oEPA
United States
Environmental Protection
Agency
Six-Year Review 3 Technical Support
Document for Disinfectants/Disinfection
Byproducts Rules

-------
Office of Water (4607M)
EP A-810-R-16-012
December 2016

-------
Disclaimer
This document is not a regulation. It is not legally enforceable and does not confer legal rights or
impose legal obligations on any party, including EPA, states or the regulated community. While
EPA has made every effort to ensure the accuracy of any references to statutory or regulatory
requirements, the obligations of the interested stakeholders are determined by statutes,
regulations or other legally binding requirements, not this document. In the event of a conflict
between the information in this document and any statute or regulation, this document would not
be controlling.

-------
This page intentionally left blank.

-------
Table of Contents
1	Introduction	1-1
2	EPA's Protocol for the Six-Year 3 Review	2-1
3	History of Disinfectants and Disinfection Byproducts Regulations	3-1
3.1	Interim Total Trihalomethanes Regulation	3-1
3.2	Stage 1D/DBPR	3-2
3.2.1	Negotiated Rulemaking	3-2
3.2.2	Proposed Stage 1 D/DBPR	3-4
3.2.3	DBP Information Collection Rule	3-4
3.2.4	The 1996 Safe Drinking Water Act Amendments, Microbial and
Disinfectants/Disinfection Byproducts (MDBP) Advisory Committee and
Notices of Data Availability	3-5
3.2.5	Final Stage 1 D/DBPR	3-5
3.3	Stage 2 D/DBPR	3-7
3.3.1	MDBP Advisory Committee and New Information	3-8
3.3.2	Proposed Stage 2 D/DBPR	3-9
3.3.3	Final Stage 2 D/DBPR	3-10
4	Health Effects	4-1
4.1	Regulated Organic DBPs	4-1
4.1.1	Toxicity Studies	4-1
4.1.2	Epidemiology and Weight of Evidence	4-24
4.1.3	Mixtures of Chlorination Organic DBPs	4-49
4.2	Regulated Inorganic DBPs	4-51
4.2.1	Bromate	4-52
4.2.2	Chlorite	4-55
4.3	Regulated Disinfectants	4-57
4.3.1	Chlorine	4-57
4.3.2	Chloramines	4-57
4.3.3	Chlorine Dioxide	4-58
4.4	Unregulated DBPs	4-59
4.4.1	Chlorate	4-59
4.4.2	Nitrosamines	4-59
4.4.3	Haloacetic Acids	4-59
4.4.4	Iodinated THMs	4-63
Six-Year Review 3 Technical Support Document	i
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
4.4.5	Haloketones	4-63
4.4.6	Haloacetaldehydes	4-64
4.4.7	Halonitromethanes	4-65
4.4.8	Haloacetonitriles	4-66
4.4.9	Haloacetamides	4-70
4.4.10	Cyanogen halides	4-71
4.4.11	Halogenated furanones	4-72
4.4.12	Halogenated benzoquinones (HBQs) and haloquinones (HQ)	4-73
4.4.13	Halogenated pyrroles	4-73
4.4.14	Aldehydes	4-74
4.5 Unregulated DBPs Data Availability Summary	4-76
5	Analytical Methods	5-1
5.1	Methods for Treatment Technique Requirement for Removal of DBP Precursors	5-6
5.1.1	Alkalinity	5-6
5.1.2	Bromide	5-7
5.1.3	Total Organic Carbon (TOC) and Dissolved Organic Carbon (DOC)	5-7
5.1.4	UV254 and Specific Ultraviolet Light Absorbance (SUVA)	5-8
5.2	Methods for Disinfection Byproducts	5-8
5.2.1	I IIY1	5-8
5.2.2	HAA5	5-10
5.2.3	Chlorite	5-12
5.2.4	Bromate	5-13
5.2.5	Unregulated DBPs	5-14
5.3	Methods for Disinfectant Residuals	5-18
5.3.1	Chlorine (Free, Combined, Total) and Chloramines	5-18
5.3.2	Chlorine Dioxide	5-18
6	Occurrence and Exposure	6-1
6.1	DBP Formation	6-2
6.1.1	Summary of Stage 1 and 2 D/DBPR Information	6-2
6.1.2	New Information since the Stage 2 D/DBPR	6-3
6.2	Occurrence of DBP Precursors	6-19
6.2.1	Organic Precursors	6-19
6.2.2	Inorganic Precursors	6-28
6.3	DBP Occurrence and Exposure	6-31
Six-Year Review 3 Technical Support Document	ii
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
6.3.1	Overview of DBP Inventory Analyses	6-32
6.3.2	Regulated Organic DBPs	6-34
6.3.3	Regulated Inorganic DBPs	6-44
6.3.4	Additional Considerations	6-50
7	Treatment	7-1
7.1	Introducti on	7-1
7.2	Background on Treatment Technologies Considered for the Stage 1 and Stage 2
D/DBPRs	7-2
7.2.1	Treatment Technique Requirements for TOC Removal	7-2
7.2.2	Treatment Technologies Considered During Rule Development	7-3
7.3	Information on Reducing DBP Formation Potentials in Treatment Plants	7-4
7.3.1	Analysis of SYR3 ICR Data for TOC Removal	7-6
7.3.2	Information on Conventional Treatment	7-11
7.3.3	Information on Non-Conventional Treatment	7-15
7.3.4	Information on Potential Add-on Physical Removal Unit Processes	7-16
7.4	Information on Source Water Management	7-22
7.5	Information on Changing Disinfection Practices in Treatment Plants and
Distribution Systems	7-24
7.6	Information on Removing DBPs after Formation in Treatment Plants and/or
Distribution Systems	7-26
8	Consideration of Other Regulatory Revisions for MDBP Rules	8-1
8.1	Stage 2 D/DBPR Consecutive System Monitoring	8-1
8.2	Stage 2 D/DBPR Compliance Monitoring - Chlorine Burn	8-2
9	References	9-1
Six-Year Review 3 Technical Support Document	iii
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Appendices
Appendix A: Additional Information for Health Effects of Regulated Organic Disinfection
Byproducts (DBPs), Regulated Inorganic DBPs and Regulated Disinfectants
(Appendix to Chapter 4)
Appendix B: Additional Information for Occurrence and Exposure to Regulated and Unregulated
Disinfection Byproducts (DBPs) (Appendix to Chapter 6)
Appendix C: Supporting Information for Treatment (Appendix to Chapter 7)
Appendix D: Consideration of Other Regulatory Revisions for MDBP Rules - Additional Issues
(Appendix to Chapter 8)
Appendix E: Additional Information Related to Chlorine Burn Analysis
Six-Year Review 3 Technical Support Document	iv
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
List of Exhibits
Exhibit 2.1: SYR Protocol Overview and Major Categories of Revise/Take No Action
Outcomes	2-2
Exhibit 3.1: Timeline for Selected Regulatory Activities Associated with Disinfectants and
Disinfection Byproducts in Drinking Water	3-1
Exhibit 3.2: Stage 1 and Stage 2 MCLs and MCLGs	3-6
Exhibit 3.3: Stage 1 and Stage 2 MRDLs and MRDLGs	3-7
Exhibit 3.4: Stage 1 D/DBPR Required Removal of Total Organic Carbon by Enhanced
Coagulation and Enhanced Softening for Subpart H Systems Using Conventional
Treatment1,2,3	3-7
Exhibit 4.1: Summary of Results from Pre-Stage 2 and Post-Stage 2 Epidemiology and Animal
Toxicity Reproductive/Developmental Studies	4-40
Exhibit 4.2: Available Quantitative Assessments for Unregulated DBPs Discussed in this
Document	4-76
Exhibit 5.1: Analytical Methods Approved in the Stage 1 and Stage 2 D/DBPRs and via the
Expedited Method Approval Process1	5-1
Exhibit 5.2: Analytical Methods for Unregulated DBPs Approved via the Expedited Method
Approval Process or Other EPA Rulemaking	5-4
Exhibit 5.3: Method Performance Metrics for EPA Methods 502.2, 524.2, 524.3, 524.4 and 551.1
- TIIYls	5-9
Exhibit 5.4: Method Performance Metrics for EPA Methods 552.1, 552.2, 552.3 and 557 and
for SM 6251 B-HAA5	5-10
Exhibit 5.5: Method Performance Metrics for EPA Methods 300.0 (Rev. 2.1), 300.1, 317.0
(Rev. 2.0), 326.0 and 327.0 (Rev. 1.1) — Chlorite	5-13
Exhibit 5.6: Method Performance Metrics for EPA Methods 300.1, 317.0 (Rev. 2.0), 321.8,
326.0,	302.0 and 557 — Bromate	5-14
Exhibit 5.7: Method Performance Metrics for EPA Methods 552.1, 552.2, 552.3 and 557-
Unregulated Brominated HAAs	5-15
Exhibit 5.8: Method Performance Metrics for Six Nitrosamines in EPA Method 521	5-16
Exhibit 5.9: Method Performance Metrics for Chlorate Using EPA Methods 300.0 (Rev. 2.1),
300.1,	317.0 (Rev. 2.0), 326.0, ASTMD6581-08 and SM4110D	5-17
Six-Year Review 3 Technical Support Document	v
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Exhibit 6.1: Regulated and Unregulated DBPs - General Information	6-3
Exhibit 6.2: Use of Disinfectants by Source Water Type and System Size for UCMR 3 Data in
EPs (select categories)	6-10
Exhibit 6.3: Use of Disinfectants by Source Water Type and System Size for UCMR 3 Data in
MRs (select categories)	6-11
Exhibit 6.4: DBP ICR and UCMR 3 Comparison — Use of Disinfectants (select categories).. 6-12
Exhibit 6.5: Singer Group Models for Estimating Unreported HAAs as a Function of Reported
11A As and TIIYls	6-18
Exhibit 6.6: Number of Systems and Population Served by Systems in the SYR3 ICR Dataset
with TOC Records, by System Type (2011)	6-20
Exhibit 6.7: Number of Systems and Population Served by Systems in the SYR3 ICR Dataset
with TOC Records, by Source Water Type (2011)	6-20
Exhibit 6.8: Number of Systems and Population Served by Systems in the SYR3 ICR Dataset
with TOC Records, by System Size and System Type (2011)	6-21
Exhibit 6.9: Raw and Finished Water Plant Means from the SYR3 ICR TOC Dataset; Surface
Water Systems (2011)	6-22
Exhibit 6.10: Cumulative Distribution of Raw Water and Finished Water Means in SYR3 ICR
TOC Dataset; Surface Water Plants (2011)	6-23
Exhibit 6.11: Raw and Finished Water Plant Mean TOC from the DBP ICR (1998) and SYR3
ICR (2011); Common Surface Water Systems	6-24
Exhibit 6.12: Raw Water Plant Mean TOC Data from Surface Water Plants in the DBP ICR
(1998, Systems Serving > 100,000 People)1	6-25
Exhibit 6.13: Finished Water Plant Mean TOC Data from Surface Water Plants in the DBP ICR
(1998, Systems Serving > 100,000 People)1	6-26
Exhibit 6.14: Raw Water Plant Mean TOC Data from GW Plants in the DBP ICR (1998,
Systems Serving > 100,000 People)1	6-26
Exhibit 6.15: Finished Water Plant Mean TOC Data from GW Plants in the DBP ICR (1998,
Systems Serving > 100,000 People)1	6-27
Exhibit 6.16: Number of Systems and Population Served by Systems in the SYR3 ICR Dataset
(2011) with DBP Records, by System Type	6-33
Exhibit 6.17: Number of Systems and Population Served by Systems in the SYR3 ICR Dataset
(2011) with DBP Records, by Source Water Type	6-33
Six-Year Review 3 Technical Support Document	vi
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Exhibit 6.18: Number of Systems and Population Served by Systems in the SYR3 ICR Dataset
(2011) with DBP Records, by System Size and System Type	6-33
Exhibit 6.19: SYR3 ICR Comparison of Individual THM4 Measurements to the MRL1	6-37
Exhibit 6.20: SYR3 ICR Comparison of Individual THM4 Measurements to the MCL (80 |ig/L)
	6-38
Exhibit 6.21: SYR3 ICR Comparison of Individual HAA5 Measurements to the MRL1	6-38
Exhibit 6.22: SYR3 ICR Comparison of Individual HAA5 Measurements to the MCL (60 |ig/L)
	6-39
Exhibit 6.23: SYR3 ICR Comparison of System Mean THM4 Measurements to the MCL (80
ug/I.)	6-40
Exhibit 6.24: SYR3 ICR Comparison of System Mean HAA5 Measurements to the MCL (60
ug/I.)	6-41
Exhibit 6.25: System Means from the SYR3 ICR THM4 Data (2011)	6-42
Exhibit 6.26: System Means from the SYR3 ICR HAA5 Data (2011)	6-42
Exhibit 6.27: SYR3 ICR Data Showing Cumulative Distribution of System Mean Concentrations
for THM4 by System Size and Source Water Type (2011)	6-43
Exhibit 6.28: SYR3 ICR Data Showing Cumulative Distribution of System Mean Concentrations
for HAA5 by System Size and Source Water Type in 2011	6-43
Exhibit 6.29: SYR3 ICR Comparison of Individual Bromate Measurements to the MRL1	6-47
Exhibit 6.30: SYR3 ICR Comparison of Individual Bromate Measurements to the MCL (10
ug/I.)	6-47
Exhibit 6.31: SYR3 ICR Comparison of Individual Chlorite Measurements to the MRL1	6-48
Exhibit 6.32: SYR3 ICR Comparison of Individual Chlorite Measurements to the MCL (1,000
ug/I.)	6-48
Exhibit 6.33: SYR3 ICR Comparison of System Mean Bromate Measurements to the MCL (10
ug/I.)	6-49
Exhibit 6.34: DBP ICR Data: Paired Monitoring Results for Chlorate and Chlorite	6-51
Exhibit 6.35: EWG Data: Paired System Daily Averages for Chlorate and Chlorite	6-52
Exhibit 6.36: System Highest Chlorate and Chlorite Levels in UCMR 3 and SYR3 ICR Datasets
(\ 73)	6-53
Six-Year Review 3 Technical Support Document	vii
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Exhibit 7.1: Required TOC Removal for Conventional Treatment Plants Using Surface Water or
GWUDI1'2'3	7-3
Exhibit 7.2: Treatment Technologies Considered for the Stage 1 and Stage 2 D/DBPRs1	7-3
Exhibit 7.3: Percent TOC Removal from Source to Filter Effluent by Surface Water Filtration
Treatment Plant Types Based on DBP ICR Dataset	7-5
Exhibit 7.4: Evaluation of TOC Compliance Monitoring Data from SYR3 ICR Dataset Relative
to 3x3 Matrix (Based on Paired TOC Data from 2006-2011)	7-8
Exhibit 7.5: TOC Removal by System Size from SYR3 ICR Dataset (Based on Paired TOC Data
from 2006-2011)	7-10
Exhibit 7.6: Treated Water TOC Levels by System Size from SYR3 ICR Dataset (Based on
Paired TOC Data from 2006-2011)	7-10
Six-Year Review 3 Technical Support Document	viii
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Acronyms
2-CAA
2-Chloroacetaldehyde
AOB
Ammonia Oxidizing Bacteria
ASTM
American Society of Testing and Materials
AWWA
American Water Works Association
BAC
Biological Activated Carbon
BAL
Bromoacetaldehyde
BAT
Best Available Technology
BCAA
Bromochloroacetic acid
BCAL
Bromochloroacetaldehyde
BCAN
Bromochloroacetonitrile
BCC
Basal Cell Carcinoma
BDCAA
Bromodichloroacetic acid
BDCAL
Bromodichloroacetaldehyde
BDCM
Bromodichloromethane
BIAA
Bromoiodoacetic acid
BIF
Bromine Incorporation Factor
BMDL
Benchmark dose level
BMX-2
3-Chloro-4-(Dibromomethyl)-5-Hydroxy-2(5H)-Furanone
BrTHMs
Brominated Trihalomethanes
CAGC
Chloramine Formed from Gaseous Chlorine
CAL
Chi oroacetal dehy de
CAOF
Chloramine Formed from Off-Site Hypochlorite
CAON
Chloramine Formed from On-Site Hypochlorite
CCC
Chlorine Chemistry Council
CCL
Contaminant Candidate List
CFR
Code of Federal Regulations
CG
Chorionic Gonadotropin
CHF
Chlorohydroxyfuranones
CHO
Chinese Hamster Ovary
CI
Confidence Interval
CLDO
Chlorine Dioxide
CLGA
Gaseous Chlorine
CLM
Chloramine
CLOF
Off-site Generated Hypochlorite Stored as Liquid
CLON
On-site Generated Hypochlorite with No Storage
CMA
Chemical Manufacturers Association
CNS
Central Nervous System
CNX
Cyanogen Halides
CPE
Cationic Polyelectrolyte
CWS
Community Water System
CYP2E1
Cytochrome P450 2E1
DBAA
Dibromoacetic Acid
DBAL
Dibromochloroacetaldehyde
DBAN
Dibromoacetonitrile
Six-Year Review 3 Technical Support Document	ix
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
DBCAL
Dibromochloroacetaldehyde
DBCM
Dibromochloromethane
DBP
Disinfectant Byproducts
D/DBPR
Disinfectants and Disinfection Byproducts Rules
DCAA
Dichloroacetic Acid
DCAL
Di chl oroacetal dehy de
DC AN
Dichloroacetonitrile
DCP
Di chl oropropanone
DHAA
Dihaloacetic Acids
DIAA
Diiodoacetic Acid
DMA
Dimethylnitrosamine
DOC
Dissolved Organic Carbon
DON
Dissolved Organic Nitrogen
Cal DPR
California Department of Pesticide Regulation
EBCT
Empty Bed Contact Time
EPA
United States Environmental Protection Agency
EWG
Environmental Working Group
FBRR
Filter Backwash Recycling Rule
FDA
Food and Drug Administration
GAC
Granular Activated Carbon
GAPDH
Glyceraldehyde-3 -Phosphate Dehydrogenase
GD
Gestational Day
GSH
Glutathione
GST
Glutathione S-Transferase
GSTM1
Glutathione S-Transferase Mu 1
GSTP1
Glutathione S-Transferase Pi 1
GSTT1
Glutathione S-Transferase Theta 1
GSTZ1
Glutathione Transferase Zeta 1
GU
Ground Water Under Direct Influence of Surface Water
GUP
Purchased Ground Water Under the Direct Influence of Surface Water
GW
Ground Water
GWP
Purchased Ground Water
GWR
Ground Water Rule
GWUDI
Ground Water Under Direct Influence of Surface Water
HAA
Haloacetic Acid
HAA5
Group of five regulated HAAs: monochloroacetic acid, dichloroacetic acid,

trichloroacetic acid, monobromoacetic acid and dibromoacetic acid
HAL
Hal oacetal dehy de
HAN
Haloacetonitrile
HBQ
Halogenated Benzoquinone
HNM
Halonitromethane
HQ
Haloquinone
HRL
Health Reference Level
IAL
Iodoacetaldehyde
IARC
International Agency for Research on Cancer
ICR
Information Collection Request
Six-Year Review 3 Technical Support Document	x
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
ICRTSD
Information Collection Request Treatment Study Database
IDSE
Initial Distribution System Evaluation
IESWTR
Interim Enhanced Surface Water Treatment Rule
IDSE
Initial Distribution System Evaluation
ILSI
International Life Sciences Institute
IRIS
Migrated Risk Information System
IRR
Incidence Rate Ratios
IUGR
Intrauterine Growth Retardation
LBW
Low Birth Rate
LCMRL
Lowest Concentration Minimum Reporting Level
LH
Luteinizing Hormone
LOAEL
Lowest Observed Adverse Effect Level
LRAA
Locational Running Annual Average
LT1
Long-Term 1 Enhanced Surface Water Treatment Rule
LT2
Long-Term 2 Enhanced Surface Water Treatment Rule
MAC
Maximum Acceptable Concentration
MBAA
Monobromoacetic Acid
MCAA
Monochloroacetic Acid
MCL
Maximum Contaminant Level
MCLG
Maximum Contaminant Level Goal
MDBP
Microbial and Disinfection Byproduct
MDL
Method Detection Limit
MGD
Millions of gallons per day
MIAA
Monoiodoacetic acid
MIEX
Magnetic Ion Exchange
MOA
Modes of Action
MRDL
Maximum Residual Disinfectant Level
MRDLG
Maximum Residual Disinfectant Level Goal
MRL
Minimum Reporting Level
MTBE
Methyl Tert-Butyl Ether
MX
3-Chloro-4(Dichloromethyl)-5-Hydroxy-2(5H)-Furanone
NAT2
N-Acetyltransferase 2
NCI
National Cancer Institute
NDBA
N-Nitrosodi-N-Butylamine
NDEA
N-Nitrosodiethylamine
NDMA
N-Nitrosodimethylamine
NDPA
N-Nitrosodi-N-Propylamine
NEMI
National Environmental Methods Index
NMEA
N-nitrosomethylethylamine
NOAEL
N o-Ob served-Adverse-Effect-Level
NODA
Notices of Data availability
NODU
No Disinfection
NOM
Natural Organic Matter
NPDWR
National Primary Drinking Water Regulations
NRWA
National Rural Water Association
NTNCWS
Non-Transient Non-Community Water System
Six-Year Review 3 Technical Support Document	xi
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
NTP
National Toxicology Program
NTU
Nephelometric Turbidity Unit
NPYR
N-Nitrosopyrrolidine
OR
Odds Ratio
OTHD
All Other Types of Disinfectant
OZON
Ozone
PAC
Polyaluminum Chloride
PAR
Population Attributable Risk
PBI
Percentage of Bromide Incorporation
PHG
Public Health Goal
PND
Postnatal Day
PTB
Pulmonary Tuberculosis
PTD
Pre-Term Delivery
PWS
Public Water Systems
QA/QC
Quality Assurance and Quality Control
RAA
Running Annual Average
RIA
Regulatory Impact Analysis
ROS
Reactive Oxygen Species
RSC
Relative Score Contribution
RSD
Relative Standard Deviation
RSSCT
Rapid Small-Scale Column Test
RR
Relative Risk
RTCR
Revised Total Coliform Rule
SAB
Science Advisory Board
see
Squamous Cell Carcinoma
sees
Scientific Committee on Consumer Safety
SCGE
Single Cell Gel Electrophoresis
SDS
Simulated Distribution System
SDWA
Safe Drinking Water Act
SDWIS
Safe Drinking Water Information System
SGA
Small for Gestational Age
SUVA
Specific Ultraviolet Light Absorbance
SW
Surface Water
SWAT
Surface Water Analytical Tool
SWP
Purchased Surface Water
SWTR
Surface Water Treatment Rule
SYR
Six-Year Review
SYR2
Second Six-Year Review
SYR3
Third Six-Year Review
TAME
Tert-amyl methyl Ether
TBAA
Tribromoacetic Acid
TBAL
Tribromoacetaldehyde
TCAA
Trichloroacetic Acid
TCAL
T ri chl oroacetal dehy de
TCAN
Trichloroacetonitrile
TCR
Total Coliform Rule
Six-Year Review 3 Technical Support Document	xii
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
TDI
Tolerable Daily Intake
TDS
Total Dissolved Solids
THAA
Trihaloacetic acids
THM
Trihalomethanes
THM4
Group of four regulated THMs: bromoform, bromodichloromethane,

dibromochloromethane and chloroform
THMFP
Trihalomethane Formation Potential
TOC
Total Organic Carbon
TON
Total Organic Nitrogen
TOX
Total Organic Halogens
TT
Treatment Technique
TTHM
Total Trihalomethanes
UCMR
Unregulated Contaminant Monitoring Rule
USGS
United States Geological Survey
UV
Ultraviolet
UVLV
Ultraviolet Light
WEOM
Water-Extractable Organic Matter
WHO
World Health Organization
WRF
Water Research Foundation
WTP
Water Treatment Plants
Six-Year Review 3 Technical Support Document	xiii
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
1 Introduction
The 1996 Safe Drinking Water Act (SDWA) Amendments require the United States
Environmental Protection Agency (EPA or the Agency) to periodically review existing national
primary drinking water regulations (NPDWRs) and determine which, if any, need to be revised.1
The purpose of the review, called the Six-Year Review (SYR), is to evaluate current information
for regulated contaminants to determine if there is new information on health effects, treatment
technologies, analytical methods, occurrence and exposure, implementation and/or other factors
that provides a health or technical basis to support a regulatory revision that will improve or
strengthen public health protection.
EPA completed and published the results of its first Six-Year Review (SYR1), on July 18, 2003
(USEPA, 2003a) and the second Six-Year Review (SYR2), on March 29, 2010 (USEPA, 2010a),
after developing a systematic approach, or protocol, for the review of NPDWRs. During SYR1,
EPA identified the Total Coliform Rule (TCR) as a candidate for revision. NPDWRs for four
additional contaminants (acrylamide, epichlorohydrin, tetrachloroethylene and trichloroethylene)
were identified as candidates for revision during the SYR2.
Under the third Six-Year Review (SYR3), EPA is reviewing the regulated chemical, radiological
and microbiological contaminants included in previous reviews, as well as the microbial and
disinfection byproducts (MDBP) regulations. Except for the 1989 Total Coliform Rule (TCR),
which was reviewed in SYR1, this is the first time EPA is conducting a SYR of the MDBP
regulations. This review includes the Stage 1 and Stage 2 Disinfectants/Disinfection Byproducts
Rules (D/DBPRs) as well as the following microbial contaminant regulations:
•	Surface Water Treatment Rule (SWTR)
•	Interim Enhanced Surface Water Treatment Rule (IESWTR)
•	Long-Term 1 Enhanced Surface Water Treatment Rule (LT1)
•	Long-Term 2 Enhanced Surface Water Treatment Rule (LT2)
•	Filter Backwash Recycling Rule (FBRR)
•	Ground Water Rul e (GWR).
Results from the review of the SWTR, the IESWTR, the LT1, the FBRR and the GWR are
discussed in a separate support document (USEPA, 2016a).
EPA is reviewing the LT2 in response to the Executive Order 13563 Improving Regulation and
Regulatory Review (also known as Retrospective Review) and as part of the SYR3 process.
1 Under the SDWA, EPA must periodically review existing national primary drinking water regulations (NPDWRs)
and, if appropriate, revise them. Section 1412(b)(9) of SDWA states: "The Administrator shall, not less often than
every 6 years, review and revise, as appropriate, each national primary drinking water regulation promulgated under
this title. Any revision of a national primary drinking water regulation shall be promulgated in accordance with this
section, except that each revision shall maintain, or provide for greater, protection of the health of persons."
Six-Year Review 3 Technical Support Document	1-1
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Results from the review of the LT2 are discussed in a separate support document (USEPA,
2016b).
The remainder of this document provides a summary of available information and data relevant
to determining which, if any, of the NPDWRs included in the Stage 1 and Stage 2 D/DBPRs are
candidates for revision under the SYR. The information cutoff date for SYR3 was December
2015. That is, information published on or before December 2015 was considered as part of the
SYR3. The Agency recognizes that scientists and other stakeholders are continuing to investigate
disinfectants and disinfection byproducts (DBPs) and publish information subsequent to this
cutoff date. While not considered as part of the SYR3, the Agency anticipates providing
consideration for that additional information in subsequent activities.
Chapter 2 of this document provides an overview of the protocol that EPA used in this review.
Chapter 3 provides an overview of the specific regulations addressed in this support document,
along with historical information about their development. Available information and data
relevant to making a determination under the SYR3 are provided in Chapter 4 (health effects),
Chapter 5 (analytical methods), Chapter 6 (occurrence and exposure), Chapter 7 (treatment) and
Chapter 8 (other regulatory revisions).
Six-Year Review 3 Technical Support Document	1-2
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
2 EPA's Protocol for the Six-Year 3 Review
This chapter provides an overview of the process the Agency used to review the NPDWRs
discussed in the SYR3. The protocol document, EPA Protocol for the Third Review of Existing
National Primary Drinking Water Regulations, contains a detailed description of the process the
Agency used to review the NPDWRs (USEPA, 2016c). The foundation of this protocol was
developed for the SYR1 based on the recommendations of the National Drinking Water
Advisory Committee (NDWAC, 2000). This SYR3 process is very similar to the process
implemented during the SYR1 and the SYR2, with some clarifications to the elements related to
the review of NPDWRs included in the MDBP rules.
Exhibit 2.1 presents an overview of the SYR protocol and major categories of review outcomes.
The protocol is broken down into a series of questions about whether there is new information
for a contaminant that suggests it is appropriate to revise one or more of the NPDWRs. The two
major outcomes of the detailed review are either:
(1)	the NPDWR is not appropriate for revision and no action is necessary at this time, or
(2)	the NPDWR is a candidate for revision.
Individual regulatory provisions of NPDWRs that are evaluated as part of the SYR are:
maximum contaminant level goals (MCLGs), maximum contaminant levels (MCLs), maximum
residual disinfectant level goals (MRDLGs), maximum residual disinfectant levels (MRDLs),
treatment techniques (TTs), other treatment technologies and regulatory requirements (e.g.,
monitoring). The MCL provisions are not applicable for evaluation of the microbial
contaminants regulations which establish TT requirements in lieu of MCLs. The MRDLG and
MRDL provisions are only applicable for evaluation of the Disinfectants and Disinfection
Byproducts Rules (D/DBPRs) as part of the SYR.
The review elements that EPA considered for each NPDWR during the SYR3 include the
following: initial review, health effects, analytical feasibility, occurrence and exposure, treatment
feasibility, risk balancing and other regulatory revisions. Further information about these review
elements are described in the protocol document (USEPA, 2016c).
Six-Year Review 3 Technical Support Document	2-1
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Exhibit 2.1: SYR Protocol Overview and Major Categories of Revise/Take No
Action Outcomes
Yes
NPDWR reviewed in recent or ongoing action?
Regulatory action ongoing
or recently completed
NPDWRs Under Review
Health effects assessment (HEA)
in process or planned? *
Yes


No

Ongoing or planned HEA
No
Outcome:
No action
at this time
Uncertain - emerging
information
Yes
No
Yes
No
No new information
Data gaps/emerging
information
Low priority - No meaningful
opportunity
Data sufficient to support
regulatory revision?
New information to suggest possible changes (i.e.,
to an MCLG, MCL, Treatment Technique and/or
other regulatory revisions)?
Meaningful opportunity for health risk reduction for
persons served by PWS and/or cost savings while
maintaining/improving public health protection?
Outcome:
Candidate
for Revision
* Contaminants with an HEA in process that have an MCL based on practical
quantitation limit and are greater than MCLG are passed to the next question to
evaluate potential to revise the MCL.
Six-Year Review 3 Technical Support Document	2-2
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
3 History of Disinfectants and Disinfection Byproducts Regulations
This chapter provides a brief history of disinfectants and disinfection byproducts (DBPs)
regulations in the United States from 1974 to 2016. A timeline of the regulatory history is shown
in Exhibit 3.1. The most recent regulation, the Stage 2 Disinfectants and Disinfection Byproducts
Rule (D/DBPR), was promulgated on January 4, 2006 (USEPA, 2006a). As explained in Chapter
2, the Initial Review Branch of the SYR protocol indicates that a regulation that was
promulgated or revised more than six years ago is eligible for regulatory review under the SYR
process. The Stage 2 D/DBPR meets this criterion and is currently being reviewed. EPA is also
reviewing the Stage 1 D/DBPR. Although some parts of the Stage 1 D/DBPR were superseded
by requirements under the Stage 2 D/DBPR, much of the Stage 1 D/DBPR is still in effect. EPA
did not review the Stage 1 D/DBPR during SYR2 because national primary drinking water
regulations (NPDWRs) under Stage 1 were the subject of a recent rulemaking (those under the
Stage 2 D/DBPR). Under the Initial Review Branch, NPDWRs for which further review of
detailed technical data is premature because the NPDWR is the subject of recent or ongoing
rulemaking or an ongoing health effects assessment may be excluded from the SYR. Excluding
such NPDWRs from review prevents duplicative Agency efforts.
Exhibit 3.1: Timeline for Selected Regulatory Activities Associated with
Disinfectants and Disinfection Byproducts in Drinking Water
M-DBP Advisory Committee established
1996
Final Information Collection Rule
2003
Six-Year Review 1
1979
Interim TTHM Regulation
'	1998
Notice of Data Availability
2016
Six-Year Review 3
2010
Six-Year Review 2
1992
Negotiating Committee established
2003	\
Proposed Stage 2 D/DBPR
1994
Proposed Stage 1 D/DBPR
2006
Final Stage 2 D/DBPR
1998
Final Stage 1 D/DBPR
1996
SDWA Amendments
1997
Notice of Data Availability
As part of the SYR, EPA is also reviewing the regulations addressing microbiological
contaminants in drinking water. These are addressed in the technical support document for the
microbial contaminant regulations and in a separate technical support document for the Long
Term 2 Enhanced Surface Water Treatment Rule (USEPA, 2016a; 2016b).
3.1 Interim Total Trihalomethanes Regulation
In 1974, researchers in the Netherlands and the United States demonstrated that trihalomethanes
(THMs) are formed as a result of drinking water chlorination (Bellar et al., 1974; Rook, 1974).
There are four common THMs: chloroform, bromoform, bromodichloromethane (BDCM) and
Six-Year Review 3 Technical Support Document	3-1
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
dibromochloromethane (DBCM). EPA subsequently conducted surveys confirming widespread
occurrence of THMs in chlorinated water supplies (Symons et al., 1975; USEPA, 1978). During
this time toxicological studies became available supporting the contention that chloroform is
carcinogenic in at least one strain of rat and one strain of mouse (NCI, 1976).
In November of 1979, EPA set an interim maximum contaminant level (MCL) for the total
concentration of the four common THMs, or total THMs (TTHM), of 100 |_ig/L (0.100 mg/L) as
an annual average (USEPA, 1979). This standard was based on the need to balance the
requirement for continued disinfection of water to reduce exposure to pathogenic
microorganisms with the need to simultaneously lower exposure to animal carcinogens like
chloroform. TTHM was also considered a surrogate measure for other chlorination DBPs.
The interim TTHM standard only applied to community water systems (CWSs) serving at least
10,000 people that add a disinfectant (an oxidant) to the drinking water during any part of the
treatment process. At the time of promulgation, about 80 percent of the small systems (i.e., those
serving fewer than 10,000 people) used ground water sources that were mostly low in THM
precursor content (USEPA, 1979), and many of them did not disinfect. Moreover, these small
systems were considered more likely to have greater risks of significant microbiological
contamination, especially if they were to reduce or eliminate chemical disinfection. Federal rules
such as the 1989 Total Coliform and Surface Water Treatment Rules did not yet exist to further
protect against microbial contamination. In 1979, the majority of outbreaks attributable to
inadequate disinfection occurred in small systems. Further, EPA determined that small systems
had limited access to the financial resources and technical expertise needed for TTHM control.
Therefore, EPA decided not to require small systems to comply with the TTHM MCL at that
time (USEPA, 1994a).
3.2 Stage 1 D/DBPR
3.2.1 Negotiated Rulemaking
EPA was required to develop rules for additional contaminants under the 1986 Amendments to
the Safe Drinking Water Act (SDWA). To solicit public comment in developing a rule, EPA
released a "strawman" rule (a pre-proposal draft) in October 1989. In this strawman rule, EPA
included a primary option of setting maximum contaminant level goals (MCLGs) and MCLs for
THMs, haloacetic acids (HAAs), chlorite and chlorate. Compliance with the MCL for HAAs was
to be based on total concentrations of five HAAs (HAA5): monochloroacetic, dichloroacetic,
trichloroacetic, monobromoacetic and dibromoacetic acids. THM4 and HAAs had shown an
association with cancer, and they were also potential indicators of the presence of other
byproducts in disinfected water that may also have adverse health effects (USEPA, 1994a). EPA
intended to monitor concentrations of TTHM and HAA5 for compliance, but their presence in
drinking water was thought to be representative of many other DBPs that may also be present in
the water; thus, a reduction in TTHM and HAA5 would indicate a reduction of total DBPs. In
addition to DBPs, the primary option in the strawman rule also included limits and health goals
for the disinfectants chlorine, chloramines and chlorine dioxide (USEPA, 1994a).
Many system operators who commented on the strawman rule were concerned about the effects
of modifying their treatment processes to meet DBP MCLs (USEPA, 1994a). These concerns
Six-Year Review 3 Technical Support Document	3-2
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
included reduced microbiological protection, creation of conditions that favored distribution
system microbiological growth (e.g., use of ozone would create biodegradable organics, and use
of chloramines would provide a nitrogenous source) and formation of residuals during treatment
that would require disposal.
EPA published a status report in June 1991 on DBP regulation development that was designed to
reflect the Agency's thinking on the strawman rule (USEPA, 1994a). The status report indicated
that EPA was considering extending coverage under the rule to all community and non-transient
non-community systems (instead of just CWSs serving at least 10,000 people, as under the 1979
TTHM rule) and was proposing a shorter list of compounds for regulation than were included in
the 1989 strawman rule.
In the status report, EPA identified risk-balancing as an issue that needed to be considered as the
rule was being developed. EPA wanted to ensure that the new regulation would not introduce
new risks. For instance, one issue was the use of alternate disinfectants to limit chlorination
byproducts. The Agency recognized that although alternate disinfection schemes (e.g., ozone and
chloramines) could greatly reduce chlorination byproducts, little was known about the
byproducts of the alternate disinfectants and their associated health risks. EPA did not want to
promulgate a standard that encouraged the shift to alternate disinfectants unless the associated
risks (including both those from byproducts and differential microbial risks from a change in
disinfectants) were adequately understood (USEPA, 1994a).
Another aspect of risk-balancing was integration with other rules, such as the Surface Water
Treatment Rule (SWTR), which had been promulgated in 1989. EPA wanted to ensure that
compliance with regulations on DBPs would not affect compliance with or protection provided
by the SWTR. Although the SWTR only mandated 3-log (99.9 percent) removal or inactivation
of Giardia and 4-log (99.99 percent) removal or inactivation of viruses, EPA guidance
recommended higher levels of treatment for poorer quality source waters. EPA was concerned
that systems would reduce microbial protection to levels nearer to the regulatory requirements by
reducing disinfection and, as a result, possibly increase microbial risks in an effort to meet DBP
MCLs. The Agency wanted to ensure adequate microbial protection while reducing risk from
DBPs (USEPA, 1994a).
EPA became interested in pursuing a negotiated rulemaking process for the development of the
D/DBPR, in large part, because no clear path for addressing all the major issues identified in the
June 1991 status report on the D/DBPR was apparent. A negotiated rule process would help
stakeholders understand the complexities of the risk-balancing issue and help reach a consensus
on the most appropriate regulation to address concerns on both DBPs and microorganisms. In
1992, EPA established the Negotiating Committee (USEPA, 1994a).
The Committee worked out an "agreement in principle" on a first round of DBP controls at its
February 1993 meeting. The "Stage 1" agreement recommended MCLs for TTHM and HAA5 at
levels the Committee deemed protective of public health: 80 and 60 |j,g/L (0.080 and 0.060
mg/L), respectively. To limit DBP precursors, the committee agreed to develop a series of
"enhanced coagulation" requirements, to vary according to systems' influent water quality and
treatment plant configurations. Members also agreed to reconvene in several years to develop a
Six-Year Review 3 Technical Support Document	3-3
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
second stage of DBP regulations, when the results of more health effects research and water
quality monitoring were available (USEPA, 1994a).
For the most part, EPA adopted the recommendations of the Negotiating Committee and the
supporting Technical Work Group for the proposed Stage 1 D/DBPR (USEPA, 1994a).
3.2.2	Proposed Stage 1 D/DBPR
The proposed Stage 1 D/DBPR was published in the Federal Register on July 29, 1994. EPA
proposed maximum residual disinfectant level goals (MRDLGs) for chlorine, chloramines and
chlorine dioxide, and MCLGs for each of the four THMs, two HAAs (dichloroacetic acid
(DCAA) and trichloroacetic acid (TCAA)), chloral hydrate, bromate and chlorite. EPA also
proposed maximum residual disinfectant levels (MRDLs) for three disinfectants (chlorine,
chloramines and chlorine dioxide), and MCLs for TTHM, HAA5 and two inorganic DBPs
(chlorite and bromate) (USEPA, 1994a). No MCLG was proposed for chlorate due to insufficient
toxicological and epidemiological data. Note that EPA proposed that the MRDL for chlorine
dioxide also apply to transient non-community systems because of the concern from short-term
exposure health effects; this was the only requirement in the proposal to apply to transient
systems. All other proposed requirements applied only to community and non-transient non-
community systems adding a chemical disinfectant (as EPA had recommended in its 1991 status
report). The proposed MRDLs and MRDLGs were similar in concept to MCLs and MCLGs;
however, since disinfectants were a necessary part of the treatment process, they could not be
considered contaminants. EPA therefore developed new terms to describe limits for
disinfectants. In addition, EPA proposed treatment techniques (TTs) (enhanced coagulation and
enhanced softening) to remove DBP precursors in systems using conventional treatment. The
proposed regulations also included monitoring, reporting and public notification requirements.
3.2.3	DBP Information Collection Rule
In 1994 EPA also proposed the Information Collection Rule (ICR) (USEPA, 1994b). The
monitoring requirements of the ICR were proposed to (1) characterize source water parameters
that influence DBP formation, (2) determine the concentrations of DBPs in drinking water, (3)
refine models for predicting DBP formation based on treatment and water quality parameters,
and (4) establish cost-effective monitoring requirements that are protective of public health. It
required systems to monitor for DBPs, along with source water parameters such as total organic
carbon (TOC), pH and alkalinity (USEPA, 1994b). The ICR also required source water
monitoring for Cryptosporidium, Giardia, viruses, Escherichia coli and total coliform bacteria in
surface water systems and systems using ground water under the direct influence of surface
water (GWUDI). The specific requirements varied by system size and source water type. The
proposed rule also required water systems, unless they met certain exclusionary criteria, to
conduct pilot- or bench-scale studies of GAC or membranes to determine the effectiveness of
these technologies for DBP removal. The ICR served as one of most important data sources
supporting the development of the Stage 2 D/DBPR and is further described in Chapter 6.
Six-Year Review 3 Technical Support Document	3-4
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
3.2.4	The 1996 Safe Drinking Water Act Amendments, Microbial and
Disinfectants/Disinfection Byproducts (MDBP) Advisory Committee and Notices of
Data Availability
The SDWA amendments of 1996 codified the risk-balancing concept. They allowed EPA to
establish an MCL "at a level other than the feasible level, if the technology, TTs and other means
used to determine the feasible level would result in an increase in the health risk from drinking
water by (i) increasing the concentration of other contaminants in drinking water or (ii)
interfering with the efficacy of drinking water TTs or processes that are used to comply with
other national primary drinking water regulations" (section 1412(b)(5)(A)). The amendments
further required that MCLs or TTs "minimize the overall risk of adverse health effects by
balancing the risk from the contaminant and the risk from other contaminants the concentrations
of which may be affected by the use of a TT or process that would be employed to attain the
maximum contaminant level or levels" (section 1412(b)(5)(B)).
Congress also required EPA to promulgate the D/DBPR in two stages as part of the 1996 SDWA
amendments (section 1412(b)(2)(C)). To help meet the statutory deadlines established by
Congress in the amendments and to maximize stakeholder participation, the Agency established
the Microbial and Disinfectants/Disinfection Byproducts (MDBP) Advisory Committee under
the Federal Advisory Committee Act in 1997 to analyze new information and data, as well as to
build consensus on the regulatory implications of this new information. The Committee
consisted of 17 members representing EPA, state and local public health and regulatory agencies,
local elected officials, drinking water suppliers, chemical and equipment manufacturers, and
public interest groups (USEPA, 2003b).
The Committee met five times, from March through July 1997, to discuss issues related to the
IESWTR and Stage 1 D/DBPR. Technical support for these discussions was provided by a
technical work group established by the Committee. The Committee's activities resulted in the
collection, development, evaluation and presentation of substantial new data and information
related to key elements of both proposed rules (USEPA, 2003b). These data were included in
two notices of data availability (NODAs) issued by EPA, as discussed below.
EPA published the two NODAs in 1997 and 1998. The 1997 NODA (USEPA, 1997a) addressed
studies on the ability of enhanced coagulation to remove TOC, new epidemiological and
toxicological information, and possible changes for the final rule regarding the point of
disinfection and disinfection benchmarking. The 1998 NODA (USEPA, 1998a) provided new
epidemiological and toxicological information and requested comment on possible changes to
some of the MCLGs in the 1994 proposal. It also requested comment on possible issues that
might arise from simultaneous compliance with the Stage 1 D/DBPR and the Lead and Copper
Rule.
3.2.5	Final Stage 1 D/DBPR
EPA finalized the Stage 1 D/DBPR (USEPA, 1998b) on December 16, 1998 (note that the final
IESWTR was also promulgated at this time). The final rule established the MCLGs and MCLs
listed in Exhibit 3.2 and the MRDLs and MRDLGs listed in Exhibit 3.3. The final rule did not
include an MCLG for chloral hydrate because it was deemed to be adequately protected for by
Six-Year Review 3 Technical Support Document	3-5
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
the other rule requirements. The final rule revised the proposed MCLG for chlorite and the
MRDLG for chlorine dioxide based on new toxicological data presented in the 1998 NOD A. All
other MCLGs and MRDLGs were promulgated as proposed. All MCLs and MRDLs were also
promulgated as proposed. The rule required systems to monitor TTHM and HAA5 at locations
representing average and/or maximum residence times in the distribution system, with the
sampling frequency and number of samples based on the population served and the number of
plants a system had. Compliance with the MCLs for TTHM, HAA5 and (for systems using
ozone) bromate, as well as with the MRDLs for chlorine and chloramines was determined based
on running annual averages (RAAs) of those samples. Compliance with the MCL for chlorite
(only for systems using chlorine dioxide) was based on the average of three samples taken in the
distribution system. For chlorine dioxide, the rule established two types of MRDL violations—
acute and non-acute, based on whether a system exceeds the MRDL at just the entrance to the
distribution system or within the distribution system as well. The rule allowed reduced
monitoring for TTHM, HAA5, chlorite and bromate under certain conditions. The final rule
applied to all community and non-transient non-community water systems (NTNCWSs) that
added a disinfectant. Systems that purchased water that had already been disinfected were not
subject to the rule.
Exhibit 3.2: Stage 1 and Stage 2 MCLs and MCLGs
DBPs
Stage 1
Stage 2
MCLG (mg/L)
MCL (mg/L) as
RAA
MCLG (mg/L)
MCL (mg/L) as
LRAA1
Chloroform
0
NA
0.07
NA
Bromodichloromethane
0
NA
0
NA
Dibromochloromethane
0.06
NA
0.06
NA
Bromoform
0
NA
0
NA
TTHM
NA
0.080
NA
0.080
Monochloracetic acid
NA
NA
0.07
NA
Dichloroacetic acid
0
NA
0
NA
Trichloroacetic acid
0.3
NA
0.02
NA
Monobromoacetic acid
NA
NA
NA
NA
Dibromoacetic acid
NA
NA
NA
NA
HAA5
NA
0.060
NA
0.060
Bromate
0
0.010
0
0.010
Chlorite
0.8
1.0
0.8
1.0
1 Locational running annual average, discussed in Section 3.3.1.
Six-Year Review 3 Technical Support Document	3-6
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Exhibit 3.3: Stage 1 and Stage 2 MRDLs and MRDLGs
Disinfectants
Stage 1
Stage 2
MRDLG (mg/L)
MRDL (mg/L)
MRDLG (mg/L) MRDL (mg/L)
Chlorine
4 (as Cb)
4 (as Cb)
Same as Stage 1
Chloramines
4 (as Cb)
4 (as Cb)
Chlorine Dioxide
0.8 (as CIO2)
0.8 (as CIO2)
Under the Stage 1 D/DBPR, the best available technology (BAT) for complying with the TTHM
and HAA5 MCLs was determined to be enhanced coagulation, enhanced softening or GAC with
a 10-minute empty-bed contact time. For bromate and chlorite, control of treatment processes
was determined to be the BAT (USEPA, 1998b).
The final Stage 1 D/DBPR also established a TT for TOC removal in plants that use
conventional treatment (USEPA, 1998b). The required percentage of TOC removal depended on
the source water TOC and alkalinity and is shown in Exhibit 3.4 below. Where meeting the
removals in the exhibit below was found to be technically infeasible, the system could apply to
the state (i.e., primacy agency) for alternative removal criteria determined by laboratory jar
testing. The final requirements were similar to the proposed requirements. However, removal
requirements in the final rule for plants with source water TOC >2.0 and up to 4.0 mg/L were
decreased as a result of studies showing that the proposed removals would be difficult to meet
for many systems and would place a significant burden on states, which would have to approve
alternative removal criteria. The TT for TOC removal is discussed further in Chapter 7.
Exhibit 3.4: Stage 1 D/DBPR Required Removal of Total Organic Carbon by
Enhanced Coagulation and Enhanced Softening for Subpart H Systems Using
Conventional Treatment123
Source Water TOC
(mg/L)
Percentage Removal Required
(Based on Source Water Alkalinity in mg/L)
0-60 mg/L
>60-120 mg/L
>120 mg/L
>2.0-4.0
35.0
25.0
15.0
>4.0-8.0
45.0
35.0
25.0
>8.0
50.0
40.0
30.0
1	Systems meeting at least one of the conditions in Section 141.135(a)(2) (i)-(vi) of the rule are not required to meet
the removals in this exhibit.
2	Softening systems meeting one of the two alternative compliance criteria in Section 141.135(a)(3) of the rule are not
required to meet the removals in this exhibit. Chapter 7 provides greater detail.
3	Systems practicing softening must meet the TOC removal requirements in the last column to the right.
3.3 Stage 2 D/DBPR
The Stage 2 D/DBPR was designed to further reduce the levels of exposure from disinfectants
and DBPs without undermining the control of microbial pathogens. The Long Term 2 Enhanced
Surface Water Treatment Rule (LT2ESWTR) was proposed at the same time as the Stage 2
D/DBPR to ensure that drinking water was microbiologically safe at the limits set for
disinfectants and DBPs. (Note that the Long Term 1 Enhanced Surface Water Treatment Rule,
Six-Year Review 3 Technical Support Document	3-7
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
which pertained to systems serving fewer than 10,000 people, was finalized in 2001, before the
effective compliance date for small systems under the Stage 1 D/DBPR.) These regulations
established removal/inactivation requirements for Cryptosporidium)
3.3.1 MDBP Advisory Committee and New Information
EPA reconvened the MDBP Advisory Committee in March 1999 to develop recommendations
on issues pertaining to the Stage 2 D/DBPR and LT2ESWTR. The Advisory Committee
collected and evaluated new information that became available after the Stage 1 D/DBPR was
published. The ICR provided new data on DBP occurrence and treatment control (note that
although these data were collected prior to the Stage 1 promulgation they did not become
available until after that rule became final); it also included new data on occurrence and
treatment of pathogens. These data were supplemented by a survey conducted by the National
Rural Water Association (NRWA), data provided by various states, data provided by the
American Water Works Association and Information Collection Rule Supplemental Surveys
(USEPA, 2003b).
Although the Stage 1 D/DBPR was projected to achieve a major reduction in DBP exposure, the
ICR data suggested that some customers would receive drinking water with elevated DBP levels
even when their distribution systems were meeting the MCLs established by the Stage 1
D/DBPR. That is, sample results at a single monitoring location could exceed 0.080 mg/L
TTHM or 0.060 mg/L HAA5, even when the RAAs were below these levels. The ICR results
also showed that Stage 1 D/DBPR monitoring sites might not be representative of the highest
DBP concentrations that occur in distribution systems (USEPA, 2003b). In addition, the new
information indicated that technologies including ultraviolet light (UV) for inactivation of
protozoa, in combination with other technologies for control of DBPs such as GAC, could be
very effective at lowering DBP levels. GAC was found to be most effective for systems with
TOC levels less than 6 mg/L. Of the plants that conducted a GAC pilot- or bench-scale treatment
study under the ICR, approximately 70 percent of the surface water plants studied could meet the
0.080 mg/L TTHM and 0.060 mg/L HAA5 RAAs, with a 20 percent safety factor (i.e., 0.064
mg/L and 0.048 mg/L, respectively) using GAC with 10 minutes of empty-bed contact time and
a 120-day reactivation frequency, and 78 percent of the plants could meet the MCLs with a 20
percent safety factor using GAC with 20 minutes of empty-bed contact time and a 240-day
reactivation frequency. The ICR treatment study results also demonstrated that nanofiltration was
a better DBP control technology (as opposed to GAC) for ground water sources with TOC
concentrations above approximately 6 mg/L (USEPA, 2003b).
After promulgation of the Stage 1 D/DBPR, new information on health effects also became
available that supported the need for the Stage 2 D/DBPR. New epidemiology and toxicology
studies evaluating bladder and rectal cancers increased the weight-of-evidence linking these
health effects to DBP exposure. The available epidemiology studies on bladder cancer related to
consumption of chlorinated drinking water allowed EPA to develop quantitative risk and benefits
estimates for that health endpoint, as discussed in greater detail in Chapter 4. Several new
reproductive and developmental studies became available, so EPA completed a more extensive
analysis of reproductive and developmental effects associated with DBPs. Both human
epidemiology studies and animal toxicology studies showed associations between chlorinated
drinking water and reproductive and developmental endpoints such as spontaneous abortion,
Six-Year Review 3 Technical Support Document	3-8
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
stillbirth, neural tube defects, pre-term delivery, intrauterine growth retardation and low birth
weight, but the data were not consistent enough to support a quantitative benefits analysis
(USEPA, 2006a).
Taking into account this new information, in 2000, the MDBP Advisory Committee developed
an agreement in principle for the Stage 2 D/DBPR (USEPA, 2000a). In the agreement, the
committee recommended maintaining the MCLs for TTHM and HAA5 at 0.080 mg/L and 0.060
mg/L, respectively, but changing the compliance calculation in two phases. The Stage 1 RAA
calculations averaged all samples collected within a distribution system over a one-year period.
The Stage 2 compliance determination would switch to a calculation based on the RAA at each
sampling location in the distribution system (referred to as the "locational" running annual
average (LRAA)). In the first phase, systems would continue to comply with the Stage 1
D/DBPR MCLs as RAAs and, at the same time, comply with MCLs of 0.120 mg/L for TTHM
and 0.100 mg/L for HAA5 calculated as LRAAs. Systems would also carry out an initial
distribution system evaluation (IDSE) to select compliance monitoring sites that accurately
reflect higher TTHM and HAA5 levels occurring in the distribution system. The second phase of
compliance would require MCLs of 0.080 mg/L for TTHM and 0.060 mg/L for HAA5 calculated
as LRAAs at individual monitoring sites identified through the IDSE. The Agreement in
Principle also provided recommendations for simultaneous compliance with the LT2ESWTR so
that the reduction of potential health hazards of DBPs did not compromise microbial protection
(USEPA, 2003b).
3.3.2 Proposed Stage 2 D/DBPR
EPA published the proposed Stage 2 D/DBPR on August 18, 2003. A summary of the key
components of the rule is included here.
The proposed rule (USEPA, 2003b) extended the applicability of the rule to include community
and non-transient non-community systems that deliver water that has been treated with a primary
or residual disinfectant other than UV light (under the Stage 1 D/DBPR, only systems that added
a disinfectant were subject to the requirements). This change was intended to account for DBPs
in consecutive systems, which purchase or otherwise obtain water from other public water
systems but do not necessarily add disinfectant themselves. Consecutive systems would be
required to comply with the revised MCLs for TTHM and HAA5 as well as the MRDLs for
chlorine and chloramines. They would not need to comply with MRDLs for chlorine dioxide or
MCLs for bromate and chlorite.
In response to new health information, the proposed Stage 2 D/DBPR revised the MCLGs for
chloroform, TCAA and monochloroacetic acid (MCAA).
In addition to the change in compliance calculation described above under the Agreement in
Principle, the proposed rule required that, in most cases, the number of TTHM and HAA5
samples collected would be based on the number of plants in a system. However, for consecutive
systems that bought all their water from other systems, the number of samples would be based on
the population served. Reduced monitoring would be permitted. For DBPs other than TTHM and
HAA5 and for disinfectants, non-consecutive systems would continue to comply with the MCLs
Six-Year Review 3 Technical Support Document	3-9
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
and MRDLs and the monitoring requirements specified in the Stage 1 D/DBPR (USEPA,
2003b).
The proposed rule required systems to conduct an IDSE based on one year of TTHM and HAA5
monitoring; specific requirements were based on source water type and system size. Instead of
collecting new data, systems also had the option of performing a study based on historical data,
distribution system modeling, or other data, and IDSE waivers were available under certain
circumstances. Systems were to submit a monitoring plan based on the IDSE results (USEPA,
2003b).
Lastly, the proposed rule required systems to evaluate "significant excursions," where individual
TTHM or HAA5 samples exceed a peak level designated by the state (note that the final rule
modified this requirement to a peak level defined by EPA). Systems would be required to
evaluate their operations to determine opportunities for reducing DBP formation and would
submit a report to the state (USEPA, 2003b). EPA proposed nanofiltration and two GAC options
as BATs for wholesale systems complying with the proposed revisions to the MCLs. It proposed
a separate BAT for consecutive systems.
3.3.3 Final Stage 2 D/DBPR
The final Stage 2 D/DBPR was published January 4, 2006. Most of the elements of the proposed
rule were retained as proposed (USEPA, 2006a). However, there were some differences, as
described below.
The MCLGs for chloroform and TCAA were finalized as proposed (USEPA, 2006a); however,
the MCLG for MCAA was revised. The final MCLGs are shown in Exhibit 3.2.
The final rule eliminated the proposed two-phase implementation period for calculating
compliance. The rule required systems to transition directly from calculating compliance as a
RAA to calculating it as a LRAA (USEPA, 2006a). The MCL values themselves remained
unchanged from the Stage 1 D/DBPR. The final rule also established monitoring requirements
based on population served by the system (the proposal had based requirements on number of
treatment plants).
The final Stage 2 D/DBPR revised the "significant excursion" requirements (USEPA, 2006a).
The rule established a threshold called the "operational evaluation level" that is determined for
each monitoring location using compliance monitoring results, above which systems would be
required to implement an operational evaluation. The operational evaluation levels for each
monitoring location are determined by the sum of the two previous quarters' TTHM (or HAA5)
results plus twice the current quarter's TTHM (or HAA5) result, at that location, divided by four
to determine an average ((Q1+Q2 +2Q3)/4)). If the average TTHM exceeds 0.080 mg/L at any
monitoring location or the average HAA5 exceeds 0.060 mg/L at any monitoring location, the
system must conduct an operational evaluation and submit a written report about the operational
evaluation to the state.
The operational evaluation includes an examination of system treatment and distribution
operational practices, including changes in sources or source water quality, storage tank
operations and excess storage capacity that may contribute to high TTHM and HAA5 formation.
Six-Year Review 3 Technical Support Document	3-10
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Systems must also identify what steps could be considered to minimize future operational
evaluation level exceedances (USEPA, 2006a).
The final rule did not modify the Stage 1 TOC precursor removal requirements, except for a
minor edit.
The BATs for the final rule are the same as for the proposed rule, except for a minor change for
small consecutive systems.
Six-Year Review 3 Technical Support Document	3-11
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
4 Health Effects
This chapter addresses the health effects of disinfectants and disinfection byproducts (D/DBPs).
This chapter is organized into multiple sections, each section addressing the health effects
associated with various types of DBPs and disinfectants. Section 4.1 focuses on the adverse
health effects that are associated with the regulated organic DBPs, specifically, trihalomethanes
(THMs) and haloacetic acids (HAAs). The health effects associated with the regulated inorganic
DBPs, bromate and chlorite, are addressed in Section 4.2. Section 4.3 addresses the health effects
associated with regulated disinfectants. Section 4.4 describes health effects information for
several "unregulated" organic DBPs. Section 4.5 presents a summary of the data available about
unregulated disinfectants.
Appendix A provides additional information on the health effects of the regulated organic and
inorganic DBPs and the regulated disinfectants. It includes additional toxicological and
epidemiological information available during the development of the Stage 1 and Stage 2
D/DBPRs, as well as some additional details on the epidemiological information that has become
available since the Stage 2 rule. Appendix A is organized in the same manner as Chapter 4.
4.1 Regulated Organic DBPs
Of the 11 DBPs regulated by EPA, 9 are organic chemicals: 4 THMs (collectively called THM4)
and 5 HAAs (collectively called HAA5). THM4 is a group of four regulated THMs: bromoform,
bromodichloromethane (BDCM), dibromochloromethane (DBCM) and chloroform. HAA5 is a
group of five regulated HAAs: monochloroacetic acid (MCAA), dichloroacetic acid (DCAA),
trichloroacetic acid (TCAA), monobromoacetic acid (MBAA) and dibromoacetic acid (DBAA).
Under the Stage 1 and Stage 2 D/DBPRs, MCLGs were established for all four THMs listed
above and for three of the five HAAs (MCAA, DCAA and TCAA).
Data from animal toxicity studies were used as the basis for establishing the MCLGs, whereas
data from the epidemiology studies were used as the basis for estimating risk reduction
associated with implementation of the rule, specifically, for the reduction of bladder cancer. Both
the animal toxicity and the epidemiology sections of this chapter focus on cancer effects and
reproductive/developmental effects. These endpoints were used to evaluate the risks associated
with exposure to the DBPs listed above. The intent of the Stage 1 and Stage 2 D/DBPRs was to
reduce human exposure not only to these nine substances but also to mixtures of DBPs formed
during disinfection of water. Reduction of human exposure to the nine substances and associated
mixtures is achieved through compliance with both the MCLs and the treatment technique (TT)
component of the Stage 1 D/DBPR. The TT is aimed at reducing precursors, such as total
organic carbon (TOC), found in source waters that leads to the formation of a vast array of
organic DBPs. This is discussed further in Chapters 6 and 7 of this support document.
4.1.1 Toxicity Studies
The relevant information from animal toxicity studies of the regulated organic DBPs is presented
in three main sections addressing: THMs, HAAs; and the mode of action relevant to
carcinogenicity.
Six-Year Review 3 Technical Support Document	4-1
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
At the time of the promulgation of the Stage 1 and Stage 2 D/DBPRs, there was a considerable
amount of information on carcinogenicity of the regulated THM and HAA DBPs based on
animal toxicity studies. Information about the reproductive and developmental effects of the
contaminants also featured prominently in the Stage 2 portion of the rule and was to a lesser
extent supported by the animal data for some of the DBPs.
EPA's Integrated Risk Information System (IRIS) conducted weight of evidence
characterizations of the carcinogenic potential for six of the regulated DBPs: four THMs -
bromoform (USEPA, 1991), BDCM (USEPA, 1993a), chloroform (USEPA, 2001a), DBCM
(USEPA, 1992a) and two HAAs - DCAA (USEPA, 2003c) and TCAA (USEPA, 201 la). Cancer
risk factors were developed for bromoform, BDCM and DBCM in support of the Stage 1
D/DBPR based on EPA's 1986 Guidelines for Carcinogen Risk Assessment (USEPA, 1986). In
2003, the cancer risk factor for DCAA was published. (USEPA, 2003c). The IRIS documents for
chloroform and DCAA were not available for the Stage 1 rule but were available for the Stage 2
rule. The TCAA IRIS document was completed after issuance of the Stage 2 rule.
Additional documentation of the carcinogenicity of bromoform, BDCM and DBCM was
evaluated for the Stage 2 rule by EPA following the 2005 EPA Guidelines for Carcinogenic Risk
Assessment resulting in changes to the information on IRIS (USEPA, 2005a). A cancer risk
factor was not derived for chloroform under the Stage 1 or Stage 2 rule based on the data that
demonstrated that the mode of action for cancer is nonlinear and the cancer classification is not
likely for doses below those leading to tissue necrosis and likely at doses that cause necrosis. For
that contaminant, the protection afforded by the Reference Dose (RfD) based on liver necrosis is
considered to also protect for cancer (USEPA, 2001a).
The MCLGs for MCAA and TCAA were supported by Criteria Documents developed by EPA
(USEPA, 2005b, 2005c). TCAA was classified as having "suggestive evidence of
carcinogenicity" supporting use of the RfD as the endpoint that is the basis of the MCLG.
MCAA was classified as having insufficient information to assess carcinogenic potential. The
most recent IRIS TCAA assessment (USEPA, 201 la) was completed after the Stage 2 rule and
replaces the EPA Criteria Document that supported the rule. The IRIS assessment characterizes
the evidence for carcinogenicity as suggestive and provides a quantitative assessment of risk
based on an EPA study that had not been published at the time of the Stage 2 Rule.
4.1.1.1 Trihalomethanes (THMs)
This section describes the basis for the MCLGs for the four regulated THMs that compose
THM4, new information that has become available since the development of the Stage 2
D/DBPRs and observations about its relevance within the context of the SYR.
Prior to the Stage 1 and Stage 2 D/DBPRs, the National Toxicology Program (NTP) completed
two-year cancer bioassays in rats and mice for chloroform (NCI, 1976), DBCM (NTP, 1985),
BDCM (NTP, 1987), and bromoform (NTP, 1989a).
An overview of relevant studies is provided in the following subsections. Additional information
on the toxicological background at the time of Stage 1 and Stage 2 regulations for cancer,
mutagenicity/genotoxicity and reproductive/developmental effects is provided in Appendix A.
Six-Year Review 3 Technical Support Document	4-2
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
4.1.1.1.1	Bromoform
Basis for the MCLG
In Stage 1 D/DBPR, EPA established an MCLG of zero for bromoform and classified
bromoform as a "probable human carcinogen" (USEPA, 1991, 1998b) based on a weight of
evidence evaluation of both the cancer and noncancer effects. Under the 2005 cancer guidelines
it was classified as "likely to be carcinogenic by all routes of exposure" (USEPA, 2005d). The
MCLG is based on a chronic animal carcinogenicity study that reported uncommon neoplasms of
the large intestines in rats (NTP, 1989b). Insufficient evidence exists regarding the mode of
carcinogenic action of bromoform, therefore, the low-dose extrapolation approach was used to be
protective of public health (USEPA, 1998b). The RfD of 0.02 mg/kg/day is based on a No-
Ob served-Adverse-Effect-Level (NOAEL) of 25 mg/kg/day from subchronic data for hepatic
lesions in male rats (NTP, 1989b) with the application of an uncertainty factor of 1000 (USEPA,
2005d).
New Information Available Since Development of Stage 2 D/DBPR
There is no new, relevant animal toxicity data for bromoform that would change its MCLG,
cancer quantification or RfD.
Relevance for SYR
No new data were identified that would change the MCLG of zero for bromoform.
4.1.1.1.2	Bromodichloromethane
Basis for the MCLG
In the Stage 1 D/DBPR, EPA established an MCLG of zero for BDCM and classified BDCM as
a "probable human carcinogen" (USEPA, 1993a, 1998b) based on a weight of evidence
evaluation of both the cancer and noncancer effects. EPA later classified BDCM as "likely to be
carcinogenic by all routes of exposure" (USEPA, 2005d) following the new cancer guidelines.
The MCLG of zero was assigned based on intestine and kidney tumor data from a chronic animal
carcinogenicity study (NTP, 1987). The low-dose extrapolation approach was used to estimate
cancer risk since there was insufficient evidence regarding the mode of action of BDCM
(USEPA, 1998b). The RfD presented on IRIS at the time of the Stage 1 D/DBPR (and which is
still currently on IRIS) is 0.02 mg/kg/day, based on a lowest observed adverse effect level
(LOAEL) of 17.9 mg/kg/day for renal cytomegaly in male mice (NTP, 1987) with the
application of an uncertainty factor of 1000 (USEPA, 1993 a). In support of the Stage 2 D/DBPR,
a criteria document for brominated THMs was used in which EPA derived an RfD of 0.003
mg/kg/day for BDCM based on degeneration of the liver in a 24-month dietary study in rats
(USEPA, 2005d, 2006a). However, for the Stage 2 D/DBPR, EPA determined that there were no
new significant health effects data suggesting the need for a change in the categorization of
BDCM as a likely human carcinogen nor for a change in the MCLG of zero (USEPA, 2003 d,
2006a).
Six-Year Review 3 Technical Support Document	4-3
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
New Information Available Since Development of Stage 2 D/DBPR
Cancer
NTP conducted a bioassay of BDCM with 50 F344N male rats and 50 B6C3F1 female mice for
each of four dose groups, using drinking water as the exposure route (NTP, 2006). The animals
were given water with BDCM concentrations of 0, 175, 350 or 700 mg/L (an estimated average
daily doses of 0, 6, 12 and 25 mg/kg respectively in rats and an estimated daily dose of 0, 9, 18
or 36 mg/kg to mice) for two years. The drinking water studies were limited to male rats and
female mice because of their sensitivity to develop neoplasms when administered BDCM by
gavage in corn oil. In the 1987 earlier gavage study there was clear evidence of cancer for both
male and female mice and rats.
Cancers or neoplastic lesions did not occur more frequently in the treated animals as a result of
exposure to BDCM in drinking water. No tumors were found in the exposed animals at levels
significantly greater than the controls. NTP concluded that BDCM in the drinking water did not
cause cancer in male rats or female mice. Toxic effects of BDCM in drinking water for male rats
included chronic inflammation. These results differed from those in the earlier corn oil gavage
study (NTP, 1987)
In 2006, Health Canada's Guideline for Canadian Drinking Water Quality: Technical Document
for THMs was published and included a maximum acceptable concentration (MAC) of 0.016
mg/L for BDCM in drinking water based on a cancer endpoint using a NTP (1987) study as the
key study and was designated as a "not-to-exceed" value as a precaution against potential
adverse reproductive effects (Health Canada, 2006). Health Canada determined that an approach
based on cancer endpoints is likely to be protective of non-cancer effects, including
reproductive/developmental effects.
In 2009, Health Canada withdrew its 2006 guideline for BDCM based on an expert panel
assessment of the NTP (2006) cancer bioassay. At the time that the expert panel was
commissioned by Health Canada, BDCM was classified in Group II: probably carcinogenic to
humans with sufficient evidence in animals and inadequate evidence in humans (Health Canada,
1994). The expert panel concluded that "The evidence from a lifetime study of [BDCM] given to
rodents by corn oil gavage is that it is an animal carcinogen. However, the second NTP lifetime
study, which tested lower doses of [BDCM] in drinking water does not support this
determination. The combined data from the two studies do not support a linear dose response. "
(Health Canada, 2008a). The expert panel concluded that the NTP (2006) drinking water study
calls into question the weight of evidence that BDCM is "probably" carcinogenic in humans, but
Health Canada's approach is to assume there is no safe exposure level. The panel pointed out
that, in shorter term studies in which BDCM was administered to rats in drinking water, aberrant
crypt foci developed in the rat large intestine and suggests that BDCM may play a role in
carcinogenesis. The panel stated that. . the possibility that a mutagenic mode of action
contributes to the carcinogenic effects of BDCM in the intestine cannot be dismissed. Therefore,
the Panel concluded that the NTP (2006) data alone are not sufficient to change the
classification to "possibly " carcinogenic in humans. "
Six-Year Review 3 Technical Support Document	4-4
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
The Health Council of the Netherlands also decided that BDCM should be considered
carcinogenic to humans and that BDCM acts by a stochastic genotoxic mechanism that is
governed by a sequence of random events (Health Council of the Netherlands, 2007). Their
recommendation corresponds to the EU classification of Group 2B: possible human carcinogen.
The NTP studies of genetic toxicology found positive results in a mouse lymphoma assay and a
small increase in sister chromatid exchanges in the presence of S9, but negative results in Ames
Salmonella Assays, Chinese Hamster Ovary Cells and for the vivo bone marrow micronuclei
assay in mice.
Reproductive/Developmental
Bielmeier et al. (2007) investigated BDCM-induced pregnancy loss in F344 rats using ex vivo
and in vitro techniques. Using ex vivo techniques, BDCM-exposed corpora lutea showed
increased progesterone secretion compared to controls, perhaps reflecting a rebound effect. In
vitro exposure to BDCM reduced luteal progesterone secretion in response to stimulation by
human chorionic gonadotropin (hCG), an analog of luteinizing hormone (LH). In earlier studies
(Bielmeier et al., 2001, 2004, see Appendix A for further elaboration), a LOAEL of 75
mg/kg/day was identified in F344 rats for full litter resorption. The ability of hCG to prevent
BDCM-induced pregnancy loss suggests an effect of BDCM on maternal LH secretion, while not
ruling out a possible effect of BDCM on luteal responsiveness. These findings suggest that
BDCM disrupts pregnancy in F344 rats via two modes: disruption of LH secretion and
diminished luteal responsiveness to LH.
The Health Canada BDCM expert panel concluded that adverse reproductive and developmental
effects of BDCM were observed only at very high, maternally toxic doses, were not consistent
between animal models and varied with method of administration (Health Canada, 2008a). The
panel stated that the weight of evidence did not support an association between adverse
reproductive and developmental effects and exposure to BDCM at levels found in drinking
water. Data are limited on potential mode(s) of action related to BDCM and adverse reproductive
and developmental toxicity.
Relevance for SYR
The outcome from the NTP (2006) study, the deliberations of the Health Canada Scientific panel
(2008a) and the Health Council of the Netherlands (2007) are relevant to the SYR of the MCLG
for BDCM. The findings from the animal studies as well as recent mechanistic data and
epidemiology findings each contributed to the review deliberations. In addition, new
pharmacokinetic information about BDCM, described in 4.1.1.3 (mode of action information
relevant to DBP carcinogenicity), is important when considering the impact of the new data on
the whether the current MCLG of zero is appropriate.
Six-Year Review 3 Technical Support Document	4-5
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
4.1.1.1.3	Dibromochloromethane
Basis for the MCLG
In the Stage 1 D/DBPR, an MCLG of 0.06 mg/L for DBCM was established by EPA based on a
weight of evidence evaluation of both the cancer and noncancer effects. At that time DBCM was
classified as a "possible human carcinogen" (USEPA, 1992a, 1998b). The MCLG was based on
the RfD, an adult tap water consumption of 2 liters/day for a 70 kg adult, and an additional risk
management factor of 10 to account for possible carcinogenicity. The assumed drinking water
contribution to total exposure was 80 percent (USEPA, 1994a). At the time of the Stage 2 Rule
an RfD of 0.02 mg/kg/day was derived based on a NOAEL of 30 mg/kg/day (adjusted to 21.4
mg/kg/day for a 5-day/week exposure) for liver effects from the subchronic portion of a NTP
(1985) study in rats and an uncertainty factor of 1000 (USEPA, 2005d). EPA used the chronic
studies of the NTP (1985) study to determine a classification of "suggestive evidence for cancer"
(USEPA, 2005d). No evidence of carcinogenicity was reported in rats, but there was equivocal
evidence of carcinogenicity in male mice and some evidence of carcinogenicity in female mice
based on an increased incidence of liver tumors. The RfD value did not change due to the lack of
significant new health effects data. EPA did not revise the MCLG for DBCM in the Stage 2
D/DBPR (USEPA, 2003d, 2006a).
New Information Available Since Development of Stage 2 D/DBPR
No new, relevant animal toxicity information was found for DBCM.
Relevance for SYR
There are no new data relevant to the SYR of the MCLG for DBCM.
4.1.1.1.4	Chloroform
Basis for the MCLG
In the Stage 1 D/DBPR, EPA finalized an MCLG of zero for chloroform based on a weight of
evidence evaluation of both the cancer and noncancer effects and classified chloroform as a
"likely human carcinogen" (USEPA, 1994a, 1998b). The MCLG was based on linear default
extrapolation until EPA completed additional deliberations with the Agency's Science Advisory
Board (SAB) on the scientific basis of the mode of action for chloroform (USEPA, 1998b). At
the same time the Agency identified 0.07 mg/L as the MCLG in a situation where a non-linear
approach was used in the evaluation of the cancer endpoint (USEPA, 1998b). For the Stage 2
D/DBPR, EPA proposed an MCLG for chloroform of 0.07 mg/L and then finalized the MCLG
of 0.07 mg/L in 2006 based on the SAB's conclusions that the nonlinear approach is most
appropriate for the risk assessment for chloroform (USEPA, 2003d, 2006a). The MCLG is based
on an RfD of 0.01 mg/kg/day, derived using a benchmark dose level (BMDL) of 1.2 mg/kg/day
for liver necrosis in dogs (Heywood et al., 1979) with an uncertainty factor of 100, adult tap
water consumption of 2 liters/day for a 70 kg adult and a relative source contribution of 20
percent for drinking water exposure (USEPA, 2006a). EPA concluded that chloroform is "likely
to be carcinogenic to humans " only under high exposure conditions that lead to cytotoxicity and
Six-Year Review 3 Technical Support Document	4-6
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
regenerative hyperplasia and that chloroform is "not likely to be carcinogenic to humans" under
conditions that do not cause cytotoxicity and cell regeneration.
New Information Available Since Development of Stage 2 D/DBPR
There was no new, relevant animal toxicity information found for chloroform.
Relevance for SYR
There are no new data relevant to the SYR for chloroform.
4.1.1.2 Haloacetic acids (HAAs)
This section describes the basis for the MCLGs for the five HAAs that comprise HAA5
(monochloroacetic acid, dichloroacetic acid and trichloroacetic acid plus monobromoacetic acid
and dibromoacetic acid). New information that has become available since the development of
the Stage 2 D/DBPR is relevant within the context of the SYR and is described below. Available
health effects information about four additional HAAs (not part of HAA5) is provided in Section
4.4.
EPA completed a health criteria document for brominated acetic acids (USEPA, 2005d) for the
Stage 2 Rule in which monobromoacetic acid (MBAA), bromochloroacetic acid (BCAA, not part
of HAA5) and dibromoacetic acid) were all identified as "not classifiable as to human
carcinogenicity" under the 1986 Carcinogen Risk Assessment Guidelines and "inadequate for an
assessment of human carcinogenic potentiaF under the 1999 Draft Guidelines for Carcinogen
Risk Assessment.
An overview of new studies is provided in the following subsections. Additional information on
the toxicological background at the time of Stage 1 and Stage 2 regulations for cancer,
mutagenicity/genotoxicity and reproductive/developmental effects is provided in Appendix A.
4.1.1.2.1 Monochloroacetic acid
Basis for the MCLG
In the Stage 1 D/DBPR, EPA did not set an MCLG for MCAA due to the lack of available health
data (USEPA, 1994a, 1998b). In the Stage 2 D/DBPR, EPA proposed an MCLG of 0.03 mg/L
and finalized an MCLG of 0.07 mg/L (USEPA, 2003d, 2005b, 2006a). The final MCLG was
based on an RfD of 0.01 mg/kg/day, using a NOAEL of 3.5 mg/kg/day for decreased body
weight, kidney, liver and spleen weights in rats (DeAngelo et al., 1997) with an uncertainty
factor of 300, adult tap water consumption of 2 liters/day for a 70 kg adult and a relative source
contribution of 20 percent for drinking water exposure (USEPA, 2005b, 2006a). The USEPA
(2005b) classified MCAA as having inadequate data to support a finding on its carcinogenicity.
Six-Year Review 3 Technical Support Document	4-7
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
New Information Available Since Development of Stage 2 D/DBPR
Cancer
Health Canada (2008b) considers MCAA unlikely to be a carcinogen to humans based on lack of
evidence. Health Canada developed a Tolerable Daily Intake (TDI) of 0.0117 mg/kg/day,
equivalent to EPA's RfD of 0.01 mg/kg/day for MCAA. The TDI is based on the same study
(DeAngelo et al., 1997), the same NOAEL of 3.5 mg/kg/day and the same uncertainty factor of
300 as used by EPA.
Relevance for SYR
There are no new data relevant to the SYR of the MCLG for MCAA.
4.1.1.2.2	Dichloroacetic acid
Basis for the MCLG
In Stage 1 D/DBPR, EPA established an MCLG of zero for DCAA based on a weight of
evidence evaluation of both the cancer and noncancer effects and classified DCAA as a
"probable or likely human carcinogen" (USEPA, 1994a, 1998b). The MCLG was based on
several studies showing liver tumors in mice and rats from lifetime exposure to DCAA in
drinking water. Insufficient evidence existed regarding the mode of carcinogenic action of
DCAA; the low-dose extrapolation approach was used to be protective of public health (USEPA,
1998b). The RfD of 0.004 mg/kg/day was based on a LOAEL of 12.5 mg/kg/day for effects on
the liver, brain and testis in dogs (Cicmanec et al., 1991) with the application of an uncertainty
factor of 3000 (USEPA, 1994a, 2003c). EPA did not revise the MCLG for DCAA in Stage 2
D/DBPR (USEPA, 2003d, 2005f, 2006a).
New Information Available Since Development of Stage 2 D/DBPR
Health Canada considers DCAA to be a probable human carcinogen based on the cancer studies
which resulted in liver tumors in rats and mice (Health Canada, 2008a). No new, relevant animal
toxicity information was found for DCAA that would alter the MCLG of zero.
Relevance for SYR
There are no data to suggest a change in the MCLG for DCAA.
4.1.1.2.3	Trichloroacetic acid
Basis for the MCLG
In the Stage 1 D/DBPR, EPA established an MCLG of 0.3 mg/L for TCAA (USEPA, 1994a,
1998b) based on developmental toxicity and limited evidence of carcinogenicity in animals. In
the Stage 2 D/DBPR, EPA proposed and finalized an MCLG of 0.02 mg/L for TCAA (USEPA,
2003d, 2005c, 2006a) derived from an RfD of 0.03 mg/kg/day, using a NOAEL for liver
histopathological changes in rats (DeAngelo et al., 1997), an uncertainty factor of 1000, an
Six-Year Review 3 Technical Support Document	4-8
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
additional risk management factor of 10 to adjust for "suggestive evidence of carcinogenicity
This MCLG was based on this RfD, using adult tap water consumption of 2 liters/day for a 70 kg
adult and a relative source contribution of 20 percent for drinking water exposure (USEPA,
2005c, 2006a).
New Information Available Since Development of Stage 2 D/DBPR
Cancer
EPA's IRIS program completed an assessment for TCAA after the completion of the Stage 2
D/DBPR (USEPA, 201 la). According to EPA's 2005 Guidelines for Carcinogen Risk
Assessment, EPA classified TCAA as having "suggestive evidence of carcinogenic potentiaF
(USEPA, 2005c). This classification was based on significantly increased incidence of liver
tumors in male and female B6C3F1 mice exposed via drinking water (DeAngelo et al., 2008;
Bull et al., 1990; Bull, 2002; Pereira, 1996; Herren-Freund et al., 1987) and a lack of treatment-
related tumors in male F344/N rats exposed via drinking water (DeAngelo et al., 1997).
As was the case with the EPA assessment, Health Canada (2008b) considers TCAA to be a
possible carcinogen based on liver tumors in mice and the lack of tumors in the male rat
(DeAngelo et al., 1997).
The USEPA (201 la) assessment established an RfD for TCAA that differs from that used in the
Stage 2 rule. The IRIS RfD of 0.02 mg/kg/day is based on a 95 percent lower confidence level on
the modeled benchmark dose for a 10 percent decrease (BMDLio) in liver necrosis in the treated
B6C3F1 mice of 18 mg/kg/day (DeAngelo et al., 2008). The study used a drinking water route of
exposure over a 60-week period. The study was published after the Stage 2 Rule.
Reproductive/Developmental
The following reproductive/developmental studies were considered in the IRIS report (USEPA,
201 la) as well as three older studies that are summarized in Appendix A.
Singh et al. (2005a, 2005b, 2006) conducted a reproductive/developmental study on inbred
Charles Foster rats that were administered TCAA via gavage on gestational days (GD) 6 through
15 at doses up to 1800 mg/kg-daily. Effects reported included decrease in maternal weight gain,
post implantation loss, reduction in mean testes weight and length of the seminiferous tubules,
reduced ovary weights, and effects on fetal brain. Maternal NOAELs and LOAELs of 1,000 and
1,200 mg/kg/day, respectively, and a developmental LOAEL of 1,000 mg/kg/day, the lowest
dose tested, were determined.
Warren et al. (2006) administered TCAA via gavage to pregnant Sprague-Dawley rats at 300
mg/kg/day on GDs 6 through 15. Mean fetal body weights were significantly reduced at this
dose, but, no statistically significant effects were noted on fetal eye development. A
developmental LOAEL of 300 mg/kg/day was determined. When Smith et al. (1989b) treated
pregnant Long-Evans rats with TCAA on GDs 6 through 15 by oral intubation at doses of 0, 330,
800, 1200 and 1800 mg/kg/day, orbit malformations were significantly increased in fetuses at
Six-Year Review 3 Technical Support Document	4-9
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
doses of 1200 and 1800 mg/kg/day. A significant increase in embryo lethality (resorbed
implants) was also observed at doses > 800 mg/kg/day.
Relevance for SYR
The new IRIS RfD and updated quantification for the cancer slope factor have the potential to
change the MCLG for TCAA. The MCLG may be derived from the noncancer RfD with
consideration of the cancer data in determination of the risk management factor applied to
Category C and suggestive carcinogens.
4.1.1.2.4	Monobromoacetic acid
Basis for the MCLG
In Stage 1 and 2 D/DBPRs, EPA did not set an RfD or MCLG for MBAA due to lack of data on
the dose-response for relevant health effects (USEPA, 1998b). Accordingly, there is no MCLG.
New Information Available Since Development of Stage 2 D/DBPR
Cancer
Health Canada (1994) considered bromoacetic acid as unclassifiable with respect to
carcinogenicity in humans based on inadequate data from animal studies and has retained this
finding based on the USEPA (2005e) assessment of "inadequate for assessment of human
carcinogenicpotentiaF (Health Canada, 2008b).
Relevance for SYR
Bromoacetic acid currently lacks an MCLG because that data were considered inadequate to
support development of an RfD or cancer classification. No new animal toxicity data were
identified under the SYR that would change this finding.
4.1.1.2.5	Dibromoacetic acid
Basis for the MCLG
In Stage 1 and 2 D/DBPRs, EPA did not set an RfD or MCLG for DBAA due to lack of
appropriate data on the dose-response for relevant health effects (USEPA, 1998b, 2005e).
New Information Available Since Development of Stage 2 D/DBPR
Cancer
NTP administered DBAA in drinking water to male and female F344/N rats and B6C3F1 mice at
daily doses up to 40 and 45 mg/kg/day in male and female rats, respectively, and 87 and 65
mg/kg/day in male and female mice, respectively (NTP, 2007c). Drinking water concentrations
were the same for males and females; the doses vary with difference in drinking water intakes
and body weights. At the end of the study, tissues from more than 40 sites were examined from
every animal. Survival was similar for animals receiving DBAA and the controls. Male rats
Six-Year Review 3 Technical Support Document	4-10
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
receiving DBAA had significantly increased rates of malignant mesotheliomas. The rates of
mononuclear cell leukemia increased in exposed female rats and, to a lesser extent, in exposed
male rats. Male and female mice exposed to DBAA had increased rates for a variety of liver
tumors; lung tumors were increased in male mice and, to a lesser extent, in female mice. NTP
concluded that there was some evidence of carcinogenic activity for mesothelioma in male rats
and mononuclear cell leukemia in female rats when administered DBAA in drinking water. NTP
concluded that there was clear evidence of carcinogenic activity based on increased incidences of
hepatocellular neoplasms in male and female mice and hepatoblastoma in male mice. An
increased incidence of lung cancer in male mice was also considered to be exposure related, and
a slight increase in lung cancer in female mice may have been related to exposure to DBAA.
Health Canada considers DBAA to be a probable carcinogen to humans based on tumors found
in several organs in rats and mice after exposure to DBAA (Health Canada, 2008b).
Mutagenicity/Genotoxicity
Positive results were reported on micronuclei formation in the blood of male mice, but not
female mice in a 13-week study on DBAA in drinking water (NTP, 2007c).
Reproductive/Developmental
In a related study, NTP conducted a 13-week study in B6C3F1 mice and F344 rats. In that study,
DBAA was administered in drinking water at concentrations of 0, 125, 250, 500, 1,000 and
2,000 mg/L, which resulted in average daily doses of approximately 10, 20, 40, 90 and 166
mg/kg/day in male rats and 16, 30, 56, 155 and 230 mg/kg/day in male mice (NTP, 2007c).
Adverse effects in male rats included retained spermatids at 40 and 90 mg/kg/day and decreased
testis weights and testicular atrophy at 166 mg/kg/day. The NOEL for testicular effects was 20
mg/kg/day in rats. In mice, the incidence of abnormal testicular morphology was significantly
increased at 115 and 230 mg/kg/day, with a NOEL for testicular effects of 56 mg/kg/day.
Relevance for SYR
The 2007 NTP study on DBAA showed clear evidence of carcinogenicity in male and female
mice and some evidence of carcinogenicity in male and female rats. These data suggest the need
for a new assessment for DBAA (NTP, 2007c).
4.1.1.3 Mode of Action Information Relevant to DBP Carcinogenicity
This section provides a summary of the studies that describe modes of action (MO A) of DBPs
that potentially lead to carcinogenicity.
4.1.1.3.1 Overview
The mode of action relates to the genotoxicity (i.e., causing DNA damage or mutation) of the
DBPs. DNA damage that is not repaired correctly by the cell can cause loss of cellular viability,
or when the cell survives, can result in clonal replication of cells carrying a DNA error, that is, a
mutation. Mutations and epigenetic changes (alterations in gene expression) are considered two
of the main genetic events that lead to tumor formation when clonal expansion of the altered cell
Six-Year Review 3 Technical Support Document	4-11
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
occurs leading to tumor growth. Thus, structural changes to DNA have the potential to lead to
damaged cells which then have the potential to form tumors. When some of the mutated tumor
cells migrate to new tissue locations the tumors are said to metastasize.
An understanding of mode of action is important when assessing whether effects observed in in
vitro assays or in experimental animals could apply to human exposures (Humpage, 2012).
Ideally, the first step in determining a MOA is to delineate the pharmacokinetics associated with
exposure to a substance, so that the relationship between external dose and the systemic
concentrations that produce an effect can be understood. Although this level of quantitative dose-
response has not been defined for most DBPs, many studies are available which describe both
MO As (changes that occur at the cellular level) and mechanism of action (changes that occur at
the molecular level) associated with exposure to DBPs.
A significant number of DBPs, both regulated and unregulated, including halofuranones,
brominated trihalomethanes (BrTHMs), brominated HAAs, haloacetonitriles, haloaldehydes,
haloketones and halonitromethanes (HNMs) can induce gene mutations (Richardson et al., 2007;
Kundu et al., 2004; Bull, 2011). Some of these DBPs are direct-acting mutagens (e.g., MX) and
some require metabolic activation (e.g., BrTHMs). Human polymorphic expression of enzymes
that are involved in the mutagenic activation or detoxification of DBPs can apparently affect
cancer risks associated with DBP exposure (Cantor et al., 2010; see Appendix A for detailed
summary of paper). The role of these enzymes will be discussed further in Section 4.1.1.3.2. The
available carcinogenicity data for the unregulated DBPs are in Section 4.4 of this chapter.
Some of the nongenotoxic MO As that have been associated with DBP exposures include the
following:
(1)	Reparative hyperplasia occurs in an effort to replace dead cells. As the rate of cell
division increases with the number of dead cells that require replacement, the DNA
replication errors increase proportionally. One hypothesis is that the tumorigenic effect of
some DBPs (e.g., chloroform) is strongly influenced by necrosis and reparative
hyperplasia, which generally occur only after exposure to high doses of these DBPs
meaning that the mode of action is nonlinear.
(2)	DNA methylation is an epigenetic mechanism that down-regulates genes without
changing their coding sequence and therefore plays an important role in DNA repair. It
modulates gene transcription and is key to histone acetylation and chromosomal stability.
DNA hypomethylation may contribute to chromosome instability and aberrant gene
expression via a nongenotoxic route (Baylin et al., 1998). The following results suggest
that DNA hypomethylation may be involved in the carcinogenic mechanism of DBPs in
the kidney, liver and colon of rodents.
a. DNA hypomethylation was associated with the induction or promotion of mouse
liver tumors by DCAA and TCAA (Tao et al., 2004a). Tao et al. (2004b) reported
that DBAA caused DNA hypomethylation in mouse liver which corresponds with
its carcinogenic and tumor promoting activity (Melnick et al., 2007).
Six-Year Review 3 Technical Support Document	4-12
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
b.	Chloroform, BDCM, DCAA and TCAA induced DNA hypomethylation in mouse
and/or rat kidney which corresponded to their carcinogenic and/or tumor
promoting activity, indicating epigenetic activity (Tao et al., 2005; USEPA,
2005d; USEPA, 2011a).
c.	When administered by gavage or in drinking water, BDCM induced DNA
hypomethylation in the colon of male F344 rats, but did not decrease DNA
methylation in the colon of male B6C3F1 mice (Pereira et al., 2004a). BDCM
also induced tumors in the colon of male and female F344 rats but not male or
female B6C3F1 mice when administered by gavage in corn oil (NTP, 1987).
Administration of BDCM in drinking water did not result in detectable increases
in colon tumors in male F344 rats (females not tested) or female B6C3F1 mice
(males not tested) when administered in drinking water (NTP, 2006). Neither of
the NTP studies evaluated the methylation status of DNA.
d.	In a separate study, Pereira et al., (2004b) examined whether supplying
methionine, an important source of methyl groups for transmethylation reactions,
to DCAA-treated mice in their drinking water would reduce the number of altered
pretumor liver foci. At the low methionine dose there was an increase in the
number of altered hepatocyte foci compared to controls, while at the higher
methionine dose the number of foci was decreased, supporting the hypothesis that
the availability of methyl groups from methionine for DNA methylation could be
important. After 44 weeks, the livers of the treated animals were removed and
examined for adenomas. The number of adenomas was decreased at both
methionine doses and there was a methionine dose related increase in DNA
methylation. However, some liver adenomas were still present. The results of this
study suggested that high dietary methionine slowed the progression of foci to
tumors. However, methionine did not totally remove the cancer risk leaving an
opportunity for other operative MO As.
(3)	Peroxisome proliferation appears to play a role in the development of liver tumors in
animals treated with TCAA by a nongenotoxic mode of action as demonstrated in a
number of long-term exposure studies in both rats and mice. Induction of liver tumors by
PPARa agonists incorporates the following key events: PPARa ligands activate PPARa
and subsequently cause an increase in hepatic peroxisomes, cell cycling/apoptosis and
lipid metabolism. These changes lead to perturbations in cell proliferation and apoptosis.
Suppression of apoptosis coupled with increased cell proliferation allows DNA-damaged
cells to persist and proliferate, resulting in preneoplastic hepatic foci and ultimately in
tumors from cells damaged by other MOAs (USEPA, 201 la).
(4)	Pals et al. (2011) and Dad et al. (2013) proposed that the monohaloacetic acids,
especially monoiodoacetic acid (MIAA), could indirectly induce DNA damage by
inhibition of glyceraldehyde-3-phosphate dehydrogenase (GAPDH), leading to a severe
reduction in cellular ATP levels by repressing the generation of pyruvate and reducing
aerobic ATP generation by way of the citric acid cycle. The hypothesis was tested in
cultured Chinese hamster ovary (CHO) cells with measurements made for cellular
Six-Year Review 3 Technical Support Document	4-13
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
genomic damage (Comet Assay) and reductions in GAPDH activity. The results indicated
that MIAAhad the greater effect on the enzyme activity followed by MBAA and MCAA.
The degree of genomic damage was correlated with the inhibition of the enzyme with
MIAA displaying the greatest toxicity and GAPDH inhibition. A loss of pyruvate leads to
mitochondrial stress, production of reactive oxygen species (ROS) and genomic DNA
damage (Pals et al., 2011; Dad et al., 2013. The inhibition of GAPDH was linked to
alkylation of a cysteine in the active site of the enzyme causing a resultant downstream
reduction in the production of pyruvate from the glycolysis pathway and thereby ATP
production via the citric acid cycle leading to cytotoxicity and generation of ROS (Dad et
al., 2013). The hypothesis was tested examining the genomic effects with and without the
addition of pyruvate to the culture. The presence of pyruvate in the culture reduced the
genomic damage as measured by the Comet Assay.
The available animal bioassay data on THMs and HAAs described above have shown some
evidence of cancer in several organs including kidney, liver, colon, large intestine, lung,
mammary gland, pancreas, mesothelioma and blood, but no evidence of bladder cancer in animal
studies. However, due to differences in physiology, metabolism and urinary pH, rodents are
generally not considered to be good models for human bladder carcinogenesis (Crallan et al.,
2006).
4.1.1.3.2 Role of Human Genetic Polymorphisms in DBP-Metabolizing Enzymes
The genotypes of some enzymes involved in the metabolism of THMs and HAAs have been
characterized and are associated with increased risk of bladder cancer in humans consuming
chlorinated drinking water (Cantor et al., 2010; see section 4.1.2.1.1 for further elaboration).
These enzymes are involved in the metabolism of many compounds, including DBPs other than
THMs and HAAs. The genes of the glutathione S-transferase (GST) superfamily of genes encode
for multifunctional enzymes that conjugate a compound or its metabolite with glutathione (GSH)
and are important in the detoxification of electrophilic molecules, including some carcinogens,
mutagens and therapeutic drugs. However, certain GST's can also activate some haloalkanes to
DNA-reactive intermediates (Thier et al., 1993, Proc. Natl. Acad. Sci. USA 90: 8576; Pegram et
al., 1997). Several genes that code for these enzymes are polymorphic, with specific genotypes
that exhibit an association with an increased cancer risk (Curran et al., 2000; Cantor et al., 2010).
Glutathione S-transferase theta-1 (GSTT1) is an enzyme that is encoded by the GSTT1 gene.
GSTT1 catalyzes the conjugation of reduced GSH to a variety of electrophilic and hydrophobic
compounds that can result in the production of DNA-reactive metabolites. For example, GSTT1-
mediated conjugation of GSH with BDCM produces an unstable GSCHCh conjugate that can
react with DNA or degrade to at least two additional DNA-reactive metabolites (Ross and
Pegram, 2003). GSTT1 is expressed in several tissues of rodents and humans and is
polymorphically expressed in human populations, with some individuals having a null genotype
(Ross and Pegram, 2004). GSTT1 is expressed in people with GSTT1(+) genotypes, which could
increase cancer risk from BrTHMs by increasing formation of mutagenic intermediates. People
with the GSTT1 null genotype have no GSTT1 activity, and therefore there is no activation of
known chemical substrates (such as BrTHMs) by this enzyme. In the Cantor et al. (2010) study,
associations between THM exposure and bladder cancer were stronger among subjects who had
the GSTT1(+) genotype. People with the GSTT1 null genotype had no increased bladder cancer
Six-Year Review 3 Technical Support Document	4-14
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
risk. (GSTT1 was considered "null" if a deletion was found in both copies of the gene; it was
considered "present" if neither or only one copy of the gene had a deletion.) BrTHMs, which had
tested negative for mutagenic activity in previous assays, were found to be mutagenic after
activation by GSTT1 in a transgenic strain of Salmonella (RSJ100) transfected with the GSTT1
gene (Pegram et al. 1997; DeMarini et al. 1997). In these studies, relative mutagenic potency
among BrTHMs was observed as follows: DBCM > bromoform > BDCM. GSTT1 activity is
abundant in the human urinary tract; Thier et al. (1998) reported renal activities approximately
twice that of the liver. Human hepatic GSTT1 is approximately equal in activity towards
electrophilic substrates compared to hepatic GSTT1 in rats, but lower than that of mice (Ross
and Pegram, 2003).
Glutathione S-transferase zeta-1 (GSTZ1) is the primary enzyme in the di-HAA metabolism
pathway (Anderson et al., 1999) and may also be involved in metabolism of brominated tri-
HAAs (Saghir and Schultz, 2005). It is encoded by the GSTZ1 gene and plays a key role in the
metabolism and clearance of HAAs. Single Nucleotide Polymorphisms (SNPs) for GSTZ1 result
in modified enzyme activity, including effects on the rate of biotransformation of di-haloacetic
acids (Board and Anders, 2011). There are four polymorphic variants of recombinant human
GSTZ with differing levels of inhibition by DCAA (Lantum et al. (2002). Cantor et al. (2010)
examined the effect of single nucleotide polymorphisms of the GSTZ1 gene on bladder cancer
incidence.
The single nucleotide polymorphism (SNP) for GSTZ1 considered by Cantor et al. (2010) had
the three genotypes of CC, CT and TT where C and T refer to the DNA bases cytosine and
thymine involved in the SNP. For their GSTZ1 analyses, Cantor et al. (2010) combined the
populations having either CT or TT genotypes, i.e, GSTZ1 CT/TT. Individuals with with
genotypes resulting in lower GSTZ1 activity (i.e., the GSTZ1 CT/TT group) were likely to have
higher sustained blood levels of HAAs and an increased bladder cancer risk in the Cantor et al.
(2010) study.
Cytochrome P450 2E1 (CYP2E1) is a member of the mixed function oxidase system and is
encoded by the CYP2E1 gene. CYP2E1 has a number of functions, including catalyzing the
primary oxidation of THMs leading to the formation of dihalocarbonyls (phosgene and its
brominated congeners), which rapidly degrade to carbon dioxide (the major oxidation product),
CO, and other minor end-products. Hepatic CYP2E1-mediated oxidation is the predominant
metabolic pathway for the THMs in rodent liver, especially as it related to chloroform (USEPA,
2001a). Although CYP2E1 is abundant in the rodent kidney (Ross and Pegram, 2004; Krajka-
Kuzniak et al., 2005), very little to no CYP2E1 activity has been found in the human kidney
(Amet et al., 1997; Baker et al., 2005). Additional pathways of parent THM metabolism, which
compete with the oxidative pathway, include reductive dehalogenation (CYP2B-mediated) and
GSH conjugation via GSTT1 leading to the formation of mutagenic intermediates. As indicated
above, the BrTHMs are much more likely to proceed through the genotoxic GSH conjugation
pathway than is chloroform, which proceeds predominately through the CYP2E1 pathway.
GSTT1-mediated conjugation of chloroform to GSH occurs only at very high chloroform
concentrations (Pegram et al., 1997). Cantor et al. (2010) examined the effect of single
nucleotide polymorphisms of the CYP2E1 gene on bladder cancer incidence. The single
nucleotide polymorphism (SNP) for CYP2E1 considered by Cantor et al. (2010) also had the
three genotypes of CC, CT and TT where C and T refer to the DNA bases cytosine and thymine
Six-Year Review 3 Technical Support Document	4-15
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
involved in the SNP. Similar to their GSTZ1 analyses, Cantor et al. (2010) combined the
CYP2E1 populations having either CT or TT genotypes, i.e, GSTZ1 CT/TT. They observed that
individuals with the CYP2E1 CC genotype had a higher incidence of bladder cancer.
Genetic polymorphisms in the genes that code for these enzymes have been studied for their
potential role in cancer susceptibility and drug response in humans. Notable, in the case-control
study by Cantor et al. (2010) mentioned above and discussed further in Section 4.1.2.1.1, a
subset of the cohort was used to investigate gene-environment interactions. Polymorphisms in
three GST genes (GSTT1, Glutathione S-Transferase Mu 1 (GSTM1) and GSTZ1), as well as in
CYP2E1 and N-acetyltransferase 2 (NAT2), were evaluated for possible association with risk of
bladder cancer from long-term exposure to DBPs in drinking water. DNA was extracted from
leukocytes or buccal cells for the genotype assays.
The association between genotypes and long-term THM exposures in humans was evaluated by
Cantor et al. (2010) to determine whether the odds ratios for various genotypes within quartiles
of THM exposure correlated with the risk for bladder cancer and whether THM odds ratios
within genotype categories differed significantly from each other. As indicated above, Cantor et
al. (2010) found that people with the GSTT1(+) genotype were at significantly greater risk for
developing bladder cancer than GSTTl-null subjects when exposed to THMs, as were those with
the GSTZ1 CT/TT or CYP2E1 CC polymorphisms. A potentially sensitive populations based on
individuals having both GSTTl-null and the GSTZ1 CT/TT polymorphism was also identified.
Cantor et al. (2010) acknowledged that while THMs are common components in disinfected
water, they may not be the most toxic or carcinogenic. Thus, it is possible that one or more of the
polymorphisms of interest could be acting through other substances whose occurrence is
correlated with THMs and explain the epidemiological associations between THMs in water and
bladder cancer in humans receiving disinfected water.
4.1.1.3.3 Mode of Action for the THMs
The following section describes the cancer MOA for the following THMs: chloroform, BrTHMs
and BDCM.
MOA for Chloroform
Chloroform produces cancer in the rodent liver and kidney by killing cells and not by a
genotoxic mechanism (Larson et al., 1996). Chloroform is metabolized by CYP2E1 and
produces phosgene, a toxic intermediate (Bull et al., 2012). Chloroform-induced tumors in the
rodent liver and kidney appear to be produced only at dose levels that result in repeated or
sustained cytotoxicity and regenerative cell proliferation from oxidative CYP2E1 metabolism
(USEPA, 2001a). As chloroform toxicity in the rodent liver and kidney becomes more severe,
the rate of cell division increases and stimulates the outgrowth of abnormal cells. The
toxicokinetic modeling to support this hypothesis is not available, therefore EPA's chloroform
mode of action assessment is based on chloroform's cytotoxicity leading to cellular necrosis.
Publications that became available after the Stage 2 rule suggest that the MOA for cancer linked
to chloroform could be more complex for the kidney. Tao et al. (2005) asserted that chloroform
caused synergistic DNA hypomethylation that increased with dose in combination with DCAA,
Six-Year Review 3 Technical Support Document	4-16
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
but not in combination with TCAA. Following sacrifice, the levels of renal DNA methylation of
the c-myc gene were measured. DNA was isolated from the kidney and methylation of DNA was
determined by dot-blot analysis and use of a monoclonal antibody for 5-methylcytosine. In male,
but not female, mouse kidneys, DCAA, TCAA and to a lesser extent chloroform decreased DNA
methylation. Coadministration of chloroform increased DCAA but not TCAA induced DNA
methylation. Tao et al. (2005) concluded that the correlation between the ability to promote
kidney tumors and the ability to induce DNA hypomethylation suggests that DNA
hypomethylation is involved in the carcinogenic mechanisms in the kidney with these DBPs.
MO A for Brominated THMs
The predominant pathway of BrTHM metabolism, like chloroform, is oxidation by CYP2E1,
producing dihalocarbonyl intermediates (X2CO) that can bind to macromolecules (especially
proteins) or be hydrolyzed to CO2 (the primary clearance mechanism). Reductive metabolism of
BrTHMs by CYPs, resulting in dihalomethyl radicals, is more extensive than for chloroform.
The oxidative and reductive pathways are generally considered to be responsible for the acute
cytotoxic effects of BrTHMs, which occur mainly in the liver and kidneys after high-dose
exposures (Pegram, 2001). The types of reactive metabolites generated by the oxidative and
reductive pathways could form adducts with the purine and pyrimidine bases in DNA (USEPA,
2005d) but this has not been demonstrated experimentally for any of the THMs.
Unlike chloroform, BrTHMs can also be metabolized by a competing pathway mediated by
GSTT1 that results in the production of highly reactive mutagenic metabolites (Pegram et al.,
1997; DeMarini et al., 1997; Ross and Pegram 2003, 2004). The metabolites of the GSTT1
pathway covalently bind DNA via formation of deoxyguanosine adducts leading to mutations
(GC —~ AT transitions) (DeMarini et al., 1997; Ross and Pegram 2003, 2004).
Agents or genetic polymorphisms that result in increased or decreased activities of the GSTT1
enzymes responsible for BrTHM metabolism can modify the risk for carcinogenicity. Those with
a null phenotype will have a lower risk than those with a homozygous positive phenotype
(Cantor et al., 2010).
Increased liver, kidney and large intestinal tumors were observed in rodent studies following oral
exposure to BrTHMs; however, scientific opinions vary regarding the causal relationship
between exposure to BrTHMs and tumors in animals as well as humans (bladder tumors).
Shokeer and Mannevik (2010) demonstrated lower hepatic activity of GSTT1 enzyme in humans
than in rodents, but haloalkanes were not tested as substrates in this study. Ross and Pegram
(2003) compared hepatic GSTT1-mediated metabolism of BDCM across species and found that
rat and human activities were similar and were both lower than in mice. Mouse liver cytosol was
13-fold more efficient in catalyzing GSH conjugation to dichloromethane than to BDCM, while
rat and human liver cytosols were three and seven fold more efficient (Reitz et al., 1989 as cited
by Ross and Pegram, 2003).
The balance between the competing CYP2E1 and GSTT1 pathways may be an important
determinant of tissue susceptibility to BrTHM-induced carcinogenesis (Ross and Pegram, 2004).
Target tissues for BrTHM-induced carcinomas in rodents had higher ratios of GSTT1 :CYP2E1
Six-Year Review 3 Technical Support Document	4-17
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
activities. Potential differences between rodent and human enzyme activities in the bladder
epithelium relative to the liver are as yet not known.
Bull (2012) proposed a pharmacokinetic analysis that suggests that in humans THMs would only
be metabolized by CYP2E1. However, the metabolic constants used in that analysis were derived
from enzyme activities in rodent livers and Salmonella, which should not be assumed to be
relevant or comparable to activities in the human urinary and intestinal tracts.
GSTT1 activity is significant in the human urinary tract (Thier et al., 1998). CYP2E1, on the
other hand, has been reported to be either not present or present only at very low levels in the
human kidney (Amet et al., 1997; Cummings et al., 2000; Baker et al., 2005). This increases the
likelihood of significant GSTT1 metabolism in the urinary tract of humans. In contrast to
humans, renal CYP2E1 levels are substantial in rodents (Ross and Pegram, 2004; Kuzniak et al.,
2005; Tabrez and Ahmad, 2010), suggesting that rodents could be less susceptible than humans
to BrTHM-induced genotoxic damage in the urinary tract but more sensitive to the phosgene-like
dibromomethaldehyde metabolites.
The chronic bioassays and the majority of animal studies with BrTHMs used oral exposures.
Because human exposures occur via multiple routes, an understanding of the volatility and
dermal permeability of the THMs is relevant. Based on data compiled from ChemlDPlus
(ChemlDPlus, 2015) and HSDB (HSDB, 2015), chloroform has the highest vapor pressure of the
chorinated/brominated THMs and is therefore more volatile than the BrTHMs, indicating it is
more likely to result in exposure via inhalation than the BrTHMs. Xu et al. (2002) examined the
skin permeability of the THMs. They found that the skin permeability coefficients, Kp, (cm/h)
were 0.16 for chloroform, 0.18 for BDCM, 0.20 for DBCM and 0.21 for bromoform. This
indicates that the BrTHMs would tend to be absorbed dermally more readily than chloroform.
However, the Kp differences are not large and the authors of this study noted that the THM Kp
values suggest that all THMs may be significantly absorbed through the skin during dermal
exposure. BrTHM pharmacokinetics are route-dependent, with dermal and inhalation exposures
leading to much higher blood levels and extra-hepatic tissue doses than oral exposure (Backer et
al., 2000; Leavens et al., 2007; Kenyon et al., 2015). This could be an important contributing
factor in the etiology of DBP-associated human bladder cancer (discussed in greater detail in
Section 4.1.2.1.1).
Based on results from a number of studies (Pegram et al., 1997; DeMarini et al., 1997; Ross and
Pegram, 2003, 2004; Leavens et al., 2007; Richardson et al., 2007; Cantor et al., 2010;
Kogevinas et al., 2010; Kenyon et al., 2015), there is a suggestion of a causal relationship
between exposure to BrTHMs (perhaps in combination with other DBPs) and bladder cancer in
humans. Additional research would help to address gaps in the current understanding of the
causal relationship. The supporting information includes the following:
• BrTHMs, and not chloroform, are mutagenic via activation by GSTT1 (Pegram et al.,
1997; DeMarini et al., 1997). The GSTT1-mediated metabolism of BDCM forms reactive
intermediates that covalently bind with deoxyguanosine bases in DNA. This evidence is
consistent with BrTHMs being mutagenic and carcinogenic (Ross and Pegram, 2003,
2004).
Six-Year Review 3 Technical Support Document	4-18
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
•	Cantor et al. (2010) reported that the association between bladder cancer and THM
exposure was stronger in GSTT1(+) subjects and that people who were GSTT1(-/-) had
no increased risk. People with susceptible genotypes for both GSTT1 and GSTZ1
(HAAs) had up to a 5.9-fold increased risk.
•	Kogevinas et al. (2010) studied swimmers exposed to DBPs in pools and showed
increases in biomarkers for genotoxicity (micronuclei and DNA damage in peripheral
lymphocytes, mutagenicity of urine (Ames assay) and micronuclei in exfoliated urothelial
cells). This provides additional evidence in support of the role of BrTHMs in producing
genotoxic effects leading to bladder cancer.
•	Stayner et al. (2014) showed an increase in micronuclei frequency in maternal blood
associated with BrTHM exposure from all sources (including ingestion, dermal and
inhalation) especially during the first and second trimesters of pregnancy and notably
from exposure due to bathing.
Bull (2012) and Hrudey et al. (2015 a,b) have noted that other DBPs in disinfected water may
co-occur with THMs and contribute to the cancer risk in humans, and that existing data are
insufficient to determine causality of the BrTHMs. Bull (2012) made the following points to
support this view, and for each one, EPA has concerns as indicated:
(1)	The same enzymes that are involved in the metabolism of THMs are involved in the
metabolism of other DBPs and in the metabolism of lipids. Thus, they could contribute to
carcinogenesis. Concern'. EPA's understanding is that the only DBP listed by Bull (2012)
that was actually shown to be a GSTT1 substrate is 1,3-dichloroacetone, which occurs at
much lower levels in drinking water than the THMs (Serrano et al., 2014).
(2)	At least one transcription factor has been shown to be modified by a GST, which
modifies the activity of the transcription factor, often playing a role in carcinogenesis.
Concern: The transcription factor example given by Bull (2012) is a modification by
GSTP1 (not GSTT1) and is therefore not relevant to the findings of Cantor et al. (2010)
or to a hypothesis involving GSTT1-mediated metabolism of BrTHMs in bladder
carcinogenesis.
(3)	The rate of THM metabolism at low blood levels from oral and dermal exposures will be
independent of the enzyme isoforms present, and other DBPs may be better substrates
than THMs for the genotypes that express the active isoforms of the GSTT1 enzyme.
Concern: This statement on the rate of THM metabolism is based on liver metabolism in
rodents and pharmacokinetic constants derived from Salmonella data which, for the
reasons stated above, should not be extrapolated to human bladder metabolism. Bull
(2012) suggested that at the low blood concentrations of BDCM in humans from
exposure through drinking water, the mutagenic metabolite of BDCM produced by
GSTT1 metabolism will be essentially zero. However, EPA contends that Bull's analysis
is based on inappropriate estimates of Km (the substrate concentration at '/2 the maximum
rate of reaction or Vmax) for these enzymes, with no consideration of Vmax). In Bull
(2012), the "human" CYP2E1 Km estimate was based on rat in vivo data which reflects
primarily liver metabolism by CYP2E1 and other CYPs, and the "human" GSTT1 Km
estimate is derived from Salmonella mutation data rather than a value derived from
Six-Year Review 3 Technical Support Document	4-19
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
studies using relevant human data. EPA further contends that Bull's analysis is not
relevant to human bladder metabolism of BrTHMs.
(4)	The differences in expression of the GSTT1 enzymes across species (humans, rats, mice)
are small and unlikely to account for interspecies differences in sensitivity. Concern'. This
statement does not consider the tissue-specific species difference in the expression of the
key enzymes (such as CYP2E1) described above.
(5)	Human GSTT1 has low activity on electrophilic substrates compared to GSTT1 activity
in rats and mice and could explain the species differences in detoxification of DBPs other
than THMs. Concern: EPA's understanding is that the statement that GSTT1 has lower
activity on electrophilic substrates in humans than in rats is not supported by information
available for halomethane substrates (Reitz et al., 1989; Ross and Pegram, 2003).
Despite EPA's articulated concerns, the points made by Bull (2012) and Hrudey et al. (2015a, b)
also support a need for additional research that would help to address gaps in the current
understanding of the causal relationship (i.e., mechanistic research) between exposure to
BrTHMs (in combination with other DBPs) and bladder cancer in humans.
Cantor et al. (2010) acknowledged that although THMs and HAAs are the most common
chemical species within the DBP mixture, they may not be the most toxic/carcinogenic. One or
more of the GST polymorphisms of interest could be acting in important ways on other DBP
compounds whose levels correlate with THM levels. However, at present, dichloroacetone is the
only additional DBP that has been identified as a GSTT1 substrate. Both the kidneys and the
bladder would receive greater internal exposure to THMs that are absorbed via dermal or
inhalation routes of exposure (see Backer et al., 2000; Leavens et al., 2007; Kenyon et al., 2015).
Although epidemiological findings suggest an increased bladder cancer risk with greater DBP
exposure from showering/bathing (Villanueva et al., 2007), it is not yet confirmed whether
BrTHMs are a causal factor in human bladder cancer. Research to support quantitative risk
assessment via these routes of exposure could further inform this question.
MO A for Bromodichloromethane (BDCM)
BDCM is generally the most prevalent and most extensively studied BrTHM. BDCM can be
metabolized by three potential pathways that give rise to reactive intermediates (Pegram, 2001;
NTP, 2006):
(1)	oxidative metabolism by cytochrome P450, primarily the CYP2E1 isoform which results
in reactive dihalocarbonyls and dihalomethyl radicals,
(2)	reductive metabolism mediated by cytochrome P450 (CYP2B isoforms), which generates
dihalomethyl radicals, and
(3)	GSTT1-catalyzed conjugation with GSH, which results in the formation of DNA reactive
species and S-dihalomethyl metabolites.
Six-Year Review 3 Technical Support Document	4-20
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
The activity of these enzymes is species dependent, tissue-specific and genetically determined.
Biotransformation during the detoxification of BDCM might explain why the acutely toxic
effects are primarily found in the liver and kidney (Health Council of the Netherlands, 2007).
BDCM is metabolized by GST-catalyzed conjugation with GSH, via GSTT1, and by the
CYP450 oxidative pathway (Pegram et al., 1997; Ross and Pegram, 2003, 2004). It is
hypothesized that bioactivation of BrTHMs catalyzed by GSTT1 could result in cell
transformations that lead to cancers. When considering the genotoxic effects that could occur at
environmental exposure levels, CYP2E1-mediated metabolism could act as a detoxification or
clearance pathway, because the ultimate product of this pathway is carbon dioxide. CYP2E1
concentrations are much higher in the liver than the kidney, resulting in detoxification of BDCM
by rat hepatic microsomes, but the efficacy of this pathway was found to be reduced in rat kidney
and large intestine (Ross and Pegram, 2004), and kidney CYP2E1 levels in humans are much
lower than in rodents. The path leading to DNA damage from BDCM could be more pronounced
in extra-hepatic tissues (due to a higher GSTT1:CYP2E1 ratio). Moreover, these tissues,
including the bladder, would be expected to receive higher doses of THMs via inhalation and
dermal routes of exposure than by the oral route where there is initial first pass metabolism in the
liver.
Differences in the distribution of BDCM among different organs after administration by gavage
versus by drinking water in rodents were addressed by pharmacokinetic modeling of BDCM.
Dose-response analyses of the carcinogenic effects using peak and cumulative rates of
metabolism via GST and CYP450 oxidative pathways in target organs were used as surrogate
dose metrics in the studies by NTP (2006) and Ross and Pegram (2004). Using a physiologically
based pharmacokinetic model for oral administration of BDCM in F344/N rats, 90 percent of
total metabolism occurs during first-pass clearance by the liver. Allocating this 90 percent of
total metabolism that occurs in the liver between the P450-mediated and GSTTT1 metabolic
pathways, approximately 99 percent occurs via the CYP450-mediated pathway and
approximately 1 percent through the GSTT1 pathway. Considering the kidney and the large
intestine, 87-88 percent of BDCM metabolism in these two organs occurs via the CYP450
pathway and 12-13 percent via the GSTT1 pathway. These organ-specific differences in the
relative importance of CYP450- and GST-mediated BDCM metabolism indicates greater relative
metabolism through the GSTT1 pathway in the kidney and large intestine than in the liver. Due
to the species and route differences in BDCM pharamacokinetics described above, humans
exposed to BDCM via dermal or inhalation exposures would be expected to experience tissue
distributions that are different from oral exposure. Extrahepatic tissues, such as organs in the
urinary tract, would receive a higher percentage of the absorbed dose following inhalation and/or
dermal exposures where more of the dose would be expected to be metabolized by GSTT1.
The products of GSTT1 metabolism have been shown to be mutagenic, which leaves open the
possibility that BrTHMs could be carcinogenic by a genotoxic mechanism in humans when
exposures are via inhalation or dermal contact because there is no first-pass metabolism by the
liver to reduce the unmetabolized BDCM reaching other tissues. After acute oral, high dose
rodent exposures to THMs, metabolites of CYP-mediated pathways can overwhelm
detoxification mechanisms, leading to cytotoxicity with a consequent increased risk for tumors.
Given the siginificant tissue-specific and species differences in the activities of key enzymes
involved in BDCM metabolism, it is important to realize that hepatic metabolism kinetics in
Six-Year Review 3 Technical Support Document	4-21
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
rodents cannot be assumed to be equivalent to human urothelial metabolism. Ross and Pegram
(2004) indicate that in the extrahepatic target tissues, GSTT1 products can be generated at low
BrTHM concentrations and account for a higher percentage of total metabolism than suggested
by Bull (2012). The case for substantial human GSTT1 metabolism in the urinary tract is even
stronger given the lack of significant CYP2E1 activity in the kidney (Amet et al., 1997;
Cummings et al., 2000; Baker et al., 2005). Inhalation and dermal exposures increase extra-
hepatic tissue concentrations due to the lack of hepatic and intestinal first-pass clearance. The
quantitative impacts of this difference on measures of internal dose in humans is discussed in
section 4.1.2.1.2.
Consistent with the findings that GSTT1-mediated metabolism of BDCM leads to DNA
modification and mutations, the Health Council of the Netherlands (2007) concluded that BDCM
could exert its carcinogenic effect by a stochastic genotoxic mechanism. According to the Dutch
Guideline to the Classification of Carcinogenic Compounds, stochastic genotoxins include
compounds that, as parent or as reactive metabolites, interact directly with DNA causing damage
such as adducts or strand breaks, leading to gene mutations or chromosome abnormalities that
occur at sites associated with carcinogenesis. As described in section 4.1.1.3.1, there is also some
evidence implicating an epigenetic carcinogenic mechanism for BDCM. BDCM has been shown
to induce DNA hypomethylation of the c-myc tumor promoter gene in B6C3F1 mice and to
cause hypomethylation in kidney DNA in male B6C3F1 mice and male F344 rats (Tao et al.,
2005), suggesting carcinogenic potential in the kidney. DNA hypomethylation occurs with
BDCM exposure in rat but not mouse colon and correlates with its carcinogenic activity in rats
and lack of carcinogenic activity in mice (Pereira et al., 2004a; George et al., 2002).
4.1.1.3.4 Mode of Action for the HAAs
The toxic potency of some of the five regulated HAAs is associated with enzyme inhibition (e.g.,
GAPDH, GSTzeta (Pals et al., 2011; Saghir and Schultz, 2005)). Dad et al. (2013) reported
inhibition of GAPDH by mono-HAAs can lead to ROS and subsequent damage to DNA.
Saghir and Schultz (2005) studied the toxicokinetics of HAA mixtures in naive and GSH
transferase zeta 1 (GSTZl)-depleted male F344 rats administered oral or IV mixtures of HAAs.
Rats were pretreated for seven days with drinking water containing DCAA to deplete GSTZ1
activity in the liver. The GSTZ1 pathway is susceptible to inactivation by exposure to DCAA
and other chlorobromo- di-HAAs. This reduction in GSTZ1 activity reduces the clearance of
chloro- and bromochloro- di-HAAs through inhibition of hepaticGSTzeta and leads to
production of alkylating metabolites from the amino acids tyrosine and phenylananine
metabolized via the GST zeta pathway. The results of low-dose exposures to HAA mixtures
suggest competitive interactions between tri- and di-HAAs. Total dose is important, as clearance
is dose dependent due to competition for GSTZ1. Polymorphic expression of GSTZ1 can affect
bladder cancer risk associated with DBP exposures, with genotypes resulting in lower GSTZ1
activity (and presumably lower HAA clearance) being associated with greater risk (Cantor et al.,
2010).
DCAA has been proposed to produce cancer by a nongenotoxic mechanism (Miller et al., 2000).
The mode of action for DCAA consists of selective stimulation of tumor cells, which arise
spontaneously, and suppression of normal division in hepatocytes, including suppressed
Six-Year Review 3 Technical Support Document	4-22
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
apoptosis, which causes small eosinophilic foci (Stauber and Bull, 1997; Miller et al., 2000).
DCAA was not associated with either liver peroxisome or hepatocyte proliferation in the studies
by DeAngelo et al. (1999). Stauber and Bull (1997) reported increased proliferation of selected
cell lines (e.g. c-Jun positive cells, following DCAA exposures).
EPA concluded that DCAA may potentially be genotoxic under in vivo exposure levels that
increase tumor incidence (USEPA, 2003c). It causes point mutations and chromosomal
aberrations at relatively high exposure levels, but mutations are viewed as exhibiting linear low-
dose response for cancer risk assessment. International Agency for Research on Cancer (IARC)
(2014) concluded that weak to moderate evidence is available to suggest that DCAA is a
genotoxic agent but that it may also act through multiple non-genotoxic mechanisms in liver
carcinogenisis. WHO (2000) concluded that there is some evidence of genotoxicity but only at
such high levels as to not be relevant for tumorigenesis. Regenerative hyperplasia is not likely to
play a role in DCAA-induced hepatocarcinogenicity (USEPA, 2003c).
DCAA caused DNA hypomethylation in male mouse kidneys, particularly hypomethylation of
the c-myc growth promoter gene (Tao et al., 2005). While Tao et al. (2005) reported that
chloroform in combination with DCAA caused synergistic hypomethylation, the data analysis
approach, as pointed out above, does not allow assessment of deviations from additivity. DCAA
did not induce renal DNA hypomethylation in female mice and does not induce kidney tumors in
female mice.
Repeated exposure to DCAA results in a decreased ability to metabolize it, attributed to DCAA's
inhibition of GSTZ which metabolizes the parent compound. DCAA induced inhibition of liver
GSTZ activity is greater in rats than in mice or humans, but this potential mode of action for its
carcinogenicity is not yet fully characterized (USEPA, 2003c). Humans with low GSTZ activity
may be more susceptible to DCAA toxicity. The carcinogenic and genotoxic effects of DCAA
are strongly associated with higher doses where DCAA metabolism is inhibited.
Many studies have found TCAA not to be genotoxic. TCAA produces liver tumors in mice, but
not in rats, and it is not considered a cancer risk at concentrations in drinking water (Bull, 2000).
Tao et al. (2005) noted that TCAA caused hypomethylation of DNA in male mouse kidneys but
not in female mice. TCAA does not induce DNA hypomethylation in female mice. While TCAA
was noted in Section 4.1.1 as being classified as a suggestive carcinogen, it has an unidentified
MOA and data supporting an important role for a nongenotoxic MOA with a strong link to the
peroxisome proliferation MOA.
Six-Year Review 3 Technical Support Document	4-23
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
4.1.2 Epidemiology and Weight of Evidence
4.1.2.1 Cancer
4.1.2.1.1 Bladder Cancer
Information Available During Development of Stage 1 and Stage 2 D/DBPRs
For the development of the benefits analysis for both the Stage 1 and the Stage 2 D/DBPRs, EPA
used five bladder cancer case-control epidemiology studies that were conducted in the 1980s and
1990s:
•	Cantor etal. (1985, 1987)
•	McGeehin et al. (1993)
•	King and Marrett (1996)
•	Freedman et al. (1997)
•	Cantor et al. (1998)
(Note that the Cantor et al. (1985) and Cantor et al. (1987) studies both used the same case-
control data.) Details for each of the above studies are summarized in Exhibit 6.3 of the EA for
the Stage 2 D/DBR (USEPA, 2005g).
These five case-control studies used similar (though not identical) exposure metrics based on
years of exposure to chlorinated drinking water (primarily chlorinated surface water) to estimate
odds ratios, although some of the studies used other metrics as well. For example, both the King
and Marrett (1996) and the Cantor (1998) studies also provided information on changes in risk
related to THM4 concentrations in the drinking water. All five studies showed an increase in the
odds ratio for bladder cancer incidence with an increased duration of exposure. Using the
published odds ratio results from these five studies, EPA calculated an estimate for the lifetime
cancer risk population attributable risk (PAR) range of 2 to 17 percent. Between 2 and 17 percent
of bladder cancers occurring in the United States could be attributed to long-term exposure to
chlorinated drinking water at the time of the Stage 1 D/DBPR. PAR is the reduction in incidence
that would be observed if the population were entirely unexposed from the presumed
contributing causative factor of incidence (in this case chlorination DBPs). Detailed explanations
of these PAR calculations, as well as for those described using additional studies, can be found in
the benefits analysis for the Stage 2 D/DBPR (USEPA, 2005g).
To support the Stage 2 D/DBPR, EPA used two additional published epidemiological studies:
•	Villanueva et al. (2003)
•	Villanueva et al. (2004)
The Villanueva et al. (2003) study was a meta-analysis that used an exposure metric of "ever
exposed to chlorinated drinking water" in the populations from the several studies included in the
analysis. Villanueva et al. (2003) calculated odds ratios from the combined results of six case-
control studies. Four of the five case-control studies used by EPA for Stage 1 D/DBPR were
used by Villanueva et al. 2003 in the meta-analysis and accounted for over 90 percent of the
Six-Year Review 3 Technical Support Document	4-24
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
weighting applied to the six studies to calculate a combined odds ratio of 1.2 (confidence bounds
= 1.1 - 1.4). The study by Freedman et al. (1997) was not included in the meta-analysis since the
underlying cohort study (Wilkins and Comstock, 1981) of the same population was included in
the meta-analysis. This study and one new study (Koivusalo et al., 1998) accounted for the
remaining weighting. Summary details of the studies used for the meta-analysis are described in
Exhibit 6.4 of the EA for the Stage 2 D/DBPR (USEPA, 2005g). EPA used the meta-analysis by
Villanueva et al. (2003) to calculate a pre-Stage 1 PAR estimate of 15.8 percent (95 percent CI
(Confidence Interval) = 8.5 - 27.2).
Villanueva et al. (2004) conducted a pooled data analysis using six studies that included both a
THM4 concentration metric and duration of exposure to chlorinated water to estimate odds ratios
for bladder cancer. Two of the six studies used in the pooled analysis of Villanueva et al. (2004),
King and Marrett (1996) and Cantor et al. (1998), were used by EPA to develop the Stage 1 and
Stage 2 D/DBPRs, as described above. Data on THMs from three of the six studies used,
(Koivusalo et al., 1998; Cordier et al., 1993; Porru unpublished) had not been previously
published; and for the sixth study, detailed THM information from part of a large U.S. study
(Cantor et al., 1987; Lynch et al., 1989) was used. Summary details of these six studies are
described in Exhibit 6.5 of the EA for the Stage 2 D/DBPR (USEPA, 2005g). For Stage 2, EPA
used the THM4 average concentrations (with additional data provided by the authors) to develop
a THM4 concentration-response relationship to predict the odds ratio as a function of average
THM4 exposure. Using a pre-Stage 1 average THM4 concentration in the U.S. of approximately
38 |ig/L, EPA derived a PAR value of 17.1 percent (95 percent CI = 2.5 - 33.1).
EPA concluded that the PAR values estimated from the three approaches noted above (i.e., the 2
-17 percent range from the five case control studies, the 15.8 percent value from the Villanueva
et al. (2003) meta-analysis and the 17.1 percent derived from the Villanueva et al. (2004) pooled
data study) provided a reasonable estimate of the percentage range of bladder cancer nationally
that is associated with chlorination DBPs in drinking water. In the Stage 2 EA (USEPA, 2005g),
EPA concluded that more evidence was available to support a potential association (though not
an established causality) between bladder cancer and DBP exposure than for other cancers
considered. At the same time, EPA acknowledged that there were gaps in the understanding of
bladder cancer etiology as it relates to chlorination DBPs that could lead to some uncertainty,
including reasons for inconsistent results across the various studies, particularly for populations
of males versus females and smokers versus nonsmokers. Males tended to have higher risks of
bladder cancer, as did smokers (USEPA, 2005g).
In summary, for the Stage 2 D/DBPR, EPA used the five case-control studies used for Stage 1,
the Villanueva et al. (2003) meta-analysis and the Villanueva et al. (2004) pooled data analysis
to obtain PAR values for pre-Stage 1 bladder cancer incidence ranging from 2 percent to 17
percent, with an indication from the more recent of the studies that the PAR values tended
toward the higher end of that range. Although these studies and the analyses performed using the
data from them to obtain the PAR values suggested an association of exposure to chlorinated
drinking water and to some extent to THM4 specifically and bladder cancer incidence, the
information was insufficient to draw a definitive conclusion regarding causality.
Six-Year Review 3 Technical Support Document	4-25
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
New Information Available Since Development of Stage 2 D/DBPR
As part of the SYR, EPA conducted a literature search to identify new epidemiology studies
about bladder cancer that became available subsequent to the promulgation of the Stage 2
D/DBPR. Eight new studies were identified: five case-control studies, two pooled and meta-
analysis studies, and one ecological study:
•	Case-control studies:
o Chang et al. (2007)
o Bove et al. (2007b)
o Villanueva et al. (2007)
o Michaud et al. (2007)
o Cantor et al. (2010)
•	Pooled data and meta-analysis studies:
o Villanueva et al. (2006)
o Costet et al. (2011)
•	Ecological study:
o Llopis-Gonzalez et al. (2011)
Overviews of these studies are presented below, with additional details provided in the relevant
sections of Appendix A. There is some overlap among these eight new studies in terms of the
populations analyzed. Specifically, of the five case-control studies, three of them (Michaud et al.,
2007; Villanueva et al., 2007; Cantor et al., 2010) are based on the same study population
enrolled in Spain between 1998 and 2001. This same Spanish population was also included in
Costet et al. (2011), one of the two new pooled and meta-analysis papers.
Costet et al. (2011) also included case-control populations from two earlier studies from Finland
(Koivusalo et al., 1998) and France (Cordier et al., 1993) that were included in the Villanueva et
al. (2004) pooled analysis study used to develop the Stage 2 D/DBPR. In addition, the new
pooled data analysis by Villanueva et al. (2006) used the same six case-control studies used by
Villanueva et al. (2004) supporting the Stage 2 D/DBPR.
Therefore, in arriving at conclusions regarding the extent to which these eight new studies
support or alter the conclusions reached in developing the Stage 2 D/DBPR, this overlap of study
populations within these eight studies and with the studies used for Stage 1 and 2 D/DBPRs
should be kept in mind. There are different implications for this overlap depending on the issue
being informed and these will be discussed in subsequent discussion of the specific studies.
Also, whereas the primary exposure metric used in the studies supporting the Stage 1 and Stage 2
D/DBPRs was duration of exposure to chlorinated drinking water, the exposure metrics used in
these new studies to estimate odds ratios for bladder cancer were THM4 exposures and to some
extent fluid consumption and duration of water use activities that lead to dermal and inhalation
exposures.
Six-Year Review 3 Technical Support Document	4-26
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
The overall conclusions from the eight new studies are as follows:
(1) The five case control studies all provide continued support suggesting an association
between exposure to DBPs, and THM4 specifically, from drinking water sources and
bladder cancer.
a. The three studies that used the same Spanish population from 1998 - 2001 (Michaud
et al., 2007; Villanueva et al., 2007; Cantor et al., 2010) each looked at different
exposure-related characteristics that provided some additional new insights to this
association. Many of the individuals examined in these studies were exposed to
drinking water with THM4 concentrations having a higher proportion of BrTHMs
than in most U.S. drinking water supplies (Hrudey et al. 2015; Regli et al. 2015).
Thus, the exposure-response relationship observed between exposure from THM4
concentrations and bladder cancer risk in these may be more pronounced than what
might be found for the general U.S. population. The polymorphism distributions for
these are not expected to be very different from those in the United States (discussed
more specifically following).
i.	Michaud et al. (2007) focused on the relationship between water intake, THM4
levels and bladder cancer. They observed that for a given THM4 concentration
exposure range the bladder cancer risk decreased with water intake. They detected
an increased odds ratio in this population subset with >26.0-49.0 [j,g/L (OR =
2.34; 95 percent CI = 1.16 - 4.71) and >49.0 ^g/L (OR = 2.06; 95 percent CI =
0.83 - 5.08) that were comparable to that reported by Villanueva et al. (2007).
However, they saw limited evidence of an interaction and no clear exposure-
response relationships when considering both exposure measures. (See Appendix
A for more detail.)
ii.	Villanueva et al. (2007) provided results showing increased risk of bladder cancer
in this population both with increased THM4 levels and increased duration of
exposure. Long-term THM4 exposure from all exposure routes was associated
with a two-fold increase in odds of bladder cancer incidence (OR = 2.1; 95
percent CI = 1.09 - 4.02) comparing those in the highest quartile of average
household THM4 level (>49 |ig/L) to those in the lowest THM4 quartile (<8
|ig/L), with a statistically significant positive trend observed in the odds of
bladder cancer for increasing quartiles of average residential THM4 level (p value
for trend<0.01). They also provided results showing that in addition to increased
risk from ingestion with increasing THM4 levels, there was evidence of higher
risks from increased time spent showering and/or bathing, and with exposure from
swimming pools. (See Appendix A for more details.)
iii.	Cantor et al. (2010) provided particularly novel information based on the Spanish
1998 - 2001 population in showing that there was an association between bladder
cancer and the presence of polymorphisms in key metabolizing enzymes that,
although not showing a clear causal relationship between DBPs and bladder
cancer, suggested a possible mechanism of action. Cantor et al. (2010) found that
people with the GSTT1(+) genotype were at significantly greater risk (OR = 2.2;
Six-Year Review 3 Technical Support Document	4-27
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
95 percent CI = 1.1 - 4.3) for developing bladder cancer when exposed to the
upper THM4 exposure quartile (>49 (J,g/L) compared to GSTTl-null participants
who had no increased risk at the same exposure level. Cantor et al. (2010) noted
that > 20 percent of their study population were joint carriers of the high risk
genotypes of the three genes evaluated and that for subjects with two of these
genotypes (GSTT1(+) and GSTZ1 CT/TT) OR increased monotonically to 5.9
percent (95 percent CI = 1.8 - 19.0) in the highest quartile of THM4 (> 49 ng/1).
Note that the two key polymorphisms, GSTT1(+) and GSTZ1 CT/TT, may be
present in approximately 80 percent and 30 percent of the U.S. population,
respectively, and that 24 percent (estimated from 0.8 x 0.3) may have both of
those polymorphisms (Regli et al., 2015). Because GSTT1 metabolizes BrTHMs,
and GSTZ1 metabolizes HAAs, these findings implicate these two prevalent DBP
classes in the etiology of DBP-associated bladder cancer.
b.	The Bove et al. (2007b) case-control study involved a New York State population of
white males and showed relatively high ORs for bladder cancer, with increased risk
associated with increased concentrations of THM4 and of individual THMs. This
study also found substantially higher risks associated with two of the species —
bromoform and BDCM — compared with DBCM and chloroform.
c.	The case-control study by Chang et al. (2007) considered a Taiwan population from
1996 to 2005 and also found increasing risk with increasing THM4 levels, where the
THM concentrations in the three groups considered were relatively low: <13.9 |j,g/L
(median 4.9); 13.9-21.1 |j,g/L (median 15.5); and >21.2 |j,g/L (median 21.2). It should
be noted that this study used an endpoint of bladder cancer deaths, not incident
bladder cancer cases and so is not directly comparable to the other studies.
(2) The two pooled data and meta-analysis studies also provided support for the association of
bladder cancer incidence and exposure to disinfected water and DBPs.
a.	The Villanueva et al. (2006) pooled data study used the same underlying study
populations as those used in the third approach for estimation of PAR values under
the Stage 2 D/DBPR. The additional insight from this study, similar to the Michaud et
al. (2007) case-control study, was consideration of joint effects of THM4 levels and
water intake. In the pooled analysis by Villanueva et al. (2006), the authors found
both decreased risk for increased tap water consumption for a given THM4
concentration range, as well as increased risk for increased THM4 level given a tap
water intake range.
b.	The Costet et al. (2011) study was a pooled data and meta-analysis of three European
populations that were considered in other case-control studies, including the Spanish
1998-2001 population as noted above and two of the European populations (from
Finland and France) that were included in the pooled data analysis of Villanueva et al.
(2004) that was used to derive one of the PAR estimates for the Stage 2 D/DBPR.
The focus of this study was to compare the combined European results with previous
results from combined North American (United States and Canada) populations to see
Six-Year Review 3 Technical Support Document	4-28
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
if there was a geographic difference in results. The authors continued to find
increased bladder cancer risk with increased THM4 concentrations and found no
difference in those risks between the North American and European populations for
comparable ranges of THM4 exposures. While populations in Europe may have
higher smoking rates than in the United States (a key causative factor associated with
bladder cancer) this confounding factor was controlled for in the respective study
populations.
(3) The eighth study was an ecological study by Llopis-Gonzalez et al. (2011) that also used a
Spanish population but one different from the previously noted Spanish study. The authors
considered various districts in and near Valencia, Spain, all having THM4 concentrations
in a range of 40 to 80 |ig/L. The authors considered an endpoint of bladder cancer mortality
rather than cancer incidence. Somewhat different from most other studies using bladder
cancer incidence and even the Chang et al. (2007) case-control study using cancer
mortality, the authors found a slight increase in risk for women, but no increased risk for
men. However, it is important to note that ecological studies are typically intended to
generate hypotheses and by themselves are less informative for drawing causal inference
compared to other study designs that examine individual-level data. In ecological studies
exposure is often characterized at the aggregate level (e.g., by county) versus the individual
level data in other study designs, and therefore there is generally greater likelihood for
unaccounted confounding factors.
Route of Exposure Considerations
Studies by Villanueva et al. (2007), Kogevinas et al. (2010) and Stayner et al. (2014) mentioned
previously underscore the significance of considering exposure to DBPs from disinfected water
by routes in addition to ingestion, notably from dermal and inhalation associated with bathing,
showering and swimming. Studies that only consider oral ingestion of drinking water may not
reflect potential risks from dermal or inhalation exposures, which are not subject to first-pass
liver metabolism and may result in relatively greater extra-hepatic distributions than oral
exposures. This has been shown in other studies as well. For example, Backer et al. (2000)
evaluated the combined effects of dermal and inhalation exposure. Mean concentrations in the
tap water used by the subjects were 20-32 |ag/l for chloroform, ~ 6 |ag/l for BDCM, ~ 1 |ag/l
DBCM and below the detection limit for bromoform. Backer et al. (2000) found that blood levels
of THMs were 4-5 times higher in people who took a 10-minute shower or bath than in people
who drank one liter of the same tap water source in 10 minutes.
In a study with human volunteers, Leavens et al. (2007) examined the relationship between oral
exposure (single 0.25 L drink, mean dose 146 ng/kg) and dermal exposure (forearm immersion
for one hour, estimated mean dose 155 ng/kg) to BDCM in water and BDCM pharmacokinetics.
Peak venous blood concentrations of BDCM ranged from 0.4 to 4.1 ng/L following oral
exposure and 39 to 170 ng/L for dermal exposure. This study demonstrates that activities
involving dermal exposure result in much higher blood concentrations and hence greater overall
distribution of BDCM to the systemic circulation compared to oral exposure.
Given the potential for multi-route exposures for some DBPs (e.g., the volatile THMs), the
aforementioned studies highlight the need to consider the impact of exposure by all relevant
Six-Year Review 3 Technical Support Document	4-29
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
routes of exposure (oral, dermal, inhalation) at internal target sites of interest. Most of the studies
conducted to date characterized exposure only relative to ingestion metrics, with the exception of
Villanuenva et al. (2007), who included information on total exposure from ingestion, bathing
and showering as well as exposures from each.
Taken together, new information available since promulgation of the Stage 2 D/DBPR (including
information pertaining to mode of action already discussed), suggests that:
•	Bladder cancer risk may be significantly associated with non-oral routes of exposure
(dermal and inhalation from bathing, showering, swimming) as well as from direct
ingestion.
•	There may be a higher risk from some of the brominated DBP species.
•	There may be a relationship between cancer risk and the presence of certain genetic
factors in the population that affect metabolism and which could point to a mechanism of
action for bladder cancer.
Regli et al. (2015) elaborated on why the above factors may contribute to increased bladder
cancer risk in populations served by systems using chlorination. Given the concern for increased
bromide levels in source drinking waters from anthropogenic sources they developed a
methodology for estimating potential increased incidence of bladder cancer incidence from
hypothetical increased levels of bromide in source waters. By better accounting for the
uncertainty of the exposure data from Villanueva et al. (2004), they refined the dose-response
function that EPA used in its benefits analysis for the Stage 2 D/DBPR to estimate potential
bladder cancer risk as a function of THM4 concentration (USEPA, 2005g). Regli et al. (2015)
estimated that for roughly every 1 ng/L increase of THM4 (due to an increase in bromide in the
source water), excess lifetime bladder cancer risk could increase by about 1 x 10"4. Although
they qualified their overall risk estimates as uncertain since causality between bladder cancer risk
and exposure from chlorinated DBPs has not been established, the authors report that there is
currently more evidence than at the time of the Stage 2 D/DBPR to suggest a basis for causality
(Regli etal., 2015).
Kenyon et al. (2015) published a refined human multi-route PBPK model for BDCM that
included chemical-specific parameters that were experimentally derived using human tissues and
data. In addition, human data from diverse sources were used to evaluate and demonstrate the
predictive capability of the model. Analyses using this model suggested a large contribution of
inhalation and especially dermal exposure (e.g., from showering) to internal dose of BDCM
reaching the systemic circulation and thus available for extra-hepatic metabolism. For example,
using a liter equivalency approach (L-eq) they estimated the BDCM concentration in a liter of
water consumed by the oral route that would be required to produce the same internal dose of
BDCM resulting from a 10-minute shower in water containing 10 |ig/L BDCM. The oral L-eq
concentrations for showering are 282, 312 and 2.1 |ig/L BDCM for maximum venous blood
concentration, area under the curve and amount metabolized in liver/hr, respectively. Based on
the hypothesis that metabolism in target tissues is important for toxicity and the development of
cancer, they found that non-oral exposures could contribute significantly to the amount of
BDCM available for metabolism in tissue and hence the potential for adverse effects. Overall,
the authors concluded that their analyses (1) demonstrated the importance of considering the
contribution of multiple routes of exposure to BDCM and similarly metabolized chemicals to
Six-Year Review 3 Technical Support Document	4-30
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
provide a more complete evaluation of potential risk of adverse health outcomes and (2) that this
refined human PBPK model could be used to estimate internal doses from real-world exposures.
The new information reviewed here strengthens the weight of evidence for the association
between bladder cancer in humans and exposure to chlorinated drinking water, with continued
indications of higher risks for both increased duration of exposure and increased concentrations
of THM4.
Notwithstanding the above, a causal relationship has not yet been established between bladder
cancer and exposure to any individual DBP or combinations of DBPs (oral, dermal, inhalation)
as noted by others (Hrudey et al., 2015). As new information continues to become available this
issue will be further informed. In this regard, the IARC advisory panel has nominated disinfected
water used for showering, bathing, swimming or drinking for evaluation for carcinogenicity,
based on ubiquitous exposure and extensive new mechanistic evidence of specific DBP toxicity,
including molecular epidemiology studies (IARC, 2014).
4.1.2.1.2 Colon/Rectal Cancer
Information Available During Development of Stage 1 and Stage 2 D/DBPRs
In the analyses supporting the Stage 1 D/DBPR, the data provided by the epidemiological studies
available at that time suggested a small increase in rectal and colon cancers from exposure to
chlorinated surface waters. The database of studies completed on colon and rectal cancers at the
time of the Stage 2 D/DBPR continued to support an association, but evidence remained mixed.
For colon cancer, one newer study supported an association (King et al., 2000) while others
showed inconsistent findings (Hildesheim et al., 1998; Yang et al., 1998). Rectal cancer study
results were mixed. Hildesheim et al. (1998) and Yang et al. (1998) supported an association
with rectal cancer whereas King et al. (2000) did not. A review of the colon and rectal cancer
epidemiological data by Mills et al. (1998) found that the evidence was inconclusive but that
there was a stronger association for rectal cancer and chlorination DBPs than for colon cancer. A
World Health Organization review (WHO 2000) reported that studies showed weak to moderate
associations with colon and rectal cancers and chlorinated surface water or THMs but that
evidence was inadequate to evaluate those associations.
EPA did not quantify the risk or risk reduction from colon or rectal cancer as part of its benefits
analysis for the Stage 2 D/DBPR but did include a brief "sensitivity analysis." Using the King et
al. (2000) study data for colon cancer in males only (showing ORs of 1.0 to 1.53), and the
Hildesheim et al. (1998) study data for rectal cancer in both sexes (showing ORs of 0.88 to 2.13),
EPA estimated that exposure to chlorinated drinking water could account for approximately 25
percent of male colon cancer and 12 percent of all rectal cancers. However, while those estimates
were provided to give some insight to the potential risk, they were not considered sufficiently
reliable to use in the risk or risk reduction analyses supporting the Stage 2 D/DBPR (USEPA,
2005g).
Six-Year Review 3 Technical Support Document	4-31
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
New Information Available Since Development of Stage 2 D/DBPR
Four studies identified since promulgation of the Stage 1 and 2 D/DBPR address colon and/or
rectal cancers: two case-control studies, one meta-analysis study and one ecological study.
•	Case-control studies:
o Bove et al. (2007a)
o Kuo et al. (2009)
•	Meta-analysis study
o Rahman et al. (2010)
•	Ecological study
o Rahman et al. (2014)
Appendix A provides detailed summaries about each of these studies relating to colon and/or
rectal cancers.
Bove et al. (2007a) compared the risk of rectal cancer with exposure to THM4 and their
individual species. THM levels varied spatially within the study county; although risk for rectal
cancer did not increase with total level of THMs, increasing levels of the component bromoform
(measured in ug/day) corresponded with an increase in odds ratios (OR = 1.85; 95 percent CI =
1.25 - 2.74) for rectal cancer. The highest quartiles of estimated consumption of bromoform
(1.69 - 15.43 ug/day) led to increased risk for rectal cancer (OR = 2.32; 95 percent CI = 1.22 -
4.39). Two other THMs were associated with an increase in risk for rectal cancer - DBCM (OR
= 1.78, 95 percent CI = 1.00-3.19) and BDCM (OR= 1.15; 95 percent CI = 1.00- 1.32).
Kuo et al. (2009) evaluated whether exposure to THM4 in drinking water is associated with the
risk of death attributed to colon cancer in 65 municipalities in Taiwan. All colon cancer deaths of
the 65 municipalities from 1997 through 2006 were obtained from the Bureau of Vital Statistics
of the Taiwan Provincial Department of Health. Controls were deaths from other causes and
were pair-matched to the cancer cases by gender, year of birth and year of death. Each matched
control was selected randomly from the set of possible controls for each cancer case. Data on
THM4 levels in drinking water in study municipalities were collected from the Taiwan
Environmental Protection Administration. The municipality of residence for cancer cases and
controls was assumed to be the source of the subject's THM4 exposure via drinking water. The
adjusted ORs for colon cancer death for those with high THM4 levels (greater than 14.8 |ig/L) in
their drinking water were 1.02 (95 percent CI = 0.87 - 1.2) and 1.04 (95 percent CI = 0.89 -
1.21) compared to the lowest group (less than 6.03 |ig/L). The results of the study showed no
statistically significant association between THM4 in drinking water at levels in this study and
risk of death from colon cancer. However, the relatively low THM4 concentrations in both the
high and low exposure groups in this study may have precluded detecting associations that might
occur at higher concentrations.
Rahman et al. (2010) identified relevant case-control and cohort studies. Separate risk estimates
for colon and rectal cancer were extracted from studies meeting the inclusion criteria. Relative
risks (RRs) from the cohort studies or odds ratios (ORs) from the case-control studies comparing
the highest exposure category with the lowest were pooled using random effects methods. A total
of 13 studies (3 cohort and 10 case-control) were analyzed. For colon cancer, the pooled RR/OR
Six-Year Review 3 Technical Support Document	4-32
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
estimates were 1.11 [95 percent CI = 0.73 - 1.70] for cohort studies, 1.33 (95 percent CI = 1.12 —
1.57) for case-control studies and 1.27 (95 percent CI = 1.08 - 1.50) combining both study
types. For rectal cancer, the corresponding RR estimates were 0.88 (95 percent CI = 0.57 - 1.35),
1.40 (95 percent CI = 1.15- 1.70) and 1.30 (95 percent CI = 1.06 - 1.59). Sensitivity analysis
showed these results were not importantly influenced by any single study. Publication bias was
not evident for the colon cancer analysis but may have been a minor issue for the rectal cancer
analysis. The results for rectal cancer may have been influenced by the quality of the studies.
Rahman et al. (2014) examined colon and rectal cancer incidence and water THM concentrations
in New South Wales, Australia._Average yearly concentrations of total and individual species of
THMs were obtained for 50 local government areas (LGAs). Indirectly-standardized incidence
rates of colon and rectal cancers in LGAs for the period 1995 to 2001 were regressed against
mean THM concentrations lagged five years, adjusting for socioeconomic status, high risk
drinking, smoking status, usual source of water and year of diagnosis, including local and global
random effects within a Bayesian framework. The statistical measure used by these authors was
the incidence rate ratios (IRRs) for an interquartile range increase in THMs, which were based
on the observed incidence of colon and/or rectal cancers relative to the expected incidence for
the 50 LGAs. Using five-year lag of exposure there was a positive association between
bromoform concentration and colo-rectal cancer in men (IRR= 1.025; 95 percent CI = 1.010 -
1.040) but not in women (IRR= 1.003; 95 percent CI = 0.987 - 1.018). The association in men
was mainly found in colon cancer with bromoform (IRR= 1.035; 95 percent CI = 1.017 - 1.053).
There was no appreciable association of colorectal cancer with other species of THMs.
Sensitivity analyses did not materially change the associations observed. The authors concluded
that a positive association was observed between colon cancer and water bromoform
concentrations in men.
Conclusions: Collectively, the post-Stage 2 studies of DBP exposure and colon and rectal cancer
risk support a continuing concern that long-term exposure to chlorination DBPs increases the
risk of colon and rectal cancers. It should be noted that the meta-analysis presented here by
Rahman et al. (2010) included studies that were completed prior to the Stage 2 D/DBPR. More
information on these studies can be found in the Stage 2 D/DBPR EA (USEPA, 2005g). Two of
these studies that reported associations on populations relying on waters with different levels of
bromoform supports the hypothesis that bromoform or other DBPs co-ocurring with bromoform
may increase the risk of colon cancer.
4.1.2.1.3 Other Cancers
Information Available During Development of Stage 1 and Stage 2 D/DBPRs
At the time of the Stage 1 D/DBPR, EPA evaluated epidemiology data for bladder, colon and
rectal cancers. The Agency did not evaluate data related to other cancers. During the
development of the Stage 2 D/DBPR, EPA reviewed studies related to other cancers as part of
the overall weight of evidence analysis (USEPA, 2005g). Studies on kidney, brain and lung
cancers and DBP exposure support a possible association (Kidney: Yang et al., 1998, Koivusalo
et al., 1998; Brain: Cantor et al., 1999; Lung: Yang et al., 1998). Definitive conclusions on other
cancers could not be made, because so few studies had examined these other endpoints. Studies
on leukemia found little or no association with DBPs (Infante-Rivard et al., 2001, 2002). Another
Six-Year Review 3 Technical Support Document	4-33
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
study did not find an association between pancreatic cancer and DBPs (Do et al., 2005). A study
researching multiple cancer endpoints found an association between THM4 exposure and all
cancer mortality when grouped together (Vinceti et al., 2004). EPA did not include quantification
of the risk or risk reduction from any of these other cancers as part of its risk and benefits
analyses for the Stage 2 D/DBPR (USEPA, 2005g).
New Information Available Since Development of Stage 2 D/DBPR
Three studies were identified that evaluated other cancer risks subsequent to promulgation of the
Stage 2 D/DBPR:
•	Chiu et al. (2010)
•	Kasim et al. (2006)
•	Karagas et al. (2008)
Appendix A provides detailed summaries about each of these studies relating to other cancers.
For cancer endpoints other than bladder, colon and rectal cancers, the post-Stage 2 epidemiology
studies provide only weak evidence of associations between drinking water DBP exposure and
pancreatic cancer, leukemia and skin cancer. Based on the available evidence, the observed
increases in risk are low and often not statistically significant. Limitations in the study designs
further diminish the strength of the evidence for positive associations.
4.1.2.1.4 Genotoxic Biomarkers
Four published studies have looked at the presence of biomarkers of genotoxicity in humans as
they relate to exposure from DBPs. One of these was published prior to the Stage 2 D/DBPR and
was mentioned briefly in the economic analysis supporting that rule (USEPA, 2005g). The other
three were published subsequent to the Stage 2 D/DBPR.
Ranmuthugala et al. (2003) reported that they found no effects on a biomarker of genotoxicity in
urinary bladder cells from THM4 exposure based on a cohort study undertaken in three
Australian communities in 1997. The three communities had varying levels of DBPs in their
water supplies (one had no measurable THM4, one had a median of 64 |ig/L and one had a
median of 138 |ig/L). The authors looked for micronuclei in bladder epithelial cells as the
biomarker. The authors considered exposure both in terms of THM4 concentrations in the water
supply, and as an intake dose (|ig/kg per day) calculated by considering individual differences in
ingestion, inhalation and dermal absorption. There were 228 participants in the study, of whom
63 percent were exposed to DBPs (at concentrations ranging from 38-157 |ig/L and doses from
3-469 |ig/kg per day). The authors reported RRs for DNA damage to bladder cells in relation to
DBPs, separately for smokers and nonsmokers as follows:
RR per 10 |ig/L:
Smokers 1.01 (95 percent CI = 0.97 - 1.06)
Nonsmokers 0.996 (95 percent CI = 0.961 - 1.032)
Six-Year Review 3 Technical Support Document	4-34
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
RR per 10 |ig/kg per day:
Smokers 0.99 (95 percent CI = 0.96 - 1.03)
Nonsmokers 1.003 (95 percent CI = 0.984 - 1.023)
The authors also found that the while the proportion of abnormal cells did not differ among the
three communities, the median unadjusted frequency of micronuclei was highest in the
unexposed community and lowest in the highest exposed community.
Ranmuthugala et al. (2003) concluded that their study provided no evidence that THM4
concentrations or intake at the levels they investigated are associated with DNA damage to
bladder cells. However, in their discussion of the study, Ranmuthugala et al. (2003) also noted
that because of the small size of the study population micronuclei might not be a sufficiently
sensitive indicator of carcinogenicity, and could also explain the lack of association in this study
between smoking and micronuclei frequency. They also noted that the higher prevalence in the
unexposed community could have been a result of a higher prevalence of smoking in that
community than in the other two.
The Villanueva et al. (2007) study, which focused on bladder cancer related to exposure from
ingestion, bathing, showering and swimming (discussed in more detail earlier), provided some
limited information on increased micronuclei related to higher THM4 exposures, but the results
were generally not statistically significant. However, the authors did note that higher associations
with micronuclei were observed for THM4 exposures from showering and bathing than
ingestion.
The Kogevinas et al. (2010) study presented results of an experimental study set in Spain that
assessed biomarkers of genotoxicity in blood, urine and exhaled air samples from 49 adult
nonsmoking volunteers before and after swimming in chlorinated water. The objective of the
study was to evaluate the genotoxicity of DBPs in swimming pool water by examining
biomarkers of genotoxicity before and after study participants swam for 40 minutes in a
chlorinated indoor swimming pool. The authors reported that the mean THM4 concentration in
the pool water was 45.4 + 7.3 |ig/L and that the pool air THM4 mean concentration was 74.1 +
23.7 |ig/m3, Biomarkers of genotoxicity included micronuclei and DNA damage (determined by
a comet assay) in peripheral blood lymphocytes before and one hour after swimming; urine
mutagenicity (determined by Ames assay) before and two hours after swimming; and
micronuclei in exfoliated urothelial cells before and two weeks after swimming. The authors
compared the biomarkers to concentrations of THM4 in exhaled breath of volunteers. The
investigators also evaluated the impact of participants' genotype on biomarker changes relative
to THM4 exposure by estimating associations and interactions with polymorphisms in genes
related to DBP metabolism and DNA repair.
On average, the concentration of THM4 in participants' exhaled breath was seven times higher
after swimming, relative to levels measured before swimming. The average THM4 levels before
and after swimming were 1.2 and 7.9 |ig/m3, respectively. The corresponding average levels for
the individual THMs were 0.7 and 4.5 |ig/m3 for chloroform, 0.26 and 1.78 |ig/m3 for BDCM,
0.13 and 1.2 |ig/m3 for DBCM, and 0.1 and 0.5 |ig/m3 for bromoform. The average number of
Six-Year Review 3 Technical Support Document	4-35
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
micronuclei-positive cells per 1,000 binucleated lymphocytes increased from 3.4 before
swimming to 4.0 after swimming; this increase was not statistically significant. The average
frequency of micronuclei in exfoliated urothelial cells and the level of urinary mutagenicity also
increased after swimming, but, again, these changes were not statistically significant. Swimming
was not associated with DNA damage detectable by the comet assay; the average amount of
DNA damage in peripheral blood lymphocytes measured through the comet assay decreased
significantly after swimming relative to before swimming (p = 0.008). An observed increase in
the frequency of micronucleated lymphocytes after swimming was positively associated with
higher exhaled concentrations of the BrTHMs but not chloroform. The P-coefficients (and 95
percent CIs) representing a change in micronucleated peripheral blood lymphocytes per 1,000
cells for a 1 -|ig/m3 change in the specific BrTHMs in exhaled breath measured after swimming
were as follows: 1.92 (95 percent CI = 0.21 - 3.63), p = 0.03 for BDCM; 1.71 (95 percent CI =
(-0.02) - 3.44), p = 0.05 for DBCM; and 5.04 (95 percent CI = 1.23 - 8.84), p = 0.01 for
bromoform. Urine mutagenicity increased significantly after swimming in association with
higher concentrations of exhaled bromoform (representing a change in urine mutagenicity for a
1 -|ig/m3 change in bromoform in exhaled air measured after swimming: P-coefficient = 5.27 (95
percent CI = 1.80 - 8.75), p = 0.004. Some effect modification by genetic polymorphisms was
observed (see below for further discussion).
This study provides insights into the relationships between swimming in chlorinated water,
biomarkers of THM exposure (THM concentrations in exhaled breath) and biomarkers of
genotoxicity. The results of Kogevinas et al. (2010) study are consistent with the hypothesis that
exposure to BrTHMs by swimming in chlorinated pools induces genotoxicity that may be
associated with cancer risk. The authors observed that only BrTHMs were associated with higher
genotoxicity; chloroform levels were not.
Kogevinas, et al. (2010) also tested several gene-environment interaction hypotheses in this
study, focusing on potential modification of the genotoxic effects of DBPs in chlorinated pool
water by variants in genes that code for enzymes thought to be important in DBP metabolism
(GSTT1, GSTZ1 and CYP2E1). In this study, subjects with the GSTT1 null genotype had lower
frequencies of micronuclei in exfoliated urothelial cells and lower urinary mutagenicity than
those with at least one functional allele. Although not statistically significant, these findings are
consistent with observations of mutagenesis in bacteria and DNA adducts in rodents. The
investigators did not observe modification of effects of THMs on micronuclei in peripheral blood
lymphocytes by GSTT1, a finding they argue is consistent with a lack of GSTT1 expression in
lymphocytes noted in other research. In contrast, individuals bearing GSTT2B +/+ gene (which
encodes the glutathione S-transferase theta 2B enzyme) had higher numbers of micronuclei in
peripheral blood lymphocytes than did other subjects. The copy-number variants encompassing
GSTT2B, which modifies GSTT2 gene expression, are in linkage disequilibrium with the
GSTT1 copy-number variants. All three of these genes are located in the same cluster and
combined effects are possible. However, the role of GSTT2 and GSTT2B genes in DBP
metabolism is not known. The authors also note that limited experimental data are available in
this study for GSTZ1 and CYP2E1 in relation to DBP exposure.
The authors reported that the evaluation of effect modification of the genotoxicity of DBPs
present in swimming pool water by genetic polymorphisms in genes active in DBP metabolism
and DNA repair was of low statistical power given the relatively small sample size and that their
Six-Year Review 3 Technical Support Document	4-36
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
findings should be interpreted with caution. Although the study had low statistical power to
identify the effects of genetic variation on responses to DBP exposures during swimming, such
variation is plausible. As such, the authors suggest that their findings of gene-environment
interactions be verified in further studies.
Stayner et al. (2014) published a genotoxic biomarker study that focused on micronuclei
frequency in maternal and cord blood lymphocytes associated with exposure to BrTHMs during
pregnancy. The study population included mothers and newborns from the island of Crete,
Greece, who became pregnant during the period of February 2007 to February 2008. There were
1,610 eligible women who agreed to participate and 1,459 were followed through delivery. A
subset of 408 donated maternal and/or cord blood for biomarker measurements. There were 214
mothers and 223 newborns (including 162 mother-child pairs) from singleton pregnancies having
micronuclei analysis of maternal and cord blood lymphocytes.
The study area, in and around the city of Heraklion, was divided into six zones according to the
source of ground water used in each area and corresponding to the six water treatment plants
supplying water to the participants. Within these zones, a total of 18 sample points were selected
(12 in Heraklion and six in rural areas), where drinking water was sampled for THMs four times
between September 2007 and January 2009 (72 samples in all). The authors indicated that they
decided to focus on BrTHMs in this study because they constituted >80 percent of the total
THMs measured. Detailed results were not presented for individual zones or sample areas. The
mean BrTHM concentration in residential water was reported to be 2.1 + 2.6 |ig/L, with a
median of 0.8 |ig/L, a minimum of 0.06 |ig/L and a maximum of 7.1 |ig/L. While these relatively
low BrTHM concentrations may have precluded detecting associations that might occur at higher
concentrations, they may also be markers for other co-occuring DBPs associated with the effects
observed.
Exposure routes to BrTHMs from water considered in this study included consumption as
drinking water, bathing, showering, swimming pool use and hand dishwashing. Information on
maternal water usage habits were combined with BrTHM water concentrations to estimate
exposure through all routes (ingestion, dermal absorption and inhalation).
Micronuclei frequency were measured in maternal and cord blood lymphocytes. Maternal blood
was collected one day after delivery and cord blood was collected from the placenta immediately
after delivery. Micronuclei detections were evaluated for both binucleated (BN) and
mononucleated lymphocytes.
The authors concluded that their study suggested that exposure to BrTHMs may increase the
frequency of micronuclei in maternal BN lymphocytes. They reported that there was no evidence
of BrTHM exposures being associated with micronuclei in maternal MONO lymphocytes nor
with micronuclei frequency in either BN or MONO lymphocytes of newborns.
The study reported an increase in the rate ratio2 of micronuclei in maternal BN lymphocytes per
1 |ig/L increase in residential tap water of 1.03 (95 percent CI = 0.99 - 1.07) over the full
2 The authors use the term "rate ratio (RR)" as an outcome measure. It is the ratio of the frequency of micronuclei in
an exposure group to the frequency in a referrant group.
Six-Year Review 3 Technical Support Document	4-37
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
pregnancy, with a slightly higher rate ratio of 1.05 for exposure during the first trimester, the
same rate ratio of 1.03 during the second trimester and no change in the rate ratio (i.e., 1.0) for
the third trimester.
They also reported an increase in the rate ratio of micronuclei in maternal BN lymphocytes per
|ig/week intake from all routes of 1.55 (95 percent CI = 0.59 - 4.09) over the full pregnancy,
with higher rate ratio increases for the first trimester of 3.14 (95 percent CI = 1.16 - 8.50) and
the second trimester of 1.68 (95 percent CI = 0.76 - 3.73); there was a reduction in the rate ratio
reported for the third trimester of 0.76 (95 percent CI = 0.40 - 1.45).
In addition, the authors reported that bathing had a particularly marked effect on micronuclei
frequency. Mothers who took baths only had an increased rate ratio of micronuclei in maternal
BN lymphocytes of 2.08 (95 percent CI = 1.09 - 3.98) compared with those who showered only.
Increased micronuclei frequency was also reported with increasing frequency of bathing per
week, duration of bathing and the product of frequency and duration.
With respect to the lack of evidence in their study that maternal exposure to BrTHMs had any
effect on micronuclei frequency in cord blood, the authors suggested that if the critical window
of exposure was during the first trimester (where the highest effects were seen for maternal blood
lymphocytes), then the cord blood lymphocytes collected at birth may not have been exposed
since the majority of the lymphocytes collected at birth in the cord blood are produced in the
third trimester (where no effects were seen for maternal blood lymphocytes). The authors also
suggested that another possible explanation was that because BrTHMs are metabolized by GST
to reactive mutagens, and that it is likely that in utero metabolism of BrTHMs is immature, there
might have been a lower exposure of reactive metabolites to the fetus.
4.1.2.2 Reproductive and Developmental Effects
During the development of the Stage 2 D/DBPR, EPA evaluated available epidemiology and
toxicology studies that looked at the relationships between exposure to chlorinated drinking
water or DBPs and adverse reproductive and developmental effects.
In the Stage 2 D/DBPR Economic Analysis, EPA stated that its evaluation of the best available
studies, particularly epidemiology studies, is that they did not at that time support a conclusion as
to whether exposure to chlorinated drinking water or DBPs caused adverse reproductive and
developmental effects, but that they did provide an indication of a potential health hazard
concern that warranted incremental regulatory action beyond the Stage 1 D/DBPR (USEPA,
2005g).
The specific reproductive and developmental endpoints that EPA focused on at that time were:
•	Fetal growth (mainly birth weight, small for gestational age (SGA) and pre-term delivery
(PTD))
•	Fetal viability (spontaneous abortion or still birth)
•	Fetal malformations (congenital anomalies)
Six-Year Review 3 Technical Support Document	4-38
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
In addition, EPA noted that there were limited studies addressing both female and male
reproductive endpoints. Possible associations between DBPs and reproductive and
developmental endpoints were found in a number of animal toxicology studies, and although the
majority of them were conducted using high doses, these studies were used to inform biological
plausibility for some of the effects observed in epidemiology studies.
The remainder of this section of the document provides a summary of both epidemiology and
animal toxicity studies addressing these several reproductive and developmental effects
endpoints that were published at the time of the Stage 2 D/DBPR and subsequent to the Stage 2
D/DBPR. In general, approximately 40 post-Stage 2 studies addressing end-points such as fetal
growth endpoints and congenital anomalies outcomes continue to support a potential health
concern, though the relationship of adverse outcomes to DBP exposure may not be known well
enough to quantify risks or benefits from reducing exposures. In addition, recent toxicological
studies on mixtures (Narotsky et al., 2011, 2013, 2015) showed diminished concern for many
reproductive and developmental endpoints (see Section 4.1.3 and Appendix A to this document
for further elaboration on the mixtures studies).
4.1.2.2.1 Epidemiology and Animal Toxicity Studies on Reproductive and Developmental
Effects
While most animal toxicity data from DBPs are derived from single-chemical studies, data about
potential adverse effects in humans come from human epidemiological studies involving
mixtures of DBPs formed during disinfection of drinking water. Both toxicological studies in
animals and human epidemiological studies have suggested that adverse reproductive and
developmental effects from DBP exposure may be of concern. These studies have not
demonstrated a causal relationship between low levels of DBPs in drinking water and
reproductive/developmental health risks in humans (Simmons, et al., 2008). EPA undertook a
multi-year research initiative involving four Agency laboratories (the "Four Lab Study") to
provide experimental data on environmentally-relevant mixtures of DBPs to help estimate the
potential health risks in humans exposed to mixtures of DBPs formed during disinfection of
drinking water (Simmons, et al., 2002, 2004). This section is intended to summarize key data that
are currently available about the adverse reproductive and developmental effects identified in
animal toxicology studies and human epidemiological studies following exposure to DBPs in
order to inform a "weight-of-evidence" assessment based on the current state of the science.
Exhibit 4.1 provides a summary of the results of the epidemiology and animal toxicity studies
published before the Stage 2 D/DBPR promulgation (and evaluated as part of the rule
development process) and subsequently to that time, addressing each of the seven primary
reproductive and developmental effects identified by EPA to be of potential concern with respect
to chlorination DBPs: birth weight; SGA; PTD; congenital anomalies; fetal loss; male
reproductive effects; and female reproductive effects.
Appendix A provides detailed summaries of the Pre- and Post-Stage 2 epidemiology
reproductive and developmental studies. It also provides brief summaries on the Pre-Stage 2
animal studies relating to reproductive and developmental effects; the Post-Stage 2 animal
studies are presented in Section 4.1.
Six-Year Review 3 Technical Support Document	4-39
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Exhibit 4.1: Summary of Results from Pre-Stage 2 and Post-Stage 2 Epidemiology
and Animal Toxicity Reproductive/Developmental Studies
Reproductive/
Developmental
Endpoint
Birth Weight
Epidemiology Studies
Animal Toxicity Studies
Pre-Stage 2
Based on 13 primary studies (Savitz et al.
(2005); Toledano et al.( 2005); Wright et al.
(2004); Wright et al. (2003); Yang (2004);
Jaakkola et al.(2001); Kallen and Robert
(2000); Dodds et al. (1999); Gallagher et al.
(1998); Kanitz et al. (1996); Bove et al. (1995);
Savitz et al.(1995); Kramer et al.(1992)) and 8
reviews (Bove et al. (2002); Graves et al.(
2001); Villanueva et al. (2001); Nieuwenhuijsen
et al. (2000); Reif et al. (2000); WHO (2000);
Craun, ed. (1998); Reif et al. (1996)), there was
some evidence, although inconsistent, for an
association between birth weight outcomes and
maternal DBP exposure.
An effect on pup weight was observed with DBCM,
BDCM and chlorite (Borzelleca and Carchman, 1982;
Christian etal., 2001a; CMA, 1996).
Post-Stage 2
Small for Gestatio
Based on 11 primary studies (Hoffman et al.
(2008a); Patelarou etal. (2011);
Grazuleviciene et al. (2011); Villanueva et al.
(2011); Hinckley et al. (2005); Lewis et al.
(2006); Yang et al. (2007); Rivera-Nunez and
Wright (2013); Kumar et al. (2013);
Danileviciute et al. (2012); Zhou et al. (2012))
and 1 meta-analysis (Grellier et al. (2010),
there is suggestive) (but not conclusive)
evidence of a small association between
increased THM or HAA in drinking water and
low birth weight outcomes.
nal Age
Pup weights were unaffected in rats given water
containing mixtures of DBPs (Narotsky et al., 2008,
2013, 2015).
Pre-Stage 2
Based on 10 primary studies (Porter et
al.(2005); Savitz et al. (2005); Infante-Rivard
(2004); Wright et al. (2004); Wright et al.
(2003); Jaakkola et al. (2001); Kallen and
Robert (2000); Dodds et al. (1999); Bove et al.
(1995); Kramer et al.(1992)) and 6 reviews
(Bove et al. (2002); Graves et al. (2001);
Villanueva et al. (2001); Reif et al. (2000);
Craun, ed. (1998); Reif et al. (1996)), there was
some evidence, although inconsistent, for an
association between SGA outcomes and
maternal DBP exposure.
Reduced fetal weight was observed at high doses of
chloroform, DCAA and TCAA (Epstein et al. (1992);
Fisher et al., (2001); Ruddick et al., (1983); Smith et
al., (1989b), (1992); Thompson et al., (1974)).
Post-Stage 2
Based on 13 primary studies (Hoffman et al.
(2008a); Patelarou etal. (2011);
Grazuleviciene et al. (2011); Costet et al.
(2012); Hinckley et al. (2005); Yang et al.
(2007); Horton et al.(2011); Summerhayes et
al. (2012); Rivera-Nunez and Wright (2013);
Kumar et al. (2013); Aggazzotti et al. (2004);
Danileviciute et al. (2012); Levallois et al.
(2012)) and 1 meta-analysis (Grellier et al.
(2010)), there is suggestive and consistent
evidence of a small positive association
between SGA and some DBP exposure
metrics.
Warren et al. (2006) showed significantly reduced
mean fetal weight with TCAA.
Six-Year Review 3 Technical Support Document	4-40
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Reproductive/
Developmental
Endpoint
Pre-Term Delivery
Epidemiology Studies
Animal Toxicity Studies
Pre-Stage 2
Based on 10 primary studies (Savitz et al.
(2005); Wright et al. (2004); Wright et al.
(2003); Yang (2004); Jaakkola et al. (2001);
Jaakkola et al. (2001); Gallagher et al. (1998);
Kanitz et al. (1996); Savitz et al.(1995); Kramer
et al. (1992)) and 6 reviews (Bove et al. (2002);
Graves et al.( 2001); Villanueva et al. (2001);
Reif et al. (2000); Craun, ed. (1998); Reif et al.
(1996)), there was no evidence of PTD
outcomes and maternal DBP exposure (and
some evidence of an inverse relationship).
No animal studies.
Post-Stage 2
Based on 10 primary studies (Hoffman et al.
(2008b); Patelarou et al. (2011); Costet et al.
(2012);	Hinckley et al. (2005); Yang et al.
(2007); Horton et al. (2011); Kumar et al.
(2013);	Rivera-Nunez and Wright (2013);
Aggazzotti et al. (2004); Lewis et al. (2007))
and 1 meta-analysis (Grellier et al. (2010)),
there is only weak evidence of PTD outcomes
and maternal DBP exposure (a few positive
findings, but more null results).
No new animal studies.
Congenital Anomalies
Pre-Stage 2
Based on 11 primary studies (Shaw et al.
(2003); Cedergren et al. (2002); Hwang et al.
(2002);	Dodds and King (2001); Kallen and
Robert (2000); Dodds et al. (1999); Klotz and
Pyrch (1999); Magnus et al. (1999); Bove et al.
(1995);	Aschengrau et al.(1993); Shaw et al.
(1991); 1 meta-analysis (Hwang and Jakkola
(2003))	and 9 reviews (Bove et al. (2002);
Graves et al. ( 2001); Villanueva et al. (2001);
Nieuwenhuijsen et al. (2000); Reif et al. (2000);
WHO (2000); Craun, ed. (1998); Reifet al.
(1996)),	there was no strong or consistent
evidence of congenital anomalies and maternal
DBP exposure (although inconsistent, the
strongest association with a specific end-point
was for neural tube defects and urinary tract
malformations; there were inconsistent results
related to cardiac anomalies).
Congenital anomalies were observed in most but not
all studies with chloroform, bromoform, BDCM,
DCAA, TCAA, MCAA, chlorite, chlorine and
trichloroacetonitrile (TCAN), and several of these
studies involved cardiac malformations (Abdel-
Rahman et al., (1982); Christ et al. (1996); Christian
et al., (2001a); Couri et al., (1982); Epstein et al.,
(1992); Harrington et al., (1995a); Johnson et al.,
(1998); Meier et al., (1985); Ruddick et al., (1983;
Smith et al., (1989b), (1990), (1992); Thompson et
al., (1974)).
Post-Stage 2
Based on seven primary studies
(Grazuleviciene et al. (2013); Righi et al.
(2012); Iszatt et al. (2011); Luben et al. (2008);
Hwang et al. (2008); Chisholm et al. (2008);
Nieuwenhuijsen et al. (2008)) and 2 meta-
analysis studies (Nieuwenhuijsen et al. (2009);
Hwang et al. (2008)), there is consistent
evidence for an association between THM
exposures and cardiac anomalies (observed in
4 of 5 studies addressing this end-point).
No new animal studies.
Six-Year Review 3 Technical Support Document	4-41
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Reproductive/
Developmental
Endpoint
Fetal Loss
Epidemiology Studies
Animal Toxicity Studies
Pre-Stage 2
Based on 10 primary studies (Savitz et al.
(2005); Toledano et al. (2005); Dodds et al.
(2004); Dodds et al. (1999); Swan et al. (1998);
Waller et al. (1998); Bove et al. (1995); Savitz
et al. (1995); Aschengrau et al. (1993);
Aschengrau et al. (1989)) and 9 review papers
(Bove et al. (2002); Graves et al.( 2001);
Villanueva et al. (2001); Nieuwenhuijsen et al.
(2000); Reif et al. (2000); WHO (2000); Craun,
ed. (1998); Mills et al. (1998); Reif et al.
(1996)), there was inconsistent, but suggestive,
evidence of an association between maternal
DBP exposure and pregnancy loss.
Based on studies with BDCM, TCAA, BCAA, sodium
chlorite and chlorine dioxide there was evidence of
litter resorption, decreased number of implantation
sites, resorbed and dead fetuses and decreased
number of live fetuses per litter (Bielmeier et al.,
(2001), (2004); Couri et al., (1982); Johnson et al.,
(1998); Narotsky etal., (1997); NTP, (1999); Suh et
al., (1983)).
Post-Stage 2
Based on 1 new primary study (Hwang and
Jaakkola (2012), which also included a meta-
analysis with 5 other studies), some evidence
is provided of increased risk of fetal loss and
exposure to THMs.
Some evidence of post-implantation loss was
observed for TCAA (Singh et al., 2005a,b, 2006).
Reviews of the potential mode of action of BDCM
induced pregnancy loss suggest it may be due to
reduced LH secretion and exposure method (gavage
vs. ad libitum drinking water) (NTP, 2006; USEPA,
2006a; Bielmeier et al., 2007). Pre-natal loss was not
observed in Sprague-Dawley rats given water
containing mixtures of DBPs (Narotsky et al., 2008,
2013, 2015) but was noted on F344 rats (Narotsky,
2011).
Male Reproductive Effects
Pre-Stage 2
Based on only 1 study (Fenster et al. (2003)),
no effects were observed on sperm motility or
sperm morphology associated with THMs in
drinking water.
Male reproductive effects were observed for BDCM,
DCAA, DBAA, BCAA, chlorite and bromate (Bhat et
al., (1991); Carlton and Smith, (1985); Christian et
al., (2002a), (2002b); Cicmanec et al., (1991); Katz et
al., (1981), Klinefelter et al., (1995); Linder et al.,
(1994a, 1994b, 1995, 1997); Toth etal., (1992); Tully
etal., (2005); Wolf and Kaiser, (1996). NTP (1998a)
found no effects on male reproductive parameters in
rats treated with with BDCM.
Post-Stage 2
Female Reproduct
Based on 5 new primary studies (Luben et al.
(2007); Iszatt et al.(2013); Zeng et al. (2013);
Nickmilder and Bernard (2011); Xie et al.
(2011)), there were no associations found with
sperm quality, although moderate decreases in
sperm levels were noted with increases in
BDCM and DBCM.
ive Effects
Male reproductive effects, including effects on the
testes and on sperm, were observed with TCAA and
DBAA (NTP, 2007c; Singh et al., 2005ab, 2006).
Effects on sperm counts and motility were observed
in Sprague-Dawley rats given water containing
mixtures of DBPs (Narotsky et al., 2013, 2015)
Pre-Stage 2
Based on 1 study (Windham et al. (2003)), it
was observed that THM exposure may affect
ovarian function; also, BrTHMs especially
DBCM were associated with shorter menstrual
cycles.
There was some evidence, though inconsistent, of
female reproductive effects. (Balchak et al., (2000);
Christian et al., (2002a); Murr and Goodman, (2005).
NTP (1998a) found no effects on female reproductive
parameters in rats treated with with BDCM.
Post-Stage 2
Based on 1 new study (MacLehose et al.
(2008)), there was no evidence of increased
time to pregnancy among women with
exposure to increasing levels of THMs.
TCAA via gavage resulted in reduced ovary weights
(Singh et al., 2005a,b, 2006). No effects on fertility or
pregnancy maintenance were observed in mixtures
of regulated DBPs (Narotsky et al., 2015).
Six-Year Review 3 Technical Support Document	4-42
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
4.1.2.2.2 Summary of Epidemiology and Animal Toxicity Studies on Reproductive and
Developmental Effects
Birth Weight: DBP-associated birth weight reductions, consistent in magnitude (from 26 to 62
grams), were reported in four epidemiology studies (Hoffman et al., 2008a; Zhou et al., 2012;
Grazuleviciene et al., 2013; Rivera-Nunez and Wright, 2013) out of five (not including
Villanueva et al., 2011) examined here among post-Stage 2 D/DBPR studies. Zhou et al. (2012)
also reported (larger) associations (-160 grams; 95 percent CI = -315 - -4) for maternal urinary
TCAA measures among a population subset with more complete questionnaire data.
Collectively, six out of eight studies to date reported statistically significant birth weight
reductions for different DBP exposures including five (Bove et al., 1995; Wright et al., 2003,
2004; Rivera-Nunez and Wright, 2013; Grazuleviciene et al., 2013) out of seven (not including
Hoffman et al., 2008a; Villanueva et al., 2011) studies that examined THM4. Notably, the
associations between specific THMs (chloroform and BDCM) and mean birth weight observed
by Rivera-Nunez and Wright (2013) largely did not persist in multi-pollutant models adjusted for
HAA5. In its earlier review of 15 articles covering fetal growth endpoints, including birth weight
outcomes, in support of the Stage 2 D/DBPR, EPA concluded that the evidence for effects on
fetal growth, including birth weight, was "inconsistent" overall, but noted that a few of the more
recent, higher quality studies provided some evidence of higher risk of low birth weight
associated with maternal DBP exposure during pregnancy (however, such evidence was limited
largely to studies of average differences in a continuous measures of birth weight, rather than
low birth weight outcomes).
The 12 studies (11 original investigations and 1 meta-analysis) published post-Stage 2 are
suggestive of small positive associations between increasing levels of THMs and HAAs in tap
water and increased risk of the adverse birth weight outcomes reviewed in this section. The
suggestive evidence provided by the studies, however, is not conclusive regarding the existence
of an increased risk of adverse low birth weight outcomes due to specific DBP exposures as
indicated by any of the following, especially at concentrations below current regulatory
standards: THM4, BrTHM, specific THMs, HAA5 and HAA9, specific HAAs and maternal
urinary TCAA as biomarkers of DBP exposure. Features of these studies limiting the weight of
evidence include the small magnitude of observed effects, errors in classification or estimation of
DBP exposures, imprecision of observed associations and inconsistent evidence of exposure-
response relationships between increasing categories of DBPs and risk of the adverse low birth
weight outcomes.
Animal toxicity studies that addressed low post-natal birth weight (pup weight) are included in
this section. Sprague-Dawley pup weights were unaffected in studies using chlorinated drinking
water, with or without ozonation; concentrated and chlorinated surface water; or water with
concentrated levels of THM4 and HAA5 (Narotsky et al., 2008, 2013, 2015). No other animal
toxicity studies published subsequent to the Stage 2 D/DBPR were identified which identified
low birth weight as an end-point. Three earlier studies were identified which reported a marginal
postnatial body weight in the F2B generation of ICR Swiss mice administered DBCM in
drinking water in a two-generation study (Borzelleca and Carchman, 1982), decreased F1 and F2
pup body weight in Sprague-Dawley rats administered chlorite in drinking water (CMA, 1996)
and decreased pup weight following parental administration of BDCM in drinking water to
Six-Year Review 3 Technical Support Document	4-43
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Sprague-Dawley rats (Christian et al., 2001a). No NOAEL or LOAEL values were identified in
the latter study due to reduced feed and water consumption of the parental females.
The weight of evidence from epidemiology studies continues to support a potential health
concern for an increased risk of the adverse birth weight outcomes; this effect was not observed
in recent animal studies using chlorinated water. Based on the limited available animal data, the
weight of evidence does not support an effect on birth weight.
Small for Gestational Age: The weight of evidence provided by the fourteen post-Stage 2
epidemiologic articles (13 primary studies and 1 meta-analysis) suggest that there is a small
positive association between exposure to DBP during pregnancy and risk of an SGA infant.
Although often failing to achieve statistical significance, there was consistency reported in RR
estimates for SGA and different exposure metrics including THM4 and DCA.
In its earlier review of 16 articles covering fetal growth endpoints in support of the Stage 2
D/DBPR, EPA concluded that the evidence for effects on fetal growth, including SGA, was
"inconsistent" overall, but noted that a few of the more recent, higher quality studies provided
some evidence of higher risk of SGA associated with maternal DBP exposure during pregnancy.
This suggestion that there are associations with SGA appears to be strengthened in that there is
more consistency across the Post-Stage 2 studies and collectively across all studies for some of
the DBP metrics. Collectively, most studies (pre- and post-Stage 2 D/DBPR) show consistently
elevated effect estimates that are small in magnitude based on high third trimester THM4
exposures.
The DCAA exposures findings measured by using water concentration data are remarkably
similar (RR range = 1.05 to 1.28) in four out of five pre- and post-Stage 2 studies examining
individual HAAs (Wright et al., 2004; Porter et al., 2005; Hinckley et al., 2005; Levallois et al.,
2012; Rivera-Nunez and Wright, 2013). One study reported identical ORs and CIs (OR = 1.4; 95
percent CI = 1.1 - 1.9) for the highest ingestion quartiles for DCAA and HAA9 (Levallois et al.,
2012), whereas another study reported higher DCAA ORs than those for HAA5 (Hinckley et al.,
2005). An additional study that examined biomonitoring data found a higher OR for SGA (1.8;
95 percent CI = 0.9 - 3.7) for detectable urinary TCAA concentrations (Costet et al., 2012).
Consistent results were not noted for other individual DBPs. However, some DBPs (including
the brominated HAAs) and other DBP mixture surrogates (e.g., total organic halides (TOX))
have not been sufficiently examined.
Animal studies generally do not use the term Small for Gestational Age (SGA) as an end-point.
Studies that used "low fetal weight" as an end-point are discussed here. Warren et al. (2006)
administered TCAA via gavage to pregnant Sprague-Dawley rats on GD 6-15. Mean fetal body
weights were significantly reduced at this dose. Earlier studies were identified which reported
low fetal weight following administration of chloroform, DCAA and TCAA. A decrease in fetal
weight was reported in pups of Sprague-Dawley rats administered chloroform by gavage during
GD 6-15 (Thompson et al., 1974; Ruddick et al., 1983). Reduced fetal body weight was observed
in two studies in which Long-Evans rats were administered DCAA by gavage on GD 6-15
(Epstein et al., 1992; Smith et al., 1992). Fetal weight was decreased in Long-Evans rats
administered TCAA by gavage on GD 6-15 (Smith et al., 1989b) and in Sprague-Dawley rats
administered TCAA by gavage on GD 6-15 (Fisher et al., 2001).
Six-Year Review 3 Technical Support Document	4-44
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
The weight of evidence from epidemiology studies continues to support a potential health
concern for an increased risk of an SGA infant born to women exposed to DBPs during
pregnancy; decreased fetal weights were observed in several animal studies following
administration of chloroform, DCAA, or TCAA to rat dams during pregnancy. Based on the data
from these animal studies, the weight of evidence supports an effect on reduced fetal weights in
rats.
Pre-Term Delivery: The 11 post-Stage 2 studies (10 primary studies and 1 meta-analysis)
provide weak evidence in support of the hypothesis that exposure to DBP during pregnancy
increases the risk of PTD. There are but a few positive findings and a larger set of null findings.
There was no consistency in the magnitude of measures of association across studies, statistically
significant associations between DBP exposure and PTD were only rarely and inconsistently
observed, and exposure-response trends were observed in only two of the studies, namely in one
assessment of PTD risk and THM4 (Yang et al., 2007) and one assessment of increasing levels
of total organic halides (Horton et al., 2011). In its earlier review of 16 articles covering fetal
growth endpoints in support of the Stage 2 D/DBPR, EPA concluded that the evidence for effects
on fetal growth, including PTD was "inconsistent" overall.
No animal toxicity studies reporting PTD as an endpoint were identified either prior to and since
the Stage 2 D/DBPR.
The weight of evidence from epidemiology studies continues to support a potential health
concern for an increased risk of PTD in women exposed to DBPs during pregnancy. The
evidence related to DBP exposures remains weak with some inconsistencies; namely, some of
the more recent studies have detected associations with different DBP metrics for both PTD (<37
weeks) and very early PTD (<32 weeks).
Congenital Anomalies: There is consistent evidence for an association between THMs and
cardiac anomalies in the post-Stage 2 epidemiology studies. Although often failing to achieve
statistical significance, there was consistent evidence of an elevated risk, with exposure-response
trends observed between THM4 and risk of cardiac defects, as well as observed elevated risks
associated with BrTHMs and chloroform. Associations between THM exposures and
cardiovascular anomalies were noted in one of three pre-Stage 2 studies that focused on this end-
point and in four of five post-Stage 2 studies that included assessment of this endpoint. THM
levels were markedly low in the one study (Righi et al., 2012) that reported an odds ratio for
major cardiac defects of 1.25 for THM exposures > 2.5 |j,g/L compared with a referant of < 1
Hg/L, but did not observe associations between THMs and major cardiac defects for THM levels
> 5 |j,g/L compared to a referant of < 5 |_ig/L. Associations with ventricular septal defects, in
particular, were noted in three of these studies.
For associations between DBP and other (non-cardiac) anomalies, the seven post-Stage 2
epidemiologic studies (Grazuleviciene et al. (2013); Righi et al. (2012); Iszatt et al. (2011);
Luben et al. (2008); Hwang et al. (2008); Chisholm et al. (2008); Nieuwenhuijsen et al. (2008))
provide no evidence or, at most, weak and inconsistent evidence. The two studies evaluating an
"any defect" endpoint did not observe exposure-response trends. The three assessments of DBP
and risk of hypospadias found no elevations in risk. Associations between DBPs (THM4,
BrTHMs, chlorite and chlorate) and urogenital, musculoskeletal, cleft palate and several other
Six-Year Review 3 Technical Support Document	4-45
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
specific anomalies were inconsistently noted in these studies, providing at most limited evidence
for an association with these defects.
In its earlier review of 12 articles covering congenital anomaly endpoints in support of the Stage
2 D/DBPR, EPA found seven studies, including a meta-analysis, supporting the hypothesis that
DBP exposure is associated with congenital anomalies; one study that showed inconsistent
results; and four studies that reported little evidence of an association between DBP and risk of
congenital anomalies. Birth defects most consistently identified as being associated with DBPs
included neural tube defects and urinary tract malformations. The post-Stage 2 studies also
reported evidence for associations between DBP and these endpoints. Hwang et al. (2008)
reported a statistically significant association of THM4 exposure (ORs between 1.2 and 1.7) with
obstructive urinary tract increases and Chisholm et al. (2008) reported an association of THM4
exposure (ORs between 1.1 and 1.4) with urogenital defects that was not statistically significant.
Nieuwenhuijsen et al. (2008) reported no increase in urinary tract defects. Although they had
very low THM4 levels in general, Righi et al. (2012) reported a small but not statistically
significant association of low level THM4 exposure (ORs between 1.2 and 1.3) and urinary tract
defects, but not at slightly higher THM4 levels. Grazuleviciene et al. (2013) reported statistically
significant association with BDCM exposure (ORs between 1.7 and 2.9) and urogenital
anomalies, as well as associations for THM4 and chloroform (ORs between 2.2 and 2.5) that
were not statistically significant. In a meta-analysis, Neiuwenhuijsen et al. (2009) reported an
increased risk for urinary tract defects (OR 1.33) comparing high to low chlorination byproduct
exposures, although the increased risk was not statistically significant.
None of the three post-Stage 2 studies (Chisholm et al., 2008; Nieuwenhuijsen et al. 2008; Righi
et al., 2012) that included neural tube defects observed associations with THM4 exposure.
No animal toxicity studies published subsequent to the Stage 2 D/DBPR were identified which
identified congenital anomalies as an end-point. Twelve animal studies were published prior to
Stage 2 D/DBPR in which congenital anomalies were observed, including increased frequency of
bilateral extra lumbar ribs, increased sternebral anomalies, increased cardiac malformations,
decreased fetal crown-rump length, increased soft tissue anomalies, increased number of
ossification sites and delayed skeletal ossification. One or more of these effects were observed
with chloroform (Thompson et al., 1974), bromoform (Ruddick et al., 1983), BDCM (Christian
et al., 2001a), DCAA (Epstein et al., 1992; Smith et al., 1992), TCAA (Smith et al., 1989b;
Johnson et al., 1998), MCAA (Smith et al., 1990), chlorite (Harrington et al., 1995a; Couri et al.,
1982), chlorine (Abdel-Rahman et al., 1982) and/or TCAN (Meier et al., 1985; Christ et al.,
1996) in Sprague-Dawley or Long-Evans rats. Delayed skeletal ossification was observed with
chlorite in New Zealand rabbits. No teratogenic effects were observed in Sprague-Dawley rats
administered chloroform by gavage in corn oil (Ruddick et al., 1983). Reproductive effects
reported for haloacetonitriles which used tricaprylin as a vehicle for gavage are not included here
because tricaprylin has been shown to be a developmental toxicant and may potentiate the effects
observed in those studies. Based on the data from these animal studies, the weight of evidence
supported an effect on the frequency of congenital anomalies.
The weight of evidence from epidemiology studies continues to support a potential health
concern for an increased risk of congenital anomalies in humans, including evidence of increased
Six-Year Review 3 Technical Support Document	4-46
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
risk of cardiac defects, neural tube defects and urinary tract malformations from exposure to
THM4; animal studies also support the risk of an increased frequency of congenital anomalies.
Fetal Loss Conclusions: The 11 epidemiology studies provide some evidence in support of the
hypothesis that exposure to DBP during pregnancy increases the risk of fetal loss/stillbirth. In its
review of the 10 studies available at the time of the Stage 2 D/DBPR, EPA concluded that
although the evidence was inconsistent overall, there was suggestive evidence of an association
between fetal loss and chlorinated water or DBP exposure. Hwang and Jaakkola (2012), the only
new publication on this endpoint since the Stage 2 D/DBPR, observed an overall small
association with fetal loss (stillbirth) in their own case-control study (OR = 1.10, 95 percent CI =
1.00 - 1.21 for medium exposure and OR = 1.06, 95 percent CI = 0.96 - 1.17 for high exposure)
and in their meta-analytic summary of the evidence contributed by five studies (Aschengrau et
al. 1993; Bove et al. 1995; Dodds et al. 1999; Toledano et al. 2005; Dodds et al. 2004) plus their
own case-control study (OR = 1.11, 95 percent CI = 1.03 - 1.19), albeit with marked
heterogeneity across studies.
In addition to the epidemiology evidence noted above on fetal loss/stillbirth, the collective
available animal toxicity data at the time of the Stage 2 D/DBPR provided evidence of fetal loss
in terms of litter resorption, post-implantation loss and decreased litter size. A study in inbred
Charles Foster rats administered TCAA via gavage on GD 6-15 at doses up to 1800 mg/kg/day
reported post implantation loss (Singh et al., 2005a,b; 2006). Prenatal survival in Sprague-
Dawley rats was unaffected in studies using chlorinated drinking water, with or without
ozonation; concentrated and chlorinated surface water; or water with concentrated levels of
THM4 and HAA5 (Narotsky et al., 2008, 2013, 2015). However, mixtures of THM4 and HAA5
contributed to pregnancy loss from a mixture of nine DBPs in F344 rats (Narotsky et al., 2013).
Studies published prior to Stage 2 D/DBPR also described full litter resorptions induced by
BDCM administered in corn oil or in an aqueous vehicle in F334 rats (Narotsky et al., 1997).
Full litter resorptions were observed following aqueous gavage to F344 rats but not to Sprague-
Dawley rats (Bielmeier et al., 2001, 2004). In a study conducted in male and female Sprague-
Dawley rats administered BCAA in drinking water, the number of live fetuses per litter and the
total number of implants per litter were significantly decreased (NTP, 2009). A study in which
TCAA was administered in drinking water to pregnant Sprague-Dawley rats resulted in
significant increases in the number of resorptions and number of implantation sites (Johnson et
al., 1998). In two studies in Sprague-Dawley rats, resorbed and dead fetuses were observed
following administration of sodium chlorite in drinking water (Couri et al., 1982) and decreases
in number of implants per litter and number of live fetuses per dam were observed following
administration of chlorine dioxide in drinking water (Suh et al., 1983).
Reviews of pregnancy loss following BDCM exposure and a discussion of the potential mode of
action were published by NTP (2006) and USEPA (2005d). Reduced LH secretion (Bielmeier et
al., 2002) and reduced luteal responsiveness to LH (Bielmeier et al., 2003) may both contribute
to BDCM-induced full litter resorption in F344 rats (US EPA, 2006a; Bielmeier et al., 2007).
However, several investigators have failed to observe full litter resorption in Sprague-Dawley
rats exposed to BDCM, suggesting that these effects may be strain-specific (Bielmeier et al.,
2001; Christian et al., 2001a; Ruddick et al., 1983). A BDCM expert panel (Health Canada,
2008a) concluded that adverse reproductive and developmental effects of BDCM were observed
Six-Year Review 3 Technical Support Document	4-47
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
only at very high, maternally toxic doses; were not consistent between animal models; and varied
with method of administration. Data for mode of action are limited.
The weight of evidence from epidemiology studies continues to support a potential health
concern for an increased risk of fetal loss/stillbirth; animal studies provide evidence of fetal loss
in terms of litter resorption, post-implantation loss and decreased litter size.
Male Reproductive Effects: Although based on a small number of studies, the weight of
evidence provided by the five post-Stage 2 epidemiologic studies suggests that there is either no
association, or, at most, a small association, between exposure to DBP and male reproductive
outcomes. The four studies of sperm quality and DBP exposure were largely negative, although
decreasing sperm concentration with increasing BrTHM exposure was observed in three studies
(Zeng et al., 2014; Iszatt et al., 2013; Luben et al., 2007). In its earlier review of one article
addressing male reproductive effects endpoints in support of the Stage 2 D/DBPR, EPA
concluded that no association was found between THM4 exposure and semen quality.
Adverse male reproductive effects were observed in animal toxicity studies conducted since
Stage 2 D/DBPR. TCAA, when administered via gavage on GD 6-15 to inbred Charles Foster
rats, resulted in a reduction in mean testes weight and length of the seminiferous tubules (Singh
et al., 2005a, b, 2006). In a 13-week study in B6C3F1 mice and F344 rats, DBAA was
administered in drinking water and resulted in testicular atrophy, reduced testicular weight,
sperm motility and sperm concentration in rats and delayed spermiation in mice and rats (NTP,
2007c). Reduced sperm counts and reduced sperm motility were reported in Sprague-Dawley
rats given concentrated and chlorinated surface water, or water with concentrated levels of
THM4 and HAA5 (Narotsky et al., 2013, 2015).
In studies conducted prior to Stage 2 D/DBPR, the following effects were observed following
BDCM administration in drinking water: sperm velocities were significantly decreased in F344
rats (Klinefelter et al., 1995); delayed sexual maturation in F1 males with reduced body weight
in Sprague-Dawley rats (Christian et al., 2002a). No effects on sperm characteristics were noted
in studies by NTP (1998a) or Christian et al., (2002a) in Sprague-Dawley rats administered
BDCM in drinking water. NTP (1998a) conducted a short-term reproductive and developmental
screening test with BDCM administered in drinking water to Sprague-Dawley rats and concluded
that BDCM was not a reproductive or developmental toxicant.
Studies were also conducted with DCAA and resulted in adverse effects on the testes, including
testicular germinal epithelial degeneration and aspermatogenesis, in Sprague-Dawley rats and
beagle dogs (Katz et al., 1981); decreased testis weight, tissue atrophy and few spermatocytes
and no mature spermatozoa in the seminiferous tubules in Sprague-Dawley rats (Bhat et al.,
1991); testicular changes, including syncytial giant cell formation and degeneration of testicular
germinal epithelium, in beagle dogs (Cicmanec et al., 1991); significant reductions in the
absolute weight of the preputial gland and epididymis and effects on sperm morphology and
decreased sperm counts in Long-Evans rats (Toth et al., 1992); and a decrease in epididymal
weight and epididymal sperm count in Sprague-Dawley rats (Linder et al., 1997).
Linder et al. (1994a, b, 1995) conducted gavage studies with DBAA in Sprague-Dawley rats
which resulted in reduced testis and epididymis weights, decreased sperm counts and
Six-Year Review 3 Technical Support Document	4-48
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
histopathological evidence of altered spermiation and reduced sperm motility. In another study,
DBAA was administered in water to Sprague-Dawley rats and resulted in altered sperm
production and some epididymal tubule changes and small or absent epididymis and small testes
(Christian et al., 2002b).
Decreased male fertility due to disruption of spermatid differentiation was observed in C57BL/6
mice following BCAA administration (Tully et al., 2005); a decrease in epididymal sperm
density was observed in rats following bromate administration in drinking water (Wolf and
Kaiser, 1996) and abnormal sperm were observed following administration of chlorite in
drinking water to Long-Evans rats (Carlton and Smith, 1985).
The weight of evidence from epidemiology studies appears not to support a potential health
concern for an increased risk of male reproductive effects; animal studies provide evidence of a
number of adverse effects, including testicular effects and decreased sperm counts.
Female Reproductive Effects: There was only one study available at the time of the Stage 2
D/DBPR on female reproductive effects. Windham et al. (2003) found that exposure to THMs
may affect ovarian function. The BrTHMs, notably DBCM, were associated with significantly
shorter menstrual cycles. There was also only one new study, MacLehose et al. (2008), since the
Stage 2 D/DBPR addressing female reproductive effects, specifically time-to-pregnancy. The
authors found no evidence of an increase in time to pregnancy with increased exposure to DBPs.
Reduced ovary weights were observed in an animal toxicity study conducted since Stage 2
D/DBPR in inbred Charles Foster rats administered TCAA via gavage on GD 6-15 (Singh et al.,
2005a, b, 2006).
Prior to Stage 2 D/DBPR, NTP (1998a) conducted a short-term reproductive and developmental
screening test with BDCM administered in drinking water to Sprague-Dawley rats and concluded
that BDCM had no effects on female reproductive parameters in rats treated with with BDCM.
Other studies observed effects on female reproductive outcomes. A study in Sprague-Dawley
rats administered BDCM in drinking water resulted in a marginal effect on estrous cyclicity in
F1 females and a small but significant delay in F1 generation sexual maturity (Christian et al.,
2002a); studies with DBAA in Sprageu-Dawley rats resulted in alterations of the estrous cycle
(Balchak et al., 2000) and in increased circulating serum estradiol levels with no observed
change in the estraou cycle (Murr and Goodman, 2005).
Insufficient epidemiology information is available to inform an updated understanding of
adverse female reproductive effects; animal studies provide marginal evidence with respect to
effects on the female reproductive system.
4.1.3 Mixtures of Chlorination Organic DBPs
Multiple studies have been conducted researching developmental and reproductive effects from
mixtures of DBPs. Narotsky et al. have conducted five studies, including two multi-generational
studies, since 2008, researching the reproductive and development effects from regulated DBPs
and from environmentally-realistic complex mixtures of DBPs formed during disinfection with
chlorine or ozone/chlorine. The most recent Narotsky et al. (2015) evaluated a drinking water
mixture of the four regulated THMs and five regulated HAAs in a multi-generational
Six-Year Review 3 Technical Support Document	4-49
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
reproductive toxicity bioassay. Some additional studies addressing DBP mixtures for other end-
points are presented in Appendix A.
Simmons et al. (2002, 2004) describes the origins of the "Four-Lab Study", so called because it
draws upon the expertise and skills from four EPA Office of Research and Development
laboratories/centers. Simmons et al. (2008) presented results from the first series of studies
proposed in an EPA peer-reviewed research plan called "Integrated Disinfection Byproducts
Mixtures Research: Toxicological and Chemical Evaluation of Alternative Disinfection
Treatment Scenarios". The research of this multidisciplinary team focused on integration of
toxicological and chemical evaluation of complex mixtures of DBPs and their
reproductive/developmental effects. These effects were identified as endpoints of concern in
epidemiologic studies and EPA determined it was feasible to conduct in vivo animal studies
investigating these endpoints. The first series of experiments studied the stability of the DBPs
and methods for concentrating them. Sprague-Dawley rats were then exposed to these
concentrates in treated water as their sole source of drinking water. These experiments
determined that using reverse osmosis membranes to concentrate water samples, with a 100-fold
concentration factor as the target reverse osmosis concentration, is the optimal approach for
treating water samples to be used in toxicological studies.
The following Narotsky et al. papers present results from the Four-Lab Study of DBP mixtures.
Narotsky et al. (2008) used an in vivo toxicity screen to evaluate the developmental effects of a
mixture of DBPs, using finished drinking water from a city in Ohio. The water was treated by
one of two methods, either conventional treatment with disinfection by chlorination or
conventional treatment with disinfection by ozonation followed by chlorination. The water was
concentrated approximately 100-fold and administered to Sprague-Dawley rats on GD 6-16. The
rat litters were examined on postnatal days (PND) 1 and 6, with no effects observed on prenatal
survival, postnatal survival, or pup weights from either the water treated by conventional
treatment/chlorination or the water treated by ozonation/chlorination.
Narotsky et al. (2011) assessed the combined toxicity of DBPs. Pregnant F344 rats were
administered mixtures of THM4, HAA5, or the full mixture of all the chemicals (nine DBPs) by
gavage on GD 6-20. All three mixtures caused pregnancy loss at > 613 |imol/kg-day. Resorption
rates were increased in the group administered HAA5 at 613 |imol/kg-day and the group
administered the nine DBPs at 307 |imol/kg-day. Eye malformations were increased in the
HAA5 group at > 308 |imol/kg-day. The authors concluded that both HAA5 and THM4
contributed to the pregnancy loss from the full mixture of nine DBPs and that the presence of
THM4 in the full mixture appeared to reduce the incidence of HAA-induced eye defects.
Narotsky et al. (2013) used a multi-generational bioassay with Sprague-Dawley rats to evaluate
an "environmentally relevant whole mixture of DBPs representative of chlorinated drinking
water." Surface water used as a source for a utility was treated, filtered and concentrated based
on TOC to 136-fold greater than the unconcentrated filtered water. The concentrated water was
chlorinated and was provided as drinking water to pregnant female rats (P0 generation) during
gestation and lactation. Weanlings (F1 generation) were also exposed to the treated drinking
water and were bred to produce an F2 generation. One set of controls received deionized water
and another set of controls received the unchlorinated concentrate. However, this latter set of
Six-Year Review 3 Technical Support Document	4-50
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
controls was discontinued on GDI 9 because of repeated clogging of the water delivery system.
The study was conducted with 2 sets of replicates each consisting of 100 animals (60 study, 40
controls), although the second replicate set was conducted for the initial part of the study only
and halted on PND 6 of the F1 generation. No evidence of toxicity was observed for the P0
dams. No adverse effects were observed for pup weight, prenatal loss, pregnancy rate, gestation
length, puberty onset in males, growth, estrous cycles, hormone levels and most neurobehavioral
endpoints. Slight, though statistically significant (at 0.05 level), effects observed included
delayed puberty for F1 females, reduced caput epidydimal sperm counts in F1 adult males and
increased incidence of thyroid follicular cell hypertrophy in P0 and adult F1 females. The
authors concluded that their multigenerational reproductive and developmental study with an
environmentally relevant mixture of DBPs yielded predominately negative results, although the
slight but significant effects noted warranted further study.
Narotsky et al. (2015) conducted a multi-generational bioassay with Sprague-Dawley rats similar
to the Narotsky et al. (2013) study above but with specific concentrations of THM4 and HAA5
rather than the "whole mixture" of DBPs. The THM4 and HAA5 levels studied reflected
concentrations that were Ox, 500x, lOOOx and 2000x the MCLs of 0.08 mg/L and 0.06 mg/1,
respectively. In this study the authors found that the mixtures up to 2000x the MCLs had no
adverse effects on fertility, pregnancy maintenance, prenatal survival, postnatal survival or birth
weights. F1 pup weights though unaffected at birth were reduced on PND 6 at the 2000x dose
and PND 21 at the lOOOx dose. Body weights were also reduced for the post-weaning F1
generation at the 2000x dose and water consumption was reduced for the post-weaning F1
generation at 500x dose. Onset of puberty was delayed for both males and females at the lOOOx
and 200x doses. Males at the 2000x dose had a small but significant increased incidence of
retained nipples and compromised sperm motility. The authors concluded that the mixture of
THM4 and HAA5 at concentrations 500x greater than the MCLs had no adverse effects and that
even at 2000x the MCLs did not affect the animal's ability to reproduce. The authors also noted
the lack of effect on prenatal survival and birth weight in this animal study contrasted with
associations reported for those end-points in some epidemiological studies. While reproduction
was unaffected, delay in the onset of puberty in both sexes and retained nipples and reduced
sperm motility in males was observed indicating some effect on endocrine physiology. The
authors commented that these latter effects may have been due to reduced water consumption
and reduced postnatal body weights.
4.2 Regulated Inorganic DBPs
This section addresses the health effects that are associated with the regulated inorganic DBPs,
specifically bromate and chlorite.
An overview of studies is provided below. Additional information for the pre-Stage 2 studies are
provided in Appendix A.
Six-Year Review 3 Technical Support Document	4-51
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
4.2.1 Bromate
Basis for the MCLG
In the Stage 1 D/DBPR, EPA established an MCLG of zero for bromate based on a weight of
evidence evaluation of both the cancer and noncancer effects indicating that bromate is a
"probable or likely human carcinogen" (USEPA, 1994a, 1998b). The MCLG was based on an
increase in kidney and thyroid tumors in several rat studies (Kurokawa et al., 1986a, 1986b;
DeAngelo et al., 1998). Insufficient evidence exists regarding the mode of carcinogenic action of
bromate; thus, the low-dose extrapolation approach was used because it is more protective of
public health (USEPA, 1998b). An EPA IRIS assessment established an RfD of 0.004 mg/kg/day
for bromate in 2001 (USEPA, 2001b) based on aNOAEL of 1.5 mg/kg/day for potassium
bromate (equivalent to 1.1 mg/kg/day bromate) for renal effects (DeAngelo et al., 1998) and the
application of an uncertainty factor of 300. The RfD value did not change in the Stage 2
D/DBPR due to the lack of significant new health effects data for systemic effects and EPA did
not revise the MCLG at that time (USEPA 2003c, 2006a).
New Information Available Since Development of Stage 2 D/DBPR
Cancer
NTP conducted non-standard, shorter-term bioassays with two different transgenic mouse
strains, in which sodium bromate was administered in drinking water for 27 and 43 weeks (NTP,
2007b). These mouse strains, Tg. AC hemizygous (gain of oncogene function {Ha ras)) and p53
haploinsufficient (loss of heterogeneity in a critical cancer gene (TrP53)) have been reported to
detect both nongenotoxic and genotoxic carcinogens and are susceptible to the rapid
development of cancer. Sodium bromate did not cause cancer in Tg. AC hemizygous or p53
haploinsufficient mice exposed to 80, 400 and 800 mg/L. NTP concluded that since sodium
bromate did cause cancer in other studies with different rodents, these transgenic mouse models
are not a sensitive means of evaluating the carcinogenicity of sodium bromate. Nonneoplastic
changes were observed in the thyroid and kidney for the Tg. AC mice.
Other
California Environmental Protection Agency (Cal EPA) (2009) set a Public Health Goal (PHG)
of 0.1 ppb for bromate based on the de minimus cancer risk level calculated from the DeAngelo
et al. (1998) study. ANOAEL of 1.1 mg/kg/day and a LOAEL of 6.1 mg/kg/day based on renal
urothelial hyperplasia were identified. The NOAEL was used as the point-of-departure for Cal
EPA's PHG. For noncancer effects, Cal EPA calculated an RfD of 0.004 mg/kg/day (identical to
EPA's RfD), based on kidney effects from the DeAngelo et al. (1998) study. WHO (2005a)
accepted the finding of carcinogenicity for bromate and established a Practical Quantification
Level of 10 [j.g/L based on analytical and technical feasibility limitations.
Mode of Action
Although bromate has been shown to be carcinogenic in animals with species differences in
sensitivity (rat>mouse>hamster), bromate has not been found to cause cancer in humans.
Possible modes of action for bromate-induced cancer, include thiol-associated (e.g., GSH-
Six-Year Review 3 Technical Support Document	4-52
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
related) oxidative damage to guanine in DNA plus the potential for accumulation of a.2Li-globulin
in male rat kidneys. These topics and others were addressed during a MOA workshop funded by
the American Water Works Association (AWW A) Research Foundation (Bull and Cotruvo,
2006). The workshop participants concluded that there was a "clear path for conducting studies
relevant to human health risks and laid out a potential research plan for filling data gaps." The
research plan included examination of the presystemic toxicokinetics of bromate at doses
representative of exposures through drinking water and studies to identify key events supportive
of a nonlinear MOA. Two possible genotoxic effects that might result from carcinogenic doses
of bromate in male rats were identified as formation of 8-hydroxyguanine DNA adducts
(oxidative damage) and production of micronuclei. Both of these effects can be non-linear with
respect to dose. Between 2006 and the present, a number of studies based on elements identified
in the workshop research plan have been conducted with results published in the peer reviewed
literature as described below.
Dodd et al. (2013) conducted study of male F344 rats exposed to potassium bromate in drinking
water at concentrations of 5 to 400 mg/L with observation periods of 2 and 13 weeks. After
sacrifice the kidneys, liver, lung, thyroid and tunis vaginalis were examined histologically; liver
kidney and thyroids weights were determined. Blood samples were analyzed for aspartate
transaminase, alanine transaminase (ALT), blood udea nitrogen, lactic dehydrogenase and
creatinine. The NOAEL was 6.2 mg/kg/day with a LOAEL of 12.6 mg/kg/day for bromate ion.
At the LOAEL there was a slight increase in kidney weight and hyaline droplets in the kidney
tubules. No effects were seen in the other organs. The hyaline droplets were present at 2 and 13
weeks.
The presence of hyaline droplets in the Dodd et al. (2013) study suggests the accumulation of
a.2Li-globulin in kidneys. Renal accumulation of a.2Li-globulin in kidneys is unique to male rats and
contributes to their carcinogenic response (Umemura and Kurokawa, 2006). This response is not
observed in humans or in female rats. However, bromate is associated with kidney tumors in
female rats, thus other MO As are likely to be involved in the tumor response. It is possible that
the cell proliferation observed in female rats could result from oxidative stress and/or
cytotoxicity. The correlation between formation of 8-oxodG and tumor response in female rats
suggests that dose-response information from the female rat is more relevant to human risk
assessment and that oxidative stress is the likely mechanism for cancer risk in humans.
The contribution of oxidative stress to bromate-induced cancer in male F344 rats was evaluated
in a drinking water study for exposure ranging from 2 to 100 weeks (Delker et al., 2006). Gene
expression analyses were performed on kidney, thyroid and mesothelial cell RNA since chronic
exposures to bromate have been shown to cause renal cell tumors in rats, hamsters and mice and
testicular mesothelial tumors in rats. The Delker et al. (2006) results suggest that the dose of
bromate must reach a threshold before tissue oxidation occurs and that gene expression profiles
may be predictive of these changes in the kidney.
Yamaguchi et al. (2008) conducted a 16-week study of potassium bromate in drinking water in
cancer prone Big Blue male rats using drinking water with concentrations of 0.02 to 500 ppb
(doses of 0.001 to 0.044 mg/kg/day). The NOAEL for the formation of 8-oxodG was 0.007
mg/kg/day and the LOAEL was 0.044 mg/kg/day. The LOAEL dose was accompanied by a
significant (p<0.05) increase in the GC:TA transversions associated with 8-oxodG nucleotides,
Six-Year Review 3 Technical Support Document	4-53
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
inflammatory cell infiltration, tubular regeneration and histological hyaline degeneration
(p<0.01). The LOAEL for hyaline degeneration was lower than that for 8-oxodG (0.002
mg/kg/day (p<0.05). The authors concluded that there could be a no effect level for bromate
mutagenicity to support a nonlinear approach for the cancer assessment, at least as it applied to
the kidney tumors. Group sizes in the Yamaguchi et al. (2008) study were five males/dose and
histological examinations were only conducted on the kidney. Thus, this study by itself is not
sufficient to demonstrate nonlinearity for the bromate-induced tumors that occur in both males
and females
Although bromate has been shown to induce genetic damage in vitro and to induce mutations in
the kidney of exposed rats, it is not clear whether bromate is a mutagenic carcinogen. The
evidence suggests that bromate's mutagenic activity is mediated by the formation of oxidative
damage to the DNA, resulting in chromosomal damage (Moore and Chen, 2006). Zhang et al.
(2011) conducted an in vitro study of cytotoxicity and cellular damage using cultures of human
and rat kidney cells in combination with assays for cell proliferation, cell morphology,
cytotoxicity and generation of ROS using a variety of techniques. The results indicated that
bromate can induce DNA damage, cell necrosis and cell cycle arrest. The ROS damage appeared
to be independent of or downstream of the DNA damage. DNA adduct formation was
accompanied by GSH depletion with the later a possible stimulus for the formation of ROS.
Bromate at concentrations in water of up to 308 mg/L was found to cause bromination of protein
tyrosines in the proteins that accumulate in the male kidney and may contribute to kidney tumors
in male rats (Koilsetty et al., 2013). The presence of a2u globulin in urine is a characteristic of
male rats but not female rats. Other tissue changes including apoptosis, levels of protein
expression and production of 8-oxodG were identified in the kidneys of both sexes. Based on
their data the authors proposed that a failure to suppress cellular apoptosis could contribute to the
mechanism for bromate-induced kidney cancer in males and females.
Relevance for SYR
The EPA IRIS assessment established the RfD of 0.004 mg/kg/day for bromate in 2001 (USEPA,
2001b), which did not change in the Stage 2 D/DBPR due to the lack of significant new health
effects data for systemic effects. There has been considerable published research on bromate
subsequent to the EPA IRIS assessment. New data on the toxicokinetics of bromate (Bull et al.,
2012) demonstrating extensive in vivo reduction to bromide and possible mechanisms associated
with injury to the kidneys, thyroid and testes (Bull and Cotruvo, 2013; Koilsetty et al., 2013)
provide important data to supplement the EPA IRIS MOA assessment (USEPA, 2001b).
However, the new data, at present, are not robust enough to alter the finding that bromate is a
likely carcinogen with a mode of action that cannot be fully determined. EPA concludes at this
time that since data are not available to conclusively demonstrate nonlinearity for the MOA for
all three observed tumors sites observed in the animal studies, there is insufficient basis for
supporting a change in the MCLG of zero.
Six-Year Review 3 Technical Support Document	4-54
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
4.2.2 Chlorite
Basis for the MCLG
In the Stage 1 D/DBPR, EPA (USEPA, 1994a) proposed an MCLG for chlorite of 0.08 mg/L
based on neurodevelopmental effects in a rat study (Mobley et al., 1990; USEPA, 1994a).
Subsequent to the proposal, EPA (USEPA, 1997b) reviewed and completed a peer review of a
two-generation reproductive study of chlorite in Sprague-Dawley rats (CMA, 1996). In this
study, male and female rats were administered sodium chlorite in drinking water at doses ranging
up to 300mg/L. Reproduction, fertility, clinical signs and histopathology were evaluated in two-
generations of offspring. In the Stage 1 D/DBPR, EPA finalized an MCLG of 0.8 mg/L for
chlorite based on the RfD (0.03 mg/kg/day, an adult tap water consumption of 2 liters/day for a
70 kg adult and an assumed drinking water contribution of 80 percent of total exposure (USEPA,
1998b). The RfD of 0.03 mg/kg/day was derived based on a NOAEL of 35 ppm (3 mg/kg/day
for the chlorite ion) for decreases in absolute brain and liver weight and lowered auditory startle
amplitude at 70 and 300 ppm and the application of an uncertainty factor of 100. Although the
RfD differed from that derived from Mobley et al. (1990) of 0.003 mg/kg/day the use of a lower
uncertainty factor in the assessment based on the CMA study yielded the same MCLG (USEPA,
1994a, 2000b). EPA did not revise the MCLG for chlorite in the Stage 2 D/DBPR (USEPA,
2003c, 2006a).
New Information Available Since Development of Stage 2 D/DBPR
Righi et al. (2012) conducted a case-control study in Northern Italy to investigate the relationship
between drinking water exposure to chlorite, chlorate and THMs and congenital anomalies. A
total of 1,917 cases of congenital anomalies (neural tube, cardiac, diaphragm and abdominal
wall, esophagus (food pipe or gullet), cleft lip and palate, respiratory, urinary tract and
chromosomal anomalies) observed in the period of 2002 to 2005 were studied. The THM
exposure levels were reported to be very low (mean 3.8 + 3.6 |ig/L), and no excess risk of
anomalies were associated with THM exposures. The levels of chlorite (mean 427 + 184 |ig/L)
and chlorate (mean 283 + 79 |ig/L), however, were relatively high. The authors reported that
women exposed to chlorite at levels > 700 |j,g/L were at higher risk of having newborns with
renal defects (OR: 3.30; 95 percent CI = 1.35 - 8.09), abdominal wall defects (OR: 6.88; 95
percent CI = 1.67 - 28.33) and cleft palate (OR: 4.1; 95 percent CI = 0.98 - 16.8); women
exposed to chlorate at levels >200 jag/1 were at higher risk of newborns with obstructive urinary
defects (OR: 2.88; 95 percent CI = 1.09 - 7.63), cleft palate (OR: 9.60; 95 percent CI = 1.04 -
88.9) and spina bifida (OR: 4.94; 95 percent CI = 1.10 - 22). The authors noted that this was the
first study showing an excess risk of different congenital anomalies associated with chlorite
and/or chlorate exposure from drinking water, and that further research using larger datasets was
needed to confirm the observed results.
In an earlier population-based, case-control study from the same area, Aggazzotti et al. (2004)
examined the association between chlorination byproducts and adverse pregnancy outcomes. The
chlorination byproducts investigated in this study were chlorate and chlorite and total and
individual THMs: chloroform, dichlorobromomethane, dibromochloromethane and bromoform.
A total of 1,194 subjects were evaluated in the study, consisting of 343 pre-term (<37 weeks)
births, 239 full-term SGA births (< 10th percentile of birth weight according to standard values
Six-Year Review 3 Technical Support Document	4-55
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
from the Italian Society of Pediatrics) and 612 controls (born >37 weeks and > 10th percentile of
birth weight). Exposure was assessed both by a questionnaire completed by the mothers on their
personal habits during pregnancy and by water samples collected at the homes of the
participants. The median concentrations of chlorate for pre-term births, full-term SGA births and
controls were: 177.00, 250.00 and 216.50 (J,g/L, respectively. No association was found between
pre-term births and exposure to chlorite or to any of the other chlorination byproducts studied.
The authors found, however, that chlorite did show an association with term-SGA births
suggesting a dose-response relationship. When chlorite levels were > 200 [j,g/L and the frequency
of bathing/showering was at least daily, they observed an OR of 1.70 (95 percent CI: 0.97 -
3.00) compared to a referent group with chlorite levels < 200 [j,g/L and a frequency of
bathing/showering less than daily. The authors noted that in their study few women consumed
tap water and they considered that the increased risk was from exposure via inhalation, noting
that while chlorite is not considered volatile, there could be chlorite present in aerosols in shower
vapors. They also noted that an alternative explanation could be that chlorite served as a proxy
for other chlorination byproducts or as a proxy for residual chlorine dioxide used as the
disinfectant (74% of the study population lived in areas where drinking water was treated with
chlorine dioxide or a combination of chlorine and chlorine dioxide).
No other new information has been identified about cancer or noncancer effects. However, new
information is available that may inform a different relative source contribution for chlorite than
was used when EPA developed the MCLG for chlorite under the Stage 1 D/DBPR. If data show
the contribution of chlorite from food or other sources than drinking water to be greater than
previously thought, then the relative source contribution (RSC) from the drinking water
component would decrease.
Data that support lowering the 80 percent RSC contribution from water would support lowering
the MCLG of chlorite. Data show that there is more dietary exposure than previously assumed
due to the increased use of chlorine dioxide and acidified sodium chlorite as disinfectants in the
processing of foods (USEPA, 2006b; WHO, 2008.) Data to support the quantification of
exposures as a result of antimicrobial uses are available in the Office of Prevention, Pesticides
and Toxic Substances Reregi strati on Eligibility Decision (USEPA, 2006b) for chlorine dioxide
and sodium chlorite and the WHO (2008) assessment of acidified sodium chlorite as a food
additive. Additional data on chlorine dioxide sanitizer uses in the United States are included in
submissions to the Food and Drug Administration (FDA) (1994). Chapter 6 provides information
about co-occurrence of chlorine dioxide, chlorite and chlorate.
Data are also available which support possible common health endpoints from exposures to
chlorate, chlorite and chlorine dioxide. Animal studies indicate that these compounds all result in
hematological and thyroid effects (Orme et al., 1985; Abdel-Rahman et al., 1984; Couri et al.,
1982; Moore and Calabrese, 1982; Bercz et al., 1982; Abdel-Rahman, 1980). Although there are
different etiologies for some of the hematological effects, the outcomes of reduced hemoglobin,
hematocrit and low red blood cell counts are the same. Less is known about the modes of action
for the thyroid effects but there is likely to be synergy when two or more of the members of the
group are present in the same matrix (e.g., food or drinking water). Limited findings in humans
support concern for exposure to mixtures for the hematological effects and impact on the kidney
during development (Lubbers et al., 1981, 1982, 1984). The high probability for co-exposures is
an important factor in considering these chemicals as a group.
Six-Year Review 3 Technical Support Document	4-56
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Relevance for SYR
New information on the relative source contribution of exposure from chorite in drinking water
and on the co-occurrence of chlorite and chlorate, along with common health endpoints of
concern, indicate a meaningful opportunity for potential risk reduction for chlorite.
4.3 Regulated Disinfectants
This section addresses the health effects that are associated with the regulated disinfectants for
which Maximum Residual Disinfectant Levels (MRDLs) and Maximum Residual Disinfectant
Level Goals (MRDLGs) have been established under the Stage 1 and Stage 2 D/DBPRs,
specifically chlorine, chloramines and chlorine dioxide.
4.3.1	Chlorine
Basis for the MRDLG
In the Stage 1 D/DBPR, EPA finalized an MRDLG of 4 mg/L for chlorine based on a weight of
evidence evaluation of both the cancer and noncancer effects and classified chlorine as "not
classifiable as to human carcinogenicity" (USEPA, 1994a, 1998b; NTP 1992a). The MRDLG
was based on the RfD of 0.1 mg/kg/day, an adult tap water consumption of 2 liters/day for a 70
kg adult and an assumed drinking water contribution of 80 percent of total exposure (USEPA,
1994a). The RfD was based on a NOAEL of 14 mg/kg/day for no treatment-related effects from
NTP (1992a), a two-year drinking water study in rats and mice, with the application of an
uncertainty factor of 100. Due to a lack of significant new health effects data available for the
Stage 2 D/DBPR, the RfD value did not change, and EPA did not revise the MRDLG for
chlorine at that time (USEPA, 2003c, 2006a).
New Information Since Stage 1 and Stage 2 D/DBPRs
No new, relevant information about cancer or noncancer effects has been identified for chlorine.
Relevance for SYR
Insufficient evidence is available to support a change in the MRDLG for chlorine.
4.3.2	Chloramines
Basis for the MRDLG
In the Stage 1 D/DBPR, EPA established a MRDLG of 4 mg/L for chloramine (USEPA, 1994a,
1998b, NTP 1992a) based on a weight of evidence evaluation of both the cancer and noncancer
effects and classified chloramines as "not classifiable as to human carcinogenicity." EPA has not
set an RfD for chloramines. The MRDLG was based on a NOAEL of 9.5 mg/kg/day for no
treatment-related effects for monochloramine from a two-year drinking water study in rats and
mice (NTP, 1990), an uncertainty factor of 100, adult tap water consumption of 2 liters/day for a
70 kg adult and an assumed drinking water contribution of 80 percent of total exposure (USEPA,
Six-Year Review 3 Technical Support Document	4-57
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
1994a). Due to the lack of significant new health effects data available for the Stage 2 D/DBPR,
EPA did not revise the MRDLG for chloramines at that time (USEPA, 2003c, 2006a).
New Information Since Stage 1 and Stage 2 D/DBPRs
No new information about cancer or noncancer effects has been identified that would change the
basis for the existing MRDLG. Since promulgation of the Stage 1 D/DBPR, EPA has developed
an EPA website (http://www.epa.gov/dwreginfo/basic-information-about-chloramines-and-
drinking-water-disinfection) that provides basic information about chloramines and chloramine-
related research and answers questions which address issues raised by the public related to
exposure to chloramines.
Relevance for SYR
Insufficient evidence is available to support a change in the MRDLG for chloramines. New
information is available about various DBPs of potential health concern, such as nitrosamines,
that may be formed in systems that use chloramination. Additional information about the
formation of nitrosamines in systems that use chloramines is provided in the Six-Year Review 3
Technical Support Document for Nitrosamines (USEPA, 2016d).
4.3.3 Chlorine Dioxide
Basis for the MRDLG
In the Stage 1 D/DBPR, USEPA (1994a) proposed a MRDLG of 0.3 mg/L for chlorine dioxide
based on an RfD of 0.01 mg/kg/day from a developmental rat study (Orme et al., 1985). EPA
(1997a) reviewed and completed a peer review of a two-generation reproductive study of chlorite
in Sprague-Dawley rats (CMA, 1996). In this study, male and female rats were administered
sodium chlorite in drinking water at doses ranging up to 300 ppm. Reproduction, fertility,
clinical signs and histopathology were evaluated in 2-generations of offspring. These data are
relevant to chlorine dioxide because chlorine dioxide is rapidly reduced to chlorite and chlorite is
oxidized to chlorate. In the Stage 1 D/DBPR, EPA finalized an MRDLG of 0.8 mg/L for chlorine
dioxide based on the same data used to derive the MCLG for chlorite (USEPA, 1998b); the RfD
of 0.03 mg/kg/day, an adult tap water consumption of 2 liters/day for a 70 kg adult and an
assumed drinking water contribution of 80 percent of total exposure. The RfD was derived based
on a NOAEL of 35 ppm (3 mg/kg/day for the chlorite ion) for decreases in absolute brain and
liver weight and lowered auditory startle amplitude at 70 and 300 ppm and the application of an
uncertainty factor of 100. In the Stage 1 D/DBPR, the final MRDLG was not changed from the
proposed value because a lower uncertainty factor (100 vs. 300) was applied with the use of the
multigeneration CMA (1996) study. EPA did not revise the MRDLG for chlorine dioxide in the
Stage 2 D/DBPR (USEPA, 2003c, 2006a).
New Information Since Stage 1 and Stage D/DBPRs
No new information has been identified about cancer or noncancer effects for chlorine dioxide.
Six-Year Review 3 Technical Support Document	4-58
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Please refer to section 4.2.2 on chlorite for a discussion about possible common health endpoints
from exposures to chlorate, chlorite and chlorine dioxide. The short half-life for chlorine dioxide
mitigates the concern from its increased use by the food industry and for other applications.
Relevance for SYR
Information about possible common health endpoints from exposures to chlorate, chlorite and
chlorine dioxide indicate meaningful opportunity for potential risk reduction for chlorine
dioxide.
4.4 Unregulated DBPs
This section provides health effects information on several organic and inorganic DBPs that are
not currently regulated in the Stage 1 or Stage 2 D/DBPRs. There are many DBPs (e.g.,
brominated HAAs, haloacetonitriles, nitrosamines, MX and chlorate) that are unregulated. The
unregulated term can be misleading in that exposures may be reduced through treatment because
of actions taken to comply with the MCL and TT requirements of the D/DBPRs.
Health effects information in this section is based on data in assessments conducted by EPA and
other agencies (e.g., ATSDR, Cal EPA, NTP and WHO) plus published articles that were not
considered under the evaluations conducted for the Stage 2 D/DBPR. Chemicals are grouped by
chemical families in the sections that follow.
4.4.1	Chlorate
Information on the health effects of chlorate is presented in Chapter 3 of Six-Year Review 3
Technical Support Document for Chlorate (USEPA, 2016e).
4.4.2	Nitrosamines
Information on the health effects of nitrosamines is presented in Chapter 3 of Six-Year Review 3
Technical Support Document for Nitrosamines (USEPA, 2016d).
4.4.3	Haloacetic Acids
Unregulated HAAs include bromochloracetic acid (BCAA), bromodichloroacetic acid
(BDCAA), dibromoacetic acid (DBAA), dibromochloroacetic acid (DBCAA), tribromoacetic
acid (TBAA) and iodinated acetic acid compounds. The USEPA (2005e) Criteria Document for
Brominated Acetic Acids includes monitoring data for monobromoacetic acid (MBAA) and
DBAA collected during the ICR monitoring for the Stage 2 rule demonstrating occurrence in
public drinking water supplies. The Criteria Document did not develop RfDs for any of the
brominated HAAs. The data for the cancer endpoint justified a classification of inadequate
information to assess carcinogenic potentiaF at that time. Subsequent to the publication of the
criteria document the NTP published the findings from bioassays of BCAA, BDCAA and
DBAA. DBAA is currently regulated as part of HAA5 but lacks an MCLG.
Plewa et al. (2002) developed an in vitro model using Chinese hamster ovary (CHO) cells for
determining a relative ranking of cytotoxic potency for a direct comparison to DBP-induced
Six-Year Review 3 Technical Support Document	4-59
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
cytotoxicity in S. typhimurium, analyzing DBPs for their ability to induce genomic DNA damage
in mammalian cells, and determining a relative rank order of their genotoxic potency and
comparing these results with data derived from Salmonella mutagenesis studies.
Using the Plewa et al. (2002) model, Plewa et al. (2010) provided a comparative systematic
analysis of chronic cytotoxicity and acute genomic DNA damaging capacity of 12 individual
HAAs in mammalian cells. In addition to the HAA5, they analyzed MIAA, diiodoacetic acid
(DIAA), bromoiodoacetic acid (BIAA), TBAA, DBCAA, BDCAA and BCAA. They identified a
rank order of chronic cytotoxicity was MIAA > MBAA > TBAA > DBCAA > DIAA > DBAA >
BDCAA > BCAA > MCAA > BIAA > TCAA > DCAA. The rank order for genotoxicity was
MIAA > MBAA > MCAA > DBAA > DIAA > TBAA > BCAA > BIAA > DBCAA. They
found that DCAA, TCAA and BDCAA were not genotoxic. The trend for both cytotoxicity and
genotoxicity is iodinated HAAs > brominated HAAs > chlorinated HAAs (Plewa et al., 2010).
The cytotoxicity for haloacids was low compared to other classes of DBPs based on results using
the CHO cell model. Using this model, the rank order from most cytotoxic to least cytotoxic for
the DBP classes was haloacetaldehydes (HALs) > haloacetamides > HNMs > haloacetonitriles >
2C-haloacids > HAAs > halomethanes. Similarly, when looking at induced genomic DNA
damage in CHO cells, the relative genotoxicity of haloacids was low compared to other classes
of DBPs. Again, using this model, the rank order from the most genotoxic to the least genotoxic
of the DBP classes showed that haloacetonitriles > haloacetamides > HNMs > HALs > HAAs >
>2C-haloacids > halomethanes (Plewa and Wagner, 2009).
Additional information about specific compounds are described below.
4.4.3.1 Bromochloroacetic acid
Cancer
The NTP (2009) completed a toxicological and carcinogenic assessment for BCAA subsequent
to the Stage 2 Rule as part of their research agenda on water disinfectants and disinfection
byproducts. BCAA was administered to F344/N rats and B6C3F1 mice in drinking water at daily
doses up to 40 and 50 mg/kg/day in male and female rats, respectively, and 90 and 60 mg/kg/day
in male and female mice, respectively in a two-year study (NTP, 2009). NTP concluded that
there is clear evidence of carcinogenic activity in rats based on an increased incidence of
malignant mesotheliomas in males, multiple fibroadenomas of the mammary gland in females
and adenomas of the large intestine in males and females. There was also clear evidence of
carcinogenic activity in mice based on increased incidences of hepatocellular neoplasms in male
and female mice and hepatoblastoma in male mice. The lowest dose in rats that demonstrated an
increased incidence of malignant mesotheliomas and pancreatic islet adenoma compared to
controls was 20 mg/kg/day. The lowest dose in the mouse study with an increase in tumors
compared to controls was 15 mg/kg/day for hepatoblastomas.
Reproductive and Developmental
NTP (2009) conducted a study in male and female Sprague-Dawley rats administered BCAA in
drinking water for various times during a 35-day period. The number of live fetuses per litter and
the total number of implants per litter were significantly decreased at the highest dose of 50
Six-Year Review 3 Technical Support Document	4-60
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
mg/kg/day. The NOAEL was 50 mg/kg/day based on decreased implants and number of live
fetuses per litter.
Tully at al. (2005) evaluated reproductive performance in juvenile and adult C57BL/6 male mice
administered BCAA for 14 days in a continuous breeding assay. Juvenile mice were exposed
from PND 8 to 21, allowed to mature and then mated. Effects on fertility were observed in mice
exposed as adults and included decreases in mean number of litters per male, percentage of litters
per female bred and total number of fetuses per male. The decreased male fertility was attributed
to disruption of spermatid differentiation.
Systemic toxicity
The NTP (2009) reported the results from a subchronic study in both male and female rats and
male and female mice. In male and female rats there were effects on liver weight and kidney
weight. The NOAEL for the effects on liver weight was 20 mg/kg/day with a LOAEL of 40
mg/kg/day in both males and females. Kidney weights were increased at a higher dose. In male
mice there was a LOAEL of 8 mg/kg/day for cell proliferation in the spleen and no NOAEL for
this effect. In females the 8 mg/kg/day was a NOAEL with a LOAEL of 17 mg/kg/day for this
effect. At higher doses effects on liver weight and body weight, plus liver periportal hepatic
cellular vacuolization, were noted.
Genotoxicity
Richardson et al. (2007) reported that BCAA had little or no toxicology data and lacked
genotoxicity data. The NTP (2009) reported that BCAA was positive for mutagenicity in S.
typhimuriam strain 100 with and without activation but not in strain 98. No micronuclei were
found in the erythrocytes from male and female mice exposed to bromochloroactic acid for 3
months (NTP, 2009).
4.4.3.2 Bromodichloroacetic acid
Cancer
NTP (2015) conducted a 2-year bioassay of BDCAA in treated F344/NTac rats and B6C3F1
mice. BDCAA administration in drinking water to rats resulted in increased incidences of
malignant mesothelioma and combined incidences of epithelial tumors of the skin in males,
increased incidences of fibroadenoma and carcinoma of the mammary gland in females, along
with adenoma or carcinoma of the Harderian gland in males and hepatocellular adenoma in
females.
In mice there was an increased incidence of hepatocellular carcinoma and hepatoblastoma in
males and females (NTP, 2015). The lowest dose to induce tumors at levels above controls in
female rats was 13 mg/kg/day for mammary gland fibroadenoma. In male rats, the lowest dose to
observe increased incidence of keratoaceanthoma, basal cell ademona or carcinoma, squamous
cell carcinoma (SCC) and other carcinogenic endpoints was 43 mg/kg/day. The lowest dose with
evidence of carcinogenicity in male and female mice was 23 mg/kg/day based on an increased
incidence of hepatocellular adenoma. The NTP (2015) concluded that there was clear evidence
of carcinogenicity in male and female rats and mice.
Six-Year Review 3 Technical Support Document	4-61
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Systemic toxicity
In the subchronic component of the NTP (2015) studies of bromodichloracetic acid, the high
dose of 72 mg/kg/day was a LOAEL for decreased testes weights and sperm motility in male rats
while in females a dose of 69 mg/kg/day (also the high dose) was a LOAEL for increased kidney
weight. The NOAELs for male and female rats were 37 and 43 mg/kg/day respectively. In male
mice the NOAEL was 30 mg/kg/day with a LOAEL of 59 mg/kg/day for increased liver weight.
In female mice there was a marginal LOAEL of 129 mg/kg/day for hepatic glycogen depletion
with a NOAEL of 70 mg/kg/day in the NTP subchronic study.
Genotoxicity
Studies of genotoxicity showed that the responses were positive for S. typhimurium strains
TA97, TA98 and TA100 in the absence of S9 and equivocal in the presence of rat S9. For E. coli
strain WP2 wvr.4/pkM101 the results were positive in the presence and absence of S9. No
significant increases in the frequencies of micronucleated normochromatic erythrocytes or the
percent of polychromatic erythrocytes (reticulocytes) were seen in blood samples from treated
mice (NTP, 2015).
4.4.3.3	Dibromochloroacetic acid
NTP (2000) performed a short-term reproductive and developmental toxicity study with
dibromochloroacetic acid administered in drinking water to rats. For the first part of the study, no
significant test-article related effects were observed at doses ranging from 30-500 ppm. There
was no estimated conversion to mg/kg/day so no NOAEL was set.
4.4.3.4	Tribromoacetic acid
Reproductive and Developmental
NTP conducted a short-term study on the reproductive and developmental effects of TBAA
(NTP, 1998b). Doses up to 400 ppm were administered in drinking water to male and female rats
(peri-conception and gestational exposure) for two weeks. No reproductive effects were observed
in males or females and evaluation of the newborn heart and brain did not reveal any treatment-
related effects. However, not data from a study that used a standard one or two generation
protocal. Thus, additional research is needed.
The International Programme on Chemical Safety (IPCS) reviewed studies on TBAA toxicity in
the Environmental Health Criteria 216 (WHO 2000). The only information provided was from a
mutagenicity study showing positive results for Ames and SOS chromotest assays.
4.4.3.5	Iodinated acids
Iodinated acids identified in drinking water in the United States include MIAAand BIAA.
Several longer-chain iodinated acids were also identified as present in treated drinking water:
(Z)- and (E)-3-bromo-3-iodopropenoic acid and (£)-2-iodo-3-methylbutenedioic acid (Plewa et
al., 2004a; Richardson et al., 2008).
Six-Year Review 3 Technical Support Document	4-62
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Cytotoxicity and Genotoxicity
MIAA is the most cytotoxic and genotoxic HAA in mammalian cell assays that has been
reported in the literature. Similar results are seen when comparing MIAA mutagenicity in S.
typhimurium and genotoxicity in CHO cells compared to MBAA and MCAA (Plewa et al.,
2004a; Richardson et al., 2008; Plewa et al., 2010).
Wei et al. (2013) examined cytotoxicity, genotoxicity and ability to transform NIH3T3 cells to
tumorigenic lines and found that prolonged exposure of NIH3T3 cells to MIAA increased the
frequencies of transformed cells with anchorage-independent growth and agglutination with
concanavalin A. They found that neither MIAA (nor iodoform) increased micronucleus
frequency, but that MIAA-transformed cells formed aggressive fibrosarcomas after inoculation
into Balb/c nude mice. They concluded that MIAA has a biological activity that is consistent
with a carcinogen and that human exposure should be of concern.
Reproductive and Development
MIAA has been shown to induce neural tube closure defects and other developmental
abnormalities in mouse embryos (Hunter and Tugman, 1995; Hunter et al., 1996). According to
Plewa and Wagner (2009), the ability to induce neural tube defects in an ex vivo mouse embryo
assay is strongly correlated with CHO cell chronic cytotoxicity and has good correlation with
CHO genotoxicity. No genotoxicity data were identified for the iodinated propenoic and
butenedioic acids.
4.4.4	Iodinated THMs
Iodinated THMs identified in drinking water in the United States include iodoform,
bromodiiodomethane, dichloroiodomethane, bromochloroiodomethane, dibromoiodomethane
and chlorodiiodomethane (Plewa et al., 2004a; Richardson et al., 2008).
Cytotoxicity and Genotoxicity
Richardson et al. (2008) found the iodinated THMs to be much less cytotoxic than the iodinated
acids, with the exception of iodoform. Iodoform was found to be mutagenic in bacteria but did
not induce chromosome aberrations in Syrian hamster embryo cells in vitro (Richardson et al.,
2007).
Of the iodinated THMs studied by Richardson et al. (2008), only chlorodiiodomethane was
found to be genotoxic. Richardson et al. (2008) noted that BrTHMs require glutathione-S-
transferase-thetal (GSTT1) mediated metabolism to form mutagenic intermediates, and it is not
known whether this is expressed in the CHO cells used in the Richardson et al. (2008)
experiment with the iodinated acetic acids.
4.4.5	Haloketones
As described by Krasner et al. (2006), EPA selected the following haloketones as priority DBPs
for a nationwide occurrence study: chloropropanone, 1,3-dichloropropanone, 1,1-
dibromopropanone, 1,1,3-trichloropropanone, l-bromo-l,l-dichloropropanone, 1,1,3,3-
Six-Year Review 3 Technical Support Document	4-63
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
tetrachloropropanone, 1,1,1,3-tetrachloropropanone, 1,1,1,3,3-pentachloropropanone and
hexachloropropanone. In this study, 1,1,3,3-pentachloropropanone and hexachloropropanone
were not analyzed because they are not stable in water. While 1,1,3,3-tetrabromopropanone was
not initially prioritized, it was identified in drinking water after the initial prioritization and was
included in the monitoring study report due to its similarity to the other priority compounds.
Several haloketone species were identified in drinking water, with the priority haloketone 1-
bromo-l,l-dichloropropanone reaching a maximum concentration of 3 [j,g/L in a distribution
sample from a plant using ozone-chlorine disinfection (Krasner et al., 2006).
Systemic effects
WHO (2003a) investigated the data for chlorinated acetones (propanones) and determined that
the data on dose-response were limited. Single doses of 1,1-dichloroacetone revealed effects on
the liver at 325 mg/kg and no toxicity was observed below 130 mg/kg (Laurie et al. 1986). No
liver toxicity was observed for 1,3-dichloracetone at doses up to 20 mg/kg, but it was shown to
potentially act as a tumor initiator in mouse skin. No guideline or regulatory value was derived
by WHO (2003a).
4.4.6 Haloacetaldehydes
Chloroacetaldehyde (CAL), dichloroacetaldehyde (DCAL), bromochloroacetaldehyde (BCAL),
trichloroacetaldehyde monohydrate (chloral hydrate) and tribromoacetaldehydes (TBAL) have
been identified in disinfected drinking water (Richardson et al., 2007). HALs are the third largest
group by weight of identified DBPs in drinking water. Jeong et al. (2015) provided a quantitative
comparison of HAL toxicity in Chinese hamster ovary cells. The rank order of HAL cytotoxicity
was found to be TBAL ~ CAL > dibromoacetaldehyde (DBAL) ~ BCAL ~
dibromochloroacetaldehyde (DBCAL) > iodoacetaldehyde (IAL) > bromoacetaldehyde (BAL) ~
bromodichloroacetaldehyde (BDCAL) > DCAL > trichloroacetaldehyde (TCAL). The HALs
were found to be highly cytotoxic compared to other DBP chemical classes. Jeong et al. found
that the rank order of HAL genotoxicity is DBAL > CAL ~ DBCAL > TBAL ~ BAL > BDCAL
> BCAL ~ DCAL > IAL. TCAL was not genotoxic (Jeong et al., 2015).
4.4.6.1 2-Chloroacetaldehyde
Cancer
2-Chloroacetaldehyde (2-CAA) was examined for carcinogenicity in rats by Daniel et al. (1992).
B6C3F1 mice were exposed to 0.1 g/L of 2-CAA (17 mg/kg/day) in a cancer bioassay. There
were significant increases for hepatic necrosis and hepatic tumors but not liver weights in the
treated rats. Only one dose was evaluated for comparison with the controls.
Genotoxicity
CAL was mutagenic in bacteria and in mammalian cells in vitro but not in mice (Richardson et
al., 2007).
Six-Year Review 3 Technical Support Document	4-64
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
4.4.6.2 Chloral hydrate (trichloroacetaldehyde monohydrate)
Cancer
USEPA (2000c) examined the toxicological effects of chloral hydrate and developed a reference
dose of 0.1 mg/kg/day based protection for central nervous system depression and
gastrointestinal irritation in humans. Chloral hydrate is used pharmacologically for sedation. The
LOAEL used in the derivation of the RfD (10.7 mg/kg/day) is based on the clinical dose used in
the sedation of adults. USEPA (2000c) considered chloral hydrate to be a weak mutagen and
clastogen based on a NTP oral gavage study. The finding for cancer under the 1996/1999
proposed cancer guidelines was that the evidence for carcinogenicity is suggestive. No
quantification for dose response was presented.
WHO (2005c) evaluated the carcinogenicity of trichloroacetaldehyde monohydrate (chloral
hydrate) and concluded that it was not classifiable for cancer based on inadequate evidence in
humans and limited evidence in animals. They derived a TDI of 0.0045/kg for systemic liver
effects based on increased proliferative lesions in the liver of mice (Geroge et al., 2000) using a
1000-fold uncertainty factor with extra uncertainty factor of 3 for the limited evidence of
carcinogenicity.
Genotoxicity
Chloral hydrate is a direct acting mutagen in vitro and it induced base-substitution mutations in
bacteria, as well as aneuploidy and micronuclei in mammals in vivo and in mammalian cells in
vitro. Chloral hydrate also induced chromosomal aberrations, gene mutations and cell
transformations in mammalian cells in vitro (Richardson et al., 2007).
4.4.7 Halonitromethanes
As described by Plewa et al. (2004b), the following HNMs were identified by EPA as target
analytes for occurrence and toxicology studies: bromonitromethane, dibromonitromethane,
tribromonitromethane, bromochloronitromethane, dibromochloronitromethane,
bromodichloronitromethane, chloronitromethane, dichloronitromethane and
trichloronitromethane (chloropicrin). ). In Section 4.4.3, EPA described the relative cytotoxicity
and genotoxicity among classes of DBPs including HNMs.
Cancer
WHO (2003b) evaluated the carcinogenic studies of chloropicrin and determined that the high
mortality in the National Cancer Instititue (NCI) (1978) bioassays and the limited endpoints
examined did not support development of a guideline value for chloropicrin.
The California Department of Pesticide Regulation (Cal EPA, 2010) released a summary of
toxicological studies with chloropicrin. Chloropicrin is used as a soil fumigant and most of the
studies involve inhalation exposures. There are few studies of oral administration. The effects
seen in a 10-day and a 90-day study in rats (Condie et al., 1994) included reduced body weights,
changes in thymus, liver and spleen weights, changes in hematological and clinical chemistry
values and histopathological lesions in the forestomach (nonglandular stomach). The NOEL in
Six-Year Review 3 Technical Support Document	4-65
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
the 90-day oral gavage study appeared to be 8 mg/kg/day based on body weight reduction,
hematological changes and histological changes in the forestomach. There is a NCI cancer study
from 1978 in rats that used the oral exposure route (gavage). There was high mortality early in
the study that caused several adjustments to dosing. Mammary gland fibroadenomas as seen in
surviving females demonstrated a dose-response trend but is confounded by the high mortality
and dosing alterations during the study. In another study (gavage in corn oil) (Lauter at al., 1995)
there was an increased incident of stomach papilloma in males and mammary fibroadenomas in
females at doses above 10 mg/kg/day (Cal EPA, 2010). No tumors were observed at 1.0
mg/kg/day.
Cytotoxicity and Genotoxicity
The HNMs are weak mutagens in S. typhimurium TA100, were potent genotoxicants in
mammalian cells and induced DNA damage in CHO cells. Dibromonitromethane is the most
cytotoxic and mutagenic HNM tested in both S. typhimurium and CHO cells (Richardson et al.,
2007). The HNMs are more cytotoxic than the corresponding HAAs. The brominated
nitromethanes and the mixed bromo- and chloro- nitromethanes were more genotoxic than the
chlorinated nitromethanes (Richardson et al., 2007).
Developmental and Reproductive
There are no oral exposure data on the developmental and reproductive effects of chlorpicrin
according to the Cal EPA (2010) assessment.
4.4.8 Haloacetonitriles
Acetonitriles, including chloro-, bromochloro-, dibromo- and trichloro- acetonitrile, are the most
commonly measured haloacetonitriles in drinking water in the United States (Richardson et al.,
2007). Several other haloacetonitriles were also measured in a recent study, including bromo-,
dibromo-, bromodichloro - and tribromoacetonitriles. ). In Section 4.4.3, EPA described the
relative cytotoxicity and genotoxicity among classes of DBPs including HANs.
WHO (2004c) developed TDIs for dichloroacetonitrile and dibromoacetonitrile (DBAN) using
studies of systemic toxicity that did not use tricaprylin for the control. A Total Daily intake
(TDI) of 2.7 |ig/kg/day for dichloroacetonitrile was set based on a LOAEL of 8 mg/kg/day for
increased relative liver weight in male and female rats in a 90-day study (Hayes et al. 1986). A
TDI of 11 |ig/kg/day for DBAN was set based on the NOAEL of 11.3 mg/kg/day for decreased
body weight in male F344 rats in a 90-day drinking water study by NTP (2001 a,b; 2002a,b).
Reproductive and Developmental
In studies summarized by WHO (2004c), dichloroacetonitrile, DBAN, bromochloroacetonitrile
and trichloroacetonitrile were linked to reproductive and developmental effects in rats. However,
many of these reproductive and developmental studies were conducted with tricaprylin as a
vehicle for gavage, and tricaprylin has subsequently been demonstrated to be a developmental
toxicant that potentiates the effects of trichloroacetonitrile (Christ et al., 1995) and, presumably,
of other HANs. Therefore, WHO (2004c) concluded that the studies using tricaprylin likely
overestimate the developmental toxicity of these HANs.
Six-Year Review 3 Technical Support Document	4-66
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Cytotoxicity and Genotoxicity
The Ames assay for DBAN was positive in TA97 and TA1535 in the presence of S9. There were
no increases in the frequencies of micronucleated erythrocytes in peripheral blood of male or
female mice from the subchronic study. DBAN also did not induce sex-linked recessive lethal
mutations in germ cells of male D. melanogaster exposed by feeding or by injection (NTP,
2010).
The brominated acetonitriles are generally not mutagenic in Salmonella, while the chlorinated
acetonitriles are mutagenic, both with and without metabolic activation (Richardson et al., 2007).
All of the haloacetonitriles tested induced DNA damage in mammalian cells. Plewa and Wagner
(2009) provided a slightly different ranking for direct acting genotoxic activity for these
chemicals, identifying chloroacetonitrile as the least genotoxic. As a class, the haloacetonitriles
are highly reactive, causing DNA damage in mammalian cells in vitro, but not inducing
mutations in bacteria (Richardson et al., 2007).
4.4.8.1 Dibromoacetonitrile
Cancer
The NTP (2010) conducted a two-year bioassay for DBAN dissolved in drinking water in male
and female F344 rats plus male and female B6C3F1 mice. As a result of this study, the NTP
concluded that there was clear evidence of carcinogenicity in male rats, male mice and female
mice. Some evidence for carcinogencity was the finding for female rats. In the male rats at the
high dose of 7 mg/kg/day there were two rare glandular stomach adenomas. The incidence of
squamous epithelial hyperplasia of the tongue was significantly increased at a dose of 7
mg/kg/day in males while both males and females exhibited hyperkeratosis of the tongue at a
dose of 4 mg/kg/day. Precancerous papilloma and keratoacanthoma of the skin displayed a
positive trend across the 2, 4 and 8 mg/kg/day doses in females. Due to a low response, this
finding was classified as equivocal for DBAN.
Tumors were also evident in the forestomach of male and female mice (squamous cell papilloma
or carcinoma, combined). They were significantly increased as compared to controls at the high
dose of 13 mg/kg/day in males and 11 mg/kg/day in females. In males, hyperplasia of the
stomach tissues was present even at the low dose of 4 mg/kg/day. Male mice of the 4 and 7
mg/kg/day dose groups had hepatocellular adenoma, hepatocellular carcinoma or hepatoblastoma
(combined), theses finding were high in all dose groups and were classified as equivocal.
Water intake was less than that of the control for both the rats and mice suggesting possible taste
aversion for the treated drinking water
Systemic
Given the carcinogenic responses in the rats and mice in the two year chronic study it is
surprising that evidence for epithelial irritation and inflammation of the oral cavity, esophagus
and forestomach was lacking in the subchronic study in both the rats and mice. The high doses of
13 and 11 mg/kg/day were NOAELs for the male and female mice, respectively. In the mice the
subchronic NOAEL was 18 mg/kg/day for males and females (NTP, 2010).
Six-Year Review 3 Technical Support Document	4-67
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Reproductive and Developmental
Meier et al. (1985) conducted a reproductive and developmental screening assay in young, male
Sprague-Dawley rats administered DBAN in drinking water on study days 6-34. No
reproductive effects or altered sperm morphology was observed. One group of female Sprague-
Dawley rats was administered the same doses of dichloroacetonitrile (DCAN) in drinking water
on study days 1-34, which included a 5-day cohabitation with the treated males (study days 13-
17) and gestation. No effects on mating, fertility, pregnancy or development were observed. A
second group of females was cohabitated with treated males and then exposed on GD 6 through
PND 1 to the same doses of DC AN. The NOAEL for female reproductive effects and for
developmental effects was the highest dose tested, 10.8 mg/kg/day. No treatment related effects
were observed for maternal body weights, or for gross necropsy or number of resorptions or
implantation sites.
4.4.8.2	Dichloroacetonitrile
Reproductive and Developmental
Meier et al. (1985) observed no effect on sperm head morphology in a study conducted in male
B6C3F1 mice treated with DC AN by gavage.
Smith et al. (1986) conducted a developmental toxicity screening study with DC AN
administered in a tricaprylin vehicle to pregnant Long-Evans rats on GDs 7-21. The percentage
of females delivering litters was significantly reduced and fetal resorptions were increased. Fetal
birth weights and postnatal pup survival were decreased. The LOAEL for developmental toxicity
was 55 mg/kg/day, which was the only dose tested.
Smith et al. (1989a) administered DC AN to pregnant Long-Evans rats by gavage in a tricaprylin
vehicle at GDs 6 through 18. Effects were noted at 25 mg/kg/day and greater, including
increased post-implantation loss and fetal resorptions; an increase in the incidence of soft tissue
malformations of the cardiovascular, digestive and urogenital systems; a decreased number of
viable litters; and decreased fetal weight and length. The NOAEL for both maternal and
developmental toxicity was 15 mg/kg/day based on increased liver weight in the dams and
decreased fetal weight and length and an increase in soft tissue malformations, respectively.
4.4.8.3	Bromochloroacetonitrile
Reproductive and Developmental
Meier et al. (1985) observed no effect on sperm head morphology in a study conducted in male
B6C3F1 mice treated with Bromochloroacetonitrile (BCAN) by gavage.
Smith et al. (1986) evaluated pups of Long-Evans rats administered BCAN on GD 7-21 in a
tricaprylin vehicle and observed significantly reduced mean birth weights and reduced body
weight gain.
Christ et al. (1995) administered BCAN to pregnant Long-Evans rats by gavage in a tricaprylin
vehicle on GD 6-18. A decrease in fetal crown-rump length and an increase in fetal
Six-Year Review 3 Technical Support Document	4-68
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
cardiovascular malformations were observed in all dose groups; an increase in total soft tissue
malformations was observed at 25 mg/kg/day and greater; an increase in full-litter resorptions,
resorbed fetuses per litter and skeletal malformations and a decrease in the number of viable
litters were observed at 45 mg/kg/day and greater. The maternal NOAEL and LOAEL were 45
and 65 mg/kg/day, respectively, based on decreased maternal weight and increased dam
mortality. The LOAEL for developmental and teratogenic effects was 5 mg/kg/day, the lowest
dose tested. It should be noted that the tricaprylin vehicle alone had significant embryotoxicity
and teratogenicity effects when compared with the water vehicle.
4.4.8.4 Trichloroacetonitrile
Reproductive and Developmental
Meier et al. (1985) observed no effect on sperm head morphology in studies conducted in male
B6C3F1 mice treated with TCAN by gavage.
Meier et al. (1985) administered TCAN by gavage in corn oil to pregnant Long-Evans rats on
GDs 6-18. An additional group of rats was given TCAN in a tricaprylin vehicle. Effects noted in
rats administered TCAN in corn oil included full-litter resorptions, decreased pregnancy rate,
maternal weight gain and liver, spleen and kidney weights; increased post-implantation loss;
decreases in live fetuses per litter, fetal body weight and fetal crown-rump length; and increases
in fetuses per litter with skeletal and soft tissue malformations. The maternal and fetal NOAELs
were 15 and 35 mg/kg/day, respectively. Fetal malformations were primarily external
craniofacial malformations and positional cardiovascular malformations when corn oil was used
as the vehicle and structural cardiovascular defects and urogenital effects when tricaprylin was
used as the vehicle.
Smith et al. (1986) conducted a developmental screening study in pregnant Long-Evans rats
administered TCAN in a tricaprylin vehicle on GDs 7-21. Effects noted included increased
maternal deaths and full-litter resorptions, fewer females becoming pregnant or delivering viable
litters and decreased pup survival and decreased weight gain in surviving pups. The only dose
tested, 55 mg/kg/day, was a LOAEL for maternal and developmental toxicity.
Smith et al. (1988) administered TCAN to pregnant Long-Evans rats by gavage in a tricaprylin
vehicle on GDs 6-18. The high dose of 55 mg/kg/day was lethal to 4 of 19 dams and caused 100
percent fetal resorption in 67 percent of surviving, pregnant dams. Increases in full-litter
resorptions and cardiovascular malformations in litters were observed in a dose-related manner at
doses of 7.5 mg/kg/day and greater. At 15 mg/kg/day and greater, post-implantation loss and
urogenital malformations increased; and at 35 mg/kg/day and greater, fetal weight decreased.
The NOAEL for teratogenic effects was the lowest dose tested, 1 mg/kg/day.
Christ (1996) administered TCAN to pregnant Long-Evans rats by gavage in corn oil on GDs 6-
18 at doses up to 75 mg/kg/day. An additional group of rats was administered 15 mg/kg/day
TCAN in tricaprylin on GDs 6-18. The following effects were noted from exposure to TCAN in
corn oil: mortality; full litter resorptions; decreased pregnancy rate; depressed maternal weight
gain; increased maternal liver, spleen and kidney weights; increased post-implantation loss;
decreased number of live fetuses per litter; and increased number of fetuses with external,
Six-Year Review 3 Technical Support Document	4-69
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
skeletal and soft tissue malformations. Decreased fetal weights and increased soft tissue and
cardiovascular malformations were observed in the rats exposed to TCAN in tricaprylin. The
NOAEL for developmental toxicity and teratogenicity was 35 mg/kg/day and the LOAEL was
55 mg/kg/day when TCAN was administered in corn oil. The LOAEL was 15 mg/kg/day, the
only dose tested, when tricaprylin was used as the vehicle.
4.4.9 Haloacetamides
Chloro-, bromo-, dichloro- trichloro-, bromo- and dibromoacetamide were identified as
contaminants in drinking water in the United States (Richardson et al., 2007). The monitoring
also detected bromochloro-, bromodichloro- and dibromochloroacetamide. A new iodinated
DBP, bromoiodoacetamide, that was not detected at the time of the Richardson et al. (2007)
publication was identified as present in disinfected water by Plewa and Wagner (2009). In
Section 4.4.3, EPA described the relative cytotoxicity and genotoxicity among classes of DBPs
including haloacetamides.
Plewa et al. (2008) provided a rank order of cytotoxicity for 13 haloacetamides (DIAcAm >
IAcAm > BAcAm > TBAcAm > BIAcAm > DBCAcAm > CIAcAm > BDCAcAm > DBAcAm
> BCAcAm > CAcAm > DCAcAm > TCAcAm). They also provided a rank order of their
genotoxicity (TBAcAm > DIAcAm approximately equal to IAcAm > BAcAm > DBCAcAm >
BIAcAm > BDCAcAm > CIAcAm > BCAcAm > DBAcAm > CAcAm > TCAcAm). DCAcAm
was shown to be not genotoxic. Plewa et al. reported that cytotoxicity and genotoxicity followed
the class order I > Br > > CI, and that, with the exception of brominated trihaloacetamides, most
of the toxicity rank order was consistent with structure-activity relationship expectations. (Plewa
et al., 2008).
Reproductive and Developmental Toxicity
The European Commission's Scientific Committee on Consumer Safety (SCCS) reviewed the
toxicity of chloroacetamide (SCCS, 2011). Though there were no guideline-compliant
developmental or reproductive studies available, the review derived a LOAEL of 24 mg/kg/day
based on maternal body weight reduction and skeletal findings in offspring when
chloroacetamide was administered from GD 14 to PND 2. The NOAEL for this effect was 3
mg/kg/day. Christian (1991) examined a variety of developmental, reproductive and systemic
toxicological endpoints of chloroacetamide. In a 13-week oral study in rats, chloroacetamide
produced testicular atrophy at doses of 12.5 mg/kg/day and above. In a 90-day developmental
toxicity study in rats, chloroacetamide did not induce any teratogenicity in doses up to 50
mg/kg/day.
Systemic toxicity
The SCCS (2011) review of chloroacetamide concluded that the human data demonstrated
allergic reactions can be elicited at concentrations lower than 0.3 percent allowable amount in
cosmetic products. The review also saw studies demonstrating that chloroacetamide causes GSH
depletion and lipid peroxidation, resulting in cell damage and morphological changes in the liver.
Six-Year Review 3 Technical Support Document	4-70
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Cytotoxicity and Genotoxicity
A study on the cytotoxicity and genotoxicity of haloacetamides in CHO cells ranked
diiodoacetamide as the most cytotoxic and trichloroacetamide as the least cytotoxic (Plewa and
Wagner, 2009). Tribromoacetamide was ranked as the most genotoxic and trichloroacetamide as
the least. Dichloroacetamide was not found to be genotoxic.
4.4.10 Cyanogen halides
4.4.10.1	Cyanogen Bromide
USEPA (1988a) established a low confidence RfD of 0.09 mg/kg/day for cyanogen bromide
based on a NOAEL for weight loss, thyroid effects and myelin loss in rats with than uncertainty
factor of 100 and a modifying factor of 5 to account for the use a study of cyanide for the
assessment. CNBr dissociates into cyanide in water. USEPA (2012a) examined cyanogen
bromide (CNBr) under the program for peer reviewed provisional threshold values and chose not
to establish an p-RfD for settings relevant to the Superfund Program due to the lack of
pharmacokinetic, dissociation rates, issues linking CNBr to simple cyanides and lack of
toxicological data specific to CNBr. The EPA (USEPA, 2012a) assessment calls attention to the
current IRIS RfD for cyanide of 0.00063 mg/kg/day for decreased cauda epididymis weight in
male #344/N rats (USEPA, 2010b) as a value that could be applicable in scenarios where
dissociation of cyanogen bromide dissociation is expected.
4.4.10.2	Cyanogen Chloride
Cyanogen chloride toxicity was evaluated by WHO (2009). They determined that since it is
unlikely to find cyanogen chloride in the water, due to rapid transformation to cyanide in water,
it was unnecessary to develop a formal guideline. In place of a TDI for cyanogen chloride, they
develop a TDI using cyanide toxicity data since it is shown that cyanogen chloride is not only
transformed in water, but also rapidly metabolized to cyanide in the body. The TDI developed
was 0.11 mg/kg/day based on cyanide toxicity information. The WHO health-based value for
long-term exposure is 0.3 mg/1 as cyanide or 0.6 mg/1 as cyanogen chloride (rounded values).
The Office of Water (USEPA, 2005i) derived a 10-day health advisory value for cyanogen
chloride of 0.1 mg/L base on a LOAEL of 14 mg/kg/day in a study of cyanide by Palmer and
Olsen (1979), with the application of a 1000-fold uncertainty factor. As mentioned above,
cyanogen chloride is transformed to cyanide in water with any residuals metabolized to cyanide
when consumed. Free cyanide in water is currently regulated with a MCL/MCLG of 0.2 mg/L
based on protection against nerve damage and thyroid problems (USEPA, 1992b). The recent
IRIS assessment (USEPA, 2010b) updated the hydrogen cyanide and cyanide salts RfD to a
value of 0.0006 mg/kg/day based on testicular effect in male rats (NTP, 1993). The new RfD
supports a lowering of the cyanide MCLG to 0.004 mg/L assuming the application of the same
water intake, body weight and RSC variables as those used when deriving the current
MCLG/MCL.
Six-Year Review 3 Technical Support Document	4-71
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
4.4.11 Halogenated furanones
3-Chloro-4-(dichloromethyl)-5-hydroxy-2(5H)-furanone (MX) has been identified in drinking
water in the United States (Richardson et al., 2007). Other halogenated furanones that have been
studied include bromine-, chlorine- and mixed halogen-substituted 4-methyl-2(5H)-furanones,
including 3-chloro-4-(dibromomethyl)-5-hydroxy-2(5H)-furanone (BMX-2) and (Z)-2,3-
dichloro-4-oxo-butenoic acid (mucochloric acid).
4.4.11.1 Mutagen X
Cancer
Mutagen X (MX) is a halogenated hydroxyfuranone that has been identified in chlorinated
drinking water. A carcinogenicity study was conducted in Wistar rats administered MX in
drinking water (Komulainen et al., 1997). Increased incidences of cholangioma (cancer of bile
ducts cells) in the liver, follicular adenoma and carcinoma in the thyroid and cortical adenomas
of the adrenal gland were observed in both sexes. The Office of Water (USEPA, 2000d)
completed a quantitative assessment of the carcinogenicity of MX classifying it as acting through
a mutagenic mode of action (USEPA, 2000d). The two-year oral exposure study by Komulainen
et al. (1997) was selected for quantitative evaluation. Data for occurrence of thyroid (follicular
adenoma and adenocarcinoma) and liver (adenoma, carcinoma, cholangioma and
cholangiocarcinoma) tumors were evaluated resulting in an oral slope factor of 3.7 (mg/kg/day)"
1. Confidence in this quantification was rated as medium. Based on the slope factor the
concentration associated with a 1 in 1,000,000 increased risk for cancer was 9.5 ng/L.
McDonald and Komulainen (2005) derived a combined cancer potency of MX for each gender
based on the incidence of all tumors. The drinking water concentration associated with a 1 in
1,000,000 increased cancer risk was calculated to be 7.8 ng/L. IARC (2004) has classified MX as
Group 2B, "possibly carcinogenic to humans."
Genotoxicity
MX is one of the most potent agents tested for mutagenicity in S. typhi murium, and the
concentration of MX in drinking water accounts for as much as 50 percent of the total
mutagenicity of these samples (Richardson et al., 2007). The addition of rat liver extract (S9
fraction) reduces its potency. It induced DNA damage, mutations and prophage induction in E.
coli. MX is also a potent genotoxicant in mammalian cells. It induced unscheduled DNA
synthesis, sister chromatid exchanges, micronuclei, chromosome aberrations, mutations and
DNA strand breaks; however, it has been found to be negative for micronucleus induction in
vivo in rodents.
Reproductive and Developmental
Huuskonen et al. (1997) administered MX to pregnant Han: Wistar rats by gavage on GDs 6-19.
There were no increases in gross, visceral, or skeletal malformations or in mortality in the
fetuses.
Six-Year Review 3 Technical Support Document	4-72
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Teramoto et al. (1998) conducted an in vitro assay using 12-day-old rat embryo midbrain and
limb bud cells to evaluate the teratogenic properties of MX. There was no or minimal effect in
the presence of S9 fraction; a significant decrease in the number of differentiated foci in the
Central Nervous System and limb bud cells was observed in the absence of S9 fraction. The in
vivo significance is not known.
4.4.11.2 Other halogenated furanones
The Office of Water (USEPA, 2000d) completed a health assessment of chlorohydroxyfuranones
(CHFs) in 2000 that summarized the available data at that time. Many of the CHF compounds
have data that identify these compounds as genotoxins with differing relative potencies, however
studies of their carcinogenic potency were lacking at that time.
Mucochloric acid toxicity was evaluated by OECD SIDS (2003) and a NOAEL for
developmental toxicity of 60 mg/kg/day was observed with no LOAEL. Systemic toxicity was
found at 30 mg/kg/day shown as reduced food consumption and bodyweight gain. Mucochloric
acid is mutagenic in S. typhimurium and CHO cells and has induced DNA damage in E. coli cells
injected into mice.
The brominated halofuranones are generally less mutagenic than MX except for BMX-2, which
caused a 140 percent increase in mutagenicity in S. typhimurium compared to MX (Richardson et
al., 2007).
4.4.12	Halogenated benzoquinones (HBQs) and haloquinones (HQ)
Cancer
The halogenated benzoquinones (HBQs), may be important bladder carcinogens in chlorinated
drinking water. They have been confirmed as DBPs in drinking water and may have
toxicological significance according to Bull et al. (2006).
IARC (1999a) evaluated haloquinones (HQ) as Group 3: not classifiable as to its carcinogenicity
to humans, based on inadequate evidence in humans and limited evidence in animals. HQ
produced benign tumors in the kidneys of male F344 rats dosed by gavage or in the diet. In the
gavage study, the tumors appeared to be the end-stage of chronic progressive nephropathy. A
nongenotoxic mode of action has been proposed based on exacerbation of spontaneously
occurring renal disease in male rats, for which there is no known relevance in humans.
4.4.13	Halogenated pyrroles
2,3,5-Tribromopyrrole was identified in drinking water in the United States (Richardson et al.,
2007). No toxicological information was identified for 2,3,5-Tribromopyrrole, bromopyrrole,
dibromopyrrole, chloropyrrole, dichloropyrrole or tricholorpyrrole.
Cytotoxicity and Genotoxicity
2,3,5-Tribromopyrrole is both highly genotoxic and highly cytotoxic in CHO cells (Richardson
et al., 2007).
Six-Year Review 3 Technical Support Document	4-73
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
4.4.14 Aldehydes
Aldehydes identified in drinking water include formaldehyde, acetaldehyde (Richardson et al.,
2007), glyoxal and methylglyoxal (USEPA, 2004). Both formaldehyde and acetaldehyde are on
the EPA fourth contaminant Candidate List (CCL4).
4.4.14.1	Formaldehyde
Cancer
USEPA (1990a) evaluated the carcinogenic potential of formaldehyde and categorized it as
"probable human carcinogen" based inhalation exposures in animals. WHO (2005b) also
classified formaldehyde in group 2A, "probably carcinogenic to humans " following inhalation
exposures. Several other agencies have examined the toxicity of formaldehyde. The California
Department of Pesticide Regulation (Cal DPR; 1997) compiled a summary of studies deemed
acceptable or unacceptable for formaldehyde toxicity and established a 2 ppm concentration
from Kerns et al. (1983) as the lowest value with an effect for nasal epithelial toxicity. USEPA
(1990) used this same inhalation study to quantify the slope factor for nasal SCC for
carcinogenicity (USEPA, 1988b).
Genotoxicity
The genotoxicity of formaldehyde has been reported in numerous studies (Richardson et al.,
2007). It induced gene mutation in bacteria, mammalian cells and rat nasal epithelium in vivo. It
was mutagenic in vitro in the presence of S9. It induced SCEs in mammalian cells and
micronuclei and chromosomal aberrations in mammalian cells and in rodents. It induced DNA
damage in bacteria and mammalian cells and DNA-protein cross-links in vitro and in rodents and
humans. It has also induced gene mutations in mouse lymphoma cells which contained large
deletions and recombinant events.
Systemic Toxicity
USEPA (1990a) generated an RfD for formaldehyde of 0.2 mg/kg/day based on reduced weight
gain a gastrointestinal histopathology in rats. The animals were exposed to concentrations of 2 to
82 mg/1 in drinking water for two years. Only the high dose had an effect. Many other agencies
have reported the various toxicological endpoints of formaldehyde (ATSDR, 1999; Health
Canada, 1997; WHO, 2002; WHO, 2005b). WHO and Health Canada used the same data as EPA
to derive their respective TDIs (0.35 mg/L Health Canada; 2.6 mg/L WHO).
4.4.14.2	Acetaldehyde
Although there is considerable toxicological information for acetaldehyde, the only risk values
determined are from inhalation exposures (USEPA, 1988b; Health Canada, 1997). The USEPA
(2004a) Office of Water Criteria Document did not establish RfDs for methylglyoxal and glyoxal
because of dose response limitations the studies that provided dose-response information.
Six-Year Review 3 Technical Support Document	4-74
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Cancer
Inhaled acetaldehyde causes tumors in the nose and trachea of hamsters and nasal cancers in rats
(USEPA, 1988b). In human it is considered a cocarcinogen with ethanol in the development of
upper digestive track cancers in alcoholics (Seitz and Stickel, 2007) especially for individuals
with certain acetaldehyde dehydogenase genotypes.
I ARC (1999b) determined that acetaldehyde is "possibly carcinogenic to humans" (Group 2B)
because there is inadequate evidence in humans for the carcinogenicity of acetaldehyde and
sufficient evidence in experimental animals. The strongest evidence for its's carcinogenicity in
humans comes from high alcohol consumers. A lifetime study of female and male rats given
drinking water containing acetaldehyde at concentrations of 0, 50, 250, 500, 1500 or 2500 mg/L
demonstrates an increase in total malignant tumors in various organs and tissues (Soffritti et al.,
2002).
Genotoxicity
The genotoxicity has been reported in numerous studies (Richardson et al., 2007). It was not
mutagenic in bacteria. It caused gene mutations, SCEs, micronuclei and chromosomal
aberrations in mammalian cells and SCEs and protein-DNA binding cross-links in rodents.
Systemic toxicity
The studies of oral exposure to acetaldehyde and few. A two-year study by Til et al. (1988)
identified irritation and changes in the GI tract after doses of >82 mg/kg/day in rats.
4.4.14.3 Glyoxal and Methylglyoxal
Cancer
USEPA (2004a) categorized the carcinogenic potential of both chemicals as cannot be
determined due to lack of human epidemiological studies and acceptable long-term animal
studies.
Genotoxicity
Glyoxal is mutagenic in the Ames assay with many strains of Salmonella typhimurium and has
been shown to cause base-pair substitutions and some frameshift mutations at G:C base pairs.
Methylglyoxal is mutagenic in bacterial systems and weakly clastogenic in rats, causing
increased micronuclei in liver and bone marrow (USEPA, 2004a).
Systemic toxicity
Both glyoxal and methyl glyoxal are associated with the formation of advanced glycosylation
end products in humans and animals as a result of crosslinking their potential to act as cross-
linking agents consequence of reacting with proteins or lipids (USEPA, 2004a). Normally this
occurs when these compounds are generated endogenously in individuals with diabetes,
atherscloersis, Alzheimer's disease and kidney failure.
Six-Year Review 3 Technical Support Document	4-75
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
4.5 Unregulated DBPs Data Availability Summary
Exhibit 4.2 shows the unregulated DBPs identified in drinking water systems that have
quantitative toxicity assessments which can be used in the assessment of risk in cases where
concentration information is available from public water systems (PWSs). The chemicals with
assessments that include RfDs or equivalents (e.g., TDI values), estimates of cancer risk
concentrations or drinking water guidelines include the following: chloral hydrate, cyanogen
chloride (based on cyanide), DBAN, dichloroacetonitrile, nitrosamines, formaldehyde and
chlorate.
Exhibit 4.2: Available Quantitative Assessments for Unregulated DBPs Discussed
in this Document
DBPs
Chemical
Reference value
Type
Biological Effect
Citation
Aldehydes
Formaldehyde
0.2 mg/kg/day
RfD
Gl tract histopath.
USEPA, 1990a
0.35 mg/L
TDI
Health Canada,
1997
2.6 mg/L
TDI
WHO, 2005b
Chlorate
Chlorate
0.03 mg/kg/day
RfD
Increased thyroid
gland follicular cell
hypertrophy
USEPA, 2006b
Cyanogen halides
Cyanogen chloride
0.11 mg/kg/day
TDI
Value developed
based on cyanide
toxicity
WHO, 2009
0.0006 mg/kg/day
RfD
(CN1-)
decreased cauda
epididymis weight
USEPA, 2010b
Haloacetoaldehydes
T richloroacetaldehyde
monohydrate (chloral
hydrate)
0.1 mg/kg/day
RfD
Liver toxicity
USEPA, 2000c
16 pg/kg/day
TDI
Liver toxicity
WHO, 2005c
Health Canada,
2008c
Haloacetonitriles
Dichloroacetonitrile
0.002.7 mg/kg/day
TDI
Increased relative
liver weight in male
and female rats
WHO 2004
Dibromoacetonitrile
0.011 mg/kg/day
TDI
Decreased body
weight in male rats
WHO, 2004
Nitrosamines
NDBA
0.03 |jg/L
E-6 conc.
Cancer
USEPA, 2016d
NDEA
0.0004 |jg/L
E-6 conc.
NDMA
0.0006 |jg/L
E-6 conc.
NDPA
0.007 |jg/L
E-6 conc.
NMEA
0.003 |jg/L
E-6 conc.
NPYR
0.002 |jg/L
E-6 conc.
Six-Year Review 3 Technical Support Document	4-76
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
In addition to the quantitative toxicity assessment information provided in Exhibit 4.2, EPA
notes the following information that has become available since the time of the Stage 2 D/DBPR
and discussed previously in this chapter. Toxicology data (subchronic, chronic) on some of the
brominated HAAs and DBAN have become available as a result of research conducted by the
NTP. These data may be used to support a quantitative evaluation of their carcinogenic effects
and possible identification of an RfD. The NTP bioassays for BCAA, BDCAA, DBAA and
DBAN are peer reviewed and published. The draft risk assessment for the halogenated furanones
(MX) requires completion and peer review.
Many of the other unregulated DBPs lack studies with dose-response to support an evaluation of
their potential to cause adverse health effects in humans. In many cases, the available data only
provide information on cytotoxicity and genotoxicity. These include the iodinated acetic acids,
iodinated trihalomethanes, HNMs, haloacetamides, halogenated benzoquinnones and
halogenated pyrazoles. Additional information is needed to more fully evaluate the health effects
of these chemicals.
Six-Year Review 3 Technical Support Document	4-77
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
5 Analytical Methods
This chapter summarizes information relevant to analytical methods for regulated and
unregulated DBPs. It provides a brief synopsis of the analytical methods developed by EPA and
others, covering methods for treatment technique (TT) requirements for removal of DBP
precursors, methods for DBPs and methods for disinfectant residuals.
For the Disinfectants and Disinfection Byproducts Rules (D/DBPR), there are no contaminants
for which the MCL/MRDL is limited by analytical feasibility. This chapter presents the
analytical methods that are currently available for D/DBPs and summarizes their performance in
cases where performance data are readily available.
Exhibit 5.1 summarizes the analytical methods developed by EPA and other method developers
(e.g., Standard Methods (SM), American Society of Testing and Materials (ASTM) International,
U.S. Geological Survey (USGS)) approved as part of the Stage 1 D/DBPR (USEPA, 1998b) and
Stage 2 D/DBPR (USEPA, 2006a), as well as those methods (referred to as alternate testing
methods) that have been approved via EPA's Expedited Method Approval process3 since Stage 2
D/DBPR promulgation. The alternate testing methods are listed in the Code of Federal
Regulations (CFR), in Appendix A to Subpart C of 40 CFR § 141.4
Exhibit 5.1: Analytical Methods Approved in the Stage 1 and Stage 2 D/DBPRs
and via the Expedited Method Approval Process1
Analyte
Approved in
Stage 1 or
Stage 2
D/DBPR2
EPA-Developed
Methods
Other Methods
Additional Methods
Approved via Expedited
Approval Since Stage 2
D/DBPR
Water Quality Parameters
Alkalinity
Stage 1
None
SM 2320 B (18th-19th ed.);
SM online 2320 B-97
ASTM D1067-92 B
USGS 1-1030-85
ASTM D1067-06 B
ASTM D1067-11 B SM 2320 B
(21st-22nd ed.)
Stage 2
None
SM 2320 B (20th ed.)
ASTM D1067-02 B
Bromide
Stage 1
300.0, Rev. 2.1;
300.1
None
None
Stage 2
317.0, Rev. 2.0;
326.0
ASTM D 6581-00
3	The Safe Drinking Water Act provides for the approval of "equally effective alternate test methods. The drinking
water Alternate Test Procedure program evaluates alternate test methods to verify that they are equally effective in
terms of method performance to approved methods in the regulations. The Expedited Method Approval process
formalizes method approvals through publication of a Federal Register notice. This process allows EPA to announce
the approval of alternate methods to laboratories and Public Water Systems in a more timely manner than traditional
rulemaking: http://water.epa.gov/scitecli/drinkingwater/labcert/analYticalmethods expedited.cfm
4	http://www.ecfr.gov/cgi-bin/text-
idx?SID=lab89b8cl4cb76ecd23585c6c2130ea2&node=pt40.23.141&rgn=div5# top
Six-Year Review 3 Technical Support Document	5-1
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Analyte
Approved in
Stage 1 or
Stage 2
D/DBPR2
EPA-Developed
Methods
Other Methods
Additional Methods
Approved via Expedited
Approval Since Stage 2
D/DBPR
Total Organic Carbon
(TOC)
Stage 1
None
SM 5310 B (19th ed.)
SM 5310 C (19th ed.)
SM 5310 D (19th ed.)
EPA 415.3, Rev. 1.2;
SM 5310 B (21st-22nd ed.);
SM 5310 C (21st-22nd ed.);
SM 5310 D (21st-22nd ed.)
Stage 2
415.3, Rev. 1.1
SM 5310 B (20th ed.)
SM 5310 C (20th ed.)
SM 5310 D (20th ed.)
SM online 5310 B-00
SM online 5310 C-00
SM online 5310 D-00

Dissolved Organic
Carbon (DOC)
Stage 1
None
SM 5310 B (19th ed.);
SM 5310 C (19thed.);
SM 5310 D (19thed.)
EPA 415.3, Rev. 1.2;
SM 5310 B (21st-22nd ed.);
SM 5310 C (21st-22nd ed.);
SM 5310 D (21st-22nd ed.)
Stage 2
415.3, Rev. 1.1
SM 5310 B (20th ed.)
SM 5310 C (20th ed.)
SM 5310 D (20th ed.)
SM online 5310 B-00
SM online 5310 C-00
SM online 5310 D-00

UV254 and Specific
Ultraviolet Light
Absorbance (SUVA)3
Stage 1
None
SM 5910 B (19th ed.) (UV254)
EPA 415.3, Rev. 1.2;
SM online 5910 B-11 (UV254);
SM 5910 B (21st-22nd ed.)
(UV254)
Stage 2
415.3, Rev. 1.1
SM 5910 B (20th ed.);
SM online 5910 B-00
(UV254)
Regulated DBPs
Trihalomethanes (THM)
Stage 1
502.2, Rev. 2.1;
524.2, Rev. 4.1;
551.1
None
EPA 524.3; EPA 524.4
Haloacetic Acids (HAA5
- MCAA, DCAA, TCAA,
MBAA, DBAA4)
Stage 1
552.1; 552.2
SM 6251 B (formerly SM
6233 B) (19thed.)
EPA 557;
SM 6251 B (21st-22nd ed.);
SM online 6251 B-07
Stage 2
552.3
SM 6251 B (20th ed.);
SM online 6251 B-94
Chlorite
Stage 1
300.0, Rev. 2.1
(monthly or daily);
300.1 (monthly or
daily)
SM 4500-CI02 E (19thed.;
daily only)
SM 4500-CI02 E (21st-22nd ed.,
daily only);
ASTM D 6581-08 A;
ASTM D 6581-08 B;
ChlordioX Plus (daily only)
Stage 2
317.0, Rev. 2.0
(monthly or daily);
326.0 (monthly or
daily);
327.0, Rev. 1.1
(daily only)
SM 4500-CI02 E (20th ed.,
daily only);
SM online 4500-CI02 E-00
(daily only);
ASTM D 6581-00
Six-Year Review 3 Technical Support Document	5-2
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Analyte
Approved in
Stage 1 or
Stage 2
D/DBPR2
EPA-Developed
Methods
Other Methods
Additional Methods
Approved via Expedited
Approval Since Stage 2
D/DBPR
Bromate
Stage 1
300.1
None
EPA 302.0; EPA 557
ASTM D 6581-08 A;
ASTM D 6581-08 B

Stage 2
321.8; 317.0, Rev.
2.0; 326.0
ASTM D 6581-00

Disinfectant Residuals
Chloramines
N/A
N/A
N/A
N/A
Chlorine
(free, combined, total)
Stage 1
None
SM 4500-CI D (19th ed.);
SM 4500-CI F (19th ed.);
SM 4500-CI G (19th ed.);
ASTM D1253-86
ASTM D1253-08;
SM 4500-CI D (21st-22nd ed.);
SM 4500-CI F (21st-22nd ed.);
SM 4500-CI G (21 st-22nd ed.);
Hach Method 10260

Stage 2
None
SM 4500-CI D (20th ed.);
SM 4500-CI F (20th ed.);
SM 4500-CI G (20th ed.);
SM online 4500-CI D-00;
SM online 4500-CI F-00;
SM online 4500-CI G-00;
ASTM D1253-86(96);
ASTM D1253-03

Chlorine (total) in
addition to those listed
above for free,
combined, total
Stage 1
None
SM 4500-CI E (19th ed.);
SM 4500-CI I (19th ed.)
EPA 334.0;
ChloroSense;
SM 4500-CI E (21st-22nd ed.);
SM 4500-CI I (21st-22nd ed.)

Stage 2
None
SM 4500-CI E (20th ed.);
SM 4500-CI I (20th ed.);
SM online 4500-CI E-00;
SM online 4500-CI I-00

Chlorine (free) in
addition to those listed
above for free,
combined, total
Stage 1
None
SM 4500-CI H (19th ed.)
EPA 334.0;
ChloroSense;
SM 4500-CI H (21st-22nd ed.);
Method D99-003 (if approved
by state)

Stage 2
None
SM 4500-CI H (20th ed.);
SM online 4500-CI H-00

Chlorine Dioxide
Stage 1
None
SM 4500-CI02 D (19th ed.);
SM 4500-CI02 E (19th ed.)
4500-CI02 E (21st-22nd ed.)
ChlordioX Plus

Stage 2
327.0, Rev. 1.1
SM 4500-CI02 D (20th ed.);
SM 4500-CI02 E (20th ed.);
SM online 4500-CI02 E-00

EPA Methods Cited:
EPA Method 300.0, Rev. 2.1 (USEPA, 1993b)
EPA Method 300.1 (USEPA, 1997c)
EPA Method 302.0 (USEPA, 2009a)
EPA Method 317.0, Rev. 2.0 (USEPA, 2001c)
EPA Method 321.8 (USEPA, 1997d)
EPA Method 326.0 (USEPA, 2002)
EPA Method 327.0, Rev. 1.1 (USEPA, 2005j)
EPA Method 334.0 (USEPA, 2009b)
EPA Method 415.3, Rev. 1.1 (USEPA,
EPA Method 415.3, Rev. 1.2 (USEPA,
EPA Method 502.2, Rev. 2.1 (USEPA,
EPA Method 524.2, Rev. 4.1 (USEPA,
EPA Method 524.3 (USEPA, 2009d)
EPA Method 524.4 (USEPA, 2013)
EPA Method 551.1 (USEPA, 1995c)
2005k) EPA Method 552.1 (USEPA, 1992c)
2009c) EPA Method 552.2 (USEPA, 1995d)
1995a) EPA Method 552.3 (USEPA, 2003e)
1995b) EPA Method 557 (USEPA, 2009e)
Six-Year Review 3 Technical Support Document	5-3
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
1	EPA's Expedited Method Approval Process was implemented in 2007, subsequent to the publication of the Final
Stage 2 D/DBPR in 2006, and includes those analytical methods that may provide opportunities for improved
performance and/or increased method sensitivity relative to the analytical methods approved in Stage 1 or Stage 2:
http://water.epa.qov/scitech/drinkinqwater/labcert/analvticalmethods expedited.cfm
2	Any analytical method approved under the Stage 1 D/DBPR was also approved for use under Stage 2. Specifically,
40 CFR §141.131 (a)(1) of the Stage 2 D/DBPR states that the analytical methods specified for compliance monitoring
are effective February 16, 1999, which is the effective date of the Stage 1 D/DBPR. The Stage 2 D/DBPR also
includes additional methods that are specified for compliance monitoring.
3	SUVA = UV254 / DOC
4	MCAA = monochloroacetic acid; DCAA = dichloroacetic acid; TCAA = trichloroacetic acid; MBAA =
monobromoacetic acid; DBAA = dibromoacetic acid
Exhibit 5.2 summarizes the analytical methods developed by EPA and approved via expedited
approval or other EPA rulemaking for the unregulated DBPs.
Exhibit 5.2: Analytical Methods for Unregulated DBPs Approved via the Expedited
Method Approval Process or Other EPA Rulemaking
Analyte
EPA-
Developed
Methods
Other
Methods
Additional Methods Approved via
Expedited Approval or Other EPA
Rulemaking1
HAAs (BCAA, BDCAA, DBCAA, TBAA2)
EPA 552.2
N/A
EPA 552.3; EPA 557
Nitrosamines
N/A
N/A
EPA 521
Chlorate
N/A
N/A
EPA 300.0, Rev. 2.1; EPA 300.1; EPA 317.0,
Rev. 2.0; EPA 326.0; SM 4110 D (21st ed.);
ASTM D 6581-00; ASTM D 6581-08
EPA Methods Cited:	EPA Method 326.0 (USEPA, 2002)	EPA Method 552.3 (USEPA, 2003e)
EPA Method 300.0, Rev. 2.1 (USEPA, 1993b) EPA Method 521 (USEPA, 2004b)
EPA Method 300.1 (USEPA, 1997c)	EPA Method 557 (USEPA, 2009e)
EPA Method 317.0, Rev. 2.0 (USEPA, 2001c) EPA Method 552.2 (USEPA, 1995d)
1	For the unregulated DBPs, methods that are approved for compliance monitoring of related regulated analytes, or
methods that have been specified for analytes listed in EPA's Unregulated Contaminant Monitoring Rule (UCMR), or
other recently-developed methods are listed.
2	BCAA = bromochloroacetic acid; BDCAA = bromodichloroacetic acid; DBCAA = dibromochloroacetic acid; TBAA =
tribromoacetic acid. The regulated HAA5 plus these four unregulated HAAs = HAA9.
The following discussion defines method performance metrics for the DBPs and disinfectant
residuals listed in Exhibit 5.1 and Exhibit 5.2. These metrics are presented in subsequent sections
of Chapter 5 for new methods approved since the Stage 2 D/DBPR was published in January
2006. This allows a comparison of method performance for those methods approved under the
Stage 1 or Stage 2 D/DBPRs and those methods approved via the Expedited Method Approval
process since the final Stage 2 D/DBPR was published. These metrics include the following:
Method detection limit (MDL) and detection limit (DL) -The MDL is defined as "the minimum
concentration of a substance that can be reported with 99 percent confidence that the analyte
concentration is greater than zero."5 The steps for determining the MDL are outlined in 40 CFR
§136, Appendix B. Over time, drinking water compliance methods have migrated away from
requiring MDL determinations in favor of confirming minimum reporting levels (see discussion
5 40 CFR § 136 Appendix B: http://www.gpo.gov/fdsvs/granule/CFR-2011 -title40-vol23/CFR-2011 -titlc40-vol23-
part 136-appB/contcnt-dct;iil.html
Six-Year Review 3 Technical Support Document	5-4
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
below). Various regulatory bodies, however, still require determination of detection limits. As a
result, most of the newer drinking water analytical methods incorporate a detection limit (DL)
determination that is defined and conducted exactly like the MDL (e.g., EPA Method 524.3).
The MDLs and DLs are shown in Exhibit 5.3 through Exhibit 5.9.
The lowest concentration minimum reporting level (LCMRL) - the LCMRL is defined as the
lowest spiking concentration such that the probability of spike recovery in the 50 to 150 percent
range is at least 99 percent (USEPA, 2010c). The LCMRL appears in recently developed
analytical methods from EPA and serves as a laboratory- and analyte-specific reporting level.
Different analysts using different equipment in different laboratories will not necessarily be able
to achieve LCMRLs that are published in EPA analytical methods; however, EPA's published
LCMRLs are an indication that low analyte concentrations can be reliably reported. The LCMRL
has been used in EPA's second and third Unregulated Contaminant Monitoring Rules (UCMR 2
and UCMR 3). With the exception of the LCMRL for chlorate, the LCMRLs summarized in
Section 5.2 are listed in the analytical methods represented in Exhibit 5.3 through Exhibit 5.9.
The LCMRLs for chlorate were developed internally by EPA during UCMR 3.
The minimum reporting level (MRL) - the MRL has evolved over time in EPA programs. In the
preamble to the proposed Stage 2 D/DBPR,6 MRLs were initially established for DBPs as part of
the 1996 Information Collection Rule. These MRLs were also proposed in the proposed Stage 2
D/DBPR and were established in the final Stage 2 D/DBPR.7 The MRLs were not determined
through a formal, statistical procedure; rather, they were based on recommendations from experts
with experience in the analysis of DBPs. The MRLs were established at concentrations at which
most laboratories could meet the precision and accuracy criteria of the analytical methods
designated for the analysis of DBPs in drinking water. These "consensus" MRLs were developed
for the trihalomethanes (THMs), the five regulated haloacetic acids (HAA5), chlorite and
bromate.
At about the same time the Stage 2 D/DBPR proposal was moving forward, EPA began
exploring development of a statistical procedure for determining laboratory- and analyte-specific
LCMRLs. In conjunction with the LCMRL, a statistically derived MRL procedure was also
developed. This MRL is determined using raw LCMRL study data and represents an estimate of
the lowest concentration of a contaminant that can be reliably measured by members of a group
of experienced drinking water laboratories (USEPA, 2007a). For six nitrosamines (N-
nitrosodimethyl amine (NDMA), A'-nitrosodi ethyl amine (NDEA), A-nitrosodi-n-propyl amine
(NDPA), A'-nitrosodi-n-butyl amine (NDBA), A^-nitrosom ethyl ethyl amine (NMEA) and N-
nitrosopyrrolidine (NPYR)), the MRL served as a national reporting level for laboratories that
participated in the analysis of drinking water samples under UCMR 2 using EPA Method 521.
For chlorate, the MRL served as a national reporting level for laboratories that participated in the
6	68 FR 49548, National Primary Drinking Water Regulations: Stage 2 Disinfectants and Disinfection Byproducts
Rule; National Primary and Secondary Drinking Water Regulations: Approval of Analytical Methods for Chemical
Contaminants, Proposed Rule, August 2003. Available on the Internet at: http://www. gpo. gov/fdsvs/pkg/FR-2003 -
08-18/pd1703-18149.pdf
7	71 FR 388, National Primary Drinking Water Regulations: Stage 2 Disinfectants and Disinfection Byproducts
Rule; Final Rule, January 2006. Available in the Internet at: htto://www. gpo. gov/fdsvs/pkg/FR-2006-0 l-04/pdf/06-
3.Pdf
Six-Year Review 3 Technical Support Document	5-5
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
analysis of drinking water samples under UCMR 3 using EPA Method 300.1, ASTM D6581-08,
or SM 4110D (21st ed.). For DBPs, MRLs that are statistically derived from LCMRL study data
are currently only available for the six nitrosamines and chlorate.
Percent recovery range and percent relative standard deviation (RSD) - the percent recovery
range demonstrates the overall accuracy of the methods for each analyte, and the percent RSDs
demonstrate the overall precision of the methods for each analyte. The data are summarized for
precision and accuracy studies documented in each analytical method, but the data do not include
holding time study nor MDL study recovery percentages or percent RSDs, since holding time
studies are less about method performance than about analyte stability and since MDL studies
are typically conducted at concentrations well below those used in precision and accuracy
studies. The precision and accuracy studies are typically performed in reagent water and/or
finished drinking water from either ground water or surface water sources. In some cases,
challenging environmental matrices are simulated by fortifying reagent water with additives such
as humic acids, or other organic or inorganic additives.
The following sections illustrate the complexity and variety in the sources of analytical methods
that were approved as part of the Stage 1 and/or Stage 2 D/DBPRs and those methods that have
subsequently been approved via EPA's Expedited Method Approval process.
EPA also used the National Environmental Methods Index (NEMI) to search for performance
metrics for methods not developed by EPA to provide some context for the data that might be
available for these other methods (see Section 5.1.1 for examples). NEMI is a database of
analytical methods and summary data for analytical methods and is run by the National Water
Quality Monitoring Council in conjunction with EPA and USGS. NEMI can be searched by
analytical method number.8 The following sections compare analytical method performance for
methods approved for the analysis of DBPs and disinfection residuals for EPA-developed
analytical methods and SMs (APHA, AWWA and WEF, 2012) that contain performance data.
Method performance data are shown in exhibits, where data for multiple EPA-developed
analytical methods are available. Issues associated with analytical methods developed by
organizations other than EPA are also discussed. The purpose of the comparison is to provide
information on the performance of analytical methods that have been approved by EPA since the
Stage 2 D/DBPR relative to those analytical methods that were approved in the Stage 1 or Stage
2 D/DBPR.
5.1 Methods for Treatment Technique Requirement for Removal of DBP Precursors
5.1.1 Alkalinity
For alkalinity, only non-EPA-developed analytical methods were approved for monitoring in the
Stage 1 and Stage 2 D/DBPRs, and only non-EPA-developed analytical methods have been
approved via EPA's Expedited Method Approval process since the final Stage 2 D/DBPR was
published.
8 https://www.nemi.gov/home/
Six-Year Review 3 Technical Support Document	5-6
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
For the Stage 1 D/DBPR, SM 2320 B (18th-19th ed.), ASTM D1067-92 B and USGS 1-1030-85
were approved for compliance monitoring. These same methods remained approved under the
Stage 2 D/DBPR including the updated version of SM 2320 B in the 20th edition of Standard
Methods for the Examination of Water and Wastewater (APHA, AWWA and WEF, 1995) and
ASTM D1067-02 B. Since publication of the Stage 2 D/DBPR in January 2006, alternate testing
methods that have been approved for alkalinity via EPA's Expedited Method Approval process
include ASTM D1067-06 B, ASTM D1067-11 B and SM 2320 B in the 21st and 22nd editions.
Although a record for ASTM D1067 is available in NEMI, there are no details regarding method
performance, and the method is not available for downloading. USGS 1-1030-85 is included in
NEMI. The only metrics available in NEMI for method performance are mean recovery,
standard deviation and percent RSD for the analysis of a single sample by 21 different
laboratories. A mean result of 26.0 mg/L as H+ was reported, with a standard deviation of 0.9
mg/L as H+ and a percent RSD of 3.5 percent.
A review of SM 2320 B indicates that the lowest concentration that can be determined is 20 mg
CaCCb/L. Lower concentrations must be determined using Part 4d of SM 2320 B. SM 2320 B is
reported to be of low bias (APHA, AWWA and WEF, 2012);9 however, percent recovery and
percent RSD are not included in the method.
5.1.2	Bromide
For bromide both EPA-developed and non-EPA-developed analytical methods were approved
for monitoring in the Stage 1 and Stage 2 D/DBPRs; however, no analytical methods have been
approved via EPA's Expedited Method Approval process.
For the Stage 1 D/DBPR, EPA Methods 300.0 (Rev. 2.1) and 300.1 were approved for
compliance monitoring. These same methods remained approved under the Stage 2 D/DBPR,
and EPA Methods 317.0 (Rev. 2.0) and 326.0, as well as ASTM D 6581-00, were also approved
for compliance monitoring under Stage 2. Since publication of the Stage 2 D/DBPR, no alternate
testing methods have been approved for bromide via EPA's Expedited Method Approval
process. Since no analytical methods have been approved for the monitoring of bromide since
publication of the Stage 2 D/DBPR, no performance data for the bromide methods are presented.
5.1.3	Total Organic Carbon (TOC) and Dissolved Organic Carbon (DOC)
For TOC and DOC, both EPA-developed and non-EPA-developed analytical methods were
approved for monitoring in the Stage 1 and Stage 2 D/DBPRs, and one EPA-developed method
and updated non-EPA methods have been approved via EPA's Expedited Method Approval
process.
9 APHA, AWWA and WEF, 2012 refers to the 22nd edition of SM. In a discussion with Dr. Glynda Smith of EPA's
Technical Support Center on March 19, 2015, she indicated that performance data do not change for SM methods
from edition to edition. If the method changes to the extent that performance changes, this is considered a major
modification and a new method number is assigned by SM. Thus, performance data from the 22nd edition are
applicable to the 18th - 21st editions also.
Six-Year Review 3 Technical Support Document	5-7
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Methods that have been approved for the monitoring of TOC are also approved for DOC; hence,
they are combined into a single section here. For the Stage 1 D/DBPR, SM 5310 B (19th ed.),
5310 C (19th ed.) and 5310 D (19th ed.) were approved for compliance monitoring. These same
methods remained approved in the Stage 2 D/DBPR, including the updated versions in the 20th
edition of Standard Methods for the Examination of Water and Wastewater. EPA Method 415.3
(Rev. 1.1) and SM online 5310 B-00, 5310 C-00 and 5310 D-00 were also approved for
compliance monitoring. Since publication of the Stage 2 D/DBPR, alternate testing methods
approved via EPA's Expedited Method Approval process for TOC and DOC have included EPA
Method 415.3 (Rev. 1.2) and SM 5310 B, SM 5310 C and SM 5310 D in the 21st and 22nd
editions. SM (APHA, AWWA and WEF, 2012) reports recovery and percent RSD only for SM
5310 D in reagent water. EPAMethods 415.3 (Rev. 1.1) and 415.3 (Rev. 1.2) report percent
recovery and percent RSD in fortified environmental waters, thus, no meaningful comparison of
relative performance can be made. In addition, EPA Methods 415.3 (Rev. 1.1) and 415.3 (Rev.
1.2) report the same method performance data (DLs, percent recovery and percent RSD), thus,
no comparison of relative performance of the two EPA-developed methods can be made.
5.1.4 UV254 and Specific Ultraviolet Light Absorbance (SUVA)
For UV254 and specific ultraviolet light absorbance (SUVA), both EPA-developed and non-EPA-
developed analytical methods were approved for monitoring in the Stage 1 and Stage 2
D/DBPRs, and one EPA-developed method and updated non-EPA methods were approved via
EPA's Expedited Method Approval process since the final Stage 2 D/DBPR was published.
SUVA is not included in the Stage 1 D/DBPR; however, UV254 is included and SM 5910 B (19th
ed.) was approved for compliance monitoring. SUVA was introduced in Stage 2 and is
determined from UV254 and DOC (SUVA = UV254 / DOC). SM 5910 B remained approved
through the Stage 2 D/DBPR, including the updated version in the 20th edition of Standard
Methods for the Examination of Water and Wastewater. EPA Method 415.3 (Rev. 1.1) and SM
online 5910 B-00 were also approved for compliance monitoring of UV254 and the determination
of SUVA. Since publication of the Stage 2 D/DBPR, alternate testing methods for UV254 and
SUVA that have been approved via EPA's Expedited Method Approval process include EPA
415.3 (Rev. 1.2), SM 5910 B in the 21st and 22nd editions and online SM 5910 B-ll. SM (APHA,
AWWA and WEF, 2012) reports multi-laboratory and single operator percent RSD only for SM
5910 B in reagent water. EPA Methods 415.3 (Rev. I.l)and415.3 (Rev. 1.2) report percent
recovery and percent RSD in fortified environmental waters, thus, no meaningful comparison of
relative performance can be made. EPA Methods 415.3 (Rev. 1.1) and 415.3 (Rev. 1.2) report
the same quality control data for DLs, percent recovery and percent RSD, thus, no comparison of
relative performance of the two EPA-developed methods can be made either.
5.2 Methods for Disinfection Byproducts
5.2.1 THM
For THMs, only EPA-developed methods were approved in the Stage 1 and Stage 2 D/DBPRs,
and only EPA-developed methods have been approved via EPA's Expedited Method Approval
process since the final Stage 2 D/DBPR was published. Since EPA methods are available at no
Six-Year Review 3 Technical Support Document	5-8
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
cost online, the availability of the analytical methods for a comparison of performance metrics is
not an issue.
For the Stage 1 D/DBPR, EPA Methods 502.2 (Rev. 2.1), 524.2 (Rev. 4.1) and 551.1 were
approved for compliance monitoring. These same methods remained approved under the Stage 2
D/DBPR. Alternate testing methods for THMs approved via EPA's Expedited Method Approval
process since Stage 2 include EPA Methods 524.3 and 524.4. Exhibit 5.3 summarizes the
available detection limits, LCMRLs, percent recoveries and percent RSDs for the five EPA-
developed analytical methods for the THMs. MRLs are from the final Stage 2 D/DBPR.
Exhibit 5.3: Method Performance Metrics for EPA Methods 502.2, 524.2, 524.3,
524.4 and 551.1 - THMs
Method/Analyte
DL/MDL
(M9/L)
LCMRL (|jg/L)
MRL (|jg/L)
Mean %
Recovery
Range
% RSD
Range
Fortifi-
cation (|jg/L)
502.2
(MDL)

Matrix:
Reagent water
Bromodichloromethane
0.02-0.10
Not determined
1.0 for each analyte
96-97
2.6-2.9
10
Bromoform
0.09-1.6
Not determined
98-106
4.0-5.2
10
Chloroform
0.01-0.02
Not determined
92-98
2.5-4.2
10
Dibromochloromethane
0.05-0.3
Not determined
99-102
2.0-3.3
10
524.2
(MDL)

Matrix:
Reagent water
Bromodichloromethane
0.03-0.08
Not determined
1.0 for each analyte
96-100
1.8-1.8
0.2, 2
Bromoform
0.12-0.20
Not determined
89-90
2.2-2.4
0.2, 2
Chloroform
0.02-0.03
Not determined
95-97
2.0-2.1
0.2, 2
Dibromochloromethane
0.05-0.07
Not determined
95-100
2.7-3.0
0.2, 2
524.3
(DL)

Matrices:
Reagent water, chlorinated ground
water, chlorinated surface water
Bromodichloromethane
0.014
0.073
1.0 for each analyte
92.8-102
1.2-8.7
0.5-10
Bromoform
0.040
0.15
78.1-92.6
2.2-8.1
0.5-10
Chloroform
0.025
0.054
80.9-99.4
1.7-8.3
0.5-10
Dibromochloromethane
0.027
0.14
86.1-97.7
1.2-7.9
0.5-10
524.4
(DL)

Matrices:
Reagent water, chlorinated ground
water, chlorinated surface water
Bromodichloromethane
0.011-0.081
0.027-0.19
1.0 for each analyte
87.3-104
1.5-8.5
0.5, 1, 10
Bromoform
0.008-0.14
0.021-0.26
80.6-103
2.4-6.8
0.5, 1, 10
Chloroform
0.015-0.070
0.032-0.16
86-103
1.7-6.1
0.5, 1, 10
Dibromochloromethane
0.006-0.10
0.016-0.23
90.7-102
1.3-5.2
0.5, 1, 10
551.1
(MDL)

Matrices:
Reagent water, fulvic acid
enhanced reagent water, high
hardness chlorinated ground water
Bromodichloromethane
0.002-0.068
Not determined
1.0 for each analyte
87-110
1.02-4.07
0.25, 1, 5
Bromoform
0.004-0.020
Not determined
82-104
0.72-2.76
0.25, 1, 5
Chloroform
0.005-0.080
Not determined
92-105
1.20-3.68
0.25, 1, 5
Dibromochloromethane
0.001-0.018
Not determined
85-106
0.71-3.38
0.25, 1, 5
Six-Year Review 3 Technical Support Document	5-9
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
A review of the performance data in Exhibit 5.3 indicates that the more recently approved
analytical methods (EPA Methods 524.3 and 524.4) are comparable to the methods that were
approved under Stage 1 and Stage 2 in terms of sensitivity, recovery and RSD. All of the
methods in Exhibit 5.3 meet the individual method requirements for percent recovery and
percent RSD.
5.2.2 HAA5
For HAA5, both EPA-developed methods and non-EPA-developed methods were approved for
monitoring during the Stage 1 and Stage 2 D/DBPRs, and both EPA-developed and updated non-
EPA methods have been approved via EPA's Expedited Method Approval process since the final
Stage 2 D/DBPR was published. However, performance data are only readily available for the
EPA-developed methods and SM 6251 B (22nd ed).
EPA Methods 552.1 and 552.2 were approved for Stage 1 analyses, and EPA Method 552.3 was
added under Stage 2. Equivalent methods approved under the Stage 1 and/or Stage 2 D/DBPRs
include SM 6251 B (formerly SM 6233 B) in the 19th and 20th editions of Standard Methods for
the Examination of Water and Wastewater and SM online 6251 B-94. Since the final Stage 2
D/DBPR was published, EPA Method 557, SM 6251 B in the 21st and 22nd editions and SM
online 6251 B-07 have been approved via the Expedited Method Approval process.
HAA5 consists of monochloroacetic acid (MCAA), dichloroacetic acid (DCAA), trichloroacetic
acid (TCAA), monobromoacetic acid (MBAA) and dibromoacetic acid (DBAA). EPA Method
552.1 includes these five HAAs along with bromochloroacetic acid (BCAA), which is not
regulated. SM 6233 B listed only the HAA5 analytes. When a standard for BCAA became
available, SM 6233 B was re-designated SM 6251 B and BCAA was added to the method with
the HAA5 analytes.10 EPA Methods 552.2, 552.3 and 557 were published in 1995, 2003 and
2009, respectively, and include nine HAAs, the five regulated contaminants plus four additional
unregulated brominated HAAs.
Exhibit 5.4 summarizes the DLs, LCMRLs, mean percent recovery values and percent RSDs for
HAA5 as listed in EPA Methods 552.1, 552.2, 552.3 and 557, along with metrics for SM 6251 B.
MRLs are from the final Stage 2 D/DBPR.
Exhibit 5.4: Method Performance Metrics for EPA Methods 552.1, 552.2, 552.3 and
557 and for SM 6251 B - HAA5
Method/
Analyte
DL/MDL
(M9/L)
LCMRL (|jg/L)
MRL (|jg/L)
Mean %
Recovery Range
% RSD
Range
Fortifi-
cation (|jg/L)
552.1
(MDL)

Matrices:
Reagent water, dechlorinated tap water, high
ionic strength water, high humic content ground
water, ozonated river water
MCAA
0.21
Not determined
2.0
46-109
1.0-15
7.5, 15
10 63 FR 69390. 1998. National Primary Drinking Water Regulations: Disinfectants and Disinfection Byproducts,
Final Rule, December 16, 1998. Available on the Internet at: http://www.gpo.gov/fdsvs/pkg/FR-1998-i2-16/pdf/98-
32887.pdf
Six-Year Review 3 Technical Support Document	5-10
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Method/
Analyte
DL/MDL
(M9/L)
LCMRL (|jg/L)
MRL (|jg/L)
Mean %
Recovery Range
% RSD
Range
Fortifi-
cation (|jg/L)
MBAA
0.24
Not determined
1.0
5-91
7.9-18
5, 10
DCAA
0.45
Not determined
1.0
59-114
0.1-14
7.5, 15
TCAA
0.07
Not determined
1.0
8-106
0.4-28
2.5, 5
DBAA
0.09
Not determined
1.0
40-103
0.7-22
2.5, 5
552.2
(MDL)

Matrices:
Reagent water, dechlorinated surface water, high
ionic strength water, high humic content ground
water
MCAA
0.273
Not determined
2.0
84.3-97
2.8-13
1.5, 3, 6
MBAA
0.204
Not determined
1.0
86.0-109
1.5-11
1, 2, 4
DCAA
0.242
Not determined
1.0
84.7-115
2.5-11
1.5, 3, 6
TCAA
0.079
Not determined
1.0
61.8-93
6.3-15
0.5, 1,2
DBAA
0.066
Not determined
1.0
71.5-112
2.8-9.2
0.5, 1,2
552.3
(DL)

Matrices:
Reagent water, chlorinated surface water,
chlorinated ground water
MCAA
0.17-0.20
Not determined
2.0
81.4-131
1.7-9.5
1, 10
MBAA
0.027-0.13
Not determined
1.0
90.7-113
1.1-4.2
1, 10
DCAA
0.020-0.084
Not determined
1.0
93.8-107
0.33-3.8
1, 10
TCAA
0.019-0.024
Not determined
1.0
89.0-107
0.52-2.1
1, 10
DBAA
0.012-0.021
Not determined
1.0
101-111
0.52-5.3
1, 10
557
(DL)

Matrices:
Reagent water, synthetic sample matrix,
chlorinated ground water, chlorinated surface
water
MCAA
0.20
0.58
2.0
95.9-109
1.7-5.2
1, 2.5, 8, 10, 15
MBAA
0.064
0.19
1.0
97.2-101
1.4-5.3
1, 2.5, 8, 10, 15
DCAA
0.055
0.13
1.0
79.6-109
1.7-9.3
1, 2.5, 8, 10, 15
TCAA
0.090
0.25
1.0
95.6-107
1.1-5.4
1, 2.5, 8, 10, 15
DBAA
0.015
0.062
1.0
84.5-111
6.0-14
1, 2.5, 8, 10, 15
SM 6251 B
(22nd ed.)
(MDL)

Matrices:
Reagent water
MCAA
0.082
Not determined
2.0
78.9-98.0
3.88-5.92
1, 5
MBAA
0.087
Not determined
1.0
70.6-99.0
2.67-4.76
1, 5
DCAA
0.054
Not determined
1.0
99.0-110
3.11-4.38
1, 5
TCAA
0.054
Not determined
1.0
92.7-101
3.06-5.49
1, 5
DBAA
0.065
Not determined
1.0
99.6-116
2.75-3.11
1, 5
A review of the data in Exhibit 5.4 indicates that the methods show an improvement in percent
recovery and percent RSD as the methods evolved from EPA Method 552.1 to 552.2 to 552.3.
All three methods include a derivatization step with acidic methanol, wherein the halogenated
carboxylic acids are converted to their corresponding methyl esters. EPA Method 557 does not
include the derivatization step; hence the issue of the efficiency of the conversion of the
carboxylic acids to the corresponding methyl esters is eliminated (however, the instrumentation
used in EPA Method 557 is much more costly than the instrumentation used in EPA Methods
552.1, 552.2 and 552.3, so simply switching to EPA Method 557 is not always feasible). How
this conversion is accomplished and how the efficacy of the conversion is monitored has changed
Six-Year Review 3 Technical Support Document	5-11
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
as the methods evolved. Since the most marked effect is on the recovery of the unregulated
HAAs, the discussion is presented further in Section 5.2.5.1.
EPA Method 552.1 demonstrates high variability in percent recovery. The very low recoveries
correspond to matrices of high ionic strength and high humic acid content. Some of the percent
RSDs are on the high side (20 percent is a typical high end of the acceptable range, although
EPA Method 551.1 requires percent recovery that is within three standard deviations of the mean
recovery); however, the percent RSDs do not show the extremes that are seen in the percent
recovery data. This suggests that low recoveries can be a problem with this method. EPA
Method 552.1 employs a solid phase extraction procedure that may have contributed to the low
recoveries. EPA Method 552.2 uses a liquid/liquid extraction procedure, and the recoveries show
marked improvement relative to EPA Method 552.1; however, they are still in the low range,
especially for DBAA and TCAA in challenging matrices.
The MDLs from SM 6251 B are similar in magnitude to those from the EPA-developed methods.
The percent recovery and percent RSD data from SM 6251 B are from fortified reagent water
samples only, so a comparison with the EPA method performance data would likely not be
meaningful.
5.2.3 Chlorite
For chlorite, both EPA-developed and non-EPA-developed analytical methods were approved
for monitoring in the Stage 1 and Stage 2 D/DBPRs, and several non-EPA-developed methods
have been approved via EPA's Expedited Method Approval process since the final Stage 2
D/DBPR was published. No meaningful comparison can be made between the Stage 1 and Stage
2-approved methods and those approved since Stage 2 for the Palintest ChlordioX Plus or ASTM
methods since performance data are not available on the NEMI website. However, the Palintest
ChlordioX Plus amperometric sensor method was approved in the June 2014 Expedited Method
Approval Action (USEPA, 2014a) for daily monitoring of chlorite as an alternative to the
approved amperometric titration methodology employed in SM 4500-C102 E.
For the Stage 1 D/DBPR, EPA Methods 300.0 (Rev. 2.1) and 300.1 and SM 4500-C102 E (for
daily checks only) in the 19th edition of Standard Methods for the Examination of Water and
Wastewater were approved for compliance monitoring. These same methods remained approved
under the Stage 2 D/DBPR. Under Stage 2, EPA Methods 317.0 (Rev. 2.0), 326.0 and 327.0
(Rev. 1.1, for daily checks only), along with SM 4500-C102 E in the 20th edition of Standard
Methods for the Examination of Water and Wastewater (daily checks only), SM online 4500-
CIO2 E-00 (daily checks only) and ASTM D 6581-00 were also approved for compliance
monitoring. Since publication of the Stage 2 D/DBPR, alternate testing methods approved via
EPA's Expedited Method Approval process for chlorite include SM 4500-C102 E in the 21st and
22nd editions of Standard Methods for the Examination of Water and Wastewater (daily checks
only), ASTM D 6581-08 A and B and Palintest ChlordioX Plus (for daily checks only). Exhibit
5.5 summarizes the DLs, LCMRLs, mean percent recovery values and percent RSDs for chlorite
as listed in EPA Methods 300.0 (Rev. 2.1), 300.1, 317.0 (Rev. 2.0), 326.0 and 327.0 (Rev. 1.1).
Method performance data are not included in SM 4500-C102 E (APHA, AWWA and WEF,
2012). The MRL is from the final Stage 2 D/DBPR.
Six-Year Review 3 Technical Support Document	5-12
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Exhibit 5.5: Method Performance Metrics for EPA Methods 300.0 (Rev. 2.1), 300.1,
317.0 (Rev. 2.0), 326.0 and 327.0 (Rev. 1.1) — Chlorite
Method/Analyte
DL/MDL
(M9/L)
LCMRL (|jg/L)
MRL
(H9/L)
Mean % Recovery
Range
% RSD
Range
Fortifi-
cation (|jg/L)
300.0 (Rev. 2.1)
(MDL)

Matrices:
Reagent water, drinking water
Chlorite
10
Not determined
20
76.0-100
N/A
0.05, 0.1, 1, 5
300.1
(MDL)

Matrices:
Reagent water, high ionic strength water,
surface water, ground water, chlorinated
drinking water, chlorine dioxide-treated drinking
water, ozonated drinking water
Chlorite
0.45-1.44
Not determined
20
84.4-105
0.41-2.15
100, 500
317.0 (Rev2.0)
(MDL)

Matrices:
Reagent water, high ionic strength water,
surface water, ground water, chlorinated
drinking water, chlorine dioxide-treated drinking
water, ozonated drinking water
Chlorite
0.45-0.89
Not determined
20
84.4-105
0.41-2.15
100, 500
326.0
(DL)

Matrices:
Reagent water, high ionic strength water, high
organic content water
Chlorite
2.0
Not determined
20
99.3-108
0.49-3.0
100, 500
327.0 (Rev. 1.1)
(DL)

Matrices:
Reagent water, chlorinated surface water,
chlorinated ground water
Chlorite
0.078-
0.11
Not determined
20
98.5-110
1.4-4.4
1, 2
A review of the performance data in Exhibit 5.5 indicates that EPA Method 327.0 (Rev. 1.1)
shows an increase in sensitivity (i.e., in the MDL/DL) relative to the other methods and both
EPA Methods 326.0 and 327.0 (Rev. 1.1) show improved recovery relative to the other methods.
However, a greater number of potentially challenging matrices were evaluated in EPA Methods
300.0 (Rev. 2.1), 300.1 and 317.0 (Rev. 2.0).
5.2.4 Bromate
For bromate, both EPA-developed and non-EPA-developed analytical methods were approved
for monitoring in the Stage 1 and Stage 2 D/DBPRs and EPA-developed and ASTM-developed
methods have been approved via EPA's Expedited Method Approval process since the final
Stage 2 D/DBPR was published. A comparison of the EPA-developed methods is presented in
Exhibit 5.6.
For the Stage 1 D/DBPR, EPA Method 300.1 was approved for compliance monitoring. This
method remained approved under the Stage 2 D/DBPR, wherein EPA Methods 317.0 (Rev. 2.0),
321.8 and 326.0, as well as ASTM D 6581-00, were also approved for compliance monitoring.
Since publication of the Stage 2 D/DBPR, alternate testing methods for bromate that have been
approved via EPA's Expedited Method Approval process include EPA Methods 302.0 and 557,
as well as ASTM D 6581-08 A and ASTM D 6581-08 B. Exhibit 5.6 summarizes the DLs,
LCMRLs, mean percent recovery values and percent RSDs for bromate as listed in EPA
Methods 300.1, 317.0 (Rev. 2.0), 321.8, 326.0, 302.0 and 557. MRLs are from the final Stage 2
D/DBPR.
Six-Year Review 3 Technical Support Document	5-13
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Exhibit 5.6: Method Performance Metrics for EPA Methods 300.1, 317.0 (Rev. 2.0),
321.8, 326.0, 302.0 and 557 — Bromate
Method/Analyte
DL/MDL
(M9/L)
LCMRL (|jg/L)
MRL
(ng/L)11
Mean %
Recovery
Range
% RSD Range
Fortifi-
cation (|jg/L)
300.1
(MDL)

Matrices:
Reagent water, high ionic strength water,
surface water, ground water, chlorinated
drinking water, chlorine dioxide-treated
drinking water, ozonated drinking water
Bromate
1.28-1.44
Not Determined
5.0
80.9-106
4.18-19.5
5, 25
317.0 (Rev. 2.0)
(MDL)

Matrices:
Reagent water, high ionic strength water, high
organic content water, surface water, ground
water, chlorinated drinking water, chlorine
dioxide-treated drinking water, ozonated
drinking water
Bromate
0.12-0.98
Not Determined
1.0
80.9-108
1.87-21.4
0.5, 5, 25
321.8
(DL)

Matrices:
Ozonated drinking water
Bromate
0.3
Not Determined
1.0
96.0-102
1.4-3.8
25
326.0
(DL)

Matrices:
Reagent water, high ionic strength water, high
organic content water
Bromate
0.17-1.2
Not Determined
1.0
92.9-110
2.0-11
1, 5, 10, 25
302.0
(DL)

Matrices:
Reagent water, synthetic sample matrix,
ground water, surface water
Bromate
0.12
0.18
1.0
89.8-104
0.84-2.6
0.5, 5
557
(DL)

Matrices:
Reagent water, synthetic sample matrix,
chlorinated ground water, chlorinated surface
water
Bromate
0.020
0.042
1.0
93.3-117
2.4-11
1, 2.5, 8, 10, 15
A review of the performance data in Exhibit 5.6 indicates that the more recently approved
analytical methods (EPA Methods 302.0 and 557) are comparable to the methods that were
approved under Stage 1 and Stage 2 in terms of recovery and RSD. However, based on the DL,
EPA Method 557 appears to be at least an order of magnitude more sensitive than the other
approved analytical methods. Thus, some improvement in method sensitivity might be expected
as a result of the approval of this method.
5.2.5 Unregulated DBPs
5.2.5.1 Unregulated Brominated HAAs
Because these contaminants are not currently regulated, there are no methods promulgated for
their analysis; however, the four unregulated brominated HAAs that augment HAA5 to HAA9,
(BCAA, bromodichloroacetic acid (BDCAA), dibromochloroacetic acid (DBCAA) and
tribromoacetic acid (TBAA)) can be quantified by the same methods as those used for HAA5
(with the exception of EPA Method 552.1, which does not include BDCAA, DBCAA, or
11 An MRL of 1.0 |ig/L must be achieved when using EPA Methods 317.0 (Rev. 2.0), 321.8 or 326.0. An MRL of
5.0 |ig/L must be achieved when using EPA Method 300.1.
Six-Year Review 3 Technical Support Document	5-14
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
TBAA). Exhibit 5.7 summarizes the DLs, LCMLRs, mean percent recoveries and percent RSDs
for the four unregulated HAAs as listed in EPA Methods 552.1, 552.2, 552.3 and 557.
Exhibit 5.7: Method Performance Metrics for EPA Methods 552.1, 552.2, 552.3 and
557 - Unregulated Brominated HAAs
Method/Analyte
DL/MDL
(M9/L)
LCMRL (|jg/L)
Mean % Recovery
Range
% RSD Range
Fortifi-
cation (|jg/L)
552.1
(MDL)
Matrices:
High ionic strength water, reagent water,
dechlorinated tap water, high humic ground water,
ozonated river water
BCAA
0.10
Not determined
85-114
0.7-16
5, 10
BDCAA
Not in method
Not in method
Not in method
Not in method
Not in method
DBCAA
Not in method
Not in method
Not in method
Not in method
Not in method
TBAA
Not in method
Not in method
Not in method
Not in method
Not in method
552.2
(MDL)
Matrices:
Reagent water, dechlorinated tap water, high ionic
strength water, high humic ground water,
BCAA
0.251
Not determined
82.5-108
2.1-9.3
1, 2, 4
BDCAA
0.091
Not determined
96.6-115
8.2-15
1, 2, 4
DBCAA
0.468
Not determined
103-114
4.0-13
2.5, 5, 10
TBAA
0.82
Not determined
96.7-126
7.6-14
5, 10, 20
552.3
(DL)
Matrices:
Reagent water, chlorinated surface water,
chlorinated ground water
BCAA
0.016-0.029
Not determined
99.5-106
0.36-3.8
1, 10
BDCAA
0.031-0.034
Not determined
87.5-117
1.1-6.1
1, 10
DBCAA
0.035-0.054
Not determined
94.4-125
1.5-8.8
1, 10
TBAA
0.097-0.11
Not determined
99.2-128
1.8-8.1
1, 10
557
(DL)
Matrices:
Reagent water, synthetic sample matrix,
chlorinated ground water, chlorinated surface water
BCAA
0.11
0.16
82.8-107
2.9-10
1, 2.5, 8, 10, 15
BDCAA
0.05
0.19
91.0-105
2.0-4.9
1, 2.5, 8, 10, 15
DBCAA
0.041
0.08
90.4-103
3.6-11
1, 2.5, 8, 10, 15
TBAA
0.067
0.27
94.0-103
1.9-5.4
1, 2.5, 8, 10, 15
At the time the Stage 1 D/DBPR was published, analytical standards for BCAA, BDCAA,
DBCAA and TBAA were not commercially available (Roberts et al., 2002).
More correctly, standards for the methyl esters of these four HAAs were not commercially
available. These methylated standards are important tools for EPA's assessment of the efficiency
of the derivatization of the various HAAs to their methyl esters, which would ideally be
conducted as part of method development. While the standards were not available at the time
EPA Method 552.2 was being developed, the standards were available during the development
of EPA Method 552.3. At that time, EPA found that under the conditions specified by EPA
Method 552.2, the methylation efficiencies for BCAA, BDCAA, DBCAA and TBAA were low.
EPA Method 552.3 uses tert-amyl methyl ether (TAME) as an alternate solvent to methyl tert-
butyl ether (MTBE). TAME has a higher boiling point than MTBE, the designated solvent in
Six-Year Review 3 Technical Support Document	5-15
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
EPA Method 552.2. The use of TAME as the extraction solvent results in more efficient
derivatization.12
As indicated in Section 5.2.2, EPA Method 557 does not include the derivatization step; hence
the issue of the efficiency of the conversion of the carboxylic acids to the corresponding methyl
esters is eliminated. However, the instrumentation used in EPA Method 557 is of much higher
cost than the instrumentation used for EPA Methods 552.1, 552.2 and 552.3, so using EPA
Method 557 to avoid issues with derivatization is not always feasible.
5.2.5.2 Nitrosamines
Because the nitrosamines are not currently regulated there are no methods promulgated for them.
Only one analytical method for drinking water, EPA Method 521, was approved for nitrosamine
monitoring under UCMR 2.
Exhibit 5.8 summarizes the DLs, LCMRLs, MRLs (these MRLs differ from the MRLs presented
earlier for other analytes; see footnote), mean percent recovery and percent RSDs for NDMA,
NDEA, NDPA, NDBA, NMEA and NPYR. The LCMRLs in Exhibit 5.8 are taken from EPA
Method 521 while the MRLs were developed by EPA for use in UCMR 2 (USEPA, 2007a).
Exhibit 5.8: Method Performance Metrics for Six Nitrosamines in EPA Method 521
Analyte
DL
(ng/L)
LCMRL
(ng/L)
MRL
(ng/L)13
Mean %
Recovery
Range14
% RSD Range
Fortification
(ng/L)
NDMA
0.28
1.6
2
83.7-94.7
3.8-12
2, 4, 10, 20
NDEA
0.26
2.1
5
84.6-95.6
6.5-14
2, 4, 10, 20
NDPA
0.32
1.2
7
77.1-97.0
3.7-10.2
2, 4, 10, 20
NDBA
0.36
1.4
4
79.7-104
2.9-16
2, 4, 10, 20
NMEA
0.28
1.5
3
81.4-91.0
4.5-9.6
2, 4, 10, 20
NPYR
0.35
1.4
2
85.2-102
4.0-12
2, 4, 10, 20
5.2.5.3 Chlorate
Because chlorate is not currently regulated there are no methods promulgated for it; however,
there are both EPA-developed methods and non-EPA-developed methods approved for related
analytes (e.g., bromide, chlorite and bromate) that are regulated in drinking water and several of
these methods can also be used for the analysis of chlorate. These include the approved EPA
12	E-mail correspondence with Dr. Glynda Smith of EPA's Technical Support Center, February 4, 2015; February 9,
2015; and February 12, 2015. Personal correspondence with Dr. Smith on February 11, 2015.
13	As determined statistically from LCMRL study data and used in UCMR 2 (USEPA, 2007a).
14	Percent recovery and percent RSD were obtained for the following matrices: reagent water, chlorinated drinking
water from a river, chlorinated drinking water from ground water and chlorinated drinking water from surface water
with high TOC.
Six-Year Review 3 Technical Support Document	5-16
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Methods 300.0 (Rev. 2.1), 300.1, 317.0 (Rev 2.0) and 326. Exhibit 5.9 summarizes the DLs,
LCMRLs, mean percent recoveries and percent RSDs for chlorate for the four approved
analytical methods that have been developed by EPA, ASTM D6581-08 which was approved for
chlorate analysis under UCMR 3 and SM 4110 D (note that the data shown are from the 22nd
edition of SM [APHA, AWWA and WEF, 2012], not the 21st, which is the approved edition for
chlorate analysis under UCMR 3). LCMRLs and the calculated MRL are only available for EPA
Method 300.1 since this was the method designated by EPA for use in UCMR 3 (USEPA,
2012b).
Exhibit 5.9: Method Performance Metrics for Chlorate Using EPA Methods 300.0
(Rev. 2.1), 300.1, 317.0 (Rev. 2.0), 326.0, ASTM D6581-08 and SM 4110 D
Method/Analyte
MDL/DL
(H9/L)
LCMRL1
(H9/L)
MRL2
(ng/L)
Mean %
Recovery Range
% RSD Range
Fortification
(ng/L)
300.0 (Rev. 2.1)
(MDL)

Matrices:
Reagent water, drinking water
Chlorate
3
N/A
N/A
97-121
N/A
0.05, 0.1, 1, 5
300.1
(MDL)

Matrices:
Reagent water, high ionic strength water, surface
water, ground water, chlorinated drinking water,
chlorine dioxide-treated drinking water, ozonated
drinking water
Chlorate
0.78-2.55
1.8-14
20
86.1-106
0.47-2.14
100, 500
317.0 (Rev. 2.0)
(MDL)

Matrices:
Reagent water, high ionic strength water, surface
water, ground water, chlorinated drinking water,
chlorine dioxide-treated drinking water, ozonated
drinking water
Chlorate
0.62-0.92
N/A
N/A
86.1-106
0.47-2.14
100, 500
326.0
(DL)

Matrices:
Reagent water, high ionic strength water, high
organic content water
Chlorate
1.7
N/A
N/A
99-111
0.66-2.8
100, 500
ASTM D6581-08
(MDL)
N/A
Matrices:
Reagent water, drinking water
Chlorate
0.32-3.49
N/A
N/A
93-107
N/A
20, 25, 180, 220,
400, 450
SM 4110 D (22nd
ed.)
(MDL)
N/A
Matrices:
Reagent water, high ionic strength water, surface
water, ground water, chlorinated drinking water,
chlorine dioxide-treated drinking water, ozonated
drinking water
Chlorate
2.55
N/A
N/A
86.1-106
0.47-2.14
100, 500
1 The LCMRLs are not from EPA Method 300.1 but were generated during UCMR 3 development and determination
of the MRL for chlorate using EPA Method 300.1.
2As determined statistically from LCMRL study data and used in UCMR 3 (USEPA, 2012b).
A review of the performance data in Exhibit 5.9 indicates that the methods are comparable in
terms of recovery and RSD; however, EPA Methods 300.1 and 317.0 (Rev. 2.0) may provide an
opportunity for better sensitivity relative to EPA Methods 300.0 (Rev. 2.1) and 326.0. Note that,
other than some differences in MDLs, the method performance data for EPA Methods 300.1,
317.0 (Rev. 2.0) and SM 4110 D (22nd ed.; APHA, AWWA and WEF, 2012) are identical.
Six-Year Review 3 Technical Support Document	5-17
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
5.3 Methods for Disinfectant Residuals
5.3.1	Chlorine (Free, Combined, Total) and Chloramines
For chlorine, only non-EPA-developed analytical methods were approved for monitoring in the
Stage 1 and Stage 2 D/DBPRs, and both EPA-developed and non-EPA-developed analytical
methods have been approved via EPA's Expedited Method Approval process since the final
Stage 2 D/DBPR was published.
For the Stage 1 D/DBPR, SM 4500-C1 D (19th ed.), SM 4500-C1 F (19th ed.) and SM 4500-C1 G
(19th ed.), ASTM D1253-86 (free, combined, total); SM 4500-C1 E (19th ed.) and SM 4500-C11
(19th ed.) (total); and SM 4500-C1 H (19th ed.) (free) were approved for compliance monitoring.
These same methods remained approved under the Stage 2 D/DBPR. Also under Stage 2, 20th
edition versions of the previously cited SMs; ASTM D1253-86(96), ASTM D1253-03, SM
online 4500-C1 D-00, SM online 4500-C1 F-00 and 4500-C1 G-00 (free, combined, total); SM
online 4500-C1 E-00 and SM online 4500-C11-00 (total); and SM online 4500-C1 H-00 (free)
were also approved for compliance monitoring. Since publication of the Stage 2 D/DBPR,
alternate testing methods for chlorine that have been approved via EPA's Expedited Method
Approval process include Hach Method 10260, ASTM D1253-08, SM 4500-C1 D (21st, 22nd ed.),
SM 4500-C1 F (21st, 22nd ed.) and SM 4500-C1 G (21st, 22nd ed.) (free, combined, total); EPA
Method 334.0, ChloroSense, SM 4500-C1 E (21st, 22nd ed.) and SM 4500-C11 (21st, 22nd ed.)
(total); EPA Method 334.0, ChloroSense, SM 4500-C1 H (21st, 22nd ed.) and Method D99-003 (if
approved by state) (free). Since only one EPA method (334.0) has been approved, method
performance data are not included in SM 4500-C1 D, F or G (APHA, AWWA and WEF, 2012)
and the other non-EPA methods are not available online, no comparison of methods approved in
the Stage 1 and Stage 2 D/DBPRs relative to those published since Stage 2 can readily be made.
Additional information about the analytical methods used for measuring the free and total
chlorine residuals in distribution system samples is provided in the Six-Year 3 Review Technical
Support Document for Microbial Contaminant Regulations (USEPA, 2016a). Within the
microbial rules, there is a requirement to maintain a detectable concentration of residual in the
distribution system, while the D/DBPR includes MRDLs for chlorine and chloramines. There
may be additional benefits to providing limits for monochloramine (see, e.g., 59 FR 38683,
USEPA 1994a) and analytical methods have been developed that may be able to accommodate
such measurements (e.g., an indophenol method that has been shown to be specific for
monochloramine). There may be opportunities to consider approaches for realizing these
additional benefits that include options that would allow utilities to use either the current
methods for free and total chlorine or the newer methods (Wahman and Pressman, 2015).
5.3.2	Chlorine Dioxide
For chlorine dioxide, both EPA-developed and non-EPA-developed analytical methods were
approved for monitoring during the Stage 1 and Stage 2 D/DBPRs and non-EPA-developed
analytical methods have been approved via EPA's Expedited Method Approval process since the
final Stage 2 D/DBPR was published.
Six-Year Review 3 Technical Support Document	5-18
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
For the Stage 1 D/DBPR, SM 4500-C102 D (19111 ed.) and SM 4500-C102 E (19111 ed.) were
approved for compliance monitoring. These same methods remained approved under the Stage 2
D/DBPR, wherein EPA Method 327.0 (Rev. 1.1), SM 4500-C102 D (20th ed.), SM 4500-C102 E
(20th ed.) and SM online 4500-C102 E-00 were also approved for compliance monitoring. Since
publication of the Stage 2 D/DBPR, alternate testing methods that have been approved via EPA's
Expedited Method Approval process for chlorine dioxide have included SM 4500-C102 E (21st,
22nd ed.) and ChlordioX Plus. Since only one EPA method has been approved, method
performance data are not included in SM 4500-C102 E (APHA, AWWA and WEF, 2012) and
the ChlordioX Plus method is not available online, no comparison of methods approved under
Stage 1 and Stage 2 vs. those published since Stage 2 can readily be made.
Six-Year Review 3 Technical Support Document	5-19
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
6 Occurrence and Exposure
This chapter summarizes information relevant to occurrence and exposure to regulated and
unregulated disinfection byproducts (DBPs). As with other aspects of the Third Six-Year Review
(SYR3), EPA limited its review of occurrence and exposure to information published through
2015. Information published since that time, while informative, was not included in this review.
Section 6.1 provides information related to DBP formation, including what was known at the
time of the Stage 2 Disinfectants and Disinfection Byproducts Rule (D/DBPR) and important
findings in the literature since the rule was promulgated.
Section 6.2 provides historical and new information related to occurrence of DBP precursors in
source water.
Section 6.3 presents historical and new information related to the occurrence of regulated and
unregulated DBPs in drinking water.
EPA used multiple data sources to evaluate occurrence and exposure to regulated and
unregulated DBPs. The SYR3 Information Collection Rule (ICR) Dataset (USEPA, 2016f),
called the "SYR3 ICR dataset," houses public water system (PWS) compliance monitoring data
collected between 2006 and 2011 for systems of all sizes. This dataset contains over 47 million
records for DBP, microbial, chemical and radiological monitoring data, with over 13 million
records passing QA/QC procedures for DBPs and microbial contaminants and indicators
(USEPA, 2016f). The SYR3 ICR dataset is regarded as the largest and most comprehensive
source of PWS compliance monitoring dataset ever compiled and analyzed by EPA's Drinking
Water Program. The SYR3 ICR dataset and general QA/QC procedures are further described in
the Analysis of Regulated Contaminant Occurrence Data from Public Water Systems in Support
of the Third Six-Year Review of Existing National Primary Drinking Water Regulations:
Chemical Phase Rules and Radionuclides (USEPA, 2016g) and The Data Management and
Quality Assurance/Quality Control Process for the Third Six-Year Review Information
Collection Rule Dataset (USEPA, 2016i).
In addition to the SYR3 ICR data, information from the DBP ICR dataset (USEPA, 2000e;
McGuire et al., 2002) was also further analyzed for understanding changes of DBP occurrence
and disinfection practices. The DBP ICR dataset was the main source of occurrence data for
development of supporting the Stage 2 D/DBPR and houses monitoring data from large public
water systems (PWSs serving a population greater than or equal to 100,000) from an 18-month
period (July 1997 to December 1998). Monitoring data for DBPs, plant treatment, source water
characteristics and disinfectant type are available within this dataset.
Appendix B provides additional information on several of the topics presented in this chapter.
The additional information presented in the appendix addresses DBP formation; precursor
occurrence analytical results based on multiple other data sources such as National Rural Water
Association (NRWA), ICR Supplemental Survey and Waterstats; discussion of EPA's Surface
Six-Year Review 3 Technical Support Document	6-1
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Water Analytical Tool (SWAT) developed to predict formation of THM415 and HAA5; detailed
discussions of the applicability of the DBP ICR and SYR3 ICR datasets; and detailed
descriptions of the QA/QC processes undertaken prior to the analysis of the SYR3 ICR data.
6.1 DBP Formation
New research since the promulgation of the Stage 2 D/DBPR has enhanced our understanding of
the key factors affecting DBP formation. This section briefly describes what was known at the
time of the Stage 2 D/DBPR and presents new information on DBP formation that is relevant to
the SYR process.
6.1.1 Summary of Stage 1 and 2 D/DBPR Information
The Stage 2 D/DBPR support documents including the Occurrence Document (USEPA, 20051),
the Economic Analysis (USEPA, 2005g) and the Technology and Costs Document (USEPA,
2005m) summarize what was known at the time regarding DBP formation.
The DBPs regulated by the Stage 2 D/DBPR include total trihalomethanes (THM4) and five
haloacetic acids (HAA5). THM4 includes all four regulated trihalomethanes (THMs):
chloroform, bromoform, bromodichloromethane (BDCM) and dibromochloromethane (DBCM).
HAA5 includes five haloacetic acids for which an adequate analytical method existed at the time
of the Stage 2 D/DBPR: monochloroacetic acid (MCAA), dichloroacetic acid (DCAA),
trichloroacetic acid (TCAA), monobromoacetic acid (MBAA) and dibromoacetic acid (DBAA).
Other groups of DBPs may also be referred to throughout this document. THM3 refers to THM4
minus chloroform. HAA6 includes all the haloacetic acids (HAAs) included in HAA5 and adds
bromochloroacetic acid (BCAA).16 HAA9 includes all nine HAAs, adding BDCAA, DBCAA
and TBAA to those included in HAA6.
Organic DBPs form by the reaction of organic matter and disinfectants (acting as oxidizing
agents) added during drinking water treatment. The Stage 2 D/DBPR Occurrence Document
(USEPA, 20051) identifies the following major factors affecting organic DBP formation:
disinfection method and dose, contact time, concentration and characteristics of precursors,
temperature and water chemistry. Information available at the time of the Stage 2 D/DBPR
showed that a variety of DBPs formed from chlorine and natural organic matter (NOM)
reactions, however, the amount of regulated organic DBP formation tended to be less upon
chloramine disinfection. Research also demonstrated that NOM containing high aromatic content
tended to increase DBP levels. Furthermore, DBP levels were found to increase with longer
disinfectant contact time and higher temperatures. Upon the understanding of various factors, a
water treatment plant model was developed to predict THM4/HAA5 levels at a national level,
15	THM4 (also referred to as TTHM) is used to recognize the regulated THMs (THM4) vs other THMs (such as
iodinated) which could be a part of the total THM mixture.
16	EPA notes that in the Fourth Unregulated Contaminant Monitoring Rule proposal (USEPA, 2015) HAA6Br
includes BCAA, bromodichloroacetic acid (BDCAA), DBAA, dibromochloroacetic acid (DBCAA), MBAA and
tribromoacetic acid (TBAA). However, for the purposes of this document, HAA6 includes the regulated species
within HAA5 and adds BCAA.
Six-Year Review 3 Technical Support Document	6-2
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
given treatment and water conditions. This model was incorporated into the SWAT, which was
used to support economic analysis for the development of the Stage 2 D/DBPR (USEPA, 2005g).
6.1.2 New Information since the Stage 2 D/DBPR
Since the promulgation of the Stage 2 D/DBPR in 2006, considerable research has been done to
better understand the formation of DBPs in drinking water, including regulated and unregulated
DBPs. Research has further examined the impacts of factors such as source water quality,
disinfection practices, treatment operations and distribution system operation and management.
There have also been advances in the development of models to predict DBP formation.
6.1.2.1 DBP Types
Numerous studies have evaluated the occurrence of different types of DBPs in PWSs, many of
which are unregulated and form after use of disinfectants other than chlorine. Over 600 different
DBPs have been identified as forming during the disinfection process and many other DBPs are
still unknown (Richardson et al., 2007; Krasner, 2009; Richardson and Postigo, 2011). Exhibit
6.1 summarizes general information about groups of regulated and unregulated DBPs, including
examples of specific DBPs within the groups, the levels at which the DBPs occur and the source
water and disinfection conditions that lead to DBP formation. These items are discussed further
in the following sections. Information is also discussed in Section 6.3 about the occurrence of
DBPs. As discussed in Section 6.1.1, EPA has proposed to monitor more brominated acetic acids
along with the precursor or precursor indicator (i.e., TOC and bromide) under the Fourth
Unregulated Contaminants Monitoring Rule (UCMR 4) (USEPA, 2015).
Exhibit 6.1: Regulated and Unregulated DBPs - General Information
DBP Group
Examples of DBPs
Relative Occurrence1
Disinfection Conditions Associated
with Formation
Regulated
trihalomethanes
Chloroform
Bromoform
Bromodichloromethane
Dibromochloromethane
Chloroform occurs at low
to mid |jg/L levels;
additional three species
occur at low |jg/L levels
(Richardson et al., 2007).
Formed by disinfection with chlorine or
chloramines. Formation tends to be less
upon chloramine disinfection.
Bromoform can also be formed in high-
bromide source waters treated with
ozone. Disinfection with chlorine dioxide
does not result in THMs; however, low
THM levels can be present due to
chlorine impurities in chlorine dioxide
(Richardson et al., 2007; Richardson
and Postigo, 2011).
Regulated
haloacetic acids
Chloroacetic acid
Bromoacetic acid
Dichloroacetic acid
Dibromoacetic acid
Trichloroacetic acid
Chloroacetic and
bromoacetic acids occur
at sub- to low |jg/L levels;
dichloroacetic,
dibromoacetic and
trichloroacetic acids
occur at low to mid- |jg/L
levels (Richardson et al.,
2007).
Formed by disinfection with chlorine,
chloramines, chlorine dioxide and
ozone, although generally formed at
highest levels upon chlorination.
Dibromoacetic acid can form when
source water contains elevated bromide
(Glaze et al., 1993; Richardson et al.,
2007; Richardson and Postigo, 2011).
Additional haloacetic
acids
Tribromoacetic acid
Bromochloroacetic acid
Bromodichloroacetic acid
Low |jg/L levels
(Obolensky, 2002;
Richardson et al., 2007).
Associated with high-bromide source
waters (Obolensky, 2002; Singer,
2006).
Six-Year Review 3 Technical Support Document	6-3
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
DBP Group
Examples of DBPs
Relative Occurrence1
Disinfection Conditions Associated
with Formation

Dibromochloroacetic acid


lodinated
trihalomethanes
Dichloroiodomethane
Bromochloroiodomethane
Dibromoiodomethane
Chlorodiiodomethane
Bromodiiodomethane
Iodoform
Sub to low |jg/L levels
(Richardson etal., 2007;
Krasner, 2009).
Can form in drinking water treated with
chlorine or chloramines when iodide is
present in source waters, however,
formation is highest when chloramines
are used with ammonia added before
chlorine (Richardson, 2003; Richardson
etal., 2007; Krasner, 2009).
lodoacids
Monoiodoacetic acid
Chloroiodoacetic acid
Bromoiodoacetic acid
Diiodoacetic acid
(E)-3-bromo-3-
iodopropenoic acid
(Z)-3-bromo-3-
iodopropenoic acid
(E)-2-iodo-3-
methylbutenedioic acid
ng/L to low |jg/L levels
(Richardson etal., 2007;
Kritsch and Weinberg,
2010).
Formed when hypoiodous acid (a result
of an iodide and oxidant reaction) reacts
with TOC. Presence of strong oxidants
such as chlorine or ozone may further
oxidize hypoiodous acid to iodate which
does not form DBPs. Weaker oxidants
like chloramines allow the hypoiodous
acid to react with TOC to form iodoacids
(Kritsch and Weinberg, 2010).
Haloacetonitriles
(HAN)
Dichloroacetonitrile
Bromochloroacetonitrile
Dibromoacetonitrile
Trichloroacetonitrile
Tribromoacetonitrile
Sub to low |jg/L levels
(Richardson etal., 2007).
Formed by treatment with chlorine,
chloramines, chlorine dioxide and
ozone; highest formation observed in
chloraminating plants (Blank et al.,
2002; Richardson et al., 2007).
Haloketones (HK)
1-bromo-1,3,3-
trichloropropanone
1-bromo-1,1-
dichloropropanone
Low |jg/L levels (Krasner
etal., 2006).
Formed by treatment with chlorine,
chloramine, chlorine dioxide and ozone
combined with either chlorine or
chloramine (Richardson and Postigo,
2011).
Halonitromethanes
(HNMs)
Chloropicrin
Bromopicrin,
Bromodichloronitromethane
Dibromochloronitromethane
Sub to low |jg/L levels
(Richardson etal., 2007)
Some compounds in this group may be
associated with use of ozone, chlorine
dioxide or UV usage (Richardson et al.,
2007; Bull et al., 2011). Formation can
be influenced by wastewater effluents
and algal blooms (Krasner, 2009; Bull et
al., 2011).
Haloacetamides
Dichloroacetamide
Dibromoacetamide
Trichloroacetamide
Sub to low |jg/L levels
(Richardson etal., 2007).
Formation associated with use of
chlorine or chloramines. There is
preliminary indication that formation
may be higher upon chloramination
(Weinberg et al., 2002; Krasner et al.,
2006; Richardson et al., 2007).
Haloacetoaldehydes
T richloroacetaldehyde
(chloral hydrate)
Dichloroacetaldehyde
Low |jg/L levels for
trichloroacetaldehyde;
sub to low |jg/L levels for
dichloroacetaldehyde
(Richardson etal., 2007;
Jeong et al., 2015).
Associated with use of chlorine,
chloramines and ozone (Krasner, 2009).
Cyanogen halides
(CNX)
Cyanogen chloride
Low ug/L levels (Bull et
al., 2011).
Formation is linked to chloramine use
(Bull etal., 2011).
Six-Year Review 3 Technical Support Document	6-4
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
DBP Group
Examples of DBPs
Relative Occurrence1
Disinfection Conditions Associated
with Formation
Oxyhalides2
Chlorate
Chlorite
Bromate
Sub to low |jg/L levels for
bromate; high |jg/L levels
for chlorite and chlorate
(Richardson etal., 2007).
Chlorate and chlorite associated with
use of chlorine dioxide or hypochlorite;
bromate is primarily a byproduct of
ozone disinfection, although some
studies have shown bromate formation
following chlorine dioxide treatment
(Richardson etal., 2007; USEPA,
2016e).
Halogenated
furanones
3-chloro-4-(dichloromethyl)-
5-hydroxy-2[5H]furanone
(MX)
Brominated MX analogs
ng/L to sub |jg/L levels
(Richardson etal., 2007)
can reach |jg/L levels
with high THM and
bromide (Krasner et al.,
2006).
Associated with use of chlorine,
chloramines and chlorine dioxide.
Halobenzoquinones
(HBQ)
2,6-dichlorobenzoquinone
2,6-dibromobenzoquinone
Sub ug/L levels (Bull.,
2012).
Formed more in the presence of
chloramines than chlorine (Bull et al.,
2009).
Nitrosamines3
/V-Nitrosodiethylamine
/V-Nitrosodimethylamine
/V-Nitrosodi-n-propylamine
A/-Nitrosopyrrolidine
/V-Nitrosodi-n-butylamine
/V-Nitrosomethylethylamine
/V-Nitrosomorpholine
A/-Nitrosopiperidine
/V-Nitrosodiphenylamine
ng/L to sub |jg/L levels
for NDMA; low ng/L
levels for other
compounds (Richardson
et al., 2007; Boyd et al.,
2011).
Shown to increase with chloramine use
and nitrogenous precursors such as
wastewater, pharmaceutical and
personal care products and drinking
water treatment chemicals (Richardson
et al., 2007; Krasner, 2009; USEPA,
2016d).
Halogenated
pyrroles
Tribromopyrrole
ng/L level (Richardson et
al., 2007)
Information not available.
Aldehydes
Formaldehyde
Acetaldehyde
Glyoxal
Methyl glyoxal
Sub to low |jg/L levels
(Richardson etal., 2007).
Mostly found with ozone use but also to
a lesser extent with chlorine dioxide
(Richardson etal., 2007; Richardson
and Postigo, 2011).
1	Section 6.3 provides additional information from the SYR3 ICR dataset about regulated DBPs.
2	For additional information on chlorate, see USEPA2016e.
3	For additional information on nitrosamines, see USEPA2016d.
6.1.2.2 Disinfection Practices
Unlike most chemical contaminants, DBPs form during treatment (i.e., disinfection or
maintenance of disinfectant residual levels). As mentioned earlier, disinfection practices
(including disinfectant types, doses and residual levels) can influence the type of DBPs that
form, as well as the concentrations at which they occur. This section summarizes information
available on the types of DBPs that have been found to occur upon different disinfection
conditions, as well as presents information on disinfectant usage at PWSs.
Six-Year Review 3 Technical Support Document	6-5
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Disinfectant Types and Doses
New research has shown that increased chlorine doses lead to increased DBP formation,
although the effect is not uniform across all DBPs. Hua and Reckhow (2008) found that higher
chlorine doses led to more trihaloacetic acids (THAA) than dihaloacetic acids (DHAA). Liu and
Reckhow (2013) found higher chlorine doses led to more chloroform and DCAA in simulated
distribution systems.
Studies have continued to show that chloramines produce less regulated THMs and HAAs than
free chlorine (Bougeard et al., 2010; Lee et al., 2007; Hua and Reckhow, 2007; Cimetiere et al.,
2010; Tian et al., 2013). Chloramines in particular produce less THM and HAA formation, but
are not as effective at producing less DHAA (Krasner, 2014). One study found that using
preformed chloramines (when ammonia is added before chlorine) produced between 7 and 18
percent of the total organic halides (TOX) that were produced by chlorination of the same water
(Reckhow et al., 2007).
Chloramines can react with nitrogenous organic compounds to form nitrosamines, as described
in the Six-Year Review 3 Technical Support Document for Nitrosamines (USEPA, 2016d). While
the formation of nitrosamines has been found to be more common with the use of chloramines,
nitrosamine formation also can occur with other disinfectants given the proper precursors.
Chloramines can also react with nitrogen-containing organic compounds to form additional
unregulated DBPs including cyanogen chloride, dichloroacetonitrile (DCAN), dichloroacetamide
and chloropicrin (Yang et al., 2010; Huang et al., 2012; Kimura et al., 2013).
New research has shown that chloramines can produce a number of unregulated DBPs, including
brominated and iodinated DBPs (Kritsch and Weinberg, 2010; Zhai et al., 2014; Richardson and
Ternes, 2014). The formation of iodinated DBPs with chloramines was found to be greater than
that with chlorine (Hua and Reckhow, 2007; Kristiana et al., 2009; Criquet et al., 2012; Jones et
al., 2012).
Recent research has verified that ozone reactions can create smaller oxygenated molecules such
as aldoketoacids, carboxylic acids and aldehydes, which can impact DBP formation when
chlorine or chloramines are used downstream (Krasner, 2014). While ozone generally forms
fewer DBP species than chlorine alone, ozone has been found to increase THM and HAA levels
compared to chlorine when some precursors are present. Bromate is of concern for systems that
use ozone with elevated bromide levels in their source water. Use of ozone to oxidize
cyanobacterial algae, followed by chlorination, was found to increase THM and HAA
concentrations compared to chlorination alone (Coral et al., 2013). DBP concentrations increased
with increased ozone dose and were correlated with release of dissolved extracellular organic
material on ozonation of the cyanobacteria.
Ozone has also been found to lead to the formation of unregulated DBPs. Researchers found that
following ozonation, levels of chloropicrin, trichloroacetaldehyde (TCAL), chloral hydrate,
cyanogen bromide, halonitromethanes (HNMs), haloacetonitriles (HANs) and haloketones (HKs)
can be elevated (Shah et al., 2012; Yang et al., 2012; Krasner, 2014; Richardson and Ternes,
2014; Xieetal., 2013).
Six-Year Review 3 Technical Support Document	6-6
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Studies have found that UV disinfection produces very few THMs or HAAs at doses typically
used for disinfection (Reckhow et al., 2010; Lyon et al., 2012; Linden et al., 2012). UV has been
linked to increases in unregulated DBPs such as chloropicrin, HNM, TCAL, chloral hydrate and
cyanogen chloride (Shah et al., 2012; Lyon et al., 2014; Krasner, 2014). Medium pressure lamps
with various chlorination strategies were found to increase chloropicrin and bromopicrin
formation more than low pressure lamps, with increases between 20 and 50 percent more than
chlorine or chloramine alone (Linden et al., 2012).
Research has continued to show that chlorite and chlorate can be present as impurities from
chlorine dioxide generation (as well as from decomposition of chlorine dioxide). Chlorite and
chlorate have been found to co-occur in hypochlorite solutions. Chlorate may be an impurity in
hypochlorite but can also be formed by the disproportionation of hypochlorite into chlorate and
chlorite. Longer storage times, higher concentrations and higher temperatures have been found to
increase chlorate concentrations in hypochlorite stock solutions (USEPA, 2016e). For more
information about the occurrence of chlorate, refer to the Six-Year Review 3 Technical Support
Document for Chlorate (USEPA, 2016e). For more information about the co-occurrence of
chlorate and chlorite, refer to Section 6.3.4 of this document.
Researchers have found that regulated DBPs may be present as impurities in disinfectants.
Emmert et al. (2011) investigated hypochlorite stock solutions at five utilities. Four of those
utilities were found to have HAAs in their hypochlorite stock solutions; THMs were also present
but in much lower concentrations. Concentrations of HAAs ranged from 56 to 627 |ig/L. When
added to water during the treatment process, the solution with the highest HAA level was enough
to be associated with a concentration of 30 |ig/L, or half the maximum contaminant level (MCL),
in the finished water. A follow-up study (Emmert et al., 2013) found HAAs in all of 30 bulk
hypochlorite samples examined but did not find THMs. The concentrations of HAAs were
enough to be associated with concentrations of 4.1 to 16.4 [j,g/L in finished water and were
higher in warmer months.
Disinfectant Usage Trends
As discussed earlier, disinfection practices can be a factor in the types of DBPs formed as well as
the levels to which they occur. Characterization of the type(s) of disinfectant(s) used and their
changes over time can be helpful for understanding the national occurrence of various DBPs and
associated disinfection practices.
The Disinfection Systems Committee under the American Water Works Association (AWW A)
has been conducting periodical national surveys (approximately every 10 years) to collect
information on disinfection practices. Their most recent survey was published in 2008 and
provides insight into disinfectant usage trends. The survey found that between 1998 and 2007,
there was a tendency for utilities to switch from using chlorine gas to hypochlorite because of
safety concerns. Chloramine usage also rose during that time period from 11 to 30 percent of all
plants surveyed, although results should be viewed with caution because there were many more
small plants included in the 1998 survey than the 2007 survey. Advanced disinfectants such as
ozone, chlorine dioxide and UV light were also found to be increasing in usage, with ozone use
rising from 6 to 9 percent, chlorine dioxide use increasing from 4 to 8 percent and UV use from 0
to 2 percent from 1998 to 2007 (AWW A, 2008). As discussed below, these trends continued
Six-Year Review 3 Technical Support Document	6-7
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
after 2007 as systems continued to comply with the Stage 1 D/DBPR/IESWTR and the more
recent Stage 2 D/DBPR/LT2ESWTR.
The Third Unregulated Contaminant Monitoring Rule (UCMR 3) dataset provides a
comprehensive set of information about disinfectant usage in the United States (USEPA, 2016h).
Spanning the period from January 2013 to December 2015,17 data in the dataset are nationally
distributed and demonstrate that systems reporting exclusive use of chloramines, as well as
systems that reported using multiple disinfectants (e.g., a system reported chlorine usage in 1
month of the UCMR and ozone in a different month), make up a significant portion of the
reporting. Under the UCMR3, disinfectant type is identified for specific monitoring locations
(Entry Point (EP) or Maximum Residence (MR)) rather than at the system level. The disinfectant
type for a given monitoring period was specific to that monitoring location rather than to the
system as a whole. As such, inferences about system-level disinfectant usage may tend to
overestimate use of a type of disinfectant in situations where that disinfectant was used only for a
portion of the UCMR monitoring program. The Six-Year Review 3 Technical Support Document
for Chlorate (USEPA, 2016e) provides further information on disinfectant usage evaluations
using the UCMR 3 dataset.
The following 11 disinfectant designation codes are used in the UCMR 3 dataset:
•	CLGA (gaseous chlorine),
•	CLOF (off-site generated hypochlorite stored as liquid),
•	CLON (on-site generated hypochlorite with no storage),
•	CAGC (chloramine formed from gaseous chlorine),
•	CAOF (chloramine formed from off-site hypochlorite),
•	CAON (chloramine formed from on-site hypochlorite),
•	CLDO (chlorine dioxide),
•	OZON (ozone),
•	ULVL (ultraviolet light),
•	OTHD (all other types of disinfectant), and
•	NODU (no disinfection).
Exhibit 6.2 and Exhibit 6.3 show information about the disinfection types reported in the
UCMR 3 dataset (as of July 2016) at the EP and MR distribution system locations, respectively.
The results are split by system size and source water type. In both EP and MR locations, more
than 30 percent of very large surface water systems (serving >100,000 people) use only
chloramines or "chlorine and chloramines," while approximately 50 to 54 percent of very large
surface water systems (serving >100,000 people) use chloramines alone or with another
disinfectant (i.e., chlorine, chloramines, ozone, chlorine dioxide, UV light and "other
disinfectant").
Exhibit 6.4 compares the disinfectant usage data from the DBP ICR and UCMR 3 datasets. A
total of 199 systems reported disinfection data in both surveys (i.e., "common systems"). In the
17 Monitoring was scheduled to occur between 2013 and 2015. Most data were received by EPA during the three-
year-long official monitoring period although the reporting of some data continued in 2016.
Six-Year Review 3 Technical Support Document	6-8
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
DBP ICR, data from 262 surface water plants were reported from these 199 systems. These
results were compared with data from the 342 EP locations and 238 MR locations associated
with surface water plants at the 199 systems in the UCMR 3 dataset (as of July 2016). The
results show an increase over that time period in the use of chlorine dioxide, ozone, UV and
chloramines. Note that although the data were from the same "common systems," the sampling
point locations in DBP ICR and UCMR 3 may not have been the same.
As discussed in USEPA (2016e) and Chapter 7 of this document, the information from multiple
datasets (i.e., DBP ICR, UCMR 2 and UCMR 3) collectively indicate that:
•	The use of disinfectants other than free chlorine (i.e., ozone, chlorine dioxide,
chloramines and UV) in treatment plants has increased over time.
•	In distribution systems, the use of chloramines has increased over time.
•	The use of hypochlorite in lieu of chlorine gas has increased over time.
These trends are important relative to the information about DBP formation and health effects.
Chapter 7 presents further discussion of potential implications with these trends, from the
treatment perspective.
Six-Year Review 3 Technical Support Document	6-9
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Exhibit 6.2: Use of Disinfectants by Source Water Type and System Size for UCMR 3 Data in EPs (select
categories)
Sampling
Location
Source
Water1
System
Size
(population
served)2
Number
of EPs
Count of EPs
Indicating
Exclusive
Use of
Chlorine
(% of Total)
m
Count of EPs
Indicating Exclusive
Use of Chloramines,
OR
both Chlorine and
Chloramines
(% of Total)
Count of EPs
Indicating Any
Instance of
Using Chlorine
(% of Total)
Count of EPs
Indicating Any
Instance of
Using
Chloramines
Count of EPs
Indicating Any
Instance of
Using Ozone
(% of Total)
Count of EPs
Indicating Any
Instance of
Using Chlorine
Dioxide
(% of Total)
Count of EPs
Indicating Any
Instance of
Using UV Light
(% of Total)
¦
Count of EPs
Indicating
Any Instance
of Using
"Other
Disinfectant"
(% of Total)
Count of EPs
Indicating
"No
Disinfectant
Used"
(% of Total)3
GW
<10,000
992
690
(69.6%)
90
(9.1%)
803
(80.9%)
108
(10.9%)
(0.1%)
3
(0.3%)
5
(0.5%)
34
(3.4%)
10,001 -
100,000
6,590
5,244
(79.6%)
602
(9.1%)
5,419
(82.2%)
620
(9.4%)
16
(0.2%)
37
(0.6%)
13
(0.2%)
97
(1.5%)
>100,000
2,256
1,947
(86.3%)
204
(9.0%)
2,017
(89.4%)
228
(10.1%)
28
(1.2%)
8
(0.4%)
2
(0.1%)
10
(0.4%)
SW
<10,000
293
155
(52.9%)
75
(25.6%)
256
(87.4%)
101
(34.5%)
19
(6.5%)
33
(11.3%)
12
(4.1%)
8
(2.7%)
10,001 -
100,000
2,257
1,240
(54.9%)
591
(26.2%)
1,594
(70.6%)
742
(32.9%)
130
(5.8%)
180
(8.0%)
92
(4.1%)
24
(1.1%)
>100,000
629
253
(40.2%)
213
(33.9%)
397
(63.1%)
317
(50.4%)
86
(13.7%)
53
(8.4%)
30
(4.8%)
5
(0.8%)
1	The source water type of the sampling location ("FacilityWaterType" in the UCMR 3 dataset) was used to develop these counts. Note: The "SW" category
includes ground water under direct influence of surface water ("GU") and mixed ("MX").
2	The population served by each system reflects the population served at the time of the UCMR 3 sample design. Refer to USEPA 2016e for full detail on UCMR3
data.
3	The counts in the "no disinfectant used" column includes only those EPs that always specified "no disinfectant used." Furthermore, any surface water facilities
identified as using no disinfection (NODU) in UCMR 3 may be a data entry error, as all surface water systems must disinfect.
Note: Based on EP locations with data posted from July 2016.
The disinfection codes used to categorize each sampling point are provided graphically in the table header above
each column. The legend to the right indicates what code or set of codes corresponds to each cell. Fully shaded cells
show codes that must be present for a sampling point to be assigned to a category and striped cells show codes that
may be present. Blank cells show codes that must not be present. Because the categories shown in this table are
neither exhaustive nor mutually exclusive, results do not add up to totals.
Layout Key
CLGA
and/or
CLOF
CAGC
and/or
CAOF

OZON
OTHD

CLDO
NODU
and/or
CLON
and/or
CAON

UVLV

Color Key

Used

May be used

Not used
Six-Year Review 3 Technical Support Document
for Disinfectants/Disinfection Byproducts Rules
6-10
December 2016

-------
Exhibit 6.3: Use of Disinfectants by Source Water Type and System Size for UCMR 3 Data in MRs (select
categories)
Sampling
Location
Source
Water1
System Size
(population
served)2
Number
of MRs
Count of EPs
Indicating
Exclusive
Use of
Chlorine
(% of Total)
Iff
Count of EPs
Indicating Exclusive
Use of Chloramines,
OR
both Chlorine and
Chloramines
(% of Total)
Count of EPs
Indicating Any
Instance of
Using Chlorine
(% of Total)
Count of EPs
Indicating Any
Instance of
Using
Chloramines
Count of EPs
Indicating Any
Instance of
Using Ozone
(% of Total)
Count of EPs
Indicating Any
Instance of
Using Chlorine
Dioxide
(% of Total)
Count of EPs
Indicating Any
Instance of
Using UV Light
(% of Total)
Count of EPs
Indicating Any
Instance of
Using "Other
Disinfectant"
(% of Total)
Count of EPs
Indicating "No
Disinfectant
Used"
(% of Total)3
GW
<10,000
710
535
(75.4%)
67
(9.4%)
596
(83.9%)
69
(9.7%)
1
(0.1%)
4
(0.6%)
5
(0.7%)
20
(2.8%)
10,001 -
100,000
3,813
3,031
(79.5%)
435
(11.4%)
3,198
(83.9%)
450
(11.8%)
27
(0.7%)
34
(0.9%)
14
(0.4%)
50
(1.3%)
>100,000
697
555
(79.6%)
113
(16.2%)
596
(85.5%)
120
(17.2%)
13
(1.9%)
0
(0%)
1
(0.1%)
2
(0.3%)
SW
<10,000
285
153
(53.7%)
74
(26.0%)
250
(87.7%)
99
(34.7%)
16
(5.6%)
30
(10.5%)
11
(3.9%)
9
(3.2%)
10,001 -
100,000
2,176
1,163
(53.4%)
604
(27.8%)
1,513
(69.5%)
750
(34.5%)
128
(5.9%)
171
(7.9%)
91
(4.2%)
28
(1.3%)
>100,000
591
189
(32.0%)
218
(36.9%)
354
(59.9%)
322
(54.5%)
85
(14.4%)
68
(11.5%)
34
(5.8%)
5
(0.8%)
1	The source water type of the sampling location ("FacilityWaterType" in the UCMR 3 dataset) was used to develop these counts. Note: The "SW" category
includes ground water under direct influence of surface water ("GU") and mixed ("MX").
2	The population served by each system reflects the population served at the time of the UCMR 3 sample design. Refer to USEPA 2016e for full detail on UCMR3
data.
3	The counts in the "no disinfectant used" column includes only those MRs that always specified "no disinfectant used." Furthermore, any surface water facilities
identified as using no disinfection (NODU) in UCMR 3 may be a data entry error, as all surface water systems must disinfect.
Note: Based on MR locations with data posted from July 2016.
The disinfection codes used to categorize each sampling point are provided graphically in the table header above
each column. The legend to the right indicates what code or set of codes corresponds to each cell. Fully shaded cells
show codes that must be present for a sampling point to be assigned to a category and striped cells show codes that
may be present. Blank cells show codes that must not be present. Because the categories shown in this table are
neither exhaustive nor mutually exclusive, results do not add up to totals.
Layout Key
CLGA
and/or
CLOF
CAGC
and/or
CAOF

OZON
OTHD

CLDO
NODU
and/or
CLON
and/or
CAON

UVLV

Color Key

Used

May be used

Not used
Six-Year Review 3 Technical Support Document
for Disinfectants/Disinfection Byproducts Rules
6-11
December 2016

-------
Exhibit 6.4: DBP ICR and UCMR 3 Comparison - Use of Disinfectants (select categories)
Among 199
Common Systems
Total
Number of
Plants/ EP
Locations
(surface
water only)3
Number of Plants / EP Locations with...
Total Number
of Plants/ MR
Locations
(surface water
only)
Number of
Plants / MR
Locations
with...
Any Instance
of Chloramines
Exclusive
Use of
Chlorine
Exclusive Use of
Chloramines, OR
both Chlorine and
Chloramines
Any Instance
of Chlorine
Dioxide
Any
Instance of
Ozone
Any
Instance of
UV Light
DBP ICR1
(01/1998-12/1998)
262 Plants
149
(56.9%)
75
(28.6%)
24
(9.2%)
14
(5.3%)
0
(0.0%)
262 Plants4
113
(43.1%)
UCMR 32
(01/2013-05/2016)
342 EP
locations
137
(40.15%)
101
(29.5%)
44
(12.9%)
50
(14.6%)
17
(5.0%)
238 MR
locations
128
(53.8%)
1	For DBP ICR, counts were generated as follows: exclusive use of chlorine = plant used no other disinfectant except chlorine (CL2); exclusive use of chloramines,
OR both chlorine and chloramines = plant used no other disinfectant except chloramine (CLM) or chloramine & chorine (CL2_CLM); any instance of chlorine
dioxide = plant used chlorine dioxide (and may have also used other disinfectants); any instance of ozone = plant used ozone (and may have also used other
disinfectants); any instance of UV light = plant used UV (and may have also used other disinfectants); any instance of chloramines = distribution disinfectant type
was chloramine with or without other disinfectants.
2	For UCMR 3, counts were generated as follows: exclusive use of chlorine = EP used no other disinfectant except chlorine (CLGA, CLOF or CLON); exclusive use
of chloramines, OR both chlorine and chloramines = EP used no other disinfectant except chloramine (CAGC, CAOF or CAON) or chloramine and chorine. (A
plant using both chloramine and chlorine would be counted in this column.); any instance of chlorine dioxide = EP used chlorine dioxide (and may have also used
other disinfectants); any instance of ozone = EP used ozone (and may have also used other disinfectants); any instance of UV light = EP used UV (and may have
also used other disinfectants); any instance of chloramines = MR used chloramine with or without other disinfectants.
3	Only DBP ICR plants served by surface water were included. Plants may have multiple EP locations. Furthermore, only UCMR 3 EP and MR locations with
source water designation "SW" were included in this analysis; those served by ground water, ground water under the direct influence of surface water ("GU") or
mixed source water ("MX") were excluded.
4	To determine the number of plants with any instance of chloramines in MR locations in DBP ICR, the disinfectant type in the distribution system was used.
Six-Year Review 3 Technical Support Document
for Disinfectants/Disinfection Byproducts Rules
6-12
December 2016

-------
6.1.2.3 Source Water Quality Research
In this section, considerations related to source water quality that affect DBP formation include
the NOM fractions (i.e., hydrophilic and hydrophobic), NOM sources (i.e., terrestrial and
aquatic), precursors in wastewater treatment plant effluent, temperature and pH.
NOM Fractions
New research conducted since development of the Stage 2 D/DBPR suggests that both
hydrophilic and hydrophobic fractions of NOM serve as DBP precursors and can influence DBP
speciation (e.g., Kim and Yu, 2005; Kanokkantapong et al., 2006; Hua and Reckhow, 2007;
Karanfil et al., 2011). Hua and Reckhow (2007), Kim and Yu (2005), Panyapinyopol et al.
(2005a; 2005b) and Chow et al. (2005) found that hydrophobic fractions of NOM increase the
formation potential of THMs, HAAs and TOX more than the hydrophilic portions. Dickenson et
al. (2008) reported, however, that, by mass, most of the byproducts of chlorination of J3-
dicarbonyl acids (i.e., aromatic structures within the hydrophilic fraction of NOM) were THMs
and DHAA. Hydrophilic portions of NOM have also been linked to nitrosamine formation
(Chuang et al., 2013; Hatt et al., 2013; Krasner et al., 2013; Wang et al., 2013). Chloride and
bromide are capable of influencing bromination rates of DBP precursors (Sivey et al., 2015). A
recent analysis of more than 30 years of published data on more than 185 NOM compounds, as
well as DBP formation reports, found that given the complexities of water quality characteristics,
NOM characteristics and DBP speciation, there is unlikely to be any one predictor of DBP
formation in drinking water (Bond et al., 2012a). Bond et al. (2012a) recommend that both
hydrophobic and hydrophilic components of NOM be removed from raw water to allow for more
effective DBP control.
NOM Sources
New research suggests that the source of NOM influences the type of DBPs that are formed.
NOM from terrestrial sources forms different types of DBPs than aquatic sources of NOM such
as algae, as described below.
Chlorination or chloramination of lignin, a key component of terrestrially derived NOM, has
been found to form TCAA and to a lesser extent DCAA (Hua et al., 2014). Aging and
biodegradation of terrestrial organic matter yields more HAA, as well as THM, than fresh
organic matter (Beggs and Summers, 2011; Reckhow et al., 2004). Terpenoids produced by
animals, plants and microorganisms have been found to contribute to THM formation (Joll et al.,
2010). Reckhow et al. (2007) found that waters with high humic content, which is indicative of
terrestrial sources, tended to form more identifiable DBPs such as THM and HAA, while waters
with low humic content formed more unknown DBPs. In a study on seven bacterial cultures
commonly found in soil and water, Ng et al. (2015) found bacterial organic matter to also be a
potential DBP precursor, as well as reduce disinfection efficiency.
Research since the Stage 2 D/DBPR has found that aquatic sources of NOM, such as algae, can
be significant contributors of THM and HAA precursors (Nguyen et al., 2005; Callinan et al.,
2013; Lui et al., 2012; Zhang et al., 2012). Callinan et al. (2013) found that THM formation
correlated well with trophic indexes of chlorophyll A and total phosphorus, indicating a
Six-Year Review 3 Technical Support Document	6-13
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
dependence of THM precursors on algal growth in lakes. Zhang et al. (2012) observed that
odorant compounds released by algae formed chloroform more efficiently than other compounds.
Despite accounting for 0.02 percent or less of total organic matter present, algal odorants were
responsible for more than 1 percent of the chloroform formed. Nguyen et al. (2005) found that
algal-derived dissolved organic carbon (DOC) formed 0.53 |imol chloroform/mg DOC, 0.27
|imol DCAA/mg DOC and 0.14 |imol TCAA/mg DOC. In addition, cyanobacteria (also referred
to as blue-green algae) presence has been correlated to precursors for a variety of DBPs
including: THMs and HAAs (Wert and Rosario-Ortiz, 2013); HAN, HNM, nitrosamines, CNX
and haloacetamides (Bond et al., 2011; Bond et al., 2012b); trichloronitromethane (Fang et al.,
2010a; Yang et al., 2011); and NDMA (Fang et al., 2010b; Li et al., 2012; Zamyadi et al., 2012).
These findings are important because algae and substances derived from algae tend to have low
SUVA values (Henderson et al., 2008; Li et al., 2012; Nguyen et al., 2005). This is contrary to
the previous recognition that NOM with high aromatic content tended to increase THM and
HAA formation potential (USEPA, 20051).
Weiss et al. (2013) conducted a study of the New York City water supply to evaluate the extent
to which source water selection strategies, based on the amount of NOM, could be used to reduce
the concentrations of DBPs in finished drinking water. Reservoir monitoring data indicated wide
variability in DBP precursors across time and source waters.
Wastewater Influences
New research has continued to show that wastewater may change the types of DBPs formed and
may influence the formation of nitrogenous DBPs. Krasner et al. (2008) suggest that the DBP
precursors in wastewater treatment plant effluent may pose more of a risk for downstream
drinking water facilities than the actual DBPs in the wastewater effluent. Wastewater treatment
facilities that practice nitrification and denitrification generally have lower levels of HAN,
haloacetaldehyde and NDMA precursors, as well as lower DOC and dissolved organic nitrogen
(DON) concentrations in their effluent than facilities without these practices (Krasner et al.,
2008). NOM in treated wastewater effluent may have a higher NDMA formation potential
compared to NOM in source drinking water without wastewater influence (Krasner et al., 2013).
Some treatment processes, including those of both wastewater and drinking water, have been
found to result in unintended consequences. For instance, while Liu and Li (2010) determined
that the biological processes in wastewater treatment plants can lower the quantity of some DBP
precursors in wastewater effluent, the wastewater treatment processes can increase formation
potential for other DBPs. Yang and Zhang (2014) found that chlorination of saline wastewater
effluents used for toilet flushing in coastal cities resulted in brominated DBP formation,
specifically halogenated pyrroles (e.g., tetrabromopyrrole, tribromochloropyrrole and
tribromopyrrole).
Rice et al. (2013) studied de facto wastewater reuse, the incidental presence of treated
wastewater in public water supplies. In 1980, EPA identified PWSs that were influenced by
upstream wastewater treatment plant discharges and found that the source water of the top 25
most affected PWSs contained between 2 and 16 percent wastewater discharges from upstream
wastewater effluents under average streamflow conditions. Rice et al. (2013) provided an update
to the original 1980 study by creating a geospatial dataset of PWSs and water treatment plants
Six-Year Review 3 Technical Support Document	6-14
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
(WTPs) across the United States and using it to determine the degree to which de facto reuse
occurs in selected cities. From 1980 to 2008, it was found that de facto reuse increased for 17 of
the 25 most heavily influenced PWSs. De facto reuse often made up significant portions of the
drinking water supplies (ranging from 7 to 100 percent) under low streamflow conditions.
Additionally, Rice and Westerhoff (2014) studied 2,056 surface water intakes from water
systems that served approximately 82 percent of the United States' population, finding that 50
percent of the intakes were potentially impacted by upstream wastewater discharges.
Research has provided new insights into the contribution of wastewater-discharged NDMA
precursors. In waters impacted by wastewater treatment plants, nitrosamine precursor
concentrations (including dimethylamine) ranged from 190 to 1,200 ng/L (Pehlivanoglu-Mantas
and Sedlak, 2006). Utilities treating surface waters impacted by wastewater flows generally show
higher nitrosamine formation compared to those treating ground water or more pristine surface
waters (Padhye et al., 2010). For more information on wastewater influences on nitrosamine
formation, see the Six-Year Review 3 Technical Support Document for Nitrosamines (USEPA,
2016d).
Wastewater effluent contains ammonia, which can also influence DBP formation and speciation.
Several studies demonstrated a decrease in total DBP formation in the presence of ammonia,
most likely due to chloramine formation (Hua and Reckhow, 2008; Yang and Shang, 2004; Fang
et al., 2010b; Matamoros et al., 2007). Sun et al. (2009) reported an increase in HAA formation
and a decrease in THM formation at elevated levels of ammonia. Research on DBP formation in
drinking water from a heavily polluted surface water in Beijing, China, suggests that for some
source waters, the presence of ammonia may significantly inhibit formation for certain types of
DBPs (including that of THM and HAA) (Tian et al., 2013).
Temperature
New research since the Stage 2 D/DBPR has provided additional insight into the role of
temperature in DBP formation. Temperature has been found to generally increase DBP
formation, but the effect varies depending on the specific DBP (Toroz and Uyak, 2005; Hua and
Reckhow, 2008; Roccaro et al., 2008). Hua and Reckhow (2008) found that THMs increased the
most with increasing temperature, followed by DHAA and THAA. Obolensky and Singer (2008)
reported that brominated DBPs were less temperature-dependent than chlorinated DBPs. Liu and
Reckhow (2015) analyzed DBP levels in hot and cold tap water originating from a municipal
water system that used free chlorine as the final disinfectant. They found that levels of THMs,
DCAA and chloropicrin were higher in the hot tap water compared to the cold tap water, though
there was no difference in the concentrations of TCAA.
Although formation of most DBPs increases with higher temperatures, some DBPs can degrade
at higher temperatures, leading to complex behavior. Liu and Reckhow (2013) reported
significant decreases in DC AN, 1,1,1-trichloropropanone, chloropicrin and 1,1-
dichloropropanone (DCP) following an increase in water temperature for 24 hours, particularly
at a higher pH of 8. The decreases in chloropicrin and DCP followed initial increases showing an
initial rapid formation followed by rapid degradation. Liu and Reckhow (2015) found that hot
tap water contained less DC AN, BDCAA, BCAA and 1,1,1-trichloropropane than cold tap
samples.
Six-Year Review 3 Technical Support Document	6-15
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
pH
New research continues to show the linkages between pH and DBP formation during
chlorination. General trends indicate increased THM formation at a higher pH, although the
influence of pH on THM formation may be more complex than previously thought and depend
on disinfectant type, precursor type and reaction time. For example, THM formation increased
from chlorine contact with carbohydrates but decreased from chlorine contact with 3-
oxopentanedioic acid at pH 8, compared to pH values of 5 and 5.5 (Bond et al., 2012a). Hua and
Reckhow (2012) reported initial increases in DBCM and bromoform in chlorinated samples for 5
days with increasing pH, but then decreasing concentrations for the remainder of the 10-day test,
when pH levels were adjusted from 7.5 to 8.3 and from 8.3 to 9.6. Based on the results of their
study, Hua and Reckhow (2012) suggest that BrTHMs may degrade under high pH levels.
Studies show mixed results on the influence of pH on HAA formation in chlorinated water.
Several researchers noted decreased TCAA formation at higher pH and no effect of pH on
DCAA formation (Bond et al., 2012a; Hua and Reckhow, 2008; Obolensky and Singer, 2008; Hu
et al., 2010; Fang et al., 2010a). Obolensky and Singer (2008) and Chu et al. (2012) found a
decrease in DHAA formation at higher pH with chlorine.
Two studies evaluated the impacts of pH on DBP formation during chloramination. Hua and
Reckhow (2008) found that TOX formation with chloramination significantly decreased under
elevated pH conditions. Both Hua and Reckhow (2008) and Pope et al. (2007) found a decrease
in DHAA formation at higher pH when chloramines were used.
6.1.2.4 Distribution System Conditions
As was recognized during the development of the Stage 2 D/DBPR, high THM4 and HAA5
levels do not necessarily occur at the location with the maximum residence time (USEPA, 20051;
USEPA, 2005g). Those factors affecting DBP formation (as discussed earlier) along with the
distribution system management practices (including localized treatment, as discussed in Chapter
7 and chlorine burn as discussed in Chapter 8) can affect temporal and spatial variation of DBP
levels throughout a distribution system. In addition, some DBPs (as organic contaminants) can
be degraded under certain conditions in a distribution system. For instance, biological and
inorganic degradation reactions have an effect on where HAA peaks occur in the distribution
system. New studies have examined the conditions under which degradation occurs. Researchers
found that monohaloacetic acids degrade most quickly, followed by DHAAs and THAAs, which
degrade slowly or not at all (Baribeau et al., 2005; Bayless and Andrews, 2008). Speight and
Singer (2005) found that degradation only occurs when no chlorine residual is present.
According to research by Baribeau et al. (2005) and Speight and Singer (2005), degradation of
HAAs proceeds more quickly at higher temperatures. Some bacterial species responsible for
HAA degradation have been identified, including Afipia and Methylohacterium (Zhang et al.,
2009a). Reaction of HAA with iron pipe walls in the distribution system has also been found to
be a mechanism for HAA degradation (Zhang et al., 2004; Arnold et al., 2010).
Six-Year Review 3 Technical Support Document	6-16
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
6.1.2.5 DBP Formation Modeling
Since promulgation of the Stage 2 D/DBPR, numerous studies have been done to develop
predictive models for DBP formation. Chowdhury et al. (2009) provided an overview of more
than 100 models for predicting DBP concentrations. While a few models have used kinetic
equations to predict DBP formation, most rely on empirical methods (Chowdhury et al., 2009).
Most models use DOC, disinfectant dose, pH, temperature and contact time as variables and use
empirical fits of DBP formation to data. Many of these models are very similar in form to the
ones used in SWAT to predict DBP formation (additional information on the SWAT model is
available in Appendix B, as well as in Chapter 7. Some models have introduced higher order
terms, such as time squared, or other variables, such as fulvic acid instead of DOC (Chowdhury
et al., 2009). Many of these models have not been calibrated using independent datasets.
Generally, models based on laboratory data can be better controlled and may be more widely
applicable, while models based on field data are more site-specific, but can show better
predictability and distribution system effects.
New DBP modelling efforts have included precursor inputs to study possible effects on DBP
formation potential (Boyer, 2015; Roccaro et al., 2015; Tang et al., 2015). Roccaro et al. (2015)
modeled the formation of THMs, HAAs and HANs in two chlorinating PWSs. DBP species as
well as NOM transformation reactions were evaluated and the authors noted formation changes
when bromide was present. Tang et al. (2015) modeled DBPs in swimming pools, in which DBP
formation was found to be caused by the continuous introduction of anthropogenic contaminants
as well as the number of pool users. Boyer (2015) evaluated previously developed THM
formation models to determine if they could be used at PWSs to predict DBP formation. The
models that contained bromide as a variable tended to under-predict THM4 concentrations;
however, the most statistically-robust models were believed to be appropriate for use at water
utilities.
In addition to models of DBP formation in water distribution systems, Chen and Westerhoff
(2010) constructed a model based on samples from wastewater treatment plants for both HAA
and THM formation. Chowdhury et al. (2011) constructed a model to predict DBP
concentrations in residential plumbing. Hao et al. (2012) used three-dimensional excitation and
emission fluorescence spectroscopy to develop a predictive model for THM and HAA formation.
As described in Chapter 3, EPA regulated HAA5, not HAA9, due to a lack of analytical
standards at that time for four species (BDCAA, DBCAA, TBAA and BCAA) (Shoaf and
Singer, 2007). To further inform an understanding of HAA9, EPA reviewed the literature on
methodologies for estimating unreported HAAs based on the reported HAA5 (or HAA6) and
THM4 concentrations. This included the review of six articles, with five papers by the "Singer
group" (Cowman and Singer, 1996; Roberts et al., 2002; Shoaf and Singer, 2007; Obolensky et
al., 2007; Obolensky and Singer, 2008) and a paper by Francis et al. (2009).
Exhibit 6.5 presents the predictive models developed by the "Singer group" for estimating the
four unregulated HAAs based on the assumption that these HAAs would form in the same
proportion as the corresponding BrTHM species in relation to chloroform on a molar basis.
Six-Year Review 3 Technical Support Document	6-17
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Exhibit 6.5: Singer Group Models for Estimating Unreported HAAs as a Function
of Reported HAAs and THMs
Equation
Roberts et al.
(2002)
Shoafand
Singer
(2007)
Obolensky and
Singer (2008)
1
[BrChAA] = Co+ Ci x ([CbAA] x ([CHBr2Cl] ^ [CHCb]]]
Co = 0
Ci = 1
n=1,844
0 0
-»¦ O
II II
->¦ O
Co = 0.422
Ci = 0.804
n=3,943 r2=0.88
2
[Br2ClAA] = Co + Ci x ([CbAA] * ([CHBr2Cl] ^ [CHCb]]]
=3 O O
II -»¦ O
J-* II II
O O
-»¦ O
II II
->¦ O
Co = 0.770
Ci =0.418
n=3,600 r2=0.67
3
[BrsAA] = Co + Ci x ([CbAA] * ([CHBrs] - [CHBrs]]]
Co = 0
Ci = 1
n=unknown
O O
-»¦ O
II II
->¦ O
Co = 1.014
Ci = 0.270
n=2,663 r2=0.23
4
[BrCIAA] = CO + CI x ([CbAA] * (([CHBrCh] + [CHBr2]] - (2 x
[CHCb]]]]
N/A
O O
-»¦ O
II II
->¦ O
N/A
N/A: model not available.
[...] indicates |jM concentration.
BrChAA = dichlorobromoacetic acid, Br2CIAA = dibromochloroacetic acid, BrsAA = tribromoacetic acid, BrCIAA =
bromochloroacetic acid, CbAA = trichloroacetic acid, CHBr3 = bromoform, CHCb = chloroform, CHBr2CI =
dibromochloromethane, CHBrCb = bromodichloromethane, CbAA = dichloroacetic acid.
Cowman and Singer (1996) examined the impact of pH and bromide concentration on the mole
fraction of total HAA subject to chlorination and chloramination, which laid a foundation for the
subsequent model developmental work. Roberts et al. (2002) developed a three-equation model
for predicting three unreported HAAs (the sum of these three species is called HAA3) using the
first 12 months of the 1997-98 DBP ICR dataset for facilities that reported all nine HAAs. (Note:
all DBP ICR plants reported HAA6; approximately 30 percent of plants reported HAA9.)
Obolensky and Singer (2008) calibrated the Roberts et al. (2002) three-equation model using the
full 18-month ICR dataset after data screening described by Obolensky et al. (2007) and applied
their model to the remaining ICR data for plants that only reported HAA6 to estimate HAA3 and
HAA9. They found that on average, HAA3 represents 13 percent of HAA9 on a molar basis
(Obolensky and Singer, 2008). Furthermore, Shoaf and Singer (2007) extended the Roberts et al.
(2002) approach by adding a fourth equation to estimate BCAA (Exhibit 6.5). A sixth paper by
Francis et al. (2009) proposed a multivariate normal model and data augmentation methods for
characterizing the distribution of groups of DBPs using datasets with left-censored and missing
data. Censored data were treated via a Bayesian Markov Chain Monte Carlo approach.
More recently, EPA proposed a data collection effort under UCMR 4 that would help to further
inform an understanding of the extent to which these unregulated HAAs are present in drinking
water (refer to 40 FR 76897, USEPA, 2015, for further information about the UCMR 4
proposal).
Six-Year Review 3 Technical Support Document	6-18
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
6.2 Occurrence of DBP Precursors
This section summarizes occurrence information for organic and inorganic DBP precursors. It
focuses on specific water quality indicators for these precursors, such as TOC as an indicator for
organic precursors.
6.2.1 Organic Precursors
6.2.1.1	Summary of Stage 1 and 2 D/DBPR Information
Prior to the development of the Stage 2 D/DBPR, the available national organic precursor
occurrence information consisted primarily of data collected as part of the DBP ICR (for systems
serving 100,000 or more people). Specifically, the DBP ICR dataset contained data for UV254,
alkalinity and TOC in surface and ground waters. The DBP ICR supplemental survey following
the DBP ICR collected precursor occurrence data from systems serving less than 100,000
(USEPA, 20051). During the development of the Stage 2 D/DBPR, TOC was considered a
surrogate for the amount of NOM and potential precursors, UV254 was considered a measure of
the aromaticity of the TOC and alkalinity was recognized as affecting TOC removal through
coagulation/sedimentation and DBP formation. During rule development for Stage 2, for large
water systems (serving at least 100,000), EPA generated summary-level information on the
concentration of these parameters as a function of source water type. Additionally, EPA used
multiple other data sources (NRWA; ICR Supplemental Survey; and Waterstats) to characterize
precursor occurrence for small and medium-sized systems. These results are presented in
Appendix B. The Occurrence Assessment for the Final Stage 2 D/DBPR also presents a national
distribution of bromide and TOC occurrence in source water (USEPA, 20051).
The Stage 1 D/DBPR requires all surface water systems (including systems using ground water
under the direct influence of surface water (GWUDI)) using conventional filtration or
precipitative softening to reduce DBP precursors.
6.2.1.2	New Information since the Stage 2 D/DBPR
For the purposes of SYR3, EPA evaluated data collected from the SYR3 ICR dataset and new
literature published since 2006. Additionally, EPA compared the DBP ICR TOC data (pre-Stage
1 D/DBPR; see introduction) to the SYR3 ICR TOC data to evaluate changes in DBP precursor
occurrence over time. EPA evaluated available TOC data on the plant-level. TOC reductions in
the context of the treatment technique (TT) requirement are discussed in Chapter 7. For
additional details on the TT requirement and how available data relate to that requirement, please
see Chapter 7.
The SYR3 ICR dataset contains TOC data for 33 states and surface water systems of all sizes.
Additionally, 34 states and primacy agencies submitted data for alkalinity. EPA excluded all data
that did not pass through its general and analyte-specific QA/QC processes to ensure that the
data were high quality; for more details on the QA/QC steps, please see Appendix B. After data
management and quality checks on the dataset were conducted, approximately 446,000 TOC and
alkalinity records from January 2006 to December 2011 were available.
Six-Year Review 3 Technical Support Document	6-19
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
It is important to note that both CWSs and non-community water systems (NCWSs) were
included in all analyses (presented in this chapter) using the SYR3 ICR precursor data.
Precursor Inventory Analyses
The results of system and population inventories of the SYR3 ICR TOC dataset in 2011 (the
most recent and complete year of data) are included below in Exhibit 6.6 through Exhibit 6.8.
For inventory information in all years (2006-2011), see Appendix B. Exhibit 6.6 depicts the
distribution of systems and population among the different system types and Exhibit 6.7 includes
the same information but distributed based on source water type. The source types are split by
ground water (includes purchasing systems), surface water (includes purchasing systems) and
GWUDI and includes purchased systems as well. Exhibit 6.8 depicts the distribution of systems
and population by both source water type, system type and aggregated by population size.
Overall, the results indicate the majority of the systems with TOC data are surface water CWSs,
serving between 3,300 and 50,000 people.
Exhibit 6.6: Number of Systems and Population Served by Systems in the SYR3
ICR Dataset with TOC Records, by System Type (2011)
Year
System Type
Systems
Population
Number
Percent
Number
Percent
2011
Community
1,775
93.9%
62,322,706
99.9%
Non-transient Non-community
116
6.1%
65,806
0.1%
Transient Non-community
0
0.0%
0
0.0%
Total
1,891
100.0%
62,388,512
100.0%
Exhibit 6.7: Number of Systems and Population Served by Systems in the SYR3
ICR Dataset with TOC Records, by Source Water Type (2011)
Year
Source Water Type
Systems
Population
Number
Percent
Number
Percent
2011
Ground Water
179
9.5%
5,068,752
8.12%
GWUDI
63
3.3%
528,104
0.85%
Surface Water
1,649
87.2%
56,791,656
91.03%
Total
1,891
100.0%
62,388,512
100.0%
Six-Year Review 3 Technical Support Document	6-20
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Exhibit 6.8: Number of Systems and Population Served by Systems in the SYR3
ICR Dataset with TOC Records, by System Size and System Type (2011)
Year
System Size
(Population Served
by the System)
Ground Water
Surface Water
Total


Number of
Systems
Population
Served
Number of
Systems
Population
Served
Number of
Systems
Population
Served
Community Water Systems
2011
< 101
16
970
74
2,185
90
3,155

101 -500
22
5,997
89
25,863
111
31,860

501 - 1,000
16
13,015
98
76,394
114
89,409

1,001 -3,300
22
42,194
339
710,311
361
752,505

3,301 - 10,000
14
92,891
430
2,622,616
444
2,715,507

10,001 -50,000
20
506,456
415
9,634,912
435
10,141,368

50,001 - 100,000
8
630,238
92
6,482,543
100
7,112,781

100,001 - 1 million
18
3,763,850
98
26,250,976
116
30,014,826

> 1 million
0
0
4
11,461,295
4
11,461,295

Total
136
5,055,611
1,639
57,267,095
1,775
62,322,706
Non-Transient Non-Community Water Systems
2011
< 101
21
977
14
769
35
1,746

101 -500
18
3,869
22
6,368
40
10,237

501 - 1,000
2
1,244
21
16,973
23
18,217

1,001 -3,300
1
1,051
15
24,755
16
25,806

3,301 - 10,000
1
6,000
1
3,800
2
9,800

10,001 -50,000
0
0
0
0
0
0

50,001 - 100,000
0
0
0
0
0
0

100,001 - 1 million
0
0
0
0
0
0

> 1 million
0
0
0
0
0
0

Total
43
13,141
73
52,665
116
65,806
Note: GWUDI systems are included in SW and purchased systems are included in each category as well.
Representativeness of SYR3 ICR Precursor Data
There were nearly 36,000 records for TOC reported in 2011 by 1,639 surface water CWSs in the
SYR3 ICR data. Overall, these systems serve approximately 57 million people. When comparing
the 2011 SYR3 data to inventory data in EPA's Safe Drinking Water Information System
(SDWIS)18 in 2011, the SYR3 total count of systems that submitted TOC data represents
18 SDWIS contains information about PWSs and their violations of EPA's drinking water regulations, as reported to
EPA by the states.
Six-Year Review 3 Technical Support Document	6-21
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
approximately 14 percent of the total SDWIS system count for surface water CWSs.19 Despite
these differences, the population served by the SYR3 ICR systems that reported TOC data in
2011 is approximately 28 percent of the entire retail population served by SDWIS surface water
CWSs in 2011, implying that the SYR3 ICR TOC data encompasses a considerable portion of
drinking water consumers from surface water CWSs. Recognizing these findings, EPA believes
that the TOC data in the SYR3 ICR are useful for informing a perspective on the national
occurrence of TOC for SYR.
While these data are helpful for informing an understanding of TOC and alkalinity occurrence on
a national level, it is important to note that there may be gaps in the available information that
could potentially lead to misrepresentation of TOC and alkalinity occurrence. For example, the
SYR3 TOC dataset does not include any surface water CWSs from some states (e.g., Texas) that,
based on data from the DBP ICR, have historically exhibited elevated levels of organic
precursors. It is also important to recognize that surface water systems not using conventional
treatment were not required to collect TOC removal data and such systems (e.g., those using
slow sand or direct filtration) would in general have lower TOC concentrations.
State-level inventory information and record counts for 2006-2011 are presented in Appendix B
for both raw and finished water. For alkalinity inventory data in all years of the SYR3 ICR data,
please see Appendix B.
National TOC Occurrence for 2011
EPA reviewed the entire SYR3 ICR TOC dataset to evaluate plant-level means for TOC in raw
and finished waters for given system size categories. These results, shown in Exhibit 6.9,
represent the most recent and complete year of the SYR3 ICR data (2011); information in all
years is provided in Appendix B. Additionally, EPA included only those plants from systems that
had "paired" data (i.e., data for raw water TOC, finished water TOC and alkalinity collected
during the same month) in the analyses below. The average plant TOC levels are shown as
cumulative distributions in Exhibit 6.10. These results do not estimate TOC removal; thus, the
reader should not construe the difference between raw and finished water TOC values to be
indicative of compliance. Please see Chapter 7 for an evaluation of TOC removal.
Exhibit 6.9: Raw and Finished Water Plant Means from the SYR3 ICR TOC
Dataset; Surface Water Systems (2011)
System Size
Year
Count of
Plants
Median
(mg/L)
Mean
(mg/L)
90%ile
(mg/L)
95%ile
(mg/L)
% Plant
Means
> 2 mg/L
% Plant
Means
> 3 mg/L
Raw Water
Serving <10,000
2011
682
2.54
3.07
5.77
6.79
66%
40%
Serving 10,000 - <100,000
2011
415
2.68
3.19
5.61
6.93
74%
40%
19 To evaluate the completeness of the SYR3 ICR TOC dataset, EPA compared the number of SW CWSs and their
associated population served in each state that submitted TOC data for the year 2011 to the number of active SW
CWSs and their associated population served according EPA's SDWIS/Fed dataset in 2011. For more details on this
comparison, refer to Appendix B.
Six-Year Review 3 Technical Support Document	6-22
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
System Size
Year
Count of
Plants
Median
(mg/L)
Mean
(mg/L)
90%ile
(mg/L)
95%ile
(mg/L)
% Plant
Means
> 2 mg/L
% Plant
Means
> 3 mg/L
Serving >100,000
2011
120
3.06
3.63
6.15
7.04
80%
52%
All
2011
1,217
2.65
3.16
5.79
6.88
70%
41%
Finished Water
Serving <10,000
2011
682
1.53
1.70
3.04
3.47
30%
10%
Serving 10,000 - <100,000
2011
415
1.60
1.69
2.61
3.02
28%
6%
Serving >100,000
2011
120
1.72
1.79
2.78
3.24
26%
8%
All
2011
1,217
1.59
1.71
2.88
3.30
29%
8%
The 2 mg/L TOC level represents the level below which TOC removal is not required in the Stage 2 D/DBPR. There
could be one or more plants per system.
Exhibit 6.10: Cumulative Distribution of Raw Water arid Finished Water Means in
SYR3 ICR TOC Dataset; Surface Water Plants (2011)
• Raw Water
• Finished Water
0123456789 10
Average TOC concentration (mg/L)
Note that the x-axis is cut-off at 10 mg/L; 15 raw water plant-level means were greater than 10 mg/L and are,
therefore, not shown here. While only raw and finished water data from plants providing both raw and finished water
data were used, the respective raw and finished water distributions for a given cumulative percentage are not paired.
Raw water data show that between 40 and 52 percent of surface water plants have average TOC
values greater than 3 rng/L depending on the system size category. Raw water data are fairly
consistent across the three size categories, with slightly higher results across the large size
category (serving 100,000 people or more).
In finished water, the data show slightly higher plant mean TOC values at the upper end of the
distribution for small systems serving less than 10,000 people compared to systems in the 2
Six-Year Review 3 Technical Support Document	6-23	December 2016
for Disinfectants/Disinfection Byproducts Rules

-------
larger size categories. For example, the percent of plant-average finished water TOC values
greater than 3 mg/L is 10 percent for small systems compared to 6 percent and 8 percent for
medium and large systems, respectively. The difference can also be seen in the percent of plant-
average TOC values greater than 2 mg/L, with 30 percent greater than 2 mg/L for small systems,
compared to 28 percent and 26 percent greater than 2 mg/L for medium and large systems,
respectively. Differences in finished water data could indicate that small systems are removing
less TOC compared to medium and large systems. Please refer to Chapter 7 for information on
TOC reductions in the context of the TT requirement.
Changes in TOC Occurrence
EPA evaluated the changes in TOC occurrence over time, using data from both the DBP ICR and
SYR3 ICR datasets. EPA used 1998 data from the DBP ICR dataset and 2011 data from the
SYR3 ICR dataset, including only data from systems that were found in both datasets (referred to
as "common systems"). As mentioned earlier, the DBP ICR only contains information from large
surface water systems serving 100,000 or more people. Thus, the common systems between the
two datasets are limited to large surface water systems.
Exhibit 6.11 below presents TOC raw and finished water plant-level summary statistics for
common systems in the DBP ICR and SYR3 ICR. The common systems were distributed across
14 states (Alabama, Alaska, Illinois, Indiana, Iowa, Kentucky, Nevada, New Jersey, North
Carolina, Oklahoma, Pennsylvania, South Carolina, Virginia and West Virginia).
Exhibit 6.11: Raw and Finished Water Plant Mean TOC from the DBP ICR (1998)
and SYR3 ICR (2011); Common Surface Water Systems
Data
Source
Year
Count1
Median
(mg/L)
Mean
(mg/L)
90%ile
(mg/L)
95%ile
(mg/L)
% Means
> 2 mg/L
% Means
> 3 mg/L
Raw Water
DBP ICR
1998
100
2.84
2.96
5.08
6.39
67%
44%
SYR3 ICR
2011
105
3.01
3.35
5.64
6.58
79%
50%
Finished Water
DBP ICR
1998
101
1.78
1.77
2.82
3.21
36%
7%
SYR3 ICR
2011
105
1.73
1.74
2.68
3.23
26%
7%
1 The 61 common water systems for raw water TOC represent 100 plants in the DBP ICR and 105 plants in the SYR3
ICR datasets and the 61 common water systems for finished water TOC represent 101 plants in the DBP ICR and
105 in the SYR3 ICR datasets.
Exhibit 6.11 demonstrates that, when looking at data from 1998 and 2011, there was an increase
in raw water TOC levels for large surface water supplies. Large system raw water TOC averaged
2.96 and 3.35 mg/L in DBP ICR and SYR3 ICR data, respectively. The finished water data
showed more variability across distribution of values, with slightly lower averages with the
SYR3 ICR data. A possible explanation for instances where there is minimal difference across
years is that even though the Stage 1 D/DBPR implementation did not begin until 2002, many
systems may have already been in the process of addressing the TOC requirements since they
were initially included with the 1994 proposed D/DBPR (see Chapter 3 for background
Six-Year Review 3 Technical Support Document	6-24
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
information on the regulation). Other possible explanations for minimal difference across years
could be that for many systems the treatment they already had in place may have been able to
remove TOC.
TOC Data from DBP ICR
The DBP ICR dataset contains TOC information spanning 18 months (July 1997-December
1998) of the pre-Stage 1 D/DBPR time period. For SYR3, EPA looked at DBP ICR data for
calendar year 1998 to minimize seasonal bias. EPA evaluated the finished water TOC data for
conventional plants (further differentiated into conventional with softening and conventional
without softening categories) and non-conventional plants (including direct filtration, slow sand
and unfiltered plants) in systems serving at least 100,000 people. For those instances where 90th
percentiles were based on less than 10 plant means and where 95th percentiles were based on less
than 20 plant means, EPA chose not to include these results because they were not believed to be
reliable estimates.
Exhibit 6.12 and Exhibit 6.13 provide a summary of raw water and finished water plant mean
TOC values for surface water plants, respectively. Results are shown separately for seven surface
water plant types as indicated in the DBP ICR.
Exhibit 6.12: Raw Water Plant Mean TOC Data from Surface Water Plants in the
DBP ICR (1998, Systems Serving > 100,000 People)1
Plant Type2
Plant Type
Code
Number
of
Surface
Water
Plants3
Percentage
of Total
Mean of
Plant
Mean
TOC, mg/L
90%ile of
Plant
Mean
TOC, mg/L
95%ile of
Plant
Mean
TOC, mg/L
% Plant
Mean
TOC
> 2 mg/L
% Plant
Mean
TOC
> 3 mg/L
Conventional
(No Softening)
CONV
258
75%
3.2
5.1
6.1
79%
45%
Conventional
(Softening)
SOFT
38
11%
4.7
7.3
7.9
95%
84%
Direct Filtration
DF
23
7%
2.4
3.2
3.8
65%
22%
In-Line Filtration
ILF
6
2%
1.4
-
-
17%
17%
Slow Sand Filtration
SSF
2
1%
1.6
-
-
50%
0%
Unfiltered
UNFILT/SW
14
4%
1.6
2.5
-
29%
0%
Other
OTHER
4
1%
2.2
-
-
75%
0%
Total
--
345
100%
3.2
5.3
6.4
77%
45%
1	If more than one TOC value was provided for a given month, EPA calculated the monthly average before calculating
the yearly average (also to minimize seasonal bias).
2	The plants with the treatment plant type code of "DIS/GW" or "MEMBRANE" (one each) were excluded. The
"Conventional (Softening)" category includes plant type codes: CMPLX/SOFT, CS/SOFT, SOFT, SPLIT/SOFT and
TS/SOFT.
3	All SW plants were included in this analysis, not just those with at least 9 months of data.
Six-Year Review 3 Technical Support Document	6-25
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Exhibit 6.13: Finished Water Plant Mean TOC Data from Surface Water Plants in
the DBP ICR (1998, Systems Serving > 100,000 People)1
Plant Type2
Plant Type
Code
Number
of
Surface
Water
Plants3
Percentage
of Total
Mean of
Plant
Mean
TOC,
mg/L
90%ile of
Plant Mean
TOC, mg/L
95%ile of
Plant
Mean
TOC, mg/L
% Plant
Mean
TOC
> 2 mg/L
% Plant
Mean
TOC
> 3 mg/L
Conventional
(No Softening)
CONV
263
75%
2.1
3.1
3.5
46%
13%
Conventional
(Softening)
SOFT
39
11%
2.6
3.7
3.9
77%
33%
Direct Filtration
DF
23
7%
1.9
2.6
2.7
48%
4%
In-Line Filtration
ILF
6
2%
1.0
-
-
0%
0%
Slow Sand Filtration
SSF
2
1%
1.2
-
-
0%
0%
Unfiltered
UNFILT/SW
14
4%
1.5
2.4
-
21%
0%
Other
OTHER
4
1%
1.3
-
-
25%
0%
Total
--
351
100%
2.1
3.2
3.7
48%
13%
1	If more than one TOC value was provided for a given month, EPA calculated the monthly average before calculating
the yearly average (also to minimize seasonal bias).
2	The plants with the treatment plant type code of "DIS/GW" or "MEMBRANE" (one each) were excluded. The
"Conventional (Softening)" category includes plant type codes: CMPLX/SOFT, CS/SOFT, SOFT, SPLIT/SOFT and
TS/SOFT.
3	All SW plants were included in this analysis, not just those with at least 9 months of data.
Conventional treatment plants without softening were the most common type of surface water
plant included in this dataset, representing approximately 75 percent of the surface water plants.
Conventional plants with softening, as well as direct filtration plants, were also common
(approximately 11 percent and 7 percent, respectively). These three plant types had higher levels
of TOC overall than the other types of treatment plants evaluated.
Exhibit 6.14 and Exhibit 6.15 provide a summary of raw water and finished water plant mean
TOC values for ground water (GW) plants, respectively. Results are shown separately for five
GW plant types as indicated in the DBP ICR.
Exhibit 6.14: Raw Water Plant Mean TOC Data from Ground Water Plants in the
DBP ICR (1998, Systems Serving > 100,000 People)1
Plant Type2
Plant Type
Code
Number
of GW
Plants3
Percentage
of Total
Mean of
Plant
Mean
TOC, mg/L
90%ile of
Plant Mean
TOC, mg/L
95%ile of
Plant
Mean
TOC, mg/L
% Plant
Mean
TOC
> 2 mg/L
% Plant
Mean
TOC
>3
mg/L
Conventional
(No Softening)
CONV
2
2%
9.4
-
-
100%
50%
Conventional
(Softening)
SOFT
23
18%
5.8
11.7
12.6
83%
61%
Six-Year Review 3 Technical Support Document	6-26
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Plant Type2
Plant Type
Code
Number
of GW
Plants3
Percentage
of Total
Mean of
Plant
Mean
TOC, mg/L
90%ile of
Plant Mean
TOC, mg/L
95%ile of
Plant
Mean
TOC, mg/L
% Plant
Mean
TOC
> 2 mg/L
% Plant
Mean
TOC
>3
mg/L
Disinfecting / GW
DIS/GW
63
50%
0.3
0.9
1.2
3%
2%
Other / GW
OTHER/GW
34
27%
0.9
2.3
3.4
12%
9%
Other
OTHER
4
3%
3.1
-
-
25%
25%
Total
--
126
100%
1.7
4.2
10.3
22%
16%
1	If more than one TOC value was provided for a given month, EPA calculated the monthly average before calculating
the yearly average (also to minimize seasonal bias).
2	The one plant with the treatment plant type code of "MEMBRANE" was excluded. The "Conventional (Softening)"
category includes plant type codes: CMPLX/SOFT, CS/SOFT, SOFT, SPLIT/SOFT and TS/SOFT.
3	All GW plants were included in this analysis, not just those with at least 9 months of data.
Exhibit 6.15: Finished Water Plant Mean TOC Data from Ground Water Plants in
the DBP ICR (1998, Systems Serving > 100,000 People)1
Plant Type2
Plant Type
Code
Number
of GW
Plants3
Percentage
of Total
Mean of
Plant Mean
TOC, mg/L
90%ile of
Plant Mean
TOC, mg/L
95%ile of
Plant
Mean
TOC, mg/L
% Plant
Mean
TOC
> 2 mg/L
% Plant
Mean
TOC
>3
mg/L
Conventional
(No Softening)
CONV
2
2%
6.2
-
-
100%
50%
Conventional
(Softening)
SOFT
23
18%
4.1
8.7
10.0
70%
43%
Disinfecting / GW
DIS/GW
63
50%
0.3
1.2
1.7
3%
0%
Other / GW
OTHER/GW
34
27%
0.9
2.3
3.4
12%
9%
Other
OTHER
4
3%
1.1
-
-
25%
0%
Total
--
126
100%
1.3
3.4
6.8
20%
11%
1	If more than one TOC value was provided for a given month, EPA calculated the monthly average before calculating
the yearly average (also to minimize seasonal bias).
2	The one plant with the treatment plant type code of "MEMBRANE" was excluded. The "Conventional (Softening)"
category includes plant type codes: CMPLX/SOFT, CS/SOFT, SOFT, SPLIT/SOFT and TS/SOFT.
3	All GW plants were included in this analysis, not just those with at least 9 months of data.
Ground water plants that disinfect ("Disinfecting / GW") were the most common type of GW
plant included in this dataset, representing approximately 50 percent of the GW plants. Other /
GW plants, as well as conventional softening plants, were also common (approximately 27
percent and 18 percent, respectively).
Literature Information on Organic Precursor Occurrence
In addition to the DBP ICR and SYR3 ICR data, information is available about organic precursor
data for individual and small groups of PWSs. For example, Potter and Wimsatt (2012) measured
TOC, DOC, UV254 and SUVA in seven source waters in Ohio, California, Minnesota and Indiana
to demonstrate compliance with quality control requirements and procedures outlined in the
Six-Year Review 3 Technical Support Document	6-27
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
approved Stage 2 D/DBPR method. The source water UV254 measurements of 0.0726 to 1.0507
cm"1 were similar to DBP ICR UV254 data that ranged from ND (non-detection) to 0.88 cm"1.
TOC measurements were in the same range as DBP ICR data with mean values ranging from
0.42 to 3.64 mg/L. Mean values for DOC ranged from 0.42 to 3.38 mg/L and the mean SUVA
values ranged from 1.95 to 3.37 L/mg-m. It is unclear how many samples were collected for
these source waters.
Selbes et al. (2015) studied nine different amino acids under different oxidation conditions and
found that the presence of amino acids in source waters can contribute to the formation of some
halogenated DBPs. However, amino acids in source waters were not determined to affect
nitrogenous DBP formation.
Since the promulgation of the Stage 2 D/DBPR, studies have demonstrated that changing climate
and weather conditions are contributing to changes in organic DBP precursor occurrence. For
example, Samson et al. (2013) demonstrated a significant correlation between climate change
variables (e.g., precipitation and temperature) and source water TOC levels using three case
study utilities that exceeded monthly TOC thresholds. In the Rocky Mountains, the mountain
pine beetle epidemic, which was sustained by warmer winter temperatures and drought
conditions, caused large scale tree die-off in one million acres of pine forest and resulted in
changes in organic precursors in drinking water sources (Mikkelson et al., 2013). Based on data
collected from 2009 to 2011, Mikkelson et al. reported a mean TOC concentration of 2.7 mg/L in
water samples from affected areas, compared to a mean TOC concentration of 0.62 mg/L in
control watersheds. Researchers have also observed increased source water DOC levels
downstream of watersheds impacted by wildfires, particularly during thunderstorms (Writer et
al., 2014) and other periods of increased flow or an increased degree of disturbance (Emelko et
al., 2013). Impacts on DBP precursors can, however, be mixed. In a controlled laboratory study,
Majidzadeh et al. (2015) observed decreased chloroform formation but increased nitrogenous
DBP formation after plant burns. In a controlled field study, Tsai et al. (2015) observed
reductions in dissolved organic matter, THMs, HANs and chloral hydrate after burn. Within
burned areas, the large loss of organic matter from the forest floor can reduce the available DBP
precursors and associated DBP formation potential, as documented by Wang et al. (2015a).
6.2.2 Inorganic Precursors
6.2.2.1 Summary of Stage 1 and 2 D/DBPR Information
Inorganic precursors relevant to DBP formation include bromide and iodide. During
development of the Stage 2 D/DBPR, EPA had occurrence information about bromide for large,
medium and small systems. These data were available in the DBP ICR for large systems; for
systems serving fewer than 100,000 people, data were available from the NRWA and ICR
Supplemental Surveys. Summary-level information about the inorganic precursor data are
included in Appendix B as well as presented in the Occurrence Document for the Stage 2
D/DBPR (USEPA, 20051). Iodide was not measured as part of ICR monitoring; thus, national-
level occurrence data for iodide was not reviewed during the development of the Stage 2
D/DBPR.
Six-Year Review 3 Technical Support Document	6-28
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
6.2.2.2 New Information since the Stage 2 D/DBPR
This section includes new information on the occurrence of bromide and iodide that has been
identified since the Stage 2 D/DBPR was promulgated in 2006.
Bromide Occurrence and Influence on DBP Formation
Since the Stage 2 D/DBPR was promulgated, new information has become available both on the
occurrence of bromide and how it influences DBP formation.
Several studies have indicated that both natural and man-made factors have contributed to
increases in bromide in many source waters. Since the Stage 2 D/DBPR was promulgated,
several studies have documented how wastewater from hydraulic fracturing activities can
contribute to increased source water bromide concentrations. The research has primarily assessed
high-total dissolved solids (TDS) wastewaters with elevated bromide concentrations that
originate from the significantly increased natural gas production in the Marcellus Shale (mainly
in Pennsylvania). Moreover, hydraulic fracturing operators in Pennsylvania have discontinued
the practice of sending wastewater from hydraulic fracturing operations to wastewater treatment
plants (States et al., 2013) and have been shifting towards treatment of those wastewaters for
reuse rather than discharging to surface water bodies (Hammer and VanBriesen, 2012).
Recent findings demonstrated that upstream increases in bromide resulted in increased THM4
and HAA5 concentrations, leading some plants to exceed the Stage 2 MCLs (Hladik et al., 2014;
Parker et al., 2014; States et al., 2013; Warner et al., 2013; McTigue et al., 2014; Xu et al.,
2008). Most drinking water treatment plants are not designed to address high concentrations of
TDS (which can include bromide and iodide), limiting their options for restricting the formation
of brominated and iodinated DBPs. Tighter restrictions in Pennsylvania on TDS in effluent from
wastewater treatment plants and centralized waste treatment facilities have led to a reduction in
in-stream bromide concentrations (Wilson and Van Briesen, 2013).
New research shows that bromide in source water is also being affected by air quality regulations
(e.g., Mercury and Air Toxic Standards; USEPA, 201 lb) that are intended to reduce metals
(including mercury), acid gases, particulates, nitrogen oxides and sulfur dioxide emissions from
coal and oil-fired electrical generating units larger than 25 megawatts. McTigue et al. (2014)
reported that the new regulations have led to an increase of the use of calcium bromide in power
plant wet scrubbers. The bromide from the coal and coal additives is discharged to receiving
streams along with the wet scrubber wastewater. McTigue et al. (2014) linked the Stage 2
D/DBPR violations at surface water treatment plants with upstream coal-fired power plants that
had installed wet scrubbers. Of 96 water treatment plants evaluated, 17 have had DBP violations
since the wet scrubbers were installed and 6 of the 17 had violations within 1 year of the
scrubber installation. One power plant installed a wet scrubber in 2007 and led to a downstream
WTP exceeding the THM4 MCL six times between 2008 and 2012. For this same water
treatment plant, data for the period 2006 to 2013 demonstrated the increased concentration of
bromoform and BDCM following installation of the wet scrubber.
Gruchlik et al. (2015) evaluated the impact of high bromide concentrations in Western Australian
drinking water sources. A survey of bromide concentrations was conducted through their study,
Six-Year Review 3 Technical Support Document	6-29
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
in which they found concentrations ranging from 400 |ig/L to 8,450 |ig/L. Bromide occurred in
both ground and surface water sources.
Increased bromide concentrations have been found to not only affect the amount of DBPs formed
but also their speciation. New research supports the previous understanding that higher source
water bromide leads to higher THM and HAA concentrations following chlorination (Yang and
Shang, 2004; Hua et al., 2006; Matamoros et al., 2007; Reckhow et al., 2007; Navalon et al.,
2008; Singer, 2010; Wert and Rosario-Ortiz, 2013). New studies have shown that bromide can
increase DBP formation when disinfectants in addition to chlorine are used. Hu et al. (2010)
reported that as source water bromide increased, THMs also increased when ozone was used
followed by chlorine. Shah et al. (2012) concluded thatNDMA formation can increase during
disinfection with chloramine with high source water bromide concentrations (e.g., >500 |ig/L).
Regli et al. (2015) estimated increased THM levels as a function of hypothetical increased source
water bromide levels. They estimated that on average across large plants in the United States,
THM levels could increase by roughly 1 |ig/L for every 10 |ig/L bromide in source water, with
increases varying greatly across plants depending on site-specific conditions.
In addition to increasing the amount of DBPs formed, new research supports the understanding
that increased bromide causes a shift toward more brominated species (Reckhow et al., 2007;
Obolensky and Singer, 2008; Pan and Zhang, 2012; Zha et al., 2014). As the bromine-to-chlorine
ratio increases, Yang and Shang (2004) and Hua et al. (2006) showed that the bromine
incorporation factor (i.e., the number of bromines substituted in each DBP) increases. High
chlorine dose, lower temperature and shorter reaction times have been found to increase the
amount of bromine incorporation into DBPs (Hua and Reckhow, 2012). Reckhow et al. (2007)
found that bromine substitution relative to chlorine substitution is not uniform across all
regulated and unregulated DBPs. Some DBPs form bromine-substituted forms more easily than
others. Bromine substitution was highest with dihaloacetonitrile followed by THM and DHAA,
with THAA having the least bromine substitution. Cornwell (2014) points out that because
bromine is heavier than chlorine, the same number of molecules of DBPs will have higher mass
concentrations as bromide increases. This could result in cases where plants near the THM4
and/or HAA5 MCL could exceed the MCL as DBPs shift to more brominated species with
higher bromide source waters.
Increased bromide incorporation in DBPs can result in an increase in HAA9 (as described in
Section 6.1.1) but result in no change or a decrease in the regulated group HAA5. Hua et al.
(2006) found that while increased bromide led to an increase in total HAAs produced, it led to a
slight decrease in the regulated HAA5. Reckhow et al. (2007) found that increasing bromide
concentrations from 0 to 30 |j,mol/L led to a decrease in HAA5 but an increase in HAA9. The
unregulated HAAs were a significant portion of HAA9 at 2 |j,mol/L bromide and were greater
than the HAA5 contribution at 10 |j,mol/L bromide. Singer (2010) found that the bromine
incorporation factor was similar between THMs and HAAs, but the HAA9 concentrations were
significantly higher than HAA5, indicating a substantial influence of brominated HAAs.
A similar trend can occur for THMs. As described above, McTigue et al. (2014) observed a
marked shift in speciation from chloroform to BrTHMs, sometimes resulting in THM4 levels
above the MCL.
Six-Year Review 3 Technical Support Document	6-30
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Iodide Occurrence and Influence on DBP Formation
Increased iodide can lead to iodinated DBPs (e.g., monoiodoacetic acid, iodobromoacetic acid,
iodobromopropenoic acid and 2-iodo-3-methylbutenedioic acid) upon disinfection. A study
(Gruchlik et al., 2015) on iodide in two Western Australian drinking water sources (ground and
surface waters) was conducted to better understand the impact and occurrence of high
concentrations of iodide in source waters. Concentrations of iodide were measured, ranging from
5 |ig/L to 593 |ig/L. In addition to inorganic iodide, organic iodine can also be a source of
precursors. For instance, iodinated x-ray contrast media have also been found to be a potential
precursor of iodinated DBPs. These iodinated media are non-toxic and are designed to pass
through the body following the x-ray procedure. Iodinated x-ray contrast media are not well
removed at wastewater treatment plants (Duirk et al., 2011). Laboratory experiments found that
reactions between iopimadol (an iodinated x-ray contrast medium), chlorine or chloramine and
NOM produced up to 212 nM (44.6 |_ig/L) dichloroiodomethane and up to 3 nM (558 ng/L)
monoiodoacetic acid (Duirk et al., 2011).
Research has led to improved understanding of the reactions of disinfectants with iodide and
other precursors to form iodinated DBPs. Compared to chlorine, chloramines appear to favor
formation of iodinated DBPs (Hua and Reckhow, 2007; Kristiana et al., 2009; Criquet et al.,
2012; Jones et al., 2012). Criquet et al. (2012) found that an increase in the bromide-to-iodide
ratio in source water, as well as exposure to free chlorine and ammonia, reduced the formation of
iodo-THMs in finished water. Hua et al. (2006) found that chlorine could oxidize iodine resulting
in less iodine incorporation at higher doses. Under chlorination, higher iodide concentrations in
source water can actually lead to lower TOX concentrations (Reckhow et al., 2007). Zha et al.
(2014) observed an increase in HAA formation in the presence of iodide followed by a decrease
in chlorinated and brominated HAA concentrations, indicating both an increased rate of HAA
formation and a shift to more iodinated HAAs.
6.3 DBP Occurrence and Exposure
For the purposes of SYR3, EPA assessed the occurrence of regulated and unregulated DBPs. As
was discussed previously, the main source of information was the SYR3 ICR dataset, which
contains PWS compliance monitoring data for THMs, HAAs, bromate and chlorite from 2006 to
2011.
This section summarizes what was known at the time of the Stage 2 D/DBPR on organic and
inorganic DBP occurrence, presents the results of new occurrence analyses using the SYR3 ICR
data for regulated DBPs and discusses new occurrence information available for unregulated
DBPs. EPA excluded all data in the SYR3 ICR dataset that did not pass through its general and
analyte-specific QA/QC processes. See Appendix B for additional details on the QA/QC steps.
Compliance with the Stage 1 D/DBPR for systems monitoring for THM4 and HAA5 is based on
a running annual average (RAA); the annual average of sample results for each treatment plant
within a given system. A key finding that led to the development of the Stage 2 D/DBPR was
that elevated concentrations of THM4 and/or HAA5 above the MCL were present at specific
locations in distribution systems and not accurately reflected by plant-wide averages (USEPA,
2006a). Systems are required under the Stage 2 D/DBPR to report the RAA at each monitoring
Six-Year Review 3 Technical Support Document	6-31
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
location, also known as the locational running annual average (LRAA). Locations with elevated
levels of DBP occurrence within treatment plants and distribution systems (e.g., maximum
residence locations where organic matter has more time to react with disinfectants) were
identified by the systems and the states in an initial distribution system evaluation (IDSE).
Systems disinfecting with ozone or chlorine dioxide are required to monitor for bromate and
chlorite, respectively. Compliance with the bromate MCL is based on monthly entry point
monitoring and compliance with the chlorite MCL is based on daily entry point and monthly
distribution system monitoring.
The timeframe of the SYR3 ICR data (2006 through 2011) covers post-Stage 1 D/DBPR
occurrence. Compliance monitoring for the Stage 2 D/DBPR did not begin until 2012, or later,
for some PWSs. As such, the SYR3 ICR dataset primarily reflects occurrence information for
systems following the effective date of the Stage 1 D/DBPR, but prior to the effective date of the
Stage 2 D/DBPR. However, some systems may have started to make operational adjustments in
anticipation of the Stage 2 requirements before their respective compliance monitoring deadlines.
In addition, for many small systems, the LRAA and RAA sampling locations will be identical
(for example, for systems with only one sampling location in the distribution system). In these
cases, the SYR3 ICR DBP dataset may be reflective of post-Stage 2 occurrence.
Exposure is characterized in the DBP analyses below by summing the population served by
systems with detections ("Phase 1" analyses) and averages ("Phase 2" analyses) above thresholds
of interest (i.e., minimum reporting levels (MRLs) and MCLs). The population served by water
from distribution system locations in which DBP levels exceeded the thresholds mentioned
above would be needed to more accurately estimate exposure; however, the population served
associated with specific sampling locations is often difficult to know and is not reported along
with other SYR3 ICR compliance monitoring records. Additionally, since non-community water
systems are included in the estimates below, the population counts may not accurately represent
the number of people who are served from non-community water systems (e.g., a campground)
where the actual number of consumers may fluctuate over a given period of time. These caveats
should be considered when reviewing the population information below, where further
information on exposure estimates is included.
In a separate effort, the American Water Work Association (AWW A) (Samson, 2015),
conducted a survey of post-Stage 2 D/DBPR occurrence for systems that serve more than
100.000	people. This survey provides a summary of the data they collected for approximately
400 systems across 44 states, covering a time period from 1980 to early 2015 (Samson, 2015).
6.3.1	Overview of DBP Inventory Analyses
The results of system and population inventory information of the entire SYR3 ICR DBP dataset
in 2011 (the most recent and complete year of ICR data) are included below in Exhibit 6.16
through Exhibit 6.18. (Inventory information from all years of the SYR3 ICR dataset (2006-
2011) is provided in Appendix B.) Exhibit 6.16 depicts the distribution of systems and
population among the different system types and Exhibit 6.17 includes the same information but
distributed based on source water type. The source types are split by ground water (includes
purchasing systems), surface water (includes purchasing systems) and purchased and non-
purchased GWUDI systems. Exhibit 6.18 depicts the distribution of systems and population by
Six-Year Review 3 Technical Support Document	6-32
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
both source water type and aggregated by population size. Overall, the results indicate that many
of the systems represented in the dataset are very small ground water systems. Nearly all systems
in the dataset are CWSs. These results are expected, based on the general system characteristics
of those that are expected to comply with the DBP regulations.
Exhibit 6.16: Number of Systems and Population Served by Systems in the SYR3
ICR Dataset (2011) with DBP Records, by System Type
Year
System Type
Systems
Population


Number
Percent
Number
Percent
2011
Community
17,484
81.3%
199,318,093
99.0%

Non-transient Non-community
4,015
18.7%
1,917,482
1.0%

Transient Non-community
8
0.0%
910
0.0%

Unknown
0
0.0%
0
0.0%

Total
21,507
100.0%
201,236,485
100.0%
Exhibit 6.17: Number of Systems and Population Served by Systems in the SYR3
ICR Dataset (2011) with DBP Records, by Source Water Type
Year
Source Water Type
Systems
Population
Number
Percent
Number
Percent
2011
Ground Water
14,558
67.7%
52,559,785
26.1%
GWUDI
380
1.8%
1,755,985
0.9%
Surface Water
6,569
30.5%
146,920,715
73.0%
Total
21,507
100.0%
201,236,485
100.0%
Exhibit 6.18: Number of Systems and Population Served by Systems in the SYR3
ICR Dataset (2011) with DBP Records, by System Size and System Type
Year
System Size
(Population
Served by the
System)
Ground Water
Surface Water
Total
Number of
Systems
Population
Served
Number of
Systems
Population
Served
Number of
Systems
Population
Served
Community Water Systems
2011
< 101
2,222
137,751
397
18,611
2,619
156,362
101 -500
3,267
836,532
932
265,835
4,199
1,102,367
Six-Year Review 3 Technical Support Document	6-33
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Year
System Size
(Population
Served by the
System)
Ground Water
Surface Water
Total


Number of
Systems
Population
Served
Number of
Systems
Population
Served
Number of
Systems
Population
Served

501 - 1,000
1,342
996,958
643
487,084
1,985
1,484,042

1,001 -3,300
1,843
3,465,602
1,528
3,072,766
3,371
6,538,368

3,301 - 10,000
1,186
6,812,788
1,369
8,369,480
2,555
15,182,268

10,001 -50,000
849
17,725,805
1,255
28,285,116
2,104
46,010,921

50,001 - 100,000
119
7,876,457
228
15,813,961
347
23,690,418

100,001 - 1 million
57
11,279,556
231
58,785,307
288
70,064,863

> 1 million
1
2,100,000
15
32,988,484
16
35,088,484

Total
10,886
51,231,449
6,598
148,086,644
17,484
199,318,093
Non-Community Water Systems
2011
< 101
1,693
92,862
93
4,604
1,786
97,466

101 -500
1,370
339,302
136
33,874
1,506
373,176

501 - 1,000
355
255,289
56
42,555
411
297,844

1,001 -3,300
219
368,505
45
82,485
264
450,990

3,301 - 10,000
31
170,278
15
89,903
46
260,181

10,001 -50,000
4
102,100
4
61,297
8
163,397

50,001 - 100,000
0
0
1
71,963
1
71,963

100,001 - 1 million
0
0
1
203,375
1
203,375

> 1 million
0
0
0
0
0
0

Total
3,672
1,328,336
351
590,056
4,023
1,918,392
Note: There is one water system with data in 2006 and 2008 that has an unknown system type. That system is not
counted in this table. In addition, GWUDI systems are included in SW and purchased systems are included in each
category.
6.3.2 Regulated Organic DBPs
6.3.2.1 Summary of Stage 1 and 2 D/DBPR Information
The primary source of national-level occurrence data used for the Stage 1 and Stage 2 D/DBPRs
was the DBP ICR dataset (as presented in the introduction). In addition to the information
summarized below, an evaluation of the DBP ICR dataset, including THM4 and HAA5
occurrence information, is available in USEPA (20051) and McGuire et al. (2002). Additional
discussions related to the analyses in McGuire et al. (2002), as well as other sources of data
(particularly for systems serving fewer than 100,000 people), are available in USEPA (20051).
Chapter 4 of McGuire et al. (2002) (McGuire and Graziano, 2002) specifically discusses the
occurrence of THMs in U.S. drinking water based on the DBP ICR dataset. Analyses of THM4
records, as well as records for the four individual species, are included, along with analyses of
Six-Year Review 3 Technical Support Document	6-34
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
variations in THM4 as a function of source water type, time of year, sampling location,
treatment, geographical location and source water quality parameters like TOC and bromide. The
results were used to estimate the Stage 1 baseline for the Stage 2 D/DBPR as described in the
Stage 2 D/DBPR Economic Analysis (USEPA, 2005g).
Chapter 5 of McGuire et al. (2002) (Obolensky, 2002) discusses the occurrence of HAAs in U.S.
drinking water based on the DBP ICR dataset. Unlike THMs, which have been the subject of
research for decades, this was the first comprehensive resource on HAA occurrence at the
national level. The chapter included results for the five species now regulated as part of HAA5,
as well as four other unregulated HAA species.
Chapter 6 in McGuire et al. (2002) (McClain et al., 2002) included information about the impact
of bromide on THM speciation. Overall, McClain et al. (2002) found that there was a shift to
brominated species as the bromide influent concentration increased.
6.3.2.2 Analysis of SYR3 ICR THM/HAA Data
Occurrence information was collected for both THM4 and HAA5 as part of the SYR3 ICR,
along with information for four individual species of THMs and five individual species of
HAAs. The information is based on data submitted by 45 states for THMs and 44 states for
HAAs, as well as several other primacy agencies, and represents systems of all sizes.
Approximately 70 percent of the systems with SYR3 ICR THM data (about 29,500) submitted
analytical results for THM4 and (4) individual THMs and approximately 74 percent of the
systems with SYR3 ICR HAA data (around 25,000) submitted analytical results for HAA5 and
all HAA5 species. After data management and quality checks on the dataset were conducted,
approximately 2.3 million THM (including results for both THM4 and individual species)
records and 1.9 million HAA (including results for both HAA5 and individual species) records
from January 2006 to December 2011 were available.
It is important to note that both community and non-community water systems were included in
all analyses (presented in this chapter) using the SYR3 ICR THM and HAA data.
Representativeness of SYR3 ICR THM and HAA Data
Inventory information and record counts for the SYR3 ICR data from 2006 to 2011 are presented
in Appendix B. These analyses indicate that the SYR3 ICR data for THMs and HAAs are
generally useful for informing an understanding of the national occurrence of those contaminants
in drinking water. The datasets for THMs and HAAs both represent a large percentage of the
total population served as compared to SDWIS 2011 data,20 systems of various sizes and source
water types and most geographical areas within the United States.
There were nearly 310,000 THM records reported in 2011 from CWSs, with the majority having
been reported by surface water CWSs. More than 17,000 CWSs are included, of which about
20 To evaluate the completeness of the SYR3 ICR DBP dataset, EPA compared the number of CWSs and their
associated population served in each state that submitted trihalomethane and haloacetic acid data for the year 2011
to the number of active CWSs and their associated population served according EPA's SDWIS/Fed dataset in 2011.
For more details on this comparison, refer to Appendix B.
Six-Year Review 3 Technical Support Document	6-35
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
two-thirds are GW systems. Overall, these systems serve close to 200 million people. When
comparing the 2011 SDWIS inventory information with 2011 SYR3 inventory information for
systems reporting THM data, the SYR3 total CWS count represents approximately 35 percent of
the total SDWIS CWS count. Despite these differences, the population served by the SYR3 CWS
that reported THM data in 2011 is approximately 67 percent of the entire retail population served
by active SDWIS CWSs in 2011, indicating that the SYR3 ICR THM data encompasses a large
portion of drinking water consumers. Note that although all SW systems must monitor for THM4
and HAA5, not all GW systems disinfect and non-disinfecting GW systems are not required to
monitor for DBPs (though non-disinfecting GW systems were included in the total SDWIS CWS
counts).
The HAA dataset is comparable with the THM dataset in terms of geographic areas, system sizes
and system types represented. However, about 3,000 fewer CWSs reported HAAs than THMs in
general; the difference came from ground water CWSs as the number of surface water CWSs
reporting THM and HAA records in 2011 is very close. Even so, the population served by
systems reporting HAA data makes up a large portion (approximately 60 percent) of the SDWIS
population served by CWSs.
Without knowing the number of SDWIS systems that are required to monitor for DBPs, it is fair
to conclude that the counts are likely biased in favor of SDWIS and that not all SDWIS systems
are regulated under the Stage 1 and Stage 2 D/DBPRs. It is also important to note that some
systems/states did provide THM and/or HAA data or their data did not pass the QA/QC review
and are, therefore, not included in the inventory tables included in this chapter and the appendix
(see Appendix B for more details on QA/QC processes). Notwithstanding, the SYR3 ICR THM
and HAA datasets are highly comprehensive, represent a variety of system sizes and types and
cover many geographic areas. Based on the outcomes of the inventory analyses, it is clear that
the SYR3 ICR is one of the largest data sources available for THM and HAA compliance
monitoring results.
The analyses presented below used the SYR3 ICR DBP dataset to assess the number of systems
(and population served by those systems) with detections and averages above thresholds of
interest (known as the Phase 1 and Phase 2 analyses, respectively - performed in the same
manner as the Stage 1 and Stage 2 analyses for the chemical phase rules [see USEPA, 2016g for
further information about occurrence analyses for the chemical phase rules]). Other analyses
included below evaluate the differences in average DBP concentrations based on source water
type and system size (cumulative distribution analyses).
Phase 1 - Comparisons of Individual Measurements Relative to the Minimum Reporting Level
(MRL) and MCL
The Phase 1 analysis uses the SYR3 ICR data to identify only systems with detection records,
which were compared to the MRL and MCL as analytical concentration thresholds. All non-
detection records were excluded from the Phase 1 analyses. The number of systems with
detections greater than the MRL and MCL varied by DBP group, with THM4 having a greater
number of systems with detections above the thresholds than HAA5. As mentioned previously,
the timeframe of the SYR3 ICR data corresponds to the period prior to full implementation of the
Stage 2 D/DBPR. Thus, the Stage 1 D/DBPR, where THM4 and HAA5 compliance was
Six-Year Review 3 Technical Support Document	6-36
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
determined as an RAA of all samples for each treatment plant, was in effect. Since the Phase 1
analyses represent counts of systems with at least one detection greater than the MCL and not
calculations of the RAA, detections above the MCL are not equivalent to violations of the MCL.
The population served is likely an overestimation of the true population exposed because of the
way this estimate was derived. In this estimate, the entire retail population served by a system
was counted, even if there was only one sample location where the concentration of THM4 or
HAA5 was greater than 80 or 60 ng/L, respectively. In these cases, the true population exposed
to such elevated concentrations would more appropriately be considered as those consumers
associated with the specific sampling locations where the elevated levels were measured.
However, the population served associated with specific sampling locations is often difficult to
know and is not reported along with other SYR3 ICR compliance monitoring records. Given
these constraints, this evaluation considered the total retail population as an upper-bound for
potential exposure to these contaminants.
Exhibit 6.19 through Exhibit 6.22 below show the Phase 1 analyses relative to the MRL and
MCL for THM4 and HAA5, respectively. The results are provided for each year of the SYR3
ICR data (2006-2011) and are split by ground water (includes purchasing systems) and surface
water (includes purchasing and GWUDI systems). The number and percentage of systems with
detections greater than or equal to the MRL (or MCL) and the population and percentage of the
population served by those systems is presented.
Exhibit 6.19: SYR3 ICR Comparison of Individual THM4 Measurements to the
MRL1
Year
Systems with at Least One Detection > MRL
(0.5 HQ/L)
Population Served by Systems with at Least One
Detection > MRL (0.5 |jg/L)
Total
Ground Water
Surface Water
Total
Ground Water
Surface Water
Number
Percent
Number
Percent
Number
Percent
Number
Percent
Number
Percent
Number
Percent
2006
14,319
81.6%
7,619
71.2%
6,700
97.8%
177,017,643
96.0%
40,948,355
87.4%
136,069,288
98.9%
2007
16,670
75.1%
9,818
64.7%
6,852
97.6%
181,767,765
95.9%
45,563,542
87.2%
136,204,223
99.2%
2008
14,115
78.9%
7,484
67.3%
6,631
98.2%
184,748,156
96.8%
42,891,929
89.4%
141,856,227
99.3%
2009
14,717
83.1%
8,062
73.7%
6,655
98.1%
188,431,430
96.7%
43,683,995
90.7%
144,747,435
98.7%
2010
17,479
77.2%
10,759
68.2%
6,720
98.0%
188,438,743
95.7%
46,435,543
87.7%
142,003,200
98.6%
2011
14,286
80.2%
7,621
69.1%
6,665
98.4%
185,903,996
96.4%
42,948,738
89.0%
142,955,258
98.9%
Average
15,264
79.4%
8,561
69.0%
6,704
98.0%
184,384,622
96.3%
43,745,350
88.6%
140,639,272
98.9%
Note: Percentages are based on the total number of systems providing at least one record for THM4.
1 Within the SYR3 ICR dataset, multiple MRLs were used for the THM4 data. However, the national modal MRL (i.e.,
mode of all state modal values) for THM4 was equal to 0.5 |jg/L.
As shown in Exhibit 6.19, averaging across the yearly results shows that almost all surface water
systems detected THM4 (approximately 98 percent), as did a majority of ground water systems
(approximately 69 percent). The population served by those systems represented almost the
entire population of systems that provided THM4 data; for surface water systems, the average
Six-Year Review 3 Technical Support Document	6-37
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
across all years is nearly 99 percent. The average percentage of all systems (both surface and
ground water) with detections greater than the MRL is close to 80 percent, indicating that the
majority of all systems that submitted THM4 data detected the analyte in any given year over the
SYR3 ICR timeframe.
Exhibit 6.20: SYR3 ICR Comparison of Individual THM4 Measurements to the MCL
(80 Mg/L)
Year
Systems with at Least One Detection > MCL (80
mq/l)
Population Served by Systems with at Least One
Detection > MCL (80 |jg/L)
Total
Ground Water
Surface Water
Total
Ground Water
Surface Water
Number
Percent
Number
Percent
Number
Percent
Number
Percent
Number
Percent
Number
Percent
2006
2,932
16.7%
652
6.1%
2,280
34.0%
59,662,823
32.4%
4,278,313
9.1%
55,384,510
40.3%
2007
2,958
13.3%
634
4.2%
2,324
33.9%
51,971,584
27.4%
4,398,346
8.4%
47,573,238
34.6%
2008
2,779
15.5%
555
5.0%
2,224
33.5%
52,505,549
27.5%
5,904,068
12.3%
46,601,481
32.6%
2009
2,583
14.6%
556
5.1%
2,027
30.5%
44,801,789
23.0%
5,568,723
11.6%
39,233,066
26.7%
2010
2,539
11.2%
579
3.7%
1,960
29.2%
51,097,106
25.9%
5,924,733
11.2%
45,172,373
31.4%
2011
2,419
13.6%
497
4.5%
1,922
28.4%
50,706,736
26.3%
3,742,539
7.8%
46,964,197
32.5%
Average
2,702
14.2%
579
4.8%
2,123
31.6%
51,790,931
27.1%
4,969,454
10.1%
46,821,478
33.0%
Note: Percentages are based on the total number of systems providing at least one record for THM4.
The Phase 1 analyses relative to the MCL for THM4 demonstrate that there is a larger percent of
at least one-time detections above 80 |ig/L in surface water systems than ground water systems;
averaging across years, nearly 5 percent of ground water systems had at least one detection over
the MCL as compared to approximately 32 percent of surface water systems across all years. By
reviewing the results across the ICR time period, one can see that there have been slight
reductions in the number of systems with detections greater than the MCL. This could be a result
of a number of factors, including system treatment changes made to more easily comply with the
Stage 1 D/DBPR and/or early system adjustments made in anticipation of more easily complying
with the Stage 2 D/DBPR when that rule would become effective.
Exhibit 6.21: SYR3 ICR Comparison of Individual HAA5 Measurements to the
MRL1
Year
Systems with at Least One Detection > MRL
(1 hq/l)
Population Served by Systems with at Least One
Detection > MRL (1 |jg/L)
Total
Ground Water
Surface Water
Total
Ground Water
Surface Water
Number
Percent
Number
Percent
Number
Percent
Number
Percent
Number
Percent
Number
Percent
2006
12,001
72.8%
5,711
57.9%
6,290
95.1%
156,393,504
91.2%
33,529,612
77.6%
122,863,892
95.7%
2007
12,823
62.7%
6,460
47.2%
6,363
94.1%
157,486,222
89.0%
34,061,983
70.9%
123,424,239
95.7%
2008
11,727
70.4%
5,481
54.1%
6,246
95.6%
162,606,241
91.0%
33,561,768
75.6%
129,044,473
96.1%
2009
12,323
74.8%
6,011
60.6%
6,312
96.3%
166,767,736
92.1%
34,953,708
79.7%
131,814,028
96.0%
Six-Year Review 3 Technical Support Document	6-38
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Year
Systems with at Least One Detection > MRL
(1 hq/l)
Population Served by Systems with at Least One
Detection > MRL (1 |jg/L)
Total
Ground Water
Surface Water
Total
Ground Water
Surface Water
Number
Percent
Number
Percent
Number
Percent
Number
Percent
Number
Percent
Number
Percent
2010
13,558
64.9%
7,234
50.8%
6,324
95.3%
165,355,659
91.8%
35,222,970
73.0%
130,132,689
98.7%
2011
11,650
71.1%
5,377
54.7%
6,273
95.6%
163,206,005
93.1%
33,215,768
76.6%
129,990,237
98.5%
Average
12,347
69.4%
6,046
54.2%
6,301
95.3%
161,969,228
91.4%
34,090,968
75.6%
127,878,260
96.8%
Note: Percentages are based on the total number of systems providing at least one record for HAA5.
1 Within the SYR3 ICR dataset multiple MRLs were used for the HAA5 data. However, the national modal MRL (i.e.,
mode of all state modal values) for HAA5 was equal to 1 |jg/L.
As demonstrated in Exhibit 6.21, averaging across years shows that almost all surface water
systems (approximately 95 percent) had detections of HAA5, as did a majority of ground water
systems (approximately 54 percent). The average across years shows that close to 70 percent of
systems detected HAA5 in any given year. The total population served by both ground and
surface water systems represented almost the entire population of systems that provided HAA5
data (approximately 91 percent).
Exhibit 6.22: SYR3 ICR Comparison of Individual HAA5 Measurements to the MCL
(60 Mg/L)
Year
Systems with at Least One Detection > MCL
(60 HQ/L)
Population Served by Systems with at Least One
Detection > MCL (60 |jg/L)
Total
Ground Water
Surface Water
Total
Ground Water
Surface Water
Number
Percent
Number
Percent
Number
Percent
Number
Percent
Number
Percent
Number
Percent
2006
1,678
10.2%
284
2.9%
1,394
21.1%
35,687,934
20.8%
1,172,846
2.7%
34,515,088
26.9%
2007
1,741
8.5%
317
2.3%
1,424
21.1%
36,451,893
20.6%
1,326,211
2.8%
35,125,682
27.2%
2008
1,535
9.2%
239
2.4%
1,296
19.8%
34,983,415
19.6%
2,971,772
6.7%
32,011,643
23.8%
2009
1,491
9.0%
240
2.4%
1,251
19.1%
31,998,268
17.7%
3,098,387
7.1%
28,899,881
21.1%
2010
1,305
6.3%
213
1.5%
1,092
16.5%
29,866,868
16.6%
957,186
2.0%
28,909,682
21.9%
2011
1,151
7.0%
163
1.7%
988
15.0%
30,375,448
17.3%
2,773,616
6.4%
27,601,832
20.9%
Average
1,484
8.4%
243
2.2%
1,241
18.8%
33,227,304
18.8%
2,050,003
4.6%
31,177,301
23.6%
Note: Percentages are based on the total number of systems providing at least one record for HAA5.
As demonstrated in the Phase 1 analysis for THM4, HAA5 average results show that the amount
of systems with detections greater than the MCL of 60 |ig/L is more prevalent in surface water
systems (approximately 19 percent) than ground water systems (approximately 2 percent).
Additionally, a slight downward trend in population exposed can be observed for surface water
systems (and less notably for ground water systems); in 2006 there were approximately 34.5
million people exposed to HAA5 levels above the MCL, greater than approximately 27.6
million people in 2011. Comparatively speaking, THM4 occurred in more systems and at higher
levels than HAA5; on average, 1,200 more systems had detections above the MCLs for THM4
than for HAA5.
Six-Year Review 3 Technical Support Document	6-39
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Phase 2 - Comparisons of Average THM4 and HAA5 Concentrations to the MCL
For Phase 2 analyses of THM4 and HAA5 data, EPA compared system mean concentrations to
the respective MCLs. The Phase 2 analyses include systems with both detection and non-
detection records; all non-detection records were set equal to zero for the calculation of system
mean concentrations. The results of the Phase 2 analyses show similarities to the Phase 1
analyses, as the number of surface water systems with averages above the MCLs is greater than
the number of GW systems. As with the Phase 1 analyses, the timeframe of the SYR3 ICR data
corresponds to the Stage 1 D/DBPR, where compliance is determined as an RAA of all samples
for each treatment plant. Since the Phase 2 analyses calculate system-wide averages, results
should not be construed as compliance under the Stage 1 D/DBPR.
Exhibit 6.23 and Exhibit 6.24 below show the Phase 2 analyses compared to the MCL for THM4
and HAA5, respectively. The summary results are provided for each year of the SYR3 ICR
dataset (2006-2011) and are split by ground water (includes purchasing systems) and surface
water (includes purchasing surface water and GWUDI systems). The number and percentage of
systems with average concentrations greater than the MCL and the population and percentage of
the population served by those systems is presented.
Exhibit 6.23: SYR3 ICR Comparison of System Mean THM4 Measurements to the
MCL (80 Hg/L)
Year
Systems with Averages > MCL (80 jjg/L)
Population Served by Systems with Averages > MCL
(80 jjg/L)
Total
Ground Water
Surface Water
Total
Ground Water
Surface Water
Number
Percent
Number
Percent
Number
Percent
Number
Percent
Number
Percent
Number
Percent
2006
933
5.3%
308
2.9%
625
9.1%
2,347,404
1.3%
436,534
0.9%
1,910,870
1.4%
2007
931
4.2%
278
1.8%
653
9.3%
2,683,938
1.4%
476,678
0.9%
2,207,260
1.6%
2008
724
4.0%
223
2.0%
501
7.4%
1,981,252
1.0%
274,524
0.6%
1,706,728
1.2%
2009
619
3.5%
197
1.8%
422
6.2%
1,476,721
0.8%
315,561
0.7%
1,161,160
0.8%
2010
633
2.8%
246
1.6%
387
5.6%
1,381,043
0.7%
266,555
0.5%
1,114,488
0.8%
2011
560
3.1%
197
1.8%
363
5.4%
1,185,915
0.6%
273,336
0.6%
912,579
0.6%
Average
733
3.8%
242
2.0%
492
7.2%
1,842,712
1.0%
340,531
0.7%
1,502,181
1.1%
Note: Percentages are based on the total number of systems providing at least one record for THM4.
It is important to note that system-level averages above the MCLs are not equivalent to violations of the MCLs.
Overall, the THM4 Phase 2 averages across yearly results show that approximately 1 percent of
the population of systems that reported THM4 in the SYR3 ICR dataset is served by a system
with an average THM4 concentration greater than 80 |ig/L. There are a greater percentage of
surface water systems (approximately 7 percent) that had averages greater than the MCL than
GW systems (approximately 2 percent). Population exposure estimates indicate that there are
slight downward trends in the population served with high THM4 levels for both ground water
and surface water systems.
Six-Year Review 3 Technical Support Document	6-40
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Exhibit 6.24: SYR3 ICR Comparison of System Mean HAA5 Measurements to the
MCL (60 Hg/L)
Year
Systems with Averages > MCL (60 jjg/L)
Population Served by Systems with Averages > MCL
(60 jjg/L)
Total
Ground Water
Surface Water
Total
Ground Water
Surface Water
Number
Percent
Number
Percent
Number
Percent
Number
Percent
Number
Percent
Number
Percent
2006
520
3.2%
145
1.5%
375
5.7%
1,257,533
0.7%
151,190
0.3%
1,106,343
0.9%
2007
494
2.4%
153
1.1%
341
5.0%
1,485,755
0.8%
119,015
0.2%
1,366,740
1.1%
2008
378
2.3%
109
1.1%
269
4.1%
826,524
0.5%
85,383
0.2%
741,141
0.6%
2009
327
2.0%
91
0.9%
236
3.6%
767,476
0.4%
81,778
0.2%
685,698
0.5%
2010
274
1.3%
92
0.6%
182
2.7%
708,673
0.4%
98,900
0.2%
609,773
0.5%
2011
216
1.3%
65
0.7%
151
2.3%
1,674,622
1.0%
99,559
0.2%
1,575,063
1.2%
Average
368
2.1%
109
1.0%
259
3.9%
1,120,097
0.6%
105,971
0.2%
1,014,126
0.8%
Note: Percentages are based on the total number of systems providing at least one record for HAA5.
It is important to note that system-level averages above the MCLs are not equivalent to violations of the MCLs.
The Phase 2 averages across yearly results for HAA5 show that the number of ground water
systems with averages greater than the MCL of 60 |ig/L decreased from 145 systems in 2006 to
65 systems in 2011. Similarly, the number of surface water systems with averages greater than
the MCL decreased from 375 systems in 2006 to 151 systems in 2011. However, despite the
reductions in the number of systems with averages above the MCL, there is variation across
years in the population served by systems with averages above the MCL (the total population
served by systems with detections greater than the MCL ranged from 708,673 in 2010 to
1,674,622 in 2011). Interestingly, the population served by systems with average concentrations
of HAA5 greater than the MCL is less than the ground water, surface water and total categories
than the population served by systems with average concentrations of THM4 greater than the
MCL. This indicates that overall exposure from drinking water above the MCLs may not be the
same for both groups of regulated organic DBPs.
The Phase 2 analyses indicate that THM4 and HAA5 average concentrations are above the
MCLs in some systems, but in general are dropping over time. It is important to note that
system-level detections and averages above the MCLs are not equivalent to violations of the
MCLs.
Cumulative Distributions of Average Concentrations for THM4 and HAA5
Using the SYR3 ICR THM4 and HAA5 data, EPA compared the occurrence of the regulated
DBP groups in systems of different sizes and source water types, using the monitoring records
from 2011, the most complete and recent year reflected in the SYR3 ICR DBP dataset. In this
analysis, an average THM4 and HAA5 concentration was calculated for each system for the
calendar year of 2011. Summary results are presented in Exhibit 6.25 for THM4 and Exhibit 6.26
for HAA5. Cumulative distributions of the average concentration per system are presented in
Six-Year Review 3 Technical Support Document	6-41
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Exhibit 6.27 for THM4 and Exhibit 6.28 for HAA5. As with the Phase 2 analyses, EPA
substituted non-detections with zero for the calculation of system mean concentration.
The resulting curves indicate that average concentrations for THM4 and HAA5 in several ground
water systems are less than those in surface water systems. For example, the THM4 cumulative
distribution plot demonstrates that approximately 50 percent of surface water systems serving
less than or equal to 10,000 people have THM4 average concentrations below 42 |ig/L. In
contrast, 50 percent of ground water systems in this same size category people have average
THM4 concentrations below 5 |ig/L. Average HAA5 concentrations at this size category are
lower; for example, about 50 percent of surface water systems serving less than or equal to
10,000 people have average HAA5 concentrations that approximately fall below 25 |ig/L. About
50 percent of ground water systems in this same size category have averages around 2 |ig/L.
Exhibit 6.25: System Means from the SYR3 ICR THM4 Data (2011)
System
Size
Year
Count of
Systems
Mean
90%ile
95%ile
% System
Means
> 80 |jg/L
Ground Water Systems
<=10,000
2011
10,052
14.5
44.4
61.6
1.9%
>10,000
2011
982
17.6
46.8
57.4
0.3%
Surface Water Systems
<=10,000
2011
5,067
43.58
73.0
85.6
6.9%
>10,000
2011
1,707
35.23
58.9
65.2
0.9%
Exhibit 6.26: System Means from the SYR3 ICR HAA5 Data (2011)
System
Size
Year
Count of
Systems
Mean
90%ile
95%ile
% System
Means
> 60 |jg/L
Ground Water Systems
<=10,000
2011
8,931
6.4
17.5
29.5
0.7%
>10,000
2011
899
7.2
21.8
28.3
0.2%
Surface Water Systems
<=10,000
2011
4,943
23.7
44.0
52.6
2.7%
>10,000
2011
1,622
21.1
38.8
46.0
1.0%
Six-Year Review 3 Technical Support Document	6-42
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Exhibit 6.27: SYR3 ICR Data Showing Cumulative Distribution of System Mean
Concentrations for THM4 by System Size and Source Water Type (2011)
Average THM4 Values in 2011
100
o	«
©	c
O	o
4)	©
©> £
> ¦©
£	©
Surface Water <1 OK
(n=5,067)
Surface Water
>10K(n=1,707)
Ground Water
>10K(n=982)
Ground water
<10K(n=10,052)
40.0
0.0
10.0
20.0
30.0
50.0
60.0
70.0
80.0
90.0
100.0
THM4 (|Jg/L)
Exhibit 6.28: SYR3 ICR Data Showing Cumulative Distribution of System Mean
Concentrations for HAAS by System Size and Source Water Type in 2011
Average HAAS Values in 2011
100
Surface Water <1 OK
(n=4,943)
Surface Water >1 OK
(n=1,622)
Ground Water
>10K (n=899)
< 0)
3e o
v> «
£ ©¦
5 co
¦S v
Ground Water <1 OK
(n=8,931)
50.0
0.0
10.0
20.0
30.0
40.0
60.0
70.0
80.0
HAA5 (pg/L)
The differences between the surface water and ground water system mean concentrations could
have resulted from a variety of factors, such as influent water quality. Given that systems using
ground water tend to have lower TOC levels than systems using surface water (as demonstrated
Six-Year Review 3 Technical Support Document	6-43
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
by DBP ICR TOC data presented in Section 6.2), these figures further support the understanding
that higher levels of precursors can, in general, result in greater levels of DBPs.
Effects from Stage 2 D/DBPR
While the SYR3 ICR did not collect information about post-Stage 2 D/DBPR occurrence, EPA
believes that the DBP occurrence estimates for THM4 and HAA5 presented in this document can
be expected to drop for those systems needing to make treatment changes to comply with the
Stage 2 D/DBPR. Samson (2015) is collecting information pertaining to post-Stage 2 D/DBPR
compliance. Samson is focused on collecting post-Stage 2 D/DBPR regulated DBP occurrence
information from systems serving more than 100,000 people. The data collected through this
project and resulting analyses may be able to serve as a comparison with the DBP ICR data from
pre-Stage 1 D/DBPR and the post-Stage 1 D/DBPR time period using the SYR3 ICR data. Such
a comparative analysis would need to consider the effects of different sampling locations
required under the Stage 2 D/DBPR.
6.3.3 Regulated Inorganic DBPs
This section summarizes occurrence and exposure information for bromate and chlorite, which
are the inorganic DBPs regulated under the Stage 1 D/DBPR. Systems were required to comply
with the regulations under the Stage 1 D/DBPR by 2004, so the timeframe in the SYR3 ICR
dataset (2006 through 2011) covers the post-Stage 1 D/DBPR occurrence.
Routine monitoring requirements for bromate require both community and non-transient non-
community water systems (NTNCWSs) using ozone for disinfection or oxidation to take one
sample per month at the entry point to the distribution system for each treatment plant using
ozone. A system is in violation of the bromate MCL if the average of samples covering any
consecutive four quarter period exceeds 10 |ig/L. Additionally, community and non-community
(includes non-transient non-community and transient non-community) water systems using
chlorine dioxide for disinfection or oxidation must monitor for chlorite daily at the entrance to
the distribution system as well as monthly in the distribution system. For any daily sample that
exceeds the chlorite MCL, the system must take additional samples in the distribution system the
following day. A system that has an average of any three sample sets exceeding 1,000 |ig/L is in
violation of the chlorite MCL. Further information on the Stage 1 D/DBPR requirements for
these contaminants is available in Chapter 3.
6.3.3.1 Summary of Stage 1 and 2 D/DBPR Information
The DBP ICR dataset contains occurrence information on bromate and chlorite occurrence and
analyses of the data are available in McGuire et al. (2002) and USEPA (20051).
Chapter 9 of McGuire et al. (2002) (Moll and Krasner, 2002) specifically discusses the
occurrence of bromate in U.S. drinking water among systems using ozone and includes the
results from the DBP ICR survey. Correlations between bromate concentrations and other
parameters, such as ozone contact time, were analyzed, but no association was found. Moll and
Krasner (2002) also discussed other surveys that examined bromate occurrence levels. One study
in particular included a utility that also participated in the DBP ICR and Moll and Krasner (2002)
Six-Year Review 3 Technical Support Document	6-44
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
found that source water bromide loading appeared to be a significant contributor to the
occurrence of bromate in finished water samples among systems using ozone.
6.3.3.2 Analysis of SYR3 ICR Bromate and Chlorite Data
Occurrence information was collected for both bromate and chlorite as part of the SYR3 ICR and
consists of data from 29 states for bromate and 28 states for chlorite, as well as from other
primacy agencies, and represents systems of all sizes. After data management and quality checks
on the dataset were conducted, approximately 8,900 bromate records and 26,000 chlorite records
from January 2006 to December 2011 were available. As mentioned earlier in the chapter, all
information on contaminant-specific QA/QC steps is provided in Appendix B.
The SYR3 ICR data are illustrative of bromate and chlorite occurrence following implementation
of the Stage 1 D/DBPR. Note that the methods used to calculate compliance for bromate and
chlorite did not change with the Stage 2 D/DBPR. Thus, although post-Stage 2 bromate and
chlorite occurrence data are not currently available, EPA believes that any changes in occurrence
due to implementation of the Stage 2 D/DBPR may not be significant.
Representativeness of SYR3 ICR Bromate and Chlorite Data
Inventory information (i.e., the number of records, the number of systems with data and the
population served by those systems) of SYR3 ICR bromate and chlorite data from 2006-2011 are
presented in Appendix B.
An important consideration for understanding the representativeness of the SYR3 ICR bromate
and chlorite data is the monitoring requirements for both analytes. As mentioned previously
(background information on the Stage 1 and Stage 2 D/DBPRs in Chapter 3), systems are only
required to monitor for bromate or chlorite if they disinfect using ozone or chlorine dioxide,
respectively. Overall, the percentage of systems using these two disinfectants in the United
States is unknown but thought to be small. Through the Stage 2 D/DBPR Economic Analysis
(USEPA, 2005g), EPA predicted baseline pre-and-post D/DBPR conditions and projected that
the majority of plants would disinfect using free chlorine or chloramines, with far fewer plants
disinfecting with ozone or chlorine dioxide. Additionally, EPA summarized the use of
disinfectants using UCMR 3 data for entry points and maximum residence time locations
(presented earlier in Section 6.1), the results of which can be characterized as post-Stage 2
D/DBPR conditions, as the monitoring took place from January 2013 to December 2015.21 Using
the UCMR 3 data, EPA found that disinfectant usage varied across source water type and system
size; with a range of 0.2 to 14.8 percent of entry points using ozone and between 0.4 and 10.3
percent of entry points using chlorine dioxide. It is important to note that although systems that
disinfect with chlorine dioxide and ozone must monitor for chlorite and bromate, respectively,
the disinfectant usage for the systems that reported these data below is unknown.
21 UCMR 3 monitoring was scheduled to occur from January 2013 through December of 2015. Some monitoring
data continued to be reported to EPA through 2016. The UCMR 3 occurrence analyses presented in this report are
based on data collected through May 2016 (released online in July 2016). The complete dataset is anticipated to
become available in early 2017.
Six-Year Review 3 Technical Support Document	6-45
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Compared to THMs and HAAs, there are fewer bromate records because only systems that use
ozone as a primary disinfectant are required to monitor for bromate. Altogether, there are 8,884
bromate records, with more than 7,000 records from surface water systems. Over 200 systems
are represented, of which about two-thirds are surface water systems. More than 23 million
people are served water by these systems, almost all of them from surface water systems. All
analyses using bromate data evaluated records at both entry point and distribution system
locations. Approximately 46 percent of the bromate records that passed QA/QC procedures are
samples taken within distribution system locations. Although systems disinfecting with ozone are
not federally required to monitor for bromate in distribution system locations, EPA chose to
include these records in analyses due to the large percentage of distribution system samples and
the possibility that some states may require distribution system monitoring for bromate.
There are fewer chlorite records (compared to THM and HAA records) because only systems
that use chlorine dioxide as a primary disinfectant are required to monitor for chlorite.
Altogether, there are almost 26,000 chlorite records, with the vast majority of records from
surface water systems. Over 200 systems are represented, of which almost 90 percent are surface
water. Over 13 million people are served water by these systems, almost entirely from surface
water systems. The chlorite dataset and analyses within this document contain records from both
entry point and distribution system monitoring locations. Overall, these results are similar to
those for bromate and indicate that the systems that provided chlorite data for the SYR3 ICR are
almost entirely surface water systems.
Taking the proportion of systems that use alternative disinfectants into account, EPA expects that
far fewer systems would report bromate and chlorite records as opposed to their organic
counterparts. This is consistent with the results of the inventory analyses for bromate and chlorite
throughout all years of the SYR3 ICR data.
Analysis Background
Similar types of inventory and occurrence analyses (except for cumulative distribution and
highest concentration analyses) were conducted for bromate and chlorite as for the organic DBPs
(discussed in Section 6.3.2.2).
Phase 1 - Comparisons of Individual Measurements Relative to the MRL and MCL
The Phase 1 analyses for regulated inorganic DBPs follow the same methodology as the Phase 1
analyses for regulated organic DBPs presented earlier. Exhibit 6.29 through Exhibit 6.32 below
show the Phase 1 analysis relative to the MRL and MCL for bromate and chlorite, respectively.
The results are provided for each year covered by the SYR3 ICR dataset and are split by ground
water (includes purchasing systems) and surface water (includes purchasing and GWUDI
systems). The number and percentage of systems with detections greater than or equal to the
MRL (or greater than the MCL) and the population and percentage of the population served by
those systems are presented.
The population served may be an overestimation of the true population exposed at a given time
because of the way this estimate was derived. In this estimate, the retail population served by a
system was counted, even if there was only one sample location where the concentration of
Six-Year Review 3 Technical Support Document	6-46
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
bromate or chlorite was greater than 10 or 1,000 |ig/L, respectively. To the extent that bromate
concentration at the tap is conserved throughout the system (which is assumed for the regulation
with no distribution sampling required), occurrence at the tap should generally reflect occurrence
within the distribution system. However, exceptions to this generalization apply such as a system
blending its distribution system water from multiple plants, not all of which may be using ozone.
Regarding chlorite, systems using chlorine dioxide must monitor daily and if only one of such
measurements exceeds the MRL and MCL, the total population served by that system would be
counted in Exhibit 6.31 and Exhibit 6.32, respectively. Also, for chlorite, the concentration
measured at a given location may be greater or less if measured at point-of-entry or within the
distribution system at different locations due to chemical reactions that can occur.
Exhibit 6.29: SYR3 ICR Comparison of Individual Bromate Measurements to the
MRL1
Year
Systems with at Least One Detection > MRL
(5 HQ/L)
Population Served by Systems with at Least One
Detection > MRL (5 jjg/L)
Total
Ground Water
Surface Water
Total
Ground Water
Surface Water
Number
Percent
Number
Percent
Number
Percent
Number
Percent
Number
Percent
Number
Percent
2006
38
33.0%
7
26.9%
31
34.8%
2,006,305
28.9%
8,359
5.5%
1,997,946
29.5%
2007
28
22.8%
7
25.9%
21
21.9%
6,363,954
45.9%
7,200
9.7%
6,356,754
46.1%
2008
48
31.6%
14
30.4%
34
32.1%
8,249,902
45.2%
12,653
12.1%
8,237,249
45.4%
2009
55
36.2%
13
34.2%
42
36.8%
9,137,433
47.3%
20,179
16.3%
9,117,254
47.5%
2010
62
41.3%
12
30.8%
50
45.0%
10,159,216
62.7%
19,839
12.4%
10,139,377
63.2%
2011
57
38.5%
15
36.6%
42
39.3%
12,985,200
61.5%
34,085
28.3%
12,951,115
61.6%
Average
48
33.9%
11
30.8%
37
35.0%
8,150,335
48.6%
17,053
14.0%
8,133,283
48.9%
Note: Percentages are based on the total number of systems providing at least one record for bromate.
1 Within the SYR3 ICR dataset, multiple MRLs were used for the bromate data. However, the national modal MRL
(i.e., mode of all state modal values) for bromate was equal to 5 |jg/L.
Exhibit 6.30: SYR3 ICR Comparison of Individual Bromate Measurements to the
MCL (10|jg/L)
Year
Systems with at Least One Detection
> MCL (10 jjg/L)
Population Served by Systems with at Least One
Detection > MCL (10 jjg/L)
Total
Ground Water
Surface Water
Total
Ground Water
Surface Water
Number
Percent
Number
Percent
Number
Percent
Number
Percent
Number
Percent
Number
Percent
2006
15
13.0%
5
19.2%
10
11.2%
661,798
9.5%
2,489
1.6%
659,309
9.7%
2007
9
7.3%
3
11.1%
6
6.3%
511,583
3.7%
935
1.3%
510,648
3.7%
2008
13
8.6%
6
13.0%
7
6.6%
1,589,208
8.7%
2,314
2.2%
1,586,894
8.7%
2009
16
10.5%
4
10.5%
12
10.5%
5,189,830
26.8%
1,230
1.0%
5,188,600
27.0%
2010
9
6.0%
2
5.1%
7
6.3%
2,396,741
14.8%
1,170
0.7%
2,395,571
14.9%
Six-Year Review 3 Technical Support Document	6-47
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Year
Systems with at Least One Detection
> MCL (10 jjg/L)
Population Served by Systems with at Least One
Detection > MCL (10 jjg/L)
Total
Ground Water
Surface Water
Total
Ground Water
Surface Water
Number
Percent
Number
Percent
Number
Percent
Number
Percent
Number
Percent
Number
Percent
2011
14
9.5%
9
22.0%
5
4.7%
498,547
2.4%
9,415
7.8%
489,132
2.3%
Average
13
9.1%
5
13.5%
8
7.6%
1,807,951
11.0%
2,926
2.4%
1,805,026
11.1%
Note: Percentages are based on the total number of systems providing at least one record for bromate.
Exhibit 6.31: SYR3 ICR Comparison of Individual Chlorite Measurements to the
MRL1
Year
Systems with at Least One Detection > MRL
(20 jjg/L)
Population Served by Systems with at Least One
Detection > MRL (20 |jg/L)
Total
Ground Water
Surface Water
Total
Ground Water
Surface Water
Number
Percent
Number
Percent
Number
Percent
Number
Percent
Number
Percent
Number
Percent
2006
93
90.3%
4
80.0%
89
90.8%
3,228,368
81.4%
63,841
56.1%
3,164,527
82.2%
2007
94
88.7%
5
71.4%
89
89.9%
3,135,052
56.1%
82,506
61.7%
3,052,546
56.0%
2008
103
77.4%
8
47.1%
95
81.9%
5,141,427
59.6%
89,445
61.7%
5,051,982
59.6%
2009
124
85.5%
7
87.5%
117
85.4%
5,506,619
77.5%
90,414
100.0%
5,416,205
77.2%
2010
133
89.3%
7
77.8%
126
90.0%
6,378,381
76.8%
63,377
98.8%
6,315,004
76.6%
2011
148
90.8%
7
77.8%
141
91.6%
7,514,795
82.4%
94,459
97.0%
7,420,336
82.3%
Average
116
87.0%
6
73.6%
110
88.3%
5,150,774
72.3%
80,674
79.2%
5,070,100
72.3%
Note: Percentages are based on the total number of systems providing at least one record for chlorite.
1 Within the SYR3 ICR dataset, multiple MRLs were used for the chlorite data. However, the national modal MRL
(i.e., mode of all state modal values) for chlorite was equal to 20 |jg/L.
Exhibit 6.32: SYR3 ICR Comparison of Individual Chlorite Measurements to the
MCL (1,000 |jg/L)
Year
Systems with at Least One Detection > MCL
(1,000 jjg/L)
Population Served by Systems with at Least One
Detection > MCL (1,000 \iglL)
Total
Ground Water
Surface Water
Total
Ground Water
Surface Water
Number
Percent
Number
Percent
Number
Percent
Number
Percent
Number
Percent
Number
Percent
2006
3
2.9%
0
0.0%
3
3.1%
174,198
4.4%
0
0.0%
174,198
4.5%
2007
4
3.8%
1
14.3%
3
3.0%
67,132
1.2%
178
0.1%
66,954
1.2%
2008
5
3.8%
0
0.0%
5
4.3%
156,392
1.8%
0
0.0%
156,392
1.8%
2009
7
4.8%
0
0.0%
7
5.1%
230,807
3.2%
0
0.0%
230,807
3.3%
2010
8
5.4%
0
0.0%
8
5.7%
284,914
3.4%
0
0.0%
284,914
3.5%
2011
10
6.1%
0
0.0%
10
6.5%
114,129
1.3%
0
0.0%
114,129
1.3%
Average
6
4.5%
0
2.4%
6
4.6%
171,262
2.6%
30
0.0%
171,232
2.6%
Note: Percentages are based on the total number of systems providing at least one record for chlorite.
Six-Year Review 3 Technical Support Document	6-48
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
The Phase 1 analyses for the regulated inorganic DBPs indicate that the average population
(across all years from 2006-2011) served by systems with detections greater than the MCL for
bromate (approximately 1.8 million) was significantly greater than the average population served
by systems with detections above the MCL for chlorite (approximately 171,262). This is likely
because the systems that use ozone to disinfect have a larger population (approximately 23.2
million) than the population served by systems that use chlorine dioxide (approximately 13.5
million), as demonstrated by the inventory tables. For the Stage 1 D/DBPR, compliance is
determined based on an average of samples for each treatment plant. Since the Phase 1 analyses
represent counts of systems with at least one detection above the MCL rather than counts of
system averages, they should not be used to estimate compliance with the Stage 1 D/DBPR.
Phase 2 - Comparisons of Average Measurements to the MCL
The Phase 2 analysis for bromate follows the same methodology as the Phase 2 analyses for
regulated organic DBPs, as presented in Section 6.3.2.2. Exhibit 6.33 below shows the Phase 2
analysis relative to the MCL for bromate. The results are provided for each year of the SYR3
ICR data and are split by ground water (includes purchasing systems) and surface water
(includes purchasing and GWUDI systems). The table presents the number and percentage of
systems with system means greater than the MCL and the population and percentage of the
population served by those systems.
Exhibit 6.33: SYR3 ICR Comparison of System Mean Bromate Measurements to
the MCL (10 Hg/L)
Year
Systems with Average > MCL (10 jjg/L)
Population Served by Systems with Average > MCL
(10 jjg/L)
Total
Ground Water
Surface Water
Total
Ground Water
Surface Water
Number
Percent
Number
Percent
Number
Percent
Number
Percent
Number
Percent
Number
Percent
2006
4
3.5%
3
11.5%
1
1.1%
10,392
0.1%
2,300
1.5%
8,092
0.1%
2007
2
1.6%
2
7.4%
0
0.0%
135
0.0%
135
0.2%
0
0.0%
2008
2
1.3%
2
4.3%
0
0.0%
137
0.0%
137
0.1%
0
0.0%
2009
2
1.3%
2
5.3%
0
0.0%
80
0.0%
80
0.1%
0
0.0%
2010
0
0.0%
0
0.0%
0
0.0%
0
0.0%
0
0.0%
0
0.0%
2011
3
2.0%
3
7.3%
0
0.0%
150
0.0%
150
0.1%
0
0.0%
Average
2
1.6%
2
6.0%
0
0.2%
1,816
0.0%
467
0.3%
1,349
0.0%
Note: Percentages are based on the total number of systems providing at least one record for bromate.
The results of the Phase 2 analysis for bromate shows that very few systems have average
concentrations above the MCL across all years of the SYR3 ICR dataset and that, relative to the
MCL, there is a lower risk of exposure than for organic DBPs. As described above for the Phase
1 analyses, compliance with bromate and chlorite MCLs under the Stage 1 D/DBPR is
determined as an average concentration for each treatment plant. Since the Phase 2 analyses
calculate system means, results should not be considered indicative of compliance under the
Stage 1 D/DBPR.
Six-Year Review 3 Technical Support Document	6-49
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Overall, the available new occurrence information for bromate and chlorite indicates that the
contaminants do occur at PWSs that have to monitor for them and detections above the MCL
occur as well. The Phase 2 analyses indicate that very few, if any, systems experienced
prolonged levels of bromate or chlorite above the MCL. Again, these occurrence estimates
should not be construed as indicative of compliance with the Stage 1 and 2 D/DBPRs.
6.3.4 Additional Considerations
EPA understands that the regulated contaminants represent only a portion of the entire universe
of DBPs and serve as proxies for the many other unregulated DBPs. Since the Stage 1 and 2
D/DBPRs, new information has become available on the relative toxicity (see Chapter 4) and
occurrence of some unregulated DBPs, including brominated and iodinated species, chlorate,
nitrogenous DBPs and halobenzoquinones.
6.3.4.1 Chlorate
Similar to nitrosamines, chlorate was included on CCL 3 and evaluated as part of Regulatory
Determinations for CCL 3. Occurrence data for chlorate were collected as part of UCMR 3,
which took place from 2013 through 2015. Detailed information on the occurrence analyses
conducted using UCMR 3 data are available in the Six-Year Review 3 Technical Support
Document for Chlorate (USEPA, 2016e).
Chlorate and chlorite are two different oxidation states of chlorine and are chemically inter-
convertible in water. While the potential common health effects of chlorate and chlorite are
discussed in Chapter 4, the co-occurrence of chlorate and chlorite in U.S. drinking water is
discussed in this section.
The data sources used to evaluate co-occurrence of chlorate and chlorite include: DBP ICR, the
Environmental Working Group (EWG) National Drinking Water Dataset and UCMR 3 data for
chlorate and SYR3 ICR data for chlorite.
DBP ICR
Under the DBP ICR, all systems serving 100,000 or more people and using chlorine dioxide or
hypochlorite solution were required to monitor chlorite and chlorate in finished water on a
monthly basis between January 1997 and June 1998 (USEPA, 2000e; McGuire et al., 2002). The
resulting dataset includes data from systems with a variety of primary disinfectants. EPA used
the dataset to extract 1,326 paired sets of monthly chlorite and chlorate monitoring results. The
samples in each pair were taken at the same time and at the same location, either at the entry
point to the distribution system or at a location in distribution system. A plot of the contaminant
concentrations in the paired samples, grouped by primary disinfectant type, is shown in Exhibit
6.34. In this analysis, non-detections are assigned a value of zero.
Six-Year Review 3 Technical Support Document	6-50
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Exhibit 6.34: DBP ICR Data: Paired Monitoring Results for Chlorate and Chlorite
2,500
HRL= 210ug/L
« 1,500
1,000 %
•	CL2 (N=67)
•	CLM (N=100)
•	CL02 (N=973)
•	03 (N=S4)
OUnknown (N=102)
MCL= 1,000 ng/L
J
•|
•O
1,000
%
Note: the Legend indicates primary disinfectant types
(#samples per disinfectant type)
1,500
Chlorate Levels ((ig/L)
2,000
2,500
The results show that there was simultaneous exceedance of the lowest chlorate Health
Reference Level (HRL) of 210 |ig/L and the chlorite MCL of 1,000 |ig/L at systems using
chloramines (in many cases presumably formed with use of hypochlorite solution) and systems
using chlorine dioxide. Note that an HRL is defined as a risk derived concentration against which
to compare the occurrence data from PWSs to determine if chlorate occurs with a frequency and
at levels of public health concern. Chlorate and chlorite also co-occurred at relatively high levels
(above the chlorate HRL of 210 ug/L and above one half the chlorite MCL or greater than 500
ug/L) when ozone was used as the primary disinfectant.
Note that the assignment of systems to primary disinfection categories was based on limited data
included in the DBP ICR dataset. Some systems (labelled "unknown") could not be categorized.
Some labelled as using chlorine, chloramines, chlorine dioxide or ozone as the primary
disinfectant may in fact have made use of a combination of disinfectants.
EWG National Drinking Water Dataset
The EWG National Drinking Water Dataset, posted online (http://www.ewg.org/tap-
water/chemical-contaminants/). includes a selection of water sampling data, covering the period
2004-2009, obtained from state water officials by EWG staff. From the EWG dataset, EPA was
able to extract 305 paired chlorate/chlorite records. These paired records are from 14 systems
(with customer bases ranging in size from 6,525 people to 289,000 people) in 5 states: Alabama,
California, Minnesota, New York and Virginia. Each pair represents a single day's average daily
chlorate concentration and average daily chlorite concentration at the system, as calculated and
Six-Year Review 3 Technical Support Document	6-51
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
reported by EWG (EWG included non-detections when calculating daily averages, assigning
them a value of zero. No information is reported by EWG about disinfection practices or
sampling locations in the distribution system. It is possible that the bulk of the EWG data may be
from systems that use chlorine dioxide that were monitoring for chlorite in compliance with the
MDBP rules. A plot of the paired concentrations, grouped by state, is shown in Exhibit 6.35.
Exhibit 6.35: EWG Data: Paired System Daily Averages for Chlorate and Chlorite
1,200
1,000
800
(*
HRL = 210 Ug/L
S 600
400
COD	?
200 @
3
o •
O o o
o
<3>
o #o cx^o o 1 oooKE>t> ckudooo arooooo arooo o
0	200	400	600	800
Chlorate Levels, ng/L
MCL= 1,000 ng/L
•	AL(N=12)
•	CA (N=18)
•	MM (N=56>
•	NV (N=2)
OVA (N =217)
1,000
1,200
The data show no daily average chlorite levels in excess of the MCL of 1,000 jig/L, which could
be attributable to compliance with the MCL under the Stage 1 D/DBPR. The distribution of daily
average chlorite levels is fairly wide (from non-detection to approximately 800 jig/L), regardless
of whether daily average chlorate levels exceed or fall below 210 |ig/L.
The EWG National Drinking Water Dataset has several limitations. It is a compilation of data
that EWG acquired from multiple sources; it is not a complete national dataset and cannot be
assumed to be representative of the nation's drinking water. The use of daily average
concentrations obscures some variability in the data. As noted above, there is no information
about sampling locations or disinfection practices associated with data in the dataset.
UCMR 3 (chlorate) and SYR3 ICR Dataset (chlorite)
The most robust and recent monitoring data on chlorate and chlorite occurrence are in the
UCMR 3 and SYR3 ICR dataset, respectively. EPA identified 73 systems that each had at least
one record in each dataset for both chlorate and chlorite. All of the SYR3 chlorite records were
from 2011. The UCMR 3 chlorate records were from 2013-2016. It is expected that most of the
73 systems employ chlorine dioxide as a primary disinfectant, as those are the systems required
Six-Year Review 3 Technical Support Document	6-52
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
to sample for chlorite. The highest chlorate and chlorite concentrations from each system are
plotted in Exhibit 6.36 below. Non-detections are assigned a value of zero. This analysis
indicates that systems reporting chlorite records below the MCL of 1,000 |ig/L may have
chlorate in concentrations significantly higher than the HRL of 210 |ig/L.
Exhibit 6.36: System Highest Chlorate and Chlorite Levels in UCMR 3 and SYR3
ICR Datasets (N = 73)
O

01
—I
01
;H
z
o
•C
u
4,000
3,500
3,000
2,500
2,000
1,500
1,000
500



















•








<—
HRL
210 ng/l











MCL
= 1,000 ng/L


•




1

¦;wv
•
•
•
•




* .1 •



—«
	
500 1,000 1,500 2,000 2,500 3,000
Chlorate Levels, |ig/L- UCMR 3 (July 2016 version)
3,500
4,000
It is important to acknowledge that there are several limitations to this analysis. Only samples
with the highest respective concentrations of chlorite and chlorate were selected for inclusion in
the analysis. Those samples were gathered in different timeframes (the SYR3 chlorite records
were from 2011, while the UCMR 3 chlorate records were from 2013-2016) and were not
necessarily taken at the same sampling point. These data provide only a crude picture of potential
co-occurrence of chlorate and chlorite in the 73 systems. Also, the 73 systems that each had at
least 1 record for both chlorate and chlorite cannot be considered nationally representative.
6.3.4.2 Nitrosamines
Five nitrosamines were included on the Third Contaminant Candidate List (CCL 3). Four
nitrosamines from CCL 3, as well as two other nitrosamines that were not on CCL 3, were later
part of the Second UCMR (UCMR 2) data collection effort and considered as part of Regulatory
Determinations for CCL 3 as candidates for a potential NPDWR. Detailed information on the
occurrence analyses conducted using UCMR 2 data are available in the Six-Year Review 3
Technical Support Document for Nitrosamines (USEPA, 2016d).
Six-Year Review 3 Technical Support Document	6-53
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
6.3.4.3 Disinfectant Residuals
In a separate but related effort, EPA evaluated disinfectant residual records (that were taken
during the same time and at the same locations as coliform samples) from the SYR3 ICR dataset
to understand what disinfectant types/levels are present upon coliform occurrence. This
information is available in the Six-Year Review 3 Technical Support Document for Microbial
Contaminant Regulations (USEPA, 2016a). Overall, analysis of the disinfectant residual data
indicated that very few records for both free and total chlorine exceeded the Maximum Residual
Disinfectant Level requirement of 4.0 mg/L.
Six-Year Review 3 Technical Support Document	6-54
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
7 Treatment
7.1 Introduction
This chapter discusses information about treatment to remove DBP precursors and DBPs that has
become available since the development of the Stage 2 D/DBPR. As with other aspects of the
SYR3, EPA limited its review to treatment information published through December 2015.
During the development of the Stage 1 and Stage 2 D/DBPRs, a variety of technologies were
evaluated for their effectiveness, applicability and unintended consequences relative to achieving
compliance with the treatment technique (TT) requirements and Maximum Contaminant Levels
(MCLs), as well as providing a basis for the Best Available Technology (BATs) (USEPA,
1998b. 2005g 2005n, 2006a, 2007).
Since the Stage 2 D/DBPR, the Agency has identified information that improves its
understanding of technologies available for lowering occurrence of and exposure to regulated
and unregulated DBPs. The information addresses the full spectrum of drinking water system
operations, including removal of organic precursors to DBPs, disinfection practices, source water
management and localized treatment. As discussed in Chapter 6, one new information source is
the SYR3 Information Collection Request (ICR) dataset. For Chapter 7, EPA analyzed this
dataset to inform the extent to which total organic carbon (TOC, as an organic precursor
surrogate) was removed from source water per the TT requirement under the Stage 1 D/DBPR.
These TOC results, along with new literature on treatment technologies, could help improve the
understanding of not only the technologies that were considered during the development of the
Stage 2 D/DBPR (i.e., enhanced coagulation, use of alternative disinfectants, granular activated
carbon (GAC) and membranes), but also those technologies not included before (e.g.,
biofiltration, localized post-treatment and source water management). EPA analyzed this new
information to assess the applicability, effectiveness and unintended consequences of these
individual treatment technologies. Overall, the information collectively indicates that: (1) greater
removals of DBP precursors are being achieved than were achieved prior to the Stage 1
D/DBPR; and (2) occurrence of DBPs can be further controlled.
This chapter is organized as follows:
Section 7.2, "Background on Treatment Technologies Considered for the Stage 1 and Stage 2
D/DBPRs," provides a brief overview of the existing TT requirement included in the Stage 1
D/DBPR for removal of TOC. This section also summarizes the treatment technologies
considered during the development of the Stage 1 D/DBPR and Stage 2 D/DBPR, respectively.
Section 7.3, "Information on Reducing DBP Formation Potentials in Treatment Plants," includes
an analysis of SYR3 ICR data on TOC removal. It also discusses the information reviewed
during the SYR3 process on: 1) conventional treatment, 2) non-conventional treatment and 3)
potential add-on physical unit processes.
Section 7.4, "Information on Source Water Management," covers literature on potential
approaches for lowering DBP formation potential in source water (e.g., source water
management, bank filtration, pre-sedimentation or pre-oxidation).
Six-Year Review 3 Technical Support Document	7-1
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Section 7.5, "Information on Changing Disinfection Practices in Treatment Plants and
Distribution Systems," focuses on formation and occurrence of different DBP groups with use of
different disinfection practices (including different disinfectant types) and discusses potential
implications of changes in disinfectant types.
Section 7.6, "Information on Removing DBPs after Formation in Treatment Plants and/or
Distribution Systems," provides information on methods for removal of DBPs after their
formation. These methods (e.g., aeration) may be applicable in treatment plants or in distribution
systems.
Appendix C provides additional information on several of the topics presented in this chapter.
7.2 Background on Treatment Technologies Considered for the Stage 1 and Stage 2
D/DBPRs
The main purpose of the Stage 1 and Stage 2 D/DBPRs is to reduce exposure to DBPs while
maintaining protection against microbial risks in public water systems (PWSs). During the
development of the Stage 1 D/DBPR, EPA determined that it was necessary to control for
organic matter in source water through a TT requirement. This TT requirement complemented
the MCLs and was designed to help remove DBP precursor material to help reduce the risks
posed by DBPs. This section briefly describes the TT requirement and the treatment technologies
considered during development of the Stage 1 D/DBPR and the Stage 2 D/DBPR, respectively.
7.2.1 Treatment Technique Requirements for TOC Removal
As described in Chapter 6, under the Stage 1 D/DBPR, PWSs using surface water or ground
water under the direct influence of surface water (GWUDI) sources and using conventional
treatment (i.e., "a series of processes including coagulation, flocculation, sedimentation and
filtration resulting in substantial particulate removal") are required to remove specified
percentages of TOC from the source water. TOC removal is achieved with enhanced coagulation
or enhanced softening unless a system meets one of several alternative compliance criteria. This
TT applies to community and non-transient non-community water systems (NTNCWSs) of all
sizes.
The TT requirement identifies the minimum percentage of TOC a conventional plant must
remove based on the raw water TOC and alkalinity levels, which are divided into three ranges,
respectively. These criteria are referred to as the "3x3 matrix" and are shown in Exhibit 7.1
(USEPA, 1998b).
Six-Year Review 3 Technical Support Document	7-2
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Exhibit 7.1: Required TOC Removal for Conventional Treatment Plants Using
Surface Water or GWUDI1,2,3
Source water TOC, mg/L
Source water alkalinity, mg/L as CaCC>3
0-60
60-120
>1204
>2.0-4.0
35.0%
25.0%
15.0%
>4.0-8.0
45.0%
35.0%
25.0%
>8.0
50.0%
40.0%
30.0%
Notes:
1	Plants meeting at least one of the alternative compliance criteria are not required to operate with
enhanced coagulation.
2	Softening plants meeting one of the alternative compliance criteria are not required to operate with
enhanced softening.
3	Compliance with the TOC removal requirement is based on a running annual average, computed
quarterly.
4	Plants practicing softening must also meet the TOC removal requirements in this column.
EPA developed the 3x3 matrix recognizing that systems would have a greater challenge
removing TOC from source waters with high alkalinity. Some types of water may not be
amenable to effective TOC removal by coagulation or softening. Alternative compliance criteria
included in the Stage 1 D/DBPR provide flexibility for complying with the TT requirements.
Those alternative criteria are described in EPA's Enhanced Coagulation Guidance Manual
(USEPA, 1999b).
7.2.2 Treatment Technologies Considered During Rule Development
Exhibit 7.2 collectively lists the treatment technologies included in the Stage 1 D/DBPR
Regulatory Impact Analysis (RIA), along with those used in the Stage 2 D/DBPR Economic
Analysis (EA). The Stage 2 D/DBPR EA (USEPA, 2005g) and its appendices (USEPA, 2005n),
along with the Technologies and Costs Document for the Stage 2 D/DBPR (USEPA, 2005m) and
the Simultaneous Compliance Guidance Manual for LT2ESWTR and Stage 2 D/DBPR (USEPA,
2007b) present a detailed description of these technologies, including their effectiveness,
applicability and unintended consequences. Water Research Foundation (WRF) studies
published since promulgation of the Stage 2 D/DBPR contain similar lists of technologies
(Schendel et al., 2009 and Becker et al., 2013).
Exhibit 7.2: Treatment Technologies Considered for the Stage 1 and Stage 2
D/DBPRs1
Stage 1 D/DBPR RIA
Treatment Technologies
Stage 2 D/DBPR EA
Treatment Technologies
Chlorine/Chloramine
Adjust Primary Disinfection
Move Points of Disinfection with Chloramines
Enhanced Coagulation
Enhanced Coagulation with Chlorine
Turbo Coagulation with Chlorine
Enhanced Coagulation with
Chloramines
Enhanced Coagulation with Chloramines
Turbo Coagulation with Chloramines
Six-Year Review 3 Technical Support Document	7-3
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Stage 1 D/DBPR RIA
Treatment Technologies
Stage 2 D/DBPR EA
Treatment Technologies
Chlorine Dioxide
Chlorine Dioxide with Chlorine
Chlorine Dioxide with Chloramines
Ozone with Chloramines
Ozone with Chlorine
Ozone with Chloramines
GAC 10
GAC10 with Chlorine
GAC10 with Chloramines
GAC10 + Chlorine Dioxide with Chlorine
GAC10 + Chlorine Dioxide with Chloramines
GAC 10 + UV (Small Systems)
GAC20
GAC20 with Chlorine
GAC20 with Chloramines
GAC20 + Chlorine Dioxide with Chlorine (Large and Medium Systems)
GAC20 + Chlorine Dioxide with Chloramines (Large and Medium Systems)
GAC20 + Ozone with Chlorine (Small Systems)
GAC20 + Ozone with Chloramines (Small Systems)
GAC20 + UV (Small Systems)
Membranes
Microfiltration/Ultrafiltration with Chlorine
Microfiltration/Ultrafiltration with Chloramines
Integrated Membranes with Chlorine (Surface Water Systems)
Integrated Membranes with Chloramines (Surface Water Systems)
Nanofiltration with Chlorine (Ground Water Systems)
Nanofiltration with Chloramines (Ground Water Systems)
1 Source: Exhibit A.7 in Appendix A of the Stage 2 D/DBPR EA (USEPA, 2005n).
7.3 Information on Reducing DBP Formation Potentials in Treatment Plants
The treatment technologies listed in Exhibit 7.2 include enhanced coagulation, granular activated
carbon (GAC) and membranes; they are intended to reduce DBP formation in treatment plants.
Enhanced coagulation is an enhanced mode of operation that assumes possible adjustments in
coagulant application and pH to achieve the minimum target TOC levels through a combination
of coagulation and sedimentation basins followed by filtration.
Exhibit 7.3 shows the percent TOC removal by surface water filtration treatment plant types
from source to filter effluent. It is based on paired TOC data from the DBP ICR dataset. As
indicated in Exhibit 7.3, prior to the Stage 1 D/DBPR, a conventional treatment train was the
most common type of treatment for surface water systems serving 100,000 or more people.
Systems with direct filtration, in-line filtration and slow sand filtration tended to have much
lower TOC levels in their source water and were not subject to the TT requirement under the
Stage 1 D/DBPR.
Six-Year Review 3 Technical Support Document	7-4
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Exhibit 7.3: Percent TOC Removal from Source to Filter Effluent by Surface Water
Filtration Treatment Plant Types Based on DBP ICR Dataset
Plant Type
Plant Type
Code
Number of
Plants
Mean Plant
Average
Raw Water
Turbidity, NTU
Mean Plant
Average
Raw Water
TOC, mg/L
Mean Plant
Average
Filtered Water
TOC, mg/L
Mean Plant
Average
%TOC
Removal1
Conventional/
Softening2
CONV/
SOFT2
272
3.5
3.5
2.2
31.2%3
Direct Filtration
DF
22
2.1
2.5
2.0
17.6%
In-Line Filtration
ILF
5
1.3
1.7
1.3
11.4%
Slow Sand
Filtration
SSF
2
1.1
1.7
1.2
28.8%
Notes:
1	%TOC removal from source water to filter effluent.
2	"Conventional/Softening" includes plant type codes in the DBP ICR database: Conv; CMPLX/SOFT; CS/SOFT;
SOFT; SPLIT/SOFT; and TS/SOFT.
3	About 24% and 8% on average came from coagulation/sedimentation and filtration, respectively.
The TOC monitoring data in the SYR3 ICR dataset enables EPA to evaluate TOC removal
relative to the 3x3 matrix criteria. This section contains analytical results on TOC removal using
SYR3 ICR data, followed by discussion of the information available since the Stage 2 D/DBPR
on conventional treatment, non-conventional treatment and potential add-on physical removal
unit processes, for reducing DBP formation potentials. Since the information presented and
discussed in this section is relatively lengthy, a summary for this section is provided below.
The analytical results from the SYR3 ICR dataset indicate a wide range of percent TOC removal
observed for each cell of the 3x3 matrix, as was anticipated when the requirements were
promulgated. The mean removal in each category of the 3x3 matrix was 6 to 19 percent higher
than the TT requirement, indicating that greater removals of DBP precursors were commonly
being achieved compared to the TT requirement. These observations are consistent with the
notion that "since the Stage 1 D/DBPR does not require that all coagulable dissolved organic
matter be removed, there is a potential for additional removal of organic matter beyond that
required by the 3x3 matrix." (McGuire et al., 2014).
Some of the TOC removal observed greater than the minimal TOC removal requirement may
reflect operational optimization of conventional treatment, including use of innovative
coagulants/coagulant aids and/or use of biofiltration. Application of biofiltration recently has
become a key research area in the water industry and there are several ongoing studies (e.g., the
biofiltration-related projects listed on the WRF website, including project numbers 4496, 4525,
4555, 4559 and 4620) that could further inform the applicability, effectiveness and unintended
consequences for use of biofiltration. Studies have shown that biological filtration can also
reduce precursors of DBPs other than THM4/HAA5 in many, though not all cases (Mitch et al.,
2009; Liao et al., 2014; Krasner et al., 2015). As noted by McGuire et al. (2014), if the removal
of precursors for DBPs other than THM4/HAA5 becomes part of the treatment goals, then
performance parameters in addition to TOC may also be needed (e.g., parameters indicating both
vulnerability and nitrosamine formation potential).
Six-Year Review 3 Technical Support Document	7-5
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
As was known during development of the Stage 1 and the Stage 2 D/DBPRs, GAC and
membranes can be added to existing treatment trains to achieve additional reductions of DBP
formation potential. One longstanding issue has been the extent to which organic precursor
removal may cause a shift of chlorinated species to more brominated species when the bromide
level is relatively high in source water (Summers et al., 1993; Symons et al., 1993). The ICR
Treatment Study database (USEPA, 2000f) provides extensive bench- and pilot-scale data by
which to evaluate the effects of GAC and membrane removal of TOC and resulting shifts in
BrTHMs. EPA's recent analysis of these data generally shows increased percent reduction of
BrTHMs as TOC removal by GAC increases (e.g., from a target effluent level of 2 mg/L to 1
mg/L) for source waters with high bromide concentrations. It also shows that bromoform
formation increases as bromide concentrations increase and that bromoform becomes the
dominating species when source water bromide concentrations exceed 200 |ig/L.
7.3.1 Analysis of SYR3 ICR Data for TOC Removal
This section presents analytical results of the SYR3 ICR data within the context of the 3x3
matrix. Appendix C of this document contains the background/inventory information,
supplemental analytical results of the paired SYR3 ICR TOC dataset and details on the creation
of the "paired TOC dataset." The main observations from the analytical results are summarized
below:
•	The data show a wide range of percent TOC removal for each combination of raw water
TOC and alkalinity levels provided in the Stage 1 D/DBPR TT requirement. The data
also indicate that the mean removal for each element of the 3x3 matrix was 6 to 19
percent greater than the requirement.
•	In the context of the 3x3 matrix, although TOC removal generally increased as the raw
water TOC levels increased, the treated water TOC levels generally still increased as the
raw water TOC levels increased. When the raw water TOC levels were greater than 8
mg/L, nearly all the plants had mean treated water TOC levels above 2 mg/L and it was
not uncommon to see the treated water TOC levels greater than 4 mg/L.
•	Regarding system sizes, while the levels of raw water TOC and alkalinity appeared
essentially no different (i.e., were independent of system size), percent TOC removals
among small systems (those serving <10,000 people) were slightly lower than in
medium systems (serving between 10,000 and 100,000) and large systems (serving
>100,000) (41 percent mean removal in small systems versus 44 percent and 45 percent
in medium and large systems, respectively).
7.3.1.1 Analytical Approach
Under the existing TT requirements, some systems may take more than one pair of TOC samples
per month and compute an average of the monitoring results each month for compliance
calculation. Compliance with the TOC removal requirements is based on a running annual
average, computed quarterly. Changes in raw water TOC and/or alkalinity levels from month to
month will cause some plants to move from one category to another in the 3x3 matrix (see
Exhibit 7.1). The required TOC removal, therefore, may change on a month-to-month basis.
Six-Year Review 3 Technical Support Document	7-6
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Such a regulatory construct makes the monthly-level analysis of the SYR3 ICR paired TOC data
more complex. To simplify the data analysis, annual averages per plant (i.e., facility in the
dataset) per calendar year were calculated, using the monthly average values for raw water TOC,
raw water alkalinity and treated water TOC. Annual average removals (percentages) of TOC
were calculated with the annual average values of raw water TOC and treated water TOC per
facility.
7.3.1.2 Summary Statistics for 3x3 Matrix
To maximize the number of records included in the data analysis, all years of data were included.
In this context, the term "Facility Years" (i.e., facilities x years) was used. It should be noted that
the use of multiple years rather than the most recent single year (i.e., 2011) can lead to an
underestimate of the levels of TOC removal achieved by the implementation of Stage 1, as 2011
shows the higher percent removal (on average) and also likely reflects the highest degree of
Stage 1 implementation (See Appendix C for more discussion). The summary statistics
associated with each TOC/alkalinity category of the 3x3 matrix are shown in Exhibit 7.4. These
statistics (based on an annual average per facility year) include the following analytical
endpoints:
(1)	The count of facility years (i.e., #Facility Years),
(2)	Percentages of facility years with percentage of TOC removal less than that required for
each of the TOC/alkalinity categories in the 3x3 matrix (i.e., %Facility Years with %
Removal < Required),
(3)	The mean, median and 90th percentile of TOC removal (i.e., Mean/Median and 90th
Percentile Removal),
(4)	Percentage of facility years with treated water TOC levels greater than 2 mg/L (i.e.,
%Facility Years with Treated TOC > 2 mg/L), and
(5)	The mean of the treated water TOC levels (i.e., Mean Treated TOC, mg/L).
These analytical endpoints were selected to represent the distribution and variation of TOC
removals in each category of the 3x3 matrix for facility years when the annual average raw water
TOC levels were greater than 2 mg/L. All facility years included in the dataset were associated
with surface water systems. Appendix C of this document shows a similar analysis for annual
average raw TOC levels < 2 mg/L (for which 99 percent of facility years included in the dataset
were associated with surface water systems).
Six-Year Review 3 Technical Support Document	7-7
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Exhibit 7.4: Evaluation of TOC Compliance Monitoring Data from SYR3 ICR
Dataset Relative to 3x3 Matrix (Based on Paired TOC Data from 2006-2011)
Raw Water
TOC, mg/L
Summary1
(Total #Facility Years = 4,793)
Raw Water Alkalinity, mg CaCOs/L
0-60
>60 to 120
>120
2.0 < TOC <4.0
#Facility Years
1,735
915
510
%Facility Years with %Removal < Required
27.8%
16.4%
9.0%
Mean Removal
41.7%
35.2%
30.4%
Median Removal
41.6%
35.1%
30.1%
90th Percentile Removal
56.2%
49.2%
47.2%
%Facility Years with Treated TOC > 2 mg/L
14.3%
26.0%
44.3%
Mean Treated TOC, mg/L
1.6
1.8
2.0
4.0 < TOC <8.0
#Facility Years
739
322
366
%Facility Years with %Removal < Required
15.6%
12.1%
4.9%
Mean Removal
54.7%
46.8%
44.1%
Median Removal
54.3%
46.3%
43.9%
90th Percentile Removal
70.0%
58.3%
61.8%
%Facility Years with Treated TOC > 2 mg/L
77.4%
91.9%
91.8%
Mean Treated TOC, mg/L
2.5
2.9
3.0
TOC > 8.0
#Facility Years
129
35
38
%Facility Years with %Removal < Required
7.0%
25.7%
2.6%
Mean Removal
66.2%
46.3%
46.9%
Median Removal
66.4%
44.2%
47.8%
90th Percentile Removal
82.2%
67.3%
63.9%
%Facility Years with Treated TOC > 2 mg/L
85.3%
100.0%
100.0%
Mean Treated TOC, mg/L
3.5
5.6
6.2
Note: Facility Years = number of facilities x number of years for the paired TOC data between 2006 and 2011.
As described in Section 7.2.1, some systems could be using some of the alternative criteria; thus,
the values of "%Facility Years with %Removal < Required" cannot be assumed to be equivalent
to the percentage of facility years with a TT violation. For instance, "%Facility Years with
%Removal < Required" is 27.8 percent (i.e., 482 facility years). This can be attributable to a
significant number of the facility years (i.e., 302) among these 482 facility years that have treated
water TOC levels less than 2 mg/L or might have elected to meet alternative criteria and be
exempted from meeting the removals specified in 3x3 matrix. The extent to which the facilities
with treated water TOC levels less than 2 mg/L actually used the alternative criteria is unknown.
Because of this uncertainty, the analytical results in the upper three boxes of the 3x3 matrix (i.e.,
for 2.0 < TOC < 4.0 mg/L) were not included in the remainder of this discussion.
Six-Year Review 3 Technical Support Document	7-8
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
A wide range of TOC removal was observed in each TOC/alkalinity category of the 3x3 matrix,
from removal percentages below the requirement to some achieving 25 percent more than the
requirement. Overall, the means and medians of removal are 6 to 19 percent more than that
required in each category included in the 3x3 matrix (refer to mean and median removals in
Exhibit 7.4 versus required removals in Exhibit 7.1). For example, in looking at the category
where TOC is in the range of 4.0 to < = 8.0 mg/L, with alkalinity of 0-60 mg/L, Exhibit 7.4
shows mean and median removals of about 54 percent, while Exhibit 7.1 shows a required
removal of 45 percent, corresponding to 9 percent more than the requirement for that category.
In addition, a comparison between the value of "90th percentile Removal" and the removal
required in each of the middle and bottom boxes indicates some facilities achieved significantly
higher removal than required (e.g., 70 percent vs 45 percent).
These observations are consistent with the notion that "since the Stage 1 D/DBPR does not
require that all coagulable dissolved organic matter be removed, there is a potential for additional
removal of organic matter beyond that required by the 3x3 matrix." (McGuire et al., 2014). As
discussed later, a TOC removal in some plants for a given category of the 3x3 matrix can be
attributable to the treatability of the water and/or an operational optimization of the conventional
treatment trains. This can be achieved with operation in a "turbo" enhanced coagulation mode
(as defined in the Economic Analysis of Stage 2 D/DBPR, USEPA, 2005g) or by following
enhanced coagulation with biofiltration.
7.3.1.3 TOC Removal by System Size
The SYR3 ICR data enables EPA for the first time to evaluate TOC removal at a national scale
among systems of different sizes. For this purpose, the systems with the paired TOC data were
grouped into three population size categories: < 10,000, 10,000-100,000 and > 100,000. As
indicated by Exhibit 7.5, small systems (< 10,000) removed slightly less TOC than medium
(10,000 -100,000) and large (> 100,000) systems—the mean removal for small systems was 41
percent versus 44 percent and 45 percent for medium and large systems, respectively. However,
the top 20 percent of performers across all system size categories achieved greater than 50
percent removal. The distributions of raw water TOC and alkalinity levels by system size are
included in Appendix C of this document; the difference in raw water values between systems of
different sizes is relatively small.
Six-Year Review 3 Technical Support Document	7-9
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Exhibit 7.5: TOC Removal by System Size from SYR3 ICR Dataset (Based on
Paired TOC Data from 2006-2011)
¦c
OJ
OJ
c
8
k_
= 100k (N=516)






































Summary Statistics (TOC Removal)





<10k
10-100k i 100k





10%iles
Mean
22.5% 27.6%
41.0% 44.0%
40.6% 43.4%
28.9%
45.3%
46.2%





Median
^wrl




90%iles
60.1% 61.3%
61.0%
0%
10%
20%
30% 40% 50% 60% 70%
Annual Average TOC Removal
80%
90%
100%
Exhibit 7.6: Treated Water TOC Levels by System Size from SYR3 ICR Dataset
(Based on Paired TOC Data from 2006-2011)
. <10k (N=2550)
* 10k - 100k (N=l 723)
*>= 100k (N=516)
in
fi
u
<3
 100k
10%iles	1,3	1.3 1.4
Mean	2.2	2.0 2.1
Median	2.0	1.8 1.8
90%iles	3.5	2.8 3.2
4	6	8	10	12	14
Annual Average Treated Water TOC, mg/l
16
18
7.3.1.4 Limitations of Paired TOC Data from SYR3 ICR
As indicated in Appendix C of this document, there are 21 states included in the paired TOC
dataset. While this dataset is substantial, EPA is not able to assess the completeness of the paired
TOC data records among these 21 states since it did not have: 1) the state inventory number of
Six-Year Review 3 Technical Support Document	7-10
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
facilities (or plants) with conventional treatment trains and 2) information about the application
and use of alternative criteria for TOC removal and record management. With respect to the
national representativeness of the TOC paired dataset, EPA also notes that some "big" states
(i.e., ones with a relatively large number of systems serving relatively large populations, and in
some cases, relatively high TOC levels in source water) are not included in the dataset, including
California, Texas and Florida.
In addition, the DBP ICR dataset indicates that, in general, ground water (GW) systems have
much lower source and finished water TOC levels (see Exhibit 6.14 and 6.15 versus Exhibits
6.12 and 6.13) than surface water systems (SW). However, some of these systems using
conventional treatment with or without softening (mostly Florida systems) had TOC levels
comparable to SWs with high TOC levels. The paired TOC dataset from the SYR3 ICR
essentially only included SW systems and provided little additional TOC occurrence data for
GW systems.
EPA's understanding is that the paired TOC data from the SYR3 ICR for SW systems is the
largest and most comprehensive dataset (since the DBP ICR dataset in 1997-1998) to indicate, at
a national level, treatment performance among plants for TOC removal and TOC levels in
treated water.
7.3.2 Information on Conventional Treatment
EPA does not have recent information on the number of water systems using conventional
treatment. However, at the time of the DBP ICR, the majority of surface water treatment plants
serving 100,000 or more people were conventional treatment plants (including the ones with
softening) (i.e., 272 out of 301 filtration plants, see Exhibit 7.3). The AWWA Disinfection
Committee (AWWA, 2000a and 2000b) reported that the small and medium SW systems also
commonly used a conventional treatment process. In addition, as indicated in Exhibit 6.14, the
DBP ICR data also showed that those ground water systems with relatively high TOC in their
source water also used conventional treatment (with or without softening).
As described in Section 7.2, the surface water conventional treatment plants that are required to
implement the TOC removals specified in the 3x3 matrix must monitor TOC in the source water
prior to any treatment, including oxidant addition. The treated water TOC also must be
monitored no later than the combined filter effluent turbidity monitoring location. Thus, the TOC
removal results from the SYR3 ICR data presented in Section 7.3.1 reflect the collective
treatment performance of the three individual treatment units (i.e., coagulation/flocculation,
sedimentation and filtration) in the conventional treatment plants. Thus, the removal of organic
matter by a conventional treatment train depends on the operating conditions for each of these
units, given source water quality.
Many plants were achieving higher percentages of TOC removal than required during the SYR3
ICR period. Such operation of enhanced coagulation may be due to what was referred to as
"turbo" enhanced coagulation in the Economic Analysis for the Stage 2 D/DBPR (see Exhibit
7.2). Also, new studies indicate that additional TOC removal can be achieved by operating the
filtration unit in a biological mode (Liao et al., 2014, 2015; Delatolla et al., 2015; Pharand et al.,
2015). This approach for operating a conventional treatment plant may enable an additive or
Six-Year Review 3 Technical Support Document	7-11
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
synergic performance of "turbo" enhanced coagulation and biofiltration, collectively resulting in
greater TOC removals.
7.3.2.1 Enhanced Coagulation
This section focuses on new information on coagulants and coagulation aids. Aside from source
water quality conditions, removal of organic matter through enhanced coagulation depends on
multiple operating factors, including pH, coagulant type and dose, coagulation aid type and dose
and hydraulic conditions in both coagulation and sedimentation basins. Hydraulic conditions and
pH are well understood from information collected during development of the Stage 1 and Stage
2 D/DBPRs, thus the focus on coagulants and coagulation aids here.
Most common coagulants are aluminum or ferric salts and TOC removal to some extent can be
increased by increasing the coagulant dose. A new coagulant is polyaluminum chloride (PAC).
Its use, in lieu of aluminum chloride or ferric chloride, could improve the efficiency of enhanced
coagulation at certain pH ranges (Hassan et al., 2010). For instance, TOC/DOC/UV254 removal
could be 10 percent more and resultant reduction of TTHM formation potential could be 20-30
percent more when PAC is used instead of ferric chloride (Hassan et al., 2010).
Tzoupanos and Zouboulis (2009) developed a composite coagulant by introducing a cationic
polyelectrolyte (CPE) into PAC. They observed more efficient coagulation, compared to the
independent applications of CPE and PAC, due to more effective particle aggregation and
reduction of overall CPE dosage.
The effectiveness of PAC for enhancing coagulation was also evaluated in several studies. Some
of these studies (Kristiana et al., 2011; Dunn and Knappe, 2013; Chu et al., 2015; Watson et al.,
2015; Plourde-Lescelleur et al., 2015) collectively demonstrated that the addition of PAC could
help with removal of both chlorination DBPs and their precursors as part of DOC. As Hanigan et
al. (2015) found PAC to be effective for removing NDMA precursors, Chu et al. (2015) showed
that the use of PAC also resulted in a significant reduction of nitrogenous DBPs (including
NDMA). Lin et al. (2015) found that PAC worked best with small molecular weight DOC. It is
worth noting that since PAC was ineffective at reducing bromide levels, enhanced coagulation
with PAC could result in a shift to more brominated DBPs as the Br":DOC ratio increases
(Watson et al., 2015).
Jiang and Wang (2004) demonstrated that potassium ferrate could perform better than ferric
sulfate for treating waters containing humic and fulvic acids for reducing UV254 absorbance,
removing dissolved organic carbon (DOC) and lowering the THM formation potential. A later
literature review by Darko et al. (2014) confirmed that the efficiency of ferrate for removing
dissolved organic matter was higher than that of traditional coagulants ferric and aluminum salts.
Ferrate ion initially can act as a strong oxidant (potentially as a disinfectant as well) and then a
coagulant after being converted to ferric ion. A study from Lim and Kim (2009) showed that the
removal rate of humic acid using ferric sulfate was improved by pretreatment with a very small
dose of ferrate. The reaction between ferrate and humic acid was completed within a minute.
However, engineering aspects of ferrate generation and any unintended consequences associated
with a field application of ferrate in water treatment may need to be further characterized.
Six-Year Review 3 Technical Support Document	7-12
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Several new coagulants have been identified and tested. Jarvis et al. (2012) compared DOC
removals using a novel zirconium oxychloride-based coagulant (Zr-Coag®) to removal using
ferric sulfate and alum in batch (jar tests) and pilot scale experiments. Results showed greater
DOC removal and lower THM4 formation potential for the Zr-Coag® treated water (100.7 +/-
15.0 |ig/L) compared to THM formation potential after ferric sulfate treatment (163.1 +/- 36.7
|ig/L), Jar test data revealed an optimum Zr-Coag® dose of between 5 and 15 mg/L at a pH of 5
to 6. A limitation of this work is that it was conducted using one source water with low turbidity
(3.5 NTU) and low alkalinity (<10 mg/L).
Organic polymers are commonly used coagulation aids. Since Wilczak et al. (2003) reported that
some polymers (such as polyDADMAC) could contain organic nitrogen and could contribute to
the NDMA precursor material, investigation of coagulation aid alternatives has been ongoing
(Cornwell et al., 2015). The Six-Year Review 3 Technical Support Document for Nitrosamines
(USEPA, 2016d) presented more detailed discussion on NDMA formation related to use of
polymers.
In addition to raw water TOC and alkalinity levels, bromide in the raw water could also be
important to DBP formation potential in treated water, particularly for brominated DBPs (which
appear more toxic than chlorinated DBPs; see Chapter 4). Studies from Kalscheur et al. (2006)
and Watson et al. (2015) showed that TOC removal did not necessarily translate to a proportional
reduction in brominated DBP formation potential in treated water. Kalscheur et al. (2006)
reported that lime softening of source water with bromide concentrations of around 160 mg/L
(three orders in magnitude higher than typical U.S. water, USEPA, 2005g) resulted in a
significant shift to brominated DBPs and actually increased THM4 formation. Similarly, Watson
et al. (2015) found that enhanced coagulation resulted in a shift to more brominated DBPs as the
ratio of bromide to DOC increased after treatment.
Overall, the information reviewed as part of the SYR3 indicates that in some cases TOC may not
be an adequate performance indicator for the enhanced coagulation process, for brominated DBP
or NDMA formation potentials. Nevertheless, the new development and application of these
innovative materials can help improve the performance of enhanced coagulation in terms of
reduced DBP formation potentials.
7.3.2.2 Biological Filtration
Biofiltration (such as slow-sand filtration) has been used for water treatment for more than 100
years (Collins et al., 1992). Operating the filtration unit in a conventional treatment plant in a
biological mode was not included in the technologies considered in the compliance decision tree
for the Economic Analysis for the Stage 2 D/DBPR (see Exhibit 7.2), due to the lack of
recognized full-scale experience and accepted design and operating parameters. Since the
promulgation of the Stage 2 D/DBPR, there has been an increased interest in applying biological
filtration to remove organic DBP precursors and trace contaminants in drinking water. The WRF
has identified biological filtration as a focus area and set a goal to "determine biofiltration
effectiveness at removing contaminants, define benefits and communicate to key stakeholders
and to provide utility guidance on optimizing biofiltration" (WRF, 2015a). WRF has initiated
numerous projects pertaining to biological filtration for removal of organic DBP precursors. The
Six-Year Review 3 Technical Support Document	7-13
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
WRF also has established the "North American Biofiltration Knowledge Base" (WRF, 2015b) to
share fundamental knowledge on: 1) the use of biological filtration in water treatment and 2) the
field operational and monitoring data from some utilities.
The 1998 DBP ICR data showed that the percent TOC removal through the filters following the
coagulation/sedimentation basins in the SW conventional treatment plants was 8 percent on
average (see the footnote of Exhibit 7.3). Lauderdale et al. (2014) found that biological filters
commonly removed 10-20 percent of organic carbon, while removals had been reported to vary
from 5 percent to 75 percent.
Some studies have found that GAC biofilters can remove more DBP precursors than
anthracite/sand biofilters (Lauderdale et al., 2014; McKie et al., 2015; Azzeh et al., 2015;
Chowdhury et al., 2010). With all of these medium types, biological filtration has been shown to
reduce THM and HAA formation potentials by higher percentages compared to DOC removal
percentages (McKie et al., 2015; Azzeh et al., 2015; Delotolla et al., 2015; Pharand et al., 2015).
While ozonation prior to biofiltration has not been demonstrated to significantly increase DOC
removal, biofiltration is known to better remove assimilable organic carbon produced by
ozonation (Krasner et al., 2012; Pharand et al., 2015). Adding nutrients has not been found to
significantly enhance biofiltration performance (McKie et al. 2015; Azzeh at al. 2015;
Lauderdale et al. 2014).
Researchers have also investigated biological filtration for its removal of precursors of
unregulated DBPs. Some studies have shown that biological filtration can reduce NDMA
precursors in many, but not in all cases. During those studies, a reduction of NDMA precursors
after biological filtration with pre-ozonation was observed (Sacher et al., 2008; Farre et al., 2011;
Liao et al., 2014). Particularly, in studies of biological filtration with GAC media, Liao et al.
(2014) found that NDMA precursor removal was greater than DOC removal. Another study
found that while NDMA formation was reduced by biological filtration in some plants, in many
plants biofiltration led to an increase in NDMA formation potentials, which probably was caused
by sloughed bacteria or soluble microbial products (Krasner et al., 2015). Thus, if removing the
precursors of DBPs in general becomes part of the treatment goals, it may be necessary to
monitor the performance of parameters other than TOC (McGuire el al., 2014).
In addition to potential removal of DBP precursors or reduction in DBP formation potentials,
biofiltration may also be capable of reducing levels of organic DBPs, particularly when GAC is
used as a medium (Wu and Xie, 2005; Johnson et al., 2009; Lou et al., 2014). By studying the
effects of empty bed contact time (EBCT) and water temperature on the removal of HAAs in a
biological activated carbon (BAC) filter with 8 days of running time, Wu and Xie (2005)
suggested a 10 minute EBCT for 4°C water and a 5 minute EBCT for water at 10°C or higher to
achieve an HAA removal efficiency of 50 percent or higher. Lou et al. (2014) demonstrated that,
at 10-60 minutes of EBCT and 24-26 °C, 30-50 percent removal of THM and greater than 80
percent removal of HAAs could be achieved through a pilot-scale BAC filter over 9-11 days.
The BAC filter was included by Johnson et al. (2009) as part of their study for localized
treatment of DBPs in distribution systems. They concluded that development of biological
activity in the GAC column could significantly prolong the unit operation. Similar to a GAC
column, however, a BAC filter would remove the chlorine residual completely, resulting in the
need to rechlorinate downstream in the distribution system.
Six-Year Review 3 Technical Support Document	7-14
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Upstream treatment can also affect biofiltration performance. For instance, McKie et al. (2015)
found that PAC coagulant reduced bioactivity on filters, possibly due to a reduction in available
phosphorus (as a nutrient). Thus, effective integration of enhanced coagulation and biofiltration
in a conventional treatment train will be vital for maximizing removal of organic matter and
control of DBP mixtures in the treated water. Sohn et al. (2007), by examining DOC removal
after individual treatment processes in a plant with coagulation, a sand filter, ozone and a
biological filter, observed that: 1) both coagulation and ozonation units removed large-
molecular-weight organic compounds better than small ones and 2) biological filtration removed
small organics better than large ones. Chu et al. (2015) investigated the overall performance of a
conventional water treatment process followed by ozonation and biological activated carbon
filtration and showed significantly higher removals of both DOC and organic nitrogen, as
compared to the conventional treatment alone.
A potential drawback of converting a traditional rapid rate filter to a biological filter is that
extracellular polymeric substances from the biological community can contribute to an increased
head loss and fouling of filter underdrains. Azzeh et al. (2015) and Lauderdale et al. (2012)
found that application of hydrogen peroxide at low doses (< 1.0 mg/L) could reduce the head
loss by up to 45 percent without compromising biological performance, although higher doses
were found to negatively impact the DBP precursor removal performance. Lauderdale et al.
(2012) concluded that the optimal hydrogen peroxide dose is site-specific and dependent on
multiple factors, such as temperature, source water and upstream treatment. Another potential
issue is that if systems convert an active filter to biofiltration by removing the chlorine residual
entering the filter, oxidized manganese that has built up on the filter media may be reduced and
released (Kohl and Dixon, 2012).
Biological filtration can have additional benefits beyond further removing organic DBP
precursors. A study by Lauderdale et al. (2014) showed that biological filtration allowed for a 50
percent reduction in coagulant dose. In addition, the use of biological filtration may produce
more biologically stable water, which can help control biofilm growth and stabilize disinfectant
residuals in distribution systems (McGuire et al., 2014).
7.3.3 Information on Non-Conventional Treatment
Based on the DBP ICR data, non-conventional treatment plant types include direct filtration, in-
line filtration, slow sand filtration, surface water unfiltered treatment and ground water
disinfection only (see Exhibit 6.14 and Exhibit 6.15 of Chapter 6). In general, TOC levels in the
source water of surface water unfiltered plants or ground water plants with disinfection only are
much lower than conventional or non-conventional filtration treatment plants (see Exhibit 6.14
and Exhibit 6.15 of Chapter 6). The discussion presented in this section focuses on direction
filtration, in-line filtration and slow sand filtration.
There is little new information on removal of DBP precursors or reduction of DBP formation
potentials by non-conventional filtration treatment plants. This may be attributable to: 1) non-
conventional treatment plants not needing to meet the TT requirements for TOC removal under
the Stage 1 D/DBPR; and 2) the DBP precursor levels (as indicated by TOC) in source water as
well as treated water in non-conventional plants are generally much lower (see Exhibit 7.3).
Six-Year Review 3 Technical Support Document	7-15
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
According to Nieminski and Perry (2015), filtration type can be categorized into high-rate versus
low-rate filtration from the perspective of hydraulic loadings. The high-rate filters include the
filters used in conventional/softening treatment, direct filtration and in-line filtration plants; the
low-rate filters include slow sand filters and bank filtration (see Section 7.4 for discussion on
bank filtration).
Exhibit 7.3 shows percent TOC removals from source water to filter effluent among different
filtration treatment plants participating in the DBP ICR. The direct filtration or in-line filtration
plants (without sedimentation basins) performed similarly, with mean TOC removals of 11-18
percent, about 50 percent less than the removals achieved by the conventional/softening
treatment plants. Yet, some of the direct filtration or in-line filtration plants had relatively high
TOC levels in the raw water (up to 4.2 mg/L) and in the filtered water (up to 2.7 mg/L),
respectively.
As with filters in conventional/softening plants, the filters in direct or in-line filtration plants may
be converted into biofilters for further removal of TOC. Exhibit 7.3 also indicates that the slow
sand filtration plants generally treat source water with low turbidity and TOC levels (even lower
than the levels seen in in-line filtration plants). Yet, relatively high percent TOC removals can be
achieved (i.e., 28 percent as a mean). This may be because slow sand filters generally have a
long running time and biological fixed-film growth can occur naturally within the filters, if the
disinfectant residuals (including free chlorine or chloramines) in the influent are absent or
sufficiently low (Collins et al., 1992; Eighmy et al., 1993). Since slow sand filtration plants
normally require a larger land area and need less operational attention as compared to other
filtration plants, they are used more frequently in small versus large or medium systems (Collins
et al., 1992; Eighmy et al., 1993).
7.3.4 Information on Potential Add-on Physical Removal Unit Processes
One approach that has been used for additional removal of DBP precursors is to include physical
removal unit processes in a treatment train (normally following filtration). These unit processes
commonly include GAC (adsorption), membranes (including microfiltration or nanofiltration)
and ion exchange. To ensure the reasonable effectiveness of these treatment units, specific pre-
treatment is typically included, particularly for surface waters. As indicated in Exhibit 7.2, GAC
and membranes were included as treatment technologies for compliance with the existing
D/DBPRs. For the Stage 2 D/D/DBPR, for systems that disinfect their source water (i.e., non-
consecutive systems), best available technologies (BATs) were defined as 1) enhanced
coagulation or enhanced softening, plus GAC 10, 2) nanofiltration and 3) GAC 20 plus chlorine
(USEPA, 2006a), based on the assessment that most water systems would be able to meet MCLs
for TTHM/HAA5 with these treatment technologies.
This section discusses new information (since development of the Stage 2 D/DBPR) on
applicability, effectiveness and unintended consequences of these unit processes.
7.3.4.1 Adsorption by GAC
The adsorbent materials used for water treatment are either carbon-based (e.g., GAC) or non-
carbon-based. Since non-GAC adsorption is mostly used to remove contaminants such as arsenic
and radionuclides, rather than DBP precursors (Schendel et al., 2009), it is not discussed here.
Six-Year Review 3 Technical Support Document	7-16
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
One longstanding concern about organic precursor removal is the extent to which it causes a shift
in chlorination DBP mixtures to more brominated species when source water bromide levels are
relatively high (Summers et al., 1993; Symons et al., 1993; Sohn et al., 2006). To further
understand this issue, EPA reassessed the data from the ICR Treatment Study Database
(ICRTSD), which contains extensive bench- and pilot-scale data on the effectiveness of GAC
and nanofiltration in controlling natural organic matter (NOM) DBP precursors (USEPA, 2000f).
This section summarizes the analytical results from this data source, along with pertinent new
literature. Additional details about EPA's analysis of the data from ICRTSD are provided in
Appendix C of this document. Overall, EPA's analysis generally shows increased percent
reduction of the sum of the BrTHMs (sometimes referred to as THM3, which is the sum of the
three BrTHM species) as TOC removal by GAC increases (e.g., from a target effluent level of 2
mg/L to 1 mg/L) for source waters with high bromide concentrations. It also shows that
bromoform formation increases as bromide concentrations increase and that bromoform becomes
the dominating species when source water bromide concentrations exceed 200 |ig/L.
7.3.4.1.1 Analysis of Data from ICRTSD for GAC
Background. The DBP ICR required surface water systems serving more than 100,000 people
with raw water TOC levels greater than 4.0 mg/L and ground water systems serving more than
50,000 people with finished water TOC levels greater than 2.0 mg/L to conduct bench or pilot
studies of GAC or nanofiltration for the control of DBP precursors (USEPA, 1996b). A total of
99 treatment studies, including 63 with GAC and 36 with nanofiltration, were conducted and the
results submitted to EPA (USEPA, 1996b; Hooper and Allgeier 2002; USEPA, 2006a). The ICR
Treatment Study represents the most extensive evaluation of GAC for DBP control under field
conditions, with a wide range of source water quality and distribution system characteristics
(Hooper and Allgeier, 2002). EPA used the ICRTSD to guide the selection of BATs in
developing the Stage 2 D/DBPR (Hooper and Allgeier, 2002; Bond and Digiano, 2004).
In the treatment studies, samples were analyzed for THM4 and HAA6 (a subset of samples were
also analyzed for HAA9) using simulated distribution system (SDS) testing to assess DBP
formation in distribution systems. The SDS test simulates the average distribution system
conditions at an individual plant, such as residence time, water pH and temperature, with free
chlorine as the primary and residual disinfectant.
Analytical Approach. The impacts of TOC removal by GAC on DBP formation were evaluated
as a function of the bromide concentration in source water. Prior to analysis, the GAC influent
and effluent water quality data were extracted from the ICRTSD based on the effluent TOC
concentration of 1 and 2 mg/L, respectively. The 1- and 2-mg/L TOC datasets contain 259 and
191 records, respectively. Data were then placed into low- and high-bromide groups based on the
median bromide concentration of 64 |ig/L for the 1-mg/L TOC dataset and 75 |ig/L for the 2-
mg/L TOC dataset. Statistical analysis (including 10th percentile, median and 90th percentile)
was performed on SDS-THM4 (the sum of four regulated THM species), SDS-THM3 (sum of
three BrTHM species), SDS-HAA9 (nine species of haloacetic acids (HAAs)) and SDS-HAA-Br
(six brominated HAAs). Bromine incorporation factor and percentage of bromide incorporation
(PBI) were calculated using equations from literature to evaluate the extent of bromine
incorporation into DBP groups (Sohn et al., 2006).
Six-Year Review 3 Technical Support Document	7-17
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Summary of Analysis. The treatment data show that: 1) the percentage removal in BrTHMs
increases as TOC removal by GAC increases from a target effluent level of 2 to 1 mg/L for
source waters with high bromide concentrations; and 2) bromoform formation increases as the
bromide concentration increases and bromoform becomes the dominating species when source
water bromide concentrations exceed 200 |ig/L for the high-bromide group. The removal of
BrTHMs is less significant for the low-bromide waters because BrTHMs were formed at lower
levels in those waters. GAC treatment resulted in a smaller PBI in treated water for both THMs
and HAAs, similar to the effect of coagulation, where a smaller percent of bromide incorporation
was observed for coagulated water (Sohn et al., 2006). Formation of brominated DBPs may have
been limited by precursor availablity at low TOC levels. Results of the GAC influent and
effluent water quality for the 1- or 2-mg/L TOC datasets are provided in Appendix C of this
document.
Limitations of ICR Treatment Study Dataset. The ICRTSD studies were conducted from July
1997 to December 1998, prior to promulgation of the Stage 1 D/DBPR. Water systems may have
optimized their treatment strategies after the promulgation of these rules, which may affect how
well the results of the GAC treatability studies represent the post-Stage 1 conditions.
Furthermore, the SDS tests used average residence time in the distribution system. Since
compliance under the Stage 2 D/DBPR is based on samples taken at locations representing
maximum residence time in the distribution system, the SDS DBP levels in the ICRTSD could
underestimate DBP formation for compliance with the Stage 2 D/DBPR.
7.3.4.1.2 GA C Literature Review
This section summarizes new information from the literature on GAC for the removal of organic
DBP precursors, inorganic DBP precursors, organic DBPs and inorganic DBPs, respectively.
Removal of Organic DBP Precursors. GAC has a long history of use in the United States for
removing certain organic compounds. Numerous studies published after development of the
Stage 2 D/DBPR provide an improved understanding of the applicability and effectiveness of
GAC for removal of organic materials. According to McGuire et al. (2014), the water system in
Cincinnati, Ohio is the first utility in the United States to install a modern GAC treatment system
with regeneration on site and a capacity of 215 million gallons per day. There are at least a dozen
additional GAC plants in the Ohio Valley, Texas, Arizona and elsewhere. Although some were
originally installed to remove NOM for DBP control, many utilities are realizing other benefits
in addition to the original purposes.
Researchers have identified and evaluated key factors to be incorporated into predictive models
for GAC unit design (Bond and Digiano, 2004; Chiu et al., 2012). Particularly, they evaluated
the relationships between GAC service life and source water type, feed water quality, GAC
particle size and EBCT. Preliminary bench-scale testing showed that prechlorination shortly
before GAC filtration resulted in lower THMs in the distribution system, compared with the use
of GAC without prechlorination (Ghosh et al., 2011).
Other studies evaluated the efficiency of GAC for removal of precursors of nitrogenous DBPs,
including NDMA (Chiu et al., 2012; Hanigan et al., 2012; Hanigan et al., 2015). These studies
looked at removal of NDMA precursors in river waters containing wastewater effluent, the effect
Six-Year Review 3 Technical Support Document	7-18
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
of the use of pre-oxidants and removal of NDMA precursors originating from polyDADMAC
coagulants.
Removal of Inorganic Precursors. Researchers have conducted bench scale testing of new
adsorbents that can remove additional bromide and iodide compared to traditional GAC and
PAC. These new adsorbents are superfine PAC, silver-impregnated activated carbon and silver-
doped carbon aerogels (Zhang et al., 2015; Ikari et al., 2015; Sanchez-Polo et al., 2006; Sanchez-
Polo et al., 2007). Superfine PAC, which has significantly smaller particle sizes than PAC,
achieved 90 percent removal of iodine following prechlorination (Ikari et al., 2015). Silver-
doped carbon aerogels likewise have been found to increase adsorption of bromide and iodide by
a factor of 3 to 12 times that of conventional GAC (Sanchez-Polo et al., 2006). However, one
potential unintended consequence of full-scale use of the silver-doped aerogel treatment method
is the possible leaching of the carbon polymer precursors.
Removal of Organic DBPs. Johnson et al. (2009) conducted a literature review on
THM4/HAA5 removal by GAC and indicated that a 70 percent or more reduction of both THM4
and HAA5 could generally be achieved. A pilot study from Babi et al. (2007) showed that the
removal capacity of GAC exhibited the order of DOC > HAAs > THMs, which was consistent
with observations from the study conducted by Kim and Kang (2008), but contrary to findings of
Xie and Zhou (2002), who indicated that GAC breakthrough of HAA5 occurred more quickly
than breakthrough of THM4. This inconsistency could be attributable to biodegradation of DOC
and HAAs occurring within the adsorption unit (Babi et al., 2007; Kim and Kang, 2008; Johnson
et al., 2009). Xie et al. (2004) concluded that GAC could be used for short-term removal of
preformed THMs (through adsorption) and long-term removal of preformed HAAs (adsorption
plus biodegradation). Booth et al. 2006 demonstrated the effectiveness GAC for removing NOM
and controlling THM4 and HAA5 in high bromide waters using free chlorine during distribution
although higher percentages of the brominated species were formed.
Xie et al. (2004) recommended retaining 5 percent of the old GAC in the column to expedite
bioactivity development after replacement for better HAA removal. Johnson et al. (2009) noted
that the design variables for the removal of THM4 and HAA5 by GAC include EBCT, use of a
pressurized or gravity flow system, type of carbon used, backwash frequency, velocity and
species being adsorbed. For THM4, the more brominated species had greater adsorption
capacities than the more chlorinated species; for HAA5, the more halogenated species had higher
adsorption capacities than mono-HAA. Removal of nitrosamines (including NDMA) by GAC is
discussed in the Six-Year Review 3 Technical Support Document for Nitrosamines (USEPA,
2016d).
Removal of Inorganic DBPs. In addition to removal of organic DBPs, GAC also exhibits some
capacity for removal of inorganic DBPs. In a full-scale study conducted by Hoehn et al. (2003),
GAC contactors were found to achieve 63 percent of average chlorite removal with influent
concentrations up to 0.8 mg/L but with very long EBCTs ranging from 48 to 130 minutes.
Several researchers observed that chlorite removal by GAC involved two steps: 1) chlorite
adsorption on GAC sites and 2) subsequent reduction to chloride. Such an observation was also
supported by some earlier studies (Gonce and Voudrias, 1994; Hoehn et al., 2003; Collivignarelli
et al., 2006). Gonce and Voudrias (1994) showed that the most effective chlorite removal
occurred at pH 5. They further indicated that chlorate was not reduced by GAC, but was only
Six-Year Review 3 Technical Support Document	7-19
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
physically and reversibly sorbed. Thus, much less removal of chlorate by GAC was observed,
compared to chlorite. Collivignarelli et al. (2006) indicated that chlorite removal was reduced
when GAC was preloaded with organic matter and specific ions (e.g., nitrate). Their study also
showed that thermally regenerated GAC demonstrated good removal for both organic matter (70-
80 percent) and chlorite (100 percent). With exhausted GAC, organic matter removal was
reduced from 40-50 percent (without chlorite) to 5-7 percent (when water was spiked with
chlorite); chlorite removal remained significant at ~ 50 percent. Huang and Cheng (2008) studied
effects of activated carbon on removal of bromate and observed that carbons with more
mesopores adsorbed more bromate. Wood-based carbons contained more mesopores than
coconut or coal-based carbons, resulting in a high removal capacity for bromate. Chen et al.
(2012) employed cationic surfactant loading to modify GAC to enhance bromate removal. With
such modified GAC, bromate was removed mostly through an ion exchange process, and the
removal increased with a decreased pH. Xu et al. (2015) prepared and tested nano-iron
hydroxide-impregnated GAC (Fe-GAC) for adsorption and reduction of bromate. They found
that while GAC alone could reduce some bromate, Fe-GAC could greatly enhance the bromate
removal capacity and removal rate, with the optimal pH being 6-8. Both Chen et al. (2012) and
Xu et al. (2015) observed that other anions (e.g., PO43" and SO42") exhibited inhibiting effects on
bromate removal.
7.3.4.2 Membranes
Membrane filtration is a separation technology that pushes or pulls water through a fixed barrier
and includes microfiltration (MF), ultrafiltration (UF), nanofiltration (NF) and reverse osmosis
(RO). The primary difference between each type of filtration is the pore size and the operating
pressure range of the membranes. The results from the ICR Treatment Study mentioned in the
previous section showed that all ground water plants, if they used NF, were able to meet the 80
Hg/L TTHM and 60 ng/L HAA5 MCLs with a 20 percent safety factor (i.e., were able to keep
TTHM and HAA5 below 64 ng/L and 48 ng/L, respectively) at the average residence time
monitoring locations (USEPA, 2005g). NF is less expensive than GAC for high-TOC ground
waters, which generally require minimal pretreatment prior to the membrane process. Also, NF is
an accepted technology for treatment of high-TOC ground waters in Florida and parts of the
Southwest (Thorsen and Flogstad, 2006).
Becker et al. (2013) indicated that MF and UF processes were essentially particle removal
processes and were not any more effective than conventional filtration processes in DBP
reduction. In contrast, high pressure membrane systems such as NF and RO could directly
remove NOM and drastically reduced DBP precursors. Netcher and Duranceau (2015)
demonstrated that settled water turbidities greater than 1 NTU could be expected to have
detrimental impacts on the efficiency of the subsequent UF process. The results from the studies
conducted by Patterson et al. (2012) and Plourde-Lescelleur et al. (2015) confirmed the
effectiveness of nanofiltration. One unintended consequence of membrane technologies may be
that they can result in removal of alkalinity from the finished water, which can lead to increased
lead and copper corrosion and may affect compliance with the Lead and Copper Rule (Becker et
al., 2013).
Six-Year Review 3 Technical Support Document	7-20
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
7.3.4.3 Ion Exchange
Ion exchange (using cation and anion resins) is a process that removes dissolved ions from water
and replaces them with other similarly charged ions. Schendel et al. (2009) also pointed out that
one of the key factors affecting the efficiency of anion exchange at removing the targeted
contaminants is the extent to which competitive ions (such as chloride and sulfate) are present in
the water. The results from the studies of Singer et al. (2009) and Watson et al. (2015) confirm
this observation.
One of the most studied ion exchange processes is magnetic ion exchange (MIEX), where the
anion exchange resin is supplemented with magnetic iron oxide to facilitate resin extraction and
regeneration. Booth et al. 2006 demonstrated MIEX to be effective in controlling THM4 and
HAA5 in high bromide waters using free chlorine during distribution although higher
percentages of brominated species were formed. New research has found that MIEX may
preferentially remove organic precursors with high SUVA (specific ultraviolet absorbance) and
low molecular weights (Plourde-Lescelleur et al., 2015; Hanigan et al., 2013; Singer et al., 2009;
Mergen et al., 2009; Drikas et al., 2011; Watson et al., 2015; Metcalfe et al., 2015). Drikas et al.
(2011) found that very hydrophobic acids represented a significant portion of the NOM removed
by MIEX. Other researchers found these types of acids could be correlated to DBP formation
(McKie et al., 2015; Chang et al., 2013). The study conducted by Metcalfe et al. (2015), with ion
exchange configurations other than MIEX, confirmed the effectiveness of anion exchange for
removing UV-absorbing materials.
Unlike other organic DBP precursor removal techniques such as enhanced coagulation, anion
exchange (including MIEX) can also remove bromide and iodide (by 21-91 percent) (Hsu and
Singer, 2010; Phetrak et al., 2014; Walker and Boyer, 2011; Xu et al., 2013; Echigo et al., 2007).
Depending on the Br":DOC in the water, however, a shift to a higher percentage of brominated
DBPs in treated water may still occur (Watson et al., 2015).
Several studies have found that the use of chloramines following ion exchange treatment can
lead to increased NDMA formation (Gan et al., 2013; Watson et al., 2015). NDMA formation
likely occurs because most anion exchange resins are composed of amines, which have been
demonstrated to be NDMA precursors (Krasner et al., 2013; Flowers and Singer, 2013). The
observed contribution to NDMA formation from MIEX is between 5 and 10 ng/L (Watson et al.,
2015; Gan et al., 2013), although researchers note that concentrations can be much higher if
preformed chloramines are used with no free chlorine contact time (Gan et al., 2013). In addition
to NDMA precursors, a recent study found that ion exchange resins could be a direct source of
nitrosamines in finished water (Watson et al., 2015).
A challenge of ion exchange treatment is disposal of the waste brine from resin regeneration.
Walker and Boyer (2011) demonstrated the use of bicarbonate as the mobile counter-ion and
sodium bicarbonate for regeneration instead of chloride-form anion exchange to address disposal
concerns.
Six-Year Review 3 Technical Support Document	7-21
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
7.3.4.4 Other Unit Processes
Many studies have been conducted to develop some innovative approaches for further removal
of DBP precursors and/or control of DBP formation. Some of these approaches have not been
applied to pilot- or full-scale plants and are not summarized here. This section only covers
oxidation and electrolysis, which have been tested at the bench- or pilot-scale and may have
applicability in full-scale treatment.
Disinfectants (typically including ozone, chlorine dioxide and free chlorine) have been applied as
oxidants at the beginning of treatment trains to chemically transform DBP precursors to forms
that result in lower DBP formation potentials. Other oxidants used include hydrogen peroxide
and permanganate. Such an operation is often referred to as pre-oxidation. The national survey
conducted by the AWWA Disinfection Committee in 2007 (AWW A, 2008) showed that 36
percent of SW systems used pre-oxidation. Matilainen and Sillanpaa (2010) provided a detailed
review of more than 50 oxidation research studies conducted between 2006 and 2009. They
noted that, although oxidation processes showed promise, circumstances in the research studies
were often impractical for full-scale plants and there were very few full-scale applications. In
addition, their review found that oxidation processes could cause a shift to lower molecular
weight compounds and that incomplete oxidation by several oxidants had been shown to increase
DBP formation potentials. Appendix C of this document includes a synopsis of some new studies
related to advanced oxidation.
Electrolysis involves the oxidation of bromide to bromine followed by volatilization of bromine
through the use of an electrical current (Kimbrough and Suffet, 2006; Kimbrough et al., 2011;
Kimbrough et al., 2012). Electrolysis has been found to remove 27-54 percent of source water
bromide when tested at the pilot scale on California State Water Project water. Important factors
for optimization include flow rate through the electrolysis reactor and contact time. For
optimization, Kimbrough and Suffet (2006) reported that the current applied should increase as
the flow becomes greater and the contact time shortens. Based on evaluation of five different
reactor bodies and varying placement of anodes within the reactors, Kimbrough et al. (2011)
concluded that full-scale reactors should maximize the surface area of the anodes and be as
shallow as possible to maximize the volatilization of bromine. A possible unintended
consequence is an increase in brominated haloacetonitriles; however, Kimbrough and Suffet
(2006) noted that the observed overall levels were very low, which made it unclear if the
increase was significant.
7.4 Information on Source Water Management
The information reviewed during the SYR3 process also reveals that some industrial activities
(e.g., hydraulic fracking or coal power generation) can increase bromide levels in drinking water
sources (see Chapter 6 of this document) and typical conventional treatment trains appear
ineffective at removal of bromide (States et al., 2013). Also, several new studies indicate that
municipal wastewater discharges and/or occurrence of algal blooms nearby water intakes can
increase the levels of DBP precursors in source water (Callinan et al., 2013; Saunders et al.,
2015). This information relates to the importance of watershed vulnerability characterization and
effective source water management practices.
Six-Year Review 3 Technical Support Document	7-22
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
A watershed vulnerability characterization that includes information about wastewater
contributions, land use (including point and non-point pollution sources) and streamflow
variations over time (for example, sewage contributions during low flow conditions) could help
to inform considerations about DBP formation potentials. For example, as noted by Krasner et al.
(2015), source waters with relatively elevated sewage contributions have been associated with
increased nitrosamine formation.
Approaches to characterizing vulnerabilities were identified in the literature. Several papers
discuss the use of fluorescence excitation/emission spectroscopy to characterize source water
DOC and track changes over time (Hua et al., 2007; Rosario-Ortiz et al., 2007; Bridgeman et al.,
2011; Pifer and Fairey, 2012). For example, Rosario-Ortiz et al. (2007) used fluorescence
analysis to distinguish between waters affected by microbial activity (e.g., by wastewater
influence) and those that were only minimally affected. Fluorescence excitation/emission
spectroscopy is non-destructive, and the potential exists for using it as an on-line, source water
monitoring and management tool (Bridgeman et al., 2011).
Weiss et al. (2013) developed a model for making source water selection decisions based on real-
time DBP precursor concentrations. Such a model could be used by utilities with multiple source
waters, intakes or intake depths. Modeling results showed that DBP precursors could be reduced
by modifying diversion decisions based on real-time DBP precursor concentrations in different
reservoirs.
Pre-treatment processes for lowering DBP formation potentials in water sources include raw
water storage/pre-sedimentation and bank filtration. Raw water storage/pre-sedimentation can
help to reduce seasonal variation of source water quality and pre-settle some particulates
(including some organic matter). The national survey conducted by the AWWA Disinfection
Committee in 2007 (AWWA, 2008) showed that 31 percent of SW systems (including those of
all sizes) used raw water storage/pre-sedimentation.
Depending on site conditions, bank filtration has been shown to be an effective method to
improve source water quality and thus reduce the treatment burden on the existing treatment
trains. Literature indicates that bank filtration can not only remove some pathogens but can
reduce the formation potentials of DBPs associated with chlorination and chloramination (Brown
et al., 2015).
Depending on site conditions (including geological conditions and land ownership), bank
filtration, if appropriately constructed and used, can improve source water quality and thus
reduce the treatment burden on a treatment train. Bank filtration appears to be more commonly
used in Europe (Wang et al., 2002).
The removal of organic compounds by bank filtration can depend on the flow path the water
takes to the collector well and the oxygen content of the water. In general, the longer the flow
path to the collector well, the better the removal. Temperature can affect both the biological
activity of the filtration and the flow dynamics (Brown et al., 2015).
Bank filtration has been found to be effective in removing low molecular weight assimilable
organic carbon from source waters (Brown et al., 2015). A few studies have examined bank
Six-Year Review 3 Technical Support Document	7-23
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
filtration for the removal of NDMA precursors and found 64 percent reduction of NDMA
formation potential, 49 to 72 percent removal of TOC and 58 to 68 percent reduction of UV
absorbance (Krasner et al., 2015). However, about 20 percent of the removal of these indices
could be attributed to dilution of the river water with ground water, rather than removal during
bank filtration. A study of four full-scale bank filtration facilities found DOC removals of
between 12 and 93 percent, with an average of 55 percent. About 10-25 percent of the removal
was due to dilution by ground water (Partinoudi and Collins, 2007).
7.5 Information on Changing Disinfection Practices in Treatment Plants and Distribution
Systems
Disinfection practices generally refer to the collective water quality and treatment conditions
under which a disinfectant or disinfectants are applied and disinfectant residuals are maintained
during treatment and distribution, including disinfectants used as pre-oxidants discussed earlier.
The Agency regularly collects information on disinfection practices, since it is critical for
understanding DBP formation and occurrence. Information about disinfection practices is
available in the 1997-1998 DBP ICR (USEPA, 2000e), the 2008-2010 UCMR 2 (USEPA,
2012c) and the 2013-2015 UCMR 3 (USEPA, 2016h). Further, it is expected that additional
information about disinfection practices will be collected as part of UCMR 4 (USEPA, 2015)). In
addition, the AWWA Disinfection Systems Committee periodically (about every 10 years since
1978) conducts a national survey in this area (AWWA, 2008). The data from these sources are
collectively presented in Chapter 6 of this document, in the Six-Year Review 3 Technical Support
Document for Nitrosamines (USEPA, 2016d) and in the Six-Year Review 3 Technical Support
Document for Chlorate (USEPA, 2016e), with the detailed discussion of various factors,
including disinfection practices, affecting formation/occurrence of different groups of DBPs. In
particular, the current distribution of disinfectant usages and changes in this distribution are
characterized and presented in Chapter 6 of this document and further discussed in the Six-Year
Review 3 Technical Support Document for Chlorate (USEPA, 2016e).
The following summary is based on those discussions, while the section "Alternative
Disinfectants" in Appendix C of this document includes a synopsis of new studies related to
common individual or combined disinfectants, including chlorines, ozone, chlorine dioxide and
UV. The section "Advanced Oxidation Processes" in Appendix C of this document also provides
a synopsis of new studies related to these disinfectants that can be used as oxidants as part of
strategies for controlling DBP formation.
Various combinations of disinfectants and precursor removal processes have been used to
achieve the DBP MCLs while also meeting the microbial standards. As predicted in the
Economic Analysis for the Stage 2 D/DBPR (USEPA, 2005g), the multiple national datasets
(including from the DBP ICR, UCMR 2 and UCMR 3) have collectively shown an increasing
trend in the number of systems using alternatives to chlorine during the past two decades. This
observation is consistent with the observation from the AWWA periodical surveys (AWWA,
2008) mentioned earlier. Numerous systems have shifted their primary disinfectant from free
chlorine to chloramines, ozone, chlorine dioxide and UV, including combinations of such
disinfectants, and have shifted to using chloramines from free chlorine as a disinfectant residual
in the distribution system.
Six-Year Review 3 Technical Support Document	7-24
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Overall, this trend implies that different organic DBPs other than chlorinated DBPs may become
more prevalent over time, especially NDMA. The same may be true for certain inorganic DBPs
such as bromate and chlorite, which as associated more strongly with alternative disinfectants.
More detailed discussion on controlling formation and occurrence of nitrosamines (including
NDMA) is presented in the Six-Year Review 3 Technical Support Document for Nitrosamines
(USEPA, 2016d). As discussed earlier, new information also informs the extent to which
different types of DBPs may be controlled depending upon where they are applied in the
treatment train and/or in combination with other disinfectants. For instance, pre-ozonation in
conjunction with biofiltration can help removal of precursors of several classes of chlorination
DBPs. EPA recognizes that the extent to which occurrence and associated health effects data
may be lacking for one group of DBP contaminants versus another, as well as for DBP mixtures,
may make treatment decisions challenging.
Regarding forms of chlorine used as free chlorine or for formation of chloramines, EPA has seen
a clearly increasing national trend toward using hypochlorite stock solution or on-site generation
of hypochlorite in lieu of chlorine gas. This shift is likely due to security concerns of transport
and storage of chlorine gas. The implications of this shift in chlorine source have become
apparent. In the UCMR 3 dataset, for example, chlorate levels are significantly higher among
systems using hypochlorite stock solution or on-site generation of hypochlorite, compared to
those using chlorine gas. The analytical results from the UCMR 3 dataset also show that the use
of chlorine dioxide can lead to a high occurrence of chlorate. The Six-Year Review 3 Technical
Support Document for Chlorate (USEPA, 2016e) presents more information on use of
hypochlorite versus chlorine gas and associated implications. In addition, the DBP ICR data
indicates that among systems serving at least 100,000 people that chlorate can co-occur with
chlorite when hypochlorite or chlorine dioxide is used. Since both hypochlorite and chlorine
dioxide are being used more frequently, they are probably being used more frequently in
conjunction with each other, which can lead to higher levels and frequencies of chlorate/chlorite
co-occurrence if no effective control measures are implemented. The water industry has provided
tools that can help utilities to manage the concentrations of chlorate in water treated with
hypochlorite stock solution. For instance, a web-based predictive tool for chlorate formation
during storage of hypochlorite solution can be found on the AWWA website
(http://www.awwa.org/resourcestools/waterandwastewaterutilitvmanagement/hvpochloriteassess
mentmodel.aspx).
Many distribution systems provide a relatively long contact time, which may inadvertently lead
to DBP formation after the treated water leaves the treatment plants. New information indicates
that system-specific models can be developed to help operators optimize DBP control strategies
in distribution systems. Behzadian et al. (2012) used an NSGA-II algorithm coupled with the
EPANET hydraulic model to concurrently optimize chlorine residuals and THM formation
resulting from booster disinfection operations. Cruickshank (2010) showed how hydraulic
models can be used to assess the impact of control strategies, such as flushing, tank turnover and
bleed water at zone boundaries, on water age. A Rhode Island utility developed an empirical
model to limit THM4 levels in water leaving a finished water storage tank (Oneby et al., 2009).
The calibrated model helped operators decide which sources of supply to use depending on
current conditions (e.g., water temperature).
Six-Year Review 3 Technical Support Document	7-25
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
7.6 Information on Removing DBPs after Formation in Treatment Plants and/or
Distribution Systems
Since the promulgation of the Stage 2 D/DBPR, EPA notes the availability of new information
on DBP removal using aeration processes through volatilization (i.e., removing volatile DBP
compounds from water by transporting them to a gas phase). As discussed by Johnson et al.
(2009), all of these DBP removal or reduction technologies may be used as a localized treatment
approach in the distribution system. Utilities treat only the flow necessary at specific locations in
the distribution system (rather than treating the entire flow at the centralized treatment plant) to
comply with the MCLs for TTHM/HAA5 under the Stage 2 D/DBPR. Since DBP removal
through biofiltration or GAC adsorption was discussed earlier in this Chapter, this section
focuses on the new information on aeration processes.
Overall, aeration can be an effective process to lower THM4 levels (more effectively for
chloroform than the brominated species), but may have little effect on HAA5 levels. It is not
clear, however, how this type of treatment will affect levels and formation potentials of "not-so-
volatile brominated DBPs" downstream and water quality stability in distribution systems
(Ghosh et al., 2015).
Three basic types of aeration systems have been identified from the mechanical perspective:
surface aeration, spray aeration and diffused aeration/air stripping (Ghosh et al., 2015; Johnson et
al., 2009; Brooke and Collins, 2011; Duranceau, 2015). Removals for individual DBPs depend
on their Henry's Law constants. Chloroform is removed to a greater extent, while brominated
species are removed to a lesser extent (Johnson et al., 2009). DBPs of low volatility Johnson et
al. (2009), based on a pilot study, indicated that while aeration (through air stripping) was
effective to lower THM4 levels, it had no effect on HAA5.
Ghosh et al. (2015) studied surface and spray aeration side by side in the clearwells in full-scale
plants. Both systems were able to achieve between 19 and 34 percent THM4 removal and the
THM4 removal efficiency of the spray aeration system was marginally better (about 5 percent).
Ghosh et al. (2015) indicated that the overall THM4 reduction could be influenced by multiple
factors, including variation of hydraulic residence times in the clearwell, formation of THM4
within the clearwell and dilution of the aeration-treated water with the incoming water. This
research team also observed that such operations could reduce the baffling factor for quantifying
disinfection credits from the clearwell.
Because water systems must meet the TTHM MCL at each sampling location as part of the Stage
2 D/DBPR, some water systems have opted to use aeration technologies to lower THM4 levels at
certain locations in their distribution systems (Ghosh et al., 2015; Duranceau, 2015). Aeration
systems may be installed inside storage facilities or located outside of storage facilities. Aeration
systems inside storage facilities commonly use diffused aeration, where air is injected at the
bottom and spray aeration, where water is sprayed through air from nozzles at the top (Brooke
and Collins, 2011). Other in-tank aeration methods include surface aeration, where aerators float
on the water's surface and low-profile aerators (Jensen et al., 2010; Johnson et al., 2009).
Aeration systems located external to storage facilities include air stripping trays or packed
towers, membrane contactors and spray/bubble vessels (Brooke and Collins, 2011; Johnson et
al., 2009).
Six-Year Review 3 Technical Support Document	7-26
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Several studies have documented the effectiveness of various forms of post-treatment aeration
for THM4 control (e.g., surface aerators in Phoenix, Arizona, as reported by Jensen et al. (2010);
spray aerators in Suisun City, California, as reported by Walfoort et al. (2008); and spray
aerators in Ballinger, Texas, as reported by Fiske et al. (2011). Brooke and Collins (2011) found
that the removal rates for individual THM species were similar for spray aeration, but removals
of chloroform were higher when a diffused aeration system was used. THM4 removal in post-
treatment aeration facilities has been found to range from 47 to 93 percent depending on several
variables, including air to water ratio, droplet travel distance, water temperature and droplet
mean diameter (Brooke and Collins, 2011).
Several studies have considered whether distribution system aeration to remove THM4 also
reduces the water's chlorine residual, which may be an unintended consequence. Individual
researchers have not found this to be a problem in full-scale installations (Johnson et al., 2009;
Sinfield and Niday, 2015). The removal of chlorine residual may require the application of
booster chlorination, which requires additional management. In contrast to the negative impacts
of residual striping, one system found that aeration improved the mixing conditions in the
storage tank, which in turn led to reduced chlorine decay in the tank, and overall a lower chlorine
dose necessary for maintaining minimum disinfectant residual levels through the entirety of the
distribution system (Sinfield and Niday, 2015). The degree of residual stripping versus reduction
in residual decay will rely on many factors, pH, temperature and the aeration system to name a
few.
One aspect of aeration processes as currently employed is that the removal of DBPs at any
midpoint in the distribution may have little impact on controlling the continual formation of
DBPs further downstream of the aeration system. While additional DBP formation does occur,
one study found that after 170 hours DBP concentrations were half of what they had been before
aeration (Johnson et al., 2009). Nevertheless, the changes in water quality after localized
treatment could vary based on site-specific conditions in individual utilities.
Six-Year Review 3 Technical Support Document	7-27
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
8 Consideration of Other Regulatory Revisions for MDBP Rules
In addition to the review of maximum contaminant level goals (MCLGs), maximum contaminant
levels (MCLs) and treatment technique (TT) established by the National Primary Drinking Water
Regulations (NPDWRs), EPA considered whether other regulatory revisions, such as monitoring
and reporting requirements, should be considered as part of the Six-Year Review (SYR) process.
The Implementation Branch of the SYR protocol decision tree requires information regarding
whether a change in a contaminant's MCL or TT, or the availability of new health effects
information, will affect the monitoring or reporting requirements for a particular contaminant.
For the Third Six-Year Review (SYR3), EPA focused this review on implementation issues that
were not already being addressed through alternative mechanisms, such as a part of a recent or
ongoing rulemaking. In addition to this criteria, EPA considered potential implementation-
related revisions if they:
(1)	Represented a potential change to an NPDWR, as defined under section 1401 of SDWA;
(2)	Were "ready" for rulemaking — that is, the problem to be resolved had been clearly
identified, along with specific options to address the problem under the current regulatory
framework; and
(3)	Would clearly improve the level of public health protection and/or provide a meaningful
opportunity for cost savings (either monetary or burden reduction) while not lessening
public health protection.
The output of the Implementation Branch is a determination regarding whether EPA should
consider revisions to the monitoring or reporting requirements of an NPDWR. It is the final
branch of the decision tree.
EPA used the protocol to evaluate which of these issues to consider under SYR3. After EPA had
a consolidated list of implementation-related issues, it shared that list with the Association of
State Drinking Water Administrators to obtain input from state drinking water agencies
concerning the significance and relevance of the issues. Implementation issues will be
considered as part of the activities associated with potential future rulemaking efforts; some of
these might be addressed through regulatory revision or clarification, while others might be
handled through guidance.
Examples of implementation issues that are related to the MDBP rules include consecutive
system monitoring and chlorine burn, both of which are described further below. Additional
implementation issues related to the MDBP rules are described in Appendix D. Implementation-
related issues for the chemical phase rules are discussed in a separate document (USEPA,
2016g).
8.1 Stage 2 D/DBPR Consecutive System Monitoring
Monitoring in some combined distribution systems may be insufficient to adequately
characterize DBP exposure. Some large, hydraulically complex combined water distribution
systems may be conducting monitoring that is not adequate to characterize exposure throughout
Six-Year Review 3 Technical Support Document	8-1
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
the distribution system. Under the Stage 2 D/DBPR, EPA provided an alternative for states to
use to modify THM4 and HAA5 monitoring requirements for consecutive and wholesale
systems, in lieu of the existing modification process under 40 CFR §141.29, which requires EPA
concurrence. As a special primacy condition (40 CFR §142.16(m)), states may apply for
approval to modify monitoring without case-by-case EPA concurrence. Such approval requires
that every system in the combined distribution system have at least one compliance monitoring
location, so that compliance determinations are based on samples taken within the individual
distribution system. EPA anticipated that states would ensure that the number of compliance
monitoring locations and frequency of sampling after modification would remain sufficient to
adequately characterize DBP exposure and protect public health.
8.2 Stage 2 D/DBPR Compliance Monitoring - Chlorine Burn
Compliance monitoring for DBPs in some systems may not fully capture DBP levels to which
customers are exposed throughout the year. Under 40 CFR §141.621(a)(2), including footnote 2,
monitoring frequency and timing are specified for surface water and ground water systems based
on a system's population size category. Systems that use chloramines as a residual disinfectant
(generally as part of a compliance strategy to meet DBP MCLs) often temporarily switch to free
chlorine as the residual disinfectant for a period (from 2-8 weeks) in order to control nitrification
in the distribution system. This practice is commonly called a "chlorine burn." During the
chlorine burn, higher levels of DBPs (i.e., THM4, HAA5 and other chlorination DBPs) are
expected to form. Systems often conduct their compliance monitoring outside of the chlorine
burn period, and therefore, potentially higher THM4 and HAA5 levels are not included in
compliance calculations. Actual exposures may be significantly higher than reported exposures
in such cases.
Additional Information Related to Chlorine Burn
The effects of chlorine burn periods on exposure to DBPs might become increasingly important
in light of the adverse health effects (reproductive and developmental) related to short-term
exposure to DBPs in chlorinated drinking water (refer to Chapter 4 for a discussion of health risk
information about reproductive and developmental toxicity). Further, such elevated
concentrations of DBPs, depending upon their levels and duration, could be important for more
accurately assessing running annual average (RRA) exposures. For example, if the burn period
were for a month, the theoretical contribution of that month's THM4 or HAA5 occurrence could
represent one-third of the occurrence for that quarter and if considered, could substantially affect
the actual average concentration for that quarter as well as that for the RAA.
Data gaps exist for several areas related to chlorine burn - e.g., the percent of the industry that
uses this practice, the frequency and length of time for which the burns are performed and the
levels of DBPs produced by short-term exposures during those periods.
To further assess these data gaps and other issues pertaining to chlorine burn practices, EPA
conducted a literature review on the potential impacts of chlorine burn on DBP formation
Six-Year Review 3 Technical Support Document	8-2
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
(USEPA, 2014b). Specifically, EPA conducted the literature review to gather available
information on the following:
•	Typical chlorine burn practices adopted by PWSs, including timing, frequency, duration,
free chlorine dose (especially relative to the dose before chlorine burn) and any other
treatment operational changes;
•	Water quality monitoring for DBP and chlorine residual levels during chlorine burn vs.
state requirements (if any);
•	Public notification practices adopted by PWSs vs. state requirements;
•	Research projects to evaluate the effect of chlorine burn on DBP formation;
•	Guidance documents or industry standards on chlorine burn practices adopted by primacy
agencies and the water industry (such as from the American Water Works Association
(AWW A) or the 10-state standards); and
•	Alternative nitrification control strategies and how they compare with the chlorine burn
practice.
The literature review shows that for utilities that use chloramine and either have had nitrification
problems, or have nitrification controls, a reported 25 to 40 percent have used free chlorine burns
to control nitrification (USEPA, 2014b). These proportions of utilities implementing chlorine
burns are similar to the 22 percent of 63 utilities reported from a 2003 survey by Harms and
Owen (2004).
In a free chlorine burn, chloramine in a distribution system or in part of a distribution system is
replaced with sufficient free chlorine to oxidize excess ammonia and eliminate substrate used by
ammonia-oxidizing bacteria. Although some utilities implement chlorine burns as a matter of
routine operations (and may do so under state requirements in some cases), chlorine burns
increase DBP formation. AWWA's manuals, M20 and M56 (AWWA 2006, 2013), provide
guidance and recommendations for minimizing nitrification and practices that minimize DBP
formation. In particular, AWWA (2013) recommends that chlorine burns be "a last resort" for
controlling nitrification given the increased THM4 and HAA formation during chlorine burns.
While free chlorine burns are primarily used to mitigate nitrification events, some utilities have
reported using free chlorine burns for biofilm reduction. AWWA (2006) reports that switching
periodically to free chlorine might also reduce growth of chloramine-resistant bacteria. The
surveys reviewed in the literature study tended to focus on medium and larger utilities.
An example of the unintended consequences of a temporary switch to use of free chlorine was
identified for a large U.S. city (Huerta et al., 2015). In that example, the city temporarily
switched its disinfectant at a treatment plant in response to concerns about nitrification in
distribution systems. Data were made available about these temporary switches, made in May
2009 and October 2014. During these events, THM4 values were often higher than 200 ppb and
sometimes higher than 300 ppb, while HAA5 values were often higher than 100 ppb and
sometimes higher than 200 ppb.
Estimation of the Effect of Chlorine Burn on DBP Levels
Data are available in the DBP ICR dataset (further discussed in Chapter 6) and the ICR
Treatment Study Data (ICRTSD, further discussed in Chapter 7) that could be used to estimate
Six-Year Review 3 Technical Support Document	8-3
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
the impact of a switch in disinfectant use, specifically from chloramine to chlorine, on the levels
of DBPs. In addition to field samples, utilities prepared simulated distribution system (SDS)
samples for their treatment plants. The SDS is a test method developed to predict the amounts of
DBPs that form based on simulated conditions in the treatment plant and distribution system.
Some of the key parameters affecting the SDS conditions are incubation time, temperature, pH
and chlorine residual. Previous studies have shown good correlation between SDS results and
field test results (McGuire et al., 2002). Samples that were disinfected using chloramine in the
ICR DBP dataset and using free chlorine in the ICRTSD could be linked using the same plant ID
number and calendar quarter to compare the difference in DBP levels under the two disinfectant
uses. In this manner, the impact on THM4 and HAA5 from a switch in disinfectant from
chloramine to chlorine could be used to mimic the impact of a chlorine burn.
EPA conducted a preliminary review of these data, based on 44 quarters from 20 plants. The
results from this review suggested that plants may observe substantial increases in THM4 levels,
and a smaller increase in HAA5 levels, when chlorine is used as the residual disinfectant
compared to chloramine. For a majority of plants, the projected increase would be approximately
30-40 |ag/L or greater but in some cases more than 80 ng/L. Additional information about this
effort are provided in Appendix E to this document.
Six-Year Review 3 Technical Support Document	8-4
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
9 References
Abdel-Rahman, M.S., D. Couri, and J.D. Jones. 1980. Chlorine dioxide metabolism in rat.
Journal of Environmental Pathology and Toxicology. 3(1-2): 421-430.
Abdel-Rahman, M.S., M.R. Berardi, and R.J. Bull. 1982. Effect of chlorine and monochloramine
in drinking water on the developing rat fetus. Journal of Applied Toxicology. 2(3): 156-159.
Abdel-Rahman, M.S., D. Couri, and R.J. Bull. 1984. Toxicity of chlorine dioxide in drinking
water. Journal of the American College of Toxicology. 3(4): 277-284.
Agency for Toxic Substances and Disease Registry (ATSDR). 1999. Toxicological Profile for
Formaldehyde. U.S. Department of Health and Human Services, Public Health Service.
Retrieved from http://www.atsdr.cdc.gov/ToxProfiles/tpl 11 .pdf
Aggazzotti, G., E. Righi, G. Fantuzzi, B. Biasotti, G. Ravera, S. Kanitz, F. Barbone, G.
Sansebastiano, M. Alberto, M.A. Battaglia, V. Leoni, L. Fabiani, M. Triassi, S. Sciacca, and
Collaborative Group for the Study of Chlorinated Drinking Waters and Pregnancy. 2004.
Chlorination by-products (CBPs) in drinking water and adverse pregnancy outcomes in Italy.
Journal of Water and Health. 2(4): 233-247.
Aida, Y., K. Yasuhara, K. Takada, Y. Kurokawa, and M. Tobe. 1992. Chronic toxicity of
microencapsulated bromodichloromethane administered in the diet to Wistar rats. Journal of
Toxicological Sciences. 17(2): 51-68.
American Public Health Association (APHA), American Water Works Association (AWW A),
and Water Environment Federation (WEF). 2012. Standard Methods for the Examination of
Water and Wastewater, 22nd Edition.
American Water Works Association (AWW A) Disinfection Committee. 2000a. Disinfection at
small systems. Journal of the American Waterworks Association. 92(5): 24-31
AWWA Disinfection Committee. 2000b. Disinfection at large and medium-size systems. Journal
of the American Water Works Association. 92(5): 32-43.
AWWA. 2006. Water Chlorination/Chloramination Practices and Principles, 2nd ed. AWWA
Manual M20. AWWA catalog no. 30020. Denver, CO: American Waterworks Association, 188
P-
AWWA Disinfection Committee. 2008. Committee report: Disinfection survey, part 1 — recent
changes, current practices, and water quality. Journal of the American Water Works Association.
100(10): 76-90.
AWWA. 2013. Nitrification Prevention and Control in Drinking Water,2nd ed. AWWA Manual
M56. AWWA catalog no. 30056-2E. Denver, CO: American Water Works Association, 320 p.
Amet, Y., F. Berthou, G. Fournier, and J.-F. Menez. 1997. Cytochrome P450 4A and 2E1
expression in human kidney microsomes. Biochemical Pharmacology. 53(6): 765-771.
Six-Year Review 3 Technical Support Document	9-1
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Anderson, W.B., P.G. Board, B. Gargano, and M.W. Anders. 1999. Inactivation of glutathione
transferase Zeta by dichloroacetic acid and other fluorine -lacking a-haloalkanoic acids.
Chemical Research Toxicology. 12(12): 1144-1149.
Andrews, J.E., H.P. Nichols, J.E. Schmid, L.E. Mole, E.S. Hunter III, and G.R. Klinefelter. 2004.
Developmental toxicity of mixtures: the water disinfection by-products dichloro-, dibromo-, and
bromochloro acetic acid in rat embryo culture. Reproductive Toxicology. 19(1): 111-116.
Arnold, W.A., R.M. Hozalski, C.R. Pearson, and K. Moore. 2010. TBAA reduction in reactors
simulating distribution system pipes. Journal of the American Water Works Association. 102(3):
99-106.
Aschengrau, A., S. Zierler, and A. Cohen. 1989. Quality of Community Drinking Water and the
Occurrence of Spontaneous Abortions. Archives of Environmental Health. 44: 283-290.
Aschengrau, A., S. Zierler, and A. Cohen. 1993. Quality of community drinking water and the
occurrence of late adverse pregnancy outcomes. Archives of Environmental Health. 48: 105—
113.
Azzeh, J., L. Taylor-Edmonds, and R.C. Andrews. 2015. Engineered biofiltration for
ultrafiltration fouling mitigation and disinfection by-product precursor control. Water Science &
Technology: Water Supply. 15(1): 124-133
Babi, K.G., K.M. Babi, A.D. Nikolaou, and T.D. Lekkas. 2007. Pilot study of the removal of
THMs, HAAs and DOC from drinking water by GAC adsorption. Desalination. 210(210): 215-
224.
Backer, L.C., D.L. Ashley, M.A. Bonin, F.L. Cardinali, S.M. Kieszak, and J.V. Wooten. 2000.
Household exposures to drinking water disinfection by-products: whole blood trihalomethane
levels. Journal of Exposure Analysis and Environmental Epidemiology. 10(4): 321-326.
Bader E.L., S.E. Hrudey, and K.L. Froese. 2004. Urinary excretion half-life of trichloroacetic
acid as a biomarker of exposure to chlorinated drinking water disinfection by-products.
Occupational and Environmental Medicine. 61(8): 715-716.
Baker, J.R., R.J. Edwards, J.M. Lasker, M.R. Moore, and S. Satarug. 2005. Renal and hepatic
accumulation of cadmium and lead in the expression of CYP4F2 and CYP2E1. Toxicology
Letters. 159(2): 182-191.
Balchak, S.K., J.M. Hedge, A.S. Murr, M.L. Mole, and J.M. Goldman. 2000. Influence of the
drinking water disinfection by-product dibromoacetic acid on rat estrous cyclicity and ovarian
follicular steroid release in vitro. Reproductive Toxicology. 14(6): 533-539.
Baribeau, H., S.W. Krasner, R. Chinn, and P.C. Singer. 2005. Impact of biomass on the stability
of HAAs and THMs in a simulated distribution system. Journal of the American Water Works
Association. 97(2): 69-81.
Six-Year Review 3 Technical Support Document	9-2
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Bayless, W. and R.C. Andrews. 2008. Biodegradation of six haloacetic acids in drinking water.
Journal of Water and Health. 6(1): 15-22.
Baylin, S.B., J.G. Herman, J.R. Graff, P.M. Vertino, and J.P. Issa. 1998. Alterations in DNA
methylation: A fundamental aspect of neoplasia. Advances in Cancer Research. 72: 141-196.
Becker, W., B. Stanford, and E.J. Rosenfeldt. 2013. Guidance on complying with Stage 2 D/DBP
regulation. Water Research Foundation. Web Report #4427.
Beggs, K.M.H. and R.S. Summers. 2011. Character and chlorine reactivity of dissolved organic
matter from a mountain pine beetle impacted watershed. Environmental Science and
Technology. 45(13): 5717-5724.
Behzadian, K., M. Alimohammadnejad, A. Ardeshir, F. Jalilsani, and H. Vasheghani. 2012. A
novel approach for water quality management in water distribution systems by multi-objective
booster chlorination. International Journal of Civil Engineering. 10: 51-60.
Bellar, T.A, J.J. Lichtenberg, and R.C. Kroner. 1974. The occurrence of organohalides in
chlorinated drinking water. Journal of the American Waterworks Association. 66(12):703-706.
Bercz, J.P., L.L. Jones, L. Garner, L. Murray, D. Ludwig, and J. Boston. 1982. Subchronic
toxicity of chlorine dioxide and related compounds in drinking water in the nonhuman primate.
Environmental Health Perspectives. 46: 47-55.
Bhat, H.K., M.F. Kanz, G.A. Campbell, and G.A.S. Ansari. 1991. Ninety-day toxicity study of
chloroacetic acids in rats. Fundamental and Applied Toxicology. 17(2): 240-253.
Bielmeier, S.R., D.S. Best, D.L. Guidici, and M.G. Narotsky. 2001. Pregnancy loss in the rat
caused by bromodichloromethane. Toxicological Sciences. 59(2): 309-315.
Bielmeier, S.R., D.S. Best, and M.G. Narotsky. 2002. Strain comparison of endocrine response
in rats to bromodichloromethane (BDCM) during pregnancy. Toxicologist. 66(S-1): 374.
Bielmeier, S.R., A.S. Murr, D.S. Best, J.M. Goldman, and M.G. Narotsky. 2003. Effects of
bromodichloromethane (BDCM) on ex vivo luteal function in the pregnant F344 rats.
Toxicologist. 72(S-1): 26-27.
Bielmeier, S.R., D.S. Best, and M.G. Narotsky, M.G. 2004. Serum hormone characterization and
exogeneous hormone rescue of bromodichloromethane-induced pregnancy loss in the F344 rat.
Toxicological Sciences. 77(1): 101-108.
Bielmeier, S.R., A.S. Murr, D.S. Best, R.A. Harrison, R.A. Pegram, J.M. Goldman, and M.G.
Narotsky. 2007. Effects of bromodichloromethane on ex vivo and in vitro luteal function and
bromodichloromethane tissue dosimetry in the pregnant F344 rat. Toxicology in Vitro. 21(5):
919-928.
Six-Year Review 3 Technical Support Document	9-3
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Blank, V., H.M. Shukairy, and J. McLain. 2002. Unregulated Organic DBPs in ICR Finished
Water and Distribution Systems. In Information Collection Rule Data Analysis, Chapter 11
Denver, CO: American Water Works Association Research Foundation.
Board, P.G. and M.W. Anders. 2011. Glutathione transferase zeta: discovery, polymorphic
variants, catalysis, inactivation, and properties of Gstzl-/- mice. Drug Metabolism Reviews. 43,
215-225.
Bond, R.G. and F.A. Digiano. 2004. Evaluating GAC performance using the ICR database.
Journal of the American Water Works Association. 96(6): 96-104.
Bond, T., J. Huang, M.R. Templeton, andN. Graham. 2011. Occurrence and control of
nitrogenous disinfection by-products in drinking water - a review. Water Research. 45(15): 4341-
4354.
Bond, T., E.H. Goslan, S.A. Parsons, and B. Jefferson. 2012a. A critical review of
trihalomethane and haloacetic acid formation from natural organic matter surrogates.
Environmental Technology Reviews. 1(1): 93-113.
Bond, T., M.R. Templeton, andN. Graham. 2012b. Precursors of nitrogenous disinfection by-
products in drinking water - a critical review and analysis. Journal of Hazardous Materials. 235:
1-16.
Bond, T., M.R. Templeton, O. Rifai, H. Ali, andN.J.D. Graham. 2014. Chlorinated and
nitrogenous disinfection by-product formation from ozonation and post-chlorination of natural
organic matter surrogates. Chemosphere. Ill: 218-224.
Booth, S., C. Fonesca, J. Sutherland, P. Carlson, and G. Amy. 2006. DBP control in high
bromide water while using free chlorine during distribution. Water Research Foundation.
Borzelleca, J.F. and R.A. Carchman. 1982. Effects of selected organic drinking water
contaminants on male reproduction. EPA 600-1-82-009 (NTIS PB82-259847). Research
Triangle Park, NC: U.S. Environmental Protection Agency. August 2007.
Bose, P. and D. A. Reckhow. 2007. The effect of ozonation on natural organic matter removal by
alum coagulation. Water Research. 41: 1516-1524.
Bougeard, C.M.M., E.H. Goslan, B. Jefferson, and S.A. Parsons. 2010. Comparison of the
disinfection by-product formation potential of treated waters exposed to chlorine and
monochloramine. Water Research. 44(3): 729-740.
Bove, F.J., M.C. Fulcomer, J.B. Klotz, J. Esmart, E.M. Dufficy, and J.E. Savrin. 1995. Public
drinking water contamination and birth outcomes. American Journal of Epidemiology. 141(9):
850-862.
Bove F.J., Y. Shim, and P. Zeitz. 2002. Drinking Water Contaminants and Adverse Pregnancy
Outcomes: A Review. Environmental Health Perspectives. llO(Suppl.l): 61-74.
Six-Year Review 3 Technical Support Document	9-4
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Bove, G.E., P. A. Rogerson, and J.E. Vena. 2007a. Case control study of the geographic
variability of exposure to disinfectant byproducts and risk for rectal cancer. International Journal
of Health Geographies. 6(18).
Bove, G.E., P. A. Rogerson, and J.E. Vena. 2007b. Case-control study of the effects of
trihalomethanes on urinary bladder cancer risk. Archives of Environmental & Occupational
Health. 62(1): 39-47.
Boyd, J.M., S.E. Hrudey, X-F. Li, and S.D. Richardson. 2011. Solid-phase extraction and high-
performance liquid chromatography mass spectrometry analysis of nitrosamines in treated
drinking water and wastewater. TrAC Trends in Analytical Chemistry. 30(9): 1210-1421.
Boyer, T.H. 2015. Meta-analysis of trihalomethane formation models and application to bromide
intrusion. In Recent Advances in Disinfection By-Products. ACS Symposium Series 1190.
Bridgeman, J., B. Magdalena, and A. Baker. 2011. The application of fluorescence spectroscopy
to organic matter characterization in drinking water treatment. Reviews in Environmental
Science and Biotechnology. 10(3): 277-290.
Brooke, E., and M.R. Collins. 2011. Post-treatment aeration to reduce THMs. Journal of the
American Water Works Association. 103(10): 84-96.
Brown, J., R. S. Summers, M. LeChevallier, H. Collins, J. A. Roberson, S. Hubbs, and E.
Dickenson. 2015. Biological drinking water treatment? Naturally. Journal of the American Water
Works Association. 107(12): 20-31.
Bull, R.J. 2000. Mode of action of liver tumor induction by trichloroethylene and its metabolites,
trichloroacetate and dichloroacetate. Environmental Health Perspectives. 108(Suppl 2): 241-59.
Bull, R.J. 2002. Contribution of dichloroacetate and trichloroacetate to liver tumor induction in
mice by trichloroethylene. Toxicology and Applied Pharmacology. 182(1): 55-65.
Bull, R.J. 2012. Toxicological Evaluation of Experimental Data that Informs the Magnitude of
Cancer Risk from Disinfection By-Products. In: Disinfection By-Products and Human Health,
eds. Hrudey, S.E. and J.W.A. Charrois, Chapter 10, pp. 179-212. London: IWA Publishing.
Bull, R.J. and J.A. Cotruvo. 2006. Research strategy for developing key information on
bromate's mode of action. Toxicology. 221(2-3): 135-144.
Bull, R.J. and J. Cotruvo. 2013. Nongenotoxic mechanisms involved in bromate-induced cancer
in rats. Journal of American Water Works Association. 105(12): E709-E720.
Bull, R.J., I.M. Sanchez, N.A. Nelson, J.L. Larson, and A.J. Lansing. 1990. Liver tumor
induction in B6C3F1 mice by dichloroacetate and trichloroacetate. Toxicology. 63(3): 341-359.
Bull, R.J., D.A. Reckhow, V. Rotello, O.M. Bull, and J. Kim. 2006. Use of Toxicological and
Chemical Models to Prioritize DBP Research. Denver, CO: AWWA Research Foundation and
American Water Works Association.
Six-Year Review 3 Technical Support Document	9-5
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Bull, R.J., G. Rice, and L.K. Teuschler. 2009. Determinants of whether or not mixtures of
disinfection by-products are similar - part a. Journal of Toxicology and Environmental Health.
72(7): 437-460.
Bull, R.J., D.A. Reckhow, A.R. Humpage, C. Joll, and S.E. Hrudey. 2011. Potential carcinogenic
hazards of non-regulated disinfection by-products: Haloquinones, halocyclopentene, and
cyclohexene derivatives, N-halamines, halonitriles, and heterocyclic amines. Toxicology. 286(1-
3): 1-19.
Bull, R.J., N. Kolisetty, X. Zhang, S. Muralidhara, O. Quinones, K. Lim, Z. Guo, J. Cotruvo, J.
Fisher, X. Yang, D. Delker, S. Snyder, and B. Cummings. 2012. Absorption and disposition of
bromate in F344 rats. Toxicology. 300, 83-91.
California Environmental Protection Agency (Cal EPA). 2009. Public Health Goal for Bromate
in Drinking Water. Office of Environmental Health Hazard Assessment, Pesticide and
Environmental Toxicology Section. Available online at:
http://oehha.ca.gov/media/downloads/water/chemicals/phg/bromatephg010110 O.pdf.
Cal EPA. 2010. Evaluation of Chloropcrin as a Toxic Air Contmaninant. Part B Human Health
Assessment. Department of Pesticide Regulation. Available online at:
http://www.cdpr.ca.gov/docs/emon/pubs/tac/part b 0210.pdf.
California Environmental Protection Agency/Department of Pesticide Regulation (Cal
EPA/DPR). 1997. Summary of toxicology data on formaldehyde. Available online at:
http://vvvvvv.cdpr.ca.gov/docs/risk/toxsums/pdfs/295.pdf.
Callinan, C.W., J.P. Hassett, J.B. Hyde, R.A. Entringer, and R.K. Klake. 2013. Proposed nutrient
criteria for water supply lakes and reservoirs. Journal of the American Water Works Association.
105(4): E157-E172.
Cantor, K.P., R. Hoover, and P. Hartge. 1985. Drinking Water Source and Bladder Cancer: A
Case-Control Study. In: Water Chlorination: Chemistry, Environmental Impact and Health
Effects, vol. 5, eds. R.L. Jolley, R.J. Bull, and W.P. Davis, 1:145-152. Chelsea, MI: Lewis
Publishers, Inc.
Cantor, K.P., R. Hoover, P. Hartge, T.J. Mason, D.T. Silverman, R. Altman, D.F. Austin, M.A.
Child, C.R. Key, L.D. Marrett, M.H. Myers, A.S. Narayana, L.I. Levin, J.W. Sullivan, G.M.
Swanson, D.B. Thomas, and D.W. West. 1987. Bladder cancer, drinking water source, and tap
water consumption: A case-control study. Journal of the National Cancer Institute. 79(6): 1269-
1279.
Cantor, K.P., C.F. Lunch, M. Hildesheim, M. Dosemeci, J. Lubin, M. Alavanja, and G.F. Craun.
1998. Drinking water source and chlorination byproducts I. Risk of Bladder Cancer.
Epidemiology. 9(1): 21-28.
Cantor, K.P, C.F. Lynch, M.E. Hildesheim, M. Dosemeci, J. Lubin, M. Alavanja, and C.F.
Craun. 1999. Drinking water source and chlorination byproducts in Iowa. III. Risk of brain
cancer. American Journal of Epidemiology. 150(6): 552-560.
Six-Year Review 3 Technical Support Document	9-6
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Cantor, K.P., C.M. Villanueva, D.Y. Silverman, J.D. Figueroa, F.X. Real, M. Garcia-Closas, N.
Malats, S. Chanock, M. Yeager, A. Tardon, R. Garcia-Closas, C. Serra, A. Carrato, G. Castano-
Vinyals, C. Samanic, N. Rothman, and M. Kogevinas. 2010. Polymorphisms in GSTT1, GSTZ1,
and CYP2E1, disinfection by-products, and risk of bladder cancer in Spain. Environmental
Health Perspectives. 118(11): 1545-1550.
Carlton, B.D. and M.K. Smith. 1985. Reproductive effects of alternate disinfectants and their
byproducts. In: Water chlorination: environmental impact and health effects, vol. 5, eds. Jolley,
R.L., et al., pp. 295-305. Chelsea, MI: Lewis Publications.
Carlton, B.D., P. Barlett, A. Basaran, K. Colling, I. Osis, and M.K. Smith. 1986. Reproductive
effects of alternative disinfectants. Environmental Health Perspectives. 69: 237-241.
Carlton, B.D., D.L. Habash, A.H. Basaran, E.L. George, and M.K. Smith. 1987. Sodium chlorite
administration in Long-Evans rats: reproductive and endocrine effects. Environmental Research.
42(1): 238-245.
Carlton, B.D., A.H. Basaran, L.E. Mezza, E.L. George, and M.K. Smith. 1991. Reproductive
effects in Long-Evans rats exposed to chlorine dioxide. Environmental Research. 56(2): 170-177.
Cedergren M.I., A.J. Selbing, O. Lofman, and B. Kallen. 2002. Chlorination byproducts and
nitrate in drinking water and risk of congenital cardiac defects. Environmental Research. 89(2):
124-130.
Chang, C.C., S.C. Ho, L.Y. Wang, and C.Y. Yang. 2007. Bladder cancer in Taiwan: Relationship
to trihalomethane concentrations present in drinking-water supplies. Journal of Toxicology and
Environmental Health, Part A. 70(20): 1752-1757.
Chang, H., C. Chen, and G. Wang, 2013. Characteristics of C-, N-DBPs formation from
nitrogen-enriched dissolved organic matter in raw water and treated wastewater effluent. Water
Research. 47(8): 2729-41.
Chemical Manufacturers Association (CMA). 1996. Sodium chlorite: Drinking water rat two
generation reproductive toxicity study. Quintiles Report Ref. CMA/17/96.
ChemlDPlus. 2015. Available on the Internet at:
https://chem.nlm.nih.gov/chemidplus/chemidlite.isp. Accessed December 3, 2015.
Chen, B., and P. Westerhoff 2010. Predicting disinfection by-product formation potential in
water. Water Research 44(13): 3755-3762.
Chen, H.W., C.Y. Chen, and G.S. Wang. 2011. Performance evaluation of the UV/H202 process
on selected nitrogenous organic compounds: Reductions of organic contents vs. corresponding
C-, N-DBPs formations. Chemosphere. 85(4): 591-597.
Chen, W., Zhang, Z., Li, Q., and Wang, H. 2012. Adsorption of bromate and competition from
oxyanions on cationic surfactant-modified granular activated carbon (GAC). Chemical
Engineering Journal. 203:319-325.
Six-Year Review 3 Technical Support Document	9-7
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Chen, B., Y. Qian, M. Wu, L. Zhu, B. Hu, and X-F. Li. 2014. Identification of precursors and
mechanisms of tobacco-specific nitrosamine formation in water during chloramination.
Environmental Science and Technology. 49(1): 459-466.
Chen, J., G.C. Douglas, T.L Thirkill, P.N. Lohstroh, S.R. Bielmeir, M.G. Narotsky, D.S. Best,
R.A. Harrison, K. Natarajan, R.A. Pegram, J.W. Overstreet, and B.L. Lasley. 2003. Effect of
bromodichloromethane on chorionic gonadotropin secretion by human placental trophoblast
cultures. Toxicological Sciences. 76(1): 75-82.
Chen, J., T.L. Thirkill, P.N. Lohstroh, S.R. Bielmeir, M.G. Narotsky, D.S. Best, R.A. Harrison,
K. Natarajan, R.A. Pegram, J.W. Overstreet, B.L. Lasley, and G.C. Douglas. 2004.
Bromodichloromethane inhibits human placental trophoblast differentiation. Toxicological
Sciences. 78(1): 166-174.
Chiu, H.F., S.S. Tsai, T.N. Wu, and C.Y. Yang. 2010. Effect modification of the association
between trihalomethanes and pancreatic cancer by drinking water hardness: Evidence from an
ecological study. Environmental Research. 110(5): 513-518.
Chiu, C-A., P. Westerhoff, and A. Ghosh. 2012. GAC removal of organic nitrogen and other
DBP precursors. Journal of the American Water Works Association. 104(7): 406-415.
Chisholm, K., A. Cook, C. Bower, and P. Weinstein. 2008. Risk of birth defects in Australian
communities with high levels of brominated disinfection by-products. Environmental Health
Perspectives. 116(9): 1267-1273.
Chlorine Chemistry Council (CCC). 2000a. Oral (Drinking Water) Range-finding
Developmental Toxicity Study of Bromodichloromethane (BDCM) in Rabbits. Protocol No.
2403-002P. Arlington, VA: Chlorine Chemistry Council.
CCC. 2000b. Oral (Drinking Water) Developmental Toxicity Study of Bromodichloromethane in
Rabbits. Protocol No. 2403-002. Arlington, VA: Chlorine Chemistry Council.
CCC. 2000c. Oral (Drinking Water) Range-finding Developmental Toxicity Study of
Bromodichloromethane in Rats. Protocol No. 2403-001P. Arlington, VA: Chlorine Chemistry
Council.
CCC. 2000d. Oral (Drinking Water) Developmental Toxicity Study of Bromodichloromethane in
Rats. Protocol No. 2403-003. Arlington, VA: Chlorine Chemistry Council.
Chow, A., S. Gao, and R. Dahlgren. 2005. Physical and chemical fractionation of dissolved
organic matter and trihalom ethane precursors: a review. Journal of Water Supply Research and
Technology. 54(8): 475-507.
Chowdhury, S., P. Champagne, and P.J. McLellan. 2009. Models for predicting disinfection by-
product (DBP) formation in drinking waters: a chronological review. Science of the Total
Environment. 407(14): 4189-4206.
Six-Year Review 3 Technical Support Document	9-8
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Chowdhury, Z., A. Traviglia, J. Carter, T. Brown, R.S. Summers, C. Corwin, T. Zearley, M.
Thurman, I Ferrara, J. Olson, R. Thacker, and P. Barron. 2010. Cost-effective regulatory
compliance with GAC biofilters. Water Research Foundation. Project #4155.
Chowdhury, S., M.J. Rodriguez, R. Sadiq, and J. Serodes. 2011. Modeling DBPs formation in
drinking water in residential plumbing pipes and hot water tanks. Water Research. 45(1): 337-
347.
Christ, S.A., E.J. Read, J.A. Stober, and M.K. Smith. 1995. The developmental toxicity of
bromochloroacetonitrile in pregnant Long-Evans rats. International Journal of Environmental
Health Research. 5(2): 175-188.
Christ, S.A, E.J. Read, J.A. Stober, and M.K. Smith. 1996. Developmental effects of
trichloroacetonitrile administered in corn oil to pregnant Long-Evans rats. Journal of Toxicology
and Environmental Health. 47(3): 233-247.
Christian, M. (eds.). 1991. Final Report on the Safety Assessment of Chloroacetamide.
International Journal of Toxicology. 10(1): 21-32.
Christian, M.S., R.G. York, A.M. Hoberman, R.M. Diener, R.M., and L.C. Fisher. 2001a. Oral
(drinking water) developmental toxicity studies of bromodichloromethane (BDCM) in rats and
rabbits. International Journal of Toxicology. 20(4): 225-237.
Christian, M.S., R.G. York, A.M. Hoberman, R.M. Diener, L.C. Fisher, and G.A. Gates. 2001b.
Biodisposition of dibromoacetic acid (DBA) and bromodichloromethane (BDCM) administered
to rats and rabbits in drinking water during range-finding reproduction and developmental
toxicity studies. International Journal of Toxicology. 20(4): 239-253.
Christian, M.S., R.G. York, A.M. Hoberman, A.M., L.C. Fisher, and W.R. Brown. 2002a. Oral
(drinking water) two-generation reproductive toxicity study of bromodichloromethane (BDCM)
in rats. International Journal of Toxicology. 21(2): 115-146.
Christian, M.S., R.G. York, A.M. Hoberman, J. Frazee, L.C. Fisher, W.R. Brown, and D.M.
Creasy. 2002b. Oral (drinking water) two-generation reproductive toxicity study of
dibromoacetic acid (DBA) in rats. International Journal of Toxicology. 21(4): 237-276.
Chu, W., N. Gao, S.W. Krasner, M.R. Templeton, and D. Yin. 2012. Formation of halogenated
C-, N-DBPs from clor(am)ination and UV irradiation of tyrosine in drinking water.
Environmental Pollution. 161: 8-14.
Chu, W., D. Yao, N. Gao, T. Bond, and M.R. Templeton. 2015. The enhanced removal of
carbonaceous and nitrogenous disinfection by-product precursors using integrated permanganate
oxidation and powdered activated carbon adsorption pretreatment. Chemosphere. 141: 1-6.
Chuang, Y-H., A. Y-C. Lin, X-H. Wang, and H-H. Tung. 2013. The contribution of dissolved
organic nitrogen and chloramines to nitrogenous disinfection by-product formation from natural
organic matter. Water Research. 47(3): 1308-1316.
Six-Year Review 3 Technical Support Document	9-9
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Cimetiere, N., F. Dossier-Berne, and J. De Laat. 2010. Effect of some parameters on the
formation of chloroform during chloramination of aqueous solutions of resorcinol. Water
Research. 44(15): 4497-4504.
Cicmanec, J.L., L.W. Condie, G.R. Olson, and S.R. Wang. 1991. 90-day toxicity study of
dichloroacetate in dogs. Fundamental and Applied Toxicology. 17(2): 376-389.
Collins, M.R., T.T. Eighmy, J.M. Fenstermacher Jr., and S.K. Spanos. 1992. Removing natural
organic matter by conventional slow sand filtration. Journal of the American Water Works
Association. 84(5): 80-90.
Collivignarelli, C., S. Sorlini, and M. Belluati. 2006. Chlorite removal with GAC. Journal of the
American Water Works Association. 98(12): 74-81.
Compton, D. 2007. Making a business case for mixed oxidants. Opflow. 33(6): 16-17.
Condie, L.W., F.B. Daniel, G.R. Olson, and M. Robinson. 1994. Ten and ninety-day toxicity
studies of chloropicrin in Sprague-Dawley rats. Drug and Chemical Toxicology. 17(2): 125-137.
Coral, L.A., A. Zamyadi, B. Barbeau, F.J. Bessetti, F.R. Lapolli, and M. Prevost. 2013.
Oxidation of microcystis aeruginosa and anabena flos-aquae by ozone: impacts on cell integrity
and chlorination by-product formation. Water Research. 47.9: 2983-2994.
Cornwell, D., N. McTigue, and R. Brown. 2014. Impact of Increased Bromide in Water Sources
on Drinking Water DBPs. AWWA WQTC.
Cornwell, D.A., S. Krasner, W.A. Mitch, and J.J. Pignatello. 2015. Investigating coagulant aid
alternatives to PolyDADMAC polymers. Project #4452.
Cordier, S., J. Clavel, J.C. Limasset, L. Boccon-Gibod, N. Le Moual, L. Mandereau, and D.
Hemon. 1993. Occupational risks of bladder cancer in France: a multicentre case-control study.
International Journal of Epidemiology. 22(3):403-l 1.
Costet, N., C.M. Villanueva, J.J.K. Jaakkola, M. Kogevinas, K.P. Cantor, W.D. King, C.F.
Lynch, M.J. Nieuwenhuijsen, and S. Cordier. 2011. Water disinfection by-products and bladder
cancer: is there a European specificity? A pooled and meta-analysis of European case-control
studies. Occupational and Environmental Medicine. 68(5): 379-385.
Costet N., R. Garlantezec, C. Monfort, F. Rouget, B. Gagniere, C. Chevrier, and S. Cordier.
2012. Environmental and urinary markers of prenatal exposure to drinking water disinfection by-
products, fetal growth, and duration of gestation in the PELAGIE birth cohort (Brittany, France,
2002-2006). American Journal of Epidemiology. 175(4): 263-75.
Couri, D. and M.S. Abdel-Rahman. 1980. Effect of chlorine dioxide and metabolites on
glutathione dependent system in rat, mouse and chicken blood. Journal of Environmental
Pathology, Toxicology, and Oncology. 3(1-2): 451-460.
Six-Year Review 3 Technical Support Document	9-10
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Couri, D., C.H. Miller, R.J. Bull, J.M. Delphia, and E.M. Ammar. 1982. Assessment of maternal
toxicity, embryotoxicity and teratogenic potential of sodium chlorite in Sprague-Dawley rats.
Environmental Health Perspectives. 46: 25-29.
Cowman, G.A. and P.C. Singer. 1996. Effect of bromide ion on haloacetic acid speciation
resulting from chlorination and chloramination of aquatic humic substances. Environmental
Science and Technology. 30(1): 16-24.
Crallan, R.A., N.T. Georqopoulos, and J. Southgate. 2006. Experimental models of human
bladder carcinogenesis. Carcinogenesis. 27(3): 374-381.
Craun, G.C. (eds.). 1998. EPA panel report and recommendations for conducting
epidemiological research on possible reproductive and developmental effects of exposure to
disinfected drinking water. Research Triangle Park, NC: U.S. Environmental Protection Agency,
National Health and Environmental Effects ResearchLaboratory.
Criquet, J., S. Allard, E. Salhi, C.A. Joll, A. Heitz, and U. von Gunten. 2012. Iodate and iodo-
trihalomethane formation during chlorination of iodide-containing waters: role of bromide.
Environmental Science and Technology. 46(13): 7350-7357.
Cruickshank, J.R. 2010. Hydraulic models shed light on water age. Opflow. 36(6): 18-21.
Cummings, A.M. and J.M. Hedge. 1998. Dibromoacetic acid does not adversely affect early
pregnancy in rats. Reproductive Toxicology. 12(4): 445-448.
Cummings, B.S., J.M. Lasker, and L.H. Lash. 2000. Expression of Glutathione-Dependent
Enzymes and Cytochrome P450s in Freshly Isolated and Primary Cultures of Proximal Tubular
Cells from Human Kidney. Journal of Pharmacology and Experimental Therapeutics. 293(2):
677-685.
Curran, J.E., S.R. Weinstein, and L.R. Griffiths. 2000. Polymorphisms of glutathione S-
transferase genes (GSTM1, GSTP1 and GSTT1) and breast cancer susceptibility. Cancer Letters.
153(1-2): 113-120.
Dad, A., C.H. Jeong, J. A. Pals, E.D. Wagner, and M.J. Plewa. 2013. Pyruvate remediation of cell
stress and genotoxicity induced by haloacetic acid drinking water disinfection by-products.
Environmental and Molecular Mutagenesis. 54(8):629-637.
Daniel, F.B., L.W. Condie, M. Robinson, J.A. Stober, R.G. York, G.R. Olson, and S.R. Wang.
1990. Comparative subchronic toxicity studies of three disinfectants. Journal of the American
Waterworks Association. 82: 61-69.
Daniel, F.B., H.P. Ringhand, M. Robinson, J.A. Stober, G.R. Olson, andN.P. Page. 1991.
Comparative subchronic toxicity of chlorine and monochloramine in the B6C3F1 mouse. Journal
of the American Water Works Association. 83: 68-75.
Six-Year Review 3 Technical Support Document	9-11
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Daniel, F.B., A.B. DeAngelo, J.A. Stober, G.R. Olson, andN.P. Page. 1992.
Hepatocarcinogenicity of chloral hydrate, 2-chloroacetaldehyde, and dichloroacetic acid in the
maleB6C3Fl mouse. Fundamental and Applied Toxicology. 19(2): 159-168.
Danileviciute, A., R. Grazuleviciene, J. Vencloviene, A. Paulauskas, and M.J. Nieuwenhuijsen.
2012. Exposure to drinking water trihalomethanes and their association with low birth weight
and small gestational age in genetically susceptible women. International Journal of
Environmental Research and Public Health. 9(12): 4470-4485.
Darko, B., J. Jiang, H. Kim, L. Machala, R. Zboril, and V.K. Sharma. 2014. Advances made in
understanding the interaction of ferrate (VI) with natural organic matter in water. Water
Reclamation and Sustainability. 183-197.
Daugherty, E., B. Mayer, and M. Abbaszadegan. 2011. Analysis of photocatalysis
implementation in water treatment systems for the removal of DBP precursors. WQTC 2011
Conference Proceedings.
DeAngelo, A.B., F.B. Daniel, B.M. Most and G.R. Olson. 1996. The carcinogenicity of
dichloroacetic acid in the male Fischer 344 rat. Toxicology. 114(3): 207-221.
DeAngelo, A.B., F.B. Daniel, B.M. Most, and G.R. Olson. 1997. Failure of monochloroacetic
acid and trichloroacetic acid administered in the drinking water to produce liver cancer in male
F344/N rats. Journal of Toxicology and Environmental Health. 52(5): 425-445.
DeAngelo, A.B., M.H. George, S.R. Kilburn, T.M. Moore, and D.C. Wolf. 1998.
Carcinogenicity of potassium bromate administered in the drinking water to male B6C3F1 mice
and F344/N rats. Toxicologic Pathology. 26(5): 587-594.
DeAngelo, A.B., M.H. George, and D.E. House. 1999. Hepatocarcinogenicity in the male
B6C3F1 mouse following a life-time exposure to dichloroacetic acid in the drinking water: dose-
response determination and modes of action. Journal of Toxicology and Environmental Health.
58(8): 485-507.
DeAngelo, A.B., D.R. Geter, D.W. Rosenberg, C.K. Crary, and M.H. George. 2002. The
induction of aberrant crypt foci (ACF) in the colons of rats by trihalomethanes administered in
the drinking water. Cancer Letters. 187(1-2): 25-31.
DeAngelo, A.B., F.B. Daniel, D.M. Wong, and M.H. George. 2008. The induction of
hepatocellular neoplasia by trichloroacetic acid administered in the drinking water of the male
B6C3F1 mouse. Journal of Toxicology and Environmental Health, Part A. 71(16): 1056-1068.
Delatolla, R., C. Seguin, S. Springthorpe, E. Gorman, A. Campbell, and I. Douglas. 2015.
Disinfection byproduct formation during biofiltration cycle: Implications for drinking water
production. Chemosphere. 136: 190-197.
Delker, D., G. Hatch, J. Allen, B. Crissman, M. George, J. Geter, S. Kilburn, T. Moore, G.
Nelson, B. Roop, R. Slade, A. Swank, W. Ward, and A. DeAngelo. 2006. Molecular biomarkers
of oxidative stress associated with bromate carcinogenicity. Toxicology. 221(2-3): 158-165.
Six-Year Review 3 Technical Support Document	9-12
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
DeMarini, D.M., M.L. Shelton, S.H. Warren, T.M. Ross, J.Y. Shim, A.M. Richard, andR.A.
Pegram. 1997. Glutathione S-transferase-mediated induction of GC to AT transitions by
halomethanes in Salmonella. Environmental and Molecular Mutagenesis. 30(4): 440-447.
Dickenson, E., R.S. Summers, J-P. Croue, and H. Gallard. 2008. Haloacetic acid and
trihalomethane formation from the chlorination and bromination of aliphatic P-dicarbonyl acid
model compounds. Environmental Science and Technology. 42(9): 3226-3233.
Do, M.T., N.J. Birkett, K.C. Johnson, D. Krewski, P. Villeneuva, and the Canadian Cancer
Registries Epidemiology Research Group. 2005. Chlorination disinfection by-products and
pancreatic cancer risk. Environmental Health Perspectives. 113(4): 418-424.
Dodd, D., D. Layko, K. Cantwell, G. Willson, and R. Thomas. 2013. Subchronic toxicity
evaluation of potassium bromate in Fisher 344 rats. Environmental Toxicology and
Pharmacology. 36: 1227-1234.
Dodds, L., W. King, C. Woolcott, and J. Pole. 1999. Trihalomethanes in public water supplies
and adverse birth outcomes. Epidemiology. 10(3): 233-7.
Dodds L. and W.D. King. 2001. Relation between trihalomethane compounds and birth defects.
Occupupational Environmental Medicine. 58: 443-446.
Dodds L., W. King, A.C. Allen, A. Armson, D.B. Fell, and C. Nimrod. 2004. Trihalomethanes in
public water supplies and risk of stillbirth. Epidemiology. 15: 179-186.
Dotson, A.D., C.E. Rodriguez, and K.G. Linden. 2012. UV disinfection implementation status in
US water treatment plants. Journal of the American Water Works Association. 104(5): 318-324.
Drikas, M., M. Dixon, and J. Morran. 2011. Long term case study of MIEX pre-treatment in
drinking water; understanding NOM removal. Water Research. 45(4): 1539-1548.
Duirk, S.E., C. Lindell, C.C. Cornelison, J. Kormos, T.A. Ternes, M Attene-Ramos, J. Osiol,
E.D. Wagner, M.J. Plewa, and S.D. Richardson. 2011. Formation of toxic iodinated disinfection
by-products from compounds used in medical imaging. Environmental Science and Technology.
45(16): 6845-6854.
Dunn, S.E., and D.R.U. Knappe. 2013. DBP Precursor and Micropollutant Removal by
Powdered Activated Carbon. Water Research Foundation. Web Report #4294.
Dunnick, J.K. and R.L. Melnick. 1993. Assessment of the carcinogenic potential of chlorinated
water: experimental studies of chlorine, chloramine, and trihalomethanes. Journal of the National
Cancer Institute. 85(10): 817-822.
Duranceau, S.J. 2015. Spray Aeration as a Trihalomethane Control Measure. 2015 Florida Water
Resources Conference Proceedings.
Six-Year Review 3 Technical Support Document	9-13
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Echigo, S., S. Itoh, and M. Kuwahara. 2007. Bromide removal by hydrotalcite-like compounds
in a continuous system. Water Science and Technology: A Journal of the International
Association on Water Pollution Research. 56(11): 117-122.
Eighmy, T.T., M.R. Collins, J.P. Malley, J. Royce, and D. Morgan. 1993. Biological Enhanced
Slow Sand Filtration for Removal of Natural Organic Matter. Denver, CO: AWWA Research
Foundation, pp. 33-69.
Emelko, M.B., U. Silins, K.D. Bladon, M. Stone, C. Williams, M. Wagner, A. Martens, and X.
Geng. 2013. The lost creek wildfire of southern Alberta, Canada: 10 years, 7 watersheds and
continued impacts. Water Research Foundation Workshop: Wildfire Readiness and Response
Workshop - Is Your Utility Prepared? April 4, 2013.
Emmert, G., C. Henson, A. Brown, and P. Simone. 2011. Investigating the presence of HAAs
and THMs in sodium hypochlorite feedstocks. Water Quality Technology Conference (WQTC)
2011 Conference Proceedings.
Emmert, G.L., P.S. Simone Jr., Y.Y. Choo, C.M. Henson, A.W. Brown, T.E. Watts III, W.E.
Stephens III, J.F. Williamson, and P.L. Ranaivo. 2013. Analysis of bulk sodium hypochlorite
feedstock for the presence of HAAs and other DBPs. Water Research Foundation. Project #4412.
Epstein, D.L., G.A. Nolen, J.L. Randall, S.A. Christ, E.J. Read, J.A. Stober, and M.K. Smith.
1992. Cardiopathic effects of dichloroacetate in the fetal Long-Evans rat. Teratology. 46(3): 225-
235.
Fan, X., Y. Tao, D. Wei, X. Zhang, Y. Lei, and H. Noguchi. 2015. Removal of organic matter
and disinfection by-products precursors in a hybrid process combining ozonation with ceramic
membrane ultrafiltration. Frontiers of Environmental Science & Engineering. 9(1): 112-120.
Fang, J. Y., J. Ma, X. Yang, and C. Shang. 2010a. Formation of carbonaceous and nitrogenous
disinfection by-products from the chlorination of microcystis aeruginosa. Water Research. 44(6):
1934-1940.
Fang, J., X. Yang, J. Ma, C. Shang, and Q. Zhao. 2010b. Characterization of algal organic matter
and formation of DBPs from chlor(am)ination. Water Research. 44(20): 5897-5906.
Farre, M.J., J. Reungoat, F.X. Argaud, M. Rattier, J. Keller, and W. Gernjak. 2011. Fate ofN-
nitrosodimethylamine, trihalomethane and haloacetic acid precursors in tertiary treatment
including biofiltration. Water Research. 45(17): 5695-5704.
Fenster L., K. Waller, G. Windham, T. Henneman, M. Anderson, P. Mendola, J.W. Overstreet,
and S.H. Swan. 2003. Trihalomethane levels in home tap water and semen quality.
Epidemiology. 14: 650-658.
Fiske, P.S., J. Oppenheimer, R. Moore, and R. Everett. 2011. In-tank aeration predicts and
reduces THMs. Opflow. 37(11): 22-24.
Six-Year Review 3 Technical Support Document	9-14
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Fisher, J.W., S.R. Channel, J.S. Eggers, P.D. Johnson, K.L. MacMahon, C.D. Goodyear, G.L.
Sudberry, D.A. Warren, J.R. Latendresse, and L.J. Graeter. 2001. Trichloroethylene,
trichloroacetic acid, and dichloroacetic acid: do they affect fetal rat heart development?
International Journal of Toxicology. 20(5): 257-267.
Flowers, R.C. and P.C. Singer. 2013. Anion exchange resins as a source of nitrosamines and
nitrosamine precursors. Environmental Science and Technology. 47:13:7365.
Food and Drug Administration (FDA). 1994. Section 178.1010 Sanitizing Solutions. 21 CFR
312.
Forssen, U., A. Herring, D. Savitz, M. Nieuwenhuijsen, P. Murphy, P. Singer, and J.M. Wright.
2007. Predictors of use and consumption of public drinking water among pregnant women.
Journal of Exposure Science and Environmental Epidemiology. 17(2): 159-169.
Francis, R.A., M.J. Small, and J.M. VanBriesen. 2009. Multivariate distributions of disinfection
by-products in chlorinated drinking water. Water Research. 43(14): 3453-3468.
Freedman, M., K.P. Cantor, N.L. Lee, L.S. Chen, H.H. Lei, C.E. Ruhl, and S.S. Wang. 1997.
Bladder cancer and drinking water: A population-based case control study in Washington
County, Maryland. Cancer Causes and Control, 8(5): 738-744.
Froese, K.L., M.I. Sinclair, and S.E. Hrudey. 2002. Trichloroacetic acid as abiomarker of
exposure to disinfection by-products in drinking water: a human exposure trial in Adelaide,
Australia. Environmental Health Perspectives. 110(7): 679-687.
Gallagher M.D., J.R. Nuckols, L. Stallones, and D.A. Savitz. 1998. Exposure to trihalomethanes
and adverse pregnancy outcomes. Epidemiology. 9: 484-489.
Gan, X., D. Kim, and T. Karanfil. 2013. MIEX® treatment of an effluent-impacted stream.
Journal of the American Water Works Association. 105(4): 195-206.
George, M.H., G.R. Olson, D. Doerfler, T. Moore, S. Kilburn, and A.B. DeAngelo. 2002.
Carcinogenicity of bromodichloromethane administered in drinking water to male F344/N rats
and B6C3F1 mice. International Journal of Toxicology. 21(3): 219-230.
Ghosh, A., S. Acquafredda, C. Seidel, C. Corwin, R. Summers, G. Lyons, andM. Xerxis. 2011.
Bench-scale testing pre-treatment and GAC synergies for the Scottsdale, AZ Chaparral WTP.
WQTC 2011 Conference Proceedings.
Ghosh, A., C. Seidel, E. Townsend, R. Pacheco, and C. Corwin. 2015. Reducing volatile
disinfection by-products in treated drinking water using aeration technologies. Water Research
Foundation. Project #4441.
Giller, S., F. Le Curieux, F. Erb, and D. Marzin. 1997. Comparative genotoxicity of halogenated
acetic acids found in drinking water. Mutagenesis. 12(5): 321-328.
Six-Year Review 3 Technical Support Document	9-15
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Glaze, W., H. Weinberg, and H.S. Weinberg. 1993. Identification and occurrence of ozonation
by-products in drinking water. Foundation and American Water Works Association. Available
online at: http://www.waterrf.org/ExecutiveSummaryLibrarv/90625 510 profile.pdf.
Gonce, N. and E.A. Voudrias. 1994. Removal of chlorite and chlorate ions from water using
granular activated carbon. Water Research. 28(5): 1059-1069.
Graves, C.G., G.M. Matanoski, and R.G. Tarfdiff. 2001 Weight of evidence for an association
between adverse reproductive and developmental effects and exposure to disinfection by-
products: a critical review. Regulatory Toxicology Pharmacology. 34: 103-124.
Grazuleviciene, R., M.J. Nieuwenhuijsen, J. Vencloviene, M. Kostopoulou-Karadanelli, S.W.
Krasner, A. Danileviciute, G. Balcius, and V. Kapustinskiene. 2011. Individual exposures to
drinking water trihalomethanes, low birth weight and small for gestational age risk: a prospective
Kaunas cohort study. Environmental Health. 10(32).
Grazuleviciene, R., V. Kapustinskiene, J. Vencloviene, J. Buinauskiene, and M.J.
Nieuwenhuijsen. 2013. Risk of congenital anomalies in relation to the uptake of trihalomethane
from drinking water during pregnancy. Occupational and Environmental Medicine. 70(4): 274-
282.
Grellier, J., J. Bennet, E. Patelarou, R.B. Smith, M.B. Toledano, L. Rushton, D.J. Briggs, and
M.J. Nieuwenhuijsen. 2010. Exposure to disinfection by-products, fetal growth, and prematurity:
A systematic review and meta-analysis. Epidemiology. 21(3): 300-313.
Gruchlik, Y., A. Heitz, C. Joll, V.G. Urs, S. Allard, J. Criquet, S. McDonald, J. Tan, L. Breckler,
F. Bradder, G. Roeszler, and D. Halliwell. 2015. Novel treatment technologies for the
minimisation of bromide and iodide in drinking water. Curtin Water Quality Research Centre:
2015-004.
Guo, Z., Y. Pei, M. Yang, Y. Zhang, J. Zhang, J. Fan, and J. Hirotsuji. 2007. Removal of
organics and control of bromate for a Southern China water supply. Journal of the American
Waterworks Association. 99(10): 110-116.
Hammer, R., and J. VanBriesen. 2012. In fracking's wake: new rules are needed to protect our
health and environment from contaminated wastewater. New York, NY: Natural Resources
Defense Council. Available online at: http://www.nrdc.org/energy/files/fracking-wastewater-
fullreport.pdf.
Hanigan, D., J. Zhang, P. Herckes, S.W. Krasner, C. Chen, and P. Westerhoff 2012. Adsorption
of N-nitrosodimethylamine precursors by powdered and granular activated carbon.
Environmental Science and Technology. 46(22): 12630-12639.
Hanigan, D., E. Inniss, and T.E. Clevenger. 2013. MIEX® and PAC for removal of hydrophilic
DBP precursors. Journal of the American Water Works Association 105(3): 84-92.
Six-Year Review 3 Technical Support Document	9-16
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Hanigan, D., J. Zhang, P. Herckes, E. Zhu, S. Krasner, and P. Westerhoff. 2015. Contribution
and removal of watershed and cationic polymer N-nitrosodimethylamine precursors. Journal of
the American Water Works Association. 107(3): 152-163.
Hao, R., H. Ren, J. Li, Z. Ma, H. Wan, X. Zheng, and S. Cheng. 2012. Use of three-dimensional
excitation and emission matrix fluorescence spectroscopy for predicting the disinfection by-
product formation potential of reclaimed water. Water Research. 46(17): 5765-5776.
Harms, L.L. and C. Owen. 2004. A Guide for the Implementation and Use of Chloramines.
Denver: American Water Works Research Foundation.
Harrington, R.M., R.R. Romano, and L. Irvine. 1995a. Developmental toxicity of sodium
chlorite in the rabbit. Journal of the American College of Toxicology. 14: 109-118.
Harrington, R.M., R.R. Romano, D. Gates, and P. Ridgway. 1995b. Subchronic toxicity of
sodium chlorite in the rat. Journal of the American College of Toxicology. 14: 21-33.
Hazardous Substances Data Bank (HSDB). 2015. Available on the Internet at:
https://toxnet.nlm.nih.gov/cgi-bin/sis/htmlgen7HSDB. Accessed December 3, 2015.
Hassan, A., N.P. Thacker, and J. Bassin. 2010. Trihalomethane formation potential in treated
water supplies in urban metro city. Environmental Monitoring Assessment. 168(l-4):489-97.
Hatt, J.W., C. Lamy, E. Germain, M. Tupper, and S.J. Judd. 2013. NDMA formation in
secondary wastewater effluent. Chemosphere. 9(1): 83-87.
Hayes, J.R., L.W. Condie, and J.F. Borzelleca. 1986. Toxicology of haloacetonitriles.
Environmental Health Perspectives. 69: 183-202.
Health Canada. 1994. Canadian Environmental Protection Act. Human health risk assessment for
priority substances. Catalogue No. En40-215/41E. Minister of Supply and Services Canada,
Ottawa. Protection Branch, Ottawa.
Health Canada. 1997. Formaldehyde in Drinking Water. Federal Provincial Territorial
Committee on Drinking Water. Available online at: http://www.hc-sc.gc.ca/ewh-
semt/alt formats/hecs-sesc/pdf/pubs/water-eau/formaldehvde/formaldehvde-eng.pdf
Health Canada. 1998. Bromate. Environment and Workplace Health. Available online at:
http://www.hc-sc.gc.ca/ewh-semt/pubs/water-eau/bromate/index-eng.php#s5 5
Health Canada. 2006. Guidelines for Canadian Drinking Water Quality: Guideline Technical
Document - Trihalomethanes. Water, Air and Climate Change Bureau, Healthy Environments
and Consumer Safety Branch. Available online at: http://www.hc-sc.gc.ca/ewh-semt/pubs/water-
eau/trihalomethanes/index-eng.php#a 1052
Health Canada. 2008a. Findings and recommendations of the BDCM expert panel meeting.
Available by request to: water_eau@hc.sc.gc.ca.
Six-Year Review 3 Technical Support Document	9-17
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Health Canada. 2008b. Guidelines for Canadian Drinking Water Quality: Guideline Technical
Document - Haloacetic Acids. Water, Air and Climate Change Bureau, Healthy Environments
and Consumer Safety Branch. Available online at: http://www.hc-sc.gc.ca/ewh-semt/pubs/water-
eau/haloaceti/index-eng. php#s 10.4.1
Health Canada. 2008c. Guidance on Chloral Hydrate in Drinking Water. Water, Air and Climate
Change Bureau, Healthy Environments and Consumer Safety Branch. Available online at:
http://healthvcanadians.gc.ca/publications/healthy-living-vie-saine/water-chloral-hvdrate-chloral-
eau/alt/water-chloral-hydrate-chloral-eau-eng.pdf
Health Council of the Netherlands. 2007. Bromodichloromethane; Evaluation of the
carcinogenicity and genotoxicity. The Hague: Health Council of the Netherlands. 2007/050SH.
Henderson, R.K., A. Baker, S.A. Parsons, and B. Jefferson. 2008. Characterization of algogenic
organic matter extracted from cyanobacteria, green algae, and diatoms. Water Research. 42(13):
3435-3445.
Herren-Freund, S.L., M.A. Pereira, M.D. Khoury, and G. Olson. 1987. The carcinogenicity of
trichloroethylene and its metabolites, trichloroacetic acid and dichloroacetic acid, in mouse liver.
Toxicology and Applied Pharmacology. 90(2): 183-189.
Heywood, R., R.J. Sortwell, P.S. Noel, A.E. Street, D.E. Prentice, F.J. Roe, P.F. Wadsworth,
A.N. Worden, and N.J. Van Abbe. 1979. Safety evaluation of toothpaste containing chloroform:
III. Long-term study in beagle dogs. Journal of Environmental Pathology Toxicology and
Oncology. 2(3): 835-851.
Hildesheim, M.E., K.P, Canbor, C.F. Lynch, M. Dosemeci, J. Lubin, M. Alavanja, and G.F.
Craun. 1998. Drinking water source and chlorination byproducts: Risk of colon and rectal
cancers. Epidemiology. 9(1): 29-35.
Hinckley, A.F., A.M. Bachand, and J.S. Reif. 2005. Late pregnancy exposures to disinfection by-
products and growth-related birth outcomes. Environmental Health Perspectives. 113(12): 1808-
1813.
Hladik, M.L., M.J. Focazio, and M. Engle. 2014. Discharges of produced waters from oil and gas
extraction via wastewater treatment plants are sources of disinfection by-products to receiving
streams. Science of the Total Environment. 466: 1085-1093.
Hoehn, R.C., C.S. Ellenberger, D.L. Gallagher, E.V. Wiseman, R.W. Benninger, and A.
Rosenblatt. 2003. C102 and by-product persistence in a drinking water system. Journal of
American Water Works Association. 95(4): 141-150.
Hoffman, C.S., P. Mendola, D.A. Savitz, A.H. Herring, D. Loomis, K.E. Hartmann, P.C. Singer,
H.S. Weinberg, and A.F. Olshan. 2008a. Drinking water disinfection by-product exposure and
fetal growth. Epidemiology. 19(5): 729-731.
Six-Year Review 3 Technical Support Document	9-18
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Hoffman, C.S., P. Mendola, D.A. Savitz, A.H. Herring, D. Loomis, K.E. Hartmann, P.C. Singer,
H.S. Weinberg, and A.F. Olshan. 2008b. Drinking Water disinfection by-product exposure and
duration of gestation. Epidemiology. 19(5).
Hooper, S. M., and S.A. Allgeier. 2002. GAC and membrane treatment studies. In Information
collection rule data analysis (M. J. McGuire, J. L. McLain, and A. Obolensky, eds). 501-538.
Denver, CO: American Water Works Association Research Foundation.
Hooth, M.J., K.S. McDorman, S.D. Hester, M.H. George, L.R. Brooks, A.E. Swank, and D.C.
Wolf. 2002. The carcinogenic response of Tsc2 mutant Long-Evans (Eker) rats to a mixture of
drinking water disinfection by-products was less than additive. Toxicologic Sciences. 69(2): 322-
331.
Horton, B.J., T.J. Luben, A.H. Herring, D.A. Savitz, P.C. Singer, H.S. Weinberg, and K.E.
Hartmann. 2011. The effect of water disinfection by-products on pregnancy outcomes in two
Southeastern US communities. Journal of Occupational and Environmental Medicine. 53(10):
1172-1178.
Hrudey, S.E., L.C. Backer, A.R. Humpage, S.W. Krasner, D.S. Michaud, L.E. Moore, P.C.
Singer, and B.D. Stanford. 2015a. Evidence for association of human bladder cancer with
chlorination disinfection by-products. Water Research Foundation and AWWA. Web Report
#4530
Hrudey, S.E., L.C. Backer, A.R. Humpage, S.W. Krasner, D.S. Michaud, L.E. Moore, P.C.
Singer, and B.D. Stanford. 2015b. Evaluating evidence for association of human bladder cancer
with drinking-water chlorination disinfection by-products. Journal of Toxicology and
Environmental Health, Part B: Critical Reviews. 18(5): 213-241.
Hsu, S and P.C. Singer. 2010. Removal of bromide and natural organic matter by anion
exchange. Water Research. 44(7): 2133-2140.
Hu, J., H. Song, and T. Karanfil. 2010. Comparative analysis of halonitromethane and
trihalomethane formation and speciation in drinking water: the effects of disinfectants, pH,
bromide, and nitrite. Environmental Science and Technology. 44(2): 794-799.
Hua, G. and D.A. Reckhow. 2007. Comparison of disinfection by-product formation from
chlorine and alternative disinfectants. Water Research. 41.8: 1667-1678.
Hua, B., K. Veum, A. Koirala, J. Jones, T. Clevenger, and B. Deng. 2007. Fluorescence
fingerprints to monitor total trihalomethanes and N-nitrosodimethylamine formation potentials in
water. Environmental Chemical Letters. 5(2): 73-77.
Hua, G. and D.A. Reckhow. 2008. DBP formation during chlorination and chloramination: effect
of reaction time, pH, dosage, and temperature. Journal of the American Water Works
Association. 100(8): 82-95.
Hua, G. and D.A. Reckhow. 2012. Effect of alkaline pH on the stability of halogenated DBPs.
Journal of the American Waterworks Association. 104(2): 107-120.
Six-Year Review 3 Technical Support Document	9-19
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Hua, G., D.A. Reckhow, and J. Kim. 2006. Effect of bromide and iodide ions on the formation
and speciation of disinfection byproducts during chlorination. Environmental Science and
Technology. 40(9): 3050-3056.
Hua, G., J. Kim, and D.A. Reckhow. 2014. Disinfection by-product formation from lignin
precursors. Water Research. 63: 285-295.
Huang, W. and Y. Cheng. 2008. Effect of characteristics of activated carbon on removal of
bromate. Water Research. Separation and Purification Technology. 59(1): 101-107.
Huang, H., Q-Y. Wu, H-Y. Hu, and W.A. Mitch. 2012. Dichloroacetonitrile and
dichloroacetamide can form independently during chlorination and chloramination of drinking
waters, model organic matters, and wastewater effluents. Environmental Science and
Technology. 46(19): 10624-10631.
Huerta, J., K. Smith, and J. Zambrano. 2015. Evaluating the Effectiveness of Controls for
Unintended Consequences Associated with a Temporary Disinfectant Switch: A Case Study.
American Water Works Association Annual Conference, ACE15.
Humpage, A.R. 2012. Mutagen X: The evolving story of an extremely potent mutagen, its
toxicology and human health risk assessment. In: Disinfection By-Products and Human Health,
eds. Hrudey, S. E. and J. Charrois. London: IWA Publishing.
Hunter, E. S., Ill and J. A. Tugman. 1995. Inhibitors of glycolytic metabolism affect neurulation-
staged mouse conceptuses in vitro. Teratology. 52: 317 (as cited in Plewa et al., 2004a).
Hunter, E. S., Ill, E.H. Rogers, J.E. Schmid, and A. Richard. 1996. Comparative effects of
haloacetic acids in whole embryo culture. Teratology. 54: 57 (as cited in Plewa et al., 2004a).
Huuskonen, H., et al. 1997. Developmental toxicity of 3-chloro-4-(dichloromethyl)-5-hydroxy-
2(5H) - furanone (MX) in rats. Human and Experimental Toxicology. 16(10): 619 (abstract
only).
Hwang, B.-F. and J.J.K. Jaakkola. 2003. Water chlorination and birth defects: A systematic
review and meta-analysis. Archives of Environmental Health. 58(2): 83-91.
Hwang, B.-F., P. Magnus, and J.J.K. Jaakkola. 2002. Risk of specific birth defects in relation to
chlorination and the amount of natural organic matter in the water supply. American Journal of
Epidemiology. 156(4): 374-382.
Hwang, B.F., J.J.K. Jaakkola, and H.R. Guo. 2008. Water disinfection by-products and the risk
of specific birth defects: a population-based cross-sectional study in Taiwan. Environmental
Health. 7(23).
Hwang, B.-F., and J.J.K. Jaakkola. 2012. Risk of stillbirth in the relation to water disinfection
by-products: A population-based case-control study in Taiwan. PLoS ONE 7(3):e33949.
Six-Year Review 3 Technical Support Document	9-20
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Infante-Rivard, C., E. Olson, L. Jacques, and P. Ayotte. 2001. Drinking Water Contaminants and
Childhood Leukemia. Epidemiology. 12(1): 3-9.
Infante-Rivard, C., D. Amre, and D. Sinnett. 2002. GSTT1 and CYP2E1 polymorphisms and
trihalomethanes in drinking water: effect on childhood leukemia. Environmental Health
Perspectives. 110(6): 591-593.
Infante-Rivard, C. 2004. Drinking water contaminants, gene polymorphisms, and fetal growth.
Environmental Health Perspectives. 112(11): 1213-1216.
Ikari, M., Y. Matsui, Y. Suzuki, T. Matsushita, andN. Shirasaki. 2015. Removal of iodide from
water by chlorination and subsequent adsorption on powdered activated carbon. Water Research
68: 227-237.
International Agency for Research on Cancer (IARC). 1991. Chlorinated Drinking-water;
Chlorination By-products; Some Other Halogenated Compounds; Cobalt and Cobalt
Compounds. In: IARC Monographs on the Carcinogenic Risk for Humans, vol. 52. Available
online at: http://monographs.iarc.fr/ENG/Monographs/vol52/mono52.pdf.
IARC. 1999a. Re-evaluation of Some Organic Chemicals, Hydrazine, and Hydrogen Peroxide.
In: IARC Monographs on the Carcinogenic Risk for Humans, vol. 71. Available online:
http://monographs.iarc.fr/ENG/Monographs/vol71/mono71.pdf.
IARC. 1999b. Acetaldehyde. In: IARC Monographs on the Carcinogenic Risk for Humans, vol.
71. Available online at: http://monographs.iarc.fr/ENG/Monographs/vol71/mono71-11 .pdf
IARC. 2004. Some Drinking Water Disinfectants and Contaminants, Including Arsenic. In:
IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, vol. 84 (as cited in
Richardson et al., 2007). Available online at
http://monographs.iarc.fr/ENG/Monographs/vol84/index.php.
IARC. 2014. Trichloroethylene, Tetrachloroethylene, and some Chlorinated Agents. In: IARC
Monographs on the Carcinogenic Risk for Humans, vol. 106. Available online at:
http://monographs.iarc,fr/ENG/Monographs/vol 106/mono 106.pdf.
International Life Sciences Institute (ILSI). 1997. An evaluation of EPA's proposed guidelines
for carcinogen risk assessment using chloroform and dichloroacetate as case studies: report of an
expert panel. Washington, DC: ILSI Health and Environmental Sciences Institute.
ILSI. 1998. Assessing the Toxicity of Exposure to Mixtures of Disinfection By-Products.
Research Recommendations. Report prepared by the ILSI Risk, Science Institute under a
cooperative agreement with U. S. EPA's Office of Water. Washington, DC.
Iszatt, N., M.J. Nieuwenhuijsen, P. Nelson, P. Elliott, and M.B. Toledano. 2011. Water
consumption and use, trihalomethane exposure, and the risk of yypospadias. Pediatrics. 127(2):
e389-e397.
Six-Year Review 3 Technical Support Document	9-21
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Iszatt, N., M.J. Nieuwenhuijsen, J. Bennett, N. Best, A.C. Povey, A.A. Pacey, H. Moore, N.
Cherry, and M.B. Toledano. 2013. Chlorination by-products in tap water and semen quality in
England and Wales. Journal of Occupational and Environmental Medicine. 70(11): 1351-0711.
Jaakkola, J., P. Magnus, A. Skrondal, B. Hwang, G. Becher, and E. Dybing. 2001. Fetal growth
and duration of gestation relative to water chlorination. Occupational and Environmental
Medicine. 58(7).
Jarvis, P., E. Sharp, M. Pidou, R. Molinder, S.A. Parsons, and B. Jefferson. 2012. Comparison of
coagulation performance and floe properties using a novel zirconium coagulant against
traditional ferric and alum coagulants. Water Research. 46(13): 4179-87.
Jensen, J., C. Siedel, S. Acquafredda, and G. Tang. 2010. Putting aeration to the test: Full-scale
demonstration results for TTHM-stripping by application of aeration strategy. WQTC 2010
Conference Proceedings.
Jeong, C.H., C. Postigo, S.D. Richardson, J.E. Simmons, S.Y. Kimura, B.J. Marinas, D. Barcelo,
P. Liang, E.D. Wagner, and M.J. Plewa. 2015. Occurrence and comparative toxicity of
haloacetaldehyde disinfection byproducts in drinking water. Environmental Science and
Technology. 49(23): 13749-13759.
Jiang, J. and S. Wang. 2004. Enhanced coagulation with potassium ferrate(VI) for removing
humic substances. Environmental Engineering Science. 20(6): 627-633.
Jo, C.H., A.M. Dietrich, and J.M. Tanko. 2011. Simultaneous degradation of disinfection
byproducts and earthy-musty odorants by the UV/H2O2 advanced oxidation process. Water
Research. 45(8): 2507-2516.
Johnson, B.A. Johnson, J.C. Lin, D. Rexing, M. Fang, J. Chan, L. Jacobsen, and P. Sampson.
2009. Localized treatment for disinfection by-products. Water Research Foundation. Project
3103.
Johnson, P.D., B.V. Dawson, and S.J. Goldberg. 1998. Cardiac teratogenicity of
trichloroethylene metabolites. Journal of American College of Cardiology. 32(2): 540-545.
Joll, C.A., M.J. Alessandrino, and A. Hetz. 2010. Disinfection by-products from halogenation of
aqueous solutions of terpenoids. Water Research. 44(1): 232-242.
Jones, D.B., A. Saglam, H. Song, and T. Karanfil. 2012. The impact of bromide/iodide
concentration and ratio on iodinated trihalomethane formation and speciation. Water Research.
46(1): 11-20.
Jorgenson, T. A., E.F. Meierhenry, C.J. Rushbrook, R.J. Bull, and M. Robinson. 1985.
Carcinogenicity of chloroform in drinking water to male Osborne-Mendel rats and female
B6C3F1 mice. Fundamental and Applied Toxicology. 5: 760-769.
Kallen B.A.J, and E. Robert. 2000. Drinking water chlorination and delivery outcome—A
registry-based study in Sweden. Reproductive Toxicology. 14(4).
Six-Year Review 3 Technical Support Document	9-22
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Kalscheur, K.N., C.E. Gerwe, J. Kweon, G.E. Speitel Jr., and D.F. Lawler. 2006. Enhanced
softening: Effects of source water quality on NOM removal and DBP formation. Journal of the
American Water Works Association. 98(11): 93-106.
Kanitz S., Y. Franco, V. Patrone, M. Caltabellotta, E. Raffo, C. Riggi, D. Timitilli, and G.
Ravera. 1996. Association between drinking water disinfection and somatic parameters at birth.
Environmental Health Perspective. 104(5).
Kanokkantapong, V., T. Marhaba, P. Pavasant, andB. Panyapinyophol. 2006. Characterization
of haloacetic acid precursors in source water. Journal of Environmental Management. 80(3):
214-221.
Karagas, M.R., C.M. Villanueva, M. Nieuwenhuijsen, C.P. Weisel, K.P. Cantor, and M.
Kogevinas. 2008. Disinfection byproducts in drinking water and skin cancer? A Hypothesis.
Cancer Causes & Control. 19(5): 547-548.
Karanfil, T., J. Hu, D.B. Jones, J.W. Addison, and H. Song. 2011. Formation of
halonitromethanes and iodo-trihalomethanes in drinking water. Water Research Foundation.
Project #4063.
Kargalioglu, Y., B.J. McMillan, R.A. Minear, and M.J. Plewa. 2002. Analysis of the cytotoxicity
and mutagenicity of drinking water disinfection by-products in Salmonella typhimurium.
Teratogen. Carcinogen. Mutagen, 22(2): 113-128.
Kasim, K., P. Levallois, K. Johnson, B. Abdous, P. Auger, and the Canadian Cancer Registries
Epidemiology Research Group. 2006. Chlorination disinfection by-products in drinking water
and the risk of adult leukemia in Canada. American Journal of Epidemiology. 163(2): 116-126.
Katz, R., C.N. Tai, R.M. Diener, R.F. McConnell, andD.E. Semonick. 1981. Dichloroacetate,
sodium: 3-month oral toxicity studies in rats and dogs. Toxicology and Applied Pharmacology.
57(2): 273-287.
Kent, F.C., K.R. Montreuil, R.M. Brookman, R. Sanderson, J.R. Dahn, and G.A. Gagnon. 2011.
Photocatalytic oxidation of DBP precursors using UV with suspended and fixed TiC>2. Water
Research. 45(18): 6173-6180.
Kenyon, E.M., C. Eklund, T. Leavens, and R.A. Pegram. 2015. Development and application of
a human PBPK model for bromodichloromethane to investigate the impacts of multi-route
exposure. Journal of Applied Toxicology. [Epub ahead of print].
Kerns, W.D., K.L. Pavkov, D.J. Donofrio, E.J. Gralla, and J.A. Swenberg. 1983. Carcinogenicity
of formaldehyde in rats and mice after long-term inhalation exposure. Cancer Res. 43: 4382-
4392.
Kim, H-C. and M.J. Yu. 2005. Characterization of natural organic matter in conventional water
treatment processes for selection of treatment processes focused on DBPs control. Water
Research. 39(19): 4779-4789.
Six-Year Review 3 Technical Support Document	9-23
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Kim, J., and B. Kang. 2008. DBPs removal in GAC filter-adsorber. Water Research. 42(1-2):
145-152.
Kimbrough, D.E., and I. H. Suffet. 2006. Electrochemical process for the removal of bromide
from California state project water. Journal of Water Supply: Research & Technology-AQUA.
55(3): 161-167.
Kimbrough, D.E., L. Boulos, S. Surawanvijit, and A. Zacheis. 2010. The reaction rates and
removal efficiency of bromide by electrolytic oxidation and volatilization. WQTC 2010
Conference Proceedings.
Kimbrough, D.E., L. Boulos, and S. Surawanvijit. 2011. Oxidation and volatilization of bromide
by electrolysis in drinking water. Journal of Water Supply: Research & Technology-AQUA.
60(3): 127-136.
Kimbrough, D.E., L. Boulos, S. Surawanvijit, P. Westerhoff, H. An, I.H. Suffet, andN.
Dunahee. 2012. Practical studies of the electrolysis and volatilization of the bromide from
drinking water to minimize bromate production by ozonation. Ozone: Science & Engineering.
34: 269-279.
Kimura, S., Y. Komaki, M.J. Plewa, and B.J. Marinas. 2013. Chloroacetonitrile andN,2-
dichloroacetamide formation from the reaction of chloroacetaldehyde and monochloramine in
water. Environmental Science and Technology. 47(21): 12382-12390.
King, W.D. and L.D. Marrett. 1996. Case-control study of bladder cancer and chlorination by-
products in treated water (Ontario, Canada). Cancer Causes Control. 7: 596-604.
King, W.D., L.D. Marrett, and C.G. Woolcott. 2000. Case-control study of colon and rectal
cancers and chlorination byproducts in treated water. Cancer Epidemiology. Biomarkers and
Prevention, 9(8): 813-818.
Klinefelter, G.R., J.D. Suarez, N.L. Roberts, and A.B. DeAngelo. 1995. Preliminary screening
for the potential of drinking water disinfection byproducts to alter male reproduction.
Reproductive Toxicology. 9(6): 571-578.
Klotz, J.B. and L.A. Pyrch. 1999. Neural tube defects and drinking water disinfection
byproducts. Epidemiology. 10(4): 383-90.
Kogevinas, M., C.M. Villanueva, L. Font-Ribera, D. Liviac, M. Bustamante, F. Espinoza, M.J.
Nieuwenhuijsen, A. Espinosa, P. Fernandez, D.M. DeMarini, J.O. Grimalt, T. Grummt, and R.
Marcos. 2010. Genotoxic effects in swimmers exposed to disinfection by-products in indoor
swimming pools. Environmental Health Perspectives. 118(11): 1531-1537.
Kohan, M.J., G. Huggins-Clark, and S.E. George. 1998. Mutagenicity of chlorinated and
brominated acetic acids. 29th Annual Meeting of the Environmental Mutagen Society, Anaheim,
CA, March 21-26. Environmental and Molecular Mutagenesis. 31(Suppl. 29): 36(abstract).
Six-Year Review 3 Technical Support Document	9-24
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Kohl, P.M., and D. Dixon. 2012. Occurrence, impacts, and removal of manganese in biofiltration
processes. Water Research Foundation. Web Report #4021.
Koivusalo, M., T. Hakulinen, T. Vartiainen, E. Pukkala, J.J. Jaakkola, and J. Tuomisto. 1998.
Drinking water mutagenicity and urinary tract cancers: a population-based case-control study in
Finland. American Journal of Epidemiology. 148(7): 704-12.
Kolisetty, N., R.J. Bull, S. Muralidhara, L. Costyn, D. Delker, Z. Guo, J. Cotruvo, J. Fisher, and
B. Cummings. 2013. Association of brominated proteins and changes in protein expression in the
rat kidney with subcarcinogenic to carcinogenic doses of bromate. Toxicology and Applied
Pharmacology. 272, 391-398.
Komulainen, H., Kosma, V. M., Vaittinen, S. L., Vartiainen, T., KalisteKorhonen, E., Lotjonen,
S., Tuominen, R. K. and Tuomisto, J. 1997. Carcinogenicity of the drinking water mutagen 3-
chloro-4-(dichloromethyl)-5-hydroxy-2(5H)-furanone in the rat. Journal of the National Cancer
Institute. 89(12): 848-856.
Kramer M.D., C.F. Lynch, P. Isacson, and J.W. Hanson. 1992. The association of waterborne
chloroform with intrauterine growth retardation. Epidemiology.
Krasner, S.W. 2009. The formation and control of emerging disinfection by-products of health
concern. Philosophical Transactions of the Royal Society A. Available online at:
http://rsta.rovalsocietypublishing.org/content/367/1904/4077.
Krasner, S.W. 2014. Balancing the control of regulated and emerging DBPs while meeting other
treatment objectives. ACS National Meeting 2014 Conference Proceedings.
Krasner, S.W., H.S. Weinberg, S.D. Richardson, S.J. Pastor, R. Chinn, M.J. Sclimenti, G.D.
Onstad, and A.D. Thruston. 2006. Occurrence of a new generation of disinfection byproducts.
Environmental Science Technology. 40(23): 7175-7185.
Krasner, S.W., P. Westerhoff, B. Chen, G. Amy, S. Nam, Z.K. Chowdhury, S. Sinha, and B.E.
Rittman. 2008. Contribution of wastewater to DBP formation. Water Research Foundation.
Krasner, S.W., W.A. Mitch, P. Westerhoff, and A. Botson. 2012. Formation and control C- and
N-DBPs in drinking water. Journal of American Water Work Association. 104. E582-E594.
Krasner, S.W., W.A. Mitch, D.L. McCurry, D. Hannigan, and P. Westerhoff. 2013. Formation,
precursors, control, and occurrence of nitrosamines in drinking water: a review. Water Research.
47(13): 4433-4450.
Krasner, S.W., R. Shirkani, P.Westerhoff, D. Hanigan, W.A. Mitch, D.L. McCurry, C. Chen, J.
Skadsen, and U. von Gunten. 2015. Controlling the Formation of Nitrosamines during Water
Treatment. Water Research Foundation Report #4730. Denver, CO.
Kristiana, I., H. Gallard, C. Joll, and J-P. Croue. 2009. The formation of halogen-specific TOX
from chlorination and chloramination of natural organic matter isolates. Water Research. 43(17):
4177-4186.
Six-Year Review 3 Technical Support Document	9-25
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Kristiana, I., C. Joll, and A. Heitz. 2011. Powdered activated carbon coupled with enhanced
coagulation for natural organic matter removal and disinfection by-product control: Application
in a Western Australian water treatment plant. Chemosphere. 83(5): 661-667.
Kritsch, K. and H. Weinberg. 2010. Iodoacetic acids in US drinking waters: occurrence and fate.
WQTC 2010 Conference Proceedings.
Kumar, S., S. Forand, G. Babcock, and S. Hwang. 2013. Total Trihalomethanes in Public
Drinking Water Supply and Birth Outcomes: A Cross-Sectional Study. Maternal and
Child Health Journal. 18(4): 996-1006.
Kundu, B., S.D. Richardson, P.D. Swartz, P.P. Matthews, A.M. Richard, and D.M. DeMarini.
2004. Mutagenicity in Salmonella of halonitromethanes: a recently recognized class of
disinfection byproducts in drinking water. Mutation Research. 562: 39-65.
Kuo, H.W., M.M. Tiao, T.N. Wu, and C.Y. Yang. 2009. Trihalomethanes in drinking water and
the risk of death from colon cancer in Taiwan. Journal of Toxicology and Environmental Health.
72: 1217-1222.
Kurokawa, Y., S. Aoki, Y. Matsushima, N. Takamura, T. Imazawa, and Y. Hayashi. 1986a.
Dose-response studies on the carcinogenicity of potassium bromate in F344 rats after long-term
oral administration. Journal of the National Cancer Institute. 77(4): 977-982.
Kurokawa, Y., S. Takayama, Y. Konishi, Y. Hiasa, S. Asahina, M. Takahashi, A. Maekawa, and
Y. Hayashi. 1986b. Long-term in vivo carcinogenicity tests of potassium bromate, sodium
hypochlorite and sodium chlorite conducted in Japan. Environmental Health Perspectives. 69:
221-236.
Krajka-Kuzniak, V., H. Szaefer, and W. Baer-Dubowska. 2005. Modulation of cytochrome P450
and phase II enzymes by protocatechuic acid in mouse liver and kidney. Toxicology. 216(1): 24-
31.
Lamsal, R., M.E. Walsh, and G.A. Gagnon. 2011. Comparison of advanced oxidation processes
for the removal of natural organic matter. Water Research. 45(10): 3263-3269.
Lantum, H.B.M., R.B. Baggs, D.M. Krenitsky, and M.W. Anders. 2002. Nephrotoxicity of
Chlorofluoroacetic Acid in Rats. Toxicological Sciences. 70(2): 261-268.
Larson, J.L., M.V. Templin, D.C. Wolf, K.C. Jamison, J.R. Leininger, S. Mery, K.T. Morgan,
B.A. Wong, R.B. Conolly, and B.E. Butterworth. 1996. A 90-day chloroform inhalation study in
female and male B6C3F1 mice: implications for risk assessment. Fundamental and Applied
Toxicology. 30, 118-137.
Lauderdale, C., P. Chadik, M.J. Kirisits, and J. Brown. 2012. Engineered biofiltration: enhanced
biofilter performance through nutrient and peroxide addition. Journal of the American Water
Works Association. 104(3): 298-310.
Six-Year Review 3 Technical Support Document	9-26
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Lauderdale, C.V., K. Scheitlin, J. Nyfennegger, G. Upadhyaya, J.C. Brown, L. Raskin, T-Z.
Chiao, and A. Pinto. 2014. Optimizing Engineered Biofiltration. Water Research Foundation.
Laurie, R.D., J.P. Bercz, T.K. Wessendarp, and L.W. Condie. 1986. Studies of the toxic
interactions of disinfection by-products. Environmental Health Perspectives. 69: 203.
Leavens, T.L., B.C. Blount, D.M. DeMarini, M.C. Madden, J.L. Valentine, M.W. Case, L.K.
Silva, S.H. Warren, N.M. Hanley, and R.A. Pegram. 2007. Disposition of bromodichloromethane
in humans following oral and dermal exposure. Toxicological Sciences. 99: 432-445.
Lee, W., P. Westerhoff, and J-P. Croue. 2007. Dissolved organic nitrogen as a precursor for
chloroform, dichloroacetonitrile, n-nitrosodimethylamine, and trichloronitromethane.
Environmental Science and Technology. 41(15): 5485-5490.
Lee, W., and P. Westerhoff. 2009. Formation of organic chloramines during water disinfection -
chlorination versus chloramination. Water Research. 43(8): 2233-2239.
Levallois, P., S. Gingras, S. Marcoux, C. Legay, C. Catto, M. Rodriguez, andR. Tardif. 2012.
Maternal exposure to drinking-water chlorination by-products and small-for-gestational-age
neonates. Epidemiology. 23(2): 267-276.
Lewis, C., I.H. Suffet, and B. Ritz. 2006. Estimated effects of disinfection by-products on birth
weight in a population served by a single water utility. American Journal of Epidemiology.
163(1): 38-47.
Lewis, C., I.H. Suffet, K. Hoggatt, and B. Ritz. 2007. Estimated effects of disinfection by-
products on preterm birth in a population served by a single water utility. Environmental Health
Perspectives. 115(2).
Li, X., J. Ma, G. Liu, J. Fang, S. Yue, Y. Guan, L. Chen, X. Liu. 2012. Efficient reductive
dechlorination of monochloroacetic acid by sulfite/UV process. Environmental Science and
Technology. 46(13): 7342-7349.
Li, W., Y. Liu, J. Duan, J. van Leeuwen, and C.P. Saint. 2015. UV and UV/H2O2 treatment of
diazinon and its influence on disinfection byproduct formation following chlorination. Chemical
Engineering Journal. 274: 39-49.
Liao, X., C. Wang, J. Wang, X. Zhang, C. Chen, S.W. Krasner, and I.H. Suffet. 2014.
Nitrosamine precursor and DOM control in an effluent-affected drinking water. Journal of the
American Water Works Association. 106(7): E307-E318.
Liao, X., C. Chen, S. Xie, D. Hanigan, J. Wang, X. Zhang, P. Westerhoff, and S.W. Krasner.
2015. Nitrosamine precursor removal by BAC: A case study of adsorption versus biotreatment.
Journal of the American Water Works Association. 107(9): E454-E463.
Lim, M. and M. Kim. 2009. Removal of natural organic matter from river water using potassium
ferrate(VI). Water, Air, and Soil Pollution. 200(1): 181-189.
Six-Year Review 3 Technical Support Document	9-27
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Lin, P., X. Zhang, J. Wang, Y. Zeng, S. Liu, and C. Chen. 2015. Comparison of different
combined treatment processes to address the source water with high concentration of natural
organic matter during snowmelt period. Journal of Environmental Sciences. 27: 51-58.
Linden, K.G., A. Dotson, H. Weinberg, B. Lyon, W.A. Mitch, and A. Shah. 2012. Impact of UV
location and sequence on by-product formation. Water Research Foundation. Available online at:
http://www.waterrf.ore/publicreportlibrarv/4019.pdf.
Linder, R.E., G.R. Klinefelter, L.F. Strader, J.D. Suarez, and C.J. Dyer. 1994a. Acute
spermatogenic effects of bromoacetic acids. Fundamental and Applied Toxicology. 22(3): 422-
430.
Linder, R.E., G.R. Klinefelter, L.F. Strader, J.D. Suarez, N.L. Roberts, and C.J. Dyer. 1994b.
Spermatotoxicity of dibromoacetic acid in rats after 14 daily exposures. Reproductive
Toxicology. 8(3): 251-259.
Linder, R.E., G.R. Klinefelter, L.F. Strader, M.G. Narotsky, J.D. Suarez, N.L Roberts, and S.D.
Perreault. 1995. Dibromoacetic acid affects reproductive competence and sperm quality in the
male rat. Fundamental and Applied Toxicology. 28(1): 9-17.
Linder, R.E., G.R. Klinefelter, L.F. Strader, J.D. Suarez, andN. Roberts. 1997. Spermatotoxicity
of dichloroacetic acid. Reproductive Toxicology. 11(5): 681-688.
Linder, K., J. Lew, B. Carter, and R. Brauer. 2006. Avoiding chlorite: chlorine and C102
together form fewer DBPs. Opflow. 32(8): 24-26.
Liu, J-L. and X-Y. Li. 2010. Biodegradation and biotransformation of wastewater organics as
precursors of disinfection by-products in water. Chemosphere. 81(9): 1075-1083.
Liu, B. and D.A. Reckhow. 2015. Disparity in disinfection by-products concentration between
hot and cold tap water. Water Research. 70: 196-204.
Liu, W., Z. Zhang, X. Yang, Y. Xu, and Y. Liang. 2012. Effects of UV irradiation and
UV/chlorine co-exposure on natural organic matter in water. Science of the Total Environment.
414: 576-584.
Liu, B. and D.A. Reckhow. 2013. DBP formation in hot and cold water across a simulated
distribution system: effect of incubation time, heating time, pH, chlorine dose, and incubation
temperature. Environmental Science and Technology. 47(20): 11584-11591.
Llopis- Gonzalez, A., S. Sagrado-Vives, N. Gimeno-Clemente, V. Yusa-Pelecha, P. Marti-
Requena, L. Monforte-Monleon, andM. Morales-Suarez-Varela. 2011. Ecological study on
digestive and bladder cancer in relation to the level of trihalomethanes in drinking water.
International Journal of Environmental Research. 5(3): 613-620.
Lou, J., C. Chang, W. Chen, W. Tseng, and H. Han 2014. Removal of trihalomethanes and
haloacetic acids from treated drinking water by biological activated carbon filter. Water, Air, &
Soil Pollution. 225:1851.
Six-Year Review 3 Technical Support Document	9-28
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Lubbers, J.R., S. Chauhan, J.K. Miller. 1981. Controlled clinical evaluations of chlorine dioxide,
chlorite, and chlorate in man. Fundamental and Applied Toxicology. 1: 334-338.
Lubbers, J.R., S. Chauhan, and J.R. Bianchine. 1982. Controlled clinical evaluations of chlorine
dioxide, chlorite and chlorate in Man. Environmental Health Perspectives. 46: 57-62.
Lubbers, J.R., S. Chauhan, J.K. Miller, and J.R. Bianchine. 1984. The effects of chronic
administration of chlorine dioxide, chlorite and chlorate to normal healthy adult male volunteers.
Journal of Environmental Patholology, Toxicology, and Oncology. 5(4-5): 229-238.
Luben, T.J., A.F. Olshan, A.H. Herring, S. Jeffay, L. Strader, R.M. Buus, R.L. Chan, D.A.
Savitz, P.C. Singer, H.S. Weinberg, and S.D. Perreault. 2007. The Healthy Men Study: An
Evaluation of Exposure to Disinfection By-Products in Tap Water and Sperm Quality.
Environmental Health Perspectives. 115(8): 1169-1176.
Luben, T.J., J.R. Nuckols, B.S. Mosley, C. Hobbs, and J.S. Reif. 2008. Maternal exposure to
water disinfection by-products during gestation and risk of hypospadias. Occupational and
Environmental Medicine. 65(6): 420-427.
Luh, J., and B.J. Marinas. 2014. Kinetics of bromochloramine formation and decomposition.
Environmental Science and Technology. 48(5): 2843-2852.
Lui, Y.S., H.C. Hong, G.J.S. Zheng, and Y. Liang. 2012. Fractionated algal organic materials as
precursors of disinfection by-products and mutagens upon chlorination. Journal of Hazardous
Materials. 209: 278-284.
Lynch, C.F., R.F. Woolson, T. O'Gorman, and K.P. Cantor. 1989. Chlorinated drinking water
and bladder cancer: effect of misclassification on risk estimates. Archives of Environmental
Health. 44(4):252-9.
Lyon, B.A., A.D. Dotson, K.G. Linden, and H.S. Weinberg. 2012. The effect of inorganic
precursors on disinfection by-product formation during UV-chlorine/chloramine drinking water
treatment. Water Research. 46(15): 4653-4664.
Lyon, B.A., R.M. Cory, and H.S. Weinberg. 2014. Changes in dissolved organic matter
fluorescence and disinfection by-product formation from UV and subsequent
chlorination/chloramination. Journal of Hazardous Materials. 264(15): 411-419.
MacLehose, R.F., D.A. Savitz, A.H. Herring, K.E. Hartmann, P.C. Singer, and H.S. Weinberg.
2008. Drinking water disinfection by-products and time to pregnancy. Epidemiology. 19(3): 451-
458.
Magnus P., J.K. Jaakkola, A. Skrondal, J. Alexander, G. Becher, T. Krogh, and E. Dybing. 1999.
Water chlorination and birth defects. Epidemiology. 10: 513-517.
Majidzadeh, H., J.J. Wang, and A.T. Chow. 2015. Prescribed fire alters dissolved organic matter
and disinfection by-product precursors in forested watersheds - Part I. A controlled laboratory
study. In Recent Advances in Disinfection By-Products. ACS Symposium Series 1190.
Six-Year Review 3 Technical Support Document	9-29
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Mao, Y., X. Wang, H. Yang, H. Wang, and Y.F. Xie. 2014. Effects of ozonation on disinfection
byproduct formation and speciation during subsequent chlorination. Chemosphere. 117: 515-520.
Matamoros, V., R. Mujeriego, and J.M. Bayona. 2007. Trihalomethane occurrence in chlorinated
reclaimed water at full-scale wastewater treatment plants in NE Spain. Water Research. 41(15):
3337-3344.
Matilainen, A., and M. Sillanpaa. 2010. Removal of natural organic matter from drinking water
by advanced oxidation processes. Chemosphere. 80(4): 351-365.
McClain, J.L, A. Obolensky, and H.M. Shukairy. 2002. Disinfection by-product speciation. In
Information Collection Rule Data Analysis, Chapter 6. Denver, CO: American Water Works
Association Research Foundation.
McCurry, D.L., S.W. Krasner, U. von Gunten, and W.A. Mitch. 2015. Determinants of
disinfectant pretreatment efficacy for nitrosamine control in chloraminated drinking water. Water
Research. 84: 161-170.
McDonald, T. A. and H. Komulainen. 2005. Carcinogenicity of the chlorination disinfection by-
product MX. Journal of Environmental Science and Health Part C - Environmental
Carcinogenesis & Ecotoxicology Reviews. 23(2): 163-214.
McDorman, K.S., M.J. Hooth, T.B. Starr, and D.C. Wolf. 2003a. Analysis of preneoplastic and
neoplastic renal lesions in Tsc2 mutant Long-Evans (Eker) rats following exposure to a mixture
of drinking water disinfection by-products. Toxicology. 187: 1-12.
McDorman, K.S., S. Chandra, M.J. Hooth., S.D. Hester, R. Schoonhoven, and D.C. Wolf. 2003b.
Induction of transitional cell hyperplasia in the urinary bladder and aberrant crypt foci in the
colon of rats treated with individual and a mixture of drinking water disinfection by-products.
Toxicologic Pathology. 31: 235-242.
McGeehin, M.A., J.S. Reif, J.C. Becher., and E.J. Mangione. 1993. Case control study of bladder
cancer and water disinfection methods in Colorado. American Journal of Epidemiology. 138.
McGuire, M.J. and N. Graziano. 2002. Trihalomethanes in U.S. drinking water NORS to ICR. In
Information Collection Rule Data Analysis, Chapter 4. Denver, CO: American Water Works
Association Research Foundation.
McGuire, M.J., J.L. McLain, and A. Obolensky (eds.). 2002. Information Collection Rule Data
Analysis. Denver, CO: American Water Works Research Foundation Research Foundation and
American Water Works Association, 600 p. Available online at
http://www.waterrf.org/PublicReportLibrarv/90947.pdf.
McGuire, M.J., X. Wu, N.K. Blute, D. Askenaizer, and G. Qin. 2009. Prevention of nitrification
using chlorite ion: Results of a demonstration project in Glendale, Calif. Journal of the American
Waterworks Association. 101(10): 47-59.
Six-Year Review 3 Technical Support Document	9-30
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
McGuire, M.J., T. Karanfil, S.W. Krasner, D.A. Reckhow, J.A. Roberson, R.S. Summers, P.
Weseteroff, and Y. Xie. 2014. Not your granddad's disinfection by-product problems and
solutions. Journal of the American Water Works Association. 106(8): 54-73.
McKie, M.J., L. Taylor-Edmonds, S.A. Andrews, andR.C. Andrews. 2015. Engineered
biofiltration for the removal of disinfection by-product precursors and genotoxicity. Water
Research. 81: 196-207.
McTigue, N.E., D.A. Cornwell, K. Graf, and R. Brown. 2014. Occurrence and consequences of
increased bromide in drinking water sources. Journal of the American Water Works Association.
106(11): 492-508.
Meier, J.R., R.J. Bull, J.A. Stober, M.C. Cimino. 1985. Evaluation of chemicals used for
drinking water disinfection for production of chromosomal damage and sperm-head
abnormalities in mice. Environmental Mutagenesis. 7: 201-211.
Melnick, R.L., A. Nyska, P.M Foster, J.H. Roycroft, and G.E. Kissling. 2007. Toxicity and
carcinogenicity of the water disinfection byproduct, dibromoacetic acid, in rats and mice.
Toxicology. 230: 126-136.
Metcalfe, D., C. Rockey, B. Jefferson, S. Judd, and P. Jarvis. 2015. Removal of disinfection by-
product precursors by coagulation and an innovative suspended ion exchange process. Water
Research. 87(15): 20-28.
Michaud, D. S., D. Spiegelman, S. Clinton, E.B. Rimm, G.C. Curhan, W.C. Willett, and E.L.
Giovannucci. 1999. Fluid intake and the risk of bladder cancer in men. The New England Journal
of Medicine. 340: 1390-1397.
Michaud, D.S., M. Kogevinas , K.P. Cantor., C.M. Villanueva, M. Garcia-Closas, N. Rothman,
N. Malats, F.X. Real, C. Serra, R. Garcia-Closas, A. Tardon, A. Carrato, M. Dosemeci, and D.T.
Silverman. 2007. Total fluid and water consumption and the joint effect of exposure to
disinfection by-products on risk of bladder cancer. Environmental Health Perspectives. 115(11):
1569-1572.
Mikkelson, K.M., E.R.V. Dickenson, R.M. Maxwell, J.E. McCray, and J.O. Sharp. 2013. Water-
quality impacts from climate-induced forest die-off Nature Climate Change. 3(3): 218-222.
Miller, J.H., K. Minard., R.A. Wind, G.A. Orner, L.B. Sasser and R.J. Bull. 2000. In vivo MRI
measurements of tumor growth induced by dichloroacetate; implications for mode of action.
Toxicology. 145, 115-125.
Mills, C.J., R.J. Bull, K.P. Cantor, J. Reif, S.E. Hrudey, and P. Huston. 1998. Workshop report.
Health risks of drinking water chlorination by-products: report of an expert working group.
Chronic Disease in Canada. 19(3): 91-102.
Mitch, W.A., S.W. Krasner, P. Westerhoff, and A. Dotson. 2009. Occurrence and formation of
nitrogenous disinfection by-products. Water Research Foundation Project #3014. Denver, CO.
Six-Year Review 3 Technical Support Document	9-31
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Mobley, S.A., D.H. Taylor, R.D. Laurie, and R.J. Pfohl. 1990. Chlorine dioxide depresses T3
uptake and delays development of locomotor activity in young rats. In: Water Chlorination:
Chemistry, Environmental Impact and Health Effects, vol. 6, pp. 347-358, eds Jolley, R.L., et al.
Chelsea, MI: Lewis Publications.
Moll, D.M. and S.W. Krasner. 2002. Bromate occurrence in disinfected water. In Information
Collection Rule Data Analysis, Chapter 9. Denver, CO: American Water Works Association
Research Foundation.
Moore, G.S., E.J. Calabrese, and D.A. Leonard. 1980. Effects of chlorite exposure on conception
rate and litters of A/J strain mice. Bulletin of Environ Contamination Toxicology. 25: 689-696.
Moore, G.S. and E.J. Calabrese. 1982. Toxicological effects of chlorite in the mouse.
Environmental Health Perspectives. 46: 31-37.
Moore, M.M. and T. Chen. 2006. Mutagenicity of bromate: implications for cancer risk
assessment. Toxicology. 221(2-3): 190-196.
Morita, H., S. Tadaki, T. Nozaka,, T. Omura., M. Haga, and A. Tanaka. 1997. Mutagenicity and
concentration of chlorination by-products in tap water in Saitama. Environmental Mutagenesis
Research. 19: 127-134.
Murr, A.S. and J.M. Goldman. 2005. Twenty-week exposures to the drinking water disinfection
by-product dibromoacetic acid: reproductive cyclicity and steroid concentrations in the female
Sprague-Dawley rat. Reproductive Toxicology. 20: 73-80.
Nagisetty, R.M., T.D. Rockaway, and G.A. Willing. 2014. Drinking water quality concerns from
chloramine induced degradation of elastomeric compounds. Journal of the American Water
Works Association. 106(9): 402-407.
Nakano, K., S. Okada, S. Toyokuni, and O. Midorikawa. 1989. Renal changes induced by
chronic oral administration of potassium bromate or ferric nitrilotriacetate in Wistar rats.
Japanese Archives of Internal Medicine. 36: 41-47.
Narotsky, M.G., B.T. Hamby, D.S. Best, and E.S. Hunter III. 1996. In vivo developmental
effects of dibromoacetic acid (DBA) and dichloroacetic acid (DCA) in mice. Teratology. 53(2).
Narotsky, M.G., R.A. Pegram, and R.J. Kavlock. 1997. Effect of dosing vehicle on the
developmental toxicity of bromodichloromethane and carbon tetrachloride in rats. Fundamental
and Applied Toxicology. 40: 30-36.
Narotsky, M.G., D.S. Best, E.H. Rogers, A. McDonald, Y.M. Sey, and J.E. Simmons. 2008.
Integrated disinfection by-products mixtures research: assessment of developmental toxicity in
Sprague-Dawley rats exposed to concentrates of water disinfected by chlorination and
ozonation/postchlorination. Journal of Toxicology and Environmental Health, Part A. 71: 1216-
1221.
Six-Year Review 3 Technical Support Document	9-32
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Narotsky, M.G., D.S. Best, A. McDonald, E.A. Godin, E.S. Hunter III, and J.E. Simmons. 2011.
Pregnancy loss and eye malformations in offspring of F344 rats following gestational exposure
to mixtures of regulated trihalomethanes and haloacetic acids. Reproductive Toxicology. 31, 59-
65.
Narotsky, M.G., G.R. Klinefelter, J.M. Goldman, D.S. Best, A. McDonald, L.F. Strader, J.D.
Suarez, A.S. Murr, I. Thillainadarajah, E.S. Hunter III, S.D. Richardson, T.F. Speth, R.J. Miltner,
J.G. Pressman, L.K. Teuschler, G.E. Rice, V.C. Moser, R.W. Luebke, and J.E. Simmons. 2013.
Comprehensive assessment of a chlorinated drinking water concentrate in a rat multigenerational
reproductive toxicity study. Environmental Science and Technology. 47(18): 10653-10659.
Narotsky, M.G., G.R. Klinefelter, J.M. Goldman, A.B. DeAngelo, D.S. Best, A. McDonald, L.F.
Strader, A.S. Murr, J.D. Suarez, M.H. George, E.S. Hunter III, and J.E. Simmons. 2015.
Reproductive toxicity of a mixture of regulated drinking-water disinfection by-products in a
multigenerational rat bioassay. Environmental Health Perspective. 123(6): 564-570.
National Cancer Institute (NCI). 1976. Report on the carcinogenesis bioassay of chloroform.
CAS No. 67-66-3. U.S. Department of Health, Education, and Welfare.
NCI. 1978. Carcinogenesis bioassay of chloropicrin. Technical Report Series No. 65.
Washington, DC: US Department of Health, Education and Welfare.
NCS Engineers. 2014. Support for Six-Year Review of Microbial/Disinfection Byproduct (DBP)
Rules; Task 5 Literature Review for Implementation of Free Chlorine Burns at Public Water
Systems and the Potential Impact on DBP Formation; Final Report. July 2014.
National Drinking Water Advisory Council (NDWAC). 2000. Working Group Meeting on
Contaminant Candidate List Regulatory Determinations and the 6-Year Review of Existing
Regulations. Available online at: https://www.epa.gov/sites/production/files/2015-
1 1/documents/march 1 2 2000 meeting on ccl and 6 year review.pdf.
Naval on, S., M. Alvaro, and H. de Garcia. 2008. Carbohydrates as trihalomethanes precursors
influence of pH and the presence of CI- and Br- on trihalomethane formation potential. Water
Research. 42(14): 3990-4000.
National Toxicology Program (NTP). 1985. Toxicology and carcinogenesis studies of
dibromochloromethane in F344/N rats and B6C3F1 mice (gavage studies). Technical Report
Series No. 282. Research Triangle Park, NC: U.S. Department of Health and Human Services.
NTP. 1987. Toxicology and carcinogenesis studies of bromodichloromethane in F344/N rats and
B6C3F1 mice (gavage studies). Technical Report Series No. 321. Research Triangle Park, NC:
U.S. Department of Health and Human Services.
NTP. 1988. Chloroform reproduction and fertility assessment in CD-I mice when administered
by gavage. NTP-89-018. NTIS PB89- 148639.Lexington, LY: Environmental Health Research
and Testing, Inc.
Six-Year Review 3 Technical Support Document	9-33
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
NTP. 1989a. National Toxicology Program. Bromoform: Reproduction and fertility assessment
in Swiss CD-I mice when administered by gavage. NTP-89-068. Research Triangle Park, NC:
National Institute of Environmental Health Sciences.
NTP. 1989b. Toxicology and carcinogenesis studies of tribromomethane (bromoform) in F344/N
rats and B6C3F1 mice (gavage studies). Technical Report Series No. 350. Research Triangle
Park, NC: U.S. Department of Health and Human Services.
NTP. 1990. NTP Technical Report on the Toxicology and Carcinogenesis Studies of Chlorinated
and Chloroaminated Water in F344/N Rats andB6C3Fl Mice (Drinking Water Studies). NTP
TR-392, National Institutes of Health.
NTP. 1992a. Toxicology and carcinogenesis studies of chlorinated water (CAS Nos. 7782-50-5
and 7681-52-9) and chloraminated water (CAS No. 10599-90-3) (deionized and charcoal-
filtered) in F344/N rats and B6C3F1 mice (drinking water studies). Technical Report Series No.
392. Research Triangle Park, NC: U.S. Department of Health and Human Services.
NTP. 1992b. Toxicology and carcinogenesis studies of monochloroacetic acid (CAS No. 79-11-
8) in F344/N rats and B6C3F1 mice (gavage studies). Technical Report Series No. 396. Research
Triangle Park, NC: U.S. Department of Health and Human Services.
NTP. 1993. Technical report on toxicity studies of sodium cyanide (CAS No. 143-33-9)
administered in drinking water to F344/N rats and B6C3FI mice. NIH Publication 94-3386.
Public Health Service, National Institutes of Health, U.S. Department of Health and Human
Services, National Toxicology Program.
NTP. 1996. Final report on the short term reproductive and developmental toxicity of
chlorodibromomethane (CAS No. 124-48-1) administered in drinking water to Sprague-Dawley
rats. NTIS/PB97-111728. Gaithersburg, MD: R.O.W. Sciences, Inc.
NTP. 1998a. Final Report on the short reproductive and developmental toxicity of
bromodichloromethane (CAS No. 75-27-4) administered in drinking water to Sprague-Dawley
rats. NTIS/PB99-111262. Research Triangle Park, NC: National Institute of Environmental.
NTP. 1998b. Short-term reproductive and developmental toxicity of tribromoacetic acid (CAS
No. 75-96-7) administered in drinking water to Sprague-Dawley rats. NTP Study No.
RDGT94009.
NTP. 2000. Short term reproductive and developmental toxicity of dibromochloro-acetic acid
(CAS No. 5278-95-5) administered in drinking water to Sprague-Dawley rats. NTIS/PB2000-
103420. Research Triangle Park, NC: National Institutes of Health, Public Health Service, U.S.
Department of Health and Human Services.
NTP. 2001a. Subchronic toxicity study of dibromoacetonitrile administered in dosed water to
B6C3F1 mice. Research Triangle Park, NC: National Toxicology Program, National Institute of
Environmental Health Sciences, National Institutes of Health.
Six-Year Review 3 Technical Support Document	9-34
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
NTP. 2001b. Subchronic toxicity study of dibromoacetonitrile administered in dosed water to
Fischer 344 rats. Research Triangle Park, NC: National Toxicology Program, National Institute
of Environmental Health Sciences, National Institutes of Health.
NTP. 2002a. Abstracts from subacute and subchronic studies of dibromoacetonitrile (DBAN) on
Fischer-344 rats and B6C3F1 mice, and update on test status. Research Triangle Park, NC:
National Toxicology Program, National Institutes of Health, National Institute of Environmental
Health Sciences. Available online at: http://ntp-server.niehs.nih.gov/.
NTP. 2002b. Pathology Working Group Chairperson's report. 13 week study of
Bromochloroacetic acid (C96019): dibromoacetonitriles (C96015): and bromodichloroacetic acid
(C096014) administered by dosed water in male and female B6C3F1 mice and male and female
F344 rats. Research Triangle Park, NC: National Toxicology Program, National Institutes of
Health, National Institute of Environmental Health Sciences.
NTP. 2006. Toxicity and carcinogenesis studies of bromodichloromethane (CAS No. 75-27-4) in
F344/N rats and B6C3F1 mice (gavage studies). Technical Report Series No. 532. Research
Triangle Park, NC: U.S. Department of Health and Human Services.
NTP. 2007a. NTP Report on the toxicology studies of dichloroacetic acid (CAS No. 79-43-6) in
genetically modified (FVB Tg.AC hemizygous) mice (dermal and drinking water studies) and
carcinogenicity studies on dichloroacetic acid in genetically modified [B6.129-Trp53tmlBrd
(N5) haploidinsufficient) mice (drinking water studies). NTP GMM 11. Research Triangle Park,
NC: U.S. Department of Health and Human Services.
NTP. 2007b. NTP report on the toxicology studies of sodium bromate in genetically modified
mice and carcinogenicity studies in genetically modified mice. NTP GMM 6. NIH Publication
No. 07-4423.
NTP. 2007c. NTP technical report on the toxicology and carcinogenesis studies of dibromoacetic
acid (CAS No. 631-64-1) in F344/N rats and B6C3F1 mice (drinking water studies). NTP
Technical Report Series No. 537. Research Triangle Park, NC: National Toxicology Program,
National Institutes of Health, Public Health Service, U.S. Department of Health and Human
Services. Available online at: http://ntp.niehs.nih.gov/index.cfm?obiectid=8831333E-FlF6-
975E-71D4F287C2229308.
NTP. 2009. NTP technical report on the toxicology and carcinogenesis studies of
bromochloroacetic acid (CAS No. 5589-96-8) in F344/N rats and B6C3F1 mice (drinking water
studies). Technical Report Series No. 549. Research Triangle Park, NC: U.S. Department of
Health and Human Services.
NTP. 2010. Toxicology and carcinogenesis studies of dibromoacetonitrile (CAS No. 3252-43-5)
in F344/N rats and B6C3F1 mice (drinking water studies). Technical Report Series No. 544.
Research Triangle Park, NC: U.S. Department of Health and Human Services.
NTP. 2015. Toxicology studies of bromodichloroacetic acid (CAS No. 71133-14-7) in F344 rats
and B6C3F1 mice and toxicology and carcinogenesis studies of bromodichloroacetic acid in
Six-Year Review 3 Technical Support Document	9-35
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
F344/NTac rats and B6C3F1/N mice (drinking water studies). Technical Report Series No. 583.
Research Triangle Park, NC: U.S. Department of Health and Human Services.
Neeman, J., R. Hulsey, D. Rexing, and E. Wert. 2004. Current Issues - controlling bromate
formation during ozonation with chlorine and ammonia. Journal of the American Water Works
Association. 96(2): 26-29.
Netcher, A.C. and S.J. Duranceau. 2015. Pretreatment of low alkalinity organic-laden surface
water prior to a coagulation-ultrafiltration membrane process. Water Research Foundation.
Project #4477.
Ng, T.W., B. Li, A.T. Chow, and P.K. Wong. 2015. Formation of disinfection by-products from
bacterial disinfection. In Recent Advances in Disinfection By-Products. ACS Symposium Series,
1190.
Nguyen, M.L., P. Westerhoff, L. Baker, Q. Hu, M. Esparza-Soto, and M. Sommerfeld. 2005.
Characteristics and reactivity of algae-produced dissolved organic carbon. Journal of
Environmental Engineering. 131(11): 1574-1582.
Nickmilder, M. and A. Bernard. 2011. Associations between testicular hormones at adolescence
and attendance at chlorinated swimming pools during childhood. International Journal of
Andrology. 34: e446-e458.
Nieminski, E.C., and S.A.L. Perry. 2015. State and provincial perspectives on the regulation of
biological filtration. Journal of the American Waterworks Association. 107(12): 32-39.
Nieuwenhuijsen, M.J., M.B. Toledano, N.E. Eaton, J. Fawell, and P. Elliott. 2000. Chlorination
disinfection byproducts in water and their association with adverse reproductive outcomes: a
review. Occupational and Environmental Medicine. 57: 73-85.
Nieuwenhuijsen M.J., M.B. Toledano, J. Bennett, N. Best, P. Hambly, C. de Hoogh, D.
Wellesley, P. A. Boyd, L. Abramsky, N. Dattani, J. Fawell, D. Briggs, L. Jarup, and P. Elliott.
2008. Chlorination disinfection by-products and risk of congenital anomalies in England and
Wales. Environmental Health Perspectives. 116(2): 216-222.
Nieuwenhuijsen, M.J., D. Martinez, J. Grellier, J. Bennett, N. Best, N. Iszatt, M. Vrijheid, and
M.B. Toledano. 2009. Chlorination disinfection by-products in drinking water and congenital
anomalies: Review and meta-analyses. Environmental Health Perspectives. 117(10): 1486-1493.
Obolensky, A., and P.C. Singer. 2008. Development and interpretation of disinfection byproduct
formation models using the information collection rule dataset. Environmental Science and
Technology. 42(15): 5654-5660.
Obolensky, A., P.C. Singer, and H. M. Shukairy. 2007. Information collection rule data
evaluation and analysis to support impacts on disinfection by-product formation. Journal of
Environmental Engineering. 133(1): 53-63.
Six-Year Review 3 Technical Support Document	9-36
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Obolensky, A. 2002. Occurrence of haloacetic acids in ICR finished water and distribution
systems. In Information Collection Rule Data Analysis, Chapter 5. Denver, CO: American Water
Works Association Research Foundation.
Oneby, M., A.L. Reid, and S. Mahmutoglu. 2009. Model sheds light on TTHM formation.
Opflow. 35(11): 20-23.
Organisation for Economic Co-operation and Development Screening Information Dataset
(OECD SIDS). 2003. Mucochloric Acid CAS: 87-56-9. Available online at:
http://www.inchem.org/documents/sids/sids/87569.pdf
Orme, J., D.H. Taylor, R.D. Laurie, and R.J. Bull. 1985. Effects of chlorine dioxide on thyroid
function in neonatal rats. Journal of Toxicology and Environmental Health. 15: 315-322.
Padhye, L., P. Wang, T. Karanfil, and C. Huang. 2010. Unexpected role of activated carbon in
promoting transformation of secondary amines to A-nitrosamines. Environmental Science and
Technology. 44(11): 4161-4168.
Palmer, I.S. and O.E. Olson. 1979. Partial prevention by cyanide of selenium poisoning in rats.
Biochemical and Biophysical Research Communications. 90(4): 1379-1386.
Pals, J.A., J.K. Ang, E.D. Wagner, and M.J. Plewa. 2011. Biological mechanism for the toxicity
of haloacetic acid drinking water disinfection byproducts. Environmental Science and
Technology. 45: 5791-5797.
Pan, Y. and X. Zhang. 2012. Four groups and new aromatic halogenated disinfection byproducts:
effect of bromide concentration on their formation and speciation in chlorinated drinking water.
Environmental Science and Technology. 47(3): 1265-1273.
Panyapinyopol, B., T.F. Marhaba, V. Kanokkantapong, and P. Pavasant. 2005a. Characterization
of precursors to trihalomethanes formation in Bangkok source water. Journal of Hazardous
Materials. 120(1): 229-236.
Panyapinyopol, B., V. Kanokkantapong, T.F. Marhaba, S.W. Wattanachira, and P. Pavasant.
2005b. Kinetics of trihalomethane formation from organic contaminants in raw water from the
Bangkhen water treatment plant. Journal of Environmental Science and Health. 40(8): 1543-
1555.
Park, S.H., L.P. Padhye, P. Wang, M. Cho, J-H. Kim, and C-H. Huang. 2015. N-
nitrosodimethylamine (NDMA) formation potential of amine based water treatment polymers:
Effects of in-situ chloramination, breakpoint chlorination, and pre-oxidation. Journal of
Hazardous Materials. 282: 133-140.
Parker, K.M., T. Zheng, J. Harkness, A. Vengosh, and W.A. Mitch. 2014. Enhanced formation of
disinfection byproducts in shale gas wastewater-impacted drinking water supplies.
Environmental Science and Technology. 48(19): 11161-11169.
Six-Year Review 3 Technical Support Document	9-37
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Partinoudi, V., and M. R. Collins. 2007. Assessing RBF reduction/removal mechanisms for
microbial and organic DBP precursors. Journal of the American Water Works Association.
99(12): 61-71.
Patelarou, E., S. Kargaki, E.G. Stephanou, M. Nieuwenhuijsen, P. Sourtzi, E. Gracia, L. Chatzi,
A. Koutis, and M. Kogevinas. 2011. Exposure to brominated trihalomethanes in drinking water
and reproductive outcomes. Occupational and Environmental Medicine. 68(6): 438-445.
Patterson, C., A. Anderson, R. Sinha, N. Muhammad, and D. Pearson. 2012. Nanofiltration
membranes for removal of color and pathogens in small public drinking water sources. Journal
of Environmental Engineering. 138(1): 48-57.
Pegram, R.A., M.E. Andersen, S.H. Warren, T.M. Ross, and L.D. Claxton. 1997. Glutathione S-
transferase-mediated mutagenicity of trihalomethanes in Salmonella typhimurium: contrasting
results with bromodichloromethane and chloroform. Toxicol. Appl. Pharmacol. 144(1): 183-188.
Pegram, R.A. 2001. Disinfection by-product pharmacokinetics. In: Microbial Pathogens and
Disinfection By-products in Drinking Water: Health Effects and Management of Risks, eds.
Craun, G.F., pp. 273-290 Washington, D.C.: ILSI Press.
Pehlivanoglu-Mantas, E. and D.L. Sedlak. 2006. The fate of wastewater-derived NDMA
precursors in the aquatic environment. Water Research. 40(6): 1287-1293.
Pereira, M.A. 1996. Carcinogenic activity of dichloroacetic acid and trichloroacetic acid in the
liver of female B6C3F1 mice. Fundamental Applied Toxicology. 31: 192-199.
Pereira, M.A., W. Wang, P.M. Kramer, and L. Tao. 2004a. DNA hypomethylation induced by
nongenotoxic carcinogens in mouse and rat colon. Cancer Letters. 212: 145-151.
Pereira, MA, W. Wang, P.M. Kramer, and L. Tao. 2004b. Prevention by methionine of
dichloroacetic acid-induced liver cancer and DNA hypomethylation in mice. Toxicological
Sciences. 77: 243-248.
Pharand, L., M.I. van Dyke, W.B. Anderson, Y. Yohannes, and P.M. Huck. 2015. Full-scale
ozone-biofiltration: Seasonally related effects on NOM removal. Journal of the American Water
Works Association. 107(12): 425-436.
Phetrak, A., J, Lohwacharin, H. Saka, M. Murakami, K. Oguma, and S. Takizawa. 2014.
Simultaneous removal of dissolved organic matter and bromide from drinking water source by
anion exchange resins for controlling disinfection by-products. Journal of Environmental
Sciences. 26(6): 1294-1300.
Pifer, A.D. and J.L. Fairey. 2012. Improving on SUVA254 using fluorescence-PARAFAC
analysis and asymmetric flow-field flow fractionation for assessing disinfection byproduct
formation and control. Water Research. 46(9): 2927-2936.
Pisarenko, A.N., B.D. Stanford, S.A. Snyder, S.B. Rivera, and A.K. Boal. 2013. Investigation of
the use of chlorine based advanced oxidation in surface water: Oxidation of natural organic
Six-Year Review 3 Technical Support Document	9-38
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
matter and formation of disinfection byproducts. Journal of Advanced Oxidation Technologies.
16(1): 137-150.
Pivarnik, M. 1998. Potential effects of maternal physical activity on birth weight: Brief review.
Medicine and Science in Sports and Exercise. 30(3): 400-406.
Plewa, M.J., Y. Kargalioglu, D. Vankerk, R.A. Minear, and E.D. Wagner. 2002. Mammalian cell
cytotoxicity and genotoxicity analysis of drinking water disinfection by-products. Environmental
and Molecular Mutagenesis. 40: 134-142.
Plewa, M.J., E.D. Wagner, S.D. Richardson, A.D. Thruston Jr., Y.T. Woo, and A.B. McKague.
2004a. Chemical and biological characterization of newly discovered iodoacid drinking water
disinfection byproducts. Environmental Science and Technology. 38(18): 4713-4722.
Plewa, M.J., E.D. Wagner, P. Jazwierska, S.D. Richardson, P.H. Chen, and A.B. McKague.
2004b. Halonitromethane drinking water disinfection byproducts: chemical characterization and
mammalian cell cytotoxicity and genotoxicity. Environmental Science and Technology. 38(1):
62-68.
Plewa, M. J., M.G. Muellner, S.D. Richardson, F. Fasano, K.M. Buettner, Y.T. Woo, A.B.
McKague, and E.D. Wagner. 2008. Occurrence, synthesis and mammalian cell cytotoxicity and
genotoxicity of haloacetamides: an emerging class of nitrogenous drinking water disinfection by-
products. Environmental Science and Technology. 42: 955- 961.
Plewa, M.J. and E.D. Wagner. 2009. Mammalian cell cytotoxicity and genotoxicity of
disinfection by-products. Water Research Foundation. Project #3089.
Plewa, M.J., J.E. Simmons, S.D. Richardson, and E.D. Wagner. 2010. Mammalian cell
cytotoxicity and genotoxicity of the haloacetic acids, a major class of drinking water disinfection
by-products. Environmental and Molecular Mutagenesis. 51(8-9): 871-878.
Plourde-Lescelleur, F., I. Papineau, A. Carriere, A. Gadbois, andB. Barbeau. 2015. NOM
removal: evaluating five process alternatives to alum coagulation. Journal of Water Supply:
Research and Technology-AQUA. 64(3): 278-289.
Pope, P.G., M. Martin-Doole, G.E. Speitel Jr., and M.R. Collins. 2007. Relative significance of
factors influencing DXAA formation during chloramination. Journal of the American Water
Works Association. 99(9): 144-156.
Porter C.K., S.D. Putnam, K.L. Hunting, and M.R. Riddle. 2005. The effect of trihalomethane
and haloacetic acid exposure on fetal growth in Maryland county. American Journal of
Epidemiology. 162: 334-344.
Potter, B.B., and J.C. Wimsatt. 2012. U.S. EPA Method 415.3: quantifying TOC, DOC, and
SUVA. Journal of the American Waterworks Association. 104(6): 358-369.
Six-Year Review 3 Technical Support Document	9-39
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Qian, Y., W. Wang, J.M. Boyd, M. Wu, S.E. Hrudey, and X-F. Li. 2012. UV-induced
transformation of four halobenzoquinones in drinking water. Environmental Science and
Technology. 47(9): 4426-4433.
Radjenovic, J., M. Jose Farre, Y. Mu, W. Gernjak, and J. Keller. 2012. Reductive
electrochemical remediation of emerging and regulated disinfection byproducts. Water Research.
46(6): 1705-1714.
Rahman, M.B., T. Driscoll, C. Cowie, and B.K. Armstrong. 2010. Disinfection by-products in
drinking water and colorectal cancer: a meta-analysis. International Journal of Epidemiology. 39:
733-745.
Rahman, M.B., C. Cowie, T. Driscoll, R.J. Summerhayes, B.K. Armstrong, and M.S. Clements.
2014. Colon and rectal cancer incidence and water trihaomethane concentrations in New South
Wales, Australia. BMC Cancer. 14: 445-454.
Ranmuthugala, G., L. Pilotto, W. Smith, T. Vimalasiri, K. Dear, and R. Douglas. 2003.
Chlorinated drinking water and micronuclei in urinary bladder epithelial cells. Epidemiology.
14(5): 617-622.
Reckhow D.A., P.L.S. Rees, and D. Bryan. 2004. Watershed sources of disinfection by-product
precursors. Water Science and Technology: Water Supply. 4(4): 61-69.
Reckhow, D.A., P. Rees, K. Nusslein, G. Makdissy, G. Devine, T. Conneely, A. Boutin, and D.
Bryan. 2007. Long term variability of BDOM and NOM as precursors in watershed sources.
Water Environment Research Foundation. Available online at:
http://www.waterrf.org/publicreportlibrarv/91186.pdf.
Reckhow, D.A., K.G. Linden, J. Kim, H. Shemer, and G. Makdissy. 2010. Effect of UV
treatment on DBP formation. Journal of the American Water Works Association. 102(6): 100-
113.
Regli, S., J. Chen, M. Messner, M. Elovitz, F. Letkiewicz, R. Pegram, T.J. Pepping, S.
Richardson, and J.M. Wright. 2015. Estimating potential increased bladder cancer risk due to
increased bromide concentrations in sources of disinfected drinking waters. Environmental
Science and Technology. 49(22): 13094-13102.
Reif, J., M. Hatch, M. Bracken, L.B. Holmes, B.A. Schwetz, and P.C. Singer. 1996.
Reproductive and developmental effects of disinfection by-products in drinking water. Environ
Health Perspect. 104(10): 1056-1061.
Reif, J. S., A. Bachand, et al. 2000. Reproductive and developmental effects of disinfection by-
products. Fort Collins, CO: Department of Environmental Health, Colorado State University.
Reitz, R. H., A.L. Mendrala, and F.P. Guengerich. 1989. In vitro metabolism of methylene
chloride in human and animal tissues: Use in physiologically based pharmacokinetic models.
Toxicology and Applied Pharmacology. 97: 230-246.
Six-Year Review 3 Technical Support Document	9-40
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Rice, J., A. Wutich, and P. Westerhoff. 2013. Assessment of de facto wastewater reuse across the
US: trends between 1980 and 2008. Environmental Science and Technology. 47(19): 11099-
11105.
Rice, J. and P. Westerhoff. 2014. Spatial and temporal variation in de facto wastewater reuse in
drinking water systems across the USA. Environmental Science and Technology. 49(2): 982-
989.
Richardson, S.D. 2003. Disinfection by-products and other emerging contaminants in drinking
water. Trends in Analytical Chemistry. 22(10) 666-684.
Richardson, S.D. and C. Postigo. 2011. Disinfection byproducts: formation and occurrence in
drinking water. Encyclopedia of environmental health. 1: 110-136
Richardson, S.D. and T.A. Ternes. 2014. Water analysis: emerging contaminants and current
issues. Analytical Chemistry. 86(6): 2813-2848.
Richardson, S. D., M.J. Plewa, E.D. Wagner, R. Schoeny, and D.M. DeMarini. 2007.
Occurrence, genotoxicity, and carcinogenicity of regulated and emerging disinfection by-
products in drinking water: a review and roadmap for research. Mutation Research. 636: 178-
242.
Richardson, S.D., F. Fasano, J.J. Ellington, F.G. Crumley, K.M. Buettner, J.J. Evans, B.C.
Blount, L.K. Silva, T.J. Waite, G.W. Luther, A.B. McKague, R.J. Miltner, E.D. Wagner, and
M.J. Plewa. 2008. Occurrence and mammalian cell toxicity of iodinated disinfection byproducts
in drinking water. Environmental and Science Technology: 42: 8330-8338.
Righi, E., P. Bechtold, D. Tortorici, P. Lauriola, E. Calzolari, G. Astolfi, M.J. Nieuwenhuijsen,
G. Fantuzzi, and G. Aggazzotti. 2012. Trihalomethanes, chlorite, chlorate in drinking water and
risk of congenital anomalies: a population-based case-control study in Northern Italy.
Environmental Research. 116: 66-73.
Rivera-Nunez, Z. and J.M. Wright. 2013. Association of brominated trihalomethane and
haloacetic acid exposure with fetal growth and preterm delivery in Massachusetts. Journal of
Occupational and Environmental Medicine. 55(10): 1125-1134.
Roberts, M.G., P.C. Singer, and A. Obolensky. 2002. Comparing total HAA and total THM
concentrations using ICR data. Journal of the American Water Works Association 94(1): 103-
114.
Roccaro, P., H-S. Chang, F.G.A. Vagliasindi, and G.V. Korshin. 2008. Differential absorbance
study of effects of temperature on chlorine consumption and formation of disinfection by-
products in chlorinated water. Water Research. 42(8): 1879-1888.
Roccaro, P., G.A. Federico, Vagliasindi, and G.V. Korshin. 2015. Bromination and chlorination
of NOM: new modeling approaches and mechanistic insights. In Recent Advances in
Disinfection By-Products. ACS Symposium Series 1190.
Six-Year Review 3 Technical Support Document	9-41
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Rodriguez-Mozaz, S., H. Weinburg, A. Saenz de Jubera, B. Lyon, J. Chu, H. Colangelo, and A.
Sykes. 2010. Disinfection byproduct formation in water treated with equivalent doses of
conventional disinfectants and mixed oxidant solutions. WQTC 2010 Conference Proceedings.
Roe, F.J.C., A.K. Palmer, and A.N. Worden. 1979. Safety evaluation of toothpaste containing
chloroform. I. long-term studies in mice. Journal of Environmental Pathology, Toxicology, and
Oncology. 2: 799-819.
Rook, J.J. 1974. Formation of haloforms during chlorination of natural water. Water Treatment
and Examination. 23(2): 234-243.
Rosario-Ortiz, F.L., S.A. Snyder, and I.H. Suffet. 2007. Characterization of dissolved organic
matter in drinking water sources impacted by multiple tributaries. Water Research. 41(18): 4115-
4128.
Ross, M.K. and R.A. Pegram. 2003. Glutathione transferase theta 1-1-dependent metabolism of
the water disinfection byproduct bromodichloromethane. Chemical Research Toxicology. 16:
216-226.
Ross, M.K. and R.A. Pegram. 2004. In vitro biotransformation and genotoxicity of the drinking
water disinfection byproduct bromodichloromethane: DNA binding mediated by glutathione
transferase theta 1-1. Toxicology Applied. Pharmacology. 195: 166-181.
Ruddick, J.A., D.C. Villeneuve, D.C., I. Chu, and V.E. Valli. 1983. A teratological assessment of
four trihalomethanes in the rat. Journal Environmental Science. Health, Part B. 18: 333-349.
Sacher, F., C.K. Schmidt, C. Lee, and U. von Gunten. 2008. Strategies for minimizing
nitrosamine formation during disinfection. Water Research Foundation. Project #2979.
Saghir, S.A. and I.R. Schultz. 2005. Toxicokinetics and oral bioavailability of halogenated acetic
acid mixtures in naive and GSTzeta-depleted rats. Toxicologic Sciences. 84: 214-224.
Saito, H., S. Isoda, M. Kato, and N. Nagaoka. 1995. Mutagenic activity of indoor swimming
pool water. Mutation Research/Environmental Mutagenesis and Related Subjects. 17: 169-177.
Samson, C. 2015 Assessing DBP occurrence: impacts of the stage 2 DBPR. WQTC Conference
Proceedings.
Samson, C., B. Rajagopalan, and S. Summers. 2013. Modeling TOC threshold exceedances for
meeting disinfection by-product drinking water regulations under the impact of climate change.
In Proceedings of International Annual Meeting of American Society of Agronomy/Crop Science
Society of America/Soil Science Society of America. November 4, 2013.
Sanchez-Polo, M., J. Rivera-Utrilla, E. Salhi, and U. von Gunten. 2006. Removal of bromide and
iodide anions from drinking water by silver-activated carbon aerogels. Journal of Colloid and
Interface Science. 300(1): 437-441.
Six-Year Review 3 Technical Support Document	9-42
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Sanchez-Polo, M., J. Rivera-Utrilla, E. Salhi, and U. von Gunten. 2007. Ag-doped carbon
aerogels for removing halide ions in water treatment. Water Research. 41(5): 1031-1037.
Saunders, J.F., A.K. Hohner, R.S. Summers, and F.L. Rosario-Ortiz. 2015. Regulating
chlorophyll a to control DBP precursors in water supply reservoirs. Journal of American Water
Work Association. 107: 11-18.
Savitz, D.A., K.W. Andrews, and L.M. Pastore. 1995. Drinking water and pregnancy outcome in
central North Carolina: Source, amount, and trihalomethane levels. Environmental Health
Perspectives. 103(6): 592-596.
Savitz, D.A., P.C. Singer, K.E. Hartmann, A.H. Herring, H.S. Weinberg, C. Makarushka, C.
Hoffman, R. Chan, and R. Maclehose. 2005. Drinking Water Disinfection By-Products and
Pregnancy Outcome. Denver, CO: AWWA Research Foundation.
Schendel, D.B., Z.K. Chowdhury, C.P. Hill, S. Summers, E. Towler, R. Balaji, R.S. Raucher,
and J. Cromwell. 2009. Technology Primers for the Simultaneous Compliance Tool. Denver,
CO: Water Research Foundation.
Scientific Committee on Consumer Safety (SCCS). 2011. Opinion on chloroacetamide.
Available online at:
http://ec.europa.eu/health//sites/health/files/scientific committees/consumer safetv/docs/sccs o
053.pdf
Seitz, H.K. and F. Stickel. 2007. Molecular mechanisms of alcohol-mediated carcinogenesis.
Nature Reviews Cancer. 7: 599-612.
Selbes, M., J. Shan, S.K. Bekaroglu, and T. Karanfil. 2015. Carbonaceous and nitrogenous
disinfection by-product formation potentials of amino acids. In Recent Advances in Disinfection
By-Products. ACS Symposium Series, 1190.
Serrano, M., M. Silva, and M. Gallego. 2014. Fast and "green" method for the analytical
monitoring of haloketones in treated water. Journal Chromatography A. 1358: 232-239.
Shah, A.D., S.W. Krasner, C. Fen, T. Lee, U. von Gunten, and W.A. Mitch. 2012. Trade-offs in
disinfection byproduct formation associated with precursor preoxidation for control of n-
nitrosodimethylamine formation. Environmental Science and Technology. 46(9): 4809-4818.
Shaw, G.M., S.H. Swan, J. A. Harris, and L.H. Malcoe. 1990. Maternal water consumption
during pregnancy and congenital cardiac anomalies. Epidemiology. 1(3):206-211.
Shaw G.M., L.H. Malcoe, A. Milea, and S.H. Swan. 1991 Chlorinated water exposures and
congenital cardiac anomalies. Epidemiology. 2: 459-460.
Shaw, G. M., D. Ranatunga, T. Quach,, E. Neri, A. Correa, and R. Neutra. 2003. Trihalomethane
exposures from municipal water supplies and selected congenital malformations. Epidemiology
14: 191-199.
Six-Year Review 3 Technical Support Document	9-43
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Shoaf, D. R. and P.C. Singer. 2007. An analysis of monitoring data for the stage 1
disinfectants/disinfection by-products rule. Journal of American Water Works Association.
99(10): 69-80.
Shokeer, A. and B. Mannervik. 2010. Residue 234 is a master switch of the alternative-substrate
activity profile of human and rodent theta class glutathione transferase Tl-1. Biochemical et
Biophysica Acta. 1800(4): 466-473.
Shuai, D., D.C. McCalman, J.K. Choe, J.R. Shapley, W.F. Schneider, and C.J. Werth. 2012.
Structure sensitivity study of waterborne contaminant hydrogenation using shape- and size-
controlled Pd nanoparticles. ACS Catalysis. 3(3): 453-463.
Simmons, J.E., S.D. Richardson, T.F. Speth, R.J. Miltner, G. Rice, K.M. Schenck, E.S. Hunter
III, and L.T. Teuschler. 2002. Development of a research strategy for integrated technology-
based toxicological and chemical evaluation of complex mixtures of drinking water disinfection
byproducts. Environmental Health Perspectives. 110(Suppl 6): 1013-1024.
Simmons, J.E., L.T. Teuschler, C. Gennings, T.F. Speth, S.D. Richardson, R.J. Miltner, M.G.
Narotsky, K.M. Schenck, E.S. Hunter III, R.C. Hertzberg, and G. Rice. 2004. Component-based
and whole-mixture techniques for addressing the toxicity of drinking water disinfection
byproducts. Journal of Toxicology Environmental Health, Part A. 67: 741-754.
Simmons, J.E., S.D. Richardson, L.T. Teuschler, R.J. Miltner, T.F. Speth, K.M. Schenck, E.S.
Hunter III, and G. Rice. 2008. Research issues underlying the four-lab study: integrated
disinfection by-products research. Journal of Toxicology Environmental Health, Part A. 71:
1125-1132.
Sinfield, L., and T. Niday. 2015. Spray aeration improves San Clemente Island drinking water.
Currents. Winter 2015. 44-51.
Singer, P. 2010. Anomalous DBP speciation patterns: examples and explanations. WQTC
Conference Proceedings.
Singer, P.C. 2006. DBPs in drinking water: additional scientific and policy considerations for
public health protection. Journal of the American Water Works Association. 98(10): 73-80.
Singer, P.C., T. Boyer, A. Holmquist, J. Morran, and M. Bourke. 2009. Integrated analysis of
NOM removed by magnetic ion exchange. Journal of the American Water Works Association.
100(1): 65-73.
Singh, R. 2005a. Testicular changes in rat exposed to trichloroacetic acid (TCA) during
organogenesis. Biomed Research. 16: 45-52.
Singh, R. 2005b. Effect of maternal administration of trichloroacetic acid (TCA) on fetal ovary
rats. Biomed Research. 16: 195-200.
Six-Year Review 3 Technical Support Document	9-44
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Singh, R. 2006. Neuroembryopathic effect of trichloroacetic acid in rats exposed during
organogenesis. Birth Defects Research Part B: Developmental and Reproductive Toxicology. 77:
47-52.
Sivey, J.D., M.A. Bickley, and D.A. Victor. 2015. Catalysis of DBP-precursor bromination by
halides and hypochlorous acid. In Recent Advances in Disinfection By-Products. ACS
Symposium Series, 1190.
Smith, M.K., H. Zenick, and E.L. George. 1986. Reproductive toxicology of disinfection by-
products. Environmental Health Perspectives. 69: 177-182.
Smith, M.K., J.L. Randall, D.R. Tocco, R.G York, J.A. Strober, and E.J. Read. 1988.
Teratogenic effects of trichloroacetonitrile in the Long-Evans rat. Teratology. 38: 113-120.
Smith, M.K., J.L. Randall, J.A. Stober, and E.J. Read. 1989a. Developmental toxicity of
dichloroacetonitrile: a by-product of drinking water disinfection. Fundamental and Applied
Toxicology. 12(4): 765-772.
Smith, M.K., J.L. Randall, E.J. Read, and J.A. Stober. 1989b. Teratogenic activity of
trichloroacetic acid in the rat. Teratology. 40: 445-451.
Smith, M.K., J.L. Randall, E.J. Read, and J.A. Stober. 1990. Developmental effects of
chloroacetic acid in the Long-Evans rat. Teratology: 41(5): 593.
Smith, M.K., J.L. Randall, E.J. Read, E.J., and J.A. Stober. 1992. Developmental toxicity of
dichloroacetate in the rat. Teratology: 46, 217-233.
Soffritti, M., F. Belpoggi, L. Lambertin, M. Lauriola, M. Padovani, and C. Maltoni. 2002.
Results of longterm experimental studies on the carcinogenicity of formaldehyde and
acetaldehyde in rats. Annals of the New York Academy of Sciences. 982: 87-105.
Sohn, J., G. Amy, and Y. Yoon. 2006. Bromide ion incorporation into brominated disinfection
by-products. Water, Air, and Soil Pollution. 174: 265-277.
Sohn, J., G. Amy, and Y. Yoon. 2007. Process-train profiles of NOM through a drinking water
treatment plant. Journal of the American Waterworks Association. 99(6): 145-153.
Sorlini, S., and C. Collivignarelli. 2005. Trihalomethane formation during chemical oxidation
with chlorine, chlorine dioxide and ozone of ten Italian natural waters. Desalination. 176(1-3):
103-111.
Speight, V.L., and P.C. Singer. 2005. Association between residual chlorine loss and HAA
reduction in distribution systems. Journal of the American Water Works Association. 97(2): 82-
91.
Speitel, G., R. Kannappan, and B. Bayer. 2011. The nitrification index: A unified concept for
quantifying the risk of distribution system nitrification. Journal of the American Water Works
Association. 103(1): 69-80.
Six-Year Review 3 Technical Support Document	9-45
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
States, S., G. Cyprych, M. Stoner, F. Wydra, J. Kuchta, J. Monnell, and L. Casson. 2013.
Marcellus shale drilling and brominated THMs in Pittsburgh, PA., drinking water. Journal of the
American Water Works Association. 105(8): 432-448.
Stauber, A.J. and R.J. Bull. 1997. Differences in phenotype and cell replication behavior of
hepatic tumors induced by dichloroacetate (DCA) and trichloroacetate (TCA). Toxicology and
Applied Pharmacology. 144:235-246.
Stayner, L.T., M. Pedersen, E. Patelarou, I. Decordier, K. Vande Loock, L. Chatzi, A. Espinosa,
E. Fthenou, M.J. Nieuwenhuijsen, E. Gracia-Lavedan, E.G. Stephanou, M. Kirsch-Volders, and
M. Kogevinas. 2014. Exposure to brominated trihalomethanes in water during pregnancy and
micronuclei frequency in maternal and cord blood lymphocytes. Environmental Health
Perspectives. 122(1): 100-106.
Suh, D.H., M.S. Abdel-Rahman, and R.J. Bull. 1983. Effect of chlorine dioxide and its
metabolites in drinking water on fetal development in rats. Journal of Applied Toxicology. 3: 75-
79.
Summerhayes, R.J., G.G. Morgan, H.P. Edwards, D. Lincoln, A. Earnest, B. Rahman, and J.R.
Beard. 2012. Exposure to trihalomethanes in drinking water and small-for-gestational-age births.
Epidemiology. 23(1): 15-22.
Summers, R.S., M.A. Benz, H. M. Shukairy, and L. Cummings. 1993. Effect of separation
processes on the formation of brominated THMs. Journal of the American Water Works
Association. 95: 88-95.
Sun, Y-X., Q-Y. Wu, H-Y. Hu, and J. Tian. 2009. Effect of ammonia on the formation of THMs
and HAAs in secondary effluent chlorination. Chemosphere. 76(5): 631-637.
Swan, S.H., K. Waller, B. Hopkins, G. Windham, L. Fenster, C. Schaefer, and R.R. Beutra.
1998. A prospective study of spontaneous abortion: Relation to amount and source of drinking
water consumed in early pregnancy. Epidemiology. 9(2): 126-33.
Symons, J.M., T.A. Bellar, J.K. Carswell, J. DeMarco, K.L. Kropp, G.G. Robeck, D.R. Seeger,
C.J. Slocum, B.L. Smith, and A.A. Stevens. 1975. National organics reconnaissance survey for
halogenated organics. Journal of the American Water Works Association. 67(11): 634-647.
Symons, J. M., S.W. Krasner, L.A. Simms, andM. Sclimenti. 1993. Measurement of THM and
precursor concentrations revisited: the effect of bromide ion. Journal of the American Water
Works Association. 85(1): 51-62.
Tabrez, S. and M. Ahmad. 2010. Cytochrome P450 system as a toxicity biomarker of industrial
wastewater in rat tissues. Food and Chemical Toxicology. 48(3): 998-1001.
Tang, H.L., R.J. Ristau II, and Y.F. Xie. 2015. Disinfection by-products in swimming pool
water: formation, modeling, and control. In Recent Advances in Disinfection By-Products. ACS
Symposium Series 1190.
Six-Year Review 3 Technical Support Document	9-46
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Tao, L., Y. Li, P.M. Kramer, W. Wang, L. Li, and M.A. Pereira. 2004a. Hypomethylation of
DNA and the insulin-like growth factor-II gene in dichloroacetic and trichloroacetic acid-
promoted mouse liver tumors. Toxicology. 196: 127-136.
Tao, L., W. Wang, L. Li, P.M. Kramer, and M.A. Pereira. 2004b. Effect of dibromoacetic acid
on dna methylation, glycogen accumulation, and peroxisome proliferation in mouse and rat liver.
Toxicologic Sciences. 82: 62-69.
Tao, L., W. Wang, L. Li, P.M. Kramer, and M.A. Pereira. 2005. DNA hypomethylation induced
by drinking water disinfection by-products in mouse and rat kidney. Toxicologic Sciences. 87(2):
344-352.
Taylor, D.H., and R.J. Pfohl. 1985. Effects of chlorine dioxide on the neurobehavioral
development of rats. In: Water chlorination: chemistry, environmental impact and health effects,
vol. 6, pp. 355-364, eds. Jolley, RL, et al. Chelsea, MI: Lewis Publications.
Teitelman, A.M., L.S. Welch, K.G. Hellenbrand, and M.B. Bracken. 1990 Effect of maternal
work activity on preterm birth and low birth weight. American Journal of Epidemiology. 131(1):
104-113.
Teramoto, S., K. Takahashi, M. Kikuta, and H. Kobayashi. 1998. Potential teratogenicity of 3-
chloro-4-(dichloromethyl)-5-hydroxy-2(5H)- furanone(MX) in micromass in vitro test. Journal
of Toxicology and Environmental Health, Part A. 53: 607- 614.
Teuschler, L.K., R.C. Hertzberg, and J.C. Lipscomb. 2000. Research Report: Risk Assessment of
Mixtures of Disinfectant By-products in Drinking Water. EPA 600-R-03-039. October 2000.
Thier, R., J. B. Taylor, S.E. Pemble, W.G. Humphreys, M. Persmark, B. Ketterer, and F.P.
Guengerich. 1993. Expression of mammalian glutathione S-transferase 5-5 in Salmonella
typhimurium TA1535 leads to base-pair mutations upon exposure to dihalomethanes.
Proceedings of the National Academy of Sciences of the United States of America. 90(18):
8576-8580.
Thier, R., E.H. Delbanco, F.A. Wiebel, E. Hallier, and H.M. Bolt. 1998. Determination of
glutathione transferase (GSTT1-1) activities in different tissues based on formation of
radioactive metabolites using 35S-glutathione. Archives of Toxicology. 72(12): 811-815.
Thompson, D.J., S.D. Warner, and V.B. Robinson. 1974. Teratology studies on orally
administered chloroform in the rat and rabbit. Toxicology and Applied Pharmacology. 29: 348-
357.
Thorsen, T., and H. Flogstad. 2006. Nanofiltration in drinking water treatment, literature review.
Research Report for Techneau Project. December 11, 2006.
Tian, C., R. Liu, H. Liu, and J. Qu. 2013. Disinfection by-products formation and precursors
transformation during chlorination and chloramination of highly-polluted source water:
significance of ammonia. Water Research. 47(15): 5901-5910.
Six-Year Review 3 Technical Support Document	9-47
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Til, H.P., R.A. Woutersen, V.J. Feron, and J.J. Clary. 1988. Evaluation of the oral toxicity of
acetaldehyde and formaldehyde in a 4-week drinking-water study in rats. Food and Chemical
Toxicology. 26(5): 447-452 (as cited inUSEPA, 1991 and 1999).
Toledano, MB., M.J. Nieuwenhuijsen, N. Best, H. Whitaker, P. Hambly, C. de Hoogh, L. Jarup,
and P. Elliott. 2005. Relation of trihalomethane concentrations in public water supplies to
stillbirth and birth weight in three water regions in England. Environmental Health Perspective.
113:225-232.
Toroz, I., and V. Uyak. 2005. Seasonal variations of trihalomethanes (THMs) in water
distribution networks of Istanbul City. Desalination. 176(1): 127-141.
Toth, G.P., R.E. Long, T.S. Mills, and M.K. Smith. 1990. Effects of chlorine dioxide on the
developing rat brain. Journal of Toxicology and Environmental Health. 31: 29-44.
Toth, G.P., K.C. Kelty, E.L. George, E.J. Read, and M.K. Smith. 1992. Adverse male
reproductive effects following subchronic exposure of rats to sodium dichloroacetate.
Fundamental and Applied Toxicology. 19: 57-63.
Tsai, K.P., M.F. Rogers, A.T. Chow, and F. Diaz. 2015. Prescribed fire alters dissolved organic
matter and disinfection by-product precursor in forested watersheds - Part II. A controlled field
study. In Recent Advances in Disinfection By-Products. ACS Symposium Series 1190.
Tully, D.B., J.C. Luft, J.C. Rockett, H. Ren, J.E. Schmid, C.R. Wood, and D.J. Dix. 2005.
Reproductive and genomic effects in testes from mice exposed to the water disinfectant
byproduct bromochloroacetic acid. Reproductive Toxicology. 19: 353-366.
Tumasonis, C.F., D.N. McMartin, and B. Bush. 1987. Toxicity of chloroform and
bromodichloromethane when administered over a lifetime in rats. Journal of Environmental
Pathology, Toxicology, and Oncology. 7: 55-64.
Tzoupanos, N.D., and A.I. Zouboulis. 2009. Novel inorganic-organic composite coagulants
based on aluminum. Desalination and Water Treatment. 13: 340-347.
Umemura, A.T. and Y. Kurokawa. 2006. Etiology of bromate-induced cancer and possible
modes of action - studies in Japan. Toxicology. 221(2-3): 154-157.
United States Environmental Protection Agency (USEPA). 1978. National Organics Monitoring
Survey (NOMS). Technical Support Division, U.S. Cincinnati: Environmental Protection
Agency, Office of Drinking Water.
USEPA. 1979. National Interim Primary Drinking Water Regulations; Control of
Trihalomethanes in Drinking Water; Final Rule. 44 FR 68624. November 29, 1979.
USEPA. 1986. Guidelines for carcinogen risk assessment. 51 FR 33992. September 24, 1986.
Six-Year Review 3 Technical Support Document	9-48
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
USEPA. 1988a. Cyanogen bromide (CASRN 506-68-3) Intergrated Risk Information System
Assessment. Available online at:
https://cfpub.epa.gov/ncea/iris2/chemicalLanding.cfm7substance nmbr=358
USEPA. 1988b. Acetaldehyde (CASRN 75-07-0) Integrated Risk Information System
Assessment. Available online at:
https://cfpub.epa.gov/ncea/iris/iris documents/documents/subst/0290 summary.pdf
USEPA. 1990a. Formaldehyde (CAS No. 50-00-0) Integrated Risk Information System
Assessment. Available online at:
https://cfpub.epa.gov/ncea/iris2/chemicalLanding.cfm7substance nmbr=419
USEPA. 1991. Integrated Risk Information System for Bromoform (CAS #75-25-3). Available
online at: http://www.epa.gov/iris/subst/0214.htm.
USEPA. 1992a. Integrated Risk Information System for Dibromochloromethane (CAS #148-48-
1). Available online at: http://vvvvvv.epa.gov/iris/subst/0214.htm.
USEPA. 1992b. National Primary Drinking Water Regulations: Synthetic Organic Chemicals
and Inorganic Chemicals; Final Rule. 57 FR 31786. July 17, 1992.
USEPA. 1992c. Method 552.1: Determination of Haloacetic Acids and Dalapon in Drinking
Water by Ion-Exchange Liquid-Solid Extraction and Gas Chromatography with an Electron
Capture Detector. Revision 1.0. Environmental Monitoring Systems Laboratory, Office of
Research and Development. August 1992.
USEPA. 1993a. Integrated Risk Information System for Bromodichloromethane (CAS #75-27-
4). Available online at: http://vvvvvv.epa.gov/iris/subst/0213.htm.
USEPA. 1993b. Method 300.0: Determination of Inorganic Anions by Ion Chromatography.
Revision 2.1. Environmental Monitoring Systems Laboratory, Office of Research and
Development. EPA 600-R-93-100. August 1993.
USEPA. 1994a. National Primary Drinking Water Regulations; Disinfectants and Disinfection
Byproducts; Proposed Rule. 59 FR 38668. July 29, 1994.
USEPA. 1994b. National Primary Drinking Water Regulations; Monitoring Requirements for
Public Drinking Water Supplies; Proposed Rule. 59 FR 6332. February 10, 1994.
USEPA. 1995a. Method 502.2: Volatile Organic Compounds in Water by Purge and Trap
Capillary Column Gas Chromatography with Photoionization and Electrolytic Conductivity
Detectors in Series. Revision 2.1. National Exposure Research Laboratory, Office of Research
and Development. EPA 600-R-95-131.
USEPA. 1995b. Method 524.2: Measurement of Purgeable Organic Compounds in Water by
Capillary Column Gas Chromatography/Mass Spectrometry. Revision 4.1. National Exposure
Research Laboratory, Office of Research and Development. EPA 600-R-95-131.
Six-Year Review 3 Technical Support Document	9-49
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
USEPA. 1995c. Method 551.1: Determination of Chlorinated Disinfection Byproducts,
Chlorinated Solvents, and Halogenated Pesticides/Herbicides in Drinking Water by
Liquid/Liquid Extraction and Gas Chromatography with Electron Capture Detection. Revision
1.0. National Exposure Research Laboratory, Office of Research and Development. EPA 600-R-
95-131.
USEPA. 1995d. Method 552.2: Determination of Haloacetic Acids and Dalapon in Drinking
Water by Liquid-Liquid Extraction, Derivatization and Gas Chromatography with Electron
Capture Detection. Revision 1.0. National Exposure Research Laboratory, Office of Research
and Development.
USEPA. 1996a. Proposed Guidelines for Carcinogen Risk Assessment. EPA 600-P-92-003C.
April 1996.
USEPA. 1996b. ICR Manual for Bench- and Pilot-Scale Treatment Studies. EPA 814-B-96-003.
April 1996.
USEPA. 1996c. DBP/ICR Analytical Methods Manual. EPA 814-B-96-002. April 1996.
USEPA. 1997a. National Primary Drinking Water Regulations; Disinfectants and Disinfection
Byproducts; Notice of Data Availability; Proposed Rule. 62 FR 59388. November 3, 1997.
USEPA. 1997b. External Peer Review of CMA Study -2- Generation, EPA Contract No. 68-
C7-0002, Work Assignment B-14, The Cadmus Group, Inc. October 9, 1997 (as cited in
USEPA, 1998a).
USEPA. 1997c. Method 300.1: Determination of Inorganic Anions in Drinking Water by Ion
Chromatography. Revision 1.0. National Exposure Research Laboratory, Office of Research and
Development. EPA 600-R-98-118.
USEPA. 1997c. Method 321.8: Determination of Bromate in Drinking Waters by Ion
Chromatography Inductively Coupled Plasma - Mass Spectrometry. Revision 1.0. National
Exposure Research Laboratory, Office of Research and Development. December 1997.
USEPA. 1998a. National Primary Drinking Water Regulations; Disinfectants and Disinfection
Byproducts; Notice of Data Availability; Proposed Rule. 63 FR 15606. March 31, 1998.
USEPA. 1998b. National Primary Drinking Water Regulations: Disinfectants and Disinfection
Byproducts; Final Rule. 63 FR 69390. December 16, 1998.
USEPA. 1998c. Assessment of Thyroid Follicular Cell Tumors. Risk Assessment Forum. EPA
630-R-97-002. March 1998.
USEPA. 1999a. Guidelines for Carcinogen Risk Assessment. Review Draft. NCEA-F-0644. July
1999.
USEPA. 1999b. Enhanced Coagulation and Enhanced Precipitative Softening Guidance Manual.
EPA 815-R-99-012. May 1999.
Six-Year Review 3 Technical Support Document	9-50
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
USEPA. 2000a. Stage 2 Microbial and Disinfection Byproducts Federal Advisory Committee
Agreement in Principle. 65 FR 83015. December 29, 2000.
USEPA. 2000b. Toxicological Review of Chlorine Dioxide and Chlorite (CAS# 10049-04-4 and
7758-19-2) in Support of Summary Information on the Integrated Risk Information System. EPA
635-R-00-007. September 2000.
USEPA. 2000c. Toxicological Review of Chloral Hydrate (CAS No. 302-17-0) In Support of
Summary Information on the Integrated Risk Information System (IRIS). EPA 635-R-00-006.
August 2000.
USEPA. 2000d. Quantitative Cancer assessment for MX and chlorohydroxyfuranons. Contract
No. 68-C-98-195. Washington, DC: Office of Water. Office of Science and Technology, Health
and ecological Criteria Division.
USEPA. 2000e. ICR Auxiliary 1 Database. EPA 815-C-00-002. April, 2000 version.
USEPA. 2000f. ICR Treatment Study Database. EPA 815-C-00-003.
USEPA. 2001a. Toxicological Review of Chloroform (CAS# 67-66-3) in Support of Summary
Information on the Integrated Risk Information System. EPA 635-R-01-001. October 2001.
USEPA. 2001b. Toxicological Review of Bromate (CAS # 15541-45-4) in Support of Summary
Information on the Integrated Risk Information System. EPA 635-R-01-002. March 2001.
USEPA. 2001c. Method 317.0: Determination of Inorganic Oxyhalide Disinfection By-Products
in Drinking Water Using Ion Chromatography with the Addition of a Postcolumn Reagent for
Trace Bromate Analysis. Revision 2.0. Technical Support Center, Office of Ground Water and
Drinking Water. EPA 815-B-01-001. July 2001.
USEPA. 2002. Method 326.0: Determination of Inorganic Oxyhalide Disinfection By-Products
in Drinking Water Using Ion Chromatography Incorporating the Addition of a Suppressor
Acidified Postcolumn Reagent for Trace Bromate Analysis. Revision 1.0. Technical Support
Center, Office of Ground Water and Drinking Water. EPA 815-R-03-007. June 2002.
USEPA. 2003a. National Primary Drinking Water Regulations; Announcement of Completion of
EPA's Review of Existing Drinking Water Standards. 68 FR 42907. July 18 2003.
USEPA. 2003b. National Primary Drinking Water Regulations: Stage 2 Disinfectants and
Disinfection Byproducts Rule; National Primary and Secondary Drinking Water Regulations:
Approval of Analytical Methods for Chemical Contaminants. 68 FR 49548. August 18, 2003.
USEPA. 2003c. Toxicological Review of Dichloroacetic acid (CAS # 79-43-6) in Support of
Summary Information on the Integrated Risk Information System. EPA 635-R-03-007. August
2003.
USEPA. 2003d. National Primary Drinking Water Regulations; Stage 2 Disinfectants and
Disinfection Byproducts; Proposed Rule. 68 FR 49548. August 18, 2003.
Six-Year Review 3 Technical Support Document	9-51
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
USEPA. 2003e. Method 552.3: Determination of Haloacetic Acids and Dalapon in Drinking
Water by Liquid-Liquid Microextraction, Derivatization, and Gas Chromatography with Electron
Capture Detection. Revision 1.0. Technical Support Center, Office of Ground Water and
Drinking Water. EPA 815-B-03-002. July 2003.
USEPA. 2004a. Drinking Water Criteria Document on Glyoxal and Methylglyoxal. OW 68-C-
98-14.1
USEPA. 2004b. Method 521: Determination of nitrosamines in drinking water by solid phase
extraction and capillary column gas chromatography with large volume injection and chemical
ionization tandem mass spectrometry (MS/MS). Version l.O.National Exposure Research
Laboratory, Office of Research and Development. EPA 600-R-05-054. September 2004.
USEPA. 2005a. Guidelines for carcinogen risk assessment. Risk Assessment Forum. EPA 630-
P-03-001B. March 2005.
USEPA. 2005b. Drinking Water Addendum to the Criteria Document for Monochloroacetic
Acid. EPA-822-R-05-008
USEPA. 2005c. Drinking Water Addendum to the Criteria Document for Trichloroacetic Acid.
EPA 822-R-05-010. November 2005.
USEPA. 2005d. Drinking Water Criteria Document for Brominated Trihalomethanes. Office of
Science and Technology. EPA 822-R-05-011. November 2005.
USEPA. 2005e. Drinking Water Criteria Document for Brominated Acetic Acids. EPA-822-R-
05-007. November 2005.
USEPA. 2005f. Drinking Water Addendum to the IRIS Toxicological review of Dichloroacetic
Acid. EPA-822-R-05-009. November 2005.
USEPA. 2005g. Economic Analysis for the Final Stage 2 Disinfectants and Disinfection
Byproducts Rule, EPA 815-R-05-010. December 2005.
USEPA. 2005i. Drinking Water Criteria Document for Cyanogen Chloride. Office of Water,
Office of Science and Technology, Health and Ecological Criteria Document.
USEPA. 2005j. Method 327.0: Determination of Chlorine Dioxide and Chlorite Ion in Drinking
Water Using Lissamine Green B and Horseradish Peroxidase with Detection by Visible
Spectrophotometry. Revision 1.1. Technical Support Center, Office of Ground Water and
Drinking Water. EPA 815-R-05-008. May 2005.
USEPA. 2005k. Method 415.3: Determination of Total Organic Carbon and Specific UV
Absorbance at 254 nm in Source Water and Drinking Water. Revision 1.1. National Exposure
Research Laboratory, Office of Research and Development. February 2005.
USEPA. 20051. Occurrence Assessment for the Final Stage 2 Disinfectants and Disinfection
Byproducts Rule. EPA 815-R-05-011. December 2005.
Six-Year Review 3 Technical Support Document	9-52
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
USEPA. 2005m. Technologies and Costs Document for the Final Long Term 2 Enhanced
Surface Water Treatment Rule and Final Stage 2 Disinfectants and Disinfection Byproducts
Rule. EPA 815-R-05-013. December 2005.
USEPA. 2005n. Economic Analysis for the Final Stage 2 Disinfectants and Disinfection
Byproducts Rule. Appendix A. December 2005.
USEPA. 2006a. National Primary Drinking Water Regulations: Stage 2 Disinfectants and
Disinfection Byproducts Rule; Final Rule. 71 FR 388. January 4, 2006.
USEPA. 2006b. Reregi strati on Eligibility Decision (RED) for Inorganic Chlorates. Office of
Prevention, Pesticides and Toxic Substances. EPA 738-R-06-014. July 2006.
USEPA. 2007a. Unregulated Contaminant Monitoring Regulation (UCMR) for Public Water
Systems Revisions, Final Rule. 72 FR 368. January 4, 2007.
USEPA. 2007b. Simultaneous Compliance Guidance Manual for the Long-Term 2 and Stage 2
DBP Rules. EPA 815-R-07-017. March 2007.
USEPA. 2009a. Method 302.0: Determination of Bromate in Drinking Water Using Two-
Dimensional Ion Chromatography with Suppressed Conductivity Detection. Version 1.0.
Technical Support Center, Office of Ground Water and Drinking Water. EPA 815-B-09-
014.September 2009.
USEPA. 2009b. Method 334.0: Determination of Residual Chlorine in Drinking Water Using an
On-Line Chlorine Analyzer. Version 1.0. Technical Support Center, Office of Ground Water and
Drinking Water. EPA 815-B-09-013. September 2009.
USEPA. 2009c. Method 415.3: Determination of Total Organic Carbon and Specific UV
Absorbance at 254 nm in Source Water and Drinking Water. Revision 1.2. National Exposure
Research Laboratory, Office of Research and Development. September 2009.
USEPA. 2009d. Method 524.3: Measurement of Purgeable Organic Compounds in Water by
Capillary Column Gas Chromatography/Mass Spectrometry. Version 1.0. Technical Support
Center, Office of Ground Water and Drinking Water. EPA 815-B-09-009. June 2009.
USEPA. 2009e. Method 557: Determination of Haloacetic Acids, Bromate, and Dalapon in
Drinking Water by Ion Chromatography Electrospray Ionization Tandem Mass Spectrometry.
Version 1.0. Technical Support Center, Office of Ground Water and Drinking Water. EPA 815-
B-09-012. September 2009.
USEPA. 2010a. National Primary Drinking Water Regulations; Announcement of the Results of
EPA's Review of Existing Drinking Water Standards and Request for Public Comment and/or
Information on Related Issues. 75 FR 15499. March 29, 2010.
USEPA. 2010b. Toxicological Review of Hydrogen Cyanide and Cyanide Salts (CAS No.
Various) in Support of Summary Information on the Integrated Risk Information System. EPA
635-R-08-016F. June 2010.
Six-Year Review 3 Technical Support Document	9-53
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
USEPA. 2010c. Technical Basis for the Lowest Concentration Minimum Reporting Level
(LCMRL) Calculator. EPA 815-R-l 1-001. December 2010.
USEPA. 2011a. Toxicological Review of Trichloroacetic acid (CAS# 76-03-9) in Support of
Summary Information on the Integrated Risk Information System. EPA 635-R-09-003D. July
2011.
USEPA. 201 lb. Regulatory Impact Analysis for the Final Mercury and Air Toxics Standards.
EPA 452-R-l 1-011.
USEPA. 2012a. Provisional Peer Reviewed Toxicity Values for Cyanogen Bromide (CASRN
506-68-3). Cincinnati, OH: Superfund Health Risk Technical Support Center, National Center
for Environmental Assessment.
USEPA. 2012b. Revisions to the Unregulated Contaminant Monitoring Regulation (UCMR 3)
for Public Water Systems, Final Rule. 77 FR 26071. May 2, 2012.
USEPA. 2012c. UCMR 2 (2008-2010) Occurrence Data. Available online at:
https://www.epa.gOv/dwucmr/occurrence-data-unregulated-contaminant-monitoring-rule#2
USEPA. 2013. Method 524.4: Measurement of Purgeable Organic Compounds in Water by Gas
Chromatography/Mass Spectrometry Using Nitrogen Purge Gas. Technical Support Center,
Office of Water. EPA 815-R-13-002. May 2013.
USEPA. 2014a. Expedited Approval of Alternative Test Procedures for the Analysis of
Contaminants Under the Safe Drinking Water Act; Analysis and Sampling Procedures, Final
Rule. 79 FR 35081. June 19, 2014.
USEPA. 2014b. Support for Six-Year Review of Microbial/Disinfection Byproduct (DBP)
Rules; Task 5 Literature Review for Implementation of Free Chlorine Burns at Public Water
Systems and the Potential Impact on DBP Formation; Final Report. Prepared by NCS Engineers
and Tetra Tech. July 2014.
USEPA. 2015. Revisions to the Unregulated Contaminant Monitoring Rule for Public Water
Systems and Announcement of a Public Meeting. 80 FR 76897. December 11, 2015.
USEPA. 2016a. Six-Year Review 3 Technical Support Document for Microbial Contaminant
Regulations. EPA 810-R-16-010. December 2016.
USEPA. 2016b. Six-Year Review 3 Technical Support Document for Long- Term 2 Enhanced
Surface Water Treatment Rule. EPA 810-R-16-011. December 2016.
USEPA. 2016c. EPA Protocol for the Third Review of Existing National Primary Drinking
Water Regulations. EPA 810-R-16-007. December 2016.
USEPA. 2016d. Six-Year Review 3 Technical Support Document for Nitrosamines. EPA-810-
R-16-009. December 2016.
Six-Year Review 3 Technical Support Document	9-54
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
USEPA. 2016e. Six-Year Review 3 Technical Support Document for Chlorate. EPA-810-R-16-
13. December 2016.
USEPA. 2016f. Third Six-Year Review Information Collection Rule (ICR) Dataset.
USEPA. 2016g. The Analysis of Regulated Contaminant Occurrence Data from Public Water
Systems in Support of the Third Six-Year Review of National Primary Drinking Water
Regulations: Chemical Phase Rules and Radionuclides Rules. EPA 810-R-16-014. December
2016.
USEPA. 2016h. Third Unregulated Contaminant Monitoring Rule Dataset. July, 2016 version.
USEPA. 2016i. The Data Management and Quality Assurance/Quality Control Process for the
Third Six-Year Review Information Collection Rule Dataset. EPA 810-R-16-015. December
2016.
Vetter, C.M., J.E. Miller, L.M. Crawford, M.J. Armstrong, J.H. Clair, M.W. Conner, L.D. Wise,
and T.R. Skopek. 1998. Comparison of motility and membrane integrity to assess rat sperm
viability. Reproductive Toxicology. 12(2): 105-114.
Villanueva, C.M., M. Kogevinas, and J.O. Grimalt. 2001. Drinking water chlorination and
adverse health effects: a review of epidemiological studies. Medicina Clinica. 117(1): 27-35.
[Spanish]
Villanueva, C.M., F. Fernandez, N. Malats, J.O. Grimalt, and M. Kogenvinas. 2003. Meta-
analysis of studies on individual consumption of chlorinated drinking water and bladder cancer.
Journal of Epidemiology Community Health. 57: 166-173.
Villanueva, C.M., K.P. Cantor, S. Cordier, J.J.K. Jaakkola, W.D. King, C.F. Lynch, S. Porru, and
M. Kogevinas. 2004. Disinfection byproducts and bladder cancer a pooled analysis.
Epidemiology. 15(3): 357-367.
Villanueva, C.M., K.P. Cantor, W.D. King, J.J.K. Jaakkola, S. Cordier, C.F. Lynch, S. Porru, and
M. Kogevinas. 2006. Total and specific fluid consumption as determinants of bladder cancer
risk. International Journal of Cancer. 118: 2040-2047.
Villanueva, C.M., K.P. Cantor, J.O. Grimalt, N. Malats, D. Silverman, A. Tardon, R. Garcia-
Closas, C. Serra, A. Carrato, G. Castano-Vinyals, R. Marcos, N. Rothman, F.X. Real, M.
Dosemeci, and M. Kogevinas. 2007. Bladder cancer and exposure to water disinfection by-
products through ingestion, bathing, showering, and swimming in pools. American Journal of
Epidemiology. 165(2): 148-156.
Villanueva, C.M., E. Gracia-Lavedan, J. Ibarluzea, L. Santa Marina, F. Ballester, S. Llop, A.
Tardon, M.F. Fernandez, C. Freire, F. Goni, X. Basagana, M. Kogevinas, J.O. Grimalt, and J.
Sunyer. 2011. Exposure to trihalomethanes through different water uses and birth weight, small
for gestational age, and preterm delivery in Spain. Environmental Health Perspectives. 119(12):
1824-1830.
Six-Year Review 3 Technical Support Document	9-55
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Vinceti, M., G. Fantuzzi, L. Monici, M. Cassinadri, G. Predieri, and G. Aggazzotti. 2004. A
retrospective cohort study of trihalomethane exposure through drinking water and cancer
mortality in northern Italy. Science of the Total Environment. 330(1-3): 47-53.
Wahman, D., L. Katz, and G. Speitel. 2006. Trihalomethane cometabolism by a mixed-culture
nitrifying biofilter. Journal of the American Water Works Association. 98(12): 48-60.
Wahman, D.G. and J.G. Pressman. 2015. Distribution system residuals - is "detectable" still
acceptable for chloramines. Journal of the American Water Works Association. 107(8): 53-63.
Walfoort, C., M.J. Messina, and D. Miner. 2008. Storage tank aeration eliminates
trihalomethanes. Opflow. 34(5): 28-29.
Walker, A. 1991. Observation and Inference. An Introduction to the Methods of Epidemiology.
Chestnut Hill, MA: Epidemiology Resources Inc.
Walker, K.M. and T.H. Boyer. 2011. Long-term performance of bicarbonate-form anion
exchange: removal of dissolved organic matter and bromide from the St. Johns River, FL, USA.
Water Research. 45(9): 2875-2886.
Waller, K., S.H. Swan, G. Delorenze, and B. Hopkins. 1998 Trihalomethanesin drinking water
and spontaneous abortion. Epidemiology. 9(2).
Wang, J., S. Hubbs, and R. Song. 2002. Evaluation of riverbank filtration as a drinking water
treatment process. WRF Report 2622.
Wang, C., X. Zhang, J. Wang, S. Liu, C. Chen, and X. Xie. 2013. Effects of organic fractions on
the formation and control of n-nitrosamine precursors during conventional drinking treatment
processes. Science of the Total Environment. 449: 295-301.
Wang, Y., J. Le Roux, T. Zhang, and J-P. Croue. 2014. Formation of brominated disinfection
byproducts from natural organic matter isolates and model compounds in a sulfate radical-based
oxidation process. Environmental Science and Technology. 48(24): 14534-14542.
Wang, J-J., R.A. Dahlgren, M.S. Ersan, T. Karanfil, and A.T. Chow. 2015a. Wildfire altering
terrestrial precursors of disinfection byproducts in forest detritus. Environmental Science and
Technology. 49(10): 5921-5929.
Wang, D., J.R. Bolton, S. Andrews, and R. Hofmann. 2015b. Formation of disinfection by-
products in the ultraviolet/chlorine advanced oxidation process. Science of the Total
Environment. 518-519: 49-57.
Water Research Foundation (WRF). 2015a. North American Biofiltration Knowledge Base.
Available from the Internet: http://www.biofiltrationknowledgebase.org/.
WRF. 2015b. Research Programs: Focus Area Program. Available from the internet:
http://vvvvvv.vvaterrf.org/the-foundation/research-programs/Pages/Focus-Area-Program.aspx.
Six-Year Review 3 Technical Support Document	9-56
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Warner, N.R., C.A. Christie, R.B. Jackson, and A. Vengosh. 2013. Impacts of shale gas
wastewater disposal on water quality in western Pennsylvania. Environmental Science and
Technology. 47(20): 11849-11857.
Warren, D., L.J. Graeter, S.R. Channel, J.S. Eggers, C.D. Goodyear, K.L. Macmahon, G.L.
Sudberry, J.R. Latendresse, J.W. Fisher, and W.H. Baker. 2006. Trichloroethylene,
trichloroacetic acid, and dichloroacetic acid: do they affect eye development in the Sprague-
Dawley rat. International Journal of Toxicology. 25(4): 279-284.
Watson, K., M.J. Farre, and N. Knight. 2015. Enhanced coagulation with powdered activated
carbon or MIEX® secondary treatment: A comparison of disinfection by-product formation and
precursor removal. Water Research. 68: 454-466.
Wei, X., S. Wang, W. Zheng, X. Wang, X. Liu, S. Jiang, J. Pi, Y. Zheng, G. He, and W. Qu.
2013. Drinking water disinfection byproduct iodoacetic acid induces tumorigenic transformation
ofNIH3T3 cells. Environmental Science and Technology. 47(11): 5913-5920.
Weinberg, H.S., S.W. Krasner, S.D. Richardson, and A.D. Thurston Jr. 2002. The occurrence of
disinfection by-products of health concern in drinking water: results of a nationwide DBP
occurrence study. National Exposure Research Laboratory. EPA 600-R-02-068.
Weiss, W.J., S.C. Schindler, S. Freud, J.A. Herzner, K.F. Hoek, B.A. Wright, D.A. Reckhow,
and W.C. Becker. 2013. Minimizing raw water NOM concentration through source water
selection. Journal of the American Water Works Association. 105(10): 596-608.
Wert, E., and M. Benotti. 2010. Identifying the role of bromamines in minimizing bromate
formation. WQTC 2010 Conference Proceedings.
Wert, E.C. and F.L. Rosario-Ortiz. 2013. Intercellular organic matter from cyanobacteria as a
precursor for carbonaceous and nitrogenous disinfection byproducts. Environmental Science and
Technology. 47(12): 6332-6340.
Wilczak, A., A. Assadi-Rad, H.H. Lai, L.L. Hoover, J.F. Smith, R. Berger, F. Rodigari, J.W.
Beland, L.J. Lazzelle, E.G. Kincannon, H. Baker, and C.T. Heaney. 2003. Formation of NDMA
in chloraminated water coagulated with DADMAC cationic polymer. Journal of the American
Waterworks Association. 95(9): 94-106.
Wilkins, J.R., III and G.W. Comstock. 1981. Source of drinking water at home and site-specific
cancer incidence in Washington County, Maryland. American Journal of Epidemiology. 114(2):
178-190.
Williams, L. and W. Persich. 2014. Programmatic treatment strategies for the control of iron,
manganese, and disinfection by-products. WQTC 2014 Conference Proceedings.
Wilson, J.M. and J.M. Van Briesen. 2013. Source water changes and energy extraction activities
in the Monongahela river, 2009-2012. Environmental Science and Technology. 47(21): 12575-
12582.
Six-Year Review 3 Technical Support Document	9-57
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Windham, G. C., K. Waller, M. Anderson, L. Fenster, P. Mendola, and S. Swan. 2003.
Chlorination by-products in drinking water and menstrual cycle function. Environmental Health
Perspectives. 111(7): 935-941.
Wolf, G.W. and L. Kaiser. 1996. Final report sodium bromate: short term reproductive and
development toxicity study when administered to Sprague Dawley rats in the drinking water.
NTP/NIEHS NO. NOI-ES-15323. Research Triangle Park, NC: National Toxicology Program,
Research Triangle Park, NC.
World Health Organization (WHO). 2000. Disinfectants and disinfectant by-products. Available
online at: http://apps.vvho.int/iris/bitstream/10665/42274/1/WHO EHC 216.pdf
WHO. 2002. Concise Chemical Assessment Document 40: Formaldehyde. Available online at:
http://www.inchem.org/documents/cicads/cicads/cicad40.htm
WHO. 2003a. Chloroacetones in Drinking Water. Available online at:
http://www.who.int/water sanitation health/water-
qualitv/guidelines/chemi cal s/chloroacetones.pdf?ua=1
WHO. 2003b. Chloropicrin in Drinking Water. Available online at:
http://www.who.int/water sanitation health/water-
q ualitv/gui del i nes/chem i cal s/chl oropi cri n. pdf? ua= 1
WHO. 2004a. Concise Chemical Assessment Document 58: Chloroform. Available online at:
http://vvvvvv.inchem.org/documents/cicads/cicads/cicad58.htm
WHO. 2004b. Brominated Acetic Acids in Drinking Water. Background Document for
Development of WHO Guidelines for Drinking Water Quality. Available online at:
http://vvvvvv.vvho.int/vvater sanitation health/dwq/chemicals/brominatedaceticacids.pdf?ua=l.
WHO. 2004c. Halogenated Acetonitriles in Drinking-water. Available online at:
http://vvvvvv.vvho.int/vvater sanitation health/vvater-
qualitv/guidelines/chemicals/halogenatedacetonit.pdf?ua=l
WHO. 2005a. Bromate in Drinking-water. Background document for development of WHO
Guidelines for Drinking-water Quality. Available online at:
http://vvvvvv.vvho.int/vvater sanitation health/dwq/chemicals/bromate260505.pdf.
WHO. 2005b. Formaldehyde in Drinking Water. Available online at:
http://vvvvvv.vvho.int/vvater sanitation health/vvater-
qualitv/guidelines/chemicals/formaldehvde 130605.pdf?ua= 1.
WHO. 2005c. Chloral Hydrate in Drinking Water. Available online at:
http://vvvvvv.vvho.int/vvater sanitation health/vvater-
ciualitv/guidelines/chemicals/chloralhvdrate 130605.pdf?ua= 1.
WHO. 2005d. Chloral Chemical Assessment. Available online at:
http://www.who.int/ipcs/publications/cicad/en/cicad25.pdf?ua=l.
Six-Year Review 3 Technical Support Document	9-58
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
WHO. 2008. Safety evaluation of certain food additives and contaminants. World Health
Organization Food Additive Series: 59.
WHO. 2009. Cyanogen Chloride Hydrate in Drinking Water. Available online at:
http://www.who.int/water sanitation health/water-qualitv/guidelines/chemicals/cvanogen-
chl ori de-b ackground. pdf?ua= 1
Wright, J.M., J. Schwartz, and D.W. Dockery. 2003. Effect of trihalomethane exposure on fetal
development. Occupational and Environmental Medicine. 60(3): 173-180.
Wright, J.M., J. Schwartz, and D.W. Dockery. 2004. The effect of disinfection by-products and
mutagenic activity on birth weight and gestational duration. Environmental Health Perspectives.
112(8): 920-925.
Writer, J.H., A. Hohner, J. Oropeza, A. Schmidt, K.M. Cawley, and F.L. Rosario-Ortiz. 2014.
Water treatment implications after the high park wildfire, Colorado. Journal of the American
Water Works Association. 106(4): 189-199.
Wu, H., and Y.F. Xie. 2005. Effects of EBCT and water temperature on HAA removal using
BAC. Journal of the American Water Works Association. 97(11): 94-101.
Wu, M., Y. Qian, J.M. Boyd, S. Leavey, S.E. Hrudey, S.W. Krasner, and X-F. Li. 2013.
Identification of tobacco-specific nitrosamines as disinfection byproducts in chloraminated
water. Environmental Science and Technology. 48(3): 1828-1834.
Xiao, Y., L. Zhang, J. Yue, R.D. Webster, and T-T. Lim. 2015. Kinetic modeling and energy
efficiency of UV/H202 treatment of iodinated trihalomethanes. Water Research. 75: 259-269.
Xie, Y.F. and H. Zhou. 2002. Use of BAC for HAA removal: Part 2, column study. Journal of
the American Waterworks Association. 94:5:126.
Xie, Y., H. Wu, and H. Tung. 2004. Haloacetic acid removal using granular activated carbon.
Water Research Foundation. Project #2825.
Xie, S.H., Y.F. Li, Y.F. Tan, D. Zheng, A.L. Liu, H. Xie, and W.Q. Lu. 2011. Urinary
trichloroacetic acid levels and semen quality: A hospital-based cross-sectional study in Wuhan,
China. Environmental Research. Ill: 295-300.
Xie, P., J. Ma, J. Fang, Y. Guan, S. Yue, X. Li, and L. Chen. 2013. Comparison of permanganate
preoxidation and preozonation on algae containing water: cell integrity, characteristics, and
chlorinated disinfection byproduct formation. Environmental Science and Technology. 47(24):
14051-14061.
Xu, X., T.M. Mariano, J.D. Laskin, and C.P. Weisel. 2002. Percutaneous absorption of
trihalomethanes, haloacetic acids, and haloketones. Toxicology and Applied Pharmacology. 184:
19-26.
Six-Year Review 3 Technical Support Document	9-59
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Xu, P., J.E. Drewes, and D. Heil. 2008. Beneficial use of co-produced water through membrane
treatment: technical-economic assessment. Desalination. 225(1): 139-155.
Xu, Z., R. Jiao, H. Liu, D. Wang, C.W.K. Chow, and M. Drikas. 2013. Hybrid treatment process
of using MIEX and high performance composite coagulant for DOM and bromide removal.
Journal of Environmental Engineering. 139(1): 79-85.
Xu, J., Gao, N., Zhao, D., Zhang, W., Xu, Q., and Xiao, A. 2015. Efficient reduction of bromate
in water by nano-iron hydroxide impregnated activated carbon (Fe-GAC). Chemical Engineering
Journal. 275:189-197.
Yamaguchi, T., M. Wei, N. Hagihara, M. Omori, H. Wanibuchi, and S. Fukushima. 2008. Lack
of mutagenic and toxic effects of low dose potassium bromate on kidneys in the Big Blue rat.
Mutation Research. 652: 1-11.
Yang, C.Y. 2004. Drinking water chlorination and adverse birth outcomes in Taiwan.
Toxicology. 198: 249-254.
Yang, X. and C. Shang. 2004. Chlorination byproduct formation in the presence of humic acid,
model nitrogenous organic compounds, ammonia, and bromide. Environmental Science and
Technology. 38(19): 4995-5001.
Yang, M. and X. Zhang. 2014. Halopyrroles: A New Group of Highly Toxic Disinfection
Byproducts Formed in Chlorinated Saline Wastewater. Environmental Science and Technology.
48(20): 11846-11852.
Yang, C.Y., H.F. Chiu, M.F. Cheng, and S.S. Tsai. 1998. Chlorination of drinking water and
cancer mortality in Taiwan. Environmental Research. 78: 1-6.
Yang, C., Z.P. Xiao, S.C. Ho, T.N. Wu, and S.S. Tsai. 2007. Association between trihalomethane
concentrations in drinking water and adverse pregnancy outcome in Taiwan. Environmental
Research. 104: 390-395.
Yang, X., C. Fan, C. Shang, and Q. Zhao. 2010. Nitrogenous disinfection by-products formation
and nitrogen origin exploration during chloramination of nitrogenous organic compounds. Water
Research. 44(9): 2691-2702.
Yang, X., W. Guo, and Q. Shen. 2011. Formation of disinfection by-products from
chlor(am)ination of algal organic matter. Journal of Hazardous Materials. 197: 378-388.
Yang, X., C. Shang, Q. Shen, B. Chen, P. Westerhoff, J. Peng, and W. Gao. 2012. Nitrogen
origins and the role of ozonation in the formation of haloacetonitriles and halonitromethanes in
chlorine water treatment. Environmental Science and Technology. 46(23): 12832-12838.
Yang, Y., Y. Komaki, S. Kimura, H. Hu, E. Wagner, B. Marinas, and M. Plewa. 2014. Toxic
impact of bromide and iodide on drinking water disinfected with chlorine or chloramines.
Environmental Science and Technology. 48: 12362-12369.
Six-Year Review 3 Technical Support Document	9-60
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Ye, T., B. Xu, Y-L. Lin, C-Y. Hu, L. Lin, T-Y. Zhang, andN-Y. Gao. 2013. Formation of
iodinated disinfection by-products during oxidation of iodide-containing waters with chlorine
dioxide. Water Research. 47(9): 3006-3014.
Zamyadi, A., Y. Fan, R.I. Daly, and M. Prevost. 2012. Chlorination of microcystis aeruginosa:
toxin release and oxidation, cellular chlorine demand and disinfection by-product formation.
Water Research. 47(3): 1080-1090.
Zeng, Q., M. Li, S.H. Xie, L.J. Gu, J. Yue, W.C. Cao, D. Zheng, A.L. Liu, Y.F. Li, and W.Q. Lu.
2013. Baseline blood trihalomethanes, semen parameters and serum total testosterone: A cross-
sectional study in China. Environment International. 54: 134-40.
Zeng Q., Y.X. Wang, S.H. Xie, L. Xu, Y.Z. Chen, L. Min, J. Yue, Y.F. Li, A.L. Liu, and W.Q.
Lu. 2014. Drinking water disinfection by-products and semen quality: A cross-sectional study in
China. Environmental Health Perspectives. 122(7): 741-746.
Zha, X.S., Y. Liu, X. Liu, Q. Zhang, R-H. Dai, L-W. Ying, J. Wu, J-T. Wang, and L. Ma. 2014.
Effects of bromide and iodide ions on the formation of disinfection by-products during ozonation
and subsequent chlorination of water containing biological source matters. Environmental
Science Pollution Research. 21(4): 2714-2723.
Zhai, H., X. Zhang, X. Zhu, J. Liu, and M. Ji. 2014. Formation of brominated disinfection
byproducts during chloramination of drinking water: new polar species and overall kinetics.
Environmental Science and Technology. 48(5): 2579-2588.
Zhang, L., W.A. Arnold, and R.M. Hozalski. 2004. Kinetics of haloacetic acid reactions with
Fe(0). Environmental Science and Technology. 38(24): 6881-6889.
Zhang, T., W. Chen, J. Ma, and Z. Qiang. 2008. Minimizing bromate formation with cerium
dioxide during ozonation of bromide-containing water. Water Research. 42(14): 3651-3658.
Zhang, P., T.M. Lapara, E.H. Goslan, Y. Xie, S.A. Parsons, and R.M. Hozalski. 2009a.
Biodegradation of haloacetic acids by bacterial isolates and enrichment cultures from drinking
water systems. Environmental Science and Technology. 43(9): 3169-3175.
Zhang, W., S. Gabos, D. Schopflocher, X.F. Li, W.P. Gati, S.E. Hrudey. 2009b. Reliability of
using urinary and blood trichloroacetic acid as a biomarker of exposure to chlorinated drinking
water disinfection byproducts. Biomarkers. 14(6): 355-65.
Zhang, X., R.J. Bull, J. Fisher, J. Cotruvo, and B. Cummings. 2011. The synergistic effect of
sodium chlorite and bromochloroacetic acid on Br03- induced renal death. Toxicology. 289(2-
3): 151-159.
Zhang, K-J., N-Y. Gao, Y. Deng, T. Zhang, and C. Li. 2012. Aqueous chlorination of algal
odorants: reaction kinetics and formation of disinfection by-products. Separation and Purification
Technology. 92: 93-99.
Six-Year Review 3 Technical Support Document	9-61
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Zhang, Y-Q., Q-P. Wu, J-M. Zhang, and X-H. Yang. 2015. Removal of bromide and bromate
from drinking water using granular activated carbon. Journal of Water and Health. 13(1): 73-78.
Zhou, W.S., L. Xu, S.H. Xie, Y.L. Li, L. Li., D. Zeng, Y.K. Du, and W.Q. Lu. 2012. Decreased
birth weight in relation to maternal urinary trichloroacetic acid levels. Science of the Total
Environment. 416: 105-110.
Zhu, M., N. Gao, W. Chu, S. Zhou, Z. Zhang, Y. Xu, and Q. Dai. 2015. Impact of pre-ozonation
on disinfection by-product formation and speciation from chlor(am)ination of algal organic
matter of Microcystis aeruginosa. Ecotoxicology and Environmental Safety. 120: 256:262.
Six-Year Review 3 Technical Support Document	9-62
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Six-Year Review 3 Technical Support
Document for Disinfectants/Disinfection
Byproducts Rules: Appendices
Six-Year Review 3 Technical Support Document
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
List of Appendices
Appendix A: Additional Information for Health Effects of Regulated Organic Disinfection
Byproducts (DBPs), Regulated Inorganic DBPs and Regulated Disinfectants
(Appendix to Chapter 4)
Appendix B: Additional Information for Occurrence and Exposure to Regulated and Unregulated
Disinfection Byproducts (DBPs) (Appendix to Chapter 6)
Appendix C: Supporting Information for Treatment (Appendix to Chapter 7)
Appendix D: Consideration of Other Regulatory Revisions for MDBP Rules - Additional Issues
(Appendix to Chapter 8)
Appendix E: Additional Information Related to Chlorine Burn Analysis
Six-Year Review 3 Technical Support Document
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Appendix A. Additional Information for Health Effects of Regulated Organic
Disinfection Byproducts (DBPs), Regulated Inorganic DBPs and Regulated
Disinfectants (Appendix to Chapter 4)
Appendix A provides additional information about the health effects of the regulated organic
disinfection by-products (DBPs), regulated inorganic DBPs and the regulated disinfectants. The
information included in Appendix A supplements information provided in Chapter 4 - Health
Effects. To aid in cross-referencing, this appendix uses the same subheading numbering and
titles that appear in Chapter 4.
A.l Regulated Organic DBPs
A. 1.1 Toxicity Studies
A.l.1.1 Trihalomethanes (THMs)
This section presents animal toxicity study information that was available during the
development of Stage 1 and Stage 2 D/DBPRs for bromoform, bromodichloromethane (BDCM),
dibromochloromethane (DBCM) and chloroform. The information includes studies of
carcinogenicity, mutagenicity/genotoxicity and reproductive/developmental effects that were
performed for each of those trihalomethanes. Details of the studies include: nominal dose, route
of exposure, duration of exposure, gender of species and strain of the species.
A. 1.1.1.1 Bromoform
Information Available During Development of Stage 1 and Stage 2 D/DBPRs
Cancer
The National Toxicology Program (NTP) conducted research on the carcinogenicity of
bromoform in 1989. Bromoform was administered by gavage in corn oil to male and female
F344/N rats and to male and female B6C3F1 mice either once (single dose of 2,000 mg/kg), for
14 days (doses ranged from 600-800 mg/kg), 13 weeks (doses ranged from 12-200 mg/kg), or 2
years (doses of 0, 100 or 200 for rats and 0, 50 or 100 for mice) (NTP, 1989a). NTP concluded
that there was clear evidence of carcinogenicity based on tumors in the large intestine (colon or
rectum) in female rats, some evidence in male rats and no evidence in mice.
Mutagenicity/Genotoxicity
Most in vitro studies with bromoform in Salmonella typhimurium were negative, with a few
studies having equivocal results, both with and without metabolic activation. Positive results
were reported for in vitro studies on DNA damage and mixed results for increased sister
chromatid exchange, chromosomal aberrations and DNA strand breaks. Bromoform was tested
in in vivo studies in rats and mice, showing positive results for sister chromatid exchange and
sex-linked recessive lethal mutations, mixed results for chromosomal aberrations and negative
Six-Year Review 3 Technical Support Document	A-l
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
results for DNA strand breaks, micronuclei formation and unscheduled DNA synthesis (NTP
1989a; USEPA, 2005d).
Reproductive/Developmental
The following reproductive and developmental studies on bromoform were reviewed by EPA
and documented in EPA's Drinking Water Criteria Document for Brominated Trihalomethanes
(USEPA, 2005d):
Ruddick et al. (1983) investigated the reproductive and developmental toxicity of bromoform
administered in doses of 50, 100 or 200 mg/kg/day by gavage in corn oil to pregnant Sprague-
Dawley rats on gestational day (GD) 6 through 15. A statistical analysis of the published data
demonstrated a significant increase in sternebral anomalies, resulting in a developmental No-
Ob served-Adverse-Effect-Level (NOAEL) and Lowest-Observed-Adverse-Effect-Level
(LOAEL) of 50 and 100 mg/kg/day, respectively. No maternal effects were observed, resulting
in a NOAEL of 200 mg/kg/day.
NTP administered bromoform at 50, 100 or 200 mg/kg/day by gavage in corn oil to Swiss CD-I
mice using a continuous breeding protocol for 7 days pre-cohabitation and 98 days cohabitation
(NTP, 1989b). The NOAEL and LOAEL for developmental and general toxicity were 100 and
200 mg/kg/day, respectively, based on postnatal survival, liver histopathology and changes in
liver and kidney weights. The maternal NOAEL and LOAEL were 100 and 200 mg/kg/day,
respectively, based on decreased body weights.
A. 1.1.1.2 Bromodichloromethane
Information Available During Development of Stage 1 and Stage 2 D/DBPRs
Cancer
Bromodichloromethane (BDCM) was found to be carcinogenic in rats and mice after
administration by gavage in corn oil (NTP, 1987). Groups of F344N female and male rats were
administered BDCM at 0, 40 or 80 mg/kg/day and groups of 50 male and female B6C3F1 mice
were administered BDCM at 0, 50 or 100 mg/kg/day, both groups received varied doses 5 times
a week for 104 weeks. Tumors were observed in the large intestine in male and female rats, in
the liver in female mice, and in the kidney in male mice and in male and female rats. The slope
factor for BDCM is based on renal tumors in male rats.
The carcinogenicity of BDCM has also been studied when administered in drinking water. In
George et al. (2002), BDCM was administered in drinking water with mean daily doses of 3.9,
20.6 and 36.3 mg/kg/day for male F344/N rats and 8.1, 27.2 and 43.4 mg/kg/day for B6C3F1
male mice. BDCM was found not to be carcinogenic in the male mice; however, it produced an
increased incidence of hepatocellular neoplasms in the male rats. Another study found an
increase in the incidence of liver neoplasms when female Wistar rats were administered BDCM
in drinking water (Tumasonis et al., 1987). Neoplasms were not observed in the kidney or large
intestine in these studies. No neoplastic effects were observed in male mice administered BDCM
in drinking water or in Wistar rats fed microencapsulated BDCM (George et al., 2002; Aida et
al., 1992). The difference in these results may be explained in part by factors such as the stability
Six-Year Review 3 Technical Support Document	A-2
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
of BDCM in feed and water, the influence of a vehicle and different rates of absorption and
metabolism following different vehicles of administration. For instance, when BDCM was
administered by gavage in corn oil at 50 and 100 mg/kg to male F344/N rats, DNA
hypomethylation in the colon was greater and more rapid than when it was administered in
drinking water at concentrations of 350 and 700 mg/L (Pereira et al., 2004a).
Hooth et al. (2002) and McDorman et al. (2003a) administered BDCM in drinking water to male
and female Tsc2 mutant Long-Evans (Eker) rats for 4 or 10 months. This particular strain of rats
is highly susceptible to the effects of renal carcinogens. No increased incidence of tumors was
observed.
Putative pre-neoplastic lesions were observed in the intestine and kidney in rats exposed to
BDCM in drinking water in studies with durations of less than one year: aberrant crypt foci were
observed in the colon of F344 and Eker rats (DeAngelo et al., 2002; McDorman et al., 2003b)
and atypical tubules and hyperplasia were observed in the kidney of Eker rats (McDorman et al.,
2003b).
Mutagenicity/Genotoxicity
In vitro studies have reported mixed results. Mutagenicity studies in Salmonella typhimurium
reported mixed results, both with and without metabolic activation, while tests in mouse
lymphoma cells were positive with metabolic activation. Mixed results were reported on tests for
sister chromatid exchange and chromosomal aberrations, both with and without metabolic
activation. Studies on DNA strand breaks showed mixed results, and primarily positive results
were reported in studies on DNA damage and micronuclei formation. In in vivo studies, results
were positive for sister chromatid exchange and chromosomal aberrations, mixed for micronuclei
formation and negative for DNA strand breaks and unscheduled DNA synthesis (USEPA,
2005d).
Reproductive/Developmental
The following reproductive and developmental studies on BDCM were reviewed in EPA's
Drinking Water Criteria Document for Brominated Trihalomethanes (USEPA, 2005d):
Ruddick et al. (1983) administered BDCM to pregnant Sprague-Dawley rats by gavage in corn
oil at doses of 50, 100 or 200 mg/kg/day on GDs 6 through 15. The maternal NOAEL and
LOAEL were 100 and 200 mg/kg/day, respectively based on significantly decreased maternal
body weight gain. There were no teratogenic effects observed. Sprague-Dawley rats maintained
their litters following BDCM exposure of 100 mg/kg on GDs 6 through 10. The small number of
litters in this study may have limited detection of significant effects at lower doses.
Klinefelter et al. (1995) observed treatment-related effects on sperm characteristics in F344 rats
during a chronic cancer bioassay in which BDCM was administered in drinking water at
approximately 0, 22 or 39 mg/kg/day. Sperm velocities were significantly decreased at 39
mg/kg/day at a 52-week interim sacrifice. No effect on sperm characteristics were observed in
reproductive studies by NTP (1998a) or Christian et al. (2002a) in Sprague-Dawley rats
administered BDCM in drinking water at concentrations similar to or higher than those used by
Klinefelter et al. (1995).
Six-Year Review 3 Technical Support Document	A-3
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Narotsky et al. (1997) administered BDCM to F334 rats by gavage in either corn oil or an
aqueous vehicle containing 10 percent Emulphor® at dose levels of 0, 25, 50 or 75 mg/kg/day on
GDs 6 through 15. Full litter resorptions were observed at doses of 50 and 75 mg/kg only if
exposure occurred on GDs 6 through 10, which is the luteinizing hormone (LH)-dependent
period of pregnancy. The developmental NOAEL and LOAEL were 25 and 50 mg/kg/day,
respectively, based on full litter resorption (NTP, 2006). The BDCM administered in the aqueous
vehicle resulted in significantly reduced maternal body weight gain at the lowest dose tested.
NTP conducted a short-term reproductive and developmental screening test with BDCM
administered in drinking water to Sprague-Dawley rats (NTP, 1998a). The study was designed to
evaluate developmental and female and male reproductive endpoints. NTP concluded that
BDCM was not a reproductive or developmental toxicant in this study at any of the doses tested,
resulting in NOAELs of 68 and 116 mg/kg/day for male and female rats, respectively, for these
endpoints.
Bielmeier et al. (2001, 2004) investigated the mode of action for full litter resorption induced by
BDCM in F344 rats in a number of studies.
•	In a study to investigate strain comparison between F334 rats and Sprague-Dawley rats,
females were dosed with BDCM by aqueous gavage in 10 percent Emulphor® on GDs 6
through 10. The incidence of full litter resorption was 62 percent in the F344 rats and 0
percent in the Sprague-Dawley rats. Surviving litters from both strains appeared normal
with no observed effect on post-natal survival, litter size or pup weight. The authors
identified a LOAEL of 75 mg/kg/day based on full litter resorption in F344 rats. A
NOAEL was not identified.
•	In a study to investigate the critical period for induction of full litter resorption in F344
rats, pregnant rats were dosed with BDCM in 10 percent Emulphor® on GDs 6 through
10 (the LH-dependent period) and GDs 11 through 15 (the LH-independent period, when
pregnancy is maintained by placental lactogens). Full litter resorption occurred in rats
dosed on GDs 6 through 10 but not in rats dosed on GDs 11 through 15.
•	LH and progesterone serum profiles were characterized during a critical period of
gestation during which BDCM was administered. A reduction in serum LH level with a
corresponding reduction in progesterone concentration was observed and the authors
suggest that BDCM alters LH secretion rather than altering luteal responsiveness alone.
However, the significant decrease in serum LH concentration is likely not the sole
determinant of pregnancy loss.
•	The ability of progesterone to prevent BDCM-induced pregnancy loss supports the
conclusion that the mode of action for pregnancy loss due to BDCM is maternally
mediated rather than the result of direct effects on the embryo. The ability of human
chorionic gonadotropin (CG), an LH agonist, to prevent BDCM-induced pregnancy loss
suggests that full litter resorption is mediated, at least in part, by an effect of BDCM on
maternal LH secretion. These results do not rule out a possible effect of BDCM on luteal
responsiveness to progesterone, as previously suggested by Bielmeier et al. (2001).
Six-Year Review 3 Technical Support Document	A-4
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
The Chlorine Chemistry Council (CCC) sponsored a number of reproductive and developmental
studies with BDCM in rats and rabbits (CCC, 2000a, 2000b, 2000c, 2000d).
•	A range-finding study was conducted in male and female Sprague-Dawley rats. BDCM
was administered in drinking water to parental rats from 14 days pre-mating and lasting
until day of sacrifice (Christian et al., 2001a). The NOAEL and LOAEL for pups was
based on decreased pup weight and decreased pup weight gain. Doses could not be
determined due to the effects of reduced water consumption and reduced feed
consumption in the parental generation females.
•	A developmental toxicity study was conducted in which female Sprague-Dawley rats
were exposed to BDCM in drinking water on GDs 6 through 21 (Christian et al., 2001a).
The developmental NOAEL and LOAEL were 45 and 82 mg/kg/day, respectively, based
on a significant number of ossification sites per fetus. The maternal NOAEL and LOAEL
were 18.4 and 45 mg/kg/day, respectively, based on reduced maternal body weight and
body weight gain.
•	A range-finding study was conducted in New Zealand White pregnant rabbits
administered BDCM in drinking water (Christian et al., 2001a). The developmental
NOAEL was approximately 76.3 mg/kg/day, which was the highest dose tested. The
maternal LOAEL was approximately 4.9 mg/kg/day, the lowest dose tested, for reduced
body weight gain.
•	A developmental toxicity study was conducted in New Zealand White pregnant rabbits
administered BDCM in drinking water on GDs 6 through 29 (Christian et al., 2001a). The
developmental NOAEL was 55.3 mg/kg/day, the highest dose tested. The maternal
NOAEL and LOAEL were 13.4 and 35.6 mg/kg/day, respectively, based on decreased
body weight gain.
•	A reproductive study was conducted in Sprague-Dawley rats administered BDCM in
drinking water on GDs 6 through 21 (Christian et al., 2002a). A marginal effect was
observed on estrous cyclicity in F1 females and a small but significant delay in F1
generation sexual maturity. The parental NOAEL and LOAEL values were 4.1-12.6 and
11.6-40.2 mg/kg/day, respectively, based on reduced body weight and body weight gain
in F0 females and F1 males and females. The reproductive NOAEL and LOAEL values
were also 4.1-12.6 and 11.6-40.2 mg/kg/day, respectively, based on delayed sexual
maturation. The study authors have questioned whether delayed sexual maturation in F1
males with reduced body weight should be considered reproductive toxicity or general
toxicity.
An in vitro model in primary cultures of human term placental trophoblasts was used to study the
effect of BDCM on chorionic gonadotropin (CG) secretion (Chen et al., 2003, 2004). BDCM
reduced secretion of immunoreactive and bioactive CG, which suggests that BDCM affects the
placenta and reduces CG production by preventing formation of syncytiotrophoblasts, the major
CG-producing cell type. The authors also showed that BDCM reduced CG secretion by primary
cultures of already-differentiated human syncytiotrophoblasts, suggesting possible effects on
both syncytiotrophoblast formation and on CG production. The authors noted that placental
trophoblasts are the sole source of CG during normal human pregnancy and play and major role
in the maintenance of the fetus.
Six-Year Review 3 Technical Support Document	A-5
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
The potential mode of action related to pregnancy loss following exposure to BDCM has been
discussed in EPA's Drinking Water Criteria Document for Br ominated Trihalomethanes
(USEPA, 2005d). Exposure of F344 rats to BDCM on GDs 8 through 9 were associated with
reduced serum progesterone levels. There was no effect on LH levels. Bielmeier et al. (2001)
suggested that BDCM exposure disrupts corpora lutea responsiveness to LH, which led to
decreased serum progesterone levels and pregnancy loss. The experiments conducted by
Bielmeier et al. (2004) were designed to re-examine maternal LH profiles during exposure to
levels of BDCM known to cause pregnancy loss, using a more sensitive assay for LH than used
by Bielmeier et al. (2001). Bielmeier et al. (2004) then demonstrated that concurrent treatment
with progesterone or with human CG, which is an LH agonist, prevented BDCM-induced
pregnancy loss. Pregnancy loss was attributed to disruption of LH secretion.
Studies suggest that reduced LH secretion (Bielmeier et al., 2002) and reduced luteal
responsiveness to LH (Bielmeier et al., 2003) may both contribute to BDCM-induced full litter
resorption in F344 rats (USEPA, 2006a). However, several investigators have failed to observe
full litter resorption in Sprague-Dawley rats exposed to BDCM, suggesting that these effects may
be strain specific (Ruddick et al., 1983; Bielmeier et al., 2001; Christian et al., 2001a). Christian
and colleagues conducted developmental toxicity studies with BDCM in which pregnant Crl (a
strain) Sprague-Dawley rats and rabbits were allowed to drink BDCM-containing water ad
libitum instead of being exposed via gavage administration. Full litter resorption was not
observed, and there were no adverse effects on embryo-fetal viability at levels up to 900 ppm
(Christian et al., 2001a). The authors suggested that the difference in sensitivity might be due to
the different reproductive performance and endocrine physiology of the species and strains or the
difference in toxicokinetics resulting from the route of exposure.
A. 1.1.1.3 Dibromochloromethane
Information Available During Development of Stage 1 and Stage 2 D/DBPRs
Cancer
DBCM was administered by gavage in corn oil to male and female F344/N rats and to male and
female B6C3F1 mice (NTP, 1985). NTP determined that there was equivocal evidence of
carcinogenicity in male mice and some evidence in female mice based on the incidence of
hepatocellular adenomas (males and females) and the combined incidence of hepatocellular
adenomas and carcinomas (females).
Mutagenicity/Genotoxicity
In vitro mutagenicity studies in S. typhimurium with DBCM showed mixed results, with positive
results primarily reported in studies without metabolic activation. Studies were primarily positive
for sister chromatid exchange and mixed for chromosomal aberrations, DNA strand breaks and
DNA damage. In vivo studies reported positive results for chromosome aberrations, sister
chromatid exchange and DNA damage and negative results for micronuclei formation, DNA
strand breaks and unscheduled DNA synthesis (NTP 1985; USEPA, 2005d).
Six-Year Review 3 Technical Support Document	A-6
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Reproductive/Developmental
Reproductive and developmental studies on DBCM were reviewed in EPA's Drinking Water
Criteria Document for Brominated Trihalomethanes (USEPA, 2005d):
Borzelleca and Carchman (1982) evaluated the toxicity of DBCM administered in drinking water
for seven weeks in a two-generation reproductive study in ICR Swiss mice. A LOAEL of 17
mg/kg/day was identified based on decreased postnatal body weight in the F2B generation but
was assumed to be "marginal" because the effects were noted in only one of the F2 litters, there
were no other adverse effects, and the number of litters and pups examined was unclear.
Ruddick et al. (1983) investigated the reproductive and developmental toxicity of DBCM
administered by gavage in corn oil to pregnant Sprague-Dawley rats on GD 6-15. The NOAEL
for developmental toxicity was 200 mg/kg/day, the highest dose tested. The study was limited by
the small number of litters.
NTP conducted a short-term reproductive and developmental toxicity screen in Sprague-Dawley
rats administered DBCM in drinking water for 35 days (NTP, 1996). No effects were observed
and the NOAELs for males and females were 28.2 and 47.8 mg/kg/day, respectively.
A. 1.1.1.4 Chloroform
Information Available During Development of Stage 1 and Stage 2 D/DBPRs
Cancer
Renal tumors (tubular cell adenoma and carcinoma) were observed in male Osborne-Mendel rats
after a 78-week exposure to 90 mg/kg/day chloroform by gavage in corn oil and hepatocellular
carcinoma was observed in all groups of male B6C3F1 mice exposed to gavage doses >138
mg/kg/day chloroform in oil for 78 weeks (NCI, 1976).
Kidney tumors were observed in ICI (a strain) mice chronically exposed to 60 mg/kg/day
chloroform by gavage, but not in those exposed to 17 mg/kg/day (Roe et al., 1979). Under the
same experimental conditions, chloroform exposure had no effect on the frequency of tumors in
C57BL, CBA and CF-1 mice. Hepatic neoplastic nodules were increased in female Wistar rats
chronically exposed to 200 mg/kg/day chloroform in drinking water (Tumasonis et al., 1987).
Liver tumors in male and female mice and kidney tumors were observed in male and female rats
and mice dosed by gavage in corn oil for 5 days a week for 78 weeks (Dunnick and Melnick,
1993).
In a two-year drinking water study by Jorgenson et al. (1985), chloroform was administered to
male Osborne-Mendel rats and to female B6C3F1 mice at concentrations up to 160 and 263
mg/kg/day in rats and mice, respectively. A significant increase in renal tumors in rats was
observed and was associated with cytotoxicity and regenerative hyperplasia. These
histopathology results support chronic renal tubule injury as the mode of action underlying the
renal tumor response. There were no liver tumors in female mice.
Six-Year Review 3 Technical Support Document	A-7
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Mutagenicity/Genotoxicity
The majority of in vitro mutagenicity studies with chloroform in S. typhimurium and Escherichia
coli, both with and without metabolic activation, were negative. Several studies showed positive
results, but these positive results may be due to the use of cytotoxic concentrations of chloroform
or use of ethanol as a diluent, resulting in formation of ethyl carbonate, an alkylating agent that
can lead to mutations. Studies were negative for in vitro tests of chromosomal aberrations and
unscheduled DNA synthesis and mixed for sister chromatid exchange. In vivo studies of
chromosomal abnormalities, sister chromatid exchange and micronuclei formation showed
mixed results, while studies on DNA damage or repair were negative. The positive results in
some of the in vivo studies could be due to the use of cytotoxic concentrations (USEPA, 2001a;
WHO, 2004a).
EPA concluded that "the weight of evidence indicates that even though a role for mutagenicity
cannot be excluded with certainty, chloroform is not a strong mutagen and that neither
chloroform nor its metabolites readily bind to DNA" (USEPA, 2001a). WHO (2004a) concluded
that the weight of evidence indicates that chloroform does not have significant genotoxic
potential, and the International Life Sciences Institute (ILSI) (1997) also concluded that the
preponderance of evidence indicates that chloroform is not strongly mutagenic.
Reproductive/Developmental
The following reproductive/developmental effects of chloroform in animal studies were
reviewed in the EPA's document Toxicological Review of Chloroform in Support of Summary
Information on the Integrated Risk Information Systems (USEPA, 2001a), the World Health
Organization's (WHO) document Concise Chemical Assessment Document 58: Chloroform
(WHO, 2004a) and Health Canada's document Guidelines for Canadian Drinking Water
Quality: Guideline Technical Document-Trihalomethanes (Health Canada, 2006).
Thompson et al. (1974) conducted a study to evaluate the teratogenicity of chloroform in
Sprague-Dawley rats and Dutch-belted rabbits. Pregnant rats were administered chloroform by
gavage at doses up to 126 mg/kg/day on GDs 6 through 15. Decreased body weight gain and
mild fatty changes in the liver were seen in the dams at 50 mg/kg/day and a significant increase
in the frequency of bilateral extra lumbar ribs and a significant decrease in fetal weight were
reported at 126 mg/kg/day. The maternal NOAEL and LOAEL were 20 mg/kg/day and 50
mg/kg/day, respectively, and the developmental NOAEL and LOAEL were 50 mg/kg/day and
126 mg/kg/day, respectively. Pregnant rabbits were administered doses up to 50 mg/kg/day via
gavage on GDs 6 throughl8, with decreased weight gain reported in the high dose dams and no
evidence of teratogenicity or fetotoxicity. The maternal NOAEL and LOAEL were 35 mg/kg/day
and 50 mg/kg/day, respectively, and the developmental NOAEL was 50 mg/kg/day.
Ruddick et al. (1983) evaluated the reproductive and developmental effects of chloroform
administered by gavage in corn oil to pregnant Sprague-Dawley rats on GDs 6 through 15 and
fetuses were examined for viability and external malformations on GD 22. Decreased body
weight gain and increased liver weights were observed in the dams at 100 mg/kg/day and lower
body weight was observed at 400 mg/kg/day in the fetuses, but no teratogenicity. Maternal
Six-Year Review 3 Technical Support Document	A-8
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
NOAEL and LOAEL values of 200 and 100 mg/kg/day, respectively, and a developmental
NOAEL of 400 mg/kg/day, the highest dose tested, were determined.
NTP conducted a reproductive and fertility study on CD-I (ICR) BR outbred albino mice.
Chloroform was administered via gavage at doses up to 41 mg/kg/day for 18 weeks during the
breeding period and the offspring were administered 41 mg/kg/day through young adulthood
(NTP, 1988). Hepatocellular degeneration in females was the only effect observed in offspring,
with no significant effects on fertility or on reproductive parameters. A NOAEL of 41 mg/kg/day
was identified for fertility (WHO, 2004a). A NOAEL or LOAEL for toxicity (hepatocellular
degeneration) was not determined because histopathology was not performed on the low- and
mid-dose mice (USEPA, 2001a).
A. 1.1.2 Haloacetic acids (HAAs)
A. 1.1.2.1 Monochloroacetic acid
Information Available During Development of Stage 1 and Stage 2 D/DBPRs
Cancer
EPA has not classified monochloroacetic acid (MCAA) for carcinogenicity. MCAA was
administered in water by gavage to F344N rats and B6C3F1 mice in two-year bioassays (NTP,
1992b). There was no evidence of carcinogenic activity in rats or mice. Similarly, MCA was not
carcinogenic in a drinking water study conducted by DeAngelo et al. (1997) in F344 rats.
Mutagenicity/Genotoxicity
Most in vitro mutagenicity studies with MCAA in S. typhimurium, E. coli and cultured
mammalian cells were negative. MCAA showed positive results in the mouse lymphoma assay
and for sister chromatid exchange without metabolic activation and negative results for
chromosomal aberrations and for sex-linked recessive mutations. An in vivo bone marrow assay
reported positive results by intraperitoneal injection and negative results by the oral or
subcutaneous routes (NTP, 1992b).
Reproductive/Developmental
The following studies were reviewed in Health Canada (2008a):
Smith et al. (1990) published an abstract of a developmental study with MCAA administered to
Long-Evans rats by gavage at doses up to 140 mg/kg/day on GD 6-15. Maternal toxicity and
heart malformations in the fetuses were observed in the high dose group, but no statistical data
were provided in the abstract.
Johnson et al. (1998) administered MCAA in drinking water to pregnant Sprague-Dawley rats at
approximately 193 mg/kg/day. No adverse reproductive, developmental or teratogenic effects
were observed; however, complete fetal examinations were not performed.
Six-Year Review 3 Technical Support Document	A-9
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
A. 1.1.2.2 Dichloroacetic acid
Information Available During Development of Stage 1 and Stage 2 D/DBPRs
Cancer
In drinking water studies, dichloroacetic acid (DCAA) caused an increased incidence of hepatic
adenomas and adenocarcinomas in male B6C3F1 mice (Daniel et al., 1992) and in male F344
rats (DeAngelo et al., 1996). A study conducted by DeAngelo et al. (1999) was used by EPA to
quantify cancer risk from ingestion of DCAA in drinking water based on a significant increase in
the incidence of hepatocellular carcinoma in male B6C3F1 mice (USEPA, 2003d). The mode of
action is not clearly understood. Extrapolation to low dose was performed by assuming a no-
threshold linear dose-response curve. EPA considers DCAA to be a likely human carcinogen
based on positive carcinogenic response in the two species of rats and mice and in both sexes and
clear evidence of a dose response (USEPA, 2003d).
Mutagenicity/Genotoxicity
Primarily negative results were reported for DCAA from in vitro mutagenicity studies in
Salmonella typhimurium, while one study of prophage 8 induction in E. coli reported positive
results. Mixed results were reported for in vitro studies on chromosomal aberrations, DNA strand
breaks, DNA repair and the mouse lymphoma mutation assay. Mixed results were also reported
for in vivo studies on micronuclei formation, DNA strand breaks and DNA adduct formation
(USEPA, 2003d).
EPA concluded that DCAA is a weak mutagen and that it induced mutations and chromosome
damage at high concentrations, but that there is uncertainty as to its genotoxicity at lower doses
(USEPA, 2003d). The information on mode of action did not support a nonlinear quantification
of risk; however, the data on mutagenicity suggest that DCAA is not a direct-acting mutagen
(USEPA, 2003d).
WHO concluded that there is some evidence that DCAA is genotoxic at high concentrations but
that these effects are not likely to be involved in the mechanism of DCAA tumorigenesis (WHO,
2000).
Reproductive/Developmental
The following reproductive and developmental studies on DCAA were reviewed by EPA, WHO,
NTP and Health Canada. (WHO, 2000; U.S EPA, 2003d; NTP, 2007a; Health Canada, 2008a):
Katz et al. (1981) investigated the reproductive effects of DCAA in Sprague-Dawley rats (3-
month gavage study) and beagle dogs (13-week capsule study). No adverse effects were noted in
female rats, while in male rats, adverse effects on the testes, including testicular germinal
epithelial degeneration and aspermatogenesis were noted at 500 and 2,000 mg/kg/day, the two
highest doses. Similar testicular effects were noted in the dogs at doses ranging from 50 -100
mg/kg/day.
Six-Year Review 3 Technical Support Document A-10
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Bhat et al. (1991) administered male Sprague-Dawley rats DCAA in drinking water for 90 days.
Decreased testes weight, tissue atrophy, and few spermatocytes and no mature spermatozoa in
the seminiferous tubules were observed in the rats at 1,100 mg/kg/day, the only dose tested.
Cicmanec et al. (1991) investigated the reproductive effects of DCAA in beagle dogs,
administering doses up to 72 mg/kg/day in gelatin capsules for 90 days. Testicular changes were
reported in the males at all doses, including syncytial giant cell formation and degeneration of
testicular germinal epithelium. Prostate glandular atrophy was also noted at 39.5 and 72
mg/kg/day. The reproductive LOAEL was 12.5 mg/kg/day, the lowest dose tested, based on
testicular changes, and this LOAEL was used as the point-of-departure for determining the RfD
of 0.004 mg/kg/day (USEPA, 2003d).
Toth et al. (1992) administered DCAA for 10 weeks by gavage to Long-Evans rats at doses up to
125 mg/kg/day. Significant reductions in the absolute weight of the preputial gland and
epididymis were noted at all dose levels and there were effects on sperm morphology and
decreased sperm counts at the higher doses. A LOAEL of 31.25 mg/kg/day, the lowest dose
tested, was identified based on the organ weight changes in the preputial gland and epididymis.
Epstein et al. (1992) conducted a series of developmental studies in pregnant Long-Evans rats
administered DCAA by gavage at doses up to 3,500 mg/kg/day for various time periods between
GDs 6-15. Reduced mean fetal body weight and increased cardiac malformations were observed
at 1,900 mg/kg/day and an increased incidence of cardiac defects were observed at 2,400 and
3,500 mg/kg/day. A developmental LOAEL of 1,900 mg/kg/day, the lowest dose tested, was
determined.
Smith et al. (1992) administered DCAA to pregnant Long-Evans rats by gavage on GD 6 -15. A
significant decrease in maternal weight gain and hypertrophy in the liver, spleen and kidneys
were reported at 140 and 400 mg/kg/day. An increase in cardiac abnormalities, reduced fetal
crown-rump length and reduced fetal body weight were observed at 400 mg/kg/day. A
statistically significant increase in soft tissue anomalies was observed at 140 and 400 mg/kg/day.
Maternal and developmental NOAELs of 14 mg/kg/day and LOAELs of 140 mg/kg/day were
determined.
Linder et al. (1997) administered DCAA orally to male Sprague-Dawley rats for up to 14 days at
doses up to 1,440 mg/kg/day to evaluate testicular toxicity. On day 14, a significant decrease in
epididymal weight was observed at 480 and 1,440 mg/kg/day, and epididymal sperm count was
decreased at> 160 mg/kg/day.
Fisher et al. (2001) administered DCAA by gavage at 300 mg/kg/day to pregnant Sprague-
Dawley rats on GD 6-15. No malformations of the heart were reported in the offspring.
Six-Year Review 3 Technical Support Document A-ll
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
A. 1.1.2.3 Trichloroacetic acid
Information Available During Development of Stage 1 and Stage 2 D/DBPRs
Cancer
TCAA increased incidences of liver tumors in male B6C3F1 mice exposed via drinking water
(Bull, 2002; Bull et al., 1990) and tumors were also observed in less-than-lifetime studies in
female B6C3F1 mice (Pereira, 1996). However, no treatment-related tumors were observed in
male F344/N rats (DeAngelo et al., 1997).
Mutagenicity/Genotoxicity
Most in vitro studies with TCAA in S. typhimurium and in mammalian systems resulted in
negative results for mutagenicity, with positive results in one study investigating SOS DNA
repair in S. typhimurium and equivocal results in a study on mouse lymphoma cells. In vivo
studies have shown mixed results. Positive results were reported for chromosomal aberrations
and mixed results for micronuclei formation and DNA strand breaks (Health Canada 2008a;
IARC 2014).
Reproductive/Developmental
The following reproductive and developmental studies on TCAA were reviewed in Health
Canada's document Guidelines for Canadian Drinking Water Quality: Guideline Technical
Document-Haloacetic Acids (Health Canada, 2008b) and in EPA's document Toxicological
Review of Trichloroacetic acid in Support of Summary Information on the Integrated Risk
Information System (USEPA, 201 la):
Smith et al. (1989b) conducted a developmental study in Long-Evans rats administered TCAA
by gavage at doses up to 1,800 mg/kg/day on GD 6-15. Maternal toxicity was observed at all
doses based on significant increases in spleen and kidney weights and developmental toxicity
was also observed at all doses based on significant decreases in mean fetal weight, significant
decreases in fetal crown-rump length and increases in the frequency of cardiac malformations.
Maternal and developmental LOAELs of 330 mg/kg/day were determined, which was the lowest
dose.
Johnson et al. (1998) administered TCAA in drinking water at 290 mg/kg/day to pregnant
Sprague-Dawley rats. A significant decrease in body weight gain of the dams and a significant
increase in the number of resorptions, number of implantation sites, and cardiac soft tissue
malformations were observed at this dose. The maternal and developmental LOAELs were 290
mg/kg/day.
Fisher et al. (2001), administered 300 mg/kg/day of TCAA via gavage to pregnant Sprague-
Dawley rats on GD 6-15. A significant reduction in maternal body weight and fetal body weight,
but no increase in cardiac malformations were reported. Maternal and developmental LOAELs
of 300 mg/kg/day were determined.
Six-Year Review 3 Technical Support Document A-12
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
A. 1.1.2.4 Monobromoacetic acid
Information Available During Development of Stage 1 and Stage 2 D/DBPRs
Reproductive/Developmental
Health Canada (2008a) and WHO (2004b) reviewed the following reproductive/developmental
study on MBAA:
Linder et al. (1994a) administered MBAA as a single gavage dose or daily gavage doses for 14
days to male Sprague-Dawley rats and no reproductive effects were observed.
A. 1.1.2.5 Dibromoacetic acid
Information Available During Development of Stage 1 and Stage 2 D/DBPRs
Mutagenicity/Genotoxicity
The results of mutagenicity and genotoxicity assays with dibromoacetic acid (DBAA) published
prior to the Stage 1 and Stage 2 D/DBPR were reviewed by WHO (2004b) and Health Canada
(2008a). Mixed results were reported from in vitro studies with dibromoacetic acid in S.
typhimurium fSaito et al., 1995; Giller et al., 1997; Morita et al., 1997; Kohan et al., 1998;
Kargalioglu et al., 2002), while positive results were reported for DNA repair in the SOS
chromotest and for DNA strand breaks, as measured in a comet assay (Giller et al., 1997; Plewa
et al., 2002). Negative results were reported for micronuclei formation in a newt micronucleus
test (Giller et al., 1997).
Reproductive/Developmental
The following reproductive and developmental studies on DBAA were reviewed in Health
Canada's Guidelines for Canadian Drinking Water Quality: Guideline Technical Document-
Haloacetic Acids (Health Canada, 2008a) and WHO's Brominated Acetic Acids in Drinking
Water (WHO, 2004b):
Linder et al. (1994a) administered single gavage doses of DBAA to male Sprague-Dawley rats
and noted adverse effects on sperm count, morphology and motility.
Linder et al. (1994b) noted a number of adverse reproductive effects in a 14-day gavage study in
male Sprague-Dawley rats administered DBAA at 0, 10, 30, 90 or 270 mg/kg/day. At the high
dose, reduced testis and epididymis weights were observed. Decreased sperm counts and
histopathological evidence of altered spermiation were observed at all doses. The LOAEL for the
14-day study was 10 mg/kg/day.
Linder et al. (1995), administered DBAA by gavage for 42 days to male Sprague-Dawley rats
and paired them with unexposed females on various treatment days and during recovery.
Treatment was stopped on day 42 due to adverse toxic effects, including labored breathing,
tremor, difficulty moving hind limbs and severe weight loss. Fertility was significantly decreased
in a time-dependent manner during treatment and until pairing on days 199-213 after treatment.
Six-Year Review 3 Technical Support Document A-13
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Sperm motility was significantly reduced after 16 or more days of treatment. No developmental
toxicity was observed in the fetuses conceived during treatment. Artificial insemination of
unexposed females with sperm from males treated for 16 or 31 days, but not for 9 days, resulted
in an effect on the reproductive competence of the females.
Linder et al. (1995) conducted a second reproductive/developmental study with Sprague-Dawley
rats using the same protocol as described above. The only significant reproductive effect at 10
mg/kg/day was a decrease in copulating pairs during study days 65-71. There was also a dose-
dependent decrease in the number of males siring two litters during the final mating period,
which was significant at 50 mg/kg/day. No significant reproductive effects were noted in
unexposed females artificially inseminated with sperm from treated males. Histopathology
results showed delayed or altered spermiation at 10 mg/kg/day and above. The NOAEL was 2
mg/kg/day.
Narotsky et al. (1996, 1997) administered DBAA by oral gavage in two developmental screening
assays in CD-I mice. Developmental effects were noted in the presence and absence of maternal
toxicity but no statistical data were provided and the final studies were not published.
Vetter et al. (1998) administered single gavage doses of DBAA to male Crl:CD (Sprague-
Dawley) BR rats. No effects on the sperm were observed, but there was mild histopathology in
the testes.
Cummings and Hedge (1998) administered DBAA by gavage to female Holtzman rats on GD 1-
8 and sacrificed on GD 9 or 20. The only effect was an increase of 170 percent in serum 17P-
estradiol, resulting in a NOAEL of 125 mg/kg/day and a LOAEL of 250 mg/kg/day.
Balchak et al. (2000) administered DBAA in drinking water for 14 days to female Sprague-
Dawley rats. Dose-related alterations of the estrous cycle were observed at doses of 90 and 270
mg/kg/day, but not at lower doses.
Christian et al. (2001b) administered DBAA in drinking water to male and female Crl:CD
Sprague-Dawley rats starting 14 days prior to cohabitation and continuing through gestation and
lactation (63-70 days of treatment). A slight but not significant reduction in mating performance
at the highest dose was the only reproductive effect noted. The parental NOAEL was 66
mg/kg/day for males and 60 mg/kg/day for females, the highest dose tested. A NOAEL for
developmental toxicity could not be determined.
Christian et al. (2002b) conducted a two-generation study in Sprague-Dawley rats administered
DBAA in drinking water. Reproductive performance and development of female rats was not
affected. Histopathology of the reproductive organs in the parental and F1 male pups revealed
altered sperm production and some epididymal tubule changes in the mid- (22.4-55.6 mg/kg/day)
and high-dose (52.4-132.0 mg/kg/day) rats and small or absent epidymides and small testes in
the F1 high dose-males. The authors identified a parental NOAEL of 4.4-11.6 mg/kg/day based
on increased liver and kidney weights.
Murr and Goodman (2005) did not observe changes in estrous cycle during a 20-week exposure
at low doses of DBAA administered in drinking water to female Sprague-Dawley rats, although
circulating serum estradiol levels were increased at weeks 3 and 11. The authors concluded that
Six-Year Review 3 Technical Support Document A-14
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
these increases in estradiol were not linked to disruption of estrous cyclicity in Sprague-Dawley
rats, which are a moderately estrogen-sensitive strain.
A.1.2 Epidemiology and Weight of Evidence
A.l.2.1 Cancer
A. 1.2.1.1 Bladder Cancer
Case-Control Studies
Bove et al. (2007b) present results of a case-control study set in western New York State (182
male bladder cancer cases and 385 male controls enrolled between the years 1978-1986) in
which estimates of total and specific THMs in tap water were combined with subjects'
residential history and estimates of long-term tap water consumption. The highest quartile of
estimated THM4 consumption (74.10-351.73 |ig/day) was associated with an increased risk of
bladder cancer, relative to the lowest quartile of THM4 consumption (0-38.04 |ig/day; odds ratio
(OR) = 2.34; 95 percent confidence interval (% CI): 1.01, 3.66). Estimates of the relative odds of
bladder cancer comparing the highest to lowest quartiles of estimated exposure to THM4 and
separately to each of the four specific THMs (bromodichloromethane, bromoform,
dibromochloromethane and chloroform) were elevated and statistically significant for all but one
of the specific THMs (dibromochloromethane). In these comparisons, the odds ratio for
bromoform was the largest in magnitude (OR = 3.05; 95% CI: 1.51, 5.69). Bladder cancer risk
was highest for those who consumed the greatest amount of water at points within the
distribution system with the oldest post-disinfected tap water (the water that had been in the
distribution system the longest following disinfection). Subjects consuming an average of 10
cups per day of water with mean water age of 188 hours post-disinfection had a greater than 5-
fold increase in the odds of bladder cancer, relative to those consuming 5 cups per day of water
with a mean water age of 13 hours post-disinfection (OR = 5.85; 95% CI: 1.93, 17.46). The
study results also suggested an exposure-response relationship between higher risk of bladder
cancer with increasing levels of THM4, bromodichloromethane and bromoform exposures.
This study adds to the weight of evidence showing a relationship between, long-term average
THM4 exposure, long-term average specific THM exposure and bladder cancer risk. The
estimated ORs may be subject to bias due to inappropriate control subject selection, exposure
measurement error and residual confounding. Exposure measurement errors, assuming they are
non-differential with respect to case status, would typically attenuate OR estimates. The study
does not address associations among sensitive populations, other than gender, and does not
assess genetic factors that may influence risk of bladder cancer associated with THM exposure.
The study results are comparable with previous studies evaluating THM exposure in men and
add to the evidence supporting the hypothesis that there is a positive association between THM
exposure and bladder cancer risk among men. The study results may not be generalizable to
populations consisting of both men and women. A strength of the study is the use of geocoding
to increase specificity and interpolation approaches to better characterize DBP formation
variability.
Six-Year Review 3 Technical Support Document A-15
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Villanueva et al. (2007) present results of a case-control study set in Spain (1,219 bladder cancer
cases and 1,271 controls enrolled between the years 1998-2001) in which lifetime personal
information on water consumption and other uses were collected. Long-term THM4 exposure
from all exposure routes was associated with a two-fold increase in odds of bladder cancer
incidence (OR = 2.1; 95% CI: 1.09, 4.02) comparing those in the highest quartile of average
household THM4 level (>49 |ig/L) to those in the lowest THM4 quartile (<8 |ig/L), with a
statistically significant positive trend observed in the odds of bladder cancer for increasing
quartiles of average residential THM4 level (ptrend<0.01). Study subjects with estimated exposure
of >35 |ig/day via ingestion had 1.35 times the odds of bladder cancer, relative to subjects that
did not drink chlorinated water (95% CI: 0.92, 1.99); this association was higher in magnitude
and statistically significant among men (OR = 1.61; 95% CI: 1.06, 2.44), but was in the opposite
direction for women (OR = 0.47; 95% CI: 0.15, 1.51).
Those in the highest quartile for showering and bathing duration had an OR for bladder cancer of
1.83 (95%) CI: 1.17, 2.87) relative to the lowest quartile, with a statistically significant positive
trend observed (pirend<0.01). Those who had "ever" swum in pools had 1.57 times the odds of
bladder cancer, relative to those that had never swum in pools (95% CI: 1.18, 2.09), with a
statistically significant trend observed with increasing the number of lifetime hours spent
swimming (ptrend = 0.02). The results are suggestive of an increased bladder cancer risk not only
with increasing exposure to THM4 via ingestion, but also due to exposure via other routes
(dermal absorption and inhalation).
The authors also examined exfoliated cells in urine for micronuclei frequency in a subset of 92
female controls, for which there were 72 with adequate samples and 44 with complete THM4
exposure data. The authors reported that women exposed to THM4 levels above the median of
26 |ig/L had a 70 percent increased probability of having a frequency of micronuclei above the
median frequency of 9/1000 compared with those exposed to THM4 below the median
concentration. The authors also noted that they observed higher associations for THM4
exposures through showering and bathing.
The study findings suggest that from the consideration of showering, bathing and swimming pool
use that dermal (and perhaps also inhalation) routes of exposure, which were not explicitly
considered by EPA in the attributable risk calculations supporting the Stage 2 rule, may
contribute to the overall risk and that these routes may lead to a higher concentration directly in
target organs than ingestion. The observation of positive associations between bladder cancers
and municipal drinking water exposure/THM4 exposure only among men is consistent with
similar gender-stratified results in other studies. The study does not address associations among
sensitive populations, other than differences in gender, and does not assess genetic factors that
may influence risk of bladder cancer associated with THM4 exposure.
The authors indicated that the positive association between micronuclei frequency and THM4
levels provides evidence of an intermediate marker of biological effect for THM4 exposure.
They noted, however, that these results were limited by the small sample size. The study results
are comparable with previous studies evaluating THM4 exposure in individuals and add to the
evidence supporting the hypothesis that there is a positive association between THM4 exposure
and bladder cancer risk.
Six-Year Review 3 Technical Support Document A-16
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Michaud et al. (2007) presented results of a case-control study set in Spain (397 bladder cancer
cases and 1,271 controls enrolled between the years 1998-2001) in which total fluid intake
(including coffee, beer, wine, liquor, champagne, soda, juices, tea, milk and water), water intake,
lifetime THM4 exposure and interaction between the different exposure metrics were assessed
with respect to risk of bladder cancer. Total fluid intake was associated with a decrease in the
relative odds of bladder cancer comparing highest to lowest quintile of fluid intake (OR = 0.62;
95% CI: 0.40, 0.95). A statistically significant inverse association was observed for water intake,
specifically, comparing highest to lowest quintile (OR = 0.47; 95% CI: 0.33, 0.66), with a
negative trend observed (ptrend < 0.0001), but not for consumption of other individual beverages
or total fluid intake not including water, but controlling for water intake in the model. The
inverse association between water intake and bladder cancer risk persisted within each level of
THM4 exposure; no statistical interaction was observed between water intake and THM4
exposure (pinteraction = 0.13). The results suggest that water intake is inversely associated with
bladder cancer risk.
The study results provide some evidence that water intake is inversely associated with risk of
bladder cancer. The study findings do not provide additional insights into the specific level of
bladder cancer risks associated with DBPs, although the study assessed and did not observe a
statistically significant interaction between THM4 and water intake. The finding of an inverse
relationship between water intake and bladder cancer risk is consistent with findings in a
prospective cohort study by the same lead author assessing bladder cancer in men (Michaud et
al., 1999) and has implications for negative confounding of THM4-bladder cancer associations in
studies that do not adjust for total water intake. The authors note a potential biological
mechanism underlying the observed inverse association between water intake and bladder
cancer: fluids flush out carcinogens or reduce their contact time with the urothelium. However,
the absence of similar inverse associations between bladder cancer risk and total fluid intake not
including water as observed in Michaud et al. (1999) prospective study noted above, and intake
of total fluids including water in this study is perplexing. The authors speculate that, in this study
population, water consumption provides a better estimate of long-term intake, as consumption is
more consistent over time, relative to consumption of other beverages.
The study results are comparable with previous studies evaluating bladder cancer risk with
respect to fluid intake and THM4 exposure in individuals and add to the evidence supporting the
hypothesis that there is a positive association between THM4 exposure and bladder cancer risk
and a negative association between water intake and bladder cancer risk.
Cantor et al. (2010) present results of a case-control study set in Spain (680 bladder cancer cases
and 714 controls enrolled between the years 1998-2001) in which lifetime personal information
on water consumption and water-related habits were collected. As noted earlier, this is a subset
of a large case-control study population used in several other studies summarized here. The
objective of this study was to investigate gene-environment interactions, including both
individual and certain combined influences of DBP exposure and polymorphisms in glutathione
S-transferase genes (specifically, GSTT1, GSTM1 and GSTZ1), cytochrome P450 genes
(CYP2E1) and N-acetyltransferase 2 (NAT2) on bladder cancer risk. THM4 exposure was
positively associated with bladder cancer, as previously reported in this study population.
Estimated ORs and 95% CIs for bladder cancer were 1.2 (95% CI: 0.8, 1.8), 1.8 (95% CI: 1.1,
2.9) and 1.8 (95% CI: 0.9, 3.5) for THM4 quartiles 2, 3 and 4, respectively, relative to quartile 1.
Six-Year Review 3 Technical Support Document A-17
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Statistically significant associations were also observed between bladder cancer and NAT2 slow
acetylator compared with rapid/intermediate genotypes (OR = 1.33; 95% CI: 1.06, 1.68) and
GSTM1 null versus GSTM1 present genotypes (OR = 1.8; 95% CI: 1.4, 2.2), similar to
associations reported previously in this study population, using different metrics (i.e., Villanueva
et al., 2007; Michaud et al., 2007; Costet et al., 2011). Statistically non-significant elevated risks
were observed for GSTT1 presence compared with null (OR = 1.21; 95% CI: 0.92, 1.59).
Statistically non-significant elevations or decreases in relative odds for three functional single-
nucleotide polymorphisms of GSTZ1 and three variants of CYP2E1 were also observed.
Associations between THM4 and bladder cancer were stronger among study subjects who were
GSTT1 +/+ or +/- relative to those who were GSTT1 null (pinteraction = 0.021), GSTZ1 rsl046428
CT/TT versus CC (pinteraction = 0.018), or CYP2E1 rs2031920 CC versus CT/TT (pinteraction =
0.035). Among a subset of subjects with forms of GSTT1 and GSTZ1 considered to increase the
risk of bladder cancer, the ORs for quartiles 2, 3 and 4 of THM4 were 1.5 (95% CI: 0.7, 3.5), 3.4
(95% CI: 1.4, 8.2) and 5.9 (95% CI: 1.8, 19.0), respectively. The authors concluded that
polymorphisms in key metabolizing enzymes modified DBP-associated bladder cancer risk,
noting that the consistency of the findings with experimental observations of GSTT1, GSTZ1
and CYP2E1 activity strengthens the hypothesis that DBPs cause bladder cancer, and points to a
potential mechanism of action, as well as classes of compounds likely to be implicated in
increasing bladder cancer risk. The study findings provide additional insights into the level of
bladder cancer risks associated with DBPs, particularly as they interact with specific genetic
polymorphisms. The study also identifies potentially sensitive populations based on genetic
polymorphisms of genes coding for key THM-metabolizing enzymes.
The study results are comparable with, and expand on, previous studies evaluating THM4
exposure in individuals and add to the evidence supporting the hypothesis that there is a positive
association between THM4 exposure and bladder cancer risk, that the mechanism of action
involves key metabolizing enzymes and that in the genes coding for these enzymes
polymorphisms appear to modify DBP-associated bladder cancer risk. The study results are also
consistent with experimental observations of GSTT1, GSTZ1 and CYP2E1 activity.1
Note that the two key polymorphisms, GSTT1(+) and GSTZ1 CT/TT, may be present in
approximately 80 percent and 30 percent of the U.S. population, respectively (Regli et al., 2015).
Moreover, about 24 percent of the population may have both polymorphisms present which
Cantor et al. (2010) found to have a highly significant increase in the odds ratios for bladder
cancer with increasing average THM4 levels in water. The Cantor et al. (2010) study results
strengthen the evidence for the hypothesis that THMs cause bladder cancer, particularly in
genetically susceptible individuals.
Chang et al. (2007) conducted a matched case-control study using data from the Taiwan
Provincial Department of Health to investigate the association between bladder cancer and
THM4 exposure in 65 municipalities in Taiwan. The case group consisted of 280 men and 123
women whose death certificate indicated bladder cancer as the cause of death. Controls were
1 See, for example, Pegram et al. 1997, Glutathione .S'-transferasc-mcdiatcd mutagenicity of trihalomethanes in
Salmonella typhimurium: contrasting results withbromodichloromethane and chloroform, Toxicol. Appl. Pharmacol
144(1): 183-188.
Six-Year Review 3 Technical Support Document A-18
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
matched to cases on gender, year of birth and year of death. Average THM4 tap water levels
were estimated using monitoring data collected by the Taiwan EPA over a two-year period and
categorized by quantile. The authors observed an elevated odds of bladder cancer for those in the
75th-90th percentile THM group (OR = 1.80; 95% CI: 1.18, 2.74) and for those in the >90th
percentile THM group (OR = 2.11; 95% CI: 1.43, 3.11), relative to those with estimated THM4
levels below the 75th percentile. The trend of increasing relative odds with increasing average
level of THM4 was statistically significant (pirend<0.001).
Although misclassification of actual THM4 exposure is likely in this study, quantiles of exposure
may be less subject to misclassification since they can help provide a general relative ranking of
exposures. The results of this study are consistent with previous studies evaluating bladder
cancer risk with respect to THM4 exposure from drinking water. Failure to adjust for smoking
leaves open the possibility that most of the excess risk of bladder cancer-specific mortality
associated with THM4 in this study may be due to smoking, under the assumption that THM4
exposure is positively associated with smoking. The magnitude of confounding by smoking is
limited by the strength of this association (if any) between smoking and THM4 exposure. This
study provides evidence in support of the hypothesis that there is a positive association between
THM4 exposure through drinking water and risk of death from bladder cancer.
Pooled Data andMeta-Analysis Studies
Villanueva et al. (2006) pooled data from six case-control studies conducted in Europe (Italy,
Finland and France) and North America (the United States and Canada) between the years 1978
and 2000, representing 2,729 cases and 5,150 controls, to evaluate whether total fluid intake of
specific fluids is associated with bladder cancer risk. In this pooled analysis, total fluid intake
was associated with a slight increased risk of bladder cancer in men (OR for a 1 liter/day
increase in total fluid intake, OR = 1.08; 95% CI: 1.03, 1.14) with a statistically significant linear
trend observed between increasing total fluid consumption and bladder cancer risk (ptrend<0.001),
but not among women (OR = 1.04; 95% CI: 0.94, 1.15; ptrend = 0.70). Men in the highest
category of total fluid intake (>3.5 liters/day) had 1.33 times the odds of bladder cancer (95% CI:
1.12, 1.58), compared to men in the lowest (< 2 liters/day). An increased risk was associated
with intake of tap water (OR = 1.46; 95% CI: 1.20, 1.78), comparing those with tap water intake
>2 liters/day to those with < 0.5 liters/day (ptrend<0.001). Again, this association was statistically
significant among men (OR = 1.50; 95% CI: 1.21, 1.88), but lower in magnitude and not
statistically significant among women (OR = 1.19; 95% CI: 0.78, 1.81). No statistically
significant associations were observed between a 1 liter/day increase in non-tap water intake and
bladder cancer in men (OR = 1.03; 95% CI: 0.95, 1.12) or women (OR = 1.03; 95% CI: 0.85,
1.24). Statistically significant associations were not observed for the continuous measure of
coffee consumption, but were observed for heavy coffee drinkers (>5 cups of coffee/day),
relative to those who reported drinking fewer than 5 cups of coffee per day (OR = 1.26; 95% CI:
1.10, 1.44), an association which was again observed to be stronger among men than among
women. As previously reported (Villanueva et al., 2004), THM4 exposure was also associated
with bladder cancer risk in this pooled analysis. The authors reported that neither coffee
exposure nor THM4 exposure confounded or modified the association between tap water intake
and bladder cancer risk. The results are suggestive of an association between long-term THM4
exposure and bladder cancer risk among men, but not women. The authors concluded that the
association of bladder cancer with tap water consumption, but not with non-tap water fluids,
Six-Year Review 3 Technical Support Document A-19
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
suggests that carcinogenic chemicals in tap water may explain the increased risk of bladder
cancer.
The results of this study provide evidence for the hypothesis that tap-water-based fluids may
increase the risk of bladder cancer among men. Statistically significant associations were not
observed between consumption of non-tap water and bladder cancer risk in this study. The
strength of this evidence is limited by possible selection bias introduced by selection of
unrepresentative control populations in the studies contributing to the pooled analysis and by
non-random patterns of missing exposure data. Information bias and failure to condition on
matching factors could have biased the observed ORs. It is unlikely that information bias,
selection bias or confounding would result in a null result for women, but not in men.
The authors present but do not discuss in detail the observed association between long-term
THM4 exposure and bladder cancer risk, as this information has been previously published. The
new studies add to the evidence of increasing bladder cancer risk with increasing levels of tap
water consumption, although the association was only observed among men but acknowledges
that the results are inconsistent with the findings from other studies. The study also supports the
hypothesis that there is a positive association between THM4 exposure and bladder cancer risk,
although this was not the focus of the current study.
Costet et al. (2011) conducted pooled and meta-analytic analyses of data from three case-control
studies set in three European countries (Finland, France and Spain), using data collected on a
total of 2,381 cases and 3,086 controls to characterize the relationship between long-term
exposure to drinking water DBPs and bladder cancer risk. The authors additionally report on a
meta-regression incorporating data from European and North American studies examining the
association between THM4 exposure and bladder cancer risk. Using meta-analytic techniques,
long-term THM4 exposure to concentrations of >50 |ig/L among men was found to be associated
with a 47 percent increase in the odds of bladder cancer, compared to THM4 exposure to
concentrations < 5 |ig/L (OR = 1.47; 95% CI: 1.05, 2.05). A linear positive trend of increasing
bladder cancer risk with increasing THM4 levels was also observed among men (ptrend = 0.01).
However, among women, long-term THM4 exposure of >50 |ig/L was associated with a
decrease in the odds of bladder cancer, compared to those with THM4 exposure < 5 |ig/L (OR =
0.70; 95%CI: 0.43, 1.14) and a linear trend was not evident (ptend = 0.27). The authors did not
further discuss the observed findings for women; note that women constituted only a small
fraction (-19 percent) of the subjects in these European studies and an even smaller fraction of
the cases (16 percent).
The results of studies on the association of bladder cancer with THM4 exposure measures were
not among the one European study and the combined European and North American studies. The
study provides evidence that estimates of the association between THM4 exposure and bladder
cancer risk observed in one European country are generalizable to other European countries, and
further, that associations observed in Europe are generalizable to the United States, and vice
versa.
The Costet et al. (2011) study results provide evidence that long-term exposure to THM4 is
associated with increased bladder cancer risk among men. Information bias and failure to
condition on matching factors could have led to biased and imprecise results. It is not possible to
Six-Year Review 3 Technical Support Document A-20
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
predict the magnitude and direction of selection bias and net bias due to confounding factors in
this study. However, the consistency of the results across studies that used different control
selection procedures and procedures to adjust for confounding factors is a strength of this work.
Ecological Study
Llopis-Gonzalez et al. (2011) conducted an ecologic study leveraging aggregate district-level
THM4 measurements and bladder cancer mortality data collected for 10 districts in Valencia,
Spain to investigate the association between THM4 exposure and bladder cancer (along with
digestive cancers). The authors reported a positive, non-monotonic relationship between ecologic
measures of THM4 and age-standardized bladder cancer mortality among women, with the
bladder cancer-specific mortality rate increasing with THM4 levels above 40 |ig/L. No
association between the district-level THM4 measurements and bladder cancer mortality was
observed for men. The authors conclude that their results suggest a possible association between
bladder cancer mortality in women and THM4 exposure at levels below the European
Community legal limit of 150 |ig/L during the eight-year study period of 2000 to 2007 (the limit
was reduced to 100 |ig/L in 2009).
This study provides weak evidence of an association between THM4 exposure and bladder
cancer mortality among women, but not men. As an ecological study, this study has considerable
limitations, including susceptibility to measurement error and potential bias due to confounding
factors. The study does not provide additional insights into the specific level of bladder cancer
risks associated with DBPs, nor to effect modification of DBP exposure by potential genetic and
other susceptibility factors. The study does little to strengthen (or weaken) the evidence for the
hypothesis that DBPs cause bladder cancer.
A. 1.2.1.2 Colon/Rectal Cancer
Case-Control Studies
Bove et al. (2007a) present results of a case-control study set in a county in western New York
State (128 white male rectal cancer cases and 253 white male controls enrolled between the years
1978-1986) in which estimates of THM4 and the four specific THMs in tap water were
combined with subjects' residential history and estimates of long-term tap water consumption to
estimate long-term average THM consumption. The primary objective of the study was to
evaluate the association between THM consumption and rectal cancer risk. The authors observed
a statistically significant positive association between rectal cancer and average bromoform
consumption ([j,g/day) (OR = 1.85; 95% CI: 1.25, 2.74). Those in the highest quartile of
estimated bromoform consumption (1.69-15.43 |ig/day) had 2.32 times the odds of rectal cancer,
relative to those in the lowest bromoform consumption quartile (0.09-0.64 (j,g/day) (95% CI:
1.22, 4.39; ptrend = 0.002). Positive associations between rectal cancer and average
dibromochloromethane consumption ([j,g/day) (OR = 1.78; 95% CI: 1.00, 3.19) and
bromodichloromethane consumption ([j,g/day) (OR = 1.15; 95% CI: 1.00, 1.32) were of
borderline statistical significance. No associations were observed between rectal cancer and
long-term average chloroform consumption ([j.g/day) (OR = 1.00; 95% CI: 0.98, 1.02) or THM4
consumption ([j,g/day) (OR= 1.01; 95% CI: 0.99, 1.03).
Six-Year Review 3 Technical Support Document A-21
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
The study does not provide evidence that long-term consumption of THM4 is associated with
rectal cancer risk. However, the study does provide some evidence supporting the hypothesis that
consumption of specific THMs, particularly bromoform (or DBPs associated with bromoform
occurrence), increases the risk of rectal cancer. The study does not address associations among
sensitive populations, other than gender, and does not assess genetic factors that may influence
risk of bladder cancer associated with THM exposure. The study results may not be
generalizable to general populations consisting of men and women.
Kuo et al. (2009) present results of a matched case-control mortality study set in Taiwan (2,195
cases of colon cancer death and 2,195 matched controls were identified between the years 1997-
2006 using death certificate data) in which THM4 were collected from 65 municipalities. The
objective of the study was to evaluate the association between exposure to DBPs and colon
cancer risk. No statistically significant associations between THM4 in drinking water and colon
cancer mortality were observed in the study. The adjusted OR for colon cancer death comparing
those in the highest category of total estimated THM4 concentration in drinking water (above the
75th percentile; >14.80 |ig/L) was 1.04 (95% CI: 0.89, 1.21) compared to the lowest exposure
group (the lowest 50th percentile; <6.03 |ig/L). The OR for those in the "medium" THM4
category (50th-75th percentile; 6.03-14.80 |ig/L) was 1.02 (95% CI: 0.87, 1.2) compared to the
same reference group. (Note that the authors did not provide any further detail on the THM4
concentration ranges in the drinking water in the Taiwan municipalities in this study, but based
on a 75th percentile of 14.80 |ig/L these levels appear to be quite low relative to levels typically
seen in chlorinated drinking water.
The results of the Kuo et al. (2009) study provide no evidence to support the hypothesis that
DBP exposure, or specifically THM4 exposure, increases the risk of colon cancer mortality. The
authors noted that their finding was consistent with most previous epidemiologic studies. Bias of
the ORs due to non-differential errors in the estimation of THM4 exposure is a potential
explanation of the negative finding. The negative finding could also reflect the relatively low
THM4 concentrations and limited differentiation in exposure among the three exposure
categories. The study does not address associations among sensitive populations. The study
results showing no statistically significant association of THM4 with colon cancer are
comparable with several, but not all previous studies evaluating THM4 exposure and colon
cancer (mortality) risk.
Meta-analysis Study
Rahman et al. (2010) conducted a meta-analysis of 13 case-control (n = 10) and cohort (n = 3)
studies published between 1978 and 2007 evaluating the risk of colorectal cancers in relation to
DBPs in drinking water. The studies in the meta-analysis included the Hildesheim et al. (1998)
and King et al. (2000) studies considered by EPA as part of the Stage 1 and Stage 2 D/DBPRs
(USEPA, 2005g) and the more recent Bove et al. (2007a) study. Three of the included studies
considered only colon cancer risk associated with DBP exposure (one cohort and two case-
control), three considered rectal cancer only (all case-control) and seven considered both colon
cancer and rectal cancer (two cohort, five case control). The authors pooled relative risks (RRs)
comparing the highest exposure category to the lowest category from each study using separate
random effects models for colon cancer and rectal cancer. The pooled RR estimates for colon
cancer comparing highest to lowest exposure groups were 1.11 (95% CI: 0.73, 1.70) for cohort
Six-Year Review 3 Technical Support Document A-22
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
studies, 1.33 (95% CI: 1.12, 1.57) for case-control studies and 1.27 (95% CI: 1.08, 1.50)
combining both study types. Corresponding RR estimates for rectal cancer were 0.88 (95% CI:
0.57, 1.35), 1.40 (95% CI: 1.15, 1.70) and 1.30 (95% CI: 1.06, 1.59). The authors did not find
evidence of a single particularly influential study. They concluded that publication bias was not
evident in the colon cancer literature but may have been a minor issue for studies of rectal
cancer. They also noted that estimated RRs for rectal cancer in association with DBP exposures
may have been influenced by the poor quality of the contributing studies. However, the authors
also noted that "although there was evidence that poor study quality affected results for rectal
cancer, removal of the four studies with very low scores on measurement only reduced the
pooled RR a little." The authors concluded that the study provides limited evidence of a positive
association between colorectal cancer and exposure to DBPs in drinking water, citing the small
number of qualifying studies and limitations in study quality as factors that hinder causal
inference.
This meta-analysis provides limited but additional evidence in support of the hypothesis that
long-term DBP exposure increases the risk of colorectal cancer. The authors noted their inability
to make a strong causal statement regarding the association between DBP exposure and risk of
colon and rectal cancers because of the small number of studies contributing to the analysis and
quality issues.
A. 1.2.1.3 Other Cancers
Chiu et al. (2010) collected data from the Taiwan Provincial Department of Health in order to
study the association between pancreatic cancer and THM4 in 53 municipalities in Taiwan.
Additionally, the authors further investigated modification of the relationship between THM4
exposure and risk of death from pancreatic cancer by calcium and magnesium levels in drinking
water. The case group consisted of 594 males and 462 females whose death certificate indicated
pancreatic cancer as the cause of death. THM4 levels were used as a marker for DBP exposure.
THM4 exposure was assessed using average THM4 levels as reported by the Taiwan EPA over a
two-year period. The authors reported an OR of 1.01 (95% CI: 0.85-1.21) for pancreatic cancer
for those with estimated THM4 exposure above the median level (4.9 |ig/L) compared to those
with estimated drinking water THM4 below that level. No interaction between THM4 levels and
calcium in drinking water was observed. However, the authors observed statistically significant
effect modification by magnesium on the multiplicative scale (p<0.05), reporting increased odds
of pancreatic cancer for those with THM4 levels above 4.9 |ig/L, relative to those with THM4
less than 4.9 |ig/L, among those with annual mean levels of magnesium in drinking water below
the median (< 5.4 mg/L) (OR = 1.42; 95% CI: 1.00, 2.02).
This study does not provide evidence that risk of death from pancreatic cancer is associated with
exposure to increased THM4 levels in drinking water. However, the results suggest a
relationship between magnesium levels in drinking water and risk of death from pancreatic
cancer associated with THM4 exposure. The findings show a statistically significant increase in
the odds of pancreatic cancer death in those with individuals exposed to less than the median
observed magnesium level of 5.4 |ig/L but greater than the median observed THM4 level of 4.9
Hg/L.
Six-Year Review 3 Technical Support Document A-23
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Kasim et al. (2006) conducted a population-based case-control study (686 cases, 3,420 controls)
to investigate the relationship between DBP exposure and adult leukemia in Canada. The study
utilized information from the Canadian National Enhanced Cancer Surveillance System.
Additional information collected on subjects' residences and drinking water source history was
linked with municipal water supply data to estimate individual chlorine DBP exposure. Overall,
the odds of leukemia were not associated with duration of THM4 exposure. However, for those
with estimated THM4 exposures greater than 40 |ig/L, the researchers observed increasing odds
of chronic myeloid leukemia with increasing duration of exposure. The OR was only statistically
significant for those with duration of exposure >31 years, the highest category of exposure
duration, and highest THM4 concentration category, > 40 |ig/L (OR = 1.72; 95% CI: 1.01, 3.08).
In contrast, they reported a protective association for chronic lymphoid leukemia for the longest
duration and highest exposure category (OR = 0.60; 95% CI: 0.41, 0.87). Increasing duration of
exposure to bromodichloromethane levels > 5 |ig/L was not associated with risk of leukemia in
this study.
The study provides limited evidence in support of the hypothesis that there is a positive
association between developing adult leukemia and increased chlorination DBP exposure in
drinking water. The results suggest that an increased risk for chronic myeloid leukemia is
associated with increased duration and level of exposure to DBPs. However, for all other
leukemia subtypes included in the study, the OR was found to decrease with increasing duration
and level of exposure to chlorine DBPs; in the instances of lymphocytic leukemia and hairy cell
leukemia, the study actually found a protective effect. While it is possible that the adult leukemia
risk associated with exposure to chlorination DBPs in drinking water differs among the disease
subtypes, the contrary findings of this study do not offer a clear indication of the overall
chlorination DBP risk associated with adult leukemia. Interpreting the findings is particularly
difficult due to the lack of other similar studies available for comparison. The authors also note
that the primary concern in their study was that a low proportion of potentially eligible cases
were included in the analysis. Of 1,997 adult leukemia cases identified, only 1,068 (53.5 percent)
are represented. Case subjects were lost because of death (292 cases), refusal of physicians to
give consent because of cases' ill health (160 cases) and refusal of the cases to participate (467
cases). The possibility of selection bias should be considered if nonrespondents differed from
those analyzed in terms of the studied risk factors.
Karagas et al. (2008) reported what they described as a "preliminary analysis" assessing the
relationship of basal cell carcinoma (BCC) and squamous cell carcinoma (SCC) with THM4
exposure using case-control data (603 BCC cases, 293 SCC cases, 540 controls) originally
gathered for an arsenic analysis. The authors reported an OR of 2.4 (95% confidence interval:
0.9, 6.7) for BCC and 2.1 (95% CI: 0.7, 7.0) for SCC, comparing those exposed from public
water sources to more than 40 |ig/L THM4 to those to a referent group with < 1 |ig/L THM4, and
those with 1 to 20 |ig/L THM4 (OR = 0.9, 95%CI: 0.6-1.5 for BCC and OR = 0.9, 95% CI: 0.6-
1.5 for SCC) and those with >20 to 40 |ig/L THM4 (OR =1.1, 95%CI: 0.7-1.8 for BCC and OR
= 1.3, 95%CI: 0.7-2.3for SCC). The authors also reported for a group served by private wells OR
= 1.1, 95%CI: 0.7-1.8 for BCC and OR = 1.1, 95%CI: 0.6-1.9 for SCC. No information was
provided on the statistical significance of these results. The authors concluded that their results
suggested that the relationship of DBPs with these forms of skin cancer warrants further
exploration.
Six-Year Review 3 Technical Support Document A-24
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
A.l.2.2 Reproductive and Developmental Effects
Birth Weight
Information Available During Development of Stage 1 and Stage 2 D/DBPRs
For the Stage 1 and Stage 2 D/DBPRs, the epidemiological evidence base addressing the
association between DBP exposure and birth weight outcomes consisted of 13 primary studies
(including 4 cross-sectional studies, 7 cohort studies and 2 case-control studies) and 8 review
papers, identified in Exhibit A.l (USEPA, 2005g).
Exhibit A.1: Studies of Birth Weight Outcomes Evaluated for Stage 1 and/or Stage
2 D/DBPRs
Study
Developmental/Reproductive
Health Outcome
Study Design
Savitz et al. (2005)
Term-Birth Weight
Prospective Cohort
Toledano et al. (2005)
LBW, very-LBW
Retrospective Cohort
Wright et al. (2004)
Term Birth Weight
Retrospective Cohort
Wright et al. (2003)
Term Birth Weight, Term-LBW
Retrospective Cohort
Yang (2004)
LBW
Cross-sectional
Jaakkola et al. (2001)
LBW
Cross-sectional
Kallen and Robert (2000)
Birth Weight
Cross-sectional
Dodds et al. (1999)
LBW
Retrospective Cohort
Gallagher et al. (1998)
LBW; Term-LBW
Retrospective Cohort
Kanitz et al. (1996)
LBW
Cross-sectional
Bove et al. (1995)
LBW
Retrospective Cohort
Savitz et al. (1995)
LBW
Case-control
Kramer et al. (1992)
LBW
Case-control
Bove et al. (2002)
LBW
Review
Graves et al. (2001)
LBW
Review
Villanueva et al. (2001)
LBW
Review
Nieuwenhuijsen et al. (2000)
LBW
Review
Reif et al. (2000)
Birth Weight; LBW
Review
WHO (2000)
LBW
Review
Craun, ed. (1998)
LBW
Review
Reif etal. (1996)
LBW
Review
Abbreviations: LBW - Low Birth Weight
The results from this collection of studies were found by EPA to be inconsistent, although a
number of these studies supported the possibility that DBP exposure is associated with adverse
Six-Year Review 3 Technical Support Document A-25
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
birth weight outcomes (particularly for birth weight decrements), including (Kramer et al., 1992;
Bove et al., 1995; Kanitz et al., 1996; Gallagher et al., 1998; Kallen and Robert, 2000; Wright et
al., 2003; Wright et al., 2004). EPA noted that the more recent studies at the time of Stage 2,
which were also higher quality studies, provided some evidence of an association between birth
weight outcomes and maternal DBP exposures during pregnancy; however, such evidence was
limited largely to studies of average differences in a continuous measure of birth weight, rather
than low birth weight outcomes.
Four studies evaluated risk of low birth weight and method of drinking water disinfection, one of
which found some evidence of an association. Kanitz et al. (1996) assessed drinking water
disinfection method (chlorine dioxide, sodium hypochlorite and chlorine dioxide/sodium
hypochlorite) in a cross-sectional study conducted in Italy. The prevalence of low birth weight
did not vary by disinfection method. Kallen and Robert (2000) study assessed drinking water
disinfection method (no chlorine, chlorine dioxide only, sodium hypochlorite only) in a cross-
sectional study conducted in Sweden. They observed a statistically significant association
between low birth weight among infants of mothers who lived in areas supplied by drinking
water disinfected using sodium hypochlorite. Jaakkola et al. (2001) assessed maternal exposure
to chlorinated drinking water during pregnancy in a cross-sectional study in Norway and found
no evidence of an association with low birth weight. Similarly, Yang (2004) study compared the
prevalence of low birth weight (LBW) in 113 municipalities supplied with chlorinated drinking
water to that of 15 municipalities that were not supplied with chlorinated drinking water (but did
not estimate DBP levels in drinking water) in a cross sectional study in Taiwan. No association
was observed between residence in an area supplied by chlorinated water and low birth weight.
The studies evaluating the risk of low birth weight outcomes and estimated THM exposures
during pregnancy generally did not observe associations between the two, although three studies
do provide some evidence for the association. Kramer et al. (1992) estimated chloroform,
bromodichloromethane (BDCM), DBCM and bromoform levels in drinking water in a case-
control study set in Iowa. No associations were observed between odds of low birth weight and
the THM exposures, with the exception of an elevated odds ratio for the association between low
birth weight and chloroform exposure that was not statistically significant. Savitz et al. (1995)
estimated maternal THM4 exposure in drinking water in a case-control study set in North
Carolina. The odds of a low birth weight outcome were not associated with estimated maternal
THM4 exposure. Dodds et al. (1999) estimated THM4 exposure during pregnancy among a
cohort of women in Nova Scotia. Dodds et al. (1999) found no evidence of an association
between THM exposure and low birth weight. Savitz et al. (2005) estimated THM4, HAA9 and
TOX exposures as well as individual brominated THM (BrTHM) and HAA species during
pregnancy in a prospective cohort study of three communities in the United States. Similarly,
Savitz et al. (2005) observed no associations with term-low birth weight. Wright et al. (2003)
estimated THM4 maternal exposures during pregnancy and for each trimester in a cross-sectional
study in Massachusetts. They did not observe statistically significant associations between
second trimester and entire-pregnancy average THM4 levels and low birth weight. In contrast,
Bove et al. (1995) estimated maternal THM4 exposure in drinking water in a cross-sectional
study in New Jersey and found a small association between THM4 levels and very low birth
weight. Gallagher et al. (1998) estimated third-trimester THM levels in drinking water in a
cohort of pregnant women in Colorado. Gallagher et al. (1998) observed strong, statistically
significant association between term-low birth weight and high third-trimester THM4 exposure
Six-Year Review 3 Technical Support Document A-26
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
and small, non-statistically significant associations between estimates of third-trimester THM4
exposure and low birth weight. Toledano et al. (2005) modelled THM4 in water zones in a large
cross sectional study of three regions in England. Statistically significant associations were
observed between THM4 and risk of LBW in one of the three regions. When assessing the
association in all three regions combined, a small increase in the risk of LBW associated with
THM4 was observed that was not statistically significant. A similarly small, non-statistically
significant increase in risk of LBW was also observed for chloroform. They did not observe
associations between LBW and BDCM or total BrTHMs.
The studies that assessed associations between DBP levels and differences in average birth,
rather than a dichotomous low birth weight outcome, provided some evidence of an association.
Wright et al. (2003) assessed associations between THM4 and both low birth weight
(summarized above) and average birth weight. They observed statistically significant
associations between second trimester and entire-pregnancy average THM4 levels and mean
birth weight (but did not observe an association between THM4 levels and their low birth weight
outcome). Wright et al. (2004) estimated THM4, chloroform, BDCM, total HAA, DCA and
TCAA levels in a large retrospective cohort study of maternal third-trimester drinking water
exposure and birth weight in Massachusetts. They observed an exposure-response relationship
between estimated THM levels and reductions in mean birth weight as well as statistically
significant reductions in average birth weight associated with BDCM and chloroform levels.
None of the review papers concluded that the weight of evidence was suggestive of a causal
relationship between DBP exposure or exposure to chlorinated drinking water during pregnancy
and risk of a low birth weight outcome. Only two reviews found some support, albeit
inconclusive, for an association between DBP exposure (Nieuwenhuijsen et al., 2000) and
exposure to chlorinated drinking water (Villanueva et al., 2001) during pregnancy and low birth
weight outcomes.
New Information Available Since Development of Stage 2 D/DBPR
EPA conducted a literature search to identify new epidemiology studies of DBP associations
with birth weight outcomes that became available subsequent to the promulgation of the Stage 2
D/DBPR. Twelve studies were identified and evaluated: four prospective birth cohort studies,
five retrospective cohort studies, one case-control study, one cross-sectional study and one meta-
analysis.
•	Prospective cohort studies:
o Hoffman et al. (2008a)
o Patelarou et al. (2011)
o Grazuleviciene et al. (2011)
o Villanueva et al. (2011)
•	Retrospective cohort studies:
o	Hinckley et al. (2005)
o	Lewis et al. (2006)
o	Yang et al. (2007)
o	Rivera-Nunez and Wright (2013)
o	Kumar et al. (2013)
Six-Year Review 3 Technical Support Document A-27
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
•	Case-control studies:
o Danileviciute et al. (2012)
•	Cross-sectional studies:
o Zhou et al. (2012)
•	Meta-analysis studies:
o Grellier et al. (2010)
Five of the 11 observational studies were conducted in the United States (Arizona (Hinckley et
al., 2005), New York (Kumar et al., 2013), Massachusetts (Lewis et al., 2006; Rivera-Nunez and
Wright, 2013)) and 3 unspecified U.S. communities (Hoffman et al., 2008a), four cities in
Europe (Spain (Villanueva et al., 2011), Lithuania (Danileviciute et al., 2012; Grazuleviciene et
al., 2011) and Crete (Patelarou et al., 2011)), and 2 in Asia (Taiwan (Yang et al., 2007) and
China (Zhou et al., 2012)). A primary birth weight outcome assessed in 6 of the 11 observational
studies that considered birth weight outcomes was low birth weight LBW defined as birth weight
<2,500 grams (Hinckley et al., 2005; Patelarou et al., 2011; Villanueva et al., 2011; Danileviciute
et al., 2012; Grazuleviciene et al., 2011; Kumar et al., 2013). Term-LBW, defined as LBW
among births >37 weeks of gestation, was assessed in 3 studies (Hinckley et al., 2005; Lewis et
al., 2006; Yang et al., 2007), and differences in mean birth weights were assessed in 5 studies
(Hoffman et al., 2008a; Villanueva et al., 2011; Zhou et al., 2012; Grazuleviciene et al., 2011;
Rivera-Nunez and Wright, 2013). [Note: Studies addressing small for gestational age (SGA)
outcomes were evaluated separately, see summary for SGA, below].
All but 2 (Patelerou et al., 2011; Zhou et al., 2012) of the 11 observational studies of birth weight
endpoints assessed associations between birth weight outcomes and total trihalomethanes
(THM4). Six of these nine studies additionally assessed specific THM concentrations
(Danileviciute et al., 2012; Grazuleviciene et al., 2011; Hinckley et al., 2005; Hoffman et al.,
2008a; Villanueva et al., 2011; Rivera-Nunez and Wright, 2013). Three of these studies
additionally assessed total BrTHM (Patelarou et al., 2011; Villanueva et al., 2011; Rivera-Nunez
and Wright, 2013), and three studies (Hinckley et al., 2005; Hoffman et al., 2008a; Rivera-Nunez
and Wright, 2013) assessed specific haloacetic acid (HAA) species as well as HAA5 and/or
HAA9. Rivera-Nunez and Wright (2013) additionally examined a DBP9 metric summing HAA5
and THM4 exposures. Five of the birth weight investigations assessed for this analysis quantified
DBP levels in water sampled quarterly (Danileviciute et al., 2012; Grazuleviciene et al., 2011;
Hinckley et al., 2005; Yang et al., 2007; Rivera-Nunez and Wright 2013), two used weekly water
samples (Hoffman et al., 2008a and Lewis et al., 2006), and the sampling frequency varied for
three other studies (Kumar et al., 2013; Patelarou et al., 2011; Villanueva et al., 2011). Zhou et
al. (2012) assessed urinary creatinine-adjusted trichloroacetic acid (TCAA) as a DBP exposure
biomarker. Nine of the 11 studies estimated tap water DBP concentrations (Danileviciute et al.,
2012; Grazuleviciene et al., 2011; Hoffman et al., 2008a; Kumar et al., 2013; Lewis et al., 2006;
Patelarou et al., 2011; Yang et al., 2007; Villanueva et al., 2011; Rivera-Nunez and Wright,
2013); four studies estimated internal THM dose (Hoffman et al., 2008; Danileviciute et al.,
2012; Grazuleviciene et al., 2011; Villanueva et al., 2011). Exposures were typically quantified
into categories (e.g., quantiles), but several studies also assessed associations between birth
weight outcomes and continuously distributed DBP exposure metrics.
In a prospective cohort study, Hoffman et al. (2008a) found limited evidence of associations and
no consistent patterns for term birth weight associations across the different study sites for the
Six-Year Review 3 Technical Support Document A-28
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
various HAAs and other DBP exposure metrics. Although not statistically significant, they did
detect some birth weight reductions consistent in magnitude with other studies for THM4
exposure (56 grams; 95% CI: -144, 32) >80 [j,g/L (vs <80 (J,g/L) and TOX exposures (40 grams;
95% CI: -109, 29) >192.7 [j,g/L (vs. < 22.4 (J,g/L). Rivera-Nunez and Wright (2013) reported
consistent birth weight reductions across all 3rd trimester HAA5 (range: 28-36 grams) and
THMBr quintiles (range: 11-23 grams), but the largest associations were detected for DBP9 (the
sum concentration of THM4 and HAA5) quintiles (range: 39-45 grams). Reductions in mean
birth weight for THM4 quintiles (range: 9-23 grams) did not persist following additional
adjustment for HAA5 levels in the multi-pollutant models. Compared to those in the lowest
quartile, Zhou et al. (2012) found lower average birth weight (range: 40 to 62 grams) of infants
whose mothers were in the highest two quartiles of creatinine-adjusted urinary TCAA
concentrations for the overall population and even larger reductions (range: 82 to 160 grams)
among a subset who had completed questionnaires with additional information on other potential
confounding variables. Although no birth weight associations were reported in a Spanish study
(Villanueva et al., 2011) across different THM metrics, Grazuleviciene et al. (2013) reported
consistent, statistically significant, mean birth weight reductions for internal dose third trimester
THM4, chloroform, bromodichloromethane and dibromochloromethane exposures in a
prospective cohort study from Lithuania.
Several of these studies observed associations between birth weight outcomes and specific DBP
species. Grazuleviciene et al. (2011) observed exposure-response relationships between internal
dose of chloroform estimated for the entire pregnancy and for each trimester and decreasing
mean birth weight (range: 53-59 grams for every 1 (J,g/day increase in chloroform). Rivera-
Nunez and Wright (2013) observed statistically significant associations between specific DBP
species (chloroform and bromodichloromethane) and mean birth weight deficits (range: 9-20
grams). These associations did not persist in multi-pollutant models (i.e., following further
adjustment for other DBP exposures), which hasn't been examined in previous studies.
Danileviciute et al. (2012) assessed entire pregnancy and trimester-specific levels of specific
DBP species (chloroform, dibromochloromethane and bromodichloromethane) and LBW risk.
They observed elevated LBW risk associated with all three species, although the associations
were statistically significant only among women with the GSTM1-0 genotype who had 3rd
trimester or entire pregnancy chloroform exposure above the median level. In a separate
investigation of the same study population, Grazuleviciene et al. (2011) observed exposure-
response relationships between internal dose of chloroform estimated for the entire pregnancy
and for each trimester and risk of LBW (OR range: 1.09-2.41) and internal dose and LBW risk.
Hinckley et al. (2005) did not observe an association between chloroform and term-LBW in their
retrospective cohort study, but they detected associations between exposure to specific HAAs
(dibromoacetic acid (DBAA), in particular) and risk of term-LBW.
The studies of birth weight endpoints evaluated for this analysis focused primarily on third
trimester DBP exposure, although most also presented association estimates for the first and
second trimesters and the entire pregnancy. Rivera-Nunez and Wright (2013) reported consistent
birth weight reductions regardless of whether second or third trimester exposure were evaluated.
Lewis et al. (2006), however, reported that fetal growth was affected by high levels (>70 jag/liter)
of THM4 exposure experienced during the second trimester, having observed elevated,
statistically significant elevations in the odds of term-LBW, relative to those with THM4 levels
Six-Year Review 3 Technical Support Document A-29
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
<40 jag/liter only during that time period. Grazuleviciene et al. (2011) observed consistent
associations between risk of LBW and of estimated categories of THM4, chloroform and
bromodichloromethane levels across all trimesters; Danileviciute et al. (2012) observed the
largest associations between THM4 and chloroform levels and LBW during the third trimester
Hinckley et al. (2005) assessed smaller exposure time windows and found evidence suggesting a
critical window for specific HAAs exposure effects on fetal growth between 33-40 weeks of
gestation.
The results from the studies of DBP exposure during pregnancy and risk of adverse birth weight
outcomes (term birth weight, LBW and term-LBW) were mixed; seven of the nine observational
studies reported at least some evidence supporting the hypothesis that DBP exposure increases
the risk of those adverse birth weight-related outcomes. There was no clear indication of greater
consistency of reported associations, nor of exposure-response trends, among studies that used
more sophisticated exposure assessment methodologies. The weight of evidence among the post-
Stage 2 D/DBPR studies is suggestive of either no associations, or at most, small positive
associations, between DBP levels in drinking water and adverse birth weight outcomes. The
evidence provided by the articles reviewed for this analysis is not conclusive regarding the
existence of an increased risk of LBW due to DBP exposure. There was little evidence of
consistency regarding the magnitude of non-null associations or exposure-response relationships.
With the exception of the Hoffman et al. (2008a) analyses and the Rivera-Nunez and Wright
(2013) multi-pollutant-adjusted models, associations between DBP exposure estimates and birth
weight outcomes were consistently observed during the third trimester and inconsistently in other
trimesters and for the entire pregnancy period.
The one meta-analysis (Grellier et al., 2010), which included four reports of THM4 associations
with LBW and four reports of THM4 and term-LBW outcomes (both sets included two studies
reviewed for the DBP Stage 2 Rule (Gallagher et al., 1998; Wright et al., 2003) and two studies
that were not (Hinckley et al. (2005); Lewis et al. (2006)), concluded that there is little or no
evidence for associations between THM4 concentrations and risk of LBW or term-LBW;
summary odds ratios corresponding to a small increase (10 (J,g/L) in THM4 were 1.00 and 1.034
for LBW and term-LBW, respectively, and were not statistically significant.
Susceptible Populations. Three of the studies (Danileviciute et al., 2012; Lewis et al., 2006;
Rivera-Nunez and Wright 2013) reviewed in this analysis reported evidence of effect
modification of associations between DBP exposure and adverse birth outcome; one by genotype
and the others by race. Danileviciute et al. (2012) jointly considered the effects of DBP exposure
and maternal genotypes and observed the strongest associations for third trimester THM4 and
chloroform-exposed women with the GSTM1-0 genotype (OR: 4.37; 95% CI: 1.36-14.08, and
OR: 5.06; 95% CI: 1.50-17.05, respectively). A corresponding increase in LBW risk associated
with third trimester THM4 and chloroform exposure among women with the GSTM1-1 genotype
were not observed (OR: 0.34; 95% CI: 0.09-1.24 and OR: 0.35; 95% CI: 0.10-1.28,
respectively). LBW odds ratio estimates associated with third trimester THM4 and chloroform
exposure were also notably higher among women with the GSTT1-0 genotype, relative to
women with the GSTT1-1 genotype, although the interaction was not statistically significant.
In their case-control study, Lewis et al. (2006) observed an increased risk of term-LBW among
women with an average THM4 exposure >70 jag/liter during the second trimester, relative to
Six-Year Review 3 Technical Support Document A-30
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
those with estimated average THM4 exposure < 40 jag/liter (OR = 1.50, 95% CI: 1.07, 2.10).
After stratifying the study population on race, however, they observed an estimated risk increase
associated with high average THM4 exposure during the second trimester of 37 percent (OR =
1.37, 95% CI: 0.80, 2.36) among Caucasians and 60 percent (OR = 1.60, 95% CI: 1.03, 2.47)
among all minority women combined (i.e., African Americans, Hispanics and Asians). However,
the combination of multiple non-Caucasian racial and ethnic groups with different susceptibility
to adverse reproductive outcomes into one group is a key limitation of this study which precludes
drawing further conclusions from these data. A much larger study in Massachusetts by Rivera-
Nunez and Wright (2013) found little evidence of effect measure modification by income,
race/ethnicity or other covariates in the multivariate linear regression models of mean birth
weight. Findings of differences in estimates of the risk of adverse birth outcomes within strata of
genotype and race are noteworthy, but should be interpreted cautiously and should be examined
in other studies.
Other notable contributions to the literature base. As noted earlier, accurate and precise
assessment of DBP exposures was a major challenge for the DBP health effects evaluations
included in this analysis, and indeed for many similar studies. Theoretically, evaluation of
biomarkers of DBP exposure can be used as an alternative strategy, one that has the potential to
overcome many of the difficulties in quantifying DBP levels in drinking water and resulting
exposure. One of the nine studies reviewed for this analysis assessed associations between birth
weight and a biomarker of DBP exposure; Zhou et al. (2012) evaluated the association between
maternal urinary TCAA and birth weight. Using a single measurement of maternal TCAA
quantified in urine sampled just prior to delivery, Zhou et al. (2012) detected an association
between urinary TCAA and birth weight; the average birth weight of infants whose mothers were
in the highest two quartiles of creatinine-adjusted urinary TCAA concentrations were much
lower (62-160 grams) than those of infants whose mothers were in the lowest quartile of TCAA
concentrations. Urinary TCAA has been demonstrated to be a reliable marker of HAA exposure
from ingestion of drinking water (Bader et al., 2004; Froese et al., 2002; Zhang et al., 2009b), but
it is unclear whether urinary TCAA would be accurate DBP surrogates for the volatile DBPs and
other non-volatile DBPs. TCAA is also not specific to DBP exposure and could result from other
environmental contaminants including trichloroethene, 1,1,1-trichloroethane and
perchloroethene, complicating the source apportionment issue. Finally, the degree to which
maternal TCAA sampled just prior to delivery, as was done in this study, adequately represents
maternal DBP exposure during a hypothesized biologically relevant DBP exposure time window
for birth weight-related indices was not evaluated by the investigators.
Post stage 2 D/DBPR, there are no new animal toxicity studies researching birth weight.
Small for Gestational Age
Information Available During Development of Stage 1 and Stage 2 D/DBPRs
For the Stage 1 and Stage 2 DBPRs, the epidemiology evidence base regarding the association
between DBP exposure and the fetal growth endpoint Small for Gestational Age (SGA,
alternatively referred to as Intrauterine Growth Retardation (IUGR)) consisted of 10 primary
studies (including 6 cross-sectional studies, 2 cohort studies and 2 case-control studies) and 6
review papers, identified in Exhibit A.2 (USEPA, 2005g).
Six-Year Review 3 Technical Support Document A-31
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Exhibit A.2: Studies of Small for Gestational Age Outcomes Evaluated for Stage 1
and 2 D/DBPRs
Study
Developmental/Reproductive
Health Outcome
Study Design
Porter et al. (2005)
IUGR
Cross-sectional
Savitz et al. (2005)
SGA
Prospective Cohort
Infante-Rivard (2004)
IUGR
Case-control
Wright et al. (2004)
SGA
Cross-sectional
Wright et al. (2003)
SGA
Cross-sectional
Jaakkola et al. (2001)
SGA
Cross-sectional
Kallen and Robert (2000)
IUGR
Cross-sectional
Dodds et al. (1999)
SGA
Retrospective Cohort
Bove et al. (1995)
SGA
Cross-sectional
Kramer et al. (1992)
IUGR
Case-control
Bove et al. (2002)
SGA
Review
Graves et al. (2001)
SGA
Review
Villanueva et al. (2001)
SGA
Review
Reif et al. (2000)
SGA
Review
Craun, ed. (1998)
SGA
Review
Reif etal. (1996)
SGA
Review
Abbreviations: SGA- Small for Gestational Age; IUGR - Intrauterine Growth Retardation.
The results from this collection of studies were found to be inconsistent, although a number of
these studies supported the possibility that DBP exposure is associated with the SGA outcome,
including (Wright et al., 2004; Wright et al., 2003, Bove et al., 1995; Kramer et al., 1992; Savitz
et al., 2005):
Kallen and Robert, (2000) assessed drinking water disinfection method (no chlorine, chlorine
dioxide only, sodium hypochlorite only) in a cross-sectional study conducted in Sweden and
Jaakkola et al. (2001) assessed maternal exposure to chlorinated drinking water during
pregnancy in a cross-sectional study in Norway. Neither of these studies found evidence of an
association between disinfection of drinking water and IUGR/SGA outcomes.
The studies evaluating the risk of SGA and estimated THM exposures during pregnancy reported
inconsistent results, with one study finding no evidence of an association, six studies finding at
least some evidence of associations of varying magnitudes and another finding an association
only in a subset of infants with a specific genetic polymorphism. Dodds et al. (1999) estimated
THM4 exposure during pregnancy among a cohort of women in Nova Scotia. They found no
evidence of an association between THM exposure and SGA. Kramer et al. (1992) estimated
chloroform, BDCM, DBCM and bromoform levels in drinking water in a case-control study set
in Iowa and observed a statistically significant increased risk of IUGR associated with
chloroform levels.
Six-Year Review 3 Technical Support Document A-32
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Bove et al. (1995) estimated maternal THM4 exposure in drinking water in a cross-sectional
study in New Jersey and found a small but statistically significant association between THM4
levels and SGA. Wright et al. (2003) estimated THM4 maternal exposures during pregnancy and
for each trimester in a cross-sectional study in Massachusetts. Wright et al. (2004) estimated
maternal third-trimester drinking water THM4, chloroform, BDCM, total HAA, DCAA and
TCAA levels in a cross-sectional study of and SGA in Massachusetts, observing an exposure-
response relationship between estimated THM levels and SGA. Savitz et al. (2005) estimated
THM4, HAA9 and TOX exposures as well as individual BrTHM and HAA species during
pregnancy in a prospective cohort study of three communities in the United States and found that
third-trimester THM4 levels above 80 |ig/L were associated with a statistically significant
doubling of the risk for SGA, compared to THM4 levels less than 80 |ig/L. Porter et al. (2005)
estimated trimester-specific and pregnancy-average exposures to specific THMs and HAAs, as
well as THM4 and HAA5 in a cross-sectional study of pregnant mothers and their infants in
Maryland. They did not observe any exposure-response trends in the odds of IUGR associated
with increasing THM4 or HAA5 levels, nor did they observe increased IUGR risk associated
with levels of specific THMs or HAAs. However, they did observe non-statistically significant
elevated risk of IUGR associated with the two highest quintiles of THM4 and statistically
significant elevated risk of IUGR associated with the two highest quintiles of HAA5.
Infante-Rivard (2004) estimated THM levels in drinking water for a case-control study in
Montreal. Although she found no association between THM levels and intrauterine growth
retardation overall, a statistically significant association was observed between THM exposure
and intrauterine growth retardation among infants with the CYP2E1 gene variant, suggesting
genetic susceptibility may modify risk of DBP effects on developmental outcomes.
None of the review papers concluded that the weight of evidence was suggestive of a causal
relationship between DBP exposure or exposure to chlorinated drinking water during pregnancy
and risk of SGA. Only one review reported some support, albeit inconclusive, for an association
between exposure to chlorinated drinking water during pregnancy and SGA (Villanueva et al.,
2001).
New Information Available Since Development of Stage 2 D/DBPR
EPA conducted a literature search to identify new epidemiology studies of DBP and SGA
outcomes that became available subsequent to the promulgation of the Stage 2 D/DBPR.
Fourteen studies were identified and evaluated: four prospective birth cohort studies, six
retrospective cohort studies, three case-control studies and one meta-analysis.
•	Prospective cohort studies:
o	Hoffman et al. (2008a)
o	Patelarou et al. (2011)
o	Grazuleviciene et al. (2011)
o	Costet et al. (2012) (also implemented a case-control sampling design)
•	Retrospective cohort studies:
o	Hinckley et al. (2005)
o	Yang et al. (2007)
o	Horton et al. (2011)
Six-Year Review 3 Technical Support Document A-33
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
o Summerhayes et al. (2012)
o Rivera-Nunez and Wright (2013)
o Kumar et al. (2013)
•	Case-control studies:
o Aggazzotti et al. (2004)
o Danileviciute et al. (2012)
o Levallois et al. (2012)
•	Meta-analysis studies:
o Grellier et al. (2010)
Five of these new SGA studies were conducted in the United States: Arizona (Hinckley et al.,
2005), Massachusetts (Rivera-Nunez and Wright, 2013), New York (Kumar et al., 2013), "two
Southern U.S. communities" (Horton et al., 2011) and "three U.S. communities" (Hoffman et al.,
2008a). Five studies were conducted in Europe: Italy (Aggazzotti et al., 2004), France (Costet et
al., 2012), Lithuania (Danileviciute et al., 2012; Grazuleviciene et al., 2013) and Crete (Patelarou
et al., 2011). One each was conducted in Australia (Summerhayes et al., 2012), Canada
(Levallois et al., 2012) and Taiwan (Yang et al., 2007). Eight of these reports also assessed other
fetal growth endpoints.
The SGA endpoint is a gestational age-adjusted measure of birth weight and a marker of fetal
growth restriction. The definitions of SGA used in this literature are heterogeneous, but
qualitatively similar. SGA is often defined as being less than or equal to the lowest 10th
percentile of weight from a reference population, commonly within categories of gender and
race/ethnicity (e.g., Rivera-Nunez and Wright, 2013). However, this SGA definition is merely
conventional, and several of the studies used alternative definitions. Aggazzotti et al. (2004),
Hoffman et al. (2008a), Horton et al. (2011), Levallois et al. (2012), and Rivera-Nunez and
Wright (2013) further restricted their case definitions to infants born after >37 weeks of
pregnancy (i.e., term-SGA). Summerhayes et al. (2012) excluded infants with gestational age <
22 weeks and >43 weeks, as well as those with birth weight >5 standard deviations from the
average for gestational age. Levallois et al. (2012) defined SGA as a neonate weighing less than
the 10th percentile weight for gestational age and gender. Hoffman et al. (2008a) defined SGA as
present among infants with birth weight below the 10th percentile specific to parity, in addition to
gender and maternal race/ethnicity. Kumar et al. (2013) defined SGA as infant birth weight
below the 10th percentile of birth weight distribution among singleton live births in New York
State for gestational age in weeks, year of birth and gender. Costet et al. (2012) defined their
SGA outcome, which they referred to as fetal growth restriction as birth weight below the fifth
percentile of the cohort's expected birth-weight distribution. Patelarou et al. (2011) defined SGA
based on weight (SGAweight) defined as a live born infant below the 10th percentile of birth weight
for gestational age in a referent population from Spain and also defined two additional endpoints
based on body length (SGAiength) and head circumference (SGAhead circumference).
Estimation of gestational age is subject to error; gestational age can be derived from maternal
report of last menstrual period (Kumar et al., 2013; Patelarou et al., 2011), estimated using
ultrasound evaluation (Grazuleviciene et al., 2011), or from clinical estimates based on either
ultrasound or the clinical examination (Rivera-Nunez and Wright, 2013; Costet et al., 2012;
Summerhayes et al., 2012; Hoffman et al., 2008a). For example, Hoffman et al. (2008a) derived
gestational age at birth using first trimester maternal report of date of last menstrual period,
Six-Year Review 3 Technical Support Document A-34
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
which was corrected by ultrasound if the two estimates of gestational age differed greater than
one week. The methodology used to estimate gestational age is not specified in six of the reports
(Aggazzotti et al., 2004; Danileviciute et al., 2012; Hinckley et al., 2005; Horton et al., 2011;
Levallois et al., 2012; Yang et al., 2007). The robustness of the study results to alternate SGA
definitions (e.g., percentile cut-points, referent population, methods used to estimate gestational
age) was examined in only one of the articles reviewed for this analysis (Summerhayes et al.,
2012) where less than the 3rd percentile weight for gestational age was also considered and for
which the authors noted there was some suggestion of a threshold, though noting that the
investigation of threshold effects was limited in their study by the lack of an unexposed
population.
Water sampled from municipal water distribution systems was used to estimate DBP exposure in
12 of the studies. Patelarou et al. (2011) obtained water samples from mothers' homes in
addition to sampling from the public water supply network. The Aggazzotti et al. (2004) study
assessed DBP levels in tap water sampled from mothers' homes. The frequency of water
sampling was typically conducted quarterly, though the sampling frequency ranged from weekly
to annually. The number of sampling sites used to assess DBP levels also varied by study. The
representativeness of the samples taken was only formally evaluated in one study (Horton et al.,
2011). Seven of 13 observational studies queried study participants' beverage consumption and
water use behaviors (Aggazzotti et al., 2004; Costet et al., 2012; Danileviciute et al., 2012;
Grazuleviciene et al., 2011; Hoffman et al., 2008a; Levallois et al., 2012; Patelarou et al., 2011).
All of the studies assessed associations between SGA and THM4; nine studies additionally
assessed specific THM concentrations, other studies additionally assessed total BrTHMs
(Patelarou et al., 2011; Rivera-Nunez and Wright, 2013). Five of the post-Stage 2 studies also
assessed HAA5 and/or HAA9 (Hinckley et al., 2005; Hoffman et al., 2008a; Horton et al., 2011;
Levallois et al., 2012; Rivera-Nunez and Wright, 2013) and three studies assessed individual
HAAs (Hinckley et al., 2005; Levallois et al., 2012; Rivera-Nunez and Wright, 2013).
Aggazzotti et al. (2004) assessed chlorite and chlorate in addition to THM4 and individual
THMs. Two studies additionally examined total organic halides (TOX) (Hoffman et al., 2008a;
Horton et al., 2011), one study examined a DBP9 metric (Rivera-Nunez and Wright, 2013), and
another study summed up all chlorinated THMs and HAAs and all BrTHMs and HAAs (Horton
etal.,2011).
Eleven of the 13 DBP studies examined tap water DBP concentrations in relation to SGA
(Aggazzotti et al., 2004; Hinckley et al., 2005; Yang et al., 2007; Hoffman et al., 2008a; Horton
et al., 2011; Summerhayes et al., 2012; Patelarou et al., 2011; Levallois et al., 2012; Costet et al.,
2012; Rivera-Nunez and Wright, 2013; Kumar et al., 2013), seven studies specifically estimated
DBP uptake based on individual-level data (Aggazzotti et al., 2004; Costet et al., 2012; Hoffman
et al., 2008a; Patelarou et al., 2011; Grazuleviciene et al., 2011; Danileviciute et al., 2012;
Levallois et al., 2012), and one study additionally examined TCAA concentrations in maternal
urine sampled early in pregnancy (Costet et al., 2012). Exposures were typically quantified into
categories (e.g., quantiles), but several studies also assessed associations between birth weight
outcomes and continuously distributed DBP exposure metrics.
In the case-control study by Aggazzotti et al. (2004), THM levels were very low, with 23 percent
of samples below the lower limit of detection. The reported frequency of use of tap water for
Six-Year Review 3 Technical Support Document A-35
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
drinking was also low (14 percent), although almost 70 percent of the participants reported using
tap water to make beverages such as coffee and tea. Chlorite and chlorate were also detected in
45 percent and 34 percent of water samples, respectively, with levels often observed to be very
high. Small elevations in odds of term-SGA for reported personal water use for home cooking
and showering/bathing were not statistically significant. THM4 exposure was not significantly
associated with term-SGA (OR: 0.63; 95%CI: 0.31-1.28), comparing subjects exposed to tap
water THM4 concentrations > 10 |ig/L and who reported bathing/showering at least daily to
those with lower THM4 concentrations or did not report bathing/showering at least daily.
Compared to those with chlorite levels <20 |ig/L or between 20 and 199 |ig/L and at lower
inhalation exposure level, those with chlorite levels >200 |ig/L and considered to have low
inhalation exposure had an odds ratio for SGA of 1.52 (95% CI: 0.91-2.54) while subjects with
chlorite levels >200 |ig/L and considered to have high inhalation exposure had an odds ratio for
SGA of 1.70 (95%CI: 0.97-3.0).
Four studies estimated associations between DBP exposures for the entire pregnancy and specific
to each trimester (Grazuleviciene et al., 2011; Danileviciute et al., 2012; Summerhayes et al.,
2012; Patelarou et al., 2011), one study examined exposure during each of the three trimesters
(Costet et al., 2012), another study examined second and third trimester exposures (Rivera-
Nunez and Wright, 2013), another five specifically assessed associations with only third
trimester exposure estimates (Aggazzotti et al., 2004; Hinckley et al., 2005; Hoffman et al.,
2008a; Horton et al., 2011; Levallois et al., 2012), and two studies examined pregnancy average
exposure estimates (Yang et al., 2007; Kumar et al., 2013).
The results from these epidemiologic studies of DBP exposure during pregnancy and risk of
adverse SGA outcomes were mixed. Three studies did not report statistically significant evidence
of increased risk of SGA associated with DBP exposure or evidence of exposure-response
relationships (Horton et al. 2011; Patelarou et al. 2011; Yang et al., 2007). Two of these studies
had very low THM levels which likely precluded evaluation of exposure-response trends.
Associations in Rivera-Nunez and Wright study (2013) for SGA noted for THM4 and BDCM
disappeared after further adjustment for HAA5 exposures, with no evidence of an exposure-
response relationship. A fifth study (Kumar et al., 2013) observed no exposure-response trend,
nor any statistically significant associations, other than a 7 percent and 10 percent increase in the
odds of SGA among infants of mothers in the second lowest and lowest (of five) THM4
exposure categories, respectively. Summerhayes et al. (2012) detected statistically significant
associations for SGA and 3rd trimester for the highest BDCM decile and the two highest THM4
and chloroform deciles. However, they did not observe any clear linear exposure-response trends
for any DBP indicator or species in any trimester. The nested case-control study by Danileviciute
et al. (2012) did not find statistically significant associations between greater than the median
levels of exposure to THM4 or specific THM species in any trimester, with the exception of a
2.2-fold increase in the odds of SGA among those who had first trimester DBCM above the
median, relative to those with DBCM less than the median level. However, all of the remaining
15 exposure metrics were consistently elevated within a range of 1.2 to 1.7 regardless of
exposure window and type of THM metric that were examined. Results similar in magnitude
(range 1.2 to 1.4) for internal dose tertiles for THM4, BDCM and CHC13 were also detected in
the overall cohort from this study base. Increased risks were noted in all BDCM categories as
well. (Grazuleviciene et al., 2011).
Six-Year Review 3 Technical Support Document A-36
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
In the study by Costet et al. (2012), THM4 ranged from 0.6 [j,g/L to 157 [j,g/L (mean (SD): 41.7
(16.1) (J,g/L). Average (SD) specific DBP levels were 9.3 (7.0) for chloroform, 8.2 (5.7) for
bromoform, 13.8 (5.5) for DBCM and 10.4 (5.4) for BDCM. Based solely on water
concentration exposure data, Costet et al. (2012) detected consistently elevated ORs (Range 1.3-
1.4) for SGA and the three upper bromoform quartiles. The authors detected consistently
elevated ORs (Range 1.5-2.4) for all three upper THM uptake quartiles for THM4, DBCM and
BDCM and also in the upper quartile of bromoform. They also reported the largest associations
with showering/bathing THM uptake exposures (OR range = 2.2-2.5).
Hinckley et al. (2005) observed small associations in ORs between SGA and continuous
measures (relative increase in odds for every 1 [j.g/L increase in DBP) of third trimester THM4,
chloroform, BDCM, DBCM and HAA5 exposures that were near the null of 1.00 and of
borderline statistical significance; ORs observed for SGA and continuous third-trimester DBAA,
DCAA and TCAA were higher (OR range: 1.01-1.6) and statistically significant for DCAA and
TCAA based on categorical and continuous exposures. Levallois et al. (2012) did not observe a
clear exposure-response relationship between SGA and third trimester THM4, and specific THM
and HAA exposures, but they did observe an elevated SGA risk among infants of mothers with
HAA5 concentrations in the highest quartile (relative to the lowest quartile) and among those
with third trimester THM4 >80 (J,g/L. They also reported consistently elevated ORs for the
highest ingestion quartile exposures for chloroform, THM4, DCAA, TCAA, HAA5 and HAA9;
but increased risks were not evident when the total THM exposure pathway estimates based on
pharmacokinetic modeling were evaluated. Hoffman et al. (2008b) observed an elevated SGA
risk associated with third trimester THM4 >80 [j,g/L (compared to <80 [ig/L), but did not detect
statistically significant SGA risk associated with the highest quintile of residential THM4
concentrations (RR = 1.3; 95% CI: 0.7-2.3) or THM4 showering and bathing estimates (RR =
1.6; 95%CI: 1.0-2.7). The authors did not observe an association with highest quintiles of
residential HAA5 concentrations (RR = 0.9; 95% CI: 0.5-1.6) or HAA5 tap water consumption
estimates (RR = 0.8; 95%CI: 0.5-1.4) but saw some suggestion of an increased risk for the
highest quintile of residential TOX concentrations (RR = 1.5; 95% CI: 0.9-2.5).
The weight of evidence based on the post-Stage 2 studies reviewed for this analysis is suggestive
of a small positive association between SGA and some between DBP exposure metrics. In
general, there was not strong support of exposure-response relationships between increased risk
of SGA and DBP exposures although there was often elevated risk noted in the highest DBP
exposure category which one would expect if there is a causal relationship. There was also some
evidence of consistency regarding the magnitude of associations for different DCAA exposure
metrics but no other strong signals were noted for individual DBP measures.
The one meta-analysis (Grellier et al., 2010) included in this report summarized the findings of
six studies of THM4 associations with SGA, two of which (Hinckley et al., 2005; Hoffman et al.,
2008a) were also reviewed for this analysis. The summary odds ratio based on the pooled
analysis for a 10 [j,g/L increase in third trimester THM4 level estimated in the meta-analysis was
statistically significant (OR = 1.01; 95% CI: 1.001-1.019). Although not statistically significant,
the summary odds ratio was the same (OR = 1.01; 95% CI: 0.971-1.051) for a 10 [j,g/L increase
in THM4 levels during the entire pregnancy based on four of the six studies reporting this
measure.
Six-Year Review 3 Technical Support Document ^4-37
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Evidence of interaction. Two of the studies reviewed in this analysis reported evidence of effect
modification of associations between DBP exposure and SGA; one by genotype and the other by
smoking.
Building on earlier work assessing modification of associations between THM exposure and
SGA by genetic polymorphisms conducted by Infante-Rivard (2004), Danileviciute et al. (2012)
jointly considered the effects of DBP exposure and polymorphisms in maternal genotypes for
glutathione S-transferases, critical enzymes in metabolic (detoxification) pathways. They found
that odds ratios for SGA associated with having DBP exposure (THM4, chloroform and
bromodichloromethane) above median levels were higher among women with GSTM1-0,
relative to those with the GSTM1-1 genotype. Similar differences in RRs for SGA associated
with dibromochloromethane exposure were not observed between those with GSTM1-0 and
GSIM1-1 genotypes. No differences in risk were observed comparing women with GSTT1-1
genotype to women with GSTT1-0 genotype. The earlier investigation of gene-by-environment
interactions conducted by Infante-Rivard (2004) assessed polymorphisms in genes coding for
CYP2E1 and 5,10-methylenetetrahydrofolate reductase and observed elevated risk for SGA
associated with the high THM4 exposure category (above the 90th percentile, corresponding to >
29.4 (J,g/L) only among those with the CYP2E1 variant, reflecting a potential genetic
susceptibility.
In their retrospective cohort study, Summerhayes et al. (2012) assessed interactions between
THM exposure and socioeconomic indicators and smoking status. They did not observe evidence
of interaction between socioeconomic status and THM exposure, but did find statistical evidence
of an interaction between smoking and THM exposure. They observed generally larger
associations between THM and SGA among infants of non-smoking mothers and weaker (i.e.,
RR estimates < 1) in infants born to smoking mothers. Interestingly, smokers also had higher
levels of THM exposure, on average, relative to non-smokers.
The differences observed in estimates of SGA risk within strata of genotype and smoking status
are noteworthy, but should be interpreted conservatively. These are novel findings but more
research is needed to elucidate whether increased risks may occur in susceptible populations.
Pre-Term Delivery
Information Available During Development of Stage 1 and Stage 2 D/DBPRs
For the Stage 1 and Stage 2 D/DBPRs, the epidemiology evidence base regarding the association
between DBP exposure and the pre-term delivery (PTD) consisted of 10 primary studies
(including 5 cross-sectional studies, 4 cohort studies and 1 case-control study) and 6 review
papers, identified in Exhibit A.3.
Exhibit A.3: Studies of Pre-Term Delivery Outcomes Evaluated for Stage 1 and/or
Stage 2 D/DBPRs
Study
Developmental/Reproductive
Health Outcome
Study Design
Savitz et al. (2005)
PTD
Prospective Cohort
Six-Year Review 3 Technical Support Document A-38
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Wright et al. (2004)
PTD
Cross-sectional
Wright et al. (2003)
PTD
Cross-sectional
Yang (2004)
PTD
Cross-sectional
Jaakkola et al. (2001)
PTD
Cross-sectional
Dodds et al. (1999)
PTD
Retrospective Cohort
Gallagher et al. (1998)
PTD
Retrospective Cohort
Kanitz et al. (1996)
PTD
Cross-sectional
Savitz et al. (1995)
PTD
Prospective Cohort
Kramer et al. (1992)
PTD
Case-control
Bove et al. (2002)
PTD
Review
Graves et al. (2001)
PTD
Review
Villanueva et al. (2001)
PTD
Review
Reif et al. (2000)
PTD
Review
Craun, ed. (1998)
PTD
Review
Reif etal. (1996)
PTD
Review
Abbreviations: PTD - Pre-Term Delivery.
The results from this collection of studies did not provide much evidence of a deleterious effect
of DBP on PTD risk. Three studies (Wright et al., 2004; Savitz et al., 1995; Jaakkola et al., 2001)
actually observed an inverse relationship between DBP (exposure to chlorinated water in the
study by Jaakkola et al., 2001) and risk of PTD - higher THM4 levels were associated with lower
risk of PTD.
Three studies evaluated PTD and method of drinking water disinfection, one of which found
some evidence of a positive association. Jaakkola et al. (2001) assessed maternal exposure to
chlorinated drinking water (and water color) during pregnancy in a cross-sectional study in
Norway and found a reduced risk of PTD among a subgroup of individuals exposed to
chlorinated water who also have water with high color content. Kanitz et al. (1996) assessed
drinking water disinfection method (chlorine dioxide, sodium hypochlorite and chlorine
dioxide/sodium hypochlorite) in a cross-sectional study conducted in Italy and found no
association between risk of PTD and disinfection method.
In contrast, Yang (2004) compared the prevalence of PTD in 113 municipalities supplied with
chlorinated drinking water to that of 15 areas that were not supplied with chlorinated drinking
water (but did not estimate DBP levels in drinking water) in a cross sectional study in Taiwan.
The author reported OR for PTD of 1.37 (95% CI: 1.20-1.56) for chlorinating versus non-
chlorinating drinking water areas and stated the results suggest there is an association between
the consumption of chlorinated drinking water and PTD risk.
The studies evaluating the risk of PTD and estimated THM exposures during pregnancy
generally did not observe any positive associations. These include Dodds et al. (1999), who
estimated THM4 exposure during pregnancy among a cohort of women in Nova Scotia, and did
not observe evidence of an association between THM exposure and risk of PTD; Wright et al.
Six-Year Review 3 Technical Support Document A-39
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
(2003) who estimated THM4 maternal exposures during pregnancy and for each trimester in a
retrospective cohort study in Massachusetts and observe no statistically significant associations
between second trimester and entire-pregnancy average THM4 levels and PTD; Gallagher et al.
(1998) in their cohort of pregnant women in Colorado did not observe any associations between
estimated third-trimester THM4 levels in drinking water and PTD; and Savitz et al. (1995) who
estimated maternal THM4 exposure in drinking water in a case-control study set in North
Carolina and, again, found no association with PTD. Similarly, Kramer et al. (1992) observed no
associations between estimated chloroform, BDCM, DBCM and bromoform levels in drinking
water and "prematurity" in their case-control study set in Iowa.
Three studies provide evidence for an inverse association between DBP and PTD. Savitz et al.
(2005) estimated THM4, HAA9 and TOX exposures as well as individual BrTHM and HAA
species during pregnancy in a prospective cohort study of three communities in the United States
and observed a weak, non-statistically significant inverse relationship between PTD and THM4.
Wright et al. (2004) estimated THM4, chloroform, BDCM, total HAA, DCA and TCAA levels
in a large cross-sectional study of maternal third-trimester drinking water exposure and birth
weight in Massachusetts. They, too, observed a reduced risk of PTD associated with increasing
THM exposures; they observed no relationship between PTD and HAAs.
None of the review papers concluded that the weight of evidence was suggestive of a causal
relationship between DBP exposure or exposure to chlorinated drinking water during pregnancy
and risk of PTD.
New Information Available Since Development of Stage 2 D/DBPR
EPA conducted a literature search to identify new epidemiology studies of DBP and PTD that
became available subsequent to the promulgation of the Stage 2 D/DBPR. Eleven studies of DBP
associations with PTD were identified and evaluated: three prospective birth cohort studies, five
retrospective cohort studies, two case-control studies and one meta-analysis.
•	Prospective cohort studies:
o Hoffman et al. (2008b)
o Patelarou et al. (2011)
o Costet et al. (2012) (also implemented a case-control sampling design)
•	Retrospective cohort studies:
o Hinckley et al. (2005)
o Yang et al. (2007)
o Horton et al. (2011)
o Kumar et al. (2013)
o Rivera-Nunez and Wright (2013)
•	Case-control studies:
o Aggazzotti et al. (2004)
o Lewis et al. (2007)
•	Meta-analysis studies:
o Grellier et al. (2010)
Six-Year Review 3 Technical Support Document A-40
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Five of these studies were conducted in the United States: Arizona (Hinckley et al., 2005),
Massachusetts (Lewis et al., 2007; Rivera-Nunez and Wright, 2013), New York (Kumar et al.,
2013), "two Southern U.S. communities" (Horton et al., 2011) and "three U.S. communities"
(Hoffman et al., 2008b); three were conducted in Europe: Italy (Aggazzotti et al., 2004), France
(Costet et al. 2012) and Crete (Patelarou et al., 2011); and one was conducted in Taiwan (Yang et
al., 2007). All of these reports also assessed other fetal growth endpoints.
The PTD endpoint was defined as a live birth occurring prior to 37 weeks gestation in all but one
of the studies; Kumar et al. (2013) defined pre-term births as live births with a gestational age of
37 weeks or less. Horton et al. (2011) and Hinckley et al. (2005) defined an additional endpoint,
very PTD, as birth at less than 32 weeks of gestation. Gestational age was derived from maternal
report of last menstrual period (Hinckley et al., 2005; Lewis et al., 2007; Kumar et al., 2013;
Patelarou et al., 2011) or estimated using ultrasound evaluation. Costet et al. (2012) and Hoffman
et al. (2008b) used a combination of these methods. For example, Hoffman et al. (2008b) derived
gestational age at birth using first trimester maternal report of date of last menstrual period,
which was corrected by ultrasound if the two estimates of gestational age differed greater than
one week. Gestational age from one study (Rivera-Nunez and Wright, 2013) was based on
clinician estimates, and the methodology used to estimate gestational age was not specified in
three of the reports (Horton et al., 2011; Yang et al., 2007; Aggazzotti et al., 2004).
One study sampled tap water from women's homes in order to estimate DBP exposure
(Aggazzotti et al., 2004), and one study assessed DBP in both representative locations of
municipal water systems and in women's homes (Patelarou et al., 2011). In the remaining eight
studies, DBP exposure was estimated in water sampled from various locations in municipal
water networks; water sampling was conducted weekly or biweekly in three studies (Horton et
al., 2011; Hoffman et al., 2008b; Lewis et al., 2007), quarterly in three studies (Rivera-Nunez
and Wright, 2013; Yang et al., 2007; Hinckley et al., 2005) and at varying intervals in two
studies (Kumar et al., 2013; Costet et al., 2012).
The number of sampling sites used to assess DBP levels varied by study, but most studies
aggregate exposure averages across all sampling sites. Four of 11 studies queried study
participants' beverage consumption and water use behaviors (Costet et al., 2012; Patelarou et al.,
2011; Hoffman et al., 2008b; Aggazzotti et al., 2004). Other than one by Patelarou et al. (2011)
which assessed only BrTHMs, all of the studies assessed associations between PTD and THM4;
five studies additionally assessed specific THM concentrations (Rivera-Nunez and Wright 2013;
Costet et al., 2012; Hoffman et al. 2008b; Hinckley et al. 2005; Aggazzotti et al., 2004), and
Rivera-Nunez and Wright (2013) additionally assessed total BrTHMs. Four studies also assessed
HAA5 or HAA9 (Rivera-Nunez and Wright, 2013; Horton et al., 2011; Hoffman et al., 2008b;
Hinckley et al. 2005), and three studies assessed specific HAA exposures (Rivera-Nunez and
Wright, 2013; Hoffman et al., 2008b; Hinckley et al. 2005). Rivera-Nunez and Wright (2013)
additionally examined a DBP9 metric summing HAA5 and THM4 exposures. Aggazzotti et al.
(2004) additionally assessed chlorite and chlorate levels. All of the studies estimated tap water
DBP concentrations in relation to PTD. Four of the 10 studies (Costet et al., 2012; Patelarou et
al., 2011; Hoffman et al., 2008b; Aggazzotti et al., 2004) combined DBP measures with
assessments of individual water use behaviors to estimate personal exposures, while the
remainder used only the aggregate DBP measures to estimate exposure. Exposures were
Six-Year Review 3 Technical Support Document A-41
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
typically quantified into categories (e.g., quantiles), but several studies also assessed associations
between birth weight outcomes and continuously distributed DBP exposure metrics.
An exposure-response trend between THM4 and PTD risk was detected in the study by Yang et
al. (2007), albeit in municipalities with very low THM levels. Horton et al. (2011) found no
statistically significant associations between pulmonary tuberculosis (PTB) and THM4 or
HAA5, but did observe a linear exposure-response trend and statistically significant elevations in
the odds of PTB associated with increasing total organic halide exposures, although the
association was only observed among women served by the water system with higher
concentrations of bromine-containing DBPs. Two studies (Kumar et al., 2013 and Rivera-Nunez;
Wright, 2013) reported some statistically significant associations between DBP and PTD,
although neither reported evidence of linear exposure-response trends. Relative to the lowest
quintile of the respective DBP, Rivera-Nunez and Wright (2013) found there was some
suggestion of associations between PTD and some DBP metrics. For example, among the highest
HAA quartile (OR = 1.13; 95% CI: 0.95 to 1.33) which seemed to be largely attributable to
DCAA quartile exposures (OR = 1.14; 95% CI: 1.03 to 1.26). Similar ORs observed in the odds
of PTD with increased levels of other summary DBP indicators and specific THM and HAA
were not statistically significant. Evidence of increasing PTD risk with increasing level of
estimated exposure (i.e., no linear exposure response) was not observed for any of the DBP
measures. Kumar et al. (2013) categorized THM4 into five groups and observed increased odds
of PTD associated with the second, third and fifth categories of estimated THM4 exposure,
relative to the lowest category, but did not observe a consistent exposure-response trend; the
greatest increase in odds (14 percent) was observed for the second lowest category of THM4 and
a statistically significant protective association (i.e., OR
-------
Congenital Anomalies
Information Available During Development of Stage 1 and Stage 2 D/DBPRs
For the Stage 1 and Stage 2 D/DBPRs, the epidemiology evidence base regarding the association
between DBP exposure and congenital anomalies consisted of 11 primary studies (including 3
cross-sectional studies, 3 cohort studies and 5 case-control studies) and 9 review or meta-analysis
papers, identified in Exhibit A.4 (USEPA, 2005g).
Exhibit A.4: Studies of Congenital Anomaly Outcomes Evaluated for Stage 1
and/or 2 D/DBPRs
Study
Developmental/Reproductive Health Outcome
Study Design
Shaw et al. f2003)
Neural Tube Defects, Oral Clefts, Heart Defects
Case-control
Cedergren et al.
(2002)
Heart Defects
Case-control
Hwang et al. (2002)
Neural Tube Defects, Oral Clefts, Heart Defects,
Respiratory System Defects, Urinary Tract
Defects
Cross-sectional
Dodds and King
(2001)
Neural Tube Defects, Oral Clefts, Heart Defects,
Chromosomal Abnormalities
Retrospective
Cohort
Kallen and Robert
(2000)
Congenital Malformations
Cross-sectional
Dodds et al. ("1999)
Neural Tube Defects, Oral Clefts, Heart Defects,
Chromosomal Abnormalities
Retrospective
Cohort
Klotz and Pyrch (1999)
Neural Tube Defects
Case-control
Magnus et al. (1999)
Neural Tube Defects, Oral Clefts, Heart Defects,
Respiratory System Defects, Urinary Tract
Defects
Retrospective
Cohort
Bove et al. f1995)
Neural Tube Defects, Oral Clefts, Heart Defects,
CNS Defects
Cross-sectional
Aschengrau et al.
(1993)
Congenital Anomalies
Case-control
Shaw et al. (1991)
Heart Defects
Case-control
Hwang and Jakkola
(2003)
Congenital Anomalies
Meta-analysis
Bove et al. (2002)
Congenital Anomalies
Review
Graves et al. (2001)
Congenital Anomalies
Review
Villanueva et al.
(2001)
Congenital Anomalies
Review
Nieuwenhuijsen et al.
(2000)
Congenital Anomalies
Review
Reif et al. (2000)
Congenital Anomalies
Review
Six-Year Review 3 Technical Support Document A-43
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Study
Developmental/Reproductive Health Outcome
Study Design
WHO (2000)
Congenital Anomalies
Review
Craun, ed. (1998)
Congenital Anomalies
Review
Reif et al. (1996)
Congenital Anomalies
Review
Abbreviations: CNS - Central Nervous System
The results from this collection of studies did not provide strong or consistent evidence of an
association between exposure to chlorinated water or DBP and birth defects. Although by no
means consistent, the evidence was stronger for an association between DBP and neural tube
defects, as evidenced in several of the original scientific papers summarized below as well as in
several of the review papers.
Four studies evaluated risk of congenital defects and method of drinking water disinfection, three
of which found at least some evidence of positive associations. An increased risk of urinary tract
and respiratory tract defects was found to be associated with chlorinated water, though other
major congenital malformations showed no association with water source or type of water
treatment (chlorination and chloramination) in a case-control study by Aschengrau et al. (1993)
from Massachusetts.
Hwang et al. (2002) conducted a large cross-sectional study in Norway, comparing exposures to
chlorinated water (and also water color levels) for mother's residence during pregnancy and risk
of neural tube defects and defects of the heart, respiratory system, oral cleft and urinary tract.
They observed associations between risk of "any birth defect", as well as cardiac, respiratory
system and urinary tract defects and exposure to chlorinated water. In contrast, Kallen and
Robert (2000) assessed drinking water disinfection method (no chlorine, chlorine dioxide,
sodium hypochlorite) in a cross-sectional study conducted in Sweden and found no associations
with prevalence of congenital defects. Magnus et al. (1999) compared presence of chlorinated
water in mothers' residences at the time of birth and neural tube defects, as well as defects of the
heard, respiratory system, urinary tract and oral cleft. They observed statistically significant
associations only between urinary tract defects and chlorination; associations were not observed
for other outcomes or all birth defects combined.
The studies evaluating the risk of birth defects and estimated THM exposures during pregnancy
remain inconsistent. Bove et al. (1995) assessed prevalence of neural tube defects, oral cleft,
central nervous system and major heart defects and observed small but statistically significant
increased risks associated with higher THM4 levels for neural tube defects, central nervous
system defects, oral cleft defects and heart defects. Klotz and Pyrch (1999) also observed an
association between highest and lowest tertile THM4 exposure levels of pregnant mothers and
subsequent risk of neural tube defects (OR = 1.6; 95 % CI: 0.9-2.7). They also reported highest
to lowest tertile results for HAA (OR = 1.2; 95 % CI: 0.5-2.6) and for HAN (OR = 1.3; 95 % CI:
0.6-2.5) which they described as "showing little relation to these defects." In a retrospective
cohort study of THM4 levels among pregnant women living in Nova Scotia and subsequent risk
of neural tube defects, oral clefts, heart defects, Dodds et al. (1999) did not observe any evidence
of associations. They did note a non-statistically significant association between THM4 and
chromosomal abnormalities. In another retrospective cohort study set in Nova Scotia, Dodds and
Six-Year Review 3 Technical Support Document A-44
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
King, (2001) evaluated associations between estimated THM, chloroform and BDCM exposure
and neural tube defects, oral clefts, heart defects and chromosomal abnormalities. Only estimated
exposure to BDCM was found to be associated with increased risk of neural tube defects and
cardiovascular anomalies. Chloroform was found to be associated with chromosomal
abnormalities.
Three studies focused on associations between DBP and overall heart defects. Cedergren et al.
(2002) examined DBP levels in the period from before inception through early pregnancy in a
Swedish case-control study. Although they identified ten specific types of cardiac defects, their
analysis focused on "any cardiac defect". They observed a statistically significant association
between chlorine dioxide in drinking water and heart defects. They also found that THM
concentrations equal to or greater than 10 [j,g/L were significantly associated with heart defects.
They did not observe any association between cardiac defects and nitrate. In two case-control
studies, however, Shaw et al. (2003) estimated THM in mothers' residences during a similar
peri-conceptional period and did not find associations or exposure-response relationships
between THM4s and conotruncal heart defects in either study. The studies were similarly
negative for neural tube defects and oral clefts. Similarly, Shaw et al., (1990, 1991) observed no
associations between cardiac anomalies and THM4 level in an earlier case-control study.
Five of the reviews/meta-analyses concluded that the evidence base evaluated provides at least
some support for an effect of DBP exposure on risk of neural tube defects (Hwang and Jakkola,
2003; Bove et al. 2002; Villanueva et al., 2001; WHO, 2000; Reif et al., 1999; Graves et al.,
2001) concluded that the findings regarding neural tube defects were inconsistent. Two reviews
(Hwang and Jakkola, 2003; Graves et al., 2001) also concluded that the evidence base supported
an association between DBP exposure and urinary system defects. Evidence of relationships
between DBP exposure and birth defects, especially for those not mentioned above, was largely
considered inconsistent, weak, insufficient and not convincing in the reviews.
New Information Available Since Development of Stage 2 D/DBPR
EPA conducted a literature search to identify new epidemiology studies of DBP and risk of
congenital anomalies that became available subsequent to the promulgation of the Stage 2
D/DBPR. Eight studies of DBP associations with congenital anomalies were identified and
evaluated: one prospective birth cohort study, three case-control studies, one cross-sectional
study, two ecological studies and two meta-analysis studies.
•	Prospective cohort studies:
o Grazuleviciene et al. (2013)
•	Case-control studies:
o Righi et al. (2012)
o Iszatt et al. (2011)
o Luben et al. (2008)
•	Cross-sectional studies:
o Hwang et al. (2008)
•	Ecologic studies:
o Chisholm et al. (2008)
o Nieuwenhuijsen et al. (2008)
Six-Year Review 3 Technical Support Document A-45
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
• Meta-analyses:
o Nieuwenhuijsen et al. (2009)
o Hwang et al. (2008)
The Luben et al. (2008) study was conducted in Arkansas; four of the studies were conducted in
Europe (Iszatt et al., 2011 and Nieuwenhuijsen et al., 2008 in the U.K., Righi et al., 2012 in Italy,
and Grazuleviciene et al., 2013 in Lithuania) and one each in Taiwan (Hwang et al., 2008) and
Australia (Chisholm et al., 2008).
The assessments of congenital anomalies in this literature were universally restricted to live
births and implemented using medical records or registry databases. The endpoints assessed in
these studies were defined variously as the occurrence of a specific anomaly (e.g., hypospadias,
cleft lip, spina bifida), the occurrence of any one of a group of anomalies (e.g., heart,
musculoskeletal, urogenital, neural tube defects), or as occurrence of "any" congenital anomaly.
Several of the studies that assessed more specific outcomes also assessed more broad categories.
All of the studies assessed THM4, and three studies (Grazuleviciene et al., 2013; Iszatt et al.,
2011; Chisholm et al., 2008) also assessed specific THM levels. Two of the studies additionally
assessed total BrTHM (Iszatt et al., 2011; Nieuwenhuijsen et al., 2008), and Luben et al. (2008)
assessed specific HAA and HAA5. Righi et al. (2012) evaluated chlorite and chlorate in addition
to THM4.
Water sampled from locations in municipal water distribution systems were used to estimated
DBP exposure in all seven observational studies reviewed. Two of the studies (Grazuleviciene et
al., 2013; Iszatt et al., 2011) combined DBP data with water use behaviors to estimate individual
DBP intake. Luben et al. (2008) assessed individual exposure in a subgroup of the study
population. Water sampling was conducted quarterly in five studies and two studies assessed
DBP measurements for the entire pregnancy using sampling frequency that was not specified.
The number of sampling sites used to assess DBP levels also varied by study.
Four studies assessed exposure during the entire pregnancy (Grazuleviciene et al., 2013; Iszatt et
al., 2011; Chisholm et al., 2008; Hwang et al., 2008); Grazuleviciene et al. (2013) also assessed
trimester-specific exposures and month-specific exposures. The Luben et al. (2008) study of
hypospadias specifically assessed exposure between weeks 6 and 16 of gestation. Four studies
assessed only first trimester DBP exposure (Nieuwenhuijsen et al., 2008; Iszatt et al., 2011;
Righi etal., 2012).
Both Chisholm et al. (2008) and Hwang et al. (2008) assessed an 'any congenital anomaly'.
Chisholm et al. (2008) observed a statistically significant OR (1.22; 95% CI: 1.01-1.48) relating
the presence of 'any' congenital anomaly to THM4 exposures; the association was observed for
those in the 'high' level of THM4 (> 130 (J,g/L), relative to those in the lowest of three categories
of THM4 exposure (< 60 (J,g/L). However, women in the middle category had a slightly lower
risk of having a child with any congenital anomaly, relative to women in the lowest category of
THM4 (i.e., no exposure-response trend). Hwang et al. (2008) observed an elevated OR (1.21;
95% CI: 1.07-1.36) among women with low THM4 (5-9 (J,g/L) relative to the reference group of
women THM4 (0-4 (J,g/L), but not among women with higher THM4 levels (>10 (J,g/L).
Six-Year Review 3 Technical Support Document A-46
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Associations between THMs and cardiovascular anomalies were noted in four of the five studies
which assessed them (Grazuleviciene et al., 2013; Nieuwenhuijsen et al., 2008; Chisholm et al.,
2008; Hwang et al., 2008), with associations with ventricular septal defects consistently observed
across the three studies which included this specific endpoint. Grazuleviciene et al. (2013)
assessed cardiac anomalies as a group, but did not specifically assess ventricular septal defects.
Hwang et al. (2008) observed an elevated OR (1.81; 95% CI: 0.98-3.35) for ventricular septal
defects only among those in the highest category of THM4 exposure (>20 (J,g/L), with no
evidence of an exposure-response relationship. In an included meta-analysis, Hwang et al. (2008)
noted ventricular septal defects as the only individual birth defect group to have a statistically
significant OR in relation to THM4 exposure (OR: 1.59; 95% CI: 1.21-2.07). In the Hwang et al.
(2008) study, the highest ORs for atrial septal defects (2.15; 95% CI: 0.70-6.60) and Tetralogy of
Fallot (1.60; 95% CI: 0.61-4.23) were observed in the low THM4 category (5-9 (J,g/L). Chisholm
et al. (2008) also observed a statistically significant increase in the odds of elevated
cardiovascular anomalies in the highest THM4 exposure category. Grazuleviciene et al. (2013)
also found evidence for an association between cardiovascular defects and high THM4 water
concentration exposures (OR: 1.54; 95% CI: 0.89-2.68). They detected ORs for congenital heart
anomalies in excess of 1.35 for all of the highest first trimester internal dose tertiles for THM4,
chloroform, BDCM and DBCM and some evidence of an exposure-response relationship with
increasing ORs across BDCM tertiles. Although their study results were largely null, with no
statistically significant trends across THM exposure categories for either their broadly defined or
more restricted sets of anomalies. Nieuwenhuijsen et al. (2008) did observe an association
between high level of THM4 exposure (> 60 (J,g/L) and ventricular septal defects (OR: 1.43; 95%
CI: 1.00-2.04) as well as between high bromoform exposure (> 4 (J,g/L) and major
cardiovascular defects (OR: 1.18; 95% CI: 1.00-1.39) and also for gastroschisis (OR: 1.38; 95%
CI: 1.00 - 1.92), an abdominal wall defect.
Chisholm et al. (2008) observed elevated odds of musculoskeletal and urogenital defects among
those in the highest category of THM4, although these odds ratios were not statistically
significant. They did not observe similarly elevated odds for integument congenital anomalies,
respiratory system defects, or nervous system defects. THM4 in this study were not assessed in a
way that was specific to a critical or biologically relevant time window of exposure. Other than
the findings mentioned above, Nieuwenhuijsen et al. (2008) observed no associations between
DBP exposure in the first 93 days of pregnancy (THM4, total BrTHM, or bromoform) and any of
the other broadly defined or more restricted sets of anomalies they assessed. Grazuleviciene et al.
(2013) reported statistically significant exposure-response relationships between congenital
musculoskeletal anomalies and DBCM tertiles based on the first and second month exposure
window (OR range 1.41 to 2.90) and for congenital urogenital anomalies based on internal dose
BDCM first trimester tertiles. Women in the highest category of THM4 exposure in the Hwang
et al. (2008) study had elevated odds of having a child with cleft palate (OR: 1.56; 95% CL1.00-
2.41). ORs for urinary tract defects were elevated across all THM4 categories (range: 1.24-1.65)
in the Hwang et al. (2008) study, with the biggest increase seen for those in the low exposure
category (5-9 (J,g/L), relative to the reference group (0-4 (J,g/L). None of the odds ratios for the
urinary tract defect endpoint achieved statistical significance.
Although Righi et al. (2012) found little evidence of associations between first trimester THM
exposures, which were generally low (3.8±3.6 (J,g/L), and birth defects, they did observe
associations between high chlorite exposure and relative odds of renal defects, abdominal wall
Six-Year Review 3 Technical Support Document A-47
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
defects and cleft palate, relative to those with the lowest category of chlorite exposure. They also
detected associations between high chlorate exposure and relative odds of obstructive urinary
defects, cleft palate and spina bifida, relative to those with the lowest category of chlorate
exposure. An exposure-response relationship was also observed for musculoskeletal anomalies
and DBCM exposure during the first and second months of pregnancy. The studies examining
the risk of hypospadias associated with THM4 (Iszatt et al., 2011; Luben et al., 2008; Hwang et
al., 2008) were all negative.
Nieuwenhuijsen et al. (2009) conducted a meta-analysis of 15 studies of DBP exposure and risk
of congenital anomalies, including four studies reviewed for this report (Nieuwenhuijsen et al.,
2008; Chisholm et al., 2008; Luben et al., 2008; Hwang et al., 2008). They found 17 percent
excess risk of all congenital anomalies combined (95% CI: 3-34 percent), comparing low
exposure to water chlorination or THM4 and a statistically significant excess risk of 58 percent
(95% CI: 21-107 percent) for ventricular septal defects. The authors did not observe evidence of
an exposure response relationship, and the finding was based on only three studies.
Nieuwenhuijsen et al. (2009) conducted separate meta-analyses for categories of birth defects
and specific anomaly endpoints if greater than two studies evaluated the same exposure index-
congenital anomaly relationship, including the following: nervous system defects including
neural tube defects, anencephalus, hydrocephalus, spina bifida, major cardiac defects, respiratory
defects, oral cleft or cleft palate defects, cleft palate only, urinary tract defects, obstructive
urinary defects and hypospadias. They observed no statistically significant relationships in these
other meta-analyses. Although not statistically significant, they observed increases in the
summary RRs for major cardiac defects (RR: 1.16; 95% CI: 0.98-1.37) and urinary tract defects
(RR: 1.33; 95% CI: 0.92-1.92) comparing high to low chlorination by-product exposure.
Fetal Loss
Information Available During Development of Stage 1 and Stage 2 D/DBPRs
For the Stage 1 and Stage 2 D/DBPRs, the epidemiology evidence base regarding the association
between DBP exposure and fetal loss consisted of 10 primary studies (including 2 cross-sectional
studies, 4 cohort studies and 4 case-control studies) and 9 review papers, identified in Exhibit
A.5 (USEPA, 2005g).
Exhibit A.5: Studies of Fetal Loss Outcomes Evaluated for Stage 1 and/or 2
D/DBPRs
Study
Developmental/Reproductive
Health Outcome
Study Design
Savitz et al. (2005)
Early and Late Pregnancy Loss
Prospective Cohort
Toledano et al. (2005)
Stillbirth
Cross-sectional
Dodds et al. (2004)
Stillbirth
Case-control
Dodds et al. (1999)
Stillbirth
Retrospective Cohort
Swan et al. (1998)
Spontaneous Abortion
Prospective Cohort
Six-Year Review 3 Technical Support Document A-48
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Study
Developmental/Reproductive
Health Outcome
Study Design
Waller et al. (1998)
Spontaneous Abortion
Prospective Cohort
Bove et al. (1995)
Fetal Deaths
Cross-sectional
Savitz et al. (1995)
Spontaneous Abortion
Case-control
Aschengrau et al. (1993)
Neonatal Death, Stillbirth
Case-control
Aschengrau et al. (1989)
Spontaneous Abortion
Case-control
Bove et al. (2002)
Spontaneous
Abortion, Fetal Death
Review
Graves et al. (2001)
Neonatal Death, Fetal Resorption
Review
Villanueva et al. (2001)
Spontaneous Abortion
Review
Nieuwenhuijsen et al. (2000)
Spontaneous Abortion, Stillbirth
Review
Reif et al. (2000)
Spontaneous Abortion, Stillbirth
Review
WHO (2000)
Miscarriage
Review
Craun, ed. (1998)
Stillbirth, Neonatal Death,
Spontaneous Abortion
Review
Mills etal. (1998)
Spontaneous Abortion
Review
Reif et al. (1996)
Stillbirth, Neonatal Death,
Spontaneous Abortion
Review
The results from this collection of studies provided relatively consistent evidence of an
association between exposure to chlorinated water or DBP and pregnancy loss.
Three studies evaluated exposure to disinfected water or water source (as opposed to evaluating
DBP levels in the water), all of which found some evidence of positive associations between
exposure to disinfected water and risk of pregnancy loss. A case-control study by Aschengrau et
al. (1989) set in Massachusetts evaluated water source (surface water versus other) among
pregnant women and observed a statistically significantly association between having a surface
water source and frequency of spontaneous abortion. A subsequent case-control study by
Aschengrau et al. (1993), also set in Massachusetts, evaluated neonatal death and stillbirth by
water source and two types of disinfection methods (chlorination or chloramination) found a
non-statistically significant increase in the prevalence of stillbirths among participants with
exposure to chlorinated (versus chloraminated) surface water. However, neonatal death was not
found to be associated with water source or disinfection method. Swan et al. (1998) compared
consumption of tap water and bottled water during early pregnancy in a cohort of women living
in three different locations in California and observed a statistically significant increase in the
frequency of spontaneous abortion at one of the three sites.
Bove et al. (1995) estimated maternal THM4 exposure in drinking water in a cross-sectional
study in New Jersey and did not find association with fetal deaths. However, many of the studies
reviewed did find associations between estimated THM exposure and pregnancy loss. Waller et
Six-Year Review 3 Technical Support Document A-49
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
al. (1998) conducted a prospective cohort study of pregnant women in California and found that
high estimated THM4 exposure (via ingestion and showering) during the first trimester of
pregnancy was associated with a statistically significant increase in the risk of spontaneous
abortion, compared to low levels of estimated THM4 intake. They also observed an exposure-
response relationship between estimated THM4 ingested and spontaneous abortion. In a
retrospective cohort study conducted by Dodds et al. (1999) in Nova Scotia, stillbirth was again
found to be statistically significantly associated with THM4, and also with specific THMs, with
higher risks observed among asphyxia-related stillbirths. In a subsequent case-control study
conducted by Dodds et al. (2004), a statistically significant association between stillbirth and
exposure to THM4, BDCM and chloroform was observed among women living in Nova Scotia
and Eastern Ontario. Toledano et al. (2005) conducted what they characterized as a "large cross-
sectional study" in England, modelling THM4 levels in three water zones and compared those
estimates to rates of stillbirth. They found a statistically significant association between THM4
and risk of stillbirth in one of the three regions (OR = 1.21, 95% CI: 1.03-1.42), although when
all three regions were combined, the elevation in risk of stillbirth was small and borderline
statistically significant (OR = 1.11; 95% CI: 1.00-1.23). Risks were also elevated for chloroform,
but no associations were observed between risk of still birth and BDCM or total BrTHMs. Savitz
et al. (2005) estimated THM4, HAA9 and TOX exposures as well as individual BrTHM and
HAA species during pregnancy in a prospective cohort study of three communities in the United
States, comparing them to risk of pregnancy loss. They did not observe an association when high
THM4 exposures were compared to low exposures. However, they did observe a statistically
significant association between BDCM and pregnancy loss (OR = 1.58; 95% CI: 1.03-2.41).
Although non-statistically significant, an increased risk of similar magnitude was seen between
DBCM and pregnancy loss (OR = 1.30; 95% CI: 0.82-2.05). They also noted increased risks
associated with pregnancy losses at greater than 12 weeks gestation for THM4, BDCM and
TOX, but concluded that most results generally did not provide support for an association. In an
earlier case-control study of THM4 concentration at the homes of pregnant women and estimated
THM4 intake set in North Carolina, Savitz et al. (1995) found a statistically significant increase
in the risk of miscarriage comparing high to low THM4 concentration, but not when comparing
THM4 intake (THM4 concentration x amount of water consumption).
Four of the reviews concluded that the evidence base evaluated provides at least some support
for an association between DBP exposure on risk of spontaneous abortion and fetal
death/spontaneous abortion (Bove et al., 2002; Villanueva et al., 2001; WHO, 2000; Mills et al.,
1998). Graves et al. (2001) concluded that there was no support for a relationship between DBP
exposure and neonatal death. Nieuwenhuijsen et al. (2000) found the evidence supporting an
association between THM exposure and spontaneous abortions/stillbirths to be weak. Craun, ed.
(1998) concluded that although some associations have been observed in epidemiologic studies,
the results do not constitute convincing evidence of a causal relationship between DBP and
stillbirth, spontaneous abortion and neonatal death. Similarly, Reif et al. (1996, 2000) concluded
that the evidence is inadequate for establishing a relationship between spontaneous abortion and
DBP exposure.
Six-Year Review 3 Technical Support Document A-50
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
New Information Available Since Development of Stage 2 D/DBPR
EPA conducted a literature search to identify new epidemiology studies of DBP and risk of fetal
loss that became available subsequent to the promulgation of the Stage 2 D/DBPR; and found
one study by Hwang and Jaakkola (2012).
Hwang and Jaakkola (2012) evaluated THM4 exposure among Taiwanese mothers of 3,289
stillbirths and 32,890 newborn control subjects in a case-control study. Water sampling was
conducted quarterly in five studies and two studies assessed DBP measurements for the entire
pregnancy using sampling frequency that was not specified. The number of sampling sites used
to assess DBP levels also varied by study. Water sampling frequency for THM4s was conducted,
at a minimum, four times per year for each water treatment plant. THM4 exposure was
calculated by calculating an average of the modeled quarterly THM4 estimates for the water
treatment plants serving each mother between the date of conception and the date of birth,
weighted by the proportion of the trimester falling into each quarterly period. Estimated THM4
exposures were categorized into four groups (0-4 [j.g/L (the reference category), 5-9 (J,g/L, 10-19
[j,g/L, 20+ (J,g/L). Covariate adjusted odds ratios were slightly elevated in the low (OR: 1.02; 95%
CI: 0.92-1.14), medium (OR: 1.10; 95% CI: 1.00-1.21) and high (OR: 1.06; 95%: 0.96-1.17)
categories. The authors also presented a meta-analytic summary odds ratio incorporating the
results of previous studies with their study and noted that it provided consistent evidence of
increased risk, but showed some heterogeneity. The summary odds ratio estimated from a
random-effects model was 1.21 (95% CI: 1.02-1.43) and interpreted by the authors as providing
consistent evidence of an increased risk of stillbirth associated with THM4 exposure, although
there was statistically significant heterogeneity observed between studies.
Male Reproductive Effects
Information Available During Development of Stage 1 and Stage 2 D/DBPRs
For the Stage 1 and Stage 2 D/DBPRs, the epidemiology evidence base regarding the association
between DBP exposure and male reproductive effects consisted of a single study conducted in
California by Fenster et al. (2003) in which THM4 levels were estimated in drinking water
sampled within three months prior to semen collection. The investigators evaluated sperm
motility and sperm morphology, but found no associations with THM4 exposures.
Information Available Since Development of Stage 2 D/DBPR
EPA conducted a literature search to identify new epidemiology studies of DBP and male
reproductive endpoints that became available subsequent to the promulgation of the Stage 2
D/DBPR. Five studies of DBP associations with male reproductive outcomes published after the
Stage 2 D/DBPR Economic Analysis (EA) (2004-2013) were identified and evaluated: one
prospective birth cohort study, one case-control study and three cross-sectional studies (USEPA,
2005g).
• Prospective cohort studies:
o Luben et al. (2007)
Six-Year Review 3 Technical Support Document AS J
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
•	Case-control studies:
o Iszatt et al. (2013)
•	Cross-sectional studies:
o Zeng et al. (2013)
o Nickmilder and Bernard (2011)
o Xieetal. (2011)
The Luben et al. (2007) study was conducted in the US; the Zeng et al. (2013) and Xie et al.
(2011) studies were conducted in China, the Iszatt et al. (2013) study was conducted in the UK,
and the Nickmilder and Bernard (2011) study was conducted in Belgium.
Four of the five studies (Iszatt et al., 2013; Zeng et al., 2013; Xie et al., 2011; Luben et al., 2007)
included assessments of semen quality (e.g., (low) sperm count, sperm morphology (percent
normal sperm), low motile sperm concentration, percent of sperm with DNA fragmentation and
percent of immature sperm). Two studies assessed serum total testosterone levels (Zeng et al.,
2013; Nickmilder and Bernard, 2011). Nickmilder and Bernard (2011) additionally assessed
serum inhibin B levels. In all of the studies, a single semen sample was provided by each subject.
Urine (and blood samples, when collected) was also sampled once for each subject. Studies that
use a single-sample to represent average, typical or usual levels of a measurement (semen quality
in this context) rely on an (often only implicit) assumption that there is little intra-individual
variability in the measurement over time.
Only two of the five studies (Iszatt et al., 2013; Luben et al., 2007) assessed DBPs in municipal
water systems. Iszatt et al. (2013) used quarterly water samples and Luben et al. (2007) used
weekly or biweekly samples. Both studies attempted to estimate exposure in the 90 days prior to
collection of the semen sample. The study by Nickmilder and Bernard (2011) assessed time
spent in pools as a proxy for DBP exposure. The remaining two studies assessed biomarkers of
DBP exposure. Xie et al. (2011) assessed urinary creatinine-adjusted TCAA concentrations and
Zeng et al. (2013) assessed whole blood levels of THM. In these two biomarker studies, semen
collection and urine/blood collection occurred on the same day.
The results from these epidemiologic studies of the association between DBP exposure and male
reproductive outcomes were largely negative. Xie et al. (2011) assessed associations between
urinary creatinine-adjusted TCAA concentration as a biomarker of DBP exposure and sperm
quality and observed no statistically significant associations, nor clear evidence of exposure-
response trends. Zeng et al. (2013) assessed concentrations of THMs in blood and found no
associations with decrements in sperm motility, or sperm velocity. They did find that moderate
levels (above the level of detection, but below the median of observable values) of BDCM and
DBCM were associated with decreased sperm linearity, compared with levels below the level of
detection. Exposure-response relationships were observed between elevated blood chloroform
and THM4 concentration and decreased sperm concentration and between elevated blood DBCM
concentration and decreased serum total testosterone. Zeng et al. (2014) detected reductions in
sperm concentration with increasing BrTHM exposure levels (-0.26 (95% CI: -0.52, -0.01). Iszatt
et al. (2013) and Luben et al. (2007) found little evidence of an association between DBP
exposure and sperm quality parameters; however, sperm concentration reductions in Luben et al.
(-0.23; 95% CI: -0.54, 0.07) for the BrTHM exposure metric were nearly identical to those
observed by Zeng et al. (2014).
Six-Year Review 3 Technical Support Document A-52
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Nickmilder and Bernard (2011) assessed associations between sperm quality parameters and
cumulative swimming pool attendance time as a proxy for exposure to chlorination byproducts in
pool water among adolescent boys. It should be noted that swimming pools are a potential source
of both DBP exposure and exposure to the disinfectants themselves. As such, the interpretation
of observed associations is limited because the effects of DBP exposure and the effects of
exposure to the disinfectants themselves cannot be distinguished. They found that, among the
adolescents assessed in the study, inhibin B concentrations were inversely associated with time
spent in indoor chlorinated pools before the age of 10, while total and free testosterone
concentrations decreased with increasing amount of time spent in indoor chlorinated pools.
Among those that swam in indoor chlorinated pools for more than 250 hours before the age of
10, or for more than 125 hours before age 7, had an almost 3-fold increase in the risk of having
low (<10th percentile) serum inhibin B and/or total testosterone, relative to those who never
swam in indoor pools during childhood. The authors did not observe associations between
cumulative time spent in indoor pools and free testosterone, LH, follicle-stimulating hormone
and dehydroepiandrosterone-sulfate, nor was low serum testosterone or inhibin B associated with
attendance of outdoor chlorinated pools or those pools disinfected with copper-silver ionization.
Female Reproductive Effects
Information Available During Development of Stage 1 and Stage 2 D/DBPR
For the Stage 1 and Stage 2 D/DBPR, the epidemiology evidence base regarding the association
between DBP exposure and female reproductive effects consisted of a single prospective cohort
study in California. Windham et al. (2003) estimated THM exposure through two routes of
exposure: showering (dermal / inhalation) and ingestion of drinking water and found that THM
exposure may affect ovarian function. BrTHMs were statistically significantly associated with
shorter menstrual cycles, especially for dibromochloromethane (DBCM). They did not observe
strong or statistically significant association between THM4 exposure and luteal phase length,
menses length, or cycle variability.
New Information Available Since Development of Stage 2 D/DBPR
EPA conducted a literature search to identify new epidemiology studies of DBP and female
reproductive endpoints and identified one study by MacLehose et al. (2008). MacLehose et al.
(2008) evaluated the association between exposure to specific DBP trihalomethanes, HAAs,
brominated-trihalomethanes, brominated-HAAs, total organic halides and
bromodichloromethane) and time to pregnancy. DBP ingestion, inhalation and absorption while
bathing or showering were estimated among newly pregnant women enrolled in the Right From
the Start prospective cohort study of reproductive outcomes conducted in three metropolitan
areas of the US. Water samples were drawn from the distribution system at each site either
weekly (for two sites) or every other week (for the site with consistently low DBP levels. Spatial
variability of DBP levels was evaluated, and levels were found to be uniform. DBP
concentrations were assigned to each menstrual cycle during which each woman attempted to
conceive. The investigators calculated four metrics each for THM4, bromodichloromethane,
brominated-trihalomethanes, brominated-HAAs and total organic halides (TOX): tap water
concentration, amount ingested through drinking, absorbed DBP through inhalation and dermal
absorption while showering or bathing (for THM4, bromodichloromethane, brominated-
Six-Year Review 3 Technical Support Document A-53
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
trihalomethanes only), and "integrated measure" of THM4, bromodichloromethane, brominated-
trihalomethanes in the bloodstream through ingestion and showering or bathing. The authors
observed no evidence of an increased time to pregnancy among women with exposure to
increasing levels of DBP.
Additional Observations on the Reproductive and Developmental Effects Based on the
Epidemiology Evidence
Because of the extensive epidemiological information base addressing reproductive and
developmental effects related to exposure to chlorination by-products, we summarize in this
section the recurrent limitations that are evident in this epidemiologic literature, which include
statistical considerations and potential sources of information bias, selection bias and
confounding arising from study design choices, exposure assessment, health endpoint assessment
and covariate assessment. It should be noted that disinfected drinking water can contain hundreds
of disinfection byproducts. The measured exposures associated with reproductive health
endpoints in these studies may directly affect the risk of reproductive outcomes; they may also
be useful though imperfect indicators for the most relevant and potentially measured or
unmeasured causative DBP exposures. In these epidemiologic studies, DBP exposure was most
often assessed by quantifying THM4. Often, specific trihalomethanes, total HAAs, specific
HAAs were also assessed. Less commonly quantified were other DBPs or DBP mixture
surrogates such as total organic halides, chlorite and chlorate levels.
Recurrent patterns across the post-Stage 2 DBP studies include the following: 1) findings of
positive associations (e.g., RRs greater than 1) between DBP indicators and adverse reproductive
outcomes that were small in magnitude and sometimes null and 2) frequent absence of observed
linear exposure-response relationships. The lack of consistent results across many of the
aforementioned outcomes could have several possible explanations. Most of the studies reviewed
included statistical adjustment for important confounders, including gestational age, maternal
age, race, body mass index, marital status, smoking and parity, and comorbidities. Several
studies attempted to additionally adjust for socioeconomic risk factors including education,
income, health insurance and adequacy of prenatal care. However, the potential for residual
confounding to bias study results and explain some of these is also a possibility as not all of the
studies reviewed included adjustment for all of these risk factors for reproductive endpoints, and
some of these potential risk factors (e.g., socioeconomic status) are difficult to measure.
Exposure misclassification (especially if non-differential) is a plausible explanation for the
patterns noted above for the lack of inconsistent results or lack of exposure-response
relationships. The limited exposure data focused on only a few surrogates to represent very
complicated DBP mixture exposure scenarios might also explain some of the mixed study results
and lead to exposure misclassification of the ideal exposure metric or truly relevant (set of) DBP
exposure(s). In addition, the available data and exposure metrics often only represent a particular
exposure route or groups of DBPs and may not be able to fully evaluate the potential for
interactions to occur between DBPs and the adverse reproductive outcomes of interest.
The quality of the DBP exposure assessment in the epidemiologic literature reviewed for this
analysis ranged from adequate to very detailed. Many of the studies evaluated objective
measures of DBP levels in drinking water, assessed water consumption behaviors prior to birth
Six-Year Review 3 Technical Support Document A-54
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
and constructed detailed metrics of DBP exposure and dose. Often, indirect estimates of DBP
exposure were derived by linking (maternal) residence to water quality monitoring data. In
studies using this exposure assessment methodology, DBP levels were often spatially
aggregated. Such area-level assessments remain the most feasible procedure for categorizing
exposure to multiple DBPs in large epidemiologic studies. Some of the studies reviewed
implemented more refined exposure assessment that combined water quality monitoring data, or
other similar assessments of DBP in centralized locations within a water system, with personal
water usage information obtained from study participants. Such methodology has the potential to
decrease exposure misclassification. However, there was no clear evidence that these studies
were more likely to observe positive associations between DBP exposure and reproductive
endpoints, relative to studies that did not assess water use behaviors. Notably, several of the
studies reviewed assessed DBP exposure using a biomarker of exposure, albeit on a smaller scale
than many of the larger studies. Development of inexpensive yet sensitive and specific
biomarkers for DBP exposure has the potential to further minimize exposure misclassification. In
their assessment of maternal urinary creatinine-adjusted TCAA as a DBP exposure biomarker in
their study of effects on birth weight, Zhou et al. (2012), observed lower average birth weight
among infants whose mothers were in the highest two quartiles of creatinine-adjusted urinary
TCAA concentrations for the overall population. They saw even larger reductions among a
subset of women who had completed questionnaires which allowed for additional adjustment of
additional covariates that may have been confounders. The study of male reproductive endpoints
and urinary TCAA conducted by Xie et al. (2011) was negative. Because TCAA and the other
HAAs are non-volatile DBPs, it is unclear to what degree maternal urinary TCAA concentrations
are valid and accurate DBP surrogates in this population, especially for the volatile DBPs. TCAA
is not specific to DBP exposure and urinary TCAA levels could reflect exposure to other
environmental contaminants. In the study of male reproductive endpoints and THM levels in
blood conducted by Zeng et al. (2013), associations were observed between moderate levels of
blood BDCM and DBCM and decreased sperm count and declined sperm linearity compared
with low levels.
Suggestive exposure-response relationships of borderline statistical significance were observed
between elevated blood TCM concentrations and decreased sperm concentration and between
elevated blood DBCM concentration and decreased serum total testosterone. Blood THM levels
are likely a more specific biomarker of exposure to volatile DBPs (e.g., THMs) across different
exposure routes compared to urinary DBP measures. However, THMs are rapidly metabolized
and may best represent baseline THM levels. Therefore, most of these measures would not
reflect the impact of recent activities that may drive average or peak exposures during critical
exposure windows. Although analysis of urinary TCAA and whole blood THM levels is more
invasive, expensive and labor intensive, they are expected to be better exposure measures
compared to assessments of DBP health effects that rely on routinely-monitored DBP
measurements.
A biologically relevant time-window for exposure has been hypothesized for many of the
reproductive health endpoints investigated in these studies. This is the period of fetal
development during which it is thought that an exposure to a toxicant may exert its influence in
development of the outcome. Many of the studies assessed DBP exposure specific to specific
trimesters of pregnancy. Several studies of congenital anomaly endpoints and of sperm quality
carried out exposure assessments targeted to well-characterized time-windows for exposure
Six-Year Review 3 Technical Support Document A-55
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
effects. However, all of the cross-sectional studies as well as several of the cohort and case-
control studies reviewed in this analysis assessed exposure at either a single point in time
(including after the birth of the infant in a few studies), or alternatively characterized exposure
for the entire pregnancy. If a true association indeed exists between DBPs and adverse
reproductive outcomes, it is possible that no association would be observed even with a precisely
measured exposure if the exposure assessment occurs outside of the biologically relevant time
window for exposure effects. In a more plausible scenario, the observed measure of association
would be attenuated if exposures assessed outside of the critical time-window for exposure are
imperfectly correlated with exposures occurring during the most biologically relevant time
periods.
Despite efforts to minimize errors in DBP exposure assessment, a certain degree of exposure
misclassification remains inevitable in DBP health effects epidemiology. The impact of this is to
decrease the sensitivity of the study to detect associations that may exist. The DBP exposure
measurement error in these epidemiologic studies is likely to be predominantly non-differential
with respect to reproductive outcomes; the use of infrequent (e.g., quarterly) water sampling,
community level (as opposed to individual-level) exposure metrics and missing exposure data all
have the potential to induce bias towards an observed null association between estimated DBP
exposure and the adverse reproductive outcomes. Although they likely can provide a sense of the
relative exposure rankings for the predominant DBPs, community level exposure metrics based
on quarterly water sampling are not likely to fully capture the full extent of spatial and temporal
variability in DBP levels over the course of a pregnancy or even smaller critical windows (e.g., a
single trimester). However, observed associations between THM4 and reproductive outcomes
were null or small and often not statistically significant, even among the studies that conducted
more frequent sampling and those that implemented exposure assessment advancements (e.g.,
Hoffman et al., 2008a, Lewis et al., 2006; Patelarou et al., 2011).
Among the studies that assessed maternal water use activities, misclassification of exposures
(both total water intake and to DBPs) is also likely due, for example, to errors in recall of water
use and water consumption outside of the home. These errors are likely to be non-differential in
the studies evaluated, and therefore would, on average, attenuate observed estimates of
association toward the null. That being said, studies that query maternal water use and
consumption have the potential to generate more accurate DBP exposure estimates, relative to
studies that rely only on DBP levels measured in municipal water samples. For example, in the
assessment of the SGA endpoint, Levallois et al. (2012), Danileviciute et al. (2012), Costet et al.
(2012), Patelarou et al. (2011), Grazuleviciene et al. (2011), Hoffman et al. (2008a) and
Aggazzotti et al. (2004) all obtained information on beverage consumption and/or water use from
study participants. It is not clear that markedly different conclusions can be drawn from this
subset of SGA studies, compared to the six studies that did not additionally survey participants'
water use. In general, there was no clear indication of greater consistency of reported
associations, nor of exposure-response trends, among studies that used more sophisticated
exposure assessment methodologies.
With respect to observational studies addressing birth weight endpoints, there is the possibility of
confounding of the observed associations by unknown and unmeasured risk factors for adverse
birth weight outcomes that are also determinants of DBP exposure, independent of birth weight.
However, all of the studies assessed for this report evaluated and adjusted for confounding by
Six-Year Review 3 Technical Support Document A-56
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
multiple known risk factors for adverse birth weight outcomes, e.g., infant sex, gestational age,
maternal age, socioeconomic indicators, prenatal care, marital status, parity, ethnicity, maternal
body mass index, maternal smoking status, passive smoking during pregnancy, maternal disease
history and alcohol consumption during pregnancy. There are two alternate scenarios of negative
confounding which would result in the observation of no association, or a small positive
association (due to observed RR being biased toward the null for a negative confounder),
between DBP exposure and adverse birth weight outcomes assuming that there truly is a causal
association between the two; the first is due to confounding by one or more factors that increase
the risk of adverse birth outcomes and decrease in magnitude or prevalence with increasing DBP
exposure, and the second is due to confounding by one or more factors that decrease the risk of
adverse birth outcomes and increase in magnitude or prevalence with increasing DBP exposure
(Walker, 1991). Correspondingly, moderate maternal physical activity has been associated with
decreased risk of fetal growth restriction (Pivarnik, 1998) and may lead to increased water
intake. If this water intake was largely due to bottled water use (which often has lower DBP
levels) among a more health conscientious population as reported in some pregnancy cohorts
(Forssen et al., 2007), then this may result in negative confounding.
Similarly, a strong risk factor like maternal smoking during pregnancy, which was unmeasured
in some studies (Yang et al., 2007; Chisholm et al., 2008; Nieuwenhuijsen et al., 2009), may also
lead to attenuation of study results due to negative confounding as it also has been shown to be
linked to increased bottled water use activities and presumably lower DBP levels (Forssen et al.,
2007). Maternal perinatal nutrition, a potential risk factor for adverse birth weight outcomes, was
not evaluated as a confounder in these studies. The direction of any potential bias due to
confounding by maternal nutrition depends on whether poor nutrition is associated with greater
or less DBP exposure. Apart from some DBP exposures, other water contaminants were not
evaluated as confounders in the studies evaluated for this report. Only two studies examined
multi-pollutant DBP models. Hoffman et al. (2008a) did adjust associations between term birth
weight and specific THMs for other THMs. Their analyses did not find consistent evidence for
an association between any DBP species and term birth weight, although the authors concluded
that these analyses were limited due to small sample sizes. A recent study by Rivera-Nunez and
Wright (2013) saw some evidence in mean birth weight reductions for HAA5 and BrTHM
exposures with and without adjustment for other DBP surrogates. The largest reduction was
noted for DBP9 exposures which may better represent a mixture metric of the predominant DBPs
that many people are exposed to. In contrast, although the authors saw increased adjusted ORs
for SGA for various DBP metrics, these diminished following further adjustment of other DBP
surrogates (i.e., THM4 or HAA5). These studies highlight the exposure assessment complexities
that warrant further research to better elucidate the relevance of previously studied DBP mixtures
relative to toxicity demonstrated in animals or other lines of evidence.
The misclassification of health endpoints in this body of scientific evidence is less problematic,
relative to the challenges posed by assessment of DBP exposures. Nevertheless, misclassification
of fetal growth and development endpoints are subject to measurement error, which is, in the
studies reviewed, likely to be non-differential with respect to exposure. For example, gestational
age measurements are estimated by use of maternal self-report of last menstrual period,
evaluation of ultrasound, clinician estimates, or a combination of these approaches. Each of these
can be subject to measurement error which can lead to outcome misclassification when used to
examine outcomes such as PTD, as well as the possibility of residual confounding in studies that
Six-Year Review 3 Technical Support Document ^4-57
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
adjust for gestational age. Another concern regarding the outcomes examined in these studies is
that having a low birth weight, or being SGA, may not necessarily be an indicator of intrauterine
growth restriction (although SGA and IUGR are often used synonymously), as babies may
simply be born constitutionally small. Thus, the use of outcome measures that incorporate
gestational age and other factors such as ethnicity into the definition (e.g., SGA) in most of these
studies should help focus on infants that are pathologically growth restricted.
Although possible, selection bias was a limited concern in the epidemiologic studies reviewed
for this analysis. Many of the existing studies use retrospective cohort designs which can
comprehensively capture whole populations. These studies have a low probability of selection
bias, since the study inclusion criteria are not likely to be differentially related to DBP exposures.
Prospective cohort studies of birth outcomes also inherently have a low likelihood of selection
bias due to minimal loss to follow-up given the short study duration. All of the case-control
studies reviewed here also took documented steps to maximize the degree to which their control
group represented the population that gave rise to the cases. Nevertheless, the studies almost
exclusively assessed reproductive outcomes only among live births. In the case-control studies,
control participants, too, were selected from among live births. In these studies, then, a selection
bias would be induced if DBP exposure or the reproductive endpoint being evaluated influence
the risk of fetal death (or elective termination of pregnancy). In such scenarios, selection bias
may induce a false negative association if a true association between DBP and the reproductive
outcome exists.
Many of the positive associations that were observed between DBP exposure indicators and
adverse reproductive outcomes in the articles assessed for this analysis were not statistically
significant. That is to say, under the assumption that there is truly no association between DBP
exposure and adverse reproductive endpoints, one expects to observe associations as large as
those that were observed, or associations of larger magnitude, greater than 5 percent of the time
due to chance alone. The statistical power of some of the studies assessed in this analysis was
limited by low DBP levels that were limited in range. Assessments of outcomes such as specific
congenital anomalies studies were further limited by small study populations and
correspondingly small numbers of cases. Although limited statistical power was a general
weakness of the epidemiologic studies reviewed, the meta-analyses conducted for several
reproductive endpoints were able to leverage the power of multiple studies to more precisely
estimate associations with THM4.
The presence of exposure-response trends, indicated as either positive associations with
continuously distributed DBP exposure metrics or as monotonically increasing RRs associated
with ordinal categories of DBP exposure (e.g., quantiles), may be evidence for there being a
causal association between DBP exposures and reproductive health endpoints. Such trends were
formally hypothesized and assessed statistically in some studies although, often, investigators
only addressed exposure-response trends informally or left consideration of exposure-response
trends to be evaluated by the reader. Setting aside statistical evidence of exposure-response
trends, even monotonically increasing RRs of reproductive endpoints with increasing levels of
DBP exposure were infrequently and inconsistently observed. Even more seldom were such
trends observed and found to be statistically significant. As noted before, the lack of exposure-
response relationships in many studies might be partially due to some sources of bias such as
information bias (resulting from exposure misclassification) or limited exposure contrasts which
Six-Year Review 3 Technical Support Document A-58
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
preclude distinct characterization of differentially exposed groups. For categorical exposures
comparisons, an additional limitation of some studies is the inability to examine a referent
population that is lowly exposed or unexposed.
Finally, it is noted that DBP levels measured in most of the studies reviewed for this report were
low and largely below current regulatory standards. Although it is important to be assessing
potential DBP health effects at such levels, the relative lack of variability and limited range of
DBP exposure constrains the statistical power of these studies, relative to studies where the DBP
exposure range is broader.
Taken together, the limitations of the epidemiologic literature assessed in this document
constrain conclusions that can be drawn. Because of these limitations, observed associations
between DBP exposure and reproductive endpoints could be higher, or lower, than the
corresponding "true" associations.
A.1.3 Mixtures of Organic Chlorination DBPs
The intent of Stage 1 and Stage 2 D/DBPRs was not only to reduce exposure to the four THMs
and five HAAs included specifically under the MCLs, but also to reduce exposure to the mixture
of organic chlorination DBPs as a group. This section provides information on animal studies
and mixtures of DBPs and an update on research that has been conducted to further understand
the toxicity of mixtures of DBPs. These mixtures include but are not limited to the nine
substances addressed by the Stage 2 D/DBPR MCLs.
In 1998, an ILSI expert panel determined that the single-chemical testing approach was not
sufficient to assess the cancer risk from DBPs (ILSI, 1998). The panel recommended a three-
tiered testing approach, focusing first on simple, defined mixtures of fewer than 10 DBPs, then
on complex mixtures which simulate disinfection scenarios, and lastly, on samples of real
drinking water. The panel suggested using three levels of studies {in vitro, short-term screening
tests or 90-day animal studies and long-term chronic bioassays), along with studies related to
chemical structure-activity relationships and mechanism of action.
After publication of the ILSI report, a collaborative research effort was undertaken by EPA's
National Center for Environmental Assessment and National Health and Environmental Effects
Research Laboratory, Virginia Commonwealth University and Tulane University (Teuschler et
al., 2000). The goal of their collective research efforts was to determine the toxicity and
carcinogenicity of mixtures of DBPs in support of human health risk assessments. This
collaborative approach resulted in new data collection, statistical analysis and methods
development. Three approaches were recommended for future research on DBP mixtures:
toxicological studies of simple defined mixtures, toxicological studies using reproducible
disinfection scenario samples and toxicological or epidemiologic studies on direct drinking water
samples.
As part of this collaborative research, two studies focused on a threshold additivity model and on
a proportional-response addition model. The threshold additivity model assessed the hepatotoxic
interactions between the THMs included in THM4. The response of specific mixtures of THMs
in water samples from 35 water treatment facilities was predicted under dose-addition based on
Six-Year Review 3 Technical Support Document A-59
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
the dose-response curves for the individual THMs. CD-I mice were exposed to the water
samples by gavage for 14 days, and the results of biochemical markers of liver toxicity of the
mixtures fell within 95 percent of those predicted by dose-addition, demonstrating that threshold
additivity is a reasonable assumption for mixtures risk assessment. The proportional-response
addition model used a generic definition of additivity that was not dependent on mechanism of
action. The proportional-response additivity model was used to estimate the proportional risks
for developmental effects from the HAA and haloacetonitrile components of two water samples,
one from the Mississippi River and one from the Ohio River. The model showed that the
concentration of the individual DBPs may not be sufficient to result in adverse effects, but the
activity of the mixture may result in additive or greater-than-additive effects.
Andrews et al. (2004) explored the developmental toxicity interactions between three HAAs -
dichloroacetic acid, dibromoacetic acid and bromochloroacetic acid - using the whole embryo
culture assay. In this in vitro assay, rat embryos at GD 9 were exposed for 48 hours to various
concentrations (50-5000 micromolar (|iM)) of the HAAs individually or in combination and
evaluated for mortality and anomalies. Individually, the HAAs resulted in a significant increase
in malformations consisting of rotational defects, heart defects, delayed caudal development,
visceral arch defects, eye defects and a low incidence of neural tube defects. There was also a
significant increase in embryo lethality at the higher doses. Of the three HAAs, dichloroacetic
acid was the least embryotoxic, exhibiting 37 percent abnormalities at 5000 |iM; dibromoacetic
acid exhibited 64 percent at 400 |iM; and bromochloroacetic acid exhibited 70 percent at 300
|iM. The authors predicted that the combined embryo toxicity of the HAAs would be additive,
and the results confirmed this prediction. Embryo toxicity from combinations of the compounds
was additive in all binary combinations as well as in the mixture of the three compounds.
Yang et al. (2014) studied the impact of two disinfectants, free chlorine vs. monochloramine, and
the impact of bromide (Br") and iodide (I") on mammalian cell toxicity of finished drinking
water. The Chinese Hamster Ovary (CHO) cell line was used for mammalian cell toxicity studies
and the CHO cell single cell gel electrophoresis (SCGE) assay was used to measure genomic
DNA damage from drinking water samples. The water disinfected with chlorine was less
cytotoxic than the water disinfected with chloramine but it was more genotoxic. The results of
the CHO cell cytotoxicity assay showed that the lowest levels of cytotoxicity were associated
with disinfection by monochloramine or free chlorine alone, and the addition of bromide and
iodide significantly increased the cytotoxicity of the chloramine or chlorine disinfection.
Similarly, the results of the SCGE assay demonstrated that the addition of bromide and iodide
significantly increased the genotoxicity of the chloramine or chlorine disinfection. The authors
concluded that the agents which resulted in cytotoxicity and genotoxicity were the generated
brominated and iodinated DBPs rather than the formation of chlorinated DBPs.
Six-Year Review 3 Technical Support Document A-60
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
A.2 Regulated Inorganic DBPs
A.2.1 Bromate
Information Available During Development of Stage 1 and Stage 2 D/DBPRs
Cancer
Bromate was carcinogenic when administered in drinking water to male and female rats.
DeAngelo et al. (1998) administered potassium bromate in drinking water to male F344 rats and
male B6C3F1 mice. Mesotheliomas, which originated from the testis, spread throughout the
peritoneal cavity. Kidney and thyroid tumors were observed in the rats. Kurokawa et al. (1986a,
1986b) conducted a study with potassium bromate in drinking water administered to male and
female F344 rats and female B6C3F1 mice. They also observed peritoneal mesotheliomas, but
did not specify the origin. Kidney and thyroid tumors were observed in male rats and kidney
tumors in female rats. It was not carcinogenic in female mice.
Mutagenicity/Genotoxicity
Mixed results were reported for bromate for in vitro mutagenicity studies in S. typhimurium.
Positive results were reported in in vitro studies on chromosomal aberrations, chromatid breaks
and micronuclei formation in mammalian cells and for positive results in the comet assay
indicative of DNA strand breaks. In vivo studies in mice reported cytogenic effects on bone
marrow cells, micronuclei formation and increases in micronucleated polychromatic erythrocytes
(Health Canada, 1998; USEPA, 2001b).
Reproductive/Developmental
USEPA (2001b) reviewed the following study on bromate:
Wolf and Kaiser (1996) administered bromate to male and female rats at doses up to 22
mg/kg/day in drinking water for various times during a 35-day period. A significant decrease in
epididymal sperm density was observed in males at 22 mg/kg/day and no effects were noted on
female reproductive end points. A reproductive NOAEL of 7.7 mg/kg/day and LOAEL of 22
mg/kg/day were identified based on the effects on sperm.
Other
Nakano et al. (1989) reported necrotic changes in the kidneys, increased blood urea nitrogen and
abnormalities in the cortical tubules of the kidneys of rats at 30 mg/kg/day in drinking water for
15 months. A LOAEL of 30 mg/kg/day was identified based on these effects (USEPA, 2001b).
DeAngelo et al. (1998) reported renal urothelial hyperplasia at 7.9 mg/kg/day in a 100-week
drinking water study in rats. A NOAEL of 1.1 mg/kg/day and a LOAEL of 7.9 mg/kg/day were
identified from this study (USEPA, 2001b).
Six-Year Review 3 Technical Support Document A-61
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
A.2.2 Chlorite
Information Available During Development of Stage 1 and Stage 2 D/DBPRs
Cancer
Kurokawa et al. (1986b) administered sodium chlorite to F344 rats and B6C3F1 mice in drinking
water for 85 or 80 weeks. No chlorite-related increases in tumor incidence were observed in the
rats and a high mortality rate in the control mice made statistical comparisons between controls
and treated mice difficult to interpret. EPA concluded that the study was inadequate for assessing
carcinogenicity due to the relatively short exposure (80 weeks) and the high incidence of early
mortality in the control mice (USEPA, 2000b). IARC (1991) evaluated the carcinogenicity data
on sodium chlorite and concluded that there was inadequate evidence for the carcinogenicity of
sodium chlorite in experimental animals.
Mutagenicity/Genotoxicity
Positive results were reported for mutagenicity of chlorite in in vitro studies in S. typhimurium,
both with and without metabolic activation. In vivo studies reported negative results for
chromosomal aberrations in micronucleus assays, in bone marrow cells and in the sperm-head
abnormality assay following gavage administration of chlorite in mice. Positive results were
reported in one micronucleus assay in bone marrow cells of mice after intraperitoneal injection
of chlorite (USEPA, 2000b).
Reproductive/Developmental
The following studies on chlorite were reviewed in USEPA (2000b) and Health Canada (2008b):
Moore et al. (1980) administered sodium chlorite to pregnant A/J mice at approximately 22
mg/kg/day in drinking water throughout gestation and lactation. No significant effects were
noted on gestation length, litter size, or number of pups dead at birth; however, significant
decreases were observed in average pup weaning weight and birth-to-weaning growth rate. A
developmental LOAEL of 22 mg/kg/day was determined.
Couri et al. (1982) conducted a developmental study in pregnant Sprague-Dawley rats exposed to
sodium chlorite in drinking water at doses up to 610 mg/kg/day on GD 8-15. Another group of
pregnant rats received 200 mg/kg/day via gavage on GD 8-15, which resulted in 100 percent
mortality. An increase in the number of resorbed and dead fetuses and decreases in crown-rump
length were reported at all dose levels, with no effects reported on postnatal growth or incidences
of soft tissue and skeletal malformations. A frank effect level of 70 mg/kg/day for resorbed and
dead fetuses and decreases in crown-rump length was determined.
Suh et al. (1983) administered chlorite to pregnant Sprague-Dawley rats at approximately 0, 0.1
or 1 mg/kg/day for 2.5 months before mating them with unexposed males, as well as during GD
0-20. No significant effects were noted on resorptions, fetus survival, fetal body weights, or
incidence of skeletal abnormalities. A developmental NOAEL of 1 mg/kg/day was determined.
Six-Year Review 3 Technical Support Document A-62
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Carlton and Smith (1985) and Carlton et al. (1987) conducted a series of reproductive/
developmental studies. In a first set of studies, male Long-Evans rats were administered doses up
to 7.5 mg/kg/day chlorite in drinking water for 56 days before mating and throughout the 10-day
mating period. Female rats were administered the same dose of sodium chlorite for 14 days
before mating, during the mating periods and throughout gestation and lactation. No dose-related
changes in fertility or in sperm parameters were observed in the parental rats; however,
significant decreases in T3 and T4 levels were observed in the offspring of rats administered 7.5
mg/kg/day. A developmental NOAEL of 0.75 mg/kg/day and LOAEL of 7.5 mg/kg/day based on
decreased hormone levels were determined. In a second set of studies, Long-Evans rats were
administered doses of chlorite of up to 27 mg/kg/day in drinking water for 72-76 days. A
significant increase in abnormal sperm was observed, with abnormalities including frayed tails,
open hooks and amorphous sperm heads. A reproductive NOAEL of 0.75 mg/kg/day and
LOAEL of 7.5 mg/kg/day were determined.
Mobley et al. (1990) exposed female Sprague-Dawley rats to approximately 3 and 6 mg/kg/day
chlorite for 10 days before mating them with unexposed males and during gestation and
lactation. Significant decreases in exploratory activity were observed in the rat pups, with a
developmental NOAEL of 3 mg/kg/day and a LOAEL of 6 mg/kg/day identified.
Harrington et al. (1995a) conducted a developmental study in New Zealand white rabbits,
administering doses up to 40 mg/kg/day chlorite in drinking water for GD 7-20. The authors
concluded that there were no treatment-related effects on pregnancy incidence, number of
implantations, number of pre-implantation losses, fetal sex ratio, number of live fetuses or fetal
visceral or structural abnormalities. Mean fetal weights were slightly decreased at 26 and 40
mg/kg/day, and skeletal variants related to incomplete fetal bone ossification were increased at
26 mg/kg/day. EPA identified a developmental NOAEL of 10 mg/kg/day and a LOAEL of 26
mg/kg/day, based on decreased fetal weight and delayed skeletal ossification (USEPA, 2000b).
The Chemical Manufacturers Association (1996) conducted reproductive/developmental studies
in Sprague-Dawley rats. Rats received drinking water containing up to 20 mg/kg/day chlorite for
males and 28.6 mg/kg/day chlorite for females for 10 weeks. Males were exposed throughout
mating, and females were exposed through mating, pregnancy and lactation for two generations.
The F1 pups were mated twice to produce the F2a and F2b generations. Reduced absolute and
relative liver weights in F0 females and F1 males and females, reduced pup survival, reduced
body weights at birth in F1 and F2 rats, lower spleen and thymus weights in F1 and F2 rats, and
lowered incidence of pups exhibiting normal righting reflex and response to an auditory startle
stimulus were observed. A developmental and toxicity NOAEL of 2.9 mg/kg/day and LOAEL of
5.9 mg/kg/day were determined based on reduced organ weights and lowered auditory startle
amplitude in the pups.
Other
Subchronic and chronic oral administration of chlorite in animals results in effects on the
stomach and organ weights, and hematotoxicity. Oral studies in rats, mice and monkeys ranging
from 30 days to 13 weeks reported hematological effects from chlorite administration (Abdel-
Rahman et al., 1984; Couri and Abdel-Rahman, 1980; Moore and Calabrese, 1982; Bercz et al.,
1982). However, USEPA (2000b) assessed the hematological effects from Abdel-Rahman et al.
Six-Year Review 3 Technical Support Document A-63
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
(1984) and Couri and Abdel-Rahman (1980) studies and stated "The lack of a consistent dose-
effect relationship, small numbers of animals, and small magnitude of effects complicate
interpretation of the results" (USEPA, 2000b). Harrington et al. (1995b) administered sodium
chlorite by gavage to Crl:CD (SD) BR rats for 13 weeks. A NOAEL of 7.4 mg/kg/day and a
LOAEL of 19 mg/kg/day for stomach lesions and increases in spleen and adrenal weights in rats
were identified from this study (USEPA, 2000b). The stomach lesions consisted of hyperplasia,
hyperkeratosis, ulceration, chronic inflammation and edema and were considered by EPA to be
noncancerous.
A.3 Regulated Disinfectants
A.3.1 Chlorine
Information Available During the Development of Stage 1 and Stage 2 D/DBPRs
Cancer
Chlorine was administered to F344/N rats and B6C3F1 mice in drinking water for two years
(NTP, 1992a). NTP concluded that there was no evidence of carcinogenicity in male rats or male
and female mice, and equivocal evidence in female rats, based on an increase in the incidence of
mononuclear cell leukemia.
Reproductive/Developmental
NTP reviewed the following two reproductive and developmental studies on chlorine (NTP,
1992a):
Abdel-Rahman et al. (1982) administered female Sprague-Dawley rats up to 100 mg/L chlorine
in drinking water for 2.5 months before conception and throughout gestation. No increase in fetal
resorptions was observed at any doses, although some soft-tissue defects were noted at 100
mg/L.
Carlton et al. (1986) conducted a reproductive/developmental study in Long-Evans rats. Chlorine
was administered by gavage at doses up to 5 mg/kg/day before breeding and throughout the 10-
day breeding cycle. No effects were noted on sperm count, sperm motility or sperm morphology
or on fertility, fetal viability or litter size.
Other
No treatment related effects were reported following 13-week drinking water studies with
chlorine in Sprague-Dawley rats at 25, 100, 175 or 200 mg/L (Daniel et al., 1990). In 90-day
drinking water studies (Daniel et al., 1991) some reductions in organ weights in B6C3Fi mice
were observed in the liver, heart and lung in male mice and in the liver, heart and spleen in
female mice at 100 and 200 mg/L (11.1 and 15.6 mg/kg/day in males and 12.9 and 15.8
mg/kg/day in females). Spleen and liver weights were reduced in male and female Sprague-
Dawley rats at 9 mg/kg/day in males and 12.1 mg/kg/day in females.
Six-Year Review 3 Technical Support Document A-64
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
A.3.2 Chloramines
Information Available During the Development of Stage 1 and Stage 2 D/DBPRs
Cancer
Chloramine was administered to F344/N rats and B6C3F1 mice in drinking water for two years
(NTP, 1992a). NTP concluded that there was no evidence of carcinogenicity in male rats or male
and female mice, and equivocal evidence in female rats, based on an increase in the incidence of
mononuclear cell leukemia.
Reproductive/Developmental
NTP (1992a) reviewed the following two reproductive and developmental studies on chloramine:
Abdel-Rahman et al. (1982) administered female Sprague-Dawley rats up to 100 mg/L
chloramine in drinking water for 2.5 months before conception and throughout gestation. No
effects were observed at any dose.
Carlton et al. (1986) conducted a reproductive/developmental study in Long-Evans rats.
Chloramine was administered by gavage at doses up to 10 mg/kg/day for 56 days before
breeding and throughout the 10-day breeding cycle. No effects were noted on sperm count,
sperm motility or sperm morphology or on fertility, fetal viability or litter size.
Other
No treatment related effects were reported following 13-week drinking water studies with
chloramine in Sprague-Dawley rats at 25, 50, 100 or 200 mg/L (Daniel et al., 1990).
A.3.3 Chlorine dioxide
Information Available During the Development of Stage 1 and Stage 2 D/DBPRs
Reproductive/Developmental
The following studies on chlorine dioxide were reviewed in USEPA (2000b) and Health Canada
(2008b):
Suh et al. (1983) administered chlorine dioxide to Sprague-Dawley rats at doses up to 10
mg/kg/day in drinking water for 2.5 months before mating and during GD 0-20. Total fetal
weights and male fetal weights were significantly increased at 10 mg/kg/day and there was a
significant trend for decreasing number of implants per litter and number of live fetuses per dam.
A developmental NOAEL and LOAEL of 1 mg/kg/day and 10 mg/kg/day, respectively, were
determined.
Orme et al. (1985) conducted a developmental study in Sprague-Dawley rats exposed to doses up
to 14 mg/kg/day chlorine dioxide in drinking water for two weeks before mating and throughout
gestation and lactation. Additional 5-day old pups (not exposed in utero) were exposed to 14
Six-Year Review 3 Technical Support Document A-65
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
mg/kg/day by gavage on PND 5-20. In the pups exposed to 14 mg/kg/day by gavage, locomotor
activity was significantly decreased, and in the pups exposed to 100 mg/L (14 mg/kg/day) in
utero, there were significant decreases in T3 and T4 levels. A developmental NOAEL of 3
mg/kg/day and a LOAEL of 14 mg/kg/day were identified.
Taylor and Pfohl (1985) administered approximately 14 mg/kg/day chlorine dioxide in drinking
water to female Sprague-Dawley rats for 14 days before breeding and throughout gestation and
lactation. A significant decrease in whole brain weight, cerebellar total DNA content and
exploratory behavior were observed in the offspring of the treated rats. The developmental
LOAEL was 14 mg/kg/day.
Toth et al. (1990) administered daily gavage doses of 14 mg/kg/day chlorine dioxide by gavage
to Sprague-Dawley rat pups on PND 1-20. No gross lesions, loss of myelin or cell changes in the
brain were observed in the pups. However, forebrain weights, protein content and DNA content
were significantly reduced in the brain at 14 mg/kg/day. A LOAEL of 14 mg/kg/day was
identified.
Mobley et al. (1990) conducted a developmental study in female Sprague-Dawley rats
administered 14 mg/kg/day chlorine dioxide in drinking water for 10 days before mating and
during the gestation and lactation periods. No significant effects were observed on litter size, but
decreased litter weights and decreased exploratory activity were observed in the offspring of the
treated rats. The developmental LOAEL was 14 mg/kg/day.
Carlton et al. (1991) administered chlorine dioxide at doses up to 10 mg/kg/day in drinking water
to Long-Evans rats for 56 days before mating and 10 days during the mating period. No
significant effects were noted on mortality, clinical signs, fertility, sperm parameters, length of
gestation, prenatal deaths, mean litter size or mean pup weights. A developmental/reproductive
NOAEL of 10 mg/kg/day was determined.
Other
Subchronic and chronic drinking water studies ranging from 30 days to 2 years resulted in nasal
lesions, aaalterations in thyroid hormone levels and hematological effects. Daniel et al. (1990)
reported a significant increase in nasal lesions at 2 mg/kg/day in Sprague-Dawley rats; however,
the toxicological significance of the nasal effect is not known and may be an artifact of treatment
(USEPA, 2000b). Bercz et al. (1982) noted a significant decrease in serum T4 levels after
administration of 9.5 mg/kg/day to monkeys for six weeks. Couri and Abdel-Rahman (1980)
found significant increases in blood glutathione (GSH) reductase levels and significant decreases
in erythrocyte GSH levels in Sprague-Dawley rats administered 0.1 - 100 mg/kg/day for one
year.
Six-Year Review 3 Technical Support Document A-66
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Appendix B. Additional Information for Occurrence and Exposure to
Regulated and Unregulated Disinfection Byproducts (DBPs)
(Appendix to Chapter 6)
This appendix summarizes information relevant to occurrence and exposure to regulated and
unregulated disinfection byproducts (DBPs), supplementing information provided in Chapter 6.
Section B.l provides what was known at the time of the development of the Stage 2
Disinfectants and Disinfection Byproducts Rule (D/DBPR) as well as new information related to
DBP formation. Section B.2 provides historical and new information related to occurrence of
DBP precursors available since the promulgation of the Stage 2 D/DBPR. Section B.3 presents
historical information on DBP occurrence, presents the results of new occurrence analyses using
the data on regulated DBPs from the Third Six-Year Review Information Collection Request
(SYR3 ICR) and discusses new occurrence information available for unregulated DBPs. Lastly,
Section B.4 describes additional quality assurance and quality control (QA/QC) process
conducted on the SYR3 ICR DBP dataset.
B.l DBP Formation
This section summarizes what was known about DBP formation at the time of the development
of the Stage 2 D/DBPR and presents new, peer-reviewed information relevant to the SYR3
process. Note that in some cases no supplemental information is available.
B.l.l Summary of Stage 1 and 2 D/DBPR Information
No additional information is provided in this appendix.
B.1.2 New Information since the Stage 2 D/DBPR
No additional information is provided in this appendix.
B.l.2.1 DBP Types
No additional information is provided in this appendix.
B.l.2.2 Disinfection Practices
This section describes disinfection methods and doses as well as contact time.
Disinfectant Types and Doses
Reactions between NOM and chlorine form a variety of halogenated DBPs, the most abundant of
which are trihalomethanes (THMs) and haloacetic acids (HAAs). Chlorine also causes the
formation of some non-halogenated DBPs, such as aldehydes and ketones. Because disinfectant
is generally the limiting reagent, dose has a large effect on DBP formation. The combination of
high doses and long residence times after booster disinfection can lead to areas of high DBP
concentrations.
Six-Year Review 3 Technical Support Document	B-l
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Studies have documented that chloramines produce significantly lower DBP levels than free
chlorine (USEPA, 20051).
The Information Collection Rule (ICR) was the main source of data for the Stage 1 and Stage 2
D/DBPRs. This database (referred to in this document as the "DBP ICR database") contains the
information collected from a survey of 296 systems comprising 512 plants (which includes 11
plants with blended source water) serving more than 100,000 people, conducted over 18 months,
from July 1997 to December 1998. It represented the largest and most comprehensive national
occurrence estimates of DBPs at that time. In addition to DBP occurrence concentrations, the
DBP ICR database also contained extensive information regarding treatment, source water
characteristics and disinfectant type. The DBP ICR database characterized the water quality at
each plant's source, at several steps in the treatment process and at several points in the
distribution system (reflecting finished water).
The DBP ICR results showed that the amount of DBP formation by chloramines was between 5
and 35 percent of the DBP formation by chlorine depending on the individual DBP species.
Studies at the time found that direct reaction between chloramines and NOM produces few
DBPs, although dichloroacetic acid (DCAA) and cyanogen chloride were produced at higher
concentrations than with the use of free chlorine (USEPA, 20051). There was no clear evidence
at the time that the reaction of NOM with chloramines leads to the formation of THMs. Most
DBPs that form during chloramination are a result of reactions between free chlorine and NOM.
The free chlorine can originate from its addition prior to ammonia or from the hydrolysis of
chloramines. Prior to promulgation of the Stage 2 D/DBPR, little was known about the
unidentified halogenated organics except that they were more hydrophilic and had a higher
molecular weight than halogenated organics produced by free chlorine.
Studies prior to promulgation of the Stage 2 D/DBPR had studied the mechanism of the
formation of A'-nitrosodi methyl amine (NDMA) from the use of chloramines in drinking water.
However, there was not enough information to draw conclusions regarding increases in NDMA
formation as systems switched from free chlorine to chloramines (USEPA, 2006a).
Chlorine dioxide does not produce significant amounts of organic halogenated DBPs. The
inorganic DBPs chlorite and chlorate, however, are byproducts of chlorine dioxide use. Both
chlorite and chlorate can be byproducts of the generation process for chlorine dioxide or can be
produced by reactions with chlorine dioxide after its addition to source water containing NOM.
Chlorite was regulated with the Stage 1 D/DBPR; chlorate was not regulated.
Research prior to the Stage 2 D/DBPR showed that ozone by itself does not form halogenated
DBPs. Ozone alters the nature of NOM and forms oxygenated DBPs such as aldehydes and
organic acids. The smaller molecules formed by ozone can be removed by biological filtration. If
chlorine is added before these smaller more reactive molecules are removed, then they can react
with chlorine to form DBPs. Ozone can also react with bromide to form bromate or
hypobromous acid, which can react to form brominated DBPs. Brominated DBPs formed after
ozonation include bromoform and cyanogen bromide, although two-thirds of the brominated
DBPs formed during ozonation had not been identified at the time.
Six-Year Review 3 Technical Support Document	B-2
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
At the time of the Stage 2 D/DBPR, no evidence suggested that the use of ultraviolet light (UV)
as a disinfectant resulted in the formation of DBPs, although little research had been performed
in the area.
The Stage 2 D/DBPR Economic Analysis (USEPA, 2005g) provided a summary of what was
known about disinfectant use based on DBP ICR data. Free chlorine only was used as a
disinfectant in 53.7 percent of plants. Chloramines preceded by free chlorine contact time were
used in 23.1 percent of plants, while 6.5 percent of plants used free chlorine as a primary
disinfectant followed by chloramines in the distribution system. Chloramine was used both as a
primary and residual disinfectant in 5.1 percent of plants. Only 2.7 percent of plants used
chlorine dioxide as a primary disinfectant followed by chlorine, while another 3.7 percent used
chlorine dioxide followed by chloramines. Three percent of plants used ozone followed by
chloramines, and another 2.1 percent used ozone followed by chlorine.
Contact Time
Information prior to the Stage 2 D/DBPR showed that DBPs form as long as disinfectant residual
and reactive DBP precursors are present. Generally, the longer the contact time, the greater the
DBP formation potential. In the presence of a disinfectant residual, both THMs and HAAs had
generally high stabilities and persisted after formation. HAAs, however, were known to
biodegrade over time when disinfectant residual was low.
B.l.2.3 Source Water Quality Research
This section describes source water characteristics that affect DBP formation, including organic
precursors, inorganic precursors such as bromide and iodide, temperature and pH.
Organic Precursors
At the time of the Stage 2 D/DBPR, studies conducted with different fractions of NOM found
that NOM with high aromatic content tended to form more DBPs than NOM with low aromatic
content. Since UV absorbance at 254 nm wavelength is correlated with aromatic content of
organics, this led to the use of specific ultraviolet absorbance (SUVA)2 as an indicator of THM
and HAA formation in source water. Waters with high SUVA were known to be more easily
treated with coagulation (USEPA, 20051).
Inorganic Precursors
Bromide in source water affects the formation of DBPs. Free chlorine can oxidize bromide to
hypobromite or hypobromous acid, which can react with NOM to form brominated DBPs.
Research at the time had shown that the rate of THM formation was higher in waters with
2 SUVA is the ultraviolet absorption at 254 nanometers divided by the concentration of dissolved organic carbon
(DOC).
Six-Year Review 3 Technical Support Document	B-3
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
increased bromide concentration. Bromine can also substitute into chlorinated DBPs in the
presence of hypobromous acid (USEPA, 20051).
Temperature
The rate of THM formation increases with temperature. Research also showed that HAA
formation increased with temperature, although the effect was less pronounced. Therefore, the
highest THM and HAA concentrations were thought to occur in the summer months. In addition,
high temperatures can accelerate chlorine depletion, resulting in less DBP formation and
biodegradation of HAA (USEPA, 20051).
pH
Research prior to the Stage 2 D/DBPR found that THM formation increases with increasing pH,
while formation of HAA and other DBPs decreases with increasing pH. The formation of more
THM at higher pH was likely due to base-catalyzed reactions. HAA formation may be altered at
high pH due to hydrolysis of precursors.
The rate of DBP formation from ozone is not affected by pH; although, the rate of ozone
decomposition increases at higher pH. Increased pH results in decreased aldehydes; although in
some situations, carbonyls could increase at higher pH. Low pH in ozonated water increases
formation of brominated DBPs. This occurs because the hypobromous acid and hypobromite
formed by reaction of bromide and ozone shift more to hypobromous acid at lower pH.
Hypobromous acid is more reactive than hypobromite in the formation of brominated DBPs
(USEPA, 20051).
B.l.2.4 Distribution System Conditions
No additional information provided in this appendix.
B.l.2.5 DBP Formation Modeling
Since promulgation of the Stage 2 D/DBPR, numerous studies have developed predictive models
for DBP formation. This section provides background on the Surface Water Analytical Tool
(SWAT) model, including its formulas for chlorination and chloramination.
Background on SWAT
Based on the research and the related national datasets available at the time, a computer program
called the SWAT was created to predict formation of four regulated THMs (THM4) and five
HAAs (HAA5) for all surface water systems serving 100,000 or more people. The program used
empirical formulas to calculate THM4, HAA5, bromate and chlorite formation at the finished
water point, average residence time and maximum residence time sites (USEPA, 2005g). For
additional details on SWAT, refer to USEPA (2005g; see Appendix A of that document).
SWAT was developed with the assistance of the M/DBP Federal Advisory Committee to model
DBP formation on a national level. The intent was that parametric inputs for many systems could
be used to develop national distributions of DBPs that in turn could be used for predicting
Six-Year Review 3 Technical Support Document	B-4
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
national compliance forecasts for different regulatory options. The predictive equations
contained in SWAT were calibrated with the national datasets generated from the DBP ICR
using a central tendency approach. SWAT was not intended to provide reliable DBP formation
predictions for specific systems, but rather to characterize national-level occurrence distributions.
Chlorination
SWAT's empirical formulas for the formation of THM4 and HAA5 used total organic carbon
(TOC), UV254, chlorine dose, bromide concentration, temperature, pH and time as variables to
determine THM4 concentration. SWAT used different equations for THM4 and HAA5
formation depending on whether the water was raw or treated. For treated water, the equation
used the product of TOC and UV254 to determine DBP formation. The equations used in SWAT
are as follows (see Appendix A in USEPA, 2005g):
THM4raw = 0.0412TOC1 098 Cl20 152Brraw0 068T0-609pHrawL601t0263
THM4treated = 23.9(TOC*UV254)0-403 Cl20-225 Br0141 1.0271"20 1.156pH"7-510 264
SWAT also predicted the concentration of the sum of HAA5. The equations for HAA5 were:
HAA5raw = 30TOC0"7 Cl20'278 Bw"0138 T0341 pHraw"0-799 tL69
HAA5treated = 41.6(TOC*UV254)0'238 Cl20585 Br"012 1.0211-20 0.932pH"7-510150
Where:
THM4 = total trihalomethanes in |ag/L
HAA5 = total 5 haloacetic acids in ng/L
TOC = TOC concentration in mg/L
Cl2 = chlorine dose in mg/L
Br = bromide concentration in |ag/L
T = temperature in degrees Celsius
t = time in hours
UV254 = UV absorbance at 254 nm
Chloramination
Empirical predictive correlations for DBP formation for systems using chloramines were not
available at the time of the Stage 2 D/DBPR. Therefore, the SWAT model assumed that THM4
formation by chloramines was 30 percent of what would have formed by chlorination and HAA5
formation was 35 percent of what would have formed by chlorination (USEPA, 2005g).
Empirical formulas were included in SWAT for THM4 and HAA5, but were not identified for
chlorite or bromate.
Six-Year Review 3 Technical Support Document	B-5
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
B.2 Occurrence of DBP Precursors
This section summarizes historical and new occurrence information on organic and inorganic
DBP precursors and precursor mixtures.
B.2.1 Organic Precursors
This section provides additional information from the Stage 1 and Stage 2 D/DBPRs and new
information available since the promulgation of the Stage 2 D/DBPR.
B.2.1.1 Summary of Stage 1 and Stage 2 D/DBPR Information
This section summarizes occurrence data that were available prior to the promulgation of the
Stage 2 D/DBPR related to organic DBP precursors. DOC and total organic nitrogen (TON)
were not measured as part of DBP ICR monitoring, thus, national-level occurrence data for these
precursors was not reviewed during the development of the Stage 2 D/DBPR.
TOC is used as an indicator of the amount of organic carbon available to react with disinfectants
to form organic DBPs. The main source of TOC data for large water systems serving at least
100,000 people was the DBP ICR database. The DBP ICR applied to surface and ground water
systems and represented data collected between July 1997 and December 1998 (USEPA, 20051).
Exhibit B. 1 summarizes DBP ICR data for water treatment plant influent TOC. Exhibit B. 1 also
summarizes DBP ICR data on UV254, a potential predictor of the tendency of a source water to
form THMs and HAAs (USEPA, 20051) and alkalinity, which affects the treatability of the
organic precursors.
Exhibit B.1: DBP ICR Large System Influent TOC, UV254 and Alkalinity Data
Parameter
Source Type
Number of
Plants
Mean of Plant
Means
Median of
Plant Means
90th
Percentile of
Plant Means
Range of
Plant Means
Total Organic Carbon (mg/L as C)

Surface
307
3.14
2.71
5.29
0.0-21.4

Ground
103
1.46
0.19
3.36
0.0-16.1
UV254 (cm1)

Surface
306
0.098
0.079
0.176
0.0-0.880

Ground
104
0.062
0.009
0.266
0.0-0.606
Alkalinity (mg/L as CaCOs)

Surface
336
81
79
165
2.75-273

Ground
121
159
156
264
1.00-415
Source: ICR AUX1 Database (USEPA, 2005I)
Notes: From the ICR AUX1 database (USEPA, 2000e). Represents distribution of plant mean data as calculated
using ICR monthly data from the last 12 months of the ICR (January 1998 - December 1998). Only plants with
reported data for at least 9 of the 12 months are included in this analysis. Does not include blended, mixed or
Six-Year Review 3 Technical Support Document	B-6
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
purchased water plants. Values below the minimum reporting level (MRL) were converted to zero to calculate plant
means.
Data for systems serving fewer than 100,000 people were available from the National Rural
Water Association (NRWA), supplemental surveys (SS) and Waterstats. Exhibit B.2 summarizes
TOC, UV254 and alkalinity data for these systems.
Exhibit B.2: Medium and Small System Influent TOC, UV254 and Alkalinity Data
Data Source/Size Category
N
Mean of Plant
Means
Median of
Plant
Means
90th Percentile
of Plant
Means
Range of
Plant
Means
Source Water TOC (mg/L as C)
NRWA Small SW Plants
96
3
2.6
5.4
0.3-9.0
ICR SS Medium SW Plants
40
3.6
3.7
5.5
0.2-7.9
ICR SS Small SW Plants
38
2.4
2.1
4.5
0.1 -7.1
WATERASTATS Medium SW Plants
102
5.6
3.2
6.4
0-200
WATERASTATS Medium GW Plants
51
2.3
0.79
7
0-25
Source Water UV-254 (cm1)
NRWA Small SW Plants
96
0.082
0.074
0.127
0.01 -0.23
ICR SS Medium SW Plants
40
0.093
0.083
0.171
0.03-0.21
ICR SS Small SW Plants
38
0.074
0.051
0.113
0.02-0.44
Source Water Alkalinity (mg/L as CaCOs)
NRWA Small Surface Water (SW) Plants
95
81
74
146
0-281
ICR Supplemental Survey (ICR SS)
Medium SW Plants
40
82
74
159
4.8-240
ICR SS Small SW Plants
38
66
55
123
4.4 - 249
Source: USEPA, 20051
Note: ICR SS data are the plant means for plants that took at least three-fourths of the total possible samples for each parameter.
Only plants that had both a winter and summer sample are included in the NRWA data for this analysis.
Exhibit B.3 shows the percent of monthly samples over the final 12 months of the DBP ICR
monitoring period (January to December 1998) that fall into specified source water TOC and
alkalinity categories. These categories are used under the Stage 1 D/DBPR to specify source
water TOC removal requirements as part of a treatment technique (TT) for DBP precursor
removal for plants using conventional treatment. Due to seasonal variation and other factors
affecting source water, the percentage removal requirements for each plant may have changed
from month to month as the influent TOC and alkalinity varied. Many samples are close to the
limits for a percentage removal group, indicating that the treatment requirements of a plant can
easily change (USEPA, 20051).
Six-Year Review 3 Technical Support Document	B-7
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Exhibit B.3: Distribution of Monthly Influent TOC (mg/L) and Monthly Influent
Alkalinity (mg/L) Samples Based on ICR Data for All Large Plants
Source Water TOC Range (mg/L)
Percentage
Alkalinity (mg/L)
Total
< 60
60 -120
> 120
< 2.0
14%
10%
16%
39%
2.0-4.0
14%
14%
13%
41%
4^
b
00
b
5%
5%
6%
16%
> 8.0
1%
0%
2%
4%
Source: ICR AUX1 Database (USEPA, 2005I)
B.2.1.2 New Information since the Stage 2 D/DBPR
This section provides new information on organic precursor occurrence from recent studies
published since the promulgation of the Stage 2 D/DBPR and presents summary inventory
information on TOC and alkalinity in the SYR3 ICR database.
Potter and Wimsatt (2012) used EPA Method 415.3 to quantify TOC, DOC and SUVA in seven
different source waters. Samples came from five surface water sources and two ground water
sources. Samples were tested on five different machines in order to determine method accuracy.
Mean TOC concentrations ranged from 0.42 to 3.64 mg/L. Mean DOC concentrations ranged
from 0.42 to 3.38 mg/L and mean SUVA values ranged from 1.95 to 3.37 L/mg-m.
Samson et al. (2013) used three case studies to develop monthly TOC thresholds defined as the
highest TOC concentration of source water that allows a conventional surface water treatment
plant to meet the DBP regulations at the maximum residence time location in the distribution
system. Statistical models were developed to relate TOC threshold exceedances with variables
including precipitation, temperature and vegetation indices.
Mikkelson et al. (2013) analyzed water-quality data from quarterly reports submitted as part of
Stage 1 and Stage 2 D/DBPR compliance by Colorado treatment plants, including five plants
impacted by the mountain pine beetle and four control plants. Mean TOC concentrations from
control sites were 0.70 mg/L and 0.62 mg/L for the years 2004 to 2008 and 2009 to 2011,
respectively. Mean TOC concentrations for infested sites were 2.5 mg/L and 2.7 mg/L for the
years 2004 to 2008 and 2009 to 2011, respectively.
Writer et al. (2014) studied the Cache la Poudre River Watershed in Colorado after the 2012
High Park Wildfire. Following the wildfire, thunderstorms in the Poudre River Watershed caused
mudslides, which resulted in sediment, ash and debris being deposited into the river. DOC
concentrations after four thunderstorms measured within the burned area at the Poudre River
drinking water intake ranged from 3.5 mg/L to 13.7 mg/L. DOC concentrations from monthly
monitoring following the thunderstorm events at both locations upstream and downstream from
the affected area ranged from 2.1 to 2.8 mg/L.
Six-Year Review 3 Technical Support Document	B-8
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Emelko et al. (2013) studied seven watersheds that were affected by the 2003 Lost Creek
Wildfire. From 2004 to 2012, Emelko et al. studied burned, burned and salvage logged,
prescribed burned and undisturbed watersheds using instrumentation in the watersheds, climate
stations and hydrometric stations. DOC was found to increase with flow and level of disturbance.
The variation in DOC peaks was greater for disturbed catchments, ranging from about 2 to 17
mg/L.
Wang et al. (2015) analyzed the water-extractable organic matter (WEOM) from burned detritus
of both moderate and high severity compared to a non-burned control. WEOM from both types
of ash had lower reactivity at 55 percent of control for THM formation and 67 percent of control
for HAA5 formation. Due to consumption of organic matter by the wildfire, the ashes contained
decreased extractable organic carbon and organic nitrogen at 27 percent of control and 19
percent of control, respectively.
B.2.2 Precursor Inventory Analyses
B.2.2.1 TOC
The results of system and population inventories of the SYR3 ICR TOC dataset in 2006-2011 are
included below in Exhibit B.4 through Exhibit B.6. Exhibit B.4 depicts the distribution of
systems and population among the different system types, and Error! Reference source not
found.includes the same information but distributed based on source water type. The source
types are split by ground water (includes purchasing systems), surface water (includes
purchasing systems), and purchased and non-purchased ground water under the direct influence
of surface water (GWUDI) systems. Exhibit B.6 depicts the distribution of systems and
population by both source water type, system type and aggregated by population size.
Exhibit B.4: Number of Systems and Population Served by Systems in the SYR3
ICR Dataset with TOC Records, by System Type (2006 - 2011)
Year
System Type
Systems
Population
Number
Percent
Number
Percent
2006
Community
1,819
89.9%
54,759,578
99.8%
Non-transient Non-community
204
10.1%
94,418
0.2%
Transient Non-community
0
0.0%
0
0.0%
Total
2,023
100.0%
54,853,996
100.0%
2007
Community
1,762
92.9%
56,135,166
99.9%
Non-transient Non-community
135
7.1%
66,133
0.1%
Transient Non-community
0
0.0%
0
0.0%
Total
1,897
100.0%
56,201,299
100.0%
2008
Community
1,853
92.8%
59,719,015
99.9%
Non-transient Non-community
144
7.2%
77,828
0.1%
Transient Non-community
0
0.0%
0
0.0%
Six-Year Review 3 Technical Support Document	B-9
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Year
System Type
Systems
Population
Number
Percent
Number
Percent

Total
1,997
100.0%
59,796,843
100.0%
2009
Community
1,842
92.4%
60,738,439
99.9%
Non-transient Non-community
151
7.6%
72,917
0.1%
Transient Non-community
0
0.0%
0
0.0%
Total
1,993
100.0%
60,811,356
100.0%
2010
Community
1,833
93.4%
61,858,371
99.9%
Non-transient Non-community
129
6.6%
71,938
0.1%
Transient Non-community
0
0.0%
0
0.0%
Total
1,962
100.0%
61,930,309
100.0%
2011
Community
1,775
93.9%
62,322,706
99.9%
Non-transient Non-community
116
6.1%
65,806
0.1%
Transient Non-community
0
0.0%
0
0.0%
Total
1,891
100.0%
62,388,512
100.0%
All Years
Community
2,479
87.4%
69,562,352
99.8%
Non-transient Non-community
357
12.6%
145,851
0.2%
Transient Non-community
0
0.0%
0
0.0%
Total
2,836
100.0%
69,708,203
100.0%
Exhibit B.5: Number of Systems and Population Served by Systems in the SYR3
ICR Dataset (2006 - 2011) with TOC Records, by Source Water Type
Year
Source Water Type
Systems
Population
Number
Percent
Number
Percent
2006
Ground Water
345
17.1%
4,807,197
8.8%
GWUDI
47
2.3%
402,823
0.7%
Surface Water
1,631
80.6%
49,643,976
90.5%
Total
2,023
100.0%
54,853,996
100.0%
2007
Ground Water
284
15.0%
5,414,297
9.6%
GWUDI
42
2.2%
360,106
0.6%
Surface Water
1,571
82.8%
50,426,896
89.7%
Total
1,897
100.0%
56,201,299
100.0%
2008
Ground Water
280
14.0%
4,625,754
7.7%
GWUDI
55
2.8%
442,967
0.7%
Surface Water
1,662
83.2%
54,728,122
91.5%
Six-Year Review 3 Technical Support Document B-10
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Year
Source Water Type
Systems
Population
Number
Percent
Number
Percent

Total
1,997
100.0%
59,796,843
100.0%
2009
Ground Water
254
12.7%
4,163,551
6.8%
GWUDI
60
3.0%
532,411
0.9%
Surface Water
1,679
84.2%
56,115,394
92.3%
Total
1,993
100.0%
60,811,356
100.0%
2010
Ground Water
244
12.4%
4,637,952
7.5%
GWUDI
60
3.1%
523,222
0.8%
Surface Water
1,658
84.5%
56,769,135
91.7%
Total
1,962
100.0%
61,930,309
100.0%
2011
Ground Water
179
9.5%
5,068,752
8.1%
GWUDI
63
3.3%
528,104
0.8%
Surface Water
1,649
87.2%
56,791,656
91.0%
Total
1,891
100.0%
62,388,512
100.0%
All Years
Ground Water
775
27.3%
7,980,533
11.4%
GWUDI
95
3.3%
639,322
0.9%
Surface Water
1,966
69.3%
61,088,348
87.6%
Total
2,836
100.0%
69,708,203
100.0%
Note: Purchased systems are included in each category.
Exhibit B.6: Number of Systems and Population Served by Systems in the SYR3
ICR Dataset (2006 - 2011) with TOC Records, by System Size and System Type
Year
Population Served
System Size
Ground Water
Surface Water
Total


Number of
Systems
Population
Served
Number of
Systems
Population
Served
Number of
Systems
Population
Served
Community Water Systems
2006
<101
53
3,094
91
3,067
144
6,161

101 -500
63
17,105
118
33,381
181
50,486

501 - 1,000
15
12,313
98
76,057
113
88,370

1,001 -3,300
34
67,859
327
683,408
361
751,267

3,301 - 10,000
22
131,679
387
2,337,185
409
2,468,864

10,001 -50,000
27
600,603
383
8,938,899
410
9,539,502

50,001 - 100,000
7
503,993
82
5,885,904
89
6,389,897

100,001 - 1 million
18
3,430,697
91
23,713,039
109
27,143,736

> 1 million
-
-
3
8,321,295
3
8,321,295
Six-Year Review 3 Technical Support Document B-ll
for Disinfectants/Disinfection Byproducts Rules
December 2016

-------
Year
Population Served
System Size
Ground Water
Surface Water
Total


Number of
Syste