oEPA
United States
Environmental Protection
Agency
Six-Year Review 3 Technical Support
Document for Microbial Contaminant
Regulations
-------
Office of Water (4607M)
EPA 810-R-16-010
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 Microbial Regulations 3-1
3.1 Surface Water Treatment Rule 3-1
3.1.1 Statutory Authority 3-1
3.1.2 Summary of the Rule 3-2
3.1.3 History of Surface Water Treatment Rule 3-3
3.1.4 The 1996 Safe Drinking Water Act Amendments, M-DBP Advisory
Committee, and Notices of Data Availability 3-9
3.2 Interim Enhanced Surface Water Treatment Rule 3-9
3.3 Filter Backwash Recycling Rule 3-10
3.4 Long-Term 1 Enhanced Surface Water Treatment Rule 3-10
3.5 Long-Term 2 Enhanced Surface Water Treatment Rule 3-10
3.6 Ground Water Rule 3-11
3.6.1 Statutory Authority 3-11
3.6.2 Summary of the Rule 3-11
3.6.3 Hi story of Ground W ater Rule 3-12
3.7 Total Coliform Rule and Revised Total Coliform Rule 3-13
3.8 Summary of Microbial Rules 3-14
4 Health Effects 4-1
4.1 SWTRs 4-1
4.1.1 MCLGs 4-1
4.1.2 Drinking Water-Associated Disease Outbreaks 4-2
4.1.3 GWUDI-Related Public Health Concerns 4-8
4.2 GWR 4-14
5 Analytical Methods 5-1
5.1 Methods for Treatment Technique Requirements Related to Raw and Finished
Water Turbidity (SWTR, IESWTR and LT1) 5-1
5.2 Methods for Measuring Disinfection Residuals (SWTR) and Disinfection
Profiling and Benchmarking (IESWTR, LT1) 5-3
5.2.1 Disinfectant Residuals 5-3
5.2.2 pH 5-9
5.2.3 Temperature 5-9
5.2.4 Heterotrophic Bacteria 5-10
Six-Year Review 3 Technical Support Document i
for Microbial Contaminant Regulations
December 2016
-------
5.3 Methods for Treatment Technique Requirements Related to Filtration
Avoidance (SWTR) 5-10
5.4 Methods for GWUDI Determination (SWTR, IESWTR and LT1) 5-12
5.4.1 Microscopic Particulate Analysis 5-12
5.4.2 Aerobic Spores 5-12
5.5 Methods for Source Water Fecal Indicator Measurement under GWR 5-19
5.6 Methods for Measuring Disinfectant Residuals in Ground Water (GWR) 5-22
6 Occurrence and Exposure 6-1
6.1 SYR3 ICR Microbial Dataset 6-2
6.2 Disinfectant Residuals in Distribution Systems 6-4
6.2.1 Chlorine Residuals for Surface Water Systems 6-8
6.2.2 Chlorine Residuals for Ground Water Systems 6-11
6.2.3 Limitations of Data Analysis 6-13
6.2.4 Considerations for Potential System-Level Analyses 6-13
6.3 Occurrence of Total Coliforms and E. coli as Function of Disinfectant Residual
Types and Levels in Distribution Systems 6-14
6.3.1 Occurrence in Surface Water 6-20
6.3.2 Occurrence in Ground Water 6-22
6.3.3 Limitations of Data Analysis 6-24
6.4 Occurrence of Total Coliforms in PWSs Using Undisinfected Ground Water 6-28
6.5 Occurrence of Viruses and Aerobic Spores in PWSs Using Undisinfected
Ground Water 6-28
7 Treatment 7-1
7.1 Introducti on 7-1
7.2 Disinfectant Residual Requirements in Distribution Systems 7-2
7.2.1 B ackground 7-2
7.2.2 Summary of Technical Review 7-3
7.2.3 Detectable Residuals for Systems Using Chloramine Disinfection 7-3
7.2.4 State Implementation of Disinfectant Residual Requirements 7-4
7.2.5 Disinfectant Residuals for Control of Legionella in Premise Plumbing
Systems 7-7
7.2.6 HPC Alternative to Detectable Residual Measurement 7-8
7.2.7 Research and Information Collection Partnership Findings 7-8
7.3 CT Criteria for Virus Disinfection 7-9
7.3.1 Background 7-9
Six-Year Review 3 Technical Support Document ii
for Microbial Contaminant Regulations
December 2016
-------
7.3.2 Summary of Technical Review 7-10
7.3.3 Basis of CT Values for Virus Inactivation in the EPA Guidance Manual 7-10
7.3.4 Information on Virus Inactivation by Free Chlorine 7-12
7.3.5 Information on Virus Inactivation by Chloramines 7-20
8 References 8-1
Six-Year Review 3 Technical Support Document iii
for Microbial Contaminant Regulations
December 2016
-------
List of Appendices
Appendix A: Data Quality Assurance/Quality Control Documentation for SYR3 ICR Microbial
Data
Appendix B: Additional Analyses on the Disinfectant Residuals in Distribution Systems
Appendix C: Additional Analyses on the Occurrence of TC+ and EC+ in Surface Water and
Ground Water Systems Compared to Disinfectant Residuals in Distribution
Systems
Appendix D: Producing a Reduced Dataset for Undisinfected Ground Water Systems
Appendix E: Analysis of the Generalized Estimating Equation (GEE) and Generalized Linear
Mixed Models (GLMM) as used to Estimate the Relative Rate of Highly Credible
Gastrointestinal Illness (HCGI) by Colford et al. (2009)
Appendix F: Occurrence of Total Conforms/A", coli in Small PWSs Using Undisinfected Ground
Water
Six-Year Review 3 Technical Support Document iv
for Microbial Contaminant Regulations
December 2016
-------
List of Exhibits
Exhibit 2.1: Six-Year Review Protocol Overview and Major Categories of Revise/Take No
Action Outcomes 2-2
Exhibit 3.1: Timeline for Selected Activities Associated with Microbial Regulations for Drinking
Water 3-1
Exhibit 3.2: NPDWRs for Microbial Rules 3-14
Exhibit 4.1: Etiology of Drinking Water-Associated Outbreaks, by Year, in the United States,
1971 to 2012 (CDC, 2015a) 4-3
Exhibit 4.2: Summary of Drinking Water-Associated Outbreaks and Assigned Deficiencies -
United States, 2003-2012 4-4
Exhibit 4.3: Cases of Legionella in the U.S., 2003-2012 4-6
Exhibit 5.1: Turbidity Analytical Methods Approved under the Surface Water Treatment Rule
(§141.74) 5-2
Exhibit 5.2: Primary Disinfectant Residual Analytical Methods Approved under the Surface
Water Treatment Rule (§141.74) 5-5
Exhibit 5.3: pH Analytical Methods Approved via the Expedited Method Approval Process ... 5-9
Exhibit 5.4: Temperature Analytical Method Approved by the Expedited Method Approval
Process 5-10
Exhibit 5.5: Heterotrophic Bacteria Analytical Methods Approved under the Surface Water
Treatment Rule (§141.74) 5-10
Exhibit 5.6: Total Coliform Bacteria Analytical Methods Approved under the Surface Water
Treatment Rule (§141.74) 5-11
Exhibit 5.7: Fecal Coliform Bacteria Analytical Methods Approved under the Surface Water
Treatment Rule (§141.74) 5-11
Exhibit 5.8: Analytical Methods Approved under the Ground Water Rule (§141.402) 5-20
Exhibit 6.1: Conceptual Overview of the Components of the SYR3 ICR Microbial Dataset 6-4
Exhibit 6.2: Counts of Chlorine Residual Data by Source Water Type, System Type and System
Size from SYR3 ICR Dataset (All Years; 2006-2011) 6-5
Exhibit 6.3: Diagram Characterizing Type of Residual Reported 6-7
Exhibit 6.4: Summary Statistics of Free and Total Chlorine Residual Concentrations in Surface
Water, by Year 6-8
Six-Year Review 3 Technical Support Document v
for Microbial Contaminant Regulations
December 2016
-------
Exhibit 6.5: Cumulative Percent of Free and Total Chlorine Residual Concentrations in Surface
Water (in 2011) 6-9
Exhibit 6.6: Free and Total Chlorine Residual - Frequency of Detection in Surface Water (All
Years; 2006-2011) 6-10
Exhibit 6.7: Summary Statistics of Free and Total Chlorine Residual Concentrations in Ground
Water, by Year 6-11
Exhibit 6.8: Cumulative Percent of Free and Total Chlorine Residual Concentrations in Ground
Water (in 2011) 6-12
Exhibit 6.9: Free and Total Chlorine Residual - Frequency of Detection in Ground Water (All
Years; 2006-2011) 6-13
Exhibit 6.10: Counts of Total Coliform and E. coli Records by Source Water Type, System Type
and System Size from SYR3 ICR Dataset (All Years; 2006-2011) 6-16
Exhibit 6.11: Summary of Total Coliform and E. coli Samples for Each Bin of Free and Total
Chlorine Residual Concentrations from SYR3 ICR Dataset (2006-2011) 6-19
Exhibit 6.12: Total Coliforms - Frequency of Detection in Surface Water as Function of
Disinfectant Types and Concentrations (2006-2011) 6-20
Exhibit 6.13: E. coli - Frequency of Detection in Surface Water (2006-2011) 6-21
Exhibit 6.14: Number of Total Coliform and E. coli Samples and Positives in Surface Water
Paired with Free and Total Chlorine Data, by Source Water Type 6-21
Exhibit 6.15: Total Coliforms - Frequency of Detection in Ground Water (2006-2011) 6-23
Exhibit 6.16: E. coli - Frequency of Detection in Ground Water (2006-2011) 6-23
Exhibit 6.17: Number of Total Coliform Samples in Ground Water Paired with Free and Total
Chlorine Data, by Source Water Type 6-24
Exhibit 6.18: Comparison of Free Chlorine Only Samples with Free Chlorine Samples Paired
with Total Chlorine 6-26
Exhibit 6.19: Comparison of Total Chlorine Only Samples with Total Chlorine Samples Paired
with Free Chlorine 6-26
Exhibit 6.20: UCMR 3 Aerobic Spore Concentration Cumulative Distribution Function 6-30
Exhibit 7.1: Distribution System Minimal Residual Requirements by States - Free Chlorine.... 7-5
Exhibit 7.2: Distribution System Minimal Residual Requirements by States - Total Chlorine.. 7-6
Exhibit 7.3: CT Values for Inactivation of HAY at 5 °C 7-11
Six-Year Review 3 Technical Support Document vi
for Microbial Contaminant Regulations
December 2016
-------
Exhibit 7.4: CT Values for Vims Inactivation with 1.0 mg/L of Free Chlorine at 5°C 7-13
Exhibit 7.5: CT Values for Virus Inactivation with 0.2 mg/L of Free Chlorine at 5°C 7-14
Exhibit 7.6: CT Values for 4-Log Inactivation of Cell-Associated and Dispersed HAV at 5°C
7-15
Exhibit 7.7: CT Values for Inactivation of Aggregated and Dispersed AD2 at 5°C and 0.2 mg/L
Free Chlorine in a River Source Water 7-15
Exhibit 7.8: CT Values for 3-Log Virus Inactivation in a River Source Water with 0.2 mg/L of
Free Chlorine 7-16
Exhibit 7.9: CT Values for Inactivation of CB5 with Free Chlorine in Recycled Water at 10°C ....
7-18
Exhibit 7.10: Comparison of CT Values for Inactivation of CB5 with Free Chlorine 7-19
Exhibit 7.11: CT Values for Virus Inactivation with 1 mg/L of Monochloramine at 5°C 7-21
Exhibit 7.12: CT Values for Monochloramine Inactivation of Aggregated and Dispersed AD2 in
River Source Water at 5°C 7-22
Exhibit 7.13: CT Values for Inactivation of AD2 by Chloramines in Recycled Water at 10°C
7-23
Six-Year Review 3 Technical Support Document vii
for Microbial Contaminant Regulations
December 2016
-------
Acronyms
Acronym
Definition
ADOH
Australian Department of Health
AGI
Acute Gastrointestinal Illness
AOC
Assimilable Organic Carbon
APHA
American Public Health Association
ASTM
American Society of Testing and Materials
AWWA
American Water Works Association
BDF
Buffered, Demand-Free Water
CCL
Contaminant Candidate List
CDC
Centers for Disease Control and Prevention
CFR
Code of Federal Regulations
DBP
Disinfection Byproduct
D/DBPR
Disinfectants and Disinfection Byproducts Rules
DNA
Deoxyribonucleic Acid
DOP
Demonstration of Performance
DPD
N, N Diethyl-1,4 Phenylenediamine Sulfate
EA
Economic Analysis
EC
E. coli
EFH
Efficiency Factor Horn
EPA
United States Environmental Protection Agency
FACTS
Free Available Chlorine Testing with Syringaldazine
FBRR
Filter Backwash Recycling Rule
GAO
Government Accountability Office
GEE
Generalized Estimating Equation
GLI
Great Lakes Instruments
GLMM
Generalized Linear Mixed Models
GUP
Purchased Ground Water Under the Direct Influence of Surface Water
GWP
Purchased Ground Water
GWR
Ground Water Rule
GWUDI
Ground Water Under Direct Influence of Surface Water
HAA
Haloacetic Acid
Six-Year Review 3 Technical Support Document viii
for Microbial Contaminant Regulations
December 2016
-------
HAV
Hepatitis A Virus
HCGI
Highly Credible Gastrointestinal Illness
HEA
Health Effects Assessment
HPC
Heterotrophic Plate Count
HSA
Hydrogeologic Sensitivity Assessment
ICR
Information Collection Request
IESWTR
Interim Enhanced Surface Water Treatment Rule
LED
Light-Emitting Diode
LT1
Long-Term 1 Enhanced Surface Water Treatment Rule
LT2
Long-Term 2 Enhanced Surface Water Treatment Rule
MCL
Maximum Contaminant Level
MCLG
Maximum Contaminant Level Goal
MCMC
Markov chain Monte Carlo
MDBP
Microbial and Disinfection Byproducts
MNV
Murine Norovirus
MPA
Microscopic Particulate Analysis
MRDL
Maximum Residual Disinfectant Level
MRDLG
Maximum Residual Disinfectant Level Goal
MUG
4-methylumbelliferyl-P-D-glucuronide
NDWAC
National Drinking Water Advisory Council
NEMI
National Environmental Methods Index
NIH
National Institute of Health
NNDSS
National Notifiable Diseases Surveillance System
NPDWR
National Primary Drinking Water Regulation
NTU
Nephelometric Turbidity Unit
ONPG-MUG
Enzyme Substrate Coliform Test/Colilert
PCCL
Primary Contaminant Candidate List
PWS
Public Water System
qPCR
Quantitative Polymerase Chain Reaction
RICP
Research and Information Collection Partnership
RMCL
Recommended Maximum Contaminant Level
RNA
Ribonucleic Acid
Six-Year Review 3 Technical Support Document ix
for Microbial Contaminant Regulations
December 2016
-------
RTCR
Revised Total Coliform Rule
SAL
Single Agar Layer
SCWA
Sonoma County Water Authority
SDWA
Safe Drinking Water Act
SDWIS
Safe Drinking Water Information System
SWP
Purchased Surface Water
SWTD
Water Source, Treatment Facility or Distribution System
SWTR
Surface Water Treatment Rule
SYR
Six-Year Review
TC
Total Coliforms
TCR
Total Coliform Rule
THM
Trihalomethane
IT
Treatment Technique
UCMR
Unregulated Contaminant Monitoring Rule
USGS
United States Geological Survey
UV
Ultraviolet
WBDO
Waterborne Disease Outbreak
WBDOSS
Waterborne Disease and Outbreak Surveillance System
Six-Year Review 3 Technical Support Document x
for Microbial Contaminant Regulations
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, needs to be
revised.1 The purpose of the review, called the Six-Year Review, is to identify those NPDWRs
for which current health effects assessments, changes in technology, analytical methods,
occurrence and exposure, implementation, and/or other factors that provides a health or technical
basis to support a regulatory revision will improve or strengthen public health protection.
EPA completed and published the results of its first Six-Year Review ("Six-Year Review 1"), on
July 18, 2003 (USEPA, 2003a) and the second Six-Year Review ("Six-Year Review 2"), on
March 29, 2010 (USEPA, 2010a), after developing a systematic approach, or protocol, for the
review of NPDWRs. During Six-Year Review 1, EPA identified the Total Coliform Rule (TCR)
as a candidate for revision. Four additional NPDWRs (acrylamide, epichlorohydrin,
tetrachloroethylene and trichloroethylene) were identified as candidates for revision during the
Six-Year Review 2.
Under the third Six-Year Review ("Six-Year Review 3"), EPA is reviewing the regulated
chemical, radiological and microbiological contaminants included in previous reviews, as well as
the microbial and disinfection byproducts (MDBP) regulations. This is the first time EPA is
conducting a Six-Year Review of 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).
In this document, the SWTR, the IESWTR and the LT1 are collectively referred to as the
SWTRs because of the close association among the three rules (IESWTR and LT1 were
amendments to the SWTR - additional information provided in Chapter 3).
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 Six-Year Review 3
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 the 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 Microbial Contaminant Regulations
December 2016
-------
process. Results from the review of the LT2 are discussed in a separate support document
(USEPA, 2016a).
The remainder of this document provides a summary of available information and data relevant
to determining if any of the microbial contaminant regulations are candidates for revision under
the Six-Year Review. The information cutoff date for Six-Year Review 3 was December 2015.
That is, information published during or before December 2015 was considered as part of the
Six-Year Review 3. The Agency recognizes that scientists and other stakeholders are continuing
to investigate microbial contaminants and publish information subsequent to this cutoff date.
While not considered as part of the Six-Year Review 3, 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 Six-Year Review 3 are provided in Chapter 4
(health effects), Chapter 5 (analytical methods), Chapter 6 (occurrence and exposure) and
Chapter 7 (treatment).
Six-Year Review 3 Technical Support Document 1-2
for Microbial Contaminant Regulations
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 Six-Year Review 3. 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, 2016b). The foundation of this
protocol was developed for the Six-Year Review 1 based on the recommendations of the
National Drinking Water Advisory Council (NDWAC; 2000). This Six-Year Review 3 process is
very similar to the process implemented during the Six-Year Review 1 and the Six-Year Review
2, 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 Six-Year Review 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 Six-Year Review
are: maximum contaminant level goals (MCLGs), maximum contaminant levels (MCLs),
maximum residual disinfectant level goals (MRDLGs), maximum residual disinfectant levels
(MRDLs), treatment techniques, other treatment technologies and regulatory requirements (e.g.,
monitoring). The MCL provisions are not applicable for evaluation of the microbial
contaminants regulations which establish treatment technique requirements in lieu of MCLs. The
MRDLG and MRDL provisions are only applicable for evaluation of the Disinfectants and
Disinfection Byproducts Rules (D/DBP) rules as part of the Six-Year Review.
The review elements that EPA considered for each NPDWR during the Six-Year Review 3
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,
2016b).
The Initial Review branch of the protocol identifies NPDWRs with recent or ongoing actions and
excludes them from the review process to prevent duplicative agency efforts (USEPA, 2016b).
The cutoff date for the NPDWRs reviewed under the Six-Year Review 3 was August 2008.
Based on the Initial Review, EPA excluded the Aircraft Drinking Water Rule, which was
promulgated in 2009, and the Revised Total Coliform Rule (RTCR) (the revision of the 1989
TCR), which was promulgated in 2013. Further, since most of the 1989 TCR requirements were
replaced by the 2013 RTCR, the 1989 TCR was excluded from the Six-Year Review 3.
Six-Year Review 3 Technical Support Document 2-1
for Microbial Contaminant Regulations
December 2016
-------
Exhibit 2.1: Six-Year Review Protocol Overview and Major Categories of
Revise/Take No Action Outcomes
Yes
No
Yes
No
No
Outcome:
No action
at this time
Uncertain - emerging
information
Yes
No
Yes
No
Ongoing or planned HEA
No new information
NPDWR reviewed in recent or ongoing action?
Regulatory action ongoing
or recently completed
NPDWRs Under Review
Data gaps/emerging
information
Low priority - No meaningful
opportunity
Health effects assessment (HEA)
in process or planned? *
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?
* 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.
Outcome:
Candidate
for Revision
Six-Year Review 3 Technical Support Document 2-2
for Microbial Contaminant Regulations
December 2016
-------
3 History of Microbial Regulations
This chapter provides a brief history of microbial contaminant regulations in the United States
from 1975 to 2016. A timeline of selected events in the statutory and regulatory history, and
regulatory review processes is shown in Exhibit 3.1. The microbial contaminant regulations
covered in this Six-Year Review include: the Surface Water Treatment Rule (SWTR), the
Interim Enhanced Surface Water Treatment Rule (IESWTR), the Long-Term 1 Enhanced
Surface Water Treatment Rule (LT1), the Long-Term 2 Enhanced Surface Water Treatment Rule
(LT2), the Filter Backwash Recycling Rule (FBRR), and the Ground Water Rule (GWR). The
Total Coliform Rule (TCR) and the Revised Total Coliform Rule (RTCR) are not being reviewed
under the Six-Year Review 3; therefore, they are described only briefly in this chapter. Note that
the LT2 is discussed in more detail in a separate support document (USEPA, 2016a).
Exhibit 3.1: Timeline for Selected Activities Associated with Microbial Regulations
for Drinking Water
1989
Final SWTR and TCR
1997
M-DBP Advisory Committee Established
1996 2002
SDWA Amendments ,/ Final LT1ESWTR
2007
TCR Distribution System Advisory Committee Established
1986
SDWA Amendments
1994
Proposed IESWTR
2000
Proposed GWR
1998
Final IESWTR
1985
Proposed recommended
MCLs for turbidity, Giardia
lamblia, and viruses
1987
Proposed TCR and SWTR
1992
Negotiating Committee Established
1996
Final Information Collection Rule
2006
GWR Notice of Data Availability
2003
Six-Year Review 1
2010
Proposed RTCR
2001
Final FBRR
2003
Proposed LT2ESWTR
Final GWR
2016
Six-Year Review 3
2013
Final RTCR
2000
Proposed LT1ESWTR and FBRR
1997
2010
2006 Six-Year Review 2
Final LT2ESWTR
IESWTR Notice of Data Availability
EPA is also reviewing the Stage 1 and Stage 2 Disinfectant and Disinfection Byproducts Rules
(D/DBPRs) as part of the Six-Year Review 3. See a separate support document for more
information about these rules (USEPA, 2016f).
3.1 Surface Water Treatment Rule
3.1.1 Statutory Authority
The 1974 Safe Drinking Water Act (SDWA) authorized EPA to protect public health by
regulating the nation's public drinking water supply. Although the SDWA was amended slightly
in 1977, 1979 and 1980, the most significant changes occurred when the SDWA was
reauthorized in 1986 and amended in 1996. To safeguard public health, the 1986 amendments
required EPA to set maximum contaminant level goals (MCLGs) and maximum contaminant
level (MCLs) for 83 contaminants.2 The 1986 amendments authorized EPA to promulgate
2 An MCLG is the level of a contaminant in drinking water below at which no known or anticipated adverse effect on the health
of persons would occur. MCLGs allow for a margin of safety and are non-enforceable public health goals.
Six-Year Review 3 Technical Support Document 3-1
for Microbial Contaminant Regulations
December 2016
-------
NPDWRs in the form of treatment techniques instead of MCLs where appropriate. EPA was also
required to establish regulations for disinfection of all public water supplies and to specify
filtration requirements for water systems that draw water from surface sources (USEPA, 1991a).
The disinfection and filtration requirements were intended to protect the public from potential
adverse health effects due to exposure to Giardia lamblia, viruses, Legionella, heterotrophic
bacteria, and other pathogens that would be removed by those treatment techniques. The 1996
amendments are discussed in more detail later in this chapter.
3.1.2 Summary of the Rule
In response to the 1986 reauthorization of the SDWA, EPA promulgated the SWTR in 1989 (for
more information about microbial rules prior to the SWTR, the reader is referred to Regli et al.,
2003). The SWTR set MCLGs for Legionella, Giardia lamblia,3 and viruses at zero since any
exposure to these microbial pathogens presents a health risk. It required most systems using
surface water or ground water under the direct influence of surface water (GWUDI) (also known
as Subpart H systems, meaning subject to the requirements of Subpart H of 40 CFR Part 141) to
remove and inactivate microbial contaminants through filtration and/or disinfection, respectively
(USEPA, 1989).
To measure the performance of filtration systems, systems were required to monitor the turbidity
of finished (treated) water. Specifically, the rule establishes treatment technique requirements for
Subpart H systems to control for Giardia lamblia and viruses by at least 99.9 percent (3-log) and
99.99 percent (4-log) removal, respectively. For a few systems with sufficiently high quality
source water and protective watershed control programs, the treatment requirement could be
achieved by using disinfection only. However, those systems must meet the 3- and 4-log
requirements through disinfection, as well as additional source water protection requirements.
The SWTR also established requirements for disinfectant residuals. In both filtered and
unfiltered systems, the residual disinfectant concentration at the entry point to the distribution
system may not be less than 0.2 mg/L for more than four hours. The main purpose of this
requirement was to ensure continuity of disinfection. The SWTR also requires a detectable
disinfectant residual or heterotrophic plate count (HPC) of 500/mL or less to be maintained
throughout the distribution system in at least 95 percent of the measurements made (USEPA,
1989). The filtration and disinfection requirements of the SWTR were intended to protect against
the potential adverse health effects of Giardia, viruses, Legionella, and heterotrophic bacteria, as
well as many other pathogenic organisms that are removed by these treatment techniques.
EPA published a guidance manual to support the SWTR; it recommends various combinations of
log-inactivation and log-removal of pathogenic organisms (USEPA, 1991b). Under the SWTR,
the state is required to develop and implement enforceable criteria by which systems demonstrate
they are achieving at least 3-log removal and/or inactivation of Giardia and 4-log removal and/or
inactivation of viruses. Essentially, all states used the recommendations of the SWTR guidance
manual to allot "credits" for filtration removal and disinfection inactivation to filtered systems,
3 The current preferred taxonomic name is Giardia duodenalis, with Giardia lamblia and Giardia intestinalis as synonyms.
However, Giardia lamblia was the name used to establish the MCLG in 1989. Elsewhere in this document this pathogen will be
referred to as Giardia spp. or simply Giardia unless discussing information on an individual species.
Six-Year Review 3 Technical Support Document 3-2
for Microbial Contaminant Regulations
December 2016
-------
which together demonstrate achievement of the removal and inactivation requirements (USEPA,
1991b).
3.1.3 History of Surface Water Treatment Rule
Prior to the 1989 SWTR, filtration and disinfection were not specifically required under federal
law, although the majority of surface water systems used these treatment technologies. However,
based on authority provided by the 1974 SDWA, EPA established interim MCLs in 1975 for
turbidity: the monthly average turbidity MCL was 1 nephelometric turbidity unit (NTU), and the
two-day average was 5 NTU (USEPA, 1976).
In November 1985, EPA proposed "recommended MCLs" (RMCLs, forerunners to MCLGs) for
turbidity, Giardia lamblia, and viruses and solicited comment on the appropriateness of
establishing RMCLs and NPDWRs for Legionella and HPC bacteria (USEPA, 1985).
In November 1987, EPA re-proposed MCLGs for Giardia and viruses (specifically enteric
viruses), proposed an MCLG for Legionella, and proposed a regulation specifying criteria under
which filtration would be required as a treatment technique (USEPA, 1987). The MCLGs for
Giardia and viruses were re-proposed to address the change in terminology (from RMCL to
MCLG) required by the 1986 SDWA amendments and to specify the types of viruses to be
included; the values themselves did not change. Along with these criteria, EPA proposed
procedures the states would use to determine which systems must install filtration. EPA also
proposed disinfection treatment technique requirements for public water systems using surface
water sources. The 1987 notice also withdrew the 1985 proposed RMCL for turbidity and instead
proposed turbidity criteria for determining whether a public water system is required to filter and
determining whether filtration alone, if required, is adequate (USEPA, 1987).
In May 1988, EPA published a notice of availability that solicited specific data, discussed
alternatives to the proposed surface water treatment requirements and solicited comment on these
alternative options (USEPA, 1988). For instance, EPA proposed alternative disinfectant residual
monitoring requirements for systems serving fewer than 500 people to allow these systems to
collect and analyze one grab sample of disinfectant residual instead of monitoring continuously.
The final SWTR was promulgated on June 29, 1989; specific components of that rule are
described in more detail in Sections 3.1.3.1 to 3.1.3.4. Additional historical information related
to the SWTR is in Regli et al. (2003).
3.1.3.1 Definitions of Surface Water and Ground Water Under Direct Influence of
Surface Water
As part of the development of the SWTR, EPA needed to clarify which systems would be
regulated under Subpart H. In particular, EPA needed to clarify when systems that could be
considered as ground water systems, were more appropriate to regulate as surface water systems
(for example, systems where the drinking water intake was in a riverbed, not in the river). Thus,
to identify a system as either ground or surface water, the SWTR defined "ground water under
the direct influence of surface water (GWUDI)." GWUDI is any water beneath the surface of the
ground with: (1) significant occurrence of insects or other macroorganisms, algae or large-
Six-Year Review 3 Technical Support Document 3-3
for Microbial Contaminant Regulations
December 2016
-------
diameter pathogens such as Giardia lamblia, or (2) significant and relatively rapid shifts in water
characteristics such as turbidity, temperature, conductivity or pH that closely correlate to
climatological or surface water conditions. The final SWTR defined GWUDI as being regulated
as surface waters because Giardia contamination of infiltration galleries, springs and wells have
been found (Hoffbuhr et al., 1986; Hibler et al., 1987). Some contamination of springs and wells
have resulted in giardiasis outbreaks (Craun and Jakubowski, 1986). Direct influence was to be
determined for individual sources in accordance with criteria established by the state (54 FR
27486, USEPA, 1989). The GWUDI designation identifies PWSs using ground water that must
be regulated as if they are surface water systems. All other PWSs using ground water are
regulated by the GWR.
The 1998 IESWTR expanded the definition of GWUDI for systems serving 10,000 or more
people to include Cryptosporidium, and the 2002 LT1 included Cryptosporidium in the GWUDI
definition for systems serving fewer than 10,000 people.
The definition of GWUDI relies heavily on water quality parameters to indicate whether the
source is at risk for Giardia or Cryptosporidium to pass from surface water to the ground water
collector. It assigns the determination to state primacy programs and includes suggested elements
of the decision-making process. The complete definition of GWUDI in 40 CFR 141.2 is:
"Ground water under the direct influence of surface water (GWUDI) means any water
beneath the surface of the ground with significant occurrence of insects or other
macroorganisms, algae, or large-diameter pathogens such as Giardia lamblia or
Cryptosporidium, or significant and relatively rapid shifts in water characteristics such
as turbidity, temperature, conductivity, or pH which closely correlate to climatological or
surface water conditions. Direct influence must be determinedfor individual sources in
accordance with criteria established by the State. The State determination of direct
influence may be based on site-specific measurements of water quality and/or
documentation of well construction characteristics and geology with field evaluation. "
The special primacy provision requirements of 40 CFR 142.16(b)(2)(B) specify that the state
application for primacy program revision approval must include a description of how the state
will accomplish the determination of which systems using a ground water source are GWUDI.
The requirements also specify that the determinations had to be completed by June 29, 1994 for
community water systems and by June 29, 1999 for non-community water systems. Federal
regulations do not include GWUDI classification re-evaluation requirements nor ongoing
monitoring of source water quality. State programs can impose such requirements.
Public Health Protection Goals from the SWTR Definition of Surface Water and GWUDI
EPA originally established the GWUDI source water classification to address the public health
concern posed by an underground source of drinking water that is subject to Giardia (and
subsequently Cryptosporidium) contamination from surface waters. Giardia and
Cryptosporidium pose significant health risks for systems using ground water closely connected
to surface water because they are not removed from water by natural filtration processes in the
course of the water's passage from surface water through the subsurface to the well. Because
Six-Year Review 3 Technical Support Document 3-4
for Microbial Contaminant Regulations
December 2016
-------
Cryptosporidium is not readily inactivated by disinfectants other than ultraviolet (UV) light, it
poses a greater public health threat than Giardia.
The 2006 GWR addresses public health protection against bacterial and viral pathogens in PWSs
that use ground water not subject to Giardia or Cryptosporidium contamination and regulated
accordingly. PWSs regulated under the GWR are not required to filter and may or may not be
disinfected. Currently, about 86,000 PWSs serving about 20 million people are undisinfected
(USEPA, 2013a).
Some PWS wells, regulated under the GWR, were subsequently found to be contaminated by or
at risk of contamination by Giardia or Cryptosporidium as a result of outbreaks (Bergmire-Sweat
et al., 1999; Lee et al., 2001; Daly et al., 2010). These wells were misclassified as ground water
and should have been determined to be GWUDI. A well misclassified as ground water rather
than GWUDI may pose a public health hazard because the well water may be inadequately
treated, receiving either no treatment or only disinfection, rather than disinfection combined with
engineered filtration or an approved alternative based on a demonstration of performance (DOP).
Reduced PWS misclassification will result in improved public health protection because fewer
people will be exposed to Giardia or Cryptosporidium via untreated or inadequately treated
(disinfected but unfiltered) ground water. Full protection against Giardia and Cryptosporidium
can only result if a well is properly regulated as GWUDI and subject to the SWTR requirements.
GWUDI Classification Principles
Although the origin of some water molecules can be ascertained based on measuring the tritiated
water component, it is otherwise impossible to identify whether a water molecule emanating
from a well originated in surface water or in ground water. Scientific studies determine water
molecule origins (water flowpath reverse tracking) by using surrogate measures such as various
combinations of stable isotopes, dissolved solids, entrained solid particles, bioindicator particles,
temperature, dyes and other dissolved and particulate tracers, and mathematical models. No
single measurement or surrogate measure is unequivocal. Thus, a scientific determination of
surface water exchange with ground water can be lengthy, time consuming, and expensive.
The transport of pathogens similar in size to Giardia (8 to 12 |im) or Cryptosporidium (4 to 6
\im) from surface water to a collector of an underground source of water requires: a) a hydraulic
connection between the waters, b) high-to-low hydraulic gradient in the direction from surface
water to the collector, however transient, and c) insufficient natural filtration to remove the
pathogens. Temporary GWUDI periods can occur due to seasonal or intermittent induced surface
water recharge or increases in average ground water velocities that decrease the amount of
natural filtration provided by the aquifer's materials. A pumping well can increase average
ground water velocity and alter or even reverse the ground water flow path if the natural
hydraulic gradient from ground water to surface water is low or if the well pumps sufficient
volumes of water to induce recharge of the subsurface aquifer by surface water.
GWUDI determination requires a regulatory decision regarding the public health risk resulting
from a poorly understood, difficult to measure, and potentially continuously changing
hydrogeologic process. Because the existing definition of GWUDI refers to "significant
occurrence" of specific types of organisms and particulates that are believed to originate from
Six-Year Review 3 Technical Support Document 3-5
for Microbial Contaminant Regulations
December 2016
-------
surface water, and "significant or relatively rapid shifts" in water characteristics that correlate to
surface water conditions, state determination of GWUDI must necessarily interpret factors
deemed "significant." This determination is made on a case-by-case basis for each water source.
Most states were required to make GWUDI determinations for large numbers of wells.
To be effective, the GWUDI determination definition and implementing methodology must be
simple and inexpensive despite the inherent complexity of the relationship between ground water
and surface water. Because of analytical limitations, assaying for Giardia or Cryptosporidium as
indicators of GWUDI is not cost effective and significant public health risk could be present
even in the absence of pathogen recovery in one or more samples (e.g., Messner and Berger,
2015). Section 5.4 provides information on the methods for GWUDI determination.
3.1.3.2 Disinfectant Residuals in the Distribution System
In the proposed SWTR, EPA proposed to require all systems using surface water (both filtered
and unfiltered) to maintain at least a 0.2 mg/L disinfection residual in at least 95 percent of the
distribution system samples taken each month. If a system failed to comply with this requirement
for any two consecutive months, it would be in violation of a treatment technique requirement.
Also, unfiltered systems failing to meet this criterion would be required to filter. The purpose of
this requirement was to limit contamination from outside the distribution system; limit growth of
heterotrophic bacteria and Legionella within the distribution system; and provide a quantitative
limit that, if exceeded, would trigger remedial action (USEPA, 1987).
EPA proposed a minimum disinfectant residual of 0.2 mg/L and concluded that such a level was
feasible for most well-operated systems (USEPA, 1989). However, public comments indicated
that, for many systems that are well operated (as evidenced by low levels of HPC in routine
monitoring), it was not feasible to maintain the proposed minimum disinfectant residual without
significantly changing existing disinfection practice (e.g., increasing existing chlorine dosages or
switching to chloramine disinfection for the distribution system).
Based on these comments and additional information about disinfection practice at the time, EPA
revised the proposed SWTR. The final SWTR requires "detectable" residuals in the distribution
system in lieu of residuals of at least 0.2 mg/L. Residual concentrations can be measured as free
chlorine, total chlorine, combined chlorine (total chlorine minus free chlorine), or chlorine
dioxide. The absence of a residual at a site within the distribution systems indicates that the
disinfectant level has been reduced, possibly as a result of localized contamination from outside
the distribution system or from organic or inorganic materials within the distribution system.
EPA recognized that the absence of a disinfectant residual at a distribution system site does not
necessarily indicate microbiological contamination; such contaminants simply may not be
present, even in the absence of a disinfectant residual. In other words, if microbial occurrence is
low, the lack of a disinfectant residual is not a concern. Thus, under the final SWTR, sites that do
not have "detectable" residuals, but have HPC measurements of 500/mL or less, are considered
equivalent to sites with "detectable" residuals for purposes of determining compliance (USEPA,
1989) (refer to Chapter 5 for a list of methods approved for measurement of HPC).
Six-Year Review 3 Technical Support Document 3-6
for Microbial Contaminant Regulations
December 2016
-------
The final rule requires a 0.2 mg/L disinfectant residual at the entry point to the distribution
system. The residual disinfectant concentration at that location may not drop below 0.2 mg/L for
more than four hours.
For systems using only surface water (or GWUDI) sources, the SWTR requires monitoring for
disinfectant residual concentrations at the same locations and at the same times as total coliforms
are sampled under the TCR, or RTCR as of April 2016. For systems that have both ground water
(which may not be disinfected) and surface waters entering the distribution system, the state may
allow monitoring for disinfectant residuals at points other than the sampling locations for total
coliforms if such points are more representative of the treated (disinfected) surface water within
the distribution system (USEPA, 1989).
For systems that cannot maintain a detectable disinfectant residual in the distribution system, if
the state determines that a system has no means for having a sample transported and analyzed for
HPC by a certified laboratory and adequate disinfection is provided by that system, the
requirement to maintain a detectable disinfection residual does not apply. The state's judgment
might be based upon considerations such as knowledge of the public water system's distribution
system, maintenance of a cross-connection control program, source water quality, and/or past
coliform monitoring results.
The SWTR requires continuous monitoring of the residual disinfectant concentration of the water
entering the distribution system for systems serving more than 3,300 people, and the lowest
value must be recorded each day, except that if there is a failure in the continuous monitoring
equipment, systems may conduct grab sampling every four hours for no more than five days
following the equipment failure (USEPA, 1989). Systems serving 3,300 or fewer people may
take grab samples rather than monitoring continuously; sample frequencies range from one to
four times per day and depend on population.
3.1.3.3 CT Values in Unfiltered Systems
The final SWTR requires 99.9 percent inactivation of Giardia and 99.99 percent inactivation of
viruses in unfiltered systems. Under the proposed SWTR, a system would have been required to
calculate CT, where "T" is disinfectant contact time, the time in minutes it takes the water to
move between the point of disinfectant application and a point before or at the first customer
during peak hourly flow, and "C" is the residual disinfectant concentration in mg/L before or at
the first customer but at or after the point where contact time is measured (USEPA, 1989).
In May 1988, EPA published a notice of data availability (USEPA, 1988) soliciting comments on
a different methodology to determine CT values for systems using ozone. This methodology
would have allowed ozone concentrations to be measured as an average across the contact basin
rather than at only the basin effluent, allowing systems to capture more accurate concentration
data and account for the fact that ozone concentrations were likely to be low at the effluent
location due to ozone's high reactivity. All the commenters who addressed this issue supported
the adoption of this provision in the final rule (USEPA, 1989). In addition, many commenters
suggested applying this provision to all disinfectants. EPA agreed that this methodology results
in a more accurate representation of actual disinfection conditions, especially in systems having
source waters with a high oxidant demand, and those systems using ozone (because it dissipates
Six-Year Review 3 Technical Support Document 3-7
for Microbial Contaminant Regulations
December 2016
-------
very rapidly). Accordingly, EPA adopted this methodology for all disinfectants in the final
SWTR (USEPA, 1989).
Although the final SWTR provides CT value tables for free chlorine, ozone, chlorine dioxide,
and chloramines for Giardia and viruses, EPA recognized that research in this field is ongoing
and included a provision in the final rule that allows unfiltered systems using a disinfectant other
than chlorine to demonstrate, by whatever means allowed by the state, that they are consistently
meeting the 99.9 and 99.99 percent removal and/or inactivation requirements (USEPA, 1989).
Such systems do not have to meet the CT values in the rule. However, note that the LT2 includes
additional CT requirements for ozone, chlorine dioxide, and UV light to allow unfiltered systems
to meet inactivation requirements for Cryptosporidium (USEPA, 2016a).
The SWTR does not require compliance with these CT value tables for filtered systems. Filtered
systems are expected to meet the removal/inactivation requirements (as discussed earlier in
Section 3.1.2) through a combination of disinfection and filtration, which will vary by system.
3.1.3.4 Filtration and Filtration Avoidance
The SWTR required filtered systems to meet turbidity criteria as part of a treatment technique.
For systems using conventional treatment or direct filtration (direct filtration is similar to
conventional filtration but does not include a sedimentation step), turbidity of samples
representative of the filtered water had to be less than or equal to 0.5 NTU in at least 95 percent
of the measurements taken each month (USEPA, 1989). The state was authorized to allow a
turbidity limit of 1 NTU if the system could demonstrate that it was still capable of achieving the
required removal and inactivation. At no time was turbidity to exceed 5 NTU. These turbidity
requirements were later modified under the IESWTR and LT1.
Under the SWTR, systems using slow sand and diatomaceous earth filtration must meet turbidity
limits of 1 NTU in at least 95 percent of samples taken each month, although slow sand systems
may apply to the state for a higher limit (USEPA, 1989). At no time can turbidity in slow sand
and diatomaceous earth systems exceed 5 NTU. These requirements were not altered by the
IESWTR or LT1.
The SWTR allowed systems using other filtration technologies not listed earlier in this section to
demonstrate through pilot studies or other means that they met the removal and inactivation
requirements (USEPA, 1989). Systems able to make such demonstrations were required to
comply with the same turbidity limits as slow sand filtration systems. These requirements were
later modified under the IESWTR and LT1 (see Sections 3.2 and 3.4).
To avoid filtration, surface water or GWUDI systems must meet certain source water quality
conditions (USEPA, 1989). Source water concentrations of fecal coliform must be 20/100 mL or
less, or total coliform concentrations must be 100/100 mL or less. Source water turbidity cannot
exceed 5 NTU except in unusual and unpredictable circumstances, and such occurrences may not
happen more than five times in ten years. The system must have redundant disinfection
components and must meet all the disinfection requirements described in earlier sections. The
system must have a watershed control program approved by the state and be subject to an annual
on-site inspection to assess the program and disinfection treatment process. The system may not
Six-Year Review 3 Technical Support Document 3-8
for Microbial Contaminant Regulations
December 2016
-------
have been identified as a source of a waterborne disease outbreak. It must have been in
compliance with the MCL for total coliforms (or the new MCL for E. coli under the RTCR as
described in Section 3.7).
3.1.4 The 1996 Safe Drinking Water Act Amendments, M-DBP 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, treatment techniques,
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 treatment techniques 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 treatment techniques "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 treatment technique or
process that would be employed to attain the maximum contaminant level or levels" (section
1412(b)(5)(B)). Section 1412(b)(8) of the SDWA, as amended on August 6, 1996, modified the
language of the 1986 amendments, directing EPA to promulgate regulations requiring
disinfection as a treatment technique as necessary for ground water systems (see Section 3.6).
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 (M-DBP) 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, 1998a).
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).
3.2 Interim Enhanced Surface Water Treatment Rule
In response to a massive 1993 cryptosporidiosis outbreak in Milwaukee, Wisconsin (and in
response to the 1996 SDWA amendments), EPA promulgated a rule that built onto the SWTR
but focused on Cryptosporidium control, called the IESWTR. Because Cryptosporidium oocysts
are not inactivated by traditional disinfectants such as chlorine, the rule instituted more stringent
filtration requirements. While cryptosporidiosis is generally a self-limiting disease, with
complete recovery in otherwise healthy persons, the disease can have very serious consequences
in sensitive populations. The IESWTR was promulgated December 16, 1998 (USEPA, 1998a). It
established an MCLG of zero for Cryptosporidium and required 99 percent (2-log) inactivation
or removal of Cryptosporidium for filtered systems serving 10,000 people or more. The rule also
Six-Year Review 3 Technical Support Document 3-9
for Microbial Contaminant Regulations
December 2016
-------
added Cryptosporidium to the definition of GWUDI and to the watershed control requirements
for unfiltered public water systems; added requirements for covering new finished water
reservoirs; required sanitary surveys for all surface water systems regardless of size; and
included disinfection benchmarking provisions to assure continued levels of microbial protection
while facilities took steps to comply with new disinfection byproduct standards under the Stage 1
D/DBPR, also promulgated in December 1998 (USEPA, 1998a, 1998b). Disinfection
benchmarking required systems with disinfection byproduct levels exceeding certain thresholds
to develop a disinfection profile (a record of log inactivation of Giardia achieved over one year)
and to then determine the lowest monthly average inactivation during that period. This
information was to be submitted to the state if the system proposed to change disinfection
practices to comply with the Stage 1 D/DBPR. Systems using ozone or chloramines for primary
disinfection were required to develop a benchmark for viruses as well.
For systems serving 10,000 or more and using conventional or direct filtration, the revised
turbidity limits were 0.3 NTU for filtered water samples in 95 percent of the samples taken each
month and no more than 1 NTU at any time. For systems using alternative filtration in which
systems demonstrate to the state that they meet the removal and inactivation requirements, the
turbidity limits were to be set by the state. The IESWTR also required continuous monitoring of
turbidity in the effluent from individual filters (USEPA, 1998a).
3.3 Filter Backwash Recycling Rule
The purpose of the FBRR, promulgated June 8, 2001, is to further protect public health by
requiring PWSs, where needed, to institute changes to the return of recycle flows to a plant's
treatment process that may otherwise compromise microbial control (USEPA, 2001). The rule
addresses a statutory requirement of the 1996 SDWA amendments to promulgate a regulation
that governs the recycling of filter backwash water within the treatment process of PWSs. It
applies to all surface water and GWUDI systems using direct or conventional filtration.
The FBRR requires that recycled filter backwash water, sludge thickener supernatant, and liquids
from dewatering processes be returned to a location such that all processes of a system's
conventional or direct filtration, as defined in 40 CFR 141.2, are employed unless the state
approved an alternate location by June 8, 2004 (40 CFR 141.76(c)).
3.4 Long-Term 1 Enhanced Surface Water Treatment Rule
The LT1, promulgated January 14, 2002, applies to public water systems that use surface water
or GWUDI and serve fewer than 10,000 persons. The LT1, with some minor variations, extends
the requirements of the IESWTR to small systems (USEPA, 2002).
3.5 Long-Term 2 Enhanced Surface Water Treatment Rule
The LT2, promulgated on January 5, 2006, requires 2- to 3-log inactivation of Cryptosporidium
in unfiltered systems and additional treatment for Cryptosporidium in filtered systems based on
the results of source water monitoring. The rule also requires covering of all uncovered finished
water reservoirs, unless systems treat reservoir effluent to provide at least 99.99 percent (4-log)
inactivation or removal of viruses, 99.9 percent (3-log) inactivation or removal of G. lamblia and
Six-Year Review 3 Technical Support Document 3-10
for Microbial Contaminant Regulations
December 2016
-------
99 percent (2-log) inactivation or removal of Cryptosporidium (USEPA, 2006a). Additional
information about the LT2 is discussed in a separate support document (USEPA, 2016a).
3.6 Ground Water Rule
3.6.1 Statutory Authority
As mentioned in Section 3.1.1, the 1986 SDWA amendments directed EPA to promulgate
regulations requiring disinfection at all PWSs, including those using ground water as a source.
Although EPA began developing a rule requiring disinfection in ground water systems in 1992,
releasing a "strawman" draft rule for comment on July 31, 1992 (USEPA, 1992), it did not
finalize the rule prior to the 1996 SDWA amendments. The 1996 SDWA amendments directed
EPA to promulgate regulations requiring disinfection as a treatment technique as necessary for
ground water systems. In addition, section 1412(b)(8) required EPA to promulgate criteria for
determining whether disinfection should be required as a treatment technique for any PWS
served by ground water. The GWR implements section 1412(b)(8) of the SDWA, as amended,
by establishing a regulatory framework for determining which ground water systems are
susceptible to fecal contamination and requiring those systems to implement corrective actions.
3.6.2 Summary of the Rule
EPA promulgated the GWR on November 8, 2006 to provide for increased protection against
microbial pathogens, specifically viral and bacterial pathogens, in PWSs that use ground water
sources. EPA was particularly concerned about ground water systems that are susceptible to fecal
contamination because these systems may be at risk of supplying water that contains harmful
microbial pathogens. Viral pathogens found in ground water systems may include enteric viruses
such as echovirus, coxsackieviruses, hepatitis A and E, rotavirus, and noroviruses (i.e., Norwalk-
like viruses). Enteric bacterial pathogens may include Escherichia coli (most is. coli is harmless
but a few strains are pathogenic, including E. coli 0157:H7), Salmonella species, Shigella
species and Vibrio cholerae.
The GWR established a risk-targeted approach to identify ground water systems susceptible to
fecal contamination and requires action to correct significant deficiencies and source water fecal
contamination in ground water systems (USEPA, 2006b). This risk-targeting strategy includes
the following:
Regular ground water system sanitary surveys to check for significant deficiencies;
A flexible program for identifying higher risk systems through TCR monitoring and state
determinations;
Ground water source monitoring to detect fecal contamination at certain ground water
systems that do not provide 4-log treatment of viruses; and
Measures to protect public health:
Six-Year Review 3 Technical Support Document 3-11
for Microbial Contaminant Regulations
December 2016
-------
- Treatment technique requirements to address sanitary survey significant deficiencies
and fecal contamination in ground water; and
- In systems providing treatment, compliance monitoring to ensure that 4-log treatment
of viruses is maintained.
Treatment technique requirements consist of implementation of one or more of the following
corrective action options: correct all significant deficiencies; provide an alternate source of
water; eliminate the source of contamination; or provide treatment that reliably achieves at least
99.99 percent (4-log) treatment of viruses (using inactivation, removal, or a state-approved
combination of 4-log virus inactivation and removal) for each ground water source (USEPA,
2006b). In addition, ground water systems must inform their customers of any uncorrected
significant deficiencies or fecal indicator-positive ground water source samples.
3.6.3 History of Ground Water Rule
Prior to the GWR, no federal regulation required either monitoring of ground water sources or
corrective action upon finding fecal contamination or identifying a significant deficiency during
a sanitary survey. In addition, a U.S. Government Accountability Office (GAO) report (1993)
found that many sanitary surveys did not evaluate one or more of the components that EPA
recommended be evaluated, and that efforts to ensure correction were often limited.
In addition, according to reports by the Centers for Disease Control and Prevention (CDC)
between 1991 (the year in which the TCR became effective) and 2000, ground water systems
were associated with 68 waterborne disease outbreaks that caused 10,926 illnesses (Moore et al.,
1993; Kramer et al., 1996; Levy et al., 1998; Barwick et al., 2000; and Lee et al., 2002). The
major deficiencies identified by the CDC as the likely cause of the outbreaks were source water
contamination and inadequate treatment (or treatment failures).
EPA began developing the GWR in 1987 with the intention of requiring across-the-board
disinfection, as directed by the 1986 SDWA amendments. A preliminary public meeting on
issues related to ground water systems was held in 1990. By 1992, EPA had developed a draft
proposed rule (a "strawman") (USEPA, 1992), which was made available for stakeholder review
upon request. Most stakeholders who commented were concerned that the rule was crafted so
that all ground water systems were assumed to be contaminated until monitoring proved
otherwise and that disinfection waivers would be difficult to obtain (USEPA, 2006b).
In response to the 1996 SDWA amendments, EPA began to consider a new approach in which
disinfection would not be mandatory for all ground water systems (USEPA, 2006b). This
approach focused primarily on establishing a reasonable means for determining if a ground water
source was vulnerable to fecal contamination (USEPA, 2006b). EPA evaluated the possibility of
developing a vulnerability assessment tool that would consider hydrogeologic information and
sources of fecal contamination.
The proposed GWR was published in the Federal Register on May 10, 2000 (USEPA, 2000).
The primary elements of the proposed GWR were sanitary surveys, triggered monitoring,
hydrogeologic sensitivity assessments (HSAs), routine source water monitoring, corrective
Six-Year Review 3 Technical Support Document 3-12
for Microbial Contaminant Regulations
December 2016
-------
action and compliance monitoring. The proposed rule would require states to identify high
priority systems through an HSA; wells located in karst, fractured bedrock, or gravel
hydrogeologic settings were considered sensitive (USEPA, 2000). These wells are potentially at
risk of fecal contamination because ground water velocities are high and fecal contamination can
travel long distances over a short time. Systems in sensitive areas would have been required to
conduct monthly routine monitoring. If a system did not have any fecal indicator-positive
samples after twelve monthly samples, the state would have been allowed to reduce routine
source water monitoring to quarterly. States would also have been allowed, after the first year of
monthly samples, to waive source water monitoring altogether for a system if the state
determined that fecal contamination of the well(s) was unlikely based on sampling history, land
use, etc. (USEPA, 2000).
Given the importance of correctly targeting systems for source water monitoring, in conjunction
with the State's desire for enough flexibility to ensure sensible decisions on a case-by-case basis,
EPA decided in the final rule to redesign the source water monitoring provision. Accordingly,
the final rule did not include a national requirement for HSAs and routine monitoring for systems
in sensitive aquifers. Rather, EPA concluded that the States are in the best position to assess
which systems would most benefit from a source water monitoring program. The final provision
was similar to routine monitoring but was identified as optional for States. EPA recommended
that States use HSAs as one tool to identify high risk systems for assessment source water
monitoring.
3.7 Total Coliform Rule and Revised Total Coliform Rule
The 1989 TCR established the MCLG of zero for total coliforms (including fecal coliforms and
E. colt). The TCR required monitoring for total coliforms in the distribution system and, if total
coliform bacteria were detected, monitoring for fecal conform/A", coli. The total coliform
monitoring frequency was determined by the population served. Under the TCR, the MCL for
total coliforms was based on the presence of total coliforms in five percent or more of samples
per month or on the presence of fecal coliform or E. coli in any sample. For systems taking fewer
than 40 samples per month, the MCL was based on the presence of coliforms in two or more
samples per month. Also, all systems had to collect repeat samples at sites that were coliform-
positive. In July 2007, EPA established the TCR Distribution System Advisory Committee to
provide advice and make recommendations to the Agency on revisions to the TCR, and on what
information about distribution systems is needed to better understand the public health impact
from the degradation of drinking water quality in distribution systems (USEPA, 2007a). The
RTCR, promulgated February 13, 2013, establishes an MCLG and an MCL for E. coli (no longer
allowing for fecal coliform measurement), a more specific indicator of fecal contamination and
potential harmful pathogens than total coliforms (USEPA, 2013a). The RTCR eliminates the
MCLG and MCL for total coliforms and instead institutes a treatment technique for coliforms
that requires assessment and corrective action. The rationale for this change is that many of the
organisms detected by total coliform methods are not of fecal origin and do not have any direct
public health implication.
Six-Year Review 3 Technical Support Document 3-13
for Microbial Contaminant Regulations
December 2016
-------
3.8 Summary of Microbial Rules
Listed in Exhibit 3.2 is a summary of the NPDWRs for the microbial rules. For each
microorganism, the rule(s) where it is referenced, MCLG, and MCL or TT are provided.
Exhibit 3.2: NPDWRs for Microbial Rules
Microorganism
MCLG
MCL or TT
Rule(s)
Giardia lamblia
Zero
TT
SWTR
Viruses
Zero (SWTR)
TT
SWTR, GWR
Legionella
Zero
TT
SWTR
Total conforms (including fecal
conforms and E. coli)
Zero (under RTCR,
only for E. coli)
TT
TCR, RTCR
Cryptosporidium
Zero
TT
IESWTR, FBRR, LT1, LT2
Heterotrophic bacteria (by the
HPC method)
N/A
TT
SWTR
Turbidity
N/A
TT
SWTR, IESWTR, LT1
Six-Year Review 3 Technical Support Document 3-14
for Microbial Contaminant Regulations
December 2016
-------
4 Health Effects
This chapter summarizes the results from EPA's review of new information related to human
health risks from exposure to microbial contaminants in drinking water. EPA evaluated whether
any new toxicological data, or waterborne endemic infection or infectious disease information,
would justify revision of a maximum contaminant level goal (MCLG) for the microbial
contaminant regulations. MCLGs are health goals set at a level at which no known or anticipated
adverse health effects occur and which allow an adequate margin of safety. EPA reviewed data
from the Waterborne Disease and Outbreak Surveillance System (WBDOSS) collected by the
Centers for Disease Control and Prevention (CDC) (http://www.cdc.gov/healthywater
/survei 11 ance/index.htm 1) and other available data that documented drinking water-associated
outbreaks. EPA also reviewed available literature on endemic disease attributable to drinking
water (e.g., Colford et al., 2009).
The review considered new information, published on or before December 2015, related to
human health risks from exposure to microbial contaminants in drinking water. The review
examined human health risks for systems regulated under the Surface Water Treatment Rules
(SWTRs) and the Ground Water Rule (GWR). Information relevant to the Six-Year Review of
the Long-Term 2 Enhanced Surface Water Treatment Rule (LT2) is provided in a separate
document (USEPA, 2016a).
4.1 SWTRs
EPA promulgated the SWTR in June 1989. It requires all water systems using surface water
sources or ground water under the direct influence of surface water (GWUDI) sources (also
known as Subpart H systems) to remove (via filtration) and/or inactivate (via disinfection)
microbial contaminants (54 FR 27486, USEPA, 1989). Under the SWTR, EPA established
NPDWRs for Giardia, viruses, Legionella, turbidity, and heterotrophic bacteria and set MCLGs
of zero for Giardia lamblia, viruses and Legionella. Under the Interim Enhanced Surface Water
Treatment Rule (IESWTR) (63 FR 69477, USEPA, 1998b) and LT1 (67 FR 1812, USEPA,
2002), EPA established an NPDWR for Cryptosporidium and set an MCLG of zero. The MCLGs
were set at zero since any exposure to these microbial pathogens presents a potential health risk.
Additional information on rule history and the SWTRs is provided in Chapter 3.
4.1.1 MCLGs
The reader is referred to the LT2 support document (USEPA, 2016a) for EPA's assessment of
health effects information related to the following pathogens: Cryptosporidium, Giardia, viruses
and other pathogens (e.g., E. coli). EPA found no new health effects information that would
suggest a need to consider a change from the MCLG of zero for Cryptosporidium, Giardia,
Legionella or viruses, or for a more stringent pathogen log reduction4 target.
4 Log reduction refers to the reduction in pathogen concentration in water through removal or inactivation. For example, a 1-log
reduction indicates the concentration is 10 times smaller (90 percent reduction), a 2-log reduction indicates the concentration is
100 times smaller (99 percent reduction).
Six-Year Review 3 Technical Support Document 4-1
for Microbial Contaminant Regulations
December 2016
-------
New dose-response data from some waterborne pathogens are available from both human and
animal exposure studies (Teunis et al., 2002a; 2002b; Armstrong and Haas, 2007; 2008; Buse et
al., 2012). Concurrently, new models seek to use the new data to provide improved infectivity,
morbidity and mortality predictions (Messner et al., 2014; USEPA, 2016a). The newer models
are specifically designed to address low dose exposure typical of drinking water rather than high
dose exposure typical of food ingestion or vaccine studies. However, because no new human
feeding studies have used low doses, any conclusions are limited despite the low uncertainty
bounds obtained in some statistical models.
4.1.2 Drinking Water-Associated Disease Outbreaks
EPA reviewed published information from the WDOSS about the occurrences and causes of
drinking water-associated outbreaks. This surveillance system is maintained by CDC, EPA, and
the Council of State and Territorial Epidemiologists, and is the primary source of data
concerning waterborne disease outbreaks (WBDOs) in the U.S. (CDC, 2015a). For an event to be
defined as a WBDO by CDC, two criteria must be met: 1) two or more persons diagnosed with
the illness must be linked epidemiologically by time, location of water exposure and case illness
characteristics, and 2) the epidemiological evidence must implicate water as the probable source
of the illness (https://www.cdc.gov/mmwr/preview/mmwrhtml/mm6431a2.htm). WBDOs also
include outbreaks associated with recreational water and other non-potable water sources. In this
document, the disease outbreak data discussed is a subset of WBDOs associated with drinking
water, or "drinking water-associated outbreaks."
Exhibit 4.1 shows the number of drinking water-associated outbreaks in the U.S. from 1971 to
2012 stratified by disease-causing agent (CDC, 2015a). When possible, CDC classifies outbreaks
as being caused by chemical, viral, fungal, parasitic, bacterial (Legionella), bacterial (non-
Legionella) or multiple agents. The reported number of outbreaks peaked during 1979 to 1983
and declined since then; this decrease may be attributed to changes in surveillance or improved
implementation of drinking water regulations, including the Total Coliform Rule (TCR) and the
SWTR beginning in 1991 (Craun et al., 2010). In addition, many water systems have made
voluntary improvements in this time frame, such as through the Partnership for Safe Water
program to reduce the risk of waterborne cryptosporidiosis (National Research Council (NRC),
2006).
CDC noted that the level of surveillance and reporting activity, as well as reporting requirements,
varies across states and localities. The capacity to investigate outbreak events and strengthen
evidence linking outbreaks to drinking water also varies across states and localities. In addition,
CDC noted that detection and investigation of drinking water-associated outbreaks might be
incomplete as it can be difficult to definitively link illnesses with drinking water because most
persons have daily exposure to tap water. For these reasons, outbreak surveillance data likely
underestimate actual values, and should not be used to estimate the total number of outbreaks or
cases of waterborne disease (CDC, 2015a).
Six-Year Review 3 Technical Support Document 4-2
for Microbial Contaminant Regulations
December 2016
-------
Exhibit 4.1: Etiology of Drinking Water-Associated Outbreaks, by Year, in the
United States, 1971 to 2012 (CDC, 2015a)
60
50
40
30
o
2
20
1971 1973 1975 1977 1979 1981
1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011
10 H
m Multiple
m Unidentified
|""1 Chemical
~ Viral
~ Parasitic
H Bacterial, non-Legionella
¦ Bacterial Legionella
Year
4.1.2.1 Deficiencies Assigned to Drinking Water-Associated Outbreaks
CDC publishes data from the WBDOSS in biennial reports (Morbidity and Mortality Weekly
Reports) which include information on water sources, deficiencies, etiology and other
characteristics associated with each drinking water-associated outbreak. CDC assigns one or
more deficiencies to outbreaks associated with drinking water, other water and unknown water
exposures (http://www.cdc.gov/healthvwater/surveillance/deficiencv-classification html). The
deficiencies provide information about how the water became contaminated, water system
characteristics and factors leading to outbreaks.
Exhibit 4.2 summarizes the data on drinking water-associated outbreaks from 2003 through
2012. It presents the total number of drinking water-associated outbreaks, the number of
outbreaks due to deficiencies related to water source, treatment facility or distribution system
(SWTD), and the number of outbreaks due to premise plumbing deficiencies (and of which,
those associated with Legionella) during this time period. Premise plumbing is the portion of the
distribution system that is inside schools, hospitals, public and private housing, and other
buildings (NRC, 2006). Note that the number of outbreaks due to other deficiency categories,
such as unknown or insufficient information, is included in the total number of outbreaks but not
presented in separate columns in Exhibit 4.2. The data presented in Exhibit 4.2 include all
outbreaks associated with drinking water systems, including public, individual, or bottled water
systems.
Six-Year Re\'iew 3 Technical Support Document 4-3
for Microbial Contaminant Regulations
December 2016
-------
Exhibit 4.2: Summary of Drinking Water-Associated Outbreaks and Assigned
Deficiencies - United States, 2003-2012
Years
(Data
Source)
Total
Number of
Drinking
Water-
Associated
Outbreaks
Number of Outbreaks
due to SWTD1 Deficiencies
Number of
Outbreaks
due to
Premise
Plumbing
Deficiency
Number
Details
2003-2004
(CDC, 2006)
30
11
• 1 untreated ground water
• 1 untreated surface water
• 2 untreated ground water and distribution system
• 3 treatment
• 3 distribution system
• 1 treatment and distribution system
12 (8 caused
by Legionella)
2005-2006
(CDC, 2008)
20
8
• 4 untreated ground water
• 2 treatment
• 2 treatment and distribution system
12 (10 caused
by Legionella)
2007-2008
(CDC, 2011)
36
21
• 13 untreated ground water
• 6 treatment
• 1 distribution system
• 1 treatment and distribution system
13 (12 caused
by Legionella)
2009-2010
(CDC, 2013)
33
13
• 8 untreated ground water
• 4 distribution system
• 1 distribution system and untreated ground water
19(19 caused
by Legionella)
2011-2012
(CDC, 2015a)
32
6
• 4 untreated ground water
• 1 untreated ground water and surface water
• 1 distribution system
23 (21 caused
by Legionella)
1 SWTD = water source, treatment facility or distribution system
CDC re-analyzed its outbreak data for untreated ground water for the years 1971-2008 and found
that untreated ground water continued to be a health risk in the U.S. (Wallender et al., 2014). The
most recent CDC Surveillance Summary (CDC, 2015a) indicates that Legionella in premise
plumbing and deficiencies in untreated ground water (most of which are public water system
(PWSs) but also include some private wells) were responsible for the majority of all outbreaks in
2011-2012 (note: EPA does not have authority to regulate private wells). CDC noted that a
reduction in outbreaks of gastrointestinal illness might be achieved when ground water systems
are properly maintained and operated to reduce or inactivate microbial contamination and that
ground water sources are further protected from fecal contamination. The report emphasized that
ground water source protection can also be improved through awareness of and compliance with
regulations such as EPA's GWR and RTCR (CDC, 2015a).
Collectively, the data indicate that, since 1971, drinking water-associated outbreaks may have
been reduced as a result of drinking water regulations. However, opportunities remain to address
disease outbreaks associated with distribution systems and untreated ground water and, at the
same time, to potentially address some of the drinking water-associated outbreaks due to little to
no disinfectant residual in the distribution system (Geldreich, 1992; Bartrand et al., 2014).
Six-Year Review 3 Technical Support Document 4-4
for Microbial Contaminant Regulations
December 2016
-------
4.1.2.2 Disease Occurrence Associated with Legionella in Premise Plumbing
One etiologic agent of particular concern for drinking water-associated outbreaks is Legionella.
Legionella bacteria can proliferate under favorable conditions at locations in the premise
plumbing and in some parts of the distribution system that are further from the central parts of
the system (often in building systems), where water has aged the longest and where there may be
little to no disinfectant residual. Further, the quality of the water delivered to building systems
and households can impact these pathogens' ability for growth and disease transmission.
Legionella spp. colonize biofilm layers, particularly those found inside large, complex plumbing
systems such as those of hospitals, hotels or long-term care facilities. This biofilm protects
Legionella from biocides and allows the bacteria to multiply to concentrations that can facilitate
transmission (CDC, 2008).
The 2003-2004 CDC Surveillance Summary was the first year that CDC reported Legionella
outbreaks with other drinking water-associated outbreaks (CDC, 2006). CDC explained that this
addition was in response to the changing epidemiology of drinking water-associated outbreaks.
In 2003-2004, 8 of 30 (26 percent) drinking water-associated outbreaks were confirmed to be
caused by Legionella.
In 2005-2006, there were 20 drinking water-associated outbreaks, 12 (60 percent) of which were
confirmed to be caused by Legionella (CDC, 2008). The report noted that the majority of
drinking water deficiencies in 2005-2006, such as those associated with biofilm growth in
plumbing systems, were associated with contamination at points outside the jurisdiction of public
water systems and which are not regulated by EPA (CDC, 2008).
CDC (2015a) provided a summary of Legionella observations over the period 2007 to 2012. The
report noted that, although the total number of annual drinking water-associated outbreaks
decreased from 2007 to 2012 (36 in 2007-2008; 33 in 2009-2010; and 32 in 2011-2012, see
Exhibit 4.2), Legionella was responsible for increasing proportions of drinking water-associated
outbreaks during this time frame (33 percent, 60 percent, and 66 percent of outbreaks,
respectively). Also, the report noted that the trend had been driven by the increasing proportion
of outbreaks associated with Legionella within community water systems (60 percent of
Legionella outbreaks in 2007-2008; 76 percent in 2009-2010, and 84 percent in 2011-2012).
In 2011-2012, among 21 Legionella outbreaks in community water systems, 14 (67 percent)
occurred in hospitals or health care facilities. The outbreak data illustrated the increased
likelihood of Legionella outbreaks at health care facilities due to the inherent vulnerability of the
population exposed, many of whom are elderly or immune-compromised. Although Legionella
outbreaks do not represent the largest number of cases from drinking water-associated outbreaks,
the outbreaks do represent the highest percentages of hospitalization and mortality, further
emphasizing the importance of controlling this etiological agent (CDC, 2015a). For example, in
the period 2011-2012, Legionella was responsible for 96 percent of hospitalizations and all
deaths from the total reported drinking water-associated outbreaks.
As discussed previously, Legionella cases reported as part of drinking water-associated
outbreaks are only those cases that meet both criteria for classification laid out by CDC for most
microbial contaminants - i.e., two or more linked cases of illness with epidemiological evidence
Six-Year Review 3 Technical Support Document 4-5
for Microbial Contaminant Regulations
December 2016
-------
by time, location of water exposure, and case illness characteristics and identifying water as the
probable cause. Lack of evidence or thorough investigation means that there may be
underreporting of cases of Legionella related to drinking water-associated outbreaks.
Separately from outbreak surveillance, CDC also produces an annual summary of notifiable
disease cases as reported by hospitals and state public health agencies to CDC. These data are
collected in the National Notifiable Diseases Surveillance System (NNDSS) and are passive data
- i.e., data voluntarily reported to CDC (same as the WBDOSS data) as opposed to active data
collected directly by CDC. However, a study of three years of active surveillance data from the
Active Bacterial Core Surveillance program showed that overall disease rates of legionellosis
were similar using this active data and the passive NNDSS data (CDC, 2015b).
Exhibit 4.3 below shows the total number of cases of legionellosis increasing steadily from 2003
to 2012. Also shown (dotted line) is the number of Legionella cases related to drinking water-
associated outbreaks over the same time period. While the number of cases linked to these
outbreak is much smaller, a similar upward trend is observable.
Exhibit 4.3: Cases of Legionella in the U.S., 2003-2012
450
400
350
300 c/)
CD
C/)
03
250 O
4—
o
200 j§
E
150 z
100
50
0
Year
Sources: CDC, 2006; 2008; 2011; 2013; 2014; 2015a
4.1.2.3 Disease Occurrence Associated with Other Pathogens in Distribution Systems
In 2011-2012, non-Legionella bacteria, parasites and viruses accounted for just 22% of all
WBDOs but 64.3% of all cases (CDC, 2015a). In particular, one outbreak of norovirus was
4,500
4,000
3,500
(/) 3,000
CD
c/)
03
o 2,500
4—
O
8 2,000
1,500
1,000
500
0
Total Number of Legionella Cases
— — — Number of Legionella cases related to
drinking water-associated outbreaks
L
/'
\
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Six-Year Review 3 Technical Support Document 4-6
for Microbial Contaminant Regulations
December 2016
-------
responsible for 119 cases or 28% of all cases over the two-year period. From 2003 to 2012 the
most commonly reported etiological agents besides Legionella were Campylobacter jejuni and
norovirus, together accounting for 20% of all WBDOs and 43% of all cases (CDC, 2006; 2008;
2011; 2013; 2015a).
In August 2013, a 4-year old boy in St. Bernard Parish, Louisiana died of meningoencephalitis.
The causative agent, Naegleria fowleri, was found to be associated with tap water from the
public water distribution system (Cope et al., 2015). Naegleria fowleri was detected in 50
percent of samples collected from the home and 25 percent of samples collected from the water
distribution system. The source water for the St. Bernard Parish water system is the Mississippi
River and treatment includes filtration, primary disinfection with chlorine and secondary
disinfection with chloramines. During sample collection, total chlorine levels throughout the
house were below the detection limit of the test (<0.02 mg/L) and water temperature in the
service line (at the outside hose bib) to the house was 29 degrees C. At 3 of the 4 distribution
system sampling locations where Naegleria fowleri was detected, there was no detectable total
chlorine residual and water temperature was >30 degrees C.
Ercumen et al. (2014) conducted a meta-analysis to investigate the relationship between
distribution system deficiencies and risk of endemic waterborne illness in consumers of tap
water. The research specifically focused on the impact of routine distribution system problems
on endemic gastrointestinal illness in populations drinking tap water versus point-of-use treated
water. The study's findings suggested that tap water consumption is associated with endemic
gastrointestinal illness in systems with malfunctioning distribution systems, including specific
distribution-related deficiencies, such as loss of pipe integrity and inadequate disinfection
residual. The authors acknowledge significant heterogeneity among study settings and water
system characteristics, even within study subgroups (Ercumen et al., 2014).
In addition to epidemic illness, endemic illness (i.e., isolated cases not associated with an
outbreak) accounts for an unknown but probably significant portion of waterborne disease and is
more difficult to recognize (USEPA, 2006c).
Although most heterotrophic bacteria in drinking water are not pathogenic to humans, a few
(Pseudomonas spp., Acinetobacter spp., Moraxella spp.) may be pathogenic to
immunocompromised consumers (Bartram et al., 2003). As part of a risk assessment analysis of
the probability of infection from drinking water, several heterotrophic bacterial species were
identified as major causes of hospital-acquired infections with a high mortality rate (Rusin et al.,
1997). However, Hunter (2003) found no epidemiological evidence that heterotrophic bacteria in
drinking water can cause disease in the general population.
4.1.2.4 Burden of Disease
The precise burden of disease is not well quantified, whether epidemic or endemic. Five
primarily waterborne diseases (giardiasis, cryptosporidiosis, Legionnaires' disease, otitis externa,
and non-tuberculous mycobacterial infection) were responsible for over 40,000 hospitalizations
per year at a cost of nearly $1 billion per year according to a recent estimate (Collier et al., 2012)
using national medical insurance claim data from CDC. Legionella and non-tuberculous
mycobacteria (NTM) together were responsible for 73 percent of the total hospitalizations and 89
Six-Year Review 3 Technical Support Document 4-7
for Microbial Contaminant Regulations
December 2016
-------
percent of the total costs. It is important to note that this is a conservative estimate, as it accounts
only for costs associated with hospitalization and not for lost wages or recovery time. A more
recent analysis focused solely on pulmonary NTM infection costs - a subset of all NTM
infections (Strollo et al., 2015). This analysis updated the case estimate for 2014 and included
antibiotics cost in its estimate, which resulted in a projected 2014 estimate of 181,000 national
annual cases at a cost of $1.7 billion.
4.1.3 GWUDI-Related Public Health Concerns
4.1.3.1 Giardia and Cryptosporidium Illnesses and Outbreaks Associated with Potentially
Misclassified Ground Water Systems
A common cause of a WBDO is a failure in the multiple barrier system designed to protect
public health. There are several summaries of Giardia and Cryptosporidium outbreaks in the
U.S. and worldwide (Baldurssun and Karanis, 2011; Chalmers, 2012; Murphy et al., 2014;
Hrudey and Hrudey, 2014).
Public water systems using ground water in the U.S. also experienced waterborne disease
outbreaks due to Giardia and Cryptosporidium (summarized in Wallender et al., 2014, Solo-
Gabriele and Neumeister, 1996). Ground water outbreaks result when natural filtration is
inadequate and disinfection treatment, if provided, is insufficient to protect public health from
epidemic disease. Endemic disease may occur for these same reasons.
Wallender et al. (2014) summarized CDC outbreak data for the years 1971-2008 and determined
that GWUDI was a "contributing factor" in 18 of 172 (10 percent) outbreaks using untreated
ground water (not including 76 outbreaks with insufficient information). Among the 248 total
untreated ground water outbreaks during this time period, Giardia and/or Cryptosporidium was
the etiological agent(s) for 16 outbreaks (six percent). Three quarters of the outbreaks involved
PWSs. These findings indicate that some of the ground water systems examined by CDC that
were not required to disinfect were contaminated with pathogens.
In reviewing the available information on outbreaks, it appears that two outbreak failure
scenarios can result from either vertical or horizontal passage of a large bolus of pathogenic
protozoa through the subsurface (sufficiently large so as to cause a recognized outbreak):
1) Untreated wells regardless of hydrogeologic setting
In the U.S. and elsewhere (e.g., Switzerland, Fiischslin et al., 2012), Giardia or Cryptosporidium
drinking water outbreaks (as compared with endemic illness due to drinking water) typically do
not occur or are not recognized to occur in disinfected wells located in porous media (sand or
sand and gravel) aquifers. One reason is that subsurface passage through sand and gravel reduces
pathogen counts and disinfection is applied as a second barrier to inactivate some of the
remaining pathogens. The subsurface natural filtration is similar to the removal achieved by slow
sand filters in an engineered filtration system. The combination of subsurface passage in material
containing unconsolidated sand and any additional treatment is typically adequate to reduce
pathogen concentrations so that only generally unrecognized endemic rather than epidemic
disease (an outbreak) may occur. This natural filtration principal is recognized as the basis for a
Six-Year Review 3 Technical Support Document 4-8
for Microbial Contaminant Regulations
December 2016
-------
demonstration of performance (DOP) of alternative filtration treatment in the LT2 Toolbox
Guidance section on Bank Filtration (USEPA, 2010c). A ground water collector that does not
provide adequate natural filtration and/or does not receive adequate disinfection is at risk for
pathogen contamination.
An outbreak that occurred due to the absence of any treatment in wells serving the PWS is
demonstrated by the giardiasis outbreak in Bartlett, New Hampshire in 2007 (Daly et al., 2010).
The PWS in Bartlett was approved by the state to supply water without treatment. The wells
were emplaced into bedrock [probably fractured metamorphic (crystalline) bedrock], which
likely facilitated infiltration of pathogen contamination. As a result of the giardiasis outbreak in
2007, it became apparent that the PWS had been incorrectly classified as ground water rather
than as GWUDI.
2) "Inadequately treated" wells in fractured bedrock or karst limestone settings
"Inadequately treated" wells refer to those that do not achieve sufficient pathogen removal and
inactivation to prevent outbreaks through natural and/or engineered filtration. "Inadequately
treated" wells in fractured bedrock or karst limestone settings are at greater risk for a Giardia or
Cryptosporidium outbreak because one of the multi-barriers (natural filtration) is not present or
not sufficient. Outbreaks associated with disinfected but inadequately naturally filtered wells are
more likely when the hydrogeologic setting is fractured bedrock or karst limestone. In these non-
porous media settings, cysts (or oocysts) can travel long distances with little attenuation. Often in
these cases, it is not recognized that one barrier (natural filtration) is underperforming or absent
due to the aquifer type.
The cryptosporidiosis outbreak in Brushy Creek, Texas in 1998 (Lee et al., 2001; Bergmire-
Sweat et al., 1999) resulted from inadequate treatment of a PWS well. The wells were located
>400 m from surface water in karst limestone and were permitted to receive disinfection with no
filtration. Because the well was located far from surface water, filtration was not required.
Because the outbreak occurred in a region where a karst limestone aquifer was present, the PWS
well was incorrectly classified as ground water rather than as GWUDI.
Although parasitic protozoan outbreaks associated with inadequately treated GWUDI wells in
alluvial (sand and gravel) aquifers have not been recognized in the U.S., they have occurred
elsewhere (e.g. Torbay, UK in 1992 and again in 1995; Hrudey and Hrudey, 2014). Even without
recognized outbreaks, inadequately treated GWUDI wells remain at risk for endemic, rather than
epidemic disease (outbreaks), due to protozoan contamination.
4.1.3.2 Randomized Controlled Intervention Study to Measure Endemic Drinking Water
Disease
EPA regulations promulgated under the SDWA are designed to protect against both endemic and
epidemic disease. Epidemic outbreaks represent the most easily identifiable, but still difficult to
recognize, "tip of the waterborne disease pyramid," meaning that for each case that actually
seeks medical care, many more are not recognized because they are either subclinical or do not
seek medical care (Craun et al., 2006). Endemic disease is much more difficult to measure
because it requires recognizing illness, identifying disease as waterborne (attributable risk) and
Six-Year Review 3 Technical Support Document 4-9
for Microbial Contaminant Regulations
December 2016
-------
measuring the disease incidence when common sources are not identified. Researchers
summarized the worldwide effort to measure endemic disease attributable to drinking water in
developed countries (Murphy et al., 2014; Sinclair et al., 2015).
Endemic disease can be estimated using a randomized, controlled trial to compare two
populations: one population that is supplied drinking water that receives no treatment or some
treatment (e.g., in the U.S., treatment required under the SDWA) and a similar or identical
population that is supplied drinking water that receives additional treatment. Three household
intervention studies were conducted; two (Sydney, Australia and Davenport, Iowa) found no
excess endemic diseases attributable to drinking water (Murphy et al., 2014; Sinclair et al., 2015)
whereas the third (i.e., Sonoma County, California) found a statistically significant attributable
risk to drinking water (Colford et al., 2009).
In the U.S., both household and community intervention randomized controlled trials have been
conducted. Household intervention studies have used a filtration and an additional in-home UV
light treatment device to provide extra treatment. Typically, these studies have a cross-over
design and some have incorporated an inactive "placebo" device so participants are unaware of
their device assignment (blinded trials) (Murphy et al., 2014). The cross-over design assigns each
household to periods with the inactive device (i.e., no additional water treatment) and periods of
additional treatment. Ideally, the household members, the plumbers doing the treatment
installation, and the researchers all are unaware to which period additional treatment is provided.
The advantage of the cross-over design is that the household members serve as their own
controls, which reduces the influence of confounding factors in differing populations. Health
effects, especially acute gastrointestinal illness (AGI), are self-reported using daily diaries.
In the analysis of the data collected during the trial, statistical analyses indicate whether the daily
relationship between AGI and additional treatment is significant. If the relationship is significant,
and accounting for differences within the population or between individuals, then an attributable
risk to water is determined. An attributable risk to water implies a causal relationship between
water and illness. Because of the randomized, controlled design, causality of the association
between water and illness is strengthened; one population is studied over differing time periods
and the only difference affecting the population is the amount of drinking water treatment at
differing times. The following text describes the only available household intervention study
conducted in a community receiving drinking water from a public water system using ground
water (Colford et al., 2009).
In the Colford et al. (2009) study, point-of-use counter top devices were installed in participant
homes. The devices were either sham treatment units (no additional treatment installed) or
additional treatment units (1 |im filters demonstrated to remove virtually all bacteria and
parasites and UV light disinfection previously demonstrated to provide 99.99 percent virus
inactivation).
Colford et al. (2009) selected Sonoma County, CA to measure endemic disease using a
randomized controlled trial (household intervention) for several reasons:
Six-Year Review 3 Technical Support Document 4-10
for Microbial Contaminant Regulations
December 2016
-------
1) A cohort of older adults in the county were already participating in a longitudinal
study of aging and physical performance (the National Institutes of Health (NIH)
funding). The Sonoma County drinking water study was also funded by NIH.
2) A large enough community to meet participant recruitment goals (988 individuals
enrolled).
3) Drinking water meeting all federal, state, and local SDWA requirements.
4) Only one drinking water source serving the community.
5) Sonoma County is a relatively short distance from the base location of the research
team (Berkeley, CA).
Colford et al. (2009) found that study participants (>55 years old) had reduced (either at a
population or individual level) health effects (measured as highly credible gastrointestinal illness,
HCGI) with the additional drinking water treatment. EPA re-assessed the Colford et al. (2009)
analysis and concluded that their statistical analysis was appropriate (see Appendix E for the
EPA statistical modeling assessment). Colford et al. determined that the Sonoma County
population had a measured 12 percent mean reduction in yearly gastrointestinal illness (12
percent attributable risk to drinking water) when receiving drinking water with extra treatment.
One interpretation of this result is that the Sonoma County drinking water is causing 12 percent
of the total yearly gastrointestinal disease burden that results from all exposure, including
daycare, food, recreational water, surfaces, children, and hospital-acquired disease pathways.
The results of Colford et al. (2009) suggest that the drinking water produced by the Sonoma
County Water Authority (SCWA) is making individuals aged 55 or older ill despite meeting all
current local, state and federal drinking water regulations. Based on the available SCWA
information (SCWA, 2013), a possible cause for the ill individuals was that none of the PWS
well water received engineered filtration to achieve adequate Giardia and Cryptosporidium
removal.
SCWA operated five horizontal wells each with a central caisson (large diameter well) and radial
laterals (slotted pipe) to capture large volumes of water. These horizontal well laterals are 50 to
60 feet below the river bed; thus the ground water flow path from the river bed to the lateral is
relatively short. The short subsurface residence time and travel distance would minimize Giardia
and Cryptosporidium removal (USEPA, 2010c) and these horizontal wells are likely GWUDI.
The 1992-1993 GWUDI determination data for the SCWA horizontal collector wells were
reported by California Department of Health Services, letter and report dated September 22,
1993 (California Department of Health Services, 1993). One well (horizontal collector well #5)
was previously determined to be potentially GWUDI and a follow-up study was undertaken by
SCWA for the state. As a result of the second study of paired total coliform river and well #5
samples and other data, the state determined that well #5 was GWUDI but awarded 2.5-log
Giardia reduction (removal) credit based on alternative filtration technology (DOP for
alternative treatment). Price et al. (1999) suggested that well #5 was GWUDI approximately 17
percent of the year. Well #5 was not used for about 2-4 months (in some years January,
Six-Year Review 3 Technical Support Document 4-11
for Microbial Contaminant Regulations
December 2016
-------
February, and March but varied from year to year) during the high river stage (>5,000 cubic feet
per second) higher risk period (SCWA, 2013). During the other 8-10 months per year, the well
was likely operated similarly to the other four horizontal collector wells, providing drinking
water that was disinfected but not subject to engineered filtration.
The GWUDI determination and alternative treatment (bank filtration) DOP for the SCWA PWS
wells were conducted based on total coliform occurrence (California Department of Health
Services, 1993) before EPA issued guidances (USEPA, 1992b; USEPA, 2010c) and were never
subsequently re-classified. EPA recommended the use of microscopic particulate analysis (MPA)
for GWUDI determination (USEPA, 1992b) and the use of aerobic spores for alternative
treatment DOP (USEPA, 2010c) (see Section 5.4 for more details). As discussed in the guidance
(USEPA, 2010c), total aerobic spores are recommended for use to demonstrate removal
/inactivation of Cryptosporidium compared to total coliforms because total coliforms have been
shown to be shorter lived and may be less mobile in the subsurface.
Assuming total coliforms have limited subsurface mobility, fewer total coliforms, as compared
with aerobic spores and/or MP A bioparticles, could arrive at the well. Thus, the use of total
coliforms as an indicator could result in a decision that the wells are more likely to be ground
water rather than GWUDI. For the well that is recognized as GWUDI, the use of total coliforms
as an indicator to determine Giardia removal credit would likely favor a decision to take the well
offline for a shorter high risk period. In contrast, the use of a more mobile and longer lived
Giardia/Cryptosporidium surrogate bioindicator (i.e., aerobic spores) would more likely favor a
decision to take the well offline for a longer high risk period, or even continuously until
additional engineered filtration is provided. If aerobic spores rather than total coliform data were
used to determine the high Giardia (and Cryptosporidium) risk period for the SCWA well #5, the
high risk period might be substantially longer than two to four months and perhaps, the well
could be at high risk much of the year.
The measured 12 percent attributable risk of AGI from drinking water identified at Sonoma
County (Colford et al., 2009) may indicate a vulnerability to the multi-barrier system that is
intended to protect public health. The SCWA system is disinfected using chlorine and the
chlorine residual level is about 0.6 mg/L at the entry point and 0.2 mg/L at the end of the
distribution system (SCWA, 2013). Disinfection is ineffective at killing Cryptosporidium
(USEPA, 2006a). Therefore, Cryptosporidium could potentially be the etiologic agent causing
AGI. Data are lacking to determine if Cryptosporidium was indeed the cause of AGI. More data
are needed to assess co-occurrence of aerobic spores and total coliforms in horizontal collector
wells to determine if these wells should be classified as GWUDI. This information could help
inform re-evaluation of GWUDI determinations to better protect public health.
4.1.3.3 Pathogenic Protozoa Occurrence in Ground Water Used for Public Drinking
Water
Traditionally, GWUDI determination under the SWTR is assumed to identify a contamination
risk from nearby surface water such as a river, stream, lake, reservoir, pond, canal, or other water
body. Thus, the assumption is that the GWUDI threat results from near-horizontal flow from the
water body to the adjacent well. EPA guidance and state programs often include a recommended
minimum set-back distance from surface water beyond which GWUDI would not be expected.
Six-Year Review 3 Technical Support Document 4-12
for Microbial Contaminant Regulations
December 2016
-------
On the other hand, the SWTR definition identifies any well as GWUDI if it is at direct risk from
Giardia, Cryptosporidium, or other coccidian parasitic protozoa independent of the type,
character or location of the source water.
However, in some hydrogeologic scenarios, such as very sensitive and vulnerable hydrogeologic
settings (e.g. karst limestone, glacier flood deposits or fractured bedrock) it is possible that
Giardia or Cryptosporidium can enter the well by vertical passage via infiltrating precipitation in
the absence of any water body. For example, the 1994 Hydro-Nine cryptosporidiosis outbreak in
Walla Walla, WA resulted from vertical passage of treated waste water containing
Cryptosporidium and used for irrigation. It is surmised that the wastewater traveled from the
surface downward through coarse, glacial and alluvial flood deposits and, as is reported in the
CDC investigation, directly along a cracked well casing with an improper seal (McKinley, 1997;
Dworkin et al., 1996).
A vertical water flow path (vertical from the surface versus horizontal from surface water) might
not be identified using the bioindicators addressed by the current MPA guidance. For example, in
the absence of a water body, it is unlikely that diatoms or other algae would be significant
bioindicators because they require a continuously moist environment.
In a karst aquifer in France, 18 ground water samples were taken from the Norville (Haute-
Normandie) public water supply well (5,000 customers, chlorine treatment) and tested for
Cryptosporidium oocysts. Thirteen of the eighteen samples were found to be Cryptosporidium
positive by solid-phase cytometry; the maximum concentration was four oocyst per 100 L
(Khaldi et al., 2011). These data show that Cryptosporidium in karst ground water includes, for
some highly vulnerable systems, Cryptosporidium occurrence resulting from poor
Cryptosporidium removal during infiltration from the surface rather than poor removal during
induced infiltration from nearby surface water. Because the SWTR definition assumes that all
Cryptosporidium in PWS wells is transported from adjacent surface water, it is silent on the issue
of Cryptosporidium transport directly from the surface, as apparently was the case in Norville,
France. Karst aquifers are a vital ground water resource in the U.S. According to the United
States Geological Survey (USGS), about 40 percent of the ground water used for drinking water
comes from karst aquifers (USGS, 2012).
Pitkanen et al. (2015) sampled 20 small (<50 population served) ground water wells once in
spring and once in autumn in Finland for Giardia and a suite of microbial indicators. Fourteen
(of nineteen) wells were undisinfected (no information on one well). They found that 4 (of 20)
wells, all undisinfected, had Giardia detection in the autumn sample. All samples, both spring
and autumn, were negative for total coliforms.
Pathogenic protozoa occurrence results within the last decade from the non-GWUDI PWSs in
the U.S. are not available because these PWSs are not required to sample ground water for
Giardia or Cryptosporidium. EPA conducted a preliminary characterization of the number of the
potentially misclassified GWUDI PWSs based on: (1) waterborne disease outbreak compilations;
(2) the SYR2 ICR and the SYR3 ICR (total coliform detections, see Section 6.4 of this
document); and (3) the Unregulated Contaminant Monitoring Rule (UCMR3) occurrence data
(aerobic spore detections and concentrations, see Section 6.5 of this document).
Six-Year Review 3 Technical Support Document 4-13
for Microbial Contaminant Regulations
December 2016
-------
4.2 GWR
EPA promulgated the GWR in 2006 (USEPA, 2006a) to provide protection against microbial
pathogens in PWSs using ground water sources. Viruses are of particular concern because they
can persist longer and can be more mobile in the subsurface than bacterial pathogens (USEPA,
2006).
The health effects associated with TC and/or other pathogen and indicator occurrence in
undisinfected PWS wells was studied for 14 communities in Wisconsin (Borchardt et al., 2012).
Borchardt et al. (2012) conducted a community intervention study in each of the 14
communities. Each undisinfected water supply was periodically treated using UV; at the same
time, drinking water samples were assayed for a suite of viral pathogens using quantitative
polymerase chain reaction (qPCR) and community members kept daily diaries to self-report
highly credible AGI. The study found that the communities and time periods with the highest
virus measures had correspondingly high AGI incidence.
Among the 14 communities, populations ranged from 1,363 to 8,300. The 14 enrolled
communities were the first to volunteer to participate among communities with populations
>1,000 and with four or fewer wells. Most wells tapped sandstone aquifers at depths from 23 to
169 m. One community is suspected of producing water from a karst aquifer. Borchardt et al.
(2012) found a statistically significant association between enterovirus and norovirus
concentrations measured by qPCR in tap water and AGI health effects in the consuming
population in the 14 communities. Adenovirus concentrations were low and not positively
associated with AGI. The estimated attributable risk to drinking water ranged from 6 percent to
22 percent, depending on the model selected. The risk may have been as high as 63 percent
among children less than 5 years old during the period when norovirus was abundant in drinking
water.
Because the qPCR method measures all viral genetic material (RNA or DNA) from both
infectious and inactive virions, viral occurrence data based on qPCR alone are not definitive.
However, the collection and analysis of health effects data, concurrent with qPCR virus
occurrence data, substantially reduces uncertainty about viral infectivity, at least at sites where
concurrent health data are available. Borchardt et al. (2012) demonstrated a statistically
significant relationship between qPCR viral occurrence and health effects. In their study, data
were collected to analyze the relationship between health effects and UV light treatment (and
qPCR virus occurrence). Additional analysis is needed to characterize the AGI attributable risk
to drinking water.
Six-Year Review 3 Technical Support Document 4-14
for Microbial Contaminant Regulations
December 2016
-------
5 Analytical Methods
This chapter summarizes the analytical methods approved for contaminant monitoring or
treatment technique requirements included under the Surface Water Treatment Rule (SWTR),
Interim Enhanced Surface Water Treatment Rule (IESWTR), Long-Term 1 Enhanced Surface
Water Treatment Rule (LT1), and Ground Water Rule (GWR). There are no methods related to
the Filter Backwash Recycling Rule (FBRR). Methods information related to Giardia and
Cryptosporidium are addressed in the LT2 support document (USEPA, 2016a), methods
information related to distribution system water monitoring required for total coliforms and E.
coli under the Revised Total Coliform Rule (RTCR) are addressed in the associated guidance
documents (USEPA, 2012), and methods information related to disinfectant Maximum Residual
Disinfectant Levels (MRDLs) are addressed in the Disinfectants and Disinfection Byproducts
Rules (D/DBPR) support document (USEPA, 2016f).
The National Environmental Methods Index (NEMI) was used to confirm details related to
methods not developed by EPA. 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 United States Geological Survey (USGS). NEMI can be searched by analytical
method number.5
5.1 Methods for Treatment Technique Requirements Related to Raw and Finished
Water Turbidity (SWTR, IESWTR and LT1)
In carrying out the combined filter effluent requirements and the individual filter effluent
requirements of the SWTR, IESWTR and LT1, systems must use methods for turbidity
measurements that were previously approved by EPA. The promulgation of the IESWTR and
LT1 did not include any changes to the approved methods for turbidity. Exhibit 5.1 summarizes
the analytical methods developed by EPA and others (e.g., Standard Methods (SM), Leek
Mitchell, PhD, Great Lakes Instruments) that are approved for turbidity monitoring as part of the
SWTR (USEPA, 1989) and modified over time (e.g., EPA approved a revised EPA Method
180.1 in August 1993), as well as those methods (i.e., alternate testing methods) that have been
approved via EPA's Expedited Method Approval process since promulgation.6 The alternate
testing methods are listed in the Code of Federal Regulations (CFR), in Appendix A to Subpart C
of 40 CFR Part 141.7
During the promulgation process of the IESWTR and LT1, EPA looked for voluntary consensus
standards with regard to calibration of turbidimeters. During the IESWTR rule development
phase, The American Society for Testing and Materials (now ASTM International) was in the
process of developing such voluntary consensus standards; however, there did not appear to be
any voluntary consensus standards available at the time IESWTR was promulgated nor were any
comments received on the topic during the LT1 promulgation process.
5 https://www.nemi. gov/home/
6 EPA's Expedited Method Approval 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/scitech/drinkingwater/labcert/analvticalmethods expedited.cfm
7 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 Microbial Contaminant Regulations
December 2016
-------
Exhibit 5.1: Turbidity Analytical Methods Approved under the Surface Water
Treatment Rule (§141.74)
Methodology
category
Method1
Method citation
Additional Information
Nephelometric
Method
Comparison of the intensity
of light scattered by the
sample under defined
conditions with the intensity
of light scattered by a
standard reference
suspension under the same
conditions.
Standard Methods for the
Examination of Water and
Wastewater 2130 B.
Standard Methods print editions
approved: 18th, 19th, 20th, 21st, 22nd
Standard Methods online versions
approved: 2130 B-01
Formazin polymer is used as the
primary standard reference
suspension.2
Nephelometric
Method
Comparison of the intensity
of light scattered by the
sample under defined
conditions with the intensity
of light scattered by a
standard reference
suspension.
EPA 180.1
"Methods for the
Determination of Inorganic
Substances in
Environmental Samples",
EPA/600/R-93/100,
August 1993.
A standard suspension (i.e.,
formazin, AMCO-AEPA-1, or Hach
StablCal) is used to calibrate the
instrument.
Laser
Nephelometry
(on-line)
Comparison of the intensity
of light scattered by the
sample under defined
conditions, with the
intensity of light scattered
by a standard reference
suspension.
Mitchell Method M5271,
Revision 1.1.
"Determination of Turbidity
by Laser Nephelometry,"
March 5, 2009.
Primary standard suspensions are
used to calibrate the instrument. A
secondary standard is monitored
periodically for deterioration using
one of the primary standards.
Laser light source: Monochromatic
source operated at a nominal
wavelength of 650 ±30nm. The light
source shall be used as a directly
received reference for the scattered
nephelometric signal.
LED
Nephelometry
(on-line)
Comparison of the intensity
of light scattered by the
sample under defined
conditions, with the
intensity of light scattered
by a standard reference
suspension.
Mitchell Method M5331,
Revision 1.1.
"Determination of Turbidity
by LED Nephelometry,"
March 5, 2009.
Primary standard suspensions are
used to calibrate the instrument. A
secondary standard is monitored
periodically for deterioration using
one of the primary standards.
LED light source: Monochromatic
source operated at a nominal
wavelength of 525 ± 15nm. The light
source shall be used as a directly
received reference for the scattered
nephelometric signal.
LED
Nephelometry
(on-line)
Comparison of the intensity
of light scattered by the
sample under defined
conditions with the intensity
of light scattered by a
standard reference
suspension.
AMI Turbiwell,
"Continuous Measurement
of Turbidity Using a
SWAN AMI Turbiwell
Turbidimeter," August
2009.
The instrumentation is installed to
read turbidity continuously. The light
source shall be a white LED emitting
visible light. The LED, all optical
elements and detectors shall have a
spectral peak response between
400 nm and 600 nm.
LED
Nephelometry
(portable)
Comparison of the intensity
of light scattered by the
sample at 90° to the beam
path, with the intensity of
light scattered by a
standard reference
suspension.
Orion Method AQ4500,
Revision 1.0.
"Determination of Turbidity
by LED Nephelometry,"
May 8, 2009.
A primary standard suspension is
used to calibrate the instrument. A
secondary standard suspension is
used as a daily calibration check
and is monitored periodically for
deterioration using a primary
standard.
Six-Year Review 3 Technical Support Document 5-2
for Microbial Contaminant Regulations
December 2016
-------
Methodology
category
Method1
Method citation
Additional Information
Great Lakes
Instruments
(GLI)
Comparison of the intensity
of light scatters by the
sample under defined
conditions with the intensity
of the light scattered by a
standard reference
suspension.
GLI Method 2, "Turbidity,"
November 2, 1992
A standard suspension of formazin,
prepared under closely defined
conditions, is used to calibrate the
instrument.
Hach
FilterTrak
Comparison of the intensity
of light scattered by the
sample under defined
conditions with the intensity
of light scattered by a
standard reference
suspension.
Hach FilterTrak Method
10133, "Determination of
Turbidity by Laser
Nephelometry," January
2000, Revision 2.0.
Calibration verification standards are
used to check instrument
performance and verify the
instrument is operating correctly.
1 This table includes methods added since the 1989 SWTR. Also includes those approved by the Expedited Method
Approval Process.
2 Formazin polymer is used as a primary turbidity suspension for water because it is more reproducible than other
types of standards previously used for turbidity analysis. Styrene divinyl benzene beads (e.g., AMCO-AEPA-1 or
equivalent) and stabilized formazin (e.g., Hach StablCalTM or equivalent) are acceptable substitutes for formazin.
5.2 Methods for Measuring Disinfection Residuals (SWTR) and Disinfection Profiling
and Benchmarking (IESWTR, LT1)
This section addresses the methods related to monitoring to ensure disinfection CT is met as well
as monitoring for disinfection profiling and benchmarking. It also addressed the approved test
methods for distribution system residuals and metrics to indicate "detectable" residuals for each
method.
5.2.1 Disinfectant Residuals
Disinfectant Residual Entering Distribution System
Methods related to primary disinfectants are also addressed in the D/DBPR support document
(USEPA, 2016f). For the most part, the methods for meeting the CT requirement and the MRDL
requirements in the D/DBPR are the same; however, there are a few differences.
The SWTR contains two methods that are not included in the D/DBPR.
The Ozone Indigo Method is not included in the D/DBPR methods section as systems are
not required to monitor for Ozone according to that rule.
• The Chlorine Dioxide (Amperometric Titration - SM 4500-C102 C) method is included
in SWTR but not in the D/DBPR. Chlorine Dioxide (Amperometric Titration - SM 4500-
C102 E) is included in the analytical methods section for both rules, but SM 4500-C102
C is not because it was outdated and inadequate for compliance sampling.
Six-Year Review 3 Technical Support Document 5-3
for Microbial Contaminant Regulations
December 2016
-------
In addition, ASTM D1253-86(96) is included in the D/DBPR as an allowable method along with
ASTM D1253-03 and ASTM D1253-08. Only ASTM D1253-03 and ASTM D1253-08 are
included in the SWTR analytical method's section. Note there is a new version of the method
that is listed on the ASTM website (ASTM D1253-14) at
http ://www. astm. org / Standards/D 1253 .htm.
All methods listed in this section (see Exhibit 5.2) can be used to ensure disinfection CT
requirements are met. Surface water systems are also required under the SWTR (141.72(a)(3)
and 141.72(b)(2)) to ensure that residual disinfectant concentration in the water entering the
distribution system is not less than 0.2 mg/L for more than 4 hours. Methods listed in this section
can all be used to ensure systems meet this requirement.
In addition, the SWTR contains a statement in 141.74(a)(2) that if approved by the state,
disinfectant residual concentrations for free chlorine and combined chlorine may also be
measured using DPD (N, N Diethyl-1,4 phenylenediamine sulfate) colorimetric test kits. Unlike
the D/DBPR, the SWTR does not contain approved methods for measuring combined chlorine.
Disinfectant Residual in Distribution System
The residual disinfectant concentration must be detectable in the distribution system in at least 95
percent of samples taken each month (40 CFR 141.72(b)(3)(i)). Water in the distribution system
with a measured heterotrophic bacteria concentration less than or equal to 500/mL, measured as
heterotrophic plate count (HPC), can satisfy the requirement for a detectable disinfectant residual
for purposes of determining compliance with this requirement. Methods listed in Exhibit 5.2 can
also be used to measure disinfectant residuals in the distribution system. A discussion of organic
chloramine issues that affect total chloramine measurements is provided in Section 7.2.3. The
HPC methods are provided in Section 5.2.4.
5.2.1.1 Ozone
Because of the unique characteristics of ozone, the procedures for determining C and T for
disinfection with ozone differ from those recommended for systems using other disinfectants.
Ozone is a powerful oxidant that reacts rapidly with organic and inorganic substances present in
the water and undergoes auto-decomposition. Therefore, its residual is much less stable than that
of other disinfectants and dissipates rapidly. The T value can be determined through a tracer
study or an equivalent method as approved by the state with air or oxygen applied during testing,
using the same feed gas rate as used during operation. The C value can be determined for
individual chambers of a contactor based on the residual measured at several points throughout
the contact chamber, or at the exit of the chamber. EPA recommends using the average dissolved
ozone concentration in the water for C.
Six-Year Review 3 Technical Support Document 5-4
for Microbial Contaminant Regulations
December 2016
-------
Exhibit 5.2: Primary Disinfectant Residual Analytical Methods Approved under the Surface Water Treatment Rule
(§141.74)
Analyte
Methodology
category
Method1
Method citation
Additional Information
Chlorine
(total)
Chlorine by
Amperometry
The amperometric method is a special
adaptation ofthe polarographic principle. Free
chlorine is titrated at a pH between 6.5 and 7.5, a
range in which the combined chlorine reacts
slowly.
Standard Methods for the
Examination of Water and
Wastewater 4500-CI D.
Standard Methods print editions
approved: 18th, 19th, 20th, 21st,
22nd
Standard Methods online versions
approved: 4500-CI D-00
Chlorine by
Amperometry
Determination of residual chlorine in water by
direct amperometric titration
ASTM D1253-03
ASTM International http://astm.orq.
Any year containing the cited
version ofthe method may be
used.
Chlorine by
Amperometry
Determination of residual chlorine in water by
direct amperometric titration
ASTM D1253-08
ASTM International httD://astm.ora.
The methods listed are the only
alternative versions that may be
used.
Amperometric
Titration (Low level
measurement
Modifies D by using a more dilute titrant and a
graphical procedure to determine the end point.
Standard Methods for the
Examination of Water and
Wastewater 4500-CI E.
Standard Methods print editions
approved: 18th, 19th, 20th, 21st,
22nd
Standard Methods online versions
approved: 4500-CI E-00
DPD Ferrous
Titri metric
DPD is used as an indicator in the titrimetric
procedure with ferrous ammonium sulfate.
Standard Methods for the
Examination of Water and
Wastewater 4500-CI F.
Standard Methods print editions
approved: 18th, 19th, 20th, 21st,
22nd
Standard Methods online versions
approved: 4500-CI F-00
Chlorine by DPD
Colorimetric
This is a colorimetric version ofthe DPD method
(4500-CI F) and is based on the same principles.
Standard Methods for the
Examination of Water and
Wastewater 4500-CI G.
Standard Methods print editions
approved: 18th, 19th, 20th, 21st,
22nd
Standard Methods online
versions approved: 4500-CI G-00
Chlorine by DPD
Colorimetric
Chlorine in the sample reacts with DPD indicator
to form a pink color that is proportional to the
chlorine concentration.2
"Hach Method 10260—
Determination of Chlorinated
Oxidants (Free and Total) in Water
Using Disposable Planar Reagent-
filled Cuvettes and Mesofluidic
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
5-5
December 2016
-------
Analyte
Methodology
category
Method1
Method citation
Additional Information
Channel Colorimetry," Hach
Company. April 2013.
Chlorine by
lodometric
Electrode
Direct potentiometric measurement of iodine
released on the addition of potassium iodide to
an acidified sample. A platinum-iodide electrode
pair is used in combination with an expanded-
scale pH meter.
Standard Methods for the
Examination of Water and
Wastewater SM 4500-CI I.
Standard Methods print editions
approved: 18th, 19th, 20th, 21st,
22nd
Standard Methods online versions
approved: 4500-CI I-00
Residual Chlorine
in Drinking Water
Using an On-Line
Chlorine Analyzer
On-line chlorine analyzer is used to continuously
monitor the chlorine concentration and is
calibrated using aqueous standards. The on-line
analyzer accuracy is periodically
verified/adjusted based on results from grab
sample analyses
EPA Method 334.0. "Determination
of Residual Chlorine in Drinking
Water Using an On-line Chlorine
Analyzer," August 2009. EPA 815-
B-09-013.
ChloroSense
Amperometric
Sensor3
Electrochemical technique known as
chronoamperometry. A reagent-free method of
analyzing water for chlorine. Portable.
ChloroSense. "Measurement of
Free and Total Chlorine in Drinking
Water by Palintest ChloroSense,"
August 2009.
Chlorine
(free)
Chlorine by
Amperometry
The amperometric method is a special
adaptation ofthe polarographic principle. Free
chlorine is titrated at a pH between 6.5 and 7.5, a
range in which the combined chlorine reacts
slowly.
Standard Methods for the
Examination of Water and
Wastewater 4500-CI D.
Standard Methods print editions
approved: 18th, 19th, 20th, 21st,
22nd
Standard Methods online versions
approved: 4500-CI D-00
Chlorine by
Amperometry
Determination of residual chlorine in water by
direct amperometric titration
ASTM D1253-03
ASTM International httD://astm.ora.
Any year containing the cited
version ofthe method may be
used.
Chlorine by
Amperometry
Determination of residual chlorine in water by
direct amperometric titration
ASTM D1253-08
ASTM International httD://astm.ora.
The methods listed are the only
alternative versions that may be
used.
DPD Ferrous
Titri metric
DPD is used as an indicator in the titrimetric
procedure with ferrous ammonium sulfate.
Standard Methods for the
Examination of Water and
Wastewater 4500-CI F.
Standard Methods print editions
approved: 18th, 19th, 20th, 21st,
22nd
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
5-6
December 2016
-------
Analyte
Methodology
category
Method1
Method citation
Additional Information
Standard Methods online versions
approved: 4500-CI F-00
Chlorine by DPD
Colorimetric
This is a colorimetric version of the DPD method
(4500-CI F) and is based on the same principles.
Standard Methods for the
Examination of Water and
Wastewater 4500-CI G.
Standard Methods print editions
approved: 18th, 19th, 20th, 21st,
22nd
Standard Methods online versions
approved: 4500-CI G-00
Chlorine by DPD
Colorimetric
Chlorine in the sample reacts with DPD indicator
to form a pink color that is proportional to the
chlorine concentration.4
"Hach Method 10260—
Determination of Chlorinated
Oxidants (Free and Total) in Water
Using Disposable Planar Reagent-
filled Cuvettes and Mesofluidic
Channel Colorimetry," Hach
Company. April 2013.
Chlorine by
Syringaldazine
(FACTS)
Measures free chlorine over the range of 0.1 to
10 mg/L. A saturated solution of syringaldazine
in 2-propanol is used.
Standard Methods for the
Examination of Water and
Wastewater 4500-CI H.
Standard Methods print editions
approved: 18th, 19th, 20th, 21st,
22nd
Standard Methods online versions
approved: 4500-CI H-00
Residual Chlorine
in Drinking Water
Using an On-Line
Chlorine Analyzer
On-line chlorine analyzer is used to continuously
monitor the chlorine concentration and is
calibrated using aqueous standards. The on-line
analyzer's accuracy is periodically
verified/adjusted based on results from grab
sample analyses
EPA Method 334.0. "Determination
of Residual Chlorine in Drinking
Water Using an On-line Chlorine
Analyzer," August 2009. EPA 815-
B-09-013.
ChloroSense
Amperometric
Sensor5
Electrochemical technique known as
chronoamperometry. A reagent-free method of
analyzing water for chlorine. Portable.
ChloroSense. "Measurement of
Free and Total Chlorine in Drinking
Water by Palintest ChloroSense,"
August 2009.
Chlorine
Dioxide
Amperometric
Method 1
The amperometric titration of chlorine dioxide is
an extension of the amperometric method for
chlorine. By performing four titrations with
phenylarsine oxide, free chlorine (including
hypochlorite and hypochlorous acid),
chloramines, chlorite, and chlorine dioxide may
be determined separately.
Standard Methods for the
Examination of Water and
Wastewater 4500-CI02 C
Standard Methods print editions
approved: 18th, 19th, 20th, 21st,
22nd
Standard Methods online versions
approved: 4500- CI02 C-00
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
5-7
December 2016
-------
Analyte
Methodology
category
Method1
Method citation
Additional Information
Amperometric
Method II
Similar to 4500-CI02 C, this procedure entails
successive titrations of combinations of chlorine
species. Subsequent calculations determine the
concentration of each species.
Standard Methods for the
Examination of Water and
Wastewater 4500-CI02 E
Standard Methods print editions
approved: 18th, 19th, 20th, 21st,
22nd
Standard Methods online versions
approved: 4500- CI02 E-00
ChlordioX Plus
Amperometric
Sensor6
Based on quantifying chemical reactions by
measuring electrical energy produced or
consumed by the reaction. Portable.
ChlordioX Plus. "Chlorine Dioxide
and Chlorite in Drinking Water by
Amperometry using Disposable
Sensors," November 2013.
DPD Method
This method is an extension of the DPD method
for determining free chlorine and chloramines in
water. Chlorine dioxide appears in the first step
of this procedure but only to the extent of one-
fifth of its total available chlorine content
corresponding to reduction of chlorine dioxide to
chlorite ion.
Standard Methods for the
Examination of Water and
Wastewater 4500-CI02 D
Standard Methods print editions
approved: 18th, 19th, 20th
Chlorine Dioxide
and Chlorite in
Drinking Water by
Visible
Spectrophotometry
Visible spectrophotometer is used to measure
the absorbance of the reagent water blank and
sample absorbance at 633 nm, which is the
absorbance maximum for Lissamine Green B in
the citric acid/glycine buffer. The absorbance
difference between the reagent water blank and
the samples is used to calculate the
concentrations of chlorine dioxide.
EPA Method 327.0, Revision 1.1,
"Determination of Chlorine Dioxide
and Chlorite Ion in Drinking Water
Using Lissamine Green B and
Horseradish Peroxidase with
Detection by Visible
Spectrophotometry," May 2005,
EPA 815-R-05-008.
Ozone
Ozone by Indigo
Colorimetric
Method
In acidic solution, ozone rapidly decolorizes
indigo. The decrease in absorbance is linear with
increasing concentration.
Standard Methods for the
Examination of Water and
Wastewater 4500-03 B
Standard Methods print editions
approved: 18th, 19th, 20th, 21st,
22nd
Standard Methods online versions
approved: 4500-03 B-97
1 This table includes methods added since the 1989 SWTR. Also includes those approved by the Expedited Method Approval Process.
2 www.hach.com/asset-qet. down load. isa?id=24364820994
3 http://www.palintest.com/products/chlorosense/
4 www.hach.com/asset-aet. down load. isa?id=24364820994
5 http://www.palintest.com/products/chlorosense/
6 http://www.palintestusa.com/products/chlordiox-plus/
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
5-8
December 2016
-------
5.2.2 pH
The pH must be monitored because disinfection effectiveness of chlorine is pH-sensitive. When
calculating CT, the pH is sampled at each monitoring point and at the same time as the residual
disinfectant concentration (during peak hourly flow).
No pH analytical method is listed in the SWTR, IESWTR, or LT1. Methods for measuring pH
are available at 141.23(k)(l)(21), and a reference to 141.23(k)(l)(21) is provided in 141.74(a)(1)
of the SWTR. The methods listed in 141.23(k)(l)(21) are EPA Methods 150.1 or 150.2, ASTM
method D1293-95, ASTM method D1293-95, 99, Method 4500-H+B in Standard Methods for
the Examination of Water and Wastewater, 18th (1992), 19th (1995), and 20th (1998) editions and
Standard Methods online version 4500-H+ B-00.
There is a reference in 141.74(a)(1) to Appendix A to Subpart C of Part 141, which provides
alternate testing methods that have been approved by EPA's Expedited Method Approval
process since promulgation. Exhibit 5.3 summarizes the analytical methods approved via EPA's
Expedited Method Approval process for pH.
Exhibit 5.3: pH Analytical Methods Approved via the Expedited Method Approval
Process
Methodology
category
pH in Water by
Potentiometry
Method
Method citation
Additional Information
Determination of the activity of the
hydrogen ions by potentiometric
measurement using a standard
hydrogen electrode and a reference
electrode.
Standard Methods for the
Examination of Water and
Wastewater 4500-H+ B.
Standard Methods print
editions approved: 21st,
22nd
pH in Water
pH meter and associated electrodes
are standardized against two
reference buffer solutions that closely
bracket the anticipated sample pH.
ASTM D 1293-12
ASTM International
http://astm.ora.
The methods listed are
the only alternative
versions that may be
used.
5.2.3 Temperature
All disinfectants, except for UV light, are temperature sensitive. CT values vary with water
temperature and, as a result, water temperature should be measured at each monitoring point and
at the same time as the residual disinfectant concentration when calculating CT. The temperature
should be recorded in degrees Celsius (°C) because the CT tables are based on temperature
measured in °C.
No analytical method for temperature is provided in the SWTR, IESWTR, or LT1. Methods for
measuring temperature are available at 141.23(k)(l)(25), but there does not appear to be a
reference to 141.23(k)(l)(25) in 141.74(a)(1) of the SWTR. The methods listed in
141.23(k)(l)(25) are Method 2550 in Standard Methods for the Examination of Water and
Wastewater, 18th (1992), 19th (1995), and 20th (1998) editions and Standard Methods online
version 2550-00.
Six-Year Review 3 Technical Support Document 5-9
for Microbial Contaminant Regulations
December 2016
-------
There is a reference in 141.74(a)(1) to Appendix A to Subpart C of Part 141, which contains
alternate testing methods that have been approved by EPA's Expedited Method Approval
process since promulgation of the particular rule. Exhibit 5.4 summarizes the analytical methods
approved via EPA's Expedited Method Approval process for temperature.
Exhibit 5.4: Temperature Analytical Method Approved by the Expedited Method
Approval Process
Methodology
category
Thermo metric
Method
Method citation
Additional Information
Measured using any standard
liquid-in-glass or electronic
thermometer with an analog or
digital readout.
Standard Methods for the
Examination of Water and
Wastewater 2550.
Standard Methods print
editions approved: 21st, 22nd
Standard Methods online
versions approved: 2550-10
5.2.4 Heterotrophic Bacteria
Exhibit 5-5 provides the heterotrophic bacteria analytical methods that must be used to satisfy
this SWTR requirement.
Exhibit 5.5: Heterotrophic Bacteria Analytical Methods Approved under the
Surface Water Treatment Rule (§141.74)
Methodology
Category
Culturable method
Method
Method citation
Additional Information
Heterotrophic Plate
Count - Pour Plate
Method
Standard Methods for the
Examination of Water and
Wastewater 9215B
Standard Methods print editions
approved: 18th, 19th, 20th, 21st,
22nd
Standard Methods online versions
approved: 9215B-00, 9215B-04
Enzymatic
detection method
Simplate
IDEXX SimPlate™ HPC Test
Method for Heterotrophs in
Water, November 2000
Note: This table includes methods added since the 1989 SWTR. Also includes those approved by the Expedited
Method Approval Process.
5.3 Methods for Treatment Technique Requirements Related to Filtration Avoidance
(SWTR)
Under the SWTR, systems that are successfully avoiding filtration must monitor their source
water quality conditions. As described in Section 3.1.3.4, a filtration avoidance system must have
a fecal coliform concentration of 20/100 mL or less, or a total coliform concentration of 100/100
mL or less in the source water (40 CFR 141.71(a)(1)). If both fecal and total coliforms are
measured, the system must meet the fecal coliform criterion. Systems are required to use an
enumeration method. The analytical methods approved under 141.852 (RTCR) are
presence/absence methods only and have not been evaluated or approved for enumeration.
Exhibit 5.6 and Exhibit 5.7 list the approved analytical methods for total coliform analytical and
fecal coliform, respectively, under the SWTR.
Six-Year Review 3 Technical Support Document 5-10
for Microbial Contaminant Regulations
December 2016
-------
Exhibit 5.6: Total Coliform Bacteria Analytical Methods Approved under the
Surface Water Treatment Rule (§141.74)
Methodology
Category
Lactose
Fermentation
Method
Method
Method citation
Additional Information
Multiple-tube
fermentation
technique for
members of the
Total Coliform group
Standard Methods for the
Examination of Water and
Wastewater 9221 A,B,C
Standard Methods print editions
approved: 18th, 19th, 20th, 21st, 22nd
Standard Methods online versions
approved: 9221A, B, C-99, 9221A,B,C-
06
Membrane
Filtration
Methods
Membrane Filter
Technique for
Members of the
Coliform Group
Standard Methods for the
Examination of Water and
Wastewater 9222 A,B,C
Standard Methods print editions
approved: 18th, 19th, 20th, 21st
Standard Methods online versions
approved:
9222 A, B, C-97
Ml agar
"New medium for the
simultaneous detection of
total coliform and Escherichia
coli in water", K.P. Brenner et
al, 1993, Appl. Environ.
Microbiol. 59:3534.
Enzymatic
detection
method
Enzyme Substrate
Coliform
Test/Colilert (ONPG-
MUG)
Standard Methods for the
Examination of Water and
Wastewater 9223
Standard Methods print editions
approved: 18th, 19th, 20th, 21st, 22nd
Standard Methods online versions
approved: 9223 B-97, 9223B-04
Note: This table includes methods added since the 1989 SWTR. Also includes those approved by the Expedited
Method Approval Process.
Exhibit 5.7: Fecal Coliform Bacteria Analytical Methods Approved under the
Surface Water Treatment Rule (§141.74)
Methodology
Category
Fecal Coliform
Procedure
(following Lactose
Fermentation
method)
Method
Method citation
Additional Information
Thermotolerant
Coliform Test: EC
Medium
Standard Methods for the
Examination of Water and
Wastewater 9221E
Standard Methods print editions
approved: 18th, 19th, 20th, 21st,
22nd
Standard Methods online versions
approved: 9221E-99, 9221E-06
Fecal Coliform
Procedure (direct
test)
Thermotolerant
Coliform Test: A1
Medium
Standard Methods for the
Examination of Water and
Wastewater 9221E
Standard Methods print editions
approved: 18th, 19th, 20th, 21st,
22nd
Standard Methods online versions
approved: 9221 E-99, 9221 E-06
Membrane Filtration
Method
Thermotolerant
(Fecal) Coliform
Membrane Filter
Procedure
Standard Methods for the
Examination of Water and
Wastewater 9222D
Standard Methods print editions
approved: 18th, 19th, 20th, 21st,
22nd
Standard Methods online versions
approved: 9222 A, B, C-97, 9222D-06
Note: This table includes methods added since the 1989 SWTR. Also includes those approved by the Expedited
Method Approval Process.
Six-Year Review 3 Technical Support Document 5-11
for Microbial Contaminant Regulations
December 2016
-------
5.4 Methods for GWUDI Determination (SWTR, IESWTR and LT1)
There are no EPA-approved methods for ground water under the direct influence of surface
water (GWUDI) determinations included in the SWTR, IESWTR or LT1. Methods for
microscopic particulate analysis (MPA) were published as a guidance (USEPA, 1992a) and may
be used for such determinations, but they are not EPA-approved. There are also no regulatory
requirements for use of total aerobic spore methods for GWUDI determination, although
Standard Method 9218 (American Public Health Association (APHA), 2012) may be used for
their analysis. Methods for MPA and total aerobic spores are discussed in this chapter because
they are commonly used and are recommended in the EPA guidance (USEPA, 2012).
5.4.1 Microscopic Particulate Analysis
The 1991 SWTR Guidance Manual (USEPA, 1991b) includes as Appendix A, the EPA
Consensus Method for Giardia Cyst Analysis that was in use at the time - in the late 1980's to
early 1990's. The analytical method described in the appendix uses source water filtration,
density-gradient sample separation, and sample concentration steps with observation of the
processed sample using Brightfield/Phase contrast microscopy to identify individual Giardia
cysts. When observing samples for the presence of cysts, one is also able to observe other
particulates in the processed water sample. The SWTR Guidance Manual discusses the potential
significance of particulates such as plant debris, diatoms and other algae, insects, and rotifers as
indicators of direct surface water influence. The SWTR Guidance Manual did not attempt to
establish a numerical GWUDI criterion based on particulate analysis. In 1992, EPA published
new MPA guidance (USEPA, 1992b), as discussed in Section 5.4.2.
5.4.2 Aerobic Spores
Standard Methods for the Examination of Water and Wastewater Method 9218, Aerobic
Endospores (APHA, 2012), uses heat treatment to inactivate any vegetative cells followed by
plating the sample onto a non-selective nutrient medium and incubating the plates at 35°C. The
endospores germinate to form bacterial colonies.
1991 GWUDI Guidance for the SWTR
The accompanying general guidance to the SWTR, including guidance on how to determine
whether sources of water are GWUDI, was published by EPA in October 1990 and revised in
March 1991 (USEPA, 1991b).
The SWTR Guidance Manual describes a multiple-step procedure for determining whether a
source should be classified as GWUDI. The steps include:
1) Perform a records review to determine if the source is obviously surface water (e.g., a
pond, lake, or stream).
2) If the source is a well, determine whether it is clearly a ground water source or whether
further analysis is needed. The construction of the well and the hydrogeology of the
aquifer including its porosity, transmissivity, and confining layers are considered. Wells
Six-Year Review 3 Technical Support Document 5-12
for Microbial Contaminant Regulations
December 2016
-------
constructed in deep, protected aquifers which are not subject to contamination from
surface water could be considered ground water.
3) If further analysis of the ground water source is needed, perform a complete review of the
system's files and perform a sanitary survey. The existing records review focuses on
source design and construction, evidence of direct surface water contamination, water
quality analysis, indicators of waterborne disease outbreaks, operational procedures such
as pumping rates, and customer complaints regarding water quality or other related
infectious illness. Existing water quality records could include total and fecal coliform
analysis, particulate analysis, and turbidity.
4) If existing records are limited or indicate a concern, conduct particulate analysis and
other water quality sampling and analysis.
The 1991 SWTR Guidance Manual includes (as Appendix A), the EPA Consensus Method for
Giardia Cyst Analysis that was in use at the time - in the late 1980's to early 1990's. The
analytical method described in the appendix uses source water filtration, density-gradient sample
separation, and sample concentration steps with observation of the processed sample using
Brightfield/Phase contrast microscopy to identify individual Giardia cysts. When observing
samples for the presence of cysts, the microscopist is also able to observe other bioparticles in
the processed water sample. The SWTR Guidance Manual discusses the potential significance of
bioparticles such as plant debris, diatoms and other algae, insects, and rotifers as indicators of
direct surface water influence. The SWTR Guidance Manual did not attempt to establish a
numerical GWUDI criterion based on bioparticle analysis.
The 1991 SWTR GWUDI determination definition and guidance is based on a new observation
about bioparticle occurrence and utility for making a GWUDI decision. At that time, there was
little or no scientific literature on Giardia (and Cryptosporidium) occurrence in ground water and
thus parasitic protozoan co-occurrence with ground water indicators or surrogates was unknown.
Some of the 1991 guidance was later shown to be inappropriate. For example, EPA (2010c) does
not recommend particle counter analysis or turbidity as measures of Cryptosporidium (and
Giardia) subsurface removal efficiency because the particles used as pathogen surrogates are not
known to be of surface water origin. On the other hand, particle counting and turbidity remain
important components for determining surface water treatment plant efficiency.
1992 Consensus Methodfor Determining GWUDI Using Microscopic Particulate Analysis
This method builds on the 1991 SWTR Guidance manual, establishes some consistency for the
MPA methodology, provides a suggested numerical score for interpretation of the findings, and
provides suggested bioparticle identity standards.
The MPA Consensus Method was designed to measure and evaluate the occurrence of a few
bioindicator groups. Interpretation of the method's results is based on three assumptions.
a) First, public water supply wells continually or sporadically induce aquifer recharge from
surface water as a result of well pumping.
Six-Year Review 3 Technical Support Document 5-13
for Microbial Contaminant Regulations
December 2016
-------
b) Second, entrained within the induced surface water recharge are organisms that typically
are found in surface water, such as diatoms and other algae, which rely on photosynthesis
to survive. Also entrained are other organisms (e.g. rotifers, insects) that that are able to
survive in shallow ground water adjacent to surface water (the hyporheic zone) in some
stages of their life cycle and in surface water during other stages.
c) Third, the longer the flow path from surface water to the pumping well, given natural
filtration materials in the aquifer (e.g., sand), the greater the amount of straining and
removal of bioindicator organisms (lower counts in each group). Wells hydraulically
connected to and supplied by recharged surface water and with limited natural filtration
are assumed to have higher bioindicator counts, thus representing some risk for Giardia
and Cryptosporidium also passing to the ground water collector from the surface water
body.
The MPA Consensus Method attempts to equate, quantitatively, the significant occurrence of
bioindicators to a risk score for GWUDI. The bioparticle groups each differ in their contributions
to the overall risk analysis and include Giardia, coccidia (which includes Cryptosporidium),
pigment-containing diatoms and chlorophyll-containing algae, some insects and insect larvae,
certain rotifers, and plant debris.
Thus, the selected bioindicator groups are expected to occur in low numbers in ground water
supplies that are not GWUDI. In the MPA scoring system, the organisms that photosynthesize
and are found in surface water are given higher weighted scores to reflect their greater
significance for indicating surface water influence. Bioindicators that live mostly in ground
water but depend on surface water for some stage of their life cycle, are given lower weighted
scores. These bioindicators (e.g. rotifers) are also expected to occur in low numbers if effectively
removed by natural filtration.
The number of each indicator observed in 100 gallons of water contributes points toward the
sample's total score and relative risk categorization. For example, an observation of from 1 to 10
diatoms in a 100 gallon sample would garner a "rare" diatom occurrence and would contribute 6
points to the total score, whereas >10 diatoms in 100 gallons would contribute 11 points toward
the total. Any occurrence of Giardia or coccidia protozoan contributes at least 20 points.
The relative risk of surface water influence (GWUDI determination) is based on the total score.
A total score of >20 is considered high risk for surface water influence, 10-19 is moderate risk,
and <9 is low risk. The points assigned to each type of particulate and the ranges of points for
each relative risk score were developed by consensus professional judgement of the authors and
their listed scientific advisors.
The selection of MPA bioindicators used and the respective weights that are applied in the
scoring process are designed to give greater significance to the most clearly surface water-related
particles (pigmented diatoms and other green algae). Higher counts of clearly surface water
organisms are given extra weight. Thus, a 'high risk' score will identify the "worst of the worst"
wells: those presenting the most public health risk for Giardia or Cryptosporidium reaching the
well from the surface water body.
Six-Year Review 3 Technical Support Document 5-14
for Microbial Contaminant Regulations
December 2016
-------
Since the guidance publication, various parties and entities have surveyed the States to determine
the GWUDI determination process in as many states as possible (e.g. Chaudhary et al, 2009).
Inspection of the surveys indicates that many states use the MPA guidance numerical criteria and
most use a numerical risk score of 15 or 16 as the defining boundary to determine which PWS
systems should be considered to be GWUDI systems.
MPA Relevance and Limitations
The MPA Consensus Method differs from the microscopic methods for Giardia or
Cryptosporidium. The use of the wound yarn (rather than cartridge) filter for MPA assay is the
primary reason the method is not standard for pathogen detection. Detection of those organisms
using the MPA Consensus Method would be due to chance, while not observing either
pathogenic protozoan in an MPA sample result does not inform public health significance.
At the time of MPA Consensus Method development, the contributors recognized that the
science was incomplete on some of the potential bioindicator particulates. For example, in
discussing the merits of aquatic crustaceans, the method states (USEPA, 1992) "The significance
of these larger organisms in ground waters is unknown at this time. " Further, the authors
acknowledge that limited recovery efficiency data were available at the time of method
development and that GWUDI determinations should not be made solely on the basis of the
results from one or two MPA samples.
The MPA method describes sample collection, analysis, and interpretation and does not consider
the aquifer type or site-specific characteristics. The MPA method encourages the use of other
pertinent information described in the SWTR Guidance Manual (such as hydrogeologic
assessments and water quality monitoring results) for determining GWUDI along with MPA
results.
The empirical use of a bioindicator particulate suite such as MPA for regulatory determination
has apparently performed well for over two decades. One reason that MPA has apparently
performed well is that it was based on hyporheic zone science, then in its infancy. However, very
early in the development of the science, it was becoming apparent that the hyporheic zone
science was diverging from GWUDI determination issues. For example, Stanford and Ward
(1992, 1993) found that that stone fly nymphs (as large as 2.0 cm long) are found in alluvial
aquifer ground water as much as 50 m away from surface water (Tobacco River infiltration
galleries used for drinking water, Eureka, MT). Stone flies are organisms with immature life
stages that live in shallow ground water. More recently, Lin et al. (2012), using microbial
geonomics, identified a rich microbial community in wells about 250 m from the Columbia
River in Hanford, WA.
As hyporheic zone science progressed, it was learned that the alluvial sands and gravels are
diverse ecosystems with rich populations. As a result, MPA risk interpretation became
increasingly dependent on organisms such as diatoms that unequivocally originate in surface
water. MPA guidance recognized this issue and only pigmented diatoms are counted. Green
pigment, indicative of recent photosynthesis suggests more recent residence in surface water as
compared with brown diatoms. Similarly, the EPA LT2 Toolbox Guidance (USEPA, 2006a)
Six-Year Review 3 Technical Support Document 5-15
for Microbial Contaminant Regulations
December 2016
-------
identifies pigmented diatom presence as a red flag indicating the possibility of Cryptosporidium
occurrence. Informal experiments reported by Wilson et al. (1996) suggest that the transition
from pigmented algae to non-pigmented cells occurs at about six months.
Currently, MPA results have value as a relatively quick, cheap and appropriate determinant of
GWUDI if the results indicate a 'high' risk based primarily on bioindicators. However, the MPA
method is not without controversy. For example, Jacangelo et al 2001, performed a small study
on MPA variability as part of their GWUDI method investigation. They found that at low risk
factors, the MPA readings were consistent between analysts in the study laboratory and readers
were also consistent with high-risk slides, although the actual scores and type and quantity of
bioindicators found by the analysts varied. For split samples analyzed by different laboratories,
varied results were obtained including, on occasion, different risk categories ranging from low to
high risk.
Newer Developments in GWUDI Determination Principles and Issues
There is overlap among the objectives of the SWTR GWUDI determination guidance (USEPA,
1991b), the EPA LT2 Toolbox Guidance on assessing alternative treatment by DOP (e.g., log
removal for riverbank filtration systems) (USEPA, 2010c), and current knowledge about
bioindicator removal by subsurface passage. Both the SWTR Guidance and the LT2 Toolbox
Guidance documents recognize that relative public health risk is assessed by bioindicator counts.
Both guidance documents use MPA, and especially diatom counts by MPA, to assess relative
risk for the purpose of GWUDI determination. However, the LT2 Toolbox Guidance primarily
recommends total aerobic bacterial spores (spores) to predict the removal of Cryptosporidium.
Spores were not included in the existing GWUDI determination guidance because their utility
was unrecognized at that time. Experience gained from implementing the LT2 Bank Filtration
guidance at Casper, WY (e.g., Gollnitz et al., 2005) and assessments of GWUDI wells at other
locations (Abbaszadegan et al., 2011) suggests that this new knowledge can be used to improve
the existing SWTR GWUDI definition and accompanying guidance.
2010 EPA Bank Filtration Guidance under LT2
Bank filtration is a surface water pretreatment process that uses the bed or bank of a surface
water body and the adjacent aquifer as a natural filter. Bank filtration systems are defined as
relying on the natural properties of the system to remove microbial contaminants. It is an
additional treatment option under the LT2 for systems using surface water to obtain additional
Cryptosporidium removal credits. The natural filtration processes of bank filtration and the
factors that govern its effectiveness are the same as those occurring in the recharge of ground
water by a surface water source.
EPA summarized the scientific literature on natural filtration principles and issues in the Bank
Filtration section of the LT2 Toolbox Guidance (USEPA, 2010c). The Bank Filtration section
describes the process of natural filtration and provides guidance on a method to determine the
appropriate Cryptosporidium log removal credit to assign for natural filtration based on a DOP.
In the guidance, EPA suggests supplementing the MPA data with data from paired ground water
and surface water samples assayed for two culturable bacterial groups: total aerobic spores and
Six-Year Review 3 Technical Support Document 5-16
for Microbial Contaminant Regulations
December 2016
-------
total coliforms. The bank filtration chapter in the guidance suggests that, because aerobic spore
removal are adequate surrogates for Cryptosporidium, any log reduction of aerobic spore counts
comparing the surface water to water from the subsurface collector can be equated to a similar
log reduction in Cryptosporidium. The guidance also suggests that presence/absence of diatoms
(determined using MP A) and total coliforms in well water presents corroborating information on
whether the total aerobic spore data at any particular site are serving as an adequate
Cryptosporidium surrogate.
In the LT2 Toolbox Guidance, EPA suggests that MPA variability may be unavoidable because
the MPA includes a range of bioindicators, each with differing surface water occurrence and
structural stability during subsurface transport. For example, a single algal chain may break-up
into numerous algal particles during subsurface transport, sampling, or laboratory handling. In
addition, some bioindicators are counted even though they are only part of the original organism,
such as plant debris or crustacean, arthropod and/or insect parts. In the LT2 Toolbox Guidance,
EPA suggests favoring bioindicators that are identifiable as whole particles, such as diatoms,
which, in the MPA protocol are counted only if they are whole (and pigmented green).
The advantage of favoring whole particles is that they can be used as surrogates for subsurface
passage Cryptosporidium removal estimates. It is more difficult to accurately estimate log-
removal by subsurface passage if a single bioparticle breaks into multiple bioparticles during
transport. Because MPA relies on a range of bioindicators of differing bioparticle stability, the
Toolbox does not recommend using MPA numerical results to determine Cryptosporidium
removal credit. In the Toolbox, EPA also emphasizes the value of aerobic spores because they
are environmentally resistant organisms that are unchanged during subsurface transport,
sampling, or laboratory analysis and are only counted if they sporulate, indicating that they are
viable.
As the result of EPA direct implementation of the Wyoming drinking water program, EPA
Region 8 has adopted aerobic spores as the primary tool for wellfield management in Casper,
WY. Wellfield management is a consequence of the agreement giving Casper 2-log
Cryptosporidium removal credit by bank filtration. The purpose of the wellfield management
plan is to use spores to inform wellfield operations such that all decisions lead to greater rather
than lesser subsurface residence time for recharging groundwater. During infiltration basin
relining operations, EPA recognized that high spore counts in wells showed that the basins were
undergoing filtration ripening effects. Thus, EPA required all relined basins to discharge to waste
for at least two weeks and until the spore counts return to background levels.
Abbaszadegan et al. (2011) supported EPA LT2 Toolbox Guidance (2010c) recommendations of
using spores to estimate Cryptosporidium removal efficiency. The authors conclude that for
aerobic spores, their size, shape, surface features, occurrence, and survival in aquifers make them
favorable surrogates for predicting Cryptosporidium removal by subsurface passage in sandy
alluvium. Figure 6.10 in Abbaszadegan et al. (2011) shows aerobic spore and MPA values from
Sioux City and Cedar Rapids, IA plotted against each other. In the plot, the 35-40 samples
appear to show a relationship such that higher spore counts correlate with higher MPA scores.
The authors conclude that a preliminary analysis suggests "that aerobic spore counts in well
Six-Year Review 3 Technical Support Document 5-17
for Microbial Contaminant Regulations
December 2016
-------
water may be a suitable indicative tool for evaluating the risk of a well as part of a GWUDI
assessment."
At the time it was developed, MPA was viewed as a qualitative indicator of the potential for
Giardia occurrence. A higher MPA score represented a greater possibility (in a
phenomenological rather than a statistically significant association) of Giardia presence. With
the development of the LT2 Toolbox Guidance for predicting Cryptosporidium removal using
spores, EPA's focus shifted to methods that predict pathogenic protozoa removal rather than risk
of occurrence. The two goals are complementary. Wells recognized as having high removal of all
large bioparticles, including both pathogenic protozoa and MPA bioindicators, are less likely to
be considered GWUDI wells because those large bioparticles are removed during subsurface
passage.
Total Aerobic Spores as Indicators of Recent Surface Water Recharge/Infiltration
One potential GWUDI bioindicator for identifying surface water influence in locations where
either horizontal or vertical flow paths predominate is total aerobic spores. Total aerobic spores
include the ubiquitous Bacillus subtilis. They originate as common soil bacteria and, because
they are long lived and environmentally resistant, are typically found at low levels in shallow
ground water and at higher levels in all surface water. Typically, total aerobic spores are
continuously washed into surface water but may also pass with infiltrating precipitation or other
waters directly from the ground surface into ground water. Where well water has elevated spore
concentrations as compared with ambient ground water (USEPA, 2010c), it is likely that these
waters are directly affected by horizontal or vertical GWUDI or other recent surface water
infiltration. At the time of SWTR promulgation and MPA guidance publication, total aerobic
spores were not included as a possible bioindicator. Although MPA protocols identify "spores"
in visual counts, these are fungal spores and not bacterial spores.
Over the past 15 years, EPA and others have gained experience using total aerobic spores. In
particular, EPA has SWTR direct implementation authority in Wyoming and has applied
knowledge gained by collecting and analyzing total aerobic spore data in GWUDI wells in
Casper, WY to other locations in the U. S. Field demonstrations have shown that the spores
perform well in demonstrating two-log removal at Casper, WY and Kennewick, WA (USEPA,
2010c). Spores also performed well in demonstrating that exceeding two log removal was not
achievable at Kearney, NE so UV light or other engineered treatment was required (State of
Nebraska, 2013). For example, as discussed above, Abbaszadegan et al. (2011) showed that high
aerobic spore counts correlated with high MPA scores. Spore sampling of PWS wells in Quebec
showed that aerobic spores were found in six of nine wells and 45 of 109 samples (Locas et al.,
2008). The authors concluded that the aerobic spore presence is an indicator of a change in water
quality and warrants further investigation to determine the source of contamination. These data
and experience suggest that adding total aerobic spores to existing GWUDI determination
methods would likely result in fewer wells being misclassified and improved public health
protection.
Although spores have been utilized as a Giardia and Cryptosporidium surrogate, there have been
relatively few laboratory studies. In a review paper, Headd and Bradford (2015) summarized and
Six-Year Review 3 Technical Support Document 5-18
for Microbial Contaminant Regulations
December 2016
-------
compared current knowledge about spores and Cryptosporidium. No similar analysis has been
performed to compare spores and Giardia.
Headd and Bradford (2015) found that aerobic spores measure approximately 1 |im (~ 0.8 to 1.5-
1.8 |iin) in diameter, compared with 4-6 |im for Cryptosporidium oocysts. Both aerobic spores
and Cryptosporidium appear to be long lived in ground water. Aerobic spores and
Cryptosporidium also exhibit similarities in surface properties which govern sand particle
attachment and release during transport through porous media. Both have roughly similar zeta
potentials at ground water pH values, both have similar isoelectric points and both have
glycoproteins on the exterior surface. Because both are long lived in ground water, spores are
expected to provide a conservative estimate of potential oocyst occurrence. Similarities in
transport properties (e.g., zeta potential, isoelectric point and surface composition) suggest that
spores could be a reasonable conservative surrogate for Cryptosporidium passage through the
subsurface (Bradford et al., 2016, accepted for publication in 2015).
Current data suggest that total aerobic spores are an appropriate and suitable surrogate for
Giardia and Cryptosporidium transport in the subsurface. Aerobic spores are long lived in the
subsurface, similar in size to cysts and oocysts, and found in sufficiently high density in ground
water and surface water so as to be easily detectable. Other than aerobic spores, there is no other
bioparticle currently known to be suitable as a Giardia and Cryptosporidium surrogate.
5.5 Methods for Source Water Fecal Indicator Measurement under GWR
Ground water systems that trigger source water monitoring as a result of a total coliform-positive
sample in the distribution system under the GWR must monitor their source water for a fecal
indicator. Depending on which fecal indicator(s) are approved by the state, the system must
monitor their source water for either E. coli, Enterococci, or Coliphage. The analytical methods
for these fecal indicators that are approved compliance with the GWR are provided in Exhibit
5.8.
Six-Year Review 3 Technical Support Document 5-19
for Microbial Contaminant Regulations
December 2016
-------
Exhibit 5.8: Analytical Methods Approved under the Ground Water Rule (§141.402)
Analyte
Methodology
category
Method
Method citation
Additional Information
Escherichia
coli
Enzymatic detection
following lactose
fermentation methods
(Standard Methods
9221B, 9221D)
Escherichia coli Procedure
Using Fluorogenic Substrate
Standard Methods for the Examination of Water
and Wastewater 9221F
Standard Methods print editions
approved: 20th, 22nd
Standard Methods online version
approved: 9221F-06
Membrane filtration
methods
Membrane Filtration with Ml
medium
EPA Method 1604: Total Coliforms and
Escherichia coli in Water by Membrane Filtration
Using a Simultaneous Detection Technique (Ml
Medium): September 2002. EPA 821-R-02-024
m-ColiBlue24 Test
Total Coliforms and E. coli Membrane Filtration
Method with m-ColiBlue24® Broth, Method No.
10029 Revision 2, August 17, 1999
Chromocult
Chromocult® Coliform Agar Presence/Absence
Membrane Filter Test Method for Detection and
Identification of Coliform Bacteria and
Escherichia coli in Finished Waters. November
2000. Version 1.0
Enzymatic detection
following membrane
filtration methods
(Standard Methods
9222B, 9222C)
MF Partition Procedures -
Nutrient Agar with MUG
(NA-MUG)
Standard Methods for the Examination of Water
and Wastewater 9222G
Standard Methods print editions
approved: 20th
Enzymatic detection
methods
Enzyme Substrate Coliform
Test
Colilert
Colilertl 8
Colisure
Standard Methods for the Examination of Water
and Wastewater 9223B
Standard Methods print editions
approved: 20th, 21st, 22nd
Standard Methods online version
approved: 9223B-04
E*Colite Test
Charm E*Colite Presence/Absence Test for
Detection and Identification of Coliform Bacteria
and Escherichia coli in Drinking Water, January
9, 1998
Readycult
Readycult Coliforms 100 Presence Absence Test
for Detection and Identification of Coliform
Bacteria and Escherichia coli in Finished Waters.
January 2007. Version 1.1
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
5-20
December 2016
-------
Analyte
Methodology
category
Method
Method citation
Additional Information
Modified Colitag
Modified Colitag™ Test Method for the
Simultaneous Detection of E. coli and other Total
Coliforms in Water (ATP D05-0035), August 28,
2009
TECTA EC/TC
Presence/Absence Method for Simultaneous
Detection of Total Coliforms and Escherichia coli
(E.coli) in Drinking Water. April 2014.
Enterococci
Multiple-Tube
Fermentation
Fecal
Enterococcus/Streptococcus
Multiple-Tube Technique
Standard Methods 9230B
Standard Methods print editions
approved: 20th, 21st, 22nd
Standard Methods online version
approved: 9230B-04
Membrane Filtration
Technique
Fecal
Enterococcus/Streptococcus
Membrane Filter Techniques
Standard Methods 9230C
Standard Methods print editions
approved: 20'
mEI medium
EPA Method 1600: Enterococci in Water by
Membrane Filtration Using membrane-
Enterococcus Indoxyl-p-D-Glucoside Agar (mEI)
EPA 821-R-02-22 (September 2002)
Enzymatic detection
methods
Enterolert
Evaluation of Enterolert for Enumeration of
Enterococci in Recreation Waters. 1996.
Budnick, G.E., Howard, R.T., and Mayo, D.R.
Appl. Environ. Microbiol. 62:3881.
Coliphage
Two Step Enrichment
Presence-Absence
Procedure
EPA Method 1601: Male-specific (F+) and
Somatic Coliphage in Water by Two-step
Enrichment Procedure; April 2001, EPA821-R-
01-030.
Fast Phage
Fast Phage Test Procedure. Presence/Absence
for Coliphage in Ground Water with Same Day
Positive Prediction. ATP Case No. D09-0007,
Version 009. November 2012.
Single Agar Layer
Procedure
EPA Method 1602: Male-specific (F+) and
Somatic Coliphage in Water by Single Agar
Layer Procedure; April 2001, EPA 821-R-01-029.
Note: This table includes those approved by the Expedited Method Approval Process.
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
5-21
December 2016
-------
5.6 Methods for Measuring Disinfectant Residuals in Ground Water (GWR)
Ground water systems that provide 4-log inactivation, removal, or a state-approved combination
of 4-log virus inactivation and removal, have notified the state that they provide 4-log virus
treatment, and have submitted results to the state that they are providing 4-log treatment must
continue to conduct compliance monitoring. The GWR requires that a system using a chemical
disinfectant to achieve the 4-log inactivation of viruses must use the analytical methods under the
SWTR in 141.74(a)(2). See Section 5.2.1 of this document for the list of methods allowed.
Six-Year Review 3 Technical Support Document 5-22
for Microbial Contaminant Regulations
December 2016
-------
6 Occurrence and Exposure
In the SWTR, EPA established requirements for disinfectant residual to control for opportunistic
pathogens in the distribution system (e.g., Legionella). The disinfectant residual concentration
entering the distribution system may not be less than 0.2 mg/L for more than four hours. A
detectable disinfectant residual or heterotrophic bacteria of 500/mL or less (measured as HPC)
must be maintained throughout the entire distribution system in at least 95 percent of the
measurements made (USEPA, 1989). Additional background about these requirements is
provided in Chapter 3 of this document. Coliform and E. coli occurrence can provide an
indication of conditions supporting bacterial growth or an intrusion event into the distribution
system. Detection of coliform bacteria is commonly associated with low distribution system
disinfectant residuals. To assess the relationship between disinfectant residual and occurrence of
indicators for pathogens in distribution systems, EPA evaluated information about chlorine
residuals and total coliforms and E. coli.
This chapter summarizes the results of EPA's occurrence analyses of regulated microbial
indicators, specifically total coliforms (TC) and E. coli (EC), and disinfectant residuals that are
measured at the same time and location using compliance monitoring data from the Third Six-
Year Review (SYR3) Information Collection Request (ICR) database (referred to as the "SYR3
ICR microbial dataset" in this report (USEPA, 2016c)). This chapter also presents the virus and
aerobic spore data collected under the third Unregulated Contaminant Monitoring Rule 3
(UCMR 3) (USEPA, 2016d). Information in this chapter is organized as follows:
Section 6.1 describes the SYR3 ICR microbial dataset - the primary data source used in
this occurrence analysis.
Section 6.2 describes the national level distribution of disinfectant residuals in
distribution systems.
Section 6.3 presents an analysis of the occurrence of TC and EC as functions of
disinfectant residual types and residual levels.
Section 6.4 presents a summary of the occurrence of TC in PWSs using undisinfected
ground water.
Section 6.5 describes the occurrence of viruses in PWSs using undisinfected ground
water based on the UCMR 3 data.
The appendices to this chapter provide additional supporting information on several topics.
Appendix A provides detailed information on the data quality assurance and quality
control (QA/QC) evaluation of the SYR3 ICR microbial dataset. It also describes the
strengths and limitations of the SYR3 ICR microbial dataset.
Appendix B provides additional disinfectant residual analytical results for surface water
and ground water systems not described in Section 6.2. Specifically, Appendix B
includes a detailed evaluation of the disinfectant residuals relative to system type and
system size, as well as seasonal changes, annual trends and geographic distribution.
Appendix C provides additional analytical results addressing the patterns of occurrence
of TC and EC related to disinfectant residuals not described in Section 6.3. Specifically,
Appendix C includes a detailed evaluation of the occurrence of positive TC and EC
Six-Year Review 3 Technical Support Document 6-1
For Microbial Contaminant Regulations
December 2016
-------
results compared to disinfectant residuals in distribution systems relative to system type
and system size, as well as seasonal changes, annual trends and geographic distribution.
Appendix D describes the process used to identify undisinfected ground water systems in
the SYR3 ICR microbial dataset.
Appendix F describes the national total conform/A", coli detection rates in PWSs using
undisinfected ground water.
6.1 SYR3 ICR Microbial Dataset
This section provides a description of the primary source of data, the SYR3 ICR database, and
describes subsets of the database that were used for the various analyses in this chapter (Sections
6.2 to 6.4). A brief description of the UCMR 3 data is provided in Section 6.5.
The SYR3 ICR database was used for EPA's occurrence analyses of the microbial and
disinfectant residual data. This database contains over 47 million records for disinfection
byproducts (DBP), microbial, chemical and radiological compliance monitoring data from
systems of all sizes. The SYR3 ICR database is the largest and most comprehensive source of
PWS compliance monitoring data to date, with over 13 million records passing QA/QC
procedures for DBPs and microbial contaminants. This database is further described in USEPA
(2016g). Details on the QA/QC steps relevant to the SYR3 ICR microbial dataset are described
in Appendix A of this document and USEPA (2016e).
As part of the SYR3 ICR, EPA requested compliance monitoring data regarding the
presence/absence of total coliforms, E. coli and/or fecal coliforms (see Chapter 3 for additional
information on compliance requirements). In addition, EPA requested data for disinfectant
residual levels in the distribution system, because water systems that disinfect are required to
monitor for the presence of a disinfectant residual when collecting coliform samples in the
distribution system. Systems must collect "routine" total coliform samples on an annual,
quarterly, or monthly basis, depending on their size and type and state requirements. Systems
serving a larger population are required to take more samples than are required for small
systems. When samples test positive for total coliforms, systems must take "repeat" samples at
and near the same location. All samples that test positive for total coliforms must also test for
either fecal coliforms or E. coli.
The SYR3 ICR database contains total coliform, E. coli and fecal coliform data from 2006
through 2011 for 46 states/entities.8 Microbial contaminant data from 34 states/entities passed
QA/QC criteria and are included in the final SYR3 ICR microbial dataset.9 A detailed
description of the QA/QC process is included in Appendix A. An initial evaluation of the dataset
8 In the SYR3 ICR microbial dataset, the term "entities" includes the following: Region 1 Tribes, Region 4 Tribes, Region 5
Tribes, Region 8 Tribes, Region 9 Tribes, American Samoa and Navajo Nation.
9 The State of Maine is included in this count of 34 states though only one record from Maine (from the year 2008) passed QA
and is included in the final SYR3 ICR microbial dataset.
Six-Year Review 3 Technical Support Document 6-2
For Microbial Contaminant Regulations
December 2016
-------
exposed a large degree of variability in the number of records provided by water systems from
state to state, as discussed in Appendix A.
Two major subsets of data from the SYR3 ICR microbial dataset were used for analysis in this
chapter. The first represents the TC/EC results paired with chlorine residual data10 (free chlorine,
total chlorine or both) (note: this is the first dataset available to evaluate the TC/EC data as
function of chlorine residual at a national level). These chlorine residual data were measured in
the field and reported with the TC/EC data. Some states provided a large amount of TC/EC data,
but only a small portion of those data were paired with chlorine residual concentrations (note:
some systems may report HPC in lieu of disinfectant residuals). As a result of the "data pairing"
and related QA/QC process, approximately 70 percent of the original chlorine residual records,
or 4 million records were used for analysis in Section 6.2; and approximately 50 percent of the
original TC records, or 4.8 million records were used for analysis in Section 6.3. Exhibit A-2 in
Appendix A documents the specific counts of records included and excluded in each step of the
QA/QC process for the SYR3 ICR microbial dataset.
A second subset of the SYR3 ICR microbial dataset was used to represent "undisinfected"
ground water systems. This subset was used in the analyses presented in Section 6.4 and
Appendix F. The methodology for identifying this subset is presented in Appendix D.
Exhibit 6.1 provides a conceptual overview of the components of the SYR3 ICR microbial
dataset, including the interrelationships between these two subsets of data. As shown in Exhibit
6.1, there is some overlap between the first and second datasets because: the first dataset includes
all the TC/EC data paired with disinfectant residual data regardless of the residual levels,
whereas in the second dataset, "undisinfected" refers to those ground water systems that either
do not practice disinfection (thus, do not report any disinfectant residual data) or have
disinfectant residuals less than 0.1 mg/L. As such, the TC/EC data paired with disinfectant
residual levels of less than 0.1 mg/L are included in both datasets. The TC/EC data that were not
paired with disinfectant residual data and not identified as undisinfected ground water systems
(i.e., within the large circle but outside of the two inner circles in Exhibit 6.1) were not included
in any of the analyses in this Chapter.
It is important to note that these analyses were conducted to help inform the Six-Year Review
and that they are not meant to assess compliance with regulatory standards.
10 Some states provided only their TC and EC data without the corresponding disinfectant residual concentrations. Some states do
not store their disinfectant residual data in their state's drinking water database alongside directly linked to their coliform results.
Six-Year Review 3 Technical Support Document 6-3
For Microbial Contaminant Regulations
December 2016
-------
Exhibit 6.1: Conceptual Overview of the Components of the SYR3 ICR Microbial
Dataset
SYR3 ICR
Microbial Dataset
TC/EC data for
disinfecting systems with
disinfectant residual
TC/EC data for
undisirtfeeted GW
systems
(Sections 6.2 & 6.3)
(Section 6.4)
'Conceptual; not to scale
Note: "Undisirifected" ground water systems refers to those that do not practice disinfection or have very low
disinfectant residuals (i.e., less than 0.1 mg/L), as described in Appendix D.
6.2 Disinfectant Residuals in Distribution Systems
This section characterizes disinfectant residual concentrations in distribution systems using the
SYR3 ICR microbial dataset. Analyses are presented separately for the two source water types
(surface water and ground water). Additional analyses, including an evaluation of potential
seasonal, annual and geographic trends, are provided in Appendix B.
Exhibit 6.2 presents an inventory of free and total chlorine data associated with total coliform
samples, and the systems providing those data, by source water type, system type and system
size for all years from 2006 to 2011. Results for each year from 2006 through 2011 are presented
in Appendix B.
Six-Year Review 3 Technical Support Document 6-4
For Microbial Contaminant Regulations
December 2016
-------
Exhibit 6.2: Counts of Chlorine Residual Data by Source Water Type, System
Type and System Size from SYR3 ICR Dataset (All Years; 2006-2011)
System
Type
Source
Water
Type
Population
Served Size
Category
Number of Systems1 with
Routine TC Samples
Number of Routine TC
Samples
Free Chlorine Total Chlorine
Free Chlorine Total Chlorine
Community
GW
<100
3,536
1,533
133,419
52,861
Water
101-500
5,268
2,731
219,433
134,599
Systems
501-1,000
1,913
1,319
89,922
76,801
1,001-4,100
2,726
1,988
264,104
181,983
4,101-33,000
1,383
1,089
525,889
310,376
33,001-100,000
120
104
213,811
123,711
>100,000
22
18
47,014
34,292
Total GW
14,968
8,782
1,493,592
914,623
SW
<100
442
234
22,847
16,182
101-500
976
660
45,633
43,295
501-1,000
502
427
25,615
29,979
1,001-4,100
1,172
1,011
125,569
122,295
4,101-33,000
1,148
901
553,971
356,958
33,001-100,000
196
137
395,702
229,562
>100,000
90
72
362,228
294,808
Total SW
4,526
3,442
1,531,565
1,093,079
Transient
GW
<100
6,290
2,500
99,287
37,852
Non-
101-500
2,184
1,029
39,883
13,275
Community
501-1,000
254
107
5,504
1,486
Water
1,001-4,100
92
37
5,442
1,099
Systems
4,101-33,000
2
0
160
0
33,001-100,000
0
0
0
0
>100,000
0
0
0
0
Total GW
8,822
3,673
150,276
53,712
SW
<100
297
133
9,644
2,089
101-500
141
37
5,690
994
501-1,000
30
11
1,025
318
1,001-4,100
17
5
1,165
84
4,101-33,000
7
1
993
21
33,001-100,000
0
0
0
0
>100,000
0
0
0
0
Total SW
492
187
18,517
3,506
Non-
GW
<100
1,721
619
38,527
11,526
Transient
101-500
1,439
532
38,334
10,111
Non-
501-1,000
379
148
11,482
3,140
Community
1,001-4,100
271
129
21,534
8,075
Water
4,101-33,000
19
10
4,505
1,221
Systems
33,001-100,000
0
0
0
0
>100,000
0
0
0
0
Total GW
3,829
1,438
114,382
34,073
SW
<100
97
38
3,644
1,411
101-500
121
65
5,271
1,954
501-1,000
32
14
1,790
858
1,001-4,100
33
23
3,815
1,645
Six-Year Review 3 Technical Support Document 6-5
For Microbial Contaminant Regulations
December 2016
-------
System
Type
Source
Water
Type
Population
Served Size
Category
Number of Systems1 with
Routine TC Samples
Number of Routine TC
Samples
Free Chlorine
Total Chlorine
Free Chlorine Total Chlorine
4,101-33,000
7
2
2,620
248
33,001-100,000
1
0
4,156
0
>100,000
0
1
0
3
Total SW
291
143
21,296
6,119
Total
GW
<100
11,547
4,652
271,233
102,239
101-500
8,891
4,292
297,650
157,985
501-1,000
2,546
1,574
106,908
81,427
1,001-4,100
3,089
2,154
291,080
191,157
4,101-33,000
1,404
1,099
530,554
311,597
33,001-100,000
120
104
213,811
123,711
>100,000
22
18
47,014
34,292
Total GW
27,619
13,893
1,758,250
1,002,408
SW
<100
836
405
36,135
19,682
101-500
1,238
762
56,594
46,243
501-1,000
564
452
28,430
31,155
1,001-4,100
1,222
1,039
130,549
124,024
4,101-33,000
1,162
904
557,584
357,227
33,001-100,000
197
137
399,858
229,562
>100,000
90
73
362,228
294,811
Total SW
5,309
3,772
1,571,378
1,102,704
1 Based on the number of unique PWSIDs, regardless of the number of records for each system.
Throughout this chapter, counts from ground water systems represent data from systems with a
primary source water type of GW (ground water) and GWP (purchased ground water). 11 Counts
from surface water systems represent data from systems with a primary source water type listed
as SW (surface water); SWP (purchased SW); GU (ground water under direct influence of
surface water); and GUP (purchased GU). In addition, counts from non-community water
systems (NCWSs) throughout this chapter represent data from non-transient non-community
water systems and transient non-community water systems. For the purposes of the analyses
presented in this report, "E. coif' and "EC" corresponds to E. coli plus fecal coliform samples
noting that the vast majority of these additional assays were for E. coli.
As shown in Exhibit 6.2, there were a similar number of samples from ground water systems as
from surface water systems. However, more than 80 percent of systems providing the free and/or
total chlorine residual data were ground water systems. Approximately 60 percent of systems
providing free and/or total chlorine residual data were community water systems (CWSs), the
remainder being either transient or non-transient non-community water systems. More small
systems than large systems provided chlorine residual data as there are more small systems than
large systems nationally. The system size category with the largest number of samples had
populations ranging from 4,101 to 33,000. This is a function of the product of the number of
systems in this range and the number of monthly routine samples these systems are required to
11 Abbreviations used in this chapter such as GW, GWP, SW, SWP, and GU were taken from SYR3 ICR database.
Six-Year Review 3 Technical Support Document 6-6
For Microbial Contaminant Regulations
December 2016
-------
take. There are many more small systems nationally than large but the smaller systems are
required to take fewer monthly samples under the TCR. The larger systems are required to take
many more monthly samples than the small systems under the TCR; however, there are fewer
large systems. The total number of samples associated with the mid-sized systems (populations
ranging from 4,101 to 33,000) ends up being the largest due to a large number of systems, as
well as a substantial amount of samples per system.
As mentioned earlier, the SYR3 ICR microbial dataset contains total coliform, E. coli and fecal
coliform data that were paired with (i.e., collected at the same time and location) field free and/or
total chlorine residual data.
Exhibit 6.3 is a diagram characterizing the type of residual reported, i.e., free chlorine only, total
chlorine only, or both free and total chlorine residual. Based on the data counts shown on Exhibit
6.3, approximately 55 percent of samples have free chlorine data only; 28 percent of samples
have total chlorine data only; and 17 percent of samples have both free and total chlorine data
reported. Because total chlorine is the sum of free chlorine and combined chlorine, samples
where free chlorine was higher than total chlorine were not included in the analysis (as noted in
Appendix A).
The SYR3 ICR database does not have a simple data field to identify the disinfectant type of free
chlorine versus chloramines for each system. In general, a water system using free chlorine in the
distribution system (chlorine system) usually reports disinfectant residual concentrations as free
chlorine; whereas a water system using chloramines (chloramine system) in the distribution
system reports total chlorine or both free and total chlorine. Since the SWTR allows water
Six-Year Review 3 Technical Support Document 6-7 December 2016
For Microbial Contaminant Regulations
Exhibit 6.3: Diagram Characterizing Type of Residual Reported
X TC/EC Data for X
Systems with
Disinfectant Residual
' free and
total residual
conc.
I Data with field
I free chlorine residual
\ concentrations only
Data with field
total chlorine residual
concentration only
•770.000
records
¦2.6 million records
1.4 million records
-------
systems using free chlorine to report disinfectant residual concentrations as free, combined, total
chlorine or both free and total chlorine, it is difficult to determine the disinfectant type solely
based on the chlorine residual data. Therefore, EPA conducted data analyses based on the type of
chlorine residual data reported (i.e., free and total chlorine), not the type of disinfectants (i.e.,
chlorine versus chloramines). The type of chlorine residual data reported (i.e., free and total
chlorine) is not necessarily indicative of the type of disinfectants (i.e., chlorine versus
chloramines) used. Given uncertainties in the chlorine residual data reporting described earlier,
those systems that reported only free chlorine data are likely representing chlorine systems; those
systems that reported only total chlorine data are likely representing chloramine systems; and
those reported both free and total chlorine data are likely representing chloramine systems.
However, EPA was unable to confirm the type of disinfectants (i.e., chlorine versus chloramines)
used by the PWSs.
6.2.1 Chlorine Residuals for Surface Water Systems
Exhibit 6.4 presents sample-level summary statistics, by year, for the free and total chlorine
residual data associated with total coliform results in surface water, including: count, 10th
percentile, median, average, 90th percentile and a count of samples greater than 4 mg/L (i.e., the
MRDL under the Stage 1 and 2 Disinfectants and Disinfection Byproducts Rules (D/DBPRs)).
For each parameter, the values are relatively stable from year to year for surface water, except
for a slight increase of the disinfectant residual level over the time of this survey. Additional
analysis of yearly trends in the data is provided later in Appendix B.
Exhibit 6.4: Summary Statistics of Free and Total Chlorine Residual
Concentrations in Surface Water, by Year
Year
Count
Chlorine Residual Concentration (mg/L)
Samples > 4 mg/L
10th
Percentile
Median
Average
90th
Percentile
Count
Percent of
Total
Free Chlorine
2006
199,834
0.20
0.82
0.96
1.83
65
0.03%
2007
233,109
0.23
0.90
1.10
2.20
768
0.33%
2008
238,586
0.24
0.97
1.15
2.40
469
0.20%
2009
247,021
0.19
0.93
1.13
2.40
608
0.25%
2010
315,366
0.20
0.97
1.11
2.20
1,053
0.33%
2011
337,462
0.23
1.00
1.14
2.20
1,012
0.30%
Total Chlorine
2006
116,248
0.60
1.51
1.70
3.00
471
0.41%
2007
124,588
0.50
1.50
1.66
3.00
491
0.39%
2008
135,662
0.50
1.43
1.63
3.00
541
0.40%
2009
177,732
0.60
1.65
1.78
3.20
1,147
0.65%
2010
260,473
0.69
1.68
1.82
3.20
1,880
0.72%
2011
288,001
0.70
1.68
1.81
3.20
2,053
0.71%
Six-Year Review 3 Technical Support Document 6-8
For Microbial Contaminant Regulations
December 2016
-------
Exhibit 6.5 provides a cumulative distribution plot presenting the free and total chlorine residual
concentrations in surface water samples, for the year 2011. The results are presented for the year
2011 only, which is the latest and largest dataset over the sample period.
Exhibit 6.5: Cumulative Percent of Free and Total Chlorine Residual
Concentrations in Surface Water (in 2011)
- Free Chlorine
- - Total Chlorine
0 1 2 3 4 5
Chlorine Residual {mg/L as CI2)
The jagged "staircase" curves are due to the presence of multiple samples with the same chlorine
residual concentration (which is itself due to the limited precision of the analytical methods and
the number of decimal places stored in the database). The cumulative distribution curve shows
that total chlorine concentrations are higher as a group than free chlorine concentrations, as
expected. The percent of samples < 0.2 mg/L is higher for free chlorine (8 percent) than total
chlorine (0.9 percent). This could reflect higher doses of chloramines often used in a PWS and/or
relative persistence of combined chlorine (i.e., the sum of the mono-, di-, and tri-chloramines)
compared to free chlorine (Kirmeyer et al., 2004; USEPA, 2007b).
Exhibit 6.6 presents the frequency of detection for the free and total chlorine residual data
associated with total coliform results in surface water. Results were generated separately for five
bins of free and total chlorine residual concentrations:
Statistic Free Chlorine Total Chlorine
100%-
Samples
337,462
288,001
10th Percentile
0.23 mg/L
0.70 mg/L
Mean Concentration
1.14 mg/L
1.81 mg/L
Median Concentration
1.00 mg/L
1.68 mg/L
90th Percentile
2.2 mg/L
3.2 mg/L
%Samples <0.2 mg/L
8%
0.9%
%Samples < 0.5 mg/L
18%
5%
Q.
E
CO
CD
O
CTi
3
C
Q>
O
<1>
CL
0)
>
•*—»
_ro
£
O
75% -
50%
25%
Six-Year Review 3 Technical Support Document 6-9
For Microbial Contaminant Regulations
December 2016
-------
- Bin 1: concentrations equal to 012;
- Bin 2: concentrations greater than 0 and less than or equal to 0.2 mg/L;
- Bin 3: concentrations greater than 0.2 mg/L and less than or equal to 0.5 mg/L;
- Bin 4: concentrations greater than 0.5 mg/L and less than or equal to 1.0 mg/L;
- Bin 5 concentrations greater than 1.0 mg/L.
The majority of surface water samples have free chlorine and total chlorine residual
concentrations of 0.5 mg/L or greater, with each successively higher bin including a larger
proportion of all samples. More samples fell into the lower bins for free chlorine compared to
total chlorine. There was a higher frequency of samples observed with values at or below 0.2
mg/L among the free chlorine samples than among the total chlorine samples.
Exhibit 6.6: Free and Total Chlorine Residual - Frequency of Detection in Surface
Water (All Years; 2006-2011)
80%
70%
60%
£
O
50%
O
0)
aj
40%
a
M-
o
>
30%
o
£
0)
D
20%
a-
£
u_
10%
0%
I.
i Free CL
i Total CL
>0-0.2 >0.2-0.5 >0.5-1.0 >1.0
Chlorine Residual Bin (mg/L)
12 Many systems reported free and/or total chlorine residual concentrations equal to 0. Those data have been retained
in this analysis, though the "0 mg/L" likely means "not detected" and not necessarily 0 mg/L. These data are
interpreted as "below detection limit" for the analyses presented in this chapter.
Six-Year Review 3 Technical Support Document 6-10
For Microbial Contaminant Regulations
December 2016
-------
6.2.2 Chlorine Residuals for Ground Water Systems
Exhibit 6.7 presents sample-level summary statistics, by year, for the free and total chlorine
residual data associated with total coliform results in ground water. Summary statistics include:
count, 10th percentile, median, average, 90th percentile and a count of samples greater than 4
mg/L. For each statistic, the values are relatively stable from year to year. Additional analysis of
yearly trends in the data is provided later in Appendix B. The chlorine residual concentrations in
ground water, as shown in Exhibit 6.7, are generally lower than those in surface water, as shown
in Exhibit 6.4.
Exhibit 6.7: Summary Statistics of Free and Total Chlorine Residual
Concentrations in Ground Water, by Year
Year
Count
Chlorine Residual Concentration (mg/L)
Samples > 4 mg/L
10th
Percentile
Median
Average
90th
Percentile
Count
Percent of
Total
Free Chlorine
2006
213,056
0.00
0.50
0.62
1.22
124
0.06%
2007
230,669
0.00
0.50
0.60
1.20
93
0.04%
2008
233,075
0.00
0.50
0.61
1.25
130
0.06%
2009
322,909
0.02
0.60
0.75
1.55
202
0.06%
2010
364,748
0.10
0.70
0.80
1.59
209
0.06%
2011
393,793
0.10
0.73
0.83
1.60
253
0.06%
Total Chlorine
2006
105,116
0.30
0.87
1.02
2.00
403
0.38%
2007
118,715
0.30
0.90
1.06
2.10
285
0.24%
2008
133,740
0.30
0.90
1.07
2.13
357
0.27%
2009
171,874
0.33
1.00
1.22
2.40
428
0.25%
2010
227,687
0.40
1.05
1.27
2.50
580
0.25%
2011
245,276
0.40
1.09
1.28
2.50
586
0.24%
Exhibit 6.8 provides a cumulative distribution plot presenting the free and total chlorine residual
concentrations in ground water samples for the year 2011. The jagged "staircase" curves are due
to the presence of multiple samples with the same chlorine residual concentration (which is itself
due to the limited precision of the analytical methods and the number of decimal places stored in
the database). Similar to the plot for surface water, the cumulative distribution curve for ground
water shows that total chlorine concentrations are higher as a group than free chlorine
concentrations.
Six-Year Review 3 Technical Support Document 6-11
For Microbial Contaminant Regulations
December 2016
-------
Exhibit 6.8: Cumulative Percent of Free and Total Chlorine Residual
Concentrations in Ground Water (in 2011)
ioo%-
, J "
V)
75%-
E
cn
CO
o
cl
is
E
O
25%-
/ f
/
/
I
( ¦>
J J
(
J (
/ /
i
/ *
/ *
i
i
t
Statistic
Free Chlorine
Total Chlorine
9
(
Samples
393,793
245,276
J
10th Percentile
0.10 mg/L
0.40 mg/L
1
Mean Concentration
0.33 mg/L
1.28 mg/L
Median Concentration
0.73 mg/L
1.09 mg/L
90th Percentile
1.6 mg/L
2.5 mg/L
%Samples < 0.2 mg/L
13%
4%
%Sarnples < 0.5 mg/L
31%
13%
— Free Chlorine
- - Total Chlorine
2 3
Chlorine Residual {mg/L as CI2)
As shown in Exhibit 6.9, the majority of ground water samples have free chlorine and total
chlorine residual concentrations of 0.2 mg/L or greater. More samples fell into the lower bins for
free chlorine compared to total chlorine. In addition, the proportion of free chlorine residual
samples in ground water decreased from the bin of concentrations greater than 0.5 mg/L to 1.0
mg/L to the bin of concentrations greater than 1.0 mg/L. There was a higher frequency of
samples observed with values at or below 0.5 mg/L among the free chlorine samples than among
the total chlorine samples.
Six-Year Review 3 Technical Support Document 6-12
For Microbial Contaminant Regulations
December 2016
-------
Exhibit 6.9: Free and Total Chlorine Residual - Frequency of Detection in Ground
Water (All Years; 2006-2011)
80%
70%
60%
c
¦2 50%
¦
« 40% ¦ aFreeCL
o _ h ¦ ¦ Total CL
|| |
| 20%
1i. i. h II II
0 >0-0.2 >0.2-0.5 >0.5-1.0 >1.0
Chlorine Residual Bin (mg/L)
6.2.3 Limitations of Data Analysis
The chlorine residual data used for this analysis were collected from 2006 through 2011. These
data do not fully reflect impacts of the implementation of the LT2, GWR and RTCR, which were
promulgated in 2006, 2010 and 2013, respectively.
As indicated previously, the SYR3 ICR microbial dataset consists of data from 34 states/entities.
EPA recognized a large degree of variability in the number of records provided by water systems
from state to state. Only the data from SDWIS states were included in the final SYR3 ICR
microbial dataset because they provided TC/EC data in a usable format that were also paired
with disinfectant residual data (USEPA, 2016e).
6.2.4 Considerations for Potential System-Level Analyses
Although not performed under the Six-Year Review 3, the SYR3 ICR dataset could be used to
evaluate impacts of potential revisions to the distribution system minimum disinfection residual
requirements (regulatory implication forecast analysis). This analysis could be conducted on a
system4evel to estimate the number and percent of public water systems that would exceed
various benchmarks and the corresponding estimations of population served by those systems.
Under the SWTR, the residual disinfectant concentration in the distribution system "cannot be
undetectable in more than five percent of the samples each month, for any two consecutive
months that the system serves water to the public." (40 CFR 141.72). The residual disinfectant
Six-Year Review 3 Technical Support Document 6-13
For Microbial Contaminant Regulations
December 2016
-------
concentration must be measured at least at the same points in the distribution system and at the
same time as total coliforms are sampled (40 CFR 141.74). The monitoring frequency for total
coliforms for community water systems is based on the population served by the system (40 CFR
141.21). For example, a system serving 25 to 1,000 people is required to collect at least one
sample per month; a system serving 17,201 to 21,500 people is required to collect at least 20
samples per month; and a system serving 3,960,001 people or more is required to collect at least
480 samples per month.
System-level analyses could be generated separately for surface water (including GWUDI), and
ground water systems, as well as for CWSs and NCWSs in different system sizes. Similar to
sample-level analyses presented in Section 6.2 of this document, system-level analyses could be
generated using only residual disinfectant records taken from the distribution system. General
considerations for potential analyses are described below:
Create a subset of data for the system-level analysis.
o Use the 2011 dataset - the latest and largest dataset over the sample period,
o Exclude the free chlorine records that are paired with total chlorine. That is, if
there are both free and total at the same time/place, only use the total chlorine
data.
o Exclude data from systems that were in violation of the TCR. Inclusion of data
from these non-compliant systems may bias results high,
o Establish criteria for defining systems to be included in the dataset. For example,
one criterion is that a system must have at least one free or total residual
disinfectant record each month for at least six months.
Establish benchmark values as potential numeric definitions for "detectable" or minimum
disinfectant residual concentrations, e.g., 0.1, 0.2, 0.3, 0.4, and 0.5 mg/L for free chlorine
and total chlorine.
Estimate the number and percent of systems that would exceed various benchmarks and
the corresponding population served by those systems:
o Calculate the percentage of records that are below a benchmark value for each
system in each month. Depending upon data availability, this could yield 12
monthly percentages for each system,
o Determine the number of systems that have monthly percentages (calculated
above) exceeding five percent for any two consecutive months,
o Determine the population served by these systems.
6.3 Occurrence of Total Coliforms and E. coli as Function of Disinfectant Residual Types
and Levels in Distribution Systems
This section analyzes the occurrence of total coliform positive results (TC+) and coli positive
results (EC+) compared to disinfectant residuals in distribution systems. All analyses are at the
sample-level and are presented separately by:
source water type (surface water and ground water);
free and total chlorine; and
Six-Year Review 3 Technical Support Document 6-14
For Microbial Contaminant Regulations
December 2016
-------
five bins of free and total chlorine residual concentrations:
- Bin 1: concentrations equal to 013;
- Bin 2: concentrations greater than 0 and less than or equal to 0.2 mg/L;
- Bin 3: concentrations greater than 0.2 mg/L and less than or equal to 0.5 mg/L;
- Bin 4: concentrations greater than 0.5 mg/L and less than or equal to 1.0 mg/L;
- Bin 5 concentrations greater than 1.0 mg/L.
Appendix C includes an evaluation of seasonal changes, annual trends, geographic distribution
and system size trends.
Exhibit 6.10 presents an inventory of routine and repeat TC and EC records that were paired with
the disinfectant residual data. It is important to note that for the TC/EC data analysis in Section
6.3, EPA applied an additional screen to the dataset: TC+ results are included only if there was a
corresponding EC/FC sample and EC/FC results are only included if they had a corresponding
TC+. (See Appendix A for more details on this QA step 11.) Thus, slightly fewer data points
were used for the TC/EC occurrence analysis in Section 6.3 compared to what is presented in
Exhibit 6.10.
More than 80 percent of systems providing TC data were ground water systems. The number of
CWSs reporting TC records was approximately twice the number of NTNCWSs and TNCWSs
combined; however, the number of routine samples reported by CWSs was an order of
magnitude greater than the number of routine samples reported by either NTNCWSs or
TNCWSs.
Exhibit 6.11 presents a breakdown of routine and repeat TC and EC positive records in the SYR3
ICR microbial dataset, with a count by bin of free and total chlorine residual concentration.
These counts are presented for all available years of data (2006 through 2011). In addition, as
described in Section 6.1, as well as Appendix A, the analyses presented in Section 6.3 are based
on a subset of the entire SYR3 ICR TC, EC and disinfectant residuals data (referred to as the
"SYR3 ICR microbial data" in this report).
13 Many systems reported free and/or total chlorine residual concentrations equal to 0. Those data have been retained
in this analysis, though the "0 mg/L" likely means "not detected" and not necessarily 0 mg/L. These data are
interpreted as "below detection limit" for the analyses presented in this chapter.
Six-Year Review 3 Technical Support Document 6-15
For Microbial Contaminant Regulations
December 2016
-------
Exhibit 6.10: Counts of Total Coliform and E. coli Records by Source Water Type, System Type and System Size
from SYR3 ICR Dataset (All Years; 2006-2011)
Routine Samples1
Repeat Samples
System
Type
Source
Water
Population Served
Size Category
Total Conforms
E. coli2
Total Coliforms
E. coli2
Type
Total #
Total #
# Positive
% Positive
# Positive
%
Total #
Total #
# Positive
% Positive
# Positive
%
Systems
Samples
Samples
Samples
Positive3
Systems
Samples
Samples
Samples Positive3
Community
GW
<100
3,775
167,744
2,816
1.68%
141
0.08%
1,101
8,298
1,336
16.10%
46
0.55%
Water
101-500
5,750
308,994
4,093
1.32%
203
0.07%
1,631
11,841
1,394
11.77%
60
0.51%
Systems
501-1,000
2,155
140,130
1,413
1.01%
53
0.04%
540
3,619
324
8.95%
25
0.69%
1,001-4,100
3,130
375,105
2,483
0.66%
110
0.03%
954
6,436
443
6.88%
10
0.16%
4,101-33,000
1,513
694,922
2,373
0.34%
110
0.02%
661
5,692
233
4.09%
18
0.32%
33,001-100,000
132
276,565
689
0.25%
31
0.01%
93
1,582
76
4.80%
9
0.57%
>100,000
22
69,714
527
0.76%
7
0.01%
21
774
50
6.46%
0
0.00%
Total GW
16,477
2,033,174
14,394
0.71%
655
0.03%
5,001
38,242
3,856
10.08%
168
0.44%
SW
<100
512
35,477
405
1.14%
40
0.11%
149
1,051
105
9.99%
9
0.86%
101-500
1,183
81,659
942
1.15%
70
0.09%
376
2,877
211
7.33%
12
0.42%
501-1,000
645
51,537
559
1.08%
33
0.06%
207
1,418
139
9.80%
8
0.56%
1,001-4,100
1,490
222,287
1,411
0.63%
75
0.03%
518
3,395
170
5.01%
12
0.35%
4,101-33,000
1,344
797,365
2,833
0.36%
134
0.02%
664
7,098
293
4.13%
4
0.06%
33,001-100,000
211
545,541
1,054
0.19%
46
0.01%
153
2,595
81
3.12%
6
0.23%
>100,000
103
538,645
1,721
0.32%
67
0.01%
82
4,020
157
3.91%
3
0.07%
Total SW
5,488
2,272,511
8,925
0.39%
465
0.02%
2,149
22,454
1,156
5.15%
54
0.24%
Transient
GW
<100
6,930
123,147
3,412
2.77%
183
0.15%
1,368
9,859
3,433
34.82%
148
1.50%
Non-
101-500
2,383
46,653
1,264
2.71%
72
0.15%
504
3,061
664
21.69%
61
1.99%
Community
501-1,000
277
6,330
86
1.36%
5
0.08%
47
239
53
22.18%
0
0.00%
Water
1,001-4,100
97
5,711
36
0.63%
3
0.05%
22
126
10
7.94%
0
0.00%
Systems
4,101-33,000
2
160
6
3.75%
0
0.00%
1
9
0
0.00%
0
0.00%
33,001-100,000
0
0
0
0.00%
0
0.00%
0
0
0
0.00%
0
0.00%
>100,000
0
0
0
0.00%
0
0.00%
0
0
0
0.00%
0
0.00%
Total GW
9,689
182,001
4,804
2.64%
263
0.14%
1,942
13,294
4,160
31.29%
209
1.57%
SW
<100
385
11,581
167
1.44%
18
0.16%
74
594
60
10.10%
15
2.53%
101-500
148
6,366
81
1.27%
6
0.09%
42
313
53
16.93%
4
1.28%
501-1,000
30
1,288
19
1.48%
1
0.08%
8
94
18
19.15%
0
0.00%
1,001-4,100
17
1,226
9
0.73%
1
0.08%
3
12
0
0.00%
0
0.00%
4,101-33,000
7
1,014
5
0.49%
0
0.00%
3
17
0
0.00%
0
0.00%
33,001-100,000
0
0
0
0.00%
0
0.00%
0
0
0
0.00%
0
0.00%
>100,000
0
0
0
0.00%
0
0.00%
0
0
0
0.00%
0
0.00%
Total SW
587
21,475
281
1.31%
26
0.12%
130
1,030
131
12.72%
19
1.84%
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
6-16
December 2016
-------
Routine Samples1
Repeat Samples
System
Source
Population Served
Total Conforms
E. coli2
Total Conforms
E. coli2
Type
Water
Size Category
Type
Total #
Total #
# Positive
0/
# Positive
%
Total #
Total #
# Positive
0/
# Positive
%
Systems
Samples
Samples
70 rOSITIVe
Samples
Positive3
Systems
Samples
Samples
70 rOSITIVe
Samples
Positive3
Non-
GW
<100
1,823
46,945
800
1.70%
47
0.10%
353
2,280
615
26.97%
26
1.14%
Transient
101-500
1,542
45,261
587
1.30%
26
0.06%
288
1,849
362
19.58%
18
0.97%
Non-
501-1,000
404
13,097
125
0.95%
8
0.06%
59
371
72
19.41%
1
0.27%
Community
1,001-4,100
288
25,371
222
0.88%
14
0.06%
81
570
92
16.14%
4
0.70%
Water
4,101-33,000
20
4,876
22
0.45%
1
0.02%
8
58
3
5.17%
0
0.00%
Systems
33,001-100,000
0
0
0
0.00%
0
0.00%
0
0
0
0.00%
0
0.00%
>100,000
0
0
0
0.00%
0
0.00%
0
0
0
0.00%
0
0.00%
Total GW
4,077
135,550
1,756
1.30%
96
0.07%
789
5,128
1,144
22.31%
49
0.96%
SW
<100
104
4,822
43
0.89%
2
0.04%
19
111
8
7.21%
0
0.00%
101-500
132
6,623
44
0.66%
3
0.05%
34
126
6
4.76%
0
0.00%
501-1,000
33
2,294
16
0.70%
2
0.09%
7
50
13
26.00%
1
2.00%
1,001-4,100
37
4,438
23
0.52%
1
0.02%
12
73
5
6.85%
0
0.00%
4,101-33,000
7
2,868
1
0.03%
0
0.00%
1
3
0
0.00%
0
0.00%
33,001-100,000
1
4,156
1
0.02%
0
0.00%
1
3
0
0.00%
0
0.00%
>100,000
1
3
3
100.00%
0
0.00%
1
3
0
0.00%
0
0.00%
Total SW
315
25,204
131
0.52%
8
0.03%
75
369
32
8.67%
1
0.27%
Total
GW
<100
12,528
337,836
7,028
6.15%
371
0.11%
2,822
20,437
5,384
77.89%
220
1.08%
101-500
9,675
400,908
5,944
5.33%
301
0.08%
2,423
16,751
2,420
53.04%
139
0.83%
501-1,000
2,836
159,557
1,624
3.32%
66
0.04%
646
4,229
449
50.54%
26
0.61%
1,001-4,100
3,515
406,187
2,741
2.17%
127
0.03%
1,057
7,132
545
30.96%
14
0.20%
4,101-33,000
1,535
699,958
2,401
4.54%
111
0.02%
670
5,759
236
9.27%
18
0.31%
33,001-100,000
132
276,565
689
0.25%
31
0.01%
93
1,582
76
4.80%
9
0.57%
>100,000
22
69,714
527
0.76%
7
0.01%
21
774
50
6.46%
0
0.00%
Total GW
30,243
2,350,725
20,954
0.89%
1,014
0.04%
7,732
56,664
9,160
16.17%
426
0.75%
SW
<100
1,001
51,880
615
3.48%
60
0.12%
242
1,756
173
27.30%
24
1.37%
101-500
1,463
94,648
1,067
3.09%
79
0.08%
452
3,316
270
29.03%
16
0.48%
501-1,000
708
55,119
594
3.26%
36
0.07%
222
1,562
170
54.95%
9
0.58%
1,001-4,100
1,544
227,951
1,443
1.89%
77
0.03%
533
3,480
175
11.86%
12
0.34%
4,101-33,000
1,358
801,247
2,839
0.88%
134
0.02%
668
7,118
293
4.13%
4
0.06%
33,001-100,000
212
549,697
1,055
0.22%
46
0.01%
154
2,598
81
3.12%
6
0.23%
>100,000
104
538,648
1,724
100.32%
67
0.01%
83
4,023
157
3.91%
3
0.07%
Total SW
6,390
2,319,190
9,337
0.40%
499
0.02%
2,354
23,853
1,319
5.53%
74
0.31%
1 A subset of these records was used in the Section 6.3 analyses (EPA removed TC+ Results that did not have a corresponding EC sample and vice versa).
2 For the analyses presented in this report, "£. coif' and "EC" corresponds to E. coli plus fecal coliform samples.
3 The "% Positive" for EC samples was calculated as the number of EC+ samples divided by the total number of TC sample
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
6-17
December 2016
-------
EPA found that there was a lower rate of occurrence of both TC and EC positives as the free or
total chlorine residual increased to higher levels. For routine samples with free chlorine, the
highest percent of samples that were TC+ or EC+ (2.3 percent and 0.11 percent, respectively)
occurred when free chlorine was equal to 0 mg/L ("not detected"). These percentages dropped by
more than half for the >0 - 0.2 mg/L bin, then appeared to flatten when free chlorine was > 0.2
mg/L. The TC+ rate was less than one percent when chlorine residuals were greater than or equal
to 0.2 mg/L of free chlorine. The trend is similar for total chlorine routine samples except that for
TC, the percent of positive samples was slightly higher for the >0 - 0.2 mg/L bin than for the 0
mg/L bin. Also, percent positive TC and EC results for the >0.2 mg/L - 0.5 mg/L bin were
slightly higher than for the >0.5 mg/L - 1.0 mg/L bin and the >1.0 bin, indicating a possible
tailing off of the TC+ and EC+ occurrence at 0.5 mg/L for total chlorine compared to tailing at
0.2 mg/L free chlorine. This relationship between chlorine residuals and occurrence of TC and
EC positives was similar to results reported by the Colorado Department of Public Health and
Environment (Ingels, 2015). In addition, this relationship is consistent with the findings of
LeChevallier et al. (1996) which stated that disinfectant residuals of 0.2 mg/L or more of free
chlorine, or 0.5 mg/L or more of total chlorine, are associated with reduced levels of coliform
bacteria.
As one might expect, the percentage of positive TC samples was much higher overall for repeat
samples (13.9 percent for free chlorine and 6.9 percent for total chlorine, on average) than for
routine samples (0.6 percent for free chlorine and 0.5 percent for total chlorine, on average).
More than 40 percent of repeat TC samples were positive when free chlorine was zero, compared
to a slightly lower repeat TC+ occurrence of approximately 29 percent when the total chlorine
was zero. Similar to routine samples, repeat TC+ occurrence declined as free and total chlorine
residual increased, with a flattening of occurrence at 0.5 mg/L for both free and total chlorine
residuals.
The highest percent of EC+ in repeat samples occurred when free chlorine was zero (2.0 percent)
and when total chlorine was >0 - 0.2 mg/L (1.01 percent). Unlike routine sample results, the
percent positive of EC repeat samples increased slightly from the >0.5 - 1.0 mg/L bin to the >
1.0 mg/L bin for both free and total chlorine.
Six-Year Review 3 Technical Support Document 6-18
for Microbial Contaminant Regulations
December 2016
-------
Exhibit 6.11: Summary of Total Coliform and E. coli Samples for Each Bin of Free and Total Chlorine Residual
Concentrations from SYR3 ICR Dataset (2006-2011)
Surface and Ground Water
Systems
Routine Samples
Repeat Samples
Group1
Disinfectant
Residual Level
(mg/L)
Total Coliforms
E. coli2
Total Coliforms
E. coll2
Total #
Samples
#
Positive
Samples
% Positive
# Positive
Samples
% Positive3
Total #
Samples
# Positive
Samples
%
Positive
# Positive
Samples
% Positive3
Free
Chlorine
0
194,354
4,463
2.3%
221
0.11%
13,6
77
5,663
41.4%
278
2.03%
CM
o
o
A
319,378
3,293
1.0%
169
0.05%
9,101
1,396
15.3%
80
0.88%
>0.2 - 0.5
602,059
3,677
0.6%
196
0.03%
11,501
753
6.5%
30
0.26%
>0.5-1.0
1,103,795
4,252
0.4%
197
0.02%
14,676
615
4.2%
16
0.11%
>1.0
1,109,384
5,057
0.5%
229
0.02%
16,010
626
3.9%
27
0.17%
Subtotal
3,328,970
20,742
0.6%
1,012
0.03%
64,965
9,053
13.9%
431
0.66%
Total
Chlorine
0
26,903
571
2.1%
48
0.18%
1,248
359
28.8%
6
0.48%
CM
o
o
A
59,370
1,339
2.3%
73
0.12%
1,292
220
17.0%
13
1.01%
>0.2 - 0.5
198,128
1,868
0.9%
77
0.04%
2,679
194
7.2%
8
0.30%
>0.5-1.0
566,203
2,636
0.5%
138
0.02%
5,365
225
4.2%
5
0.09%
>1.0
1,254,425
4,770
0.4%
235
0.02%
12,259
584
4.8%
25
0.20%
Subtotal
2,105,029
11,184
0.5%
571
0.03%
22,843
1,582
6.9%
57
0.25%
1 There is some overlap between these two groups (i.e., some TC and EC records were paired with both free and total chlorine residual concentrations).
2 As described in Section 6.1, for the purposes of the analyses presented in this report, "£. coli" and "EC" corresponds to E. coli plus fecal coliform samples.
3 The "% Positive" for EC samples was calculated as the number of EC+ samples divided by the total number of TC samples.
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
6-19
December 2016
-------
6.3.1 Occurrence in Surface Water
Exhibit 6.12 and Exhibit 6.13 present the frequency of detection of total coliform and E. coli,
respectively, over six years of data in surface water. These analyses are based on routine samples
taken in the distribution system. The exhibits show a larger proportion of TC+ and EC+ in
surface water associated with total chlorine residual data than free chlorine residual data. For
both TC and EC, there was a higher level of occurrence in the smaller chlorine residual bins.
However, the trend is relatively flat for the "free chlorine residual" group because of the bias
introduced by some records that reported zero or very low free chlorine but high total chlorine
values (e.g. in a chloramine system) (see Section 6.3.3). After those records (reported both free
and total chlorine data) are excluded, the "free chlorine only" group showed a higher level of
TC+ or EC+ rate in the smaller chlorine residual bins (e.g., 1.5 percent TC+ for the "0 mg/L" bin
and 0.9 percent TC+ for the ">0-0.2 mg/L" bin), as expected.
Exhibit 6.12: Total Coliforms - Frequency of Detection in Surface Water as
Function of Disinfectant Types and Concentrations (2006-2011)
3.0%
2.5%
2.0%
1.5%
O
^ 1.0%
0.5%
0.0%
10 mg/L
i> 0 to 0.2 mg/L
i> 0.2 to 0.5 mg/L
> 0.5 to 1.0 mg/L
i
¦ >1.0 mg/L
I
|
n
i
in
i
hi
Free Chlorine Only Free Chlorine Residual Total Chlorine Residual
Note: Routine samples only.
Six-Year Review 3 Technical Support Document 6-20
for Microbial Contaminant Regulations
December 2016
-------
Exhibit 6.13: E. coli - Frequency of Detection in Surface Water (2006-2011)
0.45%
0.40%
0.35%
0.30%
0.25%
(/)
QJ
| 0.20%
O 0.15%
LU
^ 0.10%
0.05%
0.00%
I..
10 mg/L
i> 0 to 0.2 mg/L
i> 0.2 to 0.5 mg/L
> 0.5 to 1.0 mg/L
i> 1.0 mg/L
I
Free Chlorine Only
Free Chlorine Residual
Residual
Total Chlorine Residual
Note: Routine samples only.
Exhibit 6.14: Number of Total Coliform and E. coli Samples and Positives in
Surface Water Paired with Free and Total Chlorine Data, by Source Water Type
Group
Disinfectant
Residual
Level (mg/L)
Total Coliforms in Surface Water
E. coli in Surface Water
Total #
Samples
# Positive
Samples
% Positive
# Positive
Samples
% Positive
Free
Chlorine
Only
0
11,464
177
1.54%
15
0.13%
V
0
1
o
k>
51,560
475
0.92%
23
0.04%
>0.2-0.5
161,096
760
0.47%
41
0.03%
>0.5-1.0
432,981
1,319
0.30%
72
0.02%
>1.0
559,272
2,003
0.36%
94
0.02%
Subtotal
1,216,373
4,734
0.39%
245
0.02%
Free
Chlorine
0
46,173
201
0.44%
16
0.03%
V
0
1
o
k>
113,869
668
0.59%
29
0.03%
>0.2-0.5
191,822
883
0.46%
46
0.02%
>0.5-1.0
528,813
1476
0.28%
82
0.02%
>1.0
690,577
2185
0.32%
107
0.02%
Subtotal
1,571,254
5,413
0.34%
280
0.02%
Six-Year Review 3 Technical Support Document 6-21
for Microbial Contaminant Regulations
December 2016
-------
Group
Disinfectant
Residual
Level (mg/L)
Total Coliforms in Surface Water
E. coli in Surface Water
Total #
Samples
# Positive
Samples
% Positive
# Positive
Samples
% Positive
Total
Chlorine
0
1,719
33
1.92%
7
0.41%
>0 - 0.2
20,531
426
2.07%
20
0.10%
>0.2-0.5
63,282
548
0.87%
18
0.03%
>0.5-1.0
226,637
918
0.41%
46
0.02%
>1.0
790,495
2,525
0.32%
113
0.01%
Subtotal
1,102,664
4,450
0.40%
204
0.02%
Note: Routine samples only. This exhibit presents underlying data/denominator for Exhibit 6-12 and Exhibit 6-13.
6.3.2 Occurrence in Ground Water
Exhibit 6.15 and Exhibit 6.16 present the frequency of detection of total coliform and E. coli,
respectively, over six years of data in ground water. These analyses are based on routine samples
taken in the distribution system. The "free chlorine only" data is not presented in these exhibits
because the "free chlorine residual" data did not show the data bias as observed in surface water
results (as discussed in Section 6.3.1). This would be expected because ground water systems are
less likely to use chloramine, except when ammonia is present in source water. Compared to the
surface water results shown in Exhibit 6.12 and Exhibit 6.13, the ground water exhibits show a
similar proportion of TC+ and EC+ in ground water associated with free chlorine residual data
compared to total chlorine residual data. Similar to the surface water results, for both TC and EC
occurrence in ground water, there was a higher level of occurrence in the smaller chlorine
residual bins.
Six-Year Review 3 Technical Support Document 6-22
for Microbial Contaminant Regulations
December 2016
-------
Exhibit 6.15: Total Coliforms - Frequency of Detection in Ground Water (2006-
2011)
q no/.
i 0 mg/L
9 RPL
> 0 to 0.2 mg/L
/o
> 0.2 to 0.5 mg/L
9 C\°L
¦
i > 0.5 to 1.0 mg/L
/o
in
1.0 mg/L
XJ 1 .O /O
¦5
'
O
1 n%
m
I . u /o
H
£
- -
|_
\J.\J /O
n no/.
I III
III
Free Chlorine Residual
Total Chlorine Residual
Residual
Note: Routine samples only.
Exhibit 6.16: E. coli - Frequency of Detection in Ground Water (2006-2011)
¦ 0 mg/L
¦ > 0 to 0.2 mg/L
> 0.2 to 0.5 mg/L
¦ > 0.5 to 1.0 mg/L
¦ > 1.0 mg/L
L... II...
Free Chlorine Residual Total Chlorine Residual
Residual
Note: Routine samples only.
0.45%
0.40%
0.35%
0.30%
„ 0.25%
at
>
0.20%
o
Q.
O 0.15%
LLI
s?
0.10%
0.05%
0.00%
Six-Year Review 3 Technical Support Document 6-23
for Microbial Contaminant Regulations
December 2016
-------
Exhibit 6.17: Number of Total Coliform Samples in Ground Water Paired with Free
and Total Chlorine Data, by Source Water Type
Group
Disinfectant
Residual
Level (mg/L)
Total Coliforms in Ground Water
E. coli in Ground Water
Total #
Samples
# Positive
Samples
% Positive
# Positive
Samples
% Positive
Free
Chlorine
0
148,181
4,262
2.88%
205
0.14%
V
0
1
o
k>
205,509
2,625
1.28%
140
0.07%
>0.2-0.5
410,237
2,794
0.68%
150
0.04%
>0.5-1.0
574,982
2,776
0.48%
115
0.02%
>1.0
418,807
2,872
0.69%
122
0.03%
Subtotal
1,757,716
15,329
0.87%
732
0.04%
Total
Chlorine
0
25,184
538
2.14%
41
0.16%
V
0
1
o
k>
38,839
913
2.35%
53
0.14%
>0.2-0.5
134,846
1,320
0.98%
59
0.04%
>0.5-1.0
339,566
1,718
0.51%
92
0.03%
>1.0
463,930
2,245
0.48%
122
0.03%
Subtotal
1,002,365
6,734
0.67%
367
0.04%
Note: Routine samples only. This exhibit presents underlying data/denominator for Exhibit 6-12 and Exhibit 6-13.
6.3.3 Limitations of Data Analysis
The limitations of the chlorine residual data analysis, as discussed in Section 6.2.4, also apply to
the TC/EC analyses. Inclusion of records that reported both free and total chlorine data (which
accounts for about 17 percent of total records) in the TC/EC analysis could potential create bias
to the results, as described below.
In some cases, zero or very low free chlorine and high total chlorine values are reported for the
same sample. This would be expected in chloramine systems where all the chlorine should be
combined with ammonia and reported as total. Since the primary goal of the analyses is to
evaluate TC/EC occurrence together with the occurrence of low residual values, it is possible
that including the results with both free and total chlorine residual concentrations might be
biasing TC/EC occurrence downward in the lower concentration bins for free chlorine when total
chlorine is also present (i.e., in a chloramines system). Exhibit 6.18 demonstrates the potential
impacts of this bias, showing lower TC occurrence in the free chlorine residual bins of 0 and >0
- 2 mg/L when free chlorine and total chlorine are reported together (0.5 percent for both bins)
than when free chlorine is reported alone (3.4 percent and 1.3 percent, respectively).
There are other samples where free chlorine is a significant proportion of total chlorine. It is
possible that a free chlorine system would have samples with more total than free chlorine when
they have ammonia in their source water and they do not practice breakpoint chlorination. These
Six-Year Review 3 Technical Support Document 6-24
for Microbial Contaminant Regulations
December 2016
-------
systems may experience reduced disinfection effectiveness compared to the situation where none
of the free chlorine were combined with ammonia. These samples (i.e. those with free chlorine as
a high portion of total chlorine) may also misrepresent TC/EC occurrence in the total chlorine
bins since the majority of the residual disinfectant is in the form of free chlorine. Exhibit 6.19
shows potential impacts of the bias, showing lower TC/EC occurrence in all bins for samples
with both free and total chlorine residual compared to samples where only total chlorine residual
is reported.
While Exhibit 6.18 and Exhibit 6.19 show the overall bias on all the records (i.e., surface water
and ground water results combined), the bias on the surface water results is discussed in Section
6.3.1.
Six-Year Review 3 Technical Support Document 6-25
for Microbial Contaminant Regulations
December 2016
-------
Exhibit 6.18: Comparison of Free Chlorine Only Samples with Free Chlorine Samples Paired with Total Chlorine
Free
Chlorine
Free Chlorine Only
Free Chlorine Samples Paired with
Total Chlorine
All Free Chlorine Samples
Residual
(mg/L)
# TC
# TC+
% TC+
#
EC
%
EC+
# TC
# TC+
% TC+
# EC
% EC+
# TC
# TC+
% TC+
#
EC
% EC+
Routine Samples
0
120,414
4,124
3.4%
196
0.16%
73,940
339
0.5%
25
0.03%
194,354
4,463
2.3%
221
0.11%
V
0
1
o
k>
213,808
2,810
1.3%
148
0.07%
105,570
483
0.5%
21
0.02%
319,378
3,293
1.0%
169
0.05%
0.2-0.5
493,937
3,196
0.6%
159
0.03%
108,122
481
0.4%
37
0.03%
602,059
3,677
0.6%
196
0.03%
0.5-1.0
853,445
3,667
0.4%
157
0.02%
250,350
585
0.2%
40
0.02%
1,103,795
4,252
0.4%
197
0.02%
>1.0
882,575
4,603
0.5%
192
0.02%
226,809
454
0.2%
37
0.02%
1,109,384
5,057
0.5%
229
0.02%
Sum
2,564,179
18,400
0.7%
852
0.03%
764,791
2,342
0.3%
160
0.02%
3,328,970
20,742
0.6%
1,012
0.03%
Repeat Samples
0
12,670
5,435
42.9%
270
2.13%
1,007
228
22.6%
8 0.79%
13,677
5,663
41.4%
278
2.03%
V
0
1
o
k>
7,741
1,267
16.4%
70
0.90%
1,360
129
9.5%
10
0.74%
9,101
1,396
15.3%
80
0.88%
0.2-0.5
10,107
685
6.8%
25
0.25%
1,394
68
4.9%
5
0.36%
11,501
753
6.5%
30
0.26%
0.5-1.0
12,530
545
4.3%
15
0.12%
2,146
70
3.3%
1
0.05%
14,676
615
4.2%
16
0.11%
>1.0
14,233
572
4.0%
23
0.16%
1,777
54
3.0%
4
0.23%
16,010
626
3.9%
27
0.17%
Sum
57,281
8,504
14.8%
403
0.70%
7,684
549
7.1%
28
0.36%
64,965
9,053
13.9%
431
0.66%
Exhibit 6.19: Comparison of Total Chlorine Only Samples with Total Chlorine Samples Paired with Free Chlorine
Total
Chlorine
Total Chlorine Only
Total Chlorine Samples Paired with
Free Chlorine
All Total Chlorine Samples
Residual
(mg/L)
# TC
# TC+
% TC+
#
EC
%
EC+
# TC
# TC+
% TC+
# EC
% EC+
# TC
# TC+
% TC+
#
EC
% EC+
Routine Samples
0
6,907
299
4.3%
33
0.48%
19,996
272
1.4%
15
0.08%
26,903
571
2.1%
48
0.18%
V
0
1
o
k>
47,373
1,170
2.5%
59
0.12%
11,997
169
1.4%
14
0.12%
59,370
1,339
2.3%
73
0.12%
0.2-0.5
144,827
1,587
1.1%
63
0.04%
53,301
281
0.5%
14
0.03%
198,128
1,868
0.9%
77
0.04%
0.5-1.0
344,818
2,031
0.6%
98
0.03%
221,385
605
0.3%
40
0.02%
566,203
2,636
0.5%
138
0.02%
>1.0
796,313
3,755
0.5%
158
0.02%
458,112
1,015
0.2%
77
0.02%
1,254,425
4,770
0.4%
235
0.02%
Sum
1,340,238
8,842
0.7%
411
0.03%
764,791
2,342
0.3%
160
0.02%
2,105,029
11,184
0.5%
571
0.03%
Repeat Samples
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
6-26
December 2016
-------
Total
Chlorine
Total Chlorine Only
Total Chlorine Samples Paired with
Free Chlorine
All Total Chlorine Samples
Residual
(mg/L)
# TC
# TC+
% TC+
#
EC
%
EC+
# TC
# TC+
% TC+
# EC
% EC+
# TC
# TC+
% TC+
#
EC
% EC+
0
509
144
28.3%
1
0.20%
739
215
29.1%
5
0.68%
1,248
359
28.8%
6
0.48%
V
0
1
o
k>
903
150
16.6%
8 0.89%
389
70
18.0%
5
1.29%
1,292
220
17.0%
13
1.01%
0.2-0.5
1,970
158
8.0%
2
0.10%
709
36
5.1%
6
0.85%
2,679
194
7.2%
8 0.30%
0.5-1.0
3,346
150
4.5%
5
0.15%
2,019
75
3.7%
0
0.00%
5,365
225
4.2%
5
0.09%
>1.0
8,431
431
5.1%
13
0.15%
3,828
153
4.0%
12
0.31%
12,259
584
4.8%
25
0.20%
Sum
15,159
1,033
6.8%
29
0.19%
7,684
549
7.1%
28
0.36%
22,843
1,582
6.9%
57
0.25%
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
6-27
December 2016
-------
6.4 Occurrence of Total Coliforms in PWSs Using Undisinfected Ground Water
As part of the Six-Year Review 3, EPA conducted an analysis of total conforms/A", coli data
(TC/EC) from the SYR3 microbial dataset that represents undisinfected ground water systems.
EPA analyzed data collected in 2011 for approximately 38,000 small (serving fewer than 4,101
people) undisinfected PWSs. EPA used statistical modeling to characterize distributions of TC
detection rates for each of nine groupings of PWSs based on system type (community, non-
transient non-community and transient non-community) and population served (less than 101,
101 to 1000 and 1001 to 4,100 people).
Among the three PWS types, on average, undisinfected transient PWSs have a 4.3 percent TC
detection rate as compared with 3% for undisinfected non-transient PWSs and 2.5 percent for
undisinfected community PWSs. Within each type of PWS, the smaller systems have higher
median TC detection than the larger systems. All TC-positive samples were assayed for EC.
Among TC-positive samples from small undisinfected PWSs, EC is detected in about five
percent of samples, regardless of PWS type or size. EPA evaluated the upper tail of the TC
detection rate distributions and found that significant percentages of some system types have
high TC detection rates. For example, assuming the PWSs providing data are nationally
representative, then five percent of the -52,000 small undisinfected transient PWSs in the U.S.
have TC detection rates of 20 percent or more. More details about the analysis are provided in
Appendix F.
6.5 Occurrence of Viruses and Aerobic Spores in PWSs Using Undisinfected Ground
Water
Borchardt et al. (2012) assayed 1,204 tap water samples using qPCR from 14 undisinfected
ground water systems and detected at least one virus in 287 (24 percent) samples. These results
are consistent with other PWS ground water studies (USEPA, 2006). Perhaps most significantly,
Borchardt et al. (2012) detected 51 (4 percent) norovirus positive samples with about 40
detections in the first six-month surveillance period. Significant AGI health effects were reported
during this first surveillance period, especially among children less than five years old.
EPA hypothesizes that a norovirus disease outbreak occurred in many of the 14 communities
during the first surveillance period. The outbreak was likely abetted by consumption of untreated
drinking water. Norovirus illness in the community resulted in norovirus shedding in septage and
sewage, and fecal contamination eventually arrived in untreated drinking water samples.
Consuming untreated, norovirus contaminated drinking water, likely resulted in additional health
effects in the communities. The infection cycle was halted only when most community members
not initially genetically immune, had been exposed, infected and become immune. By the time
the outbreak ended, as many as 63 percent of children under 5 years old had been exposed,
infected and ill, all during the first surveillance period
Enterovirus was detected in 109 (9 percent) samples with detections in all 4 surveillance periods
(lowest period had 10 detections). Enterovirus-related illness appears to reflect endemic, rather
than epidemic disease in the community, although the large number of enterovirus genotypes
Six-Year Review 3 Technical Support Document 6-28
for Microbial Contaminant Regulations
December 2016
-------
observed suggests that each genotype could, like norovirus, cause short period epidemics. Again,
the forthcoming scientific analysis should better document this phenomenon.
As a result of the potential public health risk associated with undisinfected PWS wells reported
by Borchardt et al. (2012) for Wisconsin, the Minnesota State Legislature requested a study of
ground waterborne viruses. The first-year results are published (Borchardt et al., 2015). Eighty-
two randomly selected (from 567 wells total) PWS wells were each sampled 6 times in a year
using qPCR and 245 virus assay results are reported (Borchardt et al., 2012); each well (but one)
sampled 3 times. Human enteric virus was detected in 41 samples from 34 wells. Seven samples
were positive for enterovirus and four samples each were positive for norovirus and rotavirus.
Nineteen samples were positive for adenovirus. [Note: one well is positive for E. coli (Borchardt
et al., 2015)]. Borchardt et al. concluded that the virus occurrence in Minnesota is, at least based
on early results, similar to previous results from Wisconsin. An epidemiological study (three
communities, three control communities) is now underway.
Under the UCMR 3, EPA sampled about 800 randomly selected undisinfected wells and
evaluated them for the presence of viruses and virus indicators using EPA Method 1615. These
data show (posted on EPA UCMR website) that only two undisinfected PWS systems were virus
positive by cell culture and no more than 16 PWSs were virus positive by qPCR.
The UCMR 3 virus results contrast significantly with the results from Borchardt et al. (2012).
One important difference is that Borchardt sampled prior to any treatment in these undisinfected
wells (e.g., softening, Fe/Mn removal). In contrast, most wells in UCMR 3 virus study were
sampled after softening or other treatment. It is unknown if the difference in sampling point
affected virus recovery. Francy et al. (2004) also sampled undisinfected PWSs for enteric virus
with sampling prior to any softening or Fe/Mn removal. They found 2 of 38 wells positive for
enterovirus by cell culture.
Available UCMR 3 data show that 252 of 793 (32 percent) of PWSs (about two thirds of wells
sampled once, others sampled twice) are aerobic spore positive. In comparison, only 41 (5
percent) and 53 (7 percent) PWSs were, respectively, enterococci or total coliform positive.
These data reflect the long lived nature of the spore as compared to the vegetative cell form of
soil bacteria.
The soil bacteria are entrained within infiltrating precipitation and are transported from the
surface or near surface to the well. These soil bacteria are found everywhere and, if found in a
well, represent a ground water recharge pathway with sufficient pore space and permeability to
permit bioparticle transport within that pathway. Because the aerobic spores are environmentally
resistant they are most likely to be found in well water as compared with the other soil bacteria,
total coliform and enterococci. The total coliform and enterococci are vegetative cells incapable
of producing a spore form and thus are more likely to be detrimentally affected by environmental
stress. The UCMR 3 virus data, as would be expected, have substantially more PWSs positive for
soil bacterial spores (32 percent) as compared with the soil bacterial vegetative cells (5 percent
enterococci and 7 percent for total coliform) because the spores are more resistant to
environmental stress.
Six-Year Review 3 Technical Support Document 6-29
for Microbial Contaminant Regulations
December 2016
-------
The UCMR 3 aerobic spore concentrations spanned three orders of magnitude, as shown in
Exhibit 6.20. For example, approximately 15 percent of detections are between 10 and 100
spore-forming units per 100 mL and approximately three percent of detections exceeded 100
spore-forming units per 100 mL. For those having concentrations of over 100 spore-forming
units per 100 mL, these PWS wells likely have a greater component of more recent surface water
and could be unrecognized GWUDI wells. These wells currently are undisinfected; treatment
such as filtration and disinfection could be warranted. For those having concentrations of
between 10 and 100 spore-forming units per 100 mL, further investigation may be warranted to
evaluate a need for disinfection or any other corrective action. We also suggest that these
undisinfected PWSs having high spore concentrations (e.g., over 100 spore-forming units per
100 mL) should be re-evaluated as possible misclassified GWUDI systems.
Exhibit 6.20: UCMR 3 Aerobic Spore Concentration Cumulative Distribution
Function
o o OO (
1 15.5 % of detects exceeded 10
O ©
0 0 OO ® °
3.2 % of detects exceeded 100
T
T
I I
500 1000
10
50 100
Spore Measurement Result
Statistical analysis (not shown) produced no significant associations between any indicator
microorganism and any pathogen or between any group of infiltration microorganisms and any
group of sewage/septage/fecal microorganisms.
Six-Year Review 3 Technical Support Document 6-30
for Microbial Contaminant Regulations
December 2016
-------
7 Treatment
This chapter summarizes the results from EPA's review of information related to the treatment of
microbial contaminants in drinking water to support the evaluation of treatment feasibility part of
the Six-Year Review. EPA conducted a scientific review of available information, published on
or before December 2015, to determine if information would suggest an opportunity to revise the
treatment technique (TT) requirements in the microbial contaminant regulations to provide
greater protection of public health. The review focused on the major provisions of the microbial
contaminant regulations where new information was identified.
7.1 Introduction
This section provides a brief overview of major TT requirements in the microbial contaminant
regulations and highlights the ones that EPA identified for further discussion in this chapter.
Additional background about these regulations is provided in Chapter 3 of this document.
SWTR. The SWTR requires all water systems using surface water or GWUDI sources (also
known as Subpart H systems) to remove (via filtration) and/or inactivate (via disinfection)
microbial contaminants to protect the public from potential adverse health effects due to
exposure to Giardia lamblia, viruses, Legionella, heterotrophic bacteria, and other pathogens
(USEPA, 1989). Specifically, it requires at least 99.9 percent (3-log) removal/inactivation of G.
lamblia and at least 99.99 percent (4-log) removal/inactivation of viruses. Other major TT
provisions include turbidity criteria for filtered systems, disinfection residual requirements prior
to point of entry to the distribution system and within the distribution system, and filtration
avoidance criteria for unfiltered systems. (USEPA, 1989). EPA published concentration x time
(CT) tables for PWSs to determine log-inactivation credit for the use of a disinfectant to meet the
disinfection TT requirements (USEPA, 1989; USEPA, 1991b).
IESWTR. The IESWTR applies to all PWSs using surface water or GWUDI, which serve 10,000
or more people. The IESWTR established TT requirements for Cryptosporidium by requiring
filtered systems to achieve at least a 99 percent (two-log) removal, tightening turbidity
performance criteria, requiring a sanitary survey for all surface water and GWUDI systems, and
setting disinfection profiling and benchmarking requirements to prevent increases in microbial
risk while systems complied with the Stage 1 D/DBPR (USEPA, 1998b).
LT1. The LT1 extended the requirements from the IESWTR for systems serving fewer than
10,000 people (USEPA, 2002).
LT2. The LT2 Rule requires 2- to 3-log inactivation of Cryptosporidium in unfiltered systems
and additional treatment for Cryptosporidium in filtered systems based on the results of source
water monitoring. The rule also requires covering of all uncovered finished water reservoirs or
for water to be treated (at least 2, 3, 4 inactivation or removal of Cryptosporidium, Giardia, and
viruses respectively (USEPA, 2006a). EPA reviewed the LT2 microbial toolbox treatment and
management strategy options for Cryptosporidium. Results of that review are provided in the
LT2 support document (USEPA, 2016a).
Six-Year Review 3 Technical Support Document 7-1
for Microbial Contaminant Regulations
December 2016
-------
Filter Backwash Recycling Rule. EPA did not identify any treatment-related topics in the FBRR
(USEPA, 2001).
Ground Water Rule. There is no distribution system disinfectant residual requirement under the
GWR (USEPA, 2006b). The CT criteria provided in the SWTRs apply to those ground water
systems that are required to disinfect. EPA did not identify any other treatment-related topics in
the GWR.
Based on the available information, EPA identified the following TT requirements of the SWTRs
(including SWTR, IESWTR and LT1) that warrant further examination in this Six-Year Review:
Requirements to maintain a minimal disinfectant residual in the distribution system
(Section 7.2), and
CT criteria for virus disinfection (Section 7.3).
7.2 Disinfectant Residual Requirements in Distribution Systems
7.2.1 Background
The term "disinfectant residual" refers to the amount of disinfectant remaining in the water after
application at some prior time, and after some amount of that applied has been exhausted. For
example, a chlorine residual is the difference between the total chlorine added and that consumed
by oxidizable matter (i.e., the chlorine demand of the water). The first step of the disinfection
process, before the water enters the distribution system, is referred to as "primary disinfection."
Primary disinfection kills or inactivates bacteria, viruses, and many other potentially harmful
organisms. Additional disinfectant can be added in a second step, called secondary disinfection,
sometime after primary disinfection but prior to entry to the distribution system or at booster
disinfection facilities in the distribution system. Secondary disinfection, with possible booster
disinfection within the distribution system, is intended to maintain a disinfectant residual
throughout the distribution system to protect drinking water quality to the customers' taps.
Distribution systems are vulnerable to contamination by a number of pathways, including the
infiltration of water external to the distribution system and by microbial growth (especially
naturally occurring bacteria such as Legionella and mycobacteria) when distribution system
conditions are favorable. Under the SWTR, the residual disinfectant concentration at the entry
point to the distribution system may not be less than 0.2 mg/L for more than four hours. The
residual disinfectant concentration in the distribution system "cannot be undetectable in more
than 5 percent of the samples each month, for any 2 consecutive months that the system serves
water to the public." (40 CFR 141.72). A detectable residual may be established by: (1) an
analytical measurement, or (2) having a heterotrophic bacteria concentration less than or equal to
500 per mL measured as heterotrophic plate count (HPC). The purpose of these disinfectant
residual requirements, as descried in the proposed SWTR (USEPA, 1987), was to:
Ensure that the distribution system is properly maintained and identify and limit
contamination from outside the distribution system when it might occur,
Limit growth of heterotrophic bacteria and Legionella within the distribution system, and
Provide a quantitative limit, which if exceeded would trigger remedial action.
Six-Year Review 3 Technical Support Document 7-2
for Microbial Contaminant Regulations
December 2016
-------
Once drinking water is disinfected to meet public health standards, the residual disinfectant level
in the distribution system must be maintained as a final barrier in protecting against waterborne
disease. Maintaining this residual disinfectant reduces bacterial growth and mitigates against
possible contamination by pathogens that may have intruded into the system. Disinfectants also
naturally degrade based on demand and water age. Operators must manage disinfectant levels on
a frequent and ongoing basis to protect consumers. One of the major purposes for requiring
distribution system residuals has historically been described as an indicator (when there is an
absence of a residual) for localized contamination and/or intrusion into the distribution system.
7.2.2 Summary of Technical Review
EPA evaluated information related to the maintenance of a minimum disinfectant level in the
distribution system and determined that a detectable concentration of disinfectant residual in the
distribution system may not be adequately protective of public health due to microbial
pathogens. This is based on concerns about analytical methods and the potential for false
positives (Westerhoff et al., 2010; Wahman and Pressman, 2015; AWWA, 2015). Maintaining a
disinfectant residual above a set numerical value in the distribution system may improve public
health protection from a variety of pathogens. Such a change could have benefits for controlling
occurrence of all types of pathogens in distribution systems, except for those most resistant to
disinfection, such as Cryptosporidium and mycobacteria. EPA noted that maintaining a
disinfectant residual above a set numerical value in the distribution system would need to also
consider impacts on the formation of DBPs (refer to the risk-balancing provisions of the SYR3
protocol).
In summer 2015, the American Water Works Association (AWWA) provided EPA with input
developed by it Disinfection Residual Strategy Panel related to the maintenance of a secondary
disinfectant residual in drinking water distribution systems (AWWA, 2015). AWWA noted that
the input primarily focused on eight topics related to the public health considerations associated
with drinking water distribution systems. These eight topics related to: analytical methods for
disinfectant residual; organic chloramines; the TCR sampling framework; minimum numerical
disinfectant residual requirements and performance objectives; institutional premise plumbing;
disinfectant residuals in ground water systems; cross-connection control; and public notification
(AWWA, 2015).
7.2.3 Detectable Residuals for Systems Using Chloramine Disinfection
For surface water systems or GWUDI systems, the SWTR requires that a disinfectant residual
cannot be undetectable in more than five percent of samples each month for any two consecutive
months (see Section 7.2.1). EPA identified two issues that have implications for the
protectiveness of allowing a detectable residual as a surrogate for bacteriological quality: organic
chloramines and nitrification. Organic chloramines affect the effectiveness of disinfectant
residuals because they:
• form during the use of free chlorine or chloramines,
interfere with commonly used analytical methods for free and total chlorine
measurements, and
Six-Year Review 3 Technical Support Document 7-3
for Microbial Contaminant Regulations
December 2016
-------
are poor disinfectants compared to free chlorine and monochloramine (Wahman and
Pressman, 2015).
Organic chloramines are known to be weaker disinfectants than free chlorine or inorganic
chloramines, showing little or no bactericidal activity. For example, the CT (concentration x
time) required to reach 99 percent (two-log) inactivation (CT99) of Escherichia coli (ATCC
25922) exposed to free chlorine, monochloramine and organic chloramines was approximately
0.5 and 10 mg C12 - min/L for free chlorine and monochloramine, respectively, while the
organic chloramine was very similar in measured residual concentration to the control
experiment and showed minimal inactivation of E. coli.
Because chloramination involves introduction of ammonia into drinking water, and
decomposition of chloramines can further release ammonia in the distribution system,
chloramine use comes with the risk of distribution system nitrification (i.e., the biological
oxidation of ammonia to nitrite and eventually nitrate). Drinking water distribution system
nitrification is undesirable and can result in water quality degradation (e.g., disinfectant
depletion, increased heterotrophic bacteria occurrence or nitrite/nitrate formation). Information
shows that maintaining a high enough level of total chlorine or monochloramine residuals in the
distribution system can help prevent both nitrification and residual depletion (Stanford et al,
2014).
7.2.4 State Implementation of Disinfectant Residual Requirements
States may adopt federal drinking water regulations or promulgate more stringent drinking water
requirements, including those for disinfectant residuals. Twenty states require a minimum free
chlorine residual of 0.2 mg/L or more (Ingels, 2015; Wahman and Pressman, 2015). Five of the
20 states set standards more stringent than 0.2 mg/L: Florida, Illinois, Iowa, and Delaware
require 0.3 mg/L; in its Emergency Distribution Disinfection Rule, Louisiana required at least
0.5 mg/L free chlorine throughout the distribution system at all times. For minimum total
chlorine residual, state requirements vary from 0.05 mg/L (New Jersey) to 1.00 mg/L or higher
(Kansas, Oklahoma, Iowa, Ohio, and North Carolina). In its Emergency Distribution
Disinfection Rule, Louisiana required a chloramine residual (measured as total chlorine) of 0.5
mg/1 throughout the distribution system at all times for systems that feed ammonia. North
Carolina has a numeric requirement for total chlorine residual but not for free chlorine residual.
Exhibit 7.1 and Exhibit 7.2 present the state requirements for free and total chlorine,
respectively, as of January 2015.
Colorado has amended its minimum disinfectant residual requirements in the distribution system
to be greater than or equal to 0.2 mg/L, effective April 1, 2016 (Ingels, 2015). Pennsylvania
recently proposed to strengthen its disinfectant residual requirements by increasing the minimum
disinfectant residual in the distribution system to 0.2 mg/L free or total chlorine (Pennsylvania
Bulletin, 2016). In March 2016, Louisiana finalized requirements that chlorinating public water
systems maintain at least 0.5 mg/L free chlorine and chloraminating systems maintain at least 0.5
mg/L total chlorine residual throughout the distribution system at all times.
In 2013, Louisiana promulgated an Emergency Distribution Disinfectant Residual Rule that
required routine, continuous disinfection of all public water systems. The rule was promulgated
Six-Year Review 3 Technical Support Document 7-4
for Microbial Contaminant Regulations
December 2016
-------
to control Naegleriafowleri, an amoeba found in several public water systems. For systems
using chlorine disinfection, the free chlorine residual at the entry point to the distribution system
had to be 0.5 mg/L for pH less than 7.0; 0.6 mg/L for pH of 7.0 to 8.0; 0.8 mg/L for pH of 8.0 to
9.0; and 1.0 mg/L for pH greater than 9.0. Disinfectant residual monitoring was required at 25
percent more sites than required by the TCR and daily residual measurements were required at
the point of maximum residence time in the distribution system. A minimum free or total
chlorine disinfectant levels of 0.5 mg/L was required to be maintained at all times in finished
water storage tanks and the entire distribution system (Louisiana Department of Health and
Hospitals, 2013). Systems using chloramine disinfection were also required to develop and
submit a nitrification control plan. The rule initially became effective on November 6, 2013 and
was renewed five times between March 2014 and July 2015. In March 2016, the Louisiana State
Sanitary Code was amended to make the Emergency Rule's requirements permanent. The March
2016 rule maintains the requirements of the Emergency Rule and strengthens monitoring
requirements for public water systems using chloramine disinfection. The Rule also requires
public water systems using chloramines to monitor for nitrification and to take corrective action
as needed, in accordance with an approved nitrification plan.
Exhibit 7.1: Distribution System Minimal Residual Requirements by States - Free
Chlorine
Residual (mgCI^/L)
X Detectable
Six-Year Review 3 Technical Support Document 7-5
for Microbial Contaminant Regulations
December 2016
-------
Exhibit 7.2: Distribution System Minimal Residual Requirements by States - Total
Chlorine
Most states have requirements for the design, construction, operation and maintenance of
distribution systems. Under the TCR/RTCR, each system must develop and monitor disinfectant
residuals according to a written sample siting plan, which is subject to state review and revision.
The AWWA Disinfection Residual Strategy Panel noted that there appears to be lack of
consistency of distribution system requirements set by states and that there is concern that the
TCR/RTCR sampling framework may not be optimized to find distribution system problem areas
for residual monitoring purposes (AWWA, 2015). The current sampling protocol for drinking
water disinfectant residual is tied to total coliform sampling sites as required under the
TCR/RTCR. Depending on drinking water distribution system hydraulics, the Panel noted that
the TCR/RTCR sampling framework may not provide an accurate assessment of when there is an
absence of a disinfectant residual at a specific location not associated with TCR/RTCR sampling
(AWWA, 2015). Areas of the drinking water distribution system that can be particularly
vulnerable to microbial contaminations are dead ends, areas near improperly functioning valves,
pressure zone boundaries, blending zones and areas of the system under lower flow conditions.
Additional considerations in selecting distribution system sampling locations may include
sensitive population areas (e.g., near schools or hospitals), pipe types relative to corrosion (e.g.,
old, unlined cast iron) and storage facilities. The Panel noted that continuous monitoring of
disinfectant residuals, which may not be feasible in some cases (e.g., in very small systems or in
locations without power or sanitary disposal), should be encouraged wherever feasible, with
priority given to system critical control points, such as areas near storage facilities, maximum
residence time sites, and sites with organic chloramine issues (AWWA, 2015).
Six-Year Review 3 Technical Support Document 7-6 December 2016
for Microbial Contaminant Regulations
Residual (mg CI2/L)
/ Detectable
-------
7.2.5 Disinfectant Residuals for Control of Legionella in Premise Plumbing Systems
Since the reporting of disease outbreaks due to Legionella began in 2001, Legionella has been
shown to cause more drinking-water-related outbreaks than any other microorganism.
Addressing premise plumbing issues is particularly challenging. Premise plumbing may be
largely outside of water utilities' operations and management control. Also, the characteristic
features of premise plumbing (e.g., low disinfectants residuals, stagnation, and warm
temperature) has a greater tendency to support growth and persistence of opportunistic
pathogens.
Studies indicate that distribution systems can play a role in influencing the transmission and
contamination of Legionella in premise plumbing systems (Lin et al., 1998; States et al., 2013).
Hospitals served by PWSs using chloramines reported fewer outbreaks of legionellosis than
those using free chlorine (Kool et al., 1999; Heffelfinger et al., 2003). Some building systems
supplied by PWSs which have switched to chloramines have seen marked reduction in the
colonization of Legionella (Flannery et al., 2006; Moore et al., 2006). One implication of these
studies is the importance of being able to reliably measure and sustain chloramine residuals to
increase the likelihood of its effectiveness at controlling Legionella in premise plumbing
systems. On the other hand, some studies have indicated that the occurrence of another pathogen,
non-tubercular Mycobacterium, may increase under chloramination conditions (Pryor et al.,
2004; Moore et al., 2006; Duda et al., 2014).
Legionella species can multiply in warm, stagnant water environments, such as in community
water storage tanks with low disinfectant residuals during warm months. Cohn et al. (2014)
observed increased incidence of legionellosis among institutions and private homes near a
community water system storage tank when the disinfectant residual in the storage tank dropped
(from greater than 0.2 mg/L to less than 0.2 mg/L) during hot summer months. Based on these
findings, the authors recommended that, regardless of total coliform occurrence, remedial actions
be taken (e.g., flushing of mains, checking for closed valves that can result in hydraulic dead-
ends, and possibly installing re-chlorination stations) when low chlorine residuals are observed
during hot summer months. They also noted that this storage tank had been cleaned subsequent
to the outbreak (Cohn et al., 2014; Ashbolt, 2015).
To help address concerns about Legionella, EPA developed a document entitled Technology for
Legionella Control in Premise Plumbing Systems: Scientific Literature Review (USEPA, 2016h).
This document summarizes information about the effectiveness of different approaches to
control Legionella in a building's premise plumbing system. EPA expects that this document will
improve public health protection by helping primacy agencies, facility operators, facility owners,
technology developers and vendors make science-based risk management decisions to control of
Legionella growth in buildings.
EPA also reviewed the scientific literature on the effectiveness of disinfectant residuals at
controlling biofilm growth. Many factors influence the concentration of the disinfectant residual
in the distribution system; and therefore, the ability of the residual to control microbial growth
and biofilm formation. These factors include the availability of nutrients (such as assimilable
Six-Year Review 3 Technical Support Document 7-7
for Microbial Contaminant Regulations
December 2016
-------
organic carbon (AOC)), the type and concentration of disinfectant, water temperature, pipe
materials, and system hydraulics.
Biofilms in distribution systems have been associated with enhanced corrosion of pipes and
deterioration of water quality. Biofilms can provide ecological niches that are suited to the
potential survival of pathogens (Walker and Morales, 1997; Baribeau et al., 2005; Behnke et al.,
2011; Wangetal., 2012; Biyela et al., 2012; Revetta et al., 2013; Ashbolt, 2015). The biofilm
can protect microorganisms from disinfectants and can enhance nutrient accumulation and
transport (Baribeau et al., 2005).
7.2.6 HPC Alternative to Detectable Residual Measurement
Under the SWTR, a system may demonstrate that its HPC levels are less than 500 per mL, at any
sampling locations, in lieu of demonstrating the presence of a detectable disinfectant residual at
that location, per primacy agency approval. Criteria used in the Netherlands for systems
operating without a distribution system disinfectant residual provides an example of an
alternative criteria than the HPC criterion. In the Netherlands, chlorine is not used routinely for
primary or secondary disinfection. They focus on maintaining a high-quality distribution system
with sufficient pressure to prevent ingress of contamination during normal operation (Smeets et
al., 2009). The leakage rate in Netherlands is low, generally less than three percent. Variable
pumps, pressure dampening devices and automated distribution control are used to minimize
pressure fluctuations and surges that could result in negative pressure in the distribution system.
Additionally, Dutch water systems use the following general approach to control microbial
activity in the distribution system without a disinfectant residual (Smeets et al., 2009):
Produce a biologically stable drinking water;
Use distribution system materials that are non-reactive and biologically stable; and
Optimize distribution system operations and maintenance practices to prevent stagnation
and sediment accumulation.
For the determination of a biologically stable water they use AOC as an indicator. Aeration and
sand filtration generally can achieve biostable water with AOC levels below 10 |ig carbon per
liter. All materials in the Netherlands have to be tested with the biofilm formation potential test
before they can be used in drinking water. The majority of the distribution systems consist of
biostable asbestos cement or polyvinyl chloride (Smeets et al., 2009).
7.2.7 Research and Information Collection Partnership Findings
EPA also reviewed key findings by the Research and Information Collection Partnership (RICP)
on drinking water distribution system issues and research and information collection priorities.
The RICP is a working group formed on the recommendation of the Total Coliform Rule
Distribution System Advisory Committee to identify specific high-priority research and
information collection activities and to stimulate water distribution system research and
information collection (USEPA, 2008).
Six-Year Review 3 Technical Support Document 7-8
for Microbial Contaminant Regulations
December 2016
-------
The RICP partners are EPA and Water Research Foundation. EPA examined information from
the 10 high priority RICP areas in the context of the Six-Year Review, particularly information
related to the effectiveness of sanitary survey and corrective action requirements under the
IESWTR. However, the RICP found limited new information that would shed light on the
frequency and magnitude of distribution system vulnerability events (e.g., backflow events,
storage tank breaches), associated risk implication, and costs for preventing such events from
occurring. The RICP report identifies potential follow-up research areas that could help to
address these gaps (USEPA and Water Research Foundation, 2016).
7.3 CT Criteria for Virus Disinfection
7.3.1 Background
Primary drinking water treatment processes (e.g., coagulation and filtration) are less effective at
removing viruses than removing other pathogen types of concern (e.g., bacteria and protozoa)
(USEPA, 1991b). Therefore, the disinfection process is very important for inactivation of
infectious viruses in drinking water. The efficacy of disinfection can be measured as a CT value.
In the Guidance Manual for Compliance with the Filtration and Disinfection Requirements for
Public Water Systems Using Surface Water Sources (USEPA, 1991b), EPA recommends a CT
value of 8 mg/L-min to achieve a 4-log inactivation of viruses with chlorine at 5 °C, pH 6-9.
EPA also recommends a CT value of 1,988 mg/L-min to achieve a 4-log inactivation of viruses
with chloramines at 5 °C, pH 8 (1991). EPA obtained these CT values from bench-scale
inactivation experiments conducted in buffered, demand-free (BDF) water with dispersed
hepatitis A virus (HAV) (USEPA, 1991b). Over the years, many studies have indicated that
HAV is less chlorine-resistant than enteroviruses, such as Coxsackie virus B5 (CB5), and also
less chloramine-resistant than adenovirus 2 (AD2). CB5 has generally been related to less severe
health effects and a lower frequency of WBDOs compared to HAV. For example, CB5 causes
hand, foot and mouth disease in children whereas HAV may cause hepatitis with significant
inflammation of the liver when infection occurs later in life (Sinclair et al., 2005). In the EPA
CCL selection process, HAV was ranked higher than enterovirus based on the WBDO,
occurrence, and health effects of the pathogens (USEPA, 2009a).
More than 100 known enteric viruses can be excreted in large numbers in human feces of
infected individuals and are potentially transmitted by water. Those of particular significance,
either due to severity or frequency of infection include HAV, enteroviruses, Norwalk-type
viruses, rotaviruses, adenoviruses and reoviruses. EPA included some of these viruses on the
Contaminant Candidate List 2 (CCL2) (i.e., adenoviruses, caliciviruses, Coxsackie virus, and
echoviruses), Contaminant Candidate List 3 (CCL3) (i.e., adenovirus, caliciviruses, enterovirus,
and HAV) and Contaminant Candidate List 4 (CCL4) (i.e.., adenoviruses, caliciviruses,
enterovirus, and HAV) as potential microbiological contaminants of concern in public drinking
water systems (USEPA, 2005; 2009b and 2016i). The publication of the CCL2 and CCL3 has
prompted new disinfection studies on the suite of CCL2 and CCL3 viruses. EPA reviewed these
new studies, along with other relevant studies, as data/information sources as part of the Six-
Year Review 3.
Six-Year Review 3 Technical Support Document 7-9
for Microbial Contaminant Regulations
December 2016
-------
The analysis did not include G. lamblia disinfection or virus disinfection by other disinfectants
(such as ozone, chlorine dioxide and UV light) because new information do not indicate a
potential for a change to CT values for those disinfectants.
7.3.2 Summary of Technical Review
EPA evaluated whether the current CT criteria based on HAV (USEPA, 1991b) are sufficiently
protective against other types of viruses. Many studies have indicated that HAV is less chlorine-
resistant than some enteroviruses, such as Coxsackie virus B5 (Black et al., 2009; Cromeans et
al., 2010; Keegan et al., 2012), and also less chloramine-resistant than adenovirus (Sirikanchana
et al., 2008; Hill and Cromeans, 2010). Based on this review, EPA identified a potential need to
update CT values for virus inactivation by free chlorine or chloramines, particularly for water
with a relatively high pH. This assessment is also relevant to the LT2 and the GWR, which refer
to the same CT tables in the original 1991 SWTR Guidance Manual (USEPA, 1991b).
7.3.3 Basis of CT Values for Virus Inactivation in the EPA Guidance Manual
7.3.3.1 Free Chlorine
EPA developed CT values (USEPA, 1991b) for 2-, 3-, and 4-log virus inactivation by chlorine
based on HAV data provided by Sobsey et al. (1988) in a research report. Sobsey et al. (1988)
conducted bench-scale experiments with dispersed HAV in buffered, demand free (BDF) water
at 5 °C. To set the CT standards, EPA grouped CT values for pH range of 6 to 9 together, created
a separate set of values for pH 10, and applied a safety factor of three to the original CT values.
The EPA CT values for chlorine incorporate a safety factor of three to account for differences
between dispersed versus aggregated HAV and the use of BDF versus environmental water. EPA
determined CT values at temperatures other than 5 °C by assuming a two-fold decrease in CT
values for every 10 °C increase. Exhibit 7.3 presents the CT values at 5 °C from the original
study and those established in the EPA guidance manual.
Six-Year Review 3 Technical Support Document 7-10
for Microbial Contaminant Regulations
December 2016
-------
Exhibit 7.3: CT Values for Inactivation of HAV at 5 °C
Log
Inactivation
Sobsey et al. (1988)
USEPA (1991)
pH 6
pH 7
pH 8
pH 9
pH 10
pH 6-9
pH 10
Inactivation of HAV by 0.5 mg/L of Free Chlorine1
2
1.18
0.70
1.00
1.25
9.8
4
30
3
1.75
1.07
1.51
1.9
14.6
6
44
4
2.33
1.43
2.03
2.55
19.3
8
60
Inactivation of HAV5 by 10 mg/L of Chloramines2
2
NA
NA
857
NA
NA
857
3
NA
NA
1,423
NA
NA
1,423
4
NA
NA
1,988
NA
NA
1,988
1 A safety factor of three was applied to the laboratory data to derive the EPA values.
2 No safety factor was applied to the laboratory data to derive the EPA values. Pre-formed chloramines were
used.
All units are mg/L-min.
NA = Not applicable because values at individual pH values was not provided.
HAV = Hepatitis A virus.
Sobsey et al. (1988) also generated data for Coxsackie virus B5 (CB5), and coliphages MS2 and
bacteriophage OX174, and demonstrated that CB5 was more resistant to disinfection than HAV
and MS2 and OX174 across the range of pH values tested (CB5 data in Section 7.3.5). EPA
selected HAV as a target virus for setting the guidelines primarily because of the severity and
frequency of the disease it caused (USEPA, 1991b).
7.3.3.2 Chloramines
EPA adopted the chloramines CT values from Sobsey et al. (1988) without applying a safety
factor to their laboratory data (Exhibit 7.3). Sobsey et al. (1988) used preformed chloramines
(i.e., chlorine mixed with ammonia before addition) in the laboratory experiments. However, in
the field application of chloramines at a drinking water facility, systems often add chlorine prior
to ammonia, which provides a very short, but sufficient contact with free chlorine to inactivate
viruses that are resistant to chloramines, such as rotaviruses (USEPA, 1991b). Because field
chloramination is more effective than the preformed chloramines used in the laboratory
experiments, EPA did not apply a safety factor to the laboratory data. Systems that apply
preformed chloramines, although not very common, should not use the EPA CT values as a
guideline because they may not be adequate for inactivating rotaviruses (USEPA, 1991b).
In their study, Sobsey et al. determined monochloramine inactivation of HAV under only one
experimental condition - pH 8, 5 °C, and 10 mg/L of monochloramine. Since systems typically
apply chloramines to drinking water at concentrations of 1 to 2 mg/L, contact times would have
to be on the order of several hundred to even a few thousand minutes to achieve 4-log
inactivation of HAV and other enteric viruses. Many water systems are not likely to achieve such
long contact times. Therefore, systems often use free chlorine as a primary disinfectant to
Six-Year Review 3 Technical Support Document 7-11
for Microbial Contaminant Regulations
December 2016
-------
achieve the EPA CT values, and chloramines as a secondary disinfectant to maintain a stable
disinfectant residual in the distribution system and minimize trihalomethane (THM) or haloacetic
acid (HAA) formation.
7.3.4 Information on Virus Inactivation by Free Chlorine
Since the development of the CT tables, researchers and others have published literature
regarding virus disinfection, inactivation kinetics, CT value estimation, and recommendations.
However, comparisons between their studies is difficult because of differences in the viruses
examined, experimental conditions investigated, and analytical/calculation methods used. In
general, the published data on different viruses tested shows the variability of inactivation
observed for a range of viruses under different conditions. It is not the intent of this document to
perform a meta-analysis of the published CT values. This document describes the relevant new
information published since 2006, and refers to only a few older papers, as needed for context.
Some researchers have examined both chlorine and chloramines disinfection in the same studies;
results for chloramines are presented in Section 7.3.5. Relative Resistance of Viruses to Chorine
Disinfection.
Black et al. (2009) conducted experiments in BDF water (5 °C, pH 7.5 and 9.0) to determine the
chlorine CT values for CB5, echovirus 1 and 12, and poliovirus 1 (PV-1). Exhibit 7.4 presents
their reported CT values, along with the EPA required values. The results of this study suggest
that commonly used concentrations of free chlorine and contact times at drinking water treatment
plants (1 mg/L for 30 to 60 min) at a pH of 7.5 would inactivate all of the study viruses and meet
the EPA requirements for 2-, 3-, and 4-log inactivation, except for CB5. At pH 9.0, CT values
exceeded the EPA values by more than double for most of the viruses. This suggests that higher
pH levels may require longer contact times to achieve adequate disinfection of CB5. The greater
resistance of CB5 to chlorine disinfection is attributed to purified CB5 aggregating rapidly at all
pH values (Jensen et. al., 1980).
Six-Year Review 3 Technical Support Document 7-12
for Microbial Contaminant Regulations
December 2016
-------
Exhibit 7.4: CT Values for Virus Inactivation with 1.0 mg/L of Free Chlorine at 5°C
PH
Log
Inactivation
Black et al. (2009)12
USEPA (1991)
CB53
E14
E124
PV-15
2
5.4
1.6
2.1
1.4
4
7.5
3
8.4
3.5
4.4
3
6
4
11.5
6.2
7.4
5.3
8
2
14
3.3
8.4
8.2
4
9.0
3
18.7
8.5
18.5
14.7
6
4
22.9
16.6
32.3
22.3
8
1 The Black et al. (2009) values do not include a safety factor, whereas the USEPA (1991) values do.
2 CT values exceeding the EPA values are in bold font.
3 Coxsackie virus (CB5).
4 Echoviruses (E1 and E12).
5 Poliovirus (PV-1).
All units are mg/L-min.
The Water Research Foundation co-sponsored a project (#3134) titled "Contaminant Candidate
List Viruses: Evaluation of Disinfection Efficacy" (Hill and Cromeans, 2010). As part of this
project, Cromeans et al. (2010) performed disinfection experiments on several human
adenoviruses (AD2, AD40 and AD41), two Coxsackie viruses (CB3 and CB5), two echoviruses
(El and El 1) and murine norovirus (MNV, studied as a surrogate for human norovirus) with 0.2
mg/L of free chlorine and 1 mg/L of monochloramine at pH 7 and 8 in BDF water at 5°C. The
results of the free chlorine inactivation are shown in Exhibit 7.5 (the results for monochloramine
are presented in Section 7.4.1). The enteroviruses (e.g., CB5) required the longest times for
chlorine inactivation and MNV the least time. CB5 required the longest exposure time, with CT
values of 7.4 and 10 for 4-log inactivation at pH 7 and 8, respectively. All the CT values
obtained met the EPA values for 2-, 3- and 4-log inactivation, except for CB5 at pH 8.
Six-Year Review 3 Technical Support Document 7-13
for Microbial Contaminant Regulations
December 2016
-------
Exhibit 7.5: CT Values for Virus Inactivation with 0.2 mg/L of Free Chlorine at 5°C
PH
Log
Inactivation
Cromeans et al. (2010)1,2 3
USEPA
(1991)
AD24
AD404
AD414
CB35
CB55
E16
E116
MNV7
7
2
0.02
<0.02
0.005
0.97
3.6
0.96
0.82
<0.02
4
3
0.06
<0.02
0.01
1.4
5.5
1.3
1.0
<0.02
6
4
0.15
<0.04
ND
2.9
7.4
1.5
1.1
<0.07
8
8
2
0.04
<0.02
<0.02
0.65
4.7
0.99
0.54
<0.02
4
3
0.12
<0.02
<0.02
1.1
7.6
1.3
0.97
<0.02
6
4
0.27
<0.04
<0.03
1.7
10
1.6
1.4
<0.08
8
1 The Cromeans values do not include a safety factor, whereas the EPA values do.
2 CT values exceeding the EPA values are in bold font.
3 ND = no data. The CT value could not be extrapolated due to asymptotic inactivation curves.
4 Adenoviruses (AD2, AD40 and AD41).
5 Coxsackie viruses (CB3 and CB5).
6 Echoviruses (E1 and E11).
7 Murine norovirus (MNV).
All units are mg/L-min.
7.3.1.4 Effect of Cell Association and Virus Aggregation
Virus particles in water can exist as single particles, as aggregates or clumps (groupings of two
or more virus particles) and associated with the host cellular material. Chlorine disinfection relies
on the ability of the chemical disinfectants to come into contact with the target organism. Where
solid particles are present, disinfection may be impeded because particles interfere with contact
between the disinfectant and the target organism (Templeton et al., 2008). There is considerable
evidence that most viruses in water are embedded in or associated with suspended solids and that
such association often interferes with virus inactivation (Sobsey et al., 1991).
Sobsey et al. (1991) compared inactivation of cell-associated and dispersed HAV by free
chlorine and chloramines. Exhibit 7.6 presents the CT values for 4-log inactivation of cell-
associated and dispersed HAV from their study. These results indicate that cell-associated HAV
was about tenfold more resistant than dispersed HAV to free chlorine at pH 6 and 8 and about
fivefold more resistant at pH 10.
Six-Year Review 3 Technical Support Document 7-14
for Microbial Contaminant Regulations
December 2016
-------
Exhibit 7.6: CT Values for 4-Log Inactivation of Cell-Associated and Dispersed
HAV at 5°C
Disinfectant
PH
Sobsey et al. (1991 )1
CT Ratio of Cell-Associated
vs. Dispersed HAV
Cell-Associated
HAV2
Dispersed
HAV2
Free Chlorine
6
29
2.3
12.6
8
27
2
13.5
10
104
19.3
5.4
Chloramines
8
1740
1225
1.4
1 In 0.01 M phosphate-buffered, halogen-demand-free reagent water.
2 Hepatitis A virus (HAV).
All units are mg/L-min.
Hill and Cromeans (2010) examined the effect of aggregation on disinfection efficacy for AD2
as it was the only study virus that successfully aggregated and retained a level of aggregation.
The CT value for chorine disinfection of aggregated AD2 was twofold that of dispersed AD2 in a
river source water (Exhibit 7.7).
Exhibit 7.7: CT Values for Inactivation of Aggregated and Dispersed AD2 at 5°C
and 0.2 mg/L Free Chlorine in a River Source Water
Log
Inactivation
Hill and Cromeans (2010)
Ratio of Aggregated
vs. Dispersed AD2
Aggregated AD2
Dispersed AD2
2
0.16
0.077
2.1
3
ND
0.15
Not applicable
4
ND
0.23
Not applicable
AD2 = Adenovirus.
ND = no data. The CT value could not be extrapolated due to asymptotic inactivation curves.
All units are mg/L-min.
7.3.4.2 Effect of Source Water Quality
Many researchers have performed disinfection studies on inactivation of viruses in purified and
BDF, reagent-grade water. Previous studies conducted with natural waters demonstrated both
increased and decreased disinfection efficacy of chlorine in these waters compared to purified or
buffered waters (Thurston-Enriquez et al., 2003). Since EPA derived the CT values from
inactivation experiments using dispersed HAV in BDF water, it is important to examine whether
these CT values are sufficient for inactivation of viruses in natural source waters.
Kahler et al. (2010) investigated the effect of source water quality on chlorine inactivation of
four viruses—CB5, El, MNV and AD2—in one untreated groundwater source and two partially
treated surface waters (obtained just prior to chemical disinfection). In all source water types,
chlorine disinfection was most effective for MNV and least effective for CB5. Inactivation of the
study viruses differed significantly between source water types, but there were no clear water
Six-Year Review 3 Technical Support Document 7-15
for Microbial Contaminant Regulations
December 2016
-------
quality characteristics trends that were associated with the lowest or highest disinfection efficacy
overall. However, CT values for CB5 in one partially treated river water exceeded the EPA CT
values (Exhibit 7.8). The results of this study demonstrate that source water quality plays a
substantial role in the inactivation of viruses and that water utilities should consider source water
quality in their disinfection practices.
Exhibit 7.8: CT Values for 3-Log Virus Inactivation in a River Source Water with
0.2 mg/L of Free Chlorine
Temp.
(°C)
PH
Kahler et al. (2010)12
USEPA (1991)
AD2
CB5
E1
MNV
5
7
0.099
5.2
0.79
0.020
6
8
0.12
7.9
1.2
0.031
6
15
7
0.063
2.0
0.48
<0.020
3
8
0.061
3.6
0.84
0.020
3
1 The Kahler et al. 2010 values do not include a safety factor, whereas the EPA values do.
2 CT values exceeding the EPA values are in bold font.
All units are mg/L-min.
AD2 = Adenovirus, CB5 = Coxsackie virus, E1 = Echovirus, MNV = Murine norovirus.
Page et al. (2009) characterized the effects of pH, temperature and other relevant water quality
parameters on the kinetics of AD2 inactivation with free chlorine. Over a pH range of 6.5 to 10,
a temperature range of 1°C to 30°C and in a variety of water types, free chlorine was highly
effective against AD2 (Page et al., 2009). They developed an inactivation model as a function of
relevant water quality parameters and disinfectant exposure. The researchers noted that the
model provided adequate representation for the free chlorine inactivation of AD2 and
comparable results to those reported in the literature for other adenovirus serotypes (Page et al.,
2009).
7.3.4.3 Determination of CT Values for CB5 in Recycled Water
In February 2013, the Australian Department of Health (ADOH) published disinfection
guidelines for recycled waters, with virus CT values adopted from recent work by the Australia
Water Quality Centre (Keegan et al., 2012). Keegan et al. conducted a detailed, comprehensive
literature review on virus inactivation in drinking water and wastewater, including the following
topics:
Viruses and viral indicators of interest in wastewater effluents;
Effects of temperature, pH, ionic strength, virus aggregation and particulates and
turbidity on disinfection;
Relative resistance of viruses to disinfection;
Matrix effects;
Use of laboratory versus environmental viruses to derive CT values;
Virus selection matrix; and
Methods for data analysis.
Six-Year Review 3 Technical Support Document 7-16
for Microbial Contaminant Regulations
December 2016
-------
Keegan et al. (2012) concluded that while the EPA CT values for viruses were appropriate for
drinking water with turbidity of < 1 NTU, research was needed to address a range of factors for
recycled waters, such as target virus, resistance, state of the virus, particle protection, turbidity
and ionic strength of the water. Therefore, the researchers conducted chlorine experiments with
CB5 on secondary treated wastewater with various turbidities (0.2, 2, 5 and 20 NTU) and pH
values (7, 8 and 9) at 10°C (Keegan et al., 2012).
The secondary treated wastewater had undergone primary sedimentation, activated sludge
treatment and clarification. The authors diluted the water with ultrapure water by 60 percent to
obtain 600 mg/L of total dissolved solids to match Victoria wastewater treatment plant
conditions. The authors adjusted the turbidity by either filtering water (0.2 NTU), diluting water
(2 NTU), or concentrating water and resuspending particulates in water (5 and 20 NTU). In
addition, the authors used BDF water at pH 9 to reproduce published data by Black et al. (2009)
(data shown in Exhibit 7.10).
Exhibit 7.9 presents the results of Keegan et al.'s chlorine experiments and the CT values
adopted in the Australian guidelines for recycled waters (ADOH, 2013). No safety factor was
used to derive the CT values in the ADOH guidelines because of the use of the secondary treated
wastewater, not BDF water. As shown in Exhibit 7.9, the CT values for the low turbidity water
(<2 NTU) in the ADOH guidelines are similar to the EPA values at pH 7, but much higher at pH
8 and 9. For example, EPA requires a CT value of 8 mg/L-min (including a safety factor of 3) to
achieve 4-log inactivation, whereas ADOH requires a CT value of 16 mg/L-min at pH 8 and 27
mg/L-min at pH 9, which are 2 and 3.4 times the EPA value, respectively.
Six-Year Review 3 Technical Support Document 7-17
for Microbial Contaminant Regulations
December 2016
-------
Exhibit 7.9: CT Values for Inactivation of CB5 with Free Chlorine in Recycled
Water at 10°C
Log
Keegan et al. (2012)14
ADOH (2013)2 4
USEPA
PH
Inactivation
Turbidity (NTU)
Turbidity (NTU)
(1991 )3
0.2
2
5
<=2
<=5
1
2.05
2.13
2.24
3
3
2
7
2
3.29
3.37
3.71
4
4
4
3
4.41
4.75
4.88
5
5
6
4
5.44
5.46
5.99
6
7
8
1
5.72
6.67
7.78
7
9
2
8
2
9.6
10.32
13.16
10
13
4
3
12.8
12.90
17.79
13
18
6
4
15.49
15.68
21.94
16
23
8
1
8.25
8.94
9.66
10
10
2
9
2
14.06
15.5
16.33
16
16
4
3
19.10
20.88
22.03
21
23
6
4
23.97
26
27.93
27
29
8
1 Chlorine dosages of 6.5 to 6.9 mg/L were used in the tests. CT values for 20 NTU turbidity tests are not shown
because they were not used in the ADOH guideline.
2 No safety factor was used.
3 HAV was used for deriving the CT values at 5°C. A safety factor of three was used.
4 CT values exceeding the EPA values are in bold font.
All units are mg/L-min.
7.3.4.4 Comparison of CT Values for CB5
Exhibit 7.10 summarizes the CT values for inactivating CB5 at various pH values, as reported by
different studies, and lists the key experimental parameters used in these studies. Among the four
studies, only Sobsey et al. (1988) reported CT values for pH 6 and 10. In general, the CT values
reported in Sobsey et al. (1998) are much higher than in other studies, which could be due to
differences in the calculation methods explained later. Comparisons of CT values for pH 7 to 9
are described as follows:
Comparison of CT values for pH 7. The two sets of CT values by Cromeans et al. (2010) and
Keegan et al. (2012) are comparable, being below the respective EPA values of 4, 6 and 8 for 2-,
3- and 4-log inactivation.
Comparison of CT values for pH 7.5. Only Black et al. (2009) reported data for this pH. For the
rest of the studies, EPA believes the CT values for pH 7.5 were derived by averaging CT values
for pH 7 and 8 {numbers in italic). All the CT values exceeded the respective EPA values of 4, 6
Six-Year Review 3 Technical Support Document 7-18
for Microbial Contaminant Regulations
December 2016
-------
and 8 for 2-, 3- and 4-log of inactivation. Keegan et al. (2012)'s data are similar to Black et al.
(2009).
Comparison of CT values for pH 8. Like for the pH 7.5 data, the Keegan et al. (2012) and the
Black et al. (2009) data are similar for pH 8. All the CT values reported for pH 8 exceeded the
respective EPA values of 4, 6 and 8 for 2-, 3- and 4-log of inactivation. Keegan's data are 55
percent to 104 percent higher than Cromeans et al. (2010) data.
Comparison of CT values for pH 9. All the CT values reported for pH 9 exceeded the respective
EPA values of 4, 6 and 8 for 2-, 3- and 44og of inactivation. Nonetheless, Keegan et al., (2012)'s
two sets of data in different water matrices are very similar to Black et al. (2009)'s. Good
correlation was observed between the 2 studies.
Exhibit 7.10: Comparison of CT Values for Inactivation of CB5 with Free Chlorine
PH
Log
Inactivation
Sobsey et al.
(1988)1
Black et al.
(2009)
Cromeans et
al.
(2010)
Keegan et al.
(2012)
USEPA
(1991 )3
CT Values5
2
3.5
4
6
3
4.4
No data
6
4
6.6
8
2
ND
ND
3.6
3.3
4
7
3
ND
ND
5.5
4.4
6
4
12
ND
7.4
5.4
8
2
ND
5.4
4.22
e.42
4
7.5
3
ND
8.4
6.62
8.62
6
4
19.12
11.5
8.T2
10.52
8
2
13
ND
4.7
9.6
4
8
3
ND
ND
7.6
12.8
6
4
26.2
ND
10.0
15.5
8
2
ND
14
ND
14.49
4
9
3
ND
18.7
ND
18.30
6
4
54
22.9
ND
22.13
8
2
206
30
10
3
ND
No data
44
4
413
60
Experimental Conditions
Six-Year Review 3 Technical Support Document 7-19
for Microbial Contaminant Regulations
December 2016
-------
PH
Log
Inactivation
Sobsey et al.
(1988)1
Black et al.
(2009)
Cromeans et
al.
(2010)
Keegan et al.
(2012)
USEPA
(1991 )3
Water Matrix
0.01 M
phosphate
BDF water
0.01 M
phosphate
BDF water
0.01 M
phosphate BDF
water
0.01 M
phosphate
BDF water
NA
Temperature (°C)
5
5
5
5
NA
Free Chlorine
Residual (mg/L)
0.5
1.0
0.2
1.5
NA
Calculation Method
C x t4
EFH model
EFH model
Estimated by
determining the
area under
chlorine decay
curve of
chlorine
concentration
vs. time
1 Data for selected pH and log of inactivation not found in the paper.
2 EPA estimated data for pH 7.5 using average values of pH 7 and 8.
3 HAV was used for deriving the CT values at 5°C. A safety factor of three was used.
4 EPA estimated the CT value by multiplying the initial chlorine concentration (0.5 mg/L) with time, which could lead to
overestimation.
CT values exceeding the EPA values are in bold font.
NA = not applicable, ND = no data available, BDF = buffered, demand-free, EFH = efficiency factor hom.
All units are mg/L-min.
Black et al. (2009) and Cromeans et al. (2010) used the "Efficiency Factor Hom" (EFH) model
to calculate CT for virus inactivation in BDF water. The EFH is a mathematical modeling
method which uses free available chlorine values determined at the beginning, middle and end of
each experiment to extrapolate CT values for viruses that do not achieve 4-log of inactivation in
the time frame being tested. In this approach, a single rate constant of chlorine decay is used to
calculate the integral. However, Keegan et al. (2012) could not use an EFH model because of the
complex decay kinetics of chlorine in wastewater caused by the interaction of chlorine with
ammonia, organic amines and other compounds in the wastewater. Therefore, they calculated the
CT directly by determining the area under the chlorine decay curve of chlorine concentration
versus time.
7.3.5 Information on Virus Inactivation by Chloramines
Systems use monochloramine primarily as a secondary disinfectant to maintain a stable
disinfectant residual in the distribution system and to minimize DBP formation and biofilm
growth. Chloramines are formed by the reaction of ammonia with aqueous chlorine and contain a
mixture of monochloramine, dichloramine and/or trichloramine. The chloramine species
distribution is controlled by pH and the chlorine to ammonia ratio. At pH 6 and above,
monochloramine predominates. Monochloramine has a slow rate of diffusion through the cell
wall; thus, it has been found to be less effective than free chlorine for virus disinfection (Baxter
et al., 2007). In contrast to free chlorine, fewer studies have investigated the disinfection kinetics
of chloramines for viruses systematically. Some researchers have examined chloramine
Six-Year Review 3 Technical Support Document 7-20
for Microbial Contaminant Regulations
December 2016
-------
disinfection alongside chlorine disinfection and found a greater variation in inactivation rates
among different viruses during monochloramine disinfection than during chlorine disinfection
(Hill and Cromeans, 2010). This section describes information from the recent studies on virus
disinfection by monochloramine.
7.3.5.1 Relative Resistance of Viruses to Monochloramine Disinfection
In the same study as discussed in Section 7.3, researchers obtained CT values for
monochloramine disinfection at pH 7 and 8 in BDF water at 5°C, as shown in Exhibit 7.11 (Hill
and Cromeans, 2010; Cromeans et al., 2010). The study viruses each exhibited notable
differences in their responses to monochloramine disinfection, both between and within virus
types. For example, within virus types, differences in monochloramine 2-log CT values were 2-
to 3-fold between CB5 and CB3, 5- to 10-fold between AD2 and AD41 and 110- to 130-fold
between El 1 and El. Overall, El is the most susceptible to monochloramine disinfection while
AD2 and El 1 are the most resistant. At pH 8 and 5°C, CT values for AD2 exceeded the EPA
recommended values for 2-, 3- and 4-log of inactivation by 12 percent to 16 percent.
Exhibit 7.11: CT Values for Virus Inactivation with 1 mg/L of Monochloramine at
5°C
PH
Log
Inactivation
Cromeans et al. (2010)12
USEPA
(1991)
AD23
AD403
AD413
CB34
CB54
E15
E115
MNV6
7
2
600
90
58
270
510
8
1,000
26
857
3
1,000
ND
190
390
710
15
1,300
70
1,423
4
1,500
ND
ND
500
900
42
1,500
150
1,988
8
2
990
360
190
240
670
8
880
36
857
3
1,600
ND
ND
330
900
18
1,200
78
1,423
4
2,300
ND
ND
420
1,100
ND
1,400
170
1,988
1 CT values exceeding the EPA values are in bold font.
2 ND = no data. The CT value could not be extrapolated due to asymptotic inactivation curves.
3 Adenoviruses (AD2, AD40 and AD41)
4 Coxsackie viruses (CB3 and CB5)
5 Echoviruses (E1 and E11)
6 Murine norovirus (MNV)
All units are mg/L-min.
Baxter et al. (2007) examined the inactivation of AD2, AD5 and AD41 by UV light, free
chlorine and monochloramine. They reported a CT value of 350 to achieve 2.5-log inactivation
of AD5 and AD41 by monochloramine at pH 8.5 and 5°C, whereas Cromeans et al. (2010)
reported a CT value of 190 to achieve 2-log inactivation of AD41 at pH 8 and 5°C (see Exhibit
7.10). Baxter et al. (2007) also reported no observed benefit in applying UV light prior to
monochloramine, in terms of enhancing the effectiveness of monochloramine. This was
Six-Year Review 3 Technical Support Document 7-21
for Microbial Contaminant Regulations
December 2016
-------
presumably due to different inactivation mechanisms of UV light (photochemical reaction with
deoxyribonucleic acid (DNA)) and monochloramine (e.g., oxidation).
7.3.5.2 Effect of Virus Aggregation
In Hill and Cromeans' study (2010), for monochloramine disinfection on a river source water,
the CT values for aggregated AD2 were 1.4 times as high as those for dispersed AD2, as shown
in Exhibit 7.12. Sobsey et al. (1991) also reported a 1.4-fold difference between monochloramine
CT values for cell-associated and dispersed HAV (Exhibit 7.6). The difference between CT
values for cell-associated and dispersed viruses using free chlorine was as much as an order of
magnitude, as shown in Exhibit 7.6. Thus, solids association/virus aggregation has a smaller
influence on HAV and AD2 inactivation by monochloramine than on inactivation by free
chlorine. This may be explained by the extended contact time required for monochloramine,
which allows more time for the chemical to permeate the aggregate.
Exhibit 7.12: CT Values for Monochloramine Inactivation of Aggregated and
Dispersed AD2 in River Source Water at 5°C
Log
Hill and Cromeans (2010)
Ratio of CT for
Inactivation
Aggregated AD2
Dispersed AD2
Aggregated vs.
Dispersed AD2
2
2,500
1,800
1.4
3
3,700
2,700
1.4
4
4,700
3,600
1.3
AD2 = Adenovirus
All units are mg/L-min.
7.3.5.3 CT Values for Adenoviruses
Although AD2 is one of the most susceptible viruses to chlorine disinfection, it is one of the
most resistant viruses to monochloramine disinfection (Hill and Cromeans, 2010). Sirikanchana
et al. (2008) performed experiments with a 0.01M buffer (phosphate or borate) solution to
investigate the effect of pH, temperature, monochloramine concentration and ammonia-nitrogen
-to-chlorine molar ratio on the inactivation kinetics of AD2 with monochloramine. Sirikanchana
et al. (2008) found the inactivation kinetics to be independent of monochloramine concentration
and ammonia-nitrogen-to-chlorine molar ratio but to have strong pH dependence, with the rate of
inactivation decreasing with increasing pH. The kinetics at pH 6 and 8 were consistent with
pseudo-first-order kinetics14, while curves at pH 10 were characterized by a lag phase followed
by a pseudo-first-order phase. The rate of inactivation also increased with increasing
temperature. The results of this study indicate that monochloramine disinfection might not
provide adequate control of adenoviruses in drinking water at high pH and low temperature.
14 If the concentration of one reactant is in great excess, it can be approximated as a pseudo-first-order reaction (treating the
reactant in excess concentration—in this case monochloramine—as a constant).
Six-Year Review 3 Technical Support Document 7-22
for Microbial Contaminant Regulations
December 2016
-------
Inactivation kinetics and pH data for AD2 in Sirikanchana et al. (2008) were also consistent with
Hill and Cromeans (2010).
Because recycled water often contains ammonia and preformed chloramines are rapidly formed
on addition of chlorine, the EPA chloramines CT values are not applicable to recycled water.
Therefore, Keegan et al. (2012) performed disinfection experiments on AD2 at various
turbidities and pH values to develop CT values for recycled water (Keegan et al., 2012; ADOH,
2013). Exhibit 7.13 presents the results of their study and the corresponding CT values in the
Australian guidelines for recycled water (ADOH, 2013). No safety factors were used to derive
the CT values in the ADOH guidelines. Their CT values are significantly higher than the EPA
values and those obtained by Sirikanchana et al. (2008), presumably due to the high turbidity and
higher chloramine dosage used in the Keegan et al. (2012) study.
Exhibit 7.13: CT Values for Inactivation of AD2 by Chloramines in Recycled Water
at 10°C
Log
Keegan et al. (2012)1
ADOH (2013)2 4
USEPA
PH
Inactivation
Turbidity (NTU)
Turbidity (NTU)
(1991 )3
2
5
<=2
<=5
1
969
1,204
977
1,201
NA
7
2
1,688
1,903
1,681
1,914
857
3
2,393
2,638
2,386
2,628
1,423
4
3,082
3,337
3,090
3,341
1,988
1
1,482
1,590
1,494
1,596
NA
8
2
2,326
2,546
2,318
2,541
857
3
3,160
3,490
3,141
3,486
1,423
4
3,949
4,426
3,965
4,431
1,988
1
2,992
4,364
3,154
4,400
NA
9
2
4,592
6,032
4,393
5,967
857
3
5,716
7,511
5,631
7,535
1,423
4
6,746
9,096
6,870
9,102
1,988
1 Chloramine dosage of 15 mg/L was used in the tests. CT values for 20 NTU turbidity tests are not shown
because they were not used in the ADOH guideline.
2 No safety factor was used.
3 HAV was used for deriving the CT values at 5°C. No safety factor was used.
4 CT values exceeding the EPA values are in bold font.
NA = not available.
All units are mg/L-min.
Six-Year Review 3 Technical Support Document 7-23
for Microbial Contaminant Regulations
December 2016
-------
8 References
Abbaszadegan, M., T. Rauch-Williams, W. P. Johnson, and S. A. Hubbs. 2011. Methods to
assess GWUDI and bank filtration performance. Water Research Foundation, Denver, CO.
American Public Health Association (APHA), American Water Works Association (AWW A),
and Water Environment Federation. 2012. Standard Methods for the Examination of Water and
Wastewater, 22nd Edition.
AWWA. 2015. Disinfection Residual Strategy Panel Memorandum: Panel Findings. July 31,
2015.
Armstrong, T.W. and C.N. Haas. 2007. A quantitative microbial risk assessment model for
legionnaires' disease: assessment of human exposures for selected spa outbreaks. Journal of
Occupational and Environmental Hygiene. 4: 634-646.
Armstrong, T.W. and C. N. Haas. 2008. Legionnaires" disease: evaluation of a quantitative
microbial risk assessment model. Journal of Water and Health. 6(2): 149-166.
Ashbolt, N.J. 2015. Environmental (saprozoic) pathogens of engineered water systems:
understanding their ecology for risk assessment and management. Pathogens. 4(2): 390-405.
Australian Department of Health (ADOH). 2013. Guidelines for Validating Treatment Processes
for Pathogen Reduction - Supporting Class A Recycled Water Schemes in Victoria. Department
of Health, Victoria.
Baribeau, H., S.W. Krasner, R. Chinn, and P C. Singer. 2005. Impact of biomass on the stability
of HA As and THMs in a simulated distribution system. Journal of the AWWA. 97(2): 69-81.
Bartram, J., J. Cotruvo, M. Exner, C. Fricker, and A. Glasmacher. 2003. Expert Consensus,
Expert Meeting Group Report. In Heterotrophic Plate Counts and Drinking-Water Safety, The
Significance of HPCs for Water Quality, IWA Publishing, London, p. 1-11.
Bartrand, T. A., J.J. Causey, and J. L. Clancy. 2014. Naegleriafowleri: an emerging drinking
water pathogen. Journal of the AWWA. 106 (10): E418-E432.
Barwick, R. S., D A. Levy, G.F. Craun, M.J. Beach, and R.L. Calderon. 2000. Surveillance for
waterborne-disease outbreaks—United States, 1997-1998. MMWR CDC Surveillance
Summary, 49(4): 1-21.
Baxter, C.S., R. Hofmann, M.R. Templeton, M. Brown, and R.C. Andrews. 2007. Inactivation of
adenovirus types 2, 5, and 41 in drinking water by UV light, free chlorine, and monochloramine.
Journal of Environmental Engineering. 133 (1): 95-103.
Behnke, S., A.E. Parker, D. Woodall, and A.K. Camper. 2011. Comparing the chlorine
disinfection of detached biofilm clusters with those of sessile biofilms and planktonic cells in
single- and dual-species cultures. Applied and Environmental Microbiology. 77(20): 7176-7184.
Six-Year Review 3 Technical Support Document 8-1
for Microbial Contaminant Regulations
December 2016
-------
Bergmire-Sweat, D., J. Morgan, K. Wilson, K. Von Alt, L. Marengo, T. Bennett, Y. M. Lee, V.
C. Tsang, W. R. MacKenzie, and B. Furness. 1999. Cryptosporidiosis at Brushy Creek:
describing the epidemiology and causes of a large outbreak in Texas, 1998. In Proceedings of the
International Conference on Emerging Infectious Diseases, Milwaukee, Wisconsin, AWWA,
Denver CO.
Biyela, P.T., R. Hodon, A. Brown, A. Alum, M. Abbaszadegan, and B.E. Rittmann. 2012.
Distribution systems as reservoirs of Naegleria fowleri and other amoebae. Journal of the
AWWA. 104(1): E66-E72.
Black, S., J.A. Thurston, and C.P. Gerba. 2009. Determination of CT values for chlorination of
resistant enteroviruses. Journal of Environmental Sciences and Health. 44(4): 336-339.
Borchardt, M. A., S. K. Spencer, B. A. Kieke Jr., E. Lambertini, and F. J. Loge. 2012. Viruses in
non-disinfected drinking water from municipal wells and community incidence of acute
gastrointestinal illness. Environmental Health Perspectives. 120(9): 1272-1279.
Borchardt, M.A., J.P. Stokdyk, A.C. Anderson, L.W. Rezania, S.K. Spencer, A.D. Firnstahl, and
T.J. Robinson. 2015. Human enteric pathogens and bovine fecal markers in non-disinfected
drinking water from community and non-community wells in Minnesota, poster abstract,
American Society for Microbiology, 115th General Meeting, New Orleans, LA.
Bradford, S.A., H. Kim, B. Headd, and S. Torkzaban. 2016. Evaluating the transport of Bacillus
subtihs spores as a potential surrogate for Cryptosporidiumparvum oocysts. Environmental
Science & Technology. 50(3): 1295-1303 (accepted for publication in 2015).
Burton P, L. Gurrin L, and P. Sly. 1998. Extending the simple linear regression model to account
for correlated responses: an introduction to generalized estimating equations and multi-level
mixed modelling. Statistics in Medicine. 17(11): 1261-91.
Buse, H.Y., M.E. Schoen, and N.J. Ashbolt. 2012. Legionella in engineered systems and use of
quantitative microbial risk assessment to predict exposure. Water Research. 46(4): 921-933.
California Department of Health Services, 1993. Surface Water Treatment Rule Demonstration
Study - Sonoma County Water Agency (SCWA). Letter from Bruce H. Burton (District
Engineer) to SCWA accompanied by a final report on Russian River Demonstration Study for
SCWA prepared by CH2M Hill. September 22.
Carriere, I and Bouyer, J. 2002. Choosing marginal or random-effects models for longitudinal
binary responses: Application to self-reported disability among older persons. MBC Medical
Research Methodology. 2(15).
Centers for Disease Control and Prevention (CDC). 2006. Surveillance for waterborne disease
outbreaks and other health events associated with recreational water - United States, 2003-2004
and Surveillance for waterborne disease outbreaks associated with drinking water and water not
intended for drinking - United States, 2003-2004. Surveillance Summaries: Morbidity and
Mortality Weekly Report. 55(SS-12).
Six-Year Review 3 Technical Support Document 8-2
for Microbial Contaminant Regulations
December 2016
-------
CDC. 2008. Surveillance for waterborne disease outbreaks and other health events associated
with recreational water and other aquatic facility-associated health events - United States, 2005-
2006 and surveillance for waterborne disease outbreaks associated with drinking water and water
not intended for drinking - United States, 2005-2006. Surveillance Summaries: Morbidity and
Mortality Weekly Report. 57(SS-9).
CDC. 2011. Surveillance for waterborne disease outbreaks and other health events associated
with recreational water - United States, 2007-2008 and surveillance for waterborne disease
outbreaks associated with drinking water - United States, 2007-2008. Surveillance Summaries:
Morbidity and Mortality Weekly Report. 60(12).
CDC. 2013. Surveillance for waterborne disease outbreaks associated with drinking water and
other nonrecreational water - United States, 2009-2010. Morbidity and Mortality Weekly
Report. 62(35): 714-720.
CDC. 2014. Summary of notifiable diseases - United States, 2012. Morbidity and Mortality
Weekly Report. 61(53): 105-107.
CDC. 2015a. Surveillance for waterborne disease outbreaks associated with drinking water -
United States, 2011-2012. Morbidity and Mortality Weekly Report. 64(31): 842-848.
CDC. 2015b. Active bacterial core surveillance for Legionellosis - United States, 2011-2012.
Morbidity and Mortality Weekly Report. 64(42): 1190-1193.
Chalmers, R. M. 2012. Waterborne outbreaks of cryptosporidiosis. Annali dell'lstituto superiore
di sanita. 48(4):429-446.
Chaudhary, K., B. Scanlon, N. Scheffer, and S. Walden. 2009. Review of the State of Art:
Ground Water Under the Direct Influence of Surface Water Programs. University of Texas at
Austin, Austin, TX (Available at: http://www.beg.utexas.edu/sce/docs/BEG GWUDI 2009.pdf).
Cohn, P.D., J.A. Gleason, E. Rudowski, S.M. Tsai, C.A. Genese, and J.A. Fagliano. 2014.
Community outbreak of legionellosis and an environmental investigation into a community
water system. Epidemiology and Infection. 143(6): 1322-1331.
Colford, Jr., J. M., J. F. Hilton, C. C. Wright, B. F. Arnold, S. Saha, T. J. Wade, J. Scott, and J.
N. S. Eisenberg. 2009. The Sonoma water evaluation trial: A randomized drinking water
intervention trial to reduce gastrointestinal illness in older adults. American Journal of Public
Health. 99(11): 1988-1995.
Collier, S. A., L.J. Stockman, L. A. Hicks, L.E. Garrison, F.J. Zhou, and M.J. Beach. 2012.
Direct healthcare costs of selected diseases primarily or partially transmitted by water.
Epidemiology and infection. 140(11): 2003-2013.
Cope, J.R., R.C. Ratard, V.R. Hill, T. Sokol, J.J. Causey, J.S. Yoder, G. Mirani, B. Mull, K.A.
Mukeijee, J. Narayanan, M. Doucet, Y. Qvarnstrom, C.N. Poole, O.A. Akingbola, J.M. Ritter, Z.
Xiong, A.J. da Silva, D. Roellig, R.B. Van Dyke, H. Stern, L. Xiao, and M.J. Beach. 2015. The
Six-Year Review 3 Technical Support Document 8-3
for Microbial Contaminant Regulations
December 2016
-------
first association of a primary amebic meningoencephalitis death with culturable Naegleria
fowleri in tap water from a U.S. treated public drinking water system. Clinical Infectious
Diseases. 60(8): e36-e42.
Craun, G.F., R.L. Calderon, and T.J. Wade. 2006. Assessing waterborne risk: an introduction.
Journal of Water and Health. 4 (Suppl. 2): 3-18.
Craun, G.F., J.M. Brunkard, J.S. Yoder, V.A. Roberts, J. Carpenter, T. Wade, R.L. Calderon,
J.M. Roberts, M.J. Beach, and S.L. Roy. 2010. Causes of outbreaks associated with drinking
water in the United States from 1971 to 2006. Clinical Microbiology Reviews. 23(3): 507-528.
Craun, G.F. and W. Jakubowski. 1986. Status of waterborne giardiasis outbreaks and monitoring
methods. Health Effects Research Laboratory, Office of Research and Development, U.S.
Environmental Protection Agency.
Cromeans, T.L., A.M. Kahler, and V.R. Hill. 2010. Inactivation of adenoviruses, enteroviruses,
and murine norovirus in water by free chlorine and monochloramine. Applied and
Environmental Microbiology. 76(4): 1028-1033.
Daly, E. R., S. J. Roy, D. D. Blaney, J. S. Manning, V. R. Hill, L. Xiao, and J. W. Stull. 2010.
Outbreak of giardiasis associated with a community drinking-water source. Epidemiology and
Infection. 138:491-500.
Duda, S., S. Kandiah, J.E. Stout, J.L. Baron, M. Yassin, M. Fabrizio, J. Ferrelli, R. Hariri, M.M.
Wagener, J. Goepfert, J. Bond, J. Hannigan, and D. Rogers. 2014. Evaluation of a new
monochloramine generation system for controlling Legionella in building hot water systems.
Infection Control and Hospital Epidemiology. 35(11): 1356-1363.
Dworkin, M. S., D. P. Goldman, T. G. Wells, J. M. Kobayashi, and B. L. Herwaldt. 1996.
Cryptosporidiosis in Washington State: An outbreak associated with well water. The Journal of
Infectious Diseases. 174: 1372-1376.
Ercumen, A., J.S. Gruber, and J.M. Colford Jr. 2014. Water distribution system deficiencies and
gastrointestinal illness: a systematic review and meta-analysis. Environmental Health
Perspectives. 122(7): 651.
Fiischslin, H. P., S. Kotzsch, and T. Egli. 2012. Cryptosporidium spp. in drinking water. Swiss
Medical Weekly. 142: wl3683.
Flannery, B., L.B. Gelling, D.J. Vugia, J.M. Weintraub, J.J. Salerno, M.J. Conroy, V.A. Stevens,
C.E. Rose, M.R. Moore, B.S. Fields, and R.E. Besser. 2006. Reducing Legionella colonization of
water systems with monochloramine. Emerging Infectious Diseases. 12(4): 588-596.
Francy, D.S., R.N. Bushon, J. Stopar, E.J. Luzano, and G.S. Fout. 2004. Environmental factors
and chemical and microbiological water-quality constituents related to the presence of enteric
viruses in ground water from small public water supplies in Southeastern Michigan. USGS
Scientific Investigations Report 2004-5219.
Six-Year Review 3 Technical Support Document 8-4
for Microbial Contaminant Regulations
December 2016
-------
Geldreich, E.E. 1992. Waterborne pathogen invasions: a case for water quality protection in
distribution. In Proceedings of AWWA Water Quality Technology Conference, pp. 1-18.
Gollnitz, W.D., J.L. Clancy, J.B. McEwen, and S.C. Garner. 2005. Riverbank filtration for
IESWTR compliance. Journal of the AWWA. 97(12):64-76.
Headd, B. and S.A. Bradford. 2015 (epub). Use of aerobic spores as a surrogate for
Cryptosporidium oocysts in drinking water supplies. Water Research. 90: 185-202. Available at:
http ://www. sciencedirect.com/sci ence/arti cl e/pi i/S0043135415304218.
Heffelfinger, J.D., J.L. Kool, S. Fridkin, V.J. Fraser, J. Hageman, J. Carpenter, and C.G.
Whitney. 2003. Risk of hospital-acquired Legionnaires' disease in cities using monochloramine
versus other water disinfectants. Infection Control and Hospital Epidemiology. 24(8): 569-574.
Hibler, C.P., C.M. Hancock, L.M. Perger, J.G. Wegrzyn, andK.D. Swabby. 1987. Inactivation of
Giardia cysts with chlorine at 0.5 °C to 5.0 °C. AWWA Research Foundation.
Hill, V.R. and T.L. Cromeans. 2010. Contaminant Candidate List Viruses: Evaluation of
Disinfection Efficacy. Water Research Foundation. Project #3134.
Hoffbuhr, J.W., J. Blair, M. Bartleson, and R. Karlin. 1986. Use of particulate analysis for source
and water treatment evaluation. Water Quality Technology Conference 1986 Conference
Proceedings.
Hrudey, S. E. and E. J. Hrudey. 2014. Ensuring Safe Drinking Water: Lessons from Frontline
Experience with Contamination. AWWA.
Hubbard AE, J. Ahern, N. L. Fleischer, M. Van der Laan, S. A. Lippman, N. Jewell, T.
Bruckner, and W. A. Satariano. 2010. To GEE or not to GEE: comparing population average and
mixed models for estimating the associations between neighborhood risk factors and health.
Epidemiology. 21(4): 467-474.
Hunter, P.R. 2003. Epidemiological and risk assessment evidence of disease linked to HPC
bacteria. Heterotrophic plate counts and drinking-water safety. IWA Publishing, London,
pp.119-136.
Ingels, T. 2015. What does undetectable mean anyway? Colorado's experience with detection of
free chlorine residuals in real world distribution systems. Presentation at the AWWA Annual
Conference.
Jacangelo, J. J., M. Watson, and N. E. Seith. 2001. Investigation of criteria for GWUDI
determination. AWWA Research Foundation and AWWA, Denver, CO.
Kahler, A.M., T.L. Cromeans, J.M. Roberts, and V.R. Hill. 2010. Effect of Source Water Quality
on Chlorine Inactivation of Adenovirus, Coxsackievirus, Echovirus, and Murine Norovirus.
Applied Environmental Microbiology. 76(15): 5159-5164.
Six-Year Review 3 Technical Support Document 8-5
for Microbial Contaminant Regulations
December 2016
-------
Khaldi, S., Ratajczak, M., Gargala, G., Fournier, M., Berthe, T., Favennec, L., and J.P. Dupont.
2011. Intensive exploitation of a karst aquifer leads to Cryptosporidium water supply
contamination, Water Research, 45(9): 2906-2914.
Keegan, A., S. Wati, andB. Robinson. 2012. Chlor(am)ine disinfection of human pathogenic
viruses in recycled waters (SWF62M-2114). Prepared by Australia Water Quality Centre for the
Smart Water Fund, Victoria.
Kirmeyer, G.J, M. LeChevallier, H. Baribeau, K. Martel, G. Thompson, L. Radder, W. Klement,
and A. Flores. 2004. Optimizing Chloramine Treatment - Second Edition. Denver, CO. Water
Research Foundation.
Kool, J.L., J.C. Carpenter, and B.S. Fields. 1999. Effect of monochloramine disinfection of
municipal drinking water on risk of nosocomial Legionnaires' disease. The Lancet. 353: 272-
277.
Kramer, M. H., B.L. Herwaldt, G.F. Craun, R.L. Calderon, and D.D. Juranek. 1996. Surveillance
for waterborne-disease outbreaks: United States, 1993-1994. U.S. Department of Health and
Human Services, Public Health Service, Centers for Disease Control and Prevention.
Krentz, C. A., K. Teschke, H. Shen, and J. Issac-Renton. 2013. The predictive value of total
coliforms in drinking water using life table analysis. Journal of Water Supply: Research and
Technology-AQUA, 62(2): 97-106.
LeChevallier, M.W., N.J. Welch, and D.B. Smith. 1996. Full-scale studies of factors related to
coliform regrowth in drinking water. Applied and Environmental Microbiology. 62(7): 2201 -
2211.
Lee, Y., P. W. Johnson, J. L. Call, M. J. Arrowood, B. W. Furness, S. C. Pichette, K. K. Grady,
P. Reeh, L. Mitchell, D. Bergmire-Sweat, W. R. MacKenzie, and V. C. W. Tsang. 2001.
Development and application of a quantitative, specific assay for Cryptosporidium parvum
oocyst detection in high-turbidity environmental water samples. The American Journal of
Tropical Medicine and Hygiene. 65(1): 1-9.
Lee, S.H., D A. Levy, G.F. Craun, M.J. Beach, and R.L. Calderon. 2002. Surveillance for
waterborne disease outbreaks—United States, 1999-2000. Morbidity and Mortality Weekly
Report 51, no. SS-08: I-28. (http://vvvvvv.cdc.gov/mmvvr/previevv/mmvvrhtml/ss5 108a I htm).
Levy, D.A., M.S. Bens, G.F. Craun, R.L. Calderon, and B.L. Herwaldt. 1998.Surveillance for
waterborne disease outbreaks—United States, 1995-1996. Morbidity and Mortality Weekly
Report 47, no. SS-05: 1 -34. (http://vvvvvv.cdc.gov/mmvvr/previevv/mmvvrhtml/00055820.htm)
Lin, Y.E., R.D. Vidic, J.E. Stout, and V.L. Yu. 1998. Legionella in water distribution systems.
Journal of the AWWA. 90(9): 112-122.
Six-Year Review 3 Technical Support Document 8-6
for Microbial Contaminant Regulations
December 2016
-------
Lin, X., J. McKinley, C.T. Resch, R. Kaluzny, C.L. Lauber, J. Fredrickson, R. Knight, and A.
Konopka. 2012. Spatial and temporal dynamics of the microbial community in the Hanford
unconfined aquifer. The ISME journal. 6(9): 1665-1676.
Louisiana Department of Health and Hospitals. 2013. Declaration of Emergency. Minimum
Disinfectant Residual Levels in Public Water Systems.
(http://dhh.louisiana.gov/assets/oph/Center-EH/engineering/Emergencv Rule/
Em ergencyRul e_N oti fi cati on. p df.)
Louisiana Department of Health and Hospitals. 2014. Declaration of Emergency. Minimum
Disinfectant Residual Levels in Public Water Systems. Available online at:
http://www.dhh.louisiana.gov/assets/oph/Center-EH/engineering/Emergency_Rule/ER_3-6-
14.pdf.
Lunn, D., D. Spiegelhalter, A. Thomas, and N. Best. 2009. The BUGS project: Evolution,
critique and future directions (with discussion), Statistics in Medicine. 28(25): 3049-3082.
Ma, J., L. Thabane, J. Kaczorowski, L. Chambers, L. Dolovich, T. Karwalajtys, and C. Levitt.
2009. Comparison of Bayesian and classical methods in the analysis of cluster randomized
controlled trials with a binary outcome: The Community Hypertension Assessment Trial
(CHAT). BMC Medical Research Methodology. 9:37.
McKinley, R. E. 1997. Hydro-Nine Cryptosporidium Outbreak Walla Walla Washington. In
Under the Microscope: Examining Microbes in Ground water: The 1996 Ground water
Foundation Symposium Proceedings, Sept. 5 & 6, 1996, Boston, MA. p. 9-20. AWWA Research
Foundation and AWWA, Denver, CO.
MessnerM.J., P. Berger, and S.P. Nappier. 2014. Fractional poisson: A simple dose-response
model for human norovirus. Risk Analysis. 34(10): 1820-9.
Messner, M.J. and P. Berger. 2016. Cryptosporidium Infection Risk: Results of New Dose-
Response Modeling. Risk Analysis. 36(10): 1969-1982 (accepted in December 2015).
Moore, A. C., B.L. Herwaldt, G.F. Craun, R.L. Calderon, A.K. Highsmith, and D.D. Juranek.
1993. "Surveillance for waterborne disease outbreaks—United States, 1991-1992." MMWR
Surveillance Summary. 42(5): 1-22.
Moore, M.R., M. Pryor, B. Fields, C. Lucas, M. Phelan, and R.E. Besser. 2006. Introduction of
monochloramine into a municipal water system: Impact on colonization of buildings by
Legionella spp. Applied and Environmental Microbiology. 72(1): 378-383.
Murphy, H. M., K. D. M. Pintar, E. A. McBean, and M. K. Thomas. 2014. A systematic review
of waterborne disease burden methodologies from developed countries. Journal of Water and
Health. 12(4): 634-655.
National Drinking Water Advisory Council (NDWAC). 2000. Working Group Meeting on
Contaminant Candidate List Regulatory Determinations and the 6-Year Review of Existing
Six-Year Review 3 Technical Support Document 8-7
for Microbial Contaminant Regulations
December 2016
-------
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.
National Research Council (NRC). 2006. Committee on Public Water Supply Distribution
Systems: Assessing and Reducing Risks. Drinking Water Distribution Systems: Assessing and
Reducing Risks. National Academies Press.
Page, M.A., J.L. Shisler, and B.J. Marinas. 2009. Kinetics of adenovirus type 2 inactivation with
free chlorine. Water Research. 43(11): 2916-2926.
Park, T. 1993. A comparison of the generalized estimating equation approach with the maximum
likelihood approach for repeated measurements. Statistics in Medicine. 12(18): 1723-1732.
Pennsylvania Bulletin. 2016. Proposed Rulemaking. Disinfection Requirements Rule. 46(8):
857-892. Available online at: http://www.pabulletin.com/secure/data/vol46/46-8/278.html.
Pitkanen, T., T. Juselius, E. Isomaki, I.T. Miettinen, M. Valve, A-L. Kivimaki, K. Lahti, and M-
L. Hanninen. 2015. Drinking water quality and occurrence of Giardia in Finnish small
groundwater supplies. Resources 4(3): 637-654.
Price, M. L., J. Flugum, P. Jeane, and L. Tribbett-Peelen. 1999. Sonoma County finds ground
water under the direct influence of surface water depends on river conditions (Abstract: 25-27).
International Riverbank Filtration Conference, November 4-6, 1999, Louisville, KY.
Pryor, M., S. Springthorpe, S. Riffard, T. Brooks, Y. Huo, G. Davis, and S.A. Sattar. 2004.
Investigation of opportunistic pathogens in municipal drinking water under different supply and
treatment regimes. Water Science and Technology. 50(1): 83-90.
Regli, S., P. Berger, and T. Grubbs. 2003. Control of Drinking Water Pathogens and Disinfection
Byproducts. Ch. 12 in Drinking Water Regulation and Health, edited by Frederick W. Pontius.
Hoboken, NJ: Wiley-Interscience.
Revetta, R.P., V. Gomez-Alvarez, T.L. Gerke, C. Curioso, J.W. Santo Domingo, and N.J.
Ashbolt. 2013. Establishment and early succession of bacterial communities in monochloramine-
treated drinking water biofilms. FEMS Microbiology Ecology. 86(3): 404-414.
Rusin, P. A., J.B. Rose, C.N. Haas, and C.P. Gerba, 1997. Risk assessment of opportunistic
bacterial pathogens in drinking water. Reviews of Environmental Contamination and
Toxicology. 152: 57-83. Springer New York.
Safe Drinking Water Act (SDWA). 2002. Title XIV of the Public Health Service Act: Safety of
Public Water Systems (Safe Drinking Water Act). Title 14. §§ 1400-1465.
Sinclair, M.I., M.E. Hellard, R. Wolfe, T.Z. Mitakakis, K. Leder, and C.K. Fairley. 2005.
Pathogens causing community gastroenteritis in Australia. Journal of Gastroenterology and
Hepatology. 20(11): 1685-1690.
Six-Year Review 3 Technical Support Document 8-8
for Microbial Contaminant Regulations
December 2016
-------
Sinclair, M., O'Toole, J., Gibney, K. and Leder, K. 2015. Evolution of regulatory targets for
drinking water quality. Journal of Water and Health 13(2):413-426.
Sirikanchana, K., J.L. Shisler, and B.J. Marinas. 2008. Inactivation kinetics of adenovirus
serotype 2 with monochloramine. Water Research. 42(6): 1467-1474.
Smeets, P.W., G.J. Medema, and J.C. Van Dijk. 2009. The Dutch secret: how to provide safe
drinking water without chlorine in the Netherlands. Drinking Water Engineering and Science. 2:
1-14.
Sobsey, M.D., T. Fuji, and P. A. Schields. 1988. Inactivation of hepatitis A virus and model
viruses in water by free chlorine and monochloramine. Water Science and Technology. 20 (11-
12): 385-391.
Sobsey, M.D., T. Fuji, and R.M. Hall. 1991. Inactivation of cell associated and dispersed
hepatitis A virus in water. AWWA Research and Technology, p 64-67.
Solo-Gabriele, H. and S. Neumeister. 1996. U.S. outbreaks of Cryptosporidiosis, Journal of the
AWWA. 88(9):76-86.
Sonoma County Water Agency (SCWA). 2013. Russian River Collector Multi-Year Operational
Analysis Radial Collector Wells 1 through 6.
Stanford, J. A. and Ward, J.V., 1992. Management of aquatic resources in large catchments:
recognizing interactions between ecosystem connectivity and environmental disturbance. In
Watershed Management, pp. 91-124. Springer, New York.
Stanford, J.A. and Ward, J.V., 1993. An ecosystem perspective of alluvial rivers: connectivity
and the hyporheic corridor. Journal of the North American Benthological Society. 12(1): 48-60.
State of Nebraska. 2013. Demonstration of Performance to Establish Natural Filtration Credits
for the City of Kearney, Nebraska. Platte River Well Field. Project #130-Cl-054.
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
AWWA. 105(8): 432-448.
Strollo, S.E., J. Adjemian, M.K. Adjemian, and D.R. Prevots. 2015. The burden of pulmonary
nontuberculous mycobacterial disease in the United States. Annals of the American Thoracic
Society. 12(10): 1458-1464.
Templeton, M., R.C. Andrews, and R. Hofmann. 2008. Particle-associated viruses in water:
Impacts on disinfection processes. Critical Reviews in Environmental Science and Technology.
38(3): 137-164.
Teunis, P. F., C.L. Chappell, and P.C. Okhuysen. 2002a. Cryptosporidium dose response studies:
variation between isolates. Risk Analysis. 22(1): 175-185.
Six-Year Review 3 Technical Support Document 8-9
for Microbial Contaminant Regulations
December 2016
-------
Teunis, P.F., C.L. Chappell, andP.C. Okhuysen. 2002b. Cryptosporidium dose response studies:
variation between hosts. Risk Analysis. 22(3): 475-485.
Thall, P.F. and S.C. Vail. 1990. Some covariance models for longitudinal count data with
overdispersion. Biometrics. 46(3): 657-671.
Thurston-Enriquez, J.A., C.N. Haas, J. Jacangelo, and C.P. Gerba. 2003. Chlorine inactivation of
adenovirus type 40 and feline calicivirus. Applied Environmental Microbiology. 69(1): 3979-
3985.
United States Environmental Protection Agency (USEPA). 1976. National Interim Primary
Drinking Water Regulations. Office of Water Supply. EPA-570/9-76-03.
USEPA. 1985. National Primary Drinking Water Regulations; Synthetic Organic Chemicals,
Inorganic Chemicals and Microorganisms; Proposed Rulemaking. 50 FR 46936. November 13,
1985.
USEPA. 1987. National Primary Drinking Water Regulations; Filtration and Disinfection;
Turbidity, Giardia lamblia, Viruses, Legionella, and Heterotrophic Bacteria; Proposed Rule. 52
FR 42178. November 3, 1987.
USEPA. 1988. National Primary Drinking Water Regulations; Filtration and Disinfection;
Turbidity, Giardia lamblia, Viruses, Legionella, and Heterotrophic Bacteria; Total Coliforms.
Notice of availability; close of public comment period. 53 FR 16348. May 6, 1988.
USEPA. 1989. National Primary Drinking Water Regulations; Filtration, Disinfection, Turbidity,
Giardia lamblia, Viruses, Legionella, and Heterotrophic Bacteria; Final Rule. Part III. 54 FR
27486. June 29, 1989. Available at:
http://water.epa.gov/lawsregs/rulesregs/sdwa/swtr/upload/SWTR.pdf. Accessed May 28, 2015.
USEPA. 1991a. Safe Drinking Water Act as Amended by the Safe Drinking Water Act
Amendments of 1986. Public Law 99-339. EPA Office of Water. EPA 570/B-91-042. Available
at: http://nepis.epa.gov/Exe/ZvPDF.cgi/60000Z6Y.PDF?Dockev=60000Z6Y.PDF. Accessed
July 22, 2015.
USEPA. 1991b. Guidance Manual for Compliance with the Filtration and Disinfection
Requirements for Public Water Systems Using Surface Water Sources. March 1991. Available
at: https://vvvvvv.epa.gov/sites/production/files/2015-10/documents/
guidance_manual_for_compliance_with_the_filtration_and_disinfection_requirements.pdf.
USEPA. 1992a. Draft Ground Water Disinfection Rule Available for Public Comment. Notice of
availability and review. 57 FR 33960. July 31, 1992.
USEPA. 1992b. Consensus Method for Determining Ground Waters Under the Direct Influence
of Surface Water Using Microscopic Particulate Analysis (MPA). EPA 910/9-92-029.
Six-Year Review 3 Technical Support Document 8-10
for Microbial Contaminant Regulations
December 2016
-------
USEPA. 1998a. National Primary Drinking Water Regulations: Disinfectants and Disinfection
Byproducts; Final Rule. 63 FR 69390. December 16, 1998.
USEPA. 1998b. National Primary Drinking Water Regulations: Interim Enhanced Surface Water
Treatment; Final Rule. 63 FR 69478. December 16, 1998.
USEPA. 2000. National Primary Drinking Water Regulations: Ground Water Rule; Proposed
Rule. 65 FR 30194. May 10, 2000.
USEPA. 2001. National Primary Drinking Water Regulations: Filter Backwash Recycling Rule;
Final Rule. 66 FR 31086. June 8, 2001.
USEPA. 2002. National Primary Drinking Water Regulations: Long-Term 1 Enhanced Surface
Water Treatment Rule; Final Rule. 67 FR 1812. January 14, 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. Long-Term 1 Enhanced Surface Water Treatment Rule Disinfection Profiling
and Benchmarking Technical Guidance Manual. EPA 816-R-03-004. Office of Water,
Washington, DC.
USEPA. 2005. Drinking Water Contaminant Candidate List 2; Final Notice. 70 FR 9071.
February 24.
USEPA. 2006a. National Primary Drinking Water Regulations: Long-Term 2 Enhanced Surface
Water Treatment Rule; Final Rule. 71 FR 654. January 5, 2006.
USEPA. 2006b. National Primary Drinking Water Regulations: Ground Water Rule; Final Rule.
71 FR 65574. November 8, 2006.
USEPA. 2006c. Occurrence and Monitoring Document for the Final Ground Water Rule. EPA
815-R-06-012. October 2006. Available online at:
https://nepis. epa.gov/Exe/ZyPDF. cgi?Dockey=P 1002XR7.txt.
USEPA. 2007a. Establishment of the Total Coliform Rule Distribution System Advisory
Committee. 72 FR 35870. June 29, 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. 2008. Total Coliform Rule/Distribution System Federal Advisory Committee.
Agreement in Principle. Available online at: https://www.epa.gov/sites/production/files/2015-
10/documents/total_coliform_rule_distribution_system_advisory_committee_agreement_in_prin
ciple_pdf.pdf.
USEPA. 2009a. Drinking Water Contaminant Candidate List 3; Final Notice. 74 FR 51850.
October 2009.
Six-Year Review 3 Technical Support Document 8-11
for Microbial Contaminant Regulations
December 2016
-------
USEPA. 2009b. Final CCL 3 Microbes: PCCL to CCL Process. EPA 815-R-09-009. Office of
Water, Washington, DC.
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. Information Collection Request, Contaminant Occurrence Data in Support of
EPA's Third Six-Year Review of National Primary Drinking Water Regulations. EPA ICR No.
2231.02, OMB Control No. 2040-0275. 75 FR 6023. February 5, 2010.
USEPA. 2010c. Long-Term 2 Enhanced Surface Water Treatment Rule Toolbox Guidance
Manual. EPA 815-R-09-016. April 2010.
USEPA. 2012. Economic Analysis for the Revised Total Coliform Rule. EPA 815-R-12-004.
USEPA. 2013a. National Primary Drinking Water Regulations: Revisions to the Total Coliform
Rule; Final Rule. 78 FR 10269. February 13, 2013.
USEPA. 2016a. Six-Year Review 3 Technical Support Document for Long-Term 2 Enhanced
Surface Water Treatment Rule. EPA-810-R-16-011.
USEPA. 2016b. EPA Protocol for the Third Review of Existing National Primary Drinking
Water Regulations. EPA-810-R-16-007
USEPA. 2016c. Third Six-Year Review (SYR3) Information Collection Rule (ICR) Database.
USEPA. 2016d. Occurrence Data for the Unregulated Contaminant Monitoring Rule. July 2016.
USEPA. 2016e. The Data Management and Quality Assurance/Quality Control Process for the
Third Six-Year Review Information Collection Rule Dataset. EPA-810-R-16-015.
USEPA. 2016f. Six-Year Review 3 Technical Support Document for Disinfectants/Disinfection
Byproducts Rules. EPA-810-R-16-012.
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.
USEPA. 2016h. Technologies for Legionella Control in Premise Plumbing Systems: Scientific
Literature Review. EPA 810-R-16-001. September 2016.
USEPA. 2016i. Drinking Water Contaminant Candidate List 4; Final Notice. 81 FR 81099.
November 17, 2016.
USEPA and Water Research Foundation. 2016. Summary Document: State of Research on High-
Priority Distribution System Issues. Jointly prepared with the Water Research Foundation.
Six-Year Review 3 Technical Support Document 8-12
for Microbial Contaminant Regulations
December 2016
-------
U.S. Government Accountability Office (GAO). 1993. Drinking Water Key Quality Assurance
Program is Flawed and Underfunded. GAO/RCED-93-97.
U.S. Geological Survey (USGS). 2012. Digital Engineering Aspects of KarstMap: A GIS
Version of Davies, W.E., J.H. Simpson, G.C. Ohlmacher, W.S. Kirk, and E.G. Newton. 1984.
Engineering Aspects of Karst: U.S. Geological Survey, National Atlas of the United States of
America, Scale 1:7,500,000. Open-File Report 2004-1352, version 1. Available online only at:
http://pubs.usgs.gov/of/2004/1352/.
Verbeke, G. and G. Molenberghs. 2000. Linear Mixed Models for Longitudinal Data. 1st Ed.
Springer series in statistics. New York: Springer.
Wahman, D.G. and J.G. Pressman. 2015. Distribution system residuals - is "detectable" still
acceptable for chloramines. Journal of the AWWA. 107(8): 53-63.
Walker, J.T. and M. Morales. 1997. Evaluation of chlorine dioxide (CIO2) for the control of
biofilms. Water Science and Technology. 35(11-12): 319-323.
Wallender, E. K., E. C. Ailes, J. S. Yoder, V. A. Roberts, and J. M. Brunkard. 2014. Contributing
factors to disease outbreaks associated with untreated ground water. Ground Water. 52(6):886-
897.
Wang, H., S. Masters, Y. Hong, J. Stallings, J.O. Falkinham, M.A. Edwards, and A. Pruden.
2012. Effect of disinfectant, water age, and pipe material on occurrence and persistence of
Legionella, mycobacteria, Pseudomonas aeruginosa, and two amoebas. Environmental Science
and Technology. 46(21): 11566-11574.
Westerhoff, P., H. Jang, M. Abbaszadegan, and A. Absar. 2010. Organic chloramine formation
and influence on disinfection efficacy and nitrification. Water Research Foundation.
Wilson, M.P., W.D. Gollnitz, S.N. Boutros, W.T. Boria. 1996. Determining Ground Water
Under the Direct Influence of Surface Water. AWWA Research Foundation, Denver, CO.
Ying, G. S. and Liu, C. C. 2006. Statistical Analysis of Clustered Data using SAS(r) System.
NESUG Conference Proceedings. 2006.
Six-Year Review 3 Technical Support Document 8-13
for Microbial Contaminant Regulations
December 2016
-------
Six-Year Review 3 Technical Support
Document for Microbial Contaminant
Regulations: Appendices
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
December 2016
-------
List of Appendices
Appendix A: Data Quality Assurance/Quality Control Documentation for SYR3 ICR Microbial
Data
Appendix B: Additional Analyses on the Disinfectant Residuals in Distribution Systems
Appendix C: Additional Analyses on the Occurrence of TC+ and EC+ in Surface Water and
Ground Water Systems Compared to Disinfectant Residuals in Distribution
Systems
Appendix D: Producing a Reduced Dataset for Undisinfected Ground Water Systems
Appendix E: Analysis of the Generalized Estimating Equation (GEE) and Generalized Linear
Mixed Models (GLMM) as used to Estimate the Relative Rate of Highly Credible
Gastrointestinal Illness (HCGI) by Colford et al. (2009)
Appendix F: Occurrence of Total Conforms/A", coli in Small PWSs Using Undisinfected Ground
Water
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
December 2016
-------
Appendix A. Data Quality Assurance/Quality Control Documentation for
SYR3 ICR Microbial Data
As part of the third Six-Year Review (SYR3), the Environmental Protection Agency (EPA)
conducted a voluntary data call-in from the states, territories and tribes to obtain the data. The
data call resulted in over 47 million compliance and water quality records collected between
2006 and 2011 delivered to EPA. The records within the SYR3 Information Collection Request
(ICR) database were collected and analyzed using a rigorous quality assurance and quality
control (QA/QC) process, which was designed to closely follow the process outlined in The Data
Management and Quality Assurance/Quality Control Process for the Third Six-Year Review
Information Collection Rule Dataset (USEPA, 2016e). See that report for the full details of the
QA/QC process, as well as data acquisition, storage, management and preparation (for analysis).
For the purposes of reviewing the microbial and disinfectant residual data during SYR3, EPA
compiled a dataset containing the records for total coliforms (TC), E. coli (EC), fecal coliform
(FC) and field free and total chlorine residual (referred to as the "SYR3 ICR microbial dataset").
This appendix provides an overview of the data management steps applicable to the SYR3 ICR
microbial data, highlighting when different approaches were used as compared to the chemical
contaminants regulated under the Chemical Phase Rules and radionuclide contaminants and
disinfection byproducts. As described below, a thorough QA/QC process was undertaken to
evaluate the microbial dataset. Note that this QA process was not entirely the same as the process
used for the Six-Year 2 data reviewed under the Revised Total Coliform Rule (RTCR).
Data Management Steps
A number of data management tasks were necessary to prepare the SYR3 datasets for QA/QC
review and, ultimately, for data analysis. Some states/entities submitted their data using the EPA-
provided Safe Drinking Water Information System (SDWIS) extract tool and other states/entities
submitted their data "as is," in several in different formats. Whenever possible, EPA restructured
the data submitted from the non-SDWIS states into the SDWIS state format.1
The SDWIS states submitted compliance monitoring data that contained TC/EC results paired
with field chlorine residual data collected at the same time and location. With the exception of
four tribal datasets (Region 1, Region 9, Navajo Nation and American Samoa), the non-SDWIS
states did not submit their microbial data in that format. Many non-SDWIS states/entities did
submit TC, EC and FC data but did not include the corresponding chlorine residual data. Other
non-SDWIS states submitted their microbial data in summary form (e.g., one summary record
for several water samples); these data were not uploaded to the SYR3 ICR database. For more
details on these data management steps see USEPA (2016e).
1 At the time of data collection for SYR3, about 75 percent of all states stored and managed at least portions of their
compliance monitoring data in the Safe Drinking Water Information System/State Version (SDWIS/State). In an
attempt to make the SYR3 data submittal process as easy for states as possible, EPA developed a SDWIS/State
Extract Tool, which ran a customized query to pull the requested data from a SDWIS/State database. Nearly all of
the states using SDWIS/State that submitted data to EPA for SYR3 used the SDWIS/State Extract Tool to extract
and compile the EPA-requested compliance monitoring data.
Six-Year Review 3 Technical Support Document A-l
for Microbial Contaminant Regulations
December 2016
-------
SYR3 ICR Database Elements
The SYR3 ICR database includes data collected from states and primacy agencies. There are
many different data elements to track items such as laboratory sample results, water system
characteristics and QA/QC processes. A more detailed description of the data and collection
efforts is available in USEPA (2016e).
For the purposes of conducting occurrence analyses, the data elements were grouped into several
tables and combined using queries to create a coherent and usable dataset. The occurrence
analyses often differ between Six-Year Review 3 contaminants, and certain elements were used
in the SYR3 ICR microbial data analyses that may not be useful or relevant to other
contaminants, and vice versa. Exhibit A.l lists each of the data elements used for conducting
microbial/disinfectant residual occurrence analyses, along with a brief description. Any fields
that were included in the original datasets but are not listed below were not relevant to
conducting the microbial/ disinfectant residual occurrence analyses presented in Chapter 6 of this
support document.
Exhibit A.1: List of Primary SYR3 ICR Dataset Elements Used for Microbial and
Disinfectant Residual Occurrence Analyses
Field Name
Description
Analyte ID
4-digit SDWIS analyte code
Analyte Name
Analyte name
State Code
Used to identify the state in which a system is located, including tribal systems.
PWSID
Public water system identification number (PWSID).
System Name
Water system name.
System Type
Water system type according to federal requirements.
C = Community water system
NC = Non-community water system
NTNC = Non-transient non-community water system
NP = Non-public water system
Retail Population Served
Retail population served by the water system.
Source Water Type
Primary water source for the water system.
GU = Ground water under direct influence of surface water
GUP = Purchased GU
GW = Ground water
GWP = Purchased GW
SW = Surface water
SWP = Purchased SW
Water Facility ID
Unique identifier for each water system facility.
Water Facility Type
Type of the water system facility.
Six-Year Review 3 Technical Support Document A-2
for Microbial Contaminant Regulations
December 2016
-------
Field Name
Description
CC = consecutive connection; CH = common headers;
CS = cistern; CW = clear well; DS = distribution system;
IG = infiltration gallery; IN = intake; NP = non-piped, purchased;
OT = other; PC = pressure control; PF = pumping facility;
RS = reservoir; SP = spring; SS = sampling station; ST = storage;
TM = transmission main (manifold); TP = treatment plant;
WH = well head; WL = well; XX = unknown
Sampling Point ID
Unique identifier for each sample point.
Sampling Point Type
Location type of a sampling point.
DS = distribution system; EP = entry point; FC = first customer;
FN = finished water; LD = lowest disinfectant residual;
MD = midpoint in the DS; MR = point of maximum residence;
PC = process control; RW = raw water source; SR = source water point; UP = unit
process; WS = water system facility point
Source Type
The type of water source, based on whether treatment has taken place.
FN = Finished, treated; RW = Raw, untreated; XX = Unknown
Sample Type Code
Sample type code.
CO = confirmation; DU = duplicate; FB = field blank;
MR = maximum residence time; MS = matrix spike;
OT = other; RP = repeat; RT = routine; RW = raw water;
SB = shipping blank; SP = special; TE = technical evaluation
Six Year ID
Unique identifier for each sample analytical result. Used as primary key to link multiple
tables.
Sample Collection Date
Sample collection date.
Detection Limit Value
Limit below which the specific lab indicated they could not reliably measure results for a
contaminant with the methods and procedures used by the lab.
Detection Limit Unit
Units of the detection limit value
Detect
Added by EPA to indicate whether the result was a detection record (1) or a non-
detection record (0), based off of the sample analytical result fields in the raw datasets.
Value
Actual numeric (decimal) value of the concentration for the chemical result. This value is
equal to zero if the analytical result is less than the contaminant's MRL.
Units
Unit of measurement for the analytical results reported. All DBP records were converted
to |jg/L for analytical purposes. All TOC and alkalinity records were converted to mg/L
for analytical purposes. Added by EPA.
Presence Indicator Code
Indicates whether results of an analysis were positive (P-Presence) or negative (A-
Absence).
Field Free Chlorine Residual Measure
Amount of free chlorine residual (mg/L) found in the water after disinfection has been
applied. These concentrations were measured in the field at the same time and location
as coliform (TC-EC-FC) samples were collected.
Field Total Chlorine Residual Measure
Amount of total chlorine residual (mg/L) found in the water after disinfection has been
applied. These concentrations were measured in the field at the same time and location
as coliform (TC-EC-FC) samples were collected.
Six-Year Review 3 Technical Support Document A-3
for Microbial Contaminant Regulations
December 2016
-------
QA/QC Steps
The SYR3 ICR QA/QC effort encountered a range of data quality issues across contaminants and
states/entities. Quality control measures were established to identify records that fit certain
criteria using a two-step process. The first round of QA/QC was established at the time of data
submission, when flags fitting exclusion criteria were run against a state's data submission.
These QA/QC steps were applied to all regulated contaminant monitoring data in the SYR3 ICR
database. See USEPA (2016e) for complete details on the first round of the SYR3 ICR QA/QC
process. Similar to the process for the chemical contaminants, radionuclides and disinfection
byproducts, the initial QA/QC steps were conducted on the SYR3 ICR microbial data.
The first round of QA/QC review resulted in the exclusion of any records that met any of the
following criteria:
Records marked with sample type codes other than routine, repeat, or confirmation;
Records not marked as being for "compliance";
Records from non-public water systems;
Records from outside of the SYR3 date range; and
• Records from systems missing inventory information.
The second round of QA/QC procedures allows for the exclusion of records that did not have
paired microbial and disinfectant residual data or do not fit within the contaminant's rule
requirement context. Additional QA/QC steps were applied that were specific to the SYR3 ICR
microbial dataset. The second round of QA/QC review resulted in the exclusion of any records
that met any of the following criteria:
Records from non-SDWIS states. SDWIS states reported field free and/or total chlorine
residual data collected at the same time and location as the TC/EC data. TC and EC data
from a total of 41 SDWIS and 5 non-SDWIS states were included in the SYR3 ICR
database. Only SDWIS states' data (and some tribal data formatted just like the SDWIS
states) were included in the final SYR3 ICR microbial dataset because those states
submitted "paired" chlorine and coliform data.
Records marked with a sample type code of confirmation. Only routine and repeat
samples were used in the analysis.
Records from water facility type codes other than distribution systems. Only data where
TYPECODE = "DS" were included in the analysis.
Free and/or total chlorine records paired with analytes other than TC/EC/FC. Only free
and/or total chlorine records associated with TC, EC, or FC samples were included in the
analysis.
Records where PRESENCEINDCODE (presence indicator code) was null or not equal
to either "A" (absent) or "P" (present).
Records with a field free chlorine concentration greater than the total chlorine
concentration.
Records without any field free chlorine or total chlorine data.
Six-Year Review 3 Technical Support Document A-4
for Microbial Contaminant Regulations
December 2016
-------
Records from Alabama, Louisiana and South Carolina. These states' data were identified
as outliers since most of the TC samples submitted by these states were TC positive
(TC+); these data were not considered to be representative of overall TC occurrence rates
in Alabama, Louisiana and South Carolina.
Records with a free chlorine concentration greater than 10 mg/L and records with a total
chlorine concentration greater than 20 mg/L. These high values were considered to be
potential outliers.
TC positive (TC+) results without a corresponding EC/FC result and EC/FC results
without a corresponding TC+ result. Note that this step was not applied to the disinfectant
residual data used in Section 6.2 analyses. However, these TC/EC/FC records were
excluded from the Section 6.3 analyses.
After applying the filter protocol to more than 12 million SYR3 ICR microbial data records and
almost 9 million free and/or total chlorine records, almost 6 million SYR3 ICR microbial data
records remained in the final dataset that was used for conducting occurrence analyses. (Note
that a subset of these 6 million records were used for the analyses in Section 6.3). Exhibit A.2
documents the specific counts of records included and excluded in each step for the three
contaminants and for the disinfectant residuals.
Six-Year Review 3 Technical Support Document A-5
for Microbial Contaminant Regulations
December 2016
-------
Exhibit A.2: QA Steps for the SYR3 ICR Microbial Data
Step
Total Coliforms
E. coli
Fecal Coliform
Field Free
Chlorine
Field Total
Chlorine
Included
Excluded
Included
Excluded
Included
Excluded
Included
Excluded
Included
Excluded
Original Records
9,953,551
1,833,281
281,642
5,273,525
3,489,849
Step 1: Initial QA (Applied to all SYR3 contaminants)1
9,766,686
186,865
1,804,329
28,952
264,090
17,552
5,181,269
92,256
3,451,496
38,353
Step 2: Removal of records from Non-SDWIS states
8,616,753
1,149,933
1,632,695
171,634
113,608
150,482
5,181,269
0
3,451,496
0
Step 3: Removal of records marked with sample type code of
"confirmation" (This analysis included "routine" and "repeat"
samples.)
8,616,074
679
1,632,093
602
113,608
0
5,181,194
75
3,451,423
73
Step 4: Removal of non-distribution system samples
8,283,060
333,014
1,396,310
235,783
108,452
5,156
4,513,013
668,181
3,217,731
233,692
Step 5: Removal of non-TC/EC/FC samples
8,283,060
0
1,396,310
0
108,452
0
4,172,134
340,879
2,680,319
537,412
Step 6: Removal of records where PRESENCE_IND_CODE
(presence indicator code) was null or not equal to either "A"
(absent) or "P" (present)
7,984,551
298,509
1,363,400
32,910
103,608
4,844
4,171,861
273
2,679,444
875
Step 7: Removal of records with field free chlorine concentration
greater than total chlorine concentration
7,829,837
154,714
1,362,436
964
103,570
38
4,016,145
155,716
2,523,728
155,716
Step 8: Removal of records without any field free or total
chlorine data
4,757,381
3,072,456
892,091
470,345
64,335
39,235
4,016,145
0
2,523,728
0
Step 9: Removal of records from AL, LA and SC
4,750,983
6,398
889,683
2,408
64,306
29
4,007,768
8,377
2,521,983
1,745
Step 10: Removal of high free chlorine concentrations > 10
mg/L; Removal of high total chlorine concentrations > 20 mg/L
4,750,432
551
889,570
113
64,304
2
4,007,235
533
2,521,771
212
Final number of records in the SYR3 microbial dataset used for
Section 6.2 analyses
4,750,432
889,570
64,304
4,007,235
2,521,771
Percent Included for Section 6.2 Analysis
47.7%
48.5%
22.8%
76.0%
72.3%
Step 11: Removal of remaining TC positive (TC+) records
without a corresponding EC/FC result or EC/FC result without a
corresponding TC+ result
4,749,332
1,100
35,889
853,681
3,781
60,523
3,423,730
583,505
2,140,638
381,133
Final number of records in the SYR3 microbial dataset used for
Section 6.3 analyses
4,749,332
35,889
3,781
3,423,730
2,140,638
Percent Included for Section 6.3 Analysis
47.7%
2.0%
1.3%
64.9%
61.3%
1 The first round of QA/QC included the basic suite of QA/QC steps that were performed on the whole dataset. This QA/QC review resulted in the exclusion of any records that met the
following criteria: (1) records marked with sample type codes other than routine, repeat, or confirmation; (2) records not marked as being for "compliance"; (3) records from non-public
water systems; (4) records from outside of the SYR3 date range; and (5) records from systems missing inventory information.
Six-Year Review 3 Technical Support Document A-6 December 2016
For Microbial Contaminant Regulations
-------
As described above, EPA deemed the current QA/QC process to be sufficient for a dataset of this
size. The QA/QC process excluded records that were identified as not being appropriate for this
analysis, yielding a final dataset to be used as a basis for analysis. The final SYR3 ICR microbial
dataset consists of compliance monitoring data received from 34 states/primacy agencies
representing a large sample of paired TC/EC and disinfectant residual data. Exhibit A.3 presents
a map of the 34 states/entities with data in the final SYR3 ICR microbial dataset.
Exhibit A.3: States/Entities with Data in SYR3 ICR Microbial Dataset
~
~
~
~
~
AmericanSamoa
District of Coiumbia
Guam
Northern Mariana Islands
Puerto Rico
Virgin Islands
Region 1 Tribes
| [ Region 2 Tribes
| | Region 3 Tribes
Region 4Tribes
Region 5Tribes
| ~~| Region 6Tribes
| | Region 7 Tribes
Region STribes
| Region 9 Navajo Nation
Region 9Tribes
| Region lOTribes
~
No data in finalSYR3 Microbial Dataset
Data in finalSYR3 Microbial Dataset
Exclude from finalSYR3 Microbial Dataset
The final SYR3 ICR microbial dataset includes almost 6 million records of paired coliform and
disinfectant residual data from 34 states over a six-year period. Data from both surface water
systems and ground water systems are included, as well as all three system types (i.e.,
community water systems, non-transient non-community water systems and transient water
systems) and all system sizes. The 34 states with data are distributed across the entire United
States. The final SYR3 ICR microbial dataset enabled the analyses to be conducted with
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
December 2016
-------
different stratifications, such as by source water type, system size, system type, etc. An
exploration of potential annual, seasonal and geographic trends was also possible using these
data.
A few limitations of the final SYR3 ICR microbial dataset are noted. An initial evaluation of the
completeness of these data exposed a large degree of variability in the number of records
provided by water systems from state to state. The number of records range from one record
from the State of Maine to almost 900,000 records from the State of Illinois. For the State of
Maine, only one record (from 2008) passed all of the QA/QC steps when preparing the final
SYR3 ICR microbial dataset. Very few TC/EC records from Maine had paired chlorine residual
data in the SYR3 ICR database. Other states' data were similar to Maine (i.e., the state provided
a large amount of TC/EC data but only a small portion of those data were paired with chlorine
residual concentrations). Furthermore, there are also many states and some regions of the U.S.
whose data are missing from the analysis. As discussed previously, only the data from SDWIS
states were included in the final SYR3 ICR microbial dataset. The SDWIS states provided
TC/EC data in a usable format that were also paired with disinfectant residual data.
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
A-8
December 2016
-------
Appendix B. Additional Analyses on the Disinfectant Residuals in
Distribution Systems
This appendix provides the analytical results for surface water and ground water systems that
were not presented within the body of Chapter 6. This appendix also includes an evaluation of
the occurrence related to disinfectant residuals in distribution systems relative to system type and
system size, as well as seasonal changes, annual trends and geographic distribution.
Exhibit B. 1 through Exhibit B.6 present a breakdown of the counts of free and total chlorine data
associated with total coliform samples, and the systems providing those data, by source water
type, system type and system size for each year of data. These results are presented for the years
2006 through 2011.
Exhibit B.1: Summary of Free and Total Chlorine Residual Data by Source Water
Type, System Type and System Size in 2006
System
Type
Source
Water
Type
Population
Served Size
Category
Number of Systems with
Routine TC Samples
Number of Routine TC
Samples
Free Chlorine
Total Chlorine
Free Chlorine Total Chlorine
CWSs
GW
<100
1,517
585
14,426
6,659
101-500
2,322
1,138
25,011
17,017
501-1,000
847
557
10,835
9,393
1,001-4,100
1,103
756
30,518
20,061
4,101-33,000
552
302
65,641
29,432
33,001-100,000
56
32
31,767
13,255
>100,000
8
3
7,472
127
Total GW
6,405
3,373
185,670
95,944
SW
<100
230
94
2,882
1,946
101-500
527
337
5,906
5,970
501-1,000
266
212
3,493
3,962
1,001-4,100
629
527
18,801
15,343
4,101-33,000
614
381
77,801
40,298
33,001-100,000
92
47
52,900
23,466
>100,000
29
21
33,839
24,266
Total SW
2,387
1,619
195,622
115,251
TNCWSs
GW
<100
3,114
898
12,654
4,392
101-500
710
262
3,424
1,117
501-1,000
79
16
480
74
1,001-4,100
36
5
562
49
4,101-33,000
2
0
80
0
33,001-100,000
0
0
0
0
>100,000
0
0
0
0
Total GW
3,941
1,181
17,200
5,632
SW
<100
170
31
1,436
253
Six-Year Review 3 Technical Support Document B-l
for Microbial Contaminant Regulations
December 2016
-------
System
Type
Source
Water
Type
Population
Served Size
Category
Number of Systems with
Routine TC Samples
Number of Routine TC
Samples
Free Chlorine
Total Chlorine
Free Chlorine Total Chlorine
101-500
76
10
676
84
501-1,000
9
2
97
20
1,001-4,100
8
0
97
0
4,101-33,000
5
0
147
0
33,001-100,000
0
0
0
0
>100,000
0
0
0
0
Total SW
268
43
2,453
357
NTNCWSs
GW
<100
769
172
3,844
1,217
101-500
573
170
3,244
1,218
501-1,000
156
37
1,098
345
1,001-4,100
87
33
1,727
704
4,101-33,000
6
2
273
56
33,001-100,000
0
0
0
0
>100,000
0
0
0
0
Total GW
1,591
414
10,186
3,540
SW
<100
40
13
462
120
101-500
55
18
669
169
501-1,000
18
8
237
162
1,001-4,100
19
12
391
189
4,101-33,000
0
0
0
0
33,001-100,000
0
0
0
0
>100,000
0
0
0
0
Total SW
132
51
1,759
640
Total
GW
<100
5,400
1,655
30,924
12,268
101-500
3,605
1,570
31,679
19,352
501-1,000
1,082
610
12,413
9,812
1,001-4,100
1,226
794
32,807
20,814
4,101-33,000
560
304
65,994
29,488
33,001-100,000
56
32
31,767
13,255
>100,000
8
3
7,472
127
Total GW
11,937
4,968
213,056
105,116
SW
<100
440
138
4,780
2,319
101-500
658
365
7,251
6,223
501-1,000
293
222
3,827
4,144
1,001-4,100
656
539
19,289
15,532
4,101-33,000
619
381
77,948
40,298
33,001-100,000
92
47
52,900
23,466
>100,000
29
21
33,839
24,266
Total SW
2,787
1,713
199,834
116,248
Six-Year Review 3 Technical Support Document B-2
for Microbial Contaminant Regulations
December 2016
-------
Exhibit B.2: Summary of Free and Total Chlorine Residual Data by Source Water
Type, System Type and System Size in 2007
System
Type
Source
Water
Type
Population
Served Size
Category
Number of Systems with
Routine TC Samples
Number of Routine TC
Samples
Free Chlorine
Total Chlorine
Free Chlorine Total Chlorine
CWSs
GW
<100
2,108
574
19,777
7,288
101-500
3,223
1,233
31,136
18,136
501-1,000
1,061
639
11,328
9,980
1,001-4,100
1,394
864
32,317
22,033
4,101-33,000
648
380
65,110
35,168
33,001-100,000
72
46
31,424
15,282
>100,000
13
12
5,116
2,424
Total GW
8,519
3,748
196,208
110,311
SW
<100
280
101
3,788
1,935
101-500
626
356
7,272
6,007
501-1,000
297
234
4,010
3,939
1,001-4,100
740
579
20,888
15,669
4,101-33,000
688
426
86,009
44,094
33,001-100,000
114
66
56,885
26,382
>100,000
50
38
48,931
25,738
Total SW
2,795
1,800
227,783
123,764
TNCWSs
GW
<100
3,531
977
13,992
4,167
101-500
940
265
4,413
1,123
501-1,000
105
19
585
67
1,001-4,100
42
5
720
60
4,101-33,000
2
0
67
0
33,001-100,000
0
0
0
0
>100,000
0
0
0
0
Total GW
4,620
1,266
19,777
5,417
SW
<100
173
16
1,460
192
101-500
83
9
763
57
501-1,000
15
4
131
29
1,001-4,100
9
2
141
2
4,101-33,000
6
1
173
18
33,001-100,000
0
0
0
0
>100,000
0
0
0
0
Total SW
286
32
2,668
298
NTNCWSs
GW
<100
932
168
4,974
1,039
101-500
776
143
5,073
913
501-1,000
217
37
1,645
337
1,001-4,100
127
28
2,484
602
4,101-33,000
9
1
508
96
33,001-100,000
0
0
0
0
>100,000
0
0
0
0
Total GW
2,061
377
14,684
2,987
SW
<100
52
8
590
72
101-500
61
13
698
137
501-1,000
22
7
281
144
1,001-4,100
23
10
568
171
Six-Year Review 3 Technical Support Document B-3
for Microbial Contaminant Regulations
December 2016
-------
System
Type
Source
Water
Type
Population
Served Size
Category
Number of Systems with
Routine TC Samples
Number of Routine TC
Samples
Free Chlorine
Total Chlorine
Free Chlorine Total Chlorine
4,101-33,000
1
0
200
0
33,001-100,000
1
0
321
0
>100,000
0
1
0
2
Total SW
160
39
2,658
526
Total
GW
<100
6,571
1,719
38,743
12,494
101-500
4,939
1,641
40,622
20,172
501-1,000
1,383
695
13,558
10,384
1,001-4,100
1,563
897
35,521
22,695
4,101-33,000
659
381
65,685
35,264
33,001-100,000
72
46
31,424
15,282
>100,000
13
12
5,116
2,424
Total GW
15,200
5,391
230,669
118,715
SW
<100
505
125
5,838
2,199
101-500
770
378
8,733
6,201
501-1,000
334
245
4,422
4,112
1,001-4,100
772
591
21,597
15,842
4,101-33,000
695
427
86,382
44,112
33,001-100,000
115
66
57,206
26,382
>100,000
50
39
48,931
25,740
Total SW
3,241
1,871
233,109
124,588
Exhibit B.3: Summary of Free and Total Chlorine Residual Data by Source Water
Type, System Type and System Size in 2008
System
Type
Source
Water
Type
Population
Served Size
Category
Number of Systems with
Routine TC Samples
Number of Routine TC
Samples
Free Chlorine
Total Chlorine
Free Chlorine Total Chlorine
CWSs
GW
<100
2,122
611
19,635
8,062
101-500
3,256
1,310
30,325
20,084
501-1,000
1,097
653
11,357
11,328
1,001-4,100
1,462
874
32,509
24,211
4,101-33,000
690
399
66,799
38,883
33,001-100,000
72
48
30,168
16,525
>100,000
16
10
5,152
2,932
Total GW
8,715
3,905
195,945
122,025
SW
<100
301
98
3,764
2,037
101-500
664
338
7,370
5,984
501-1,000
319
226
4,061
4,327
1,001-4,100
778
539
20,573
16,605
4,101-33,000
701
441
86,742
48,029
33,001-100,000
126
69
60,460
27,661
>100,000
59
43
49,113
29,994
Total SW
2,948
1,754
232,083
134,637
TNCWSs
GW
<100
3,655
1,133
15,054
5,923
101-500
1,080
328
4,873
1,594
Six-Year Review 3 Technical Support Document B-4
for Microbial Contaminant Regulations
December 2016
-------
System
Type
Source
Water
Type
Population
Served Size
Category
Number of Systems with
Routine TC Samples
Number of Routine TC
Samples
Free Chlorine
Total Chlorine
Free Chlorine Total Chlorine
501-1,000
113
20
655
113
1,001-4,100
40
5
666
62
4,101-33,000
2
0
9
0
33,001-100,000
0
0
0
0
>100,000
0
0
0
0
Total GW
4,890
1,486
21,257
7,692
SW
<100
194
26
1,465
311
101-500
98
9
866
68
501-1,000
18
3
163
19
1,001-4,100
13
0
172
0
4,101-33,000
7
0
193
0
33,001-100,000
0
0
0
0
>100,000
0
0
0
0
Total SW
330
38
2,859
398
NTNCWSs
GW
<100
1,011
185
5,332
1,441
101-500
823
140
5,300
1,314
501-1,000
224
33
1,669
363
1,001-4,100
138
31
2,837
809
4,101-33,000
10
3
735
96
33,001-100,000
0
0
0
0
>100,000
0
0
0
0
Total GW
2,206
392
15,873
4,023
SW
<100
58
13
566
95
101-500
67
18
751
199
501-1,000
23
8
309
127
1,001-4,100
24
11
579
206
4,101-33,000
1
0
479
0
33,001-100,000
1
0
960
0
>100,000
0
0
0
0
Total SW
174
50
3,644
627
Total
GW
<100
6,788
1,929
40,021
15,426
101-500
5,159
1,778
40,498
22,992
501-1,000
1,434
706
13,681
11,804
1,001-4,100
1,640
910
36,012
25,082
4,101-33,000
702
402
67,543
38,979
33,001-100,000
72
48
30,168
16,525
>100,000
16
10
5,152
2,932
Total GW
15,811
5,783
233,075
133,740
SW
<100
553
137
5,795
2,443
101-500
829
365
8,987
6,251
501-1,000
360
237
4,533
4,473
1,001-4,100
815
550
21,324
16,811
4,101-33,000
709
441
87,414
48,029
33,001-100,000
127
69
61,420
27,661
>100,000
59
43
49,113
29,994
Total SW
3,452
1,842
238,586
135,662
Six-Year Review 3 Technical Support Document B-5
for Microbial Contaminant Regulations
December 2016
-------
Exhibit B.4: Summary of Free and Total Chlorine Residual Data by Source Water
Type, System Type and System Size in 2009
System
Type
Source
Water
Type
Population
Served Size
Category
Number of Systems with
Routine TC Samples
Number of Routine TC
Samples
Free Chlorine
Total Chlorine
Free Chlorine Total Chlorine
CWSs
GW
<100
2,533
828
24,830
8,791
101-500
4,098
1,835
41,508
23,046
501-1,000
1,494
963
16,806
13,640
1,001-4,100
2,135
1,371
50,892
31,497
4,101-33,000
1,028
683
94,543
53,510
33,001-100,000
88
60
34,748
19,776
>100,000
17
8
8,032
5,507
Total GW
11,393
5,748
271,359
155,767
SW
<100
335
147
4,081
2,675
101-500
749
481
7,854
6,807
501-1,000
380
326
4,447
5,151
1,001-4,100
851
788
20,767
20,754
4,101-33,000
785
597
85,783
59,941
33,001-100,000
131
82
62,258
35,915
>100,000
66
48
54,371
44,898
Total SW
3,297
2,469
239,561
176,141
TNCWSs
GW
<100
4,121
1,437
18,351
6,789
101-500
1,407
544
8,458
2,530
501-1,000
146
51
1,090
319
1,001-4,100
51
14
936
147
4,101-33,000
1
0
1
0
33,001-100,000
0
0
0
0
>100,000
0
0
0
0
Total GW
5,726
2,046
28,836
9,785
SW
<100
198
49
1,667
373
101-500
112
15
1,045
107
501-1,000
21
5
194
55
1,001-4,100
13
2
202
37
4,101-33,000
6
0
134
0
33,001-100,000
0
0
0
0
>100,000
0
0
0
0
Total SW
350
71
3,242
572
NTNCWSs
GW
<100
1,203
317
7,594
2,260
101-500
1,027
250
7,744
1,923
501-1,000
270
68
2,292
515
1,001-4,100
188
63
4,112
1,304
4,101-33,000
16
5
972
320
33,001-100,000
0
0
0
0
>100,000
0
0
0
0
Total GW
2,704
703
22,714
6,322
SW
<100
67
20
663
191
101-500
101
44
1,043
390
501-1,000
24
11
354
138
Six-Year Review 3 Technical Support Document B-6
for Microbial Contaminant Regulations
December 2016
-------
System
Type
Source
Water
Type
Population
Served Size
Category
Number of Systems with
Routine TC Samples
Number of Routine TC
Samples
Free Chlorine Total Chlorine
Free Chlorine Total Chlorine
1,001-4,100
25
11
634
227
4,101-33,000
3
1
566
72
33,001-100,000
1
0
958
0
>100,000
0
1
0
1
Total SW
221
88
4,218
1,019
Total
GW
<100
7,857
2,582
50,775
17,840
101-500
6,532
2,629
57,710
27,499
501-1,000
1,910
1,082
20,188
14,474
1,001-4,100
2,374
1,448
55,940
32,948
4,101-33,000
1,045
688
95,516
53,830
33,001-100,000
88
60
34,748
19,776
>100,000
17
8
8,032
5,507
Total GW
19,823
8,497
322,909
171,874
SW
<100
600
216
6,411
3,239
101-500
962
540
9,942
7,304
501-1,000
425
342
4,995
5,344
1,001-4,100
889
801
21,603
21,018
4,101-33,000
794
598
86,483
60,013
33,001-100,000
132
82
63,216
35,915
>100,000
66
49
54,371
44,899
Total SW
3,868
2,628
247,021
177,732
Exhibit B.5: Summary of Free and Total Chlorine Residual Data by Source Water
Type, System Type and System Size in 2010
System
Type
Source
Water
Type
Population
Served Size
Category
Number of Systems with
Routine TC Samples
Number of Routine TC
Samples
Free Chlorine
Total Chlorine
Free Chlorine Total Chlorine
CWSs
GW
<100
2,603
975
26,126
10,761
101-500
4,129
2,062
44,071
27,820
501-1,000
1,577
1,068
19,086
16,072
1,001-4,100
2,267
1,606
56,510
40,331
4,101-33,000
1,175
839
113,404
73,342
33,001-100,000
103
69
41,646
28,179
>100,000
16
14
9,805
10,876
Total GW
11,870
6,633
310,648
207,381
SW
<100
327
170
4,111
3,711
101-500
770
530
8,534
8,987
501-1,000
384
362
4,780
6,065
1,001-4,100
821
866
22,271
26,151
4,101-33,000
877
720
105,667
77,859
33,001-100,000
152
111
78,231
54,877
>100,000
85
60
84,004
80,354
Six-Year Review 3 Technical Support Document B-7
for Microbial Contaminant Regulations
December 2016
-------
System
Type
Source
Water
Type
Population
Served Size
Category
Number of Systems with
Routine TC Samples
Number of Routine TC
Samples
Free Chlorine
Total Chlorine
Free Chlorine Total Chlorine
Total SW
3,416
2,819
307,598
258,004
TNCWSs
GW
<100
4,084
1,406
18,796
8,089
101-500
1,459
598
8,755
3,128
501-1,000
174
62
1,244
449
1,001-4,100
71
27
1,222
342
4,101-33,000
1
0
3
0
33,001-100,000
0
0
0
0
>100,000
0
0
0
0
Total GW
5,789
2,093
30,020
12,008
SW
<100
222
99
1,685
492
101-500
118
24
1,140
319
501-1,000
25
9
207
104
1,001-4,100
14
4
240
33
4,101-33,000
6
1
130
2
33,001-100,000
0
0
0
0
>100,000
0
0
0
0
Total SW
385
137
3,402
950
NTNCWSs
GW
<100
1,238
371
7,764
2,676
101-500
1,089
336
8,005
2,354
501-1,000
284
97
2,328
748
1,001-4,100
211
98
5,066
2,213
4,101-33,000
16
6
917
307
33,001-100,000
0
0
0
0
>100,000
0
0
0
0
Total GW
2,838
908
24,080
8,298
SW
<100
70
24
633
409
101-500
99
49
1,057
495
501-1,000
21
10
293
144
1,001-4,100
25
14
806
402
4,101-33,000
4
1
616
69
33,001-100,000
1
0
961
0
>100,000
0
0
0
0
Total SW
220
98
4,366
1,519
Total
GW
<100
7,925
2,752
52,686
21,526
101-500
6,677
2,996
60,831
33,302
501-1,000
2,035
1,227
22,658
17,269
1,001-4,100
2,549
1,731
62,798
42,886
4,101-33,000
1,192
845
114,324
73,649
33,001-100,000
103
69
41,646
28,179
>100,000
16
14
9,805
10,876
Total GW
20,497
9,634
364,748
227,687
SW
<100
619
293
6,429
4,612
101-500
987
603
10,731
9,801
501-1,000
430
381
5,280
6,313
1,001-4,100
860
884
23,317
26,586
4,101-33,000
887
722
106,413
77,930
33,001-100,000
153
111
79,192
54,877
Six-Year Review 3 Technical Support Document B-8
for Microbial Contaminant Regulations
December 2016
-------
System
Type
Source
Water
Type
Population
Served Size
Category
Number of Systems with
Routine TC Samples
Number of Routine TC
Samples
Free Chlorine
Total Chlorine
Free Chlorine
Total Chlorine
>100,000
85
60
84,004
80,354
Total SW
4,021
3,054
315,366
260,473
Exhibit B.6: Summary of Free and Total Chlorine Residual Data by Source Water
Type, System Type and System Size in 2011
System
Type
Source
Water
Type
Population
Served Size
Category
Number of Systems with
Routine TC Samples
Number of Routine TC
Samples
Free Chlorine
Total Chlorine
Free Chlorine
Total Chlorine
CWSs
GW
<100
2,706
936
28,625
11,300
101-500
4,204
2,014
47,382
28,496
501-1,000
1,615
1,052
20,510
16,388
1,001-4,100
2,278
1,614
61,358
43,850
4,101-33,000
1,165
851
120,392
80,041
33,001-100,000
95
67
44,058
30,694
>100,000
22
14
11,437
12,426
Total GW
12,085
6,548
333,762
223,195
SW
<100
319
182
4,221
3,878
101-500
783
550
8,697
9,540
501-1,000
379
371
4,824
6,535
1,001-4,100
833
876
22,269
27,773
4,101-33,000
883
720
111,969
86,737
33,001-100,000
160
108
84,968
61,261
>100,000
77
62
91,970
89,558
Total SW
3,434
2,869
328,918
285,282
TNCWSs
GW
<100
4,171
1,444
20,440
8,492
101-500
1,529
620
9,960
3,783
501-1,000
193
65
1,450
464
1,001-4,100
79
26
1,336
439
4,101-33,000
0
0
0
0
33,001-100,000
0
0
0
0
>100,000
0
0
0
0
Total GW
5,972
2,155
33,186
13,178
SW
<100
229
44
1,931
468
101-500
115
26
1,200
359
501-1,000
28
9
233
91
1,001-4,100
16
1
313
12
4,101-33,000
6
1
216
1
33,001-100,000
0
0
0
0
>100,000
0
0
0
0
Total SW
394
81
3,893
931
NTNCWSs
GW
<100
1,277
361
9,019
2,893
101-500
1,137
329
8,968
2,389
501-1,000
300
102
2,450
832
1,001-4,100
213
94
5,308
2,443
Six-Year Review 3 Technical Support Document B-9
for Microbial Contaminant Regulations
December 2016
-------
System
Type
Source
Water
Type
Population
Served Size
Category
Number of Systems with
Routine TC Samples
Number of Routine TC
Samples
Free Chlorine
Total Chlorine
Free Chlorine Total Chlorine
4,101-33,000
15
6
1,100
346
33,001-100,000
0
0
0
0
>100,000
0
0
0
0
Total GW
2,942
892
26,845
8,903
SW
<100
71
24
730
524
101-500
91
48
1,053
564
501-1,000
21
10
316
143
1,001-4,100
25
13
837
450
4,101-33,000
7
2
759
107
33,001-100,000
1
0
956
0
>100,000
0
0
0
0
Total SW
216
97
4,651
1,788
Total
GW
<100
8,154
2,741
58,084
22,685
101-500
6,870
2,963
66,310
34,668
501-1,000
2,108
1,219
24,410
17,684
1,001-4,100
2,570
1,734
68,002
46,732
4,101-33,000
1,180
857
121,492
80,387
33,001-100,000
95
67
44,058
30,694
>100,000
22
14
11,437
12,426
Total GW
20,999
9,595
393,793
245,276
SW
<100
619
250
6,882
4,870
101-500
989
624
10,950
10,463
501-1,000
428
390
5,373
6,769
1,001-4,100
874
890
23,419
28,235
4,101-33,000
896
723
112,944
86,845
33,001-100,000
161
108
85,924
61,261
>100,000
77
62
91,970
89,558
Total SW
4,044
3,047
337,462
288,001
1 There is some overlap between the free chlorine and total chlorine groups (i.e., some TC records were associated
with both free and total chlorine residual concentrations). See Section 6.1.1 for a more detailed description about the
records that were associated with both free and total chorine residual concentrations and the possible implications on
the data analysis.
The remaining exhibits of this appendix (Exhibit B.7 through Exhibit B.24) present an evaluation
of the occurrence relative to system type, system size, seasonal changes, annual trends and
geographic distribution. The majority of these analyses focus on data for the year 2011, primarily
for community water systems only.
System Type
Exhibit B.7 and Exhibit B.8 are cumulative distribution plots presenting the free and total
chlorine residual concentrations in surface water community water systems and non-community
water systems, respectively, for the year 2011. Exhibit B.9 and Exhibit B.10 are cumulative
distribution plots presenting the free and total chlorine residual concentrations in ground water
community water systems and non-community water systems, respectively, for the year 2011.
Six-Year Review 3 Technical Support Document B-10
for Microbial Contaminant Regulations
December 2016
-------
Exhibit B.7 (samples from surface water CWSs) is very similar to the results from all surface
water systems (CWSs and NCWS, as shown in Exhibit 6.5), as expected, given that the vast
majority of the surface water samples came from CWSs (e.g., 337,462 free chlorine samples
from surface water systems with 328,918 of those free chlorine samples coming from surface
water CWSs). In general, the mean and median concentrations are similar between CWS and
NCWS, but the percent of samples < 0.2 mg/L and < 0.5 mg/L are higher in NCWSs for both
free and total chlorine. The results seem unusual since NCWS typically have less spread out
distribution systems compared to CW7Ss, and thus the water should have shorter time in the
distribution system and less disinfectant residual decay. The lower values in NCWSs compared
to CWSs may reflect different operational strategies, reporting errors, or the fact that the NCWS
dataset is so small compared to CWSs.
Exhibit B.7: Cumulative Percent of Free and Total Chlorine Residual
Concentrations in Surface Water CWSs (in 2011)
— Free Chlorine
- - Total Chlorine
2 3
Chlorine Residual (mg/L as CI2)
100%*
a)
S,
Q_
E
CD
co
o
CD
CT)
£
c
CD
O
CD
CL
J2
E
O
75%-
50%-
25%-
Statistic Free Chlorine Total Chlorine
Samples 328.918 285,282
10th Percentile 0.24 mg/L 0.70 mg/L
Mean Concentration 1.14 mg/L 1.82 mg/L
Median Concentration 1.00 mg/L 169 mg/L
90th Percentile 2.2 mg/L 3.2 mg/L
%Samples < 0.2 mg/L 8% 0.9%
%Samptes < 0.5 mg/L 18% 5%
Six-Year Review 3 Technical Support Document B-11
for Microbial Contaminant Regulations
December 2016
-------
Exhibit B.8: Cumulative Percent of Free and Total Chlorine Residual
Concentrations in Surface Water NCWSs (in 2011)
100%-
-------
Exhibit B.10: Cumulative Percent of Free and Total Chlorine Residual
Concentrations in Ground Water NCWSs (in 2011)
100%-
¦R 75%
CO
a;
Q.
E
03
CO
0
a)
CD
TO
1
QJ
2
a)
Q_
to
D
E
O
50%
25%
— Free Chlorine
- - Total Chlorine
Statistic
Free Chlorine
Total Chlorine
Samples
60,031
22.081
10th Percentile
0.00 mg/L
0.00 mg/L
Mean Concentration
0.72 mg/L
0.98 mg/L
Median Concentration
0.50 mg/L
0.85 mg/L
90th Percentile
1.8 mg/L
2.1 mg/L
%Samptes < 0.2 mg/L
29%
20%
%Samples <0.5 mg/L
46%
32%
2 3
Chlorine Residual (mg/L as CI2)
System Size
Exhibit B. 11 presents summary statistics, by system size, for the free and total chlorine residual
data associated with total coliform results in surface water CWSs for the year 2011. Exhibit B.12
presents similar information for ground water CWSs for the year 2011. Summary statistics
include: count, 10th percentile, median, average and 90th percentile.
The SW results for free chlorine show that the median concentration is slightly lower for systems
serving < 1,000 people, although there is a slight increase in the smallest size category (systems
serving < 100 people). The 10th percentile values for SW are highest for systems serving 4,100 -
100,000. The 10Th percentile concentrations for the two smallest systems size categories for SW,
as well as the largest size category, were less than 0.2 mg/L. Results for total chlorine are
similar, with the lowest 10th, median, mean and 90th percentile values generally occurring in
systems serving 100 - 1,000 people.
Six-Year Review 3 Technical Support Document B-13
for Microbial Contaminant Regulations
December 2016
-------
Exhibit B.11: Summary Statistics of Free and Total Chlorine Residual
Concentrations in Surface Water CWSs (in 2011), by System Size
System Size
(Population
Served)
Count
Chlorine Residual Concentration (mg/L)
10th Percentile
Median
Average
90th
Percentile
Free Chlorine
<100
4,221
0.10
0.83
1.05
2.27
101-500
8,697
0.10
0.64
0.77
1.51
501-1,000
4,824
0.20
0.79
0.90
1.80
1,001-4,100
22,269
0.20
0.90
0.95
1.74
4,101-33,000
111,969
0.30
1.00
1.04
1.80
33,001-100,000
84,968
0.34
0.96
1.10
2.10
>100,000
91,970
0.14
1.10
1.38
3.10
Total
328,918
0.24
1.00
1.14
2.20
Total Chlorine
<100
3,878
0.60
1.80
2.01
3.60
101-500
9,540
0.50
1.56
1.73
3.30
501-1,000
6,535
0.30
1.30
1.47
2.90
1,001-4,100
27,773
0.50
1.50
1.60
2.90
4,101-33,000
86,737
0.70
1.54
1.69
2.90
33,001-100,000
61,261
0.65
1.70
1.80
3.20
>100,000
89,558
0.79
2.02
2.05
3.40
Total
285,282
0.70
1.69
1.82
3.20
Six-Year Review 3 Technical Support Document B-14
for Microbial Contaminant Regulations
December 2016
-------
Exhibit B.12: Summary Statistics of Free and Total Chlorine Residual
Concentrations in Ground Water CWSs (in 2011), by System Size
System Size
(Population
Served)
Count
Chlorine Residual Concentration (mg/L)
10th
Percentile
Median
Average
90th
Percentile
Free Chlorine
<100
28,625
0.00
0.60
0.73
1.51
101-500
47,382
0.10
0.70
0.79
1.58
501-1,000
20,510
0.18
0.74
0.84
1.60
1,001-4,100
61,358
0.20
0.84
0.91
1.68
4,101-33,000
120,392
0.21
0.80
0.87
1.60
33,001-100,000
44,058
0.22
0.70
0.74
1.28
>100,000
11,437
0.24
0.91
1.17
2.60
Total
333,762
0.20
0.77
0.85
1.60
Total Chlorine
<100
11,300
0.20
0.99
1.06
2.05
101-500
28,496
0.30
1.00
1.10
2.00
501-1,000
16,388
0.37
1.00
1.14
2.04
1,001-4,100
43,850
0.43
1.06
1.24
2.30
4,101-33,000
80,041
0.50
1.10
1.34
2.60
33,001-100,000
30,694
0.58
1.20
1.49
2.80
>100,000
12,426
0.88
2.00
1.86
2.80
Total
223,195
0.46
1.10
1.31
2.50
Temporal/Seasonal Analysis
To assess any potential seasonal variations in the data, EPA calculated the frequency of
detection, by month, for the free and total chlorine residual data associated with total coliform
results for surface water CWSs in the year 2011. Results were generated separately for five bins
of free and total chlorine residual concentrations (Exhibit B. 13 and Exhibit B. 14, respectively):
(1) concentrations equal to 0 mg/L; (2) concentrations greater than 0 and less than or equal to 0.2
mg/L; (3) concentrations greater than 0.2 mg/L and less than or equal to 0.5 mg/L; (4)
concentrations greater than 0.5 mg/L and less than or equal to 1.0 mg/L; and (5) concentrations
greater than 1.0 mg/L.
The free chlorine and total chlorine data (for SW CWSs in 2011) exhibited the same general
seasonal patterns. In all months, the higher the chlorine bin, the larger the number of records of
free and total chlorine, and consequently the percent of total samples. Also, the proportion of
samples in each of the five chlorine bins varied slightly over the course of the year, with a
slightly larger percentage of samples in the middle three chlorine bins in the summer and fall
Six-Year Review 3 Technical Support Document B-15
for Microbial Contaminant Regulations
December 2016
-------
months, and a slightly larger percentage of samples in the largest bin (greater than 1.0 mg/L) in
the winter and early spring months.
EPA also calculated the frequency of detection, by month, for the free and total chlorine residual
data associated with total coliform results for ground water CWSs in the year 2011. Results were
generated separately for five bins of free and total chlorine residual concentrations (Exhibit B.15
and Exhibit B. 16, respectively).
Six-Year Review 3 Technical Support Document B-16
for Microbial Contaminant Regulations
December 2016
-------
Exhibit B.13: Free Chlorine Residual - Frequency of Detection in Surface Water CWSs (in 2011), by Month
Free Chlorine
Bin
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Number of Records
0
848
841
848
822
840
971
815
846
875
573
624
593
>0-0.2
1,566
1,586
1,417
1,418
1,441
1,800
2,011
2,143
2,191
2,115
1,894
1,712
>0.2-0.5
2,126
2,200
2,200
2,456
2,564
3,233
3,511
3,625
3,552
3,241
2,819
2,417
>0.5 -1.0
8,343
8,271
8,729
9,028
9,449
9,730
9,387
9,523
8,950
8,825
8,659
8,565
>1.0
15,301
15,005
15,194
13,860
12,709
11,879
11,647
11,351
11,524
11,989
12,973
13,293
Total
28,184
27,903
28,388
27,584
27,003
27,613
27,371
27,488
27,092
26,743
26,969
26,580
Percent of Total
0
3.0%
3.0%
3.0%
3.0%
3.1%
3.5%
3.0%
3.1%
3.2%
2.1%
2.3%
2.2%
>0-0.2
5.6%
5.7%
5.0%
5.1%
5.3%
6.5%
7.3%
7.8%
8.1%
7.9%
7.0%
6.4%
>0.2-0.5
7.5%
7.9%
7.7%
8.9%
9.5%
11.7%
12.8%
13.2%
13.1%
12.1%
10.5%
9.1%
>0.5 -1.0
29.6%
29.6%
30.7%
32.7%
35.0%
35.2%
34.3%
34.6%
33.0%
33.0%
32.1%
32.2%
>1.0
54.3%
53.8%
53.5%
50.2%
47.1%
43.0%
42.6%
41.3%
42.5%
44.8%
48.1%
50.0%
Total
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
B-17
December 2016
-------
Exhibit B.14: Total Chlorine Residual - Frequency of Detection in Surface Water CWSs (in 2011), by Month
Total
Chlorine Bin
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Number of Records
0
16
14
20
17
9
12
8
24
11
14
9
11
>0-0.2
168
133
160
148
214
334
436
519
544
436
378
331
>0.2-0.5
727
652
725
779
959
1,288
1,384
1,650
1,515
1,341
1,098
949
>0.5 -1.0
3,570
3,372
3,799
4,003
4,289
4,825
4,832
5,003
4,744
4,907
4,517
4,062
>1.0
18,883
18,893
18,449
18,168
17,532
17,346
16,774
17,188
17,517
17,961
18,500
19,115
Total
23,364
23,064
23,153
23,115
23,003
23,805
23,434
24,384
24,331
24,659
24,502
24,468
Percent of Total
0
0.1%
0.1%
0.1%
0.1%
0.0%
0.1%
0.0%
0.1%
0.0%
0.1%
0.0%
0.0%
>0-0.2
0.7%
0.6%
0.7%
0.6%
0.9%
1.4%
1.9%
2.1%
2.2%
1.8%
1.5%
1.4%
>0.2-0.5
3.1%
2.8%
3.1%
3.4%
4.2%
5.4%
5.9%
6.8%
6.2%
5.4%
4.5%
3.9%
>0.5 -1.0
15.3%
14.6%
16.4%
17.3%
18.6%
20.3%
20.6%
20.5%
19.5%
19.9%
18.4%
16.6%
>1.0
80.8%
81.9%
79.7%
78.6%
76.2%
72.9%
71.6%
70.5%
72.0%
72.8%
75.5%
78.1%
Total
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
B-18
December 2016
-------
Exhibit B.15: Free Chlorine Residual - Frequency of Detection in Ground Water CWSs (in 2011), by Month
Free Chlorine
Bin
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Number of Records
0
896
877
914
865
837
803
849
989
980
897
916
803
>0-0.2
2,320
2,169
2,287
2,304
2,342
2,546
2,635
2,798
2,738
2,743
2,644
2,502
>0.2-0.5
5,220
5,108
4,939
5,316
5,444
5,760
6,115
6,505
6,284
6,059
5,795
5,417
>0.5 -1.0
9,974
10,105
10,122
9,737
9,461
9,916
9,920
9,973
9,811
10,139
10,007
10,216
>1.0
9,355
9,286
9,143
9,124
8,856
8,646
8,403
7,830
8,494
8,791
8,826
9,011
Total
27,765
27,545
27,405
27,346
26,940
27,671
27,922
28,095
28,307
28,629
28,188
27,949
Percent of Total
0
3.2%
3.2%
3.3%
3.2%
3.1%
2.9%
3.0%
3.5%
3.5%
3.1%
3.2%
2.9%
>0-0.2
8.4%
7.9%
8.3%
8.4%
8.7%
9.2%
9.4%
10.0%
9.7%
9.6%
9.4%
9.0%
>0.2-0.5
18.8%
18.5%
18.0%
19.4%
20.2%
20.8%
21.9%
23.2%
22.2%
21.2%
20.6%
19.4%
>0.5 -1.0
35.9%
36.7%
36.9%
35.6%
35.1%
35.8%
35.5%
35.5%
34.7%
35.4%
35.5%
36.6%
>1.0
33.7%
33.7%
33.4%
33.4%
32.9%
31.2%
30.1%
27.9%
30.0%
30.7%
31.3%
32.2%
Total
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
B-19
December 2016
-------
Exhibit B.16: Total Chlorine Residual - Frequency of Detection in Ground Water CWSs (in 2011), by Month
Total
Chlorine Bin
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Number of Records
0
86
91
83
88
80
77
77
81
98
110
103
84
>0-0.2
450
457
412
454
475
545
661
665
650
592
508
480
>0.2-0.5
1,724
1,599
1,711
1,702
1,774
1,966
1,981
2,299
2,258
2,140
1,959
1,868
>0.5 -1.0
5,682
5,469
5,710
5,773
5,707
6,129
6,009
6,228
6,256
6,216
6,157
6,043
>1.0
10,185
10,337
10,520
10,090
9,987
10,009
9,917
10,094
9,736
9,942
10,186
10,425
Total
18,127
17,953
18,436
18,107
18,023
18,726
18,645
19,367
18,998
19,000
18,913
18,900
Percent of Total
0
0.5%
0.5%
0.5%
0.5%
0.4%
0.4%
0.4%
0.4%
0.5%
0.6%
0.5%
0.4%
>0-0.2
2.5%
2.5%
2.2%
2.5%
2.6%
2.9%
3.5%
3.4%
3.4%
3.1%
2.7%
2.5%
>0.2-0.5
9.5%
8.9%
9.3%
9.4%
9.8%
10.5%
10.6%
11.9%
11.9%
11.3%
10.4%
9.9%
>0.5 -1.0
31.3%
30.5%
31.0%
31.9%
31.7%
32.7%
32.2%
32.2%
32.9%
32.7%
32.6%
32.0%
>1.0
56.2%
57.6%
57.1%
55.7%
55.4%
53.4%
53.2%
52.1%
51.2%
52.3%
53.9%
55.2%
Total
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
B-20
December 2016
-------
Annual Trends Analysis
To assess any potential trends over the six years of data in the SYR3 ICR microbial dataset, EPA
calculated the frequency of detection, by year, for the free and total chlorine residual data
associated with total coliform results for surface water CWSs. Results were generated separately
for five bins of free and total chlorine residual concentrations (Exhibit B.17 and Exhibit B.18,
respectively).
For free chlorine, the number of samples in each bin tended to increase over the six year period,
with the largest free chlorine bin (greater than 1.0 mg/L) making up an increasingly larger
proportion of all samples over the course of the six years (Exhibit B.17). In the last two years of
data (2010 and 2011), nearly twice as many samples were greater than 1 mg/L compared to 37
percent greater than 1 mg/L in 2006. Interestingly, the percent of samples with reported zero free
chlorine residual increased in 2009 and 2010 (3.8 percent and 3.4 percent, respectively)
compared to 2.7 percent, 2.2 percent, and 2.2 percent for 2006, 2007 and 2008. The percent of
samples in the >0 - 0.2 mg/L bin generally decreased, however, from 2006 - 2011.
Exhibit B.17: Free Chlorine Residual - Frequency of Detection in Surface Water
CWSs, by Year
Free
Chlorine Bin
2006
2007
2008
2009
2010
2011
# Records
0
5,250
5,025
5,009
9,093
10,560
9,496
>0-0.2
16,069
16,277
15,916
18,323
21,117
21,294
>0.2-0.5
29,156
30,133
28,510
28,724
33,198
33,944
>0.5 -1.0
73,428
81,631
77,246
77,141
102,297
107,459
>1.0
71,719
94,717
105,402
106,280
140,426
156,725
Total
195,622
227,783
232,083
239,561
307,598
328,918
% of Total
0
2.7%
2.2%
2.2%
3.8%
3.4%
2.9%
>0-0.2
8.2%
7.1%
6.9%
7.6%
6.9%
6.5%
>0.2-0.5
14.9%
13.2%
12.3%
12.0%
10.8%
10.3%
>0.5 -1.0
37.5%
35.8%
33.3%
32.2%
33.3%
32.7%
>1.0
36.7%
41.6%
45.4%
44.4%
45.7%
47.6%
Total
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
Similar to the free chlorine data, the percent of samples in the largest total chlorine bin increased
slightly between 2006 and 2011 (see Exhibit B.18). In 2010 and 2011, approximately 75 percent
of the samples were in the >1.0 mg/L bin, compared to 65 - 70 percent in the same bin in 2006 -
2008. Unlike the free chlorine data, the percent of total chlorine data in the lowest two bins (0
and >0 - 0.2 mg/L) generally decreased between 2006 and 2011.
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
B-21
December 2016
-------
Exhibit B.18: Total Chlorine Residual - Frequency of Detection in Surface Water
CWSs, by Year
Total
Chlorine Bin
2006
2007
2008
2009
2010
2011
# Records
0
289
447
175
211
260
165
>0-0.2
2,547
2,982
3,755
3,597
3,285
3,801
>0.2-0.5
7,155
9,030
10,316
10,343
12,218
13,067
>0.5 -1.0
24,935
28,945
33,211
36,204
49,612
51,923
>1.0
80,325
82,360
87,180
125,786
192,629
216,326
Total
115,251
123,764
134,637
176,141
258,004
285,282
% of Total
0
0.3%
0.4%
0.1%
0.1%
0.1%
0.1%
>0-0.2
2.2%
2.4%
2.8%
2.0%
1.3%
1.3%
>0.2-0.5
6.2%
7.3%
7.7%
5.9%
4.7%
4.6%
>0.5 -1.0
21.6%
23.4%
24.7%
20.6%
19.2%
18.2%
>1.0
69.7%
66.5%
64.8%
71.4%
74.7%
75.8%
Total
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
EPA also calculated the frequency of detection, by year, for the free and total chlorine residual
data associated with total coliform results for ground water CWSs. Results were generated
separately for five bins of free and total chlorine residual concentrations (Exhibit B.19 and
Exhibit B.20, respectively).
Exhibit B.19: Free Chlorine Residual - Frequency of Detection in Ground Water
CWSs, by Year
Free
Chlorine Bin
2006
2007
2008
2009
2010
2011
# Records
0
9,937
10,218
10,815
16,162
11,917
10,626
>0-0.2
28,092
29,636
29,625
29,475
29,528
30,028
>0.2-0.5
52,970
55,819
54,226
64,209
65,582
67,962
>0.5 -1.0
64,590
69,679
68,391
88,897
110,544
119,381
>1.0
30,081
30,856
32,888
72,616
93,077
105,765
Total
185,670
196,208
195,945
271,359
310,648
333,762
% of Total
0
5.4%
5.2%
5.5%
6.0%
3.8%
3.2%
>0-0.2
15.1%
15.1%
15.1%
10.9%
9.5%
9.0%
>0.2-0.5
28.5%
28.4%
27.7%
23.7%
21.1%
20.4%
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
B-22
December 2016
-------
Free
Chlorine Bin
2006
2007
2008
2009
2010
2011
>0.5 -1.0
34.8%
35.5%
34.9%
32.8%
35.6%
35.8%
>1.0
16.2%
15.7%
16.8%
26.8%
30.0%
31.7%
Total
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
Exhibit B.20: Total Chlorine Residual - Frequency of Detection in Ground Water
CWSs, by Year
Total
Chlorine Bin
2006
2007
2008
2009
2010
2011
# Records
0
1,488
1,414
1,516
1,578
1,199
1,058
>0-0.2
4,446
4,756
5,466
5,509
6,479
6,349
>0.2-0.5
18,477
18,614
19,543
19,619
22,394
22,981
>0.5 -1.0
36,871
43,022
47,662
52,884
67,420
71,379
>1.0
34,662
42,505
47,838
76,177
109,889
121,428
Total
95,944
110,311
122,025
155,767
207,381
223,195
% of Total
0
1.6%
1.3%
1.2%
1.0%
0.6%
0.5%
>0-0.2
4.6%
4.3%
4.5%
3.5%
3.1%
2.8%
>0.2-0.5
19.3%
16.9%
16.0%
12.6%
10.8%
10.3%
>0.5 -1.0
38.4%
39.0%
39.1%
34.0%
32.5%
32.0%
>1.0
36.1%
38.5%
39.2%
48.9%
53.0%
54.4%
Total
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
Geographic Analysis
To assess any potential geographic trends in the data, EPA calculated the frequency of detection,
by state, for the free and total chlorine residual data associated with total coliform results for
surface water CWSs in 2011. Results were generated separately for five bins of free and total
chlorine residual concentrations (Exhibit B.21 and Exhibit B.22, respectively). Similar tables for
ground water results are presented in Exhibit B.23 and Exhibit B.24.
Twenty states currently have requirements for minimum free chlorine residual in the distribution
system; these requirements apply to systems using free chlorine. Twelve of the 20 states have
data in the SYR3 ICR microbial dataset. Illinois and Iowa require a minimum of 0.3 mg/L; and
10 states require a minimum of 0.2 mg/L free chlorine residual (Indiana, Kansas, Kentucky,
Missouri, North Carolina, Nebraska, Ohio, Oklahoma, Texas and West Virginia). State
requirements for minimum total chlorine residual also vary; the following states require a 1.00
mg/L or higher minimum total chlorine residual (Kansas, Oklahoma, Iowa, Ohio and North
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
B-23
December 2016
-------
Carolina). Total chlorine residual requirements apply to systems using chloramines, but may also
apply to systems using free chlorine.
There is a wide range of records for each state that makes it difficult to make comparisons. As
shown in Exhibit B.21, the four states/entities with the largest number of SW CWSs submitting
free chlorine data with their total coliform samples in 2011 were North Carolina, Oklahoma,
Texas and Virginia. All four states had more than 300 systems with data. Two of these four
states also provided the most samples overall for free chlorine (North Carolina and Virginia).
There are no free chlorine data from Nebraska and Arkansas, and few data from American
Samoa, Hawaii, Kansas, Navajo Nation and Rhode Island. American Samoa reported only 20
results for 5 systems, and all were in the "0 mg/L" bin.
Iowa and Texas had the most samples in the 0 mg/L bin, with 11.5 percent and 42.9 percent
respectively. These states also had a high percentage of samples in the > 0 - 0.2 mg/L bin
compared to other states. (Currently, the States of Iowa and Texas require a minimum free
chlorine residual in the distribution system of 0.3 mg/L and 0.2 mg/L, respectively.) States with a
low percentage of samples in the 0 bin but high percent in the > 0 - 0.2 mg/L include Alaska,
Iowa, Illinois, Kansas, Missouri, Navajo Nation, Nevada, New York, Rhode Island, Vermont and
Region 5 and 8 Tribes. (Four of these states require a minimum free chlorine residual in the
distribution system; Iowa and Illinois require a minimum residual of 0.3 mg/L while Kansas and
Missouri require a minimum residual of 0.2 mg/L.) The reason for high occurrence of low free
chlorine residual sample results in states that have a minimum requirement is unclear. It is
possible that the minimum residual requirements came after 2011. It is also possible that systems
in those states using chloramines reported free chlorine data along with total chlorine data.
For total chlorine (Exhibit B.22), the four states/entities with the largest number of SW CWSs
submitting total chlorine data with their total coliform samples in 2011 were Illinois, Kansas,
Texas and West Virginia. All four states had more than 280 systems with data. The three states
submitting the most samples overall for total chlorine were Illinois, Ohio and Texas. There were
no total chlorine data from Region 4 or Region 5 Tribes, as well as Alaska, American Samoa,
Idaho, Navajo Nation, Oregon and Virginia. Several other states/entities provided very few total
chlorine residual results.
Within the three states with the most data (IL, OH and TX), the majority of samples (almost 65
percent in IL, more than 71 percent in OH, and almost 95 percent in TX) had total chlorine
concentrations that were greater than 1 mg/L. (The State of Ohio requires a minimum total
chlorine residual in the distribution system of 1 mg/L.) These states had relatively low
percentages of samples in the >0 - 0.2 mg/L bin compared to some other states. New York and
Region 8 Tribes had the largest percentage of samples in the > 0 - 0.2 mg/L bin; however, these
percentages were based on a low number of samples overall (36 samples and 52 samples for
New York and Region 8 Tribes, respectively).
EPA also calculated the frequency of detection, by state, for the free and total chlorine residual
data associated with total coliform results for ground water CWSs in 2011. Results were
generated separately for five bins of free and total chlorine residual concentrations (Exhibit B.23
and Exhibit B.24).
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
B-24
December 2016
-------
Exhibit B.21: Free Chlorine Residual - Frequency of Detection in Surface Water CWSs (in 2011), by State
State12
No. of
Systems
Number of Records Within Each Free Chlorine Bin
Percent of Records Within Each Free Chlorine Bin
Total
0
>0 - 0.2
>0.2-0.5
>0.5-1.0
>1.0
Total
0
>0 - 0.2
>0.2-0.5
>0.5-1.0
>1.0
AK
110
2,742
4
433
1,278
775
252
100.0%
0.1%
15.8%
46.6%
28.3%
9.2%
AR
0
0
0
0
0
0
0
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
AS
5
20
20
0
0
0
0
100.0%
100.0%
0.0%
0.0%
0.0%
0.0%
CT
58
26,286
207
2,375
5,331
11,274
7,099
100.0%
0.8%
9.0%
20.3%
42.9%
27.0%
HI
3
35
0
0
18
16
1
100.0%
0.0%
0.0%
51.4%
45.7%
2.9%
IA
82
11,148
1,280
2,509
607
3,727
3,025
100.0%
11.5%
22.5%
5.4%
33.4%
27.1%
ID
64
1,754
2
142
435
750
425
100.0%
0.1%
8.1%
24.8%
42.8%
24.2%
IL
260
30,168
471
5,664
4,786
15,349
3,898
100.0%
1.6%
18.8%
15.9%
50.9%
12.9%
IN
53
1,208
0
46
176
558
428
100.0%
0.0%
3.8%
14.6%
46.2%
35.4%
KS
3
7
0
1
0
3
3
100.0%
0.0%
14.3%
0.0%
42.9%
42.9%
KY
187
29,390
2
107
960
7,411
20,910
100.0%
0.0%
0.4%
3.3%
25.2%
71.1%
MO
160
6,243
2
691
631
1,523
3,396
100.0%
0.0%
11.1%
10.1%
24.4%
54.4%
MT
52
3,212
0
46
399
1,648
1,119
100.0%
0.0%
1.4%
12.4%
51.3%
34.8%
NC
321
37,696
37
1,458
2,380
11,406
22,415
100.0%
0.1%
3.9%
6.3%
30.3%
59.5%
NE
0
0
0
0
0
0
0
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
NM
34
5,950
10
528
1,004
3,202
1,206
100.0%
0.2%
8.9%
16.9%
53.8%
20.3%
NN
7
310
0
32
119
122
37
100.0%
0.0%
10.3%
38.4%
39.4%
11.9%
NV
13
465
0
49
168
213
35
100.0%
0.0%
10.5%
36.1%
45.8%
7.5%
NY
239
3,778
42
409
1,105
1,418
804
100.0%
1.1%
10.8%
29.2%
37.5%
21.3%
OH
236
51,039
19
1,216
3,583
19,003
27,218
100.0%
0.0%
2.4%
7.0%
37.2%
53.3%
OK
366
14,245
388
528
1,489
3,089
8,751
100.0%
2.7%
3.7%
10.5%
21.7%
61.4%
OR
205
21,547
16
493
4,370
12,639
4,029
100.0%
0.1%
2.3%
20.3%
58.7%
18.7%
Rl
8
331
1
116
1
68
145
100.0%
0.3%
35.0%
0.3%
20.5%
43.8%
TX
356
16,131
6,924
2,742
837
1,155
4,473
100.0%
42.9%
17.0%
5.2%
7.2%
27.7%
VA
330
54,745
65
808
2,308
7,591
43,973
100.0%
0.1%
1.5%
4.2%
13.9%
80.3%
VT
67
1,965
0
253
575
829
308
100.0%
0.0%
12.9%
29.3%
42.2%
15.7%
WV
71
1,270
0
13
84
425
748
100.0%
0.0%
1.0%
6.6%
33.5%
58.9%
WY
103
5,056
4
445
831
2,214
1,562
100.0%
0.1%
8.8%
16.4%
43.8%
30.9%
Tribes - 01
2
1,091
0
77
166
686
162
100.0%
0.0%
7.1%
15.2%
62.9%
14.8%
Tribes - 04
1
37
0
0
0
11
26
100.0%
0.0%
0.0%
0.0%
29.7%
70.3%
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
B-25
December 2016
-------
State12
No. of
Systems
Number of Records Within Each Free Chlorine Bin
Percent of Records Within Each Free Chlorine Bin
Total
0
>0 - 0.2
>0.2-0.5
>0.5-1.0
>1.0
Total
0
>0 - 0.2
>0.2-0.5
>0.5-1.0
>1.0
Tribes - 05
2
25
0
12
3
6
4
100.0%
0.0%
48.0%
12.0%
24.0%
16.0%
Tribes - 08
24
613
2
70
143
140
258
100.0%
0.3%
11.4%
23.3%
22.8%
42.1%
Tribes - 09
12
411
0
31
157
208
15
100.0%
0.0%
7.5%
38.2%
50.6%
3.6%
Total
3,434
328,918
9,496
21,294
33,944
107,459
156,725
100.0%
2.9%
6.5%
10.3%
32.7%
47.6%
1 This column presents the standard 2-letter state abbreviations with the exception of "AS" for American Samoa and "NN" for Navajo Nation.
2 All states/entities that provided any free and/or total chlorine data are listed in this table. A few states/entities submitted only free or only total chlorine data; thus, their total number of
systems with data in this table is listed as zero.
Exhibit B.22: Total Chlorine Residual - Frequency of Detection in Surface Water CWSs (in 2011), by State
State12
No. of
Systems
Number of Records Within Each Total Chlorine Bin
Percent of Records Within Each Total Chlorine Bin
Total
0
>0 - 0.2
>0.2-0.5
>0.5-1.0
>1.0
Total
0
>0 - 0.2
>0.2-0.5
>0.5-1.0
>1.0
AK
0
0
0
0
0
0
0
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
AR
268
22,130
3
1,938
4,654
8,878
6,657
100.0%
0.0%
8.8%
21.0%
40.1%
30.1%
AS
0
0
0
0
0
0
0
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
CT
12
3,370
7
290
218
500
2,355
100.0%
0.2%
8.6%
6.5%
14.8%
69.9%
HI
1
9
0
1
1
0
7
100.0%
0.0%
11.1%
11.1%
0.0%
77.8%
IA
89
14,439
0
50
370
2,817
11,202
100.0%
0.0%
0.3%
2.6%
19.5%
77.6%
ID
0
0
0
0
0
0
0
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
IL
328
36,344
4
267
1,964
10,620
23,489
100.0%
0.0%
0.7%
5.4%
29.2%
64.6%
IN
66
2,091
1
106
260
447
1,277
100.0%
0.0%
5.1%
12.4%
21.4%
61.1%
KS
282
11,428
8
94
159
541
10,626
100.0%
0.1%
0.8%
1.4%
4.7%
93.0%
KY
74
14,836
0
6
122
1,277
13,431
100.0%
0.0%
0.0%
0.8%
8.6%
90.5%
MO
199
15,148
0
29
182
938
13,999
100.0%
0.0%
0.2%
1.2%
6.2%
92.4%
MT
17
1,587
0
70
166
764
587
100.0%
0.0%
4.4%
10.5%
48.1%
37.0%
NC
152
21,656
6
104
243
679
20,624
100.0%
0.0%
0.5%
1.1%
3.1%
95.2%
NE
2
3,703
0
0
0
1
3,702
100.0%
0.0%
0.0%
0.0%
0.0%
100.0%
NM
10
3,119
0
43
177
2,227
672
100.0%
0.0%
1.4%
5.7%
71.4%
21.5%
NN
0
0
0
0
0
0
0
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
NV
1
36
0
0
1
29
6
100.0%
0.0%
0.0%
2.8%
80.6%
16.7%
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
B-26
December 2016
-------
State12
No. of
Systems
Number of Records Within Each Total Chlorine Bin
Percent of Records Within Each Total Chlorine Bin
Total
0
>0 - 0.2
>0.2-0.5
>0.5-1.0
>1.0
Total
0
>0 - 0.2
>0.2-0.5
>0.5-1.0
>1.0
NY
5
36
1
10
7
18
0
100.0%
2.8%
27.8%
19.4%
50.0%
0.0%
OH
247
54,550
7
114
2,068
13,491
38,870
100.0%
0.0%
0.2%
3.8%
24.7%
71.3%
OK
255
15,852
80
397
1,257
2,591
11,527
100.0%
0.5%
2.5%
7.9%
16.3%
72.7%
OR
0
0
0
0
0
0
0
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Rl
4
14
0
6
4
4
0
100.0%
0.0%
42.9%
28.6%
28.6%
0.0%
TX
495
45,928
44
15
183
2,092
43,594
100.0%
0.1%
0.0%
0.4%
4.6%
94.9%
VA
0
0
0
0
0
0
0
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
VT
42
1,785
0
39
167
392
1,187
100.0%
0.0%
2.2%
9.4%
22.0%
66.5%
WV
293
16,139
2
133
794
3,355
11,855
100.0%
0.0%
0.8%
4.9%
20.8%
73.5%
WY
18
948
0
80
62
192
614
100.0%
0.0%
8.4%
6.5%
20.3%
64.8%
Tribes - 01
1
81
0
8
5
66
2
100.0%
0.0%
9.9%
6.2%
81.5%
2.5%
Tribes - 04
0
0
0
0
0
0
0
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Tribes - 05
0
0
0
0
0
0
0
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Tribes - 08
7
52
2
1
3
3
43
100.0%
3.8%
1.9%
5.8%
5.8%
82.7%
Tribes - 09
1
1
0
0
0
1
0
100.0%
0.0%
0.0%
0.0%
100.0%
0.0%
Total
2,869
285,282
165
3,801
13,067
51,923
216,326
100.0%
0.1%
1.3%
4.6%
18.2%
75.8%
1 This column presents the standard 2-letter state abbreviations with the exception of "AS" for American Samoa and "NN" for Navajo Nation.
2 All states/entities that provided any free and/or total chlorine data are listed in this table. A few states/entities submitted only free or only total chlorine data; thus, their total number of
systems with data in this table is listed as zero.
Exhibit B.23: Free Chlorine Residual - Frequency of Detection in Ground Water CWSs (in 2011), by State
State12
No. of
Systems
Number of Records Within Each Free Chlorine Bin
Percent of Records Within Each Free Chlorine Bin
Total
0
>0 - 0.2
>0.2-0.5
>0.5-1.0
>1.0
Total
0
>0 - 0.2
>0.2-0.5
>0.5-1.0
>1.0
AK
107
2,381
22
900
673
520
266
100.0%
0.9%
37.8%
28.3%
21.8%
11.2%
AR
0
0
0
0
0
0
0
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
AS
11
1,086
60
130
311
505
80
100.0%
5.5%
12.0%
28.6%
46.5%
7.4%
CT
486
9,115
3,578
947
2,109
2,017
464
100.0%
39.3%
10.4%
23.1%
22.1%
5.1%
HI
37
740
5
206
315
195
19
100.0%
0.7%
27.8%
42.6%
26.4%
2.6%
IA
842
25,882
580
2,580
4,754
10,696
7,272
100.0%
2.2%
10.0%
18.4%
41.3%
28.1%
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
B-27
December 2016
-------
State12
No. of
Systems
Number of Records Within Each Free Chlorine Bin
Percent of Records Within Each Free Chlorine Bin
Total
0
>0 - 0.2
>0.2-0.5
>0.5-1.0
>1.0
Total
0
>0 - 0.2
>0.2-0.5
>0.5-1.0
>1.0
ID
270
8,429
425
3,210
3,639
878
277
100.0%
5.0%
38.1%
43.2%
10.4%
3.3%
IL
1,016
59,121
238
5,326
16,732
28,398
8,427
100.0%
0.4%
9.0%
28.3%
48.0%
14.3%
IN
314
5,346
6
462
1,727
2,453
698
100.0%
0.1%
8.6%
32.3%
45.9%
13.1%
KS
7
436
0
1
53
272
110
100.0%
0.0%
0.2%
12.2%
62.4%
25.2%
KY
56
4,491
0
15
123
1,771
2,582
100.0%
0.0%
0.3%
2.7%
39.4%
57.5%
MO
737
22,136
1
2,427
2,591
6,892
10,225
100.0%
0.0%
11.0%
11.7%
31.1%
46.2%
MT
131
2,168
6
239
1,135
594
194
100.0%
0.3%
11.0%
52.4%
27.4%
8.9%
NC
1,332
24,511
395
1,221
4,409
9,676
8,810
100.0%
1.6%
5.0%
18.0%
39.5%
35.9%
NE
0
0
0
0
0
0
0
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
NM
369
9,566
711
1,236
3,442
3,117
1,060
100.0%
7.4%
12.9%
36.0%
32.6%
11.1%
NN
119
2,284
4
651
711
690
228
100.0%
0.2%
28.5%
31.1%
30.2%
10.0%
NV
92
2,728
60
424
774
1,295
175
100.0%
2.2%
15.5%
28.4%
47.5%
6.4%
NY
509
5,809
280
538
1,890
1,943
1,158
100.0%
4.8%
9.3%
32.5%
33.4%
19.9%
OH
692
35,920
67
1,682
5,104
19,025
10,042
100.0%
0.2%
4.7%
14.2%
53.0%
28.0%
OK
363
7,321
323
642
2,163
2,450
1,743
100.0%
4.4%
8.8%
29.5%
33.5%
23.8%
OR
317
12,825
66
2,071
3,324
4,490
2,874
100.0%
0.5%
16.1%
25.9%
35.0%
22.4%
Rl
38
386
261
86
39
0
0
100.0%
67.6%
22.3%
10.1%
0.0%
0.0%
TX
2,804
62,686
2,380
219
4,165
13,656
42,266
100.0%
3.8%
0.3%
6.6%
21.8%
67.4%
VA
660
13,270
623
982
2,675
3,830
5,160
100.0%
4.7%
7.4%
20.2%
28.9%
38.9%
VT
255
3,938
110
1,647
1,485
594
102
100.0%
2.8%
41.8%
37.7%
15.1%
2.6%
WV
52
776
0
4
240
286
246
100.0%
0.0%
0.5%
30.9%
36.9%
31.7%
WY
164
3,108
384
611
985
936
192
100.0%
12.4%
19.7%
31.7%
30.1%
6.2%
Tribes - 01
0
0
0
0
0
0
0
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Tribes - 04
9
209
0
0
7
77
125
100.0%
0.0%
0.0%
3.3%
36.8%
59.8%
Tribes - 05
67
1,733
22
566
757
242
146
100.0%
1.3%
32.7%
43.7%
14.0%
8.4%
Tribes - 08
63
1,235
19
272
405
377
162
100.0%
1.5%
22.0%
32.8%
30.5%
13.1%
Tribes - 09
166
4,126
0
733
1,225
1,506
662
100.0%
0.0%
17.8%
29.7%
36.5%
16.0%
Total
12,085
333,762
10,626
30,028
67,962
119,381
105,765
100.0%
3.2%
9.0%
20.4%
35.8%
31.7%
1 This column presents the standard 2-letter state abbreviations with the exception of "AS" for American Samoa and "NN" for Navajo Nation.
2 All states that provided any free and/or total chlorine data are listed in this table. A few states submitted only free or only total chlorine data; thus, their total number of systems with
data in this table is listed as zero.
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
B-28
December 2016
-------
Exhibit B.24: Total Chlorine Residual - Frequency of Detection in Ground Water CWSs (in 2011), by State
State12
No. of
Systems
Number of Records Within Each Total Chlorine Bin
Percent of Records Within Each Total Chlorine Bin
Total
0
>0 - 0.2
>0.2-0.5
>0.5-1.0
>1.0
Total
0
>0 - 0.2
>0.2-0.5
>0.5-1.0
>1.0
AK
0
0
0
0
0
0
0
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
AR
438
22,662
6
2,413
5,536
9,955
4,752
100.0%
0.0%
10.6%
24.4%
43.9%
21.0%
AS
0
0
0
0
0
0
0
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
CT
12
153
13
94
30
15
1
100.0%
8.5%
61.4%
19.6%
9.8%
0.7%
HI
1
4
0
0
2
2
0
100.0%
0.0%
0.0%
50.0%
50.0%
0.0%
IA
872
28,190
156
217
1,504
9,167
17,146
100.0%
0.6%
0.8%
5.3%
32.5%
60.8%
ID
1
12
0
0
11
1
0
100.0%
0.0%
0.0%
91.7%
8.3%
0.0%
IL
861
24,068
28
463
2,376
8,794
12,407
100.0%
0.1%
1.9%
9.9%
36.5%
51.5%
IN
300
7,748
25
524
2,202
3,668
1,329
100.0%
0.3%
6.8%
28.4%
47.3%
17.2%
KS
581
19,392
13
739
1,971
5,594
11,075
100.0%
0.1%
3.8%
10.2%
28.8%
57.1%
KY
9
356
0
0
3
150
203
100.0%
0.0%
0.0%
0.8%
42.1%
57.0%
MO
746
23,669
0
427
1,206
5,397
16,639
100.0%
0.0%
1.8%
5.1%
22.8%
70.3%
MT
34
464
8
37
289
102
28
100.0%
1.7%
8.0%
62.3%
22.0%
6.0%
NC
328
6,573
322
69
372
1,309
4,501
100.0%
4.9%
1.0%
5.7%
19.9%
68.5%
NE
0
0
0
0
0
0
0
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
NM
129
4,951
43
673
2,260
1,606
369
100.0%
0.9%
13.6%
45.6%
32.4%
7.5%
NN
0
0
0
0
0
0
0
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
NV
16
360
0
3
41
262
54
100.0%
0.0%
0.8%
11.4%
72.8%
15.0%
NY
13
19
14
5
0
0
0
100.0%
73.7%
26.3%
0.0%
0.0%
0.0%
OH
722
35,900
61
170
2,965
16,955
15,749
100.0%
0.2%
0.5%
8.3%
47.2%
43.9%
OK
147
2,755
80
125
518
784
1,248
100.0%
2.9%
4.5%
18.8%
28.5%
45.3%
OR
0
0
0
0
0
0
0
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Rl
18
162
114
41
4
3
0
100.0%
70.4%
25.3%
2.5%
1.9%
0.0%
TX
1,028
40,851
103
86
797
5,696
34,169
100.0%
0.3%
0.2%
2.0%
13.9%
83.6%
VA
2
3
0
1
2
0
0
100.0%
0.0%
33.3%
66.7%
0.0%
0.0%
VT
84
670
57
220
229
106
58
100.0%
8.5%
32.8%
34.2%
15.8%
8.7%
WV
179
4,118
1
30
628
1,778
1,681
100.0%
0.0%
0.7%
15.3%
43.2%
40.8%
WY
4
6
0
0
5
1
0
100.0%
0.0%
0.0%
83.3%
16.7%
0.0%
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
B-29
December 2016
-------
State12
No. of
Systems
Number of Records Within Each Total Chlorine Bin
Percent of Records Within Each Total Chlorine Bin
Total
0
>0 - 0.2
>0.2-0.5
>0.5-1.0
>1.0
Total
0
>0 - 0.2
>0.2-0.5
>0.5-1.0
>1.0
Tribes - 01
0
0
0
0
0
0
0
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Tribes - 04
0
0
0
0
0
0
0
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Tribes - 05
0
0
0
0
0
0
0
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Tribes - 08
15
100
14
11
28
28
19
100.0%
14.0%
11.0%
28.0%
28.0%
19.0%
Tribes - 09
8
9
0
1
2
6
0
100.0%
0.0%
11.1%
22.2%
66.7%
0.0%
Total
6,548
223,195
1,058
6,349
22,981
71,379
121,428
100.0%
0.5%
2.8%
10.3%
32.0%
54.4%
1 This column presents the standard 2-letter state abbreviations with the exception of "AS" for American Samoa and "NN" for Navajo Nation.
2 All states that provided any free and/or total chlorine data are listed in this table. A few states submitted only free or only total chlorine data; thus, their total number of systems with
data in this table is listed as zero.
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
B-30
December 2016
-------
Appendix C. Additional Analyses on the Occurrence of TC+ and EC+ in
Surface Water and Ground Water Systems Compared to Disinfectant
Residuals in Distribution Systems
This appendix provides the analytical results for surface water and ground water systems that
were not presented within the body of the chapter in Section 6.3, related to the occurrence of
TC+ and EC+ results compared to disinfectant residuals in distribution systems. This appendix
includes an evaluation of the occurrence relative to system type and system size, as well as
seasonal changes, annual trends and geographic distribution. All analyses in this section are
based on routine samples taken in the distribution system.
System Type
Exhibit C.l and Exhibit C.2 present the frequency of detection of total coliforms over the six
years of data in community water systems (CWSs) and non-community water systems (NCWSs;
includes non-transient non-community and transient non-community water systems),
respectively, that were served by surface water. Results were generated separately for five bins
of free and total chlorine residual concentrations.
For free chlorine, a higher percentage of samples were TC+ when the residual was 0 and >0 -
0.2 mg/L for NCWSs (1.7 percent and 1.9 percent, respectively), compared to CWSs (0.4 percent
and 0.5 percent, respectively). The percent TC+ results for the lower free chlorine bins were
obscured by the records that reported zero or very low free chlorine but high total chlorine values
(e.g., in a chloramine system) (see Sections 6.3.1 and 6.3.3), particularly for CWSs using surface
water. These CWSs are more likely to use chloramines than NCWSs. For total chlorine, percent
positive total coliform results were slightly higher for CWSs compared to NCWSs. It is difficult
to draw conclusions regarding relative occurrence, however, because of the smaller sample size
of the NCWS dataset (see Exhibit C.5).
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-l
December 2016
-------
Exhibit C.1: Total Coliforms - Frequency of Detection in Surface Water CWSs
(2006-2011)
2.5%
¦ 0 mg/L
¦ > 0 to 0.2 mg/L
Free Chlorine Residual Total Chlorine Residual
Residual
Exhibit C.2: Total Coliforms - Frequency of Detection in Surface Water NCWSs
(2006-2011)
-) CO/
Z-. -J/O
-) no/
¦ 0 mg/L
¦ > 0 to 0.2 mg/L
¦ > 0.2 to 0.5 mg/L
1 > 0.5 to 1.0 mg/L
¦ > 1.0 mg/L
Z..W/0
¦
1
¦
±.D /0
(/)
0)
>
to 1 fW
1
¦
0 l.U/o
Q_
U
I—
NO
0 s %
llll—
U.J /o
o no/
III
w. w /o
Free Chlorine Residual
Residual
Total Chlorine Residual
Exhibit C.3 and Exhibit C.4 present the frequency of detection of E. coli over the six years of
data in surface water CWSs and NCWSs, respectively. For free chlorine, the rate of EC+ was
higher for surface water NCWSs compared to CWSs for all disinfectant residual bins. For total
chlorine, there were no EC+ sample results in NCWSs when total chlorine was equal to zero (or
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-2
December 2016
-------
"below detection limit"), but 0.35 percent and 0.26 percent EC+ results when total chlorine was
greater than 0 - 0.2 mg/L, and greater than 0.2 mg/L - 0.5 mg/L respectively. It is difficult to
draw conclusions regarding relative occurrence, however, because of the smaller sample size of
the NCWS data (see Exhibit C.5). EC+ rates were higher for NCWSs than for CWSs for all
disinfectant residual bins when NCWS data were available.
Exhibit C.3: E. coli - Frequency of Detection in Surface Water CWSs (2006-2011)
0.50%
0.45%
0.40%
0.35%
v 0.30%
>
% 0.25%
o
Q_
(1 n ?n°>£
¦ 0 mg/L
¦
¦ > 0 to 0.2 mg/L
¦ > 0.2 to 0.5 mg/L
i > 0.5 to 1.0 mg/L
¦ > 1.0 mg/L
w U.ZU/O
LU
NO
°N H1
U. -LD /0
0.10%
0.05%
0.00%
Free Chlorine Residual
Residual
Total Chlorine Residual
Exhibit C.4: E. coli - Frequency of Detection in Surface Water NCWSs (2006-2011)
0.50%
10 mg/L
0-45% B>0 to 0.2 mg/L
0.40% ¦ > 0.2 to 0.5 mg/L
0.35% > 0.5 to 1.0 mg/L
0.30% ¦ > 1-0 mg/L
ju 0.25%
g 0.20%
Q.
y 0.15%
° 0.10%
0.05%
0.00%
Free Chlorine Residual Total Chlorine Residual
Residual
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-3
December 2016
-------
Exhibit C.5: Number of Total Coliform Surface Water Samples Paired with Free
and Total Chlorine Data, by System Type (underlying data/denominator for
Exhibit C.1, Exhibit C.2, Exhibit C.3 and Exhibit C.4)
System
Type
Group
Chlorine Bin
Total Coliforms
E. coli
% Positive
# Positive
Samples
% Positive
# Positive
Samples
%
Positive
CWSs
Free Chlorine
0
44,396
170
0.38%
14
0.03%
> 0 to 0.2 mg/L
108,975
576
0.53%
24
0.02%
> 0.2 to 0.5 mg/L
183,651
817
0.44%
38
0.02%
> 0.5 to 1 mg/L
519,173
1,416
0.27%
77
0.01%
> 1 mg/L
675,250
2,089
0.31%
98
0.01%
Total
1,531,445
5,068
0.33%
251
0.02%
Total Chlorine
0
1,535
30
1.95%
7
0.46%
> 0 to 0.2 mg/L
19,966
416
2.08%
18
0.09%
> 0.2 to 0.5 mg/L
62,127
534
0.86%
15
0.02%
> 0.5 to 1 mg/L
224,826
910
0.40%
46
0.02%
> 1 mg/L
784,585
2,489
0.32%
111
0.01%
Total
1,093,039
4,379
0.40%
197
0.02%
NCWSs
Free Chlorine
0
1,777
31
1.74%
2
0.11%
> 0 to 0.2 mg/L
4,894
92
1.88%
5
0.10%
> 0.2 to 0.5 mg/L
8,171
66
0.81%
8
0.10%
> 0.5 to 1 mg/L
9,640
60
0.62%
5
0.05%
> 1 mg/L
15,327
96
0.63%
9
0.06%
Total
39,809
345
0.87%
29
0.07%
Total Chlorine
0
184
3
1.63%
0
0.00%
> 0 to 0.2 mg/L
565
10
1.77%
2
0.35%
> 0.2 to 0.5 mg/L
1,155
14
1.21%
3
0.26%
> 0.5 to 1 mg/L
1,811
8
0.44%
0
0.00%
> 1 mg/L
5,910
36
0.61%
2
0.03%
Total
9,625
71
0.74%
7
0.07%
Exhibit C.6 and Exhibit C.7 present the frequency of detection of total coliforms over the six
years of data in ground water CWSs and NCWSs, respectively.
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-4
December 2016
-------
Exhibit C.6: Total Coliforms - Frequency of Detection in Ground Water CWSs
(2006-2011)
4.5%
4.0%
3.5%
3.0%
jS 2.5%
£ 2.0%
u
h-
55 1.5%
1.0%
0.5%
0.0%
10 mg/L
I > 0 to 0.2 mg/L
> 0.2 to 0.5 mg/L
> 0.5 to 1.0 mg/L
I > 1.0 mg/L
!¦¦¦¦ 1
1
Free Chlorine Residual
Total Chlorine Residual
Residual
Exhibit C.7: Total Coliforms - Frequency of Detection in Ground Water NCWSs
(2006-2011)
4.5%
4.0%
¦ 0 mg/L
¦ > 0 to 0.2 mg/L
¦ > 0.2 to 0.5 mg/L
1 > 0.5 to 1.0 mg/L
3.5%
3.0%
1
1
¦ > 1.0 mg/L
a! 2.5%
>
£ 2.0%
u
1
¦
55 1.5%
1.0%
0.5%
n no/
III
1
III
w.w/o
Free C
hlorine Residual
Residual
Total C
:hlorine Residual
Exhibit C.8 and Exhibit C.9 present the frequency of detection of E. coli over the six years of
data in ground water CWSs and NCWSs.
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-5
December 2016
-------
Exhibit C.8: E. coli - Frequency of Detection in Ground Water CWSs (2006-2011)
0.35%
0.30%
0.25%
> 0.20%
o
0.15%
u
LU
SO
0.10%
0.05%
0.00%
I
Free Chlorine Residual
10 mg/L
I > 0 to 0.2 mg/L
I > 0.2 to 0.5 mg/L
> 0.5 to 1.0 mg/L
I > 1.0 mg/L
¦—
Total Chlorine Residual
Residual
Exhibit C.9: E. coli - Frequency of Detection in Ground Water NCWSs (2006-2011)
0.35%
0.30%
0.25%
$ 0.20%
>
O 0.15%
D.
U
LU
s? 0.10%
0.05%
0.00%
10 mg/L
I > 0 to 0.2 mg/L
I > 0.2 to 0.5 mg/L
> 0.5 to 1.0 mg/L
I > 1.0 mg/L
lllll
Free Chlorine Residual
Total Chlorine Residual
Residual
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-6
December 2016
-------
Exhibit C.10: Number of Total Coliform Ground Water Samples Paired with Free
and Total Chlorine Data, by System Type (underlying data/denominator for
Exhibit C.6, Exhibit C.7, Exhibit C.8 and Exhibit C.9)
System
Type
Group
Chlorine Bin
Total Coliforms
E. coli
% Positive
# Positive
Samples
% Positive
# Positive
Samples
%
Positive
CWSs
Free Chlorine
0
69,658
1,478
2.12%
58
0.08%
> 0 to 0.2 mg/L
176,170
1,733
0.98%
83
0.05%
> 0.2 to 0.5 mg/L
360,672
2,088
0.58%
119
0.03%
> 0.5 to 1 mg/L
521,400
2,243
0.43%
83
0.02%
> 1 mg/L
365,256
2,210
0.61%
88
0.02%
Total
1,493,156
9,752
0.65%
431
0.03%
Total Chlorine
0
8,252
156
1.89%
25
0.30%
> 0 to 0.2 mg/L
32,999
676
2.05%
33
0.10%
> 0.2 to 0.5 mg/L
121,624
1,099
0.90%
49
0.04%
> 0.5 to 1 mg/L
319,226
1,525
0.48%
78
0.02%
> 1 mg/L
432,493
1,907
0.44%
99
0.02%
Total
914,594
5,363
0.59%
284
0.03%
NCWSs
Free Chlorine
0
78,523
2,784
3.55%
147
0.19%
> 0 to 0.2 mg/L
29,339
892
3.04%
57
0.19%
> 0.2 to 0.5 mg/L
49,565
706
1.42%
31
0.06%
> 0.5 to 1 mg/L
53,582
533
0.99%
32
0.06%
> 1 mg/L
53,551
662
1.24%
34
0.06%
Total
264,560
5,577
2.11%
301
0.11%
Total Chlorine
0
16,932
382
2.26%
16
0.09%
> 0 to 0.2 mg/L
5,840
237
4.06%
20
0.34%
> 0.2 to 0.5 mg/L
13,222
221
1.67%
10
0.08%
> 0.5 to 1 mg/L
20,340
193
0.95%
14
0.07%
> 1 mg/L
31,437
338
1.08%
23
0.07%
Total
87,771
1,371
1.56%
83
0.09%
System Size
To assess any potential variations in the SYR3 ICR microbial data due to system size, EPA
calculated the frequency of detection for the total coliform results in surface water systems over
the six years for each of seven system size categories presented in the NPDWR Revisions to the
Total Coliform Rule: < 100; 101 - 500; 501 - 1,000; 1,001 - 4,100; 4,101 - 33,000; 33,001 -
100,000; and > 100,000. Results were generated separately for five bins of free chlorine residual
concentrations for CWSs and NCWSs (Exhibit C.l 1 and Exhibit C.12, respectively). Results
were also generated separately for five bins of total chlorine residual concentrations for CWSs
and NCWSs (Exhibit C.16 and Exhibit C.17, respectively).
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-7
December 2016
-------
For free chlorine samples, there is a general trend of higher TC+ occurrence in surface water
CWSs and NTNCWSs serving 4,100 or fewer compared to larger systems. For surface water
CWSs, the percent TC+ is highest in either the 0 mg/L bin or the >0 - 0.2 mg/L bin for surface
water CWSs serving 33,000 people or fewer. This relationship does not hold, however, for the
medium and large systems, most likely due to the bias of chloramine systems that reported zero
or very low free chlorine but high total chorine. Medium and large surface water systems are
more likely to use chloramines than small systems. For surface water NCWSs, there is a notable
peak in TC+ occurrence in the 0 mg/L bin (8.57 percent). This peak is based on data from 7
systems that comprise the 35 overall samples in this system size category. One of these 7
systems reported the three TC+ samples. The rest of the data in these exhibits show a slight trend
in increasing TC+ occurrence with decreasing system size.
Exhibit C.11: Surface Water CWSs: Percent of TC Positives Paired with Free
Chlorine Data by System Size (2006-2011)1
¦ Free Chlorine = 0
¦ Free Chlorine >0 to 0.2 mg/L
¦ Free Chlorine >0.2 to 0.5 mg/L
I Free Chlorine >0.5 to 1 mg/L
¦ Free Chlorine >1 mg/L
nhi Jin _iin
1,001-4,100 4,101-33,000 33,001-100,000 >100,000
1 Different scales were used for the Percent of TC Positives (y-axis) for the CWS results (in this exhibit) compared to
the NCWS results in the next exhibit (Exhibit C.12) to enable a closer look at the CWS results.
3.0%
2.5%
<100 101-500 501-1,000
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-8
December 2016
-------
Exhibit C.12: Surface Water NCWSs: Percent of TC Positives Paired with Free
Chlorine Data by System Size (2006-2011)
9.0%
8.0%
7.0%
6.0%
5.0%
4.0%
3.0%
2.0%
1.0%
0.0%
III lllll
III! I.ll
I Free Chlorine = 0
I Free Chlorine >0 to 0.2 mg/L
I Free Chlorine >0.2 to 0.5 mg/L
Free Chlorine >0.5 to 1 mg/L
I Free Chlorine >1 mg/L
<100
101-500
501-1,000
1,001-4,100
4,101-33,000 33,001-100,000
>100,000
EPA also calculated the frequency of detection for the E. coli results in surface water systems
over the six years for each of seven system size categories. Results were generated separately for
five bins of free chlorine residual concentrations for CWSs and NCWSs (Exhibit C.13 and
Exhibit C.14, respectively).
The system size trends for EC+ occurrence are similar to, but not exactly the same as, the system
size trends for TC+ occurrence. For free chlorine samples in CWSs and NCWSs (Exhibit C.13
and Exhibit C.14, respectively), there is a general trend of higher EC+ occurrence in systems
serving 1,000 or fewer compared to larger systems. It is also important to note the small sample
size for NCWSs; there were few, if any, samples from NCWSs serving more than 4,100 people.
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-9
December 2016
-------
Exhibit C.13: Surface Water CWSs: Percent of EC Positives Paired with Free
Chlorine Data by System Size (2006-2011)
0.40%
0.35%
0.30%
l/l
CD
¦| 0.25%
"in
o
Q_
y o.2o%
0
1 0.15%
u
d)
Q_
0.10%
0.05%
0.00%
I Free Chlorine = 0
I Free Chlorine >0 to 0.2 mg/L
I Free Chlorine >0.2 to 0.5 mg/L
Free Chlorine >0.5 to 1 mg/L
I Free Chlorine >1 mg/L
ill. !¦ ¦
<100
101-500
501-1,000
1,001-4,100
4,101-33,000 33,001-100,000
>100,000
Exhibit C.14: Surface Water NCWSs: Percent of EC Positives Paired with Free
Chlorine Data by System Size (2006-2011)1
0.18%
0.16%
0.14%
£ 0.12%
.>
8 o.io%
Q_
U
^ 0.08%
c
§ 0.06%
CD
Q_
0.04%
0.02%
0.00%
I Free Chlorine = 0
I Free Chlorine >0 to 0.2 mg/L
I Free Chlorine >0.2 to 0.5 mg/L
Free Chlorine >0.5 to 1 mg/L
I Free Chlorine >1 mg/L
<100
101-500
501-1,000
1,001-4,100 4,101-33,000 33,001-100,000 >100,000
1 Different scales were used for the Percent of EC Positives (y-axis) for the NCWS results (in this exhibit) compared
to the CWS results in the previous exhibit (Exhibit C. 13) to enable a closer look at the NCWS results.
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-10
December 2016
-------
Exhibit C.15: Number of Total Coliform Surface Water Samples Paired with Free
Chlorine Data, by System Size and System Type (underlying data/denominator for
Exhibit C.11, Exhibit C.12, Exhibit C.13 and Exhibit C.14)
System Size
Total # TC SW Samples Paired with Free Chlorine
0
> 0 to 0.2
mg/L
> 0.2 to 0.5
mg/L
> 0.5 to 1
mg/L
> 1 mg/L
Total
CWSs
<100
1,951
1,980
4,009
6,702
8,187
22,829
101-500
2,722
6,805
11,317
14,604
10,145
45,593
501-1,000
1,315
2,375
5,239
8,723
7,961
25,613
1,001-4,100
4,822
8,512
20,271
41,809
50,150
125,564
4,101-33,000
12,654
28,935
63,836
197,220
251,292
553,937
33,001-100,000
9,597
25,726
51,180
152,577
156,614
395,694
>100,000
11,335
34,642
27,799
97,538
190,901
362,215
NCWSs
<100
1,386
2,145
3,089
3,334
3,331
13,285
101-500
300
1,698
2,884
3,215
2,863
10,960
501-1,000
35
252
455
776
1,297
2,815
1,001-4,100
48
561
1,100
1,382
1,889
4,980
4,101-33,000
6
193
392
440
2,582
3,613
33,001-100,000
2
45
251
493
3,365
4,156
>100,000
0
0
0
0
0
0
Results were also generated separately for five bins of total chlorine residual concentrations for
CWSs and NCWSs (Exhibit C.16 and Exhibit C.17, respectively). For total chlorine samples,
there is also a general trend of higher TC+ occurrence in systems serving 4,100 or fewer
compared to larger systems. The one notable exception is for large surface water NCWSs serving
> 100,000; those systems show a very high TC+ occurrence for the two total chlorine residual
categories with concentrations greater than 0.5 mg/L. As noted in Exhibit C.20, the NCWS
results in this system size category are based on results from a single non-transient non-
community purchased surface water system in Texas; the only three total coliform results from
this system were TC+.
For total chlorine results in surface water CWSs, the highest percent of TC+ samples occurred in
either the 0 mg/L bin or the >0 - 0.2 mg/L bin, with a general trending downward of TC+
samples with increasing total chlorine concentrations for all size categories.
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-ll
December 2016
-------
Exhibit C.16: Surface Water CWSs: Percent of TC Positives Paired with Total
Chlorine Data by System Size (2006-2011)1
10.0%
9.0%
8.0%
tfl
cu
¦| 7.0%
Vi
o
£ 6.0%
I—
2 5.0%
c
u
gJ 4.0%
Q_
3.0%
2.0%
1.0%
0.0%
Hi ¦ ill ¦ -li .
ITotal Chlorine = 0
ITotal Chlorine >0to 0.2 mg/L
ITotal Chlorine >0.2 to 0.5 mg/L
Total Chlorine >0.5 to 1 mg/L
ITotal Chlorine >1 mg/L
la— ll .
<100
101-500
501-1,000
1,001-4,100 4,101-33,000 33,001-100,000 >100,000
1 Different scales were used for the Percent of TC Positives (y-axis) for the CWS results (in this exhibit) compared to
the NCWS results in the next exhibit (Exhibit C.17) to enable a closer look at the CWS results.
Exhibit C.17: Surface Water NCWSs: Percent of TC Positives Paired with Total
Chlorine Data by System Size (2006-2011)1
11.0%
10.0%
9.0%
8.0%
1.0%
6.0%
5.0%
4.0%
3.0%
2.0%
1.0%
0.0%
¦llll
ll
I
ITotal Chlorine = 0
ITotal Chlorine >0 to 0.2 mg/L
ITotal Chlorine >0.2 to 0.5 mg/L
Total Chlorine >0.5 to 1 mg/L
ITotal Chlorine >1 mg/L
<100
101-500
501-1,000
1,001-4,100
4,101-33,000 33,001-100,000
>100,000
1 Due to the small number of records in the >100,000 population size category for NCWSs (a total of 3 samples as
can be seen in Exhibit C.20), the percent of TC+ was equal to 100%. However, the upper bound on the y-axis in this
plot was set equal to 11 percent to enable a closer look at the other NCWS results.
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-12
December 2016
-------
The EC results paired with total chlorine concentrations are presented in Exhibit C. 18 and
Exhibit C.19 for CWSs andNCWSs, respectively. For total chlorine samples in SW CWSs
(Exhibit C. 18), there is not a strong system size trend in EC+ occurrence. For total chlorine
samples in SW NCWSs (Exhibit C.19), there are no distinguishable system size trends. As
mentioned earlier, it is also important to note the small sample size for NCWSs; there were few,
if any, samples from NCWSs serving more than 4,100 people.
For both free and total chlorine data, the trend of higher occurrence in the lower disinfectant
residual bins is not as evident in the EC+ dataset compared to the TC+ results.
Exhibit C.18: Surface Water CWSs: Percent of EC Positives Paired with Total
Chlorine Data by System Size (2006-2011)
1.2%
1.0%
0.8%
0.6%
S 0.4%
0.2%
0.0%
Ha- ¦!-
I _ li
ITotal Chlorine = 0
ITotal Chlorine >0to 0.2 mg/L
ITotal Chlorine >0.2 to 0.5 mg/L
Total Chlorine >0.5 to 1 mg/L
ITotal Chlorine >1 mg/L
<100
101-500
501-1,000
1,001-4,100 4,101-33,000 33,001-100,000 >100,000
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-13
December 2016
-------
Exhibit C.19: Surface Water NCWSs: Percent of EC Positives Paired with Total
Chlorine Data by System Size (2006-2011)1
¦ Total Chlorine = 0
¦ Total Chlorine >0 to 0.2 mg/L
¦ Total Chlorine >0.2 to 0.5 mg/L
Total Chlorine >0.5 to 1 mg/L
¦ Total Chlorine >1 mg/L
I
101-500 501-1,000 1,001-4,100 4,101-33,000 33,001-100,000 >100,000
1 Different scales were used for the Percent of EC Positives (y-axis) for the NCWS results (in this exhibit) compared
to the CWS results in the previous exhibit (Exhibit C. 18) to enable a closer look at the NCWS results.
Exhibit C.20: Number of Total Coliform Surface Water Samples Paired with Total
Chlorine Data, by System Size and System Type (underlying data/denominator for
Exhibit C.16, Exhibit C.17, Exhibit C.18 and Exhibit C.19)
System Size
Total # TC SW Samples Paired with Total Chlorine
0
> 0 to 0.2
mg/L
> 0.2 to 0.5
mg/L
> 0.5 to 1
mg/L
> 1 mg/L
Total
CWSs
<100
170
465
1,263
3,218
11,061
16,177
101-500
367
2,030
4,489
8,652
27,750
43,288
501-1,000
96
1,658
3,781
6,202
18,242
29,979
1,001-4,100
185
3,250
11,573
27,056
80,229
122,293
4,101-33,000
554
4,855
19,807
80,943
250,792
356,951
33,001-100,000
117
5,988
15,078
46,287
162,092
229,562
>100,000
46
1,720
6,136
52,468
234,419
294,789
NCWSs
<100
155
224
364
642
2,115
3,500
101-500
19
206
328
667
1,728
2,948
501-1,000
7
37
216
194
722
1,176
1,001-4,100
3
98
247
297
1,084
1,729
4,101-33,000
0
0
0
10
259
269
0.9%
0.8%
0.7%
« 0.6%
CD
>
1 0.5%
Q_
£ 0.4%
S 0.3%
CD
Q_
0.2%
0.1%
0.0%
<100
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-14
December 2016
-------
System Size
Total # TC SW Samples Paired with Total Chlorine
0
> 0 to 0.2
mg/L
> 0.2 to 0.5
mg/L
> 0.5 to 1
mg/L
> 1 mg/L
Total
33,001-100,000
0
0
0
0
0
0
>100,000
0
0
0
1
2
3
1 The NCWS results in this system size category are based on results from a single non-transient non-community
water purchased surface water system in Texas that serves more than 200,000 people; the only three total coliform
results from this system were TC+.
EPA calculated the frequency of detection for the total coliform results in ground water systems
over the six years for each of seven system size categories and presented the results separately
for five bins of free chlorine residual concentrations for CWSs and NCWSs (Exhibit C.21 and
Exhibit C.22, respectively). Similar results for E. coli results in ground water systems are
presented for CWSs (Exhibit C.23) and NCWSs (Exhibit C.24).
Exhibit C.21: Ground Water CWSs: Percent of TC Positives Paired with Free
Chlorine Data by System Size (2006-2011)
4.5%
4.0%
3.5%
3.0%
2.5%
2.0%
1.5%
1.0%
0.5%
0.0%
h
I Free Chlorine = 0
I Free Chlorine >0 to 0.2 mg/L
I Free Chlorine >0.2 to 0.5 mg/L
Free Chlorine >0.5 to 1 mg/L
I Free Chlorine >1 mg/L
¦ llll
.... I. J
<100
101-500
501-1,000
1,001-4,100
4,101-33,000 33,001-100,000
>100,000
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-15
December 2016
-------
Exhibit C.22: Ground Water NCWSs: Percent of TC Positives Paired with Free
Chlorine Data by System Size (2006-2011)
4.5%
4.0%
3.5%
3.0%
2.5%
2.0%
1.5%
1.0%
0.5%
0.0%
I Free Chlorine = 0
I Free Chlorine >0 to 0.2 mg/L
I Free Chlorine >0.2 to 0.5 mg/L
Free Chlorine >0.5 to 1 mg/L
I Free Chlorine >1 mg/L
llll I
<100
101-500
501-1,000
1,001-4,100 4,101-33,000 33,001-100,000 >100,000
Exhibit C.23: Ground Water CWSs: Percent of EC Positives Paired with Free
Chlorine Data by System Size (2006-2011)
0.16%
0.14%
0.12%
$ 0.10%
>
| 0.08%
u
LU
° 0.06%
c
CD
u
£ 0.04%
0.02%
0.00%
I Free Chlorine = 0
I Free Chlorine >0 to 0.2 mg/L
I Free Chlorine >0.2 to 0.5 mg/L
Free Chlorine >0.5 to 1 mg/L
I Free Chlorine >1 mg/L
ll. I
ll
<100
101-500
501-1,000
1,001-4,100 4,101-33,000 33,001-100,000 >100,000
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-16
December 2016
-------
Exhibit C.24: Ground Water NCWSs: Percent of EC Positives Paired with Free
Chlorine Data by System Size (2006-2011)
0.30%
0.25%
0.20%
0.15%
S o.io%
0.05%
0.00%
I
I
I Free Chlorine = 0
I Free Chlorine >0 to 0.2 mg/L
I Free Chlorine >0.2 to 0.5 mg/L
Free Chlorine >0.5 to 1 mg/L
I Free Chlorine >1 mg/L
<100
101-500
501-1,000
1,001-4,100 4,101-33,000 33,001-100,000 >100,000
Exhibit C.25: Number of Total Coliform Ground Water Samples Paired with Free
Chlorine Data, by System Size and System Type (underlying data/denominator for
Exhibit C.21, Exhibit C.22, Exhibit C.23 and Exhibit C.24)
System Size
Total # TC GW Samples Paired with Free Chlorine
0
> 0 to 0.2
mg/L
> 0.2 to 0.5
mg/L
> 0.5 to 1
mg/L
> 1 mg/L
Total
CWSs
<100
21,156
16,583
30,794
37,393
27,445
133,371
101-500
18,744
30,108
53,828
65,618
51,038
219,336
501-1,000
4,427
11,138
22,651
28,850
22,787
89,853
1,001-4,100
7,306
28,940
60,626
88,287
78,845
264,004
4,101-33,000
13,129
61,175
126,350
190,510
134,630
525,794
33,001-100,000
4,700
21,048
57,573
96,151
34,312
213,784
>100,000
196
7,178
8,850
14,591
16,199
47,014
NCWSs
<100
59,359
14,408
19,810
20,937
23,249
137,763
101-500
13,693
9,585
17,538
18,352
19,010
78,178
501-1,000
3,052
1,760
3,722
4,585
3,866
16,985
1,001-4,100
2,056
3,379
7,731
7,841
5,965
26,972
4,101-33,000
363
207
764
1,867
1,461
4,662
33,001-100,000
0
0
0
0
0
0
>100,000
0
0
0
0
0
0
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-17
December 2016
-------
EPA also calculated the frequency of detection for the total coliform results in ground water
systems over the six years for each of seven system size categories, presented separately for five
bins of total chlorine residual concentrations for CWSs and NCWSs (Exhibit C.26 and Exhibit
C.27, respectively). Similar results for E. coli results in ground water systems are presented for
CWSs (Exhibit C.28) and NCWSs (Exhibit C.29).
Exhibit C.26: Ground Water CWSs: Percent of TC Positives Paired with Total
Chlorine Data by System Size (2006-2011)1
14.0%
12.0%
10.0%
8.0%
6.0%
4.0%
2.0%
0.0%
ITotal Chlorine = 0
ITotal Chlorine >0 to 0.2 mg/L
ITotal Chlorine >0.2 to 0.5 mg/L
Total Chlorine >0.5 to 1 mg/L
ITotal Chlorine >1 mg/L
m\
ll
h. ii
ii.. j
Ii. II... -i—
I..
<100
101-500
501-1,000
1,001-4,100
4,101-33,000 33,001-100,000
>100,000
1 Different scales were used for the Percent of TC Positives (y-axis) for the CWS results (in this exhibit) compared to
the NCWS results in the next exhibit (Exhibit C.27) to enable a closer look at the CWS results.
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-18
December 2016
-------
Exhibit C.27: Ground Water NCWSs: Percent of TC Positives Paired with Total
Chlorine Data by System Size (2006-2011)
35.0%
30.0%
25.0%
20.0%
15.0%
10.0%
5.0%
0.0%
¦ Total Chlorine = 0
¦ Total Chlorine >0to 0.2 mg/L
¦ Total Chlorine >0.2 to 0.5 ms/L
¦ Total Chlorine >0.5 to 1 mg/L
¦ Total Chlorine >1 mg/L
ill.. Il Ii
<100
101-500
501-1,000
1,001-4,100 4,101-33,000 33,001-100,000 >100,000
Exhibit C.28: Ground Water CWSs: Percent of EC Positives Paired with Total
Chlorine Data by System Size (2006-2011)
0.45%
0.40%
0.35%
0.30%
0.25%
0.20%
0.15%
0.10%
0.05%
0.00%
h
II
ii ih.
I
ITotal Chlorine = 0
ITotal Chlorine >0 to 0.2 mg/L
ITotal Chlorine >0.2 to 0.5 mg/L
Total Chlorine >0.5 to 1 mg/L
ITotal Chlorine >1 mg/L
<100
101-500
501-1,000
1,001-4,100 4,101-33,000 33,001-100,000 >100,000
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-19
December 2016
-------
Exhibit C.29: Ground Water NCWSs: Percent of EC Positives Paired with Total
Chlorine Data by System Size (2006-2011)
0.60%
0.50%
0.40%
0.30%
S 0.20%
0.10%
0.00%
I
III ill
I II
ITotal Chlorine = 0
ITotal Chlorine >0 to 0.2 mg/L
ITotal Chlorine >0.2 to 0.5 mg/L
Total Chlorine >0.5 to 1 mg/L
ITotal Chlorine >1 mg/L
<100
101-500
501-1,000
1,001-4,100 4,101-33,000 33,001-100,000 >100,000
Exhibit C.30: Number of Total Coliform Ground Water Samples Paired with Total
Chlorine Data, by System Size and System Type (underlying data/denominator for
Exhibit C.26, Exhibit C.27, Exhibit C.28, and Exhibit C.29)
System Size
Total # TC GW Samples Paired with Total Chlorine
0
> 0 to 0.2
mg/L
> 0.2 to 0.5
mg/L
> 0.5 to 1
mg/L
> 1 mg/L
Total
CWSs
<100
4,267
3,560
10,294
14,582
20,154
52,857
101-500
2,441
8,048
23,007
43,377
57,716
134,589
501-1,000
256
4,538
12,477
25,830
33,695
76,796
1,001-4,100
634
6,294
25,609
65,953
83,488
181,978
4,101-33,000
494
9,000
34,396
113,493
152,990
310,373
33,001-100,000
149
1,455
14,152
48,104
59,849
123,709
>100,000
11
104
1,689
7,887
24,601
34,292
NCWSs
<100
11,856
3,334
7,887
10,672
15,616
49,365
101-500
4,755
1,966
3,581
5,358
7,725
23,385
501-1,000
179
186
627
1,382
2,252
4,626
1,001-4,100
128
330
1,089
2,652
4,975
9,174
4,101-33,000
14
24
38
276
869
1,221
33,001-100,000
0
0
0
0
0
0
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-20
December 2016
-------
System Size
Total # TC GW Samples Paired with Total Chlorine
0
> 0 to 0.2
mg/L
> 0.2 to 0.5
mg/L
> 0.5 to 1
mg/L
> 1 mg/L
Total
>100,000
0
0
0
0
0
0
Temporal/Seasonal Analysis
To assess any potential seasonal variations in the SYR3 ICR microbial data, EPA calculated the
frequency of detection, by month, for the total coliform results in surface water systems for each
calendar month in each year (2006 - 2011). Results were generated separately for five bins of
free and total chlorine residual concentrations (Exhibit C.31 and Exhibit C.32, respectively).
Note that this analysis does not differentiate between seasonal and non-seasonal systems.
TC+ occurrence was generally higher in warmer months regardless of residual concentration,
indicating a strong seasonal trend. The seasonal effect was stronger for free chlorine samples
than it was for total chlorine samples. For both free and total chlorine samples, the highest
percent of TC+ occurred in either the 0 mg/L bin or the >0 - 0.2 mg/L bin regardless of season
(with the exception of free chlorine data in January), with a general trending downward of TC+
samples with increasing residual concentrations, particularly in the total chlorine data (Exhibit
C.32).
Exhibit C.31: Surface Water PWSs: Percent of TC Positives Paired with Free
Chlorine Data by Month (2006-2011)1
¦ Free Chlorine = 0
¦ Free Chlorine >0 to 0.2 mg/L
¦ ¦ Free Chlorine >0.2 to 0.5 mg/L
¦ Free Chlorine >0.5 to 1 mg/L
¦ Free Chlorine > lmg/L
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
1 Different scales were used for the Percent of TC Positives (y-axis) for the free chlorine results (in this exhibit)
compared to the total chlorine results in the next exhibit (Exhibit C.32) to enable a closer look at the free chlorine
results.
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-21
December 2016
-------
Exhibit C.32: Surface Water PWSs: Percent of TC Positives Paired with Total
Chlorine Data by Month (2006-2011)
4.0%
3.5%
3.0%
2.5%
2.0%
1.5%
1.0%
0.5%
0.0%
1 ll
1
In.
1... III..
11.1
III
II
ll
II
II
I Total Chlorine = 0
¦ Total Chlorine >0 to 0.2 mg/L
I Total Chlorine >0.2 to 0.5 mg/L
¦ Total Chlorine >0.5 to 1 mg/L
¦ Total Chlorine > lmg/L
ll
ll.
Jan Feb Mar Apr May Jun
Jul
Aug Sep Oct Nov
llll
Dec
EPA also calculated the frequency of detection, by month, for the E. coli results in surface water
systems for each calendar month in each year (2006 - 2011). Results were generated separately
for five bins of free and total chlorine residual concentrations (Exhibit C.33 and Exhibit C.34,
respectively).
EC+ occurrence in surface water systems was generally higher in warmer months regardless of
residual concentration, indicating a seasonal trend. Similar to the TC+ results, the seasonal effect
was stronger for paired EC+/free chlorine samples than it was for paired EC+/total chlorine
samples. As with previous analyses in this section, the trend of higher occurrence in the lower
disinfectant residual bins is not as evident in the EC+ dataset (Exhibit C.33 and Exhibit C.34)
compared to the TC+ results (Exhibit C.31 and Exhibit C.32).
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-22
December 2016
-------
Exhibit C.33: Surface Water PWSs: Percent of EC Positives Paired with Free
Chlorine Data by Month (2006-2011)
0.20%
0.18%
0.16%
0.14%
0.12%
0.10%
0.08%
0.06%
0.04%
0.02%
0.00%
ll. Illl
I Illl llll
iJ
I Free Chlorine = 0
I Free Chlorine >0 to 0.2 mg/L
I Free Chlorine >0.2 to 0.5 mg/L
Free Chlorine >0.5 to 1 mg/L
I Free Chlorine > lmg/L
I I l . I ll. I
Jan
Feb
Mar
Apr
May Jun
Jul
Aug
Sep
Oct
Nov
Dec
Exhibit C.34: Surface Water PWSs: Percent of EC Positives Paired with Total
Chlorine Data by Month (2006-2011)
2.0%
1.8%
1.6%
1.4%
1.2%
1.0%
0.8%
0.6%
0.4%
0.2%
0.0%
,.i.. . I Ll.
Jan
Feb
Mar
Apr
May
Jun
Jul
l__.
Aug
ITotal Chlorine = 0
ITotal Chlorine >0to 0.2 mg/L
ITotal Chlorine >0.2 to 0.5 mg/L
Total Chlorine >0.5 to 1 mg/L
ITotal Chlorine > lmg/L
Sep
Oct
Nov
Dec
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-23
December 2016
-------
Exhibit C.35: Number of Total Coliform Samples in Surface Water Paired with Free
and Total Chlorine Data, by Month (underlying data/denominator for Exhibit C.31,
Exhibit C.32, Exhibit C.33 and Exhibit C.34)
Month
Total # TC SW Samples Paired
0
> 0 to 0.2
mg/L
> 0.2 to 0.5
mg/L
> 0.5 to 1
mg/L
> 1 mg/L
Total
Free Chlorine
Jan
3,967
8,004
12,559
40,661
61,063
126,254
Feb
3,480
7,562
12,262
40,651
62,031
125,986
Mar
3,844
7,437
12,913
42,002
61,597
127,793
Apr
3,797
8,281
14,192
44,804
59,392
130,466
May
4,169
8,938
15,264
46,445
56,057
130,873
Jun
4,091
10,291
17,697
46,986
53,907
132,972
Jul
3,869
11,068
19,175
46,229
53,429
133,770
Aug
4,152
11,651
19,782
46,000
52,327
133,912
Sep
4,045
11,163
19,337
44,592
53,722
132,859
Oct
3,611
10,866
18,529
44,157
55,441
132,604
Nov
3,584
9,763
16,396
43,299
58,700
131,742
Dec
3,564
8,845
13,716
42,987
62,911
132,023
Total Chlorine
Jan
151
1,038
3,443
15,986
64,615
85,233
Feb
118
810
3,265
15,591
64,332
84,116
Mar
127
838
3,544
16,438
66,089
87,036
Apr
134
1,030
3,969
17,856
65,849
88,838
May
128
1,239
5,013
19,192
64,877
90,449
Jun
148
1,793
6,059
20,467
63,316
91,783
Jul
179
2,452
6,786
20,718
61,999
92,134
Aug
185
2,730
7,440
21,521
65,211
97,087
Sep
160
2,617
7,222
20,638
65,528
96,165
Oct
128
2,434
6,545
20,643
67,588
97,338
Nov
110
2,066
5,511
19,272
69,217
96,176
Dec
151
1,484
4,485
18,315
71,874
96,309
EPA calculated the frequency of detection, by month, for the total coliform results in ground
water systems for each calendar month in each year (2006 - 2011). Results were generated
separately for five bins of free and total chlorine residual concentrations (Exhibit C.36 and
Exhibit C.37, respectively).
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-24
December 2016
-------
Exhibit C.36: Ground Water PWSs: Percent of TC Positives Paired with Free
Chlorine Data by Month (2006-2011)
6.0%
5.0%
4.0%
3.0%
u
g 2.0%
1.0%
0.0%
llll
Jan
Il.i Iim ll i liii
Feb Mar Apr May
I Free Chlorine = 0
I Free Chlorine >0 to 0.2 mg/L
I Free Chlorine >0.2 to 0.5 mg/L
Free Chlorine >0.5 to 1 mg/L
I Free Chlorine > lmg/L
y
in
Jun
Jul
Aug Sep Oct Nov Dec
Exhibit C.37: Ground Water PWSs: Percent of TC Positives Paired with Total
Chlorine Data by Month (2006-2011)1
4.5%
4.0%
3.5%
3.0%
2.5%
2.0%
1.5%
1.0%
0.5%
0.0%
lii ill.. llliT
III
II
Jan
Feb
Mar Apr May Jun
i I
Aug Sep
ITotal Chlorine = 0
ITotal Chlorine >0to 0.2 mg/L
ITotal Chlorine >0.2 to 0.5 mg/L
Total Chlorine >0.5 to 1 mg/L
ITotal Chlorine > 1 mg/L
III
L
Oct
Nov
Dec
1 Different scales were used for the Percent of TC Positives (y-axis) for the total chlorine results (in this exhibit)
compared to the free chlorine results in the previous exhibit (Exhibit C.36) to enable a closer look at the total chlorine
results.
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-25
December 2016
-------
EPA calculated the frequency of detection, by month, for the E. coli results in ground water
systems for each calendar month in each year (2006 - 2011). Results were generated separately
for five bins of free and total chlorine residual concentrations (Exhibit C.38 and Exhibit C.39,
respectively).
Exhibit C.38: Ground Water PWSs: Percent of EC Positives Paired with Free
Chlorine Data by Month (2006-2011)
0.35%
0.30%
0.25%
tfl
cu
0.20%
w
O
Q_
£ 0.15%
o
c
§ 0.10%
o
Q_
0.05%
0.00%
III I lli I illn
III
III
I Free Chlorine = 0
I Free Chlorine >0 to 0.2 mg/L
I Free Chlorine >0.2 to 0.5 mg/L
Free Chlorine >0.5 to 1 mg/L
I Free Chlorine > lmg/L
11 IIli 11 ¦
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-26
December 2016
-------
Exhibit C.39: Ground Water PWSs: Percent of EC Positives Paired with Total
Chlorine Data by Month (2006-2011)
¦ Total Chlorine = 0
1.0%
¦ Total Chlorine >0to 0.2 mg/L
0-9% ¦ Total Chlorine >0.2 to 0.5 mg/L
0.8% I Total Chlorine >0.5 to 1 mg/L
q 70/ I BTotal Chlorine > lmg/L
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Exhibit C.40: Number of Total Coliform Samples in Ground Water Paired with Free
and Total Chlorine Data, by Month (underlying data/denominator for Exhibit C.36,
Exhibit C.37, Exhibit C.38 and Exhibit C.39)
Month
Total # TC GW Samples Paired
0
> 0 to 0.2
mg/L
> 0.2 to 0.5
mg/L
> 0.5 to 1
mg/L
> 1 mg/L
Total
Free Chlorine
Jan
12,262
15,562
31,367
46,128
33,860
139,179
Feb
10,639
15,263
31,090
47,354
35,088
139,434
Mar
11,378
15,788
31,603
47,546
34,791
141,106
Apr
12,660
16,293
32,779
47,922
34,861
144,515
May
11,887
16,414
33,326
47,221
33,817
142,665
Jun
12,672
17,439
34,708
47,552
33,479
145,850
Jul
12,692
18,425
36,527
48,235
33,582
149,461
Aug
13,258
18,914
37,484
48,546
33,540
151,742
Sep
14,184
18,927
37,577
48,711
34,961
154,360
Oct
13,689
18,413
36,787
49,006
36,763
154,658
Nov
11,309
17,457
34,429
47,914
36,704
147,813
Dec
11,551
16,614
32,560
48,847
37,361
146,933
Total Chlorine
Jan
2,074
2,719
9,921
25,977
35,657
76,348
Feb
1,614
2,687
9,412
25,555
37,114
76,382
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-27
December 2016
-------
Month
Total # TC GW Samples Paired
0
> 0 to 0.2
mg/L
> 0.2 to 0.5
mg/L
> 0.5 to 1
mg/L
> 1 mg/L
Total
Mar
1,856
2,819
9,780
26,714
38,736
79,905
Apr
2,260
2,960
10,321
27,653
37,862
81,056
May
1,891
3,174
10,971
27,949
38,181
82,166
Jun
2,053
3,366
11,503
28,492
37,754
83,168
Jul
2,478
3,571
12,331
29,186
37,602
85,168
Aug
1,997
3,924
13,155
29,899
39,250
88,225
Sep
2,532
3,713
12,888
29,860
39,423
88,416
Oct
2,636
3,561
12,535
30,218
40,025
88,975
Nov
1,825
3,353
11,379
29,327
40,506
86,390
Dec
1,968
2,992
10,650
28,736
41,820
86,166
Annual Trends Analysis
To assess any potential yearly trends over the six years of data in the SYR3 ICR microbial data,
EPA calculated the frequency of detection for the total coliform results in surface water systems.
Results were generated separately for five bins of free and total chlorine residual concentrations
(Exhibit C.41 and Exhibit C.42, respectively).
For free chlorine, the data show a general trend of higher TC+ occurrence in 2007 and 2008,
with a general trending downward of the data in 2009 through 2011. The trend of increasing TC+
occurrence with decreasing free chlorine residual is also stronger in 2006 - 2008 than it is for
2009 - 2011. For total chlorine, the results are opposite with higher overall occurrence
happening in the last two years of the dataset (2010 and 2011) compared to 2006 - 2009. Total
chlorine results show a more consistent trend in TC+ occurrence and disinfectant residual
concentration, with higher TC+ occurrence in the 0 mg/L or the >0 - 0.2 mg/L bin for all years,
and a general trending down off occurrence in the higher residual bins.
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-28
December 2016
-------
Exhibit C.41: Surface Water PWSs: Percent of TC Positives Paired with Free
Chlorine Data by Year (2006 - 2011)
1.2%
1.0%
£ 0.8%
u
H
o
| 0.6%
0.4%
0.2%
0.0%
I Free Chlorine = 0
I Free Chlorine >0 to 0.2 mg/L
I Free Chlorine >0.2 to 0.5 mg/L
Free Chlorine >0.5 to 1 mg/L
I Free Chlorine > lmg/L
2006
2007
2008
2009
2010
2011
Exhibit C.42: Surface Water PWSs: Percent of TC Positives Paired with Total
Chlorine Data by Year (2006 - 2011)
¦ Total Chlorine = 0
¦ Total Chlorine >0to 0.2 mg/L
qqo/0 ¦ Total Chlorine >0.2 to 0.5 mg/L
Total Chlorine >0.5 to 1 mg/L
3.5%
ITotal Chlorine > 1 mg/L
> 3.0%
s
£ 2.5%
I—
¦2 2.0%
0J
o
£ 1.5%
1.0%
0.5%
0.0%
Li
I
2006
2007
2008
2009
2010
2011
EPA also calculated the frequency of detection for the E. coli results in surface water systems.
Results were generated separately for five bins of free and total chlorine residual concentrations
(Exhibit C.43 and Exhibit C.44, respectively). Exhibit C.43 and Exhibit C.44 do not show a
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-29
December 2016
-------
uniform trend in percent EC+ in surface water from 2006 to 2011, with each of the two residual
type bins for free and total chlorine showing different trends and peak years.
Exhibit C.43: Surface Water PWSs: Percent of EC Positives Paired with Free
Chlorine Data by Year (2006 - 2011)
0.14%
0.12%
0.10%
l/l
cu
E 0.08%
in
o
Q_
U
£ 0.06%
o
c
£ 0.04%
0 to 0.2 mg/L
I Free Chlorine >0.2 to 0.5 mg/L
Free Chlorine >0.5 to 1 mg/L
I Free Chlorine > lmg/L
iL
III I III!
2006
2007
2008
2009
2010
2011
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-30
December 2016
-------
Exhibit C.44: Surface Water PWSs: Percent of EC Positives Paired with Total
Chlorine Data by Year (2006 - 2011)
0.9%
0.8%
0.7%
0.6%
0.5%
0.4%
0.3%
0.2%
0.1%
0.0%
I
I
ITotal Chlorine = 0
ITotal Chlorine >0to 0.2 mg/L
ITotal Chlorine >0.2 to 0.5 mg/L
Total Chlorine >0.5 to 1 mg/L
ITotal Chlorine > lmg/L
2006
2007
2008
2009
2010
2011
1 Different scales were used for the Percent of EC Positives (y-axis) for the total chlorine results (in this exhibit)
compared to the free chlorine results in the previous exhibit (Exhibit C.43) to enable a closer look at the total chlorine
results.
Exhibit C.45: Number of Total Coliform Samples in Surface Water Paired with Free
and Total Chlorine Data, by Year (underlying data/denominator for Exhibit C.41,
Exhibit C.42, Exhibit C.43 and Exhibit C.44)
Year
Total # TC SW Samples Paired
0
> 0 to 0.2
mg/L
> 0.2 to 0.5
mg/L
> 0.5 to 1
mg/L
> 1 mg/L
Total
Free Chlorine
2006
5,553
16,634
30,298
74,514
72,823
199,822
2007
5,306
17,047
31,352
82,933
96,456
233,094
2008
5,282
16,759
29,806
78,654
108,073
238,574
2009
9,479
19,226
30,163
78,864
109,277
247,009
2010
10,813
22,048
34,692
104,195
143,587
315,335
2011
9,740
22,155
35,511
109,653
160,361
337,420
Total Chlorine
2006
314
2,615
7,329
25,159
80,823
116,240
2007
459
3,039
9,201
29,105
82,778
124,582
2008
192
3,857
10,496
33,419
87,694
135,658
2009
265
3,693
10,508
36,541
126,725
177,732
2010
291
3,411
12,424
50,045
194,284
260,455
2011
198
3,916
13,324
52,368
218,191
287,997
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-31
December 2016
-------
EPA calculated the frequency of detection, by year, for the total coliform results in ground water
systems. Results were generated separately for five bins of free and total chlorine residual
concentrations (Exhibit C.46 and Exhibit C.47, respectively).
Exhibit C.46: Ground Water PWSs: Percent of TC Positives Paired with Free
Chlorine Data by Year (2006 - 2011 )1
4.0%
3.5%
3.0%
2.5%
2.0%
1.5%
1.0%
0.5%
0.0%
I Free Chlorine = 0
I Free Chlorine >0 to 0.2 mg/L
I Free Chlorine >0.2 to 0.5 mg/L
Free Chlorine >0.5 to 1 mg/L
I Free Chlorine > lmg/L
ll
ll
111
2006
2007
2008
2009
2010
2011
1 Different scales were used for the Percent of TC Positives (y-axis) for the free chlorine results (in this exhibit)
compared to the total chlorine results in the next exhibit (Exhibit C.47) to enable a closer look at the free chlorine
results.
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-32
December 2016
-------
Exhibit C.47: Ground Water PWSs: Percent of TC Positives Paired with Total
Chlorine Data by Year (2006 - 2011)
5.0%
4.5%
4.0%
3.5%
3.0%
2.5%
2.0%
1.5%
1.0%
0.5%
0.0%
ITotal Chlorine = 0
ITotal Chlorine >0 to 0.2 mg/L
ITotal Chlorine >0.2 to 0.5 mg/L
Total Chlorine >0.5 to 1 mg/L
ITotal Chlorine > lmg/L
l l I ill..
[ill
2006
2007
2008
2009
2010
I
2011
EPA calculated the frequency of detection, by year, for the E. coli results in ground water
systems. Results were generated separately for five bins of free and total chlorine residual
concentrations (Exhibit C.48 and Exhibit C.49, respectively).
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-33
December 2016
-------
Exhibit C.48: Ground Water PWSs: Percent of EC Positives Paired with Free
Chlorine Data by Year (2006 - 2011)
0.30%
0.25%
0.20%
a. 0.15%
u
g o.io%
0.05%
0.00%
lll.l ill I
1.1
I Free Chlorine = 0
I Free Chlorine >0 to 0.2 mg/L
I Free Chlorine >0.2 to 0.5 mg/L
Free Chlorine >0.5 to 1 mg/L
I Free Chlorine > lmg/L
2006
2007
2008
2009
2010
2011
Exhibit C.49: Ground Water PWSs: Percent of EC Positives Paired with Total
Chlorine Data by Year (2006 - 2011)
¦ Total Chlorine = 0
¦ Total Chlorine >0to 0.2 mg/L
¦ Total Chlorine >0.2 to 0.5 mg/L
1 Total Chlorine >0.5 to 1 mg/L
¦ Total Chlorine > 1 mg/L
¦
1
1
1
ill ¦ IIbiI
ll .
1..
1 . ill..
2006 2007 2008 2009 2010 2011
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-34
December 2016
-------
Exhibit C.50: Number of Total Coliform Samples in Ground Water Paired with Free
and Total Chlorine Data, by Year (underlying data/denominator for Exhibit C.46,
Exhibit C.47, Exhibit C.48 and Exhibit C.49)
Year
Total # TC GW Samples Paired
0
> 0 to 0.2
mg/L
> 0.2 to 0.5
mg/L
> 0.5 to 1
mg/L
> 1 mg/L
Total
Free Chlorine
2006
22,134
31,605
57,915
68,311
33,038
213,003
2007
23,348
33,785
62,516
75,457
35,484
230,590
2008
24,624
33,991
61,121
74,992
38,255
232,983
2009
29,960
34,859
73,936
99,620
84,445
322,820
2010
24,641
35,249
75,647
122,993
106,101
364,631
2011
23,474
36,020
79,102
133,609
121,484
393,689
Total Chlorine
2006
3,609
4,988
20,023
38,897
37,594
105,111
2007
3,947
5,043
19,773
44,780
45,165
118,708
2008
5,056
5,996
21,258
50,196
51,218
133,724
2009
4,704
6,495
22,158
56,811
81,704
171,872
2010
3,765
8,114
25,503
72,378
117,916
227,676
2011
4,103
8,203
26,131
76,504
130,333
245,274
Geographic Analysis
To assess any potential geographic trends in the SYR3 ICR microbial data, EPA calculated the
frequency of detection, by state, for the total coliform results in surface water, ground water and
all systems. Results for all five bins of free and total chlorine residual concentrations were
combined; the percent of TC+ for all systems (SW and GW) are presented in Exhibit C.51.
A total of 34 states/entities provided TC data for surface water and/or ground water systems.
Twenty-eight of those states/entities provided sample data with TC positives. States in the upper
three categories of TC+ measures are located in all parts of the United States. However, a
potential geographic pattern of occurrence is obscured by the lack of data from 23 states. For
example, there are very limited data for the southern Rockies and no data for the Upper Midwest
or the southeast portion of the country. The seven states with the highest occurrence of TC
positives in all systems (SW and GW systems) are Arkansas, Connecticut, Nevada, New York,
Rhode Island, Texas and Vermont. Of these seven states, Texas is the only one that requires a
minimum free chlorine residual in the distribution system; that minimum requirement is equal to
0.2 mg/L. Eight of the 12 states with minimum free chlorine requirements that have data in the
SYR3 ICR microbial dataset were positive for TC in less than 0.5 percent of samples. Refer to
the geographic analysis section of Appendix B for a list of the states with minimum requirements
for free and total chlorine residual in the distribution system.
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-35
December 2016
-------
Exhibit C.51: All PWSs (SW + GW): Percent of TC Positives (2006 -2011)
t.
¥
W
| AmericanSamoa
[ Region 1 Tribes
[ Region 7 Tri bes
| District of Columbia
[ Region 2 Tribes
Region STribes
j Guam
[ Region 3Tribes
j Region 9 Navajo Nation
[ Northern Mariana Islands
Region 4Tribes
| Region 9Tribes
| Puerto Rico
Region STribes
| Region lOTribes
j Virgin Islands
| Region 6Tribes
n
No data in finalSYR3 Microbial Dataset
~
0%samplesTC+
~
>0 - 0.2% sam plesTC+
n
>0.2 -0.5% samp)esTC+
¦
>0.5-1% samples TC+
¦
>1% samples TC+
Exhibit C.52: Number of TC Samples and Percent of TC+, by State (underlying
data for Exhibit C.51)
State/
Region
All Systems (SW + GW)
Surface Water Systems
Ground Water Systems
Total #
TC
Samples
# TC+
Samples
% TC+
Samples
Total #
TC
Samples
# TC+
Samples
% TC+
Samples
Total # TC
Samples
# TC+
Samples
% TC+
Samples
AK
35.396
183
0.52%
19,760
113
0.57%
15,636
70
0.45%
AR
259,507
3,165
1.22%
120,999
1,131
0.93%
138,508
2,034
1.47%
AS
2,241
0
0.00%
0
0
0.00%
2,241
0
0.00%
CT
266,775
3,122
1.17%
156,547
192
0.12%
110,228
2,930
2.66%
HI
8,773
33
0.38%
886
0
0.00%
7,887
33
0.42%
IA
269,118
967
0.36%
88,576
151
0.17%
180,542
816
0.45%
ID
76,066
576
0.76%
16,035
72
0.45%
60,031
504
0.84%
IL
799,476
1,597
0.20%
354,102
499
0.14%
445,374
1,098
0.25%
IN
47,764
72
0.15%
9,218
20
0.22%
38,546
52
0.13%
KS
200,874
1,788
0.89%
69,318
586
0.85%
131,556
1,202
0.91%
KY
296,791
866
0.29%
259,455
669
0.26%
37,336
197
0.53%
ME
1
0
0.00%
1
0
0.00%
0
0
0.00%
MO
107,306
576
0.54%
37,327
119
0.32%
69,979
457
0.65%
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
€-36
December 2016
-------
State/
Region
All Systems (SW -t
GW)
Surface Water Systems
Ground Water Systems
Total #
TC
Samples
# TC+
% TC+
Total #
TC
Samples
# TC+
% TC+
Total # TC
# TC+
% TC+
Samples
Samples
Samples
Samples
Samples
Samples
Samples
MT
43,519
142
0.33%
27,425
46
0.17%
16,094
96
0.60%
NC
323,550
1,584
0.49%
144,120
518
0.36%
179,430
1,066
0.59%
NE
21,785
0
0.00%
21,785
0
0.00%
0
0
0.00%
NM
115,640
462
0.40%
34,475
55
0.16%
81,165
407
0.50%
NN
15,403
0
0.00%
1,734
0
0.00%
13,669
0
0.00%
NV
17,446
253
1.45%
3,393
43
1.27%
14,053
210
1.49%
NY
68,342
890
1.30%
23,428
411
1.75%
44,914
479
1.07%
OH
167,129
292
0.17%
97,530
54
0.06%
69,599
238
0.34%
OK
253,936
2,403
0.95%
185,230
1,185
0.64%
68,706
1,218
1.77%
OR
205,086
954
0.47%
121,722
277
0.23%
83,364
677
0.81%
Rl
6,957
426
6.12%
2,230
33
1.48%
4,727
393
8.31%
TX
449,773
5,873
1.31%
130,763
1,826
1.40%
319,010
4,047
1.27%
VA
327,504
1,780
0.54%
236,258
696
0.29%
91,246
1,084
1.19%
VT
40,616
529
1.30%
14,528
92
0.63%
26,088
437
1.68%
WV
137,475
549
0.40%
101,569
295
0.29%
35,906
254
0.71%
WY
56,240
338
0.60%
32,446
86
0.27%
23,794
252
1.06%
Region 1 Tribes
2,310
0
0.00%
2,303
0
0.00%
7
0
0.00%
Region 4 Tribes
2,192
13
0.59%
208
1
0.48%
1,984
12
0.60%
Region 5 Tribes
10,642
62
0.58%
285
1
0.35%
10,357
61
0.59%
Region 8 Tribes
9,153
89
0.97%
3,127
13
0.42%
6,026
76
1.26%
Region 9 Tribes
24,422
0
0.00%
2,254
0
0.00%
22,168
0
0.00%
Total
4,669,208
29,584
0.63%
2,319,037
9,184
0.40%
2,350,171
20,400
0.87%
EPA also calculated the frequency of detection, by state, for the E. coli results in surface water,
ground water and all systems. Results for all five bins of free and total chlorine residual
concentrations were combined; the percent of EC+ for all systems (SW and GW) are presented
in Exhibit C.53.
A total of 34 states/entities provided TC data in surface water and/or ground water for this
analysis. Twenty-seven of those states/entities provided sample data identified EC positives.
States in the upper three categories of EC+ measures are located in all parts of the United States.
However, a potential geographic pattern of occurrence is obscured by the lack of data from 23
states. For example, there are very limited data for the southern Rockies and no data for the
Upper Midwest or the southeast portion of the country. The four states with the highest
occurrence of EC positives are Connecticut, Kansas, New York and Rhode Island and. Of these
four states, Kansas is the only one that requires a minimum free chlorine residual in the
distribution system; that minimum requirement is equal to 0.2 mg/L. Seven of the 12 states with
minimum requirements that have data in the SYR3 ICR microbial dataset were positive for EC in
less than 0.02 percent of samples.
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-37
December 2016
-------
Exhibit C.53: All PWSs (SW + GW): Percent of EC Positives (2006 -2011)
~
AmericanSamoa
~
Region 1 Tribes
~
Region 7 Tribes
~
District of Colu mbi a
~
Region 2 Tribes
~
Region S Tribes
~
Guam
~
Region 3Tribes
~
Region 9 Navajo Nation
~
Northern Mariana Islands
~
Region 4Tribes
~
Region 9 Tribes
~
Puerto Rico
~
Region 5Tribes
~
Region 10 Tribes
~
Virgin Islands
~
Region 6Tribes
n
No data in finalSYR3 Microbial Dataset
~
0% samples EC+
~
> 0 -0.01% sampl es EC+
~
>0.01-0.02%sampl es EC+
¦
> 0.02 -0.05% sam pi es EC+
¦
> 0.05% sa mples EC+
Exhibit C.54: Number of TC Samples and Percent of EC+, by State (underlying
data for Exhibit C.53)
State/
Region
All Systems (SW + GW)
Surface Water Systems
Ground Water Systems
Total #
EC
Samples
# EC+
Samples
% EC+
Samples
Total #
EC
Samples
# EC+
Samples
% EC+
Samples
Total # EC
Samples
# EC+
Samples
% EC+
Samples
AK
35,396
12
0.03%
19,760
11
0.06%
15,636
1
0.01%
AR
259,507
119
0.05%
120,999
41
0.03%
138,508
78
0.06%
AS
2,241
0
0.00%
0
0
0.00%
2,241
0
0.00%
CT
266,775
172
0.06%
156,547
19
0.01%
110,228
153
0.14%
HI
8,773
3
0.03%
886
0
0.00%
7,887
3
0.04%
IA
269,118
102
0.04%
88,576
17
0.02%
180,542
85
0.05%
ID
76,066
24
0.03%
16,035
4
0.03%
60,031
20
0.03%
IL
799,476
86
0.01%
354,102
41
0.01%
445,374
45
0.01%
IN
47,764
4
0.01%
9,218
1
0.01%
38,546
3
0.01%
KS
200,874
136
0.07%
69,318
39
0.06%
131,556
97
0.07%
KY
296,791
46
0.02%
259,455
35
0.01%
37,336
11
0.03%
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
€-38
December 2016
-------
State/
Region
All Systems (SW -t
GW)
Surface Water Systems
Ground Water Systems
Total #
EC
Samples
# EC+
% EC+
Total #
EC
Samples
# EC+
% EC+
Total # EC
# EC+
% EC+
Samples
Samples
Samples
Samples
Samples
Samples
Samples
ME
1
0
0.00%
1
0
0.00%
0
0
0.00%
MO
107,306
27
0.03%
37,327
14
0.04%
69,979
13
0.02%
MT
43,519
6
0.01%
27,425
2
0.01%
16,094
4
0.02%
NC
323,550
52
0.02%
144,120
14
0.01%
179,430
38
0.02%
NE
21,785
0
0.00%
21,785
0
0.00%
0
0
0.00%
NM
115,640
27
0.02%
34,475
1
0.00%
81,165
26
0.03%
NN
15,403
0
0.00%
1,734
0
0.00%
13,669
0
0.00%
NV
17,446
6
0.03%
3,393
2
0.06%
14,053
4
0.03%
NY
68,342
79
0.12%
23,428
30
0.13%
44,914
49
0.11%
OH
167,129
13
0.01%
97,530
1
0.00%
69,599
12
0.02%
OK
253,936
91
0.04%
185,230
46
0.03%
68,706
45
0.07%
OR
205,086
60
0.03%
121,722
20
0.02%
83,364
40
0.05%
Rl
6,957
14
0.20%
2,230
1
0.05%
4,727
13
0.28%
TX
449,773
206
0.05%
130,763
55
0.04%
319,010
151
0.05%
VA
327,504
76
0.02%
236,258
35
0.02%
91,246
41
0.04%
VT
40,616
14
0.03%
14,528
5
0.03%
26,088
9
0.03%
WV
137,475
26
0.02%
101,569
10
0.01%
35,906
16
0.04%
WY
56,240
17
0.03%
32,446
4
0.01%
23,794
13
0.05%
Region 1 Tribes
2,310
0
0.00%
2,303
0
0.00%
7
0
0.00%
Region 4 Tribes
2,192
1
0.05%
208
0
0.00%
1,984
1
0.05%
Region 5 Tribes
10,642
0
0.00%
285
0
0.00%
10,357
0
0.00%
Region 8 Tribes
9,153
4
0.04%
3,127
1
0.03%
6,026
3
0.05%
Region 9 Tribes
24,422
0
0.00%
2,254
0
0.00%
22,168
0
0.00%
Total
4,669,208
1,423
0.03%
2,319,037
449
0.02%
2,350,171
974
0.04%
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
C-39
December 2016
-------
Appendix D. Producing a Reduced Dataset for Undisinfected Ground
Water Systems
Total Coliform (TC), E. coli (EC) and Fecal Coliform (FC) data, as received, included records
for individual samples. To better understand how TC and fecal indicator (either EC or FC)
positive rates varied by system size, system type, sample type and disinfection practice, EPA
needed to identify which systems were using ground water without disinfection. Then, to
simplify statistical modeling of the TC and EC positives2 rates, the data for each system and
month were reduced to a small number of summary counts: (a) the total number of routine
samples assayed, (b) the number of routine samples testing positive for TC, (c) the total number
of TC positive routine samples tested for EC and (d) the number of routine samples testing
positive for EC. This appendix summarizes the processes for identifying undisinfected ground
water systems and producing summary counts for small undisinfected ground water systems.
EPA analyzed the occurrence of total coliforms in PWSs using undisinfected ground water and
presented the results in Section 6.4 and Appendix F of this document.
Identification of Undisinfected Ground Water Systems
Ground water systems may use disinfectants in different ways. Many do not disinfect at all.
Some may add a disinfectant, typically free chlorine or UV, at the source to achieve 4-log virus
treatment under the Ground Water Rule (GWR). Others may provide this level of treatment but
do not monitor to qualify for 4-log treatment under the GWR. Some may add some chlorine at
the source to oxidize then remove iron and manganese. Other ground water systems may use
chlorine (or, less frequently, chloramines) as a residual disinfectant to provide some public
health protection and improve water quality in the distribution system. Individual ground water
sources within a system may receive different levels of treatment.
EPA conducted an analysis to evaluate the possible differences in coliform occurrence between
disinfected and undisinfected ground water systems. The SYR3 ICR microbial dataset does not
contain a simple data field that identifies the disinfection status of ground water systems. EPA
developed an approach for categorizing the ground water systems from SDWIS states with total
coliforms as disinfected or undisinfected systems. The steps in this process are described below.
1. Identify the subset of GW systems from the SDWIS states with total coliform data.
A total of 83,535 systems (from SDWIS states) submitted total coliform results (2006 to
2011) that passed the list of initial QA/QC checks.3 Of those, 72,582 are GW systems.
2 There are some systems that take a fecal coliform (FC) sample following a TC+ result rather than an EC sample;
thus, FC counts were also included.
3 These initial QA/QC checks included the identification of the following: (1) records marked with sample type
codes other than routine, repeat or confirmation; (2) records marked as not being for compliance; (3) records from
non-public water systems; (4) records from outside of the SYR3 date range; and (5) records from systems missing
inventory information. All of these data were excluded from the process of identifying undisinfected GW systems.
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
D-l
December 2016
-------
2. Use the treatment process data in the SYR3 ICR dataset, as well as SDWIS, to identify
disinfected GW systems.
a. Assume any systems listing 4-log treatment of viruses in the treatment table are
disinfected GW systems.
b. Assume any additional systems that do not report 4-log treatment of viruses but report
that they are disinfected and report chlorine, chloramines, UV, chlorine dioxide are also
disinfected GW systems.
3. Of the remaining systems that were not identified as disinfected GW systems in step 2,
evaluate field free and total chlorine data.
a. Assume any systems with at least one free or total chlorine record >0.1 mg/L are
disinfected GW systems.
b. Assume the remaining systems with no field disinfectant residual data and no disinfected
information in their treatment type are undisinfected GW systems. (Note that the list of
GW systems identified as undisinfected may include systems with free or total chlorine
records < 0.1 mg/L.)
Data Reduction
The SYR3 ICR dataset contains TC, EC, and FC data from 2006 through 2011 for 46 states (41
SDWIS and 5 non-SDWIS states4). The basic suite of QA/QC steps were conducted on the TC,
EC, and FC data. For more details on these QA/QC steps, refer to USEPA (2016e), The Data
Management and Quality Assurance/Quality Control Process for the Third Six-Year Review
Information Collection Rule Dataset. This QA/QC review resulted in the exclusion of any
records that met the following criteria:
records marked with sample type codes other than routine, repeat, or confirmation;
records not marked as being for "compliance";
records from non-public water systems;
records from outside of the SYR3 date range; and
• records from systems missing inventory information.
Additional QA/QC steps were applied that were specific to TC, EC, and FC.5 All records
identified as follows were excluded from the analysis:
4 About 75% of all states currently store and manage at least portions of their compliance monitoring data in the
Safe Drinking Water Information System/State Version (SDWIS/State). The majority of states using SDWIS/State
that submitted data to EPA used a SDWIS Query Extract Tool, developed and provided by EPA, to extract and
compile the EPA-requested compliance monitoring data. The states not using SDWIS/State submitted their
compliance monitoring data "as is," resulting in a variety of formats of datasets submitted to EPA. Furthermore, not
all of the requested data from the non-SDWIS states was in a format usable to EPA for the SYR3 analyses.
5 Note that a detailed QA was not conducted to ensure that all repeat samples had a corresponding routine sample.
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
D-2
December 2016
-------
• Records where PRESENCEINDCODE (presence indicator code) was null or not equal
to either "A" (absent) or "P" (present);
• TC positive (TC+) results without a corresponding EC or FC result;6
• EC and FC results without a corresponding TC+ result; and
• Records from facility type codes other than distribution systems (i.e., only data where
TYPE CODE = "DS" were included in the analysis).
Rather than including a record for each sample assayed, the reduced dataset includes, for each
water system and month, counts of the routine and repeat samples assayed and found to be
positive for TC, EC and FC. Field names for these counts all begin with as shown in Exhibit
D.l.
Exhibit D.1: Descriptions of Field Names in the Undisinfected Ground Water
Systems Reduced Dataset
Field Name
Description
PWSID
Public water system identification number (PWSID)
Month
Month (1 through 12)
Year
Year (2006 through 2011)
Retail Population Served
Retail population served by the water system
Water system type according to federal requirements
System type
C = Community water system
NTNC = Non-transient non-community water system
NC = Transient non-community water system
Water source for the water system.
Source Water Type
GW = Ground Water (included in this category were systems using GW or Purchased GW
[GWP])
SW = Surface Water (included in this category were systems using SW, Purchased SW
[SWP], Ground water Under Direct Influence of Surface Water [GU], and Purchased GU
[GUP])
Disinfecting?
An indication if the system disinfects its water (Y = Yes; blank = No). All systems with a
source water type = "SW" were assumed to be disinfecting. Note: An explanation of the
determination of the ground water systems' disinfection status is included on pages 2 and
3 of this document.
#TC Samples (routine)
The count of routine total coliform (TC) samples
# TC+ Samples (routine)
The count of routine TC positive samples
# EC Samples (routine)
The count of routine E. coli (EC) samples
# EC+ Samples (routine)
The count of routine EC positive samples
# FC Samples (routine)
The count of routine fecal coliform (FC) samples
# FC+ Samples (routine)
The count of routine FC positive samples
# TC Samples (repeat)
The count of repeat TC samples
# TC+ Samples (repeat)
The count of repeat TC positive samples
6 TC+ results were linked with EC and FC samples if they had the same water system ID, water system facility ID,
sample point ID, sample collection date, lab assigned ID, and sample ID. Only the SDWIS states had data in all of
these fields to enable this linkage between TC+ and EC/FC data.
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
D-3
December 2016
-------
Field Name
Description
# EC Samples (repeat)
The count of repeat EC samples
# EC+ Samples (repeat)
The count of repeat EC positive samples
# FC Samples (repeat)
The count of repeat FC samples
# FC+ Samples (repeat)
The count of repeat FC positive samples
In the final "reduced" dataset, there are data for a total of 80,692 water systems located in 39
states. Exhibit D.2 provides an extract of the information included in the final "reduced" dataset.
Note that not all systems have results for all 12 months of each year. Furthermore, there were
some repeat samples that occurred in a different month than their corresponding routine sample;
thus, some system/month/year combinations have counts of repeat samples but no routine
samples.
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
D-4
December 2016
-------
Exhibit D.2: Extract of Reduced Data for Three Systems
PWSID
Month
Year
Population
Served
(Retail)
System
type
Source
Water
Type
Disin-
fecting?
#TC
Samples
(RT)
#TC+
Samples
(RT)
#EC
Samples
(RT)
#EC+
Samples
(RT)
#FC
Samples
(RT)
#FC+
Samples
(RT)
#TC
Samples
(RP)
#TC+
Samples
(RP)
#EC
Samples
(RP)
#EC+
Samples
(RP)
#FC
Samples
(RP)
#FC+
Samples
(RP)
CT0640011
4
2006
388,700
C
sw
Y
218
6
6
18
CT0640011
4
2007
388,700
C
sw
Y
214
3
3
9
CT0640011
4
2008
388,700
c
sw
Y
260
1
1
3
CT0640011
4
2009
388,700
c
sw
Y
196
3
3
9
CT0640011
4
2010
388,700
c
sw
Y
201
1
1
3
CT0640011
4
2011
388,700
c
sw
Y
208
IA3353088
6
2006
6,415
c
GW
Y
7
IA3353088
6
2007
6,415
c
GW
Y
7
IA3353088
6
2008
6,415
c
GW
Y
7
IA3353088
6
2009
6,415
c
GW
Y
7
IA3353088
6
2010
6,415
c
GW
Y
7
IA3353088
6
2011
6,415
c
GW
Y
7
SC1720001
1
2006
15,141
c
GW
1
1
1
1
SC1720001
5
2006
15,141
c
GW
2
2
2
5
SC1720001
7
2007
15,141
c
GW
2
2
2
6
SC1720001
8
2010
15,141
c
GW
1
1
1
4
SC1720001
9
2007
15,141
c
GW
1
1
1
3
SC1720001
9
2008
15,141
c
GW
1
1
1
3
SC1720001
10
2009
15,141
c
GW
1
1
1
3
Six-Year Review 3 Technical Support Document
for Microbial Contaminant Regulations
D-5
December 2016
-------
Appendix E. Analysis of the Generalized Estimating Equation (GEE) and
Generalized Linear Mixed Models (GLMM) as used to Estimate the
Relative Rate of Highly Credible Gastrointestinal Illness (HCGI) by
Colford et al. (2009)
Summary of the Colford et al. (2009) Results
The goal of the Colford et al. paper "was to estimate the efficacy of an in-home water filter to
reduce the risk of highly credible gastrointestinal illness (HCGI) among older adults living in a
community whose tap water met or exceeded current US drinking water standards."
Colford et al. reported that they found a 12% mean reduction in gastrointestinal illness episodes
per year among an elderly population in households using a filter. This finding is based on the
GEE model estimate of the device (active filter v sham filter) rate ratio to be 0.88, with 95%
confidence interval (0.77, 1.00). Hence, the 95% confidence interval for the estimated reduction
is (0%>, 23%). It should be noted that the upper bound of this wide confidence interval suggests
the plausibility that filter use has no reduction (i.e., 0%).
The paper also presents results from the GLMM model which estimates the device rate ratio of
episodes per year to be 0.85, with 95% confidence interval (0.76, 0.94). Exhibit E. 1 compares the
GEE and GLMM device rate ratio confidence intervals, and shows the similarity of the two
intervals, despite the GEE upper confidence limit attaining the value 1.00 while the GLMM
upper confidence limit does not. Such consistency between the two models is expected as the
two models have similar specifications (both models have the same relationship between
predictor variables and gastrointestinal illness, differing essentially only in the way variability is
formulated in the regressions). Colford et al. note, as in Diggle et al (1994), GEE provides 'a
marginal, population-averaged inference' and GLMM provides 'an individual-specific
inference.' This distinction is not important since, as in Hubbard et al (2010), in certain linear
and log-linear (e.g. Poisson as used in Colford et al.) models the parameter estimates from the
GEE and GLMM Poisson regression have equivalent interpretation towards individual averages
and population averages.
The study collected self-reported occurrences of gastrointestinal illness which was then used to
define highly credible gastrointestinal illness (HCGI).
Summary of the Colford et al. (2009) Statistical Approach and Assumptions
The study goal was to estimate the efficacy of an in-home water filter (device) to reduce the risk
of highly credible gastrointestinal illness (HCGI). Each household was to use an active device in
one cycle (6 months) and a sham device in another cycle; the name cross-over study is applied to
such designs where the household is exposed to various treatments in consecutive periods. GEE
and GLMM models were developed with device as the only predictor variable (unadjusted
models). Models were also developed that adjust for a set of covariates, namely, gender, age,
Six-Year Review 3 Technical Support Document E-l
for Microbial Contaminant Regulations
December 2016
-------
self-reported health, number of medications, irritable bowel syndrome at baseline, diarrhea at
baseline, and daily water consumption (adjusted models).
The GEE and GLMM both model the same mathematical relationship shown in Equation (1),
wherey represents either the episodes of HCGI or the days of HCGI, t represents the time (or
person-days) within a cycle, x' represents the predictor variables (in this case, device, cycle,
gender, age, self-reported health, number of medications, irritable bowel syndrome at baseline,
diarrhea at baseline, and daily water consumption), and /? represents the corresponding
parameter estimates for these predictors.
log(E(y)) - log(t) = x'(3 (1)
GEE and GLMM differ in their expression and estimation of the statistical variation in the data.
The statistical variation in GLMM is modeled as three variance components, 1) residual (each
person/cycle randomly varies), 2) person, and 3) household. The statistical variation in GEE is
modeled as a correlation structure among the residuals (e.g. residuals among persons within a
household will be correlated while persons in different households will be uncorrelated). Both
GEE and GLMM apply a Poisson regression to episodes or days of HCGI. Without going into
detail, loosely GLMM uses maximum likelihood to solve for the mathematical relationship and
variance components, while GEE uses weighted least squares to solve for the mathematical
relationship and correlation structure. In any event, the similarities of the two methodologies
dictates that the corresponding results are expected to be consistent.
The paper initially focuses on 8 analyses; combinations of 1) episodes and days of HCGI, 2)
GEE and GLMM, and 3) unadjusted and adjusted (for additional covariates) analyses. The
Davenport Study (Colford et al. 2005) and other studies have found the duration of an HCGI
episode to be reasonably short and constant, hence days of HCGI is highly correlated with
episodes of HCGI. The study states episodes to be the 'primary outcome' and prevalence (i.e.
days) to be the 'secondary outcome, presumably since days, instead of episodes, adds an extra
dimension of variability. Further, literature on longitudinal data suggests adjusting for covariates
related to the outcome will explain more of the between person variation, and result in smaller
standard error estimates for the parameter estimates. Hence, the paper's focus is further narrowed
to the adjusted GEE and GLMM models. The results of these two models has been extracted
from the paper and presented in Exhibit E.8.
The study goal is addressed by the device parameter estimates in Exhibit E.8 and depicted in
Exhibit E. 1. The GEE model estimates the rate ratio to be 0.88, with 95% confidence interval
(0.77, 1.00). The interpretation of this rate ratio, as discussed previously in the study summary, is
an expected 12% reduction in HCGI episodes in households using a filter. The GLMM model
estimates the rate ratio to be 0.85, with 95% confidence interval (0.76, 0.94). The point
estimates, 0.88 and 0.85, do not differ greatly, nor do the respective lower confidence limits and
upper confidence limits. Hence, the GEE and GLMM findings are consistent with one another,
despite the GEE upper confidence limit attaining the value 1.00 while the GLMM upper
confidence limit does not.
Six-Year Review 3 Technical Support Document E-2
for Microbial Contaminant Regulations
December 2016
-------
Review of Modeling
The rationale for presenting GEE and GLMM models in Colford et al. could be questioned.
While the findings of the GEE and GLMM models are consistent with one another, a reader may
question why both are presented and which is the more appropriate.
The literature is rich with comparisons of the GEE and GLMM methodologies. Generally, the
two methods are found to achieve similar results and only occasionally is one method highly
recommended over the other. Exhibit E.3 through Exhibit E.7 show results of several studies
where the GEE and GLMM models have been compared. These studies, like Colford et al., show
the GEE and GLMM findings to be consistent with one another.
Colford et al. correctly assert that GEE provides 'a marginal, population-averaged inference' and
that GLMM provides 'an individual-specific inference'. Diggle et al. (1994) and other papers
espouse on this distinction. This distinction is important for logistic regression, but not so for
Poisson regression as in this study. The parameter estimates from the GEE and GLMM Poisson
regression have equivalent interpretation towards individual averages and population averages.
One often noted distinction between GEE and GLMM models is that the GEE methodology is
more robust than the GLMM methodology. Specifically, Hubbard et al. (2010) and several other
papers note that even if the correlation structure modeled in GEE is wrong, the standard error
estimates can be valid. Further, the GEE approach does not require distributional assumptions
concerning the variance components. The person and household random effects in the GLMM
models of Colford et al. are assumed to follow a normal distribution. This normality assumption
can be difficult to dispute or verify. Thus Colford et al. show some preference to the GEE model
due to its robustness.
The random effects in a GLMM model, when substantiated, can result in much smaller standard
error estimates for the parameter (Park 1993). In Exhibit E.6, the GLMM slope standard error
estimate (0.033) is half the magnitude of the GEE slope standard error estimate (0.065). This is
not the case for the Colford et al. analysis, so no preference towards the GLMM model may be
conferred. More specifically, if the variance component assumption for GLMM correctly
modeled HCGI among individual within households, then the standard errors estimates for the
parameter estimates would likely be much smaller in the GLMM models than in the GEE
models. Since estimates from Colford et al. are not much smaller, this cast some doubt on the
validity of the normality assumptions for GLMM.
How well the model explains the observed data can be assessed by goodness-of-fit statistics,
such as Akaike's information criterion (AIC). Such statistics are useful to compare two GLMM
models or two GEE models, but have limited utility in comparing a GEE model to a GLMM
model. The reason being most such statistics are designed to measure the improvement of fit
between incremental changes in a given model form. There is no 'incremental' difference
between GEE and GLMM. Alternatives such as cross-validation and Bayesian methods, could be
applied to address whether the GEE or GLMM significantly explains the data better.
Six-Year Review 3 Technical Support Document E-3
for Microbial Contaminant Regulations
December 2016
-------
Carriere and Bouyer (2002) conclude GEE models are 'easy to implement and represent a first
solution' and that GLMM 'although more complex, uses all available data and are more suitable
for explicative studies.' Feng et al. (2001) states 'GLMM works well but requires full
distributional assumptions, GEE is too liberal'. Several papers also note that GLMM is less
restrictive concerning missing data; the statistical nomenclature is that GEE requires a missing
completely at random (MCAR) assumption, while GLMM requires only missing at random
(MAR). Colford et al. note that over 80% of households completed both cycles of the study.
However, they also note data from 157 households were discarded due to mislabeled devices.
Colford et al. only present the findings from the GEE model in the Discussion section of their
paper. While not specifically noted in the paper, the robustness of the GEE approach may be the
reason for their preference of GEE over GLMM. Because there is nothing in the GLMM model
results that indicate that it would be preferred and because the two approaches are consistent
with one another, emphasizing the GEE results seems reasonable.
Their conclusion is an expected 12% reduction in HCGI episodes among an elderly population of
households using a filtration device compared to an elderly population of households using no
device. However, the sample size of the study is small (557 households) and the 95% confidence
interval for the estimated reduction includes 0%, so we cannot definitively conclude that there is
a reduction based on the GEE results. Further, in their subgroup analysis, the reduction was
estimated as negligible in cycle 1 and 25% in cycle 2 (possibly indicating the device effect is a
surrogate for some other effect). Moreover, the cycle effect is much larger in magnitude than the
device effect (i.e., the GEE model finding is that there is 45% more HCGI in cycle 1 compared
to cycle 2, whereas the device effect is only 12%). The authors note that such a cycle effect has
been reported by others. This cycle effect could possibly be described as a Hawthorne effect (i.e.,
that knowledge of being in a study has an effect). In addition to the GEE model parameter
estimate, this Hawthorne effect is evident in Exhibit E.2. Exhibit E.2 clearly shows decreasing
incidence of HCGI over time on study. Alternative analyses are suggested in the next section.
Other effects that are larger than the device effect in the GEE results include gender, irritable
bowel syndrome at baseline and diarrhea at baseline.
Suggestions for Further Analyses/Research
Further analysis of this rich dataset could provide additional insights. For example, Exhibit E.2
below shows a clear effect due to time on study. The GEE model crudely incorporates time on
study using the variable cycle. A more elaborate relationship between time on study and HCGI
could be postulated. Additionally, there is a possible seasonal effect (i.e. drinking water in winter
or during drought is maybe more likely to cause HCGI). Modeling such temporal effects would
be of interest. It may also be possible to abstract additional covariates. For example, Beaudeau et
al. (2014) consider the turbidity of the drinking water supply at a point in time as a potential
predictor of HCGI, though it may not be possible to obtain such past data for the Sonoma water
system in the Colford et al. study.
The GEE and GLMM models in this paper each use one of the many variance structures
available. Alternative variance structures could be tested. Also, there are alternative modeling
Six-Year Review 3 Technical Support Document E-4
for Microbial Contaminant Regulations
December 2016
-------
approaches; including Bayesian methods, transition models, and survival analysis. Diggle et al.
has some discussion relating these approaches. Just as GEE and GLMM have common themes,
there are common themes among most of the alternative modeling approaches, hence it seems
likely the alternatives would shed little new light.
Colford et al. models both episodes and days of HCGI. The correlation of episodes and days
indicates the length of HCGI episodes is likely reasonably constant. Regardless, an analysis with
length of HCGI episode seems interesting. It may be that HCGI episodes are shorter in
households using a filtration device, or other hypotheses could be contemplated.
Summary Responses to the Questions Posed in the Technical Direction from EPA
The following provides summary responses to the four EPA questions based on the discussion
presented above:
What role doe the GEE and GLMM mathematical models have in estimating HCGI attributable
incidence to drinking water?
GEE and GLMM are both viable methods to estimate incidence of HCGI (both use Poisson
regression, but modified to account for the longitudinal dataset). Both models provide incidence
rate ratio estimates for device use, which translate into estimates of potential reduction in HCGI
for households using filters.
Why are both models presented; and what are the strengths and weaknesses of each? and
what assumptions are used in the models?
While 8 analyses (combinations of episodes and days of HCGI, GEE and GLMM, and
unadjusted and adjusted) are presented by Colford et al. (Table 3 in the paper), unadjusted
models should not be considered for policy inference since important covariates are not
accounted for. Models with days of HCGI can be discounted since days and episodes of HCGI
are correlated and using days essentially only adds another level of variation to modeling. GEE
estimation of variation is more robust than GLMM. The study goal was to assess filter use
impact on an elderly population, and GEE is designed to provide inference on a population.
GLMM is designed to provide individual inference, but for Poisson regression, GLMM
parameters have both a population and individual interpretation. These slight advantages of GEE
may make it the preferred model to be used for policy inference.
Goodness-of-fit statistics to compare GEE and GLMM were not available. Even if they were
available, they would not indicate statistical superiority of one model over the other. Other
methods, such as cross-validation or Bayesian methods, would be unlikely to statistically
demonstrate superiority of one model over the other.
Assuming that models are used to address factors not specifically accounted for by the study
design, what are the specific factors and are there ways other than the use of GEE and GLMM
that could be used to better understand the health diary data?
Six-Year Review 3 Technical Support Document E-5
for Microbial Contaminant Regulations
December 2016
-------
Both GEE and GLMM point estimates of device effect on HCGI are not adversely affected by
factors not specifically accounted for by the study design. The standard errors may be reduced
(and the corresponding confidence intervals tighter) if the study design could incorporate
additional important factors. Plausible other factors include water quality measures (e.g.,
turbidity) or weather measures (e.g., heat waves or rainy season may effect HCGI incidence).
Future study designs should identify and measure covariates more specifically that might relate
to HCGI due to other sources than drinking water.
Other statistical models (e.g., survival analysis) also model incidence, but would not be expected
to yield any better understanding. Finally, time on study is modeled as simply cycle 1 (first 6
month period using one device) and cycle 2 (next 6 month period using the other device).
Improved measures of time on study are suggested, thereby accounting for the Hawthorne effect.
However, it is unclear whether substantial improvements to the results from the two models used
by Colford et al. would be achieved.
Six-Year Review 3 Technical Support Document E-6
for Microbial Contaminant Regulations
December 2016
-------
Exhibit E.1: GEE and GLMM point estimates and 95% confidence intervals for
device effect
¦
—¦ ¦
0.4
0.5
0.6
0.7
0.8
0.9
1
1.1
Note: Pink squares are GLMM, and blue diamonds are GEE. Colford et al. 2009.
Exhibit E.2: Weekly changes in the number of episodes (per person-year) of
highly credible gastrointestinal illness during the Sonoma Water Evaluation Trial,
2001-06. Colford et al. 2009
¦ Active, Cycle 1
o Sham, Cycle 1
• Active, Cycle 2
~ Sham, Cycle 2
Weeks since enrollment
Six-Year Review 3 Technical Support Document E-7 December 2016
for Microbial Contaminant Regulations
-------
Exhibit E.3: GEE and GLMM model of Thall & Vail (1990) data
GEE
GLMM
P (s.e.)
P (s.e.)
intercept
1.35 (0.16)
1.00 (0.15)
time
0.11 (0.12)
0.11 (0.05)
treatment
0.027 (0.22)
-0.023 (0.20)
interaction
-0.10 (0.21)
-0.10 (0.07)
Note: Poisson regression of seizure counts for 59 patients on treatment (placebo or progabide) at 4
time points.
Exhibit E.4: GEE and GLMM model of Pothoff &Roy data (Verbeke and
Molenberghs 2000)
GEE
GLMM
P (s.e.)
P (s.e.)
intercept girls
17.18 (1.25)
17.18 (1.29)
intercept boys
16.21 (1.04)
16.25 (1.07)
slope girls
0.49 (0.10)
0.49 (0.10)
slope boys
0.80 (0.09)
0.80 (0.09)
Note: Poisson regression of growth data for 11 girls and 16 boys at 4 ages.
Exhibit E.5: GEE and GLMM model of Chroidal Neovascularization Prevention
Trial data (Ying & Liu 2006)
GEE
GLMM
P (s.e.)
P (s.e.)
intercept
-1.00 (0.23)
-1.27 (0.21)
laser treatment
0.054 (0.32)
0.054 (0.23)
Note: Poisson regression of visual acuity for 156 patients with one eye laser treated and other eye
control over 4 years.
Exhibit E.6: GEE and GLMM model of Sly et al data (Burton et al 1998)
GEE
GLMM
P (s.e.)
P (s.e.)
intercept
57.2 (27.1)
59.2 (26.2)
slope time
0.247 (0.065)
0.247 (0.033)
Normal regression of peak expiratory flow measured daily for 12 asthmatic boys over 3 months.
Six-Year Review 3 Technical Support Document E-8
for Microbial Contaminant Regulations
December 2016
-------
Exhibit E.7: GEE, GLMM, and Bayesian GLMM model of Community Hypertension
Assessment Trial (Ma et al 2009)
GEE
GLMM
Bayesian GLMM
Odds Ratio (95% CI)
Odds Ratio (95% CI)
Odds Ratio (95% CI)
intervention v. control
1.14 (0.72, 1.80)
1.10 (0.65, 1.86)
1.12 (0.64, 1.95)
Note: logistic regression of normal blood pressure (v. high) for 1540 elderly patients randomized to intervention or
control over 1 year.
Exhibit E.8: Model results for episodes of highly credible gastrointestinal illness
using GEE and GLMM analysis for an active vs. a sham device in the Sonoma
Water Evaluation Trial, 2001-06
GEE
GLMM
Outcome / Model Specification
RR*
(95% Cl)+
RR*
(95% CI)*
Episodes of highly credible gastrointestinal illness
Adjusted Estimate
Device (Active vs. Sham)
0.88
(0.77, 1.00)
0.85
(0.76, 0.94)
Cycle (1 vs 2)
1.45
(1.29, 1.66)
1.47
(1.32, 1.64)
Male (vs female)
0.76
(0.60, 0.98)
0.64
(0.51, 0.79)
Age (per 10 years)
0.93
(0.83, 1.06)
0.88
(0.75, 1.03)
Self-reported health (vs. Excellent)
Good
0.74
(0.53, 1.03)
0.68
(0.54, 0.86)
Fair
0.87
(0.54, 1.41)
0.81
(0.52, 1.27)
Poor
0.87
(0.52, 1.45)
1.49
(0.64, 3.48)
Number of medications
1.09
(1.05, 1.13)
1.10
(1.06, 1.15)
Irritable bowel syndrome at baseline
1.49
(1.08, 2.06)
1.80
(1.24, 2.61)
Diarrhea at baseline
2.58
(1.93, 3.45)
4.62
(3.69, 5.80)
Total water consumption (per 8-ounce glass)
1.03
(0.98, 1.07)
1.02
(0.97, 1.06)
* RR : Rate ratio (episodes of illness)
f 95% Confidence Intervals for GEE models estimated using exchangeable correlation & robust SEs
$ 95% Confidence Intervals. All GLMM specifications include random intercepts for individual and household.
Note: Extracted from Table 3 of Colford et al. 2009 paper.
Six-Year Review 3 Technical Support Document E-9
for Microbial Contaminant Regulations
December 2016
-------
Appendix F. Occurrence of Total Coliforms / E. coli in Small PWSs Using
Undisinfected Ground Water
F.l. Data Source and Groups
The total conform/A", coli (TC/EC) data used in this analysis originated as part of a data
extraction effort for a large suite of contaminants compiled into a database and provided to EPA
that included the year 2011. Previously, EPA had similarly analyzed TC/EC data from the year
2005 and reported the results in the Revised Total Coliform Rule (RTCR) Economic Analysis
(EA) (USEPA, 2012). EPA has extracted TC/EC data from the 2011 dataset. The TC/EC data
were compiled from PWS or state reports (states often perform TC/EC assays for PWSs). Some
states did not follow uniform procedures in building the TC/EC database. For example, some
states may have only entered total coliform positive data into the database. Data from these states
(i.e., Louisiana, Alabama, and South Carolina) with anomalous record keeping procedures were
removed from this dataset.
Data provided to EPA by most states did not include a data field to indicate disinfection.
However, there typically is a data field for ancillary information such as chlorine residual in the
distribution system. For the 2011 data, EPA developed a multi-step decision tree to identify
undisinfected systems by a process of elimination, using this ancillary information. (See
Appendix D of this document for a description of this process.) Undisinfected PWSs in the 2005
data may have been identified or verified by merging two differing state and national datasets, a
costly step not undertaken for the 2011 data. After applying the decision tree to arrive at a set of
undisinfected systems, EPA did not test the results to evaluate the decision tree result. In
comparing the 2005 and 2011 datasets, EPA observed small differences in ancillary information
in 2005 versus 2011. Also, there is not complete overlap between the 2005 and 2011 data (i.e.,
the same states do not report the same data in the same way in both years).
The complete dataset used in this analysis consists of TC records from about 38,000
undisinfected systems for 2011 (note that the 2005 data analyzed in the RTCR EA included TC
records from about 60,000 undisinfected systems). For modeling purposes, these data were
divided into 27 basic subsets of systems (3 system types, 3 water types and 3 size ranges).
Exhibit F.1: Undisinfected Small Ground Water Systems from SYR3 ICR Dataset
Used for TC Analyses
Size Ranges
Community Water
Systems
Non-Transient Non-
Community Water
Systems
Transient Non-
Community Water
Systems
<101 people served
2,262
2,246
18,538
101 - 1,000 people served
2,450
2,378
9,539
1,001 - 4,100 people served
492
182
143
Six-Year Review 3 Technical Support Document F-l
for Microbial Contaminant Regulations
December 2016
-------
This section discusses data from all three system types and size ranges and one water type—
undisinfected ground water (Exhibit F.l).
EPA assumed that each system has four detection rates: 1) TC detection rates in routine samples;
2) TC detection rates in repeat samples, 3) EC detection rates, given TC detection, in routine
samples and 4) EC detection rates, given TC detection, in repeat samples. In the following
analysis, EPA analyzed only routine samples.
F.2. Data Analysis
This analysis addresses only PWSs that use undisinfected ground water. Because the disinfection
barrier is absent, any public health benefit might be greatest in unprotected undisinfected PWSs.
The purpose of this data analysis is to identify and characterize the groups of PWSs that have
high TC detection rates.
The limited amount of data for the individual small PWSs prevented us from precisely estimating
any particular system's detection rate, but such data from a large number of systems supported
estimation of distributions of detection rates. To estimate the distributions of detection rates,
EPA assumed that each system has two unobserved detection rates: 1) TC detection rates in
routine samples; and 2) EC detection rates, given TC detection in routine samples. For each
system, the observed fraction (number of detects/number of assays) is an imprecise estimate of
the unobserved detection rate.
Routine TC detection rates vary from system to system, even among systems of the same type
and size. The beta distribution serves well to describe these varying rates. EPA did not directly
observe the system-specific detection rates or their distributions, but instead estimated the
parameters of these distributions using the data, summarized as the number of routine TC assays
(N) and the number of routine TC positives (K) for each system. Assuming that the assays for a
particular system are each independent, identically distributed Bernoulli trials, the number of TC
detections for the system is a binomial random variable with parameters N and the unobserved
detection rate, p. For example, if a system were to assay 12 routine TC samples and find 2 to be
positive, the ratio K / N = 1 / 6 would be an imprecise estimate of the detection probability. In
modeling, EPA used the counts K and N, rather than their ratio to inform the likelihood function.
Likelihood is a function of the data (expressed in terms of K and N for each system) and the beta
distribution parameters a and (3. Below, in Equation 1, log likelihood (LL) is expressed as a
function of the data from NSys systems and beta distribution parameters a and (3. NSys is the
number of systems with data and i is an index for systems. Log likelihood is the sum of the NSys
system-specific log likelihoods. Logarithms were used to avoid computational
overflow/underflow issues.
Equation 1
NSys
LL(oc,P) := ^
In
i = 1
T(a + P) a-1 Np-1
p -(1 - p)
r(a)T(p) F
Ni!
Ki!-(Ni-Ki)!
K; Ni-K;
•P (1 -P) I*
Six-Year Review 3 Technical Support Document F-2
for Microbial Contaminant Regulations
December 2016
-------
In Equation 1, a and P are model parameters, Y is the gamma function, p is the unobserved
probability of a TC-positive, K; is the number of TC-positive assay results, Ni is the number of
TC assays, and Ni - K, is the number of negative assay results for system i. The integral is
evaluated for each of the NSys systems having data. The integral simplifies to the result shown
in parentheses in Equation 2:
Equation 2
TT, n, ^ . N^S , f r(q + P)-r(q + Kj)-r(p + Nj - Kj)^|
M4a'P) • 2j I r(a + p +Ni)-r(a)T(P) J
i=l V v / /
C is a constant that depends only on the data, as shown in Equation 3.
Equation 3
cJyj, fr1)
^ lr(Ki + l)-r(Ni-Ki+ l) J
Equation 2 can be expressed in terms of function lbeta (the natural logarithm of the beta
function), as shown in Equation 4:
Equation 4
NSys
y.-'"-Pl: c' ^ (lbeta(a + Ki,P+Ni-Ki)-beta(a,P))
i = 1
Parameterizing the beta distribution as u = ln(a /13) and v = ln(a + |3), and using a wide flat prior
over u and v, these new parameters were estimated in a Bayesian framework using the above
likelihood function. Markov Chain Monte Carlo (MCMC) samples of parameter pairs were
produced using R (The R Foundation for Statistical Computing,
http://www.r-proiect.org/foundation/main.html). using the "LearnBayes" package simcontour
function (Albert, 2007). Results were checked by generating independent MCMC samples using
R and OpenBUGS (Lunn et al., 2009).
F.3 Results
Parameter Estimation
Some of the results are shown below to illustrate the data, data analysis, statistical modeling and
results. To best illustrate the differences between the three system types, Exhibit F.2 displays
only results for the smallest systems: those serving 25 to 100 people.
Six-Year Review 3 Technical Support Document F-3
for Microbial Contaminant Regulations
December 2016
-------
Exhibit F.2: MCMC Samples Predicting TC Detection in PWS Subsets: Three PWS
Types Serving Smallest Populations (25 - 100 People)
° Community
A Non-Transient Non-Community
+ Transient Non-Community
u = ln(alpha / beta)
The exhibit shows three MCMC samples, each MCMC sample consisting of 1,000 pairs of
parameters u and v. Each plotted point is a parameter pair (u, v) describing a realistic beta
distribution of TC detection rates that is consistent with the data. The X-axis (u) is log odds for
mean TC positive detection probability. The mean TC positive detection probability (mean(p))
can be derived from u as follows:
mean(p) = a / (a + (3) = eu / (1 + eu)
Three of these mean values (2 percent, 3 percent, and 4 percent) are shown as vertical dashed
lines in Exhibit F.2. Log odds associated with these percentages are negative because the values
are less than 0.5. For example, the log odds associated with 2 percent is the natural logarithm of
0.02 / 0.98, which is -3.89. The Y-axis (v) is a precision parameter. In terms of the conventional
parameters (a and (3), u is the log of the mean odds (a /(3) and v is the log of the sum a + (3. The
conventional parameters can be determined from u and v as follows:
a = eu + v/(l +eu)
(3 = ev / (1 + eu)
Six-Year Review 3 Technical Support Document F-4
for Microbial Contaminant Regulations
December 2016
-------
A wide scatter of points indicates large uncertainty, due to having fewer data to support the
estimate. A tight set of plotted points indicates smaller uncertainty due to a larger dataset (i.e.,
many systems, assays and PWSs with multiple assays). The filled circle in the center of each
cluster is the sample mean of u and v for the cluster. The means for parameter u correspond to an
average detection rate of 2.5 percent for community water systems, 3.0 percent for non-transient
water systems, and 4.3 percent for transient non-community water systems. Vertical dashed lines
correspond to log odds for average TC detection rates of 2 percent, 3 percent and 4 percent. For
example, the tight cluster of beta distributed probabilities for the transient systems (plus signs)
all have average TC detection rates above 4 percent.
Exhibit F.2 shows the following:
The most precise parameter estimates are for the transient PWSs, as they have the tightest
cluster of points in the figure. This is not surprising, given the large number of transient
PW S s (see Exhibit F. 1).
The least precise parameter estimates are for the non-transient PWSs, as they have the
greatest scattering of points. The numbers of non-transient PWSs and community PWSs
are similar, but monitoring tends to be more frequent for community PWSs and as a
result, there are more data, supporting a more precise estimate for community PWSs.
On average, the highest TC detection rates are for transients, followed by non-transient
and community PWSs.
Community PWSs have the lowest between-system variance (greatest between-system
precision). Transient PWSs and non-transient PWSs have greater between-system
variance, v, suggesting that these groups have more PWSs with detection rates much
greater than the means.
Exhibit F.3 is a similar display showing clusters of beta-distributed probabilities for TC detection
rate distributions of all three size groupings of undisinfected transient PWSs. The figure shows
that, among these PWSs, smaller systems have higher average TC detection rates than larger
systems. A similar result was reported by EPA (USEPA, 2012) for PWSs that use disinfected
ground water or surface water.
Six-Year Review 3 Technical Support Document F-5
for Microbial Contaminant Regulations
December 2016
-------
Exhibit F.3: MCMC Samples Predicting TC Detection in PWS Subsets: Transient
PWS Types Serving Three Small Population Subsets
+ <101
101 to 1000
* 1001 to 4100
u = ln(alpha I beta)
In Exhibit F.3, estimates for the largest transient PWSs (serving 1,001 to 4,100 people) are
widely dispersed due to the small number of systems in this subset. The exhibit shows that the
average detection rate for the larger systems is low (between 1 percent and 2 percent), compared
to the smaller transient PWSs. MCMC samples for the two smallest subsets overlap and are
precise (tightly clustered) due to the large numbers of systems in these subsets. MCMC sample
means are displayed as small open circles. Based on the mean of u, the average detection rates
are 4.3 percent for systems serving fewer than 101, 4.1 percent for those serving 101 to 1,000
people, and 1.3 percent for systems serving 1,001 to 4,100 people.
Six-Year Review 3 Technical Support Document F-6
for Microbial Contaminant Regulations
December 2016
-------
Exhibit F.4 and Exhibit F.5 show estimates for community and non-transient PWSs. Again, the
effect of system size is shown. Smaller systems have greater average detection rates and more
between-system variability.
Exhibit F.4: MCMC Samples Predicting TC Detection in PWS Subsets: Community
PWS Types Serving Three Small Population Subsets
-4.8 -4.6 4.4 -4.2 -4.0 -3.8 -3.6
u = h(alpha / beta)
Six-Year Review 3 Technical Support Document F-7
for Microbial Contaminant Regulations
December 2016
-------
Exhibit F.5: MCMC Samples Predicting TC Detection in PWS Subsets: Non-
Transient PWS Types Serving Three Small Population Subsets
LD
-4.6 -4.4 -4.2 -4.0 -3.8 -3.6 -3.4
u = ln(alpha / beta)
F.4 Detection Rates and Risk
The public health significance of TC detection is uncertain. However, TC detection has utility as
a relative risk marker, perhaps indicating infiltration of recent precipitation. Even within sets of
PWSs with low average detection rates, individual PWSs can have detection rates in the upper
tail of the distribution, and much greater than the average detection rate. Thus, EPA
hypothesized that public health hazard is high for PWSs having high TC detection rates.
To illustrate the relative hazard, Exhibit F.6 and Exhibit F.7 show the distribution of routine TC
detection rates for the smallest of the community, non-transient and transient PWSs using
undisinfected ground water. The cumulative distribution functions shown are based on MCMC
sample mean parameter values (u and v). Exhibit F.6 shows that, among the three PWS types,
transient PWSs have the largest percentage with high TC detection rates (e.g., above 15 percent
or any other potential hazard marker percentage).
Six-Year Review 3 Technical Support Document F-8
for Microbial Contaminant Regulations
December 2016
-------
Exhibit F.6: Detection Rate Distribution Functions for Small (<101) PWSs Based
on MCMC Sample Mean Parameter Values
#8.3% of transient systems have positive rates of at least 15%
Q_
14.1 % of transient systems have positive rates of at least 10%
o _
* — 25.5% of transient systems have positive rates of at least 5%
Transient Non-Community
Non-Transient Non-Community
Community
o _
0.00
0.05
0.10
0.15
0.20
0.25
0.30
Routine TC Detection Rate
Exhibit F.7 shows the fraction of PWSs having routine TC detection rates above selected values.
A significant fraction of smaller systems have high TC detection rates. For example, 5 percent of
transient non-community systems serving populations fewer than 101 individuals (about 2,500
undisinfected systems) have TC detection rates of at least 20 percent.
The smallest rate in Exhibit F.7 (5 percent) is of special interest because observing 5 percent or
more positives in a month triggers an assessment under the RTCR in systems that assay 40 or
more samples per month (larger PWSs). For the smaller PWSs that assay fewer than 40 samples
per month, two TC positive samples trigger an assessment. Notice that about one in four of the
smallest transient non-community systems are estimated to have positive rates of 5 percent or
more.
Exhibit F.7: Routine Total Coliform Detection Rates in Undisinfected PWS
Systems Serving <101 People
Detection Rate
Community
Non-Transient Non-
Community
Transient Non-
Community
5% or more
16% of systems
17% of systems
25% of systems
10% or more
6.5% of systems
9.5% of systems
14% of systems
15% or more
2.5% of systems
5.7% of systems
8.3% of systems
20% or more
1.0% of systems
3.5% of systems
5.0% of systems
30% or more
0.16% of systems
1.4% of systems
1.8% of systems
Six-Year Review 3 Technical Support Document F-9
for Microbial Contaminant Regulations
December 2016
-------
EPA re-analyzed the 2005 TC data and analyses (60,000 wells from a slightly differing set of
states, with undisinfected wells determined by merged databases, using the same analytical
solution used for the 2011 data). EPA found that, for the 2005 data, the maximum likelihood
estimate of the average TC detection rate was 6 percent, as compared with 5 percent for the 2011
data, for the transient PWSs serving populations less than 101 individuals. For this same
grouping of PWSs, in re-examining the tail of the distribution, EPA found 4.6 percent (for 2005
data) versus 5 percent (for 2011 data) of PWSs had a TC detection rate of 20 percent or more.
In the 2011 data, about 5 percent of TC detections were positive for E. coli. This rate appears
relatively unchanged between 2005 and 2011. However, the response to an E. coli detection has
changed due to the promulgation of the Ground Water Rule and RTCR. As a result, an E. coli
detection may require a corrective action to find the fecal contamination source and end the
contamination. Treatment, such as installing disinfection may be required by the state. Because
the number of E. coli detections is small as compared with TC detections, EPA was unable to
determine precise estimates of E. coli detection rate distributions. EPA found no significant
differences in the average EC detection rates across PWS sizes and types.
Six-Year Review 3 Technical Support Document F-10
for Microbial Contaminant Regulations
December 2016
------- |