonia was
completely converted to nitrite within 24 hours, and the pH was decreased to below 7. So the final
©2010 Water Research Foundation. ALL RIGHTS RESERVED
78 | Effect of Nitrification on Corrosion in the Distribution System
pH of non-GAC treated water was similar to the pH observed after GAC treatment (Figure 8.5).
For monochloramine, chlorine, and no disinfectant conditions, the pH of water not treated by GAC
stayed higher (Figure 8.5).
Nitrification occurrence and pH change in pipes
Some of the nitrification was occurring in the containers and GAC before the pipes,
whereas in other cases the nitrification and pH changes were occurring in the pipes. For example,
in the phase III work where pH had been re-adjusted to 7.8 before water was fed into pipes, in
waters containing free ammonia at least some nitrification occurred in all pipes in both non-GAC
and GAC treated water. PVC had much higher nitrification activity than lead, solder and brass
containing rigs (Table 8.2). This is reasonable considering that > 200 ppb copper was present in
copper containing rigs, which is adequate to inhibit nitrification (Loveless and Painter 1968;
Skinner and Walker 1961; Zhang et al. 2009a). In addition, ammonia consumption and pH were
found to be highly correlated in PVC pipes (R2 = 0.91), indicating nitrification was a major cause
of pH drop (Figure 8.10). But in lead, solder and brass containing rigs, besides nitrification, there
are many other factors that can contribute to pH change in the rigs. For example, the increase in pH
due to copper corrosion might counter the pH decrease by nitrification (equation 8.7).
2Cu + 4H+ + O2 → 2Cu2+ + 2H2O
(8.7)
So the extent of nitrification and pH change during stagnation were therefore less strongly
correlated in these pipe rigs (R2 = 0.28, 0.32 and 0.52 in copper-lead, copper solder and
copper-brass containing rigs, respectively). Indeed, amongst all the different plumbing materials,
only the pH in PVC and copper-lead rigs consistently dropped (Table 8.3).
For waters containing monochloramine, ammonia consumption was also observed in every
pipe (Table 8.2). But due to the low levels of conversion and the reaction between nitrite and
chloramine, it is not possible to conclude whether nitrification actually occurred in the pipes with
chloramine. On one hand, nitrification may have occurred, because ammonia was consumed, and
nitrite/nitrate produced could be removed by reacting with monochloramine or the pipe surface.
But on the other hand, it is possible the low levels of ammonia removal could be due to natural loss
of ammonia to the air or uptake by heterotrophic bacteria. In any case, nitrification in the presence
of chloramine was much lower than was observed in corresponding pipes with ammonia.
Nitrification Associated Effect on Lead Release
The higher nitrification extent in the pipe with ammonia tended to cause relatively high
lead leaching versus all other conditions in copper-lead rigs, since pH was dropped by nitrification
in these pipes (Figure 8.11).
Nitrification Associated Effect on Copper Release
During phase I study, copper release in the water with free ammonia was lower than that
with other disinfectants > 99% confidence (p < 0.01); but in phase III study, after nitrification
occurred, copper release with ammonia was increased to levels similar to the other disinfectant
conditions (Figure 8.12). Again, this effect was correlated to the lower pH in the water with
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 8: Effect of Nitrification and GAC Filtration on Copper and Lead Leaching | 79
ammonia, which was partly caused by nitrification. Copper release trends in copper-solder and
copper-brass rigs did not change significantly after nitrification occurrence.
Field Studies
In parallel field work, samples were collected from homes with chronic lead and copper
corrosion problems occurring at the ends of distribution systems. All of the homes were in systems
using chloramine disinfectants. Measurements were made of water chemistry after overnight
stagnation for comparison to that obtained after extensive flushing.
As was observed in the laboratory work, large pH drops were occurring while the water sat
stagnant in the pipes during overnight stagnation (Table 8.4), contributing to localized lead and
copper corrosion problems. In Hawaii homes, detection of acid producing bacteria and high HPC
levels were associated with these pH drops (Marshall and Edwards 2004), whereas in Indiana and
Washington D.C., nitrification (indicated by increased nitrite) was detected.
Some of the largest pH drops and highest lead values observed in the field occurred in
homes using GAC whole house filters. Ironically, these homes are not sampled during EPA Lead
and Copper Rule Monitoring, due to concerns that the filters might produce water with atypically
low lead.
In the Indiana study, the consumers had significant blue staining and frequently observed
“blue water.” A follow-up bench scale test was conducted to examine the effects of countering the
pH drop that was occurring in the home plumbing during overnight stagnation by addition of
caustic (sufficient to keep final pH at 7.8). After a period of several weeks, copper release after 24
hours stagnation had stabilized at 6 mg/L in the water without the pH adjustment versus 1 mg/L in
the water with the pH adjustment.
While other researchers have made similar observations regarding the potential adverse
consequences of pH depression during stagnation for copper leaching (Webster et al. 2000), the
increasing use of chloramination and in-house filters can be expected to make such situations more
prevalent. Clearly, corrosion control strategies at utilities need to consider this potential
complication when determining target pH values for optimal corrosion control.
SUMMARY AND CONCLUSIONS
Bio-chemical reactions can drop pH 1 or 2 units during stagnation in premise plumbing
systems, undermining control of lead and copper corrosion of drinking water. Carbon dioxide
produced by biological activity in the pipes and on GAC filters, or acidity production by
nitrification were dominant contributors to pH depression. The largest pH drops occurred in PVC
pipes, because overall corrosion and buffering for copper and other metallic plumbing tended to
raise pH of the bulk water.
GAC filtration significantly decreased the pH, removed organic matter and disinfectant
residuals. GAC stimulated both heterotrophic bacterial growth and nitrification when
monochloramine was used. Complete conversion of ammonia to nitrate occurred in waters
pre-treated by GAC, such that nitrite production was undetectable and could not be used to
indicate nitrification.
The pH decrease by GAC treatment and nitrification increased copper release in all
conditions. But for lead release, the net result of GAC treatment was highly dependent on the
lead-bearing pipe material. Specifically, GAC treatment markedly increased lead release in
©2010 Water Research Foundation. ALL RIGHTS RESERVED
80 | Effect of Nitrification on Corrosion in the Distribution System
copper-lead rigs; however, in copper-solder and copper-brass rigs the beneficial effect of other
changes by GAC can sometimes outweigh the detrimental effect of decreased pH, thereby
decreasing lead release.
In Blacksburg water, occurrence of nitrification in pipes during overnight stagnation had
no discernable direct adverse impacts on metal leaching, other than those attributable to pH
depression. Nitrification was not significant in premise plumbing during the first 11 months of
testing when water in the pipe rigs was completely changed twice a week. Nitrification was rapidly
established if water was pre-treated using a GAC filter, or if a portion of the water was allowed to
sit stagnant for one day before mixing with freshly prepared water. In the time period before
nitrification became established, average lead release was not significantly affected by disinfectant
types. Average copper release was lowest with free ammonia in copper-lead rigs relative to those
exposed to no disinfectant, chlorine or chloramine. In copper brass rigs, copper release followed
the order of no disinfectant < ammonia < chlorine < monochloramine.
At three different utilities using chloraminated water, homes with lead and copper leaching
problems had significant pH depression during overnight stagnation. Though an exhaustive study
was not done, the problematic situations tended to occur in situations with infrequent water use,
GAC filters or at the end of the distribution system.
FIGURES AND TABLES
Silicone Stopper
Copper Pipe Wired Galvanic Connection
Spacer Lead, Solder or Brass Fixture
Figure 8.1 Pipe Rig Setup
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 8: Effect of Nitrification and GAC Filtration on Copper and Lead Leaching | 81
1250 ml
1250ml,
Fill pipes
Mix with 500 ml “old water”
stagnant
in the bottle
Blacksburg water
+
disinfectant Adjust pH to 7.8
2500 ml
GAC filter
1250ml,
Fill pipes
1250 ml Phase II: No pH adjustment before
filling pipe
Phase III: pH adjusted back to 7.8
before filling pipe
Mix with 500 ml “old
water” stagnant in the
GAC bed
Figure 8.2 Water Preparation for Phase II and Phase III study
Total Lead Release, ppb
1000
480 405
306 395
ammonia
monochloramine
chlorine
no disinfectant
569
445 493
284
100
7.5 6.7
10
9.7
4.9
1.4
2.2 2.2
1.8
1
Copper-Lead
Copper-Solder
Copper-Brass
PVC
Figure 8.3 Average lead release after a stagnation period of 3.5 days in the pipes
Data was the average of 16 measurements over the 11 months exposure time.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
82 | Effect of Nitrification on Corrosion in the Distribution System
Figure 8.4 Average copper release after a stagnation period of 3.5 days in the pipes
Data was the average of 14 measurements over the 11 months exposure time and error bars
indicates 95% confidence interval of these measurements.
Initial pH = 7.8
Figure 8.5 pH values after water passing through GAC filtration or after mixing fresh water
with aged water
Initial pH = 7.8. Data was the average of three measurements and error bars indicates 90 %
confidence interval of these measurements.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 8: Effect of Nitrification and GAC Filtration on Copper and Lead Leaching | 83
Figure 8.6 Nitrification in PVC pipes with chloramine disinfectant
Y axis indicates ammonia loss after 24 hours stagnation time in the pipe. Higher ammonia loss
indicates more nitrifier activity.
Figure 8.7 Representative nitrogen balance 24 hours after water was fed into PVC pipes
Ammonia consumption is 1 mg/L-N with non-GAC treated water, and 0.9 mg/L-N with GAC
treated water.
Figure 8.8 Effect of GAC treatment on lead release (Phase II)
Y axis shows Ratio of lead release with GAC treated waters / non-GAC treated waters. Error bars
represent 90% confidence interval.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
84 | Effect of Nitrification on Corrosion in the Distribution System
Figure 8.9 Effect of GAC treatment on copper release (Phase II)
Y axis shows Ratio of copper release with GAC treated waters / non-GAC treated waters. Error
bars represent 90% confidence interval.
Figure 8.10 Final pH and ammonia consumption correlation in PVC pipes after 24 hours
stagnation
Initial pH = 7.8.
Figure 8.11 Lead release in copper-lead rig, non-GAC
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 8: Effect of Nitrification and GAC Filtration on Copper and Lead Leaching | 85
Figure 8.12 Copper release in copper-lead rig, non-GAC
Table 8.1
Water quality comparison before and after filtration through GAC
pH
ammonia
Monochloramine
Chlorine
No disinfectant
Disinfectant
alkalinity,
UV254,a HPC,
Ammonia,
residual,
mg/LTOC, ppb
mg/L-N
cfu/ml
bs
mg/L-Cl2
CaCO3
before
7.94
31
1412
0.030
43600
after
7.03
28
161
0.004
39000
before
7.91
29
1307
0.052
7425
0.95
0.90
4.00
0.80
0.80
after
6.88
25
137
0.005
14900
0.09
before
7.86
28
1410
0.030
125
1.51
0.04
after
7.10
26
63
0.005
3950
before
7.94
32
1408
0.031
104750
after
6.97
29
275
0.009
58900
©2010 Water Research Foundation. ALL RIGHTS RESERVED
86 | Effect of Nitrification on Corrosion in the Distribution System
Table 8.2
Average ammonia consumption % after 24 hours stagnation in the indicated pipe material
ammonia
monochloramine
PVC
97.0
16.8
ammonia
monochloramine
PVC
96.5
75.0
non-GAC
Copper- CopperLead
Solder
29.9
24.2
26.1
23.3
GAC
Copper- CopperLead
Solder
21.2
19.8
11.0
19.9
CopperBrass
32.9
19.2
CopperBrass
21.9
18.4
Table 8.3
Final pH after 24 hours stagnation in the indicated pipe material
Initial pH = 7.8
ammonia
monochloramine
chlorine
no disinfectant
PVC
7.24
7.58
7.55
7.79
ammonia
monochloramine
chlorine
no disinfectant
PVC
7.18
7.41
7.59
7.84
non-GAC
Copper- CopperLead
Solder
7.39
7.51
7.68
7.60
7.63
7.68
7.56
7.56
GAC
Copper- CopperLead
Solder
7.57
7.61
7.64
7.68
7.70
7.73
7.74
7.76
CopperBrass
7.85
7.92
7.75
7.93
CopperBrass
7.79
7.78
7.91
7.96
Table 8.4
Field studies at three utilities
Initial
pH
Nitrite
pH after Increase,
stagnation mg/L-N
Indiana
7.90
7.20
0.25
Hawaii
Washington D. C.
7.89
7.90
6.77
6.70
0.00
0.50
©2010 Water Research Foundation. ALL RIGHTS RESERVED
CHAPTER 9
UTILITY INTERVIEW AND CASE STUDIES
Helene Baribeau, Yan Zhang and Marc Edwards
The case studies were designed to document effects of nitrification on corrosion, and
effects of corrosion on nitrification, in drinking water distribution systems and premise
(buildings/homes) plumbing. The case studies also attempted to correlate nitrification occurrence
with concentrations of macro-/micro-nutrients in water and in proximity of certain distribution
system materials. To accomplish these goals, the following tasks were conducted:
• Interview participating utilities to collect background information
• Review existing utility data to conduct a preliminary assessment
• Collect samples in selected distribution systems
On the basis of the above information, Case Studies and utility guidelines were prepared.
METHODOLOGY
Utility Interviews
The interviews were conducted using a questionnaire that collected background
information about the participating utilities. The questionnaire is presented in Appendix E. Twelve
of the 14 participating utilities returned filled out the preliminary questionnaire.
Review of Existing Water Quality Data
Existing water quality data were examined to assess nitrification issues for distribution
system and premise plumbing. The water quality parameters of interest requested at sampling
locations including entry points (POEs) to the distribution system (after addition of all chemicals)
in the distribution system, and premise plumbing were requested for the following: water
temperature, pH, total chlorine, chlorine-to-ammonia-N weight ratio, ammonia concentration (free
and/or total ammonia), nitrite concentration, nitrate concentration, heterotrophic plate counts
(HPC), total organic carbon (TOC) concentration, dissolved organic carbon (DOC) concentration,
Trihalomethanes
(THMs)
concentration
(chloroform,
bromodichloromethane,
dibromochloromethane, and bromoform), alkalinity, lead concentration, copper concentration,
calcium concentration, total dissolved solids (TDS), dissolved oxygen concentration, chloride
concentration, sulfate concentration, potassium concentration, aluminum concentrations, zinc
concentration, iron concentration, manganese concentration, magnesium concentration, phosphate
concentration, silica concentration.
Distribution System Sampling
Samples collection was coordinated from participating utilities as per Table 9.1.
87
©2010 Water Research Foundation. ALL RIGHTS RESERVED
88 | Effect of Nitrification on Corrosion in the Distribution System
To the extent possible, samples were collected at the POE(s) to the distribution system, as
well as at two to three homes throughout the chloraminated section of the distribution system.
First-draw and flushed samples were collected from homes presenting higher risks of lead and
copper contamination, and/or homes near areas where nitrification has been experienced. Samples
were collected from faucets within homes (not from outside hose bibs), with a focus on homes
without home water treatment devices. Residents were asked to let the water stagnate in the pipe
for at least 8 hours prior to sample collection, i.e., they could not use the sampled faucet. The first
sample collected was for the bacteriological tests. First-draw samples for the physical and
chemical parameters were collected immediately after. After collecting the first-draw samples, the
flushed samples were collected after at least 5 minutes of flushing. For each first-draw and flushed
samples, samples were unfiltered and filtered through 0.45-µm to examine the potential
occurrence and effect of particles. Preferably, samples were collected by utility staff; alternatively,
homeowners were asked to collect the samples. In either case, clear instructions were provided.
Samples were analyzed for ammonia, nitrite, nitrate, lead and copper release and other basic water
quality (pH, chlorine, alkalinity, temperature, HPC, MPN etc). The specifics of the analytical
methods were presented in earlier chapters.
RESULTS AND DISCUSSION
Utility Interviews
Twelve utilities filled out the questionnaire. These utilities represented different
geographic areas of the U.S. and included one participant from Canada (Figure 9.1; one respondent
requested to remain anonymous, and is thus not represented on the figure). They were of different
size, serving populations varying from 4,330 to 1,600,000 persons (Figure 9.2). The size of their
distribution system also varied: they had 1 to 11 POEs (average of 3 POEs), and 2 to 175 storage
tanks. The respondents’ total distribution system storage capacity is illustrated on Figure 9.3.
Eleven of the 12 participants were able to provide estimated proportion of various pipe
materials in their distribution system as a function of pipe diameter, and results are summarized in
Figures 9.4. Because a number of respondents were not able to distinguish between iron pipes that
were lined and unlined, no distinction was made between these two types of pipes. Likewise, no
distinction was made between polyvinyl chloride (PVC) and high-density polyethylene (HDPE)
pipes because a number of respondents did not have the ability to distinguish between the
prevalence of these materials. Results show that the majority of the pipes are 6 to 8 inches (15 to 20
cm) in diameter, and are mainly made of lined or unlined cast iron. Responses received regarding
the age of pipes in the distribution system are difficult to illustrate because respondents provided
either an average or a range of ages. Overall, results suggest that asbestos cement pipes appear to
be 40 to 50 years age on average, all of the cast iron pipes (lined or unlined) are older than 45
years, but all of the ductile iron pipes are younger than 50 years. Steel pipe ages vary from 0 to 80
years, and as expected, PVC and HDPE pipes are all less than 50 years age.
Regarding premise plumbing, nine of the 12 participants provided service pipe material
information. Results showed that copper was the material of choice, with 70-90% of the service
lines (average of 83%). Two respondents reported relatively high proportion of HDPE (25 to 30%
of the service lines), and three respondents reported lead service lines (15 to 18%). The percentage
of homes built before the lead solder ban in 1986 and served by the respondents is illustrated on
Figure 9.5.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 9: Utility Interview and Case Studies | 89
The respondents showed a varied experience with chloramination, as illustrated on
Figure 9.6. Only 2 of the 12 participants were periodically switching to free chlorine as distribution
system residual disinfectant, and they switched once per year for a 1-month (4-week) period, and
were targeting a free chlorine residual at the POE of the distribution system of 3.5 to 4.0 mg/L Cl2.
Nitrification Experience
Only one of the 12 respondents reported no incidence of nitrification in their distribution
system. 56% reported detection of significant nitrification once per season, although a fair number
of respondents (33%) reported observing nitrification once per week (Figure 9.7). Most
respondents have never monitored for nitrification in premise plumbing. Two of them stated that
they have monitored to at least some extent, and did not detect problems, and two others reported
that they have observed nitrification in premise plumbing on a seasonal basis.
Respondents were queried about the proportion of nitrification episodes that occur in
specific pipe materials and diameters in their distribution system. Only three respondents
answered this question. Two of them reported that 80 to 100% of the nitrification episodes
occurred in 6 to 12 inches unlined cast iron pipes, and the rest of the episodes in 6 to 12 inches
cement-mortar lined ductile iron pipes. The third respondent stated that 60 to 90% of their
nitrification episodes occurred in 6 to 8 inches asbestos cement pipes, and the remaining episodes
are unknown. None of the respondents were able to provide information regarding the pipe
material of the premise plumbing where nitrification has been detected, most likely because
utilities were not regularly monitoring for this issue, as mentioned above.
Respondents were asked to rate a number of methods used to monitor, prevent, and correct
nitrification; results are summarized on Figures 9.8, 9.9, and 9.10, respectively. The three
preferred methods to monitor for nitrification were decrease in total chlorine residual, increase in
nitrite concentration, and decrease in pH. Although respondents were specifically queried about
the frequency of monitoring for nitrification purposes, responses provided appear to be related to
general water quality monitoring (Figure 9.9). To prevent nitrification, the most common methods
reported are cycle, mix and clean storage reservoirs; flush the affected area of the distribution
system; increase water quality monitoring; and increase chloramine residual at the plant effluent
(Figure 9.10). Interestingly, a number of nitrification prevention methods were classified as
“essential”, but not widely used. These methods include implementation of a corrosion control
program, increase pH, flush the entire distribution system, decrease ammonia residual at the POE,
and use chlorination or chloramination booster stations. The most common methods used to
correct nitrification are to flush the affected area of the distribution system, increase chloramine
residual, and drain and clean storage reservoirs (Figure 9.11).
The wide use of increased chloramine residual to correct nitrification is a concern, because
utility experience has shown that increasing the chloramine dose does not correct nitrification as it
increases ammonia availability once chloramine degrades. As mentioned above, a number of
corrective methods were classified as “essential”, but not widely used: change distribution system
hydraulics (e.g., eliminate dead-ends), increase pH, modify Cl: NH3 ratio and decrease ammonia
residual at the POE.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
90 | Effect of Nitrification on Corrosion in the Distribution System
Corrosion Issues
All respondents were on reduced monitoring for compliance with the Lead and Copper
Rule (LCR), except one. Nine of the 12 respondents adjusted the pH to 7.2 to 9.7 for corrosion
control, typically using sodium hydroxide, although sodium carbonate and calcium oxide are used
by 3 of those 9 respondents. Six respondents used a corrosion inhibitor: 5 used a phosphate base
agent, and another used stannous chloride.
Only 2 of the 12 respondents were able link occurrence of nitrification with the presence of
corrosion phenomenon. For these respondents, red water was observed from cast iron pipes during
nitrification episodes.
Review of Existing Water Quality Data
Water quality data was received from 10 of the participating utilities, but data from only a
handful of utilities could be used because often times, nitrification is not monitored regularly, and
when nitrification indicators are analyzed, sampling sites or times differ from those used for LCR
compliance. Nitrification or corrosion indicators are practically never monitored at customer’s homes.
Portland Water Bureau (City of Portland), Oregon
The Portland Water Bureau (Portland) did not provide sufficient water quality data in their
distribution system to examine potential associations between nitrification and corrosion indicators.
However, Portland’s staff provided results of an earlier study with a similar goal. That study
involved sample collection at one location in Portland’s building; samples were collected before and
after an 8-hour period during which water was not used. The service lines examined were made of
copper. Samples were collected on six occasions at one site, and once at a second site.
Results confirmed that nitrification was occurring in the copper service lines during water
stagnation, as shown by increases in free ammonia concentration (increases ranged from 0.07 to
0.17 mg/L N) and nitrite concentration (from <0.005 to 0.036 mg/L N), and decrease in chlorine
residual (from 1 mg/L before stagnation to below detection limit). Portland’s staff hypothesized
that the increase in free ammonia concentration was due to the loss in chloramine residual. During
water stagnation, copper concentrations increased from 0.008 to 0.020 mg/L, but lead
concentrations remained low in all samples (<0.006 mg/L).
Bangor Water District, Maine
Bangor Water District (Maine) provided nitrification and corrosion data, but the sampling
sites were different for both sets of data. The data were thus analyzed separately.
For nitrification, monthly averages of the 40 Total Coliform Rule (TCR) sampling sites for
nitrite, total chlorine and pH from 2003 through 2007 were provided. These data show that from
January to June, nitrite concentrations remain below 0.02 mg/L N, but increase significantly during
the summer months (up to 0.13 mg/L), indicating severe nitrification occurrence. Then the nitrite
levels decrease in October through December. Total chlorine residuals follow a similar pattern, with
residuals greater than ~1.8 mg/L Cl2 from January to June, and then the concentrations dropped
significantly during the summer months (June through October) (down to below 1 mg/L). In
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 9: Utility Interview and Case Studies | 91
October, the chlorine levels gradually increase to > 1.8 mg/L. The pH values in the distribution
system remain fairly consistent at ~9.5 (data not shown).
The 90th percentile for lead and copper data measured in the distribution system from 1992
through 2007, and detailed data for June 2007 were provided. Figure 9.12 shows that the 90th
percentile of lead concentrations were generally below 8 µg/L, although the maximum 90th
percentile measured was 15 µg/L in June 2002, which is just at the USEPA action level. Detailed
data for June 2007 (Figure 9.13) show that even though the 90th percentile lead concentration was
well below the action level, very high lead levels (up to 36 µg/L) were measured for a few samples
in the distribution system. As for copper, 90th percentile concentrations were well below the
USEPA’s action level of 1.30 mg/L, with a maximum value of 0.28 mg/L in June 2007
(Figure 9.14). Similar observations can be made from the detailed data of June 2007 (Figure 9.15),
where the maximum copper concentration measured was 0.63 mg/L.
Pinellas County Utilities, Florida
After conversion to chloramines in 2002, Pinellas County Utilities (Pinellas) observed
increases in coliform counts and HPCs. Nitrification in storage reservoirs began within four
months of the conversion, despite operational changes to increase water turnover. Nitrification in
reservoirs was accompanied by significant drops in pH, DO, and alkalinity. However, there was no
significant increase in lead and copper concentrations.
Pinellas provided data collected at three POEs and a number of distribution system
sampling sites and storage reservoirs supplied by these POEs. Data were collected from January
2004 through September 2007.
Very mild nitrification occurred in the section of the distribution system supplied by the
Keller 1 POE most of the time (90th percentile nitrite < 0.03 mg/L N) (Figure 9.16). However, the
maximum nitrite concentrations (0.3 mg/L-N) suggest that severe nitrification has happened at
times at the Seaview sampling site, with nitrite concentration as high as 0.30 mg/L N. Lead and
copper concentrations did not correlate with nitrification, as the levels were higher at the POE,
then decrease in the distribution system.
For Keller 2 POE and corresponding distribution system sampling sites, nitrification did
not happen during the study period, with maximum nitrite concentration of 0.03 mg/L N measured
in the distribution system. Most metals showed stable concentrations in the distribution system,
with median concentrations of 0 to 0.2 µg/L for lead, and 0 to 0.01 mg/L for copper.
In Cypress 60 POE and corresponding distribution system sampling sites, nitrification
appears to have occurred only at Fort De Soto sampling site, although nitrite concentrations
remained low, with a maximum of 0.06 mg/L N. Again, metal concentrations were not affected by
nitrification, stable levels were observed at different locations, with median lead concentrations of
0 to 0.2 µg/L, and median copper concentrations of 0 to 0.01 mg/L.
To date, overall, no correlation was found between nitrification and metal release in
Pinellas County, FL.
City of Greenville, North Carolina
Historically, the City of Greenville (Greenville) has had very few problems with lead,
although a high profile problems briefly arose from 2004-2006. During this incident, Greenville’s
staff noticed that particulate lead and occurrences with high lead were highly isolated. Even within
©2010 Water Research Foundation. ALL RIGHTS RESERVED
92 | Effect of Nitrification on Corrosion in the Distribution System
the same apartment, one faucet could produce water with very high lead concentration whereas
another faucet may not show high levels. Very low pH values (in the order of 3.0 and even less)
have been measured at the surface of the leaded solder due to galvanic corrosion, and it is possible
that these low pHs are the cause of lead solder corrosion in Greenville’s water. It was also
considered possible that nitrification was contributing to lower pHs as well.
In an attempt to solve the problem, Greenville switched from polyaluminum chloride
(PACl) to alum. Lead values appeared to decrease significantly in a number of homes shortly after
the switch, and within a few months they dropped below the action level. It seems unlikely that a
switch to alum from PACl could have changed the incidence of nitrification, so the practical data
seemingly confirm that nitrification was not a factor.
To further examine these issues, Greenville provided nitrification and corrosion related
data collected at the POE of Greenville’s distribution system. Because nitrite data were not
provided (only the sum of nitrite and nitrate concentrations were provided), it is difficult to assess
whether nitrification occurred. The higher concentrations of nitrite + nitrate could be due to
naturally occurring nitrate in the water. Although the median lead concentration was 0 µg/L, the
90th percentile was higher than the action level (23 µg/L), and the maximum concentration was
very high (2.62 mg/L). Copper concentrations were generally low, with median and maximum
concentrations of 0.09 and 1.82 mg/L, respectively. The minimum, median and maximum ORP
were 3, 468 and 746, respectively.
Distribution System Sampling
Samples were collected in six distribution systems.
City of St. Paul, Minnesota
In the past, the City of St. Paul (St. Paul) experienced occurrences of higher concentrations
of lead and copper, but these problems were resolved by using tin chloride. St. Paul is now on
reduced monitoring for compliance with the TCR. The use of tin chloride also decreased the
occurrence of nitrification, although distribution system water quality data suggest that
nitrification may be occurring at a number of locations.
Samples were collected at the POE to the distribution system, and at three homes.
Significant nitrification occurred in two of the sampling homes (Nos. 4 and 273-Figure 9.17). For
site No. 4, nitrification was only observed in the first draw sample but not the flush sample,
indicating that water stagnation or the service line/premise plumbing material was the triggering
factor. At site No. 273, evidence of nitrification lasted through the flushing period, indicating that
it was occurring in the water main. No significant nitrification occurred at site No. 223.
Correspondingly, at site No. 4 and No. 273, nitrifier BART tests indicated the presence of nitrifiers
(1000 cfu/ml), while no nitrifiers were detected at POE and Site No. 223.
As expected, at site No. 273, nitrification decreased pH and DO, and accelerated
disinfectant decay (Figure 9.18). pH and chlorine was also decreased at site No. 4, but at a lesser
extent (Figure 9.18), which was consistent with its lower ammonia loss (Figure 9.17). However,
HPC was not significantly affected by nitrification; low levels of HPC (≤140 cfu/ml) were
detected at all sampling locations.
The service lines at all three locations were made of lead piping, so the detection of lead is
not surprising (Figure 9.19). The interesting point is that higher total lead levels were present at
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 9: Utility Interview and Case Studies | 93
site No. 4 and site No. 273 where significant nitrification occurred. Copper concentrations were
also high at Sites Nos. 4, but not site No. 273.
As discussed earlier, nitrification appeared to be most severe at Site No. 273, and also
occurred at Site No. 4, but not at Site No. 223. Interestingly, lower levels of tin were observed at
Site No. 273, whichis believed to at least it’s true in this system inhibit nitrification in this system.
Lower levels of iron, copper, zinc and magnesium were also observed at Site No. 273, and these
elements have been shown to be toxic to nitrification at high levels (Zhang et al. 2009a). Since the
three sampling sites have the same pipe materials, it is possible that the difference of these nutrient
levels might account for the different nitrification extent (Table 9.2). Also, higher levels of iron
were associated with Site No. 273, indicating the presence of iron material in the preceding
distribution system, which is likely to enhance nitrification occurrence.
Anonymous Utility
This utility has experienced nitrification, and is on reduced monitoring for compliance with
the LCR.
Samples were collected at the entry point to the distribution system (water treatment plant
effluent), and at three homes. Nitrification occurred during stagnation (1st draw sample) at all three
homes, and most significantly at Home No. 1 (Figure 9.20). Home No. 2 had more incomplete
nitrification (higher nitrite) than the other two locations. Without stagnation (flushed sample),
nitrification persisted in home No. 1, but was reduced in home No. 2 and No. 3. Different from
ammonia loss trend, nitrifier MPN was highest at Home No. 2 (170000 MPN/100 ml). Low levels
of nitrifier MPN were detected at home No. 1 and No. 3 (200-300 MPN/100 ml). Nitrifier MPN at
the entry point was non-detectable. BART tests also showed the same trend as MPN test, with
higher levels in No. 2 than No. 1 and No. 3.
The water quality change by nitrification corresponds well with nitrification activity, in
that Home No. 1 had the largest pH, DO and total chlorine decrease (Figure 9.21). At Home No. 1
and No. 2, HPC was increased by two orders of magnitude, while at Home No. 3, HPC was slightly
decreased. It’s interesting to point out that at all three sampling homes, the chlorine residual in first
draw samples were in the form of chloramines (free chlorine < 0.4), while in flush samples,
60-70% of the total chlorine was in the form of free chlorine.
Higher lead, copper and zinc concentrations were observed in the first draw samples at all
three homes, with the highest levels in Home No. 1 (Figure 9.22). This is consistent with the more
severe nitrification and lower pH, DO, total chlorine in Home No. 1, indicating a clear correlation
between nitrification and metal release.
As discussed earlier, nitrification appeared to be most severe at Home No. 1. In terms of
nutrient differences, home No. 1had higher potassium (7000 ppb vs. 4000 ppb), silicate (1900 ppb
vs. 1400 ppb) and lower calcium (32000 ppb vs. 25000 ppb). However, the correlation between
these nutrient levels and nitrification extent cannot be confirmed.
Irvine Ranch Water District, California
At the Irvine Ranch Water District (IRWD), samples were collected at two entry points
(surface water and ground water sources) and three homes on November 1, 2007. Nitrification was
occurring at all three sites, as indicated by ammonia decrease (Figure 9.23). Different from other
utilities, the most significant nitrification was occurring in the flush sample at Site B. This was
©2010 Water Research Foundation. ALL RIGHTS RESERVED
94 | Effect of Nitrification on Corrosion in the Distribution System
possibly because nitrifiers were introduced from the POE water, rather than developed during
stagnation. Indeed, in POE-well water, nitrifiers were detected by both nitrifier MPN (2200
MPN/100 ml) and BART test (1000 cfu/ml).
Corresponding to the ammonia loss, a significantly lower chlorine residual was observed in
the flush sample at Site B (Figure 9.24). The flush sample at Site B also had lower DO and pH,
although the overall DO (0.6 mg/L) and pH drop (0.2 unit) was small compared to the other utility.
Bacterial counts were also performed in the first-draw samples collected from the homes, and in
the flushed samples collected at the POEs. HPCs were surprisingly high at the POEs to the
distribution system (8500 and 78000 cfu/ml for the well and surface water); these concentrations
were even higher than those measured in the first draw samples of the three homes studied
(200-3000 cfu/ml). Using the culturing technique, nitrifiers were also detected at high
concentration at the well head.
All three service lines were made of copper piping. The difference between Sites A and B
could be associated with the presence of higher TOC concentration at Site B, which tends to
decompose chloramines at a faster rate. The fact that Site B had the longest service line may have
also explained the more severe nitrification and increased chloramine decay at that site.
In terms of metal release, high lead concentrations were observed at Site C, up to 175 ppb
was observed in the first draw filtered sample (Table 9.3). Higher copper and zinc levels were
present in the first draw samples of Site B and Site C, indicating the possible release of these
metals from brass materials. Site B flush sample also had slightly higher iron levels than the other
samples, which might be related to its higher nitrification activity.
Portland Water Bureau (City of Portland), Oregon
At the Portland Water Bureau (Portland), samples were collected at the POE to the
distribution system, at one of the employees’ home (Site No. 1; the service line is made of copper
with partially galvanized and lead solder), at a groundwater (GW) pump station that has shown
signs of nitrification (Site No. 2; the service line was built in 1984 and is made of copper pipe), and
at the water control center (Site No. 3; the service line was built in 1992 and is made of copper).
Sampling was conducted on November 20, 2007. Nitrification occurred at all their sites, with Site
No. 2 had the most significant nitrification (Figure 9.25). Correspondingly, nitrifier BART test
detected 1000 cfu/ml nitrifiers in Site No. 2 and 1000 cfu/ml nitrifiers in Site No. 1 and No. 3.
However, nitrifier MPN test did not detect nitrifiers at any of the sampling locations.
Corresponding to nitrification activity, at all three sites, DO was reduced from 10 mg/L
down to 7- 9 mg/L, with the lowest DO (7 mg/L) in the No. 2 flush sample. Compared to POE, pH
was generally reduced by 0-0.5 unit (Figure 9.26). However, at Site No.2 where the most
significant nitrification occurred, pH was increased by 1.1-1.9 unit. The increase of pH at this site
could be possibly due to the corrosion of concrete lined pipe. For chlorine levels, at all three sites,
there was non-detectable chlorine in the first draw samples. In the flush sample, Site No. 3 had the
lowest chlorine (0.4 mg/L) compared to the other two sites (0.9 and 1 mg/L). Bacterial counts were
performed in the first-draw samples collected from the homes, but in the flushed samples collected
at the POE. Very high HPC was detected at Site No. 2, where nitrification appeared to be the most
severe. Lowest HPC appeared to be at Site No. 3 where nitrification was lowest.
When comparing with the POE, copper and lead levels were significantly higher at Site No.
3 (Figure 9.27). Iron levels in the first draw samples at all three sites were also higher than POE.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 9: Utility Interview and Case Studies | 95
As mentioned above, nitrification appeared more severe at Site No. 2, but was detected also
at Sites Nos. 1 and 3. When examining the metal concentrations in light of these results, calcium
and iron concentrations were higher at Site No. 2. Considering the high pH at Site No. 2, this might
indicate the presence of concrete lined iron pipe.
Bangor Water District, Maine
Water treated by Bangor Water District (Bangor) first flows through a long transmission
main, then through the City of Bangor, and lastly to a consecutive system – Hampden Water
District, which results are presented immediately below.
Samples were collected from the POE to the distribution system and two homes on
November 13, 2007. Site No. 1 was fed by two copper service lines coming off of a 10-inch
asbestos cement main. Site No. 2 was supplied by a copper service line coming off of an 8-inch
ductile iron main. Both sites showed similar extent of nitrification in the first draw sample (Figure
9.28). However, at Site No. 1, nitrification lasted through flushing, but at Site No. 2, nitrification
was largely reduced after flushing. For first draw samples, higher nitrifier level was detected at
Site No. 1 (1100 MPN/ml) than POE and Site No. 2 (20 and 70 MPN/100 ml). However, nitrifier
BART test only detected the presence of nitrifiers at Site No. 2 (1000 cfu/ml).
Corresponding to the nitrification activity, pH was reduced at both sites, but at a lesser
extent in the flush sample at Site No. 2 (Figure 9.29). Total chlorine was also decayed to very low
levels, with Site No. 1 decayed more significantly. At Site No. 1, DO was reduced from 12 mg/L at
POE to below 10 mg/L in both first draw and flush samples. However, no DO reduction was
observed at Site No. 2. In terms of HPC, very low levels were detected (< 100 cfu/ml) at POE.
High levels of HPC were observed at Site No. 1 (7300 cfu/ml) and No. 2 (16500 cfu/ml), which is
correlated with the nitrification occurrence at these sites.
High lead levels were observed in the first draw samples of Site No. 2. Site No. 1 also had
slightly elevated lead levels (Figure 9.30). High copper levels were observed at both sites as well.
Nitrification and its resulted pH and total chlorine reduction are suspected to be responsible for
these elevated metal concentrations.
Hampden Water District, Maine
Hampden Water District (Hampden) is a retailer system of Bangor. Hampden distributes
Bangor’s water without further treatment, except for additional disinfection (booster
chloramination). Samples were collected at POE to Hampden’s distribution system, and at three
homes on November 28, 2007.
Complete nitrification was observed at site # 1 and # 2, as indicated by 100% conversion of
ammonia to nitrate. Surprisingly, the flush samples at these two sites showed either higher (#1) or
similar (#2) activity than the first draw sample (Figure 9.31), indicating nitrification was not
caused by stagnation. Site # 3 also had slight nitrification, but only in the flush sample.
Considering that the service line at the three sites are made of cast iron, plastic, and copper pipe,
respectively, it is not surprising to see the higher nitrification activity at site # 1 and # 2.
Water quality changes at the three sites corresponds very well with the nitrification activity
(Figure 9.32). For example, at site # 1 and # 2, where complete nitrification occurred, pH was
reduced by 1.5 unit, chlorine was completely decayed, and HPC was increased to 105 cfu/ml. At
site # 3, pH, chlorine decay and HPC was less affected.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
96 | Effect of Nitrification on Corrosion in the Distribution System
Existing data in Hampden indicated a number of high lead occurrences. In this sampling
event, high copper, zinc and lead levels were observed at site # 2 (Table 9.4), with up to 657 ppb
lead observed in the first draw sample. Since the service line at site # 2 was made of plastic, these
high metal concentrations might originate from the corrosion of brass faucet. Site # 3 also had
higher lead (11 ppb) than the POE, but the levels were below the action limit.
CONCLUSIONS
Utility Interviews
Twelve utilities of various sizes and representing different geographic areas filled out the
questionnaire. The majority of the pipes found in the respondents’ distribution systems was 6 to
8 inches (15 to 20 cm) in diameter, and are mainly made of lined or unlined cast iron.
Most respondents have experienced nitrification once per season, although a fair number of
respondents (33%) reported observing nitrification once per week. Little information was
available regarding the occurrence of nitrification in specific pipe materials. Two respondents
reported that 80 to 100% of the nitrification episodes occurred in 6 to 12 inches unlined cast iron
pipes, and the rest of the episodes in 6 to 12 inches cement-mortar lined ductile iron pipes. One
respondent stated that 60 to 90% of their nitrification episodes occurred in 6 to 8 inches asbestos
cement pipes, and the remaining episodes are unknown. None of the respondents were able to
provide information regarding the pipe material of the premise plumbing where nitrification has
been detected.
All respondents were on reduced monitor for compliance with the LCR, except one. Only
one of the respondents did not control corrosion; most of the others adjusted the pH using sodium
hydroxide or alternatively sodium carbonate. Six of the twelve respondents used a corrosion
inhibitor, mainly phosphate base agents. Only 2 of the 12 respondents were able to discuss
occurrences of nitrification in the presence of corrosion. For these respondents, red water is
observed in cast iron pipes during these episodes.
Review of Existing Water Quality Data
Among the utility providing existing water quality data, many utilities did not monitor
nitrification regularly. In the few utilities that monitor nitrification, nitrification sampling sites
differ from those used for LCR compliance. For example, in Bangor Water District (Maine), both
significant nitrification and high lead events were observed, but no correlation can be drawn due to
the different sampling sites for nitrification and lead and copper rule monitoring. In one utility
(Portland, OR), limited data has indicated nitrification accompanying elevated lead and copper
levels, but this correlation cannot be confirmed.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 9: Utility Interview and Case Studies | 97
Distribution System Samplings
Samples collected from the homes in the six distribution systems all showed signs of
nitrification. Within a system, the extent of nitrification in different sampling homes varied. Many
factors might contribute to the nitrification difference. For example, in St. Paul, MN, tin chloride
was used to inhibit nitrification in this system, and the home with the worst nitrification had the
lowest level of tin. In Hampden, MN, higher levels of nitrification were observed in the homes
with iron and plastic service line than that with copper service line, indicating a potential
relationship between the pipe materials and nitrification extent.
The water quality change by nitrification corresponds very well with nitrification activity.
In most systems, pH and DO were decreased most significantly at the homes with the highest level
of nitrification. Acceleration of chlorine decay and increase of HPC were often observed at these
sites. In general, within a system, copper and lead levels were correlated with the nitrification
extent. That is, the worst lead, copper and zinc contamination event occurred in homes with worst
nitrification situation, except for one utility-Portland, OR. At this utility, the home with the highest
nitrification activity increased the pH by 1.1 unit instead of decreasing pH; and lead and copper
contamination was less than the other homes with lower nitrification. This was possibly due to the
presence of concrete lined iron pipe, as indicated by high calcium and iron concentrations.
So, overall, the distribution system sampling strongly suggested the correlation between
nitrification, water quality change (pH reduction, HPC increase, etc) and corrosion.
FIGURES AND TABLES
Figure 9.1 Geographic location of the respondents
©2010 Water Research Foundation. ALL RIGHTS RESERVED
98 | Effect of Nitrification on Corrosion in the Distribution System
≥1,000,000
persons
17%
<100,000
persons
25%
500,000999,999
persons
17%
100,000499,999
persons
41%
Figure 9.2 Number of persons served by each respondent
≥500MG
8%
<25 MG
25%
200-499 MG
25%
50-199 MG
17%
25-49 MG
25%
Figure 9.3 Total storage capacity in the distribution system
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Proportion of Pipes
Chapter 9: Utility Interview and Case Studies | 99
50%
45%
40%
35%
30%
25%
20%
15%
10%
5%
0%
>12 inches
10-12 inches
6-8 inches
≤4 inches
80%
Proportion of Pipes
70%
60%
Other or Unknown
50%
PVC and HDPE
40%
Steel
Concrete
30%
Ductile Iron
20%
Cast Iron
10%
Asbestos Cement
0%
≤4 inches 6-8 inches
10-12
inches
>12 inches
Figure 9.4 Proportion of pipe material for specific pipe diameter categories in the
distribution system
©2010 Water Research Foundation. ALL RIGHTS RESERVED
100 | Effect of Nitrification on Corrosion in the Distribution System
91-100%
11%
31-40%
11%
51-60%
22%
81-90%
22%
71-80%
34%
Figure 9.5 Percentage of homes built before 1986 and served by the respondents
≥50 years
25%
<10 years
25%
40-49 years
0%
30-39 years
8%
20-29 years
0%
10-19 years
42%
Figure 9.6 Number of year of experience with chloramination
Once per
week
33%
Once per
season
56%
Once per 2
weeks
11%
Figure 9.7 Frequency of nitrification occurrences in the respondents’ distribution system
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 9: Utility Interview and Case Studies | 101
0%
20%
40%
60%
Decrease in Total Chlorine Residual
Increase in Nitrite
Decrease in pH
Increase in Coliforms
Increase in HPC-Agar
Temperature
Increase in Nitrate
Decrease in Free Ammonia Residual
Increase in Free Ammonia Residual
Change in Cl2:NH3-N Ratio
Increase in HPC-R2A
Change in Chloramine Speciation
Decrease in Total Ammonia…
Increase in Total Ammonia Residual
Decrease in Dissolved Oxygen
Increase in AOB
Increase in AOC
80%
100%
Not effective
Slightly effective
Effective
Very effective
Essential
Figure 9.8 Methods used to monitor for nitrification
0%
20%
40%
60%
Increase in Coliforms
Temperature
Increase in HPC-Agar
Increase in HPC-R2A
Decrease in Total Cl2 Residual
Decrease in pH
Change in Cl2:NH3-N Ratio
Decrease in Free Ammonia Residual
Increase in Free Ammonia Residual
Increase in Nitrite
Increase in Nitrate
Decrease in Total Ammonia Residual
Increase in Total Ammonia Residual
Change in Chloramine Speciation
Decrease in Dissolved Oxygen
Increase in AOB
Increase in AOC
80%
100%
Every 4 hours
Daily
Weekly
Every 2 weeks
Monthly
Quarterly
Annually
Figure 9.9 Frequency of monitoring
©2010 Water Research Foundation. ALL RIGHTS RESERVED
102 | Effect of Nitrification on Corrosion in the Distribution System
0%
20%
40%
60%
Cycle Tank
Flush Area
Tank Mixing
Increase Monitoring
Increase Chloramine
Corrosion Control
Increase pH
Decrease Ammonia
Flush Entire DS
Chloramine Booster
Chlorine Booster
Modify Cl2:NH3-N
Switch to Chlorine
Clean Tank
Change Hydraulics
Blend
Chloramine Stability
Modify Treatment
Clean DS
Chlorine Dioxide
Chlorite
80%
100%
Not effective
Slightly effective
Effective
Very effective
Essential
Figure 9.10 Methods used to prevent for nitrification
0%
20%
40%
60%
80%
100%
Flush Area
Increase Chloramine
Drain Tank
Clean Tank
Breakpoint Area
Change Hydraulics
Increase pH
Decrease Ammonia
Modify Cl2:NH3-N
Not effective
Breakpoint Entire…
Slightly effective
Flush Entire DS
Effective
Chlorine Dioxide
Very effective
Essential
Chlorite
Figure 9.11 Methods used to correct nitrification
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 9: Utility Interview and Case Studies | 103
20
18
16
EPA Action Level = 15 µg/L
15
Lead (µg/L)
14
12
12
10
10
8
6
8
7
6
6
6
6
4
2
0
Dec-92
Jun-93
Jun-94
Jun-95
Jun-96
Jun-99
Jun-02
Jun-04
Jun-07
1
2
3
4
5
6
7
8
9
Figure 9.12 90th percentile of lead concentrations measured in Bangor’s distribution system
40
35
30
Lead (µg/L)
25
20
15
EPA Action Level = 15 µg/L
10
90th Percentile
Result = 6 µg/L
5
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
Sample Site Number
Figure 9.13 Lead concentrations measured in Bangor’s distribution system in June 2007
©2010 Water Research Foundation. ALL RIGHTS RESERVED
104 | Effect of Nitrification on Corrosion in the Distribution System
1.40
EPA Action Level = 1.30 mg/L
1.20
Copper (mg/L)
1.00
0.80
0.60
0.40
0.28
0.14
0.18
Jun-99
Jun-02
Jun-04
Jun-07
6
7
8
9
0.16
0.20
0.09
0.10
0.08
0.08
Dec-92
Jun-93
Jun-94
Jun-95
Jun-96
1
2
3
4
5
0.08
0.00
Figure 9.14 90th percentile of copper concentrations measured in Bangor’s distribution
system
1.40
EPA Action Level = 1.30 mg/L
1.20
Copper (mg/L)
1.00
0.80
0.60
90th Percentile
Result = 0.28 mg/L
0.40
0.20
0.00
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
Sample Site Number
Figure 9.15 Copper concentrations measured in Bangor’s distribution system in June 2007
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Lead (mg/L)
Chapter 9: Utility Interview and Case Studies | 105
0.010
0.009
0.008
0.007
0.006
0.005
0.004
0.003
0.002
0.001
0.000
Keller-1 POE
US-19 Old
Post
US-19 Neb
Seaview
0.035
Copper (mg/L)
0.030
0.025
0.020
0.015
0.010
0.005
0.000
Keller-1 POE
US-19 Old
Post
US-19 Neb
Seaview
Figure 9.16 Lead and copper levels at the Pinellas’ Keller 1 POE and corresponding
distribution system sampling locations
Note: the bottom and top of the boxes represent the 10th and 90th percentile values,
respectively, while the end of the lines represent the minimum and maximum values
Figure 9.17 Nitrification activity in first draw samples in St. Paul, MN
©2010 Water Research Foundation. ALL RIGHTS RESERVED
106 | Effect of Nitrification on Corrosion in the Distribution System
Figure 9.18 Water quality in first draw samples in St. Paul, MN
Note: Free chlorine at all locations were below 0.5 mg/L.
Figure 9.19 Lead levels in St. Paul, MN
Figure 9.20 Nitrification activity in the anonymous utility
Figure 9.21 Water quality in first draw samples in the anonymous utility
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 9: Utility Interview and Case Studies | 107
Figure 9.22: Lead, copper and zinc levels in the anonymous utility
©2010 Water Research Foundation. ALL RIGHTS RESERVED
108 | Effect of Nitrification on Corrosion in the Distribution System
Figure 9.23 Nitrification activity at IRWD utility
Note: For Site A and C, first draw data were presented, flush data were similar to first draw.
Because the POEs were continuously flowing, only flushed samples could be collected.
Figure 9.24 Chlorine decay at IRWD utility
Note: For Site A, first draw data were presented; flush data were similar to first draw. No
data available at site C due to resident sampling.
Figure 9.25 Nitrification activities in first draw samples at Portland utility
Note: Nitrification activity was slightly lower in flush samples.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 9: Utility Interview and Case Studies | 109
Figure 9.26 pH and HPC at Portland utility
Figure 9.27 Lead and copper levels at Portland utility
©2010 Water Research Foundation. ALL RIGHTS RESERVED
110 | Effect of Nitrification on Corrosion in the Distribution System
Figure 9.28 Nitrification activity in the first draw samples in Bangor utility
Figure 9.29 pH and Total chlorine in the first draw samples in Bangor utility
Figure 9.30 Lead and copper levels in Bangor utility
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Chapter 9: Utility Interview and Case Studies | 111
Figure 9.31 pH and Total chlorine in the first draw samples in Hampden utility
Note: For Site # 2 and # 3, data shown were first drawn sample.
Figure 9.32 Water quality change in Hampden Water District
©2010 Water Research Foundation. ALL RIGHTS RESERVED
112 | Effect of Nitrification on Corrosion in the Distribution System
Table 9.1
Utilities that participated in the distribution system samplings
Participating Utility
Sampling Date
City of St. Paul, Minnesota
October 30, 2007
Irvine Ranch Water District, California
November 1, 2007
Portland Water Bureau (City of Portland), Oregon
November 20, 2007
Bangor Water District, Maine
October 21, 2007;
Re-sampled on November 13, 2007
Hampden Water District, Maine
November 28, 2007
Anonymous utility
November 29, 2007
Table 9.2
Nutrient levels at different sampling locations
1st draw fIltered
1st draw ufiltered
POE
flush filtered
flush ufiltered
1st draw fIltered
1st draw ufiltered
Resident #4
flush filtered
flush ufiltered
1st draw fIltered
1st draw ufiltered
Resident # 273
flush filtered
flush ufiltered
1st draw fIltered
1st draw ufiltered
Resident #223
flush filtered
flush ufiltered
Fe, ppb
4
6
2
13
25
3
8
8
26
188
58
133
8
124
8
13
Cu,ppb
1
1
1
1
55
771
37
4
10
15
7
10
11
31
2
4
Zn, ppb
4
4
5
5
64
8
58
4
13
21
13
12
66
269
7
10
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Mg, ppb
8035
8040
8025
8178
6144
4325
6224
6614
6494
6433
6473
6461
7501
7459
7844
7816
Sn, ppb
66
83
69
89
29
117
7
14
8
31
4
31
7
137
44
57
Chapter 9: Utility Interview and Case Studies | 113
Table 9.3
Metal Release at IRWD utility
flush filtered
POE-Surface
flush unfiltered
flush filtered
POE-Well
flush unfiltered
1st draw filtered
1st draw unfiltered
A
flush filtered
flush unfiltered
1st draw filtered
1st draw unfiltered
B
flush filtered
flush unfiltered
1st draw filtered
1st draw unfiltered
C
flush filtered
flush unfiltered
Fe, ppb
10
5
6
10
92
7
14
3
2
6
29
33
0.0
2.1
0.8
2.9
Cu, ppb
1
2
12
15
55
65
16
19
265
311
26
39
140
164
10
21
Zn, ppb
8
6
6
6
15
18
9
9
206
227
8
8
114
122
13
20
Table 9.4
Metal Release in Hampden Water District
POE
#1
#2
#3
filtered
unflitered
filtered
unflitered
filtered
unflitered
filtered
unflitered
Cu, ppb
32
59
39
81
402
444
157
187
Zn, ppb
14
59
7
12
175
168
10
24
Pb, ppb
4
7
1
1
12
658
8
12
Note: data shown were for first draw samples.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Pb, ppb
0.5
0.9
0.7
1.2
0.5
1.1
0.7
1.0
1.6
2.4
0.5
1.5
175
47
23
33
114 | Effect of Nitrification on Corrosion in the Distribution System
©2010 Water Research Foundation. ALL RIGHTS RESERVED
APPENDIX A
NITRIFICATION IN DRINKING WATER SYSTEMS
Yan Zhang, Nancy Love, and Marc Edwards
Reprinted with permission from Critical Review in Environmental Science & Technology, 39(3):
153-208. Copyright 2009 Taylor and Francis
ABSTRACT
Nitrification is increasingly a concern in United States potable water distribution systems.
This paper reviews research on nitrification as it relates to the ammonia levels and unique
environments present in potable water distribution systems. Factors affecting nitrification
occurrence, nitrification impacts on water quality and corrosion, nitrification monitoring and
control methods are emphasized. The potential role of nitrogen cycling via coupled microbial and
electrochemical reactions is also described.
INTRODUCTION
The United States Environmental Protection Agency (USEPA) Stage 1 and Stage 2
Disinfectants and Disinfection By-Products Rule (D/DBPR) has been actively negotiated since
1992. The use of combined chlorine (e.g., chlorine with ammonia or chloramine) rather than free
chlorine at water utilities is often the lowest cost means of complying with D/DBPR. About 30%
of surface water treatment plants currently use chloramines versus 20% in 1990, and as many as
40-65% of surface water treatment plants might ultimately use chloramines after implementation
of Stage 2 (Seidel et al. 2005; USEPA 2005)
In addition to reduced formation of Disinfectant By-Products (DBP), use of chloramines
produces less chlorinous and chloro-phenolic tastes and odors relative to chlorine (Feben 1935).
Chloramines are also generally considered to be much more persistent in water than free chlorine
(Neden et al. 1992b; Norton and LeChevallier 1997). However, recent laboratory studies and
practical utility experience have revealed that chloramines can sometimes decay as fast as chlorine
(Powell 2004; Zhang 2005; Zhang and Edwards 2007). Chloramines are also a much less effective
disinfectant when compared to the same concentration of free chlorine, but the D/DBPR was
crafted to ensure "that the reduction of potential health hazards of DBPs does not compromise
microbial protection (Federal-Register 2003).” The lower disinfectant efficacy of chloramines is
partly offset by use of higher doses (compared to chlorine) and its greater stability (Neden et al.
1992b; Norton and LeChevallier 1997).
The ammonia released during chloramine decay can trigger nitrification incidence and
reduced pH or higher nitrite. Nitrification problems in drinking water systems and their link to
chloramination were identified as early as the 1930’s (Feben 1935; Hulbert 1933). A recent
telephone survey indicated that two thirds of medium and large US systems that chloraminate
experience nitrification to some degree (Wilczak et al. 1996). In Australia, using
most-probable-number (MPN) procedure, nitrifying bacteria were detected in 64% of samples
from five chloraminated water supplies and in 21% of samples containing more than 5 mg/L of
monochloramine (Cunliffe 1991).
115
©2010 Water Research Foundation. ALL RIGHTS RESERVED
116 | Effect of Nitrification on Corrosion in the Distribution System
Nitrification and nitrifying bacteria are also widespread in drinking water systems in
Finland (boreal region). Fifteen drinking water distribution systems were tested for
ammonia-oxidizing bacteria (AOB) and nitrite-oxidizing bacteria (NOB) using MPN and
Nitrification Potential methods, both AOB and NOB were detected at the end of the distribution
systems in at least eleven systems including non-disinfected and chlorinated systems (Lipponen et
al. 2002).
NITRIFICATION AND PHYSIOLOGY OF NITRIFYING BACTERIA
Nitrification is usually accomplished through a two-step microbiological process. The first
step is the oxidation of ammonia to nitrite by ammonia-oxidizing bacteria (AOB), also known as
ammonia oxidizers. The stoichiometry of this reaction, assuming a biomass yield of 0.2 mg
biomass formed per mg ammonia oxidized on a theoretical oxygen demand basis
(Grady et al. 1999), is:
NH4+ + 1.9 O2 + 0.069 CO2 + 0.0172 HCO3H2O + 1.97 H+
0.0172 C5H7O2N + 0.983 NO2- + 0.966
(A.1)
The second step in nitrification is the oxidation of nitrite to nitrate carried out by
nitrite-oxidizing bacteria (NOB), also known as nitrite oxidizers. The stoichiometry of this
reaction, assuming a biomass yield of 0.1 mg biomass formed per mg nitrite oxidized on a
theoretical oxygen demand basis (Grady et al. 1999), is:
NO2- + 0.00875 NH4+ + 0.035CO2 + 0.00875 HCO3- + 0.456 O2 + 0.00875 H2O
C5H7O2N + 1.0 NO3(A.3)
0.00875
Ammonia oxidizers have been reclassified based on phylogenetic methods. AOB relevant
to freshwater or low salinity systems fall into the Nitrosomonas (including ‘Nitrosococcus
mobilis’) and Nitrosospira (including Nitrosolobus and Nitrosovibrio) lineages and are members
of the β subclass of the proteobacteria (Purkhold et al. 2003). The most thorougly studied ammonia
oxidizer is Nitrosomonas europaea, because it has been available as a pure culture for more than
50 years (Engel and Alexander 1958; Lewis and Pramer 1958; Meiklejohn 1950). The rapid
development of more reliable, culture-independent methods has allowed researchers to identify
which bacterial species are present in natural and engineered systems, although generalizations as
to which strains predominate in which systems have not been successfully established (Geets et al.
2006). For example, while Nitrosomonas europaea has been found to be the predominant AOB in
some wastewater treatment biofilm systems (Gieseke et al. 2003), Nitrosomonas oligotropha
(Lydmark et al. 2006) and bacteria of the genus Nitrosospira (Coskuner and Curtis 2002) have also
been found to predominate in biofilm and activated sludge bioreactors, respectively. Using culture
independent analysis, Regan et al. (2003) found Nitrosomonas oligotropha to predominate in
chloraminated systems whereas Eichler et al. (2006) found Nitrosospira briensis to predominate in
chlorinated distribution system water. Although these cultures are of different genera, both are
known to be among the AOB that have high affinity for ammonia, thereby enhancing their ability
to survive under low ammonia environments (Bollmann et al. 2002; Bollmann et al. 2005).
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix A: Nitrification in Drinking Water Systems | 117
Like all microbes, the growth of nitrifiers can be limited by nutrients, and the extent of this
limitation is often described by the Monod equation:
μ = μmax * Ss/(Ks + Ss),
(A.2)
where μ is the specific growth rate at substrate concentration SS, μmax is the maximum specific
growth rate, and Ks is the half-saturation coefficient which is defined as the substrate concentration
at which μ is equal to half of μmax. Lower Ks values reflect cultures that have a higher affinity for a
given substrate. The half-saturation coefficient for AOB found in drinking water distribution
systems, Nitrosomonas oligotropha and Nitrosospira briensis, are below 5 μM (<0.07 mg/L as N)
(Bollmann et al. 2005; Koops and Pommerening-Roser 2001). In contrast, the KS value for pure
culture strains that are typically found in ammonium rich environments (such as wastewater
treatment plants) are an order of magnitude higher (Koops and Pommerening-Roser 2001).
Accordingly, the lower half-satuation constant of Nitrosomonas oligotropha and Nitrosospira
briensis versus Nitrosomonas europaea might explain the relative dominance of these strains at
the relatively low ammonium concentrations (between 0 to 1 mg/L as N) (Odell et al. 1996) often
present in drinking water. Furthermore, one study by Gieseke et al. (2001) suggested that
Nitrosomonas oligotropha might also have a higher affinity for oxygen than other Nitrosomonas
species, which would enable it to better compete with heterotrophs for oxygen in multispecies
biofilms such as those found in potable water systems. Nitrosomonas oligotropha was speculated
to retain more ribosomes than Nitrosomonas europaea during inactivation (Gieseke et al. 2001),
which might partly explain in situ detection of Nitrosomonas oligotropha rather than
Nitrosomonas europaea by fluorescent in situ hybridization (FISH), though there is no evidence
for this hypothesis yet (Gieseke et al. 2001).
The NOB are found in four phylogenetic groups, two of which are associated with marine
environments (Koops and Pommerening-Roser 2001). Two non-marine genera of NOB have been
widely reported. The first genus of NOB is Nitrobacter, which belongs to the α-proteobacteria
and was historically identified as the predominant NOB in many natural and engineered
environments (Painter 1977). More recently, the other genus was identified through
culture-independent methods, which showed that the phylogenetically distinct Nitrospira are more
common in nitrifying wastewater bioreactors (Burrell et al. 1998; Juretschko et al. 2002),
freshwater aquaria (Hovanec et al. 1998) and in both pilot (Regan et al. 2002) and full-scale
(Regan et al. 2003) drinking water distribution systems than previously thought. In the drinking
water distribution systems studied by Regan and colleagues, Nitrospira moscoviensis was most
prominent with few Nitrobacter detected. This is not surprising given that this strain was first
isolated from a corroded iron pipe in a heating system (Ehrich et al. 1995). Recently, researchers
(Daims et al. 2006; Maixner et al. 2006) showed that Nitrospira can be segregated into
sublineages, and that sublineage II grows more effectively than sublineage I in environments with
a low concentration of nitrite, further supporting the notion that subgroups of Nitrospira can
tolerate and may actually prefer low nitrite growth environments, such as those found in drinking
water distribution systems.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
118 | Effect of Nitrification on Corrosion in the Distribution System
FACTORS AFFECTING NITRIFICATION IN DRINKING WATER SYSTEMS
In order for nitrifying bacteria to grow, certain physical factors and chemical substrates are
required. These factors are likely to affect growth of nitrifying bacteria in drinking water
distribution systems (Wolfe and Lieu 2001).
Presence of Ammonia
Nitrifying bacteria are chemolithotrophic bacteria with limited ability to utilize organic
compounds. Ammonia, nitrite and (in some species) urea are the only exogenous energy source for
nitrifying bacteria, although the presence of fructose can provide an energy benefit to N. europaea
(Hommes et al. 2003).
In drinking water systems, ammonia can be present from the untreated drinking water,
released by chloramines decay, or even formed in the distribution system from reactions between
nitrate and metals (Table A.1). The initial chlorine to ammonia ratio used to form chloramines
affects the free ammonia levels at the start of the distribution system (Fleming et al. 2005). Excess
free ammonia present at lower ratios of chlorine to ammonia (< 4:1 mass ratio) tend to encourage
nitrification (Karim and LeChevallier 2006; Skadsen 1993).
We note that abiotic corrosion reactions with distribution system pipe materials could serve
to regenerate ammonium from nitrate at metal pipe surfaces. Complete conversion of nitrate to
ammonium has been detected in a new galvanized steel pipe during stagnation (McIntyre and
Mercer 1993). These reactions could result in coupled nitrogen cycling between abiotic (Reaction
10 and 11 in Table A.1) and biotic reactions (equation A.1 and A.3).
4 Fe° + NO3- + 10 H+ = 4 Fe 2+ + NH4+ + 3H2O
(A.4)
Consequently, even a small amount of ammonia or nitrate has the potential to support a
relatively high level of nitrifying bacteria growth on iron or lead pipe. That is, reaction of nitrate
with metallic pipes can create a continuous supply of ammonium for nitrifiers.
Dissolved Oxygen (DO)
Nitrifying bacteria are obligate aerobes. During nitrification, 4.33 mg O2 is consumed for
every mg-N of ammonia oxidized to nitrate according to stoichiometry (Grady et al. 1999).
The oxygen half saturation constant of Nitrosomonas and Nitrobacter is reported to lie
between 0.3 and 1.3 mg/L including values established under both pure cultures and activated
sludge conditions (Sharma and Ahlert 1977). Considering diffusion resistance, values of 0.5 and
0.68 mg/L have been adopted as typical for Nitrosomonas and Nitrobacter, respectively (Grady et
al. 1999). The limiting dissolved oxygen concentration in activated sludge reactors for nitrifying
bacteria ranges from 0.5 to 4 mg/L (Stenstrom and Song 1991) and 2 mg/L has been established as
the limiting concentration in other studies (Painter 1977). These values can vary greatly depending
on the culture purity, cell residence time and mass transfer resistance conditions (Stenstrom and
Song 1991).
Drinking water distribution systems are generally well oxygenated and have sufficient
oxygen for nitrifier growth in the bulk water. However, dead ends, stagnation and biofilms, and
corrosion may create micro-anaerobic environments where oxygen could be limiting. Ammonia
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix A: Nitrification in Drinking Water Systems | 119
oxidizers may still survive at low DO as they are tolerant of low oxygen or anoxic environments
(Kowalchuk and Stephen 2001). They can also use nitrate or nitrite as an alternative electron
acceptor when oxygen is low (Alleman and Preston 2005). This process is not thought to support
cell growth, but it may provide enough energy to allow survival under anaerobic conditions
(Kowalchuk and Stephen 2001).
Cell Attachment
In drinking water systems, most microbes are attached to pipe surfaces in film/slime layers
on the pipe. This layer is often referred to as a biofilm, but this living layer is also associated with
and integrated into inorganic layers present on the pipe termed “scale.” The inorganic materials
can include characteristic rust layers from the underlying material as well as deposits/precipitates
from the water. Nitrifiers can exist as free living cells or attach to a surface as a biofilm and they
have been repeatedly shown to prefer surfaces under many circumstances. Based on modeling
results, Furumai and Rittman suggested that growth in the biofilm is advantageous because it
prevents detachment loss (Furumai and Rittmann 1994); it is also possible that biofilm growth
makes the cells less susceptible to inactivation by disinfectants. Attached nitrifiers in biofilms are
more active than free living cells and more resistant to substrate limitation (Keen and Prosser
1987; Stein and Arp 1998). Biofilms and aggregates were also reported to facilitate nitrification
outside the optimal pH ranges for growth (Tarre and Green 2004).
The presence of nitrifiers in biofilms helps explain the persistance of nitrifiers in drinking
water systems (Wilczak et al. 1996). Cunliffe (1991) and Skadsen (1993) speculated that greater
detention times favor the formation of biofilms observed in dead-ends of distribution systems and
water storage reservoirs. These are often sites of persistent and highly active nitrification (Cunliffe
1991; Skadsen 1993). Also, sediment areas (Ike et al. 1988), and plumbing systems in premises
(Edwards et al. 2005) can support more nitrifier growth due to low disinfectant residuals and more
surface area for attachment. Sediment and tubercles in distribution pipes may also exert a chlorine
demand and further facilitate nitrifier growth (Odell et al. 1996). Attachment to solid surfaces can
also enhance bacteria including nitrifiers’ resistance to potential toxins like disinfectants
(Wolfe et al. 1988).
Temperature
Temperature can exert an effect on biological reactions in two ways: by influencing the
rates of enzymatically catalyzed reactions and by affecting the rate of substrate diffusion to the
cells (Grady et al. 1999). Kinetic parameters of nitrification are affected by temperature following
the Arrhenius equation in a certain range, the commonly used equation in environmental
engineering field is:
k1 = k2 θ (T1-T2)
(A.5)
where k1 and k2 are the temperature dependent kinetic parameters at temperature T1 and T2, θ is the
temperature coefficient (Grady et al. 1999). For the temperature range in which this equation is
valid, the maximum specific growth rate for nitrification doubles for every 6 ˚C increase in
temperature if an average reported θ value (θ =1.12) was used (Characklis and Gujer 1979).
However, higher temperature increases ammonia half-saturation coefficients for nitrifiers (Knowles
©2010 Water Research Foundation. ALL RIGHTS RESERVED
120 | Effect of Nitrification on Corrosion in the Distribution System
et al. 1965), which would decrease the growth rate according to the Monod equation. Nevertheless,
nitrifier growth and ammonia utilization at 20 ˚C is significantly lower than at 30 ˚C, and
experiments at these two temperatures noted that this temperature increase doubled substrate
consumption rates and biomass yield (Groeneweg et al. 1994). Protein denaturation becomes
important at temperature ranges above the range over which equation 5 applies (Neufeld et al. 1986;
Neufeld et al. 1980). When the inhibiting effect of denaturation exceeds the higher biological growth
rate associated with increasing temperature, nitrifier growth slows or stops. Early work by Gibbs
(1920) and Sawyer and Bradney (1946) reported 55 ˚C as very effective for inactivating nitrifiers.
Other researchers have found that 45 ˚C inactivated nitrifiers (Neufeld et al. 1980).
Nitrifiers are known to grow in a wide temperature range of 4- 45 ˚C although the upper
limit is unclear (Painter 1977; Wolfe and Lieu 2001). However, the optimal growth for nitrifiers
occurs in a very narrow temperature range. In a pure culture study, the optimum temperature for
Nitrosomonas was reported to be 35 ˚C (Grunditz and Dalhammar 2001; Painter 1977). For
Nitrobacter, the optimum lies between 35 - 42 ˚C (Grunditz and Dalhammar 2001; Laudelot and
Tichelen 1960; Painter 1977). In a mixed culture of AOB and NOB found in activated sludge, the
optimum temperature for nitrification occurrence was 30 ˚C (Neufeld et al. 1986; Neufeld et al.
1980). Different optimum temperatures for nitrification have been found by other researchers,
including 15 ˚C (Charley et al. 1980), 21- 28 ˚C (Wong-Chong and Loehr 1975), and no optimum
in a range of 15 to 35 ˚C (Shammas 1971).
Not surprisingly, in drinking water systems nitrification is highly influenced by water
temperature. In their survey of nitrification occurrence in drinking water distribution systems,
Wilczak et al. (1996) concluded that nitrification incidences were higher during summer or when
temperatures were greater than 15 ˚C. Wolfe et al. (1988 & 1990) and Ike et al. (1988) also found
the number of AOB was approximately 100 to 1000 times higher in the summer than in the winter.
No AOB were detected in water storage tank in the study of Wolfe et al. (1990) when the water
temperature was less than 18 ˚C; however, above this value, the number of AOB generally
increased in relation to temperature. Elevated temperature also increases the rate of chloramine
decay and therefore provides more free ammonia for nitrifier growth (Nowlin et al. 2001). But on
the other hand, elevated temperature increased the disinfectant efficiency of chloramine (Lieu et
al. 1993), which could help control nitrifiers in at least some cases.
In drinking water systems, the optimal temperature for nitrification and nitrifier growth has
been cited as 25 to 30 ˚C (Odell et al. 1996; Skadsen 1993; Wolfe and Lieu 2001). Analysis from a
South Australia study indicated that nitrifying bacteria grew in distribution systems with
temperatures ranging from 10 to 34 ˚C (Cunliffe 1991). But even some distribution system sites
sampled under cold water conditions (below 10 ˚C) also showed evidence of nitrification such as
increase in nitrite concentrations (Wilczak et al. 1996).
Light
Nitrifiers are very sensitive to visible and ultraviolet irradiation and even fluorescent
lighting (Wolfe and Lieu 2001). Ammonia oxidation by Nitrosomonas europaea is inhibited by
sunlight (Alleman et al. 1987) and UV light (Hooper and Terry 1974). Alleman and Preston (2005)
also pointed out that when nitrifiers grow as attached biofilm, bacterial layering will provide
considerable shading and shelter the organisms deeper in the biofilm from stressful light.
Nitrifying bacteria have the capability to recover from light inhibition. Recovery after exposure to
light occurs in four to six hours (Alleman et al. 1987; Hooper and Terry 1974). Light is almost
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix A: Nitrification in Drinking Water Systems | 121
completely absent in drinking water distribution system pipes, although plastics including PVC
allow some light transmittance to the pipe interior and sunlight falls on open reservoirs.
Nonetheless, nitrification in open reservoirs has been noted (Odell et al. 1996).
pH
Many studies have reported the effect of pH on nitrification (Table A.2). Nitrification
occurs over a fairly wide pH range, as low as 4.6 (Wolfe and Lieu 2001) and as high as 11.2
(Prakasam and Loehr 1972). Nitrite oxidizers exist in a narrower range of pH values than ammonia
oxidizers. The optimum pH for nitrification is between 7 and 8. Accordingly, the neutral to slightly
alkaline pHs often found in the bulk water of drinking water systems accommodate nitrification
(Odell et al. 1996; Wilczak et al. 1996).
It is commonly accepted that pH affects nitrification by changing the concentration of free
ammonia, because free ammonia rather than ammonium ions have been identified as the substrate
for ammonia oxidation (Suzuki et al. 1974). Consideration of equilibrium (NH3 + H+ → NH4+ with
a pKa of 9.25) illustrates that very little free ammonia exists at a pH below 6, at which point
nitrification has been noted to cease (Painter 1977; Stein and Arp 1998). However, Groeneweg et
al. (1994) observed that ammonia oxidation rate of Nitrosomonas europaea was not constant in a
pH range of 5-11 when applying a constant free ammonia concentration, indicating that other
mechanisms besides free ammonia concentrations are accountable for the pH effect on
nitrification. Groeneweg and colleagues preferred the theory of Bock et al. (1991) as an
explanatory hypothesis, which states that at low pH, most energy is consumed to maintain a
favorable pH environment inside the cell, thus there is no energy left for growth.
Tarre and Green (2004) recently suggested that the effect of pH has little to do with free
ammonia concentration; rather, the reduced growth that had been observed at lower pH arose from
an inorganic carbon limitation due to CO2 stripping occurring in aerated batch culture. Using pure
oxygen aeration in a continuous flow reactor, Tarre and Green (2004) demonstrated a very high
nitrification rate of 5.6 g-N/L-day by Nitrosomonas at low pH in both a biofilm reactor (4.3 ± 0.1
pH) and a suspended biomass reactor (3.2-3.8 pH). This rate is similar to that reported for
nitrifying reactors at optimum pH. Somewhat consistent with the Tarre and Green hypothesis,
Groeneweg et al. (1994) observed a lower optimal pH for nitrification than others (Table A.2)
when using CO2-enriched air for aeration at a constant free ammonia concentration. Resolving this
debate is important relative to better understanding mitigation of nitrification in potable water
systems, since altering bulk water pH is an economically attractive strategy for reducing
nitrification occurrence (Harrington et al. 2002; Skadsen 2002). However, adjusting pH might not
be effective since the pH near the biofilm on iron pipe walls can be much lower (> 3 pH units) than
in the bulk water (McNeill and Edwards 2001).
In drinking water systems, pH can affect nitrification not only by affecting the growth of
nitrifiers, but also by affecting ammonia release from chloramine decay and chloramine
inactivation rate on nitrifiers (Harrington 2002; Oldenburg et al. 2002). The effect of pH on
ammonia release, disinfection decay and nitrifying bacterial growth exhibit contrary trends, so the
optimal pH for nitrification occurrence is not always the same as the optimal pH for nitrifier
growth (Harrington et al. 2002).
©2010 Water Research Foundation. ALL RIGHTS RESERVED
122 | Effect of Nitrification on Corrosion in the Distribution System
Alkalinity and Inorganic Carbon
A large amount of alkalinity is consumed during the oxidation of ammonia to nitrate: 8.62
mg HCO3- (14 mg/L as CaCO3) is consumed for every mg NH4+-N oxidized. Most alkalinity that is
consumed results from neutralization of hydrogen ions during the oxidation of ammonia, and only
a small amount of alkalinity is lost by incorporation of inorganic carbon into cell biomass (Grady
et al. 1999). Alkalinity levels which are adequate to stop pH from dropping below a preferred
range favor nitrification (Alleman and Preston 2005). Potable water typically has 1 mg/L NH3-N
or less, and alkalinity is usually maintained well above 14 mg/L as CaCO3 for corrosion control.
Therefore, minimum required levels of alkalinity are sufficient to support complete conversion of
ammonia to nitrate. Though Wilczak et al. (1996) reported that nitrification did not significantly
impact alkalinity; this conclusion might have been different if the waters studied had less than 30
mg/L as CaCO3.
Nitrifying bacteria are autotrophic bacteria, and they use carbon dioxide as a carbon source
which is fixed via the Calvin cycle. Inorganic carbon is not limiting to ammonia oxidizers even in
wastewaters (Kowalchuk and Stephen 2001) unless very high levels of nitrogenous substances are
present (Jun et al. 2000). When the ratio of inorganic carbon to ammonia, expressed as mg as
CaCO3/ mg ammonia-N ratio is above 8.25, the nitrite oxidizing process is the rate limiting
process; while below this level, the ammonia oxidizing process is often the limiting step (Sakairi et
al. 1996). This effect of inorganic carbon on the growth of AOB and NOB is more significant in
biofilm systems (Tokutomi et al. 2006). In drinking water systems, a ratio of at least 14 (minimum
14 mg/L as CaCO3 alkalinity and maximum 1 mg/L-N ammonia) is maintained, hence, nitrite
oxidation is suspected to be the rate limiting step. This explains the accumulation of nitrite in
distribution systems when nitrification occurs.
Organic Carbon and Competition between Heterotrophic Bacteria and Nitrifying Bacteria
Organic substrates are ubiquitous in the aquatic environment (Rittmann and Manem 1992)
and even predominantly autotrophic nitrifying bacteria can incorporate exogenous organic
compounds like acetate, 2-oxaloglutrarate, succinate, amino acids and sugars into biomass (Clark
and Schmidt 1966; Clark and Schmidt 1967; Martiny and Koops 1982; Wallace et al. 1970). In
fact, all AOB can metabolize selected organic compounds to a limited extent in the presence of an
inorganic energy source (Wolfe and Lieu 2001). Nitrosomonas europaea has been reported to
grow on fructose or pyruvate instead of CO2 as the sole carbon source (Hommes et al. 2003).
Beneficial effects of organic compounds on the growth of nitrifying bacteria have been
reported. Pyruvate and peptone increased cell yield of a Nitrosomonas strain, while formate and
acetate shortened the lag phase and enhanced cell yield of a Nitrospira strain (Krummel and Harms
1982). The time for a culture of Nitrospira marina to double in cell number decreased from 90
hour when grown autotrophically to 23 hours when grown in medium supplemented with
pyruvate, yeast extract and petone (Watson et al. 1986). All the above studies with AOB were
conducted with ammonia present as an energy source, although Pan and Umbreit (1972a) grew
Nitrosomonas europaea in the presence of glucose and in the absence of ammonia using a dialysis
system to remove toxic metabolic products. The cell yield using glucose as an energy source was
less than with ammonia; moreover, the results have been difficult to reproduce and have been
criticized by others (Hommes et al. 2003; Krummel and Harms 1982).
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix A: Nitrification in Drinking Water Systems | 123
Overall organic carbon is generally considered to have a negative effect on nitrification.
This effect can be due to the toxic nature of the organic material itself or due to heterotrophic
bacteria out-competing nitrifiers for essential co-nutrients (Hockenbury et al. 1977; Sharma and
Ahlert 1977).
Nitrifiers and heterotrophs compete for surfaces, dissolved oxygen, ammonium and other
nutrients. Nitrifiers must reduce inorganic carbon to the oxidation state of cellular carbon, a
reaction that is energy intensive and results in lower yield and growth rates compared to
heterotrophs (Rittmann and Manem 1992). Nitrifiers are therefore at a competitive disadvantage
versus heterotrophs at high levels of organic carbon. Based on their lower half saturation constants
(Km), heterotrophs were generally considered to have a greater affinity for dissolved oxygen
(Grady et al. 1999) and ammonium (Rosswall 1982) than nitrifiers, although studies have
indicated that specific affinity (Vmax/Km) rather than Km should be considered when determining
the competition by heterotrophs and nitrifiers for oxygen (Bodelier and Laanbroek 1997). When
coexisting in a biofilm, faster growing heterotrophs are believed to dominate at the biofilm’s
surface while the slower growing nitrifiers exist deep inside the film because bacteria with higher
growth rate can survive the higher detachment rate at the biofilm surface (Furumai and Rittmann
1994); therefore mass transport of key constituents such as oxygen through the biofilm to nitrifiers
can be an issue since it must first pass through a layer of heterotrophs (Furumai and Rittmann
1994; Rittmann and Manem 1992). Heterotrophs can also out-compete nitrifiers for ammonia
when organic carbon is high (Jansson 1958).
If disinfectant were not considered, the net result is that below a critical organic
carbon-to-nitrogen ratio, nitrifiers and heterotrophs coexist. Furthermore, under extreme
conditions when no dissolved organic carbon is present in the water, heterotrophs are completely
dependent on the lysis and extracellular products of nitrifiers as a source of organic carbon. At
higher organic carbon to nitrogen ratios or if ammonia is limited, nitrifiers’ numbers are strongly
reduced by heterotrophs (Ohashi et al. 1995; Verhagen and Laanbroek 1991).
Conceptually, it is therefore expected that regimes of organic carbon and ammonia can be
demarcated in which nitrifiers dominate in the system, co-exist with heterotrophs, or are negligible
based on the levels of organic carbon relative to ammonia. A rough approximation can be made
based on the relative growth rate of nitrifiers or by using the data of Verhagen and Laanbroek
(1991) (Figure A.1 and Supporting Information). If C and N concentrations are below the line
deducted by the Monod Model, nitrifiers would be dominant in the system. If C and N
concentrations are above this line but below the line deduced from Verhagen and Laanbroek’s
study, both nitrifying and heterotrophic bacteria grow in the system, but heterotrophic bacteria
dominate. If C and N concentrations are above the line of Verhagen and Laanbroek, nitrifying
bacteria would be negligible in the system. However, the work of Verhagen and Laanbroek was
performed at very high ammonia concentration (> 28 mg/L-N) and organic carbon (72- 360
mg/L-C) relative to the conditions in drinking water system. The predictions would be more
accurate in predicting the occurrence of nitrification in drinking water systems if a critical
carbon/nitrogen ratio can be established under conditions more relevant to typical assimilable
organic carbon concentrations (between 0.01 – 1 mg/L) and NH3-N levels (between 0 to 1 mg/L as
N) in drinking water systems.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
124 | Effect of Nitrification on Corrosion in the Distribution System
BENEFICIAL INTERACTIONS BETWEEN HETEROTROPHIC BACTERIA AND
NITRIFYING BACTERIA
Heterotrophic bacteria are always found in association with nitrifying bacteria when
nitrification occurs in water distribution systems. This is probably because the two groups can
have synergistic effects under some circumstances.
Heterotrophs can produce organic compounds that stimulate nitrifiers and degrade organic
compounds that are inhibitory to the nitrifiers. Hockenbury et al. (1977) reported filtrate from
heterotrophic bacteria had a slightly (e.g., 17%) beneficial effect on nitrifying bacteria growth, and
they proposed that heterotrophs might excrete metabolic products useful to nitrifiers. Pan and
Umbreit (1972b) also reported Nitrosomonas and Nitrobacter growth could be stimulated by a
particular heterotrophic bacteria group, as illustrated by shortening of the lag phase and extension
of log phase. It was suggested that this stimulation was due to removal of toxic metabolites. Also,
heterotrophs can produce extracellular polymers that can improve the aggregation of nitrifiers into
biofilms and protect nitrifiers from detachment by pre-dominating in the outer layer of
multi-species biofilms (Furumai and Rittmann 1994; Rittmann and Manem 1992;
Rittmann et al. 1994).
The production of soluble microbial products (SMP) by nitrifying bacteria can provide a
sole or supplemental organic substrate for heterotrophic bacteria. For Nitrobacter at stable and
complete substrate oxidation, about 0.025 mg COD is produced for every mg N oxidized, and for
Nitrosomonas, 0.073 mg COD is produced for every mg N oxidized (Rittmann et al. 1994). SMP
produced by nitrifiers can therefore enhance heterotroph accumulation and stability, especially
when inputs of organic substrates are low (Kindaichi et al. 2004; Rittmann et al. 1994).
Heterotrophic growth is sustained by SMP formed by nitrifiers when there is no organic-carbon
substrate supplied (Furumai and Rittmann 1994; Gieseke et al. 2005; Kindaichi et al. 2004). When
the input ratio of COD to NH4+-N is 1, more than 40% of the COD utilization by the heterotrophs is
nitrifier-generated SMP (Furumai and Rittmann 1992; Furumai and Rittmann 1994; Rittmann et
al. 1994), and Kindaichi et al. (2004) noted that an autotrophic nitrifying biofilm fed with only
NH4+ as an energy source and no organic carbon was composed of 50% nitrifiers (AOB and NOB)
and 50% heterotrophic bacteria. In a drinking water distribution system, occurrence of AOB and
Heterotrophic Plate Count (HPC) can be linearly related (Wolfe et al. 1990). This would obviously
be expected if HPC were dependent on AOB for fixed organic carbon.
Confusing matters somewhat, it is known that certain heterotrophic bacteria can carry out
nitrification, e.g. Alcaligenes faecalis, Pseudomonas, Thiosphaera pantotropha, fungi and some
algae (Bock et al. 1992; Focht and Verstraete 1977; Killham 1986; Verstraete and Alexander 1973;
Verstraete and Alexander 1986; Watson et al. 1989), although at a slower rate and using different
mechanisms than autotrophic nitrifiers. Recently, ammonia oxidizing Archaea have been
identified in a marine environment and in wastewater treatment plants, proving that bacteria are
not the only organisms that can carry out nitrification (Könneke et al. 2005; Park et al. 2006).
Additionally, Archaea are suspected to play a major role in nitrification at low DO.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix A: Nitrification in Drinking Water Systems | 125
Micro-nutrient Limitation and Inhibition
All microbial growth depends on the availability of nutrients. Microbial growth can be
represented by a simple overall equation:
Carbon + Nitrogen + Phosphorus + Trace Nutrients (K, Cu, Fe, etc.) + O2
→ Aerobic Microbial Growth
(A.6)
Nitrifiers also follow this growth pattern. The nutrients in a potable water system that are
present at relatively low levels (and which therefore limit the maximum number of bacteria that
can grow) can be limiting (Edwards et al. 2005; Madigan et al. 2000; Morton et al. 2005). In terms
of the key nutrients of carbon and ammonia, it is expected that inorganic carbon is rarely limiting
(Jun et al. 2000), whereas ammonia is frequently found to be limiting (Leu et al. 1998; Rittmann
and Soneyink 1984; Wolfe and Lieu 2001).
Phosphorus in the environment occurs mostly as orthophosphate (PO43-), but phosphites
[hypophosphate PO23- and phosphate PO33-] and phosphides [PH3] are also present at trace levels
in water distribution systems (Morton et al. 2003). Much of the total phosphorus in natural waters
is not readily available for bacterial metabolism, and phosphorus has been reported to limit
heterotrophic bacterial growth in areas of the US, Finland, Japan and China (Marshall and
Edwards 2004b; Miettinen et al. 1997; Sang et al. 2003; Sathasivan and Ohgaki 1999). Van der Aa
et al. (2002) and Kors et al. (1998) reported that nitrification processes used to remove ammonia at
water treatment plants were limited when phosphorus was almost completely removed or below 15
μg/L-P at low temperatures. Nitrification rates recovered in both plants after dosing phosphate at
35-50 and 100-150 μg/L-P. Van der Aa et al. (2002) proposed that at least 10 μg/L-P phosphate is
necessary for nitrification to remove 1 mg/L NH3-N ammonia from the raw water. The reported
minimum phosphorus concentration for nitrification in raw water treatment in the literature range
from 3 to 20 μg/L-P, and very little nitrite oxidation by Nitrobacter occurred when phosphorus was
less than 50 μg/L-P (Aleem and Alexander 1960; Van Droogenbroeck and Laudelout 1967). Other
studies found that growth of both Nitrosomonas europaea and Nitrobacter winogradskyi depend
on cell phosphate content, which is related to the history of the nitrifier culture (Van
Droogenbroeck and Laudelout 1967). Although this does not affect the growth as much as the
phosphate concentration in the growth medium, it is predicted that the effect of culture history
would be more pronounced if nitrifiers are growing under very low ammonium/nitrite and
phosphate concentrations as opposed to the high ammonium/nitrite (> 300 mg/L-N) and phosphate
concentration range (> 310 mg/L) used in the study (Van Droogenbroeck and Laudelout 1967).
In addition to the macronutrients of inorganic carbon, ammonia and phosphorus, nitrifier
growth also requires trace nutrients such as potassium, calcium, copper, magnesium and other
constituents. Trace nutrients have four possible metabolic impacts dependent on concentration
(Figure A.2). In deficiency, bacterial activity can be limited if an essential nutrient concentration is
too low. Deficiencies of trace elements like potassium (Fransolet et al. 1988), boron, calcium,
chromium, cobalt, copper, iron, magnesium, manganese, molybdenum, nickel, vanadium, and zinc
can decrease growth or cause bacterial death (Reeves et al. 1981). As the trace nutrient concentration
increases, bacterial activity may be restored and reach an optimum dependent on the circumstance of
growth (e.g., fixed film versus batch culture, pH condition). At excess concentrations, however,
detrimental effects may result. Many trace nutrients, including metals, can stimulate growth at low
concentrations and cause toxicity at high concentrations (Sato et al. 1986).
©2010 Water Research Foundation. ALL RIGHTS RESERVED
126 | Effect of Nitrification on Corrosion in the Distribution System
For nitrifying bacteria, metal inhibition can occur by blocking the enzyme function (Martin
and Richard 1982). Active cultures are more sensitive than resting cultures (Ibrahim 1989), and
inhibition worsens with decreased detention time and organic matter concentration (Disalvo and
Sherrard 1980). Nitrosomonas was found to be at least as sensitive as Nitrobacter in terms of
inhibition by heavy metals (Lee et al. 1997).
A survey was recently completed of 330 raw drinking waters in the U.S. to determine the
concentration of various metal and phosphorus constituents (AWWARF 2004). While the
concentration of these constituents is sometimes altered somewhat by treatment process, analysis
of that data indicates many trace elements occur in ranges that would be limiting in some cases and
toxic in others (Table A.3). Each of the key nutrients that might be limiting are discussed in the
sections that follow, since this could be one key to understanding why nitrifier growth is rampant
in some potable water systems but not others.
It is also important to note that the concentrations of many nutrients are altered as a result
of water treatment and/or distribution, either purposefully or as a result of corrosion processes. It is
therefore possible that decisions made about treatment and inhibitor dosing could strongly
influence the extent of nitrification occurring in the distribution system.
Copper
Copper (II) is believed to be a key component of the ammonia monooxygenase (AMO)
enzyme, which is essential for ammonia oxidation (Ensign et al. 1993; Richardson and Watmough
1999) and AOB growth. For a pure culture of Nitrobacter (NOB) 48 ppb copper has been
recommended (Tang 1992). However, excess copper is known to be toxic to nitrifiers as in the case
for heterotrophs (Braam and Klapwijk 1981). Copper levels as low as 2 ppb have been found to be
toxic to nitrifying bacteria in waters with low chelating capacity (Waara and Wilander 1985). The
inhibition of nitrification with copper is sometimes reversible through addition of copper chelating
compounds within 24 hours (Braam and Klapwijk 1981).
The effect of copper on nitrification was also found to be dependent on the free copper
2+
Cu concentration (Braam and Klapwijk 1981). The free copper concentration increases at lower
pH and the dependence of toxicity with pH probably results from this chemical change (Loveless
and Painter 1968).
Certain complexes of copper may increase toxicity. For example, the effect of copper on
nitrification is dependent on the ammonia concentration, as copper inhibition increases as
ammonia concentration increases. The level of copper that caused 50% inhibition of N. europaea
decreased from 0.5 to 0.01 ppm as the total ammonia concentration rose from 3 to 23 mg/L-N. This
was attributed to higher toxicity of copper-ammine complexes [Cu (NH3)X2+] (Sato et al. 1988).
The effect of copper also varies based on the purity of the nitrifying culture. Concentrations
of copper necessary to inhibit nitrification in activated sludge are much higher than those required to
give the same effect in pure culture, possibly because of the biomass level [Mixed Liquor Suspended
Solids (MLSS) concentration] and likely associated sorption of free copper (Tomlinson et al. 1966).
For example, at pH 7.6, 4 ppm copper had 75% inhibitory effect on ammonia oxidation in pure
culture, whereas about 200 ppm copper is required to achieve 75% inhibition in activated sludge
having a concentration of 1350 to 1700 mg of dry matter/L (Tomlinson et al. 1966).
The range of copper that stimulates or inhibits nitrification is therefore expected to vary
widely based on nitrifier strain and growth conditions (Table A.3).
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix A: Nitrification in Drinking Water Systems | 127
Phosphate, pH interaction in relation to copper effect on nitrification
Phosphate and pH interact to control free and total copper (II) concentrations (Edwards et
al. 2002; Schock et al. 1995), and would therefore be expected to exert a key control over
nitrification in premise plumbing in which copper is the most common pipe material. Copper pipes
are covered with thin patina (rust) or scale layers, the identity of which controls the free copper in
water during overnight stagnation. The scale layer could be composed of Cu(OH)2 scale, CuO,
malachite (Cu2CO3(OH)2) or other compounds.
About 56% of utilities add phosphorus to water (0.2 to 3 mg/L) to form a low solubility
Cu3PO4 scale to attempt and reduce copper solubility (McNeill and Edwards 2002). When Cu(OH)2
is present as in the case of new pipe, free Cu 2+ released to the water is higher than when phosphorus
is added, especially at lower pHs (Figure A.3). Specifically, free Cu 2+ is predicted to be 27 mg/L at
pH 6 and decrease 10 fold for every 0.5 pH unit increase in pH. Consequently, in this type of pipe
and water, nitrification would be inhibited by toxic levels of free copper at lower pH and might be
limited by the absence of free copper at higher pH. Using thresholds of 0.1 ppm copper based on the
completely mixed suspended growth tests mentioned earlier, it is predicted that copper toxicity is
likely below pH 7 and growth might be limited by low free copper above pH 7.5.
When phosphorus is present in water along with a scale-forming, relatively insoluble
Cu3(PO4)2 solid, free Cu 2+ is much lower in water at lower pH (Figure A.3). Thus, phosphate
corrosion inhibitor might be expected to reduce copper-induced toxicity to nitrifiers in copper pipe
when pH is below 7, especially at higher phosphorus doses. The potential dual stimulating action
of phosphorus as a nutrient and also as an inhibitor of copper-induced toxicity is deserving
additional attention.
Nickel
Addition of low levels of nickel (II) to soil stimulated nitrification either directly or by
inhibiting competitors of nitrifiers (deCantanzaro and Hutchinson 1985). High concentrations of
nickel could inhibit nitrifier growth by inducing an extended lag phase. Nickel toxicity is most
evident for N. europaea at low ammonia concentration and high nickel concentration (Sato et al.
1986). The threshold levels of nickel toxicity also vary largely and are dependent on bacterial
strain and growth conditions (Table A.3). Nickel is a major component of stainless steel (Lula
1986), the corrosion of stainless steel plumbing material might contribute nickel to drinking water
and MCL regulates nickel at 0.1 mg/L (USEPA 2001a)
Chromium
Chromium is a commonly studied metal ion for its toxicity in the nitrification process and
there are a wide range of impacts on chromium inhibition (Table A.3). Chromium (III) was used in
some studies (Harper et al. 1996; Skinner and Walker 1961), Chromium (II) in others (Martin and
Richard 1982), and chromium (VI) (Tomlinson et al. 1966) in still others. The different oxidation
states might partly explain the ranges of effects observed (Table A.3), although the definitive work
on this has not yet been conducted. Like Nickel, chromium is a major component of stainless steel
(Lula 1986), and MCL regulates chromium at 0.1 mg/L (USEPA 2001a)
©2010 Water Research Foundation. ALL RIGHTS RESERVED
128 | Effect of Nitrification on Corrosion in the Distribution System
Zinc
Zinc (II) can cause toxicity by precipitating phosphorus nutrients (Harper et al., 1996). No
stimulatory effect has been found for pure Nitrosomonas culture in a range of 0.005-0.5 ppm Zn,
while 0.08-0.5 ppm Zn has been noted to inhibit nitrification (Loveless and Painter 1968). For
Nitrobacter, the threshold inhibitory level is higher than for Nitrosomonas culture; up to 1 ppm Zn
had no inhibition (Waara and Wilander 1985). Recent lab study on non-purified Nitrosomonas
europaea also showed clear inhibition of nitrification by 0.15 ppm Zn2+ (Zhang and Edwards
2005). However, another study with a full scale drinking water treatment plant concluded that no
nitrification would be expected if 0.5 ppm Zn2+ were added (Bott 2005). Zinc is often added to
water supplies in conjunction with phosphorus corrosion inhibitors and its potential role in
inhibition or stimulation of growth is therefore of high interest.
Iron
Iron is also considered to comprise an active binding site on AMO (Zahn et al. 1996).
Limited growth of nitrifiers was observed without the addition of supplemental iron (Sato et al.
1988). However, Ensign et al., 1993 found that the addition of Fe (II) or Fe (III) had no stimulating
effect on AMO activity in a pure Nitrosomonas europaea culture, and addition of up to 2 ppm-Fe
as ferrous sulfate had no effect on the growth of Nitrosomonas. Conversely, addition of chelated
iron (Skinner and Walker 1961) or exogenous siderophores (biologically-generated iron chelating
agents) (Wei et al. 2006) have been shown to enhance AOB growth. In fact, although
Nitrosomonas europaea is incapable of producing its own siderophores (Chain et al. 2003), its
ability to utilize exogenous siderophores may explain why nitrifiers grow better in the presence of
other bacteria if they provide iron chelating agents that can support AOB growth. Finally, the
solubility of ferrous and ferric iron is reduced by addition of phosphorus, analogous to the
preceding discussion of copper.
Lead
As with the other heavy metals, lead complexation and solubility play a role in the
observed effects, and this most likely accounts for the different inhibitory levels of lead reported
for various pure culture experiments (Table A.3). In one study of activated sludge, up to 100 ppm
lead (II) had no inhibiting effect on nitrifier growth rate because it was virtually all precipitated or
complexed (Martin and Richard 1982). In another study by Ibrahim (1989), addition of 0.1, 0.5, 1
and 2 ppm Pb caused 18, 29, 53 and 78% inhibition of respiration rate for active nitrifiers grown
with adequate substrate and 12, 25, 37 and 50% for resting nitrifiers grown with limited substrate,
respectively. The toxicity of lead was higher than that observed for nickel in the same study. As
was the case with copper, solubility of free lead can be dramatically reduced by the presence of
phosphorus in water.
Molybdenum
Nitrite oxidoreductase
(enzyme involved
in
nitrite oxidation)
is
a
molybdenum-iron-sulphur protein, containing 0.12 or 0.7 molybdenum per 400,000 molecular
weight (Ferguson 1998; Prosser 1986). A minimum of 2000 atoms of molybdenum are required
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix A: Nitrification in Drinking Water Systems | 129
for the synthesis of one Nitrobacter cell (Finstein and Delwiche 1965). Zavarzin (1958) proposed
that molybdenum participates directly in the enzymatic oxidation of nitrite, but a later study found
no evidence for this (Prosser 1986). Nonetheless, addition of molybdenum (VI) produced an
11-fold increase in both nitrite utilization and cell mass development of Nitrobacter over five days
(Finstein and Delwiche 1965). Stimulating effects were found at concentrations as low as 10-9 M
(0.1 ppb) by Finstein and Delwiche (1965) and 10-7 M (10 ppb) by Zavarzin (1958). Zavarzin also
found that molybdenum (VI) stimulated nitrification most effectively at pH 7.7 and actually
inhibited nitrification at pH 8.8; moreover, the stimulating effect of molybdenum (VI) occurred
only in the presence of 5.6 ppm iron as FeSO4. There is currently no drinking water maximum
contaminant level for molybdenum, although the irrigation standard is 0.010 ppm Mo (Chakrabarti
and Jones 1983).
Tungsten
Tungsten (VI) is reported to stimulate nitrite oxidation at a concentration of 10-5 mM (1.84
ppb-W); however, there was no stimulation when 10-4 mM molybdenum and 10-5 mM tungsten
were added together (Zavarzin 1958). Contrary to molybdenum (VI), tungsten (VI) stimulated
nitrite oxidation most effectively at pH 8.8 while it inhibited oxidation at pH 7.7 (Zavarzin 1958).
Addition of 10-5 M (1.84 ppm-W) tungsten slightly inhibited substrate oxidation of Nitrobacter
pure culture (Finstein and Delwiche 1965).
Sodium, Calcium, Magnesium, Potassium, Cadmium and Cobalt
Several studies report the effects of sodium, calcium, magnesium, potassium and cobalt on
nitrification (Table A.3). Sodium in natural water is typically in the range of 3 to 300 ppm. These
concentrations allow ready growth of nitrifiers based on the minimum required level, but might be
less than optimal. The magnesium levels found in natural water are physiologically relevant to
nitrifying bacteria growth as it can limit growth at the lower range and inhibit growth at the higher
range. The ability of magnesium to limit growth was more dramatic at lower pH (Loveless and
Painter 1968). Cadmium also can inhibit nitrification in a manner similar to copper, but of lesser
magnitude (Sato et al. 1986).
The effect of mixed trace elements appears complex. For example, a mixture of copper
(decrease growth rate) and nickel (cause extended lag phase) affected both growth rate and lag
phase, but the inhibition intensity is decreased compared to copper or nickel alone. A mixture of
cadmium (decrease growth rate) and nickel (cause extended lag phase) also affected both growth
rate and lag phase, but the growth rate was further reduced compared to cadmium alone while the
lag phase was shortened compared to nickel alone (Sato et al. 1986).
Disinfectant Residual
Monochloramine disinfectant is formed by combination of free chlorine with ammonia at a
mass ratio of approximately 4 mg/L Cl2 per 1 mg/L NH3-N or less. About ¾ of utilities in the U.S.
using chloramine target the residual between 1 and 3 mg/L (Wilczak et al. 1996). Monochloramine
can be viewed as having two competing effects on AOB: inactivation resulting from the presence
of chloramine disinfectant (biocide) and growth from the presence of ammonia (food) (Edwards et
al. 2005; Fleming et al. 2005). That is, although AOB are approximately 13 times more resistant to
©2010 Water Research Foundation. ALL RIGHTS RESERVED
130 | Effect of Nitrification on Corrosion in the Distribution System
monochloramine than free chlorine disinfectant (Wolfe et al. 1990), monochloramine is
nonetheless a disinfectant. However, when monochloramine disappears through a variety of
reactions (Table A.1), free ammonia is often formed. Due to the difference of strain, temperature,
chlorine to ammonia ratios, water chemistry, enumeration method and pH values, reported CT99
values in different studies range from 3 to 19, 000 mg Cl2⋅min/L (Cunliffe 1991; Oldenburg et al.
2002; Wolfe et al. 1990).
Fleming et al., 2005 proposed that nitrification occurrence in a distribution system depends
on the relative concentration of chlorine (biocide) and free ammonia (food). A nitrification
potential curve was constructed for a pilot scale study. Based on this curve, a threshold total
chlorine value of 1.6 mg/L was established, above which, nitrification would be prevented
regardless of the free ammonia concentration and below which, nitrification occurrence depends
on the ratio of chlorine and free ammonia. Based on the slope of the nitrification potential curve,
we can further conclude that nitrification is prevented when the biocide to food mass ratio is above
8. Similarly, in a Florida utility, nitrification rarely occurred when chlorine residual was above 1
mg/L and biocide to food ratio was more than 5 (Liu et al. 2005).
Nitrification has been observed to occur more frequently in systems with low chloramine
doses (Lieu et al. 1993; Odell et al. 1996; Skadsen 1993; Wilczak et al. 1996). Chloramine levels
typically used for potable water disinfection (1 to 2 mg/L according to Wolfe et al., 1990) should
be sufficient to eliminate nitrifiers.
On the other hand, nitrification occurred in the Ann Arbor, Michigan system which had an
average of 5 to 6 mg/L monochloramine and an upper dose up to 8 mg/L (Skadsen 1993). In a
study in Australia, 20.7 % of samples with monochloramine concentrations greater than 5 mg/L
had nitrifying bacteria detected (Cunliffe 1991). These observations are not contradictory to the
theory of Fleming et al. (2005). The high dose of chloramines in Ann Arbor reservoir was applied
after nitrification started, so chloramine decay was accelerated by the nitrite produced and the
disinfectant concentration in the distribution system was much lower than expected (Fleming et al.
2005; Odell et al. 1996). Nitrifiers might stay viable under high disinfectant dose for a period of
time as proposed in Cunliffe’s study (1991). Also, the nitrification potential curve and the
necessary chloramine level to prevent nitrification vary depending on the specific water quality
and treatment practices in different systems (Fleming et al. 2005; Lieu et al. 1993). Although the
Fleming et al. (2005) study used real water, nitrifier growth in a pilot system is still vastly different
from a real system where biofilm could be better established and different pipe materials could
exert a protective effect by reacting with and destroying the chloramine disinfectant (Table A.1).
Distribution Materials and Corrosion Control Strategies
Different pipe materials can be expected to strongly influence occurrence of nitrification
since the pipe can serve as a source of trace nutrients, toxic metals, attached growth and
disinfectant destruction. Heterotrophic bacteria exhibit the best growth on reactive surfaces such
as iron pipes, whereas PVC pipes typically have lower levels of bacterial growth (Camper et al.
2003). Cement and epoxy were intermediate in terms of support for attached heterotrophs (Camper
et al. 2003). Iron pipes are problematic, not only because of the beneficial surface, but also because
it destroys chlorine disinfectant (Camper et al. 2003). Likewise, for nitrifying bacteria, it was
speculated that iron tubercles in distribution pipes may exert chlorine demand and facilitate
nitrifier growth (Odell et al. 1996). Iron corrosion by-products can accelerate chloramine decay
and release ammonia for nitrifying bacteria growth:
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix A: Nitrification in Drinking Water Systems | 131
1/2NH2Cl + H+ + Fe2+→ Fe3+ + 1/2NH4+ +1/2 ClFe0 from iron pipe can also recycle ammonia from nitrate:
4 Fe° + NO3- + 10 H+ = 4 Fe 2+ + NH4+ + 3H2O
(A.7)
(A.4)
The ammonia thus generated could be used by nitrifiers or lost through breakpoint
chlorination when chlorine is present (Figure A.4). Under some circumstances, iron corrosion
could provide essential micronutrients necessary for nitrifying bacteria growth, such as
phosphorus impurities in the metal or iron dissolution from scale (Morton et al. 2005).
Concrete materials can leach lime and therefore have a higher surface pH (> pH 12 for fresh
concrete) than bulk water (AWWARF and DVGW-TZW 1996). Leaching of lime and other trace
nutrients from concrete is very dependent on water chemistry (Guo et al. 1998). Concrete lined pipes
had the lowest AOB and HPC in a study at two California utilities using chloramines. The authors
suggested it was because of the high pH due to concrete leaching (Steward and Lieu 1997). But
cement lined ductile iron pipe supported a higher heterotrophic biomass than did unlined iron in
Pinellas county, Florida (LePuil et al. 2003). It has been speculated that concrete might accelerate
bacterial growth by facilitating chloramine autodecomposition by surface catalysis (Woolschlager
and Soucie 2003), and previous modeling work did suggest that cement lined pipes exerted a
substantial and surprising chloramine demand (Woolschlager et al. 2001). While the results are
seemingly contradictory in the two studies, they might be attributed to fresher concrete or lower
alkalinity in the earlier study which caused a high pH rise and inhibited nitrifier growth.
For lead pipes and materials containing lead such as brass, two contrary effects on
nitrifying bacteria growth could exist. As in the case of iron, corrosion reactions between lead and
nitrate could recycle ammonia from nitrate and support a large nitrifier population (Reaction 10,
Table A.1). On the other hand, soluble lead leached during corrosion might be able to inhibit
nitrification, although lead inhibition on nitrification was only observed in soil and not in aquatic
environments (Loveless and Painter 1968; Shkelqim and Malcolm 2002).
Zinc which leaches from galvanized iron pipes and copper leaching from copper pipes could
both serve as a trace nutrient source at low levels and toxic elements at high levels for nitrifiers. The
net effect of a pipe material depends on the corrosion intensity and water quality. Nitrification has
been reportedly stimulated in copper pipes in Willmar, Minnesota (Murphy et al. 1997b).
Filtration Treatment Strategies
Granular activated carbon (GAC) and powdered activated carbon (PAC) are applied for
enhanced removal of organic matter and taste and odor problems. As early as 1935, Feben
recognized that filter beds receiving water with ammonia offer a nearly ideal environment for
nitrifying bacteria. GAC has high porosity and can increase nutrient adsorption and bacterial
attachment (Rollinger and Dott 1987). Activated carbon can also accelerate chloramine decay
through the following reactions:
(A.8)
NH2Cl + H2O + C* → NH3 + H+ + Cl- + CO*
+
2NH2Cl + CO* → N2 + H2O + 2H + 2Cl + C*
(A.9)
Note: C* and CO* indicate active carbon and surface oxide on the carbon.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
132 | Effect of Nitrification on Corrosion in the Distribution System
Growth of nitrifying bacteria was easily initiated in GAC beds when chloramine
concentrations were less than 1 mg/L-Cl2, and the established nitrifying bacteria could easily
tolerate higher chloramine concentrations of 4 mg/L-Cl2 (Fairey et al. 2004). Nitrification was also
reported to occur immediately after a change in treatment from sand to GAC in Ann Arbor,
Michigan (Skadsen 1993). It was suspected that GAC helped to convert chloramines to excess free
ammonia and enabled nitrifying bacteria to proliferate (Tokuno 1997; Vahala 2002). Higher rates
of nitrification also occurred after installing GAC in a full scale distribution system in Finland due
to decreased competition for space between nitrifying and heterotrophic bacteria, the introduction
of a shelter inside the carbon fines and a slightly higher excess ammonia concentration (Vahala
2002; Vahala and Laukkanen 1998).
EFFECTS OF NITRIFICATION
Increased Nitrite and Nitrate
Nitrification converts ammonia to nitrite and nitrate. Nitrite and nitrate are regulated by the
USEPA and have primary Maximum Contaminant Levels (MCLs) of 1 and 10 mg/L as N,
respectively. The concentration of nitrite in typical surface and groundwater is far below 0.1 mg/L.
Theoretically, the ammonia from 4 mg/L chloramines (chlorine to ammonia ratio is 4:1) could be
converted to 1.1 mg/L-N nitrite (USEPA 2005). Thus, the US EPA MCL of 4 mg/L chloramine
also serves to prevent exceedence of the drinking water MCL for nitrite even under the worst case
in which all chloramine ammonia is converted to nitrite. In drinking water systems, increases in
nitrite and nitrate are usually on the order of 0.05 – 0.5 mg/L-N, but increases above 1 mg/L-N
have been noted in stagnant parts of some distribution systems(USEPA 2006; Wilczak et al. 1996).
Nitrite accumulation can possibly cause the potentially fatal condition of methemoglobinemia,
which is also called “blue baby syndrome” (Peavy et al. 1985). Nitrite has also been proven to
accelerate chloramine decay (Margerum et al. 1994; Valentine 1984). Stoichiometrically, 1
mg/L-N nitrite could consume 5 mg/L-Cl2 monochloramine (Table A.1, Reaction 6).
In most systems, 50 μg/L nitrite-nitrogen is set as a criterion indication of nitrification
(Kirmeyer 1995; Wilczak et al. 1996). Breakpoint chlorination is applied in the Metropolitan
Water District of Southern California when nitrite exceeds this level (Wolfe et al. 1988), but a
lower warning level of 25 μg/L was adopted for the Ann Arbor system because nitrification had
become a significant problem at this level (Skadsen 1993).
pH and alkalinity
The oxidation of ammonia to nitrite and then to nitrate by nitrifying bacteria reduces pH
and consumes large amounts of alkalinity. Reduction of pH and alkalinity can lead to violation of
USEPA Lead and Copper Rule either through failure to maintain designated optimal water quality
parameters, or through an action level exceedence at the tap (Odell et al. 1996). The extent of pH
and alkalinity changes caused by nitrification will depend on the amount of water buffering
capacity and nitrifying bacteria activity. In most drinking water systems, significant pH and
alkalinity changes were not observed (pH changes mostly ranged within ± 0.3) (Wilczak et al.
1996). Reduction of alkalinity by precipitation of CaCO3 on pipes and increasing of alkalinity by
dissolution of cement further complicate matters.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix A: Nitrification in Drinking Water Systems | 133
Heterotrophic Bacterial Growth and Coliform Growth
Heterotrophic bacterial growth is generally monitored by Heterotrophic Plate Count (HPC)
in drinking water systems (Clesceri et al. 1998). Increase in HPC has always been noted to
accompany nitrification occurrence (Powell 2004; Skadsen 1993; Wilczak et al. 1996; Wolfe et al.
1990). The HPC increase could be the result of the depletion of chloramines or supported growth
by organic carbon released from nitrifying bacteria (Wilczak et al. 1996). High HPC may be
associated with a variety of problems including proliferation of undesirable organisms and
aesthetic-compromised water quality (Yang et al. 2004). Systems experiencing nitrification might
not be able to meet the total coliform rule (TCR) (USEPA 2001b). Disinfectant residual decrease
during nitrification can lead to increased growth of coliform. Also, when certain control practices
are implemented, they may attack biofilm in the distribution system and lead to positive total
coliform samples. Two Florida utilities reported periodic violations of the TCR when free chlorine
was used for nitrification control after extended periods of chloramination (Wilczak et al. 1996).
Rapid Decay of Chloramines
The first indication of nitrification is often a difficultly in maintaining a constant
chloramine residual (Cunliffe 1991; Odell et al. 1996; Skadsen 1993; Wilczak et al. 1996; Wolfe et
al. 1988). A possible explanation for the accelerated chloramine decay could be: first, nitrite can
accelerate chloramines decay; second, ammonia oxidation can shift the equilibrium of
monochloramine formation so that chloramine is hydrolyzed as free ammonia (Cunliffe 1991).
The disappearance of the chloramine residual may cause a violation of disinfectant residual
standards in the Surface Water Treatment Rule. This rule requires the detection of a disinfectant
residual in at least 95% of monthly distribution system samples (Yang et al. 2004). Compliance
with the D/DBP rule may also be affected when utilities change their chloramine chemistry by
increasing the chloramine dose or increasing the chlorine to ammonia ratio. Systems using
chloramines at levels as high as 7 or 8 mg/L are at a high risk of violating D/DBP rules (Harms and
Owen 2004; Skadsen 1993).
Dissolved Oxygen (DO)
DO has been noted to decrease in distribution systems when nitrification is occurring
(Odell et al. 1996; Wilczak et al. 1996), sometimes by as much as 7 mg DO/mg NH3-N. This
significantly exceeds the stoichiometric level of 4.33 mg DO/mg NH3-N (Grady et al. 1999).
Although other reactions such as corrosion might remove oxygen, this high depletion indicated a
probability of recycled ammonia due to reactions with pipe materials (Table A.1). This recycling
reaction could regenerate ammonia and cause excessive nitrification and oxygen consumption.
Another explanation might be due to increased aerobic respiration by increased HPC.
Corrosion
Corrosion could be affected by nitrification through decreased pH and alkalinity (Table
A.4), and increases in bacteria growth are generally believed to increase corrosion (McNeill and
Edwards 2001). Nitrite levels produced by nitrification in drinking water system might stimulate
corrosion rate by shifting the redox potential (Rozenfeld 1981) Increased corrosion can result in
©2010 Water Research Foundation. ALL RIGHTS RESERVED
134 | Effect of Nitrification on Corrosion in the Distribution System
violation of USEPA Lead and Copper Rule (LCR), increased customer complaints about red water
(high level of iron release in drinking water due to iron pipe corrosion), and taste and odor
problems (Seidel et al. 2005).
Profound adverse impacts of nitrification on corrosion of lead pipe in drinking water were
first noted more than 100 years ago (Garret 1891) and the issue of nitrification in chloraminated
water supplies was reasonably described 70 years ago (Feben 1935; Hulbert 1933; Larson 1939).
Recently, work performed in Pinellas County, Florida by Powell et al. (2004) highlighted some
concerns related to iron corrosion control and red water. Furthermore, elevated copper levels at the
tap were clearly tied to activity of nitrifying bacteria in Willmar homes (Murphy et al. 1997b), and
nitrification was implicated in higher lead leaching in Ottawa (Douglas et al. 2004) and to some
extent in Washington D.C. (Edwards and Dudi 2004). The severe degradation of concrete by
nitrification is well-established in other industries (Bastidas and Sánchez-Silva 2006; Kaltwasser
1976). The destruction of concrete materials by nitrification is mainly from a loss of binding
material by acidification (Meincke et al. 1989).
NITRIFICATION MONITORING IN DRINKING WATER SYSTEMS
Biological Monitoring
MPN
Relatively little is known about the occurrence of nitrifying bacteria in drinking water due to
the difficulty in isolating and enumerating these organisms in environmental samples (Wolfe et al.
1988). Nitrifiers are most frequently enumerated by a most-probable-number (MPN) technique,
using a media selective for ammonia or nitrite oxidizers. The MPN tubes are generally incubated for
at least three weeks for ammonia oxidizers and up to 15 weeks for nitrite oxidizers. Unfortunately,
low recovery efficiencies, ranging from 0.1 to 5 % have been obtained with the MPN technique
(Wolfe et al. 1988). Despite this limitation, this technique is still the most common method used by
many researchers for nitrifier quantification (Fleming et al. 2005; Wolfe et al. 1990).
Molecular Techniques
Because of limitations in culturing, the use of nucleic acids to detect and identify AOB has
increased in the past few years (Regan 2001). For identification of nitrifiers, 16S rRNA (or 16S
rDNA), which is present in all bacteria but has regions of highly variable nucleotide sequences, is
commonly targeted. amoA is a gene that is only present in AOB, and it has been used increasingly
for AOB detection and characterization since 1995 (Regan 2001; Wagner et al. 1998; Wagner et al.
1995). Fluorescent in situ hybridization (FISH), typically targeting 16S rRNA involves fixation of
cells, hybridization with fluorescent-tagged complementary probes and analysis of fluorescence
signal (Regan 2001; Wagner et al. 1998; Wagner et al. 1995). Although FISH has been found to be
an effective tool for determining cell activity for some species, for many AOB, the rRNA and
mRNA content under starved conditions (low nitrifying activity) was not lower than that under
nutrient abundant conditions (high nitrifying activity), therefore, FISH, which detects nitrifier
activity by detecting 16S rRNA, is not particularly effective for determining AOB activity
(Bollmann et al. 2005; Wagner et al. 1995). Also, false positives might result due to lack of
specificity and false negatives may result due to the stringency issues with probes and low
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix A: Nitrification in Drinking Water Systems | 135
fluorescent responses (Yilmaz and Noguera 2004). The method is also limited due to the limitation
of existing databases (Regan 2001).
Polymerase Chain Reaction (PCR) is also used to identify and quantify nitrifiers, targeting
either 16S rDNA or amoA gene sequences (Baribeau 2006; Purkhold et al. 2003; Regan et al.
2007). PCR involves DNA extraction, amplification and sequencing or quantification of target
gene products (Purkhold et al. 2003; Regan et al. 2007). Erroneous information might also result in
PCR analysis during sample collection, cell lysis, DNA extraction and amplification
(Wintzingerode et al. 1997).
If one is interested in determining the diversity of nitrifiers, denaturing gradient gel
electrophoresis (DGGE), terminal restriction fragment length polymorphism (T-RFLP), cloning
and sequence analysis can be used (Baribeau 2006; Qin et al. 2007; Regan et al. 2002).
Fluorescent antibodies
Fluorescent-antibody (FA) technique can also be used to detect and enumerate nitrifier
strains (Belser and Schmidt 1978; Volsch et al. 1990) . This approach does not require lengthy
incubation periods like in culturing techniques, but it requires different FAs for different nitrifier
strains and analysis of a single sample could require excessive work (Baribeau 2006; Belser and
Schmidt 1978), also it can have non-specific binding of FA to extracellular polymeric substances
(Szwerinski et al. 1985) .
Cell mass
Nitrifier concentration can be determined directly by counting cell numbers under a
microscope using a counting chamber (Keen and Prosser 1987; Murphy et al. 1997b; Regan et al.
2007). This method however only counts total cell number and it is difficult to distinguish nitrifiers
in a mixed culture.
Optical density (turbidity) could also be used to measure biomass concentration. A good
linear relationship was found between dry matter concentration and optical density (Groeneweg et
al. 1994). In the Ann Arbor, Michigan system, a small negative correlation between turbidity and
monochloramine (monochloramine decreased during nitrification) was observed (Skadsen 1993).
Turbidity increased or remained the same in most of the systems investigated in the Wilczak et al.
(1996) survey.
Nitrification Indicator
Ideally, utilities would have a method of detecting nitrification early to allow the
implementation of control strategies before water quality degrades significantly. However,
traditional MPN methods are imperfect and require lengthy incubation times. The water industry
therefore relies on surrogate water quality parameters as early warning indicators (Feben 1935;
Regan 2001). Every indicator has a rationale for its use in nitrification detection (Table A.5), but
may not be applicable in certain systems (Table A.6). So, no single water quality parameter by
itself is a perfect indicator of nitrification, and several important factors need to be tracked
including nitrite, nitrate, chloramine dosage and residual, ammonia concentration, pH, HPC, and
dissolved oxygen (Wilczak et al. 1996).
©2010 Water Research Foundation. ALL RIGHTS RESERVED
136 | Effect of Nitrification on Corrosion in the Distribution System
CONTROL AND PREVENTION OF NITRIFICATION
The first and most important step in addressing nitrification is to conduct sufficient
monitoring to document normal fluctuations and identify abnormal values as early as possible and
then initiate control methods before severe nitrification problems occur (Harms and Owen 2004).
Different control methods might work differently in a specific environment (Table A.7).
Optimize Chloramine Dose
The easiest and most cost-effective means of nitrification control is to optimize the
chloramine dosage at the treatment plant (Lieu et al. 1993). This includes optimizing the ratio of
chlorine to ammonia or increasing chloramine dose. Both of these methods are not effective
control methods once nitrification has begun, but they can be used as preventative measures.
Increasing the ratio of chlorine to ammonia can reduce the amount of free ammonia available
as an energy source for nitrifiers. Free ammonia is almost completely eliminated at a 5:1 weight ratio
of Cl2: NH3-N (Kirmeyer et al. 1993). Ratios of 4 to 4.5: 1 or 5: 1 have been suggested to control
nitrification (Harms and Owen 2004). Optimizing chlorine to ammonia ratio (68%) is the most
common reported nitrification control strategy in a recent survey (Seidel et al. 2005). Increasing the
ratio of chlorine to ammonia may have precluded nitrification in Garvey Reservoir in 1986 (Wolfe et
al. 1988) but evidence suggests that nitrification can take place even when only small amounts of
ammonia are available (Odell et al. 1996). No trend was found between the ratio of chlorine to
ammonia and nitrification incidence in an industry survey (Wilczak et al. 1996), suggesting that this
strategy is not always successful. This is further illustrated by the experience of Ann Arbor, where a
ratio of 4.75:1 was unsuccessful due to poor control of ammonia (Skadsen 1993).
Nitrification can theoretically be prevented if the total chlorine concentration in a specific
location within the distribution system is greater than the value predicted when AOB growth and
inactivation rates are equal (Fleming et al. 2005). This method can only prevent the onset of
nitrification and the suggested value is between 2-4 mg/L (Harms and Owen 2004). However, it is
not effective when nitrification has begun, even with a high dose up to 8 mg/L (Fleming et al.
2005; Odell et al. 1996; Skadsen 1993). Also, long-term operation of increasing chloramine dose is
not feasible under the Stage 1 DBPR which sets a Maximum Residual Disinfectant Level (MRDL)
of 4 mg/L for chloramine (Harms and Owen 2004).
Adding booster stations to increase chloramine residual can be another control method
(Wolfe and Lieu 2001). Uncontrolled blending of chlorinated and chloraminated water could
occur near the booster station, and in some cases, cause breakpoint chlorination, increases in DBP
levels, or decreases in disinfectant residuals (USEPA 2005).
Regardless of chlorine dose and chlorine: ammonia ratio in the main distribution system,
problems can still occur in buildings, since chloramine can rapidly decay through reactions with
copper pipe (Nguyen and Edwards. 2005).
Breakpoint Chlorination
Breakpoint chlorination is the process of switching from chloramines to free chlorine
within the distribution system for a period of time (Harms and Owen 2004). Also known as a
“chlorine burn,” breakpoint chlorination is probably the most effective control measure for
nitrification once an episode is under way (Odell et al. 1996). About ¼ of utilities surveyed
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix A: Nitrification in Drinking Water Systems | 137
periodically switched to free chlorine to control nitrification; mostly once per year with the
duration ranging from less than a week to more than a month (Seidel et al. 2005). Breakpoint
chlorination has been found effective in different studies (Odell et al. 1996; Skadsen 1993) but the
switch to free chlorine can increase HPC or coliform growth, either due to free chlorine’s poor
ability to disinfect particle-associated bacteria from GAC (Skadsen 1993) or biofilm sloughed off
by the change (Odell et al. 1996). Other disadvantages to prolonged use of free chlorine are the
formation of disinfection by-products and consumer complaints about chlorinous taste (Ferguson
et al. 2005; Harms and Owen 2004).
Decrease Water Age
Operational practices that reduce water age can minimize nitrification. Increased turnover
is practiced in water reservoirs and flushing is practiced in distribution systems to reduce water age
(Odell et al. 1996; Wolfe et al. 1988). Flushing is the second most common practice (54%) for
nitrification control (Seidel et al. 2005). Flushing can also remove tubercles and sediments and
eliminate dead-ends, thus allowing the disinfectant better access to biofilms containing nitrifiers
(Harms and Owens, 2004). Flushing has been proved effective for short term nitrification control
(Odell et al. 1996; Skadsen 1993), and it has to be implemented frequently and regularly to be
effective (Wolfe and Lieu 2001). However, it is not possible for water utilities to control water age
in buildings, since flush times are under the control of residents (Edwards et al. 2005).
Upgrading Properties
Utilities can employ a number of operational and design measures including installing
recirculation facilities on standpipes and elevated storage, designing new reservoirs with inlet and outlet
pipes to prevent short-circuiting, retrofitting reservoirs with baffles to improve circulation, looping
dead-end mains and implementing regular flushing programs in problem areas (Odell et al. 1996).
Chlorite Ion (ClO2-)
Oxidation of ammonia by Nitrosomonas europaea is strongly inhibited by NaClO2 (Ki, 2
μM) (Hynes and Knowles 1983). McGuire et al. (1999) proposed the addition of chlorite or use of
chlorine dioxide (ClO2) to produce chlorite to inhibit the oxidation of ammonia and nitrite. Results
in full scale distribution systems proved that chlorite ion is likely to suppress nitrification.
However, the most recent studies by (Passantino 2003) and Karim & LeChevallier (2006) showed
that nitrification was not controlled by 0.5 mg/L chlorite especially after long-term application.
Chlorite and chlorine dioxide are also known to cause respiratory problems and irritation when
present at high concentrations and they are also regulated compounds (ATSDR 2005). The MCL
for chlorite in drinking water is 1 mg/L and the Maximum Residual Disinfectant Level (MRDL)
for chlorine dioxide is 0.8 mg/L (USEPA 2001a). Philosophically, it is difficult to justify addition
of one disinfection by-product (chlorite) to water in order to allow use of a disinfection strategy
designed to remove other disinfection by-products. Doing so implies a high degree of confidence
in the lower potential health detriments of chlorite versus detriments from the disinfection
by-products formed from free chlorine.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
138 | Effect of Nitrification on Corrosion in the Distribution System
Removing Organic Matter
Organic matter can react with chloramines through oxidation reactions (Reaction 4,
Table A.1) and different studies have proven the role of Natural Organic Matter (NOM) in
accelerating chloramine decay (Margerum et al. 1994; Song et al. 1999; Tomas 1987). Measures
designed to slow the decay of chloramine and its associated release of ammonia should
theoretically reduce the occurrence of AOB growth (Harrington et al. 2002). From this aspect,
removing organic compounds at the treatment plant has the potential to be effective for long-term
improvement of distribution system nitrification. In the Harrington et al. (2002) pilot scale study,
removal of NOM by enhanced coagulation delayed the onset of nitrification compared to
conventional coagulation. From another perspective, heterotrophic growth is suppressed by
removing organic matter, and nitrification might be promoted considering the competition
between nitrifiers and heterotrophs. Removal of organic matter to low levels is not emphasized in
the U.S. to the same extent as in some European countries, and further investigation is needed
(Odell et al. 1996).
Adjusting pH
Water with a high pH may reduce nitrification by creating suboptimal growth conditions
for AOB (Skadsen 2002). High pH can also reduce the rate of chloramine decay and formation of
free ammonia. Tomas (1987) stated that the rate of chloramine decay approximately doubles for a
drop of 0.7 pH units. Elevating the pH of the finished water to greater than 9.3 reduced the
frequency of nitrification in Ann Arbor (Skadsen 2002). Other water utilities also reported success
in controlling nitrification by increasing water pH to 9 (Gates and Lavinder 1997; Kirmeyer 1995).
Raising the pH can also help control corrosion and reduce lead and copper leaching to the water. It
also controlled taste and odor problems by promoting monochloramine versus dichloramine
formation (Skadsen 2002). On the other hand, high pH has been shown to decrease the
effectiveness of chloramine for inactivating AOB (Harrington et al. 2002; Oldenburg et al. 2002),
so the effect of pH in controlling nitrification is site specific (Oldenburg et al. 2002). In a pilot
scale study, a rough rank of pH onset time of nitrification at three different pHs is pH 8.5, 8.9 and
7.9 from earliest to latest (Harrington et al. 2002).
Control Nutrient Level
Theoretically, nitrifying bacteria growth could be limited when a specific nutrient is less
than optimal or inhibited when a nutrient is in excess. The levels of some nutrients present in U.S.
drinking waters are in a range relevant to nitrifying bacteria growth (Table A.3), so reducing a
nutrient level to limit nitrifying bacteria growth or increasing nutrient level (for nutrients not toxic
to human health) to inhibit nitrifying bacteria growth can be a possible approach to control
nitrification. This method has not been considered in drinking water systems and more research
should be conducted. Previous discussion in this paper also highlighted the potential importance of
this mode of control in copper pipe and galvanized plumbing often found in buildings.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix A: Nitrification in Drinking Water Systems | 139
REGULATIONS REGARDING CONTROL OF NITRIFICATION
The state of Florida has required utilities to monitor for nitrite in distribution systems since
the 1980s (Wolfe et al. 1988). EPA Phase II inorganic contaminant regulations require water
systems to sample for nitrite and nitrate at each entry point to the distribution system on at least an
annual basis. Additional monitoring is required on a quarterly basis for at least one year following
any one routine sample in which the measured concentration is greater than 50% of the MCL
(USEPA 2001a). Some consideration has been given to require more nitrite monitoring within
chloraminated distribution systems (USEPA 2003b). Chlorination is required annually by North
Carolina. The duration is from a few days to a few months (Harms and Owen 2004).
CONCLUSIONS
More water utilities are switching to chloramine as a secondary disinfectant to comply with
disinfectant rules and nitrification is a major concern in this implementation. Drinking water
systems provide a favorable environment for the growth of nitrifying bacteria. Nitrification
occurring in drinking water systems can cause water quality deterioration, corrosion, and difficulty
maintaining disinfectant residuals. Monitoring for nitrification could be conducted through a
culturing method, molecular method or from water quality indicators, but every method has
limitations. This poses a significant challenge to understand and control nitrification in drinking
water systems. Until now, optimizing disinfectant dosing, controlling water quality and improving
water facilities are the three major considerations for nitrification control; but these methods are all
site-specific and no consistently efficient methods have been proposed.
FIGURES AND TABLES
Negligible Nitrifiers
Heterotrophs dominant, but Nitrifiers
also grow in the system
Nitrifiers dominant
Figure A.1 Conceptual model illustrating coexistence of nitrifiers and heterotrophs
The top line (triangle) is plotted by using C/N = 10 of Verhagen and Laanbroek, 1991 directly, and
0.17 N
the bottom line (circle) is plotted by equation C =
, which is deducted by using
0.806 + 0.586 N
Monod equation and letting nitrifer growth rate equal to heterotroph growth rate (Appendix 1).
©2010 Water Research Foundation. ALL RIGHTS RESERVED
140 | Effect of Nitrification on Corrosion in the Distribution System
Inhibiting
Toxic
Nitrifier
Deficient Optimal
Nutrient
Figure A.2 Four states of nitrifier activity changing with metal concentration
Figure A.3 Free Cu (II) concentrations in relation to pH and phosphorus levels
Note: species calculated using 0.4 mM alkalinity, solubility constants cited from Schock
and Lytle, 1995.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix A: Nitrification in Drinking Water Systems | 141
NO3‐ Nitrifiers, 2
NO2 ‐ Nitrifiers, 2
NH2Cl
5, chloramine decay
by ferrous iron
11, Ammonia cycling by
elemental iron
NH4+ Fe 3+ + e -
Fe2+ + 2ePO23-, PO33-
Fe0
Figure A.4 Disinfectant, nitrifiers interaction and nutrient release on iron pipe surface
Note: Numbers marked in red represent reactions listed in Table A.1.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
142 | Effect of Nitrification on Corrosion in the Distribution System
Table A.1
Important reactions influencing nitrogen in the distribution system
Reaction Description
1. Formation of chloramines
Overall Reaction
NH3 + HOCl→ NH2Cl + H2O
Reference
Snoeyink and Jenkins 1980
NH4+ + 1.9 O2 + 0.069 CO2 + 0.0172 HCO30.0172 C5H7O2N + 0.983 NO2- + 0.966 H2O +
Grady et al. 1999
+
1.97 H
2. Ammonium and nitrite
utilization by nitrifiers
-
NO2 + 0.00875 NH4+ + 0.035CO2 + 0.00875
HCO3- + 0.456 O2 + 0.00875 H2O
C5H7O2N + 1.0 NO3
0.00875
Grady et al. 1999
-
3. Formation of ammonia via
Vikesland et al. 1998
3NH2Cl → N2+ NH3 + 3Cl- +3H+
chloramines decay
4. Release of ammonia through 1/10C H O N + NH Cl + 9/10 H O→ 4/10CO +
5 7 2
2
2
2
Woolschlager et al. 2001
oxidation of organic matter by
+
1/10HCO3 + 11/10NH4 +Cl
chloramines
5. Release of ammonium through
reaction of chloramines with
1/2NH2Cl + H+ + Fe2+→ Fe3+ + 1/2NH4+ +1/2 Cl- Vikesland and Valentine 2000
corrosion products at pipe surfaces
6. Release of ammonia through
Valentine 1984
NH2Cl + NO2-+ H2O→ NH3 + NO3- + HCl
oxidation of nitrite by chloramines
7. Breakpoint reactions with free
2NH2Cl + HOCl→ N2 + H2O + 3HCl
Snoeyink and Jenkins 1980
chlorine
8. Direct metabolism by ammoniaWoolschlager et al. 2001
NH2Cl+ O2→ NO2 + 2H+ + Cloxidizing bacteria
9. Ammonia oxidation by
Jetten et al. 1997
NH3+ NO2-→ N2 + H2O
anaerobic bacteria
Pb + NO3-→ NO2- + PbO
10. Cycling of ammonia through
corrosion with metallic lead
11. Cycling of ammonia through
corrosion with metallic iron
Uchida and Okuwaki 1998
-
Uchida and Okuwaki 1998
3Pb + H2O + 2NO2-→ N2 + 3PbO + 2OH-
Uchida and Okuwaki 1998
4 Fe° + NO3- + 10 H+ = 4 Fe 2+ + NH4+ + 3H2O
Huang and Zhang 2005
-
3Pb + 2H2O + NO2 → NH3 + 3PbO + OH
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix A: Nitrification in Drinking Water Systems | 143
Table A.2
pH-dependent growth range and optimal pH for Nitrifying Bacteria
pH Range
Optimal pH
Nitrifiers strain or Nitrification Occurrence
Environment
6-10
7-8
Not specific strain, but in pure culture
4.6-9
7.5-8.1
Not specified
<11
6.7-7.0
Pure Nitrosomonas europaea cultured in
suspension, fermenter sparged with CO2
enriched air
Ammonia oxidation rate
Groeneweg et al. 1994
7.2-9.8
7.2-8.5
Nitrification in Drinking water system
Utility Survey
Odell et al. 1996
6.5-9.5
n.d.
Nitrification in Drinking water system
Utility Survey
Wilczak et al. 1996
n.d.
7-8
Nitrosomonas
n.d.
7.5-8
Nitrobacter
n.d.
8.1
n.d.
7.9
n.d.
7.82
high nitrification rate at
pH below 4
Nitrification Indication Parameters
N/A, generalized from Previous
literature
N/A, generalized from Previous
literature
N/A, generalized from Previous
literature
N/A, generalized from Previous
literature
Pure Nitrosomonas suspension culture,
grown in fermenter with aeration
Pure Nitrobacter suspension culture,
grown in fermenter with aeration
Nitrobacter winogradskyi suspension
culture, grown in continuous dialysis
Mixed culture in both suspended growth
and biofilm growth with oxygen
Ammonia oxidation rate
Nitrite oxidation rate
References
Painter 1977
Wolfe and Lieu 2001
USEPA 2005
USEPA 2005
Grunditz and Dalhammar
2001
Grunditz and Dalhammar
2001
Oxygen Uptake Rate
Boon and Laudelout 1962
Ammonia oxidation rate
Tarre and Green 2004
©2010 Water Research Foundation. ALL RIGHTS RESERVED
144 | Effect of Nitrification on Corrosion in the Distribution System
Table A.3
Limiting and Inhibiting Range of Various Nutrients and Their Occurrence Range in US
Raw Drinking Water
Element
Stimulating
Concentration Range
Culture purity (Growth
(specific
Condition)
stimulating/inhibting extent)
0.1 ppm
Nitrosomonas europaea
Tomlinson 1966
0.005-0.03 ppm
Nitrosomonas europaea
Loveless and Painter 1968
> 0.1 ppm
Nitrosomonas europaea
Skinner and Walker 1961
0.05- 0.56 ppm
Nitrosomonas europaea
Loveless and Painter 1968
4 ppm had 75 % inhibition
0.01 ppm had 5 % inhibition
Cu (II)
Inhibiting
Inhibiting
Stimulating
Mg (II)
K (I)
Cr
Inhibiting
Inhibiting
Inhibiting
Tomlinson 1966
0- 1 ppm
Waara and Wilander 1985
Not specified
Martin and Richard 1982
1 ppm
Activated sludge
Activated sludge, long term
exposure
Activated sludge , short term
exposure
Pettet 1956
200 ppm had 75 % inhibition
Ni (II)
Nitrosomonas europaea ,
short term exposure
Wastewater Treatment
Inoculum
> 0.4 ppm
> 20 ppm
Stimulating
Occurrence in US. Raw
Drinking Water 1
References
Tomlinson 1966
Tomlinson 1966
640 ppm
Pure Nitrosomonas
Meikleohn 1954
0.5 ppm
Activated sludge
Disalvo & Sherrard 1980
> 0.25 ppm
Nitrosomonas europaea
Skinner and Walker 1961
11.8 ppm complete inhibition
Pure Nitrosomonas
Meikleohn 1954
> 0.1 ppm
Not specified
Martin and Richard 1982
0.7 ppm
Activated sludge
Harper 1996
1 ppm
Activated sludge
Pettet 1956
1 ppm had 20 % inhibition
Activated sludge
Ibrahim 1989
12 ppm had 88 % inhibition
Activated sludge
Martin and Richard 1982
23.6 ppm
Activated sludge, AOB
Tomlinson 1966
295 ppm
Activated sludge, NOB
Tomlinson 1966
12.5-50 ppm
Nitrosomonas europaea
Loveless and Painter 1968
> 50 ppm
Nitrosomonas europaea
Loveless and Painter 1968
12 g/L complete inhibition
Pure Nitrosomonas
Meikleohn 1954
19.5 g/L complete inhibition
Pure Nitrosomonas
Meikleohn 1954
> 0.25 ppm Cr (III)
Nitrosomonas europaea
Skinner and Walker 1961
0.3 ppm Cr (III)
Activated sludge
Harper 1996
> 1 ppm (II)
Not specified
Martin and Richard 1982
1 ppm had 10 % inhibition
> 100 ppm Cr (VI) had 75 %
inhibition
Activated sludge
Ibrahim 1989
Activated sludge
Tomlinson 1966
High ammonia, Low
biodegradable organic
leachate
Harper 1996
> 0.3 ppm Cr (III)
0- 36 ppb
0- 107 ppm
0- 41.5 ppm
0- 47.1 ppb
(continued)
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix A: Nitrification in Drinking Water Systems | 145
Table A.3 (Continued)
Element
Concentration Range
Culture purity (Growth
(specific
Condition)
stimulating/inhibting extent)
Up to 1 ppm had no effect
Nitrosomonas europaea
Skinner and Walker 1961
0.08- 0.5 ppm
Loveless and Painter 1968
> 1 ppm
Nitrosomonas europaea
Wastewater Treatment
Inoculum
Activated sludge
11 ppm had 25 % inhibition
Activated sludge
Martin and Richard 1982
> 10 ppm
Not specified
Martin and Richard 1982
0.06-0.15 % (600-1500 ppm)
Nitrosomonas europaea
Loveless and Painter 1968
0.7 % (7000 ppm)
Nitrosomonas europaea
Loveless and Painter 1968
11.5 g/L complete inhibition
Increasing from 0.5 ppm to 0.6
ppm chelated Fe (II)
Pure Nitrosomonas
Meikleohn 1954
Nitrosomonas europaea
Skinner and Walker 1961
3 ppm had 80 % inhibition
Zn (II)
Inhibiting
References
Waara and Wilander 1985
Pettet 1956
0.65 ppm
Stimulating
Na (I)
Inhibiting
Stimulating
Fe
Co (II)
Pb (II)
Cd (II)
Mo (VI)
59 ppm had 60% inhibition
0.005-0.5 ppm, no effect
> 0.5-1 ppm
Not specified
> 1 ppm, 1.7 ppm had 90%
inhibition
2. 1 ppm
0.1 ppm had 18% inhibition
0.25 ppm
Wastewater Treatment
Inoculum
Pure Nitrosomonas
Activated sludge
Nitrosomonas europaea
< 75 ppb had < 10% inhibition,
500 ppb had 85% inhibition
Wastewater Treatment
Inoculum
14.3 ppm had 42% inhibition
Activated sludge
Unspecified Nitrobacter
Strain
Unspecified Nitrobacter
Strain
Unspecified Nitrobacter
Strain
6 ppm
Inhibiting
560 ppm
Up to 1 ppm had no effect
0.08-0.5 ppm
Inhibiting
0.59 ppm
Inhibiting
Inhibiting
Stimulating
0.096-960 ppb
Stimulating
1.84 ppb
Inhibiting
1.84 ppm
W (VI)
P (V)
stimulating
> 3-20 ppb-P
Nitrosomonas europaea
> 50 ppb-P
Nitrobacter agilis
0.05 ppm is enough
0.3 ppm is necessary
0.5- 20 ppm is enough
8000 ppm complete inhibition
Nitrosomonas europaea
Meikleohn 1954
Meikleohn 1954
Skinner and Walker 1961
Loveless and Painter 1968
Finstein and Delwiche
1965
Inhibiting
0- 500 ppm
0- 22 ppm
0- 7.0 ppb
Meikleohn 1954
Loveless and Painter 1968
Martin and Richard 1982
Waara and Wilander 1985
0- 8.3 ppb, but corrosion
of lead pipes could
contribute lead to the water
Meikleohn 1954
Ibrahim 1989
Skinner and Walker 1961
Waara and Wilander 1985
0- 0.3 ppb
Martin and Richard 1982
Finstein and Delwiche
1965
0- 75 ppb
Zavarzin 1958
Finstein and Delwiche
1965
Van Droogenbroeck and
Laudelout 1967
Aleem and Alexander 1960
Mn
Ca (II)
0- 614 ppb
Meikleohn 1954
Pure Nitrosomonas and
Nitrobacter
Pure Nitrosomonas
Nitrosomonas europaea
Nitrosomonas europaea
Unspecified Nitrobacter
Strain
Pure Nitrosomonas
Nitrosomonas europaea
Optimal
Occurrence in US. Raw
Drinking Water 1
0- 1 ppm
0- 1.2 ppm
Nitrosomonas europaea
Pure Nitrosomonas
Skinner and Walker 1961
Boltjes 1935
Loveless and Painter 1968
Meikleohn 1954
1: (Parks et al. 2004)
©2010 Water Research Foundation. ALL RIGHTS RESERVED
0- 210 ppm
146 | Effect of Nitrification on Corrosion in the Distribution System
Table A.4
Key effects of nitrification on water quality and corrosion
Change due
to
nitrification
Possible
direct/indirect effect
on water quality
Nitrite
Nitrite MCL,
increased lead from
brass, disinfectant
loss
production
Possible effect on
corrosion-materials
degradation
Increased microbial corrosion,
nitrite catalyzed stress
corrosion failures and attack
on grain boundaries (brass)*
Other Concerns
Samples for nitrite MCL not
collected in premise plumbing,
so potential problem can be
missed.
Taste & Odor from
corrosion, red and
blue water
complaints.
Divergence from targeted
optimal corrosion control
relative to finished water
Lead and Copper Rule, Toxicity
from Blue Water typically
occurs in new homes not tested
in LCR
Concern over
pathogen re-growth
and loss of
disinfection
More microbial corrosion,
likely link to some cases of
pinhole leaks in copper tube
Total coliform rule, HPC action
levels, Legionella and
Mycobacterium as emerging
issues
Rapid decay
of chloramine
Failure to maintain
residual at distant
parts of distribution
system
Effect on corrosion rates
dependent on relative
corrosivity of chloramine vs.
decay products
Decreased
DO
Low redox in iron
pipe associated with
more red water
Lower pH,
alkalinity and
DIC
Higher HPC
Highly corrosive sulfate
reducing bacteria are
anaerobic
Chlorine residual not routinely
monitored in premise plumbing,
where it controls opportunistic
pathogens
Lead and Copper Rule, color,
taste and odor complaints
*(Guo et al. 2002; Larson et al. 1956; Pugh et al. 1966; Sundberg et al. 2003)
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix A: Nitrification in Drinking Water Systems | 147
Table A.5
Nitrification Indicator and rationale for its use
Nitrification Indicator
Nitrite
Rationale
References
Nitrite production corresponds with the viable cell Engel and Alexander 1958; Lieu et
count of Nitrosomonas europaea , but may lag
al. 1993;Wilczak et al. 1996
behind MPN method
Nitrogen species and
Decreased total ammonia concentrations
Wilczak et al. 1996
balance
accompanies nitrite and nitrate increase during
nitrification
Difficulty in maintaining
nitrification causes rapid loss of chloramine
Cunliffe 1991; Odell et al. 1996;
disinfectant residual
residual
Skadsen 1993; Wilczak et al. 1996;
Wolfe et al. 1988
Hill 1946; Larson 1939
pH and alkalinity
Nitrification reduces pH and consume large amount
of alkalinity
HPC
HPC increase accompany nitrification and a linear Wolfe et al. 1990; Skadsen 1993;
relationship exists between AOB and HPC when
Odell et al 1996
above 350 cfu/ml
DO
4.33 mg oxygen is consumed per mg ammonia
Grady et al. 1999; Wilczak et al.
oxidized during complete nitrification;
1996; Odell et al. 1996
Table A.6
Situations where nitrification indicators do not always work
Indicator
Pitfalls of Nitrification Indicators
Disinfectant residual Copper pipes-chloramines decay fast with/without nitrification; dead endschloramines are too low to compare decay with/without nitrification
pH
Concrete pipes-release high alkalinity; Iron pipe-consume H+ through
equation (4); high alkalinity water
Alkalinity
Concrete pipes-release high alkalinity; Iron and other pipes- nitrification
effect is trivial compared to scale deposition during corrosion; high
alkalinity water
Ammonia
Iron/lead pipes-recycle nitrification product back to ammonia; ammonia
loss through nitrification is insignificant compared to other pathways
Nitrite/Nitrate
Iron/lead pipes-recycle nitrite/nitrate back to ammonia; GAC-removes
Nitrite/Nitrate ions
Nitrogen balance
Situations where organic nitrogen is present are hard to quantify
HPC
Dead ends-cannot tell if HPC increase is caused by chloramines depletion
or nitrification; HPC distribution between water and biofilms on pipes
varies due to water quality, flushing program, etc
DO
Dead ends-DO is affected more by the water flowing frequency; situations
where pipe corrosion affects DO more than nitrification
©2010 Water Research Foundation. ALL RIGHTS RESERVED
148 | Effect of Nitrification on Corrosion in the Distribution System
Table A.7
Nitrification control methods: effectiveness and rationale
Control Method
Increase chlorine to
ammonia ratio
Increase chloramine dose
Likely effect in premise
plumbing
Sometimes effective
Sometimes effective
Breakpoint chlorination
Sometimes effective
Decrease water age
(flushing)
Enhance facility property
(better equipment design)
Using chlorite
Always effective, but for
short term
Always effective
Remove organic matter
Adjust pH
Effective but use with
caution
Sometimes effective
Sometimes effective
Control nutrient level
Sometimes effective
Rationale
Reduces ammonia concentration, but nitrifiers could grow even
with small amount of ammonia
Effective at preventing onset of nitrification, but not effective at
controlling it once it has begun since nitrite increases
chloramine decay rate
Chlorine can inactivate nitrifiers, and no ammonia provided
compared to chloramine, but sometimes chlorine completely
decays before it gets to premise plumbing
Nitrifiers can be flushed from system, but could reestablish
between flushing intervals
Better design will mean fewer stagnation and dead end areas
Chlorite can control nitrification, but can be toxic in high levels
Reduces chloramine decay and heterotrophic growth
Competing effects of increasing pH: 1. reducing nitrifier growth,
reducing chloramine decay; 2. reducing chloramine inactivation
effect
Some nutrient levels may be in a range that is impractical to
control
SUPPORT INFORMATION-CONCEPTUAL MODEL TO PREDICT REGIONS OF
NITRIFIER AND HETEROTROPHIC EXISTENCE
1. Conceptual model based on Verhagen and Laanbroek’s theory that the existence of
nitrifiers depends on whether heterotrophic bacteria are limited by carbon or nitrogen.
If heterotrophic growth is limited by carbon, nitrifiers can coexist with heterotrophs and
the heterotrophic growth rate is:
C
, where µmax is the maximum specific growth rate of heterotrophs, Ks is
µ1 =µmax •
Ks + C
the half saturation constant for carbon, C is the carbon concentration.
If heterotrophic growth rate is limited by ammonia nitrogen, no ammonia is available for
nitrifiers and therefore nitrifier growth is negligible. In this case the heterotrophic growth rate is:
N
, where µmax is the maximum specific growth rate of heterotrophs, KN
µ2 =µmax •
KN + N
is the half saturation constant for ammonia nitrogen, N is the ammonia concentration.
µ1= µ2 when the carbon-to-nitrogen ratio is at a critical value:
C KS
=
N KN
By applying known kinetic parameters to a system, the critical C/N ratio can be calculated,
and nitrification occurrence in a system can be predicted, specifically, nitrification is not likely to
occur if C/N is above the critical value and nitrification is possible if C/N is below the critical
value. Verhagen and Laanbroek, 1991 measured a critical C/N ratio of about 10. Plot this line for
carbon and nitrogen levels in drinking water (Figure A.1), and if the C and N concentrations are
above this line, nitrifiers are negligible in the system. If C and N concentrations are below this line,
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix A: Nitrification in Drinking Water Systems | 149
both nitrifying bacteria and heterotrophic bacteria grow in the system, and which one dominates
can be defined by the following equation:
2. Compare relative growth rate of nitrifiers and heterotrophs-Dominancy of growth
depends on the relative growth rate of heterotrophic bacteria and nitrifiers.
At 20ºC, nitrifier growth is limited by ammonia, assuming µmax = 0.034 hr-1, Ks = 1.3 mg/L
(Grady et al., 1999),
N
µ =0.034 •
, where µmax is the maximum specific growth rate for nitrifiers, Ks is
1 .3 + N
the half saturation constant, N is the ammonia concentration.
Heterotrophic growth is limited by carbon, assuming µmax = 0.62 hr-1, Ks = 5 mg/L as COD
(Grady et al., 1999),
C
µ =0.62 •
, where C is the organic carbon concentration
5+C
Let nitrifer growth rate equal heterotroph growth rate:
0.17 N
C=
0.806 + 0.586 N
Plot this line for carbon and nitrogen levels in drinking water (Figure A.1), and if C and N
concentrations are above this line, heterotrophic bacteria will overgrow nitrifiers and dominate the
system. If C and N concentrations are below this line, nitrifying bacteria will overgrow
heterotrophs and dominate the system.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
150 | Effect of Nitrification on Corrosion in the Distribution System
©2010 Water Research Foundation. ALL RIGHTS RESERVED
APPENDIX B
NITRIFICATION IN PREMISE PLUMBING: ROLE OFPHOSPHATE, PH,
AND PIPE CORROSION
Yan Zhang, Allian Griffin, Marc Edwards
Reprinted with permission from Environmental Science & Technology 2008, 42 (12), 4280-4284.
Copyright 2008 American Chemical Society
ABSTRACT
Nitrification in PVC premise plumbing is a weak function of pH over the range 6.5-8.5 and
is insensitive to phosphate concentrations 5-1000 ppb. Lead pipe enhanced nitrification relative to
PVC, consistent with expectations that nitrifiers could benefit from ammonia recycled from nitrate
via lead corrosion. Relatively new copper pipe (< 1.5 years old) did not allow nitrifiers to establish,
but nitrifiers gradually colonized over a period of months in brass pipes, when copper
concentrations were reduced by pH adjustment or orthophosphate. Nitrifiers were inhibited by
trace copper, but not by lead levels up to 8,000 ppb. In some systems using chloramines, brass in
plastic plumbing systems might be more susceptible to lead/copper leaching, and accelerated
dezincification, due to lower pH values resulting from nitrification.
INTRODUCTION
Preventing water quality degradation in potable water premise plumbing systems is an
underappreciated challenge and a high priority for future research (Edwards et al. 2003; Snoeyink
et al. 2006). Maintenance of safe water stored in buildings is of world-wide concern in situations
where potable water is distributed to homes from central treatment plants or even produced on-site
(Snoeyink et al. 2006; Sobsey et al. 2003). The type of degradation to water quality that can occur
during storage and its ultimate significance to public health is controlled by a complex interplay of
materials, corrosion, microbiology, aquatic chemistry and other factors.
In the United States utilities are increasingly using chloramine to comply with regulations
for disinfection by-products (Seidel et al. 2005). The ammonia formed via chloramine decay can
support autotrophic microbial nitrification. Organic carbon and acid produced via nitrification can
stimulate growth of heterotrophic bacteria, contribute to loss of disinfectant, and also create
problems with lead and copper contamination from corrosion of premise plumbing systems
(Edwards and Dudi 2004; Harrington 2002; Murphy et al. 1997b; Powell 2004; Skadsen 1993;
Wilczak et al. 1996; Wolfe et al. 1990; Zhang et al. 2009a). Nitrification also creates nitrite which
has a relatively low maximum contaminant level (MCL) of 1 mg/L-N. Prevention of nitrification is
therefore deemed to have desirable public health consequences.
With the exception of the Lead and Copper Rule (LCR) (USEPA 1991), routine
distribution system monitoring for nitrification products and other contaminants in the United
States stipulates extensive pre-flushing of water from premise plumbing lines before sample
collection. As a result, even if water samples are collected from building taps, they rarely quantify
the extent of water quality degradation occurring during storage. Only a few reports have
discussed occurrence of nitrification during storage in premise plumbing (Murphy et al. 1997a),
151
©2010 Water Research Foundation. ALL RIGHTS RESERVED
152 | Effect of Nitrification on Corrosion in the Distribution System
and neither the prevalence of the problem or key factors that contribute to its occurrence are
understood.
A recent literature review predicted that the incidence of nitrification in premise plumbing
systems would be at least partly controlled by water chemistry and materials usage (Zhang et al.
2009a). Copper and brass plumbing materials invariably leach copper to water supplies at
concentrations dependent on numerous factors including pipe age and water chemistry. In the
range of copper known to occur in water of premise plumbing systems (0-5000 ppb), slight
increases in copper (≈ 1-10 ppb) from corrosion could stimulate nitrification, whereas higher
levels of copper (above about 100 ppb) could inhibit nitrification (Zhang et al. 2009a). Corrosion
of copper and brass plumbing used in many buildings may therefore be preventing degradation of
water quality from nitrification in at least some instances, and well-intentioned actions to
minimize copper leaching to water (increasing pH, adding orthophosphate corrosion inhibitors or
using only plastic pipes) might increase the likelihood of establishing nitrifiers.
Lead materials, in contrast, are capable of converting products of nitrification (nitrite and
nitrate) back to ammonia via anodic corrosion reactions (Pb0 Æ Pb+2 + 2e-) (Uchida and Okuwaki
1998). Zinc in galvanized pipe has also been reported to convert nitrite to ammonia under drinking
water conditions (Kunzler and Schwenk 1983). It is therefore possible that nitrifier growth on lead
and zinc alloy surfaces would be favored relative to more inert surfaces such as polyvinyl chloride
(PVC), exacerbating nitrification occurrence and undermining efforts to mitigate lead
contamination of water supplies from pH adjustment (Edwards and Dudi 2004; Zhang et al. 2009a)
The goal of this work is to investigate the interplay between materials selection, pH, and
orthophosphate corrosion inhibitor dosing on nitrification occurrence as it occurs in home
plumbing systems.
EXPERIMENTAL SECTION
Pipe Materials
Pipe materials included PVC (Polyvinyl chloride), copper, lead and non-leaded brass.
PVC, copper and lead pipe sections were 30 cm length × 1.9 cm diameter and brass pipe sections
were 30 cm in length × 1.3 cm diameter. PVC and copper pipes were purchased from a local
hardware store, and pure lead and brass pipes were specially fabricated. Water in the pipes was
changed every 3.5 days (twice a week) using a “dump and fill” protocol to simulate typical water
use and replenish nutrients for microbial nitrifier growth. These pipes were maintained at room
temperature throughout the entire experiment.
Water Chemistry
Pipes were first conditioned using synthesized potable water. This synthesized water
contained MgSO4 (1 ppm-Mg), CaCl2 (4.9 ppm-Ca), KCl (10 ppm-K), Na2HPO4 (1 ppm-P),
NaHCO3 (2 mM) and other trace nutrients (5 ppb Cu2+, 1.7 ppb Mo6+, 0.1 ppb Co2+, 5.6 ppb Mn2+,
2.6 ppb Zn2+ and 0.1 ppm Fe2+). The pH was adjusted to 8.6 before filling up the pipes. This water
simulates typical drinking water conditions and provides essential nutrients for nitrifying bacteria.
After a year exposure and conditioning of the pipes, no nitrification/nitrifier activity could be
detected, as would be expected given that nitrifiers were never inoculated to the rig and no
ammonia was in the water. Pipes were then exposed to the same water with addition of 2 ppm-N
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix B: Nitrification in Premise Plumbing: Role ofPhosphate, Ph, and Pipe Corrosion | 153
(NH4)2SO4 and 10 % dechlorinated Blacksburg, VA tap water to provide opportunity for nitrifiers
to establish (Supporting Information-Figure S.1).
After complete nitrification was established in PVC, brass and lead pipes (as measured by
100 % conversion of ammonia to nitrite or nitrate), the inoculation with Blacksburg, VA water was
stopped and only the synthesized water was used thereafter. Use of this water was continued for
three additional weeks at initial pH 8.6. Thereafter, pipes were divided into three groups destined
to receive either 5, 60 or 1000 ppb PO4-P. These levels are representative of phosphate occurrence
in potable water supplies considering use of orthophosphate as a corrosion inhibitor. A total of 60
pipes were used (4 materials × 3 phosphate concentrations × 5 replicates). After exposing the pipes
to water with an initial pH of 8.6 for one month, pH was decreased stepwise in 0.5 unit increments
(from 8.6 to 6.5) with a one-month exposure at each pH level.
Analytic Methods
Water samples were collected before and after introduction to the pipes each week so that
nitrification activity and associated water quality changes could be quantified. Nitrifier activity
was tracked by measuring loss of ammonia, production of nitrite and nitrate and reduction of pH.
pH was monitored by using pH electrode according to Standard Method 4500 H+ B (Clesceri et al.
1998). NH4-N was measured with salicylate method using a HACH DR/2400 spectrophotometer,
according to Standard Method 4500 NH3 (Clesceri et al. 1998). NO2-N and NO3-N were measured
using DIONEX, DX-120 ion chromatography (IC), according to Standard Method 4110 (Clesceri
et al. 1998). Nitrifier density was monitored once at the third week of each pH level (no monitoring
was conducted at pH 8.6) by a five-tube Most Probable Number (MPN) procedure (Wolfe et al.
1990). Five tenfold serial dilutions were used, resulting in concentrations of 10-1 to 10-5. Each tube
contained 9 ml medium inoculated with 1 ml of sample or 1 ml from the preceding dilution
(Supporting Information Figure S.2). The medium and dilution water used was modified from
(Wolfe et al. 1990) by adding 0.084 g NaHCO3, 0.0001 g Na2MoO4•2H2O, 0.000172 g
MnSO4•2H2O, 0.000004 g CoCl2•6H2O, 0.0001 g ZnSO4•7H2O and 0.0000133 g CuCl2•2H2O;
and no phenol red solution was added. The medium pH was adjusted to 8 and was sterilized by
autoclaving before use. The tubes were incubated at 30 °C in the dark for three weeks and then
tested for nitrite/nitrate using sulfanilic acid and N,N-dimethyl-α-naphthylamine, and zinc dust
(Mac Faddin 2000). MPN index was calculated according to standard method 9221 C (Clesceri et
al. 1998), which gives most probable nitrifier numbers and corresponding 95 % confidence limits.
The lowest reportable MPN index was 20/100 mL (Regan et al. 2007).
Bulk water heterotrophic bacteria were monitored with Heterotrophic Plate Count (HPC)
according to Standard Method 9215 (Clesceri et al. 1998) using the spread plate method with R2A
medium. Total Organic Carbon (TOC) was analyzed using a SIEVERS 800 Total Organic
Analyzer according to Standard Method 5310C (Clesceri et al. 1998). Dissolved oxygen was
quantified at each pH level according to Standard Method 4500 O G (Clesceri et al. 1998) using a
Dissolved Oxygen Meter YSI Model 58. Soluble and total metal release was also quantified.
Soluble metal concentration was operationally defined by filtration though a 0.45 µm pore size
syringe filter. Total metal release was quantified by digesting samples with 2% nitric acid in a 90
°C oven. Metal concentrations (Cu, Zn, Pb) and phosphorus levels were quantified using an
Inductively Coupled Plasma Mass Spectrophotometer (ICP-MS) according to Standard Method
3125-B (Clesceri et al. 1998).
©2010 Water Research Foundation. ALL RIGHTS RESERVED
154 | Effect of Nitrification on Corrosion in the Distribution System
RESULTS AND DISCUSSIONS
Nitrification in Different Pipe Materials
Nitrifier Establishment
At pH 8.6, nitrifier activity (as measured by ammonia loss during 3.5 day stagnation) was
established first in lead pipes, followed by PVC and then brass (Figure B.1). Ammonia loss was
100 % after just 10 days in lead pipes and 38 days in PVC pipes (Figure B.1). Ammonia loss in
brass pipes was non-detectable initially, but markedly increased after 122 days and eventually rose
to 100% after 143 days. In copper pipes, ammonia loss was never above 30% throughout the 164
days’ inoculation period, consistent with expectations of copper toxicity to nitrifiers (Zhang et al.
2009a).
The increase of nitrifier activity with time in brass pipes occurred only after levels of
copper leaching to water dropped below about 0.1 ppm as the pipe metal aged (Figure B.2).
Interestingly, total lead release up to 8000 ppb and soluble lead release up to 220 ppb in the lead
pipes seemed to have no inhibitory effect on nitrification. It is unclear whether the enhanced
nitrification observed in the early stages of the experiment on lead pipe versus PVC was
attributable to benefits of nutrient cycling via lead corrosion or other possible factors such as
increased surface roughness that aid colonization of the pipe surface.
Changes in Water Chemistry
The occurrence of nitrification significantly impacted water chemistry during stagnation
(Supporting Information-Table S.1). For example, when the initial pH of water was 8.0, final pH
decreased by 0.75 unit in PVC and 0.5 unit in bulk water from lead pipes. Consideration of
nitrification stoichiometry, acid production and water chemistry (Zhang et al. 2009a) predicts a pH
drop of up to 0.9 unit in this water from complete nitrification, so observed drops in pH are
consistent with expectations, although other chemical reactions impacting pH are also occurring
(Zhang and Edwards 2007). Dissolved oxygen (DO) decreased by up to 2.5 mg/L in PVC, lead and
brass pipes, while there was very little change in DO in copper pipes. This study did not find higher
bulk water HPC associated with nitrification (Supporting Information-Table S.1), contrary to
some earlier studies (Powell 2004; Skadsen 1993; Wilczak et al. 1996; Wolfe et al. 1990). But the
morphology of bulk water bacterial colonies plated from water in copper pipes, where no
nitrification was observed, was clearly different from those plated from the lead, PVC and brass
pipes (Supporting Information-Figure S.3).
TOC was present in the initial water at 8-60 ppb and in some situations increased more than
ten-fold after 3.5 days of stagnation (Figure B.3). Amongst the four pipe materials, brass had the
largest increase in TOC, followed by lead and then copper and PVC (Figure B.3). The TOC levels
detected in brass and lead pipes far exceeds any reasonable expectation of TOC production due to
nitrification alone even with cycling of nutrients, and suggests that other forms of autotrophic
growth or TOC creation may be occurring. That might include evolution of H2 substrate from
metal corrosion or other pathways. The higher levels of TOC could reflect a short-term sloughing
of biofilm on the days of sampling.
It has previously been observed that distribution system materials with higher surface area
and higher corrosion rates support more biofilm growth and biofilm sloughing (Camper et al.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix B: Nitrification in Premise Plumbing: Role ofPhosphate, Ph, and Pipe Corrosion | 155
2003). Likewise, in this work for lead and brass, it is possible that higher rates of metal corrosion
could translate to more cycling of nitrate and more biofilm. Consistent with that idea, TOC tended
to be lower in the presence of more phosphate corrosion inhibitor for lead and brass (Edwards et al.
2002; McNeill and Edwards 2002; Schock et al. 1995). If lower phosphate acted by inducing a
nutrient limitation to nitrification (Van Droogenbroeck and Laudelout 1967) the opposite trend
should have been observed (Figure B.3).
Effect of pH and Phosphate Levels on Metal Release
pH and phosphate can directly impact nitrifier growth, since the optimum pH for
nitrification is generally believed to be 7.5-8 and phosphate is an essential nutrient (Zhang et al.
2009a). But pH and phosphate also influence total metal leaching to water via corrosion of lead,
copper and brass, as well as the percentage of metal that is present as the free ion (e.g., Cu+2, Pb+2
and Zn+2) in water (Edwards et al. 2002; Schock et al. 1995). Thus, phosphate and pH were
expected to influence nitrification occurrence in premise plumbing where copper is the most
common pipe material. Currently, about 50% of U.S. utilities use some form of phosphate
corrosion inhibitor (McNeill and Edwards 2002).
Models were recently used to predict the role of phosphate and pH in controlling free
[Cu2+] levels in water and their likelihood to impact nitrification (Zhang et al. 2009a). As pH
decreases, free Cu+2 deemed especially toxic to nitrifiers increases. But at a given pH, phosphate
dosing to water can decrease the concentration of total copper and Cu+2 by reducing the solubility
product of corrosion products on the pipe surface. Zhang et al, 2009 used solubility models (Zhang
et al. 2009a) and predicted that pH levels below about 7.3 would be highly toxic for nitrifiers in
copper pipe, but the presence of high orthophosphate (e.g., 1 ppm at PO4-P) nitrification could
allow nitrification at much lower pH levels.
Copper Pipes
At all phosphorus levels tested, total copper release increased as pH decreased (Figure
B.4), consistent with expectations from earlier research (AWWARF and DVGW-TZW 1996).
Soluble copper was always between 48 % - 95 % of the total copper (data not shown). At a given
pH, total and soluble copper release was lower at the condition with 1000 ppb-P phosphate (Figure
B.4), also consistent with expectations from earlier work (AWWARF and DVGW-TZW 1996;
Edwards et al. 2002; Schock et al. 1995).
Brass Pipes
Total copper release from brass pipes in this research was not strongly affected by either
pH or phosphorus levels (Table B.1). Although average soluble copper release from brass tended
to decrease slightly at higher pH or if 1000 ppb-P was dosed, this trend was not statistically
significant at greater than 95% confidence (Supporting Information-Table S.2). In other studies,
such as that of Lytle and Schock, total copper release from brass at pH 7 was significantly higher
than that at pH 8.5 (Lytle and Schock 1996). They also found orthophosphate either decreased or
had no effect on total copper leaching depending on the brass types (Lytle and Schock 1996).
Zinc toxicity to nitrifiers has been reported (Zhang et al. 2009a). Total and soluble zinc (61
%- 98 % of total zinc) release in brass pipes was significantly increased (up to 9 times) when pH
©2010 Water Research Foundation. ALL RIGHTS RESERVED
156 | Effect of Nitrification on Corrosion in the Distribution System
was decreased from 8.6 to 6.5; however, zinc release was not affected by different phosphorus
levels (Table B.1), consistent with Lytle and Schock’s results (Lytle and Schock 1996). The trends
in zinc leaching were statistically significant at > 95 % confidence.
Lead Pipes
In lead pipes, phosphorus levels had a more dramatic effect on lead release than did pH
changes. Regardless of inhibitor dose, total and soluble lead release did not increase when the
initial pH was decreased from 8.5 to 7.5, but soluble lead release increased markedly (Supporting
Information-Table S.3) when initial pH was reduced to 6.5-7.0. The effects of pH and
orthophosphate on soluble lead were roughly consistent with solubility model predictions
(AWWARF and DVGW-TZW 1996; Lytle and Schock 1996; Schock 1989). Soluble and total
lead release were reduced by addition of 1000 ppb orthophosphate (Supporting
Information-Table S.3), but was less significantly impacted by 60 ppb orthophosphate compared
to the lowest dose of 5 ppb, consistent with earlier work and solubility models (AWWARF and
DVGW-TZW 1996; Lytle and Schock 1996; Schock 1989).
Combined Effect of pH and Phosphate Levels on Nitrifier Activity
PVC Pipes
Reported optimal pH for nitrifier growth generally falls between 7.5-8 (Wolfe and
Lieu 2001) . In PVC pipes average nitrifier MPN was 0.5 – 1 log lower at pH 6.5 than at pH 8
(Table B.2), however, this decrease was not statistically different at 95 % confidence (Supporting
Information-Table S.3). The decrease in pH did not reduce ammonia loss in PVC pipes, since
complete ammonia conversion to nitrite and nitrate occurred even at pH 6.5. Thus, nitrifiers can
remain active in PVC premise plumbing at much lower pH ranges than have been previously
reported (Wolfe and Lieu 2001).
At phosphorus levels below about 50 ppb-P, it has been reported that nitrification was
inhibited via nutrient limitation (Van Droogenbroeck and Laudelout 1967). However, in this study
using PVC pipe, nitrifier MPN (Table B.2) and nitrifier activity (indicated by ammonia loss) was
very high even at 5 and 60 ppb-P levels. This might be due to the higher levels of phosphorus used
(1000 ppp-P) in the six month inoculation period before the phosphorus level in the medium was
reduced to 5 ppb or 60 ppb. The high cell phosphate content in nitrifier biofilm (Van
Droogenbroeck and Laudelout 1967) or pipe surface deposits containing phosphate might serve as
reservoirs for bioavailable phosphorus after the level introduced to the pipe was reduced. The
effects of high cell phosphate content has also been reported to be more significant at lower
ammonia/nitrite levels (Van Droogenbroeck and Laudelout 1967).
Copper Pipes
In copper pipes, nitrifiers were only detected at pH 8 (Table B.2) with 60 or 1000 ppb
phosphate. MPN indexes were two orders of magnitude higher at 1000 ppb versus 60 ppb
phosphate (Table B.2). Both trends were consistent with expectations for lower copper toxicity to
nitrifiers in the presence of phosphate and higher pH. Average ammonia loss was less than 10 % at
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix B: Nitrification in Premise Plumbing: Role ofPhosphate, Ph, and Pipe Corrosion | 157
all pH and phosphorus levels, illustrating the role of copper pipe in preventing nitrification in
premise plumbing relative to PCV or lead.
Brass Pipes
In brass pipes, average nitrifier MPN was 1 – 2.9 log lower at pH 6.5 than at pH 8
(Table B.2) and this trend was significant at > 95% confidence (Supporting Information-Table
S.4). The more dramatic effect of lower pH in brass relative to PVC was possibly due to the
increased concentration of Cu+2. As was the case with copper, nitrifier MPN was lower at 5 and 60
ppb-P than at 1000 ppb-P. Corresponding to the lower MPN, as pH decreased, ammonia loss also
decreased at lower pH and phosphate levels (Figure B.5).
Lead Pipes
Nitrifier MPN from lead pipes either stayed the same or decreased slightly as pH
decreased, and no significant difference was observed among the three phosphorus levels
(Table B.2). Similar to PVC, complete ammonia conversion to nitrite and nitrate occurred even at
pH 6.5. Average nitrifier MPN in lead pipes was slightly higher than that in PVC pipes
(Table B.2); however, this trend was not statistically different at 95 % confidence (Supporting
Information-Table S.4).
Interplay of Phosphorus, pH and Pipe Materials on Nitrification
Appropriate levels of phosphorus and pH are necessary for nitrifier growth, but phosphorus
and pH also influence nitrifiers indirectly by controlling corrosion rates and metal release.
Dependent on the pipe material and specific circumstance, metal leaching can be observed to
completely inhibit nitrification (e.g., newer copper or brass) or possibly stimulate nitrification
(e.g., lead pipe) in at least some cases.
Attempts were made to universally correlate predicted free copper, soluble copper or total
copper to nitrifier activity observed in the brass and copper pipe during stagnation, over the range
of conditions tested in this study. The best overall relationship was observed for nitrifier activity
versus soluble copper in brass pipes (Figure B.6). Consistent with prior batch studies (Zhang and
Edwards 2005), nitrifier activity was markedly inhibited above about 80 ppb soluble copper.
Attempts were also made to develop correlations based on predicted Cu+2 using the approach in
Zhang et al., 2009a (Zhang et al. 2009a), but this approach did not provide improved predictive
ability. This is likely due to uncertainties in the type of solids that are present which actually
control Cu+2 solubility, differences between the pH at the pipe surface versus that measured for the
bulk water, and other limitations.
Overall, the indirect effect of pH and phosphate on nitrifiers as exerted through control of
corrosion and metal toxicity, was more important than the direct effects. That is, levels of
phosphate and pH tested had stronger impacts on nitrifier activity in copper and brass pipes than in
PVC and lead pipes where no metal toxicity was observed.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
158 | Effect of Nitrification on Corrosion in the Distribution System
IMPLICATIONS FOR DRINKING WATER SYSTEMS
The interplay between water chemistry and microbial growth in premise plumbing is
obviously very complex and system dependent. Consequently, many presumptions about “worst
case” locations for sampling of water quality degradation in water distribution/premise plumbing
systems should be examined carefully. For example, based on the analysis presented in this paper,
first draw lead leaching in systems using chloramine can be expected to be much worse in some
new homes plumbed completely with PVC or plastic pipe, rather than with copper. This is because
the lower pH levels resulting from enhanced nitrifier colonization of the PVC pipe systems might
adversely impact lead and copper leaching from brass used in consumer faucets, when compared
with lead leaching in identical new homes plumbed with copper pipe. Indeed, recent field data
collected by the authors and others (Kimbrough 2007) has demonstrated that first draw lead, nickel
and zinc can be higher in homes plumbed completely with plastic pipe (and leaded brass fixtures)
relative to homes plumbed with copper pipe. All these systems used chloramine, and therefore,
enhanced nitrification and larger pH drops potentially occurring from nitrification in the mainly
plastic versus copper pipe systems could be partly responsible. It is particularly ironic that copper
levels in first draw water collected from homes plumbed with PVC and brass faucets, can often
exceed those observed in similar homes plumbed with copper pipe/brass faucets (Kimbrough
2007). The pH drop occurring during overnight stagnation should be examined as a key
contributing factor to this phenomena in future research.
Likewise, in homes plumbed with PVC and other plastics, brass is still frequently used to
connect pipe fittings and in valves. The lower pH levels resulting from nitrification could
contribute to premature failure of this brass via dezincification and other corrosion reactions,
relative to what is commonly observed when the same fittings are used with copper pipe. There
have recently been very high rates of brass fitting failures occurring in plastic premise plumbing
systems in certain U.S. localities (Kimbrough 2007).
ACKNOWLEDGEMENT
This work was supported by American Water Works Association Research Foundation
(AWWARF) and the United States Environmental Protection Agency (USEPA). Matching funds
to that grant were also provided by the Copper Development Association (CDA). The opinions,
findings, conclusions or recommendations are those of the authors and do not necessarily reflect
the views of AWWARF, USEPA or the CDA. The authors would also like to thank Meredith
Raetz for help in maintaining the experiment.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix B: Nitrification in Premise Plumbing: Role ofPhosphate, Ph, and Pipe Corrosion | 159
FIGURES AND TABLES
Figure B.1 Ammonia loss % after 3.5 day stagnation periods with each pipe material
Figure B.2 Ammonia loss % after 3.5 day stagnation and total Cu release from brass pipes
©2010 Water Research Foundation. ALL RIGHTS RESERVED
160 | Effect of Nitrification on Corrosion in the Distribution System
Figure B.3 TOC levels after 3.5 day stagnation
Note: Each data point is the average of five replicate samples. Error Bars indicate 95 % confidence
interval.
Figure B.4 Copper release in copper pipes
Note: Each data point is the average of at least four weeks testing. Error Bars indicate 95 %
confidence interval.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix B: Nitrification in Premise Plumbing: Role ofPhosphate, Ph, and Pipe Corrosion | 161
Figure B.5 Average ammonia loss % versus pH at different Phosphorus levels in brass pipes
Note: Data reported was the average of at least four weeks testing. Error Bars indicate 95 %
confidence interval of these testings.
Figure B.6 Correlation of ammonia loss % vs. soluble Cu in copper and brass pipes
©2010 Water Research Foundation. ALL RIGHTS RESERVED
162 | Effect of Nitrification on Corrosion in the Distribution System
Table B.1
Metal release in brass pipes
Phosphorus5
60 1000
5
60
1000
5
60
1000
ppb
Total Cu, ppb
Soluble Cu, ppb
Total Zn, ppb
pH
6.5
74
91
74
52
63
53
12168 10966 10250
7
70
81
71
53
47
47
6591
6108
6691
7.5
83
72
83
47
33
35
4623
2561
4528
8
75
76
70
47
37
32
3420
3995
3941
8.6
92
80
69
35
34
20
2336
1860
2845 Note: each data reported was the average of at least four weeks testing
Table B.2
Logarithm of nitrifier MPN at different phosphorus and pH levels in different pipe
materials
P-ppb
PVC
Copper
Brass
Lead
5
60
1000
5
60
1000
5
60
1000
5
60
1000
pH
8
5.5
5.5
5.7
< 1.3
2.2
4.0
5.4
5.4
5.7
5.7
6.2
6.2
7.5
4.5
5.2
5.1
< 1.3
< 1.3
< 1.3
4.4
4.3
4.7
5.7
6.0
5.4
7
4.5
4.8
4.5
< 1.3
< 1.3
< 1.3
4.5
3.2
5.4
5.7
5.7
5.4
©2010 Water Research Foundation. ALL RIGHTS RESERVED
6.5
4.5
5.0
4.7
< 1.3
< 1.3
< 1.3
2.5
2.7
4.7
4.5
5.4
5.2
Appendix B: Nitrification in Premise Plumbing: Role ofPhosphate, Ph, and Pipe Corrosion | 163
SUPPORTING INFORMATION
Stagnant
Blacksburg tap water
Dechlorinated
Add 2 mg/L NH3-N
pH to 8
Adjust
Used for
inoculation
Figure S.1: Nitrifier Inoculation using Blacksburg Tap Water
Description: The nitrifier growth in Blacksburg, VA tap water was stimulated and
maintained in a plastic container by adding dechlorinated tap water and ammonia and adjusting pH
to 8 twice a week; a portion of this water was used to inoculate pipes while the rest was kept
stagnant to mix with the next batch of dechlorinated tap water.
1
1
1
1
Tubes used for
dilution
Sample
1 ml
10-1
10-2 1 ml 1 ml 10‐3 1 ml 10-4
1 ml
All tubes contain 9 ml dilution water MPN testing tubes
10-1
10-2
10-3
10-5
10-4
Figure S.2: Serial dilution used for Nitrifier MPN procedure (Adapted from (Vikesland et
al. 2006) ).
©2010 Water Research Foundation. ALL RIGHTS RESERVED
164 | Effect of Nitrification on Corrosion in the Distribution System
Figure S.3: Bulk Water Heterotrophic Plate Colonies from the Four Pipe Materials
Table S-1: Typical Water Chemistry
Initial
PVC
Copper
Brass
Lead
pH
DO, mg/L-O2 HPC, cfu/ml
8.00 ± 0.10 8.30 ± 0.20
140000
7.25 ± 0.28 6.75 ± 0.60
120000
8.20 ± 0.34 8.16 ± 0.19
130000
7.77 ± 0.23 5.94 ± 1.12
120000
7.51 ± 0.19 6.58 ± 0.57
110000
Table S.2
95 % Confidence Interval for Metal Release in Brass Pipes
Phosphorus5
60
1000
5
60
1000
5
60
1000
ppb
Total Cu, ppb
Soluble Cu, ppb
Total Zn, ppb
pH
6.5
25
22
48
24
12
35
2876
2003
2267
7
11
51
15
12
12
11
1988
1928
1552
7.5
23
29
21
1010
897
625
8
9
23
23
4
9
3
760
986
775
8.6
29
15
17
16
14
17
667
554
526 ©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix B: Nitrification in Premise Plumbing: Role ofPhosphate, Ph, and Pipe Corrosion | 165
Table S.3
Total and Soluble Lead Release in Lead Pipes
Phosphorusppb
8.6
5
60
1000
2516 ± 1166
1751 ± 1731
1528 ± 777
5
60
1000
93 ± 10
83 ± 3
72 ± 16
pH
8
7.5
7
Average total Pb release, ppb
6.5
1831 ± 2003 1193 ± 1371 913 ± 353 1155 ± 973
1545 ± 1279 2637 ± 5675 2255 ± 3228 928 ± 606
643 ± 149
336 ± 119
292 ± 126
326 ± 261
Average soluble Pb release, ppb
88
138 ± 35
194 ± 50
95 ± 9
73 ± 10
80
98 ± 20
152 ± 48
66 ± 4
66
78 ± 11
110 ± 25 Note: each data reported was the average of at least four weeks testing, ± indicate 95 % confidence
interval
Table S.4
Logarithm of 95 % Confidence Limit of Nitrifier MPN at Different Phosphorus and pH
Levels in Different Pipe Materials
pH
8
7.5
7
6.5
P-ppb 95 % confidence 95 % confidence 95 % confidence 95 % confidence
Limit
Limit
Limit
Limit
Lower Upper Lower Upper Lower Upper Lower Upper
5
5.0
6.1
4.0
5.0
4.0
5.1
4.0
5.1
PVC
60
5.0
6.1
4.8
5.7
4.5
5.3
4.6
5.4
1000
5.3
6.3
4.7
5.6
4.0
5.1
4.3
5.2
< 1.3
< 1.3
< 1.3
< 1.3
5
< 1.3
< 1.3
< 1.3
Copper
60
1.8
2.6
< 1.3
< 1.3
< 1.3
1000
3.6
4.5
5
5.0
6.0
4.0
4.9
2.0
2.9
2.0
3.1
Brass
60
5.1
5.8
4.0
4.8
3.6
4.5
2.0
3.0
1000
5.3
6.3
4.3
5.2
5.0
6.0
4.3
5.2
5
5.3
6.3
5.3
6.3
5.3
6.3
4.0
5.1
Lead
60
5.8
6.7
5.5
6.5
5.3
6.3
5.0
6.0
1000
5.8
6.7
5.0
6.0
5.0
6.0
4.8
5.7 ©2010 Water Research Foundation. ALL RIGHTS RESERVED
166 | Effect of Nitrification on Corrosion in the Distribution System
©2010 Water Research Foundation. ALL RIGHTS RESERVED
APPENDIX C
CULTURE INDEPENDENT QUANTIFICATION OF AMMONIA
OXIDIZING BACTERIAL IN DRINKING WATER SYSTEMS
Ameet Pinto and Nancy Love
METHODOLOGY
Sample collection
300 ml of samples were collected from the four different pipe materials used in Chapter 4
study polyvinyl chloride (PVC), copper, lead, and unleaded brass. Additionally, 1 liter water
samples were also harvested from the 8 different pipe materials used in Chapter 5 and Chapter 6
study, including PVC, cast iron coupons, copper, epoxy coated copper (CuE), new lead pipes, old
lead pipes, galvanized iron (GI) and stainless steel (SS), as well as water from the reservoir tank.
Samples were filtered through 0.22 µm polycarbonate filters (Whatman) and stored at -80°C until
DNA extraction.
DNA Extraction
In order to quantify DNA extraction efficiency, all samples were co-extracted with
reference filters with immobilized pure culture Pseudomonas aeruginosa PAO1 (PA). A batch
culture of PA was grown at 37°C in minimal media (MM) containing (in grams per liter) KH2PO4
(3.4), K2HPO4 (4.35), NH4Cl (1.0), and CH3COONa (1.65) and (in milligrams per liter) EDTA
(186), MgSO4 _ 7H2O (150), MnSO4 _ 4H2O (4.5), NaMoO4 _ 2H2O (0.5), H3BO3 (0.15), CaCl2
(20), ZnCl2 (1.5), CuCl2 _ 2H2O (0.5), CoCl2 _ 6H2O (1.5), and FeCl2 _ 4H2O (11). The pH of the
media was adjusted to 7.0 (Muller et al. 2007). The culture was grown in 250 ml Erlenmeyer flasks
in a volume of 100 ml. The culture was harvested in the stationary phase and direct microscopic
cell counts were conducted. A total of 2.15x107± 2.2x106 PA cells were immobilized on several
0.2 µm polycarbonate filters (IsoporeTM membrane filters, Millipore. Billerica, MA) and stored in
2 ml tubes at -80°C until DNA extraction.
Three different DNA extraction protocols were tested for this study. These protocols
included cell lysis by repetitive-freeze thaw cycles, bead beating shear, and enzymatic treatment.
The final protocol used for this study combined bead beating shear and enzymatic treatment to
provide optimum DNA extraction efficiency. The DNA extraction protocol used in this study is a
modified version of the one used by Regan et al. (Regan et al. 2003). Sample filters were cut into
two halves and processed independently. One half of each sample filter was combined with a
reference filter of immobilized PA cells and processed as follows: 1) 300 µl of Tris-EDTA (TE)
buffer (10 mM Tris, 1 mM EDTA, pH 8.0) and pre-sterilized glass beads were added to the
extraction tube, 2) non-ionic surfactant Triton-X 100 was added to a final concentration of 1%.
The contents of the sample tubes were then homogenized by bead beating at 46,000 rpm for 60
seconds. The homogenized samples were incubated at 37°C with lyzosyme (final concentration:
5.6 mg/ml) for 30 minutes, followed by another 30 minute incubation period at 70°C with
proteinase K (final concentration: 1.1 mg/l). The extraction tubes were then centrifuged at 13,000
rpm for 10 minutes. The cell lysate was transferred to a fresh sterile microcentrifuge tube and
167
©2010 Water Research Foundation. ALL RIGHTS RESERVED
168 | Effect of Nitrification on Corrosion in the Distribution System
centrifuged again to remove any associated cell or filter debris. The extracted DNA was then
isolated from the cell lysate using Wizard® DNA cleanup system according to manufacturer’s
instructions (Promega corporation, Madison, WI. Catalog number: A7280). This DNA clean up kit
relies on the resin binding of DNA, capture of the resin on a filter, and elution of the DNA by warm
TE buffer. DNA extracted from the two filter halves were then pooled together, concentrated again
using the Wizard® DNA cleanup system and finally separated into five aliquots and stored at
-80°C until further analyses.
DNA Extraction Efficiency
DNA extraction efficiency for each sample was quantified by comparing PA cell count in
the extracted genomic DNA to the total PA cells immobilized onto the reference filters. A recently
developed primer set targeting the regA gene, which regulates the synthesis of Toxin A in P.
aeruginosa, was used for PA quantification (Lee et al. 2006). Primer sequence is shown in
Table C.1. Primers were ordered from Integrated DNA Technologies (Coraville, IA). Serial
dilutions of genomic DNA extracts from pure culture P. aeruginosa varying from 106-101 cells
were used to generate standard curves for PA quantification. The real-time PCR mix consisted of
12.5 µl of Power SYBR® green PCR master mix (Applied Biosystems Inc. Foster City, CA.
Catalog number: 4367659), 5 µl of diluted sample or pure culture DNA template, 1 µM of each
primer, and sterile nanopure water to a total volume of 25 µl. All samples were analyzed in
triplicate. Real-time PCR reactions were executed in an Applied Biosystems 7300 Real-Time PCR
System with a temperature profile of 50°C for 2 minutes, 95°C for 10 minutes and 40 cycles of
95°C for 15 seconds and 60°C for 60 seconds. Dissociation curves were generated at the end of
each real-time PCR run to ensure purity of the generated PCR amplicons. Each real-time PCR run
included a negative control of genomic DNA extracted for pure culture Nitrosomonas europaea.
The primer set showed 93-94% PCR efficiency, as seen in Figure C.1 and Figure C.2.
Ammonia Oxidizing Bacteria Quantification
Genomic DNA was extracted from pure culture N. europaea, an ammonia oxidizing
bacteria (AOB). Reference amplicons for AOB specific real-time PCR were generated by
targeting a 465 bp region of the 16S rRNA gene of N. europaea, using primer set CTO189f and
CTO654r (Table C.2) (Kowalchuk et al. 1997). The forward primer CTO189f is a 2:1 (molar ratio)
combination of primers CTOP189a/b and CTO 189c (Kowalchuk et al. 1997). The PCR mix
consisted of 25 µl Taq PCR mastermix (Qiagen Inc. Valencia, CA. Catalog number: 201443), 0.2
µM of each primer, 1 µl of genomic DNA, and sterile nanopure water to a final volume of 50 µl.
PCR reactions were conducted in a Hybaid PCR Spring Thermal Cycle (Thermo Fisher Scientific
Inc, Waltham, MA). The temperature profiles for these PCR reactions were as follows: 93°C for
60 seconds, followed by 35 cycles of 92°C for 30 seconds, 57°C for 60 seconds and 68°C for 45
seconds (+1 second per cycle), and finally 68°C for 5 minutes (Kowalchuk et al. 1997). The
presence of the 465 bp amplicons was confirmed by running the entire final PCR product through
1.2% agarose gel. Bands corresponding to the 465 bp were cut from the gel and purified using
Qiaquick Gel extraction kit (Qiagen Inc. Valencia, CA. Catalog number: 28704).
The purified amplicons were processed again through a 1.5 % agarose gel and the target
amplicons were extracted and purified again to ensure product purity. The target amplicon
quantity in the purified mix was quantified spectrophotometrically. Serial dilutions of this purified
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix C: Culture Independent Quantification of Ammonia Oxidizing Bacterial in Drinking Water Systems | 169
amplicon varying from 1012-102 were used as standard for real-time PCR. Primer set CTO189f and
RT1r was used for quantification of sample AOB (Table C.2) (Hermansson and Lindgren 2001).
This primer set showed PCR efficiency between 104-107% (Figure C.3). The real-time PCR mix
consisted of 12.5 µl of Power SYBR® green PCR master mix (Applied Biosystems Inc. Foster
City, CA. Catalog number: 4367659), 5 µl of diluted sample DNA or standard template, 0.3 µM of
each primer, and sterile nanopure water to a total volume of 25 µl. Real-time PCR temperature
profiles were identical to the ones used for PA quantification. All samples were analyzed in
triplicate. Dissociation curves were generated at the end of each real-time PCR run to ensure purity
of the generated PCR amplicons. Each real-time PCR run included a negative control of genomic
DNA extracted from pure culture P. aeruginosa. The sample AOB quantities were determined by
adjusting the real-time PCR results for sample specific DNA extraction efficiency.
Evaluating the Effectiveness of Fluorescent Staining for Sample Bacteria Quantification
The efficiency of fluorescent techniques such as fluorescent in-situ hybridization (FISH)
was also tested. 200 ml samples were collected from five independent PVC plumbing set-ups. The
samples were concentrated by repetitive centrifugation at 5000 rpm to a final volume of 0.5 ml.
The 0.5 ml of concentrated sample was fixed in 3% paraformaldehyde at 4°C for 4 hours as
detailed previously (Amann et al. 2001), followed by two centrifugations and washes with
phosphate-buffered saline (PBS, 130 mM NaCl, 10 mM phosphate buffer, pH 7.2). Samples were
then stored in a mix of 50% ethanol and 50% PBS. Fixed samples were then immobilized onto 12
well slides, stained with 4',6-diamidino-2-phenylindole (DAPI) and dried. DAPI is a DNA staining
dye and dried slides were mounted in one part Vectashield® H-1000 (Vector Laboratories,
Burlingame, CA. Catalog number: H-1000) mixed with two parts Citifluor AF-1 (Electron
Microscopy Sciences. Hatfield, PA. Catalog number: AF-1), and covered with full length cover
slips using nail polish and the edges were sealed with Parafilm® M. Previously fixed N. europaea
cells were used a positive control for this procedure. Phase contrast and fluorescence images were
taken by an Axioskop 2 Epiflourescence Microscope (Carl Zeiss Inc. Thornwood, NY). In order to
evaluate the effectiveness of fluorescent microscopy for sample bacterial quantification, the % of
phase contrast image stained by the DAPI was determined by image analyses. Images were
analyzed using ImageJ (National Institute of Health). For each image the backgrounds were
subtracted using rolling ball radius of 50 pixels, followed by conversion to 8-bit grayscale and
threshold based on phase contrast images. Relative areas stained by DAPI were determined for 10
independent images obtained from 2 different slide wells for each sample, including the positive
control.
RESULTS AND DISCUSSION
DNA Extraction Efficiency
DNA extraction efficiencies were calculated by dividing total PA cell count obtained by
real-time PCR by the total cells immobilized on co-extracted reference filters. The extraction
efficiencies for all samples were 35±12%. Figure C.4 and C.5 depict the extraction efficiencies for
samples from both bench scale and large scale rigs. The efficiencies for the large scale rigs do not
present standard errors, as only one sample was collected and analyzed for each set-up.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
170 | Effect of Nitrification on Corrosion in the Distribution System
AOB Quantification by Real-Time PCR
The total AOB quantities for all samples were significantly low (<108 cells/ml). Table C.3
lists the AOB quantities for the samples from Chapter 4bench scale study at different phosphate
levels. Brass pipes consistently exhibited higher AOB concentrations as compared to other
materials, except for lead pipes at a 5 ppb phosphate concentration. Interestingly, the AOB
concentrations in the copper pipes seem to increase at higher phosphate concentrations.
Table C.4 lists the AOB concentrations in the large scale rigs. The cell concentrations were
extremely low, ranging between 103 to 108 cells/ml. In fact, the AOB concentration in the PE rig
was below the detection limit of 100 cells/ml.
Effectiveness of Fluorescent Staining for Sample Bacteria Quantification
DAPI staining results confirmed that low cell density of the samples. Figure C.6 indicate
the sparseness of DAPI signal confirming the presence of DNA as compared to the particulate
material in the samples. Figure C.7 shows a similar comparison for pure culture N. europaea cells
with a cell density of 108 cells/ml.
Quantitative analyses of the % DAPI staining indicating the presence of DNA showed that
for the five samples analyzed the average stained area was 39 ± 16%. Table C.5 lists the % area
stained by DAPI for all five processed samples.
This low staining efficiency indicates that the water samples processed for this study are
not amenable to analyses by molecular techniques such as FISH. It is important to note, that only
200 ml of samples were harvested for this analyses. With increase in sample volume, a greater
number of cells can be potentially harvested thus increasing the effectiveness of this technique.
DATA INTERPRETATION
All the AOB cell concentrations detected in this study were below 108 cells/ml, with some
samples exhibiting as low as 105 cells/ml. Though the DNA extraction efficiencies were greater
than 20% for all processed samples, the AOB quantification relies on the assumption that cell lysis
efficiency does not vary between reference cells, in this case P. aeruginosa, and target cells, i.e.
ammonia oxidizing bacteria. Zoentendal et al. (Zoetendal et al. 2001) have shown that different
DNA extraction protocols show vastly variable detections limits for different bacterial species.
Low cell numbers in the sample matrix of interest can further exacerbate this obstacle. To address
this shortcoming, large number of sample replicates need to be collected and different DNA
extraction protocols need to be adopted, to confidently quantify the nitrifying bacteria.
However, the approach adopted in this study can be used to address presence/absence of
AOB. Moreover, since the same protocol has been adopted for all the samples, with relatively
similar DNA extraction efficiency; a comparative evaluation can be made of AOB quantities
between different plumbing systems. Previously published MPN analysis data (Zhang et al.
2008a), indicates that AOB were more prevalent in PVC and lead plumbing system as compared to
the brass and copper systems. This difference was attributed to leaching of copper into the water
matrix. However, culture independent evaluation of water samples from the same systems
confirms the presence of AOB in the brass and copper systems. In fact, the AOB quantities in the
copper and brass pipes are significantly higher than the PVC system for the same phosphate
content. These results are contradictory to MPN data presented in Chapter 4. However, a larger
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix C: Culture Independent Quantification of Ammonia Oxidizing Bacterial in Drinking Water Systems | 171
amount of samples need to analyzed before any conclusion can be drawn. And it was quite possible
that the sample used was contaminated due to the lack to sterile sample handling (Chapter 4).
CONCLUSION
This study has provided a reliable protocol for bacterial DNA extraction from drinking
water samples, which typically harbor low bacterial cell numbers. This protocol uses a
combination of physical shear and enzymatic lysis to extract DNA from filter immobilized cells.
DNA purification for cell lysate relies on resin binding of DNA. This approach showed better
reliability as compared to traditional co-precipitation with glycogen as presented previously
(Regan et al. 2003b). DNA extraction efficiencies were calculated for all samples, by including the
co-extraction of a reference filter with pure bacterial culture. Quantification calculations for target
species were adjusted for the extraction efficiency variability between the samples.
The real-time PCR approach confirmed the presence of ammonia oxidizers in all plumbing
systems. The culture independent quantitative analyses showed different trends in nitrifier cell
abundance as compared to previous MPN data. But caution should be exercised while interpreting
the data presented here. Accurate estimation of trends requires a large amount of sample processing,
with replicate experiments. This is especially true while interpreting microbial abundance trends for
samples with extremely low cell numbers, such as the ones presented in this study.
FIGURES AND TABLES
34
2
R = 0.999
Slope= -3.50
PCR efficiency = 93%
Threshold cycle (Ct)
32
30
28
26
24
22
20
18
10 2
10 3
10 4
10 5
10 6
Pseudomonas aeruginosa cell number
10 7
Figure C.1 Standard curve for primer set PAER-f/PAER-r used for quantification of P.
aeruginosa: example 1
©2010 Water Research Foundation. ALL RIGHTS RESERVED
172 | Effect of Nitrification on Corrosion in the Distribution System
36
R2 = 0.998
Slope= -3.47
PCR efficiency = 94%
34
Threshold cycle (Ct)
32
30
28
26
24
22
20
18
16
100
101
102
103
104
105
106
107
Pseudomonas aeruginosa cell number
Figure C.2 Standard curve for primer set PAER-f/PAER-r used for quantification of P.
aeruginosa: example 2
40
R2 = 0.996
Slope= -3.16
PCR efficiency = 107%
Threshold cycle (Ct)
35
30
25
20
15
10
5
0
102
103
104
105
106
107
108
109
1010
1011
1012
Nitrosomonas europaea 16S rDNA copy number
1013
1014
Figure C.3 Standard curve for primer set CTO189f/RT1r used for quantification of AOB
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix C: Culture Independent Quantification of Ammonia Oxidizing Bacterial in Drinking Water Systems | 173
DNA extraction efficiency (%)
60
50
Brass Copper Lead PVC 40
30
20
10
0 Figure C.4 DNA extraction efficiencies for laboratory scale plumbing systems
100
80
60
GI rig
Old lead rig
Cast Iron rig
Old Lead Rig PVC rig SS rig
Tank rig
Cu rig
DNA extraction efficiency (%) 40
20
0
Figure C.5 DNA extraction efficiencies for large scale rig samples
©2010 Water Research Foundation. ALL RIGHTS RESERVED
174 | Effect of Nitrification on Corrosion in the Distribution System
Figure C.6 DAPI staining results for a 400x concentrated sample
Note: Blue stain indicates the presence of DNA. Note the large amount of unstained area as
compared to presence of particulate matter.
Figure C.7 DAPI staining results for a N. europaea cells at 108 cells/ml
Note: Blue stain indicates the presence of DNA. Note the large amount of stained area.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix C: Culture Independent Quantification of Ammonia Oxidizing Bacterial in Drinking Water Systems | 175
Table C.1
Primers used for P. aeruginosa quantification
Primer name
Sequence
PAER-f
TGC TGG TGG CAC AGG ACA T
PAER-r
TTG TTG GTG CAG TTC CTC ATT G
Target gene
Reference
regA
(Lee et al. 2006)
Table C.2
Primers used for N. Europaea amplicon generation and total AOB quantification
Primer name
Sequence
Target gene
CTO 189a/b
GGA GRA AAG CAG GGG ATC G
16S rRNA
CTO 189c
GGA GGA AAG TAG GGG ATC G
16S rRNA
CTO 654 r
CTA GCY TTG TAG TTT CAA ACG C
16S rRNA
RT1r
CGT CCT CTC AGA CCA RCT ACT G
16S rRNA
Reference
(Kowalchuk et al.
1997)
(Hermansson and
Lindgren 2001)
Table C.3
Ammonia oxidizing bacteria concentrations for laboratory scale set-ups
Phosphate content
(ppb)
AOB concentration
(cells/ml)
5
4.27 x 107 ± 1.45 x 106
60
3.07 x 107 ± 1.05 x 106
1000
6.41 x 107 ± 2.19 x 106
5
2.91 x 106 ± 9.92 x 104
60
3.15 x 107 ± 1.07 x 106
1000
1.08 x 107 ± 3.68 x 105
5
4.03 x 106 ± 1.38 x 105
60
1.21 x 107 ± 4.14 x 105
1000
N/A*
5
1.40 x 106 ± 4.78 x 104 60
2.93 x 106 ± 8.15 x 104 1000
N/A*
Material
Brass
Copper
PVC
Lead ©2010 Water Research Foundation. ALL RIGHTS RESERVED
176 | Effect of Nitrification on Corrosion in the Distribution System
Table C.4
AOB concentration for large scale rigs
Rig type
AOB concentration
GI rig
6.25 x 106 ± 6.40 x 104
Old lead rig
BDL*
Cast iron rig
1.09 x 105 ± 1.11 x 103
New lead rig
1.38 x 105 ± 1.41 x 104
PVC rig
1.06 x 105 ± 1.08 x 103
SS rig
7.46 x 105 ± 7.763 x 103
Tank rig
7.62 x 106 ± 7.80 x 104
Cu rig
7.75 x 107 ± 5.88 x 105
*BDL: below detection limit of 102 cells/ml
Table C.5
DAPI staining for water samples for five different PVC plumbing systems
Sample
% DAPI stain
A
55 ± 23
B
48 ± 19
C
24 ± 50
D
19 ± 12
E
40 ± 9
Positive control: N. europaea
108 ± 15
©2010 Water Research Foundation. ALL RIGHTS RESERVED
APPENDIX D
LEAD CONTAMINATION OF POTABLE WATER DUE TO
NITRIFICATION
Yan Zhang, Allian Griffin, Mohammad Rahman, Ann Camper, Helene Baribeau and Marc
Edwards
Reprinted with permission from Environmental Science & Technology 43(6): 1890-1895.
Copyright 2009 American Chemical Soecity
ABSTRACT
Nitrification can increase levels of soluble lead in potable water by reducing pH. The
magnitude of the pH effect depends on the initial alkalinity, extent of nitrification and associated
acid production. At 100 mg/L alkalinity as CaCO3, complete nitrification did not significantly
decrease pH (pH stayed > 7.5) or increase lead contamination from lead pipe, but at 15 mg/L
alkalinity, nitrification decreased the pH by 1.5 units (pH reduced to < 6.5) and increased soluble
lead contamination by 65 times. Lower pH values from nitrification also leached 45% more lead
and 81% more zinc from leaded brass connected to PVC pipes versus copper pipes. Particulate
lead leaching was high, but did not seem to vary dependent on nitrification. Production of nitrite
and nitrate, or reductions in inorganic carbon or dissolved oxygen via nitrification, did not
significantly influence lead leaching.
INTRODUCTION
As more utilities in the United States switch to chloramines for residual disinfection of
potable water (Seidel et al. 2005; Zhang et al. 2009b), there is increasing concern about potential
costs and health implications of corrosion induced by nitrification (Zhang et al. 2009a).
Nitrification, the conversion of ammonia to nitrite (NO2-) and then nitrate (NO3-) by nitrifying
bacteria, can impact corrosion by decreasing pH, alkalinity and dissolved oxygen (Zhang et al.
2009a). Production of organic carbon, and accelerated disinfectant decay, might also stimulate
growth of corrosion-influencing microbes which could also affect corrosion (Wilczak et al. 1996;
Zhang et al. 2009a). A 1996 survey indicated that two thirds of the medium and large utilities that
use chloramines report nitrification problems in water mains (Wilczak et al. 1996), and it is very
likely that even a greater percentage have nitrification issues if premise plumbing is considered
(Zhang et al. 2008a; Zhang et al. 2009a).
There is some anecdotal evidence of corrosion problems triggered by chloramines and
nitrification in at least some circumstances. For example, recent work in Pinellas County, Florida
highlighted some concerns related to iron corrosion control and red water (Powell 2004).
Likewise, elevated copper levels at the tap were suspected to be linked to action of nitrifying
bacteria in Willmar, Minnesota homes (Murphy et al. 1997a). Nitrification also co-occurred with
higher lead leaching in Ottawa (Douglas et al. 2004), Washington D.C., Durham and Greenville,
NC homes (Edwards and Dudi 2004; Edwards and Triantafyllidou 2007; Triantafyllidou et al.
2007). However, in most studies, the link between nitrification and corrosion is not confirmed, and
the mechanism by which nitrification affects corrosion was not postulated. Only in the Ottawa
177
©2010 Water Research Foundation. ALL RIGHTS RESERVED
178 | Effect of Nitrification on Corrosion in the Distribution System
study (Douglas et al. 2004), the higher lead was suspected to be the result of nitrification causing a
pH decrease, but whether this pH decrease occurred upstream or in the lead pipe was not clear.
Given the high costs and health implications of corrosion to utilities and consumers
(Edwards and Dudi 2004; Triantafyllidou et al. 2007), and further considering that prior research
emphasized nitrification problems occurring in the main distribution system whereas virtually all
lead and copper plumbing materials are located after the property line (Committee on Public Water
Supply Distribution Systems: Assessing and Reducing Risks 2006), it is important to better
understand effects of nitrification on corrosion and metal release under conditions found in
premise plumbing. Moreover, the work attempts to scientifically verify anecdotal links established
between nitrification and increased lead leaching through a well-controlled laboratory study.
EXPERIMENTAL SECTION
Water Chemistry
Lead pipes of 1.9 cm diameter were fabricated and then cut into 30 cm lengths. The lead
pipes were aged by exposure to a synthesized water for one year without nitrification, and then
exposed to water with ammonia (and resulting nitrification) for 15 months as described elsewhere
(Zhang et al. 2008a). No disinfectant had ever been added to the pipes. Thirty pipes were exposed
at 5, 60 and 1000 ppb orthophosphate-P (10 at each phosphate level). The ten replicate pipes were
separated into three groups. The first group continued as a control (4 pipes), and the second group
was modified by addition of free chlorine to a final concentration of 10 mg/L total chlorine (3
pipes). The added chlorine reacted with the existing ammonia to form monochloramine. The third
group of pipes was modified by addition of 1 mg/L chlorite (3 pipes). The high chloramine
(Fleming et al. 2005) and chlorite (McGuire et al. 1999; McGuire et al. 2006) levels were used to
attempt to inhibit nitrification that was allowed to proceed unimpeded in the control. The pH of
each type of water was adjusted to 8 before filling up the pipe. The alkalinity of the water was
dropped stepwise from 100 mg/L down to 30, 15 and then 0 mg/L alkalinity by decreasing the
amount of NaHCO3 added. Each alkalinity level was maintained for sufficient time for
nitrification, pH and lead leaching to stabilize. Water in the pipes was changed every 3.5 days
(twice a week) using a “dump and fill” protocol to simulate infrequent water use and replenish
nutrients for microbial nitrifier growth. The pipes were maintained at room temperature.
Analytical Methods
Nitrifier activity was tracked by measuring loss of ammonia, production of nitrite and
nitrate and reduction of pH. pH was monitored by using pH electrode according to Standard
Method 4500-H+ B (Clesceri et al. 1998). Total ammonia (= NH3 + NH4+) was measured using a
salicylate method with a HACH DR/2400 spectrophotometer, according to Standard Method
4500-NH3 (Clesceri et al. 1998). NO2--N and NO3--N were measured using DIONEX, DX-120 ion
chromatography, according to Standard Method 4110 (Clesceri et al. 1998). Soluble and total
metal release was also quantified. Soluble metal was operationally defined as the portion of metal
passing through a 0.45 µm pore size syringe filter. Total metal release was quantified by digesting
samples with 2% nitric acid for 24 hours in a 90 °C oven. Metal concentrations were quantified
using an Inductively Coupled Plasma Mass Spectrophotometer (ICP-MS) according to Standard
Method 3125-B (Clesceri et al. 1998).
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix D: Lead Contamination of Potable Water due to Nitrification | 179
Case Studies
Two types of case studies were conducted at utilities and at bench scale at the University of
Montana.
Water Utility Studies
Six participating utilities coordinated the collection of samples at three consumer homes
before and after stagnation. Samples were analyzed for ammonia, nitrite, nitrate, lead and copper
release and other basic water quality parameters (pH, chlorine, alkalinity, temperature, HPC, etc).
More details of the sampling procedures are included in the Supporting Information.
Montana Bench Test
A plumbing rig was constructed to directly test the effect of pipe material (PVC versus
Copper) on nitrification and resulting lead contamination of water by leaded brass. A brass rod
(0.64 cm diameter × 10 cm length, C35300 alloy with 2% lead) was machined and placed inside a
PVC or copper pipe (1.3 cm diameter × 61 cm length) to simulate the situation in homes with
PVC/copper plumbing and leaded brass faucets. The rod was not in electrical contact with the PVC
or copper pipe. Each experiment was run in triplicate using a synthesized potable water containing
nitrifying bacteria. The synthesized water contained (NH4)2SO4 (2.13 mg/L-N), initial pH of 8.15,
Na2HPO4 (1 mg/L-P), NaHCO3 (35 mg/L as CaCO3), Elliot Humics (4 mg/L) (International
Humic Substances Society) and other salts described elsewhere (Rahman 2008). Water in the
pipes was changed every Monday, Wednesday and Friday and samples were analyzed for
ammonia, pH, lead and zinc release as described above.
RESULTS AND DISCUSSIONS
Inhibition of Nitrification with Monochloramine/Chlorite vs. Control Condition
Addition of either choramine or chlorite effectively inhibited nitrification in the lead pipes.
Ammonia conversion to nitrate and nitrite (i.e., ammonia loss) decreased from > 80% to < 40%
upon addition of choramine or chlorite (Figure D.1).
In most water systems, the bicarbonates and carbonate mainly represent the alkalinity of
water. Based on nitrification stoichiometry, 14 mg/L as CaCO3 alkalinity is consumed for every
mg NH4+-N oxidized (Grady et al. 1999), so it was expected that nitrification could be somewhat
limited by a lack of alkalinity at the 6% of U.S. utilities with water containing less than 15 mg/L
alkalinity as CaCO3 (AWWA. 1996). But in this study, for the control pipes, virtually complete
nitrification occurred even when the added carbonate alkalinity was 15 mg/L (Figure D.1). This is
possible because most bicarbonate alkalinity is consumed for maintaining pH, and only a small
amount is incorporated into cell biomass (Grady et al. 1999). Since lead corrosion can increase pH
and alkalinity, the actual limiting alkalinity would be lower than that predicted from stoichiometry.
To confirm that at least some inorganic carbon is required, testing was conducted with no added
alkalinity, in which case nitrification was abruptly halted even in the control (Figure D.1). Thus, it
is unlikely that nitrification will be inorganic carbon limited at water utilities, unless alkalinity is
well below 15 mg/L as CaCO3.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
180 | Effect of Nitrification on Corrosion in the Distribution System
All other factors being equal, a lesser amount of nitrification translates to a higher final pH in
the pipe, because less acid is produced via nitrification (Zhang et al. 2009a). However, due to the
relatively high buffering of the water at 100 mg/L alkalinity, pH values in the pipes with
monochloramine/chlorite were only slightly higher (p ≥ 0.04) relative to the control condition, even
though there was less nitrification (Figure D.1). But when alkalinity (and buffering) was decreased
to 30 mg/L and then 15 mg/L, the final pH values with monochloramine/chlorite were much higher
(p < 0.001) than in the control (Figure D.1). Thus, the extent of pH reduction by nitrification depends
not only on the extent of nitrification, but also the initial alkalinity level. At 0 alkalinity, where
nitrification stopped in the controls, pH values in the pipes with monochloramine/chlorite were
slightly lower relative to that without inhibitors (p ≥ 0.005) (Figure D.1).
The above results were examined relative to predicted trends using solubility and chemical
reaction models. Predictions of final pH based only on bulk water chemistry and the expected acid
production by nitrification, confirms that the extent of the pH drop should increase with decreasing
alkalinity (Supporting Information-Figure S.1). In fact, the predicted pH drops correlate very well
with actual monitored pH drops (R2 = 0.81-1), although the actual pH drop is typically only
60-90% of the predicted pH drop. It is not uncommon to see differences between the predicted and
actual pH drops due to other reactions (i.e., scale dissolution and corrosion) that tend to increase
the pH (Zhang et al. 2008b).
Effect of Nitrification on Lead Release
Reductions in pH due to nitrification have been hypothesized to increase lead release
(Garret 1891; Odell et al. 1996; Zhang et al. 2009a), although there has been no research that
directly confirmed this hypothesis (Zhang et al. 2008b). In this study, a head to head comparison of
pipes with nitrification inhibitors to those without confirmed that nitrification increased lead
release, but the extent of the effect is highly dependent on the initial alkalinity level (Figure D.2).
Specifically, at 100 mg/L alkalinity, lead release was not increased by nitrification, as indicated by
similar or even higher total and soluble lead release in the pipes with monochloramine/chlorite
versus the condition without nitrification inhibitors (Table S.1) (Figure D.2). At 30 mg/L
alkalinity, nitrification increased total lead release up to 5 times and soluble lead release up to 21
times (p ≤ 0.002-Table S.1) (Figure D.2). At 15 mg/L alkalinity, total lead release was increased up
to 5.5 times and soluble lead release up to 65 times (p ≤ 0.00008-Table S.1) (Figure D.2). These
trends are in agreement with the fact that the pH was reduced most significantly by nitrification at
the lower alkalinity levels (Figure D.1). When alkalinity dropped to 0, lead release in the control
pipes dropped to levels similar (or even slightly lower) than those observed with
monochloramine/chlorite (Figure D.2, Table S.1), proving that the nitrification inhibitors
themselves were not decreasing soluble or total lead.
It was also considered possible that soluble microbial product (SMP) produced by
nitrification might be increasing lead solubility. However, based on data considered from earlier
studies (Rittmann et al. 1994), nitrification of 2 mg/L ammonia would produce 20 ppb-C SMP
from autotrophic growth, which is not sufficient to increase soluble lead by 1000 ppb due to
complexation.
Lead in drinking water can be present as truly dissolved soluble species (e.g., Pb+2,
Pb(CO3-2)2, or as colloids/particulates that can include detached solder, brass or lead scale
(McNeill and Edwards 2004; Triantafyllidou et al. 2007). In the pipes with high nitrification (no
inhibitor), average soluble lead was gradually increased from 60, 220 to 850 ppb as alkalinity
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix D: Lead Contamination of Potable Water due to Nitrification | 181
levels were dropped from 100, 30 to 15 mg/L, respectively (Figure D.3). When alkalinity was
dropped from 15 to 0 mg/L as CaCO3, soluble lead decreased to 160 ppb, because the final pH was
much higher (Figure D.1). Particulate lead stayed constant at around 200 ppb (Figure D.3). In the
pipes with low nitrification (monochloramine/chlorite), neither soluble nor particulate lead was
significantly changed (Figure D.3). Thus, nitrification increased the soluble lead release, but had
no detrimental impact on the particulate lead release (Figure D.3).
Solubility models were used as described elsewhere to better understand the interplay
between pH drops, alkalinity, buffering, DIC and lead solubility (AWWARF and DVGW-TZW
1996; Schock 1989). Percent changes in soluble lead in response to various 0.5 unit pH drops were
calculated as a function of initial alkalinity and pH change (Figure D.4). In the absence of lead
phosphate solids, soluble lead is predicted to markedly increase when pH drops, and the
percentage increase is very significant at lower alkalinities (Figure D.4). For example, predicted
soluble lead levels increase by 120% (from 470 to 1035 ppb) when pH drops from 7 to 6.5 at 15
mg/L alkalinity. But if the water has 100 mg/L alkalinity, the pH drop from 7 to 6.5 only increases
predicted soluble lead by 40% (from 320 to 447 ppb) (Figure D.4).
The solubility model trends and observed impacts of lower pH on lead release are also
consistent with utility experience and other recent data (Zhang et al. 2008a). That is, the effects of
pH on lead contamination and lead solubility are relatively weak at the higher alkalinities. For
example, the 90 percentile lead levels reported by utilities are not a strong function of pH if
alkalinity is > 30 mg/L, but lower pH markedly increases 90%’ile lead if alkalinity is < 30 mg/L
(Dodrill and Edwards 1995).
In situations where a Pb3(PO4)2 solid controls solubility, a pH drop from 8 down to 7.5 or
even 7.0 has relatively little effect on soluble lead, especially at higher alkalinities (Figure D.4) . In
fact, the lower pH can even decrease lead solubility in some cases. In such systems, the final pH
must drop down to about pH 6.5 before solubility is markedly increased (Figure D.4), and even so,
the changes in absolute lead concentrations are low relative to the 15 ppb action level. For
example, at 15 mg/L alkalinity, soluble lead levels are predicted to be 9 and 16 ppb at pH 8 and 6.5,
respectively. The key conclusion is that pH drops from nitrification might not increase lead
solubility at utilities dosing orthophosphate corrosion inhibitors, if the dose is sufficient to form
Pb3(PO4)2 or similar solids on the pipe surfaces.
The models were also used to examine the possible role of decreased Dissolved Inorganic
Carbon (DIC) due to autotrophic growth. Consideration of this factor had only slight predicted
impacts on soluble lead (Figure D.4), supporting the idea that the primary effect of nitrification is
exerted through its impact on pH.
Direct Comparison of Model Results to Experimental Data
For the experiments conducted herein, the initial phosphate in the water was virtually all
removed during stagnation in the lead pipe. Even when the initial water contained 1000 ppb
phosphate, only 10 ppb phosphate was present in the water after stagnation (i.e., > 99% had been
removed by reactions with the lead pipe wall). Assuming even a very low corrosion rate of 0.1
uA/cm2, the Pb+2 formed per unit volume of water in the pipe exceeds the phosphate that is present
by more than an order of magnitude. Hence, the extreme loss of phosphate from the water, by
processes such as sorption to newly forming lead hydroxyl-carbonates scale is not unexpected.
This explains why the data in Figure D.2, behave more in keeping with trends expected for a
system without phosphate inhibitor, because the added phosphate was effectively removed and
©2010 Water Research Foundation. ALL RIGHTS RESERVED
182 | Effect of Nitrification on Corrosion in the Distribution System
Pb3(PO4)2 was under-saturated in these tests. In a practical situation with more frequent water
changes and flow, it is highly likely that Pb3(PO4)2 solid would have formed. Thus, all results in
our tests, are representative of expectations at utilities that do not dose a high amount of
orthophosphate to the water.
The solubility model predictions (based on measured final pH, final total lead, total
phosphate and final alkalinity) are in good agreement with the actual data on soluble lead for
situations with high nitrification (R2 = 0.76) (Figure D.5). The conclusion is that the complex
response of soluble lead to initial alkalinity (Figure D.2) is completely consistent with solubility
model predictions that consider lead-carbonate solid formation, lead carbonate complexation and
other factors (AWWARF and DVGW-TZW 1996; Dodrill and Edwards 1995; Edwards et al.
1999; Schock 1989). This good correlation also suggests that the complexation by SMP, is not the
major contributor to the increased soluble lead after nitrification.
It is important to note that higher alkalinity has a dual benefit in preventing problems with
soluble lead due to nitrification. First, as a buffer, the extent of the pH drop due to a given amount
of ammonia conversion is reduced. This is obvious based on the pH (-log of average [H+]) of 7.7,
6.92 and 6.19 at 100, 30 and 15 mg/L alkalinity, respectively (Figure D.1). Second, a given pH
drop of 0.5 units also has a much lesser impact on soluble lead at higher alkalinity (Figure D.4).
Lead Solubility with Nitrification Inhibitors or at Lower Phosphate Levels
In nearly all tests with lower levels of phosphate, or with nitrification inhibitors, relatively
little nitrification occurred. As a result, the final pH was higher, and soluble lead levels were lower
versus the control condition dosed at 1 mg/L-P. In these cases the solubility model did not predict
trends quite as well as for cases where extensive nitrification occurred (e.g., Supporting
Information Figure S.2). This is probably because the soluble lead was a relatively small fraction
(10-42%) of the total lead present, and soluble lead data therefore has higher error due to
significant interference from colloids less than 0.45 um in size (McNeill and Edwards 2004) when
the fraction of soluble lead is low relative to total lead.
Case Studies
Water Utility Studies
Samples collected from homes at six utilities using chloramines indicated extensive
nitrification during stagnation via a decrease of ammonia (Table D.1) and increase of nitrite/nitrate
species. Similar to our earlier field studies (Zhang et al. 2008b), significant pH drops (0.34-2) were
also observed (Table D.1). When the actual pH drop is compared to the pH drop predicted based on
ammonia consumption, the actual pH drop could either be higher or lower than the predicted pH
drop (Table D.1). This was not surprising considering the complexity of the reactions due to other
bacteria that can occur in pipes during stagnation (Figure D.2). However, there are also cases
where the predicted pH drop was very close to the actual pH drop; for example, site #1 in the
anonymous and Portland utility. For samples collected at a given utility, the magnitude of the pH
drop amongst different sampling sites was generally consistent with the ranking of ammonia
consumption (Table D.1). An exception was site #3 in Portland, OR, where the ammonia
consumption was high, but pH was observed to increase rather than decrease. This was possibly
due to the presence of concrete lined pipes in the system, which can leach lime to water, as is
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix D: Lead Contamination of Potable Water due to Nitrification | 183
seemingly confirmed by higher calcium levels in water samples from this site compared to the
other two (5 ppm versus 2 ppm).
For all utilities investigated, the sampling sites with the smallest ammonia consumption
and pH drop usually had the least lead, copper or zinc contamination (#3 site within each utility).
Overall, the level of lead, copper and zinc contamination was somewhat controlled by the
magnitude of the pH drop caused by nitrification. Because the plumbing materials present in each
home will vary markedly in terms of lead content and other important issues, perfect correlations
between pH drop and lead leaching are not expected.
Montana Bench Test
Consistent with expectations based on prior research (Zhang et al. 2008a), complete
nitrification occurred in PVC pipes after two months, but never occurred in copper pipes even after
9 months due to copper toxicity (total copper concentration was 1500 ± 130 ppb in copper pipes
and 75 ± 5 ppb in PVC pipes). In PVC pipes, nitrification decreased the final pH to 6.77, while in
copper pipes, final pH was increased to 9.96, possibly due to copper corrosion (Zhang et al. 2008b)
(Table D.2). Not surprisingly, the average lead and zinc released from the leaded brass in the PVC
pipes were 45% and 81% higher, respectively, than was observed in copper pipes (Table D.2). This
type of effect also may have been observed at one of the utilities sampled (Table D.1-Hampden,
MN). Higher copper, lead and zinc were observed in homes with plastic rather than copper service
plumbing. Overall, these data offer clear proof that in situations where nitrification is occurred and
leaded brass faucets are present, the first draw lead might be higher in homes plumbed with
PVC/plastic pipe versus homes plumbed with copper pipe (Zhang et al. 2008a).
IMPLICATIONS FOR DRINKING WATER SYSTEMS
Nitrification can sometimes increase soluble lead contamination of potable water by
reducing pH. The extent of the pH reduction and impact on soluble lead release depends on the
amount of nitrification, the initial alkalinity and initial pH level. In considering the possible effects
of nitrification on lead solubility for a given circumstance, the “worst case” pH drop from
nitrification (assuming 100% conversion of total ammonia) for a utility (Figure D.6) can be
calculated. For example, at a utility with initial pH of 8.0 and 46 mg/L initial alkalinity and 1 mg/L
NH3-N, pH is expected to drop up to 1 unit due to nitrification (Figure D.6). The lead solubility
models (i.e., Figure D.4) can then be used to predict that the 1 unit drop in pH from 8 to 7 could
increase soluble lead by 50%.
This analysis indicates that the initial pH also plays an important dual role. First, systems
with pH between 7.5 and 8 are more likely to have active nitrification (Wolfe and Lieu 2001).
Second, a given amount of nitrification activity would induce a much larger pH drop in systems
with initial pH of 8-8.5, since buffer intensity is at minimum at pH = 8.3 in carbonate systems
(Snoeyink and Jenkins 1980). For example, comparing site # 1 at St Paul, MN and the anonymous
utility, with similar initial alkalinities (42 and 44 mg/L) and ammonia loss (0.4 and 0.41 mg/L), the
predicted pH drop due to nitrification is 0.5 unit in St Paul, MN (initial pH of 9) and 0.61 unit in the
anonymous utility (initial pH of 8.3) (Table D.1). The actual pH drop in St. Paul, MN was even
smaller-0.12 pH unit (Table D.1).
©2010 Water Research Foundation. ALL RIGHTS RESERVED
184 | Effect of Nitrification on Corrosion in the Distribution System
The overall evaluation indicates that serious problems with lead leaching from nitrification
are not expected at the alkalinities and initial pHs encountered at many water utilities. Hence, use
of Figure D.6 and Figure D.4 can serve as an important screening tool, to consider a utility’s
susceptibility to problems with higher soluble lead as a result of nitrification. Certainly, utilities
like Ottawa, Canada (initial pH 8.5 and initial alkaliiinity 35 mg/L) (Douglas et al. 2004) are
predicted to be very susceptible to elevated soluble lead from nitrification. Similar approach can be
taken for evaluating copper leaching problems that might be induced by pH drops due to
nitrification, but contrary to lead release, copper solubility is most strongly affected by pH changes
in higher alkalinity waters, like Willmar, Minnesota (Murphy et al. 1997a).
FIGURES AND TABLES
Figure D.1 Ammonia loss % and final pH, versus time when initial alkalinity was reduced in
stages from 100 to 0 mg/L
Note: All pipes were dosed with 1 mg/L as P.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix D: Lead Contamination of Potable Water due to Nitrification | 185
Figure D.2 Total and soluble lead release versus time in lead pipes at 1 mg/L as P
Figure D.3 Average soluble and particulate lead release at different alkalinity levels and 1
mg/L as PO4-P
Note: Error bars indicate 95% confidence interval.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
186 | Effect of Nitrification on Corrosion in the Distribution System
Figure D.4 Predicted % soluble lead increase due to pH decrease (Assuming 2000 ppb lead)
Note: Lower graph assumes solubility control by Pb3(PO4)2 solids.
Figure D.5 Actual soluble lead vs. predicted soluble lead in lead pipes at the control
condition with 1 mg/L-P
Note: Pb3(PO4)2 or similar solids were not predicted to form due to high removal of
phosphate during stagnation.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix D: Lead Contamination of Potable Water due to Nitrification | 187
Figure D.6 Resulting pH drop based on ammonia oxidized at different initial alkalinity and
pH
Table D.1
Water utility studies results
#1
#2
#3
#1
Anoymous # 2
#3
#1
Portland,
#2
OR
#3
Bangor, # 1
ME
#2
#1
Hampden,
#2
ME
#3
St. Paul,
MN
Service Ammonia
decrease,
Pipe
mg/L
Material
0.4
Lead
0.8
0.18
0.41
0.34
0.38
0.23
Copper
0.18
0.32
0.6
Copper
0.62
cast iron
0.35
plastic
0.44
copper
0.08
POE Alkalinity, Predicted Actual Total
Total
Total
pH
mg/L
pH drop pH drop Pb, ppb Cu, ppb Zn, ppb
9
8.3
7.8
9.73
9.6
42
38
39
44
49
51
10
13
11
22
22
26
25
25
0.5
1
0.2
0.62
0.54
0.54
0.57
0.57
0.82
0.35
0.35
0.16
0.2
0.03
0.12
0.72
-0.18
0.6
0.3
0
0.5
0.3
-1.1
0.71
0.58
1.8
1.99
0.15
261
58
8
7
2
5
7
21
4
4.8
49
1
658
12
771
15
31
550
215
180
312
944
177
331
228
81
444
187
8
21
269
486
52
135
13
34
11
70
6
12
168
24
Note: Data reported represents the 1st draw sample after overnight stagnation. Calculations
are made by comparing water quality to those collected at the POE.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
188 | Effect of Nitrification on Corrosion in the Distribution System
Table D.2
Montana bench test results
Pipe Material
Montana
Bench Test
PVC
Copper
Ammonia
loss, %
71±3
6±5
Final pH
6.77 ± 0.06
9.77 ± 0.13
Total Pb, Total Zn,
ppb
ppb
67 ± 3 763 ± 33
46 ± 4 421 ± 30 Note: Data reported were the average of 36 measurements over four weeks time. ±indicates
95% confidence interval
SUPPORTING INFORMATION
Figure S.1: Actual pH drops versus predicted pH drops at 1 mg/L-P.
Note: Predicted pH drops were modeled using MINIQL software. Acid produced from
nitrification in each case is based on measured ammonia loss, nitrate/nitrite production and the
following equations (Equation S.1 and Equation S.2). This produced acid is added to the initial
acid (acid needed to adjust the initial pH to 8), and the result is used to calculate the final pH.
NH4+ + 1.9 O2 + 0.069 CO2 + 0.0172 HCO3- Æ 0.0172 C5H7O2N + 0.983 NO2- + 0.966
H2O + 1.97 H+
Equation S.1
+
NO2 + 0.00875 NH4 + 0.035CO2 + 0.00875 HCO3- + 0.456 O2 + 0.00875 H2O Æ 0.00875
Equation S.2
C5H7O2N + 1.0 NO3-
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix D: Lead Contamination of Potable Water due to Nitrification | 189
Figure S.2: Actual soluble lead versus predicted soluble lead at 1 mg/L-P.
Table S.1
p-value for the paired t-test on total and soluble lead levels (1 mg/L-P)
Soluble Lead
chlorine vs. no inhibitor
chlorite vs. no inhibitor
Total Lead
chlorine vs. no inhibitor
chlorite vs. no inhibitor
100
0.45
0.2
100
0.46
0.04
Alkalinity
30
15
0.0002
0.00006
0.002
0.00008
Alkalinity
30
15
0.34
0.007
0.0003
0.0002
0
0.039
0.003
0
0.11
0.006 S.1 WATER UTILITY STUDY DETAILS
Samples were collected from a selection of six participating utilities. At each utility, samples
were collected at the POE(s) to the distribution system, as well as at two to three homes throughout the
chloraminated section of the distribution system. First-draw and flushed samples were collected from
homes presenting higher risks of lead and copper contamination, and/or homes near areas where
nitrification has been experienced. Samples were collected from faucets within homes (not from
outside hose bibs), and from homes without after-market home water treatment devices. Residents
were asked to let the water stagnate in the pipe for at least 8 hours prior to sample collection, i.e., they
could not use the sampled faucet. Five samples are collected at each location: 1st draw bacterial test
sample, 1st Draw Unfiltered, 1st Draw Filtered, Flush Unfiltered, Flush Filtered. The first sample
collected was for the bacteriological tests (50ml for HPC and MPN test). First-draw samples (250 ml)
for the physical and chemical parameters were collected immediately after. After collecting the
first-draw samples, the flushed samples (250 ml) were collected after at least 5 minutes of flushing.
Sixty ml of both first draw and flush samples were filtered through 0.45-µm to produce filtered
samples. Preferably, samples were collected by utility staff; alternatively, homeowners were asked to
collect the samples. In either case, clear instructions were provided.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
190 | Effect of Nitrification on Corrosion in the Distribution System
©2010 Water Research Foundation. ALL RIGHTS RESERVED
APPENDIX E
UTILITY INTERVIEW QUESTIONNAIRE
Helene Baribeau
The purpose of this questionnaire is to obtain background information on your system to
help build case studies and utility guidelines that will appear in the final report of this project. Case
studies will be developed from analysis of your corrosion data (collected from the Lead and
Copper Rule compliance monitoring) and nitrification data (as available). We are also interested in
incorporating any special studies that may have been conducted in your distribution system in
relation to corrosion, nitrification, or the interplay between nitrification and corrosion. We may
contact you to further discuss any important issues indicated on your questionnaire.
Please respond to this questionnaire as accurately as possible. If a question would require
unreasonable search time, feel free to skip it and add a note in this regard (e.g., “too difficult to
find”). If you use both free chlorinated and chloraminated waters in your distribution system,
please respond only for the portion of your system that is chloraminated. Note that all responses
will remain anonymous. The survey addresses the following topics:
Section A – Background Information
Section B – Nitrification
Section C – Corrosion
Section D – Closing Remarks
Thank you very much for your time and support to our project!
If you have any questions, please contact:
Hélène Baribeau, Ph.D., P.E.
Carollo Engineers, P.C.
199 South Los Robles Avenue, Suite 530
Pasadena, California 91101
Phone: 626-535-0180, ext.238
Cell:
714-716-6715
Fax:
626-535-0185
E-mail: [email protected]
191
©2010 Water Research Foundation. ALL RIGHTS RESERVED
192 | Effect of Nitrification on Corrosion in the Distribution System
SECTION A: BACKGROUND INFORMATION
Question A1
Contact information for your utility:
Utility Name:
_______________________________________________
Contact Person: _______________________________________________
Title / Responsibility
_______________________________________________
Address1:
_______________________________________________
Address2:
_______________________________________________
City, State, Zip: _______________________________________________
Phone:
_______________________________________________
Cell:
_______________________________________________
Fax:
_______________________________________________
Email:
_______________________________________________
Question A2
Start date of chloramination (month, year):
Start date:
_____________________________________________________
Question A3
Do you periodically switch to free chlorine as distribution system residual disinfectant?
No
Yes
How often do you switch from chloramine to free chlorine?
Monthly
Quarterly
Seasonally
Once per year
Twice per year
As needed. Specify: _______________________
When you switch to free chlorine, how long do you apply it?
____________________________________________
When you switch to free chlorine, what is the target chlorine residual leaving your facility?
_____________________________
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix E: Utility Interview Questionnaire | 193
Question A4
Provide an estimate of the size and characteristics of your drinking water distribution system:
Number of service connections:
Number of persons served:
Number of storage facilities (i.e., tanks and reservoirs):
Total storage capacity in tanks and reservoirs:
Number of pressure zones:
Number of entry points in the distribution system during typical
operation:
Question A5
Provide the estimated proportion (percentage) of each of the following pipe materials and diameters
in your distribution system:
Pipe diameter
Pipe material
≤4 inches
6 to 8 inches
10 to 12 inches
Asbestos cement
Cast iron (unlined)
Cast iron
(cement-mortar lined)
Ductile iron (unlined)
Ductile iron
(cement-mortar lined)
Concrete
Steel
Polyvinyl chloride
(PVC)
Other: ________________
©2010 Water Research Foundation. ALL RIGHTS RESERVED
>12-inches
194 | Effect of Nitrification on Corrosion in the Distribution System
Question A6
Provide the estimated age of each of the following pipe materials and diameters in your distribution
system:
Pipe diameter
≤4 inches
Pipe material
6 to 8 inches
10 to 12 inches
>12 inches
Asbestos cement
Cast iron (unlined)
Cast iron (cement-mortar lined)
Ductile iron (unlined)
Ductile iron (cement-mortar
lined)
Concrete
Steel
Polyvinyl chloride (PVC)
Others: _______________
Question A7
Provide the estimated proportion (percentage) of each of the following pipe materials for the premise
(building/home service lines) plumbing of your distribution system:
Lead:
___________________________________________
Copper:
___________________________________________
Polyvinyl chloride (PVC):
___________________________________________
Galvanized iron:
___________________________________________
Stainless steel:
___________________________________________
Others (specify):
___________________________________________
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix E: Utility Interview Questionnaire | 195
Question A8
What percentage of homes in your system was built before the lead solder ban in 1986?
0 to 10%
11 to 20%
21 to 30%
31 to 40%
41 to 50%
51 to 60%
61 to 70%
71 to 80%
81 to 90%
91 to 100%
Don’t know
SECTION B : NITRIFICATION
Question B1
Has your system experienced nitrification in the distribution system?
No
Yes On average, how often do you respond to nitrification, either by increasing monitoring or
implementing more drastic measures such as flushing or breakpoint chlorination:
Once per week
Once per 2 weeks
Once per month
Once per 2 months
Once per season
Other. Specify: ________________________________
©2010 Water Research Foundation. ALL RIGHTS RESERVED
196 | Effect of Nitrification on Corrosion in the Distribution System
Question B2
Has your system experienced nitrification in premise (building/home service lines) plumbing?
We have never monitored for nitrification in premise plumbing using stagnant first draw samples.
No
Yes How was it brought to your attention?
___________________________________________________________________
On average, how often do you detect nitrification in premise plumbing:
Once per week
Once per 2 weeks
Once per month
Once per 2 months
Once per season
Other. Specify: ________________________________
Question B3
Provide the estimated occurrence of nitrification in the following pipe materials and diameters in
your distribution system (as a percentage of total number of occurrences):
Pipe diameter
≤4 inches
Pipe material
6 to 8 inches
10 to 12 inches
Asbestos cement
Cast iron (unlined)
Cast iron (cement-mortar lined)
Ductile iron (unlined)
Ductile iron
(cement-mortar lined)
Concrete
Steel
Polyvinyl chloride (PVC)
Others: ________________
©2010 Water Research Foundation. ALL RIGHTS RESERVED
>12 inches
Appendix E: Utility Interview Questionnaire | 197
Question B4
Provide the estimated occurrence of nitrification in the following pipe materials in the premise
(building/home service lines) plumbing (as a percentage of the number of occurrences in premise
plumbing):
Lead:
___________________________________________
Copper:
___________________________________________
Polyvinyl chloride (PVC):
___________________________________________
Galvanized iron:
___________________________________________
Stainless steel:
___________________________________________
Others (specify):
___________________________________________
©2010 Water Research Foundation. ALL RIGHTS RESERVED
198 | Effect of Nitrification on Corrosion in the Distribution System
Question B5
Based on YOUR experience, rate the following methods used to monitor for nitrification:
Rate the effectiveness of each method: 0: Do not apply this method, 1: Not effective in your system, 2:
Slightly effective in your system, 3: Effective in your system, 4: Very effective in your system, or
5: Essential.
Indicate the monitoring frequency used in YOUR distribution system: 1: Every 4 hours, 2: Daily, 3:
Weekly, 4: Bi-weekly, 5: Monthly, 6: Quarterly, 7: Annually, or 8: Not analyzed.
Also, indicate the number of sampling locations.
Parameter
Effectiveness
Monitoring
frequency
Change in chlorine-to-ammonia-N ratio
Decrease in total chlorine residual
Change in chloramine speciation
Decrease in dissolved oxygen
Decrease in free ammonia residual
Increase in free ammonia residual
Decrease in total ammonia residual
Increase in total ammonia residual
Increase in nitrite
Increase in nitrate
Decrease in pH
Increase in ammonia-oxidizing bacteria
Increase in coliforms
Increase in HPC (R2A)
Increase in HPC (agar)
Temperature
Increase in assimilable organic carbon (AOC)
Others (specify):
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Number of
locations
Appendix E: Utility Interview Questionnaire | 199
Question B6
What methods have YOU used to prevent nitrification? 0: Do not apply this method, 1: Not
effective in your system, 2: Slightly effective in your system, 3: Effective in your system, 4: Very
effective in your system, or 5: Essential.
Blend water:
Use chloramination booster stations:
Use chlorination booster stations:
Improve chloramine stability (e.g., implement GAC, change coagulation
process):
Increase chloramine residual at the plant effluent(s):
Decrease ammonia residual at the entry point(s):
Modify chlorine-to-ammonia ratio:
Modify treatment process other than disinfection:
Increase water pH:
Flush the affected area of the distribution system:
Flush the entire distribution system:
More aggressive cleaning such as swabbing or pigging:
Change distribution system hydraulics, eliminate dead-ends:
Implement pipe corrosion control program:
Increase reservoir/tank mixing:
Reduce reservoir/tank storage (cycle reservoir or increase turnover rate):
Clean reservoir/tank:
Seasonal switch to free chlorine:
Use chlorine dioxide (specify the point of addition): ___________
Add chlorite:
Increase water quality monitoring:
Others (specify): ______________________________________
©2010 Water Research Foundation. ALL RIGHTS RESERVED
200 | Effect of Nitrification on Corrosion in the Distribution System
Question B7
What methods have YOU used to correct nitrification after it takes place? 0: Do not apply this
method, 1: Not effective, 2: Slightly effective, 3: Effective, 4: Very effective, or 5: Essential.
Increase chloramine residual:
Modify chlorine-to-ammonia ratio:
Decrease ammonia residual at the entry point:
Increase water pH:
Partially drain reservoir/tank:
Breakpoint chlorinate reservoir/tank and problem area(s):
Breakpoint the entire chloraminated area of the system:
Clean reservoir/tank:
Flush the affected area of the distribution system:
Flush the entire distribution system:
Change distribution system hydraulics, eliminate dead-ends:
Use chlorine dioxide (specify the point of addition): ___________
Add chlorite:
Others (specify): ______________________________________
SECTION C : CORROSION
Question C1
Is your system on reduced monitoring for the Lead and Copper Rule compliance?
Yes
No
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix E: Utility Interview Questionnaire | 201
Question C2
What corrosion control treatment practice(s) do you currently use? Check all that apply.
Corrosion control is not practiced
pH adjustment.
If you adjust pH, your target pH is: ________________________
What method do you use to modify the pH: __________________
_____________________________________________________
Add corrosion inhibitor. If so, which one (if it is a blend, please specify the
composition)? _________________________________________
Dose of the corrosion inhibitor? ____________________________
Target residual in the distribution system? ___________________
Raise the dissolved inorganic carbonate (DIC). If so, what compound do you use?
____________________________________________________
Dose of compound used to raise DIC? ______________________
Question C3
Has your system experienced nitrification (as characterized per Question B5 above) in the presence
of corrosion problems:
No
Yes. If so, please fill out the following table:
Corrosion problem in the
presence of nitrification
Distribution system material that was suspected
to be corroded at a higher rate and/or affected
by nitrification
Higher incidence of red
water
Higher incidence of blue
water
High lead concentration
High copper concentration
Others (specify):
________________
________________
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Frequency of
occurrence
(once per …)
202 | Effect of Nitrification on Corrosion in the Distribution System
SECTION D: CLOSING REMARKS
Question D1
Comments about any of the answers provided above.
______________________________________________________________________
______________________________________________________________________
______________________________________________________________________
______________________________________________________________________
______________________________________________________________________
Question D2
Has your system conducted special studies related to nitrification and/or corrosion in your
distribution system and/or premise (building/home) plumbing?
No
Yes.
Provide a brief description of these studies: _________________
_____________________________________________________
_____________________________________________________
_____________________________________________________
Question D3
Please provide the following data, IF EASILY AVAILABLE, collected at the entry points to the
distribution system (after addition of all chemicals) and in the distribution system, including premise
(buildings/homes) plumbing. Three years of REPRESENTATIVE data received in electronic files would
be preferable:
¾ Water temperature (indicate °F or °C)
¾ pH
¾ Total chlorine residual (mg/L Cl2)
¾ Chlorine-to-ammonia-N weight ratio (mg:mg)
¾ Ammonia concentration (specify free or total ammonia, and specify mg/L as N or mg/L as NH3 or
NH4+)
¾ Nitrite concentration (specify mg/L as N or mg/L NO2-)
¾ Nitrate concentration (specify mg/L as N or mg/L NO3-)
¾ Heterotrophic plate counts (HPC) (specify the incubation media [R2A or agar], the incubation
temperature [indicate °F or °C], and incubation time)
¾ Total organic carbon (TOC) concentration (mg/L C)
¾ Dissolved organic carbon (DOC) concentration (mg/L C)
¾ Trihalomethanes (THMs) concentration (µg/L) (chloroform, bromodichloromethane,
dibromochloromethane, and bromoform)
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Appendix E: Utility Interview Questionnaire | 203
¾ Alkalinity (mg/L CaCO3)
¾ Lead concentration (specify units)
¾ Copper concentration (specify units)
¾ Calcium concentration (specify mg/L, or mg/L as CaCO3)
¾ Total dissolved solids (TDS) (specify units)
¾ Dissolved oxygen concentration (mg/L O2)
¾ Chloride concentration (mg/L Cl-)
¾ Sulfate concentration (specify mg/L as S or mg/L SO42-)
¾ Potassium concentration (mg/L K)
¾ Aluminum (mg/L Al)
¾ Zinc (mg/L Zn)
¾ Iron concentration (specify units)
¾ Manganese (specify units)
¾ Magnesium (specify units)
¾ Phosphate concentration (specify mg/L PO4 or mg/L P)
¾ Silica concentration (specify mg/L as SiO2 or Si)
Question D4
How long did it take you to complete this survey (without considering interruptions)?
Less than 1h00
1h00 to 2h00
2h00 to 3h00
3h00 to 4h00
4h00 to 5h00
5h00 to 6h00
6h00 to 7h00
7h00 to 8h00
8h00 to 9h00
9h00 to 10h00
©2010 Water Research Foundation. ALL RIGHTS RESERVED
204 | Effect of Nitrification on Corrosion in the Distribution System
©2010 Water Research Foundation. ALL RIGHTS RESERVED
REFERENCES
Aleem, M.I.H. & Alexander, M. 1960. Nutrition and physiology of nitrobacter agilis. Applied and
Environmental Microbiology, 8:2:80.
Alleman, J.E., Keramida, V. & Panteakiser, L. 1987. Light-induced nitrosomonas inhibition.
Water Research, 21:4:499.
Alleman, J.E. & Preston, K. 2005. Behavior and physiology of nitrifying bacteria.
<http://bridge.ecn.purdue.edu/~alleman/w3-articles/nitrifier-physiology/nitrifier-behavior
.html>.
Allen, M.J., Edberg, S.C. & Reasoner, D.J. 2004. Heterotrophic plate count bacteria - what is their
significance in drinking water? International Journal of Food Microbiology, 92:3:265.
Atlas, D., Coombs, J. & Zajicek, O.T. 1982. The corrosion of copper by chlorinated drinking
waters. Water Research, 16:5:693.
Amann, R., B. M. Fuchs, and S. Behrens. 2001. The identification of microorganisms by
fluorescence in situ hybridisation. Current Opinion in Biotechnology 12:231-236.
ATSDR.
2005.
Toxicity
of
chlorite
and
chlorine
dioxide.
<http://www.atsdr.cdc.gov/toxprofiles/phs160.html>.
AWWA. 1996. Water stats survey/database. Denver, CO.
AWWARF. 2004. Occurrence survey for boron and hexavalent chromium AWWARF.
AWWARF & DVGW-TZW. 1996. 2nd. Internal corrosion of water distribution systems. Denver,
CO:
American
Water
Works
Association
Research
Foundation,
DVGW-Technologiezentrum Wasser.
Baribeau, H. 2006. Fundamentals and control of nitrification in chloraminated drinking water
distribution systems, M56.
Bastidas, E. & Sánchez-Silva, M. 2006. Coupled reliability model of biodeterioration, chloride
ingress
and
cracking
for
reinforced
concrete
structures.
<http://dspace.uniandes.edu.co:5050/dspace/bitstream/1992/906/1/Balkema_Emilio_Bast
idas.pdf >.
Belser, L.W. & Mays, E.L. 1980. Specific inhibition of nitrite oxidation by chlorate and its use in
assessing nitrification in soils and sediments. Applied and Environmental Microbiology,
39:3:505.
Belser, L.W. & Schmidt, E.L. 1978. Serological diversity within a terrestrial ammonia-oxidizing
population. Applied and Environmental Microbiology, 36:4:589.
Benninger, R.W. & Sherrard, J.H. 1978. Nitrification and alkalinity relationship in activated
sludge. Journal of Water Pollution Control Federation 50:2132.
Both, G.J., Gerards, S. & Laanbrock, H.J. 1990. Enumeration of Nitrite-Oxidizing Bacteria
In Grassland Soils Using A Most Probable Number Technique: Effect of Nitrite
Concentration and Sampling Procedure. FEMS Microbio Ecol 74:277.
Bock, E., Koops, H.P., Ahlers, B. & Harms, H. 1992. Oxidation of inorganic nitrogen compounds
as energy source. . The prokaryotes.
Bock, E., Koops, H.P., Harms, H. & Ahler, B. 1991. The biochemistry of nitrifying organism. .
Variations in autotrophic life:360.
Bodelier, P.L.E. & Laanbroek, H.J. 1997. Oxygen uptake kinetics of pseudomonas chlororaphis
grown in glucose- or glutamate-limited continuous cultures. Archives of Microbiology,
167:6:392.
205
©2010 Water Research Foundation. ALL RIGHTS RESERVED
206 | Effect of Nitrification on Corrosion in the Distribution System
Bollmann, A., Bar-Gilissen, M.J. & Laanbroek, H.J. 2002. Growth at low ammonium
concentrations and starvation response as potential factors involved in niche differentiation
among ammonia-oxidizing bacteria. Applied and Environmental Microbiology,
68:10:4751.
Bollmann, A., Schmidt, I., Saunders, A.M. & Nicolaisen, M.H. 2005. Influence of starvation on
potential ammonia-oxidizing activity and amoa mrna levels of nitrosospira briensis.
Applied and Environmental Microbiology, 71:3:1276.
Boon, B. & Laudelout, H. 1962. Kinetics of nitrite oxidation by nitrobacter winogradskyi. The
Biochemical Journal, 85:3:440.
Both G.J., Gerards S, Laanbrock H.J. 1990. Enumeration of Nitrite-Oxidizing Bacteria In
Grassland Soils Using A Most Probable Number Technique: Effect of Nitrite
Concentration and Sampling Procedure. FEMS Microbio Ecol 74:277-286.
Bott, C. 2005. Evaluation of zinc inhibition of nitrification and bnr phases iii: Final report.
<http://www.epa.gov/dclead/Final_Zinc_Arlington_WPCP_Study_VMI_9-20-05.pdf>,
[7/25].
Boulay, N. & Edwards, M. 2001. Role of temperature, chlorine, and organic matter in copper
corrosion by-product release in soft water. Water Research, 35:3:683.
Braam, F. & Klapwijk, A. 1981. Effect of copper on nitrification in activated-sludge. Water
Research, 15:9:1093.
Bremer, P.J., Webster, B.J. & Wells, D.B. 2001. Biocorrosion of copper in potable water. Journal
American Water Works Association, 93:8:82.
Burrell, P.C., Keller, J. & Blackall, L.L. 1998. Microbiology of a nitrite-oxidizing bioreactor.
Applied and Environmental Microbiology, 64:5:1878.
Campbell, H.S. 1950. Pitting Corrosion in Copper Water Pipes Caused by Films of
Carbonaceous Material Produced during manufacture. Journal of Industrial Metal 77:345.
Camper, A.K. 2004. Involvment of Humic Substances in Regrowth. International journal of
food microbiology., 92:335.
Camper, A.K., Brastrup, K., Sandvig, A., Clement, J., Spencer, C. & Capuzzi, A.J. 2003. Effect of
distribution system materials on bacterial regrowth. Journal American Water Works
Association, 95:7:107.
Cantor, A.F., Bushman, J.B., Glodoski, M.S., Kiefer, E., Bersch, R. & Wallenkamp, H. 2006.
Copper pipe failure by microbiologically influenced corrosion. Materials Performance,
45:6:38.
Cantor, A.F., Park, J.K. & Vaiyavatjamai, P. 2003. Effect of chlorine on corrosion in drinking
water systems. Journal American Water Works Association, 95:5:112.
CDC. 2008 Possible link of poor pou devices, plumbing to disease. Plumbing & Mechanical
Chain, P., Lamerdin, J., Larimer, F., Regala, W., Lao, V., Land, M., Hauser, L., Hooper, A., Klotz,
M., Norton, J., Sayavedra-Soto, L., Arciero, D., Hommes, N., Whittaker, M. & Arp, D.
2003. Complete genome sequence of the ammonia-oxidizing bacterium and obligate
chemolithoautotroph nitrosomonas europaea. Journal of Bacteriology, 185:9:2759.
Chakrabarti, T. & Jones, P.H. 1983. Effect of molybdenum and selenium addition on the
denitrification of wastewater. Water Research, 17:8:931.
Characklis, W.G. & Gujer, W. 1979. Temperature dependency of microbial reactions. Kinetics of
wastewater treatments:111.
Charley, R.D., Hooper, D.G. & Lee, A.G. 1980. Nitrification kinetics in activated sludge at various
temperatures and dissolved oxygen concentrations. Water Research, 14:1387.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
References | 207
Clark, C. & Schmidt, E.L. 1966. Effect of mixed culture on nitrosomonas europaea simulated by
uptake and utilization of pyruvate. Journal of Bacteriology, 91:1:367.
Clark, C. & Schmidt, E.L. 1967. Growth response of nitrosomonas europaea to amino acids.
Journal of Bacteriology, 93:4:1302.
Clesceri, L.S., Greenberg, A.E. & Eaton, A.D. 1998. 20. Standard methods for the examination of
water and wastewater. Washington, DC: American Public Health Association; American
Water Works Association and Water Environment Federation.
Committee on Public Water Supply Distribution Systems: Assessing and Reducing Risks, N.R.C.
2006. Drinking water distribution systems: Assessing and reducing risks. Washington,
DC: The National Academy Press.
Coskuner, G. & Curtis, T.P. 2002. In situ characterization of nitrifiers in an activated sludge plant:
Detection of nitrobacter spp. Journal of Applied Microbiology, 93:3:431.
Cunliffe, D.A. 1991. Bacterial nitrification in chloraminated water-supplies. Applied and
Environmental Microbiology, 57:11:3399.
Daims, H., Maixner, F., Lucker, S., Stoecker, K., Hace, K. & Wagner, M. 2006. Ecophysiology
and niche differentiation of nitrospira-like bacteria, the key nitrite oxidizers in wastewater
treatment plants. Water Science and Technology, 54:1:21.
Davidson, D. 1996. Effects of arthrobacters, acidovorax delafieldii, and bacillus megaterium
colonization on copper solvency in a laboratory test loop. Biofouling, 9:279.
deCantanzaro, J.B. & Hutchinson, T.C. 1985. Effects of nickel addition on nitrogen
mineralization, nitrification, and nitrogen leaching in some boreal forest soils. Water, Air,
and Soil Pollution, 24:153.
DeGraca,
A.
2005.
San
francisco's
chloramine
conversion
experience.
<http://www.nyc.gov/html/dep/pdf/conference/ppdegraca.pdf>, [Sep, 23rd].
Disalvo, R.M. & Sherrard, J.H. 1980. The stimulation of nitrification at low nickel concentrations.
Industrial Wastewater Treatment-Nitrification:175.
Dodrill, D. & Edwards, M. 1995. Corrosion control on the basis of utility experience. Journal
American Water Works Association, 87:7:74.
Douglas, I., Guthmann, J., Muylwyk, Q. & Snoeyink, V. 2004. Corrosion control in the city of
ottawa—comparison of alternatives and case study for lead reduction in drinking water.
Eleventh Canadian National Conference and Second Policy Forum on Drinking Water
Dudi, A. 2004. Reconsidering lead corrosion in drinking water: Product testing, direct chloramine
attack and galvanic corrosion. Master's. Virginia Tech, Blacksburg,VA.
Edwards, M., Bosch, D., Loganathan, G.V. & Dietrich., A.M. 2003. The future challenge of
controlling distribution system water quality and protecting plumbing infrastructure:
Focusing on consumers. IWA Leading Edge Conference
Edwards, M. & Boulay, N. 2001. Organic matter and copper corrosion by-product release: A
mechanistic study. Corrosion Science, 43:1:1.
Edwards, M. & Dudi, A. 2004. Role of chlorine and chloramine in corrosion of lead-bearing
plumbing materials. Journal American Water Works Association, 96:10:69.
Edwards, M. & Ferguson, J.F. 1993. Accelerated testing of copper corrosion. Journal American
Water Works Association, 85:10:105.
Edwards, M., Hidmi, L. & Gladwell, D. 2002. Phosphate inhibition of soluble copper corrosion
by-product release. Corrosion Science, 44:5:1057.
Edwards, M., Jacobs, S. & Dodrill, D. 1999. Desktop guidance for mitigating pb and cu corrosion
by-products Journal American Water Works Association, 91:5:66.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
208 | Effect of Nitrification on Corrosion in the Distribution System
Edwards, M., Marshall, B., Zhang, Y., Abbott, S. & Strickhouser.., A. 2005. Unintended
consequences of chloramine hit home. AWWA WQTC. Quebec City, Canada.
Edwards, M., Schock, M.R. & Meyer, T.E. 1996. Alkalinity, ph, and copper corrosion by-product
release. Journal American Water Works Association, 88:3:81.
Edwards, M. & Triantafyllidou, S. 2007a. Chloride-to-sulfate mass ratio and lead leaching to
water. Journal American Water Works Association, 99:7:96.
Edwards, M. & Triantafyllidou, S. 2007b. Chloride to sulfate mass ratio and lead leaching to
water. Journal American Water Works Association, 99:7:96.
Ehrich, S., Behrens, D., Lebedeva, E., Ludwig, W. & Bock, E. 1995. A new obligately
chemolithoautotrophic, nitrite-oxidizing bacterium, nitrospira moscoviensis sp. Nov. And
its phylogenetic relationship Archives of Microbiology, 164:1:16.
Eichler, S., Christen, R., Holtje, C., Westphal, P., Botel, J., Brettar, I., Mehling, A. & Hofle, M.G.
2006. Composition and dynamics of bacterial communities of a drinking water supply
system as assessed by rna- and DNA-based 16s rrna gene fingerprinting. Applied and
Environmental Microbiology, 72:3:1858.
Engel, M.S. & Alexander, M. 1958. Growth and autotrophic metabolism of nitrosomonas
europaea. Journal of Bacteriology, 76:217.
Ensign, S.A., Hyman, M.R. & Arp, D.J. 1993. In vitro activation of ammonia monooxygenase
from nitrosomonas europaea by copper. Journal of Bacteriology, 175:7:1971.
Fairey, J.L., Katz, L.E. & Speitel, G.E. 2004. Monochloramine destruction in gac beds.
Proceedings of the 2004 AWWA WQTC. .
Feben, D. 1935. Nitrifying bacteria in water supplies. . Journal American Water Works
Association, 27:4:439.
Federal-Register. 1991. Lead and copper-final rule, 56:110:26460
Federal-Register. 2006. National primary drinking water regulations: Stage 2 disinfectants and
disinfection
byproducts
rule.
<http://www.epa.gov/fedrgstr/EPA-WATER/2006/Janurary/Day-04/w03/htm>, [2/8].
Ferguson, B.A., DiGiano, F.A. & Chandran, K. 2005. Impact of temporary switches from
monochloramine to free chlorine on water quality in distribution systems. . 2005 AWWA
annual conference. San Francisco, CA, 2005.
Ferguson, S.J. 1998. Nitrogen cycle enzymology. Current Opinion in Chemical Biology, 2:2:182.
Finstein, M.S. & Delwiche, C.C. 1965. Molybdenum as a micronutrient for nitrobacter. Journal of
Bacteriology, 89:1:123.
Flannery, B., Gelling, L.B., Vugia, D.J., Weintraub, J.M., Salerno, J.J., Conroy, M.J., Stevens,
V.A., Rose, C.E., Moore, M.R., Fields, B.S. & Besser, R.E. 2006. Reducing legionelia
colonization of water systems with monochloramine. Emerging Infectious Diseases,
12:4:588.
Fleming, K.K., Harrington, G.W. & Noguera, D.R. 2005. Nitrification potential curves: A new
strategy for nitrification prevention. Journal American Water Works Association, 97:8:90.
Fleming, K.K., Harrington, G.W. & Noguera, D.R. 2008. Using nitrification potential curves to
evaluate full-scale drinking water distribution systems. Journal American Water Works
Association, 100:10:92.
Focht, D.D. & Verstraete, W. 1977. Biochemical ecology of nitrification and denitrification.
Advances in microbial ecology, , 1:135.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
References | 209
Fransolet, G., Depelchin, A., Villers, G., Goossens, R. & Masschelein, W.J. 1988. The role of
bicarbonate in bacterial-growth in oligotrophic waters. Journal American Water Works
Association, 80:11:57.
Furumai, H. & Rittmann, B.E. 1992. Advanced modeling of mixed populations of heterotrophs
and nitrifiers considering the formation and exchange of soluble microbial products. Water
Science and Technology, 26:3-4:493.
Furumai, H. & Rittmann, B.E. 1994. Evaluation of multiple-species biofilm and floc processes
using a simplified aggregate model. Water Science and Technology, 29:10-11:439.
Garret, J.H. 1891. The action of water on lead. London, England. : H.K. Lewis.
Gates, J. & Lavinder, S. 1997. Action plan for improving chloramine residuals and minimizing the
occurrence of nitrification in the florida keys aqueduct authority’s distribution system.
AWWA water quality and technology conference. .
Geets, J., Boon, N. & Verstraete, W. 2006. Strategies of aerobic ammonia-oxidizing bacteria for
coping with nutrient and oxygen fluctuations. FEMS Microbiology Ecology, 58:1:1.
Gibbs, W.M. 1920. The isolation and study of nitrifying bacteria. Soil Science, 8:427.
Gieseke, A., Bjerrum, L., Wagner, M. & Amann, R. 2003. Structure and activity of multiple
nitrifying bacterial populations co-existing in a biofilm. Environmental Microbiology,
5:5:355.
Gieseke, A., Nielsen, J.L., Amann, R., Nielsen, P.H. & de Beer, D. 2005. In situ substrate
conversion and assimilation by nitrifying bacteria in a model biofilm. Environmental
Microbiology, 7:9:1392.
Gieseke, A., Purkhold, U., Wagner, M., Amann, R. & Schramm, A. 2001. Community structure
and activity dynamics of nitrifying bacteria in a phosphate-removing biofilm. Applied and
Environmental Microbiology, 67:3:1351.
Grady, C.P.L., Daigger, G.T. & Lim, H.C. 1999. 2nd edition. Biological wastewater treatment.
New York. : Basel. Marcel Dekker, Inc.
Groeneweg, J., Sellner, B. & Tappe, W. 1994. Ammonia oxidation in nitrosomonas at nh3
concentrations near km- effects of ph and temperature. Water Research, 28:12:2561.
Goeres, M.D., et al. 2005. Statistical Assessment of a Laboratory Method for Growing Biofilms
Microbiology 151:757-762.
Grunditz, C. & Dalhammar, G. 2001. Development of nitrification inhibition assays using pure
cultures of nitrosomonas and nitrobacter. Water Research, 35:2:433.
Guo, Q., Toomuluri, P.J. & Eckert, J., J.O. . 1998. Leachability of regulated metals from
cement-mortar linings. Journal American Water Works Association, 90:3:62.
Guo, X., W. , Chu, K.G. & Qiao, L. 2002. Relativity between corrosion-induced stress and stress
corrosion cracking of brass in an ammonia solution. Journal of University of Science and
Technology Beijing 9:6:431.
Hallama, N.B., Westa, J.R., C.F. Forstera, Powellb, J.C. & Spencerc, I. 2002. The decay of
chlorine associated with the pipe wall in water distribution systems. Wat. Res., 36:3479.
Hammes, F., Berney, M., Wang, Y.Y., Vital, M., Koster, O. & Egli, T. 2008. Flow-cytometric total
bacterial cell counts as a descriptive microbiological parameter for drinking water
treatment processes. Water Research, 42:1-2:269.
Harms, L.L. & Owen, C.A. 2004. Keys for controlling nitrification. AWWA WQTC.
Harper, S.C., Manoharan, R., Mavinic, D.S. & Randall, C.W. 1996. Chromium and nickel toxicity
during the biotreatment of high ammonia landfill leachate. Water Environment Research,
68:1:19.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
210 | Effect of Nitrification on Corrosion in the Distribution System
Harrington, G.W. 2002. Ammonia release from chloramine decay: Implications for the prevention
and control of nitrification in distribution systems. Denver, Colorado:AWWA Reserach
Foundation and AWWA.
Harrington, G.W., Noguera, D.R., Kandou, A.I. & Vanhoven, D.J. 2002. Pilot-scale evaluation of
nitrification control strategies. Journal American Water Works Association, 94:11:78.
Hockenbury, M.R., Daigger, G.T. & Grady, C.P.L. 1977. Factors affecting nitrification. Journal of
the Environmental Engineering Division-Asce, 103:1:9.
Hermansson, A., and P.-E. Lindgren. 2001. Quantification of Ammonia-Oxidizing Bacteria in
Arable Soil by Real-Time PCR. Applied and Environmental Microbiology 67:972-976.
Hill, J.C. 1946. Bacterial oxidation of ammonia in circulating water. Journal American Water
Works Association, 38:980.
Hockenbury, M.R. & Grady, C.P.L. 1977. Inhibition of nitrification - effects of selected organic
compounds. Journal of Water Pollution Control Federation, 49:5:768.
Hommes, N.G., Sayavedra-Soto, L.A. & Arp, D.J. 2003. Chemolithoorganotrophic growth of
nitrosomonas europaea on fructose. Journal of Bacteriology, 185:6809.
Hong, P.K. & MacCauley, Y. 1998. Corrosion and leaching of copper tubing exposed to
chlorinated drinking water. Water Air Soil Pollution, 108:457.
Hooper, A.B. & Terry, K.R. 1974. Photoinactivation of ammonia oxidation in nitrosomonas.
Journal of Bacteriology, 119:899.
Hovanec, T.A., Taylor, L.T., Blakis, A. & Delong, E.F. 1998. Nitrospira-like bacteria associated
with nitrite oxidation in freshwater aquaria. Applied and Environmental Microbiology,
64:1:258.
Huang, Y.H. & Zhang, T.C. 2005. Effects of dissolved oxygen on formation of corrosion products
and concomitant oxygen and nitrate reduction in zero - valent iron systems with or without
aqueous fe2+ Water Research, 39:9:1751.
Hubert, C., Nemati, M., Jenneman, G. & Voordouw, G. 2005. Corrosion risk associated with
microbial souring control using nitrate or nitrite. Applied Microbiology and Biotechnology,
68:2:272.
Hulbert, R. 1933. Ammonia-chlorine treatment yields nitrite in effluent. Engrg. News-Record.
Hynes, R.K. & Knowles, R. 1983. Inhibition of chemoautotrophic nitrification by sodium-chlorate
and sodium-chlorite - a reexamination. Applied and Environmental Microbiology,
45:4:1178.
Ibrahim, A.M. 1989. Heavy metal inhibition of resting nitrifying bacteria. 43rd Purdue university
industrial waste conference proceedings.:309.
Ike, N.R., Wolfe, R.L. & Means, E.G. 1988. Nitrifying bacteria in a chloraminated drinking-water
system. Water Science and Technology, 20:11-12:441.
Jansson, S.L. 1958. Tracer studies on nitrogen transformations in soil with special attention to
mineralization-immobilization relationships. Kungl. Lantbrukshögskolans Annaler
14:101.
Jetten, M.S.M., Horn, S.J. & van Loosdrecht, M.C.M. 1997. Towards a more sustainable
wastewater treatment system. Water Science & Technology, 35:9:171.
Jones, D.A. 1996. 2nd. Principles and preventions of corrosion. Upper Saddle River, NJ: Simon &
Schuster.
Jun, B.H., Tanji, Y. & Unno, H. 2000. Stimulating accumulation of nitrifying bacteria in porous
carrier by addition of inorganic carbon in a continuous-flow fluidized bed wastewater
treatment reactor. Journal of Bioscience and Bioengineering, 89:4:334.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
References | 211
Juretschko, S., Loy, A., Lehner, A. & Wagner, M. 2002. The microbial community composition of
a nitrifying-denitrifying activated sludge from an industrial sewage treatment plant
analyzed by the full-cycle rrna approach. Systematic and Applied Microbiology, 25:84.
Kaltwasser, H. 1976. Destruction of concrete by nitrification. Applied Microbiology and
Biotechnology 3:3:185.
Karim, M.R. & LeChevallier, W.R. 2006. Alternative treatments for nitrification control in
chloraminating utilities. AWWA annual conference.
Keen, G.A. & Prosser, J.I. 1987. Steady-state and transient growth of autotrophic nitrifying
bacteria. Archives of Microbiology, 147:1:73.
Kielemoes, J., Deboever, P. & Verstraete, W. 2000. Influence of denitrification on the corrosion of
iron and stainless steel powder. Envir. Sci. Technol., 34:4:663.
Killham, K. 1986. Heterotrophic nitrification. Nitrification.
Kimbrough, D.E. 2007. Brass corrosion as a source of lead and copper in traditional and all-plastic
distribution systems Journal American Water Works Association, 99:8:70.
Kindaichi, T., Ito, T. & Okabe, S. 2004. Ecophysiological interaction between nitrifying bacteria
and heterotrophic bacteria in autotrophic nitrifying biofilms as determined by
microautoradiography-fluorescence in situ hybridization. Applied and Environmental
Microbiology, 70:3:1641.
Kirmeyer, G.J., Foust, G.W., Pierson, G.L., Simmler, J.J. & LeChevallier, M.W. 1993. Optimizing
chloramine treatment. . Denver Colorado:AWWARF and AWWA.
Kirmeyer, G.J.e.a. 1995. Nitrification occurrence and control in chloraminated water systems.
Denver, Colo, :AWWARF and AWWA.
Knowles, G., Downing, A.L. & Barret, M.J. 1965. Determination of kinetic constants for nitrifying
bacteria in mixed culture with the aid of an electronic computer. Journal of General
Microbiology 38:263.
Könneke, M., Bernhard, A.E., de la Torre, J.R., Walker, C.B., Waterbury, J.B. & Stahl, D.A. 2005.
Isolation of an autotrophic ammonia-oxidizing marine archaeon. Nature, 437:543.
Kool, J.L., Bergmire-Sweat, D., Butler, J.C., Brown, E.W., Peabody, D.J., Massi, D.S., Carpenter,
J.C., Pruckler, J.M., Benson, R.F. & Fields, B.S. 1999a. Hospital characteristics associated
with colonization of water systems by legionella and risk of nosocomial legionnaires'
disease: A cohort study of 15 hospitals. Infection Control and Hospital Epidemiology,
20:12:798.
Kool, J.L., Carpenter, J.C. & Fields, B.S. 1999b. Effect of monochloramine disinfection of
municipal drinking water on risk of nosocomial legionnaires' disease. Lancet,
353:9149:272.
Koops, H.P. & Pommerening-Roser, A. 2001. Distribution and ecophysiology of the nitrifying
bacteria emphasizing cultured species. FEMS Microbiology Ecology, 37:1:1.
Kors, L.J., Moorman, J.H.N., Wind, A.P.M. & van der Hoek, J.P. 1998. Nitrification and low
temperature in a raw water reservoir and rapid sand filters. Water Science & Technology,
37:2:169.
Korshin, G.V., Ferguson, J.F. & Lancaster, A.N. 2000. Influence of natural organic matter on the
corrosion of leaded brass in potable water. Corrosion Science, 42:1:53.
Korshin, G.V., Ferguson, J.F. & Lancaster, A.N. 2005. Influence of natural organic matter on the
morphology of corroding lead surfaces and behavior of lead-containing particles. Water
Research, 39:5:811.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
212 | Effect of Nitrification on Corrosion in the Distribution System
Korshin, G., Perry, S. & Ferguson, J. 1996. Influence of NOM on Copper Corrosion. Journal
American Water Works Association, 88:7:36.
Korshin, G.V., Perry, S.A.L., Edwards, M.A. & Ferguson, J.F. 1994. Influence of natural organic
matter on corrosion of copper in potable waters. Water Research: Proceedings American
Water Works Association Annual Conference. :797.
Kowalchuk, G. A., J. R. Stephen, W. De Boer, J. I. Prosser, T. M. Embley, and J. W. Woldendorp.
1997. Analysis of ammonia-oxidizing bacteria of the beta subdivision of the class
Proteobacteria in coastal sand dunes by denaturing gradient gel electrophoresis and
sequencing of PCR-amplified 16S ribosomal DNA fragments. Applied and Environmental
Microbiology 63:1489-1497.
Kowalchuk, G.A. & Stephen, J.R. 2001. Ammonia-oxidizing bacteria: A model for molecular
microbial ecology. Annual Review of Microbiology, 55:485.
Kristiansen, H. 1982. Corrosion of copper in soft water with different content of humic substances
and various temperatures. VATTEN, 38:181.
Krummel, A. & Harms, H. 1982. Effect of organic matter on growth and cell yield of
ammonia-oxidizing bacteria. Archives of Microbiology 133:1:50.
Kunzler, R. & Schwenk, W. 1983. Factors influencing nitrate reduction by galvanized and
non-galvanized plain steels. GWF-Wasser/Abwasser, 124:8:378.
Larson, T.E. 1939. Bacteria, corrosion and red water. Journal American Water Works Association,
31:1186.
Larson, T.E., King, R.M. & Henley, L. 1956. Corrosion of brass by chloramine. Journal American
Water Works Association:84.
Laudelot, H. & Tichelen, L.V. 1960. Kinetics of the nitrite oxidation by nitrobacter winogrdski.
Journal of Bacteriology, 79:392.
Lechevallier, M.W., Lowry, C.D. & Lee, R.G. 1990. Disinfecting biofilms in a model
distribution-system. Journal American Water Works Association, 82:7:87.
Lechevallier, M.W., Lowry, C.D., Lee, R.G. & Gibbon, D.L. 1993. Examining the relationship
between iron corrosion and the disinfection of biofilm bacteria. J. AWWA., 85:7:111.
Lee, D. Y., K. Shannon, and L. A. Beaudette. 2006. Detection of bacterial pathogens
oligonucleotide microarray in municipal wastewater using an ind real-time quantitative
PCR. Journal of Microbiological Methods 65:453-467.
Lee, C. 2004. Effect of various dissolved species on anaerobic iron corrosion. Master's. Virginia
Tech, Blacksburg, VA.
Lee, Y.W., Ong, S.K. & Sato, C. 1997. Effects of heavy metals on nitrifying bacteria. Water
Science and Technology, 36:12:69.
Lehtola, M.J., Miettinen, K.T., Keinanen, M.M., Kekki, T.K., Laine, O., Hirvonen, A., Vartiainen,
T. & Martikainen, P.J. 2004. Microbiology, chemistry and biofilm development in a pilot
drinking water distribution system with copper and plastic pipes. Water Research,
38:17:3769.
LePuil, M., Randall, A.A. & Taylor, J.S. 2003. Final report: Biofilm measurements in the pinellas
county drinking water distribution system. :University of central florida.
Leu, H.G., Lee, C.D., Ouyang, C.F. & Tseng, H.T. 1998. Effects of organic matter on the
conversion rates of nitrogenous compounds in a channel reactor under various flow
conditions. Water Research, 32:3:891.
Lewis, R.F. & Pramer, D. 1958. Isolation of nitrosomonas in pure culture. Journal of
Bacteriology, 76:524.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
References | 213
Lieu, N.I., Wolfe, R.L. & Means, E.G. 1993. Optimizing chloramine disinfection for the control of
nitrification. Journal American Water Works Association, 85:2:84.
Lipponen, M.T.T., Suutari, M.H. & Martikainen, P.J. 2002. Occurrence of nitrifying bacteria and
nitrification in finnish drinking water distribution systems. Water Research, 36:17:4319.
Liu, S.B., Taylor, J.S., Randall, A.A. & Dietz, J.D. 2005. Nitrification modeling in chloraminated
distribution systems. Journal American Water Works Association, 97:10:98.
Loveless, J.E. & Painter, H.A. 1968. The influence of metal ion concentrations and ph value on the
growth of a nitrosomonas strain isolated from activated sludge. Journal of General
Microbiology 52:1.
Lovley, D.R. & Phillips, E.J.P. 1987. Competitive mechanisms for inhibition of sulfate reduction
and methane production in the zone of ferric iron reduction in sediments. Applied and
Environmental Microbiology, 53:11:2636.
Lula, R.A. 1986. Stainless steel-revised from an introduction to stainless steel by j. G. Parr and a.
Hanson. Metals Park, Ohio: American Society for Metals.
Lydmark, P., Lind, M., Sörensson, F. & Hermansson, M. 2006. Vertical distribution of nitrifying
populations in bacterial biofilms from a full-scale nitrifying trickling filter. Environmental
Microbiology, 8:11:2036.
Lykins, J.B.W., Clark, R.M. & Goodrich, J.A. 1992. Point-of-use/point-of-entry for drinking
water treatment: Lewis Publishers.
Lytle, D.A. & Schock, M.R. 1996. Stagnation time, composition, ph and orthophosphate effects on
metal leaching from brass.
Lytle, D.A. & Schock, M.R. 2005. Formation of pb(iv) oxides in chlorinated water. Journal
American Water Works Association, 97:11:102.
Mac Faddin, J.F. 2000. 3rd. Biochemical tests for identification of medical bacteria. Philadelphia:
Lippincott Williams & Wilkins.
Madigan, M.J., Martinko, J.M. & Parker, J. 2000. Nutrition and metabolism. Brock biology of
microorganisms.
Maixner, F., Noguera, D.R., Anneser, B., Stoecker, K., Wegl, G., Wagner, M. & Daims, H. 2006.
Nitrite concentration influences the population structure of nitrospira-like bacteria.
Environmental Microbiology, 8:8:1487.
Margerum, D.W., Schurter, L.M., Hobson, J. & Moore, E.E. 1994. Water chlorination chemistry nonmetal redox kinetics of chloramine and nitrite ion. Environmental Science &
Technology, 28:2:331.
Marshall, B. & Edwards, M. 2004a. Evaluation of lead corrosion control strategies in maui
water-final report:Virginia Tech.
Marshall, B. & Edwards, M. 2004b. Evaluation of lead corrosion control strategies in maui
water.Final report.
Martin, G. & Richard, Y. 1982. Elements of the inhibition of nitrifying bacteria. Water Science
and Technology, 14:6-7:781.
Martiny, H. & Koops, H.-P. 1982. Incorporation of organic compounds into cell protein by
lithotrophic, ammonia-oxidizing bacteria. Antonie van Leeuwenhoek 48:327.
McFeters, G.A. 1990. Drinking water microbiology-progress and recent development. New York,
NY: Springer-Verlag.
McGuire, M.J., Lieu, N.I. & Pearthree, M.S. 1999. Using chlorite ion to control nitrification.
Journal American Water Works Association, 91:10:52.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
214 | Effect of Nitrification on Corrosion in the Distribution System
McGuire, M.J., Pearthree, M.S., Blute, N.K., Arnold, K.F. & Hoogerwerf, T. 2006. Nitrification
control by chlorite ion at pilot scale. Journal American Water Works Association, 98:1:95.
McIntyre, P. & Mercer, A.D. 1993. Corrosion and related aspects of materials for potable water
supplies. London, UK: Institute of Materials.
McNeill, L.S. & Edwards, M. 2001. Iron pipe corrosion in distribution systems. Journal American
Water Works Association:88.
McNeill, L.S. & Edwards, M. 2002. Phosphate inhibitor use at us utilities. Journal American
Water Works Association, 94:7:57.
McNeill, L.S. & Edwards, M. 2004. Importance of pb and cu particulate species for corrosion
control. Journal of Environmental Engineering-Asce, 130:2:136.
Meiklejohn, J. 1954. Some aspects of the physiology of the nitrifying bacteria Autotrophic
micro-organisms : Fourth symposium of the society for general microbiology held at the
institution of electrical engineers 68.
Meiklejohn, J. 1950. The isolation of nitrosomonas europaea in pure culture. The Journal of
General Microbiology, 4:185.
Meincke, M., Krieg, E. & Bock, E. 1989. Nitrosovibrio spp., the dominant ammonia-oxidizing
bacteria in building sandstone. Applied and Environmental Microbiology, 55:8:2108.
Miettinen, I.T., Vartiainen, T. & Martikainen, P.J. 1997. Phosphorus and bacterial growth in
drinking water. Applied and Environmental Microbiology, 63:8:3242.
Morton, S.C., Glindemann, D. & Edwards, M. 2003. Phosphates, phosphites, and phosphides in
environmental samples. Environmental Science & Technology, 37:6.
Morton, S.C., Zhang, Y. & Edwards, M.A. 2005. Implications of nutrient release from iron metal
for microbial regrowth in water distribution systems. Water Research, 39:13:2883.
Murphy, B., O’Connor, J.T. & O’Connor, T.L. 1997a. Willmar, minnesota battles copper
corrosion-part 2. Public Works, 128:12:44.
Murphy, B., O’Connor, J.T. & O’Connor, T.L. 1997b. Willmar, minnesota battles copper
corrosion part 1. Public Works, :65.
Neden, D.G., Jones, R.J., Smith, J.R., Kirmeyer, G.J. & Foust, G.W. 1992a. Comparing
chlorination and chloramination for controlling bacterial regrowth. Journal American
Water Works Association, 84:7:80.
Neden, D.G., Jones, R.J., Smith, J.R., Kirmeyer, G.J. & Foust, G.W. 1992b. Comparing
chlorination and chloromination for controlling bacterial regrowth. Journal American
Water Works Association, 84:7:80.
Neufeld, R., Greenfield, J. & Rieder, B. 1986. Temperature, cyanide and phenolic nitrification
inhibition. Water Research, 20:5:633.
Neufeld, R.D., Hill, A.J. & Greenfield, J. 1980. Factors influencing biological nitrification of steel
industry wastewaters. . AlChe Symposium Series. , 77:209:151.
Nguyen, C. & Edwards., M. 2005. Chemistry of rapid chloramine decay in water contacting
copper plumbing. AWWA WQTC.
Nguyen, C.K. 2005. Interactions between copper and chlorine disinfectants: Chlorine decay,
chloramine decay and copper pitting. Master. Virginia Tech, Blacksburg, VA.
Norton, C.D. & LeChevallier, M.W. 1997. Chloramination: Its effect on distribution system water
quality. Journal American Water Works Association, 89:7:66.
Nowlin, M., Summers, R.S. & Work, L. 2001. Chlorine and chloramine decay in drinking water:
The
impact
of
temperature.
<http://ceae.colorado.edu/environ/JoAnn/REU/2001/Melanie/>, [Dec, 30th].
©2010 Water Research Foundation. ALL RIGHTS RESERVED
References | 215
Odell, L.H., Kirmeyer, G.J., Wilczak, A., Jacangelo, J.G., Marcinko, J.P. & Wolfe, R.L. 1996.
Controlling nitrification in chloraminated systems. Journal American Water Works
Association, 88:7:86.
Ohashi, A., deSilva, D.G.V., Mobarry, B., Manem, J.A., Stahl, D.A. & Rittmann, B.E. 1995.
Influence of substrate c/n ratio on the structure of multi-species biofilms consisting of
nitrifiers and heterotrophs. Water Science and Technology, 32:8:75.
Oldenburg, P.S., Regan, J.M., Harrington, G.W. & Noguera, D.R. 2002. Kinetics of nitrosomonas
europaea inactivation by chloramine. Journal American Water Works Association,
94:10:100.
Painter, H.A. 1977. Microbial transformations of inorganic nitrogen. Progress in Water
Technology, 8:4/5.
Pan, P. & Umbreit, W.W. 1972a. Growth of mixed cultures of autotrophic and heterotrophic
organisms. Canadian Journal of Microbiology 18:153.
Pan, P. & Umbreit, W.W. 1972b. Growth of obligate autotrophic bacteria on glucose in a
continuous flow-through apparatus. Journal of Bacteriology, 109:1149.
Park, H.D., Wells, G.F., Bae, H., Criddle, C.S. & Francis, C.A. 2006. Occurrence of
ammonia-oxidizing archaea in wastewater treatment plant bioreactors. Applied and
Environmental Microbiology, 72:5643.
Parker, D.S. 1975. Process Design Manual For Nitrogen Control.
Parks, J., Edwards, M., Vikesland, P., Fiss, M. & Dudi, A. 2008. Autogenous healing of concrete
in the drinking water industry. Denver, CO:AWWA Research Foundation.
Parks, J.L., McNeill, L., Frey, M., Eaton, A.D., Haghani, A., Ramirez, L. & Edwards, M. 2004.
Determination of total chromium in environmental water samples. Water Research,
38:12:2827.
Passantino, L.e.a. 2003. Pipe-loop studies evaluating nitrification and disinfection by-product
formation following the conversion to chloramines. American Water Works Association
Annual Conference.
Peavy, H.S., Rowe, D.R. & Tchobanoglous., G. 1985. Environmental engineering. New York,
NY: McGraw-Hill, Inc.
Pettet, A.E.J. 1956. Effects in metal finishing wastes on sewage purification. . Journal and
proceedings / The Institute of Sewage Purification.:36.
Pettet, A.E.J. 1956. Effects in metal finishing wastes on sewage purification. . Journal and
proceedings / The Institute of Sewage Purification.:36.
Powell, R. 2004. Implementation of chloramination by a florida utility: The good, the bad, and the
ugly. AWWA WQTC conferences.
Prakasam, T.B.S. & Loehr, R.C. 1972. Microbial nitrification and denitrification in concentrated
wastes. Water Research, 6:7:859.
Prosser, J.I. 1986. Nitrification. Oxford Washington DC: Published for the Society for General
Microbiology by IRL Press.
Pugh, E.N., Montague, W.G. & Westwood., A.R.C. 1966. Stress corrosion cracking of copper.
Corrosion Science, 6:345.
Purkhold, U., Wagner, M., Timmermann, G., Pommerening-Röser, A. & Koops, H.-P. 2003. 16s
rrna and amoa-based phylogeny of 12 novel betaproteobacterial ammonina-oxidizing
isolates: Extension of the dataset and proposal of a new lineage within the nitrosomonads.
International Journal of Systematic and Evolutionary Microbiology, 53:5:1485.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
216 | Effect of Nitrification on Corrosion in the Distribution System
Qin, Y.-Y., Li, D.-T. & Yang, H. 2007. Investigation of total bacterial andammonia-oxidizing
bacterial community composition in a full-scale aerated submerged biofilm reactor for
drinkingwater pretreatment in china. FEMS Microbiol. Lett., 268:126.
Rahman, M.S. 2008. Nitrification in premise plumbing systems. Montana State University,
Bozeman, MT.
Reed, B.E., Arunachalam, S. & Thomas, B. 1993. Removal of lead and cadmium from aqueous
waste streams using granular activated carbon (gac) columns. Water Environment
Federation 66th Annual Conference and Exposition.
Reeves, M.W., Pine, L., Hutner, S.H., George, J.R. & Harrell, W.K. 1981. Metal requirements of
legionella pneumophila. Journal of Clinical Microbiology, 13:4:695.
Regan, J. 2001. Microbial ecology of nitrification in chloraminated drinking water distribution
systems. Ph.D. University of Wisconsin-Madison.
Regan, J.M., Cho, A.-Y., Kim, S. & Smith, C.D. 2007. Monitoring ammonia-oxidizing bacteria in
chloraminated distribution systems. Denver, CO:AWWA Research Foundation and
AWWA.
Regan, J.M., Harrington, G.W., Baribeau, H., De Leon, R. & Noguera, D.R. 2003. Diversity of
nitrifying bacteria in full-scale chloraminated distribution systems. Water Research,
37:1:197.
Regan, J.M., Harrington, G.W. & Noguera, D.R. 2002. Ammonia- and nitrite-oxidizing bacterial
communities in a pilot-scale chloraminated drinking water distribution system. Applied
and Environmental Microbiology, 68:1:73.
Rehring, J.P. & Edwards, M. 1996. Copper corrosion in potable water systems: Impacts of natural
organic matter and water treatment processes. Corrosion, 52:4:307.
Reiber, S. 1991. Galvanic stimulation of corrosion on lead-tin solder sweated joints. Journal
American Water Works Association, 83:7:83.
Reiber, S. & Edwards, M. 1997. Utility experience and the lead and copper rule. Drinking Water
Research, 7:3:2.
Rennie, R.J. & Schmidt, E.L. 1977. Immunofluorescence Studies of Nitorbacter Populations in
Soils Canadian Journal of Micorbiology 23:1011-1017.
Richardson, D.J. & Watmough, N.J. 1999. Inorganic nitrogen metabolism in bacteria. Chemical
Biology 3:207.
Rittmann, B.E. & Manem, J.A. 1992. Development and experimental evaluation of a steady-state,
multispecies biofilm model. Biotechnology and Bioengineering, 39:9:914.
Rittmann, B.E., Regan, J.M. & Stahl, D.A. 1994. Nitrification as a source of soluble organic
substrate in biological treatment. Water Science and Technology, 30:6:1.
Rittmann, B.E. & Soneyink, V.L. 1984. Achieving biologically stable dinking water. Journal
American Water Works Association 106.
Rollinger, Y. & Dott, W. 1987. Survival of selected bacterial species in sterilized activated carbon
filters and biological activated carbon filters. Applied and Environmental Microbiology,
53:4:777.
Rosswall, T. 1982. Microbiological regulation of the biogeochemical nitrogen cycle. Plant and
Soil, 67:15.
Rozenfeld, I.L. 1981. Corrosion inhibitors:160.
Sakairi, M.A.C., Yasuda, K. & Matsummura, M. 1996. Nitrogen removal in sea water using
nitrifying and denitrifying bacteria immobilized in porous cellulose carrier. Water Science
& Technology, 34:7-8:267.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
References | 217
Sang, J.Q., Zhang, X.H., Zhou, H.H. & Wang, Z.S. 2003. Limiting phosphorus to control
bio-stability in drinking water Advances in Water Science (China), 14:6:720.
Sathasivan, A., Fisher, I. & Kastl, G. 2005. Simple method for quantifying microbiologically
assisted chloramine decay in drinking water. Environmental Science & Technology,
39:14:5407.
Sathasivan, A. & Ohgaki, S. 1999. Application of new bacterial regrowth potential method for
water distribution system- a clear evidence of phosphorus limitation. Water Research,
33:1:137.
Sato, C., Leung, S.W. & Schnoor, J.L. 1988. Toxic response of nitrosomonas europaea to copper
in inorganic medium and wastewater. Water Research, 22:9:1117.
Sato, C., Schnoor, J.L. & McDonald, D.B. 1986. Characterization of effects of copper, cadmium
and nickel on the growth of nitrosomonas europaea. . Environmental Toxicology and
Chemistry, 5:403.
Sawyer, C.N. & Bradney, L. 1946. Modernization of the bod test for determining the efficiency of
sewage treatment processes. Science of Sewage Works, 18:1113.
Schock, M.R. 1989. Understanding corrosion control strategies for lead. Journal American Water
Works Association, 81:7.
Schock, M.R., Harmon, S.M., Swertfeger, J. & Lohmann, R. 2001. Tetravalent lead: A hitherto
unrecognized control of tap water lead contamination. AWWA Water Quality Technology
Conference.
Schock, M.R., Lytle, D.A. & Clement, J.A. 1995. Effect of ph, dic, orthophosphate, and sulfate on
drinking water cuprosolvency.
Scholz, M. & Xu, J. 2002. Comparison of constructed reed beds with different filter media and
macrophytes treating urban stream water contaminated with lead and copper. Ecological
Engineering, 18:3:385.
Seidel, C.J., Mcguire, M.J., Summers, S. & Via, S. 2005. Have utilities switched to chloramines?
Journal American Water Works Association, 97:10:87.
Shammas, N.K. 1971. Optimization of biological nitrification. Ph.D. University of Michigan.
Sharma, B. & Ahlert, R.C. 1977. Nitrification and nitrogen removal. Water Research, 11:897.
Sherrard, J.H., et al. 1980. Influence of the COD:NH4+-N Ratio on Nitrification and Alkalinity in
the Activated Sludge Process Journal of Civil Engineering Design 2:1:31.
Shkelqim, C. & Malcolm, E.S. 2002. Critical concentrations of copper, nickel, lead, and cadmium
in soils based on nitrification. Communications in Soil Science and Plant Analysis,
V33:1:19.
Silhan, J., Corfitzen, C.B. & Albrechtsen, H.J. 2006. Effect of temperature and pipe material on
biofilm formation and survival of escherichia coli in used drinking water pipes: A
laboratory-based study. Water Science and Technology, 54:3:49.
Skadsen, J. 1993. Nitrification in a distribution system. Journal American Water Works
Association:7:95.
Skadsen, J. 2002. Effectiveness of high ph in controlling nitrification. Journal American Water
Works Association, 94:7:73.
Skinner, F.A. & Walker, N. 1961. Growth of nitrosomonas europaea in batch and continuous
culture. Archiv. Fur. Mikrobiologie, 38.
Snoeyink, V.L. & Jenkins, D. 1980. Water chemistry. New York: John Wiley & Sons.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
218 | Effect of Nitrification on Corrosion in the Distribution System
Snoeyink, V.S., Hass, C.H., Boulos, P.F., Burlinghame, G.A., Camper, A.K., Clark, R.N.,
Edwards, M.A., LeChevalier, M.W., McMullen, L.D., Moe, C.L., Nieminski, E.C., Smith,
C.D., Spath, D.P. & Valentine, R.L. 2006. Drinking water distribution systems: Assessing
and reducing risks. . Washington, DC :: National Academies Press.
Sobsey, M.D., Hanzel, T. & Venczel., L. 2003. Chlorination and safe storage of household
drinking water in developing countries to reduce waterborne disease. Water Science &
Technology 47:3:221.
Song, D.J., Sheikholeslami, A., Hoover, L.L., Turner, K.A., Lai, H.H. & Wilczak, A. 1999.
Improvement of chloramine stability through ph control, toc reduction and blending at
ebmud, california. AWWA Annual Conference.
Srinivasan, S., Harrington, G.W., Xagoraraki, I. & Goel, R. 2008. Factors affecting bulk to total
bacteria ratio in drinking water distribution systems. Water Research, 42:13:3393.
Stein, L.Y. & Arp, D. 1998. Ammonium limitation results in the loss of ammonia-oxidizing
activity in nitrosomonas europaea. Applied and Environmental Microbiology, 64:4:1514.
Stenstrom, M.K. & Song, S.S. 1991. Effects of oxygen transport limitation on nitrification in the
activated sludge process. Research Journal WPCF 63:3:208.
Steward, M.H. & Lieu, N.I. 1997. Nitrification in chloraminated drinking water and its association
with biofilms. AWWA WQTC.
Strickhouser, A.E. 2007. Legionella pneumophila in domestic hot water systems: Evaluation of
detection methods and environmental factors affecting survival. Virginia Tech,
Blacksburg, VA.
Sundberg, R., Holm, R., Hertzman, S., Hutchinson, B. & Lindh-Ulmgren., E. 2003.
Dezincification (da) and intergranular corrosion (iga) of brass-influence of composition
and heat treatment. Metall., 57:721.
Suzuki, I., Kwok, S.C. & Dular, U. 1974. Ammonia or ammonium ion as substrate for oxidation
by nitrosomonas europaea cells and extracts. Journal of Bacteriology, 120:1.
Szwerinski, H., Gaiser, S. & Bardtke, D. 1985. Immunofluorescence for the quantitative
determination of nitrifying bacteria: Interference of the test in biofilm reactors. Applied
Microbiology and Biotechnology 21:125.
Tang, N.H. 1992. Osar parameters for toxicity of organic chemicals to nitrobacter. Journal of
Environmental Engineering-ASCE, 118:1.
Tarre, S. & Green, M. 2004. High rate nitrification at low ph in suspended- and attached-biomass
reactors. Applied and Environmental Microbiology, 70:11:6481.
Thomas, P.M. 1987. Formation and decay of monochloramine in south australia water supply
systems. 12th Federal Convention. Australia Water and Wastewater Association.
Tokuno, S.-I. 1997. Granular activated carbon filtration and nitrification. Water Engineering &
Management, 144:2:22.
Tokutomi, T., Kiyokawa, T., Shibayama, C., Harada, H. & Ohashi, A. 2006. Effect of inorganic
carbon on nitrite accumulation in an aerobic granule reactor. Water Science & Technology,
53:12:285.
Tomas, P.M. 1987. Formation and decay of monochloramine in south australia water supply
systems. 12th Federal Convention. Australia Water and Wastewater Association.
Tomlinson, T.G., Boon, A.G. & Trotman, N.A. 1966. Inhibition of nitrification in the activated
sludge process of sewage disposal. The Journal of Applied Bacteriology, 29:2.
Triantafyllidou, S., Parks, J. & Edwards, M. 2007a. Lead particles in potable water. Journal
American Water Works Association, 99:6:107.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
References | 219
Triantafyllidou, S., Parks, J. & Edwards., M. 2007b. Lead particles in potable water. Journal
American Water Works Association, 96:6:107.
Uchida, M. & Okuwaki, A. 1998. Decomposition of nitrate by in situ buff abrasion of lead plate.
Hydrometallurgy, 49:297.
USEPA.
1991.
Lead
and
copper
rule:
A
quick
reference
guide.
<http://www.epa.gov/safewater/lcrmr/pdfs/qrg_lcmr_2004.pdf>, [Nov, 20th].
USEPA. 2001a. National primary drinking water regulations, 14:141.
USEPA.
2001b.
Total
coliform
rule.
<http://www.epa.gov/safewater/disinfection/tcr/regulation.html>, [July, 7th].
USEPA. 2003a. Analysis and findings of the gallup organization’s drinking water customer
satisfaction
survey.
<http://www.epa.gov/safewater/ccr/pdfs/tools_survey_gallup_customersatification2003.p
df >, [7/25].
USEPA. 2003b. Consideration of other regulatory revisions for chemical contaminants in support
of the six-year review of the national primary drinking water regulations, EPA
815-R-03-005.
USEPA. 2005. Nitrification. <http://www.epa.gov/safewater/tcr/pdf/nitrification.pdf >, [12/21].
USEPA. 2006. Epa review of dc wasa lead data. <http://www.epa.gov/dclead/>, [4/25].
Vahala, R. 2002. Two step granular activated carbon filtration in drinking water treatment. Ph.D. .
Helsinki University of Technology
Vahala, R. & Laukkanen, R. 1998. Nitrification in gac-filtered, uv-disinfected and chlorinated
distribution system. Water Supply, 16:3/4:245.
Valentine, R.L. 1984. Disappearance of monochloramine in the presence of nitrite. Water
chlorination chemistry, environmental impact and health effects, 5.
Van der Aa, L.T.J., Kors, L.J., Wind, A.P.M., Hofman, J.A.M.H. & Rietveld, L.C. 2002.
Nitrification in rapid sand filter: Phosphate limitation at low temperatures. Water Science
and Technology: Water Supply, 2:1:37.
Van Droogenbroeck, R. & Laudelout, H. 1967. Phosphate requirement of the nitrifying bacteria.
Antonie van Leeuwenhoek, 33:287.
Vasconcelos, J.J., Rossman, L.A., Grayman, W.M., Boulos, P.F. & Clark, R.M. 1997. Kinetics of
chlorine decay. Journal American Water Works Association, 89:7:54.
Vasquez, F.A., Heaviside, R., Tang, Z. & Taylor, J.S. 2006. Effects of free chlorine and
chloramines on lead release in a distribution system. Journal American Water Works
Association, 98:2:144.
Verhagen, F.J.M. & Laanbroek, H.J. 1991. Competition for ammonium between nitrifying and
heterotrophic bacteria in dual energy-limited chemostats. Applied and Environmental
Microbiology, 57:11:3255.
Verstraete, W. & Alexander, M. 1973. Heterotrophic nitrification in samples of natural
ecosystems. Environmental Science & Technology, 7:1.
Verstraete, W. & Alexander, M. 1986. Heterotrophic nitrification. Nitrification.
Vikesland, P.J., Love, N., Chandran, K., Fiss, E.M., Rebodos, R., Zaklikowski, A.E., DiGiano,
F.A. & Ferguson, B.A. 2006. Seaonal chlorination practices and impacts to
chloraminating utilities:AWWARF.
Vikesland, P.J., Ozekin, K. & Valentine, R.L. 2001. Monochloramine decay in model and
distribution system waters. Water Research, 35:7:1766.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
220 | Effect of Nitrification on Corrosion in the Distribution System
Vikesland, P.J. & Valentine, R.L. 2000. Reaction pathways involved in the reduction of
monochloramine by ferrous iron. Environmental Science & Technology, 34:1:83.
Vikesland, P.J., Ozekin, K. & Valentine, R.L. 1998. Effect of natural organic matter on
monochloramine decomposition: Pathway elucidation through the use of mass and redox
balances. Environmental Science & Technology, 32:10:1409.
Volk, C.A., Volk, C.B. & Kaplan, L.A. 1997. Chemical Composition of Biodegradable Dissolved
Organic Matter in Streamwater. Limnol. Oceanogr. , 42:39-44.
Volsch, A., Nader, W.F., Geiss, H.K., Nebe, G. & Birr, C. 1990. Detection and analysis of two
serotypes of ammonia-oxidizing bacteria in sewage plants by flow cytometry. Applied and
Environmental Microbiology, 56:8:2430.
Waara, K.O. & Wilander, A. 1985. Effects of heavy metals on nitrification studied with chemostat
technique. Verh. Internat. Verein. Limnol., 22.
Wagner, I. 1993. Influence of operating conditions on materials and water quality in dirnking
water distribution systems. Corrosion and Related Aspects of materials for portable water
supplies.:161.
Wagner, M., Noguera, D.R., Juretschko, S., Rath, G., Koops, H.P. & Schleifer, K.H. 1998.
Combining fluorescent in situ hybridization (fish) with cultivation and mathematical
modeling to study population structure and function of ammonia-oxidizing bacteria in
activated sludge. Water Science and Technology, 37:4-5:441.
Wagner, M., Rath, G., Amann, R., Koops, H.P. & Schleifer, K.H. 1995. In-situ identification of
ammonia-oxidizing bacteria. Systematic and Applied Microbiology, 18:2:251.
Wallace, W.S., Knowles, E. & Nicholas., D.J.D. 1970. Intermediary metabolism of carbon
compounds by nitrifying bacteria. Archives of Microbiology, 70:26.
Watson, S.W., Bock, E., Harms, H., Koops, H.P. & Hooper, A.B. 1989. Nitrifying bacteria.
Batimore, M.D.: Williams and Wilkins
Watson, S.W., Bock, E., Valois, F.W., Waterbury, J.B. & Schlosser, U. 1986. Nitrospira marine
gen. Nov. Sp. Nov.: A chemolithotrophic nitrite-oxidizing bacterium. Archives of
Microbiology, 144:1.
Webster, B.J., Werner, S.E., Wells, D.B. & Bremer, P.J. 2000. Microbiologically influenced
corrosion of copper in potable water systems—ph effects. Corrosion, 56:9:942.
Wei, X., Vajrala, N., Hauser, L., Sayavedra-Soto, L.A. & Arp, D.J. 2006. Iron nutrition and
physiological responses to iron stress in nitrosomonas europaea. Archives of Microbiology
186:2:107.
Westerhoff, P. & James, J. 2003. Nitrate removal in zero-valent iron packed columns. Water
Research, 37:1818.
Wilczak, A., Jacangelo, J.G., Marcinko, J.P., Odell, L.H., Kirmeyer, G.J. & Wolfe, R.L. 1996.
Occurrence of nitrification in chloraminated distribution systems. Journal American Water
Works Association, 88:7:74.
Wintzingerode, F.v., Goëbel, U.B. & Stackebrandt, E. 1997. Determination of microbial diversity
in environmental samples: Pitfalls of pcr-based rrna analysis. FEMS Microbiology Reviews
21:213.
Wolfe, R.L. & Lieu, N.I. 2001. Nitrifying bacteria in drinking water. . Encyclopedia of
Environmental Microbiology:2167.
Wolfe, R.L., Lieu, N.I., Izaguirre, G. & Means, E.G. 1990. Ammonia-oxidizing bacteria in a
chloraminated distribution-system - seasonal occurrence, distribution, and disinfection
resistance. Applied and Environmental Microbiology, 56:2:451.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
References | 221
Wolfe, R.L., Means, E.G., Davis, M.K. & Barrett, S.E. 1988. Biological nitrification in covered
reservoirs containing chloraminated water. Journal American Water Works Association,
80:9:109.
Wolfe, R.L., Ward, N.R. & Olson, B.H. 1984. Inorganic chloramines as drinking-water
disinfectants - a review. Journal American Water Works Association, 76:5:74.
Wong-Chong, G.M. & Loehr, R.D. 1975. The kinetics of microbial nitrification. Water Research,
9:12:1099.
Woolschlager, J., Rittmann, B., Piriou, P., Klene, L. & Schwartz, B. 2001. Using a comprehensive
model to identify the major mechanisms of a chloramines decay in distribution systems.
Water Science and Technology: Water Supply, 1:4.
Woolschlager, J. & Soucie, W. 2003. Chlorine reactivity with concrete pipe surfaces. World Water
and Environmental Resources.
Wqa. 2006. Heterotrophic bacteria in dinking water from pou & poe devices.
<http://www.wqa.org/pdf/technical/HPC200[5].pdf>, [October].
Yang, J., Harrington, G.W. & Noguera, D.R. 2004. Nitrification modeling in pilot-scale
chloraminated dinking water distribution systems. AWWA WQTC
Yilmaz, L.S. & Noguera, D.R. 2004. Mechanistic approach to the problem of hybridization
efficiency in fluorescent in situ hybridization. Applied and Environmental Microbiology,
70:12:7126.
Zahn, J.A., Arcerio, D.M., Hooper, A.B. & A., D.A. 1996. Evidence for an iron center in the
ammonia monooxygenase from nitrosomones europaee. . FEBS Letters, 397:1:35.
Zavarzin, G.A. 1958. On the inducer of the second phase of nitrification-ii: The effect of heavy
metals on nitrification. Microbiology, 27.
Zhang, Y. 2005. Relative effects of water chemistry on aspects of iron corrosion. Master. Virginia
Tech.
Zhang, Y. & Edwards, M. 2005. Water chemistry and growth of nitrifying bacteria in water
distribution systems. AWWA WQTC.
Zhang, Y. & Edwards, M. 2007. Anticipating effects of water quality changes on iron corrosion
and red water. Journal of Water Supply: Research & Technology- AQUA, 56:1:55.
Zhang, Y., Griffin, A. & Edwards, M. 2008a. Nitrification in premise plumbing: Role of
phosphate, ph and pipe corrosion Environmental Science & Technology, 42:12:4280.
Zhang, Y., Griffin, A., Rahman, M., Camper, A. & Edwards, M. 2009a. Lead contamination of
potable water due to nitrification. Environmental Science & Technology, 43:6:1890.
Zhang, Y., Love, N. & Edwards, M. 2009b. Nitrification in drinking water systems. Critical
Reviews in Environmental Science & Technology, 39:3:153.
Zhang, Y., Triantafyllidou, S. & Edwards, M. 2008b. Effect of nitrification and gac filtration on
copper and lead leaching in home plumbing systems ASCE Journal of Environmental
Engineering, 134:7:521.
Zoetendal, E. G., K. Ben-Amor, A. D. L. Akkermans, T. Abee, and W. M. de Vos. 2001. DNA
Isolation Protocols Affect the Detection Limit of PCRApproaches of Bacteria in Samples
from the Human Gastrointestinal Tract. Systematic and Applied Microbiology 24:405-410.
©2010 Water Research Foundation. ALL RIGHTS RESERVED
222 | Effect of Nitrification on Corrosion in the Distribution System
©2010 Water Research Foundation. ALL RIGHTS RESERVED
ABBREVIATIONS
ACS
AMON
AOB
AOC
AWWA
AwwaRF
CDA
COD
CuE
DIC
D/DBPR
DGGE
DO
DS
FA
FISH
GAC
GW
HDPE
HPC
ICP-MS
IRWD
LCR
MCL
MPN
MRDL
NOB
NOM
PAC
PACl
PCR
PE
PEX
POE
PVC
RO
SMP
TCR
TDS
THM
T-RFLP
TOC
USEPA
American Chemical Society
Ammonia Monooxygenase
Ammonia Oxidizing Bacteria
Available Organic Carbon
American Water Works Association
Awwa Research Foundation
Copper Development Association
Chemical Oxygen Demand
Epoxy coated Copper
Dissolved Inorganic Carbon
Disinfectants and Disinfection By-Products Rule
Denaturing Gradient Gel Elecrophoresis
Dissolved Oxygen
Distribution System
Fluorescent-Antibody
Fluorescent In Situ Hybridization
Granular Activated Carbon
Groundwater
High-Density Polyethylene
Heterotrophic Plate Counts
Inductively Coupled Plasma Mass Spectrophotometer
Irvine Ranch Water District
Lead and Copper Rule
Maximum Contaminant Level
Most Probable Number
Maximum Residual Disinfectant Level
Nitrite Oxidizing Bacteria
Natural Organic Matter
Project Advisory Committee
Polyaluminum Chloride
Polymerase Chain Reaction
Polyethylene
Cross Linked Polyethylene
Point of Entry
Polyvinylchloride
Reverse Osmosis
Soluble Microbial Products
Total Coliform Rule
Total Dissolved Solids
Trihalomethanes
Terminal Restriction Fragment Length Polymorphism
Total Organic Carbon
United States Environmental Protection Agency
223
©2010 Water Research Foundation. ALL RIGHTS RESERVED
224
©2010 Water Research Foundation. ALL RIGHTS RESERVED
Effect of Nitrification on Corrosion in the Distribution System
6666 West Quincy Avenue, Denver, CO 80235-3098 USA
P 303.347.6100 • F 303.734.0196 • www.WaterResearchFoundation.org
4015
1P-3C-4015-01/10-FP