570391001
o
..: l
•T 0
GUIDANCE MANUAL
FOR
COMPLIANCE WITH THE
FILTRATION AND DISINFECTION REQUIREMENTS
FOR
PUBLIC WATER SYSTEMS
USING
SURFACE WATER SOURCES
MARCH 1991 EDITION
I
5'
o
o
UJ
a
o\
\o
iH
o
00
Z
U
SCIENCE AND TECHNOLOGY BRANCH
CRITERIA AND STANDARDS DIVISION
OFFICE OF DRINKING WATER
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C.
-------
GUIDANCE MANUAL
FOR
COMPLIANCE WITH THE
FILTRATION AND DISINFECTION REQUIREMENTS
FOR
PUBLIC WATER SYSTEMS
USING
SURFACE WATER SOURCES
for
Science and Technlogy Branch
Criteria and Standards Division
Office of Drinking Water
U.S. Environmental Protection Agency
Washington, D.C.
Contract No. 68-01-6989
by
Malcolm Pirnie, Inc. HDR Engineering, Inc.
100 Eisenhower Drive 5175 Hillsdale Circle
Paramus, New Jersey 07653 Eldorado Hills, CA 95630
October, 1990
-------
Acknowledgments
Preparation of this document involved important contributions from
many people in two consulting engineering firms, several private
consultants, and the United States Environmental Protection Agency
(USEPA). Malcolm Pirnie, Inc., with technical contributions
provided by HDR Engineering Inc., conducted the day-to-day work
under contract with the USEPA.
Principal authors from Malcolm Pirnie were David J. Hiltebrand and
Linda Averell Wancho. Personnel from HDR involved in this work
were Jerry Troyan and Perri P. Garfinkel. -Additional personnel
from Malcolm Pirnie who contributed to either the technical content
or the preparation of the manual included: John E. Dyksen, James K.
Schaefer, Scott L. Phillips and Peter B. Galant.
Private consultants who contributed to the document included:
Dr. Appiah Amirtharajah, Georgia Institute of Technology, Atlanta,
GA; Dr. Ovadia Lev, Hebrew University, Jerusalem, Israel;
Dr. Vincent Oliveri, formerly John Hopkins University, Baltimore,
MD; Dr. Phillip C. Singer, University of North Carolina, Chapel
Hill, NC; Dr. Mark Wiesner, Rice University, Houston, TX.
Preparation of the document was overseen by Stig Regli, the USEPA
project officer. Valuable technical review and major contributions
to the text were provided by Thomas Grubbs, Office of Drinking
Water, USEPA, and Leigh Woodruff, Region X, USEPA.
Special thanks are given to the following individuals working for
USEPA whose review and comment on numerous drafts greatly
contributed to the evolution of this document: John Davidson,
Office of Policy, Planning and Evaluation; Edwin Geldreich,
Drinking Water Research Division, Office of Research and
Development; John Hoff, formerly with Drinking Water Research
Division, Office of Research and Development; Walt Jakubo.wski,
Environmental Monitoring and Support Laboratory, Office of Research
and Development; Dr. Gary Logsdon, formerly with Drinking Water
Research Division, Office of Research and Development; Kevin
Reilly, Region I; Margaret Silver, Office of General Council; and
Jim Westrick, Technical Support Division, Office of Drinking Water.
Appreciation is also expressed to state public health officials,
representatives of the drinking water industry, academicians, and
the American public for their participation in submitting timely
and insightful comments without which this document would not have
been possible.
Some of the appendices have primary authors which are noted on the
corresponding cover pages.
-------
TABLE OF CONTENTS
1. INTRODUCTION 1-1
2. GENERAL REQUIREMENTS 2-1
2.1 Application 2-1
2.1.1 Types of Water Supplies 2-2
2.1.2 Determination of Applicable Sources 2-3
2.2 Treatment Requirements 2-13
2.3 Operator Personnel Qualifications 2-14
3. COMPLIANCE FOR SYSTEMS NOT FILTERING 3-1
3.1 Source Water Quality Criteria 3-1
3.1.1 Coliform Concentrations 3-2
3.1.2 Turbidity Levels 3-4
3.2 Disinfection Criteria 3-6
3.2.1 Inactivation Requirements 3-6
3.2.2 Determination of Overall Inactivation
for Residual Profile, Multiple
Disinfectants and Multiple Sources
and Multiple Sources 3-19
3.2.3 Demonstration of Maintaining a Residual 3-29
3.2.4 Disinfection System Redundancy 3-32
3.3 Site-Specific Conditions 3-34
3.3.1 Watershed Control Program 2-35
3.3.2 On-site Inspection 3-3"
3.3.3 No Disease Outbreaks 3-4C
3.3.4 Monthly Coliforn, MCL 3-41
3.3.5Total Trihalomethane (TTHM) Regulations 3-42
4. DESIGN AND OPERATING CRITERIA FOR FILTRATION AND
DISINFECTION TECHNOLOGY 4-1
4.1 Introduction 4-1
4.2 Selection of Appropriate Filtration Technology 4-1
4.2.1 General Descriptions 4-2
4.2.2 Capabilities 4-3
4.2.3 Selection 4-8
4.3 Available Filtration Technologies 4-9
4.3.1 Introduction 4-9
4.3.2 General 4-10
4.3.3 Conventional Treatment 4-13
4.3.4 Direct Filtration • 4-17
4.3.5 Slow Sand Filtration 4-18
4.3.5 Diatomaceous Earth Filtration 4-21
4.3.7 Alternate Technologies 4-22
4.3.8 Nontreatment Alternatives 4-24
4.4 Disinfection 4-25
4.4.1 General 4-25
4.4.2 Recommended Removal/Inactivation 4-25
4.4.3 Total Trihalomethane (TTHM) Regulations 4-29
-------
TABLE OF CONTENTS (Continued)
Page
5. CRITERIA FOR DETERMINING IF FILTRATION AND DISINFECTION
ARE SATISFACTORILY PRACTICED 5-1
5.1 Introduction 5-1
5.2 Turbidity Monitoring Requirements 5-1
5.2.1 Sampling Location 5-1
5.2.2 Sampling Frequency 5-2
5.2.3 Additional Monitoring 5-3
5.3 Turbidity Performance Criteria 5-3
5.3.1 Conventional Treatment or
Direct Filtration 5-5
5.3.2 Slow Sand Filtration 5-7
5.3.3 Diatomaceous Earth Filtration 5-7
5.3.4 Other Filtration Technologies 5-8
5.4 Disinfection Monitoring Requirements 5-8
5.5 Disinfection Performance Criteria 5-9
5.5.1 Minimum Performance Criteria Required
Under the SWTR 5-9
5.5.2 Recommended Performance Criteria 5-10
5.5.3 Disinfection By-Product Considerations 5-12
5.5.4 Recommended Disinfection System Redundancy 5-14
5.5.5 Determination of Inactivation by
Disinfection ' 5-14
5.6 Other Considerations 5-28
6. REPORTING 5-1
5.1 Reporting Requirements for Public Water Systems
Not Providing Filtration 5-1
6.2 Reporting Requirements for Public Water Systems
Using Filtration 5-3
7. COMPLIANCE 7-1
7.1 Introduction 7-1
7.2 Systems Using a Surface Water Source
(Not Ground Water Under the Direct
Influence of Surface Water) 7-1
7.3 Compliance Transition with Current NPDWR
Turbidity Requirements 7-3
7.4 Systems Using a Ground Water Source Under
the Direct Influence of a Surface Water 7-4
7.5 Responses for Systems not Meeting -the SWTR Criteria 7-6
7.5.1 Introduction 7-5
7.5.2 Systems Not Filtering 7-6
7.5.3 Systems Currently Filtering 7-8
B. PUBLIC NOTIFICATION 8-1
-------
TABLE OF CONTENTS fContlnued^
LIST OF FIGURES
Figure Following
No. Description Page
2-1 Steps to Source Classification 2-4
3-1 Determination of Inactivation for Multiple
Disinfectant Application to a Surface Water
Source 3-22
3-2 Individually Disinfected Surface Sources Combined
at a Single Point 3-25
3-3 Multiple Combination Points for Individually
Disinfected Surface Sources 3-25
4-1 Flow Sheet for a Typical Conventional Water
Treatment Plant 4-13
4-2 Flow Sheet for Typical Softening Treatment Plants 4-14
4-3 Flow Sheet for a Typical Direct Filtration Plant 4-17
4-4 Flow Sheet for a Typical Direct Filtration Plant 4-17
with Flocculation
LIST OF APPENDICES
Appendix Description Page
A Use of Particulate Analysis for Source and Water
Treatment Evaluation A-l
B Institutional Control of Legionella B-l
C Determination Of Disinfectant Contact Time C-l
D Analytical Requirements of the SWTR and a Survey
of the Current Status of Residual Disinfectant
Measurement Methods for all Chlorine Species and Ozone D-l
E Inactivation Achieved by Various Disinfectants E-l
F Basis for CT Values F-l
G Protocol for Demonstrating Effective Disinfection G-l
H Sampling Frequency for Total Coliforms in the
Distribution System H-l
I Maintaining Redundant Disinfection Capability 1-1
-------
TABLE OF CONTENTS (Continued^
9. EXEMPTIONS 9-1
9.1 Overview of Requirements 9-1
9.2 Recommended Criteria 9-2
9.3 Compelling Factors 9-3
9.4 Evaluation of Alternate Water Supply Sources 9-7
9.5 Protection of Public Health 9-7
9.6 Notification to EPA 9-11
LIST OF TABLES
Table Following
No. Description Page
2-1 Survey Form for the Classification of Drinking
Water Sources 2-7
4-1 Removal Capabilities of Filtration Processes 4-3
4-2 Generalized Capability of Filtration Systems to
Accommodate Raw Water Quality Conditions 4-8
6-1 Source Water Quality Conditions for Unfiltered
Systems 6-4
5-2 Long Term Source Water Quality Conditions for
Unfiltered Systems 6-4
6-3 CT Determination for Unfiltered Systems -
Monthly Report to Primacy Agency 5-4
5-4 Disinfection Information for Unfiltered Systems -
Montly Report to Primacy Agency 5-4
6-5 Distribution System Disinfectant Residual Data for
Unfiltered and Filtered Systems - Monthly Report to
Primacy Agency 5-4
6-5 Monthly Report to Primacy Agency for Compliance
Determination - Unfiltered Systems 6-4
6-7 Daily Data Sheet for Filtered Systems 6-4
5-8 Monthly Report to Primacy Agency for Compliance
Deterimination-Filtered Systems 5-4
7-1 Requirements for Unfiltered Systems 7-3
7-2 Requirements for Filtered Systems 7-3
-------
TABLE OF CONTENTS
LIST OF APPENDICES
"Spendix Description pjne
J Watershed Control Program j-1
K Sanitary Survey K_l
L Small System Considerations L-l
M Protocol for Demonstration of Effective Treatment M-l
N Protocols for Point-of-Use Treatment Devices N-l
0 Guidelines to Evaluate Ozone Disinfection 0-1
-------
1. INTRODUCTION
This Guidance Manual complements the filtration and disinfection
treatment requirements for public water systems using surface water
sources or ground water under the direct influence of surface water
promulgated in 40 CFR Part 141, Subpart H. In this manual, these
requirements are referred to as in the Surface Water Treatment Rule
(SWTR).
The purpose of this manual is to provide guidance to United States
Environmental Protection Agency (USEPA) Regional Offices, Primacy Agencies
and affected utilities in the implementation of the SWTR, and to help
assure that implementation is consistent. For example, the SWTR sets
treatment requirements which apply to a large range of source water
conditions. The guidance manual suggests design, operating and perform-
ance criteria for specific surface water quality conditions to provide the
optimum protection from microbiological contaminants. These recommenda-
tions are presented as advisory guidelines only; unlike the provisions of
the SWTR, these recommendations are not mandatory requirements, In many
cases, it will be appropriate to tailor requirements to specific
circumstances; the guidance manual is designed to give the Primacy Agency
flexiDility in establishing the most appropriate treatment requirements
for the systems within their jurisdiction.
Throughout this document, the term "Primacy Agency" refers to a
State with primary enforcement responsibility for public water systems or
"primacy," or to mean EPA in the case of a State that has not obtained
primacy.
In order to facilitate the use of this manual, it has been
structured to follow the framework of the SWTR as closely as possible.
Brief descriptions of the contents of each section of this manual are
presented in the following paragraphs.
Section 2
This section provides guidance for determining whether a *ater
supply source is subject to the requirements of the SWTR including the
determination of whether a ground water source is under the direct
1-1
-------
influence of surface water,i.e. at risk for the presence of Giardia cysts
or other large microorganisms. The overall treatment requirements of the
SWTR are also presented, along with recommendations for the qualifications
of operator personnel.
Section 3
For systems which are subject to the requirements of the SWTR and
which do not currently provide filtration, this section provides guidance
to the Primacy Agency for determining if a given system:
Meets the source water quality criteria
Meets the disinfection requirements including:
99.9 and 99.99 percent inactivation of Giardia cysts and
viruses and application of the CT (disinfectant residual
concentration x contact time) concept
Point of entry to distribution system requirements
Distribution system requirements
Provision for disinfection system redundancy
Maintains an adequate watershed control program
Meets the on-site inspection requirements
Has not had an identified waterborne disease outbreak
Complies with the requirements of the revised Total Coliform
Rule
Complies with Total Trihalomethane (TTHM) Rule
Section 4
This section pertains to systems which do not meet the requirements
to avoid filtration outlined in Section 3 and therefore are required to
install filtration. Guidance is given for the selection of an appropriate
filtration technology based on the source water quality and the capabili-
ties of various technologies to achieve the required performance criteria.
In addition, recommended design and operating criteria are provided for
different filtration technologies.
1-2
-------
Section 5
Section 5 presents guidance to the Primacy Agency for determining
compliance with the turbidity and disinfection performance requirements,
and in turn, whether filtration and disinfection are satisfactorily
practiced. Recommendations are made for the level of disinfection to be
provided in order to meet the overall treatment requirements of the SWTR.
This section describes how to evaluate the adequacy of disinfection using
CT or other methods.
Section 6
Section 6 provides guidelines to the Primacy Agency for establishing
the reporting requirements associated with the SWTR. The requirements
include report content and frequency, and are applicable to both filtering
and nonfiltering systems.
Section 7
This section provides an overview of the schedule for Primacy
Agencies and utilities to meet the requirements of the SWTR. Examples are
presented to provide guidance for corrective measures which can be taken
by systems which are not in compliance with the treatment requirements.
Section 8
This section presents guidance on public notification. Included are
examples of events which would require notification, language for the
notices and the methods of notification.
Section 9
Section 9 provides guidance to the Primacy Agency for determining
whether a system is eligible for an exemption. The criteria for
eligibility for an exemption include:
Compelling factors (economic or resource limitations)
No available alternate source
Protection of public health
This section also provides guidance for evaluating the financial
capabilities of a water system, reviewing the availability of alternate
sources and suggests interim measures for protecting public health.
1-3
-------
Appendices
The manual also contains appendices which provide more detailed
guidance in specific areas. These include:
Appendix A - EPA Consensus
Method for Giardia cyst Analysis
Several procedures are available for Giardia cyst analysis in water.
In 1983 the USEPA held a conference to establish a consensus on the
procedure to be used in the future. This consensus method would promote
uniformity in testing and provide a basis for future conparisons. The
consensus method and the background data used to develop it are presented
in this appendix.
Appendix B - Institutional
Control of Legionella
Filtration and/or disinfection provides protection from Legionel la.
However, it does not assure that recontamination or regrowth will not
occur in the hot water or cooling systems of buildings within the
distribution system. This appendix provides guidance for monitoring and
treatment which can be used by institutional systems for the control of
Legionella.
Appendix C - Determination of Disinfectant
Contact, Time
In many cases, the determination of disinfectant contact times
needed to evaluate the CT of a water system will necessitate the use of
tracer studies. This appendix provides guidance for conducting these
studies. In some cases it may not be practical to conduct a tracer study.
For such cases guidance is given for estimating the detention time based
on the physical configuration of the system.
1-4
-------
Appendix D - Analytical Requirements
of the SWTR and A Survey of the Current
Status of Residual Disinfectant
Measurement Methods for all Chlorine
Species and Ozone
This appendix includes a listing of the analytical methods required
under the SWTR. An executive summary of a report on the analytical
methods used to measure the residual concentrations of the various
disinfectants is included. The reliability and limitations of each of the
methods are presented.
Appendix E - Inactivations Achieved
by the Various Disinfectants
This appendix presents the log inactivations of Giardia cysts and
viruses which are achieved at various CT levels by chlorine, chlorine
dioxide, chloramines and ozone. Inactivations of viruses achieved by UV
absorbance are also included.
Appendix F - Basis for CT Values
This appendix provides the background and rationale utilized in
developing the CT values for the various disinfectants. Included is a
paper by Clark and Regli, 1990, in which a mathematical model was used in
the determination of CT values for free chlorine.
Appendix G - Protocol for Demonstrating
Effective Disinfection
This appendix provides the recommended protocols for demonstrating
the effectiveness of chloramines, chlorine dioxide and ozone as primary
disinfectants.
Appendix H - Sampling Frequency for
Total Colifonns in the Distribution System
The sampling frequency required by the revised Total Colifonn Rule
54 FR 27544 (June 29, 1989) is presented in this appendix.
1-5
-------
Appendix I - Maintaining
Redundant Disinfection Capability
This appendix details the conditions and equipment which should be
maintained by a system using chlorine, chlorine dioxide, ozone or
chloramines to assure that compliance with the SWTR requirement for
redundant disinfection is met.
Appendix J - Watershed Control Program
This appendix provides a detailed outline of a watershed program.
This program may be adjusted by the Primacy Agency to serve the specific
needs of a particular water system.
Appendix K - Sanitary Survey
This appendix provides guidance for conducting a comprehensive
sanitary survey of a supply source and its treatment and delivery to the
consumer. Suggested elements of an annual on-site inspection are included
in Section 3.
Appendix L - Small System Considerations
This appendix describes difficulties which may be faced by small
systems in complying with the SWTR along with guidelines for overcoming
these difficulties.
Appendix M - Protocol for the
Demonstration of Effective Treatment
This appendix presents pilot study protocols to evaluate the
effectiveness of an alternate filtration technology in meeting the
performance requirements of the SWTR. It presents the use of particle
size analysis for demonstrating the actual removal of Giardia cyst
achieved by a treatment train. Guidance for conventional and direct
filtration plants to demonstrate that adequate filtration is being
maintained at effluent turbidities between 0.5 and 1 Nephelometnc
Turbidity Unit (NTU) is also included.
1-6
-------
Appendix N - Protocol for
Point-of-Use Treatment Devices
In some limited cases, it may be appropriate to install point-of-use
(POD) or point-of-entry (POE) treatment devices as an interim measure to
provide protection to the public health. This appendix provides a
protocol for evaluating and determining the efficacy of POU/POE treatment
devices.
Appendix 0 - Guidelines to
Evaluate Ozone Disinfection
The CT evaluation used for other disinfectants is inappropriate for
ozone. This appendix presents alternative methods for evaluating the
disinfection effectiveness of ozone systems.
1-7
-------
2. '-NFRM REQUIREMENTS
2. 1 ^DDl ication
The SWTR pertains to all public water systems wnich utilize a surface
Abater source or ground water source under the direct influence of surface
water. The SWTR defines a surface water as all waters which are open to
the atmosphere and subject to surface runoff. Ground water under the
direct influence of surface water is defined as: any water beneath tne
surface of the ground with (i) significant occurrence of insects or other
macroorgani sms , algae, organic debris, or large-diameter pathogens such as
Giardia lamblia, or (ii) significant and relatively rapid shifts in water
characteristics such as turbidity, temperature, conductivity, or pH which
closely correlate to cl imatological or surface water conditions. Direct
influence must be determined for each individual source in accordance with
criteria established by the Primacy Agency. The Primacy Agency criteria
may provide for documentation of well construction and geology, with field
evaluation, or site-specific measurements of water quality as explained in
Section 2.1.2.
Saline water sources such as the ocean are not generally considered
to be subject to the requirements of the SWTR because of the low survival
time of pathogens in a saline environment (Geldreich, 1989). Pathogens
generally can only survive a few hours in saline water and any remaining
pathogens should be removed or inactivated during desalination. However,
-: -'s JP to the Primacy Agency's discretion to determine which systems
must meet the- SWTR requirements. In cases where there is a sewage
discharge located near the water intake, it may be appropriate for the
?rimacy Agency to require the system to comply with the SWTR.
The traditional concept that all water in subsurface aquifers is free
from pathogenic organisms is based upon soil being an effective filter
that removes microorganisms and other relatively large particles by
straining and antagonistic effects (Bouwer, 1978). In most cases
pathogenic bacteria retained in the soil find themselves in a hostile
environment, are not able to multiply and eventually die. However, some
jnderground sources of drinking water may be subject to contamination by
pathogenic organisms from the direct influence of nearby surface waters.
Only those subsurface sources which are at risk to contamination from
Giardia cysts will be subject to the requirements of the SWTR. G
2-1
-------
cysts generally range in size from 7 to 12 urn. Suosurface sources /.men
:nay De at ris* to contamination from bacteria and enteric viruses, out
Anicn are not at risk from Giardia cysts will be regulated either under
tne 7ota1 Coliform Rule or fortncoming disinfection treatment requirements
for ground Caters. EPA intenas to promulgate disinfection requirements
for ground water systems in conjunction with regulations for disinfection
by-products by 1992.
2.1.1 Types of Water SUDD!ies
Surface Waters
Surface water supplies that are often used as sources of drinking
water include two major classifications, running and quiescent waters.
Streams, rivers and brooks are examples of running water, *hile lakes,
reservoirs, imooundments and ponds are examples of quiescent waters. The
exposure of surface waters to the atmosphere results in exposure to
precipitation events, surface water runoff and contamination with micro
and macroorganisms resulting from activities in their surrounding areas.
These sources are subject to the requirements of the SWTR.
Systems with rain water catchments not subject to surface runoff
;e.g. roof catchment areas) are not considered vulnerable to contamination
from animal populations which carry protozoan cysts pathogenic to humans
and are thus not subject to the SWTR requirements. However, such systems
snou:d at least provide disinfection to treat for potential bacterial and
;iral contamination coming from bird populations.'
Ground Caters under Direct Infuence of Surface Water
Ground *ater sources whicn may be subject to contamination with
:athogemc organisms from surface waters include, sorings, infiltration
gai'eries, wells or other collectors in subsurface aquifers. The
*oilowing section presents a recommended procedure for determining wnether
a source will be subject to the requirements of the SWTR. These
determinations are to be made for each individual source. If the
determination will involve an evaluation of water quality, eg. particulate
analysis, it is important that these analyses be made on water taken
One study (Markwell and Shortridge, 1981) indicates that a
cycle of waterborne transmission and maintenance of influenza
virus may exist within duck communities, and that it is
conceivable for virus transmission to occur in this manner to
other susceptible animals, including humans.
2
-
-------
directly from tne source sna not on :1ended *ater or water *-:in :re
distribution system.
2.1.2 Determination of Applicable Sources
The Primacy Agency has the responsibility for determining which -\ater
suoplies must meet the requirements of the SWTR. However, it ^s the
responsibility of the water purveyors to provide th-3 Primacy Agency with
the information needed to make this determination. This section provides
guidance to the Primacy Agency for determining which water supplies are
surface waters or ground waters directly influenced by a surface water and
are thereby subject to the requirements of the SWTR. Following the
determination that the source is subject to the SWTR, the requirements
enumerated in Sections 2.2 and 2.3 must be met.
The Primacy Agency must develop a program for evaluating ground *ater
sources for direct influence by December 30, 1990. All community ground
water systems must be evaluated by June 29, 1994, while all non-community
systems must be evaluated by June 29, 1999. Primacy Agencies with an
approved Wellhead Protection (WHP) Program, may be able to use the WHP
program's requirements which include delineation of wellhead protection
areas, assessment of sources of contamination and implementation cf
management control measures. These same requirements can be -sea -"or
meeting the requirements of the watershed control program for ground /.ate*-
^rder the direct influence of a surface water.
A multiple step aoproach has been developed as the recommended r.etncd
of determining whether a ground water source is under direct influence cf
a surface warter. This approach includes the review of information
garnered during sanitary surveys. As defined by the USEPA, a sanitary
s-rvey is an on-site review of the water source, facilities, equipment
operation and maintenance of a public water system for the purpose of
evaluating the adequacy of such source, facilities, equipment, operation
and maintenance for producing and distributing safe drinking water.
Sanitary surveys are required under the Total Coliform Rule and may be
reauired under the forthcoming disinfection requirements for ground water
systems as a condition for obtaining a variance or for determining the
level of disinfection required. Therefore, it is recommended that the
determination of direct influence be correlated with the sanitary surveys
conducted under these other requirements.
2-3
-------
A. Source Evaluation Protocol
As illustrated on Figure 2-1, the determination of ^nether a source
-, s SuDject to tne requirements of tne SVnR may "ivc'ive one or more of tne
folI owing steos:
1. A review of the records of the system's source(s) to determine
wnether the source is obviously a surface water, i.e. pona,
lake, streams, etc.
2. If the source is a well, determination of whether it is clearly
a ground water source, or whether T"urther analysis is needed
to determine possible direct surface water influence.
3. A complete review of the system's files followed by a field
sanitary survey. Pertinent information to gather in the file
review and field survey includes: source design and construc-
tion; evidence of direct surface water contamination; water
Quality analysis; indications of wateroorne disease outbreaks;
operational procedures (i.e. pumping rates, etc.); and customer
complaints regarding water quality or water related infectious
i1Iness.
4. Conducting particulate analyses and other water quality
sampling and analyses.
Step 1. Records Review
A review of information pertaining to eacn source should be carries
out to -centify those sources which are obvious surface waters. "These
Aould include ponds, lakes, streams, rivers, reservoirs, etc. If tre
source is a surface water, then tne S'wTR would apply, and criteria 11 tre
ru^e wou;d reea to De applied to Getermine if filtration is necessary. ~.~
the source is n-ot an obvious surface water, then further analyses, as
presepted in Steps 2, 3, or 4, are needed to determine if the SWTR wi'l
appl>. If tre source is a well (vertical or horizontal), go to Step 2.
If the source is a spring, infiltration gallery, or any other subsurface
source, proceed to Step 3 for a more detailed analysis.
Step 2. Review of Well Sources
while most well sources have historically been considered to ce
ground water, recent evidence suggests that some wells, especially shaTow
wells constructed near surface waters, may be directly influenced oy
surface water. One approach in determining whether a well is subject to
contamination by surface water would be to evaluate the water quality :'
the well by the criteria in Step 4. However, this process is rather :-,-ne
2-4
-------
s2
0 0 =
— = " 0)
CO ^ — 3 £
« > TJ 5= re
r 5 * -? fn -
0)
>
- 12 f\
*> '= °- 3 £ U
2 o w w S5
3 Z = 0 CQ
5 1§
^5
>
''. 8.
1 "
^ en «
^ c «
0 g = _
^ tn to
c£^ f
^M C
^? OJ
< 2
>
\
\_
x*~N
1 O 1
W
5 * 2
o» « "- - «
C = C 0 >
•—CO C 3 tm
a 0 « -o =
co c 2.5^-
«25 to^c5^
82 5 -o =
" = -2 S S
3 •— 5 CO CO
0 C £ CO
to — cc
r __ _ _ — ..
>
>
Q.
a.
<
^
I
*
o
Z
(/)
0)
o
Q
CC
H-
to
i
f
/*
i ,_,, ,
/
k
•V X"
P Q
k
X
D
( 1 Issl laKft
>»^^- 52S— Scu^^-S
— >> ^ 3-— COCO wSs^T?
*Ss ^ spi i§«sl
5S5.1_ p.-SEo^-StSSlS
« S 8 1 i-S8i2« is^ s-s
sll 1 |*?> islsi
ocS = 1 g 5 s Is^o
"-" § < 8 S Ili5u
\ <•> a to =
.^
w
b. -—
/»T\ s «>\
a) 2 t/f
/\ ^B *•
« 0 « o
^ in * £S
3 S 9 o
W 8 5 '"" i
= 1=11
l*tfSi
X3 J* i
O « CO
i
yj
>
C
a
•41
* 1 (01
V x^
r
?
j
"a.
a.
<
CC
(f
)
i
-J
i
o
K
<
XSSIFIC
PS TO SOURCE CL/
LU
i
f—
to
1
y-~
1
CM
LU
g
u.
-------
consuming and labor intensive. In an attempt to reduce the effort "eeaed
to evaluate well sources, a set of criteria has oeen Developed to identify
-veils in deep, /veil protected aquifers which are not suDject to contamina-
tion from surface water. While these criteria are not as definitive as
water quality analysis, it is believed that they provide a reasonable
degree of accuracy, and allow for a relatively rapid determination for a
large number of well sources in the U.S.
Wells with perforations or a well screen less than or equal to 50
feet in depth are considered to be shallow wells, and should be evaluated
for direct surface influence according to steps 3 and/or 4. For wells
greater than 50 feet in depth, State or system files should be reviewed
for the criteria listed below:
1. The well construction should include:
A surface sanitary seal using bentonite clay, concrete
or other acceptable material.
A well casing that penetrates a confining bed.
A well casing or collector laterals that are only
perforated or screened below a confining bed.
The importance of evaluating the hydrogeology of wells or
collectors, even those more than 200 feet from a surface water,
cannot be overstated. The porosity and transmissivity of
aquifer materials, hydrologic gradients, and continuity of
confining layers above screens or perforations may need to be
considered in detail for some sources. Porous aquifer material
is more likely to allow surface water to directly influence
groOrrrd water than finer grained materials. In addition, nigh
well pumping rates may alter the existing hydrologic gradient.
Ground water flow direction may change sucn that surface water
is drawn into a collector, whereas under low pumping rates it
may not. Evaluating pumping rate effects and other hydrogeolo-
gic information must be done on a site specific basis.
If information on well construction or hydrogeology are
incomplete or raise questions regarding potential surface water
influence, a more detailed analysis in steps 3 and 4 should be
considered.
2. The casing or nearest collector lateral should be located at
least 200 feet from any surface water.
3. The water quality records should indicate:
2-5
-------
No record of total coliform or -'seal coliform contamina-
tion in untreated samples collected over tne past tnree
years.
No history of turpidity proplems associated with the
source.
No history of known or suspected outbreak of Giardia. or
other pathogenic organism associated with surface water
(e.g. Cryptosporidium), which has been attributed to that
source.
If data is available for particulate matter in the well there
should be:
No evidence of particulate matter associated *itn
surface water.
If data is available for turbidity or temperature from the «ell
and a nearby surface water there should be:
No turbidity or temperature data which correlates
to that of a nearby surface water.
.N'ells that meet all of the criteria listed above are not subject to
the requirements of the S'w'TR, and no additional evaluation is reeded.
.•.'ells tnat do not meet all the requirements listed reouire *urtre"
evaluation in accordance with Steps 3 and/or 4 to determine whether cr not
they are directly influenced by surface water.
Step 3. On-site Inspection
For sources other than a *ell source, the State or system files
should 5e reviewed for the source construction and -\ater quality
conditions as listed in Step 2. Reviewing historical records in State cr
system files is a valuable information gathering tool for any source.
However, the results may be inconclusive. A sanitary survey in the field
may be helpful in establishing a more definite determination of whether
the water source is at risk to pathogens from direct surface *ater
influence.
Information to obtain during an on-site inspection include:
Evidence that surface .%ater enters the source through defects
in the source such as >ack of a surface seal on wells, infi'-
2-6
-------
tration gallery 'atera's exposed to sj^face ,\ater, springs open
to tne atmosphere, surface runoff entering a spring or otner
collector, etc.
Distances to obvious surface water sources.
If the survey indicates that the well is subject to direct surface
water influence, the source must either be reconstructed as explained
later in this section or it must be treated in accordance with the
requirements for the SwTR. If the survey does not show conclusive
evidence of direct surface water influence, the analysis outlined in Step
4 should be conducted.
The Washington State Department of Social and Health Services has
developed a form to guide them and provide consistency in their evaluation
of sources for surface water influence (Notestine & Hudson, 1988). Table
2-1 provides a copy of this form as a guide for evaluating sources.
Step 4. Paniculate Analysis and Other Indicators
a. Surface *'ater Indicators
Paniculate analysis is intended to identify organisms whicn cn'y
occur in surface waters as opposed to ground waters, and wnose presence ^n
a ground water /vould clearly indicate that at least some surface water 135
:een mixed with it. The EPA Consensus Method in Appendix A can be .sea
~or 3jardia cyst analysis.
In 1986-Hoffbuhr et. al. listed six parameters identifiable in a
particulate analysis which were believed to be valid indicators of sjrt'ace
contain1, nar'on of ground water. These were: diatoms, rotifers, coccicr, a,
plant debris, insect parts, and Giardia cysts. Later work by Notestine
and Hudson (1988) found that rnicrobiologists did not all define plant
debris in the same way, and that deep wells known to be free of airect
surface water influence were shown by particulate analysis to contain
"plant debris" but none of the other five indicators. Their work suggests
that "plant debris" may not currently be a useful tool in determining
direct surface water influence, but may be in the future when a stancara
definition of "plant debris" is developed. Therefore, it is recommended
that only the presence of the other five parameters; diatoms and certa-n
other algae, rotifers, coccidia, insect parts, and Gia^dia. be used ;s
2-7
-------
TABLE 2-1
SURVEY FORM "OR THE CLASSIFICATION OF DRINKING dATER SOURCES
General
Utility Name (ID*)
2. Utility Person(s) Contacted
3. Source Type (As snown on state inventory)
Spring Horizontal Well Vertical
Infiltration System Shallow Well Well
4. Source Name Year constructed
5. Is this source used seasonally or intermittently? No Yes
If yes, are water quality problems the reason? No res
5. Has there ever been a waterborne disease outbreak associated with
this source? Yes No If yes, explain
Have there been turbidity or bacteriological MCL violations witmn
the last five years associated with this source? No res
If yes, describe frequency, cause, remedial action (s) taKen
3. nave there been consumer complaints within the past five years
associated with this source? No Yes If yes, discuss
nature, frequency, remedial action taken
9. Is there any evidence of surface water intrusion (pH, temperature,
conductivity, etc. changes) during the year? Yes No
Ifyes, describe
If not, submit supporting data.
10. Sketch of source in plan view (on an additional sheet)
-1-
-------
Sha11cw Wei 1s
1. Does the /veil meet good sanitary practices regarding location, con-
struction, seal etc. to prevent the entrance of surface water?
Yes No If no, describe the deficiencies
2. What is the depth of the well? (ft)
Elevation of top of casing? (ft msl)
Elevation of land surface? (ft msl)
3. Hydrogeology (Attach copy of well log or summarize it on reverse)
a. Depth to static water level? (Feet)
b. Drawdown? (Feet)
c. What is the depth to the highest screen or perforation?
(Feet)
d. Are there impervious layers above the mghest screen of
perforation?
Yes No Unknown
If yes, please describe
4. Is there a permanent or intermittent surface water within 200 feet
of the well? Yes No If yes, describe (type, distance
etc.) and submit location map
rthat is the elvation of normal pool?
Elevation of 100 year flood level?
Elevation of Dottom of lake or riverT
Additional comments:
-2-
-------
Sori
a. what is the size of the catcnment area (acres)?
b. Give a general description of the area (terrain; vegetation;
soil etc.)
2. What is the vertical distance between the ground surface and the
nearest point of entry to the spring collector(s) (feet)?
3. How rapidly does rainfall percolate into the ground around the
spring?
Percolates readily; seldom if ever any runoff.
Percolates readily but there is some runoff in heavy rain.
Percolates slowly. Most local rainfall ponds or runs off.
Other
4. Does an impervious layer prevent direct percolation of surface water
to the collector(s)? Yes No Unknown
5. Is the spring properly constructed to prevent entry of surface
«ater? Yes No
6. Sediment
a. Is the spring box free of debris and sediment? Yes No
D. when was it last cleaned (Date)
How often does it need to be cleaned? (month)
How much sediment accumulates between cleaning?
inches)
Additional comments:
-3-
-------
[nitration Systems
,vhat are the shortest distances (vertical and horizontal separating
the collector from the nearest surface water? (Feet)
2. Does turbidity of the source vary 0.2 NTU or more throughout the
year? Yes No Not measured
If yes, describe how often and how much (pH, temperature,
conductivity, etc.)
3. Additional Comments
Survey Conducted By: Date:
Decision? Surface Impacted Source Yes No If
further evaluation needed (paniculate analysis, etc.)
-------
indicators of direct surface contamination. In audition, 7 jm) organisms /vhich are clearly of surface water origin sucn
as Diphilobothrium are present, these snould also be considered as
iraicators of direct surface water influence.
b. Interpretation
Since standard methods have not been developed specifically for
particulate analysis, there has not been consistency in the way samples
have been collected and analyzed. Differences in the degree of training
and experience of the microbiologists has added further to the difficulty
in comparing results from sample to sample, and system to system. The
current limitations in sample collection and analytical procedures must be
considered when interpreting the results. Until standardized methods are
developed, the EPA Consensus Method included in Appendix A is recommended
as the analytical method for particulate analysis. The following is a
discussion of the significance of finding the six indicators identified
above.
Identification of a Giardia cyst in any source water should be
considered conclusive evidence of direct surface water influence. The
seated presence of diatoms in source water should be considered as
conclusive evidence of direct surface water influence. However, it is
•-iportant that this determination be based on live diatoms, and not e^ptv
5''ica skeletons which may only indicate the historical presence of
s-rface water.
Bluegreen, green, or other chloroplast containing a'igae reduire
sjn light for"~their metabolism as do diatoms. For that reason tre-r
^eoeated presence in source water should also be considered as conclusive
ev-cence of direct surface water influence.
Hoffbuhr (1986) indicates that rotifers and insect parts are
indicators of surface water. Others have pointed out though that rotifers
do not require sunlight, and not all rotifers require a food source sacn
as algae which originates in surface water.' Their nutritional require-
ments may be satisfied by organic matter such as bacteria, or decomposing
soil organic material, not necessarily associated with surface *ater.
More precise identification of rotifers, i.e. to the species level, -s
necessary to determine the specific nutritional requirements of tne
rotifer(s) present. Further information on identifying rotifer species
and on which species require food sources originating in surface *ater
2-8
-------
would be vaiuao'e, but is not readily availaole at this time, .-nthcut
knowledge of .\nich species is present, the finding of rotifers indicates
that the source 's either a) directly influenced by surface water, or D)
it contains organic matter sufficient to support the growth of rotifers.
It could be conservatively assumed based on this evidence alone that such
a source is directly influenced by surface water. However, it is
recommended that this determination be supported by other evidence, eg.
the source is near a surface water, turbidity fluctuations are signifi-
cant, etc.
Insects or insect parts likewise may originate in surface water, from
the soil, or they may be airborne in uncovered sources. If insects are
observed in a particulate analysis sample, it should be confirmed if
possible that there is no other route by which insects could contaminate
the source other than surface water. For example, if a spring is sampled,
and the cover is not well constructed, it is possible that insects found
in a sample were airborne rather than waterborne. Insects which spend a
portion of their lifecycle in water are the best indicators of direct
surface water influence, for example, larvae of mayflies, stoneflies,
damselflies, and dragonflies. Terrestrial insects should not be ruled out
as surface water indicators though, since their accidental presence ;n
surface water is common.
riowell, (1989) has indicated that some insects may burrow and the
finding of eggs cr burrowing larvae (eg. chironormds) may not be good
indicators of direct surface water influence. For some insects this rcay
be true, but the-distance which insects burrow in subsurface sediments is
expected to be small, and insect larvae are generally large in comparison
to G ^ a r_d i a cysts. Until further research suggests otherwise, it -is
recommended that insects or insect parts be considered strong evidence of
surface water influence if not direct evidence in and of themselves. The
strength of this evidence would be increased if the source in question is
near a surface water, and particulate analysis of the surface water found
similar insects.
Coccidia are intracellular parasites which occur primarily in verte-
brates, eg. animals and fish, and live in various tissues and organs
including the intestinal tract (eg. Cryptosporidium). Though not
frequently identified by normal particulate analysis techniques, coccidia
are good indicators of direct surface water contamination since they
2-9
-------
'-equire a vertebrate host or hosts and are generally large in size (10 -
20 urn or greater). Cryptosporidium is commonly found in surface water,
but due to its small size (4-6 urn) it is not normally identified without
specific antibody staining techniques.
Other macroorganisms (>7 urn) which are parasitic to animals and fish
may be found and are good indicators of surface water influence. Examples
include, but are not limited to, helminths (e.g., tape worm cysts),
ascaris, and Diphyllobothrium.
c. Sampling Method
A suggested protocol for collecting samples is listed below.
Sampl ing -Procedure
Samples should be collected using the equipment outlined in the
EPA Consensus Metnod included in Appendix A.
- , Location
Samples should always be collected as close to the source as
possible, and prior to any treatment. If samples must be taken
after disinfection, samples should be noted and analyzed as
soon as possible.
Number
A minimum of two samples should be collected during the period
the source is most susceptible to surface water influence.
Such critical periods will vary from system to system and will
need to be determined case by case. For some systems, it may
be one or more days following a significant rainfall (eg. 2"
in 24 hours). For other systems it may be a period of maximum
flows and stream turbidities following spring snowmelt, or
during the summer months when water tables are elevated as a
resiTlt of irrigation. In each case, particulate samples should
be collected when the source in question is most effected. A
surrogate measure such as source turbidity or depth to water
table may be useful in making the decision to monitor. If
there is any ambiguity in the particulate analysis results,
additional samples should be collected when there is the
greatest likelihood that the source will be contaminated by
surface water.
Volume
Sample volume should be between 500 and 1000 gallons, and
should be collected over a 4 to 8 hour time period. It is
preferable to analyze a similar (+/- 10%) volume of water for
all sources, preferably a large volume, although this may not
always be possible due to elevated turbidity or sampling
logistics. The volume filtered should be recorded for all
samples.
2-10
-------
d. Other Indicators
A number of other indicators could be used to provide supportive
evidence of surface influence. While particulate analysis probably
orcvides the most direct evidence that pathogens from surface water could
be migrating into a ground water source, other parameters such as
turoidity, temperature, pH and conductivity could provide supportive, but
less direct, evidence.
Turbidity fluctuations of greater than 0.5 - 1 NTU over the course
of a year may be indicative of surface water influence. Considerable
caution should be used when evaluating turbidity changes though, since the
turbidity could be caused by very small particles (< lum) not originating
in a surface water or it could be that larger particles are being filtered
out and only the very smallest particles migrate into the water source.
Only ground water sources at risk to contamination from Giardia or other
large pathogens (> 7 urn) are subject to the SWTR requirements.
Temperature fluctuations may also indicate surface water influence.
Fortunately these are easy to obtain and if there is a surface water
within 500 feet of the water source, measurements of both should be
recorded for comparison. Large changes in surface water temperature
closely followed by similar changes in source temperature would be
indicative of surface water influence. Also, temperature changes (in
degrees F) of greater than 15 to 20% over the course of a year appear to
be a characteristic of some sources influenced by surface water (Randall,
1970). Changes in other chemical parameters such as pH, conductivity,
haraness,etc. e-ould also be monitored. Again, these would not give a
direct indication of whether pathogens originating in surface water were
present, but could indicate whether the water chemistry was or was not
similar to a nearby surface water and/or whether source water chemistry
changed in a similar pattern to surface water chemistry. At this time no
numerical guidelines are available to differentiate what is or is not
similar, so these comparisons are more qualitative than quantitative.
B. Seasonal Sources
Some sources may only be used for part of the year, for example
during the summer months when water usage is high. These sources should
not be excluded from evaluation and, like other sources, should be
evaluated during their period(s) of highest susceptibility. Particular
2-11
-------
attention should be given to those sources which appear to ba directly
influenced by surface water during part of the year. There may be times
during which these subsurface water sources are not influenced by surface
-vater and other times when they are part or all surface water. If that is
the case, then it is critical that careful testing be done prior to,
during and at the end of the use of the source. This should be done over
several seasons to account for seasonal variation. In practice, it is
preferable to use sources which are less vulnerable to contamination since
susceptible sources will necessitate ongoing monitoring and close
attention to operation.
C. Modification of Sources
Sources directly influenced by surface water may be altered in some
cases to eliminate the surface water contamination. Primacy Agencies may
elect to allow systems with such sources to modify the construction of the
source and/or the area surrounding the source in an effort to eliminate
surface water contamination. Since this could be expensive and take
considerable time to evaluate for effectiveness, careful consideration
should be given to the decision to modify a source. In deciding whether
source modification is appropriate, systems and Primacy Agencies snoula
consider the following points:
Is the cause of the surface water contamination known? If the
specific cause or point of surface water contamination is not
known, it will not be possible to determine an effective
control strategy. Further, there may be several reasons why
the source is susceptible to direct surface water influence.
For~example, an infiltration gallery may receive surface water
because some of its laterals are exposed in the bed of a nearby
stream, and also because laterals distant from the stream are
shallow and are affected by surface runoff. Simply modifying
or eliminating one or the other set of laterals in this case
would not entirely eliminate surface water influence.
What is the likelihood that modification of the source will be
effective? Assuming that the source of contamination has been
identified, the expected effectiveness of control measures
should be evaluated. If the cause is relatively evident, a
crack in a well casing or an uncovered spring box for example,
then there is a high degree of confidence that an effective
solution could be developed. Should the nature of the contami-
nation be more diffuse, or widespread, then the merits of
spending time and money to modify the source should be careful-
ly considered. In the case of the example above, eliminating
the use of the laterals under the stream will solve part of the
2-12
-------
problem. However, without considerably more hydrogeologic
information about the aquifer and the placement of the otner
laterals, it is not clear what, if any, control measures would
effectively eliminate direct surface water influence in those
laterals distant from the stream.
If a source is identified as being directly influenced by surface
water, and it is decided to attempt to modify it, interim disinfection
practices which will ensure at least 99.9% inactivation of Giardia should
be considered. Methods and levels of disinfection which can be used to
achieve such removals can be found in S141.72 (a) of the SWTR and in
Section 3.2 of this manual.
A partial listing of types of modifications which could be undertaken
includes:
Diverting surface runoff from springs by trenching, etc.
Redeveloping springs to capture them below a confining layer.
Covering open spring collectors.
Reconstructing wells to install sanitary seals, and/or to
screen them in a confined (protected) aquifer.
Repairing cracks or breaks in any type of source collector trat
allows the entry of surface contaminants.
Discontinue the use of infiltration laterals which intercept
surface water.
An extended period of monitoring should follow reconstruction (eg,
through at least two years or critical periods) to evaluate whether the
source is still directly influenced by surface water. Preferably
particulate analysis would be used to make such evaluations, but it may te
helpful to use simpler measures, such as temperature and turbidity, as
screening tools. Longer term monitoring at critical times may also be an
appropriate agreement between the system and the Primacy Agency if there
is still doubt about the long term effectiveness of the solution.
If modification is not feasible, another alternative to avoid having
to comply with the SWTR may be to develop a new well either deeper or at
a different location.
2-13
-------
2.2 Treatment Requirements
According to the SvJTR, all community and noncommunity public water
systems which use a surface water source or a ground water under the
direct influence of a surface water must achieve a minimum of 99.9 percent
(3-log) removal and/or inactivation of Giardia cysts, and a minimum of
99.99 percent (4-log) removal and/or inactivation of viruses. In the SWTR
and this manual, "viruses" means viruses of fecal origin which are
infectious to humans by waterborne transmission. Filtration plus
disinfection or disinfection alone may be utilized to achieve these
performance levels, .depending on the source water quality and site
specific conditions. The SWTR establishes these removal and/or inactiva-
tion requirements based on G1 ardia and viruses because this level of
treatment will also provide protection from heterotrophic plate count
(HPC) bacteria and Legionella' as required in the SDWA amendments.
Guidelines for meeting the requirements of the SWTR are provided in
the remainder of this manual as outlined in Section 1. All systems must
meet the operator qualifications presented in Section 2.3.
2.3 Operator Personnel Qualifications
The SWTR requires that all systems must be operated by qualified
personnel. It is recommended that the Primacy Agency set standards for
operator qualifications, in accordance with the system type and size. In
order to accomplish this, the Primacy Agency should develop a method of
evaluating an operator's competence in operating a water treatment system.
Primacy Agencies which do not currently have a certification program are
thereby encouraged to implement such a program. An operator certification
program provides a uniform base for operator qualifications and an
organized system for evaluating these qualifications.
It is recommended that plant operators have a basic knowledge of
science, mathematics and chemistry involved with water treatment and
supply. The minimum requirements for at least one key staff member should
include an understanding of:
In the SWTR and this manual "Legionella" means a genus of
bacteria, some species of which have caused a type of pneumonia
called Legionnaires Disease; the etiologic agent of most cases
of Legionnaires Disease examined has been L_.. pneumophila.
2-14
-------
The principles of ^ater treatment and distribution and their
characteristics
The uses of potable water and variations in its demand
The importance of water quality to public health
The equipment, operation and maintenance of the distribution
system
The treatment process equipment utilized, its operational
parameters and maintenance
The principles of each process unit (including the scientific
basis and purpose of the operation and the mechanical compo-
nents of the unit)
Performance criteria such as turbidity, total coliform, fecal
coliform, disinfectant residual, pH, etc. to determine opera-
tional adjustments
Common operating problems encountered in the system and actions
to correct them
The current National Primary Drinking Water Regulations, the
Secondary Drinking Water Regulations and monitoring and
reporting requirements
- 'Methods of sample collection and sample preservation
Laboratory equipment and tests used to analyze samples (where
appropriate)
The use of laboratory results to analyze plant efficiency
Record keeping
Customer relations
Budgeting and supervision (where appropriate)
Training in the areas listed above and others is available through
the American Water Works Association (AWWA) training course series for
*ater supply operations. The course series includes a set of four
training manuals and one reference book as follows:
Introduction to Water Sources and Transmission (Volume 1)
Introduction to Water Treatment (Volume 2)
Introduction to Water Distribution (Volume 3)
Introduction to Water Quality Analyses (Volume 4)
2-15
-------
Reference Handbook: Basic Science Concepts and ADD 1 '.canons
Instructor Guide and Solutions Manual for Volumes 1, 2, 3 and
4
These manuals are available through the American Water Works Associa-
tion, 6666 West Quincy Avenue, Denver, Colorado 80235 USA, (303) 794-7711.
The State of California also offers a series of training manuals for
water treatment plant operators prepared by the California State
University School of Engineering in Sacramento. The manuals include:
1. Water Supply System Operation. (1 Volume)
2. Water Treatment Plant Operation. (2 Volumes)
These operator training manuals are available from California State
University, Sacramento, 6000 J Street, Sacramento, California 95819, phone
(915) 454-6142.
Completion of an established training and certification program will
provide the means of assuring that the operators have received training in
their respective area, and are qualified for their position. The
education and experience requirements for certification should be
commensurate with the size and the complexity of the treatment system. At
the present time, some states have instituted a certification program
wnile others have not. Following is a summary of the basic contents of a
certification program, which can serve as a guide to the Primacy Agency in
developing a complete program.
Board of examiners for the development and implementation of
the program.
Classification of treatment facilities by grade according to
the size and technology of the facilities.
Educational and experience requirements for operators of the
various treatment facilities according to grade.
A written/oral examination to determine the knowledge, ability
and judgement of the applicants with certification obtained
upon receiving a passing grade.
Renewal program for the license of certification, including the
requirement of additional coursework or participation in
workshops.
The certification program should provide technica^y qualified
personnel for the operation of the plant.
2-15
-------
The extensive responsibility which is placed on the operating
personnel warrants the development of an outline of the responsibilities
and authority of the personnel members to aid them in the efficient
operation of the plant. The major responsibilities which should be
delegated in the outline of responsibilities include: the normal
day-to-day operations, preventive maintenance, field engineering, water
quality monitoring, troubleshooting, emergency response, cross-connection
control, implementation of improvements, budget formulation, response to
complaints and public/press contact. A reference which the Primacy Agency
may utilize in developing the outline is "Water Utility Management
Practices" published by AWWA.
2-17
-------
3. CRITERIA FOR SYSTEMS NOT FILTERING
The provisions of the Surface Water Treatment Rule (SWTR) require
that filtration must be included in the treatment train unless certain
criteria are met. These criteria are described in this chapter. They
include:
Source Water Quality Conditions
1. Colifonn concentrations (total or fecal).
2. Turbidity levels.
Disinfection Criteria
1. Level of disinfection.
2. Point of entry disinfection.
3. Distribution system disinfection.
4. Disinfection redundancy or automatic shutoff.
Site-Specific Criteria
1. Watershed control program.
2. On-site inspections.
3. No waterborne disease outbreaks.
4. Complies with the total coliform MCL.
5. Complies with the Total Trihalomethane (TTHM) regulation.
Currently this only applies to systems serving more than
10,000 people.
The purpose of this section is to provide guidance to the Primacy
Agency for determining compliance with these provisions.
3.1 Source Hater Quality Criteria
The first step in determining if filtration is required for a given
surface water supply is to determine whether the supply meets the source
water quality criteria as specified in the SWTR. If the supply does not
meet the source water quality criteria, changes in operation to meet the
site-specific criteria may improve the water quality so that the source
3-1
-------
criteria will be met. However, if the Primacy Agency believes that the
source water quality criteria and/or the site-specific criteria cannot be
met, or that filtration is appropriate regardless, the Primacy Agency may
require the installation of filtration without a complete evaluation to
determine whether the system meets all the criteria required to avoid
filtration.
Sampling Location
The SWTR requires that source water samples be collected at a loca-
tion just prior to the "point of disinfectant application," i.e., where
the water is disinfected and no longer subject to surface runoff. For
example, a system which has multiple reservoirs in series, where each of
the reservoirs has previously been disinfected and receives surface
runoff, must take the raw water sample(s) just prior to the point of
disinfection or disinfection sequences used for calculating the CT
[disinfectant residual (mg/L) x contact time (min.)]. Disinfected water
in reservoirs receiving surface runoff cannot be counted toward CT credit.
It is also not appropriate for systems to monitor the source water after
the "point of disinfectant application" even if disinfection from this
point is not used for calculating CT credit.
3.1.1 Coliform Concentrations: The SWTR states that, to avoid
filtration, a system must demonstrate that either the fecal coliform
concentration is less than 20/100 ml fir the total coliform concentration
is less than 100/100 ml in the water prior to the point of disinfectant
application in 90 percent of the samples taken during the six previous
months. Where monitoring for both parameters has been or is conducted,
the rule requires that only the fecal coliform limit be met. However, EPA
recommends that the analytical results for both total coliforms and fecal
colifomis be reported. In addition, if the turbidity of a surface water
source is greater than 5 NTU and the surface source is blended with a
ground water source to reduce the turbidity, EPA recommends that the high
turbidity water prior to blending meet the fecal coliform source water
qua!ity criteria.
Elevated coliform levels in surface water indicate higher probabili-
ties of fecal contamination, some of which could be protected from
exposure to disinfection by embodiment in particulate matter. Blending of
3-2
-------
the surface water with ground water to reduce coliform levels may obscure
the indication of such possible effects. Thus, EPA does not recommend
blending to reduce coliform levels in the source water. Furthermore, EPA
does not recommend blending to reduce turbidity levels in cases where
elevated fecal contamination may be masked.
Ongoing monitoring is required to ensure that these requirements are
continually met. The samples may be analyzed using either the multiple
tube fermentation method or the membrane filter test (MF) as described in
the 16th Edition of Standard Methodj.
Sampling Frequency
Minimum sampling frequencies are as follows:
Population Served Coliform Samples/Week
<500 1
501-3,300 2
3,301-10,000 3
10,001-25,000 4
>25,000 5
Grab samples must be taken on different days. In addition, one
sample must be taken every day during which the turbidity exceeds 1 NTU,
unless the Primacy Agency determines that the system, for logistical
reasons outside the system's control, cannot have the sample analyzed
within 30 hours of collection. If taken, these samples count towards the
weekly sampling requirement. Also, under the Total Coliform Rule, systems
must take one coliform sample in the distribution system near the first
service connection within 24 hours after a source water turbidity
measurement exceeds 1 NTU. This measurement must be included in the total
colifonn compliance determination. The purpose of these requirements is
to ensure that the monitoring occurs during worst case conditions.
The initial evaluation of the source water quality is based on the
data from the previous 6 months. After the initial evaluation, systems
must continue to conduct sampling each month to demonstrate compliance
with the source water quality criteria on an ongoing basis. If the
criterion has not been met, the system must filter.
Use of Historical Data Base
3-3
-------
Some systems may already monitor their source water for total and/or
fecal coliform concentration. The resulting historical data base may be
sufficient for the Primacy Agency to make the initial determination of
whether the system meets the source water quality criteria. The
historical data base is considered sufficient for making this determina-
tion if:
The raw water sampling location is upstream of the point of
disinfectant application as previously defined.
The monthly samples represent at least the minimum sampling
frequency previously mentioned.
The sampling period covers at least the previous six months.
3.1.2 Turbidity Levels; To avoid filtration, the turbidity of the
water prior to disinfection cannot exceed 5 NTU, on an ongoing basis,
based on grab samples collected every four hours (or more frequently) that
the system is in operation. A system may substitute continuous turbidity
monitoring for grab sample monitoring if it validates such measurements
for accuracy with grab sample measurements on a regular basis, as
specified by the Primacy Agency.1 If a public water system uses contenti-
ons monitoring, it must use turbidity values recorded every four hours (or
some shorter regular time interval) to determine whether it meets trie
turbidity limit for raw water. A system occasionally may exceed the 5 NTU
limit and still avoid filtration as long as (a) the Primacy Agency
determines that each event occurred because of unusual or unpredictable
circumstances -and (b) as a result of this event, there have not been more
than two such events in the past twelve months the system served water to
the public or more than five such events in the past 120 months the system
Validation should be performed at least twice a week based on the
procedure outlined in Part 214A in the 16th Edition of Standard
Methods. Although the 17th Edition is-available, the 16th Edition is
that which is referred to in the rule. Improper installation of
continuous monitors may allow for air bubbles to enter the monitor
resulting in false turbidity spikes. To avoid air bubbles reaching the
turbidimeter, the sample tap should be installed below the center line
of the pipe and an air release valve may be included on the sample
1 ine.
3-4
-------
served water to the public. An "event" is defined as a series of
consecutive days in which at least one turbidity measurement each day
exceeds 5 NTU.
It is important to note that every event, i.e., exceedance of the 5
NTU limit, regardless of whether the system must filter as a consequence,
constitutes a violation of a treatment technique requirement. For
example, if the turbidity exceeded 5 NTU in at least one measurement each
day for three consecutive days, this would constitute one event and one
treatment technique violation. If this was the third event in the past 12
months the system served water to the public, or the sixth event in the
past 120 months the system had served water to the public, the system
would also be required to install filtration. In all cases, the system
must inform the Primacy Agency when the turbidity exceeds 5 NTU as soon as
possible, but no later than the end of the next business day.
The Primacy Agency should evaluate additional data from the utility
to determine the significance of the event with respect to the potential
health risk to the community and determine whether a boil water notice is
necessary. The additional data may include raw water fecal coliform
levels, duration and magnitude of the turbidity excursion, nature of the
turbidity (organic or inorganic), disinfectant residual entering tne
system during the excursion and/or coliform levels in the distribution
system following the excursion. Boil water notices are not required uncer
the SUTR, they may be issued at the discretion of the Primacy Agency.
In order to determine if the periods with turbidity greater than
5 NTU are unusual or unpredictable, it is recommended that in addition to
the historical turbidity data, the water purveyor should collect and
provide to the Primacy Agency current and historical information on flows,
reservoir water levels, climatological conditions, and any other informa-
tion that the Primacy Agency deems relevant. The Primacy Agency will then
evaluate this information to determine if the event was unusual or
unpredictable. Examples of unusual or unpredictable events include
hurricanes, floods and earthquakes. High turbidity events may be avoided
by:
Use of an alternate source which is not a surface water and
does not have to meet the requirements of the SWTR.
3-5
-------
Use of an alternate source which is not a surface water and
does not have to meet the requirements of the SWTR.
Use of an alternate source which is a surface water and which
does meet the requirements of the SWTR.
Utilization of stored water to supply the community until the
source water quality meets the criteria.
3.2 Disinfection Criteria
3.2.1 Inactivation Requirements
To avoid filtration, a system must demonstrate that it maintains
disinfection conditions which inactivate 99.9 percent of Giardia cysts and
99.99 percent of viruses every day of operation except any one day each
month. If the disinfection conditions provide less than these inactiva-
tions during more than one day of the month, the system is in violation of
a treatment technique requirement. If the system incurs such a violation
during any two months in the previous 12 months, the system must install
filtration, unless one of the violations was caused by unusual and
unpredictable circumstances as determined by the Primacy Agency. Systems
with three or more violations in the previous 12 months must install
filtration regardless of the cause of the violation. To demonstrate
adequate inactivations, the system must monitor and record the disinfec-
tant^) used, disinfectant residual(s), disinfectant contact time(s), pH
(for chlorine), and water temperature, and use these data to determine if
it is meeting the minimum total inactivation requirements in the rule.
A number of disinfectants are available, including ozone, chlorine,
chlorine dioxide and chloramines. The SWTR prescribes CT [C, residual
disinfectant concentration (mg/L) x T, contact time (min)] levels for
these disinfectants which will achieve different levels of inactivation
under various conditions. The disinfectant(s) used to meet the inactiva-
tion requirements is identified as the primary disinfectant throughout the
remainder of this document.
To determine compliance with the inactivation requirements, a system
must calculate the CT value(s) for its disinfection conditions during peak
hourly flow once each day that it is delivering water to its customers.
For the purpose of calculating CT value, T is the time (in minutes) it
3-6
-------
takes the water, during peak hourly flow, to move between the point of
disinfectant application and a point where, C, residual disinfectant
concentration is measured prior to the first customer. Residual
disinfectant concentration is the concentration of the disinfectant (in
mg/L) at a point before or at the first customer. Contact time in
pipelines must be calculated based on plug flow (i.e., where all water
moves homogeneously in time between two points) by dividing the internal
volume of the pipeline by the peak hourly flow rate through that pipeline.
Contact time within mixing basins, settling basins storage reservoirs, and
any other tankage must be determined by tracer studies or an equivalent
method as determined by the Primacy Agency. The contact time determined
from tracer studies to be used for calculating CT is T10. T10 is the
detention time corresponding to the time for which 90 percent of the water
has been in contact with at least the residual concentration, C. Guidance
for determining contact times for basins is provided in Appendix C.
The first customer is the point at which finished water is first
consumed. In many cases this will include the treatment plant itself.
This definition of first customer pertaining to the point of first
consumption assures that the water has received the required disinfection
to provide protection from microorganisms for all consumers. Peak hourly
flow should be considered as the greatest volume of water passing through
the system during any one hour in a consecutive 24 hour period. Thus, it
is not meant to be the absolute peak flow occurring at any instant during
the day.
Systems with only one point of disinfectant application ^ay
determine the total inactivation based on one point of residual measure-
ment prior to the first customer, or on a profile of the residual
concentration after the point of disinfectant application. Methods of
disinfection measurement are presented in Appendix D. The residual
profile and the total inactivation is calculated as follows:
Measure the disinfectant residual, C, at any number of points
within the treatment train.
Determine the travel time, T, between the point of disinfec-
tant application and the point where C is measured for the
first section. For subsequent measurements of "C," T is the
3-7
-------
time it takes for water to move from the previous "C" measure-
ment point to this point of measurement.
Calculate CT for each point of residual measurement (CT. ).
Determine the inactivation ratio (CT.JiC/CT39 3) for each sec-
tion.-
Sum the inactivation ratios for each section, i.e. C,T,/CT,9 9
+ C,T,/CT99 9 + CnTn/CT99 9 to determine the total inac'tivation
ratio".
If the total inactivation ratio (sum (CTcm/CT99 9)) is equal to or greater
than 1.0, the system provides greater than 99.9 percent inactivation of
Giardia cysts), and the system meets the disinfection performance re-
quirement. Further explanation of CT calculations is presented in Section
3.2.2.
Systems need only calculate one CT (CTCJ|£) each day, for a point at
or prior to the first customer; alternatively they have the option of cal-
culating numerous CTs after the point of disinfectant application but
prior to the first customer to determine the inactivation ratio. Profil-
ing the residual gives credit for the higher residuals which exist after
the disinfectant is applied but before the first customer. Profiling the
residual may not be necessary if one CT is calculated (CTcm), and this
exceeds the applicable CT99 s. In this case, the system is meeting the
disinfection performance requirement. For systems with a very low oxidant
demand in the water and long contact times, this approach may be the most
practical to usjs.
For systems with multiple points of disinfectant application, such
as ozone followed by chlorine, or chlorine applied at two different points
in the treatment train, the inactivation ratio of each disinfectant
section prior to the first customer is used to determine the total
inactivation ratio. The disinfectant residual of each disinfection
CT99 9 is the CT value required to achieve 99.9 percent or 3-log Giardia
cyst inactivation for the conditions of pH, temperature and residual
concentration for each section. A section is the portion of the system
with a measurable contact time between two points of disinfection
application or residual monitoring.
3-8
-------
section and the corresponding contact time must be measured at some point
prior to the subsequent disinfection application point(s) to determine the
inactivation ratio for each section, and whether the total inactivation
ratio is 1.0 or more. For example, if the first disinfection section
provided an inactivation ratio of 2/3 (or 99 percent inactivation) and the
second disinfection section provided an inactivation ratio of 1/3 (or 90
percent inactivation), the total inactivation ratio would equal 1.0 (2/3
+ 1/3 = 1) indicating that 99.9% inactivation was provided and the
disinfection requirements are met. Further explanation of the determina-
tion of total inactivation provided is contained in Section 3.2.2.
Maintaining Inactivation Level
The SWTR establishes CTs for chlorine, chlorine dioxide, ozone and
chloramines which will achieve 3-log inactivations of Giardia cysts and at
least 4-log inactivation of viruses. Appendix E presents CTs for these
and other log inactivations. A system must demonstrate compliance with
the inactivation requirements based on conditions occurring during peak
hourly flow. Since a system generally can only identify peak hourly flow
after it has occurred, hourly residual measurements during the day are
suggested. If the sampling points are remote, or manpower is limited and
collection of hourly grab samples is impractical, continuous monitors may
be installed. In cases where continuous monitors are impractical, the
Primacy Agency may establish an acceptable monitoring program on a
case-by-case basis; where possible this should be based on historical flow
patterns. Measurements for the hour of peak flow can then be used in
calculating CT. The pH (for systems using chlorine) and temperature must
be determined daily for each disinfection sequence prior to the first
customer.
Since the system's inactivation is determined during peak hourly
flow, the disinfectant dosage applied to meet CT requirements may not be
necessary during lower flow conditions. Continuing to apply a disinfec-
tant dosage based on the peak hourly flow could possibly result in
increased levels of disinfectant by-products, including TTHMs and
increased costs. Under lower flow conditions, a higher contact time is
available and a lower residual may provide the CT needed to meet the
inactivation requirements. The system may therefore choose to adjust the
3-9
-------
disinfectant dose with changes in flow. The system should, however,
maintain a disinfectant residual which will still provide a 3-log
inactivation of Giardia cysts and a 4-log inactivation of viruses at
non-peak hourly flows. The system should therefore evaluate the residual
needed to provide the required inactivation under different flow
conditions and set the dosage accordingly. The following provides an
example of maintaining the required inactivation.
Example
A 5 mgd non-filtering system disinfecting with free chlorine at one
point of application, has a contact time of 165 minutes during a peak flow
of 5 MGD. The flow varies from 1 to 5 MGD. The pH and temperatures of
the water are 7 and 5 C, respectively. At a residual of 0.9 mg/L, a CT of
148 mg/L-min is required to meet the disinfection requirements. The CT
for 0.9 mg/L residual is determined by straight line interpolation between
0.8 mg/L and 1.0 mg/L residuals. Under lower flow conditions, the
available contact time is longer and a lower residual would provide the
required disinfection. Based on existing contact time and using the
appropriate CT tables (in this case, Table E-2) in Appendix E for a 3-log
Giardia cyst inactivation, the required disinfection would be provided by
maintaining the following chlorine residuals for the indicated flow:
Contact CT (mg/L-min) Free Chlorine
Flow fHGD^ Time (min^ Required Residual (mq/L^
5 165 148 0.9
4 206 145 0.7
3 275 143 0.5
2 - 412 139 0.4
1 825 139 0.2
This table indicates the variation of residuals needed for the
system to provide the required inactivation. For chlorine, the disinfec-
tant residual cannot be adjusted in direct proportion to the flow because
the CT needed for disinfection is dependent upon the residual. Since it
is not practical to continuously adjust the residual and, since a
disinfection level for a 3-log Giardia cyst inactivation must be
maintained under all flow conditions, it is suggested that the flow
variation at the utility be divided into ranges and the residual needed at
3-10
-------
the higher flow rate of each range be maintained for all flows within the
range to ensure the required disinfection. The following flow ranges and
residuals are suggested for the system:
Free Chlorine
Flow Range (MGD) Residual fmg/L)
1 - 1.9 0.4
2 - 3.9 0.6
4-5 0.9
By maintaining these residuals, the utility is ensuring the provision of
the required disinfection while minimizing the disinfectant application,
which should result in lower disinfection by-products and costs.
Although these residuals will meet the inactivation requirements,
maintaining a residual in the distribution system must also be considered.
If no other point of disinfection exists prior to the distribution system,
the residual for disinfection must be maintained at a level which will
also provide a residual throughout the distribution system. The complete
range of flows occurring at the plant should be evaluated for determining
the required residual. A utility may establish the residual requirements
for as many flow ranges as is practical.
The CTs determined from the daily system data should be compared to
the values in the table for the pH and temperature of the water, to
determine if the required CT has been achieved. Only the analytical
methods prescribed in the SWTR, or otherwise approved by EPA, may be usea
for measuring disinfectant residuals. Methods prescribed in the SWTR are
listed in Append4x D. The Appendix also contains a paper which describes
monitoring methods for various disinfectants and conditions.
The Primacy Agency should make periodic checks on its utilities to
assure that they are maintaining adequate disinfection at non-peak flow
conditions.
Meeting the Inactivation Requirement Using Free Chlorine
When free chlorine is used as a disinfectant, the efficiency of
inactivation is influenced by the temperature and pH of the water. Thus,
the measurement of the temperature and pH for the determination of the CT
is required. The SWTR provides the CT requirements for free chlorine at
3-11
-------
various temperatures and pHs which may occur in a source water. These
values are presented in Table E-l through Table 1-7 in Appendix E. The
basis for these values is discussed in Appendix F. For free chlorine, a
3-log inactivation of Giardia cysts will provide greater than a 4-log
inactivation of viruses, thus meeting the SWTR inactivation requirements.
As indicated in Table E-2, a raw water temperature of 5 C, a pH of
7.0, and a residual chlorine concentration of 1.4 mg/L require a CT of 155
mg/L-min to provide a 3-log inactivation of Giardia cysts. Therefore, to
meet the inactivation requirement under these conditions with one point of
residual measurement, a contact time of 111 minutes [(155 mg/L-min)/ (1.4
mg/L)] prior to the first customer would be required.
Meeting the Inactivation Requirement Using Chloramines
Chloramines are a much weaker oxidant than free chlorine, chlorine
dioxide and ozone. The CT values for Chloramines presented in Table E-12
are based on disinfection studies using preformed Chloramines and in vitro
excystation of Giardia muris cysts (Rubin, 1988). No safety factor was
applied to the laboratory data on which the CT values were based since EPA
believes that chloramination, conducted in the field, is more effective
than using preformed Chloramines.
In the laboratory testing using preformed Chloramines, ammonia and
chlorine were reacted to form Chloramines before the addition of the
microorganisms. Under field conditions, chlorine is usually added first
followed by ammonia addition further downstream. Also, even after the
addition of ammonia, some free chlorine residual may persist for a period
of time. Therefore, free chlorine is present for a period of time prior
to the formation of Chloramines. Since this free chlorine contact time is
not duplicated in the laboratory when testing with preformed Chloramines,
the CT values obtained by such tests may provide conservative values when
compared to those CTs actually obtained in the field with chlorine applied
before ammonia. Also, other factors such as mixing in the field (versus
no mixing in the laboratory) may contribute to disinfection effectiveness.
For these reasons, systems using Chloramines for disinfection may
demonstrate effective disinfection in accordance with the procedure in
Appendix G in lieu of meeting the CT values in Appendix E.
3-12
-------
If a system uses chloramines and is able to achieve the CT values
for 99.9 percent inactivation of Giardia cysts, it is not always
appropriate to assume that 99.99 percent or greater inactivation of
viruses was also achieved. New data indicate that Hepatitis A virus is
more sensitive than Giardi? cysts to inactivation by preformed chloramines
(Sobsey, 1988). The CT values required to achieve 99.99 percent
inactivation of Hepatitis A with preformed chloramines are lower than
those needed to achieve 99.9 percent inactivation of Giardia cysts. These
data contrast with other data which indicate that rotavirus is more
resistant than Giardia cysts to preformed chloramines (Hoff, 1986).3
However, rotavirus is very sensitive to inactivation by free chlorine,
much more so than Hepatitis A (Hoff, 1986;4 Sobsey, 1988). If chlorine
is applied prior to ammonia, the short term presence of free chlorine
would be expected to provide at least 99.99 percent inactivation of
rotavirus prior to the addition of ammonia and subsequent formation of
chloramines. Thus, EPA believes it is appropriate to use Hepatitis A
data, in lieu of rotavirus data, as a surrogate for defining minimum CT
values for inactivation of viruses by chloramines, under the condition
that chlorine is added to the water prior to the addition of ammonia.
A system which achieves a 99.9 percent or greater inactivation of
Giardia cysts with chloramines can be considered to achieve at least 99.99
percent inactivation of viruses, provided that chlorine is added to the
water prior to the addition of ammonia. Table E-13 provides CT values for
achieving different levels of virus inactivation. However, if ammonia is
added first, the CT values in the SWTR for achieving 99.9 percent
inactivation of Giardia cysts cannot be considered adequate for achieving
99.99 percent inactivation of viruses.
Under such cases of chloramine production, the SWTR requires systems
to demonstrate through on-site challenge studies, that the system is
CT values in excess of 5,000 are required for a 4-log inactivation of
rotavirus by preformed chloramines but no minimum CT values have been
determined.
CT values ranging from 0.025 to 2.2 achieve 99 percent inactivation of
rotavirus by free chlorine at pH = 6 -10 and 4 - 5°C (Hoff, 1986).
3-13
-------
achieving at least a 4-log inactivation of viruses. Guidance for
conducting such studies is given in Appendix G. Once conditions for
achieving a 4-log inactivation of viruses has been established, the
Primacy Agency should require systems to report their disinfection
operating conditions on an ongoing basis. These conditions should verify
that the system is operating at CT values in excess of that needed to
achieve a 4-log virus inactivation or 3-log Giardia cyst inactivation,
whichever is higher.
Meeting the Inactivation Requirement Using Chlorine Dioxide
Under the SWTR, the CT values for the inactivation of Giardia cysts
using chlorine dioxide are independent of pH. Under the SWTR the only
parameter affecting the CT requirements associated with the use of
chlorine dioxide is temperature. Table E-S in Appendix E presents the
chlorine dioxide CT values required for the inactivation of Giardia cysts
at different temperatures. The basis for these CT values is discussed in
Appendix F. Systems which use chlorine dioxide are not required to
measure the pH of the disinfected water for the calculation of CT. For
chlorine dioxide, a 3-log inactivation of Giardia cysts'will generally
result in greater than a 4-log virus inactivation, and assure meeting the
SWTR inactivation requirements. However, for chlorine dioxide, unlike
chlorine where this relationship always holds true, at certain tempera-
tures, the 4-log virus CTs may be higher than the 3-log Giardia cyst CTs.
The Primacy Agency may allow lower CT values than those specified in
the SWTR for individual systems based on information provided by the
system. Protocols for demonstrating effective disinfection at lower CT
values is provided in Appendix G.
As indicated in Tables E-8 and E-9, the CT requirements for chlorine
dioxide are substantially lower than those required for free chlorine.
However, chlorine dioxide is not as stable as free chlorine or chloramines
in a water system and may not be capable of providing the required
disinfectant residual throughout the distribution system. In addition,
out of concern for toxicological effects, EPA's current guideline is that
the sum of the chlorine dioxide, chlorate and chlorite residuals, be less
than 1.0 mg/L at all consumer taps. This guideline may be lowered as more
health effects data become available. These concerns further reduce the
3-14
-------
feasibility of using chlorine dioxide as a secondary disinfectant for
distribution systems. Therefore, the use of chlorine dioxide as a primary
disinfectant may result in the need for the application of a secondary
disinfectant, such as chlorine or chloramines, that will persist in the
distribution system and provide the required residual protection.
Meeting the Inactivation Requirement Using Ozone
Another disinfectant to inactivate Giardia cysts and viruses is
ozone. As with chlorine dioxide, under the SWTR, the CT values for ozone
are independent of pH. Tables E-10 and E-ll present the CT requirements
for ozone at different source water temperatures. The basis for the CT
values for ozone is given in Appendix F. As for free chlorine, a 3-log
Giardia cyst inactivation with ozone will result in greater than a 4-log
virus inactivation. Unlike chlorine, for cases where only a 1-log or
lower Giardia inactivation is needed with ozone, the CT values for virus
inactivation may be higher than the CT for Giardia. The Primacy Agency
may allow lower CT values for individual systems based on information
provided by the system that demonstrates that CT values lower than those
specified in the rule achieve the same inactivation efficiencies (see
Appendix G).
Ozone is extremely reactive and dissipates quickly after applica-
tion. Therefore, a residual5 can only be expected to persist a short time
The residual must be measured using the Indigo Trisulfonate Method
(Bader & Soigne, 1981) or automated methods which are calibrated in
reference to the results obtained by the Indigo Trisulfonate method, on
a regular basis as determined by the Primacy Agency. The Indigo
Trisulfonate method is included in the 17th Edition of Standard
Methods. This method is preferable to current standard methods because
of the selectivity of the Indigo Trisulfonate indicaor in the presence
of most interferences found in ozonated waters. The ozone degrades an
acidic solution of indigo trisulfonate in a 1:1 proportion. The
decrease in absorbance is linear with increasing ozone concentrations
over a wide range. Malonic acid can be added to block interference
from chlorine. Interference from permanganate, produced by the
ozonation of manganese, is corrected by running a blank in which ozone
is destroyed prior to addition of the indigo reagent. The samples can
be analyzed using a spectrophotometer at a 600 nm wavelength which can
detect residuals as low as 2 ug/L or a visual color comparison method
which can measure down to 10 ug/L ozone. Although currently availacie
monitoring probes do not use the Indigo Trisulfonate Method, they can
3-15
-------
after application. In addition, the application of ozone to water is
dependent on mass transfer. For these reasons, the method of CT
determination used for the other disinfectants is impractical for ozone.
The CT:alc must be determined for the ozone contactor alone. The contactor
will have some portions where the ozone is applied and other portions of
the contactor where ozone is no longer applied, which are referred to as
the reactive flow chambers.
For many ozone contactors, the residual in the contactor will vary
in accordance with the method and rate of application, the residual will
be nonuniform and is likely to be zero in a portion of the contactor. As
previously indicated, the CT value is based on the presence of a known
residual during a specific contact time. Thus disinfection credit is only
provided for the time when a residual is present. Besides the nonuniform-
ity of the residual, monitoring the residual will be difficult because of
the ozone's high reactivity and the closed design of the contactors.
In addition to the difficulty in determining the ozone residual for
the CT calculation, the contact time will vary between basins depending
on their flow configuration. Several types of devices are available for
adding ozone to water including porous diffusers, submerged turbines,
injector, packed towers and static mixers. Each type of device can be
used in either single or multiple chamber contactors. The flow through a
single chamber turbine contactor will approximate a completely mixed unit,
while flow through a single chamber diffused contactor, or a multiple
chamber diffused contactor, will more closely represent plug flow. This
variation in flow in different contactors makes the use of Tlo inappropri-
ate for some contactors.
The differences between ozone contactors and other disinfection
systems resulted in the development of several approaches for determining
the inactivation provided by ozone, including:
Evaluation of C and T
Segregated Flow Analysis (SFA)
Continuously Stirred Tank Reactor (CSTR)
Site Specific Evaluation
be calibrated via this method.
3-16
-------
The method which is appropriate for a particular system will depend on
system configuration and the required level of inactivation. Another
significant difference is that ozone may be applied to provide only a
portion of the overall 3-log Giardia cyst and 4-log virus inactivation
with the remainder of the inactivation provided by another disinfectant.
Appendix 0 provides details for selecting the appropriate method of
evaluation for specific conditions.
The evaluation of C and T involves separate determination of the
ozone residual concentration, C, and the contact time, T, in the
contactor. C can be determined for individual chambers of a contactor
based on the residual measured at several points throughout the chamber,
or at the exit of the chamber. The T value can be determined through a
tracer study or an equivalent method as approved by the Primacy Agency
with air or oxygen applied during testing, using the same feed gas rate as
used during operation. Appendix 0 provides details for the CT approach.
SPA is based on the results of a tracer study used in conjunction
with the measured ozone residual to determine the survival of microorgan-
isms exiting the contactor. The survival corresponds to a certain
inactivation. Guidelines for this approach are included in Appendix 0.
The CSTR approach is applicable for contactors which have a high
degree of mixing. Experience has shown that for contactors such as
turbine units, the ozone residual is generally uniform throughout the
contactor. The ozone residual measured at the exit if the contactor is
used in an equation for CSTRs to determine the inactivation provided.
Appendix 0 provides details for conducting CSTR analysis.
Site specific evaluations may include:
Measurement of an observable parameter to correlate with C
Mathematical model for disinfection efficiency
Microbial indicator studies for disinfection efficiency
to more closely determine the inactivation provided in a particular
system. Appendix 0 provides details for applying site specific evalua-
tions.
3-17
-------
Summary
Many systems which do not provide filtration will have difficulty in
providing the contact time necessary to satisfy the inactivation
requirements prior to the first customer. For example, a system using
free chlorine at a water temperature of 5 C, a pH of 7.0 and a chlorine
residual of 1.4 mg/L would require 111 minutes of contact time to meet the
inactivation requirement. Potential options for these systems include:
Installation of storage facilities to provide the required
contact time under maximum flow conditions.
Use of an alternate primary disinfectant such as ozone or
chlorine dioxide which has CT values lower than those required
for free chlorine for the required inactivation.
For some systems, the difficulty in obtaining the required
inactivation may only be a seasonal problem. A system that has raw water
temperatures which reach 20 C during the summer months at a pH of 7.0, may
have sufficient contact time to meet the CT of 56 mg/L-min (Table E-5) at
a chlorine concentration of 1 mg/L. However, assuming the same pH and
chlorine concentration, it may not have sufficient contact time to meet
the CT requirement at 5 C, 149 mg/L-min (Table E-2), or at 0.5 C,
210 mg/L-min (Table E-l). Under those conditions, a system could choose
to use ozone or chlorine dioxide on a seasonal basis, since they are
stronger disinfectants requiring a shorter contact time.
As indicated in Table E-12, the CT values for chloramines may be
impractical to attain for most systems. Systems which currently utilize
chloramines as a primary disinfectant may need to use either free chlor-
ine, chlorine dioxide or ozone in order to provide the required disin-
fection. However, systems using chloramines as a primary disinfectant may
chose to demonstrate the adequacy of the disinfection. Appendix G
presents a method for making this demonstration.
Meeting the Inactivation Requirement Using Alternate Disinfectants
For systems using disinfectants other than chlorine, chloramines,
chlorine dioxide, or ozone, the.effectiveness of the disinfectant can be
demonstrated using the protocol contained in Appendix G. The protocol in
Appendix G.3 for batch testing should be followed for any disinfectant
3-18
-------
which can be prepared in an aqueous solution and will be stable throughout
the testing. For disinfectants which are not stable, the pilot study
protocol outlined in Appendix G.4 should be followed.
3.2.2 Determination of Overall Inactivation for Residual Profile,
Multiple Disinfectants and Multiple Sources
For systems which apply disinfectant(s) at more than one point, or
choose to profile the residual from one point of application, the total
inactivation is the sum of the inactivation ratios between each of the
points of disinfection or between each of the residual monitoring points,
respectively. The portion of the system with a measurable contact time
between two points of disinfection application or residual monitoring will
be referred to as a section. The calculated CT (CTcalc) for each section
is determined daily.
The CT needed to fulfill the disinfection requirements is CT99 9,
corresponding to a 3-log inactivation of Giardia cysts and greater than or
equal to a 4-log inactivation of viruses (except for chloramines and
sometimes chlorine dioxide as explained in Section 3.2.1). The inactiva-
tion ratio for each section is represented by CTCIIC/CT99 9, as explained in
Section 3.2.1, and indicates the portion of the required inactivation
provided by the section. The sum of the inactivation ratios from each
section can be used to determine the overall level of disinfection
provided. Assuming inactivation is a first order reaction, the inac-
tivation ratio corresponds to log and percent inactivations as follows:
C_L3i./£!99 9 - Log Inactivation Percent Inactivation
0.17 = 0.5 log = 68%
0.33 = 1 log = 90%
0.50 » 1.5 log = 96.8%
0.67 = 2 log 99%
0.83 * 2.5 log = 99.7%
1.00 = 3 log = 99.9%
1.33 * 4 log 99.99%
3-19
-------
CT99 9 can be determined for each section by referring to Tables E-l
through E-13 in Appendix E, using the pH (when chlorine is the disinfec-
tant) and temperatures of the water for the respective sections. These
tables present the log inactivation of Giardia cysts and viruses achieved
by CTs at various water temperatures and pHs.
Log inactivations are additive, so:
0.5 Log + 1.0 Log » 1.5 Log or
0.17CT99 , + 0.33CT999 = 0.5CT99 9
If the sum of the inactivation ratios is greater than or equal to
one, the required 3-log inactivation of Giardia cysts has been achieved.
An inactivation ratio of at least 1.0 is needed to demonstrate compliance
with the Giardia cyst inactivation requirements for unfiltered systems.
The total log inactivation can be determined by multiplying the sum
of the inactivation ratios (sum (CTCilc/CT99 9)), by three. The total log
inactivation can be determined in this way because CT99 9 is equivalent to
a 3-log inactivation. The total percent inactivation can be determined as
follows:
y * Iflfi - Iflfl Equation (1)
101
where: y = % inactivation
x = log inactivation
For example:
x = 3.0 log inactivation
y * 100 - 100 * 99.9 % inactivation
10""
As explained in Section 3.2.1, the CTcllc determined for each disin-
fection section is the product of the disinfectant residual in mg/L and
the detention time in minutes through the section at peak hourly flow.
However, for many water systems, peak hourly flow will not necessarily
occur simultaneously in all sections. The extent to which the occurrence
of peak hourly flow will vary between sections of the system depends on
3-20
-------
the characteristics of an individual system including its size, storage
capacity within the distribution system, the number of sources, and
hydraulic capacities between different sections. In order to simplify the
determination of peak hourly flow for the system, it should be taken as
peak hourly flow in the last section of the system prior to the first
customer.
The CT values for all the sections should be calculated for the flow
and the residuals occurring during the hour of peak flow in the last
section. The most accurate way to determine the flow in a particular
section is through the use of a flow meter. However, some sections of the
system may not have a flow meter. The following guidelines can be used to
determine the flow to be used in calculating CT:
For sections which do not have meters, the flow should be
assumed to be the higher of the two flows occurring in the
closest upstream and downstream sections with meters.
In cases where a section contains a pipeline and a basin with
the flow meter located prior to the basin, the metered flow
does not represent the discharge rate of the basin. The
difference in inlet and discharge rates from a basin will
impact the water level in the basin. As explained in Appendix
C, falling water levels will result in lower T10 values.
To assure that the detention time of a basin is not
overestimated, the discharge flow from a basin should be
used in lieu of the influent flow, unless the influent
flow is higher.
To estimate the discharge flow from a basin the closest
flow meter downstream of the basin should be used.
The following example presents the determination of the total
percent inactivation for multiple points of disinfection, with variation
in flow between sections.
Example
A community of 6,000 people obtains its water supply from a lake
which is 10 miles from the city limits. Two 0.2 MG storage tanks are
located along the 12-inch transmission line to the city. The water is
disinfected with chlorine dioxide at the exit from the lake and with
chlorine at the discharge from the first and second storage tanks. The
3-21
-------
average water demand of the community is 1 MGD with a peak hourly demand
of approximately 2 MGD. For the calculations of the overall percent
inactivation, the supply system is divided into three sections as shown on
Figure 3-1.
Section 1 - from the lake to the discharge from the first storage
tank,
Section 2 - from the discharge from the first storage tank to the
discharge from the second tank
Section 3 - from the discharge of the second storage tank to the
first customer
The overall inactivation is computed daily for the peak hourly flow condi-
tions. Sections 1 and 3 contain flow meters to monitor the water being
withdrawn from the lake and the water being delivered to the distribution
system as shown on Figure 3-1. On the day of this example calculation,
the peak hourly flow in section 3 was 2 MGD. During this hour, water was
being withdrawn from the lake at a rate of 1.5 mgd. Considering the
placement of flow meters, the flow of 2 mgd measured in section 3 should
be used for calculating CT for that section. Since section 2 does not
have a flow meter, the meter in section 3 serves as a measure of the
discharge from storage tank 2 and should be the flow used in the
calculation of CT for section 2. The flow meter in section 1 records the
flow through the transmission main which should be used in the calculation
of CT for the pipeline. However, this meter does not represent the
discharge from-storage tank 1. Since the water is being pumped to the
distribution system at a higher rate than the flow entering storage tank
1, the flow of 2 mgd measured in section 3 should be used for calculating
the CT for storage tank 1.
The pH, temperature and disinfectant residual of the water were
measured at the end of each section just prior to the next point of
disinfection and the first customer during the hour of peak flow. The
water travels through the 12-inch transmission main at 177 ft/min at
3-22
-------
STORAGE
TANK 1
STORAGE
TANK 2
111 CUSTO
FM
T
CHLORINE
DIOXIDE
CHLORINE
CHLORINE
SECTION
SECTION
SECTION I
1
FIGURE 3-1-DETERMINATION OF INACTIVATION FOR
MULTIPLE DISINFECTANT APPLICATION
TO A SURFACE WATER SOURCE
-------
.0
Section 1
15,840
1.5
2.0
89
116
205
chlorine
dioxide
0.1
5
8
Section 2
26,400
2.0
2.0
111
114
225
chlorine
0.2
5
8
Section 3.
10,560
2.0
45
0
45
chlorine
0.4
5
8
1.5 MGO. The detention times of the storage tanks were read off the T
vs. Q plots generated from tracer studies conducted on the storage tanks
(see Appendix C). The data for the inactivation calculation are as
follows:
length of pipe (ft)
flow (mgd)
pipe
tank
contact time (min)
pipe
tank
total
disinfectant
residual (mg/L)
temperature (C)
pH
This information is then used in conjunction with the CT99 9 values in
Appendix E to determine the (CTcalc/CT99 9) in each section as follows:
Section 1 - Chlorine dioxide
CT:a!. - 0.1 mg/L x 105 minutes = 20.5 mg/L-min
From Table E-8 at a temperature of 5 C and pH * 8,
CT59 3 is 26 mg/L-min
CT:a =/CT99 9 = 20.5 mq/L-min = 0.79
26 mg/L-min
Section 2 - Chlorine
CT,il: = 0.2 mg/L x 225 minutes * 45 mg/L-min
From Table E-2 at a temperature of 5 C and pH * 8,
CT99 9 is 198 mg/L-min
CT.J)C/CT99 9 * 45 mg/L-min » 0.23
198 mg/L-min
6 Q = 1.5 X 106 gal/dav X 1 ft3 X dav » 177 ft/rain
A (1 ft' /4) 7.48 gal 1440 min
3-23
-------
;
Section 3 - Chlorine
CT:j,: = 0-4 mg/L-min x 45 min = 18 mg/L-min
From Table E-2 at a temperature of 5 C and pH * 8,
CT99 9 is 198 mg/L-min
CT-m/CTs9 9 = 18 mg/L-min = 0.09
198 mg/L-min
The sum of CTCJlc/CT9g 9 is equal to 1.11, which is greater than 1,
therefore, the system meets the requirements of providing a 3-log
inactivation of Giardia cysts. The log inactivation provided is:
x = 3 x CJ_CJIC * 3 x 1.11 « 3.33
The percent inactivation can be determined using equation 1.
y = 100 - IflQ = 100 - IfiO = 100 - 0.05 = 99.95% inactivation
IF33 2,138
The system meets the requirement of providing a 99.9 percent inactivation
of Giardia cysts.
The SWTR also requires that the public be provided with protection
from Legionella as well as Giardia cysts and viruses. Inactivation levels
have not been set for Legionella because the required inactivation of
Giardia cysts will provide protection from Legionella.7 However, this
level of disinfection cannot assure that all Legionella will be inacti-
vated and that no recontamination or regrowth in recirculating hot water
systems of buildings or cooling systems will occur. Appendix B provides
Kuchta et al. (1983) reported a maximum CT requirement of 22.5 for a
99 percent inactivation of Legionella in a 21 C tap water at a pH of
7.6-8.0 when using free chlorine. Using first order kinetics, a 99.9
percent inactivation requires a CT of 33.8. Table A-5 presents the CTs
needed for free chlorine to achieve a 99.9 percent inactivation of
Giardia cysts at 20 C. This table indicates that the CT required for
a 3-log inactivation of Giardia at the temperature and pH of the
Legionella test ranges from 67 to 108 depending on chlorine residual.
These CT's are two to three times higher than that which is needed to
achieve a 3 log inactivation of Legionella.
3-24
-------
guidance for monitoring and treatment to control Legione1 la in institu-
tional systems.
The above discussion pertains to a system with one source with
sequential disinfection. Another system may blend more than one source,
and disinfect one or more of the sources independently prior to blending.
System conditions which may exist include:
All the sources are combined at one point prior to supplying
the community but one or more of the sources are disinfected
prior to being combined, as shown on Figure 3-2.
Each source is disinfected individually and enters the
distribution system at a different point, as shown on Fig-
ure 3-3.
For all systems combining sources, the first step in determining the
CT should be to determine the CT.m provided from the point of blending
closest to the first customer using the contact time and residual at peak
hourly flow for that portion of the distribution system. This corresponds
to section D on Figure 3-2 and section E on Figure 3-3. If the CTca:. for
section D or E provides the required inactivation, no additional CT credit
is needed and no further evaluation is required. However, if the CT for
section D or E is not sufficient to achieve the required inactivation,
then the inactivation ratio (CT;alc)/(CT99 9) should be determined for each
section to determine the overall inactivation provided for each source.
The total inactivation must be greater than or equal to one for all
sources in order to comply with the requirements for 3-log inactivation of
Gi a rdi a cysts. "
On Figure 3-2, sections A, B, C and D contain sampling points a, b,
c and d, respectively. The sum of the inactivation ratios for sections
A+D, B+D and C+D must each be greater than or equal to one for the
disinfection requirements to be met.
The total inactivation for each source on Figure 3-2 should be
determined as follows:
Source I
Determine CTcalc for sections A and D based on the residual
measurements at sample points a and d, and the travel time
3-25
-------
Source II
Source III
through each section under peak hourly flow conditions for the
respective section.
Determine CT99 , for the pH and temperature conditions in each
section using the tables in Appendix E
Calculate the inactivation ratios (CTciic/CT99 9) for sections
A and D.
Calculate the sum of the inactivation ratios for sections A
and D to determine the total inactivation for source I.
If the sum of the inactivation ratios is greater than or equal
to 1.0, the system has provided the required 3-log Giardia
cyst inactivation.
Determine CT.J|C for section B based on the residual measured
at sample point b and the travel time through the section
under peak hourly flow conditions.
Determine CT99 9 for section B for the pH and temperature
conditions in the section using the appropriate tables in
Appendix E.
Calculate the inactivation ratio (CTC|U/CT99 9) for section B.
Add the inactivation ratios for sections B and D to determine
the total inactivation for source II.
If the sum of the inactivation ratios is greater than or equal
to 1.0, the system has provided the required 3-log Giardia
cyst inactivation for the source.
Determine CT,4lc for section C based on the residual measured
at sample point c and the travel time through the section
under peak hourly flow conditions.
Determine CT99 9 for section C for the pH and temperature
conditions in the section using the appropriate tables in
Appendix E.
Calculate the inactivation ratio (CTC||C/CT99 9) for section C.
Add the inactivation ratios for sections C and D to determine
the total inactivation for Source III.
3-26
-------
1st CUSTOUEF
FIGURE 3- 2 INDIVIDUALLY DISINFECTED
SURFACE SOURCES COMBINED
AT A SINGLE POINT
1 at CUSTOMER I
OltlNPffCTANT
APPLICATION
COMBINATION POINT
SAMPLING POINTS
FIGURE 3-3 MULTIPLE COMBINATION PO
FOR INDIVIDUALLY DISINFEC
SURFACE SOURCES
-------
If the sum of the inactivation ratios is greater than or equal
to 1.0, the system has provided the required 3-log Giardia
cyst inactivation for the.source.
The determination of the total inactivation for each source may
require more calculations for systems such as that on Figure 3-3 than on
Figure 3-2. On Figure 3-3 sections A, B, C, D, and E contain sampling
points a, b, c, d, and e respectively. In order to minimize the
calculations needed, the determination of the total inactivation should
begin with the source closest to the first customer.
The total inactivation for each source on Figure 3-3 should be
determined as follows:
Source III
Determine CT for sections C and E based on the residual
measurement at sample points c and e and the detention time in
each section under peak hourly flow conditions for the
respective section.
Determine CT99 , for the pH and temperature conditions in each
section using the tables in Appendix E.
Calculate the inactivation ratios (CT,JIC/CT99 9) for sections
C and E.
Calculate the sum of the inactivation ratios for sections C
and E to determine the total inactivation for source III.
If the sum of the inactivation ratios is greater than or equal
to 1.0, the system has provided the required 3-log Giardia
cyst inactivation for source III.
Source II
Determine CT , for section D based on the residual measured
at sample point d and the detention time through the section
under peak hourly flow conditions.
Determine CT,, , for section D for the pH and temperature
conditions in the section using the appropriate tables in
Appendix E.
Calculate the inactivation ratio (CTelle/CT99 9) for section D.
Add the inactivation ratios for sections D and E to determine
the overall inactivation.
3-27
-------
If the sum of the inactivation ratios is greater than or equal
to 1.0, the system has provided the required 3-log GJardia
cyst inactivation for source II, as well as source I since the
water from each of these sources are combined prior to
sections D and E.
If the total inactivation ratio for sections D and E is less
than 1.0, additional calculations are needed. Proceed as
follows for source II.
Determine CT.3IC for section B based on the residual measured
at sample point b and the detention time through the section
under peak hourly flow conditions.
Determine CT99 9 for section B for the pH and temperature
conditions in the section using the appropriate tables in
Appendix E.
Calculate the inactivation ratio (CTcm/CT99 9) for section B.
Add the inactivation ratios for sections B, D and E to
determine the total inactivation for source II.
If the sum of the inactivation ratios is greater than or equal
to 1.0, the system has provided the required 3-log Giardia
cyst inactivation for the source.
Source I
As noted in the determination of the inactivation provided for
source II, if the sum of the inactivation ratios for sections D and E is
greater than or equal to 1.0, the system has provided the required 3-log
Giardia cyst inactivation. However, if this sum is less than 1.0
additional calculations will be needed to determine the overall inactiva-
tion provided "for source I. The calculations are as follows:
Source I
Determine CT,1IC for section A based on the residual measured
at sample point a and the detention time in the section under
peak hourly flow conditions.
Determine CT99 9 for section A for the pH and temperature
conditions in the section using the appropriate tables in
Appendix E.
Calculate the inactivation ratio (CTcalc/CT99 g) for section A.
Add the inactivation ratios for sections A, D, and E to
determine the total inactivation for source I.
3-28
-------
If the sum of the inactivation ratios is greater than or equal
to 1.0, the system has provided the required 3-1og Giardia
cyst inactivation for the source.
3.2.3 Demonstration of Maintaining a Residual
The SWTR establishes two requirements concerning the maintenance of
a residual. The first requirement is to maintain a minimum residual of
0.2 mg/L entering the distribution system. The second is to maintain a
detectable residual throughout the distribution system. The disinfectant
used to meet these requirements is identified as the secondary disinfec-
tant throughout the remainder of this document. These requirements are
further explained in the following sections.
Maintaining a Residual Entering the Distribution System
To avoid filtration, the disinfectant residual in water entering the
distribution system cannot be less than 0.2 mg/1 for more than four hours,
with one exception noted below. Systems serving more than 3,300 persons
must monitor continuously. If there is a failure in the continuous
monitoring equipment, the system may substitute grab sampling every four
hours for up to five working days following the failure of the equipment.
Systems serving 3,300 or fewer people may monitor continuously or take
grab samples at the frequencies prescribed below:
System Size by Population Samples/day*
<500 1
50l-1,000 2
1,001-2,500 3
2,501-3,300 4
'Samples cannot be taken at the same time.
The sampling intervals are subject to Primacy Agency review and
approval.
If at any time the residual disinfectant concentration falls below 0.2
mg/1 in a system using grab sample monitoring, the system must continue to
take a grab sample every four hours until the residual disinfectant
concentration is equal to or greater than 0.2 mg/1. For all systems, if
the residual concentration is not restored to at least 0.2 mg/1 within
four hours after a value of less than 0.2 mg/1 is observed, the system is
3-29
-------
in violation of a treatment technique requirement, and must install
filtration. However, if the Primacy Agency finds that the exceedance was
caused by an unusual and unpredictable circumstance, the Primacy Agency
may choose not to require filtration. EPA expects Primacy Agencies to use
this provision sparingly; it is intended to encompass catastrophic events,
not infrequent large storm events. In addition, any time the residual
concentration falls below 0.2 mg/1, the system must notify the Primacy
Agency. Notification must occur as soon as possible, but no later than
the end of the next business day. The system also must notify the Primacy
Agency by the end of the next business day whether or not the residual was
restored within four hours.
Failure of a monitoring or reporting requirement does not trigger a
requirement to filter although they are violations.
Maintaining a Residual Within the System
To avoid filtration, the disinfectant residual in the distribution
system cannot be undetectable in more than five percent of the samples in
a month, for any two consecutive months that the system serves water to
the public. Systems may measure HPC instead of disinfectant residual.
Sites with HPC concentrations of less than or equal to 500/ml are
considered equivalent to sites with detectable residuals for the purpose
of determining compliance. Public water systems must monitor for the
presence of a disinfectant residual (or HPC levels) at the same frequency
and locations as total coliform measurements taken pursuant to the Total
Colifonn Rule. However, if the Primacy Agency determines, based on site-
specific considerations, that a system has no means for having a sample
transported and analyzed for HPC by a certified laboratory within the
requisite time and temperature conditions (Method 907, APHA, 1985), but
that the system is providing adequate disinfection in the distribution
system, this requirements does not apply to that system.
For systems which use both surface and ground water sources, the
Primacy Agency may allow the system to take disinfectant residual or HPC
samples at points other than the total colifonn sampling locations if it
determines that such points are more representative of treated (disin-
fected) water quality within the distribution system.
3-30
-------
Disinfectant residual can be measured as total chlorine, free
chlorine, combined chlorine or chlorine dioxide (or HPC level). The SWTR
lists the approved analytical methods for these analyses. For example,
several test methods can be used to test for chlorine residual in the
water, including amperometric titration, DPD colorimetric, DPD ferrous
titrimetric method and iodometric method, as described in the 16th Edition
of Standard Methods.8 Appendix 0 provides a review and summary of
available techniques to measure disinfectant residuals.
If a system fails to maintain a detectable disinfectant residual or
an HPC level of less.than or equal to 500/ml in more than 5 percent of the
samples during a month, for any two consecutive months the system serves
water to the public, the system is in violation of a treatment technique
requirement. In addition, this system must install filtration unless the
Primacy Agency determines that the violation was not due to a deficiency
in treatment of the source water (e.g., the violation was due to a
deficiency in the distribution system, such as cross-connection contamina-
tion or failure in the pipeline).
The absence of a detectable disinfectant residual in the distribu-
tion system may be due to a number of factors, including:
Insufficient chlorine applied at the treatment plant
Interruption of chlorination
A change in chlorine demand in either the source water or the
distribution system
Long standing times and/or, long transmission distances
Available options to correct the problem of low disinfectant
residuals in distribution systems include:
Routine flushing
Also, portable test kits are available which can be used in the field
to detect residual upon the approval of the Primacy Agency. These kits
may employ titration or colorimetric test methods. The colorimetric
kits employ either a visual detection of a residual through the use of
a color wheel, or the detection of the residual through the use of a
hand held spectrophotometer.
3-31
-------
Increasing disinfectant doses at the plant
Cleaning of the pipes (either mechanically by pigging or by
the addition of chemicals to dissolve the deposits) in the
distribution system to remove accumulated debris which may be
exerting a disinfectant demand;
Flushing and disinfection of the portions of the distribution
system in which a residual is not maintained; or
Installation of satellite disinfection feed facilities within
the distribution system.
For systems unable to maintain a residual, the Primacy Agency may
determine that it is not feasible for the system to monitor HPC and judge
that disinfection is adequate based on site-specific conditions.
Additional information on maintaining a residual in the system is
available in the AWWA Manual of Water Supply Practices and Water
Chlorination Principles and Practices.
3.2.4 Disinfection System Redundancy
Another requirement for unfiltered water supply systems is
disinfection facility redundancy. A system providing disinfection as the
only treatment is required to assure that the water delivered to the
distribution system is continuously disinfected. The SWTR requires either
redundant disinfection equipment with auxiliary power and automatic
start-up and alarm; or an automatic shutoff of delivery of water to the
distribution system when the disinfectant residual level drops below 0.2
mg/L. In order to fulfill the requirement of providing redundant
disinfection facilities, the following system is recommended:
All components have backup units with capacities equal to or
greater than the largest unit on-line.
A minimum of two storage units of disinfectant which can be
used alternately • e.g., two cylinders of chlorine gas, two
tanks of hypochlorite solution
Where the disinfectant is generated on-site, such as ozone,
backup units with a capacity equal to or greater than that of
the largest unit on-line.
Automatic switchover equipment to change the feed from one
storage unit to the other before the first empties or becomes
inoperable
3-32
-------
Feed systems with backup units with capacities equal to or
greater than the largest unit on-line.
An alternate power supply such as a standby generator with the
capability of running all the electrical equipment at the
disinfection station. The generator should be on-site and
functional with the capability of automatic start-up on power
failure
Systems providing disinfection may have several different configura-
tions for type and location of disinfectant application. The following
guidelines are provided to assist Primacy Agencies and utilities in
determining the need for redundancy. Possible disinfection configurations
include:
one disinfectant used for primary and secondary disinfection
- one point of application
- multiple points of application
two different disinfectants used for primary and secondary
disinfection
In many cases one disinfectant will be used to fulfill both the
total inactivation and residual requirements. One or more application
points may be used to accomplish this. When one application point is used
to meet both the primary and secondary disinfection requirements, the
system is required to include redundant disinfection facilities.
When multiple points of application are used, redundancy is
recommended fOF-the disinfection facilities at each point of application
which is essential to meet the total inactivation requirements. In
addition, to assure the maintenance of a residual entering and throughout
the distribution system, either:
the last point of application prior to the distribution system
should have redundancy, or
the point of application immediately prior to this point
should have redundancy and sufficient capacity to assure a
residual entering the distribution system.
Systems may also use two different disinfectants, one to fulfill the
inactivation requirements and the second to maintain a residual. An
3-33
-------
example of this would include a system using ozone as a primary disinfec-
tant and chloramines as a secondary disinfectant. EPA recommends that:
the disinfection facilities at each point of disinfectant
application in the primary system essential in providing the
overall inactivation requirements include redundancy, and
the secondary disinfection facilities include redundancy,
unless the disinfectant used for primary disinfection can
provide a residual for the distribution system as well. If
the primary disinfectant can be used for residual maintenance,
the last point of primary disinfectant application should
include redundancy and sufficient capacity to assure a
residual entering the distribution system.
Appendix I contains more specific information to assist the Primacy
Agency in establishing requirements for providing redundant disinfection
facilities.
Providing automatic shutoff of water delivery requires approval by
the Primacy Agency. The Primacy Agency must determine that this action
will not result in an unreasonable risk to health or interfere with fire
protection. This determination should include the evaluation of the
system configuration to protect against negative pressures in the system,
and providing for high demand periods including fire flow requirements.
Automatic shutoff should be allowed only if systems have adequate
distribution system storage to maintain positive pressure for continued
water use.
3-34
-------
3.3 SITE-SPECIFIC CONDITIONS
In addition to meeting source water quality criteria and disinfec-
tion criteria, nonfiltering systems using surface water supplies must meet
the following criteria:
Maintain a watershed control program
Conduct a yearly on-site inspection
Determine that no waterborne disease outbreaks have occurred
Comply with the revised annual total coliform MCL
Comply with TTHM regulations (currently applies to systems
serving >10,000 people)
Guidelines for meeting these other criteria are presented in the following
sections.
3.3.1 Watershed Control Program
A watershed control program is a surveillance and monitoring program
which is conducted to protect the quality of a surface water source. An
aggressive and detailed watershed control program is desirable to
effectively limit or eliminate potential contamination by human viruses.
A watershed program may impact parameters such as turbidity, certain
organic compounds, viruses, total and fecal coliforms, and areas of wild-
life habitation. However, the program is expected to have little or no
impact on parameters such as naturally occurring inorganic chemicals.
Limiting human activity in the watershed may reduce the likelihood of
animals becoming infected with pathogens and thereby reduce the transmis-
sion of pathogens by wildlife. Preventing animal activity near the source
water intake prior to disinfection may also reduce the likelihood of
pathogen occurrence at the intake.
The effect of a watershed program is difficult to quantify since
many variables that influence water quality are beyond the control or
knowledge of the water supplier. As a result, the benefit of a watershed
control program or specific control measures must in many cases be based
on accumulated cause and effect data and on the general knowledge of the
impact of control measures rather than on actual quantification. The
effectiveness of a program to limit or eliminate potential contamination
by human viruses will be determined based on: the comprehensiveness of
the watershed review; the ability of the water system to effectively carry
3-35
-------
out and monitor the management decisions regarding control of detrimental
activities occurring in the watershed; and the potential for the water
system to maximize land ownership and/or control of land use within the
watershed. According to the SWTR, a watershed control program should
include as a minimum:
A description of the watershed including its hydrology and
land ownership
Identification, monitoring and control of watershed character-
istics and activities in the watershed which may have an
adverse effect on the source water quality
A program to gain ownership or control of the land within the
watershed through written agreements with land owners, for the
purpose of controlling activities which will adversely affect
the microbiological quality of the water
An annual report which identifies special concerns in the
watershed and how they are being handled, identifies activi-
ties in the watershed, projects adverse activities expected to
occur in the future and how the utility expects to address
them.
Appendix J contains a more detailed guide to a comprehensive
watershed program.
In preparing a watershed control program, surface water systems
should draw upon the State watershed assessments and nonpoint source (NPS)
pollution management programs required by S319 of the Clean Water Act.
Information on these programs is available from State water quality
agencies or EPA's regional offices. Assessments identify NPS pollutants
in water and assess the water quality. Utilities should use the
assessments when evaluating pollutants in their watershed. Surface water
quality assessments can also be obtained from the lists of waters prepared
under $304(1) of the Clean Water Act, and State biennially prepared
S305(b) reports.
State NPS management programs identify best management practices
(BMPs) to be employed in reducing NPS pollution. These management
programs can be incorporated in the watershed program to protect against
degradation of the source water quality.
3-36
-------
For systems using ground water sources under the influence of
surface water, the control measures delineated in the Wellhead Protection
(WHP) program encompass the requirements of the watershed control program,
and can be used to fulfill the requirements of the watershed control
program. Guidance on the content of State Wellhead Protection Programs
and the delineation of wellhead protection areas is given in: "Guidance
for Applicants for State Wellhead Protection Program Assistance Funds
Under the Safe Drinking Water Act," June, 1987, and "Guidelines for
Delineation of Wellhead Protection Areas," June, 1987, available from the
EPA office of Ground-Water Protection (WH-550G).
As a minimum, the WHP program must:
Specify the duties of State agencies, local governmental
entities and public water supply systems with respect to the
development and implementation of Programs;
Determine the wellhead protection area (WHPA) for each
wellhead as defined in subsection 1428(e) based on all
reasonably available hydrogeologic information, ground-water
flow, recharge and discharge and other information the State
deems necessary to adequately determine the WHPA;
Identify within each WHPA all potential anthropogenic sources
of contaminants which may have any adverse effect on the
health of persons;
Describe a program that contains, as appropriate, technical
assistance, financial assistance, implementation of control
measures, education, training and demonstration projects to
protect the water supply within WHPAs from such contaminants;
Present contingency plans for locating and providing alternate
drinking water supplies for each public water system in the
event of well orwellfield contamination by such contaminants;
Consider all potential sources of such contaminants within the
expected wellhead area of a new water well which serves a
public water supply system; and
Provide for public participation.
3.3.2 Qn-site Inspection
The watershed control program and on-site inspection are inter-
related preventive strategies. On-site inspection is actually a program
which includes and surpasses the requirements of a watershed program.
3-37
-------
While the watershed program is mainly concerned with the water source,
on-site inspection includes some additional requirements for source water
quality control and is also concerned with the disinfection facilities.
As defined by the EPA, an on-site inspection includes review of the water
source, disinfection facilities and operation and maintenance of a public
water system for the purpose of evaluating the adequacy of such systems
for producing safe drinking water.
The SWTR requires an annual on-site inspection to evaluate the
watershed control program and disinfection facilities. The inspection
must be performed by a party approved by the Primacy Agency. The inspec-
tion should be conducted by competent individuals such as sanitary and
civil engineers, sanitarians, and technicians who have experience and
knowledge in the operation, maintenance, and design of water systems, and
who have a sound understanding of public health principles and waterborne
diseases. Guidance for the contents of an inspection are included in the
following paragraphs. Appendix K presents guidelines for a sanitary
survey which includes and surpasses the requirements of an on-site
inspection.
As the first step in determining which SWTR requirements, if any, a
source is subject to, EPA recommends that utilities conduct a detailed,
comprehensive sanitary survey. Appendix K presents a comprehensive list
of water system features that the person conducting the survey should be
aware of and review as appropriate. This initial investigation estab-
lishes the quality of the water source, its treatment and delivery to the
consumer. EPA"recommends that this comprehensive evaluation be repeated
every three years for systems serving 4,100 people or less and every five
years for systems serving more than 4,100 people. Also, under the Total
Coliform Rule, ground water systems which take less than 5 coliform
samples per month must conduct such sanitary surveys within every 5 or 10
years depending upon whether the source is protected and disinfected.
The annual on-site inspection to fulfill the SHTR requirements
should include as a minimum:
3-38
-------
1. Source Evaluation
a. Review the effectiveness of the watershed control
program (Appendix J).
b. Review the physical condition and protection of the
source intake.
c. Review the maintenance program to insure that all
disinfection equipment is appropriate and has received
regular maintenance and repair to assure a high operat-
ing reliability.
2. Treatment Evaluation
a. Review improvements and/or additions made to disinfec-
tion processes during the previous year to correct
deficiencies detected in earlier surveys.
b. Review the condition of disinfection equipment.
c. Review operating procedures.
d. Review data records to assure that all required tests
are being conducted and recorded and disinfection is
effectively practiced (CT calculations should be spot
checked to ensure that they were done correctly).
e. Identify any needed improvements in the equipment,
system maintenance and operation, or data collection.
In addition to these requirements, a periodic sanitary survey is
recommended for all systems, including those with filtered and unfiltered
supplies. The s.anitary survey should include the items listed in 1 and 2
above as well as:
3. Distribution System Evaluation
a. Review the condition of storage facilities.
b. Determine that the system has sufficient pressure
throughout the year.
c. Verify that system equipment has received regular
maintenance.
d. Review additions/improvements incorporated during the
year to correct deficiencies detected in the initial
inspection.
3-39
-------
e. Review cross connection prevention program, including
annual testing of backflow prevention devices.
f. Review routine flushing program for effectiveness.
g. Evaluate the corrosion control program and its impact on
distribution water quality.
h. Review the adequacy of the program for periodic storage
reservoir flushing.
i. Review practices in repairing water main breaks to
assure they include disinfection.
4. Management/Operation Evaluation
a. Review the operations to assure that any difficulties
experienced during the year have been adequately
addressed.
b. Review staffing to assure adequate numbers of properly
trained and/or certified personnel are available.
c. Verify that a regular maintenance schedule is followed.
d. Audit systems records to verify that they are adequately
maintained.
e. Review bacteriological data from the distribution system
for coliform occurrence, repeat samples and action
response.
3.3.3 No Disease Outbreaks
Under the provisions of the SWTR, a surface water system which does
not filter must not have been identified as a source of waterborne
disease, or if it has been so identified, the system must have been
modified sufficiently to prevent another such occurrence, as determined by
the Primacy Agency. If a waterborne disease outbreak has occurred and the
outbreak was or is attributed to a treatment deficiency, then the system
must install filtration unless the system has upgraded its treatment
system to remedy the deficiency which led to the outbreak and the Primacy
Agency has determined that the system is satisfying this requirement. If
the Primacy Agency has determined the disease outbreak was the result of
a distribution system problem rather than a source water treatment
deficiency, the system is not required to install filtration.
3-40
-------
In order to determine whether the above requirement is being met,
the responsible federal, state and local health agencies should be
surveyed to obtain the current and historical information on waterborne
disease outbreaks which may have occurred within a given system. Whether
conducted by the Primacy Agency or submitted by the water purveyor, this
information should include:
1. Source of the Information:
a. Name of agency
b. Name and phone number of person contacted
c. Date of inquiry
2. Outbreak Data
a. Known or suspected incidents of waterborne disease
outbreaks
b. Date(s) of occurrence(s)
c. Type or identity of illness
d. Number of cases
3. Status of Disease Reporting:
a. Changes in regulations; e.g., giardiasis was not a
reportable disease until 1985
4. If a Disease Outbreak has Occurred:
a. Was the reason for the outbreak identified; e.g.,
inadequate disinfection?
b. Did the outbreak occur while the system was in its
current configuration?
c. - Was remedial action taken?
d. Have there been any further outbreaks since the remedial
action was taken?
If a review of the available information indicates that the system
or network for disease reporting is inadequate within the Primacy Agency's
area of responsibility, efforts should be made to encourage the appropri-
ate agencies to upgrade the disease reporting capabilities within the
area.
3-41
-------
3.3.4 Monthly Coll form MCL
To avoid filtration, a system must comply with the MCL for total
coliforms, established in the Total Coliform Rule, for at least 11 out of
the previous 12 months the system served water to the public on an ongoing
basis, unless the Primacy Agency determines that failure to meet this
requirement was not caused by a deficiency in treatment of the source
water. If the Primacy Agency makes such a determination, the system is
not required to install filtration. The Total Coliform Rule requires
systems using surface water or ground water under the influence of surface
water which do not filter to collect a sample at or near the first
customer each day that the turbidity level exceeds 1 NTU within 24 hours
of learning of the result and to analyze the sample for the presence of
total coliforms. (If the Primacy Agency determines that it is not
possible for the system to have such a sample analyzed within 24 hours,
this time limit may be extended on a case-by-case basis.) This sample may
be used to fulfill the routine compliance monitoring requirements of the
Total Coliform Rule. The results of the additional sample must be
included in determining whether the system is in compliance with the
monthly MCL for total coliforms.
3.3.5 Total Trihalomethane (TTHM) Regulations
For the system to continue to use disinfection as the only
treatment, it must comply with the total trihalomethane (TTHM) MCL
regulation. The current regulation established an MCL for total TTHM of
0.10 mg/L for systems serving a population greater than 10,000. Both the
MCL and the system population covered may be reduced in the future, and
this should be considered when planning disinfectant application.
One alternative to meet the CT requirements of the SWTR is to
increase the disinfectant dose. For many systems, a higher chlorine dose
will result in increased formation of TTHMs. Changes in disinfection
practice should maintain TTHM levels of less than 0.10 mg/L. In lieu of
increasing chlorine dose, use of an alternate disinfectant which produces
fewer TTHMs could be considered. Alternate disinfectants include the use
of ozone or chlorine dioxide as primary disinfectants with chlorine or
chloramines as secondary (residual) disinfectants. It is important to
note that EPA also will promulgate regulations for disinfectants and
3-42
-------
disinfection by-products which may limit application of some of these
disinfectants. EPA recommends that Primacy Agencies keep informed through
communication with EPA on interim guidance on how to avoid conflict for
systems to comply with both the SWTR and the forthcoming regulations on
disinfectants and disinfection by-products. Any changes which appear to
not meet the by-product regulations should not be implemented.
3-43
-------
4. DESIGN AND OPERATING CRITERIA FOR
FILTRATION AND DISINFECTION TECHNOLOGY
4.1 Introduction
To comply with the SWTR, public water systems must include filtra-
tion, or some other approved particulate removal technology, in their
treatment process unless they are able to satisfy certain conditions.
Those conditions include compliance with source water quality criteria and
site-specific criteria. Guidance for determining whether these conditions
are met is provided in Section 3 of this manual. Systems unable to
satisfy these conditions must provide particulate removal and meet
criteria pertaining to operation, design and performance. These criteria
are specified in part in the definitions of technologies in the SWTR and
more specifically as determined by the Primacy Agency.
This section provides guidance both for those water systems which
currently do not have filtration equipment and must add it, and for
systems which have existing filtration processes. Guidance on additional
alternatives for small systems is presented in Appendix L.
This section includes guidance on the following topics:
a. Filtration Technology: Descriptions, capabilities, design
criteria and operating requirements for each technology, and
a listing of major factors to be considered in their
selection, including raw water quality considerations.
b. Disinfection: Descriptions of the most applicable disin-
fection technologies used with filtration systems, and a
presentation of the relative effectiveness of these disinfec-
tion technologies with respect to inactivation of bacteria,
cysts and viruses.
c. Alternate Technologies: Descriptions of some currently
available alternate filtration technologies.
d. Other Alternatives: Includes a description of some nontreat-
ment alternatives including regionalization and use of an
alternate source.
4.2 Selection of Appropriate Filtration Technology
Filtration is generally provided by passing water through a bed of
sand, a layer of diatomaceous earth or a combination bed of coarse anthra-
4-1
-------
cite coal overlaying finer sand. Filters are classified and named in a
number of ways. For example, based on application rate, sand filters can
be classified as either slow or rapid; yet these two types of filters
differ in many more characteristics than just application rate. They
differ in their removal process, bed material, method of cleaning, and
operation. Based on the type of bed material, filters can be classified
as sand, diatomaceous earth, dual-media (coal-sand) or even multi-media
in which a third layer of high density sand is used.
4.2.1 General Descriptions
Current technologies specified by the SWTR are:
a. Conventional Treatment: A series of processes including
coagulation, flocculation, sedimentation and filtration.
b. Direct Filtration: A series of processes including coagula-
tion (and perhaps flocculation) and filtration, but excluding
sedimentation.
c. Slow Sand Filtration: A process which involves passage of raw
water through a bed of sand at low velocity, generally less
than 0.4 meters/hour (1.2 ft/hr), resulting in substantial
particulate removal by physical and biological mechanisms.
d. Diatomaceous Earth Filtration: A process that meets the
following conditions:
A precoat cake of diatomaceous earth filter media is
deposited on a support membrane (septum)
The water is filtered by passing it through the cake on
the septum; additional filter media, known as body feed,
is continuously added to the feed water in order to
maintain the permeability of the filter cake.
4-2
-------
e. Alternate Technologies: Any filtration process other than
those listed above. Available alternate filtration technolo-
gies include, but are not limited to:
Package Plants1
Cartridge Filters
4.2.2 Capabilities
Filtration processes provide various levels of turbidity and
microbial contaminant removal. When properly designed and operated and
when treating source waters of suitable quality, the above filtration
processes are capable of achieving at least a 2-log (99 percent) removal
of Giardia cysts and at least a 1-log (90 percent) removal of viruses
without disinfection (Logsdon, 1987b; USEPA, 1938b; Roebeck, 1962). The
exception is cartridge filters which may not provide effective virus
removal. A summary of the removal capabilities of the various filtration
processes is presented in Table 4-1.
As indicated in Table 4-1, conventional treatment without disinfec-
tion is capable of achieving up to a 3-log removal of Giardia cysts and
up to a 3-log removal of viruses. Direct filtration can achieve up to a
3-log removal of Giardia cysts and up to a 2-log removal of viruses.
Achieving the maximum removal efficiencies with these treatment processes
requires the raw water to be properly coagulated and filtered. Factors
which can adversely affecy removal efficiencies include:
Raw water turbidities less than 1 NTU
Cold water conditions
Non-optimal or no coagulation
Improper filter operation including:
1 Depending upon the type of treatment units in place, historical
performance and/or pilot plant work, these plants could be categorized
as one of the technologies in a-d above at the discretion of the State.
Several studies have already indicated that some package plants
effectively remove Giardia cysts. If such plants provided adequate
disinfection so that the complete treatment train achieves at least a
3-log removal/inactivation of Giardia cysts and a 4-log removal/inacti-
vation of viruses, use of this technology would satisfy the minimum
treatment requirements.
4-3
-------
No filter to waste
Intermittent operation
Sudden rate changes
Poor housekeeping
Operating the filters after turbidity breakthrough
Studies of slow sand filtration have shown that this technology
(without disinfection) is capable of providing greater than a 3-log
removal of Giardia. cysts and greater than a 3-log removal of viruses.
Factors which can adversely affect removal efficiencies include:
Poor source water quality
Cold water conditions
Increases in filtration rates
Decreases in bed depth
Improper sand size
Inadequate ripening
Diatomaceous earth (DE) filtration can achieve greater than a 3-log
removal of Giardia cysts when sufficient precoat and body feed are used.
However, turbidity and total colifonn removals are strongly influenced by
the grade of DE employed. Conversely, DE filtration is not very effective
for removing viruses unless the surface properties of the diatomaceous
earth have been altered by pretreatment of the body feed with alum or a
suitable polymer. In general, DE filtration is assumed to achieve only
a 1-log removal of viruses unless demonstrated otherwise. Factors which
can affect the removal of Giardia cysts and viruses include:
Precoat thickness
Amount of body feed
Grade of DE
Improper conditioning of septum
- Improper pretreatment of the body feed
Package plants can be used to treat water supplies for communities
as well as for recreational areas, parks, construction camps, ski resorts,
military installations and other facilities where potable water is not
available from a municipal supply. Operator requirements vary signifi-
cantly with specific situations. Under unfavorable raw water conditions,
4-4
-------
TABLE 4-1
REMOVAL CAPABILITIES OF FILTRATION PROCESSES''1'
Process
Conventional
Direct Filtration
Slow Sand Filtration
Diatomaceous Earth
Filtration
Giardia;2)
Cvsts
atment 2
n 2
tion 2
- 3
- 3
- 3(5>
Yi
1
1
1
Log Removals
ruses
. 2(3)
- 3<4>
Total a]
Col i form
>4
1 - 3
1 - 2
2 - 3
(5)
1 - 2
(2)
1 - 3
Note:
1. Without disinfection.
2. Logsdon, 1987b.
3. Roebeck si il 1962.
4. Poynter and Slade, 1977.
•+. ruynicr aim Jiauc, i7//.
5. These technologies generally achieve greater than a 3-log removal.
-------
package plants could demand full-time attention. Package plants are most
widely used to treat surface supplies for removal of turbidity, color and
coliform organisms prior to disinfection. They are currently available
in capacities up to 6 mgd.
Colorado State University conducted a series of tests on one package
plant over a 5-month period during the winter of 1985-86 (Horn and
Hendricks, 1986). Existing installations in Colorado had proven effective
for turbidity removal, and the tests at the university were designed to
evaluate the system's effectiveness in removing coliform bacteria and
Giardia cysts from low turbidity, low temperature source waters. The test
results showed that the filtration system could remove greater than
99 percent of Giardia cysts for waters which had less than 1 NTU turbidity
and less than 5 C temperatures, as long as proper chemical treatment was
applied, and the filter rate was 10 gpm/ft2 or less. In addition, an
alternate water source having a turbidity ranging from 3.9 to 4.5 NTU was
used in 12 test runs with coagulant doses ranging from 15 to 45 mg/l. The
effluent turbidities from these runs were consistently less than 0.5 NTU.
Surveys of existing facilities indicate that while package plants
may be capable of achieving effective treatment, many have not consistent-
ly met the interim MCL for turbidity, and in some cases, coliforms were
detected in the filtered water (Morand et al., 1980; Morand and Young,
1983). The performance difficulties were primarily the result of the
short detention time inherent in the design of the treatment units, the
lack of skilled operators with sufficient time to devote to operating the
treatment facilities, and the wide-ranging variability in quality of the
raw water source. For instance, raw water turbidity was reported to often
exceed, 100 NTU at one site. Improvements in operational techniques and
methods at this site resulted in a substantial improvement in effluent
quality. After adjustments were made, the plant was capable of producing
a filtered water with turbidities less than 1 NTU, even when influent
turbidities increased from 17 to 100 NTU within a 2-hour period, as long
as proper coagulation was provided.
One of the major conclusions of these surveys was that package water
treatment plants manned by competent operators can consistently remove
4-5
-------
-------
turbidity and bacteria from surface waters of a fairly uniform quality.
Package plants applied where raw water turbidities are variable require
a high degree of operational skill and nearly constant attention by the
operator. Regardless of the quality of the raw water source, all package
plants require at least a minimum level of maintenance and operational
skill and proper chemical treatment if they are to produce satisfactory
water quality.
Cartridge filters using microporous filter elements (ceramic, paper
or fiber) with pore sizes as small as 0.2 urn may be suitable for producing
potable water from raw water supplies containing moderate levels of
turbidity, algae and microbiological contaminants. The advantage to small
systems of these cartridge filters is that, with the exception of
disinfectant, no other chemicals are required. The process is one of
strictly physical removal of small particles by straining as the water
passes through a porous cartridge. Other than occasional cleaning or
cartridge replacement, operational requirements are not complex and do not
require skilled personnel. However, the SWTR does require each surface
water system to be operated by a qualified operator, as determined by the
Primacy Agency. Such a system may be suitable for some small systems
where, generally, only maintenance personnel are available for operating
water supply facilities. However, the use of cartridge filters should be
limited to low turbidity source waters because of their susceptibility to
rapid headless buildup. For example, manufacturer's guidelines for
achieving reasonable filter run lengths with certain polypropylene filter
elements are that the raw water turbidity be 2 NTU or less (USEPA, 1988b).
Long (1983) analyzed the efficacy of a variety of cartridge filters
using turbidity measurements, particle size analysis, and scanning
electron microscope analysis. The filters were challenged with a
suspension of microspheres averaging 5.7 urn in diameter which is smaller
than a Giardia cyst. The microspheres were applied at a concentration of
40,000 to 65,000 spheres per ml. Ten of 17 cartridge filters removed over
99.9 percent of the microspheres.
In tests using live infectious cysts from a human source, cartridge
filters were found to be highly efficient in removing Giardia cysts
4-6
-------
(Hibler, 1986). Each' test involved challenging a filter with 300,000
cysts at a concentration of 10,000 cysts/ml. The average removal for five
tests was 99.86 percent, with removal efficiencies ranging from 99.5 per-
cent to 99.99 percent.
The application of cartridge filters to small water systems using
either cleanable ceramic or disposable polypropylene cartridges appears
to be a feasible method for removing turbidity and most microbiological
contaminants. However, data regarding the ability of cartridge filters
to remove viruses are not available. Since disinfection by itself could
achieve a 4-log inactivation of viruses, if the cartridge filter removes
greater than or equal to 3 logs of Giardia. then the filter plus
disinfection would achieve the overall minimum requirements, regardless
of whether only negligible Giardia inactivation is achieved (e.g., less
than 0.5 log). However, consideration should be given to the feasibility
of providing multiple barriers of treatment for each target organism,
i.e., some Giardia and virus removal by each barrier (i.e., some removal
by filtration and some inactivation by disinfection) as protection in case
one of the barriers fails. The efficiency and economics of the process
must be closely evaluated for each situation. Pretreatment in the form
of roughing filters (rapid sand or multi-media) or fine mesh screens may
be needed to remove larger suspended solids which, if not removed, could
cause the rapid buildup of headloss across the cartridges (USEPA, 1988a).
In general, conventional treatment, direct filtration, slow sand
filtration and diatomaceous earth filtration can be designed and operated
to achieve the-maximum removal of the water quality parameters indicated
in Table 4-1. However, for the purpose of selecting the appropriate
filtration and disinfection technologies and for determining design
criteria, these filtration processes should be assumed to achieve a 2-log
removal of-Siardia cysts and a 1-log removal of viruses. This conserva-
tive approach will assure that the treatment facility has adequate
capability to respond to non-optimum performance due to changes in raw
water quality, plant upsets, etc. The balance of the required removals
and/or inactivation of Giardia cysts and viruses would be achieved through
the application of appropriate disinfection.
4-7
-------
The performance of alternate technologies such as cartridge filters,
and possibly package plants, depending upon the unit under consideration
cannot be stated with certainty at this time. Because of these perform-
ance uncertainties, pilot studies should be used to demonstrate their
efficacy for a given water supply.
4.2.3 Selection
For any specific site and situation, a number of factors will
determine which filtration technology is most appropriate. Among these
are: raw water quality conditions, space and personnel availability, and
economic constraints. A discussion of the impact of raw water quality on
the technology selection is presented here. The impact of site-specific
factors and economic constraints is presented in the USEPA document
"Technologies and Costs for the Removal of Microbial Contaminants from
Potable Water Supplies" (USEPA, 1988b).
Raw Water Quality Conditions
The number of treatment barriers provided .should be commensurate
with the degree of contamination in the source water. The four technolo-
gies specified in the SWTR vary in their ability to meet the performance
criteria when a wide range of raw water quality is considered. While the
numerical values of raw water quality that can be accommodated by each of
the four technologies will vary from site to site, general guidance can
be provided. General guidelines for selecting filtration processes, based
on total coliforra count, turbidity, and color are presented in Table 4-2.
It is not recommended that filtration systems other than those listed in
Table 4-2 be used when the general raw water quality conditions exceed
the values listed, unless it has been demonstrated through pilot testing
that the technology can meet the performance criteria under the raw water
quality conditions expected to occur at the site.
The filtration processes listed in Table 4-1 are capable of
achieving the required performance criteria when properly designed and
operated if they are treating a source water of suitable quality (i.e.,
generally within the ranges indicated in Table 4-2). One of the causes
4-8
-------
TABLE 4-2
GENERALIZED CAPABILITY OF FILTRATION SYSTEMS
TO ACCOMMODATE RAH WATER QUALITY CONDITIONS
General Restrictions
Total
Coliforms Turbidity Color
Treatment (1/100 m]\ (NTU1 (CU)
Conventional with
predisinfection <20,000(3) No restrictions^' <75'2)
Conventional without
predisinfection <5,000(3) No restrictions(3) <75'-Z)
Direct filtration
with flocculation <500k3) <7-14(1) <40(4)
In-line filtration <500(3) <7-14a) <10(3)
Slow sand filtration <800l5) <10(5) <5(3)
Diatomaceous earth
filtration <50(3) <5{3) <5;3)
Notes;
1. Depends on algae population, alum or cat ionic polymer
coagulation -- (Cleasby et al., 1984.)
2. USEPA, 1971.
3. letter-man, 1986.
4. Bishop et al., 1980.
5. Slezak and Sims, 1984.
-------
of filtration failures is the use of inappropriate technology for a given
raw water quality (Logsdon, 1987b). These criteria are general guide-
lines. Periodic occurrences of raw water coliform, turbidity or color
levels in excess of the values presented in Table 4-2 should not preclude
the selection or use of a particular filtration technology. For example,
the following alternatives are available for responding to occasional raw
water turbidity spikes:
a. Direct Filtration
Continuous monitoring and coagulant dose adjustment
More frequent backwash of filters
Use of presedimentation
b. Slow Sand Filtration
Use of a roughing filter
Use of an infiltration gallery
c. Oiatomaceous Earth Filtration
Use of a roughing filter
Use of excess body feed
For the above alternatives, EPA recommends that pilot testing be
conducted to demonstrate the efficacy of the treatment alternative.
The characteristics of each filtration technology are a major factor
in the selection process. Significant characteristics include performance
capabilities (contaminant removal efficiencies), design and construction
requirements, and operation and maintenance requirements. Details
regarding each of the four filtration technologies are presented in the
following section.
4.3 Available Filtration Technologies
4.3.1 Introduction
As indicated in the preamble to the SWTR, the historical responsi-
bility of the States to establish design and operating criteria for public
drinking water plants will continue. The purpose of the following
sections is to provide guidance on how the design and operating criteria
may need to be changed in order to assure that the performance criteria
in the SWTR are met.
4-9
-------
The design criteria for the various filtration technologies found
in the 1987 edition of Recommended Standards for Water Works (Great Lakes,
1987) are the minimum design criteria that a majority of states are
currently following.2 These standards are referred to as Ten States
Standards in the remainder of this manual. The design criteria contained
in the Ten States Standards have not been duplicated here. Rather, the
reader is referred to the Ten States Standards directly. EPA recommends
the following additions and/or changes to the Ten State Standards in order
to assure compliance with the performance criteria of the SVTR.
4.3.2 General
The following recommendations apply to all filtration plants:
a. All filtration plants should provide continuous turbidity
monitoring of the effluent turbidity from each individual
filter.3•* If continuous monitoring is impractical, routine
monitoring of individual filters is recommended as a minimum.
b. All filtration systems should be concerned with the peak
turbidity levels in the filtered water after backwashing and
2 Based upon the results of a survey conducted for the American Water
Works Association Research Foundation (AWWARF), some 38 states use the
Ten States Standards entirely or in modified form (AWWARF, 1986).
Although "this is not a requirement of the SWTR, it is recommended
because of the possibility that not all filters in a treatment plant
will produce the sane effluent turbidity. This nay be due to a variety
of conditions that include bed upsets, failure of media support or
underdrain systems, etc. Although the combined effluent from all the
filters may meet the turbidity requirements of the SWTR, the turbidity
level from an individual filter may substantially exceed the limits.
This may result in the passage of Giardia cysts or other pathogens.
Validation should be performed at least twice a week based on the
procedure outlined in Part 214A in the 16th Edition of Standard Methods.
It should be noted that improper installation of continuous monitors
may allow for air bubbles to enter the monitor resulting in false
turbidity spikes. To avoid air bubbles reaching the turbidimeter the
sample tap should be installed below the center line of the pipe and
an air release valve may be included on the sample line.
4-10
-------
make every attempt to operate the filters to minimize the
magnitude and duration of these turbidity spikes.5
Individual filters should be monitored as discussed in Section
4.3.2.a and when excessive turbidity spikes are found, corrective actions
taken. During these turbidity peaks, Giardia cysts and other pathogens
may be passed into the finished water. There is evidence that a 0.2 to
0.3 NTU increase in the turbidity during the first period of the filter
run can be associated with rises in Giardia cyst concentrations by factors
of twenty to forty (Logsdon, 1985). Special studies should be conducted
to determine the extent of the turbidity spike problems.
There are basically four approaches available for correcting
problems with turbidity spikes after backwashing. These are as follows
(Bucklin, et al 1988):
Proper chemical conditioning of the influent water to the
filter can minimize the magnitude and duration of these
turbidity spikes. This could include proper control of the
primary coagulant chemicals such as alum or iron compounds.
In some cases filter aids using polymers may be needed to
control the turbidity spikes.
Gradually increasing the filtration rate in increments when
placing the filter in operation. Starting the filter at a low
flow rate and then increasing the flow in small increments
over 10 to 15 minutes has been shown to reduce the turbidity
spikes in some cases (Logsdon, 1987).
Addition of coagulants to the backwash water has also been
shown to reduce the extent of turbidity spikes after backwash.
Typically the same primary coagulant used in the plant is
addeb1 to the backwash water. Polymers alone or in combination
with the primary coagulant may also be used.
Filter-to-waste may be practiced where a portion of the
filtered water immediately after starting the filter is
wasted. This is only possible where the filter system has
For most high rate granular bed filters, there is a period of
conditioning, or break-in immediately following backwashing, during
which turbidity and particle removal is at a minimum, referred to as
the break-in period. The turbidity peaks are thought to be caused by
remnants of backwash water within the pores of and above the media
passing through the filter, and/or floe breakup during the filter
ripening period before it can adequately remove influent turbidity.
4-1.1
-------
provided the necessary valves and piping to allow this
procedure. There is some concern whether or not this practice
is beneficial. The extra valve operations needed for filter-
to-waste can disrupt the filter flow rate to the extent that
they create their own turbidity spikes. Some knowledge of the
time actually needed for filter-to-waste is also needed before
it can be determined that this is an effective procedure for
controlling turbidity spikes. If the length of time the
filter-to-waste is practiced is less than that before the
turbidity spike passes, the disruption caused by the valve
operation may actually increase the turbidity spike.
Different plants and the individual filters within the plant may
have different turbidity spike characteristics. The four approaches
presented above, therefore, must be evaluated on a case-by-case basis.
Special studies will be required to identify those filters with the
turbidity spike problems and assist in selecting which of the four
approaches is best for correcting the problem. It has been generally
found that turbidity spikes can be minimized through one or a combination
of the first three approaches.
In order to establish filter-to-waste operating guidelines, the
following procedure is suggested:
Review the effluent turbidity data for each filter and deter-
mine which filter historically has the highest effluent tur-
bidity.
Following backwashing of the filter with the poorest perfor-
mance, place that filter into service and collect crab samples
every 5 to 10 minutes for a period of at least 60 minutes.6
Analyze the grab samples for turbidity and determine how long
the filter must be in operation before the effluent turbidity
drops
to less than or equal to 0.5 NTU
or 1 NTU in cases where a filtered water turbidity of
less than or equal to 1 NTU is allowed.
6 Continuous turbidity monitoring can be used in place of grab sampling.
4-12
-------
Limited information exists on the typical magnitude and duration of
peak turbidity levels after backwashing and what levels are considered
acceptable to assure that these turbidity spikes are not associated with
passage of Giardia cysts. Information from plant scale tests, showing
the typical magnitude and duration of these turbidity spikes is available
from two plants (Bucklin et al.. 1988). Studies conducted at these plants
over a year showed that these peaks occurred within the first few minutes
after the filter was placed back in operation, their effects lasted for
several hours, and varied in magnitude from 0.08 to 0.35 NTU on average.
For existing plants without provisions for filter-to-waste, the
decision to add the necessary piping to provide this capability should be
made only after carefully evaluating the other three approaches. If the
results of special studies show that the other three options are not
effective in minimizing the turbidity spikes then the expense of adding
the filter-to-waste capabilities may be justified.
For new plants the capability of filter-to-waste may be required by
the Primacy Agency or should be considered. By having this capability,
additional flexibility will be available for turbidity spike control.
This flexibility may also be useful for other filter maintenance functions
such as after media replacement or when heavy chlorination of the filter
is needed after maintenance.
4.3.3 Conventional Treatment
Conventional treatment is the most widely used technology for
removing turbidity and microbial contaminants from surface water supplies.
Conventional treatment includes the pretreatnent steps of chemical
coagulation, rapid nixing, flocculation and sedimentation followed by
filtration. These conventional treatment plants typically use aluminum
and iron compounds in the coagulation processes. Polymers may also be
used to enhance the coagulation and filtration processes. A flow sheet
for a conventional treatment plant is presented on Figure 4-1.
Lime softening is a treatment process used to remove hardness and
turbidity from surface waters. Treatment is typically accomplished with
conventional process units. The lime softening process removes the
4-13
-------
calcium and magnesium from the water by precipitating them as calcium
carbonate and magnesium hydroxide. Turbidity levels in the water are also
reduced by this process. Lime and possibly soda ash is added to the raw
water to raise its pH to a point at which these precipitates are formed
and then removed from the water during sedimentation and filtration. Lime
softening may be used for the removal of carbonate hardness in the pH
range of 9 to 10 through a single stage process. Two-stage lime/soda ash
softening at a pH of 10 to 12 can be used for the removal of non-carbonate
hardness and magnesium. Two-stage softening includes recarbonation to
neutralize the caustic alkalinity, reducing the pH to the range of 8.5 to
9.5. A flow sheet for typical one- and two-stage softening plants is
presented on Figure 4-2.
Each of these three conventional treatment processes uses filtration
following sedimentation. Three different types of filters are used. Sand
filters, normally found in older plants, use a single media of sand to
form a filter bed, and are generally designed with a filtration rate of
2 gpm/ft2. Newer plants normally use dual-media or mixed media filters.
Dual media filters use a combination of anthracite coal along with a sand
to form the filter bed. Mixed media filters use coal, sand, and a third
material to form the filter bed. Dual and mixed media filters can be
designed to operate at higher filtration rates than sand filters, i.e.,
4 to 6 gpm/ft2.
Design (Criteria
The minimum design criteria in the Ten State Standards for
conventional treatment are considered sufficient for the purposes of
complying with the SHTR with the following addition:
The criteria for sedimentation should be expanded to include
other methods of solids removal including dissolved air
flotation. Plate separation and upflow-sol ids contact
clarifiers included in the 1987 Ten State Standards should
also be considered.
Operating Requirements
In addition to the operating requirements in the Ten State
Standards, a coagulant should be used at all times the treatment plant is
4-14
-------
COAGULANTS
RAPIO MIX
30 SEC 2 IMM
DETENTION
— ^
FLOCCULATION
20-45 MIN
— ^
SEONMENTATIOM
1-4 HOURS
FILTRATION
RAPID SAND 2 9pm ft?
DUAL AND TRI-MIXEO
MCCXA 4-< gpm ft?
FIGURE 4-1-FLOW SHEET OF A TYPICAL CONVENTIONAL
WATER TREATMENT PLANT
-------
SINGLE STAGE SOFTENING I1!
LIME
|aA»io MIX
JO SEC 2 M4M
3ETCNTION
•— ^
FUOCCULATION
:o-4$ MIN
i— ^i
[21
SEOIMCMTATION
« HOURS
— ^
FILTH AT IQN
flAPIO SAMO. 2 •••/««*
DUAL ANO MULTI
MCQIA 4.4 «««/n 1
1] PH RANGE 9-10
[2] OR ALTERNATE SOLIDS REMOVAL PROCESS
TWO STAGE SOFTENING [1
LIME
INFLUENT
SODA ASH
\
FLOCCULATOR-
CLARIFIER /
RECARBONATION
1J PH RANGE 10-12
FLOCCULATOR-\
CLARIFIER /'
SOFTENED WATER
FIGURE 4-2-FLOW SHEET OF TYPICAL SOFTENING TREATMENT PLANTS
-------
^ ;
in operation.' Conventional and direct filtration plants must be monitored
carefully because failure to maintain optimum coagulation can result in
poor filter performance and breakthrough of cysts and viruses.8 Although
the detention time provided by the settling basins results in some margin
of safety, the loss of coagulation control at the chemical feed or rapid
mix points may not be noticed until the poorly coagulated water reaches
the filters, after the process has failed. Failure to effectively monitor
and control filter operation can result in undetected poor filter
performance (Logsdon, 1987a; Logsdon, 1987b).
Effective operation of a conventional treatment plant requires
careful monitoring and control of:
Chemical Feed
Rapid Mix
Flocculation
Sedimentation
Filtration
For the purposes of the SWTR, the requirements for effective
operation of a conventional water treatment plant can be summarized as
follows:
a. The application of a coagulant and the maintenance of
effective coagulation and flocculation at all times when a
treatment plant is in operation.9 Proper process control
7 Dependable removal of Giardia cysts can not be guaranteed if a water
is filtered without being properly coagulated (Logsdon, 1987b; Al-Ani
et al., 1985). This is true even if the raw water turbidity is less
than 1 NTU.
As indicated in the preamble to the proposed SWTR, 33 percent of the
reported cases of giardiasis in waterborne disease outbreaks were
attributed to improperly operated filtration plants.
Some conventional water treatment plants which treat low turbidity
source waters (<1 NTU) reportedly discontinue the application of
coagulant(s) during periods of low turbidity since the raw water already
meets the turbidity MCL. However, studies have shown that cyst removal
for low turbidity waters is the most difficult to achieve and requires
optimum pretreatment including coagulation to achieve effective removals
4-15
-------
procedures should be used at the plant to assure that chemical
feeds are adjusted and maintained in response to variations
in raw water temperature and turbidity.
Maintenance of effective filtration will require proper
operation procedures to meet the turbidity requirements of the
SWTR. Proper operation should include:
Proper chemical conditioning of the water ahead of the
filter to assure adequate turbidity removal through the
filter.
Control of the flow rates and elimination of rapid
changes in flow rate applied to the filter.
Backwashing of filters before the filtered water quality
is degraded to the point that the plant fails to meet
the turbidity requirements of the SWTR. The criteria
on which to base initiation of backwash will have to be
determined for each plant. Experience with operation
cycles including run times and headless data may serve
as the basis for this site specific criteria.
After backwash bringing the clean filters back on line
so that excessive turbidity spikes in the filtered water
are not created. Section 4.3.2.B of this manual
discusses these turbidity spikes and approaches
available to minimize them.
Filters removed from service generally should be backwashed
upon start-up. However, in some cases, it may be impractical
to backwash filters each time they are removed from service.
Accordingly, the Primacy Agency may choose to allow start-up
without backwash ing under certain conditions on a site-by-site
basis. In making this decision, the following should be
considered:
_ - the length of time the filter was off-line
performance of the filter while being put on-line
The filter should be brought back on-line in such a way that
no turbidity spikes that could be associated with passage of
(Al-Ani et al., 1985).
4-16
-------
Giardia cysts and other pathogens occur. If problems with
turbidity spikes are found when starting up dirty filters,
special studies should be used to evaluate if any of the
approaches discussed in Section 4.3.2.B of this manual are
effective in minimizing the turbidity spikes.
4.3.4 Direct Filtration
A direct filtration plant can include several different pretreatment
unit processes depending upon the application. In its simplest form, the
process includes only in-line filters preceded by chemical coagulant
application and mixing. The mixing step, particularly in pressure
filters, can be satisfied by influent pipeline turbulence. In larger
plants with gravity filters, an open rapid-mix basin with mechanical
mixers typically is used. Figure 4-3 illustrates the unit processes of
a typical direct filtration plant.
Another variation of the direct filtration process consists of the
addition of a coagulant to the raw water followed by rapid mixing and
flocculation, as illustrated on Figure 4-4. The chemically conditioned
and flocculated water is then applied directly to a dual- or multi-media
filter (USEPA, 1988b).
Design Criteria
The 1987 edition of the Ten State Standards recommends pilot studies
to determine most design criteria. For the purposes of implementation of
the SWTR this requirement is considered sufficient with the following
exception:
a. A coagulant must be used at all times when the treatment plant
is in operation.
10
10 Optimum coagulation is critical for effective turbidity and microbiolog-
ical removals with direct filtration (Al-Ani et al., 1985).
4-17
-------
Operating Requirements
Operating considerations for direct filtration plants are essential-
ly identical to those for conventional treatment plants. The major
difference is that a direct filtration plant will not have a clarifier,
and may or may not have a flocculation or contact basin. In addition, EPA
recommends that all direct filtration plants, both new and existing, be
required to make provisions to minimize the break-in time of a filter
being put on-line.11
As with conventional treatment, the initiation of backwashing a
filter should first be based on filter effluent turbidity values, then by
headloss and run time. Effluent turbidity monitoring equipment should be
set to initiate filter backwash at an effluent value of 0.5 NIL) or less,
in order to meet filtered water quality requirements. Also, any filters
removed from service should be backwashed upon start-up. In some cases,
it may not be practical to backwash filters every time they are removed
from service. This decision should be made by the Primacy Agency on a
case-by-case basis, based on the same considerations as for conventional
systems.
4.3.5 Slow Sand Filtration
Slow sand filters differ from single-media rapid-rate filters in a
number of important characteristics. In addition to the difference of
flow rate, slow sand filters:
a. Function using biological mechanisms as well as physical-che-
mical mechanisms
b. Use smaller sand particles
c. Are not backwashed, but rather are cleaned by removing the
surface media
d. Have much longer run times between cleaning
11 As with conventional treatment, direct filtration produces a relatively
poor quality filtered water at the beginning of filter runs and
therefore a filter-to-waste period is recommended. In some cases, the
addition of a filter aid or bringing filters on-line slowly will be
appropriate (Cleasby et al., 1984).
4-18
-------
COAGULANTS
INFLUENT'
RAPID MIX
30 SEC 2 MIN
DETENTION
DUAL OR MIXED
MEDIA FILTER
4-5 gpm ft *
FIGURE 4-3 FLOW SHEET FOR A TYPICAL
DIRECT FILTRATION PLANT
COAGULANTS
•NFLUENT
.
RAPID MIX
30 SEC • 2 MIN
DETENTION
—
FLOCCULATION
15-30 MIN
— -^
DUAL OR MIXED
MEDIA FILTER
4-5 gpm ft 2
FIGURE 4-4-FLOW SHEET FOR A TYPICAL DIRECT
FILTRATION PLANT WITH FLOCCULATION
-------
e. Require a ripening period at the beginning of each run
Although rapid rate filtration is the water treatment technology
used most extensively in the United States, its use has often proved
inappropriate for small communities since rapid-rate filtration is a
technology that requires skilled operation by trained operators. Slow
sand filtration requires very little control by an operator. Consequent-
ly, use of this technology may be more appropriate for small systems where
source water quality is within the guidelines recommended in Section
4.2.3.
As indicated in this section, slow sand filtration also may be
applicable to other source water quality conditions with the addition of
pretreatment such as a roughing filter or presedimentation.
Design Criteria
The minimum design criteria presented in the Ten State Standards for
slow sand filters are considered sufficient for the purposes of implemen-
tation of the SWTR with the following exceptions:
a. Raw water quality limitations should be changed to reflect the '
values given in Table 4-2.12
b. The effective sand size should be between 0.15mm and 0.35mm
rather than the current 0.30 mm to 0.45 mm.13
Additional guidance on the design of slow sand filtration is
available in the design manual entitled Slow Sand Filtration for Community
Hater Supplies Technical Paper 24. 1987 published by the International
12 Without pretreatment, limitations exist in the quality of water that
is suitable for slow sand filtration (Logsdon, 1987b; Cleasby et al.,
1984; Bellamy et al., 1985; Fox et al., 1983).
13 Significant decreases in total coliform removals were shown at effective
sand sizes less than 0.35 mm (Bellamy et al., 1985). As defined in the
AWWA Standard for Filtering Material, effective size is the size opening
that will pass 10 percent by weight of a sample of filter material.
4-19
-------
Reference Centre for Community Water Supply and Sanitation (IRC),
P.O. Box 5500, 2280 HM Rijswijk, the Netherlands.
Operating Requirements
Maintenance of a slow sand filter involves two periodic tasks:
a. Removal of the top 2 to 3 cm (0.8 to 1.2 inches) of the
surface of the sand bed when the headloss exceeds 1 to 1.5 m.:<
b. Replacement of the sand when repeated scrapings have reduced
the depth of the sand to approximately one-half of its design
depth (Bellamy et al., 1985).
Following scraping, slow sand filters produce poorer quality
filtrate at the beginning of a run, and a filter-to-waste or ripening
period of one to two days is recommended before use to supply the system.
The ripening period is an interval of time immediately after a scraped
filter is put back on-line, when the turbidity or particle count results
are significantly higher than the corresponding values for the operating
filter. During this time, the microorganisms multiply and attain
equilibrium in the "schmutzdecke." Filter effluent monitoring results
should be used to determine the end of the ripening period. For example,
a turbidimeter could be set at 1.0 NTU or less to initiate start of the
filter run.
When repeated scrapings of the sand have reduced the depth of the
sand bed to approximately one-half of its design depth, the sand should
be replaced. „ Filter bed depths of less than 0.3 to 0.5 m (12 to 20
inches) have been shown to result in poor filter performance (Bellamy et
al., 1985). The replacement procedure should include removal of the
remaining sand down to the gravel support, the addition of new sand to one
half of the design depth and placement of the sand previously removed on
top of the new sand.15
14 Removal of this top layer of the "Schmutzdecke" should restore the
filter to its operational capacity and initial headloss.
15 This procedure results in clean sand being placed in the bottom half
of the filter bed and biologically active sand in the top half reducing
the amount of time required for the curing period. It also provides
4-20
-------
The amount of time for the biological population to mature in a new
sand filter (also called curing) and to provide stable and full treatment
varies. The World Health Organization (1980) reported that curing
requires a few weeks to a few months. Fox et al., (1983) found that
"about 30 days" were required to bring particle and bacterial effluents
down to a stable level. All researchers agree that a curing time for a
new filter is required before the filter operates at its fullest potential
(Bellamy et al., 1985).
4.3.6 Diatomaceous Earth Filtration
Diatomaceous earth (DE) filtration, also known as precoat or
diatomite filtration, is appropriate for direct treatment of surface
waters for removal of relatively low levels of turbidity and microorgan-
isms.
Diatomite filters consist of a layer of DE about 3 mm (1/8 inch)
thick supported on a septum or filter element. The thin precoat layer of
DE must be supplemented by a continuous body feed of diatomite, which is
used to maintain the porosity of the filter cake. If no body feed is
added, the particles filtered out will build up on the surface of the
filter cake and cause rapid increases in headless. The problems inherent
in maintaining the proper film of DE on the septum have restricted the use
of diatomite filters for municipal purposes, except under certain
favorable raw water quality conditions, i.e., low turbidity and good
bacteriological quality. Specific upper limits of raw water quality
parameters are-not well-defined because diatomaceous earth filtration
performance depends on the nature, as well as the concentration, of the
raw water particles and the grades of diatowite employed. Logsdon (1987b)
reported that filtered water turbidities above 1 NTU and short filter runs
were observed for several diatomaceous earth plants having maximum raw
water turbidities above 20 NTU.
for a complete exchange of sand over time, alleviating potential
problems of excessive silt accumulation and clogging of the filter bed
(Bellamy et al., 1985).
4-21
-------
Design Criteria
The minimum design criteria presented in the Ten State Standards for
diatomaceous earth filtration are considered sufficient for the purposes
of compliance with the SWTR with the following exceptions:
a. The recommended quantity of precoat is 1 kg/m2 (0.2 pounds per
square foot) of filter area, and the minimum thickness of the
precoat filter cake is 3mm to 5mm (1/8 to 1/5-inch).ls
b. Treatment plants should be encouraged to provide a coagulant
coating (alum or suitable polymer) of the body feed.17
Operating Requirements
Operating requirements specific to DE filters include:
Preparation of body feed and precoat
Verification that dosages are proper
Periodic backwashing and disposal of spent filter cake
Periodic inspection of the septum(s) for cleanliness or damage
Verification that the filter is producing a filtered water
that meets the performance criteria
4.3.7 Alternate Technologies
The SWTR allows the use of filtration technologies other than those
specified above provided that the system demonstrates to the Primacy
Agency using pilot studies or other means that the filtration technology
when combined with disinfection achieves at least 3-log Piardia cyst and
4-log virus removal/inactivation. Such technologies must also meet the
turbidity performance criteria for slow sand filtration. Guidance for
16 Studies have shown that a precoat thickness of 1 kg/m2 (0.2 lbs/ft:) was
most effective in Giardia cyst removal and that the precoat thickness
was more important than the grade size in cyst removal (DeWalle et al.,
1984; Logsdon et al., 1981; Bellamy et al., 1984).
17 Although enhancement of the DE is not required for Giardia cyst removal,
coagulant coating of the body feed has been found to significantly
improve removals of viruses, bacteria and turbidity. (Brown et al.,
1974; Bellamy et al., 1984).
4-22
-------
conducting pilot studies to demonstrate this effectiveness is provided in
Appendix M of this manual.
Reverse osmosis is a membrane filtration method which is used for
desalination and/or the removal of organic contaminants. The treatment
process is effective for the removal of Giardia cysts and viruses and no
demonstration is necessary.
Alternate filtration technologies which are currently available
include, but are not limited to:
Package Plants
Cartridge Filters
Package plants in principle are not a separate technology from the
preceding technologies. However, in many cases they are different enough
in design criteria, and operation and maintenance requirements that they
should be considered as an alternate technology. The package plant is
designed as a factory-assembled, skid-mounted unit generally incorporating
a single, or at most, a few process units. A complete treatment process
typically consists of chemical coagulation, flocculation, settling and
filtration. Package plants generally can be applied to flows ranging from
about 25,000 gpd to approximately 6 mgd (USEPA, 1988b). In cases where
the Primacy Agency believes that the design criteria of the package plant
corresponds to the design criteria of the processes established earlier
in this section (i.e., that the package plant qualifies as conventional
or direct filtration), the requirement of pilot testing may be waived.
The application of cartridge filters using either cleanable ceramic
or disposable polypropylene cartridges to snail water systems may be a
feasible method for removing turbidity and some microbiological contami-
nants, such as Giardia cysts although no data are available regarding
their ability to remove viruses. Pilot studies are required to demon-
strate the efficacy of this technology for a given supply. However, if
the technology was demonstrated to be effective through pilot plant
studies at one site, then the technology could be considered to be
effective at another site which had similar source water quality
conditions. Therefore, pilot plant testing at the new site might not be
necessary.
4-23
-------
It is important to note that the demonstration of achieving the 3-
log Giardia cyst and 4-log virus removal/inactivation requirements
includes disinfection. Thus, if a cartridge filter is demonstrated to
achieve a 3-log removal of Giardia cysts and it is determined by CTs that
the disinfection achieves at least a 4-1og virus inactivation, the
effectiveness of the technology would be demonstrated. The technology
must also maintain turbidities less than 1 NTU in 95 percent of the
monthly samples. Meeting this turbidity requirement assures a high
probability that turbidity will not interfere with disinfection and that
the inactivation efficiencies predicted by the CTs are reliable.
Design Criteria
After any necessary pilot studies are conducted and a successful
demonstration of performance has been made, design criteria should be
established and approved by the Primacy Agency. Eventually, a sufficient-
ly large data base will become available, making it easier to apply the
alternate technologies to other water supplies of similar quality.
Operating Requirements
After any necessary pilot studies are conducted and a successful
demonstration of performance has been made, operating requirements should
be established and approved by the Primacy Agency.
4.3.8 Nontreatment Alternatives
Under certain circumstances, some systems may have other alterna-
tives available. These alternatives include regionalization and the use
of alternate sources.
For small water systems which must provide filtration, a feasible
option way br to join with other small or large systems to form a region-
al water supply system. In addition, alternative water sources located
within a reasonable distance of a community which would allow the system
to meet the requirements of the SWTR and other applicable drinking water
regulations, may be developed to provide a satisfactory solution to a
community water quality problem. The availability of alternative ground
4-24
-------
water sources will depend upon the size and location of the system and
the costs involved.
4.4 Disinfection
4.4.1 General
The SWTR requires that disinfection be included as part of the
treatment of surface water for potable use. As noted earlier, EPA
recommends that the number of treatment barriers be commensurate with the
degree of contamination in the source water in accordance with Table 4-2.
For example, as indicated in Table 4-2, when the total coliforms in the
source water are greater than 5,000/100 ml, conventional treatment with
predisinfection is recommended. However, the selection of appropriate
disinfection requires consideration of other factors in addition to than
those included in Table 4-2. These considerations include:
a. Source water quality and the overall removal/inactivation of
Gjardia cysts and viruses desired.
b. Likelihood of TTHM formation.
c. Potential need for an oxidant for purposes other than
disinfection including control of taste, odor, iron,
manganese, color, etc.
4.4.2 Recommended Removal/inactivation
The SWTR requires a minimum 3-log removal/inactivation of Giardia
cysts and a minimum 4-log removal/inactivat ion of viruses:
a. We 1,1-operated conventional treatment plants which have been
optimized for turbidity removal can be expected to achieve at
least a 2.5-log removal of Giardia cysts.
b. Well-operated diatomaceous earth, slow sand filtration and
direct filtration plants can be expected to achieve at least
2—log removal of Giardja cysts.
EPA recommends that:
a. Conventional filtration systems provide sufficient disinfec-
tion to achieve a minimum of 0.5-log Giardia cyst and 2-log
virus inactivation.
4-25
-------
b. Slow sand filtration systems provide sufficient disinfection
to achieve a minimum of 2-log Giardia cyst and 2-log virus
inactivation.
c. Systems using diatomaceous earth and direct filtration, or
other filtration methods, should provide sufficient disinfec-
tion to achieve a minimum of 1-log Giardia cyst and 3-log
virus inactivation.
Further guidance on the disinfection level to be provided is
contained in Section 5. CT values for achieving these inactivations are
presented in Appendix E. As indicated in this Appendix:
a. A comparison of Tables E-l through E-6 with Table E-7
indicates that systems which achieve a 0.5-log inactivation
of Giardia cysts, using free chlorine, will achieve greater
than a 4-log inactivation of viruses.
b. Ozone and chlorine dioxide are generally more effective at
inactivating viruses than Giardia cysts. However, as
indicated in Tables E-8 through E-ll, there are some
conditions under which the disinfection needed to provide the
recommended virus inactivation is higher than that needed for
the recommended Giardia cyst inactivation. Therefore, a
system using ozone or chlorine dioxide for disinfection must
check the CT values needed to provide the recommended
inactivation of both Giardia cysts and viruses and provide the
higher of the two disinfection levels. Systems may demon-
strate their efficiency for overall removal/inactivation using
the protocol in Appendices G and M.
c. As indicated in Tables E-12 and E-13, chloramines are much
less effective for inactivating Giardia cysts and viruses than
the other disinfectants. Also, chloramines may be applied to
the system in several ways, either with chlorine added prior
to ammonia, ammonia added prior to chlorine or preformed. For
systems applying chlorine ahead of ammonia, the required level
of disinfection may be determined as follows:
determine the CT needed to provide the required
inactivation of Giardia and viruses and provide the
higher of the two levels or
4-26
-------
follow the protocol in Appendix G to demonstrate
effective inactivation to allow lower levels of
disinfection.
For systems applying ammonia ahead of chlorine or preformed
chloramines, the EPA recommends that the system demonstrate
effective virus inactivation according to the protocol in
Appendix G, since the CT values for virus inactivation in
Table E-13 only apply to the addition of chlorine prior to
ammon i a.
Although the SWTR requires a minimum of a 3-log removal/inactivation
of Giardia cysts and a minimum of a 4-log removal/inactivation of viruses,
it may be appropriate for the Primacy Agency to require greater removals/-
inactivations depending upon the degree of contamination within the source
water.
Rose (1988) conducted a survey of water sources to characterize the
level of Giardia cyst occurrence for "polluted" and "pristine" waters.
Polluted waters are defined as waters in the vicinity of sewage and
agricultural wastes, while pristine waters are those originating from
protected watersheds with no significant sources of microbiological
contamination from human activities. EPA believes that treatment should
be provided to assure less than one case of microbiologically-caused
illness per year per 10,000 people. In order to provide this level of
protection, 3, 4 or 5-log Giardia cyst removal/inactivation should be
provided for ttie following source water qualities:
Giardia Cyst Removal/inactivation Required Based11 19
on Source Hater Cvst Concentration
Giardia Inactivation 3-log 4-log 5-log
Allowable daily avg
cyst concentration/100 L <1 >1-10 >10-100
(geometric mean)
18 Rose, 1988.
19 10"4 annual risk per person based on consumption of 2 liters of *ater
daily.
4-27
-------
According to these guidelines, systems with sewage and agricultural
discharges to the source water should provide treatment to achieve an
overall 5-log removal/inactivation of Giardia cysts, while the minimum
required 3-log removal/inactivation is sufficient for sources with no
significant microbiological contamination from human activities. A 4-log
removal/inactivation of cysts should be provided for source waters whose
level of microbiological contamination is between these two extremes. The
location of discharges or other activities polluting the water supply with
respect to the location of the intake should also be considered in
determining the level of removal/inactivation needed. For instance, long
travel times and substantial dilution of a discharge will lessen the
impact of the discharge on the intake water quality, in which case less
of an increase in the overall treatment recommended above, would be
warranted. It is important to note that these levels of treatment for
different generalized source water characterizations are presented only
as guidelines. The Primacy Agency could develop disinfection requirements
based on these or other guidelines. It could also require systems with
available resources to conduct raw water monitoring for Giardia cyst
concentrations to establish the appropriate level of overall treatment and
disinfection needed.
The Primacy Agency may also review the nature of occurrence of
Giardia-sized particles in the raw water supply and the association with
turbidity occurrence. If it can be demonstrated that a higher degree of
removal of particles in the size range of Giardia is accomplished when
turbidity levels and associated Giardia levels are elevated, then a log
removal credit higher than 3 could be allowed for that particular
treatment plant, during such occurrences. This credit should correspond
to the log particle removal efficiencies accomplished, as determined by
particle counting data, or turbidity data if properly qualified. In all
cases, a minimum of 0.5 log reduction of Giardia should be achieved by
disinfection in addition to the removal credit allowed for by other
treatment.
4-28
-------
Until a risk analysis for exposure to viruses is developed, a rough
guideline for virus removal/inactivation, can be considered as follows:
a. For a 4-log Giardia. cyst removal/inactivation, a 5-log virus
removal/inactivation is recommended.
b. For 5-log Giardia removal/inactivation, a 6-log virus
removal/inactivation is recommended.
These guidelines assume that virus occurrence in the source water
is roughly proportional to Giardia cyst occurrence, and that
viruses occur at higher concentrations in source waters, or
are more infectious than Giardia cysts and
infections from viruses may have more health risk significance
than Giardia cysts.
Based on these assumptions, higher levels of protection are warranted.
To meet the levels of inactivation recommended here, significant
changes in the system may be required. To avoid changes in the system
which may result in conflicts with future regulations, the Primacy Agency
may wish to establish interim disinfection levels to provide protection
of the public health prior to the promulgation of the disinfection
by-product regulations and then reconsider whether these levels are still
appropriate after the disinfection by-product regulations are promulgated.
Guidance for establishing interim disinfection requirements is provided
in Section 5.5.
4.4.3 Total Tribalomethane fTTHEn Regulations
In addition to complying with disinfection requirements, systems
must meet the requirements of the TTHM regulations. Currently, this
regulation includes an MCL for TTHMs of 0.10 mg/L for systems which serve
greater than 10,000 people. EPA expects to issue new regulations with a
lower MCL in the near future. These regulations may also pertain to
systems serving less than 10,000 people. Therefore, the selection of an
appropriate disinfectant or disinfection strategy must include consid-
eration of current and future regulations.
4-29
-------
5. CRITERIA FOR DETERMINING IF FILTRATION
AND DISINFECTION ARE SATISFACTORILY PRACTICED
5.1 Introduction
Under the SWTR, new and existing filtration plants must meet
specified monitoring and performance criteria in order to assure that
filtration and disinfection are satisfactorily practiced. These criteria
include:
Turbidity monitoring requirements
Turbidity performance criteria
Disinfection monitoring requirements
Disinfection performance criteria
The overall objective of these criteria is to provide control of:
Giardia cysts; viruses; turbidity; HPC; and Legionella by assuring a high
probability that:
a. Filtration plants are well-operated and achieve maximum
removal efficiencies of the above parameters.
b. Disinfection will provide adequate inactivation of Giardia
cysts, viruses, HPC and Legione!la.
5.2 Turbidity Monitoring Requirements
5.2.1 Sampling Location
The purpose of the turbidity requirements for systems which use
filtration is to indicate:
a. Giardia cyst and general particulate removal for conventional
treatment and direct filtration
b. General particulate removal for diatomaceous earth filtration
and slow sand filtration
c. Possible interference with disinfection for all filtration
processes
To accomplish the purposes of the turbidity requirements, the SWTR
requires that the turbidity samples be representative of the system's
5-1
-------
filtered water. The sampling locations which would satisfy this
requirement include:
a. Combined filter effluent prior to entry into a clearwell,
b. Clear-well effluent;
c. Plant effluent or immediately prior to entry into the distri-
bution system; or
d. Average of measurements from each filter effluent.
The selection of sampling locations for demonstrating compliance
with the turbidity performance criteria is left to the system or the
preference of the Primacy Agency.
5.2.2 Sampling Frequency
The turbidity of the filtered water must be determined:
a. At least once every four hours that the system is in opera-
tion, or
b. The Primacy Agency may reduce the sampling frequency to once
per day for systems using slow sand filtration or filtration
treatment other than conventional treatment, direct filtration
or diatomaceous earth filtration, if it determines that less
frequent monitoring is sufficient to indicate effective
filtration performance. For systems serving 500 or fewer
people, the Primacy Agency may reduce the sampling frequency
to once per day regardless of the type of filtration used if
it determines that less frequent monitoring is sufficient to
indicate effective filtration performance.
A system may substitute continuous turbidity monitoring for grab
sample monitoring if it validates the continuous measurement for accuracy
on a regular basis using a protocol approved by the Primacy Agency. EPA
recommends that the calibration of continuous turbidity monitors be
verified at least twice per week according to the procedures established
in Method 214A of the 16th Edition of Standard Methods.1
1 Although the 17th Edition of Standard Methods is available, the 16th
Edition is referred to in the SWTR. Continuous turbidity monitors must
be installed properly to prevent air bubbles from reaching the monitor.
5-2
-------
5.2.3 Additional Monitoring
As indicated in Section 4.3.2, EPA recommends that systems equip
each filter with a continuous turbidity monitor. This recommendation is
not part of the requirements of the SWTR and is not required for
establishing compliance. Rather, it is recommended as a tool for systems
to use to better monitor their treatment efficiency and to provide a
method for detecting a deterioration in filter performance.
If continuous monitoring of each filter effluent cannot be
implemented, then EPA recommends that at least the following be conducted
on a quarterly basis:
a. Monitor each filter, either by grab samples or continuous
monitors, through the course of a routine cycle of operation,
i.e.: from restart to backwash
b. Visually inspect each filter where appropriate for indications
of physical deterioration of the filter
These are general suggestions. The Primacy Agencies are encouraged
to work with the systems to determine the best overall monitoring
program(s) for their particular filtration plants in order to assess the
status of the filter units. Each filter within a system should be
maintained so that each filter effluent meets the turbidity performance
criteria for the combined filter effluent (i.e., the turbidity limits
specified in the SWTR).
5.3 Turbidity Performance Criteria
The SWTR establishes turbidity performance criteria for each of the
filtration technologies. As previously indicated, these criteria provide
an indication of:
a. Effective particle and microbial removal
b. Potential for interference with disinfection
In filtration, effective particle removal depends on both physical
and chemical factors. The particles to be removed must be transported to
the surface of the media and they must attach to the media. When
efficient particle removal does not occur, the deterioration of filter
5-3
-------
performance can be due to either physical problems with the filters or
problems with the treatment chemistry.
Physical problems which can result in a deterioration of filter
performance include:
a. Media loss
b. Media deterioration
c. Mud ball formation
d. Channeling or surface cracking
e. Underdrain failure
f. Cross-connections
In addition, the treatment chemistry has a significant impact on
filtration. Specifically, effective particle removal is a function of
the:
a. Raw water chemistry and the changes induced by the chemicals
added
b. Surface chemistry of the particles to be removed
c. Surface chemistry of the media
Consequently, when a filter experiences particle or turbidity breakthrough
prior to the development of terminal headless, the search for alternatives
to correct the problem must include not only an evaluation of the
potential physical causes but the treatment chemistry as well. Generally
this involves an evaluation of one or more of the following:
a. Alternate coagulant type and/or dose
b. Alternate coagulant aid or flocculant aid type and/or dose
c. Need for an alternate oxidant type and/or dose
d. Need for a filter aid or alternate dose
5-4
-------
5.3.1 Conventional Treatment or Direct Filtration
The minimum turbidity performance criteria for systems using
conventional treatment or direct filtration are:
a. Filtered water turbidity must be less than or equal to 0.5 NTU
in 95 percent of the measurements taken every month.
b. Filtered water turbidity levels of less than or equal to 1 NTU
in 95 percent of the measurements taken every month may be
permitted on a case-by-case basis if the Primacy Agency
determines that the system (filtration with disinfection) is
capable of achieving the minimum overall performance require-
ments of 99.9 percent removal/inactivation of Giardia cysts at
the higher turbidity level. Such a determination could be
based upon an analysis of existing design and operating
conditions and/or performance relative to certain water quali-
ty characteristics. The design and operating conditions to be
reviewed include:
the adequacy of treatment prior to filtration,
the percent turbidity removal across the treatment
train, and
level of disinfection.
Water quality analysis which may also be used to evaluate the
treatment effectiveness include particle size counting before
and after the filter. Pilot plant challenge studies simulat-
ing full scale operation may also be used to demonstrate
effective treatment. Depending on the source water quality
and system size, the Primacy Agency will determine the extent
of the analysis and whether a pilot plant demonstration is
needed. For this demonstration, systems are allowed to
include disinfection in the determination of the overall
performance by the system.2
c. Filtered water turbidity may not exceed 5 NTU at any time.
The Primacy Agency can assume that conventional treatment plants
that are meeting the minimum performance criteria are achieving at least
a 2.5-log removal of GJarojia cysts and at least a 2-log removal of viruses
prior to disinfection.3
Recommended protocol for this demonstration is presented in Appendix M.
The literature indicates that well operated conventional treatment
plants can achieve up to 3-log reduction of Giardia cysts and viruses
(Logsdson, 1987b and Roebeck et al., 1962). Limiting the credit to
2.5-logs for Giardia cysts and 2-logs for viruses provides a margin of
safety by requiring more disinfection. This is consistent with the
5-5
-------
The Primacy Agency can assume that direct filtration plants that are
meeting the minimum performance criteria are achieving at least a 2-log
removal of Giardia cysts and a 1-log removal of viruses.4
Although the minimum turbidity performance criterion allows for a
maximum filtered water turbidity of 0.5 NTU, treatment facilities using
conventional treatment or direct filtration, whose raw water supplies have
turbidity levels of 1 NTU or less, should be encouraged to achieve
filtered water turbidity levels of less than 0.2 NTU.5
Primacy Agencies may allow systems which believe that they are
actually achieving greater than a 2- or 2.5-log Giardia cyst removal to
demonstrate the actual removal achieved using the protocol outlined in
Appendix M. It is reasonable to expect that systems using conventional
treatment for high turbidity source water (e.g., turbidities in excess of
100 NTU), and which optimize chemical treatment prior to filtration, may
be achieving a 3-log or greater Giardia cyst removal if their filter
effluent is substantially below the 0.5 NTU turbidity limit. Softening
plants using conventional processes and 2-stage treatment processes may
also achieve a 3-log Giardia cyst removal/inactivation. The high pH of
softening may result in inactivation of Giardia cysts and viruses which
can be demonstrated according to the protocol outlined in Appendix G.
Appendix M can be used to demonstrate the Giardia cyst removal achieved.
multiple barrier concept.
* Literature indicates that well operated direct filtration plants can
achieve up to a 3-log removal of Giardia cysts and up to a 2-log
removal of viruses (Logsdon, 1987b; Roebeck et al., 1962). Limiting
the credit to 2-log for Giardia cysts and 1-lpg for viruses provides a
margin of safety by requiring more disinfection. This is consistent
with the multiple barrier concept.
5 Research has demonstrated that filter effluent turbidities substantial-
ly lower than 0.5 NTU are needed to obtain effective removals of
Giardia cysts and viruses with low turbidity source waters (Logsdon,
1987b; Al-Anl et al., 1985).
5-6
-------
5.3.2 Slow Sand Filtration
For systems using slow sand filtration, the turbidity performance
requirements are:
a. The filtered water turbidity must be less than or equal to
1 NTU in 95 percent of the measurements for each month.
b. At the discretion of the Primacy Agency, a higher filter
effluent turbidity may be allowed for well operated plants
(Section 4.3.5) on a case-by-case basis, if there is no
interference with disinfection and the turbidity level never
exceeds 5 NTU. Noninterference with disinfection could be
assumed if the finished water entering the distribution system
is meeting the coliform MCL and HPC levels are less than 10/ml
during times of highest turbidity.
c. Filtered water turbidity may not exceed 5 NTU at any time.
Slow sand filtration plants, with appropriate design and operating
conditions and which meet the minimum turbidity performance criteria can
be considered to be well operated and achieving at least a 2-log removal
of Giardia cysts and 2-log removal of viruses without disinfection.6
Primacy Agencies may allow systems which believe that they are actually
achieving greater than a 2-log Giardia cyst removal to demonstrate the
actual removal achieved using the protocol outlined in Appendix M.
5.3.3 Diatomaceous Earth Filtration
For systems using diatomaceous earth filtration, the turbidity
performance criteria are:
a. The filtered water turbidity must be less than or equal to
1 NTU in 95 percent of the measurements for each month.
b. The turbidity level of representative samples of filtered
water must at no time exceed 5 NTU.
Diatomaceous earth systems, with appropriate design and operating
conditions and which meet the minimum turbidity performance criterion can
As indicated in Section 4, pilot studies have shown that with proper
nurturing of the schmutsdecke, operation at a maximum loading rate of
0.2 m/hr will provide optimum removal of Giardia cysts and viruses
(Logsdon, 1987b; Bellamy et al., 1985).
5-7
-------
be considered to be well operated and achieving at least 2-log removal of
Giardia cysts and at least 1-log removal of viruses without disinfection.
Systems which believe that they are actually achieving greater than a 2-
log Giardia cyst removal may demonstrate the actual removal achieved using
the protocol outlined in Appendix M.
5.3.4 Other Filtration Technologies
The turbidity performance criteria for filtration technologies other
than those presented above, are the same as for slow sand filtration. The
Giardia cyst removal achieved by these systems must be demonstrated to the
Primacy Agency. The protocol outlined in Appendix M may be used as a
basis for this demonstration.
Reverse osmosis is a membrane filtration method used to remove
dissolved solids from water supplies. Desalination is a typical use of
the process. Application to potable water treatment is limited to
extremely high quality raw water supplies of low turbidity (1 NTU or
less), or following pretreatment to produce a supply of low turbidity.
The membrane excludes particles larger than 0.001 to 0.0001 urn
range, thereby effectively removing bacteria, Giardia cysts and viruses.
Credit can be given for at least a 3-log Giardia cyst and 4-log virus
removal, with no demonstration. It should be noted that this removal
credit assumes the membranes are in tact with no holes in the membranes
allowing the passage of organisms.
5.4 Disinfection Monitoring Requirements
Each system must continuously monitor the disinfectant residual of
the water as it enters the distribution system and record the lowest
disinfectant residual each day. If there is a failure in the continuous
monitoring equipment, the system may substitute grab sample monitoring
every 4 hours for up to 5 working days following the equipment failure.
Systems serving 3300 people or fewer may take grab samples in lieu of
continuous monitoring at frequencies as follows:
5-8
-------
System Population Samples/Day
<500 1
501-1,000 2
1,001 - 2,500 3
2,501 - 3,300 4
The grab samples must be taken at different times during the day,
with the sampling intervals subject to Primacy Agency review and approval.
If the residual concentration falls below 0.2 mg/L, the system must take
another sample within 4-hours and notify the Primacy Agency as soon as
possible, but no later than the end of the next business day, even if the
residual is restored to 0.2 mg/L or greater within 4 hours. If the
residual is not restored to 0.2 mg/L or greater within 4 hours, the system
is in violation of a treatment technique requirement. Each system must
also measure the disinfectant residual in the distribution system at the
same frequency and locations at which total coliform measurements are made
pursuant to the requirements in the revised Total Coliform Rule (54 FR
27544; June 29, 1989). For systems which use both surface and ground
water sources, the Primacy Agency may allow substitute sampling sites
which are more representative of the treated surface water supply.
5.5 Disinfection Performance Criteria
5.5.1 Minimum Performance Criteria Required bv the SWTR
For systems which provide filtration, the disinfection requirements
of the SWTR are:
a. Disinfection must be provided to ensure that the total
treatment processes of the system (including filtration)
achieves at least a 3-log removal/inactivation of Giardia cyst
and a 4-log removal/inactivation of viruses. The Primacy
Agency must determine what level of disinfection is required
for each system to meet this criterion.
b. The system must demonstrate by continuous monitoring and
recording that a disinfectant residual in the water entering
the distribution system is never less than 0.2 mg/L for more
than 4 hours. If at any time the residual falls below 0.2
mg/L for more than 4 hours the system is in violation. The
5-9
-------
system must notify the Primacy Agency whenever the residua]
falls below 0.2 mg/L before the end of the next business day.
c. The system must demonstrate detectable disinfectant residuals
or HPC levels of 500 or fewer colonies/ml in at least 95
percent of the samples from the distribution system each month
for any two consecutive months.
5.5.2 Recommended Performance Criteria,
Disinfection must be applied to assure that the overall treatment
provided achieves at least a 3-log removal/inactivation of Giardia cyst
and a 4-log removal/inactivation of viruses. As outlined in Section 5.3,
well operated filter plants achieve at least a 2 to 2.5-log removal of
Giardia cysts and between a 1 to 2-log removal of viruses. EPA therefore
recommends that the Primacy Agencies adopt additional disinfection perfor-
mance criteria that include:
a. As a minimum, primary disinfection requirements that are
consistent with the overall treatment requirements of the
SWTR, or preferably;
b. Primary disinfection requirements as a function of raw water
quality as outlined in Section 4.4.
Recommended Minimum Disinfection
The required minimum primary disinfection is the disinfection
needed for the entire treatment process to meet the overall treatment
requirement of 3-log Giardia and 4-log virus removal/inactivation. The
following table provides a summary of the expected minimum level of
treatment performance in well operated filter systems and the recommended
level of disinfection.
Expected Recommended Disinfection
Log Removals (Log Inactivations)
Filtration Giardia Viruses Giardia Viruses
Conventional 2.5 2.0 0.5 2.0
Direct 2.0 1.0 1.0 3.0
Slow Sand 2.0 2.0 1.0 2.0
Diatomaceous
Earth 2.0 1.0 1.0 3.0
5-10
-------
In cases where the system believes that the treatment processes are
achieving greater removals than those listed above, the actual removal
provided by the processes can be demonstrated through the procedures
outlined in Appendix M. However, EPA recommends that, despite the
removals demonstrated, systems should provide a minimum of 0.5 log Giardia
cyst inactivation to supplement filtration and maintain a second treatment
barrier for microorganisms.
Recommended Disinfection as a Function of Raw Hater Quality
Although the SWTR requires the overall treatment to provide a
minimum of a 3-log Giardia cyst and a 4-log virus removal /inactivation, it
may be appropriate for the Primacy Agency to require greater removal s/-
inactivations depending on the degree of contamination in the source water
as presented in Section 4.4. Following is a summary of the recommended
overall treatment which should be provided based on an estimate of the
Giardia cyst concentration in the source water:
Allowable daily avg
cyst concentration/100 L
^geometric mean^ <1 >1-1Q >10-100
Giardia cyst Removal/Inacti vation 3-log 4-log 5-log
Virus Removal/Inactivation 4-log 5-log 6-log
If a slow sand filtration plant must achieve a 4-log removal/inacti-
vation of Giardia cysts and a 5-log removal/inactivation of viruses, and
credit for 2-log Giardia cyst and 2-log virus removal by filtration is
granted, disinfection for a 2-log Giardia cyst inactivation and 3-log
virus inactivation would be needed to meet the overall removal/inacti-
vation. However, Primacy Agencies may allow systems which use particle
size analysis outlined in Appendix M to demonstrate greater than a 2-log
Giardia cyst removal to provide less than 2-log Giardia cyst inactivation
through disinfection.
5-11
-------
5.5.3 Disinfection Bv-Product Considerations
Although the EPA suggests increased levels of disinfection for
various^source water conditions, a utility should not implement such a
change without considering the potential conflict with the requirements of
existing or future disinfection by-product regulations. EPA intends to
promulgate National Primary Drinking Water Regulations to regulate levels
of disinfectants and disinfection by-products when it promulgates
disinfection requirements for ground water systems (anticipated in 1992).
EPA is concerned that changes required in utilities' disinfection
practices to meet the recommended inactivations for the SWTR might be
inconsistent with treatment changes needed to comply with the forthcoming
regulations for disinfectants and disinfection by-products. For this
reason, EPA recommends that Primacy Agencies exercise discretion,
sensitive to potential disinfection by-product concerns, in determining
the level of disinfection needed for filtered systems to meet the overall
treatment requirements specified in the rule or recommended based on
source water quality.
Until the promulgation of the disinfection by-product regulation,
EPA recommends that the Primacy Agency allow more credit for Giardia cyst
and virus removal by filtration than otherwise recommended if a) the
Primacy Agency determines that a system is not currently at a significant
risk from microbiological contamination at the existing level of
disinfection and b) less stringent interim disinfection conditions are
necessary for the system to modify its disinfection process to optimally
achieve compliance with the SWTR as well as the forthcoming disinfection
by-product regulations. The following paragraphs outline the recommended
disinfection levels for systems meeting the above conditions.
For well-operated conventional filtration plants that meet the
minimum turbidity requirements at all times, the Primacy Agency may
consider giving the system credit for 3-log Giardia cvst removal (in lieu
of the generally recommended 2.5-log credit). Also, for well-operated
direct filtration plants, the Primacy Agency may consider giving the
system credit for 2.5-log Giardia cyst removal in lieu of the generally
5-12
-------
recommended 2.0-log credit. EPA recommends that these additional credits
be given for conventional or direct filtration only if:
a. The total treatment train achieves 1) at least 99 percent
turbidity removal, or filtered water turbidities are consis-
tently less than 0.5 NTU, whichever is lower,7 21 2) a 99.9
percent removal of particles in the size range of 5 to 15 urn
is demonstrated as outlined in Appendix M;3 and
b. The level of heterotrophic plate count (HPC) bacteria in the
finished (disinfected) water entering the distribution system
is consistently less than 10/ml.
Systems using slow sand filtration or diatomaceous earth filtration
may be given interim credit for up to 3-log Giardia cyst removal if the
system meets the recommended conditions listed above for conventional
systems. Pilot plant studies have demonstrated that these technologies,
when well operated, generally achieve at least 3.0-log removals (USEPA,
1988a).
The EPA believes that interim level of disinfection requirements may
be appropriate in some cases depending upon source water quality,
reliability of system operation and potential increased health risks from
disinfection by-products. EPA intends to regulate disinfectants and
disinfection by-products in 1992. At this time it will become apparent
how systems with disinfection by-product problems can optimally meet the
disinfection requirements of the SWTR and the disinfection by-products
regulations, concurrently.
7 For example, a system with a raw water turbidity averaging 20 NTU
maintaining a filtered water turbidity less than 0.2 NTU can be granted
3-log Giardia cyst removal credit with no further demonstration.
8 In cases where the Primacy Agency has a data base which shows a
correlation between turbidity and Giardia cysts removal, turbidity may
be used in lieu of particle size analysis. Turbidity removal require-
ments should be set to assure 99.9 percent Giardia cyst removal. A
correlation between turbidity and Giardia cyst removal was shown in a
study reported by Hendricks et al (1984).
5-13
-------
5.5.4 Recommended Disinfection System Redundancy
The SWTR does not require a redundant disinfection system for
filtered supplies. However, in order to assure the continuous provision
of disinfection to meet the overall removal/inactivation requirements and
to maintain a residual entering the distribution system, EPA recommends
that redundant disinfection equipment be provided. As contained in the
1987 edition of Ten State Standards, where disinfection is required for
protection of the supply, standby equipment is required. Automatic
switchover should be provided as needed, to assure continuous disinfectant
application.
Recommendations for providing redundant disinfection are outlined in
Section 3.2.4 and detailed in Appendix I.
5.5.5 Determination of Inactivation bv Disinfection
The desired level of inactivation can be achieved by disinfection at
any point in the treatment or distribution system prior to the first
customer. Disinfection provided prior to filtration is referred to as
pre-disinfection while disinfection after filtration is referred to as
post-disinfection. As presented in Section 3.2, the inactivation of
Giardia cysts and viruses provided by disinfection are indicated by CT
values.
The SWTR defines CT as the residual disinfectant concentration(s) in
mg/L multiplied by the contact time(s) in minutes. The contact time is
measured from the point of disinfectant application to the point of
residual measurement or between points of residual measurement. The
inactivation efficiency can be determined by calculating CT at any point
along the process after disinfectant application prior to the first
customer.
A system may determine the inactivation efficiency based on one
point of residual measurement prior to the first customer, or on a profile
of the residual concentration after the point of disinfectant application.
The residual profile is generated by monitoring the residual at several
points between the point(s) of disinfectant application and the first
customer. The system can then use the method described in Section 3.2 for
determining the total inactivation credit. Profiling the residual allows
5-14
-------
for credit of significantly higher residuals which may exist before the
water reaches the first customer. Methods for determining various
disinfectant residuals are described in Appendix 0.
In pipelines, the contact time can be assumed equivalent to the
hydraulic detention time and is calculated by dividing the internal volume
of the pipeline by the peak hourly flow rate through the pipeline. In
mixing basins and storage reservoirs, the hydraulic detention time
generally does not represent the actual disinfectant contact time because
of short circuiting. The contact time in such chambers should be
determined by tracer studies or an equivalent demonstration. The time
determined from the tracer study to be used for calculating CT is T10. T,0
represents the time that 90 percent of the water (and microorganisms
within the water) will be exposed to disinfection within the disinfectant
contact chamber. Guidance for determining detention time in contact
chambers is provided in Appendix C.
The residual disinfectant concentration should be measured daily,
during peak hourly flow, for each disinfectant section prior to the first
customer in the distribution system. Unless a system knows from
experience when peak flow will occur, a system can only identify peak
hourly flow after it has occurred. Therefore, EPA suggests that residual
measurements be taken every hour. If it is not practical to take grab
samples each hour, the system may take grab samples during the period peak
flow is expected to occur, or continuous monitors may be used. The
measurements taken during the hour of peak flow can then be used to
determine the CT for each section (CTcau). The determination of CTs is
explained in Section 3.2.1.
Although the inactivation maintained in the system is determined
during peak hourly flow, the disinfectant dosage applied to maintain this
inactivation may not be necessary under lower flow conditions. Under
lower flow conditions, a higher contact time is generally available and
the CT needed to meet the required inactivation may be met with a lower
residual concentration. Continuing to apply a disinfectant dosage based
on the peak hourly flow may provide more disinfection than is needed,
increasing costs and possibly resulting in increased levels of disinfec-
tant by-products. However, the system should also maintain the required
5-15
-------
inactivation levels at non-peak hourly flows. The system should therefore
evaluate the dose needed to provide the CT necessary for maintaining the
required inactivation under different flow conditions and set the dosage
accordingly. The following example provides guidelines for determining
flow ranges and disinfection levels to maintain the required disinfection.
Example
A 20-mgd direct filtration plant applying free chlorine as a
disinfectant has a contact time of 27.5 minutes under peak flow condi-
tions. As noted in Section 5.3, well-operated direct filtration plants
achieve 2-log Giardia cyst removal and 1-log virus removal. Therefore,
disinfection for 1-log Giardia cyst inactivation and 3-log virus
inactivation is recommended. The pH and temperature of the water are 7
and 5 C, respectively. Using Table 1-2, a CT of 55 is required to achieve
1-log Giardia cyst inactivation at a residual of 2 mg/L. This level of
treatment is more than adequate for 3-log inactivation of viruses
requiring a CT of 6, as indicated in Table E-7. However, under low flow
conditions the available contact time is longer, and lower residuals are
needed to provide the same level of inactivation. Based on the calculated
contact time under various flow rates and the CT values in Table E-2,
adequate disinfection would be provided by maintaining the following
chlorine residuals for the indicated flows:
CT90
Free Chlorine
Residual frog/L)
2.0
1.5
1.0
0.5
CT90 corresponds to a 1-log inactivation. If a different level of
inactivation were needed, CT values for that inactivation would be read
from the tables corresponding to the pH and temperature of the water.
Section 3.2.2 lists the percent inactivations corresponding to
log inactivations, i.e., 0.5-log equals 68 percent requiring
CT68.
5-16
Flow fMGD)
20
15
10
5
Contact
the (min}
27.5
36
54
108
(mg/L-min)
Required
55
52.5
50
47
-------
In cases where the residual, pH or temperature of the water is
an intermediate value not reported in the tables, linear
(straight-line) interpolation may be used. I
For example, in the above listing, 0.5 mg/L residuals are not
included in the Appendix E tables. The CT90 value was
determined by interpolating between the <0.4 mg/l value of 46
mg/L-min and the 0.6 mg/l value of 48 mg/L-min.
CT values for intermediate pH and temperature values may also
be interpolated; or
The CT values for the higher pH or lower temperature listed in
the table may be used instead of interpolation.
CT99 9 tables in the SWTR can be used to calculate the CT
required to achieve any log inactivation by:
log inactivation
CTrequired = required x CT,, .
3.0 log
The variation in CT required with respect to the residual for
chlorine makes it impractical for the utility to continually change the
disinfectant dose as the flow changes. Therefore, EPA suggests that the
flow variation at the utility be divided into ranges and the residual
needed at the higher flow of the range be maintained for all flows within '
the range to assure adequate disinfection. The following flow ranges and
residuals at the given pH and temperature are suggested for this plant:
Free Chlorine
Flow Range fMGD) Residual fmg/U
5-10 1.0
10-15 1.5
15-20 2.0
In this way, the utility is assuring the provision of the required
disinfection while minimizing the disinfectant costs and possibly lowering
disinfection by-products.
Although these residuals will meet the required CT, maintaining a
residual in the distribution system must also be considered. If there is
no other point of disinfection prior to the distribution system, the
residual for disinfection must be maintained at a level which will also
5-17
-------
provide a residual throughout the distribution system. The complete range
of flows occurring at the plant should be evaluated for determining the
required residual. The utilities may establish the residual needs for as
many flow ranges as is practical.
The Primacy Agency should make periodic checks to assure that the
utility is maintaining adequate disinfection at both peak and non-peak
flow conditions.
In contrast to this close control of disinfectant addition and CT
monitoring, for filtered systems which have long detention times and
regularly exceed the CT requirements for the inactivation needed, it may
be unnecessary for the system to calculate CTs each day of operation.
Unlike unfiltered systems where CTs must be calculated each day, for
filtered systems, monitoring the residual at the end of the contact time
may be sufficient to indicate that the required disinfection is provided.
However, this results in much higher CTs in the summer than is needed,
which adds to.costs and possibly unnecessary increased production of
disinfection by-products. The following example outlines one scenario for
which this would apply.
Example
A utility disinfects with chlorine ahead of a reservoir prior to
direct filtration. The Primacy Agency may give a well-operated direct
filtration plant credit for 2-log Giardia cyst removal and 1-log virus
removal. Therefore, 1-log Giardia cyst and 3-log virus inactivation
through disinfection is needed. For free chlorine, the CTs for 1-log
Giardia cyst inactivation exceed the CTs for 3-log virus inactivation.
Therefore, CTs for Giardia cyst inactivation are the controlling CTs. The
following water quality conditions occur in the reservoir during the year:
pH 7 - 7.5
Temperature (° C) 5-20
Chlorine residual (mg/L) 0.2 - 0.8
The required CT for chlorine increases with:
increasing residual,
increasing pH, and
decreasing temperature
5-18
-------
Thus, for a residual of 0.8 mg/L the CT needed for a 1-log Giardia
cyst inactivation is as follows: I
fiH Temperature (C\ mg/L-min
7.5 5 58 (Table E-2)
7 20 18 (Table E-5)
Tracer studies conducted on the reservoir indicated a T10 of 150
minutes at the system's maximum flow. For the maximum CT of 58 mg/L-min
required, the minimum residual needed to'meet this requirement is 0.4
mg/L, calculated as:
58 mo/L-min = 0.4 mg/L
150 min
At a residual of 0.4 mg/L, CT90 is 55 mg/L-min. Thus, any residual >0.4
mg/L will provide the needed disinfection throughout the year and the
Primacy Agency may require the system to report only the residual
maintained, reducing the effort needed to determine effective disinfec-
tion. Maintaining this residual in the summer, however, provides much i
higher CTs than needed, possibly resulting in unnecessary costs and
increased disinfection by-products.
Meeting the Recommended Inactivation Using Free Chlorine
As previously indicated in Section 3.2.1, the effectiveness of free
chlorine as a disinfectant is influenced by both the temperature and pH of
the water and by the concentration of chlorine. The inactivation of
Giardia cysts by free chlorine at various temperatures and pHs are
presented in Appendix E (Table E-l through Table E-6). The CT values for
the inactivation of viruses by free chlorine are presented in Table E-7.
To determine whether a system is meeting these inactivations, the
free chlorine residual, pH and temperature must be measured, at one point
or several points prior to the first customer, where contact time(s) is
measured. The contact time should be determined from the point of
application of the disinfectant to the point(s) where the residual is
measured for determining CTs prior to the first customer. The CTs
5-19
-------
actually achieved in the system should then be compared to the values in
the table for the pH and temperature of the water at the point(s) of
residual measurement. Guidance on calculating the CT for chlorine is
presented in Section 3.2.1.
Meeting the Recommended Inactivation Using Chlorine Dioxide
CT values for the inactivation of Giardia cysts by chlorine dioxide
are presented in Table E-8 and the CT values for the inactivation of
viruses are presented in Table E-9. As shown in Tables E-8 and E-9, the
only parameter affecting the CT requirements for chlorine dioxide is
temperature. However, the disinfection efficiency of chlorine dioxide may
be significantly increased at higher pHs. Since the CT values in Tables
E-8 and E-9 were based on data at pH 7 and 6, respectively, and chlorine
dioxide appears to be more effective at higher pHs, systems with high pHs
may wish to demonstrate that CT values lower than those presented in
Tables E-8 and E-9 may achieve the desired level of inactivation.
Chlorine dioxide residuals are short-lived. Therefore, sampling and
residual analysis at various points in the treatment process downstream of
the point of application may be necessary to establish the last point at
which a residual is present. Subsequent sampling and residual analyses
conducted upstream of this point can be used to determine the CT credit by
using the demonstrated detention time between the point of application and
the sampling location. Methods for calculating CT values are presented in
Section 3.2. Systems using chlorine dioxide may conduct pilot studies to
demonstrate effective disinfection in lieu of calculating CT, or for
determining that lower CT values than those in Appendix E are appropriate.
Guidelines for conducting these studies are presented in Appendix G.
Meeting the Recommended Inactivation using Ozone
CT values for the inactivation of Giardia cysts by ozone are
presented in Table E-10 for various temperatures and inactivation rates.
As indicated in this table, the CTs required for inactivation with ozone
are substantially lower than those required for free chlorine. This
reflects the fact that ozone is a more powerful disinfectant. The CT
requirements for inactivation of viruses using ozone are presented in
Table E-ll. In cases where only a 1-log or lower Giardia cyst inactiva-
tion is needed, the CT values for virus inactivation may be higher than
5-20
-------
the CTs for Giardia cysts. Because of the reactivity of ozone, it is
unlikely that a residual will exist for more than a few minutes. As a
result, the application of a persistent disinfectant such as chlorine or
chloramines is needed to maintain the required disinfectant residual in
the distribution system. Guidance for calculating CT values for ozone are
presented in Section 3.2.1 and Appendix 0. In lieu of calculating the CT
for an ozone contactor or demonstrating that lower CTs are effective, the
disinfection efficiency can be demonstrated through pilot studies as
presented in Appendix G.
Meeting the Recommended Inactivation Requirements using Chloramines
CT values for the inactivation of Giardia cysts by chloramines are
presented in Table E-12. The high CT values associated with the use of
chloramines may be unachievable for some systems. In these cases,
chlorine, ozone, or chlorine dioxide should be used for primary disinfec-
tion, and chloramines for residual disinfection, as necessary. Table E-13
presents CT values for the inactivation of viruses with chloramines. This
table is only applicable for indicating virus inactivation efficiencies if
chlorine is added prior to ammonia. Systems which add ammonia prior to
chlorine or ammonia and chlorine concurrently, can determine viral
inactivation efficiencies using the protocol given in Appendix G. For
systems applying chloramines to meet the virus inactivation requirements,
EPA recommends that they also monitor for HPC in the finished water, as
presented in Section 5.6. Systems also may demonstrate effective
disinfection with chloramines in lieu of calculating CT, or to determine
that lower CT values than those indicated in Appendix E are appropriate.
The protocols outlined in Appendix G can be used for this demonstration.
Further guidance on chloramines is given in Section 3.2.1.
Meeting the Inactivation Requirement
Usiny Ultraviolet (UV^ Radiation
Ultraviolet radiation is a method of disinfection which can be
applied to meet the virus inactivation requirements of the SWTR.
UV disinfectant dose, expressed in terms of UV intensity and
exposure time/unit area (mW-sec/cm2) incorporates the elements of the CT
concept and therefore can be considered as analogous or equivalent to a CT
value. UV disinfection usually employs commercially available units
5-21
-------
designed to deliver doses of 25 to 35 mW-sec/cm2. The dose can be
increased by reducing water flow rate and/or by adding additional units in
series. UV disinfection efficiency differs from that of chemical
disinfectants in that it is not affected by water temperature. UV
radiation does not effectively penetrate solids and is absorbed by certain
dissolved substances. Therefore, turbidity and other water quality
factors are important determinants of UV disinfection efficiency, and UV
should be applied after turbidity removal.
CT values for the inactivation of Giardia cysts by UV are not
included in Appendix'E. The results of two studies (Rice and Hoff, 1981;
Carlson e_t il, 1985) indicate that Giardia cysts are extremely resistant
to inactivation by UV with doses greater than 60 mW-sec/cm2 causing less
than 80% inactivation. Because UV appears to be very ineffective for
Giardia cyst inactivation and in the absence of sufficient data showing
the doses needed to inactivate 0.5 to 3.0 logs of cysts, UV must be used
in combination with other disinfectants to provide evidence of effective
cyst inactivation.
CT values for the inactivation of viruses by UV are presented in
Table £-14. Units used for UV disinfection should be equipped with fail-
safe devices that will provide automatic shutdown of water flow if UV dose
decreases to levels lower than those specified in Table E-14.
Meeting the Inactivation Requirement Using Alternate Disinfectants
For system using disinfectants other than chlorine, chloramines,
chlorine dioxide, or ozone, the effectiveness of the disinfectant can be
demonstrated using the protocol contained in Appendix G. The protocol in
Appendix G.3 for batch testing should be followed for any disinfectant
which can be prepared in an aqueous solution and will be stable throughout
the testing. For disinfectants which are not stable, the pilot study
protocol outlined in Appendix G.4 should be followed.
Examples for Determining the Disinfection to be Providej)
1) Recommended 0.5-1og Giardia. 2-1og Virus Inactivation
A community of 70,000 uses a river as its drinking water source.
Ozonation prior to a conventional treatment plant is used to treat the
5-22
-------
water. The source has a protected watershed with limited human activity
and no sewage discharge. The river water has the following water quality
characteristics:
Turbidity 10 - 200 NTU
Total estimated Giardia cyst level <1/100 /I
pH 7.0 - 7.5
Temperature 5-15
The treatment plant has a design capacity of 15 mgd and treats an
average flow of 10 mgd. A three chamber ozone contactor precedes the
rapid mix. Alum and polymer are added as a coagulant and coagulant aid,
respectively. The finished water turbidity at the plant is maintained
within the range of 0.1 to 0.2 NTU. Chloramines are applied after the
filters, but prior to the clearwells, to maintain a residual entering and
throughout the distribution system.
Based on the raw water quality and source water protection, an
overall 3-log Giardia cyst and 4-log virus removal/inactivation is
appropriate for this water source. However, as noted in Section 5.3,
Primacy Agencies may credit well operated conventional filtration plants
with 2.5-log Giardia cyst removal and 2-log virus removal. Therefore,
disinfection for 0.5-1og Giardia cysts and 2-log viruses is recommended to
meet the overall treatment requirements of the SWTR.
On the day of this example calculation, the peak hourly flow rate of
the plant was 13 mgd. The contact time of the ozone basin, T10 determined
from tracer study data is 6 minutes for this flow. The water had a pH of
7 and a temperature of 5 C on the day of the calculation. For ozone under
these conditions of pH and temperature, the following CTs are needed for
the required inactivation (Tables E-10, E-ll):
0.5-log Giardia 2-log virus
CT 0.3 0.6
The CT values indicate that viruses are the controlling parameter for
disinfection and the overall inactivation provided will be calculated
based on viruses. The overall virus inactivation provided by the ozone
contactor is determined as follows:
Average
5-23
-------
Residual ' T,0 CT..lie CT99 9 CT.llc/CT99 9
Chamber C (mg/L) (minutes) fm'g/U (mg/L-min.l
1 0.1 2 0.2 0.9 0.22
2 0.2 2 0.4 0.9 0.44
3 0.2 2 0.4 0.9 0.44
The sum of CTcalc/CT9g 9 is 1.1. This corresponds to more than a 3-1og virus
inactivation determined as 3 X CTC1|C/CT39 9 » 3 X 1.1 « 3.3-log. Therefore,
the system exceeds the recommended inactivation.
2) Recommended l-1og Giardia Cvst. 2-log Virus Inactivation
A 2 MGD slow sand filtration plant treating reservoir water, fed by
mountain streams with no nearby wastewater discharges, provides drinking
water for a community of 8,000 people. The water quality at the intake
has the following water quality characteristics:
Turbidity 5-10 NTU
Total coliforms Not measured
Total estimated Giardia cyst level <1/100 L
pH 6.5 - 7.5
Temperature 5 - 15 C
The filtered water turbidity ranges from 0.6 - 0.8 NTU. Considering
the source water quality and plant performance, an overall 3-log Giardia
cyst and 4-log virus removal/inactivation is considered sufficient for
this system. As noted in Section 5.3, the Primacy Agency may credit slow
sand plants with 2-log Giardia cyst and 2-log virus removal. Therefore
disinfection for 1-log Giardia cyst and 2-log virus inactivation is
recommended for the system to meet the overall treatment requirements.
Chlorine is added prior to the clearwells to provide disinfection.
The clearwells have a capacity of 80,000 gallons. A one mile, 16-inch
transmission main transports the water from the treatment plant to the
first customer. The inactivation provided is determined daily for the
peak hourly flow conditions." Tracer studies have been conducted to
determine the T10 for the clearwells for different flow rates. For the
purposes of calculating the inactivation the system is divided into two sections.
Section 1 - clearwell
5-24
-------
Section 2 - transmission main
The flowrate at peak hourly flow from the clearwell was 1.5 mgd on
the day of this example. At this flowrate, the T,0 of the clearwell is 67
minutes, as determined from the results of the tracer studies. At this
flowrate, water travels through the transmission main at 99 ft/min. The
data for the calculation of the inactivation is as follows:
Section 1 Section 2
length of pipe (ft) 0 5280
contact time (min)
pipe 0 53
basin 67 0
total 67 53
disinfectant chlorine chlorine
residual (mg/L) 1.0 0.6
temperature C 55
pH 7.5 7.5
For free chlorine, a 1-log Giardia cyst inactivation provides greater than
a 4-log virus inactivation; therefore, Giardia cyst inactivation is the
controlling parameter, and the inactivation provided is determined based
on Giardia cysts. The calculation is as follows:
Section 1 _ - Chlorine
CT,m = 1.0 mg/L x 67 minutes * 67 mg/L-min
From Table E-2, at a temperature of 5 C and a pH of 7.5, CT55 9 is
179 mg/L-min
CTcau/CT99 a B 67 mg/L-min * 0.37
w 179 mg/L-min
Section 2 - Chlorine
CTC|U = 0.6 mg/L x 53 minutes = 32 mg/L-min
From Table E-2, at a temperature of 5 C and a pH of 7.5, CT39 9 is
171 mg/L-min
CTC1U/CT99 9 » 32 mg/L-min.' 0.19
171 mg/L-min
The sum of CTcllc/CT99 9 is equal to 0.56. This is equivalent to a 1.7-log
Giardia cyst inactivation determined as 3-log x CTcllc/CT99 9 « 3 x 0.56 =
5-25
-------
1.7-logs. Therefore, the system exceeds the disinfection recommended to
meet the overall treatment requirements.
3) Recommended 2-1og Giardia Cvst. 4-log Virus Inactivation
A community of 30,000 people uses a reservoir treated by direct
filtration «for its water supply. The reservoir is fed by a river which
receives the discharge from a wastewater treatment plant 10 miles upstream
of the reservoir. The reservoir water quality is as follows:
Turbidity 5 - 15 NTU
Total coliforms 100 - 1000/100 ml
Total estimated Giardia cyst level 5/100 L
pH 6-7
Temperature 5 - 15 C
Based on the source water quality, an overall removal/inactivation
of 4-log Giardia cyst and 5-log virus is recommended as outlined in
Section 4.4.
The source water flows by gravity to a 3 MG storage reservoir prior
to pumping to the water treatment plant. Chloramines are produced by
first adding chlorine then ammonia to the water within the inlet of the
storage reservoir. Chlorine dioxide is added to the filtered water prior
to the clear-wells. Chloramines are applied after the clearwells to
maintain a residual in the distribution system. The system design flow is
8 mgd with an average flow of 5 mgd. For the calculation of the overall
inactivation, the system is divided into 2 sections.
Section 1 - the storage reservoir and the transmission to the
treatment plant
Section 2 - the clearwells
The overall inactivation for the system is computed daily at the
peak hourly flow conditions. The pH, temperature, and disinfectant
residual is measured at the end of each section prior to the next point of
disinfectant application and the first customer. The flow is measured in
the transmission main entering the plant and exiting the clearwells. On
the day of this example calculation, the peak hourly flow was 6 mgd in the
transmission mains entering and leaving the plant. If the flowrates were
different, the T,0 corresponding to the respective flowrate would be used
5-26
-------
in the calculation. Guidance for determining CTs when flowrates vary
within a system is given in Section 3.2. The water velocity through the
20-inch transmission main is 256 ft/min at a flow of 6 mgd. Tracer
studies were conducted on the storage reservoir and clearwells. As
determined from the testing the detention times, T,0, of* the basfns at a
flow of 6 mgd are 380 and 130 minutes for the storage reservoir and
clear-wells, respectively. The data for the calculation of inactivation is
as follows:
Section 1 Section 2
length of pipe (ft) • 4500 0
contact time (min)
pipe 18 0
basin 380 130
total 398 130
disinfectant chloramines chlorine dioxide
residual (mg/L) 1.5 0.2
temperature C 5 5
pH 7 7
For each of the disinfectants used, the following CTs are needed for
2-log Giardia and 4-log virus inactivation for the pH and temperature
conditions of the system.
CT for 2-log CT for 4-log
Giardia Virus
chloramines 1430 1988
chlorine dioxide 17 33.4
The CT required for the virus inactivation is higher than that
needed for Giardia inactivation for each of the disinfectants. Since the
viruses are the controlling parameter, the inactivation calculation will
be based on the viruses. The calculation is as follows:
Section 1 - Chloramines
CTcllc » 1.5 mg/L x 398 minutes « 597 mg/L-min
From Table E-13, at a temperature of 5 C and a pH of 7, CT99 99 is
1988 mg/L-min
CTc.ic/CT99 99 • 597 mq/L-min » 0.3
5-27
-------
1988 mg/L-min
Section 2 - Chlorine Dioxide
CT.alc = 0.2 mg/L x 130 minutes = 26 mg/L-min
From Table E-9, at a temperature of 5 C and a pH of 7, CT39 99 is
33.4 mg/L-min
CT:au/CT99 99 s 26 mp/L-min = 0.78
33.4 mg/L-min
The sum of CTC4U/CT9999 is equal to 1.08, which is equivalent to a 4.3-log
inactivation of viruses, determined as follows:
x = 4-log x crca1c = 4 x 1.08 = 4.3-1ogs
CTg9 99
Therefore, the system provides sufficient disinfection to meet the overall
recommended treatment performance.
5.6 Other Considerations
Monitoring for heterotrophic plate count (HPC) bacteria is not
required under the SWTR. However, such monitoring may provide a good
operational tool for:
Measuring microbial breakthrough
Evaluating process modifications
Detecting loss of water main integrity
Detecting bacterial regrowth conditions within the distribu-
tion system
Determining interference with the coliform measurements (AWWA,
1987)
Therefore, EPA recommends routine monitoring for HPC in the plant
effluent and within the distribution system whenever the analytical
capability is available in-house or nearby. Systems which do not have
this capability should consider using a semi-quantitative bacterial water
sampler kit, although this is not acceptable for compliance monitoring.
5-28
-------
As discussed in the preamo,e to the SWTR, EPA believes that it is
inappropriate to include HPC as a treatment performance criterion in the
rule since small systems would not have in-house analytical capability to
conduct the measurement, and they would need to send the samples to a
private laboratory. Unless the analysis is conducted rapidly, HPC may
multiply and the results may not be representative.
EPA recommends an HPC level of less than 10/ml in the finished water
entering the distribution system and levels of less than 500/ml throughout
the distribution system.
Legionella is another organism which is not included as a treatment
performance criterion. Inactivation information on Legionella is limited.
EPA believes that treatment which complies with the SWTR will remove
and/or inactivate substantial levels of Legionella which might occur in
source waters, thereby reducing chances that Legionella will be trans-
ported through the system and reducing the possibility that growth might
occur in the distribution system or hot water systems within homes and
institutions. Since Legionella are similar in size to coliform organisms,
removals by filtration should be similar to those reported for total
coliforms. In addition, the available disinfection information indicates
that the CT requirements for inactivation of Legionella are lower than
those required for the inactivation of Giardia cysts. EPA recognizes,
that regardless of the treatment provided, some Legionella may enter
plumbing and air conditioning systems and subsequently multiply (Muraca et
al., 1986). EPA believes that these concerns are best addressed through
guidance contained in Appendix B.
5-29
-------
6. REPORTING
6.1 Reporting Requirements for Public Water Systems
Not Providing Filtration
The SWTR requires unfiltered systems to prepare monthly reports for
the Primacy Agency to determine compliance with the requirements for:
source water fecal and/or total coliform levels
source water turbidity levels
disinfection level
disinfectant residual entering the distribution system
disinfectant residuals throughout the distribution system.
The monthly reports must be prepared and submitted to the Primacy
Agency within 10 days after the end of the month. The utility must
maintain a daily or monthly data log used to prepare the monthly reports.
Tables 6-1 through 6-5 are examples of daily data sheets which the
utilities may find useful for logging the data needed to prepare reports
for the Primacy Agency.
Table 6-6 presents a concise format which can be used by the system
for the monthly reports to the Primacy Agency. Tables 6-3 and 6-4 must
also be submitted with the monthly report. After the initial 12 months
of reporting, the Primacy Agency may remove the requirement for reporting
the information contained in Table 6-3 if it is satisfied that the system
is computing compliance with the CT requirements correctly. The
individual sample results summarized in the monthly reports should be kept
on file at the utility for a minimum of 5 years.
In addition to the monthly reporting requirements for source water
quality conditions and disinfection information, systems with unfiltered
supplies are also required to submit annual reports for the watershed
control program and the on-site inspection, within 10 days after the end
of the federal fiscal year.
The Primacy Agency will review the reports to determine whether the
system is in compliance. A possible report format for the watershed
control program is:
6-1
-------
1. Summarize all activities in the watershed(s) for the previous
year.
2. Identify activities or situations of actual and potential
concern in the watershed(s).
3. Describe how the utility is proceeding to address activities
creating potential health concerns.
EPA recommends that the Primacy Agency submits the annual watershed
reports to the State Water Quality Managers. The reports will be useful
in updating statewide assessments and management programs.
The SWTR requires each system to provide the Primacy Agency with a
report of the on-site inspection unless the inspection is conducted by the
Primacy Agency. EPA suggests that:
1. A report of the inspection containing the findings, suggested
improvements and dates by which to complete improvements is
to be prepared following the initial system review. When and
how system has resolved problems identified in the previous
report should also be included.
2. To lessen the burden on utilities, a report containing results
of the general survey should be submitted in subsequent years.
In addition to these reporting requirements, the SWTR requires that
the reporting requirements of the Total Trihalomethane Regulation and the
Coliform Rule also be met.
Records of waterborne disease outbreaks also must be maintained.
In the event &f a waterborne disease outbreak, as defined in part 141.2
of the SWTR, the Primacy Agency must be notified by the end of the next
business day.
The report of the outbreak should contain:
1. Date of occurrence
2. Type of illness
3. Number of cases
4. System conditions at the time of the outbreak, including
disinfectant residuals, pH, temperature, turbidity, and
bacteriological results.
The records of an outbreak should be maintained permanently or until
filtration is installed.
6-2
-------
6.2 Reporting Requirements for Public Water Systems Using Filtration
The SWTR requires filtered water systems to submit monthly reports
to the Primacy Agency for determination of compliance with the require-
ments for:
treated water turbidity
disinfectant residual entering the distribution system
disinfectant residuals throughout the distribution system
Tables 6-7 and 6-8 present a format which the utility can use as a daily
data log and to submit monthly reports to the Primacy Agency.
Recommended Reporting Not Required by the SWTR
The Primacy Agency may also want filtered water systems to report
some information associated with recommendations made in this manual which
are not requirements of the SWTR. EPA recommends that filtered water
systems:
1. Report the log inactivation of Giardia cysts and viruses,
required by the Primacy Agency.
2. Report point of application for all disinfectants used.
3. Report the daily CT(s) used to calculate the log inactivation
of Giardia cysts and viruses.
4. If more than one disinfectant is used, report the CT(s) and
inactivation(s) achieved for each disinfectant and the total
percent inactivation achieved.
5. Note any difference between the measured CT(s) and the CT
required to meet the overall minimum treatment performance
requirement specified by the Primacy Agency.
Tables 6-3 and 6-4 can be used to maintain the records necessary for
numbers 2 through 5.
This information can be used to determine the disinfection level
maintained by the system to assure that the overall removal/inactivation
required is maintained.
6-3
-------
The Primacy Agency may make provisions to minimize the reporting
requirements for systems with reservoirs, large amounts of storage or long
transmission mains which provide a long disinfectant contact time. Since
these systems typically provide inactivation in excess of that needed, the
Primacy Agency may require the system only to report the minimum daily
residual at the end of the disinfectant contact time. The CT maintained
can then be estimated based on this residual and the contact time under
the system design flow. This method of CT determination will eliminate
the need for the system to determine the contact time under maximum flow
conditions each day.
6-4
-------
Month .
Year
TABLE 6-1
1
SOURCE WATER QUALITY CONDITIONS FOR UNFILTERED SYSTEMS
(For system use only)
System/Treatment Plant.
PWSID
TjrbiO:r\ \Icjs.r;
3
Coliform Measurements
Maximum
1 i
Date '
No.
Fecal
of Samples
1 :
-i i
Total
No
Fecal( < =
ol' Samples Meeting Specified
20/100 mL)
Total ( < =
Limits
100/100 mL)
Turbidity E..
iNTUI ••.. -
i
10
12
13
15
25
• 23
29
30
31
1
Totals '
1
1
i
Maximum daily :-:?., • =
Total number of :-:- - - - -
Notes-
Samples are taken from the source water immediately prior to the first disinfection point included in the CT deter-
As specified in 40 CFR 141 74(b)(l), a fecal or total coliform sample must be taken on each day that the
system operates and a source water turbidity measurement exceeds 1 NTU.
For each day that the maximum turbidity exceeds 5 NTU. the date should also be entered for the day that the Sta:j
of this exceedance. e.g., "7 3-22 Apr"
A "yes" response is required each day the maximum turbidiry exceeds 5 NTU and the previous day did not Th.s
of the beginning of a rurbidity "event" The :otaJ nunber of ">es" responses equals the number of turbidity cse--
-------
TABLE 6-2
1.ONG-TERM SOURCE WATER QUALITY CONDITIONS FOR
UNFILTERED SYSTEMS
(For system use only)
Year
Systern/Treaiment Plant
PWSID
Coliform Measurements
i No of Samples
Month Fecal Total
January i
February
j
March
April ;
May
No oi Samples Meeting Specified Limits
FecaiK = 20.100 nL)
Total (< = 100/100 mL)
i
i
Turbiditv Measjrer.^r.rs
Days with N^i-rcr .,
Turbidity T-rru.^
>5STU E.^:.:,
i
i
June <
-•-:,
August
i
Ssptenoer
Octooer
NoemDer
December :
-.
1
Total
-------
TABLE 6-3
CT DETERMINATI
Month
YSMT
ON FOR UNFILTERED SYSTEMS -- MC
System/T realm
PWSID
Disinfectant/Sequence of Application
i
! Disinfectant
j Concentration.
Date ' C (mg/L)
1
2 i
3
4
5
6
i
S
3 ,
Disinfectant ! 4
Contact Time. i CTcalc
T (min.) ! (=CxT)
[
1
•
1
!
1.2
>NTHLY REPORT TO PRIMACY AGENCY
ent Plant
3.5
PH
1
9 : i
10
3 i
Water ;
Temp 6
(deg C) CT99 9 ' 'CTcalc CT99 9'.
I
1
i |
1
;
! i
a i
12
13
• 14
is ;
16
1
i
j
1
|
]
17
IS , ;
19
20 i
2!
I
_.
1 «
i I
24 ' I
25
26
2?
25
29
' 30
31
! i
— i i
' i !
i i
'
i
1
Prepared by
Date
^'otes
1 To be included in the monthly report for it least 12 months after the initiation of reporting. After that time, the Pr-.rar.
rr.ay no longer require thi* form.
2 Use a separate form for each disinfectant/sampling site Enter disinfectant and sequence position, e.g., "ozone/ i»t ^r
3 Measurement taken at peak hourly flow.
4 CTcalc = C (mg/L) x T (min ).
5 OnJv required if the disinfectant is free chlorine.
6 FromTables 1 1 - 1 6. 2 Land 3 1.40CFR 141 •4,t,i3)
-------
TABLE 6-4
DISINFECTION INFORMATION
FOR UNFILTERED SYSTEMS -- MONTHLY REPORT TO PRIMACY AGENCY
Month
Year
1
Mimnun Dismtwt-ant Residual
at Poim-ot'-Entry to
Date Distribution System !mg/L)
1
2
3
a
5
6
Sysiemn'rcatraent PI
PWSID
ant
'CTcaic/CT99 9) (from Table 6-3)
Disinfectant
1st
2nd
3rd
Sequence
4th
5th
i i i
3
6th
1
2 SUM (CTcalc CT99 9' <1
SUM (CTcalc/CT99 9) (YesorN'oi
I !
1
1 1
9 ' : <
:o
1 1
1
> : i i
:• ; ! i
1 i i \
:5
.1 ' !
-
.i :
,y
i
1^ i ! ! !
- • I !
2 - ' ' i
_3
""^ i'ii j
_, ^ : i
15
:*
2 5
19
30
3 i
i
i |
Prepared by
Due
Votes
1 If iess than 0 2 mg/L. the lowest level and duration of the period must be reported, e.g., "0.1-3 hrs.".
2 To determine SUM (CTc4lc/CT99 9), add (CTcalc/CT99.9) values from the firs disinfectant sequence to the last
3 If SUM .'CTcaic/CT99 9) < 1, a treatment technique violation has occurred, and a "yes" response must be entered
-------
TABLE 6-5
DISTRIBUTION SYSTEM DISINFECTANT RESIDUAL DATA FOR UNFILTERED AND FILTERED SYSTEMS
MONTHLY REPORT TO PRIMACY AGENCY
Month
Year
Date No of Sites Where
Disinfectant Residual
. *as Measured (=4)
;
T ,
1
4
5
o
-
3
9
"-,
i •>
! ;
4
( >
t *1
-
:s
i -i
; y
2C
_ ,
--
_ _*
I-
^ >
26
** "*
:s
29 '
• 30
• 31
Total <3 =
Syjtem/Treatraent Pltnt _
pwsro
No. of Sites Where no - No. of Sites Where
Disinfectant Residual Disinfectant Residual
Measured, but HPC ! Not Detected, no HPC
Measured (=b) | Measured (»c)
i
i
i
'
^
i
i
I
1
I
1
i
1
!
b= 'C =
No of Sites Where
Disinfectant Residual
Not Defected.
HPC > 500/ml( = d)
d =
No of Sires Where
Disinfectant Residual
Not Measured,
HPC > 500nU=e>
i
e =
100
Prepared by.
Date .
-------
TABLE
Month .
Year_
MONTHLY REPORT TO PRIMACY AGENCY FOR
COMPLIANCE DETERMINATION -- UN FILTERED SYSTEMS
System/Treatment Plant .
pwsro
Source Water QuaJiry Conditions
\ Cumulative number of months for which results are reported
For source water cohforn monitoring ;N'o of months)
For turbidity monitoring .
I
Coliform Criteria
Previous 6 months':
Percentage of samples < •
Percentage of samples < '
Is F < 90% '. Yes:
(No of monthsi
No of Samples
No of Samples Meeting Specified Limits
Fecal
TotaJ
FecaJ (<» 20/100 mL)
Total «= IOC ,JGnL,
z =
: 20/100 mL fecal coliforms. F = yw x 100
• 100/100 mL total coliforms. T = z/x x 100
No: N/A is T < 90% ?• Yes: _
.No:
.N/A:
Turbidity Criteria
Maximum turbidity level for reporting (current) month = NTU
Enter the month 120 months prior to the reporting month or January 1991 (whichever is later)_
Dates of 5 NTU Exceedances Since Latest Month Recorded Above
| Beginning Dace
I
j
i
Duration (days)
Date Reported
Disinfection Criteria
•\ Point-ol'-Emrv Minimum Disinfectant Residual Criteria
Davs the Residual was <0 2 mg/L
Dav
Duration of Low Level (hrs
Date Reponed
to Primacy Agency
B Distribution S>stem DisinfectanrReiidual Criteria
The value of a, b. c, d, and t from Table 6-5, as specified in 40 CFR 141 75 (b)(2)(in)(A)-(E)
a = b = , c = , d = , e =
a + b
For previous month, V = _____ %
C Disinfection Requirement Criteria
< Record the date and value of SUM (CTcak/CT99.9) for any SUM (CTcalc/CT99.9) < 1 (from Table 6-4):
If none, enter "none"
; | Date ! SUM (CTcalc/CT99 9)|
Prepared by .
Date .
Notes
The current 6-month cumulauves are required to determine whether compliance with the eohform :r
has been achieved. These totals are calculated from the previoui 6-month cumulative*, the current
month's, and totals from the earliest of 5 previous months.
-------
TABLE 6-7
DAILY DATA SHEET FOR FILTERED SYSTEMS
(For iy*era u»e only)
Month .
Year
System/Treatment Plant.
Filtration Technology
PWSID
1
Minimum Disinfectant Residual
at Pomt-of-Entry to
Date1 Distribution System (rng'L)
I
i
3
1
Maximum Filtered Water Turbidity
Filter
*
.i |
5
Combined Filter
Effluent
Clearweil
Effluent
Plant
Effluent
3 : 4 5
1 No of Turbidity No ot' Turbidity
No of Turbidity Measurements < =i Measurements
Measurements ! Specified Limit ' > 5 N'T I'
1 !
|
|
!
i
i
!
0 I
"* ! '
3 ;
5 i • ;
:o
i ;
i: 1
13
U
;i i
A \
' J
i i
t
t
!
1
j
Totals-
For multiple disinfectants, this column mu« only be completed for the la« disinfectant added prior to entering the distribution
svstem If less than 0 2 mg/L, the duration of the period must be reported, e.g., "0.1-3 hn".
For systems using conventional treatment, direct filtration, or technologies other than slow sand or diatomaceous earht filtration.
turbidity measurements may be taken at the combined filter effluent, clearwell effluent, or plant effluent prior to entry into the
distribution system. The turbidity may also be measured for each individual filter with a separate sheet maintained for each
For continuous monitors count each 4-hour period as 1 sample
Depending on the filtration technology employed, the number of turbidity samples meeting the following levels must be recorded
conventional treatment or direct filtration-0 5 NTU. slow sand filtration-1 NTU, diatomaceous earth filtration-1 NTU The S'ste na>
specify dlternate performance levels for conventional treatment or direct filtration, not exceeding 1 NTU, and slow sand :"J:ra:..:n.
not enceedmg 5 NTU, in which ease the number of turbidity measurements meeting these levels must be recorded
IP reco'dirg the number of turbidity measurements exceeding 5 NTU. the turbidity values should also be recorded, eg ' 5 3 5 2. 3
-------
T \BLE6-8
Month .
Year
MONTHLY REPORT TO PRIMACY AGENCY FOR
COMPLIANCE DETERMINATION - FILTERED SYSTEMS
System/Treatment Plant
Type of Filtration
Turbidity Limit
PWSID
Turbiditv Performance Criteria
A Total number of filtered water turbiditv measurements =
B Total number of Tillered water turbidity measurements that are less than or equal to the specified limits
for the filtration technology employed « ___^_
C The percentage of turbidity measurements meeting the specified limits = B/A x 100 = _ / _
100
D Record the date and turbidity value for any measurements exceeding 5 NTU' If none, enter "none"
! Date i Turbiditv. NTU
| i
i , 1
1
! '
DjsiRt'ecnon Performance Criteria
\ Pomt-of-Entry Minimum Disinfectant Residual Criteria
: Minimum Disinfectant Residual \ Minimum Disinfectant Residual
'at Pomt-of-Entry ! |at Pomt-of-Entry |
Date to Distribution System (mg/L) • Date 'to Distribution System f rng/L)
i
Minimum Disinfectant Res:;,.*!
iat Pomt-of-Entry
Date to Distribution Svsten — s L
1 ,11 ' ' 21 :
2 ' 12
3 13 j
4 14
5 ; 15
6 16
; 17
S .. 18
9 19 ,
10 20 •
22
23 i
24 i
25
26 ,'
L 27 '
28 '
29 i
30 i
! 3l
Days the Residual *as <0 2 ng L
Day
Duration of Low Level (hrs ) Date Reported to Primacy Agency
1
i
B Distribution System Disinfectant Residual Criteria
! The value of a, b, c. d, and e from Table 6-5, as specified in 40 CFR 141 75 (b)(2)(ui)(a)-(e):
, b
c =
*100
d =
_, e =
a - b
For previous month, V
Prepared by,
Date .
-------
7. COMPLIANCE
7.1 Introduction
This section provides guidance on when and how the requirements of
the SWTR will go into effect, including determinations made by Primacy
Agencies.
7.2 SYSTEMS USING A SURFACE WATER SOURCE (NOT GROUND WATER
UNDER THE DIRECT INFLUENCE OF SURFACE WATERS
The SDWA requires, within 18 months following the promulgation of
a rule, that Primacy Agencies promulgate any regulations necessary to
implement that rule. Under S1413, these rules must be at least as
stringent as those required by EPA. Thus, Primacy Agencies must
promulgate regulations which are at least as stringent as the SWTR by
December 30, 1990. By December 30, 1991, each Primacy Agency must
determine which systems will be required to filter. If filtration is
required, it must be installed within 18 months following the determina-
tion or by June 29, 1993, whichever is later. In cases where it is not
feasible for a system to install filtration in this time period, the
Primacy Agency ~ay allow an exemption to extend the time period (see
Section 9).
If a Primacy Agency fails to comply with this schedule for adopting
the criteria and applying them to determine who must filter, systems must
comply with the "objective" or self-implementing criteria (i.e., the
requirements that are clear on the face of the rule and do not require the
exercise of Primacy Agency discretion). Unfiltered supplies must comply
beginning December 30, 1991 and filtered supplies beginning June 29, 1993.
Monitoring requirements for unfiltered systems must be met beginning
December 30, 1990 unless the Primacy Agency has already determined that
filtration is necessary. This coincides with the Agency's requirement to
promulgate regulations for making filtration decisions by that date under
the SDWA. Primacy Agencies may specify which systems should conduct the
monitoring necessary to demonstrate compliance with the criteria for
avoiding filtration. For some systems where an historical data base
exists, and where it is apparent that the system would exceed the source
7 - 1
-------
water quality criteria (or that some other criteria would not be met, such
as an adequate watershed control program), no monitoring may be necessary
for the Primacy Agency to determine that filtration is required. If a
particular system (and/or the Primacy Agency) knows that it cannot meet
the criteria for avoiding filtration, there is no reason to require that
system to conduct the source water monitoring prior to the formal decision
by the Primacy Agency that filtration is required. This is true because
the only purpose of that monitoring would be to demonstrate whether or not
the criteria to avoid filtration are being met.
In reviewing the data for determining which systems must filter, the
Primacy Agency will have to decide on a case-by-case basis the conditions
which will require filtration. For example, a system may not meet the
specified CT requirements for the first few months of monitoring and
upgrades its disinfection to meet the CT requirements in subsequent
months. In this case, the Primacy Agency could conclude that the system
will be able to meet this criterion for avoiding filtration. The time
periods specified for in the criteria to avoid filtration (e.g., six
months for total coliforms, one year and ten years for turbidity and one
year for CT requirements) do not begin until December 30, 1991 unless the
Primacy Agc-cy specifies an earlier date.
Beginning December 30, 1991 the requirements for avoiding filtration
specified in S141.7l(a) and (b) and the requirements of S141.71(c) and
5141.72(a) go into effect unless the Primacy Agency already has determined
that filtration is required. Beginning December 30, 1991, if a system
fails to meet a"hy one of the criteria for avoiding filtration, even if the
system were meeting all the criteria up to that point, it must install
filtration and comply with the requirements for filtered systems includ-
ing the general requirements in S141.73 and the disinfection requirements
in S141.72(b), within 18 months of the failure. Whenever a Primacy Agency
determines that filtration is required, it may specify interim require-
ments for the period prior to installation of filtration treatment.
Following the determination that filtration is required, the system
must develop a plan to implement its installation. The plan must include
consideration for the following:
7 - 2
-------
Providing uninterrupted water service throughout the
transition period
Siting for the future facility
Financing options and opportunities
Scheduling of design and construction
Systems which are unable to install filtration within the specified time
frame may apply for an exemption to extend the period for installing
filtration.
Table 7-1 summarizes the requirements for the SWTR for unfiltered
systems noting conditions which require the installation of filtration.
It is important to note that only treatment technique violations trigger
the requirement to install filtration while violations of monitoring,
reporting or analytical requirements do not. The monitoring requirements
for unfiltered supplies are presented in Section 3 and the reporting
requirements are presented in Section 6.
All systems with filtration in place must meet the treatment
technique requirements specified in S141.73 (filtration criteria) and
S141.72(b) (disinfection criteria), and the monitoring and reporting
requirements specified in S141.74(c) and S141.75(b), respectively,
beginning June 29, 1993. Table 7-2 summarizes the SWTR requirements for
filtered systems, including conditions needed for compliance with
treatment requirements. Monitoring requirements for filtered supplies are
enumerated in Section 5 and reporting requirements are presented in
Section 6.
7.3 Compliance Transition with Current NPDWR Turbidity Requirements
The current (interim) NPDWR for turbidity under 5141,13 (MCI
requirements) and S141.22 (monitoring requirements) will apply for
unfiltered systems until December 30, 1991 unless the Primacy Agency
determines that filtration is required. In cases where filtration is re-
quired, the interim NPDWR applies until June 29, 1993 or until filtration
is installed, whichever is later. Unfiltered supplies will also be
7 - 3
-------
subject to the turbidity monitoring requirements of S141.74(b)(2)
beginning December 30, 1990 coincidently with the interim requirements.
Beginning June 29, 1993, the turbidity performance criteria for filtered
systems (S141.73), and the monitoring requirements under S141.74 will
apply.
7.4 Systems Using a Ground Water Source
Under the Direct Influence of Surface Water
Part of the Primacy Agency's program revisions to adopt the SWTR
must include procedures for determining, for each system in the Primacy
Agency served by a ground water source, whether that source is under the
direct influence-of surface water. By June 29, 1994 and June 29, 1999,
each Primacy Agency must determine which community and non-community
public water supplies, respectively, use ground water which is under the
direct influence of surface water. EPA recommends that these determina-
tions be made in conjunction with related activities required by other
regulations (e.g., sanitary surveys pursuant to the final coliform rule,
vulnerability assessments pursuant to the volatile organic chemicals rule,
the forthcoming disinfection requirements for ground water systems). In
addition, EPA-approved wellhead protection programs required under the
Safe Drinking Water Act Section 1428 may contain methods and criteria for
determing zones of contribution, assessments of potential contamination,
and management of sources of contamination. These programs may be used
as a partial basis for the vulnerability assessment and for making tne
determination <5f (a) whether a system is under the direct influence of
surface water and (b) if direct influence is determined, whether there is
adequate watershed control to avoid filtration. Guidelines for developing
and implementing a wellhead protection program are found in "Guidelines
for Applicants for State Wellhead Protection Program Assistance Funds
under the Safe Drinking Water Act" (U.S. EPA, 1987a).
A system using a ground water source under the influence of surface
water that does not have filtration in place must begin monitoring and
reporting in accordance with S141.74(b) and S141.75(a), respectively, to
determine whether it meets the criteria for avoiding filtration beginning
December 30, 1990 or six months after the Primacy Agency determines tHat
7 - 4
-------
«> o
&
£ ^O
csi _ *«
3 «•
o o c
v JB »» ••
w 6 v
— »<••. —
- E *• w a
o —
e e
o —
o a
- ao c as c c - c o « - *• o
» £ 3 «; = 3 » =s e - — *•» e
- - C C
w i* e o
~| -O w g
c o * e
_ g fi *
,
** ^ a, e
« - -o fi
ii
. 00
— i* * N^. —*
w —1 w e tst tsj
•* O| * r-« e "—' —
-------
e a.
c •»
£ a M •> g A — 9 „ — — —t^ — ^*!
— ^ o — — —fc— u e — 3*a«
- e^*>>- aa— — o «« * »i o —
C C — o X
i a • « a
w • — •» OP
«* a e «
— m — e ••
> * ^ - 6 o
— Oh. O — A I
=gf?£.!
?x «O
** C O
X — o
o
««
e
-------
E
o-
I:
e -
i:
o •»
tM
c
~
o ft*
g *i
E c
-------
e i
= 6
o E =
= E
— tst
**
•» e:
E « «- E
•* « c »
^ _ » m v>
_ e o — —
s6:
30 —
-------
the ground water source is under the influence of surface water, whichever
is later. Within 18 months following the determination that a system is
under the influence of surface water, the Primacy Agency must determine,
using the same criteria that apply to systems using a surface water
source, whether the system must provide filtration treatment. As for
systems using a surface water source, the Primacy Agency must evaluate the
data on a case-by-case basis to determine conditions which will trigger
the need for filtration.
Beginning December 30,1991 or 18 months after the determination that
a system is under the direct influence of surface water, whichever is
later, the criteria for avoiding filtration in S141.71(a) and (b) and the
requirements for unfiltered systems in S141.71(c) and S141.72(a) go into
effect, unless the Primacy Agency has determined that filtration is
required. As with systems using a surface water source, subsequent
failure to comply with any one of the criteria for avoiding filtration
requires the installation of filtration treatment. Thus, beginning
December 30, 1991 or 18 months after the Primacy Agency determines that
a system is using a ground water source under the direct influence of
surface water, whichever is later, a system which fails to meet any one
of the criteria to avoid filtration must install filtration and comply
with the requirements for filtered systems within 18 months of the failure
or by June 29, 1993, whichever is later. As for unfiltered systems,
systems under the direct influence of surface water may apply for an
exemption to extend the^time period for installing filtration.
Any systeln using a ground water source that the Primacy Agency
determines is under the direct influence of surface water end that already
has filtration in place at the time of the Primacy Agency determination
must meet the treatment technique, monitoring and reporting requirements
for filtered systems beginning June 29, 1993 or 18 months after the
Primacy Agency determination, whichever is later.
7 - 5
-------
7.5 Responses for Systems not; Mating SWTR Criteri?
7.5.1 Introduction
Systems which presently fail to meet the SWTR criteria may be able
to upgrade the system's design and/or operation and maintenance in order
to achieve compliance. The purpose of this section is to present options
which may be followed to achieve compliance.
7.5.2 Systems Not Filtering
Systems not filtering must meet the criteria to avoid filtration
beginning December 30, 1991 and on a continuing basis thereafter or
install filtration. Systems not filtering can be divided into two
categories:
A. Those systems not currently meeting the SWTR criteria but with
the ability to upgrade to meet them.
B. Those systems not able to meet the SWTR criteria by December
30, 1991. If the installation of filtration is not possible
by June 29, 1993 the system may request an exemption and take
interim measures to provide safe water to avoid violation of
a treatment technique requirement.
Systems in Category A
Example A - Response Situation
Condition; System is not meeting the source water fecal and/or
total coliform concentrations but has not received judgment on the
adequacy-of its watershed control.
Response Options;
Monitor for fecal coliforms rather than total coliforms if
this is not already done. Fecal coliforms are a direct
indicator of fecal contamination where total coliforms are
not. If total coliform levels are exceeded but fecal levels
are not, the system meets the criteria.
Take appropriate action in the watershed to assure fecal and
total coliform concentrations are below the criteria, such as
elimination of animal activity near the source water intake.
7 - 6
-------
Example B - Response Situation
Condition! System meets the source water quality criteria,
watershed control requirements, and is maintaining a disinfectant
residual within the distribution system, but is not able to meet the
CT requirements due to lack of contact time prior to the first
customer.
Response Options:
Increase the application of disinfectant while monitoring THM
levels to ensure they remain below the MCL.
Add additional contact time through storage to obtain an
adequate CT.
Apply a more effective disinfectant such as ozone.
Systems in Category B
Example A - Response Situation
Condition; System meets the source water turbidity but not the
fecal colifonn requirements. A sewage treatment plant discharges
into the source water. A determination has been made that the
system does not have adequate watershed control.
Response Options;
Purchase water from a nearby surveyor or use an alternate
source such as ground water if available.
Take steps to install filtration, applying for an exemption
(time delay) as presented in Section 9 where appropriate.
Example B
Condition; The source water exceeds a turbidity of 5 NTU for more
than two periods in a year under normal weather and operating
conditions.
Response Potions;
Purchase water from a nearby purveyor or use an alternate
source such as ground water if available.
Take steps to install filtration, applying for an exemption
(time delay) as presented in Section 9 where appropriate.
7 - 7
-------
In the interim prior to adoption of either of the above options,
certain protective measures may be appropriate. One protective
measure which can be used would be the issuance of a public notice
to boil all water for consumption during periods when the turbidity
exceeds 5 NTU. If such a notice is issued, the utility should
continue sampling the distribution system for chlorine residual and
total coliforms, and initiate measurement of HPCs in the distribu-
tion system. These data and the raw water turbidity should be used
to determine when to lift the boil water notice.
The notice could be lifted when:
The historical (prior to high turbidity) disinfectant residual
concentration is reestablished in the distribution system;
The total coliform requirements are met;
The HPC count is less than 500/ml; and
The turbidity of the raw water is less than 5 NTU.
7.4.3 Systems Currently Filtering
Systems which are currently filtering must meet the SWTR criteria
within 48 months of the SWTR to be in compliance, after which the criteria
must be continually met for the system to be in compliance.
Example A - Response Situation
Condition; A direct filtration plant is treating a surface water
which is not compatible with this treatment process. The system is
not achieving its required turbidity performance or disinfection
criteria.
Response Potions;
Optimize coagulant dose.
Reduce filter loading rates.
Evaluate the effect on performance of installing flocculation
and sedimentation ahead of the filters.
7 - 8
-------
Example B - Response Situation
Condition: A filtration plant is using surface water which is
compatible with its treatment system. The system is not achieving
disinfection performance criteria required by the Primacy Agency to
achieve a 1-log inactivation of Giardia cysts; however, it is
meeting the requirements of the Total Coliform Rule.
Response Options:
Increase disinfectant dosage(s).
Install storage facilities to increase disinfectant contact
time.
Ensure optimum filtration efficiency by:
Use of a filter aid.
Reduction in filter loading rates.
More frequent backwashing of filters.
The Primacy Agency may grant additional removal credit for optimum
filtration.
EPA intends to promulgate National Primary Drinking Water Regula-
tions to regulate levels of disinfectants and disinfectant by-product when
it promulgates disinfection requirements for ground water systems
(anticipated in 1992). EPA is concerned that changes required in
utilities' disinfection practices to meet the required inactivations for
the SWTR might be inconsistent with treatment changes needed to comply
with the forthcoming regulations for disinfectants and disinfection
by-products. For this reason, the EPA is allowing Primacy Agencies
discretion in determining the level of disinfection required for filtered
systems to meet the overall treatment performance requirements specified
in the rule or recommended based on source water quality.
During the interim period, prior to promulgation of the disinfection
by-product regulation, EPA recommends that the Primacy Agency allow more
credit for Giardia cyst and virus removal than generally recommended.
This interim level is recommended in cases where the Primacy Agency
determines that a system is not currently at a significant risk from
microbiological concerns at the existing level of disinfection and that
7 - 9
-------
-------
-------
8. PUBLIC NOTIFICATION
The SWTR specifies that the public notification requirements of the
Safe Drinking Water Act (SDWA) and the implementing regulations of 40 CFR
Paragraph 141.32 must be followed. These regulations divide public
notification requirements into two tiers. These tiers are defined as
follows:
1. Tier 1:
a. Failure to comply with MCL
b. Failure to comply with prescribed treatment technique
c. Failure to comply with a variance or exemption schedule
2. Tier 2:
a. Failure to comply with monitoring requirements
b. Failure to comply with a testing procedure prescribed
by a NPDWR
c. Operating under a variance/exemption. This is not
considered a violation but public notification is
required.
The SWTR classifies violations of Sections 141.70, 141.71(c),
141.72 and 141.73 (i.e., treatment technique requirements as specified in
Section 141.76) as Tier 1 violations and violations of Section 141.74 as
Tier 2 violations. Violations of 141.75 (reporting requirements) do not
require public notification.
There are certain general requirements which all public notices must
meet. All notices must provide a clear and readily understandable
explanation of the violation, any potential adverse health effects, the
population at risk, the steps the system is taking to correct the
violation, the necessity of seeking alternate water supplies (if any) and
any preventative measures the consumer should take. The notice must be
conspicuous, not contain any unduly technical language, unduly small print
or similar problems. The notice must include the telephone number of the
owner or operator or designee of the public water system as a source of
additional information concerning the violation where appropriate. The
notice must be bi- or multilingual if appropriate.
In addition, the public notification rule requires that when
providing information on potential adverse health effects in Tier 1 public
8 - 1
-------
a deferral is necessary for the system to upgrade its disinfection process
. op ,„ ,y achieve compliance „,«, the SWT* as „,„ as the fort
i. Section 5.5.3
guidelines for establishing interim disinfection
requirements.
7 - 10
-------
notices and in notices on the granting and continued existence of a
variance or exemption, the owner or operator of a public water system must
include certain mandatory health effects language. For violations of
treatment technique requirements for filtration and disinfection, the
mandatory health effects language is:
Microbiological Contaminants
The United States Environmental Protection Agency (EPA) sets drinking
water standards and has determined that microbiological contaminants are
a health concern at certain levels of exposure. If water is inadequately
treated, microbiological contaminants in that water may cause disease.
Disease symptoms may include diarrhea, cramps, nausea, and possibly
jaundice and any associated headaches, and fatigue. These symptoms,
however, are not just associated with disease-causing organisms in
drinking water, but also may be caused by a number of factors other than
your drinking water. EPA has set enforceable requirements for treating
drinking water to reduce the risk of these adverse health effects.
Treatment such as filtering and disinfecting the water removes or destroys
microbiological contaminants. Drinking water which is treated to meet EPA
requirements is associated with little to none of this risk and should be
considered safe.
Further, the owner or operator of a community water system must give
a copy of the most recent notice for any Tier 1 violations to all new
billing units or hookups prior to or at the time service begins.
The medium for performing public notification and the time period
in which notification must be sent varies with the type of violation and
is specified in Section 141.32. For Tier 1 violations (i.e., violations
of Sections 141^.70, 141.71, 141.72 and 141.73), the owner or operator of
a public water system must give notice:
1. By publication in a local daily newspaper as soon as possible
but in no case later than 14 days after the violation or
failure. If the area does not have a daily newspaper, then
notice shall be given by publication in a weekly newspaper of
general circulation in the area, and
2. By either direct mail delivery or hand delivery of the notice,
either by itself or with the water bill not later than 45 days
after the violation or failure. The Primacy Agency may waive
this requirement if it determines that the owner or operator
has corrected the violation within the 45 days.
8 - 2
-------
Although the SWTR does not specify any acute violations, the Primacy
Agency may specify some Tier 1 violations as posing an acute risk to human
health; for example these violations may include:
1. A waterborne disease outbreak in an unfiltered supply.
2. Turbidity of the water prior to disinfection of an unfiltered
supply or the turbidity of filtered water exceeds 5 NTU at any
time.
3. Failure to maintain a disinfectant residual of at least 0.2
mg/1 in the water being delivered to the distribution system.
For these violations or any others defined by the Primacy Agency as
"acute" violations, the system must furnish a copy of the notice to the
radio and television stations serving the area as soon as possible but in
no case later than 72 hours after the violation. Depending upon circum-
stances particular to the system, as determined by the Primacy Agency, the
notice may instruct that all water should be boiled prior to consumption.
Following the initial notice, the owner or operator must give notice
at least once every three months by mail delivery (either by itself cr
with the water bill), or by hand delivery, for as long as the violation
or failure exists.
There are two variations on these requirements. First, the owner
or operator of a community water system in an area not served by a daily
or weekly newspaper must give notice within 14 days after the violation
by hand delivery or continuous posting of a notice of the violation. The
notice must be in a conspicuous place in the area served by the system and
must continue for as long as the violation exists. Notice by hand
delivery must be repeated at least every three months for the duration of
the violation.
Secondly, the owner or operator of a noncoitrounitv water system
(i.e., one serving a transitory population) may give notice by hand
delivery or continuous posting of the notice in conspicuous places in the
area served by the system. Notice must be given within 14 days after the
violation. If notice is given by posting, then it must continue as long
8 - 3
-------
as the violation exists. Notice given by hand delivery must be repeated
at least every three months for as long as the violation exists.
For Tier 2 violations (i.e., violations of 40 CFR 141.74, analytical
and monitoring requirements) notice must be given within three months
after the violation by publication in a daily newspaper of general
circulation, or if there is no daily newspaper, then in a weekly
newspaper. In addition, the owner or operator shall give notice by mail
(either by itself or with the water bill) or by hand delivery at least
once every three months for as long as the violation exists. Notice of
a variance or exemption must be given every three months from the date it
is granted for as long as it remains in effect.
If the area is not served by a daily or weekly newspaper, the owner
or operator of a community water system must give notice by continuous
posting in conspicuous places in the area served by the system. This must
continue as long as the violation does or the variance or exemption
remains in effect. Notice by hand delivery must be repeated at least
every three months for the duration of the violation or the variance of
exemption.
For noncommunity water systems, the owner or operator may give
notice by hand delivery or continuous posting in conspicuous places;
beginning within 3 months of the violation or the variance or exemption.
Posting must continue for the duration of the violation or variance or
exemption and notice by hand delivery must be repeated at least every
3 months during this period.
The Primacy Agency may allow for owner or operator to provide less
frequent notice for minor monitoring violations (as defined, by the
Primacy Agency if EPA has approved the Primacy Agency's substitute
requirements contained in a program revision application).
To provide further assistance in preparing public notices, several
examples have been provided. However, each situation is different and
may call for differences in the content and tone of the notice. All
notices must comply with the general requirements specified above.
8 - 4
-------
Example 1 - Tier 1 Violation-Unfilled $upp1y
Following is an example of a Tier 1 violation which may be
considered by the Primacy Agency to pose an acute risk to human health.
A system which does not apply filtration experiences a breakdown in
the chlorine feed systems and the switchover system fails to activate the
backup systems. A number of hours pass before the operator discovers the
malfunction. The operator, upon discovery of the malfunction, contacts
the local television and radio stations and announces that the public is
receiving untreated water. The announcement may read as follows:
We have just received word from the Aswan Water Board that a
malfunction of the disinfection system has allowed untreated water
to pass into the distribution system. Thus, this system providing
drinking water is in violation of a treatment technique requirement.
The United States Environmental Protection Agency (EPA) sets
drinking water standards and has determined that microbiological
contaminants are a health concern at certain levels of exposure.
If water is inadequately treated, microbiological contaminants in
that water may cause disease. Disease symptoms may include
diarrhea, cramps, nausea, and possibly jaundice and any associated
headaches, and fatigue. These symptoms, however, are not just
associated with disease-causing organisms in drinking water, but
also may be caused by a number of factors other than your drinking
water. EPA has set enforceable requirements for treating drinking
water to reduce the risk of these adverse health effects. Treatment
such as filtering and disinfecting the water removes or destroys
microbiological contaminants. Drinking water which is treated to
meet EPA requirements is associated with little to none of this risk
and should be considered safe.
The temporary breakdown in disinfection may have allowed micro-
organisms to pass into the distribution system. The operation of
the syst«ra has been restored so that no further contamination of
the distribution system will occur. Any further changes will be
announced.
Additional information is available at the following number:
235-WATER.
A direct mailing of the notice is provided within 45 days of the
occurrence.
Example 2 - Tier 1 Violation-Unfiltered Supply
Following is an example of a Tier 1 violation which may be
considered by the Primacy Agency to pose an acute risk to human health.
8 - 5
-------
A system supplies an unfiltered surface water to its customers.
During a period of unusually heavy rains caused by a hurricane in the
area, the turbidity of the water exceeds 5 NTU. The turbidity data during
which the heavy rains occur is as follows:
Dav 1 NTU Dav 2 NTU Dav 3 NTU Dav 4 NTU Dav 5 NTU
0.4 0.8 0.7 0.7 7.6
0.4 0.5 0.4 7.6 3.1
0.5 0.5 0.4 11.3 2.7
0.7 0.4 0.5 9.6 0.7
1.1 0.4 0.4 7.2 0.8
0.9 0.6 0.6 5.0 0.5
The following public notice was prepared and submitted to the local
newspaper, television and radio stations within 72 hours of the first
turbidity exceedence of 5 NTU.
The occurrence of heavy rains in our watershed is causing a rise in
the turbidity of the drinking water supplied by Fairfax Water
Company.
Turbidity is a measurement of particulate matter in water. It is
of significance in drinking water because irregularly shaped
particles can both harbor microorganisms and interfere directly with
disinfection which destroys microorganisms. While the particles
causing the turbidity may not be harmful or even visible at the
concentrations measured, the net effect of a turbid water is to
increase the survival rate of microorganisms contained in the water.
This is of concern because several diseases are associated with
waterborne microorganisms.
Because of the high turbidity levels, the Fairfax system is in
violation of a treatment requirement set by the Environmental
Protection Agency (EPA).
The United States Environmental Protection Agency (EPA) sets
drinking water standards and has determined that microbiological
contaminants are a health concern at certain levels of exposure.
If water is inadequately treated, microbiological contaminants in
that water may cause disease. Disease symptoms may include
diarrhea, cramps, nausea, and possibly jaundice and any associated
headaches, and fatigue. These symptoms, however, are not just
associated with disease-causing organisms in drinking water, but
also may be caused by a number of factors other than your drinking
water. EPA has set enforceable requirements for treating drinking
water to reduce the risk of these adverse health effects. Treatment
such as filtering and disinfecting the water removes or destroys
microbiological contaminants. Drinking water which is treated to
8 - 6
-------
meet EPA requirements is associated with little to none of this risk
and should be considered safe.
In order to protect yourself from illness, all water from the
Fairfax system used for drinking, cooking and washing dishes should
be boiled at a rolling boil for one minute.
The system is being closely monitored and a notice will be issued
when the water returns to an acceptable quality and no longer needs
to be boiled.
The utility continues sampling the distribution system for chlorine
residual and total coliforms, and initiates measurement of the HPCs in the
distribution system. The notice is lifted when all the following are met:
The historical (prior to high turbidity) disinfectant residual
concentration is reestablished in the distribution system.
The total coliform requirements are met.
The HPC count is <500/ml.
The turbidity of the raw water is less than 5 NTU.
The Primacy Agency most decide whether the turbidity event was unusual or
unpredictable and whether filtration should be installed.
Example 3 - Tier 1 Violation - Filtered Supply
A conventional treatment plant is treating a surface water. A
malfunctioning alum feed system resulted in an increase of the filter
effluent turbidities. The effluent turbidity was between 0.5 and 1.0 NTU
in 20 percent of the samples for the month. The utility issued a notice
which was published in a local daily newspaper within 14 days after the
violation. The notice read as follows:
During the previous month, the Baltic Water Treatment Plant
experienced difficulties with the chemical feed system. The
malfunctions caused an effluent turbidity level above 0.5 NTU in 20
percent of the samples for the month. The current treatment
standards require that the turbidity must be less than 0.5 NTU in
95 percent of the monthly samples. The Baltic drinking water system
has thus been in violation of a treatment technique requirement.
The United States Environmental Protection Agency (EPA) sets
drinking water standards and has determined that microbiological
8 - 7
-------
contaminants are a health concern at certain levels of exposure.
If water is inadequately treated, microbiological contaminants in
that water may cause disease. Disease symptoms may include
diarrhea, cramps, nausea, and possibly jaundice and any associated
headaches, and fatigue. These symptoms, however, are not just
associated with disease-causing organisms in drinking water, but
also may be caused by a number of factors other than your drinking
water. EPA has set enforceable requirements for treating drinking
water to reduce the risk of these adverse health effects. Treatment
such as filtering and disinfecting the water removes or destroys
microbiological contaminants. Drinking water which is treated to
meet EPA requirements is associated with little to none of this risk
and should be considered safe.
The chemical, feed and switchover components of the system have been
repaired and are in working order and turbidity levels are meeting
the standard. It is unlikely that illness will result from the
turbidity exceedences previously mentioned because continuous
stringent disinfection conditions were in effect and the system was
in compliance with other microbiological drinking water standards
pertaining to microbiological contamination. However, a doctor
should be contacted in the event of illness. For additional
information call, 1-800-726-WATER.
8 - 8
-------
9. EXEMPTION^
9.1 Overview of Requirements
Section 1416 of the Safe Drinking Water Act allows a Primacy Agency
to exempt any public water system within its jurisdiction from any
treatment technique requirement imposed by a national primary drinking
water regulation upon a finding that:
1. Due to compelling factors (which may include economic
factors), the public water system is unable to comply with the
treatment technique requirement;
2. The public water system was in operation on the effective date
of the treatment technique requirement or, for a system that
was not in operation by that date, only if no reasonable
alternative source of drinking water is available to the new
system; and
3. The granting of the exemption will not result in an unreason-
able risk to health.
If a Primacy Agency grants a public water system an exemption, the
Agency must prescribe, at the time the exemption is granted, a schedule
for:
1. Compliance (including increments of progress) by the public
water system with each treatment technique requirement with
respect to which the exemption was granted; and
2. Implementation by the system of such control measures as the
Primacy Agency may require during the period the exemption is
in .effect.
Before prescribing a schedule, the Primacy Agency must provide
notice and opportunity for a public hearing on the schedule. The schedule
prescribed must require compliance by the public water system with the
treatment technique requirement as expeditiously as practicable, but in
no case later than one year after the exemption is issued (except that,
if the system meets certain requirements, the final date for compliance
may be extended for a period not to exceed three years from the date the
exemption is granted). For systems serving less than 500 service
9 - 1
-------
connections, and meeting certain additional requirements, the Primacy
Agency may renew the exemption for one or more additional two-year
periods.
Under the SWTR, no exemptions are allowed from the requirement to
provide disinfection for surface water systems, but exemptions are
available to reduce the degree of disinfection required. Exemptions from
the filtration requirements are available. The following sections present
guidelines for evaluating conditions under which exemptions are appropri-
ate.
9.2 Recommended Criteria
In order to obtain an exemption from the SWTR, a system must meet
certain minimum criteria to assure no unreasonable risk to health. These
should be applied before looking at other factors such as economics.
Recommended minimum criteria for assuring no unreasonable risk to health
exists are listed below.
Systems which do not provide filtration
Practice disinfection to achieve at least a 2-log inactivation
of Giardia cysts; or comply with the disinfection requirements
for the distribution system as defined in Section 141.72(b)
of the SWTR.
Comply with the monthly coliform MCL; or provide bottled water
(or another alternate water source) or point of use treatment
devices for their customers in which representive samples
comply with all the MCL National Primary Drinking Water
Regulations.
EPA recommends that in order to obtain an extension to the initial
1 year exemption period in addition to the required elements in Section
1416, the system would need to be in compliance with the monthly coliform
MCL, satisfy the above disinfection criteria and not have any evidence of
waterborne disease outbreaks attributable to the system at the end of that
first exemption period. If at any point during the extended exemption
period the system did not meet these conditions, the exemption should be
withdrawn and the system should be subject to an enforcement action.
9 - 2
-------
Systems whi^h provide filtration
Practice disinfection to achieve at least a 0.5 log inactiva-
tion of Giardia cysts; or comply with the disinfection
requirements for the distribution system as defined in
Section 141.72 of the rule.
Comply with the monthly coliform MCL; or provide bottled water
(or another alternate water source) or point of use treatment
devices for their customers in which represent!ve samples
comply with all the MCL National Primary Drinking Water
Regulations.
^
Take all practical steps to improve the performance of i
filtration system.
ts
In order to obtain an extension to the initial exemption period, in
addition to the required elements in Section 1416, the system should be
in compliance with the colifortn MCL, satisfy the above disinfection
criteria and not have any evidence of waterborne disease outbreaks
attributable to the treatment system at the end of that first exemption
period. If at any point during the extended exemption period the system
did not meet these conditions, the exemption should be withdrawn and the
system should be subject to an enforcement action. In addition, the
system must continue to be taking steps to improve the performance of its
filtration system to achieve the criteria specified in the SWTR.
Once these minimum requirements are applied, the Primacy Agency
should look at the other factors as described in Sections 9.3, 9.4, and
9.5.
9.3 Compelling Factors
Compelling factors are often associated with small systems. The
major compelling factor tends to be economic. In some cases the
compelling factor may not be solely economic, but rather the contractual
and physical infeasibility of having a required treatment installed within
the time period specified in the regulation. For example, it may not be
feasible for a very large system to install filtration by June 1993 if
required. In such cases exemptions are also appropriate. Additional
considerations for small systems are presented in Appendix L.
9 - 3
-------
If system improvements necessary to comply with the SWTR incur costs
which the Primacy Agency determines pose an economic barrier to acquisi-
tion of necessary treatment, the system fulfills the criteria of
demonstrating a compelling hardship which makes it unable to meet the
treatment requirements. In such cases, the EPA believes it is reasonable
to grant an exemption if the system also meets the criteria in 9.4 and
9.5.
The USEPA document, "Technologies and Costs for the Removal of
Microbial Contaminants from Potable Water Supplies," contains costs
associated with available treatment alternatives (USEPA, 1988b). Costs
found in this document, or those generated from more site-specific
conditions, can be used as the basis for determining the ability of a
system to afford treatment. The total annual water production costs per
household for a system can be estimated based on the household water usage
and the production costs per thousand gallons. As estimated in the above
cited USEPA document, each cent per thousand gallons of treated water is
approximately equivalent to $1 per year per household if a household water
usage of 100,000 gallons per year is assumed.1 This estimate will need to
be adjusted according to water usage for cases where the household usage
differs from 100,000 gallons per year.
The following examples are presented to provide guidance in
estimating costs for a system to upgrade its system or install filtration.
This cost information could be used for determining whether a system might
be eligible for an exemption.
Example 1
A water system which supplies an average daily flow of 0.05 mgd to
a small urban community receives its water supply from a lake. The system
currently provides disinfection with chlorine but does not provide
filtration. The system reviewed its source water quality and found the
characteristics to be as follows:
1 This is the national average residential household consumption reported
in: Final Descriptive Summary - 1986 Survey of Community Water Systems.
October 23, 1987. USEPA: Office of Drinking Water.
9 - 4
-------
Total coliforms 1,000/100 ml
Turbidity 10 - 13 NTU
Color 6 - 9 CU
Based upon the criteria in the SWTR, this source requires filtration
and a review of the water quality criteria presented in Table 4-2
indicates that the treatment technique best suited to these source
conditions is conventional treatment. A conventional package treatment
plant with a capacity of 0.068 MGD may be purchased and put on line at
a cost of $277/household-year not including real estate, piping or raw
water pumping costs which may be significant depending on the plant
location.2 EPA has estimated that, on average, these costs might add
another 50% depending on site specific factors (USEPA, 1989)
Thus the cost estimate for implementing filtration indicates that
the increase in the average annual household water bill would be
approximately $277 plus the cost of real estate, piping, and raw water
pumping as needed. The incomes of people in the community and the current
water bills can be reviewed by the Primacy Agency along with these
estimated costs to determine if an undue economic hardship is incurred by
these treatment methods. Upon determination that an economic hardship is
incurred, the Primacy Agency may grant an exemption from filtration,
provided that the system can assure the protection of the health of the
community. However, if the water supply system for a nearby community
meets the drinking water standards ajnd there is the ability to hook up to
that system, an exemption generally should not be granted unless such
costs also presented an economic hardship.
Example 2
A large urban community, with a median annual income of $25,000 per
family, is supplied with water from lakes and reservoirs. The community
places an average daily demand of 3 mgd on the supply system. The
watershed of the system is moderately populated and used for farming and
Table VI-3 ("Technologies and Costs for the Removal of Microbial
Contaminants From Potable Water Supplies," USEPA, 1988b) lists the total
costs as 277.4 cents/1000 gal. Estimated costs for real estate, piping
and raw water pumping as a function of site specific conditions are
available in Table E-l, E-2, and E-3 of this same document.
9 - 5
-------
grazing. The system currently provides filtration using diatomaceous
earth filtration and disinfection with chloramines.
A review of the source and finished water quality was conducted to
evaluate the plant's performance. The source water quality was determined
to be:
Total coliforms 30 - 40/100 ml
Turbidity 2 - 3 NTU
Color 1 - 2 CU
Diatomaceous earth is therefore an acceptable filtration method.3
However, review of the finished water showed that a residual in the
distribution system is only maintained 80 percent of the time. In
addition to this, coliforms were detected in 10 percent of the samples
taken over the twelve month period. Inspection of the chlorination
equipment showed the equipment is deteriorated. Review of the monthly
reports showed that the coliforms appeared in the distribution system
shortly after the chlorinators malfunctioned. This observation led to the
conclusion that new disinfection facilities were needed.
The source water quality and available contact time after disinfec-
tion were then used to determine the most appropriate disinfectant for the
system. As described in Section 5.5, ozone, chlorine or chlorine dioxide
can be used as primary disinfectants given these conditions. A prelimi-
nary review of costs for applying the various disinfectants showed
chlorine to be the most economical at a cost of $2.8/household/year*
(USEPA, 1988b). This cost does not include backup equipment; however,
even with providing duplicate equipment doubling this cost to $5.6/house-
hold/ year, the improvement incurs minimal cost and the Primacy Agency
should not grant the system an exemption based on economic hardship.
As determined from Table 4-2 of Section 4.
Table VI-12 (USEPA, 1988b) lists a total cost of 2.8 cents/1000 gal for
a plant capacity of 5.85 mgd.
(2.8 cents) (Sl/household-vear) « $2.8/household-year
(1,000 gal) (cents/1000 gal)
9 - 6
-------
9.4 Evaluation of Alternate Water ^uaplv Sources
Systems which would incur very high costs for installing a required
treatment to comply with the SWTR, should evaluate the possibility of
using an alternate source. These alternate sources include:
The use of ground water
Connection to a nearby water purveyor
Use of an alternate surface water supply
When considering the use of ground water, the purveyor must
determine the capacity of the underlying aquifer for supplying the demand.
The water quality characteristics of the aquifer must be evaluated to
determine what treatment may be needed to meet existing standards. The
cost of the well construction and treatment facilities must then be
determined and converted into a yearly cost per household.
The connection to a nearby purveyor involves contacting the purveyor
to determine their capacity and willingness to supply the water. Once it
has been determined that the alternate source meets all applicable
drinking water standards, the cost of the transmission lines, distribution
system, and other facilities (e.g. disinfection, repumping, etc.) must
then be determined and amortized into a yearly cost per household.
If the cost for using an alternate source is found by the Primacy
Agency to present an economic hardship, and the purveyor can demonstrate
that there will be no unreasonable risk to health, the Primacy Agency may
grant an exemption to the SWTR for the purveyor and develop a schedule of
compliance.
9.5 Protection of Public Health
Systems which apply for an exemption from the SWTR must demonstrate
to the Primacy Agency that the health of the community will not be put at
risk by the granting of such an exemption. A system should be able to
provide adequate protection for the public health by meeting the minimum
suggested EPA requirements in Section 9.2. However, a Primacy Agency may
specify additional measures or criteria a system must meet to protect
public health, depending on the particular circumstances. Systems *itn
currently unfiltered surface water supplies which fail to meet the source
9 - 7
-------
water quality criteria will be required to install filtration as part of
their treatment process. However, it may take 3 to 5 years or more before
the filtration system can be designed, constructed and begin operation,
thereby justifying the granting of an exemption. During this period,
possible interim measures which the system could take to further satisfy
the Primacy Agency's concern include one or more of the following:
a. Use of higher disinfectant dosages without exceeding the TTHM
MCL (even for systems not currently subject to this MCL)
b. Installation of a replacement or additional disinfection
system which provides greater disinfection efficiency and
which can be integrated into the new filtration plant
c. Increasing the monitoring and reporting to the Primacy Agency
d. Increasing protection of the watershed
e. Increasing the frequency of sanitary surveys
f. Temporarily purchasing water from a nearby water system
g. For small systems, temporary installation of a mobile
filtration (package) plant
h. Increasing contact time by rerouting water through reservoirs
In some cases systems may be able to increase their disinfection
dosages during the interim period to provide additional protection against
pathogenic organisms. This alternative should be coupled with a
requirement for increased monitoring for colifortns, HPC and disinfectant
residual within the distribution system. However, disinfectant dosage
should not be increased if this would result in a violation of the TTHM
MCL, even for systems not currently subject to this MCL.
Systems which are planning to install filtration may be able to
utilize a more efficient disinfectant that can later be integrated into
the filter plant. Currently ozone and chlorine dioxide are considered to
be the most efficient disinfectants.
For all systems which do not meet the source water quality criteria
ajui must install filtration, EPA recommends that during the interim period
the Primacy Agency increase its surveillance of the system and require
9 - 8
-------
increased monitoring and reporting requirements to assure adequate
protection of the public health.
Any required increases in watershed control and/or on-site
inspections will not alleviate the need for more stringent disinfection
requirements and increased monitoring of the effectiveness of the system
employed. Their purpose would be to identify and control all sources of
contamination so that the existing system will provide water of the best
possible quality.
For some systems, it may be possible to purchase water from a nearby
system on a temporary basis. This may involve no more than the use of
existing interconnections or it may require the installation of temporary
connections.
Trailer mounted filtration units (package plants) are sometimes
available from state agencies for emergencies or may be rented or leased
from equipment manufacturers.
Systems may also be required to supply bottled water or install
point-of-entry (POE) treatment devices. For the reasons listed below,
these alternatives should only be utilized if the previously mentioned
alternatives are not feasible:
In many states bottled water is subject only to the water
quality requirements of the FDA as a beverage and not to the
requirements of the Safe Drinking Water Act.
Point-of-entry treatment devices are not currently covered by
performance or certification requirements which would assure
their effectiveness or performance.
If the installation of POE devices is required, the selection of the
appropriate treatment device should be based upon a laboratory or field
scale evaluation of the devices. A guide for testing the effectiveness
of POE units in the microbiological purification of contaminated water is
provided in Appendix N.
Several issues arise with the use of POE devices. These include
establishing who or what agency (1) has the responsibility for ensuring
compliance with standards; (2) retains ownership of the treatment units;
(3) performs monitoring, analyses and maintenance; and (4) manages the
9 - 9
-------
treatment program and maintains insurance coverage for damage and liabil-
ity. It should also be considered that there is no significant increase
in risk over centrally treated water.
These issues should be borne in mind when POE as a treatment
alternative is being considered.
Systems with currently unfiltered surface water supplies which meet
the source water quality criteria, but do not meet one or more of the
other requirements for watershed control, sanitary survey, compliance with
annual coliform MCL or disinfection by-product regulation(s), will be
required to install filtration unless the deficiencies can be corrected
within 48 months of promulgation of the SWTR. Interim protection measures
include those previously listed.
Systems with currently unfiltered surface water supplies which meet
the source water quality criteria and the site specific criteria but which
do not meet the disinfection requirements, will be required to install
filtration unless the disinfection requirements (adequate CT and/or
disinfection system redundancy) can be met. During the interim period,
available options include:
a. Temporary installation of a mobile treatment plant
b. Temporary purchase of water from a nearby purveyor
c. Increased monitoring of the system
d. Installation of temporary storage facilities to increase the
disinfectant contact time
Currently filtered supplies which fail to meet the turbidity or
disinfection performance criteria presented in Section 5 will be required
to evaluate and upgrade their treatment facilities in order to attain
compliance. During the interim period available options for improving the
finished water quality include:
a. Use of a filter aid to improve filter effluent turbidities
b. Increased disinfectant dosages
c. The addition of an alternate disinfectant is an option after
the disinfection by-products rule is promulgated
9 - 10
-------
d. Reduction in filter loading rates with subsequent reduction
in plant capacity
e. Installation of temporary storage facilities to increase
disinfectant contact time
9.6 Notification to EPA
The SDWA requires that each Primacy Agency which grants an exemption
notify EPA of the granting of this exemption. The notification must
contain the reasons for the exemption, including the basis for the finding
that the exemption will not result in an unreasonable risk to public
health and document the need for the exemption.
9 - 11
-------
REFERENCES
-------
REFERENCES
Ali-Ani, M.; McElroy, J. M.; Hibler, C. P.; Hendricks, D. W. Filtration
of Giardja Cysts and other Substances, Volume 3: Rapid Rate Filtration.
EPA-600/2-85-027, U.S. Environmental Protection Agency, WERL, Cincinnati,
Ohio, April, 1985.
American Public Health Association; American Water Works Association;
Water Pollution Control Federation. Standard Methods for the Examination
of Water and Wastewater. 16th ed., pp. 134-6, 298-310, 827-1038, 1985.
American Public Health Association; American Water Works Association;
Water Pollution Control Federation. Standard Methods for the Examination
of Water and Wastewater. 17th ed., 1989.
American Water Works Association. Manual of Water Supply Practices and
Water Chlorination Principles and Practices, 1973.
American Water Works Association Research Foundation (AWWARF). A Summary
of State Drinking Water Regulations and Plan Review Guidance. June, 1986.
Bader, H.; Hoigne, J. Determination of Ozone in Water by the Indigo
Method, Water Research 15; 449-454, 1981.
Bellamy, W. D.; Lange, K. P.; Hendricks, D. W. Filtration of Giardia Cvsts
and Other Substances. Volume 1: Diatomaceous Earth Filtration.
EPA-600/2-84-114, U.S. Environmental Protection Agency, Cincinnati, Ohio,
1984.
Bellamy, W. D.; Silver-man, G. P.; Hendricks, D. W. Filtration of Giardia
Cysts and Other Substances. Volume 2: Slow Sand Filtration. EPA-600/2-
-85-026, U.S. Environmental Protection Agency, MERL, Cincinnati, Ohio,
April, 1985.
Bishop, S.; Craft, T. F.; Fisher, D. R.; Ghosh, M.; Prendiville, P.W.;
Roberts, K. J.; Steimle, S.; Thompson, J. The Status of Direct Filtration,
Committee Report. J.AWWA, 72(7):405-411, 1980.
Bouwer, H. Ground Water Hvdroloov. McGraw Hill Book Co., New York,
pp. 339-356, 1978.
Brown, T. S.; Malina, J. F., Jr.; Moore, B. D. Virus Removal by Diatoma-
ceous Earth Filtration - Part 1 & 2. J.AWWA 66(2):98-102, (12):735-738,
1974.
Bucklin, K.; Amirtharajah, A.; Cranston, K. Characteristics of Initial
Effluent Quality and Its Implications for the Filter-to-Waste Procedure.
AWWA Research Foundation Report. November, 1988.
-1-
-------
Carlson, D.A.; Seabloom, R.W.; DeWalle, F.D.; Wetzler, T.F.; Evgeset, J.;
Butler, R.; Wangsuthachart, S.; Wang, S. Ultraviolet Disinfection of
Water for Small Water Systems. EPA/600/2-85-092, U.S. Environmental
Protection Agency, Water Engineering Reserach Laboratory, Drinking Water
Research Division, Cincinnati, Ohio, September, 1985.
Clark, R.M.; Regli, S. A Mathematical and Statistical Analysis for the
Inactivation of Giardia Iambiia by Free Chlorine. Submitted to the
Journal of Environmental Science Engineering, 1989.
Clark, R.; Regli, S.; Black, D. Inactivation of Giardia 1 amb1ia by Free
Chlorine: A Mathematical Model. Presented at AWWA Water Quality
Technology Conference. St. Louis, Mo., November 1988.
Cleasby, J. L.; Hilmoe, D. J.; Dimitracopoulos, C. J. Slow-Sand and Direct
In-Line Filtration of a Surface Water. J.AWWA, 76(12):44-55, 1984.
DeWalle, F. 8.; Engeset, J.; Lawrence, W. Removal of Giardia Iambiia
Cysts by Drinking Water Plants. EPA-600/S2-84-069, United States En-
vironmental Protection Agency, MERL, Cincinnati, Ohio, May 1984.
Fox, K. R.; Miltner, R. J.; Logsdon, G. S.; Dicks, D. L.; Drolet, L. F.
Pilot Plant Exploration of Slow Rate Filtration. Presented at the AWWA
Annual Conference Seminar, Las Vegas, Nevada, June 1983.
Fujioka, R.; Kungskulniti, N.; Nakasone, S. Evaluation of the Presence
- Absence Test for Coliforms and the Membrane Filtration Method for
Heterotrophic Bacteria. AWWA Technology Conference Proceedings, November,
1986.
Geldreich, E. Personal communication to Linda Averell, Malcolm Pirnie
Engineers, Paramus, New Jersey, July 1989.
Great Lakes-Upper Mississippi River Board of State Public Health and
Environmental Managers Committee. Recommended Standards for Water Works.
1987 Edition. -
Hendricks, D.; Al-Ani, M.; Bellamy, W.; Hibler, C.; McElroy, J. Surrogate
Indicators for Assessing Removal of Giardia Cysts, AWWA Water Quality
Technology Conference, 1984.
Hoff, J. C. Inactivation of Microbial Agents bv Chemical Disinfectants.
EPA-600/S2-86-067, U.S. Environmental Protection Agency, Water Engineer-
ing Research Laboratory, Drinking Water Research Division, Cincinnati,
Ohio, September 1986.
Hoffbuhr, J. W.; Blair, J.; Bartleson, M.; Karlin, R. Use of Particulate
Analysis for Source and Water Treatment Evaluation. AWWA Water Quality
Technology Conference Proceedings, November 1986.
Horn, J. B.; Hendricks, D. W. Removals of Giardia Cysts and other
Particles from Low Turbidity Waters Using the Culligan Multi-Tech Filtra-
-2-
-------
tion System. Engineering Research Center, Colorado State University,
Unpublished, 1986.
Joost, R. D.; Long, B. W.; Jackson, I. Using Ozone as a Primary
Disinfectant for the Tucson CAP Water Treatment Plant, presented at the
IOA/PAC Ozone Conference, Monroe, MI, 1988.
Kuchta, J. M.; States, S. J.; McNamara, A. M.; Wadowsky, R. M.; Yee, R. B.
Susceptibility of Legionella pneumophila to Chlorine in Tap Water. Appl.
Environ. Microbiol., 46(5): 1134-1139,' 1983.
Letter-man, R. D. The Filtration Requirement in the Safe Drinking Water
Act Amendments of 1986. U.S. EPA/AAAS Report, August 1986.
Logsdon, G. S.; Symons, J. M.; Hoye, Jr., R. L.; Arozarena, M. M.
Alternative Filtration Methods for Removal of Giardia Cysts and Cyst
Model. J.AWWA, 73:111-118, 1981.
Logsdon, G.; Thurman, V.; Frindt, E.; Stoecker, J. Evaluating Sedimenta-
tion and Various Filter Media for Removal of Giardig Cysts. J. AWWA,
77:2:61, 1985.
Logsdon, G. S. Report for Visit to Carrollton, Georgia, USEPA travel
report, February 12, 1987a.
Logsdon, G. S. Comparison of Some Filtration Processes Appropriate for
Giardia Cyst Removal. USEPA Drinking Water Research Division; Presented
at Calgary Giardia Conference, Calgary; Alberta, Canada, February 23-25,
1987b.
Long, R. L. Evaluation of Cartridge Filters for the Removal of Giardia
lamblia Cyst Models from Drinking Water Systems. J. Environ. Health,
45(5):220-225, 1983.
Markwell, D. D., and Shortridge, K. F. Possible Waterborne Transmission
and Maintenance of Influenza Viruses in Domestic Ducks. Applied and
Environmental Microbiology. Vol. 43, pp. 110-116, January, 1981.
Morand, J.( M.; C. R. Cobb; R. M. Clark; Richard, G. S. Package Water
Treatment Plants, Vol. 1, A performance Evaluation. EPA-600/2-80-008a,
USEPA, MERL, Cincinnati, Ohio, July 1980.
Morand, J. M,; Young, M. J. Performance Characteristics of Package Water
Treatment Plants, Project Summary. EPA-600/52-82-101, USEPA, MERL,
Cincinnati, Ohio, March, 1983.
Muraca, P.; Stout, J. E.; Yu, V. L. Comparative Assessment of Chlorine,
Heat, Ozone, and UV Light for Killing Legionella pneumophila Within a
Model Plumbing System. Appl. Environ. Microbiol., 53(2):447-453, 1987.
Notestine, T., Hudson, J. Classification of Drinking Water Sources as
Surface or Ground Waters. Final Project Report. Office of Environmental
-3-
-------
Health Programs. Washington Department of Social and Health Services,
August 1988.
Poynter, S. F. B.; Slade, J. S. The Removal of Viruses by Slow Sand
Filtration, Prog. Wat. Tech. Vol. 9, pp. 75-88, Pergamon Press, 1977.
Printed in Great Britain.
Randall, A.D. Movement of Bacteria From a River to a Municipal Well - A
Case History. American Water Works Association Journal. Vol. 62, No. 11,
p.716-720., November 1970.
Rice, E.W.; Hoff, J.C. Inactivation of Giardia Iambi la cysts by
Ultraviolet Radiation. Appl. Environ. Microbiol. 42: 546-547, 1981.
Robeck, G. G.; Clarke, N. A.; Dostal, K. A. Effectiveness of Water
Treatment Processes in Virus Removal. J. AWWA, 54(10):1275-1290, 1962.
Robson, C.; Rice, R.; Fujikawa, E.; Farver, B. Status of U.S. Drinking
Water Treatment Ozonation Systems, presented at IOA Conference, Myrtle
Beach, SC, December 1988.
Rose, J. Cryptosporidium in Water; Risk of Protozoan Waterborne Trans-
mission. Report prepared for the Office of Drinking Water, U.S. EPA,
Summer, 1988.
Rubin, A. Factors Affecting the Inactivation of Giardia Cysts by
Monochloramine and Comparison with other Disinfectants. Water Engineering
Research Laboratory, Cincinnati, OH, March 1988a.
Rubin, A. "CT Products for the Inactivation of Giardia Cysts by Chlorine,
Chloramine, Iodine, Ozone and Chlorine Dioxide" submitted for publication
in J. AWWA, December 1988b.
Slezak, L.; Sims, R. The Application and Effectiveness of Slow Sand
Filtration in the United States. J.AWWA, 76(12):38-43, 1984.
Sobsey, M. Detection and Chlorine Disinfection of Hepatitus A in Water.
CR-813-024. EPA Quarterly Report. December 1988.
Stolarik, G.; Christie, D. Projection of Ozone C-T Values Los Angeles
Aqueduct Filtration Plant, 1988.
U. S. Environmental Protection Agency, Office of Drinking Water, Criteria
and Standards Division. Manual for Evaluating Public Drinking Water
Supplies, 1971.
U. S. Environmental Protection Agency, Office of Drinking Water. Public
Notification Handbook for Drinking Water Suppliers, May 1978.
U. S. Environmental Protection Agency, Office of Ground Water Protection.
Guidelines for Applicants for State Wellhead Protection Program Assistance
Funds Under the SDWA, June 1987a.
-4-
-------
U. S. Environmental Protection Agency, Office of Ground Water Protection.
Guidelines for Delineation of Wellhead Protection Area, June 1987b.
U. S. Environmental Protection Agency, Office of Drinking Water. Workshop
on Emerging Technologies for Drinking Water Treatment, April, 1988a.
U. S. Environmental Protection Agency, Office of Drinking Water.
Technologies and Costs for the Removal of Microbial Contaminants from
Potable Water Supplies, October, 1988b.
World Health Organization Collaborating Center. Slow Sand Filtration of
Community Water Supplies in Developing Countries. Report of an Interna-
tional Appraiser Meeting, Nagpur, India, Bulletin Series 16, September
15-19, 1980.
-5-
-------
APPENDIX A
EPA CONSENSUS METHOD
FOR GIARDIA CYST ANALYSIS
-------
To be^o: till •ccv^-.C'up-i ;v. Caitoi-j, Ja.y iAzic :•>".:: 2 d:-i ^e a i:e Rc^n 12 Lafco-.a^v-.:/. T^.e ^:-^i.
C i^^-xtczc r.ci "
Methods of Testing for {Tiardia in '.'.ater (George i'Jay .sccnceios, F.e;:;r.^i
Micrcbiologis:. Region 10 Laj:r-:::
Manchester, '.'.ashingtcnj
Background:
Although recent development of an excystation technique by Drs. Bin^har,
Meyer, Rice and Schaefer could in future lead to developing cultural .-nerr.cci,
at this tir.e nn leliable methods exist for culturing Ciardia cysts iror, -,>a:e:
sairples. At present, the only practical method for determining the presence
of cysts in water is by direct r.icroscopic examination cf sartple concentrate:
Microscopic detection in water-sample concentrates isn't an ideal process.
Finding and identifying the cysts relies almost entirely on the training,
skill, experience and persistence of the examiner. (And it is a skill not
widespread ar.ong water-supply laboratories.) But despite its limitations,
microscopic identification is currently the best method ue have.
Years ago, the basic assumption was made that in order to find Giardia cv~:5
in water, some form of sample concentration was necessary. .As early as l.:3^,
labs ^ere using membrane filters with a porosity of 0.45 jjr.. V.ith :e.. e^ceptirns,
these attempts uere unsuccessful. The center for Disease Centre! has :r:e-
partic^late filtration, with diatomaceous earth as the '..odium. This re:"\ed
the cvsis from the water, but the cysts couldn't be separated from the
particles of diatomaceous earth.
'.••ith the recer.t increase in the incidence of waterbome giardiasis, furtr.er
efforts nave seen made to improve the detection method. .An ideal metr.c,- •.,;-!-
be one mat recovers all cysts in a water sample rapidly, cheaply and si~ly;
allows rapid detection, identification and quantification; and provides
information on the viability of and/or infectivity potential cf cysts detected..
Unfortunately, no such method exists. The methods presently available
can be broadly separated into two general stages: primary concentration a:
processing (see Table 1 on next page), and detection and identification
(see Table 2 on next page).
-------
Methods of Testing for Ciarciia in A'ater (Continued...)
HETHC3
"e-norare
Ctl iulenc
PelyCdrsinjte
!293:afli-5uni)
P«rv:.i|ie *i l t.-iti
vauti-
etc.)
f; ana
-S?nS. 1956
'/per, OuFfjin 4 nenry
1562, i^njuo'isnea)
Sna« et al. 197?
T.T979
et.jj.. 1983
3re*er, wriqnt Sute UN.
oi isneoj
. C3HS no,
l isneo)
MSLLTS
1 91! /mm ?
i: PS:. i5-;ac3 3*1
ICtii .
Generjlly 5002
Out poor eUm
Gooa '43ij recovery,
use.
Generally uiisucesjf.
, CA Ovtriil recovery 20'
y':-:;;-o.s *a"'woytn Seam.
• • ':e-s 196C, EPA-Ci
, 7 i ».n srlsn i pol>5rs!yiene)
1979 & Recovery 3-15i
£urac:icn ave. 531
"i!' '.oore Csrj.
.npyoii»n»a;
lay oe '.sefyi 'sr
•jsnings
3'.«a''«, j. of «asrt., 1982 C'.a-irs 75i -ecsve^/
;«"S.sl isnee) ?rsis srlcr f :ers
TAI
HETHQD
Itwunofluorescen
11535, CSCn$ LID, Berkley. CA
1933
R-S'J.'S
Gooa prep.. Cross
;FA
S«ucn, E?».Ctncinn»t1
Sti11 unaer stuOy
S
-------
Methods of Testing for Giardia in Water ("Continued...';
Copies of Table 1 and Table 2 are also shown in Appendix C, along with
further detail about the methods.
EPA Consensus Method:
In September, 198C, the EPA convened a workshop en Giardia methodology in
Cincinnati. Its main purpose was to identify the best available rr.etr.cdclogy,
and to agree on a reference method. The five labs in attendance reco^nne-
that any proposed method would be based in large part on opinions and personal
preferences rather than on hard data, but that agreeing on a consensus r.e--.;c
would promote uniformity and provide a basis for future comparisons. Cur
lab has modified the EPA consensus method slightly for our use. This method
is outlined below.
Filter unwound into quarters
I
Rinsed in distilled water with polysorbate 20
I
Settled overnight, or centrifuged
i
Collect sediment and add 2' Formaldehyde in PBS
i
Settled overnight, or centriruged
Collect sediment
i
Sucrose or . „
Percoll-sucrose
gradient
^licroscopic observation of the entire
concentrate (Bright fie Id/Phase- contrast)
-------
AP?EM'I\
CGN'CENTRATINC, PROCESSING, DETECTING AND IDENTIFY I \G
riARTTA CYSTS IN '.•,'ATER
Tk
ac^-iou/id -t /:;(:• -via -tec/: iacoc-x
fe, "Mctni'ii ;•,' Teittr.g fo-i G-c^iica -c>i
c-Ki, a^d •';-. a;: :'u.t£-o:c o-' the mcd<.«<.id E?A C-
J W *. trf
-------
£MDIXC: CONCEMTRATING, P?OCESS!VJG
nE'ECTING AMD IDENTIFYING GIAPHIA 'CYSTS 'N WATE7
"ETHQD
PJVFS'IGATOP (S)
P.£SULTS
DO! vcirM nate
'293mm-Sum)
2. Participate Filtration
(diatomaceous earth,
etc. )
Alga* (Foe^st) C
4. Aninnic and Cationic
Exchange Resins
Ralston
Chang and Kabler
DyDer, DuFrain and Henry Eng
19«?, (unpublished)
Shaw et al , 1977
Tube F1' 1 ters
Holman et a]_, 1983
DHHS, VJaTliington
Brewer, Wright State UN.
(unpublished)
Riggs, CDHS Lab, Berkley, CA
(unoublished)
Generally unsuccessful
Passing 1 gal/mi o a
10 PSI. 15-1800 gal
total
Generally gooc1
but poor eluation
Good rapid recovery,
but 1 i mi tod in fie! d
use
Generally unsuccessful
Overall recovery 20-80
percent
C" -i i t o r 2
'" anr( ijn orlon anc^
00"! voro1 vl ena)
^ssette
F i Tterwashing Apparatus
J?i-jjbowski, Erickson, 1979 and
1980, EPA-Cincinnati
Milliner* Corp.
(unpuM ish
DuWalle, U. cf Wash., 1982
i sh*d)
Recovery 3-15 oe^c
Extraction ave. 38
percent
'lay be useful -or
process:ng fi1ter
v/ashings
Claims "5 percent
recovery fron orl
filters"
TABLE 1
-------
SETECTI^G AND :DE.kr:-Y?jG GIASDIA CY$~ IN V
PRIMARY CONCENTRATION AMP PROCESSES METHODS
1. MEV3RAME FILTER (MH METHODS
a. rqiulosic (!m'xed esters of ce1
Cham and fabler ;n 1955
First to use '-'F for cyst recover/. Recovered 20-42 oercent 3t
concentration of 3, 5, and 10 cyst/gal. - nn c>'St *oun-t at
\ cyst/gal.
M»tnrM was used 1-n 1965 Colorado outbreak 'Moore, _e_t aj_, 1959)
? liter size water samples from 10 sites. No cysts~~wei"e de*2:
Use i* cellulosic filters have generally not been successfu1 *
Demonstrating cysts vn drinkinq w^ter.
Polycarbonate
1. Luchtel and folleaqes {n I960 used 293 mm, 5.0 jm pore size
nudeooor-? f»C) filters to concentrat0 fomali n-fixe^. G.
c.vsts fron 20 L tap water samples. Recovery rates o£ ap5"ro
75 percent were reoortel.
2. pyper of OuFrain ^nd Henry Engineers c^in good recovery witn
nucleoporp filt?r at =» Tow rate of 1 gal. /mi n., not cv-sr II
oassing 15-1800 gal. in iust over 24 hours.
Even with these claims v ^j^^ and Luchte1, the 'IF '1ethod h^s j"'
(Asoen, 19^^) been successful in demonstrating cysts in water--;--:
Decause:
1. Inability to process a sirf";cient volume.
2. Inability to remove "ysts from M^ter.
"?. Cysts weren't present at time of sampling during cr afte" c>t
a. S°.t|f) - CDC (Shaw, 1977) used high-vol filtration through
TarF filter (280,000 qal. tota^ove" 10 days) - was backf1ush3^ --.: '
gal. drums and coagulated w/alum. Concentration fed to beagle D..::'5;
an^ after treatment (chees^oth to wire screening to 30 »m '•'•' t:
centrifuge) was examined microsconically. First time cysts obs«srvr: •
water suooly after concentration.
b. niatomaceous earth (HE) - CDC iJuranek, 1979) used DE to remove :-s:;
fro- seeded '.vat er. PFob^m was fat cysts couldn't be removes •'-:- ~~
Darticles. 3rewer (1983) claims 5.2-31.1 percent recovery frcri I:
backwash. Retention thro'ign 3 'o"*^ (celit'j ^05, HyFl o-Sucer:e' j- •
c°lite 5^0) at cyst concentration Banging from 6-15,000 cyst/^. : ' •
rjnge hetweon 66-100 percent.
-------
, PROCESS IMP,,
IDENTIFYING GIARHIA CYSTS IN UAT-
V.GAE CENTRIFUGE
a. '/.'as found to recover more cysts (10X) than a series of MF-f liters and
nylon screens: 3 vs. 1 day by MF.
b. Mav be impractical in field because of power requirement.
c. If 'jsed in lab, 1 large single sample collected in the field could niss
cyst.
* . May *ind application for concentration cysts from orlon filter washings.
ArO'lIC AMD CATIOUIC EXCHANGE RESIMS (Brewer - unpublished)
a. ^ased on hypothesis t*at cysts could be attracted to charged surfaces,
cvsts have a charg0 of approximately 25mV at pH 5.5 which increases in
3lectro-neqativity as the pH rises to 8.0.
b. Charge attraction techniques have been used for concentration of both
bacteria and viruses in water.
c. Five °xchanQe resins were tested:
(1. 4° percent recovery fron anionic Oowex 1-XY colunns
(2. 38 percent recovery from cationic Dowex 50W-X8 columns
*. Conoared to parallel tests w/diatomaceous earth, exchange resins 'ess
qcficient in retention.
RALC'QH E?Oyv-FIBF.RGLAS$ TUBE FILTERS
a. Riggs of CSHD, Viral and Rick. Lab., can ^ilter 500 gallons dr^'nXT-g
water thru VO" - 9 wn Balston tube filter.
b. Backflushes w/1 L 3 percent beef extract or solution of 0.5 percent
potassiun citrate.
c. Concentration is centrifuged w/40 oercent potassium citrate and -id-' 2
layer filtered thru *> u polycarbonate filters.
d. Uses direct irriunofluorescence antibody technique for detection and
identification.
e. Claims ?0-30 percent efficiency in collection, coprocessing and ::.
YARN'-IOVEN nEPTH FILTERS
a. In 1975 E°A develooed a concentration-extraction method invoiv
volumes of water thru .Tiicroporous yarnv/oven orlon-fiber fil'ers
b. This nethod has been tenatively adopted as the "method of chc-:-:
concentrating cysts from water Sij
-------
0£TECTIV."3 AND ^ENT
, PROCESSING,
GIA^OIA CYS'S IN '.'ATE*
C4;
7.
8.
c. Since initial studies which showed onlv 3-15 percent recovery wi tn a mean
o* 6.3 percent an* a 53 percent extraction rate, several chang2s have Deen
made which may have increase;! the retention rate to >20 percent.
1. Gone from 7 to 1 urn porsity filter
2. Limited the ra*e of flow to 1/2 gallon/min
3. Limited the pressure head to 10 PSI
4. Have gone to po
-------
AMD IDENTIFYING dlARPIA CYSTS IN W.VER
HETECTIOM METHODS
l.a. niRECT CLUORESCEMT ANTIBODY (QRA^ TECHNIQUE
1. Riggs has oroduced a high titer purified inrnune ssra to Gi'ardi'a lamblia
cysts in guinea pigs and labeled it with Huorecein isotfiio cyanate" iera
is puri'ied thru NfyOH and DEAE sefadex fractionation.
2. Obtained cross reactions with Chilonastix mesnili cysts but claims it can
be easily distinguished from Giardia by its smaller size.
i.b. INDIRECT FLUORESCENT ANYBODY (IFA) TECHNIQUE
1. Sauch using IFA with immune sera from rabbits (unourif ieri) . It is reacted
with commercially available fluorescent-labeled goat anti-rabbit gamna
globulin.
2. Some cross-reactions with certain algal cells.
I.e. MONOCLONAL AMTIROOIES
1. Using clones of hybridoma cell lines obtained by fusing mouse myeloma
cells with sn^en cells from mice (BALB/c) immunized with G. lamsi ia
troohozoites.
2. ^ro^uced eight monoclonal antibodies evaluated by IFA against 3oth trcnns
3PH cysts.
a. 3/8 stained the ventral d-'sk
b. 2 staine^ the nuclei
c. 2 stained cytoplasm^'c granules
ri. 2 stained membrane comnon^nts
3. Variaoility in staining may be due to differences in stages of eicystrer.t.
4. Preliminary results ind;>;ate nonoclonal ABs may give rapid and specific ID
o* cysts.
5. Rx may be too specific, not reacting with all human forms of G_. lanbli
may have to go to polyclonal
?.. EL ISA
a. Hungar at John Hopkins (unpublished) has produced a Detection method by
ELISA using a intact "sandwich" technique in 96-wel1 microtiter olatos.
b. Using antisera from 2 different animals (may present problen).
c. He«d a minimum of 12 cysts/we!1 for color Rx .
-------
APPENDIX B
INSTITUTIONAL CONTROL OF LEGIONELLA
-------
APPENDIX B
INSTITUTIONAL CONTROL OF LEGIQNELLA
LegionelU is a genus name for bacteria commonly found in lake and
river waters. Some species of this genus have been identified as the
cause of the disease legionellosis. In particular, Legione 11 a pneumophila
has been identified as the cause of Legionnaires disease, the pneumonia
form of legionellosis and with Pontiac Fever, a nonpneumonia disease.
Outbreaks of legionellosis are primarily associated with inhalation of
water aerosols or, less commonly, with drinking water containing
Legionella bacteria with specific virulence factors not yet identified.
Foodborne outbreaks have not been reported (USEPA, 1985).
As discussed in this document, treatment requirements for disinfec-
tion of a municipal water supply are thought to provide at least a 3 log
reduction of Legionella bacteria (see Section 3.2.2). However, some
recontamination may occur in the distribution system due to cross
connections and during installation and repair of water mains. It has
been hypothesized that the low concentrations of Leoionella entering
buildings due to these sources may colonize and regrow in hot water
systems (USEPA, 1985). Although all of the criteria required for
colonization are not known, large institutions, such as hospitals, hotels,
and public buildings with recirculating hot water systems seem to be the
most susceptible. The control of LegloneVIa in health care institutions,
such as hospitals, is particularly important due to the increased
susceptibility of many of the patients. The colonization and growth of
Legionella in drinking water primarily occurs within the consumer's
plumbing systems after the water leaves the distribution system.
Therefore, the control of these organisms must be the consumer's
responsibility. This appendix is intended to provide guidance to these
institutions for the detection and control of the Legionella bacteria.
B.I MONITORING
It is suggested that hospitals, and other institutions *un
potential for the growth of Legionella. conduct routine monitoring of
8-1
-------
their hot water systems at least quarterly.1 The analytical procedures
for the detection of these organisms can be found in Section 912.1
"LeglQne11a.ceag" of the 16th edition of Standard Methods. Samples should
be taken at, or closely following, the hot water storage reservoir and
from a number of shower heads. It is recommended that showers with the
least frequent usage be included in the sampling program. Follow-up
testing is suggested for all positive indications prior to the initiation
of any remedial measures. If the the presence of Leoionella is confirmed,
then remedial measures should be taken. Although the regrowth of
Lec-ionella is commonly associated with hot water systems, hot and cold
water interconnections may provide a pathway for cross contamination. For
this reason, systems detecting Leoionella in hot water systems should also
monitor their cold water systems.
B.2 TREATMENT
Because the primary route of exposure to Legionella is probably
inhalation, rather than ingestion, it is recommended that disinfection
procedures include an initial shock treatment period to disinfect shower
heads and hot water taps where the bacteria may colonize and later become
airborne. The shock treatment period should also include disinfection of
hot water tanks. After this time, a point-of entry treatment system can
be installed to provide continual disinfection of the hot water system.
B-2.1 Initial Disinfection
The most applicable method for the initial disinfection of shower
heads and water taps is heat eradication. The fittings can be removed and
held at temperatures greater than 60 C for at least 24 hours. Disinfec-
tion of fittings can also be achieved by soaking or rinsing with a strong
chlorine solution. When soaking the fittings, a minimum chlorine strength
of 50 mg/L should be used for a period of no less than 3 hours. Rinsing
Monitoring frequency based on the reported rate of Legionella regrowth observed
during disinfection studies (USEPA, 1985).
B-2
-------
with chlorine should be performed with more concentrated solutions. Care
must be taken not to corrode the finished surface on the fittings.
Commercially available bleaches, for example, are typically 5.25 percent
chlorine by weight.
B.2.2 Long-Term Disinfection
Heai - Numerous studies have shown that increasing the hot water
temperature to 50 - 70 C over a period of several hours may help to reduce
and inhibit Legionella populations. However, some instances of regrowth
after 3 to 6 months have been reported. In these cases, the authors have
concluded that a periodic schedule of short-term temperature elevation in
the hot water may be an effective control against legionellosis (USEPA,
1985; Muraca, 1986). Disinfection by this method also requires periodic
flushing of faucets and shower heads with hot water. Although heat
eradication is easily implemented and relatively inexpensive, a disadvan-
tage is the potential need for periodic disinfection. The potential for
scalding from the unusually hot water also exists (USEPA, 1985; Muraca, et
al. 1986).
Chlorination - Several studies have suggested that a free chlorine
residual of 4 mg/L will eradicate Legionellq growth. There is, however,
a possibility for recontamination in areas of the system where the
chlorine residual drops below this level. A stringent monitoring program
is therefore required to ensure that the proper residual is maintained
throughout the system and under varying flow conditions. It may also be
necessary to apply a large initial chlorine dose to maintain the 4 mg/L
residual. This may cause problems of pipe corrosion and, depending on
water quality, high levels of trihalomethanes (THMs).
Ozone - Ozone is the most powerful oxidant used in the potable
water industry. One study indicated that an ozone dosage of 1 to 2 mg/L
was sufficient to provide a 5 log reduction of legionella (Muraca, et al.
1986). Ozone is generated by passing a high voltage current of electrici-
ty through a stream of dry air or oxygen. The use of high voltage
electricity requires proper handling to avoid creating hazardous
conditions. The ozone is applied by bubbling the ozone containing gas
through the water in a chamber called a contactor.
B-3
-------
One of the disadvantages of this system is its complexity. it
requires a dry air or oxygen source, a generator, and a contactor sized to
provide 2 to 5 minutes of contact time and an ambient ozone monitor. All
materials in contact with the ozone must be constructed of special ozone
resistant materials to prevent leakage. Leak detection is also required
because of the toxic nature of ozone and possible explosive conditions if
pure oxygen is used for generation.
Another disadvantage of ozonation is the rapid decomposition of
ozone residuals. The half-life of ozone in drinking water is typically
around 10 minutes. This makes it difficult, if not impossible, to
maintain a residual throughout the water system and may require the use of
a supplementary disinfectant such as chlorine or heat. For these reasons
it is not thought that ozonation is viable for institutional applications.
Ultraviolet Irradiation - Ultraviolet (UV) light, in the 254
nanometer wavelength range can be used as a disinfectant. UV systems
typically contain low-pressure mercury vapor lamps to maximize output in
the 254 nm range. Water entering the unit passes through a clear cylinder
while the lamp is on, exposing bacteria to the UV light. Because UV light
can not pass through ordinary window glass, special glass or quartz
sleeves are used to assure adequate exposure.
The intensity of UV irradiation is measured in microwatt-seconds per
square centimeter (uW-s/cm2). Several studies have shown a 90 percent
reduction of Legionella with a UV dosage of 1000 - 3000 uW-s/cm2, compared
to 2000 to 5000 uW-s/cm2 for E. coli. Salmonella and Pseudomonas (USEPA,
1985). In another study, a 5 log reduction of Legionella was achieved at
30,000 uW-s/cm2; and the reduction was more rapid than with both ozone and
chlorine disinfection (Muraca, et al. 1986).
The major advantage of UV disinfection is that it does not require
the addition of chemicals. This eliminates the storage and feed problems
associated with the use of chlorine, chlorine dioxide and chloramines. In
addition, the only maintenance required is periodic cleaning of the quartz
sleeve and replacement of bulbs. UV monitors are available which measure
the light intensity reaching the water and provides a signal to the user
when maintenance is required. These monitors are strongly suggested for
any application of UV irradiation for disinfection. It should be noted,
3-4
-------
however, that these monitors measure light intensity which may not be
directly related to disinfection efficiency. The UV lamps should
therefore not be operated past the manufacturers use rating even with a
continuous UV monitor installed.
Another disadvantage of UV disinfection, as with ozonation, is that
a residual is not provided. A supplementary disinfectant may therefore be
required to provide protection throughout the system. In addition-,
turbidity may interfere with UV disinfection by blocking the passage of
light to the microorganisms.
B.3 OTHER CONTROL METHODS
In addition to chemical and heat disinfection, there are system
modifications which can be made to inhibit Legionella growth. Many
institutions have large hot water tanks heated by coils located midway in
the tank. This type of design may result in areas near the bottom of the
tank which are not hot enough to kill Leoionella. Designing tanks for
more even distribution of heat may help limit bacterial colonization. In
addition, sediment build-up in the bottom of storage tanks provides a
surface for colonization. Periodic draining and cleaning may therefore
help control growth. Additionally, other studies have found that hot
water systems with stand-by hot water tanks used for meeting peak demands,
still tested positive for Legione11 a despite using elevated temperature
(55 C) and chlorination (2 ppm) (Fisher-Hoch, et al. 1984.) Stringent
procedures for the cleaning, disinfection and monitoring of these stagnant
tanks should be set up and followed on a regular basis.
In another study, it was reported that black rubber washers and
gaskets supported Legione11 a growth by providing habitats protected from
heat and chlorine. It was found, after replacement of the black rubber
washers with Proteus 80 compound washers, that it was not possible to
detect Leqionella from any of the fixtures (Colbourne, et al. 1984).
B-5
-------
B.4 CONCLUSIONS
LeQionella bacteria have been identified as the cause of the disease
legionellosis, of which the most serious form is Legionnaires Disease.
Although conventional water treatment practices are sufficient to provide
disinfection of Legionella. regrowth in buildings with large hot water
heaters, and especially with recirculating hot water systems, is a
significant problem. This problem is of particular concern to health care
institutions, such as hospitals, where patients may be more susceptible to
the disease.
This guideline suggests a program of 'quarterly monitoring for
Legionella. If the monitoring program suggests a potential problem with
these organisms, a two stage disinfection program is suggested consisting
of an initial period of shock treatment followed by long term disinfec-
tion.
Four methods of disinfection for the control of Legionella were
presented in this appendix; heat, chlorination, ozonation, and ultraviolet
irradiation. All four of the methods have proven effective in killing
Legionella. Ultraviolet irradiation and heat eradication are the
suggested methods of disinfection due, primarily, to advantages in
monitoring and maintenance. However, site specific factors may make
chlorination or ozonation more feasible for certain applications. In
addition, it is recommended that all outlets, fixtures and shower heads be
inspected and all black rubber washers and gaskets replaced with materials
which do not support the growth of Legione11 a organisms.
One problem associated with the application of point-of-entry
treatment systems is the lack of an approved program for certifying
performance claims. However, the National Sanitation Foundation (NSF),
Ann Arbor, MI an unofficial, non-profit organization, does have a testing
program to verify disinfection efficiencies and materials of construction.
Certification by the NSF, or other equivalent organizations, is desirable
when selecting a treatment system.
B-6
-------
References
Colbourne, J.; Smith, M. G.; Fisher-Hoch, S. P. and Harper, D. Source of
Legionella pneumophila Infection in a Hospital Hot Water System:
Materials Used in Water Fittings Capable of Supporting L. pneumophila
Growth. In: Thornsberry, C.; Balows, A.; Feeley, J. C. and Jakubowski,
w. Legione 11 a - Proceedings of the 2nd International Symposium. American
Society for Microbiology, pp. 305-307, 1984.
Fisher-Hoch, S. P.; Smith, M.G.; Harper, D. and Colbourne, J. Source of
Legionella pneumonia in a Hospital Hot Water System, pp. 302-304 in
Thornsberry, C.; Balows, A.; Feeley, J.C. and Jakubowski, W. Legionella
Proceedings of the 2nd International Symposium, American Society for
Microbiology, pp. 302-304, 1984.
Muraca, P.; Stout, J. E. and Yu, V. L. Comparative Assessment of
Chlorine, Heat, Ozone, and UV Light for Killing Lepionella pneumophila
Within a Model Plumbing System. Appl. Environ. Microbiol. 53(2):447-453,
1986.
U.S. Environmental Protection Agency, Office of Drinking Water. Control
of Legionella in Plumbing Systems, Health Advisory (1985).
B-7
-------
APPENDIX C
DETERMINATION OF DISINFECTANT
CONTACT TIME
-------
APPENDIX C
DETERMINATION OF DISINFECTANT CONTACT TIME
As indicated in Section 3, for pipelines, all fluid passing through
the pipe is assumed to have a detention time equal to the theoretical or
mean residence time at a particular flow rate. However, in mixing basins,
storage reservoirs, and other treatment plant process units, utilities
will be required to determine the contact time for the calculation of CT
through tracer studies or other methods approved by the Primacy Agency.
For the purpose of determining compliance with the disinfection
requirements of the SWTR, the contact time of mixing basins and storage
reservoirs used in calculating CT should be the detention time at which 90
percent of the water passing through the unit is retained within the
basin. This detention time was designated as Tlo according to the
convention adopted by Thirumurthi (1969). A profile of the flow through
the basin over time can be generated by tracer studies. Information
provided by these studies is used for estimating the detention time, T,0,
for the purpose of calculating CT.
This appendix is divided into two sections. The first section
presents a brief synopsis of tracer study methods, procedures, and data
evaluation. In addition, examples are presented for conducting hypo-
thetical tracer studies to determine the T10 contact time in a clearwell.
The second section presents a method of determining T10 from theoretical
detention times in systems where it is impractical to conduct tracer
studies.
C.I Tracer Studies
C.I.I Flow conditions
Although detention time is proportional to flow, it is not generally
a linear function. Therefore, tracer studies are needed to establish
detention times for the range of flow rates experienced within each
disinfectant section.
As discussed in Section 3.2, a single flow rate may not characterize
the flow through the entire system. With a series of reservoirs,
C-l
-------
cleanvells, and storage tanks flow will vary between each portion of the
system.
In filter plants, the plant flow is relatively uniform from the
intake through the filters. An increase or reduction in the intake
pumping capacity will impart a proportional change in flow through each
process unit prior to and including the filters. Therefore, at a constant
intake pumping rate flow variations between disinfectant sections within
a treatment plant, excluding clear-wells, are likely to be small, and the
the design capacity of the plant, or plant flow, can be considered the
nominal flow rate through each individual process unit within the plant.
Clearwells may operate at a different flow rate than the rest of the
plant, depending on the pumping capacity.
Ideally, tracer tests should be performed for at least four flow
rates that span the entire range of flow for the section being tested.
The flow rates should be separated by approximately equal intervals to
span the range of operation, with one near average flow, two greater than
average, and one less than average flow. The flows should also be
selected so that the highest test flow rate is at leaste 91 percent of the
highest flow rate expected to ever occur in that section. Four data
points will assure a good definition of the section's hydraulic profile.
The results of the tracer tests performed for different flow rates
should be used to generate plots of T10 vs. Q for each section in the
system. A smooth line is drawn through the points on each graph to create
a curve from which T10 may be read for the corresponding Q at peak hourly
flow conditions. This procedure is presented in Section C.I.8.
It may not be practical for all systems to conduct studies at four
flow rates. The number of tracer tests that are practical to conduct is
dependent on site-specific restrictions and resources available to the
system. Systems with limited resources can conduct a minimum of one
tracer test for each disinfectant section at a flow rate of not less than
91 percent of the highest flow rate experienced at that section. If only
one tracer test is performed, the detention time determined by the test
may be used to provide a conservative estimate in CT calculations for that
section for all flow rates less than or equal to the tracer test flow
rate. T10 is inversely proportional to flow rate, therefore, the T10 at a
C-2
-------
flow rate other than that which the tracer study was conducted (Tlos) can
be determined by multiplying the T10 from the tracer study (T,OT) by the
ratio of the tracer study flow rate to the desired flow rate, i.e.,
Tios * TIOT x QT/OD where
T!l3S » T,3 at system flow rate
Tior s Tio at tracer flow rate
QT » tracer study flow rate
Q0 » system flow rate
The most accurate tracer test results are obtained when flow is
constant through the section during the course of the test. Therefore,
the tracer study should be conducted at a constant flow whenever
practical. For a treatment plant consisting of two or more equivalent
process trains, a constant flow tracer test can be performed on a section
of the plant by holding the flow through one of the trains constant while
operating the parallel train(s) to absorb any flow variations. Flow
variations during tracer tests in systems without parallel trains or with
single clearwells and storage reservoirs are more difficult to avoid. In
these instances, Tlo should be recorded at the average flow rate over the
course of the test.
C.I.2 Other Tracer Studv Considerations
In addition to flow conditions, detention times determined by tracer
studies are dependent on the water level in the contact basin. This is
particularly pertinent to storage tanks, reservoirs, and clear-wells which,
in addition to being contact basins for disinfection are also often used
as equalization storage for distribution system demands. In such
instances, the water levels in the reservoirs vary to meet the system
demands. The actual detention time of these contact basins will also vary
depending on whether they are emptying or filling.
For some process units, especially sedimentation basins which are
operated at a near constant level, that is, flow in equals flow out, the
detention time determined by tracer tests is valid for calculating CT when
the basin is operating at water levels greater than or equal to the level
at which the test was performed. If the water level during testing is
C-3
-------
higher than the normal operating level, the resulting concentration
profile will predict an erroneously high detention time. Conversely,
extremely low water levels during testing may lead to an overly conserva-
tive detention time. Therefore, when conducting a tracer study to
determine the detention time, a water level at or slightly below, but not
above, the normal minimum operating level is recommended.
For many plants, the water level in a clear-well or storage tank
varies between high and low levels in response to distribution system
demands. In such instances, in order to obtain a conservative estimate of
the contact time, the tracer study should be conducted during a period
when the tank level is falling (flow out greater than flow in). This
procedure will provide a detention time for the contact basin which is
also valid when the water level is rising (flow out less than flow in)
from a level which is at or above the level when the T10 was determined by
the tracer study. Whether the water level is constant or variable, the
tracer study for each section should be repeated for several different
flows, as described in the previous section.
For clearwells which are operated with extreme variations in water
level, maintaining a CT to comply with inactivation requirements may be
impractical. Under such operating conditions, a reliable detention time
is not provided for disinfection. However, the system may install a weir
to ensure a minimum water level and provide a 'reliable detention time.
Systems comprised of storage reservoirs that experience seasonal
variations in water levels may perform tracer studies during the various
seasonal conditions. For these systems, tracer tests should be conducted
at several flow rates and representative water levels that occur for each
seasonal condition. The results of these tests can be used to develop
hydraulic profiles of the reservoir for each water level. These profiles
can be plotted on the same axis of T10 vs. Q and may be used for calculat-
ing CT for different water levels and flow rates.
Detention time may also be influenced by differences in water
temperature within the system. For plants with potential for thermal
stratification, additional tracer studies are suggested under the various
seasonal conditions which are likely to occur. The contact times
determined by the tracer studies under the various seasonal conditions
C-4
-------
should remain valid as long as no physical changes are made to the mixing
basin(s) or storage reservoir(s).
As defined in Section 3.2.2, the portion of the system with a
measurable contact time between two points of disinfection or residual
monitoring is referred to as a section. For systems which apply
disinfectant(s) at more than one point, or choose to profile the residual
from one point of application, tracer studies should be conducted to
determine T10 for each section containing process unit(s). The T10 for a
section may or may not include a length of pipe and is used along with the
residual disinfectant concentration prior to the next disinfectant appli-
cation or monitoring point to determine the CTCJ|C for that section. The
inactivation ratio for the section is then determined. The total
inactivation and log inactivation achieved in the system can then be
determined by summing the inactivation ratios for all sections as
explained in Section 3.2.2.
For systems that have two or more units of identical size and
configuration, tracer studies only need to be conducted on one of the
units. The resulting graph of T10 vs. flow can be used to determine T,.
for all identical units.
Systems with more than one section in the treatment plant may
determine T10 for each section
by individual tracer studies through each section, or
by one tracer study across the system
If possible, tracer studies should be conducted on each section to
determine the T10 for each section. In order to minimize ihe time needed
to conduct studies on each section, the tracer studies should be started
at the last section of the treatment train prior to the first customer and
completed with the first section of the system. Conducting the tracer
studies in this order will prevent the interference of residual tracer
material with subsequent studies.
However, it may not always be practical for systems to conduct
tracer studies for each section because of time and manpower constraints.
In these cases, one tracer study may be used to determine the T values
C-5
-------
for all of the sections at one flow rate. This procedure involves the
following steps:
1. Add tracer at the beginning of the furthest upstream disinfec-
tion section.
2. Measure the tracer concentration at the end of each disinfec-
tion section.
3. Determine the T10 to each monitoring point as outlined in the
data evaluation examples presented in Section C.I.7.
4. Subtract Tlo values of each of the upstream sections from the
overall Tlo value to determine the Tlo of each downstream
section.
This approach is valid for a series of two or more consecutive
sections as long as all process units within the sections experience the
same flow condition. This approach is illustrated by Hudson (1975) in
which step-dose tracer tests were employed to evaluate the baffling
characteristics of flocculators and settling basins at six water treatment
plants. At one plant, tracer chemical was added to the rapid mix, which
represented the beginning of the furthest upstream disinfection section in
the system. Samples were collected from the flocculator and settling
basin outlets and analyzed to determine the residence-time characteristics
for each section. Tracer measurements at the flocculator outlet indicated
an approximate T10 of 5 minutes through the rapid mix, interbasin piping
and flocculator. Based on tracer concentration monitoring at the settling
basin outlet, an approximate T10 of 70 minutes was determined for the
combined sections, including the rapid mix, interbasin piping, floccu-
lator, and settling basin. The flocculator T10 of 5 minutes was subtracted
from the combined sections' T10 of 70 minutes, to determine the T,0 for the
settling basin alone, 65 minutes.
This approach may also be applied in cases where disinfectant
application and/or residual monitoring is discontinued at any point
between two or more sections with known T10 values. These Tlo values may
be summed to obtain an equivalent T]0 for the combined sections.
For ozone contactors, flocculators or any basin containing mixing,
tracer studies should be conducted for the range of mixing used in the
C-6
-------
process. In ozone contactors, air or oxygen should be added in lieu of
ozone to prevent degradation of the tracer. The flow rate of air or
oxygen used for the contactor should be applied during the study to
simulate actual operation. Tracer studies should then be conducted at
several air/oxygen to water ratios to provide data for the complete range
of ratios used at the plant. For flocculators, tracer studies should be
conducted for various mixing intensities to provide data for the complete
range of operations.
C.I.3 Tracer Study Methods
This section discusses the two most common methods of tracer
addition employed in water treatment evaluations, the step-dose method and
the slug-dose method. Tracer study methods involve the application of
chemical dosages to a system and tracking the resulting effluent
concentration as a function of time. The effluent concentration profile
is evaluated to determine the detention time, Tlo.
While both tracer test methods can use the same tracer materials and
involve measuring the concentration of tracer with time, each has distinct
advantages and disadvantages with respect to tracer addition procedures
and analysis of results.
The step-dose method entails introduction of a tracer chemical at a
constant dosage until the concentration at the desired end point reaches
a steady-state level. Step-dose tracer studies are frequently employed in
drinking water applications for the following reasons:
the resulting normalized concentration vs. time profile is
directly used to determine, T10, the detention time required
for calculating CT
very often, the necessary feed equipment is available to
provide a constant rate of application of the tracer chemical
One other advantage of the step-dose method is that the data may be
verified by comparing the concentration versus elapsed time profile for
samples collected at the start of dosing with the profile obtained when
the tracer feed is discontinued.
C-7
-------
Alternatively, with the slug-dose method, a large instantaneous dose
of tracer is added to the incoming water and samples are taken at the exit
of the unit over time as the tracer passes through the unit. A disadvan-
tage of this technique is that very concentrated solutions are needed for
the dose in order to adequately define the concentration versus time
profile. Intensive mixing is therefore required to minimize potential
density-current effects and to obtain a uniform distribution of the
instantaneous tracer dose across the basin. This is inherently difficult
under water flow conditions often existing at inlets to basins. Other
disadvantages of using the slug-dose method include:
the concentration and volume of the instantaneous tracer dose
must be carefully computed to provide an adequate tracer
profile at the effluent of the basin
the resulting concentration vs. time profile cannot be used to
directly determine T,0 without further manipulation
a mass balance on the treatment section is required to
determine whether the tracer was completely recovered
One advantage of this method is that it may be applied where
chemical feed equipment is not available at the desired point of addition,
or where the equipment available does not have the capacity to provide the
necessary concentration of the chosen tracer chemical. Although, in
general, the step-dose procedure offers the greatest simplicity, both
methods are theoretically equivalent for determining T,0. Either method
is acceptable for conducting drinking water tracer studies, and the choice
of the method may be determined by site-specific constraints or the
system's experience.
C.I.4 Tracer Selection
An important step in any tracer study is the selection of a chemical
to be used as the tracer. Ideally, the selected tracer chemical should be
readily available, conservative (that is, not consumed or removed during
treatment), easily monitored, and acceptable for use in potable water sup-
plies. Historically, many chemicals have been used in tracer studies that
do not satisfy all of these criteria, including potassium permanganate,
alum, chlorine, and sodium carbonate. However, chloride and fluoride are
C-8
-------
the most common tracer chemicals employed in drinking water plants that
are nontoxic and approved for potable water use. Rhodamine WT can be used
as a fluorescent tracer in water flow studies in accordance with the
following guidelines:
Raw water concentrations should be limited to a maximum
concentration of 10 mg/L.
Drinking water concentations should not exceed 0.1 ug/L.
Studies which results in human exposure to the dye must be
brief and infrequent.
Concentrations as low as 2 ug/L can be used in tracer studies
because of the low detection level in the range of 0.1 to 0.2
ug/L.
The use of Rhodamine B as a tracer in water flow studies is not recom-
mended by the EPA.
The choice of a tracer chemical can be made based, in part, on the
selected dosing method and also on the availability of chemical feeding
equipment. For example, the high density of concentrated salt solutions
and their potential for inducing density currents, usually precludes
chloride and fluoride as the selected chemical for slug-dose tracer tests.
Fluoride can be a convenient tracer chemical for step-dose tracer
tests of clearwells because it is frequently applied for finished water
treatment. However, when fluoride is used in tracer tests on clarifiers,
allowances should be made for fluoride that is absorbed on floe and
settles out of water (Hudson, 1975). Additional considerations when using
fluoride in tracer studies include:
it is difficult to detect at low levels
many states impose a finished water limitation of 1 mg/L
the federal secondary and primary drinking water
standards (MCLs) for fluoride are 2 and 4 mg/L, respec-
tively
The use of fluoride is only recommended in cases where the feed equipment
is already in place for safety reasons.
C-9
-------
In instances where only one of two or more parallel units is tested,
flow from the other units would dilute the tracer concentration prior to
leaving the plant and entering the distribution system. Therefore, the
impact of drinking water standards on the use of fluoride and other tracer
chemicals can be alleviated in some cases.
C.I.5 Tracer Addition
The tracer chemical should be added at the same point(s) in the
treatment train as the disinfectant to be used in the CT calculations.
C.I.5.1 Step-dose Method
The duration of tracer addition is dependent on the volume of the
basin, and hence, its theoretical detention time. In order to approach a
steady-state concentration in the water exiting the basin, tracer addition
and sampling should usually be continued for a period of two to three
times the theoretical detention time (Hudson, 1981). It is not necessary
to reach a steady state concentration in the exiting water to determine
T,0, however, it is necessary to determine tracer recovery. It is
recommended that the tracer recovery be determined to identify hydraulic
characteristics or density problems. Generally, a 90 percent recovery is
considered to provide reliable results for the evaluation of T,0.
In all cases, the tracer chemical should be dosed in sufficient
concentration to easily monitor a residual at the basin outlet throughout
the test. The required tracer chemical concentration, is generally depen-
dent upon the^ nature of the chosen tracer chemical, including its
background concentration, and the mixing characteristics of the basin to
be tested. Recommended chloride doses on the order of 2& mg/L (Hudson,
1975) should be used for step-method tracer studies where the background
chloride level is less than 10 mg/L. Also, fluoride concentrations as low
as 1.0 to 1.5 mg/L are practical when the raw water fluoride level is not
significant (Hudson, 1975). However, tracer studies conducted on systems
suffering from serious shortcircuiting of flow may require substantially
larger step-doses. This would be necessary to detect the tracer chemical
and to adequately define the effluent tracer concentration profile.
C-10
-------
C.I.5.2 Slug-dose Method
The duration of tracer measurements using the slug-dose method is
also dependent on the volume of the basin, and hence, its theoretical
detention time. In general, samples should be collected for at least
twice the basin's theoretical detention time, or until tracer concentra-
tions are detected near background levels. In order to get reliable
results for Tlo values using the slug-dose method, it is recommended that
the total mass of tracer recovered be approximately 90 percent of the mass
applied. This guideline presents the need to sample until the tracer
concentration recedes to the background level. The total mass recovered
during testing will not be known until completion of the testing and
analysis of the data collected. The sampling period needed is very site
specific. Therefore, it may be helpful to conduct a first run tracer test
as a screen to identify the appropriate sampling period for gathering data
to determine T10.
Tracer addition for slug-dose method tests should be instantaneous
and provide uniformly mixed distribution of the chemical. Tracer addition
is considered instantaneous if the dosing time does not exceed 2 percent
of the basin's theoretical detention time (Marske and Boyle, 1973). One
recommended procedure for achieving instantaneous tracer dosing is to
apply the chemical by gravity flow through a funnel and hose apparatus.
This method is also beneficial because it provides a means of standardiza-
tion, which is necessary to obtain reproducible results.
The mass....of tracer chemical to be added is determined by the desired
theoretical concentration and basin size. The mass of tracer added in
slug-dose tracer tests should be the minimum mass needed to obtain
detectable residual measurements to generate a concentration profile. As
a guideline, the theoretical concentration for the slug-dose method should
be comparable to the constant dose applied in step-dose tracer tests,
i.e., 10 to 20 mg/L and 1 to 2 mg/L for chloride and fluoride, respective-
ly. The maximum mass of tracer chemical needed is calculated by
multiplying the theoretical concentration by the total basin volume. This
is appropriate for systems with high dispersion and/or mixing. This
quantity is diluted as required to apply an instantaneous dose, and
minimize density effects. It should be noted that the mass applied ib not
C-ll
-------
likely to get completely mixed throughout the total volume of the basin.
Therefore, the detected concentration might exceed theoretical concentra-
tions based on the total volume of the basin. For these cases, the mass
of chemical to be added can be determined by multiplying the theoretical
concentration by only a portion of the basin volume. An example of this
is shown in Section C.I.7.2 for a slug-dose tracer study. In cases where
the tracer concentration in the effluent must be maintained below a
specified level, it may be necessary to conduct a preliminary test run
with a minimum tracer dose to identify the appropriate dose for determin-
ing Tlo without exceeding this level.
C.I.6 Test Procedure
In preparation for beginning a tracer study, the raw water
background concentration of the chosen tracer chemical must be estab-
lished. The background concentration is essential, not only for aiding in
the selection of the tracer dosage, but also to facilitate proper
evaluation of the data.
The background tracer concentration should be determined by
monitoring for the tracer chemical prior to beginning the test. The
sampling point(s) for the pre-tracer study monitoring should be the same
as the points to be used for residual monitoring to determine CT values.
The monitoring procedure is outlined in the following steps:
If the tracer chemical is normally added for treatment,
discontinue its addition to the water in sufficient time to
permit the tracer concentration to recede to its background
level before the test is begun.
Prior to the start of the test, regardless of whether the
chosen tracer material is a treatment chemical, the tracer
concentration in the water is monitored at the sampling point
where the disinfectant residual will be measured for CT
calculations.
If a background tracer concentration is detected, monitor it
until a constant concentration, at or below the raw water
background level is achieved. This measured concentration is
the baseline tracer concentration.
C-12
-------
Following the determination of the tracer dosage, feed and monitoring
point(s), and a baseline tracer concentration, tracer testing can begin.
Equal sampling intervals, as could be obtained from automatic
sampling, are not required for either tracer study method. However, using
equal sample intervals for the slug-dose method can simplify the analysis
of the data. During testing, the time and tracer residual of each
measurement should also be recorded on a data sheet. In addition, the
water level, flow, and temperature should be recorded during the test.
C.I.6.1 Step-dose Method
At time zero, the tracer chemical feed will be started and left at
a constant rate for the duration of the test. Over the course of the
test, the tracer residual should be monitored at the required sampling
point(s) at a frequency determined by the overall detention time and site
specific considerations. As a general guideline, sampling at intervals of
2 to 5 minutes should provide data for a well-defined plot of tracer
concentration vs. time. If on-site analysis is available, less frequent
residual monitoring may be possible until a change in residual concentra-
tion is first detected. As a guideline, in systems with a theoretical de-
tention time greater than 4 hours, sampling may be conducted every 10
minutes for the first 30 minutes, or until a tracer concentration above
the baseline level is first detected. In general, shorter sampling
intervals enable better characterization of concentration changes;
therefore, sampling should be conducted at 2 to 5-minute intervals from
the time that a concentration change is first observed until the residual
concentration reaches a steady-state value. A reasonable sampling
interval should be chosen based on the overall detention time of the unit
being tested.
If verification of the test is desired, the tracer feed should be
discontinued, and the receding tracer concentration at the effluent should
be monitored at the same frequency until tracer concentrations correspond-
ing to the background level are detected. The time at which tracer feed
is stopped is time zero for the receding tracer test and must be noted.
The receding racer test will provide a replicate set of measurements which
can be compared with data derived from the rising tracer concentration
versus time curve. For systems which currently feed the tracer chemical,
C-13
-------
the receding curve may be generated from the time the feed is turned off
to determine the background concentration level.
C.I.6.2 Slug-dose Method
At time zero for the slug-dose method, a large instantaneous dose of
tracer will be added to the influent of the unit. The same sampling
locations and frequencies described for step-dose method tests also apply
to slug-dose method tracer studies. One exception with this method is
that the tracer concentration profile will not equilibrate to a steady
state concentration. Because of this, the tracer should be monitored
frequently enough to ensure acquisition of data needed to identify the
peak tracer concentration.
Slug-dose method tests should be checked by performing a material
balance to ensure that all of the tracer fed is recovered, or, mass
applied equals mass discharged.
C.I.7 Data Evaluation
Data from tracer studies should be summarized in tables of time and
residual concentration. These data are then analyzed to determine the
detention time, T10, to be used in calculating CT. Tracer test data from
either the step or slug-dose method can be evaluated graphically,
numerically, or by a combination of these techniques.
C.I.7.1 Step-dose Method
The graph-ical method of evaluating step-dose test data involves
plotting a graph of dimensionless concentration versus time and reading
the value for T,0 directly from the graph at the appropriate dimensionless
concentration. Alternatively, the data from step-dose tracer studies may
be evaluated numerically by developing a semi-logarithmic plot of the
dimensionless data. The semi-logarithmic plot allows a straight line to
be drawn through the data. The resulting equation of the line is used to
calculate the T10 value, assuming that the correlation coefficient
indicates a good statistical fit (0.9 or above). Scattered data points
from step-dose tracer tests are discredited by drawing a smooth curve
through the data.
C-14
-------
An illustration of the T10 determination will be presented in an
example of the data evaluation required for a clearwell tracer study.
C.I.7.2 Slug-dose Method
Data from slug-dose tracer tests is analyzed by converting it to the
mathematically equivalent step-dose data and using techniques discussed in
Section C.I.7.1 to determine Tlo. A graph of dimensionless concentration
versus time should be drawn which represents the results of a slug-dose
tracer test. The key to converting between the data forms is obtaining
the total area under the slug-dose data curve. This area is found by
graphically or numerically integrating the curve. The conversion to
step-dose data is then completed in several mathematical steps involving
the total area.
A graphical technique for converting the slug-dose data involves
physically measuring the area using a planimeter. The planimeter is an
instrument used to measure the area of a plane closed curve by tracing its
boundary. Calibration of this instrument to the scale of the graph is
required to obtain meaningful readings.
The rectangle rule is a simple numerical integration method which
approximates the total area under the curve as the sum of the areas of
individual rectangles. These rectangles have heights and widths equal to
the residual concentration and sampling interval (time) for each data
point on the curve, respectively. Once the data has been converted, T,0
may be determined in the same manner as data from step-dose tracer tests.
Slug-dose, concentration profiles can have many shapes, depending on
the hydraulics of the basin. Therefore, slug-dose data points should not
be discredited by drawing a smooth curve through the data prior to its
conversion to step-dose data. The steps and specific details involved
with evaluating data from both tracer study methods are illustrated in the
following examples.
Example for Determining T10 in a Clearwell
Two tracer studies employing the step-dose and slug-dose methods of
tracer addition were conducted for a clearwell with a theoretical
detention time, T, of 30 minutes at an average flow of 2.5 MGD. Because
fluoride is added at the inlet to the clearwell as a water treatment
chemical, necessary feed equipment was in place for dosing a constant
C-15
-------
concentration of fluoride throughout the step-dose tracer test. Based on
this convenience, fluoride was chosen as the tracer chemical for the
step-dose method test. Fluoride was also selected as the tracer chemical
for the slug-dose method test. Prior to the start of testing, a fluoride
baseline concentration of 0.2 mg/L was established far the water exiting
the clearwell.
Step-dose Method Test
For the step-dose test a constant fluoride dosage of 2.0 mg/L was
added to the clearwell inlet. Fluoride levels in the clearwell effluent
were monitored and recorded every 3 minutes. The raw tracer study data,
along with the results of further analyses are shown in Table C-l.
The steps in evaluating the raw data shown in the first column of
Table C-l are as follows. First, the baseline fluoride concentration,
0.2 mg/L, is subtracted from the measured concentration to give the
fluoride concentration resulting from the tracer study addition alone.
For example, at elapsed time = 39 minutes, the tracer fluoride concentra-
tion, C, is obtained as follows:
C = C - C
measur efl ban i i in
= 1.85 mg/L - 0.2 mg/L
= 1.65 mg/L
This calculation was repeated at each time interval to obtain the data
shown in the third column of Table C-l. As indicated, the fluoride
concentration rises from 0 mg/L at t = 0 minutes to the applied fluoride
dosage of 2 mg/L, at t = 63 minutes.
The next step is to develop dimensionless concentrations by dividing
the tracer concentrations in the second column of Table C-l by the applied
fluoride dosage, Co - 2 mg/L. For time = 39 minutes, C/Co is calculated
as follows:
C/Co * (1.65 mg/L)/(2.0 mg/L)
= 0.82
The resulting dimensionless data, presented in the fourth column of
Table C-l, is the basis for completing the determination of Tlo by either
the graphical or numerical method.
C-16
-------
TABLE C-l
CLEARHELL DATA--STEP-DQSE TRACER TEST0 : 3)
t. minutes
0
3
6
9
12
15
18
21
24
27
30
33
36
39
42
45
48
51
54
57
60
63
Fluoride Concentration
Measured. mg/L
0.20
0.20
0.20
0.20
0.29
0.67
0.94
1.04
1.44
1.55
1.52
1.73
1.93
1.85
1.92
2.02
1.97
1.84
2.06
2.05
2.10
2.14
Tracer. mg/L Dimensi
0
0
0
0
0.09
0.47
0.74
0.84
1.24
1.35
1.32
1.53
1.73
1.65
1.72
1.82
-1.77
1.64
1.86
1.85
1.90
1.94
onless. C/Co
0
0
0
0
0.045
0.24
0.37
0.42
0.62
0.68
0.66
0.76
0.86
0.82
0.86
0.91
0.88
0.82
0.93
0.92
0.95
0.96
Notes:
1. Baseline cone. = 0.2 mg/L, fluoride dose = 2.0 mg/L
2. Measured cone. = Tracer cone. + Baseline cone.
3. Tracer cone. = Measured cone. - Baseline cone.
-------
In order to determine T,0 by the graphical method, a plot of C/Co vs.
time should be generated using the data in Table C-l. A smooth curve
should be drawn through the data as shown on Figure C-l.
T10 is read directly from the graph at a dimensionless concentration
(C/Co) corresponding to the time for which 10 percent of the tracer has
passed at the effluent end of the contact basin (T10). For step-dose
method tracer studies, this dimensionless concentration is C/Co * 0.10
(Levenspiel, 1972).
T10 should be read directly from Figure C-l at C/Co = 0.1 by first
drawing a horizontal line (C/Co = 0.1) from the Y-axis (t * 0) to its
intersection with the smooth curve drawn through the data. At this point
of intersection, the time read from the X-axis is T10 and may be found by
extending a vertical line downward to the X-axis. These steps were
performed as illustrated on Figure C-l, resulting in a value for Tlo of
approximately 13 minutes.
For the numerical method of data analysis, several additional steps
are required to obtain T10 from the data in the fourth column of Table C-l.
The forms of data necessary for determining T10 through a numerical
solution are 1oglo(l-C/Co) and t/T, the elapsed time divided by the
theoretical residence time. These are obtained by performing the required
mathematical operations on the data in the fourth column of Table C-l.
For example, recalling that the theoretical detention time, T, is 30
minutes, the values for loglo (1-C/Co) and t/T are computed as follows for
the data at t j* 39 minutes:
loglo(l-C/Co) * Iog10 (1-0.82)
• 1og,0 (0.18)
= -0.757
t/T « 39 min/30 min » 1.3
This calculation was repeated at each time interval to obtain the
data shown in Table C-2. These data should be linearly regressed as
log,0(l-C/Co) versus t/T to obtain the fitted straight-line parameters to
the following equation:
C-17
-------
FIGURE C-1
C/Co vs. Time
Graphical Analysis for T10
o
O
O
0.1
10
20 30 40
TIME (MINUTES)
60
-------
TABLE C-2
DATA FOR NUMERICAL DETERMINATION OF T
10
I/I 1aflia(l-C/Co)
0 0
0.1 0
0.2 0
0.3 0
0.4 -0.020
0.5 -0.116
0.6 -0.201
0.7 -0.237
0.8 -0.420
0.9 -0.488
1.0 -0.468
1.1 -0.629
1.2 -0.870
1.3 -0.757
1.4 -C.854
1.5 -1.046
1.6 -0.939
1.7 -0.745
1.8 -1.155
1.9 - -1.125
2.0 -1.301
2.1 -1.532
-------
log,0(l-C/Co) = m(t/T) + b (1)
In equation 1, m and b are the slope and intercept, respectively,
for a plot of log10(l-C/Co) vs. t/T. This equation can be used to
calculate T10, assuming that the correlation coefficient for the fitted
data indicates a good statistical fit (0.9 or above).
A linear regression analysis was performed on the data in Table C-2,
resulting in the following straight-line parameters:
slope = m * -0.774
intercept * b * 0.251
correlation coefficient = 0.93
Although these numbers were obtained numerically, a plot of
log10(l-C/Co) versus t/T is shown for illustrative purposes on Figure C-2
for the data in Table C-2. In this analysis, data for time = 0 through 9
minutes were excluded because fluoride concentrations above the baseline
level were not observed in the clearwell effluent until t = 12 minutes.
Equation 1 is then rearranged in the following form to facilitate a
solution for T,0:
T,0/T = (log,0 (1 - 0.1) - b)/m (2)
In equation 2, as with graphical method, T10 is determined at the
time for which- C/Co » 0.1. Therefore, in equation 2, C/Co has been
replaced by 0.1 and t (time) by T.Q. To obtain a solution for Tu, the
values of the slope, intercept, and theoretical detention time are
substituted as follows:
T10/30 min. » (loglo(l - 0.1) - 0.251)/(-0.774)
T10 = 12 minutes
In summary both the graphical and numerical methods of data
reduction resulted in comparable values for Tlo. With the numerical
method, T10 was determined as the solution to an equation based on the
C-18
-------
straight-line parameters to a linear regression analysis of the tracer .
study data instead of an "eyeball" estimate from a data plot. *
Slug-dose Method Test
A slug-dose tracer test was also performed on the clear-well at a
flow rate of 2.5 mgd. A theoretical clearwell fluoride concentration of
2.2 mg/L was selected. The fluoride dosing volume and concentration were
determined from the following considerations:
Dosing Volume
The fluoride injection apparatus consisted of a funnel and a
length of copper tubing. This apparatus provided a constant
volumetric feeding rate of 7.5 liters per minute (L/min) under
gravity flow conditions.
At a flow rate of 2.5 mgd, the clearwell has a theoretical
detention time of 30 minutes. Since the duration of tracer
injection should be less than 2 percent of the clearwell's
theoretical detention time for an instantaneous dose, the
maximum duration of fluoride injection was:
Max. dosing time * 30 minutes x .02 = 0.6 minutes
I
At a dosing rate of 7.5 L/min, the maximum fluoride dosing
volume is calculated to be:
Max. dosing volume = 7.5 L/min. x 0.6 minutes = 4,5 L
For this tracer test, a dosing volume of 4 liters was select-
ed, providing an instantaneous fluoride dose in 1.8 percent of
the-theoretical detention time.
Fluoride Concentration
The theoretical detention time of the clearwell, 30 minutes,
was calculated by dividing the clearwell volume, 52,100
gallons or 197,200 liters, by the average flow rate through
the clearwell, 2.5 mgd.
Assuming the tracer is completely dispersed throughout the
total volume of the clearwell, the mass of fluoride required
to achieve a theoretical concentration of 2.2 mg/L is calcu-
lated as follows:
Fluoride mass (initial) = 2.2 mg/L x 197,200 L x J__g = 434g
1000 rag
C-19
-------
FIGURE C-2
1-C/Co vs. t/T
Numerical Analysis for T10
O
0.01
2.5
Slope. m • -0.774
Intercept, b-0.251
CorritatJon Co«ffld«nt • 0.93
-------
The concentration of the instantaneous fluoride dose is
determined by dividing this mass by the dosing volume, 4
liters:
Fluoride concentration * 434 g = 109 g/L
4 L
Fluoride levels in the exit to the clearwell were monitored and
recorded every 3 minutes. The raw slug-dose tracer test data are shown in
Table C-3.
The first step in evaluating the data for different times is to
subtract the baseline fluoride concentration, 0.2 mg/L, from the measured
concentration at each sampling interval (Table C-3). This is the same as
the first step used to evaluate step-dose method data and gives the
fluoride concentrations resulting from the tracer addition alone, shown in
the third column of Table C-3. As indicated, the fluoride concentration
rises from 0 mg/L at t = 0 minutes to the peak concentration of 3.6 mg/L
at t = 18 minutes. The exiting fluoride concentration gradually recedes
to near zero at t = 63 minutes. It should be noted that a maximum
fluoride concentration of 2.2 mg/L is based on assuming complete mixing of
the tracer added throughout the total clear-well volume. However, as shown
in Table C-3, the fluoride concentrations in the clean-veil effluent
exceeded 2.2 mg/L for about 6 minutes between 14 and 20 minutes. These
higher peak concentrations are caused by the dispersion of tracer
throughout only a portion of the total clearwell volume. If a lower
tracer concentration is needed in the effluent because of local or federal
regulations, the mass to be added should be decreased accordingly.
The dimensionless concentrations in the fourth column of Table C-3
were obtained by dividing the tracer concentrations in the third column by
the clear-well's theoretical concentration, Co » 2.2 mg/L. These
dimensionless concentrations were then plotted as a function of time, as
is shown by the slug-dose data on Figure C-3. These data points were
connected by straight lines, resulting in a somewhat jagged curve.
The next step in evaluating slug-dose data is to determine the total
area under the slug-dose data curve on Figure C-3. Two methods exist for
finding this area -- graphical and numerical. The graphical method is
C-20
-------
CM
—i C
-— o
<•? uj
02
(Jl
1-0
I
O)
OOOOO —
ococotr>uiLnro«-*r'i)>-«cvj'— • o •— ' -~ o
'-'OOOOOOOOOOOOOOO
CNJ
o
o
i'
o
o
T O 00 CTi CSI
OOOO-^f^fl — ^--^
oooooooooo
c —• z
3 o
o
13
0>
3
o - o
u — o
4- ZTT3
i- E —
3
J*.
O O
-------
FIGURE C-3
C/Co vs. Time
Conversion of Slug-to Step-Dose Data
.3 p
r
i
5 r
h
-> •
10 20 30 40 50
TIME (MINUTES)
60 70
A
-------
based on a physical measurement of the area using a planimeter. This
involves calibration of the instrument to define the units conversion and
tracing the outline of the curve to determine the area. The results of
performing this procedure may vary depending on instrument accuracy and
measurement technique. Therefore, only an illustration of the numerical
technique for finding the area under the slug-dose curve will be presented
for this example.
The area obtained by either the graphical or numerical method would
be similar. Furthermore, once the area is found, the remaining steps
involved with converting the data to the step-dose response are the same.
Table C-4 summarizes the results of determining the total area using
a numerical integration technique called the rectangle rule. The first
and second columns in Table C-4 are the sampling time and fluoride
concentration resulting from tracer addition alone, respectively. The
steps in applying these data are as follows. First, the sampling time
interval, 3 minutes, is multiplied by the fluoride concentration at the
end of the 3-minute interval to give the incremental area, in units of
milligram minutes per liter. For example, at elapsed time, t = 39
minutes, the incremental area is obtained as follows:
Incremental area = sampling time interval x fluoride cone.
= (39-36) minutes x 0.4 mg/L
= 1.2 mg-min/L
This calculation was repeated at each time interval to obtain the data
shown in the third column of Table C-4.
If the data had been obtained at unequal sampling intervals, then
the incremental area for each interval would be obtained by multiplying
the fluoride concentration at the end of each interval by the time
duration of the interval. This convention also requires that the
incremental area be zero at the first sampling point, regardless of the
fluoride concentration at that time.
As is shown in Table C-4, all incremental areas were summed to
obtain 59.4 mg-min/L, the total area under the slug-dose tracer test
curve. This number represents the total mass of fluoride that was
C-21
-------
detected during the course of the tracer test divided by the average flow
rate through the clear-well.
To complete the conversion of slug-dose data to its equivalent
step-dose response requires two additional steps. The first involves
summing, consecutively, the incremental areas in the third column of Table
C-4 to obtain the cumulative area at the end of each sampling interval.
For example, the cumulative area at time, t * 27 minutes is found as
follows:
Cumulative area = 0 + 0 + 0 + 0 + 3 + 10.2 + 10.8 + 5.4 + 5.7 + 3.6
= 38.7 mg-min/L
The cumulative areas for each interval are recorded in the fourth column
of Table C-4.
The final step in converting slug-dose data involves dividing the
cumulative area at each interval by the total mass applied. Total area
based on applied mass is calculated as follows:
Total area mass applied/average flow = 434 g x 1000 ifl/6,570 L
g min
» 66.1 trig-mi n
L
For time = 39 minutes, the resulting step-dose data point is calculated as
follows:
C/Co = 49.5 mg-min/L / 59.4 mg-min/L
= 0.83
The result of performing this operation at each sampling interval is the
equivalent step-dose data. These data points are shown in the fifth
column of Table C-4 and are also plotted on Figure C-3 to facilitate a
graphical determination of T10. A smooth curve was fitted to the step-dose
data as shown on the figure.
T10 can be determined by the methods illustrated previously in this
example for evaluating step-dose tracer test data. The graphical method
illustrated on Figure C-3 results in a reading of T10 = 15 minutes.
C-22
-------
O) i/l
— O
I
•— CL
3 O>
O--M
LU 1/5
o o o o o o o o o o o o o o o o o o o o o -«
«
t—
o
0)
,2 §
i
0)
en
<— c
C I
il
oj
U ^3
C 0)
""
•o
o
3
0*
3
C
°i
-------
C.I.7.3 Additional Considerations
In addition to determining T10 for use in CT calculations, slug-dose
tracer tests provide a more general measure of the basin's hydraulics in
terms of the fraction of tracer recovery. This number is representative
of short-circuiting and dead space in the unit resulting from poor
baffling conditions and density currents induced by the tracer chemical.
A low tracer recovery is generally indicative of inadequate hydraulics.
However, inadequate sampling in which peaks in tracer passage are not
measured will result in an under estimate of tracer recovery. The tracer
recovery is calculated by dividing the mass of fluoride detected by the
mass of fluoride dosed.
The dosed fluoride mass was calculated previously and was 434 grams.
The mass of detected fluoride can be calculated by multiplying the total
area under the slug-dose curve by the average flow, in appropriate units,
at the time of the test. The average flow in the clearwell during the
test was 2.5 mgd or 6,570 L/min. Therefore, the mass of fluoride tracer
that was detected is calculated as follows:
Detected fluoride mass = total area x average flow
= 59.4 mg-mjn x 1 g x 6,570 [,
L 1000 mg min
= 390 g
Tracer recovery is than calculated as follows:
Fluoride recovery » detected mass/dosed mass x 100
* 390 g / 434 g x 100
= 90 %
This is a typical tracer recovery percentage for a slug-dose test, based
on the experiences of Hudson (1975) and Thirumurthi (1969).
C.I."8 "flow Dependency of T10
For systems conducting tracer studies at four or more flows, the T,0
detention time should be determined by the above procedures for each of
the desired flows. The detention times should then be plotted versus
flow. For the example presented in the previous section, tracer studies
C-23
-------
were conducted at additional flows of 1.1, 4.2, and 5.6 MGD. The T10
values at the various flows were: *
1.1 25
2.5 13
4.2 7
5.6 4
T10 data for these tracer studies were plotted as a function of the flow,
Q, as shown on Figure C-4.
If only one tracer test is performed, the flow rate for the tracer
study should be not less than 91 percent of the highest flow rate
experienced for the section. The hydraulic profile to be used for
calculating CT would then be generated by drawing a line through points
obtained by multiplying the T10 at the tested flow rate by the ratio of the
tracer study flow rate to each of several different flows in the desired
flow range.
For the example presented in the previous section, the clearwell
experiences a maximum flow at peak hourly conditions of 6.0 mgd. The
highest tested flow rate was 5.6 mgd, or 93 percent of the maximum flow.
Therefore, the detention time, T,0 = 4 minutes, determined by the tracer
test at a flow rate of 5.6 mgd may be used to provide a conservative
estimate of T,0 for all flow rates less than or equal to the maximum flow
rate, 6.0 mgd. The line drawn through points found by multiplying T'" =
4 minutes by th« ratio of 5.6 mgd to each of several flows less than 5.6
mgd is also shown on Figure C-4 for comparative purposes with the
hydraulic profile obtained from performing four tracer studies at
different flow rates.
C.2 Determination of T10 Without Conducting a Tracer Study
In some situations, conducting tracer studies for determining the
disinfectant contact time, T10, may be impractical or prohibitively
expensive. The limitations may include a lack of funds, manpower or
equipment necessary to conduct the study. For these cases, the Primacy
Agency may allow the use of "rule of thumb" fractions representing the
C-24
-------
FIGURE C-4
Detention Time vs. Flow
35
30
25
CO*
LJJ „_
h- 20
Z
2
a ^
10
X
Average X
4-Flow profile
1-Flow profile
Maximum
Extrapolation
345
FLOW (MGD)
6
-------
ratio of T10 to T, and the theoretical detention time, to determine the
detention time, T10, to be used for calculating CT values. This method for
finding T10 involves multiplying the theoretical detention time by the rule
of thumb fraction, T10/T, that is representative of the particular basin
configuration for which Tlo is desired. These fractions provide rough
estimates of the actual T10 and are recommended to be used only on a
limited basis.
Tracer studies conducted by Marske and Boyle (1973) and Hudson
(1975) on chlorine contact chambers and flocculators/settling basins,
respectively, were used as a basis in determining representative T10/T
values for various basin configurations. Marske and Boyle (1973) performed
tracer studies on 15 distinctly different types of full-scale chlorine
contact chambers to evaluate design characteristics that affect the actual
detention time. Hudson (1975) conducted 16 tracer tests on several
flocculation and settling basins at six water treatment plants to identify
the effect of flocculator baffling and settling basin inlet and outlet
design characteristics on the actual detention time.
C.2.1 Impact of Design Characteristics
The significant design characteristics include: length-to-width
ratio, the degree of baffling within the basins, and the effect of inlet
baffling and outlet weir configuration. These physical characteristics of
the contact basins affect their hydraulic efficiencies in terms of dead
space, plug flow, and mixed flow proportions. The dead space zone of a
basin is basin volume through which no flow occurs. The remaining volume
where flow occurs is comprised of plug flow and mixed flow zones. The
plug flow zone is the portion of the remaining volume in which no mixing
occurs in the direction of flow. The mixed flow zone is characterized by
complete •ixing in the flow direction and is the complement to the plug
flow zone. All of these zones were identified in the studies for each
contact basin. Comparisons were then made between the basin configura-
tions and the observed flow conditions and design characteristics.
The ratio T10/T was calculated from the data presented in the studies
and compared to its associated hydraulic flow characteristics. Both
studies resulted in T10/T values which ranged from 0.3 to 0.7. The results
C-25
-------
of the studies indicate how basin baffling conditions can influence the
T,0/T ratio, particularly baffling at the inlet and outlet to the basin.
As the basin baffling conditions improved, higher T10/T values were
observed, with the outlet conditions generally having a greater impact
than the inlet conditions.
As discovered from the results of the tracer studies performed by
Marske and Boyle (1973) and Hudson (1975), the effectiveness of baffling
in achieving a high T10/T fraction is more related to the geometry and
baffling of the basin than the function of the basin. For this reason,
T10/T values may be defined for three levels of baffling conditions rather
than for particular types of contact basins. General guidelines were
developed relating the TIO/T values from these studies to the respective
baffling characteristics. These guidelines can be used to determine the
T10 values for specific basins.
C.2.2 Baffling Classifications
The purpose of baffling is to maximize utilization of basin volume,
increase the plug flow zone in the basin, and minimize short circuiting.
Some form of baffling at the inlet and outlet of the basins is used to
evenly distribute flow across the basin. Additional baffling may be
provided within the interior of the basin (intra-basin) in circumstances
requiring a greater degree of flow distribution. Ideal baffling design
reduces the inlet and outlet flow velocities, distributes the water as
uniformly as practical over the cross section of the basin, minimizes
mixing with the water already in the basin, and prevents entering water
from short circuiting to the basin outlet as the result of *ind or density
current effects. Three general classifications of baffling conditions --
poor, average, and superior -- were developed to categorize the results of
the tracer studies for use in determining T10 from the theoretical
detention time of a specific basin. The T10/T fractions associated with
each degree of baffling are summarized in Table C-5. Factors representing
the ratio between T10 and the theoretical detention time for plug flow in
pipelines and flow in a completely mixed chamber have been included in
Table C-5 for comparative purposes. However, in practice the theoretical
T10/T values of 1.0 for plug flow and 0.1 for mixed flow are seldom
C-26
-------
achieved because of the effect of dead space. Conversely, the T10/T values
shown for the intermediate baffling conditions already incorporate the
effect of the dead space zone, as well as the plug flow zone, because they
were derived empirically rather than from theory.
As indicated in Table C-5, poor baffling conditions consist of an
unbaffled inlet and outlet with no intra-basin baffling. Average baffling
conditions consist of intra-basin baffling and either a baffled inlet or
outlet. Superior baffling conditions consist of at least a baffled inlet
and outlet, and possibly some intra-basin baffling to redistribute the
flow throughout the basin's cross-section.
The three basic types of basin inlet baffling configurations are:
a target-baffled pipe inlet, an overflow weir entrance, and a baffled
submerged orifice or port inlet. Typical intra-basin baffling structures
include: diffuser (perforated) walls; launders; cross, longitudinal, or
maze baffling to cause horizontal or vertical serpentine flow; and
longitudinal divider walls, which prevent mixing by increasing the
length-to-width ratio of the basin(s). Commonly used baffled outlet
structures include free-discharging weirs, such as sharpcrested and
V-notch, and submerged ports or weirs. Weirs that do not span the width
of the contact basin, such as Cipolleti weirs, should not be considered
baffling as their use may substantially increase weir overflow rates and
the dead space zone of the basin.
C.2.3 Examples of Baffling
Examples of these levels of baffling conditions for rectangular and
circular basins are explained and illustrated in the following section.
Typical uses of various forms of baffled and unbaffled inlet and outlet
structures are also illustrated.
The-plan and section of a rectangular basin with po.oj baffling con-
ditions, which can be attributed to the unbaffled inlet and outlet pipes,
is illustrated on Figure C-5. The flow pattern shown in the plan view
indicates straight-through flow with dead space occurring in the regions
between the individual pipe inlets and outlets. The section view reveals
additional dead space from a vertical perspective in the upper inlet and
lower outlet corners of the contact basin. Vertical mixing also occurs as
C-27
-------
PLAN
SECTION
FIGURE C-5 POOR BAFFLING CONDITIONS --
RECTANGULAR CONTACT BASIN
-------
71
i ' •
71
A
PLAN
SECTION
FIGURE C-6 AVERAGE BAFFLING CONDITIONS
RECTANGULAR CONTACT BASIN
-------
/
XI
H /
"J
-A
PLAN
V
i — f
x
/ / A
SECTION
FIGURE C-7 SUPERIOR BAFFLING CONDITIONS
RECTANGULAR CONTACT BASIN
-------
PLAN
SECTION
FIGURE C-8 POOR BAFFLING CONDITIONS
CIRCULAR CONTACT BASIN
-------
PLAN
SECTION
FIGURE C-9 AVERAGE BAFFLING CONDITIONS
CIRCULAR CONTACT BASIN
-------
Superior baffling conditions are attained in the basin configuration
shown on Figure C-10 through the addition of a perforated inlet baffle and
submerged orifice outlet ports. As indicated by the flow pattern, more of
the basin's volume is utilized due to uniform flow distribution created by
the perforated baffle. Short circuiting is also minimized because only a
small portion of flow passes directly through the perforated baffle wall
from the inlet to the outlet ports.
C.2.4 Additional Considerations
Flocculation basins and ozone contactors represent water treatment
processes with slightly different characteristics from those presented in
Figures C-5 through C-10 because of the additional effects of mechanical
agitation and mixing from ozone addition, respectively. Studies by Hudson
(1975) indicated that a single-compartment flocculator had a T10/T value
less than 0.3, corresponding to a dead space zone of about 20 percent and
a very high mixed flow zone of greater than 90 percent. In this study,
two four-compartment flocculators, one with and the other without
mechanical agitation, exhibited T,0/T values in the range of 0.5 to 0.7.
This observation indicates that not only will compartmentation result in
higher T,0/T values through better flow distribution, but also that the
effects of agitation intensity on T,0/T are reduced where sufficient
baffling exists. Therefore, regardless of the extent of agitation,
baffled flocculation basins with two or more compartments should be
considered to possess average baffling conditions (T10/T » 0.5), whereas
unbaffled, single-compartment flocculation basins are characteristic of
poor baffling conditions (T10/T = 0.3).
Similarly, multiple stage ozone contactors are baffled contact
basins which-show characteristics of average baffling conditions. Single
stage ozone contactors should be considered as being poorly baffled.
However, circular, turbine ozone contactors may exhibit flow distribution
characteristics which approach those of completely mixed basins, with a
T10/T of 0.1, as a result of the intense mixing.
In many cases, settling basins are directly connected to the
flocculators. Data from Hudson (1975) indicates that poor baffling
conditions at the flocculator/settling basin interface can result in
C-29
-------
PLAN
SECTION
FIGURE C-10 SUPERIOR BAFFLING CONDITIONS
CIRCULAR CONTACT BASIN
-------
References
Hudson, H. E., Jr.. "Residence Times in Pretreatment", J. AWWA, pp. 45-52,
January, 1975.
Hudson, H. E.( Jr.. Water Clarification Processes; Practical Design and,
Evaluation. Van Nostrand Reinhold Company, New York, 1981.
Levenspiel, 0.. Chemical Reaction Engineering. John Wiley & Sons, New
York, 1972.
Marske, D. M. and Boyle, J. D.. "Chlorine Contact Chamber Design - A Field
Evaluation", Water and Sewage Works, pp. 70-77, January, 1973.
Thirumurthi, D.. "A Break-through in the Tracer Studies of Sedimentation
Tanks", J. WPCF, pp. R405-R418, November, 1969.
C-31
-------
APPENDIX D
ANALYTICAL REQUIREMENTS OF THE SWTR AND
A SURVEY OF THE CURRENT STATUS OF RESIDUAL DISINFECTANT
MEASUREMENT METHODS FOR ALL CHLORINE SPECIES AND OZONE
-------
APPENDIX D
ANALYTICAL REQUIREMENTS
Only the analytical method(s) specified in the SWTR, or otherwise
approved by EPA, may be used to demonstrate compliance with the
requirements of the SWTR. Measurements of pH, temperature, turbidity, and
residual disinfectant concentrations must be conducted by a party approved
by the Primacy Agency. Measurements for total coliforms, fecal coliforms,
and heterotrophic bacteria as measured by the heterotrophic plate count
(HPC), must be conducted by a laboratory certified by the Primacy Agency
or EPA to do such analysis. Until laboratory certification criteria are
developed for the analysis of HPC and fecal coliforms, any laboratory
certified for total coliform analysis is acceptable for HPC and fecal
coliform analysis. The test methods to be used for various analyses are
listed below:
(1) Fecal coliform concentration - Method 908C (MPN Procedure),
908D (Estimation of Bacterial Density), or 909C (Membrane
Filter Procedure) as set forth in Standard Methods for the
Examination of Water and Wastewater. American Public Health
Association, 16th edition.
(2) Total coliform concentration - Methods 908A, B, D (MPN
Procedure) or 909A, B (Membrane Filter Procedure) as set forth
in Standard Methods for the Examination of Water and
Wastewater. American Public Health Association, 16th edition;
Autoanalysis Colilert (EPA refers to this as Minimal Medium
ONPG-MUG Method), as set forth in Applied and Environmental
Microbiology, American Society for Microbiology, Volume 54,
No. 6, June 1988. pp. 1595-1601.
(3) Heterotorphic Plate Count - Method 907A (Pour Plate Method),
as set forth in Standard Methods for the Examination of Water
and Wastewater. American Public Health Association, 16th
edition.
(4) Turbidity - Method 214A (Nephelometric Method) as set forth in
Standard Methods for the Examination of Water and Wastewater.
American Public Health Association, 16th edition.
(5) Residual disinfectant concentration - Residual disinfectant
concentrations for free chlorine and combined chlorine must be
measured by Method 408C (Amperometric Titration Method),
Method 408D (DPD Ferrous Titrimetric Method), Method 408E (DPD
Color-metric Method), or Method 408F (Leuco Crystal Violet
D-l
-------
Method) as set forth in Standard Methods for the Examination
O-f Water and Wastewater. American Public Health Association,
16th edition. Disinfectant residuals for free chlorine and
combined chlorine may also be measured by using DPD
colorimetric test kits if approved by the Primacy Agency.
Disinfectant residuals for ozone must be measured by the
Indigo Trisulfonate Method (Bader, H., Hoigne, J.,
"Deternination of Ozone in Water by the Indigo Method; A
Submitted Standard Method;" Ozone Science and Engineering,
Vol. 4, pp. 169-176, Pergamon Press Ltd., 1982), or automated
methods which are calibrated in reference to the results
obtained by the Indigo Trisulfonate Method, on a regular
basis, as determined by the Primacy Agency. This method is
described in section of the manual. (Note: This method is
included in the 17th edition of Standard Methods for the
Examination of Water and Wastewater. American Public Health
Association; the Idiodometric Method in the 16th edition may
not be used.) Disinfectant residuals for chlorine dioxide
must be measured by Method 410B (Amperometric Method) or
Method 410C (DPD Method) as set forth in Standard Methods for
the Examination of Water and Wastewater. American Public
Health Association, 16th edition.
(6) Temperature - Method 212 as set forth in Standard Methods for
the Examination of Water and Wastewater. American Public
Health Association, 16th edition.
(7) pH - Method 423 as set forth in Standard Methods for the
Examination of Water and Wastewater, American Public Health
Association, 16th edition.
References
Edberg et al, "National Field Evaluation of a Defined Substrate Method for
the Simultaneous Enumeration of Total Coliforms and Escherichia Coli from
Drinking Water:"- Comparison with the Standard Multiple Tube Fermentation
Method," Applied and Environmental Microbiology, Volume 54, pp. 1595-1601,
June 1988.
0-2
-------
PREFACE
The AWWA paper entitled "A survey of the current status of residual
disinfectant measurement method for all chlorine species and ozone" will
be included in the final document. It has not been included here for the
sake of brevity. However, the publication is available from the AWWA
Customer Services Department, 6666 W. Quincy Avenue, Denver, Co. 80235;
Telephone (303) 794-7711. The document publication number is 90529.
The above publication summarizes the AWWA Research foundation's 816
page publication entitled " Disinfectant Residual Measurement Methods",
publication number 90528. This document is also available from the
customer services department listed above.
D-3
-------
A SURVEY OF THE CURRENT STATUS CF RESIDUAL DISINFECTANT
MEASUREMENT METHODS FOR ALL CHLORINE SPECIES AND OZONE
Gilser: Gorion
Department of Chemistry
Miami University
Oxford, CH 45055
William J. Csooer
Drinxing Watap R"eseapc.n Ca.ncar
Flcrica Incsrnacional Um'vers*'/
Miami , "Icriaa 23139
Rip 3. Rica
Rica, Inccrcorarad
Asnton, yarylana 2C361
oilbert E. Pacay
"esart-ent of Chemistry
Miami University
AWWA Raseapc.*! "cunaaticn
5ccc W. Ouincy Avenue
"envoi- r" 3P2"S
^«-'''« | WW wUC«*tf
fiovemcep 1937
by tne American Water WOPICS Association
-------
DISCLAIMER
This study was furcec :y trie American Watar Wcr
-------
FOREWORD
'ill's resort is oart of tr.e on-going rssearc.i orccm of t.ie Air.i'A ^ssaarc.i
Foundation. The rssaarc.n cescnoec in tre fell owing paces was furceo- oy
t.ie Foundation 1.1 benalf of its -emBers and suoscrieers 1.1 oartieular arc
t.ie water sues!;/ industry in general. Selectaa for funning ay AWUARF's
Soars of Trustees, t.ie project was identified as a practical, prior-it/ neec-
of Me incustry. It is hosed tr.at t.iis puolicaticn will receive wice arc
serTous attant:on and tnat its findings, conclusions, anc raccraencaticns
will be acoliaa in csnnumties inrougnout "ne Unitsd Statas ana Car.aca.
The .Resaarc.n Founaat:cn was craatad by :re watar sucoly iicustry a
canter for ccocerative rasaarcn and development. The Foundation i
coes not conduct rasearcn; it functions as a planning and ^anacsne
agency, awarding contract: to ctner institutions, SUCH as water u
universities, engineering fins, and ctner organizations. The sc:
and tacnnical expertise of
t~e staff
is further ennancaa by industr
volunteers wno sarve on Project Advisory Cerrmittees and on otr.er s
committees arc councils. •-n extensive planning process involves m
hundreds of water professionals in t.ie imoortant tas* cf '
-------
PREFACE
7:iis socument surrniar: :ss :r:a AWWA Researcn Founcaticn's 315 :acs
puolication "Cisinfactant Resiaual Measurement Met.ncas.' "nat
puolicjtion (?'jfilic3t:cn Numoer 90523) can be orsarac frtsi :na -UWA
Cjstcner Services Deoaptsenc, 6566 w. Quincy Avenue, Denver, CC 3C225;
talesnone, (203) 794-7711.
71-!e ourpose of this summary document is to provide tr.e wa*ar utnity
laooratory analyst witn cuicanca in selecting ais:nfactant rastcual
•neasurament metnoas. Eitner tnis document or trie full racort is
reccrrr.enaed as a ccmoanicn to Standard Metnoas for trte Examination of
Watar ana
-------
ACKNOWLEDGEMENTS
The autrcps wisn to exsrsss tneip acapeci at: on to t.~e Amep'can Watap WOP'<
Assoc: acion - Resaarcn Founaat:on for the opportunity :o carry oui :ms
'ai rav:ew of :r;e li
, :r.e auc.icrs would like to pay tn'buta :o t.ie really
peoola •- all t.^ose wno did t.ta.e origlna] WOPK and mace tr.is sacondapy
SCUPCS of inforaaf.cn possible.
Finally, tfie autfiops wisn to exopass tneip apppeciaticn :o :.ie mercers of
:ne Prcjec1: Acviscpy Ccranttaa:
1) ;'ap'< Captap, Ph.D.
Rcci
-------
EXECUTIVE SUMMARY
BACKGROUND
The objective of this Report is to review and summarize all disinfectant ra-
sidual aeasurement techniques currently available for free chlorine (along •-••_cr.
the various chloraaines), combined chlorine, chlorite ion, chlorine dioxice,
chlorate ion, and ozone.
Presently, both chlorine dioxide and ozone are gaining considerable favor as
alternatives to chlorine disinfection (1). The analytical chemistry for these
disinfectants when compared with chlorine is even more complex and less readily
understood as evidenced by various surveys (2-5) and detailed research carried
out in various laboratories (6-10).
Chlorine dioxide is manufactured at the site of its use by reactions involv-
ing sodium chlorite, chlorate ion, chlorine gas and/or hypochlorite ion and sul-
furic acid or hydrochloric acid (11-12). Consequently, chlorate ion, chlorite
ion, hypochlorite ion and/or hypochlorous acid frequently will be found occur-
ring as by-products or unreacted starting materials. These materials are strong
oxidizing agents which are very reactive and behave in many ways similar to
chlorine dioxide itself.
There are more than 2,000 wat«r treatment plants today using ozone, and less
than half of them are applying ozone solely for disinfection. The large major-
ity of water treatment plants use ozone as a chemical oxidant. Many of t.-.e
plants applying ozone for disinfection also are using ozone, in the sase slant,
for chemical oxidation. Analyses for residual ozone in water are asplicaole
only in the treatment plant, either in the ozone contactor(s) or at t.-.eir
outlets. Residual ozone is never present in the distribution system; however,
its by-products may be.
There have been numerous attempts to evaluate the relative advantages ar.d
disadvantages associated with the measurement of free and combined chlorine.
Different criteria are frequently used for the evaluation of the analytical
measurements and ..often suggestions for the improvement of test procedures have
gone largely ignored. No comprehensive and objective review of the literature
appears to be available. This Report is aimed at providing such a review along
with guidance and recommendations as to what criteria water utilities should use
in selecting residual monitoring techniques based on circumstances by category.
OBJECTIVES
1. To review and summarize all residual measurement techniques
currently available for free chlorine-- caking into account
the roles of chloraaines.
2. To review and suaaarize all residual measurement techniques
currently available for combined chlorine.
-------
3. To briefly review the present understanding of the chlorir.e-
amaonia chemistry and in particular, in relationship to che
measurement of chlorine and combined chlorine.
•*. To review and summarize all residual measurement techniques
currently available for chlorine dioxide, chlorite ion and
chlorate ion.
5. To review and summarize the analytical procedures currently
used by operating water utilities to control ozone treatment
processes, considering disinfection as well as the many oxid-
ative applications of ozone in water treatment applications.
6. To discuss common interferences associated with the measurement
of each of the disinfectants/oxidants described above (free
chlorine, combined chlorine, chlorite ion, chlorine dioxide,
chlorate ion, and ozone).
7. To provide guidance and recommendation for water utilities in
selecting residual monitoring techniques for each of the above
disinfectants/oxidants.
8. To recommend future research for development of monitoring and
analytical methods to improve accuracy, and reduce tise ar.d cost
requirements for the measurement of the above disinfectants.
In che full report, we present as complete as possible an examination of :.-.e
world-wide body of literature on analytical methods used by the water utility
industry in order to elaborate on the various problems, advantages, cisadvan-
tages and known interferences for each of the currently used analytical zetr.ccs
Foremost in our objectives has been a better understanding of the reliabil-
ity of various measurements which have been carried out. Since there are inr.er-
ent limitations in all measurements, it becomes apparent chat there are specific
needs for some indication of the reliability of the result, i.e., what is c.-.e
precision and accuracy of the reported value, and are these acceptable?
The volatility of mosc of the disinfectants makes sampling and sasoie
handling a major contributor to imprecision and inaccuracies. "Standard
additions" is a questionable technique; it should be avoided if possible, since
the pipetting and dilution process causes potential loss of disinfectant.
The relative usefulness of each method, along with descriptions of known
interferences such as turbidity, organic matter, ionic materials, solids, color,
buffering capacity, as well as the nature of the sample and the time between
collection of the sample and the actual analysis, are described in this report.
It must be emphasized, however, that almost invariably each of the methods
described is based on the total oxidizing capacity of the solution being
analyzed and is readily subject to interferences from the presence of other
potential oxidizing agents and/or intermediates from concomitant chemical
reactions. Under ideal conditions some of the methods are accurate to be::er
-------
chan ±1%--especially in the absence of common interferences--whereas ::r.er
methods are alaosc semi-quantitative in nature with many common spec.es
intertering with both Che precision and accuracy of the measurements.
V« have also included chlorate ion as a residual species in chat sr.lv
recently have reliable analytical methods begun to appear in the literature
(5,6,10). . Ve also report on the chemistry of the chlorine-ammonia system and
the associated breakpoint reactions, because one of the most common inteferer.ces
in the aeasurement of free chlorine is monochloraoine.
The most important development for this report has been the decision to in-
clude a preliminary section describing an "idealized" analytical method. The
need for this section became apparent as our writing progressed describing eacn
of the analytical methods for chlorine. Specific items included in this "ideal-
ized" method are accuracy, precision, reproduciblity, lack of interferences,
ease of use of the method, lack of false positive values, and so forth.
The benefit of the "idealized" analytical method is to allow individual com-
parisons and to allow the choice between various methods based on ir.divicual
aethod shortcomings. For example, a particular method might have as its or.lv
difficulty interference by manganese and iron. In many circuastar.ces. this r.-se
of interference might be a major problem. However, should the water surplv
under consideration not have any manganese or iron, it is quite likely chat :r.e
aethod might be very usable--and as a matter of face well aight be the best
aethod of choice.
In other cases, speed of analysis rather than potential interferences or
choice of some other important characteristic) might be the guiding factor :r.
choosing an analytical method. In this way rational choices can be cade :sseci
on potential and/or real difficulties and/or interferences and as compared co in
"idealized" aethod -- rather tnan a possibly controversial existing zetr.oc.
Table I has been constructed as a quick reference guide to the availaale
methods for the determination of water disinfection chemicals and bvprocuccs
Included are pertinent analytical characteristics such as detection lisits,
working range, interferences, accuracy and precision estimates. The current
status of the method, as gleaned from this report. is given. The r.ecessarv
operator skill level is given to aid the laboratory manager in assessing c.-.e
manpower requirements for a particular method. Additional informal::. or.
concerning the reasons for the current status is contained in the Recommendation
Section of the Executive Suaaary and the coisplete report.
As each of the methods is described in detail in the full report, specific
conclusions are drawn--along with appropriate recoanendations- -by comparing cr-.e
method against the "idealized" analytical method for that species.
One additional benefit of this direct comparison is the possibility of aid-
ing or subtracting a method to the list of Standard Methods for rhe Exanirat-.c^.
of 'Jater ard Wastewater. (13), based on a rational set of criteria. It snou^d
also be possible in the future to decide on the viability of various rretr.ccs
based on their meeting specific criteria rather than based only on corr.pansors
between analytical laboratories (and personalized subjective reactions :o :r.e
various mecnods themselves
-------
TA3LZ I. CHARACTERISTICS AND COMPARISONS OF ANALYTICAL METHODS'5
Species' DETECTION WORKING EXPECTED EXPECTED
TYPE OF TEST MEASURED LIMIT RANGE ACCURACY PRECISION SKI1
(METHOD)* DIRECTLY (mg/L) (mg/L) (± %) (± %) LIVE
k
FREE CHLORINE
f
"Ideal- Cl, + HOC1/OC1- 0.001 0.001 - 10 0.5 O.I
UV/VISI3LE C12 + HOC1/OC1- - 1 ( 1-100 NR NR 3
Continuous C12 + HOC1/OC1' 1.5 1.5-300 NR NR 3
AMPEROMETRIC TITRATION:
Forward C12 -(• HOC1/OC1' 0.00181 > 10 NF NT 2
0.02 - 0.032 > 10 NF 3-50 2
Sack C17 t- HOC1/OC1' 0.002 > 10 3-50 NF 2
Continuous C12 + HOC1/OC1' 0.005 > 10 N*R 1.0 2/3
IODOMETRIC TITRATICN:
Standard (Total Chlorine) 0.073 0.1-10 Ml :,-R :
0.354 0.5-10 N-R NR ;
DPD
FAS Tit'n C12 -*• HOC1/OC1' 0.004s 0.01-10 NT 2-7
0.0114 0.01-10 NT 2-7
Color'atrc Cl, •»• HOC1/OC1' 0.01s 0.01-10 1-15 1 - U
Steadifac Cl, + HOC1/OC1" 0.01s 0.01 - 10 NF NR 1/2
Black and
Whittle Cl, * HOC1/OC1- 0.01 0.25-3 NF NR
Whittle &
Lapceff Cl, * HOC1/OC1' 0.01 0.25-10 NR 0 - 10
-------
TASLI I. CHARACTERISTICS (conc'd)
STABILITY
REAGENT PRODUCTS INTERFERENCES ?H RANGE
FIELD
TEST AUTOMATED
CURRENT
5 YRS
NA
NA
1-2 yrs
1-2 yrs
1-2 yrs
1-2 yrs
1 yr
1 yr
powder
stable*
powder
stable8
powder
stable8
powder
stable8
months
months
> 1 DAY
NA
NA
NA
NA
NA
NA
10 nin
or less
10 min
or less
30 oin
30 min
30 min
30 min
NR
Ml
NONE
ClNHj - C13N
backgnd Abs
C1NH, - CljN
C1NH, - CljN
C1NH, - CljN
C1NH, - CljN
C1NH, - C13N
All oxidizing
species
All oxidizing
species
C1NH, - CljN
oxid species
C1NH, • CljN
oxid species
C1NH, - C13N
oxid species
C1NH, - C1,H
oxid species
rl *TH n v
uivTiA • w i • fct
oxid species
Oxidizing
species
Independent
pH Dependent
pH Dependent
pH Dependent
pH Dependent
pH Dependent
pH Dependent
pH Dependent
pH Dependent
Requires
buffer
Requires
buffer
Requires
buffer
Requires
buffer
Requires
buffer
Buffering
YES
NO
NO
YES
YZS
YES
YES
NO
NO
NO
NO
YES
YES
YES
YES
YES
NO
YES
YES
YES
YES
YES
NO
NO
NO
NO
NO
NO
NO
NO
RECOMMENDED
RECOMMENDED
(LAS TEST)
CONT'D STVDY
RECOMMENDED
RECOMMENDED
RECOMMENDED
RECOMMENDED
(LAS TEST:
RECOMMENDED
(LAS TEST:
RECOMMENDED
(LAS TEST)
RECOMMENDED
(LAS TEST)
RECOMMENDED
(FIELD TEST)
RECOMMENDED
(FIELD TEST)
ABANDON
RECOMMENDED
(LAS TEST)
-------
:ABL£ :. CHARACTERISTICS AND COMPARISONS OF ANALYTICAL METHOD
TYPE OF TEST
(METHOD)*
Species* DETECTION WORKING EXPECTED EXPECTED
MEASURED LIMIT RA^NGE ACCURACY PRECISION S:
-------
TABLZ I. CHARACTERISTICS (cons'd)
STABILITY
REACE.NT PRODUCTS
2 years7 30 min
at high
2 years7 30 min
ac high
NF NF
NF NF
NF 15-20 min
NF 55 min
INTERFERENCES
Oxidizing
CL, species
Oxidizing
C17 species
Oxidizing
species
Oxidizing
species
Oxidizing
species
Oxidizing
species
pH RANGE
Buffering
critical
Buffering
critical
Buffering
required
Buffering
required
NR
NR
FIELD
TEST
YES
YES
YES
YES
NO
NO
CURRENT
AUTOMATED STAG'S
NO RECOMMENDED
NO RECOMMENDED
NO ABANDON
NO A3ANDCN
NO ABANDON
NO A3ANDCN
Ml
NR
NR
1 sec
sec
sec
None
Oxidizing
species
Non«
Independent NO POSSIBLE A3ANDON
pH Dependent NO POSSIBLE CONT'D STUD
pH Dependent NO YES CONT'D srJD
NA
NA
NA
NA Oxidizing
Cas species
NA Oxidizing
species, Cl'
3 months NA
Dependent POSSIBLE POSSIBLE
on pH
CONT'D STUDY
NR
POSSIBLE POSSIBLE CONT'D STUDY
Oxidizing
species Cl"
NA Oxidizing
species, Cl"
pH Dependent YES YES RECOMMENDED
Buffer POSSIBLE POSSIBLE
required
-------
TABLE I. CHARACTERISTICS AND COMPARISONS OF ANALYTICAL .".ETKCCS0 icor.r'i
Species* DETECTION '-ORKING EXPECTED EXPECTED
TYPE OF TEST MEASURED LIMIT RANGE ACCURACY PRECISION SKIL
(METHOD)* DIRECTLY (mg/L) (mg/L) (± %) (± %) LEVE
TOTAL CHLORINE•
•Ideal" Cl, + HOC1/CC1- 0.001 0.001-10 0.5 0.1
NH,C1 NHC1, NCI,
AMPEROMETRIC TITRATION:
Forward Cl, -t- HOC1/OC1' 0.00181 > 10 NF NF 2
NH,C1 NHC1, NCI,
Cl, + HOC1/OC1- 0.02 -0.032 > 10 NF 3 - 50 2
NHjCl NHC1, NC13
Sack Cl, * HOC1/OC1- 0.002 > 10 3-30 NF 2
NH,C1 NHC1, NCI,
Concinuous Cl, -t- HOC1/OC1" 0.005 > 10 NR 1.0 2/3
NH3C1 .N*HCla NCI,
IODOMETRIC TITRATION:
Standard Cl, f HOCl/OCl' 0.073 0.1-10 MR rrR ;
NH,C1 NHC1, NCI3
Clj f HOC1/OC1- 0.35« 0.5 - ICO N'R N"R :
NH,C1 NHC1, NClj
DPD
FAS Tic'n Cl, + HOC1/OC1- 0.0043 0.01 - 10 NF 2-7 l
NH,C1 NHC1, NCI,
ci, + Hoci/ocr 0.11* 0.01-10 NF 2-7 i
NH,C1 NHC1, NCI,
Color'mcrc Cl, t- HOCl/OCl' 0.001s 0.01 - 10 1 - 15 1 - lu
NH,C1 NHC1, NCI,
LCV
Black 6.
Vhiccle Cl, f HOCl/OCl- 0.005 0.25-3 NF A - 10 1
NHJC1 NHC1, NCI,
-------
TABLE I. CHARACTERISTICS (cont'd)
STABILITY FIELD CURRENT
REAGENT PRODUCTS INTERFERENCES pH RANGE TEST AUTOMATED STATUS
5 YRS
1 - 2 yrs
1 - 2 yrs
1 - 2 yrs
L - 2 yrs
1 yr
1 yr
powder
stable*
powder
stable*
powder
stable8
months
> 1 DAY NONE
NA Oxidizing
Species
NA Oxidizing
Species
NA Oxidizing
Species
NA Oxidizing
Species
10 ain All oxidizing
species
10 min All oxidizing
species
30 min . Oxidizing
Species
30 min Oxidizing
Species
30 min Oxidizing
Species
.'«*R Oxidizing
Species
Independent YES
of pH
pH Dependent YES
pH Dependent YES
pH Dependent YES
pH Dependent YES
?H Dependent NO
pH Dependent NO
Requires NO
buffer
Requires YES
buffer
Requires YES
buffer
Requires YES
buffer
YES RECOMMENDED
YES RECOMMENDED
YES RECOMMENDED
YES •3.ECCwrVi'EN'DrD
YES RECOMMENDED
NO RECOMMENDED
(LA3 "ST^
NO RECOMMENDED
(LAB TEST}
NO RECOMMENDED
(LAB TEST)
NO RECOMMENDED
(FIELD TES'
NO RECOMMENDED
(FIELD TES
NO ABANDON
-------
TASLZ I. CHARACTERISTICS AND COMPARISONS OF ANALYTICAL METHC2S'9 'c=n:'i^
TYPE OF TEST
(METHOD)*
Species' DETECTION WORKING EXPECTED EXPECTED
MEASURED LIMIT RANGE ACC'JRACY PRECISION SKILL'
DIRECTLY (mg/D (aig/L) (± %) (± %) LEVEL
Vhiccle &
Lapceff Cl_ -t- HOC1/OC1' 0.01 0.25-10 NF i - 10 2
NHJC1 NHC12 NCI,
FACTS
Color'acre Cl, + HOC1/OC1" 0.1
NH,C1 SHCla NCI,
0.25 -10 5 - 20 1 - 11
Specc'phoco Cl, -t- HOC1/OC1- 0.012 0.05 - 10 SF
NC1
ELECTRODE METHODS
Pot'=ecric Cl, -t- HOC1/OC1' 0.005 0.01 - 1 1 - 6 7 - 10
NH2ci jmcij NCI, •
MONOCHLORAMINE9
"Ideal'
0.001
0.001-10 0.5 0.1
L'V/VISISLE
NH,C1
- 1
100
NTl
AMPERCMETRIC TITRATION:
Forward NH2C1
Back NH2C1
NR
MR
> 10
> 10
NF 0-10
NF NF
DPD
FAS Tic'n
Color'acre
NH,Cl
NHjCl
NR
NR
0.01 - 10
0.01 - 10
NF 2 • 7 1
NF 5 - 75 1
10
-------
TABLE 1. CHARACTERISTICS (conc'd)
STABILITY
REAGENT PRODUCTS INTERFERENCES pH RANGE
FIELD CURRENT
TEST AUTOMATED STATUS
aonths .S"R
Oxidizing
Species
Buffering YES
NO RECOMMENDED?
(LAB TEST)
2 YRS
2 YRS
30 min
at high
C13
30 min
ac high
Cl,
Oxidizing
Species
Oxidizing
species
Buffering YES NO
critical
RECOMMENDED
Buffering YES
critical
NO
3 months NA
Oxidizing pH Dependent YES
Species, Cl*
YES RECOMMENDED
5 YRS > 1 DAY
NONE
Cndependent YES
NA CljNH - CljN
backgnd Abs
pH Dependent NO
NO
RECChru~'l.'~r"'
1-2 yrs NA
1-2 yrs NA
C1,NH - CljN pH Dependent YES YES
C1,KH - CljN pH Dependent YES YES
RECOMMENDED
RECOMMENDED
powder 30 min
stable9
powder 30 min
stable4
C1MH, - CljN
oxid species
C1NH, - CljN
oxid species
Requires
buffer
Requires
buffer
NO
YES
NO
NO
RECOMMENDED
(LAB TEST)
RECOMMENDED
(FIELD TES1
11
-------
TABLE .. CHARACTERISTICS AND COMPARISON'S OF ANALYTICAL METHODS3 [ccr.t'd;
TYPE OF TEST
(METHOD)*
Species* DETECTION VORXING EXPECTED EXPECTED
MEASURED LIMIT RANGE ACCURACY PRECISIC'J
DIRECTLY (mg/L) (rag/L) (± %) (* %)
Whittle &
Lapteff
NH,C1
0.25-10 NF
ELECTRODE METHODS
Silver iodide
Voltaooetrie
NH7Cl
.VR
0.1 - 10
DICHLDRAMINE^
•Ideal-
0.001 0.001 - 10 0.5
0. 1
UV/VISI3LE
NHC1,
- 1
1 - 100
AMPERCMETRIC TITRATICN:
Forward N"HC12
Back NHC1,
.VR
> 10
> 10
NF 3
3-50 NT
DPD
FAS Tic'n
Color'mtrc
NHC1.
NHC1,
NR
0.01
0.01
10
10
NF
NF
0 - 100
LCV
Whittle &
Lapccff
NHC1.
NR
0.25 - 10
NF
10 - 150 2
-------
TABLE I. CHARACTERISTICS (cont'd)
STABILITY
REAGENT PRODUCTS INTERFERENCES pH RANGE
FIELD
TEST AUTOMATED
CURREN:
STATUS
months
NR
Oxidizing
species
Requires
buffer
YES
NO
g rr^v»v«r>"«r-»
NA
NA Oxidizing Requires POSSIBLE POSSIBLE CONT'D STUDY
species buffer
5 YRS > 1 DAY
NA
NA
NONE
C1NH, & CljN
backgnd Abs
Independent
of pH
YES
?H Dependent NO
YES RECOMMENDED
NO RECOMMENDED
,LAB TEST:
1-2 vrs
1-2 vrs
S'A C1NH, & C13N pH Dependent YES
NA C1NH, & CljM ?H Dependent YES
YES RECCMMENDE2
YES RECOMMENDED
powder 30 ain
stable4
powder 30 ain
stable*
months
NR
INH, & CljN
oxid species
INK, & C13N
oxid species
Oxidizing
species
Requires
buffer
Requires
buffer
Requires
buffer
NO
YES
YES
NO
NO
NO
RECOMMENDED
(LAB TEST)
RECOMMENDED
(FIELD TEST:
RECOMMENDEDD
(LAB TEST)
13
-------
TABLE I. CHARACTERISTICS AND COMPARISONS OF ANALYTICAL METHODS* (cor.fi
Species' DETECTION ','CRKING EXPECTED EXPECTED
TYPE OF TEST MEASURED LIMIT RANGE ACCURACY PRECISION SKILL'
(METHOD)* DIRECTLY (ng/L) (mg/L) (± %) (± %) LEVEL
TRICHLORAMIKE9
"Id«al" NCI, 0.001 0.001 - 10 0.5 0.1 i
UV/VISI3LE NCI, NR NR :TR N"R 3
AMPEROMETRIC TITRATION:
Forward NC13 NR > 10 NF 5 - ICO 2
DPD
FAS TiC'n NCI, NR 0.01 - 10 Ml Ml 1
Color'acre NC13 • NR 0.01 - 10 Ml Ml
LCV
Vhitrle & NC13 Ml 0.25 • 10 NR N"R ;
Lapceff
CHLORINZ DIOXIDE
•Ideal' CIO, 0.001 0.001-10 0.5 0.1 1
ICDOMETRIC " CIO, 0.002 0.002 - 95 1 - 2 1-2 2
AMPEROMETRIC CIO,10 0.012 0.02 - ?? 1 - 15 1 - 15 3
DPD CIO,10'11 0.008 0.008 - 20
UV
Manual CIO, 0.05 0.05-500 5
FIA CIO, 0.25 0.25 - 142 2
-------
TABLE I. CHARACTERISTICS (cont'd)
STABILITY
REAGENT PRODUCTS INTERFERENCES
5 YRS
NA
1-2 yrs
powder
stable8
powder
stable8
xonchs
5 YRS
1 VR
good
solid
stable8
none
none
> 1 DAY NONE
NA GINK, - CUNH
backgnd Abs
HOC 1/0 CL-
NA C1NH7 - C1,NH
30 Bin C1NH, - C12NH
oxid species
30 min C1NH, - C1,NH
oxid species
NR Oxidizing
species
> 1 DAY NONE
Subject to Oxidizing
oxidation species
Subject to Metal ions &
oxidation nitrite ion
< 30 oin Oxidizing
species
none Other UV
absorbers
none none
FIELD CURRENT
pH RANGE TEST AUTOMATED STATUS
Independent YES
pH Dependent NO
pH Dependent NO
Requires NO
buffer
Requires YES
buffer
Requires YES
buffer
Independent YES
2-5 NO
7 NO
7 NO
Independent NO
Independent NO
YES RECOMMENDED
NO RECCMMENDED
(LAS TIST/
YES RECCXMENDED
/T .3 ••— C"" ^
1 t>O . ^0 . ;
NO RECCMME1IDED
r.\3 TEST)
NO RECOMMENDED
(LA3 TEST^
NO RECCXXENDED
i, — -k3 T i 3 . /
YES RECOMMENDED
NO NOT RECOMMENDED
NO CURRENTLY USED
NO NOT RECOMMENDED
YES RECOMMENDED
(H3 TEST)
YES RECOMMENDED
(LAS TEST)
15
-------
TABLE I. CHARACTERISTICS AND COMPARISONS CF ANALYTICAL METHODS'* (ccr.c'
Species' DETECTION VCRKING EXPECTED EXPEC'
TPE OF TEST MEASURED LIMIT RANGE ACCURACY PRECI.
(METHOD)' DIRECTLY (mg/L) (ag/L) (± %) (± '
ACVK12
CHLOROPHENOL RED
o-TOLIDINE
INDIGO BLUE
CHEMILUMINESCENCE
Luoinol
CDFIA13
ELECTROCHE.M.
CIO, 0.04 0-25 NR .N"R
CIO, 0.003 0.003 - 1 10 5
CIO, 0.1 NR NR Ml
CIO, 0.01 NR N*R 1.5
CIO, 0.3 0.3-1 NR 3
CIO, 0.005 0.005-74 2
PC Microelec. CIO, + CIO,* -1.3 NR 7 N"R
Vic. Carbon
Volcaa. Mem.
CIO, 32 NR NR NR
CIO, 0.25 NR N-R NR
Rotating Vole.
Membrane CIO, 0.30 0.30-3 NR 5.4
CHLORITE ION
•Ideal- CIO," 0.001 0.001 - 10 0.5 0,1
AMPEROMETRIC
lodomecric CIO,' 0.05 0.05 - 95 5 5
IODOMETRIC
Sequential CIO,' 0.011 > 1 1 1
Modified CIO,' 0.3 0.5-20 0.5 1-3
DPD CIO," 0.01 0.01 • 10 5 5
2/3
16
-------
TA3LE I. CHARACTERISTICS (conc'd)
STABILITY
REAGENT PRODUCTS INTERFERENCES ?H RANGE
FIELD CURREN1
TEST AUTOMATED STATUS
NR
6 months
NR
good
1 DAY
1 DAY
none
none
none
NR
NR
NR
good
< 1 sec
< 1 sec
none
none
none
minimal
unknown
Oxidizing
species
0, Cl,
NR
ci,
cio7-
cio,-
HOC1
3.1-3.4 NO
7 YES
ra NO
> 4 NO
NR NO
> 12 NO
5-5.5 NO
3.5-7 NO •
7.8 NO
NO
NO
NO
NO
NO
YES
NO
NO
NO
~ <^ »»^« » -^ ^ ^» «^«
-v.i ^ J i i , -
NOT RECOMMEND
NOT RECOMMEND
NOT RECOMMEND
NOT ?,ECCMyrND
RECOMMENDED
:ONT'D STUD
CONT ' D S i wD
uONT ' D ST". D
CCNT'D Sr.'D
i
ID
ID
•""S
«•/
12
;Y
Y
'•
none
none
HOC1
5 - 5.5
NO
NO
5 YRS
> I DAY
NONE
Independent YES
YES
RECOMMENDED
1 YR
Subject co
oxidation
Oxidizing
species
2 - 5
NO
NO NOT RECOMMENDED
good
good
Solid
stable8
Subject to
oxidation
Subject to
oxidation
< 30 min
Metal ions &
nitite ion
Metal ions &
nitite ion
Oxidizing
species
NO NO RECOMMENDED AT
HIGH CONC.
NO NO CONT'D STUDY
NO NO NOT RECOMMENDID
17
-------
TA3L2 I. CHARACTERISTICS AND COMPARISONS CF ANALYTICAL METHODS5
TYPE CF TEST
(METHOD)'
CHLORATE ION
•Ideal"
IODOMETRIC
Sequencial
Modified-
FIA
DPD
OZONE
•Ideal'
ARSENIC BACK
TITRATICN
FACTS
DPD
INDIGO
Spect'phoco
Species'
MEASURED
DIRECTLY
c:o,-
cio3-
ClOj-
ciOj-
cio3-
°3
°3
o,
°3
Qj
°3
DETECTION
LIMIT
(ag/L)
0.001
0.064
0.3
0.03
0.01
0.01
0.002
0.002
0.02
0.1
0.001
0.006
0.1
WORKING EXPEC
RANGE ACCUR,
(ag/L) (±
0.001 - 10 0
> 1
0.3 - 20
0.08 - 0.3 3
0.01 - 10
0.01-10 0
0.5 - 100
0.5-65
0.5-5 5
0.2-2 5
0.01 - .1
0.05 - .5
> 0.3
•»«•<
* w<
AC
%)
.5
2
1
. 5
5
. 5
-
-
•
•
1
1
1
c •.•••• «
(± %)
3.1
1 - 3
35 1
0.5
0.5
0.5
13
-------
TABLE I. CHARACTERISTICS (cont'd)
STABILITY
REAGENT PRODUCTS INTERFERENCES pH RANGE
FIELD CURREN:
TEST AUTOMATED STATUS
5 YRS
good
good
1 year
Solid
stable*
5 YRS
1 YR
1 YR
2 YRS
Solid
stable8
good
good
good
> 1 DAY
Subject to
oxidation
Subject to
oxidation
1 day
< 20 min
> 1 DAY
subject co
oxidation
subject to
oxidation
no fading
first 5 aiin
< 30 min
good
good
good
NONE Independent YES
Metal ions & 7 NO
nitrite ion
Metal ions & 2 NO
nitrite ion
Oxidizing < 1 NO
species
Oxidizing 7 NO
species
NONE Independent YES
All ozone < 2 NO
by products
and oxidants
Oxidizing 6.8 NO
species
Oxidizing 6.6 NO
species
Oxidizing 6.4 NO
species
Cl,, Mn ions 2 NO
Br, I,
Cl,, Mn ions 2 NO
Br, I,
Cl,, Mn ions 2 NO
YES RECOMMENDED
NO RECOMMENDED AT
HIGH CCNC.
NO CONT'D STUDY
YES USED AFTER ALL
'"'D " "* 3 " — «x.-.
NO NOT RECOMMENDED
VIS R£CC— C-
NO ABANDON
NO CONT'D STUDY
NO NOT RECOMMENDED
NO SOT RECOMMENDED
YES RECOMMENDED
YES RECOMMENDED
YES RECOMMENDED
Br2 I.
19
-------
TA3LE I. CHARACTERISTICS AND COMPARISONS 0? ANALYTICAL METHODS5 (conc'd;
(±
Species'
V?E OF TEST MEASURED
(METHOD)* DIRECTLY
INDIGO (cont'd)
Visual 03
GDFIA 03
LCV 03
ACVK 03
o-TOLIDINE 03
3ISTERPYRIDINE 03
CARMINE INDIGO 03
ELECTROCHEM
Aaperonecric Total
oxidants
Amperometric
iodometric - Total
Oxidants
Bare electrode 03
Membrane elect. 0,
Differential
Pulse Dropping
Mercury 03
Differential
Pulse Polar-
ography 03
Pocenciometnc 03
DETECTION FORKING EXPECT!
LIMIT RANGE ACCURAf
(mg/L) (ag/L) (± %
0.1 0.01 - 0.1
> 0.1
0.03 0.03 - 0.4
other ranges
possible
0.005 NR
0.25 O.C5 - 1
NOT QUANTITATIVE
' 0.004 0.05 - 20
< 0.5 NR
- 1 NR
- 0.5 NR
0.2 NF
0.062 NF
NR NR
0.003 NR
NR NR
5
5
1
SR
N-R
N"R
2.7
NR
5
5
5
5
Nl
Nl
V
0.5
NR
5
5
NH
20
-------
TA3LZ I. CHARACTERISTICS (conc'd)
STABILITY
REAGENT PRODUCTS INTERFERENCES pH RANGE
FIELD CURREN
TEST AUTOMATED STATJS
good
good
good
Stable
MR
NR
Good
NR
r.one
1 YR
none
none
none
none
none
good
good
good
Stable
NR
NR
Good
NR
NA
Subject to
oxidation
NR
NR
NR
NR
NR
Cl, , Mn ions 2
Br, I,
Cl, . Mn ions 2
Br, I2
Cl, at > Img/L 2
S2' SO3' Cr*" 2
Mn > 1 mg/L 2
Cl, > 10 mg/L
Metal ions, N02' 2
C12 < 7
NR 2
Oxidizing 2
species
Oxidizing 4-4.5
species
NR NR
NR NR
NR NR
NR 4
NR NR
YES
YES
NO
NO
NO
YES
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO RECOMMENDED
NO RECOMMENDED
YES COMPARISON
STUDIES
NEEDED
NO CONT'D STUDY
NO CONT'D STUDY
NO A3ANDCN
YES RECOMMENDED
NO CO NT ' D STUDY
YES RELATIVE
MONITORING
NO NOT RE CC MM END ED
YES CONT'D STUDY
POSSIBLE CCNT'D STUDY
NO RESEARCH LA3
NO CCNT'D STUDY
YES CCNT'D STUDY
21
-------
TABLI I. CHARACTERISTICS AND COMPARISONS OF ANALYTICAL METHODS5
•YPE OF TEST
(METHOD)*
Species' DETECTION -ORKING EXPECTED EXPECTED
MEASURED LIMIT RANGE ACCURACY PRECISION
DIRECTLY (mg/L) (fflg/L) (± %) 'r %)
UV
0.02
> 0.02
0.51'
ISOTHERMAL
PRESSURE CHANCE
x 10-s 4 x 10-s - 10 0.5
0.5
OZONE GAS PHASE
"Ideal"
uv
1 1 - 50,000 I
0.5 0.5 - 50,000 2
2.5
Stripping
Absorption
lodoaetry 03
Chemiluainescence 03
Gas phase titration 03
Rhodaaine S/
Gallic Acid 03
Aaperoaetry Os
0.002 0.5 - ICO 1-35 1-2 2
0.005 0.005 - 1 7 5 1,2
0.005 0.005 • 20 . 3 35 2
0.001 Ml Ml 5
Ml Ml Ml Ml
' for page nuabers in the full report, refer co che Alphabetical Ir.ciex
f direct determination of the species measured without interferences
* Operator Skill Levels: 1 - minimal, 2 - good technician,
3 - experienced chemist
SA Not applicable
Ml Not reported
NF Not found
1 Using research grade electrochemical equipment
2 Using commercial cicracor
3 Speccrophocomecric endpoint detection
>4 Visual endpoinc detection.
5 Using tesc kic
6 Liquid reagent is unstable
7 Scablility is very dependent on the purity of the 2-propanol used
-------
TABLE I. CHARACTERISTICS (cor.t'd)
STABILITY
REAGENT PRODUCTS INTERFERENCES pH RANGE
none
none
good
Other
Absorber
none
Ir.deoendent
Independent
FIELD
TEST AUTOMATED STATUS
N'O
YES ESTABLISH
MOLAR ABSORB -
TIVITY
YES COMPARISON
STUDY
none
none
none
none
none
none
Independent YES YES
NA YES YES
RECOMMENDED
RECOMMENDED
good good
stable < 1 sec
stable stable
problems
none
none
SO, NO;
none
none
MR
NR
NA
NA
NA
NA
NA
YES NO ABANDON
YES YES RECOMMENCED
YES NO NOT RECOMMENDED
YES POSSIBLE SOT RECOMMENDED
YES YES NOT RECOMMENDED
8 Total Chlorine is all chlorine species with +1 oxidation state
9 Very little^accual work has been carried out on selective determination
of chloraaine*. The values reported are from extrapolated studies char
had objectives other than the selective determination of chloraair.es.
Most methods are indirect procedures which are not recommended
10 Indirect method
11 1/5 of CIO, determined
12 Acid chrome violec potassiua (ACVK)
13 Gas diffusion flow injection analysis (GOFIA)
14 Based on current molar absorbtivity and proper saaple handling tecniques
Current best estimates of molar absorbtivity of 2900-3300 give a.
possible error of > 10%.
® Taken from Cordon, Cooper, Rice, and Pacey, AWA-RF Review on
"Disinfectant Residual Measurements Methods" (1987)
-------
Chapter 4 (Indexed Reference Citations) has been included in this report in
order to assist readers in locating particular papers of interest. 7r.e -3
categories for chlorine, chloranines. and the oxy-chlorine species, along vitn
the additional 60 categories for ozone, should make the task of finding ir. •
dividual papers of interest considerably less cumbersome. Papers vnich deserve
several methods have been included in each of the appropriate categories. .-.11
togetr.er. the 1,400 references cited in Chapters 1-3 r.uzoer ;ore than 2. 120
individual citations when distributed in the indexed fora of Chapter V
Chapter 5 is an alphabetical listing of the individual references citations
Finally, a detailed Index has been included in order to assist readers in
locating subjects of specific interest, '-e hope the readers -'ill find t.-.ase
additional chapters as useful as have we in preparing this report.
RECOMMENDATIONS
General Statements on Comparisons.
There have been and will continue to be reports of aethods corranson. 7r.e
of the most important considerations for a method is accuracy, i 2. tr.e asili;-.-
of the method to determine the correct concentration of a disinfectant :n
solution. An equally important consideration is precision, i.e. r.ow veil cicas
the analytical method reproducibly measure the saam concentration. Frequer.tlv
experiments are conducted to determine the "equivalency" of the zetr.ccs. ~r--z
such results, methods may be found to be equivalent, but the only aralvti:a.
considerations tested were accuracy, as judged by a Referee .a.etr.oc. —£
precision, judged for each metnod based on the experimental design.
No considerations were given to specificity or analyst preference. Vac ere
of the most difficult tasks in the area of disinfection analytical metr.oss
development is comparison testing. It is recomn«nd«d that a protocol ;e
developed to initiate comparison of the disinfectants. This protocol should
include all of the factors delineated in the "Ideal Method" and snould be
undertaken in both laboratory controlled conditions and at selected wacer
treatment plants around the country.
Chlorine Chemistry.
Clearly, the conversion to moles, equivalents, or normality fron units of
mg/L (as C12) or mg/L (as other oxidants) can easily be confused (and
confusing). Our recommendation is that all oxidizing agents be reported in molar
units (M) and, if necessary, in mg/L of that oxidizing agent as measured (i e.
ag/L (as C17) or mg/L (as C10S") or mg/L (as C10S") . Furthermore, we recotsreend
that oxidizing equivalents per mole of oxidant be reported to minimize
additional potential confusion. For example, when CIO, is reduced to CIO.',
this corresponds to one equivalent/mole; on the other hand, when CIO; is reduced
to Cl" , this corresponds to five equivalents/mole. A sugary of molecular
weights and oxidizing equivalents for the various chlorine species, oxychlorire
species and ozone is given in Table II.
-------
TA3L£ II. EQUIVALENT WEIGHTS FOR CALCULATING CONCENTRATIONS
BASIS OF MASS.
Molecular Zaui
Species
Chlonr.t
Monochlo raaine
Dichloraaine
Trichloraaine
Chlorine dioxide
Chlorine dioxide
Chlorite ion
Chlorate ion
Ozone
Ozone
'-eight
g/nol
70.906
51.476
85.921
120.366
67.452
67.452
67.452
33.451
47.998
47.998
Electrons
Transferred
2
2
4
6
1
5
4
6
2
6
ON THE
valent
'-eight
s/eq
35.
25.
2i
20.
67.
13.
4 1"
13.
23.
3.
453
733
430
061
452
490
363
909
999
000
Several mechanisms have been proposed for the decomposition of dichloramir.e ,
but the complete aecnanism ac the breakpoint has not been resolved. Cleariv. the
chemistry is complicated and varies markedly with solution composition. -\
detailed understanding of the specific reactions involved requires a detailed
knowledge of the concentration of all chloraaune species in the system.
Nitrogen-containing organic compounds may b« present in surface water and
ground-water. Because of analytical complexities, very few detailed studies
have been undertaken to determine the individual compounds present and the
concentration at which they exist. KJeldahl nitrogen analysis is used
frequently, but this dots noc provide any dtcailtd information with regard to
individual compounds. The area of organic nitrogen and the determination of
specific compounds in natural waters is one of the increasing interest ar.d
requires considtrably aore research in characterization and methods development.
Ultraviolet Methods.
In general, because the molar absorptivities are quite low for chlorine and
chloramine species, ultraviolet methods are noc considered useful in routine
monitoring of chlorine residuals. In addition co the low molar absorptivities.
there is often background absorbance that may interfere with the measurement in
various natural waters. However, these measurements are of use in standardising
the chlorine species in distilled waters and are often used in experimental vortc
25
-------
related to chlorine speculation. This method docs have considerable pocert.a.
for the determination of relacively high concentrations of halogens.
particularly in relatively clean water. This method night find use ir.
monitoring chlorine species in water treatment plants. However, with a r.ora
elaborate aultiwavelength spectrophotoneter and computer-controlled spectral
analysis, it mignt be possible co analyze several halogens simultaneously.
Ir is also possible that additional methods using permeable nembrar.es ccu.d
be developed for the simultaneous determination of chlorine species in aqueous
solution. Additional work is necessary in this area. Alchougn the rsoLar
absorptivities of the species is not of a magnitude as to lend it to the routine
determination of the dilute (less than 10"s M) chlorine and chlonr.e-araor.ia
species, it is potentially helpful in determining the concentration of stansara
solutions. Absorption spectrophotometric analysis has and will continue to ;e
very important in the area of chlorine chemistry. It can be used in the
unambiguous determination of relacively high concentrations of rr.e species in
relatively purt water.
Continuous Aarperomecric Titration Method.
Interferences appear to be reduced using the continuous arr.perometnc zetr.sc.
because the reagents are added to the saople just prior to contacting tr.e
indicating electrode. Thus, when compared to the amperometric titration. tr.e
amount of interference by iodate ion, bromate ion. copper(II), ironilll), ana
manganese( IV) is reduced by approximately one-tenth. N'o reports appear to ;e
available in the literature -on the determination of mixed oxidants using the
amperometric method. Such experiments need to be carried out. In addition. few
experiments have been reported which clearly demonstrate that the electrodes
remain uncontarainated for drinking water or waste water systems. In tr.e absence
of such comparisons, the accuracy of any electrode proceoure -av :a
quescionaole.
However, the amperometric deration determination of chlorine species ra-
mains the standard for routine laboratory measurements. Given proper analvst
training and experience, the commercially available instrumentation is sensitive
and precise. This method should remain as the method for laboratory use and
accuracy comparisons. It requires more analyst experience than coloriaetric
methods, but can b« relied on to give very accurate and precise measurements
It should be noted that car* oust be exercised when using one titrator for tr.e
measurement of both free and combined chlorine. Small quantities of iodide ion
can lead to errors wh«n differentiating between free and combined chlorine.
Careful rinsing with chlorine demand free wattr (COFV) is a must! Additional
development of automated back-titration equipment with the goal of lowering the
limit of detection and improving the reproducibility would be highly beneficial.
lodometric Titration Method.
The iodometric titration is useful for determining high concentrations of
total chlorine. The most useful range is 1 ng/L (as C17) or greater. It is a
common oxidation-reduction titration analytical method and provides a reference
procedure for total chlorine. Although not necessarily used routinely, r.ost
laboratories use it as a reference method and it is not likely ever to be
eliminated from use.
26
-------
Coloriaecric Methods.
Ic is reported in Standard ^ethods (13) that nitrogen trichloride can be
measured using the D?D metnod; however, the method has not been confirmed by
independent investigations and should be used only as a qualitative method
Additional research is necessary to determine the effectiveness of the j?D
method for nitrogen trichloride. The effect of the presence of nercur-.:
chloride in the reagents for air.iaizir.g the breakthrough of laor.ochLoraraine ir.ro
the free chlorine reading vith the DPD methoa has been shovn. It is very
important that the addition of mercuric chloride to tne buffer be followed to
minimize the direct reaction of monochloramine with DPD. This phenomenon is not
thoroughly understood. This effect should be scudied more thoroughly and tr.e
principle may be applicable to all of the colorimetric methods.
The use of thioacetamide was evaluated for monochloraaine (using 2?D-
Steadifac). It was shovn under these conditions to eliainate any positive
inteference in the free residual measurement. These results are not as ye:
understood, but the implication is chat the chemistry of oxidation is different
for monochloramine and free chlorine. These results suggest that more work is
necessary to better define the reactions involved, and this may lead to a aore
usable analytical procedure. This procedure is recommended for use in waters
that are suspected to be relatively high in combined chlorine.
The DPD-Ethyl Acetate Extraction Procedure is a modification of the 2?D
chemistry. The method is based on the oxidation of iooide ion by active
chlorine followed by extraction of the iodine species into ethyl acetate. This
procedural modification may be of use in the determination of total residual
chlorine in both the field and laboratory. Additional work is necessary Before
it can be used to any great extent. It does not appear to offer SUBS tart:2!
advantages to the already well tested colorimetnc method for Laboratsrv
measurements.
The DPD methods have become the most widely used procedures for the measure-
ment of chlorine. This is not likely to change. The DPD color reagent,
liquid form, has been shovn to be quite unstable and is not recommended for use.
It is sensitive to oxidation by oxygen and thus requires a control measurement.
Clearly, it is better to use dry reagents.
Leuco Crystal Vlolec, LCV.
No studies have been reported that examine the interference of chlorine
dioxide and/or ozone in the LCV method. It is anticipated that these oxidancs
would interfere in the method, and studies should be conducted to quantify these
potential interferents.
Syringaldazine; FACTS.
A study using syringaldazine in a continuous method to differentiate free
from coaoine chlorine has been reported. It vas concluded that it could be used
and was useful in controlling free chlorination. Further work would have to be
conducted to use this or any coloriaetric method in continuous analyzers.
-------
Cheailufflinescence.
Several papers have appeared that detail the reaction of hydrogen reroxiie
and hypochlorous acid and the resulting chemiluninescence. The recnanisrs .".as
been relatively well established and the chemiluainescence is thougr.t to occur
as a result of the formation of singlet oxygen. The light emitted is reel ,fJ5
run), and occurs aost readily in alkaline solution. This reaction is ratr.er
insensitive to low concentrations and is not suitable for the determination of
hypochlorous acid in aqueous solution. However, the studies tr.at have bear.
reported can serve as a guide for those interested in pursuing other -ethods far
the determination of hypochlorous acid by chemiluainescence. It is not sensitive
enough to be considered as an analytical method for chlorine in vater traacrer.t.
A study has been reported chat details the use of luainoi for the
measurement of hypochlorite ion. The optimum pH for analysis was between 9.0 ar.d
11.0 Luainol also has been used for the determination of hydrogen peroxide.
4, 5 ,6 , 7 , - tetramethoxyluminol is 30 I more sensitive than luair.ol. Zither of
these compounds may be more sensitive in the determination of free chlorine. As
these compounds have not been tried it appears that acditional studies are
necessary. From the limited data available, it appears that this reaction .-.as
considerable promise as an analytical aethod. It say very veil be t.-.e rose
sensitive method to date.
Ic is reported thac lophine, in a reaction with hypochlorite ion. produces
light. Very few details were given in the study for this reaction. It appears
that lophine also may be good as. a chemiluainescence reaction systera for free
chlorine. Additional work should be undertaken to better characterise t.-.e
details of this reaction.
Luminol and some of its derivatives, or lophine, raay be well suites :or r.-.a
very sensitive measurements of chlorine species. Additional researcr. s.-.ou.c -e
undertaken to develop the use of chemiluainescence for use in t.-.e ieterr.ir.acior.
of chlorine in water. The potential exists for rapid, sispie, ani specific
methods for chlorine and possibly other oxidants. Vith the acver.t of firer
optic sensors and their application in chemiluainescence rethocis, t-is
technology will be important in the future.
Fluorescence.^
The use of rhodamine B has been reported as a low level fluoroaetnc ~etr.od
for the determination of bromine. This method is qualitatively specific for
bromine, although chlorine will react to decrease the fluorescence. The advant-
age of this method is chac it is capable of determining oxidants at very low
concentrations. This method could be applied to chlorine analysis by first
using the free chlorine to oxidize the bromide ion to bromine, an irreversible
reaction. followed by the determination of bromine. This aetr.od was rot
developed fully and very little work has been undertaken since tr.e first
publication. It does appear to have considerable potential and future researcr.
in the area of methods development should not exclude additional work on t-.-. s
fluorometric procedure.
23
-------
Other Electrode Methods.
Additional studies are required to better understand the limitations of
membrane electrode methods. It appears that they say have prominent roles to
play in chlorine residual measurements in the future.
In a series of experiments carried out for :he determination of fraa
chlorine in tap vater, it was observed that there was a statistical!'.'
significant difference between the results of the amperometric titracion and tr.e
membrane electrodes. It was thougnt to be a problem in the membrane electroces.
However, on reconsideration, it is possible that the electrodes were actuallv
giving a free chlorine reading and the amperomecric titration was reading t.-.a
sua of free and organically combined chlorine. The study was conducted on water
which is relatively high in organic nitrogen. It is possible that consideraoLe
chlorine is present as organically combined chlorine ano interferes in the
amperometric titracion procedure, but does noc interfere with the aeraorar.e
electrode measurements. This question oust b« resolved. Carefully designed
experiments to expicitly resolve these differences would be most appropriate.
There have been no reports of experiments using bare-electrode acperometnc
analyzers where other oxidants such as chlorine dioxide, chlorite ion, chlorate
ion or ozone have been tested with the bare-electrode. Additional studies are
required to expand these bare-electrode aaperometric studies to quantitata
interferences with oxidants otner than those tested, and to expar.c to oc.-.er
natural waters.
Since the accuracy of the potentiometric electrodes is affected. .:
temperature corrections are noc used, it is recommended that temperature :e
either controlled or rseasured simultaneously. Additional indepeneent reasura-
ments of accuracy should be undertaken for the potentiometric eiectroces.
It appears that the potentiosetric electrode can be used for :r.e
determination of total residual oxidant. It is suitable for continuous
measurements and appears to give results that are acceptable when ccstarec. :c
the acperomecric titracor.
General Suzmary and ReconaendaCions for Chlorine.
In comparing air of the methods to the "Ideal Method" we find that none co~e
very close to our ideal standard. Continued development of the various methods
will, however, com* closer and closer to the ideal.
For the present, the aaperometric titration techniques will remain :.-.e
laboratory standard used for the basis of comparisons of accuracy. These
methods, with proper precautions can differentiate between the common inorganic
chorine/chlorine ammonia species, and in general suffer from as few inter-
ferences as any of the methods.
Of the three cotsaon colorimetric procedures. DPD, LCV, and FACTS, the DPD is
by far the most cotaaonly used method. From the available literature it is clear
that the DPD procedure has a nunber of weaknesses. In particular, the colored
product is a free radical which limits the stability of the colored reaction
product. The direct reaction with rsonochloraffline. to form a product :cen::cal
29
-------
to ;.-.e reaction with free chlorine, is also a drawback. This pros lea car. :e
reduced by the addition of thioacetamide. Liquid reagent instability precl'_-es
their use in most cases; care should be taken to determine blanks frequently.
The present LCV method that appears in Standard Methods (13) is outdated ar.s.
has been substantially improved upon by Whittle and Lapteff (Li). This remod.
allows for the differentiation of the common free and combined inorganic
chlorine species. However, because only one comparison study has :aen
conducted, additional collaborative testing is recommended.
The FACTS test procedure appears to be very useful for the determination of
free chlorine in the presence of relatively high concentrations of coaomea in-
organic chlorine. A severe drawback of the FACTS test procedure is the insolu-
bility of the syringaldazine in either 2-propanol or water. This leads to cif-
ficulties in reagenc preparation, and presumably to the color stability proalea
encountered at the higher concentrations of chlorine (greater tnan 6 • 3 az/L
(as C12)). Although a method for the use of the FACTS test for total chlorine
has been reported, it should be tested further.
Electrode methods have been developed employing several different concerts.
The membrane electrodes appear to have potential as specific aethods for h;.-po-
chlorous acid. Common interferences are other nonionized molecules sue.-, as
chlorine dioxide and ozone. Potentiometric electrodes for the determination of
total chlorine are icproving in both detection liait and staoility. These
electrodes appear to have promise in the area of process control. Their
inclusion as methods for routine-use in the laboratory and field is warranted
Both fluorescence and chemiluainescence methods also show prcaisa for me
specific determination of free chlorine at very low concentrations, '-imir. mis
area of spectrofluorometric methods, there is considerable worn vec co re
initiated. Continued development work is warranted and recommended in mis
promising area.
From the review of analytical procedures for the determination of cr.lorire
in aqueous solution, it is readily apparent that only a few of the aetr.ods are
used routinely. Nevertheless, there is certain to be a continued interest in
developing new and better methods of analysis. We would strongly recoeaend mat
new aethods be presented in terms of the "Ideal Method" and that whenever pos-
sible, comparisons with real samples and interlaboratory comparisons be =ade.
Flow injection analytical techniques are becoming very common. Continuea
development should lead to the automation of many colorimetric and fluororsetric
analytical methods for the measurement of free and combined chlorine and irs
various species in water. With the current emphasis on automation, the eethocs
that are to be developed and those already developed can readily exceed present
standards of accuracy and precision. Automation will also lead to ooeracor
independent methods and should lead to improvements in process control and
monitoring.
Chlorine Analytical Methods Comparative Studies.
The reader is cautioned against accepting the results of any or all of me
above tests without some reservations. Where possible we have tried co add co^-
30
-------
aencs, parenthetically, based upon our knowledge of che field. '.- is verv :--
portant in reviewing data froa comparison cescs chac the analyse be aware of the
objectives of the comparison testing. For example, a test ;?.ay be juczea
unacceptable because of an unacceptable lower limit of detection chat is bevor.c
che need for concern for other investigators.
In general when testing several test procedures it is important co idencify
che objective of che testing. Squally important is Che use of che daca. In
reporting che results of the above tests, it should be kept in nind chat r.any
manufacturers of chemicals for analytical methods and Test Kits change their
procedures as a result of the testing. The concerned analyst needs co determine
if the results are still valid. This change is not necessarily applicable to
other studies where the chemistry of an analytical method is examined. Ir.
general, the more the test studies chemistry and not merely che test procedures.
the more applicable the results are for future reference.
Another area of confusion concerns precision and accuracy. An analytical
method may be judged acceptable based on the precision of the results, while che
same aethod may give poor accuracy. These statistical parameters are separaca
and must be tested using different experimental designs. Comparisons with c.-.e
•Ideal Method* would require that both be at acceptable levels.
In general, there is a lack of comprehensive studies co betcer understand
che chemistry associated with the individual test procedures. Investigations of
this nature are necessary on a continuing basis, because of the advances in ana-
lytical instrumentation and our-continued improvements in understanding che ce-
cails of the underlying chemistry.
Chlorine Dioxide Analytical Methods.
The iodometric aethod is a questionable method even for carefully ccr.crclled
researcn laboratory chlorine dioxide standards. In real samples wnere a large
number of potential interferences can exist, che method is destined ro proauce
erroneous results. Newer, sore species specific methods are better cr.oices.
Any cethod which determines concentrations by difference is potentially
inaccurate and subject to large accuaulative errors--both in ceras of accuracy
and precision. The-subtraction of two large numbers to produce a small number
means chat the errors associated with those large numbers are propagated co che
small nuaoer. The result in many cases is chat the error is larger chan cr.e
smaller number, therefore, giving meaningless information. Methods such as
this, which obtain values by differences, should be avoided.
The OPD aethod uses the difference method in the evaluation of concen-
trations. The direct measurement of species by means of a more reliable ar.d
accurate method to determine chlorine dioxide is needed. The same questions
raised about the OPD aethod for chlorine also apply here.
Ultraviolet spectrophotometry, utilizing continuous flow automated methods.
has a great potential for accurate and precise measurements with che added
advantage of ease of operation and high sample throughput. Flow injection
analysis methods (FIA) should be carefully evaluated against existing nechods
for accuracy and precision. The method should be field tested and che pocenc;ai
31
-------
problem of memorane reliability should be evaluated for long terra operators.
Additional bench studies using continuous flow aethods with chemilumir.escar.t
detection must be carried out. The superior selectivity of this aethod needs :o
be utilized. Comparison lab testing and field study should be carried out.
Chlorite/Chlorate Ion Analytical Method*.
The iodometric/amperometric methods are indirect determinations of chlorite
ion and cannoc be recommended. The DPD raethod for chlorite ion can not :e
recommended because ic is unreliable.
The iodometric sequential methods appear to be very workable on samoles
containing greater than 1 ag/L chlorite ion or chlorate ion with good precision
and accuracy resulting. The method requires considerable operator skill ana
experience to obtain good precision and accuracy for samples containing less
than 1 mg/L chlorite ion or chlorate ion. The method should be field testae;
with other methods using both high and low ratios of chlorate ion to chlorite
ion. The method should be used with caution on low level samples of crir.kir.s
water and/or wastewacer, although direct methods requiring less specialised
skills are preferred.
Interlaboratory comparisons should be carried out for the Codified
iodometric method for the direct analysis of chlorite ion and chlorate ion. The
detailed effects of various potential interferences need to be evaluated.
The argentometric titration method is to be recoaeended only for relative'.-/
high concentrations of oxy-chlorine species (10-100 mg/L) but aay be verv ^serul
in establishing inter-laboratory bench mark comparisons at these hiir. concen-
tration ranges. No such comparisons are currently available.
A highly precise, automated FIA method for low level chlorate ion reeds :o
be developed possibly using various masking agents such as glycine, oxalic acid,
aalonic acid, and nitrite ion to initially remove other possible oxv-haiogen
interfering species. The method appears to be very promising in that it can ce
used to directly determine low level chlorate ion concentrations.
Difficulties Vith Ozone Measurements: Heed For Ideal Method.
As a consequence of the nature of ozone, its continuous self-decomposition,
volatility from solution, and the reaction of ozone and its decomposition
products with many organic and inorganic contaminants in water. the deter-
mination of dissolved residual ozone is very difficult. A detailed knowledge of
the mechanism of aqueous ozone decomposition and the potential role of the
various highly reactive intermediates, is imperative in order to accurately
evaluate the analytical methods (15). In this context it should be noted that
most ozone methods are modifications of chlorine residual methods •-•hich
determine total oxidants in the solution. Therefore, ozone decomposition
products such as hydrogen peroxide and the like are also measured.
lodometry can be used as an example of the difficulties encountered .r.
making aqueous ozone measurements (16). Iodide ion is oxidized to :odir.e bv
ozone in an unbuffered potassiura iodide solution. The pH then is adjusted to :
-------
with sulfuric acid and the liberated iodine is titrated with soaiuzi cr.iosuj.fa-a
co a search end point. The ozone/iodine stoichiomecry for this reaction has oeen
found =o range from 0.65 Co 1.5. Factors affecting the stoichioraetry induce
pH, buffer composition, buffer concentration, iodide ion concentration. sampling
techniques, and reaction time. The pH during the initial ozone/iodide icn
reaction and the pH during the iodine deterainacion have been shown to aarkea'.y
alter the ozone/iodine scoichioaecry. The formation of iodate ion and hydroeen
peroxide have been implicated specifically as factors affecting the ozone/iocir.e
stoichio::etry .(17) . Modifications in the iodine determination include char.zes
in end point detection, pH, and back-titration techniques, "one of these
modifications has been demonstrated to be totally sacisfaccory.
The biggest difficulty in interpreting the existing ozone literature is that
no one method has been accepted as the Referee Method. Therefore, coraoanson
between several different methods can create false conclusions about tr.e
accuracy of the methods. The method most often used for comparative purposes ir.
the research laboratory is UV measurement of ozone at 260 no. Even with this
method there" is apparent confusion over the molar absorptivity for aqueous
ozone, with the values ranging from 2900 to 3600 M"lca'1 (16).
All analytical methods reported, particularly chose of early vintage, should
be reevaiuated, considering the recent information about oxidative by-procucts
from ozone decomposition and the ozonation process itself. Sore of these
factors may not have been considered during development of tr.e origi.-.ai
analytical procedures. Certainly, more detailed information and comparisons
should be available. Because of che difficulties of establisnir.g a reliable
Referee Method we propose chat che existing and future setnocs oe cor.parec
against an "Ideal Method". This "Ideal Method" would incorporate all o: :.-.e
characteristics that are desired for an ozone method, caking into account all
other potential interferences, decomposition products, and saapies arijir.at;.-^
from various sources. Finally, automation, while not an absolute necessitv -3.r.
add co che selectivity and ideal nature of a method for ozone determination.
Ozone Measurement: Gas Phas*.
The many uses of ozonation in the treatment of drinking water are control lei
by monitoring a number of parameters. Dissolved residual ozone is only one of
these parameters, and its measurement controls only disinfection conducted after
filtration, but before addition of a residual disinfectant for the distribution
system. However, it is very clear that the cost, efficiency, safety arc
improvements in design of ozone water purification systems is extremely
dependent on che accurate determination of gas phase ozone. Therefore
analytical methods muse be developed that will accurately measure ozone in cr.e
gas phase and residual ozone in the aqueous phase. At this point it LS
unrealistic to believe thac one single method will be acceptable for both sarole
matrices.
lodomecry, UV absorption and chemilurainescence are the three =ost coT.-.on
methods employed for gas phase measurements (16). Each of these has been acpliei
to determine che afflounc of ozone present in generator exit gases, when scrippec
from solution to the gas phase, or the amount of ozone in a contactor e\naust
gas.
33
-------
These techniques of aonitoring concentrations in contactor exr.aus: gases are
quite promising as a method of controlling the production of adequate quantises
of ozone. This provides considerable savings in electrical energy coses for
ozone generation. Direct incer-comparisons of the various gas pnase ieasurer:er.t
techniques are needed in order to evaluate accuracy.
Determination of stripped ozone in the gaseous state was reported in ~r.s
15th Edition of Standard Methods (13) for measuring ozone dissolved in wa;er.
However, in addition to the procedure being subject to the sane liaitations of
UV absorption and chemiluainescence procedures in aqueous solution, t.-.e effects
of the gas stripping process itself aust be taken into consideration.
Although the iodometric stripping/aqueous absorption method has been
approved in Standard Methods (13), we question the accuracy of the metnod. .-.11
evidence would suggest that the method is problematic. Even though the
impurities are substantially left behind by the stripping, the actual procedure
and the continual decomposition of ozone does introduce inaccuracies ir.to :his
method. This method can be used as a relative measure of ozone for control
purposes.
This basic stripping approach followed by absorption in acueous solution
(and colorimetric measurement) may deserve to be studied further. However, rr.e
biggest potential proolem appears to be that at high concentrations of stone :.-.e
coloriaetric compounds may react by a mechanism different from rr.at used for
residual ozone measurements. Research should be concentrated on t.-.e reagents
that have already exhibited ozone selectivity.
lodometry (Aqueous Phase).
If the performance of ozone in a specific creataent application is -.0: De-
pendent only on the ozone, but is instead a collective function of its reactive
decomposition products as well, then iodometry can give a representative ana
reproducible reading of the tocal oxidancs. For example, aost European srinkir.g
water treatment plants employing ozonation as the priaary disinfectant, have
relied on iodometric measurements as the basis for insuring adequate
disinfection, attaining a residual "ozone" level of 0.4 mg/L in :r.e first
contact chamber and maintaing this level for at least four minutes).
However, it is now abundantly clear that the 0.4 mg/L value is a measure of
the amount of total oxidants present, and not necessarily ozone alone.
Therefore, either the absolute level of ozone required to attain the expected
degree of disinfection is lower than 0.4 mg/L over the required period of tiae.
or some of the decomposition/oxidation products formed upon ozonation also have
disinfecting properties, or both. Clearly, detailed experiments r.eed to be
carried out to demonstrate the efficacy of disinfection by the decomposition
products of ozone. Similar efficacy data for ozone decomposition products could
be developed for other uses of ozone (e.g., chemical oxidation) when measurement
of residual ozone levels must be made to control the process. Such data would
help to justify the continued use of iodometry to measure "total oxidants',
rather than only ozone.
Historically, iodometry has been used as the reference method for de:er-
mining ozone, and against which other analytical procedures nave been
-------
"standardised". It is now quice clear thac because of its lack of seieccivi:v
che use of iodomecry should be liaiced to that of only a control procedure. Ir.
cerms of ozonacion processes, measurement-for control purposes-of che produccicr.
race of ozone generators and bacterial disinfection/viral inactivacion say be
based upon iodomecry, provided the user recognizes che many lioitacions of the
method. The reevaluacion of chis method must be carried ouc with the specific
goal being to define when che method is reliable and the situations where it is
not accurate.
Many authors have tactfully pointed out the many disadvantages of iodometrv,
leaving it to the reader to decide whether or not to use the procedure. In a
detailed conparison of eight analytical methods for the determination o:
residual ozone it was concluded (16):
"No iodometric method is recommended for the determination
of ozone in aqueous solution because of the unreliability
of the method and because of the difficulty of the com-
parison of results obtained wtch minor modifications in
the iodometric method itself."
Arsenic(III) Direct Oxidation.
In the direct oxidation of arsenic(III), ozone reacts vich inorganic
arsenic(III) at pH 4-7, the pH is adjusted to 6.5-7 and the excess arsenic(III)
species is back-titrated with standard iodine to a starch end point. Values for
residual ozone determined by the .arsenic direct oxidation method and by ;.-.e
indigo method agreed within 61 of the UV values. The primary advantages of the
arsenic direct oxidation procedure are minimal interferences, good precision :n
che hands of experienced operators, and apparently good overall accuracy. This
procedure continues to be recommended along with the indigo method. Additional
comparisons of this method should be made with the indigo method under various
conditions.
SyringaldLazine, FACTS.
The FACTS procedure, which was developed for the selective determination of
free available chlorine (hypochlorous acid + hypochlorite ion) in the presence
of combined chlorine (chloraaines), has been adapted for the determination of
residual ozone (19). In this procedure, an aqueous solution of ozone is added
to a solution of potassium iodide, and the liberated iodine is added to a 2-
propanol solution of syringaldazine ac pH 6.6. The resulting color is measured
spectrophotooetrically ac 530 na.
The FACTS procedure has the major advantage of providing a spectrophoto-
setrie procedure for the determination of ozone. However, the major limitations
of the FACTS method are still those of the iodometric procedure. Due to the
observed changes in slope and intercept which are problems caused by the
interferences, self-decomposition of ozone, and stoichiometry, this method could
be reviewed in order to fully evaluate its potential usefulness. However.
considering :he other colorimetric methods that are available further
development of the FACTS method does not seem to give any promise of tr.e
improved selectivity that is needed.
35
-------
N,N-Di«chyl-p-phenyienediaaine. DPD.
The DPD procedure is based on "he ozone oxidation of iodide ion present ir.
excess phosphace buffer ac pH 6.4 co produce iodine, which then oxidises che DPD
cacion co a pink burster cation wnich is measured speccrophotooecrically, or
titrated. . The interferences include all oxidants capable of oxidising iodide
ion co iodine, including ozone decomposition products, halogens, and manganese
oxides (20).
One advantage of the DPD method is thac determinations can be aade by
ferrous ammonium sulfate (FAS) titriaetry, spectrophocoraetrically or by a color
comparator. Ozone concentrations of less than or equal ro 2 ag/L can be
determined colorimetrically. Clearly, the procedure requires the difference of
differences and is limited by the saa* factors which limit iodometry, specific-
ally the presence of materials which can oxidizt iodide ion to iodine.
Although evaluation of this procedure versus the standard ultraviolet and
indigo procedures would seem to be necessary to make a more educated decision
about the continued use or abandonment of this method, the recommencacion is
chat other colorimetric methods are considerably aore reliable :han DPD.
Therefore development or testing is neither recommenced nor considered necessary
at this time.
Indigo Trisulfonate.
The indigo method is subject to fewer interferences than most coloriaecric
methods and fewer interferences than all iodometric procedures (21-23). AC r.-i
2, chlorite, chlorate, and perchlorace ions, and hydrogen peroxide do -cc
decolorize Indigo Reagent when observed within a few hours ar.s --hen :.-.e
concentrations of the interferents are within a factor of 10 of thac of che
ozone co be determined.
Ozone decomposition products and the products of ozonolysis of organic
solutes do not appear to interfere. However, chlorine, bromine, and iodir.e do
cause some interference, as do the oxidized forms of manganese. The acdicicr. of
aaIonic acid to the sample* will mask the interference of chlorine.
For the Indigo Trisulfonate Method, it should be noted that when che
ultraviolet absorption method is used to standardize the indigo method (or ary
method) for ozone, the choice of molar absorptivity is very critical. It is
recommended that the equations of Hoigne continue to be used since chey are
based on a molar absorptivity of 2950 M"lcm~l. If and when a different value
for molar absorptivity is reported and confirmed, the (calibration) equations
would have to be appropriately changed. In this manner, all current
measurements using the indigo method would continue to be comparable.
The advantages of the indigo procedure is that it is based on a measure of
discoloration which is rapid and stoichiometric. This analytical procedure is
recommended for use over any other procedure for che determination of residual
ozone. Its primary attributes are its sensitivity, selectivity, accuracy,
precision, speed, and simplicity of operation.
36
-------
The gas diffusion flow injeccion analysis (CD-FIA) procedure eliminates t.-.e
interference of oxidized foras of manganese, and aarkedly reduces the interfer-
ence of chlorine (24). Other than interference of chlorine which can be recuces
to zero by addition of malonic acid, there are no known interferences to r.-.e
determination of ozone by this*GD-FlA procedure using the indigo method.
The primary advantages of the CD-FIA procedure are its accuracv,
selectivity, lack of interferences, reproducibility, and rapidity. Thus, ;-e
method is well suited for laboratory research studies and for use as an
automated analytical procedure.
More studies should be conducted with specific gas-perseable membranes.
particularly with respect to repeated and/or continuous exposure to ozone solu-
tions. The use of FIA equipment in a process control environment also nust be
evaluated. The GD-FIA indigo procedure might well be adopted as the analytical
method of choice.
o-Tolidine
The o-tolidine method (addition of 1-2 drops of o-tolidine solution to
ozone-containing water to develop the yellow color) is very sirole. and easily
adapted to field color comparators, suitable for unskilled analysts. However.
this advantage cannot compensate for the lack of quantitation of the zechod, r.or
for the carcinogenicity of the reagent (o>tolidine). The recocaencation is to
abandon this method.
Carmine Indigo.
The carmine indigo procedure has been used in Canadian water wortcs plar.ts
for the past 15 years. The ozone containing water is titrated with a solution
of carmine indigo until a faint blue color persists indicating that all of the
ozone has been destroyed. Specific interferences are unknown, but ar.v oxidar.t
capable of decolorizing the carmine indigo dye most likely will interfere.
Effects of interferents should be determined, as should precision, accuracv,
and effects of reagenc scorage and pH. The method should be studied in direct
comparison with other methods, such as the indigo and UV absorption methods.
Automation of this method could lead to improved selectivity for ozone.
Aoperoaetry.
With bare electrode amperemeters, either the solution or the electrode is
rotated to establish a diffusion layer, and the electrical current measured is
directly proportional to Che concentration of dissolved oxidant (25). Commer-
cial amperomecric analyzers give satisfactory results provided there is no
oxidant other than ozone present in the sample. In many situations they provide
adequate monitoring of total oxidant. The bare electrode system has good
sensitivity, and is applicable as a continuous nonselective monitor for ozone.
'-hen other oxidants such as chlorine, chlorine dioxide, bromine, and iodine are
present, the technique has difficulties. The exact nature and magnitude of
these interferences requires additional research.
37
-------
Due co "he accumulation of surface impurities at che electrode surfaces, all
bare amperometric eleccrode systems are subject to Loss of sensitivity vith use.
With uncovered electrode surfaces, fouling has been observed to be a significant:
problem as was the case in earlier tests with oxygen electrodes. Additionally.
the response is influenced by numerous surface-active agents and also halogens
and oxygen.
An improvement in the development of araperometric methods for ozone analysis
has been the application of gas-permeable membranes for increasing selectivity
and preventing electrode fouling (26-27). These Teflon membrane electrodes
exhibit less than 2% interference (in terms of current response) from bromine,
hypobroraous acid, chlorine dioxide, hydrogen peroxide, nitrogen trichloride, ana
hypochlorous acid (26-27).
This type of araperomecric membrane sensor needs to be developed further
based on the exhibited selectivicies. The most disturbing attribute is tr.e
temperature dependence. If different membranes could maintain selectivity vnila
minimizing the temperature effect, this type of sensor could become higr.ly
recommended.
The application of positive voltage potentials and the use of polymeric =en-
branes that are selectively permeable to gases has enhanced the opportunity far
selective measurement of ozone. This is a very significant improvement over
bare aaperooetric electrodes as well as most older colorimetric/spectrophoto-
metric and titriaetric methods. With an applied voltage of +0.6 V (vs SCE) at
the cathode, only the most -powerful oxidizing agents can overcome the
"resistance" of this anodic voltage and cause electron flow cathodicai.lv thrcuer.
the electrochemical circuit. This general approach should continue to se -sec
in future electrochemical developments.
Other Electrochemical Methods.
In the differential pulse polarography procedure (DPP), a predeterrv.r.ad
amount of phenylarsine oxide (PAO) is added in excess to an ozone solution -a
reduce the levels of dissolved ozone present. Excess PAO Chen is measured
quantitatively by pulse polarography. The DPP method may under scr.e
circumstances be useful in the research laboratory. The prospeccs of its use in
the plant or field-are not as promising since a higher degree of operator stciil
is required.
Potentiometry involves the cathodic reduction of dissolved ozone. The
diffusion-limiting current measured is proportional to che concentration of
ozone in the water. Further evaluation of potentiometric systems r.ay be in
order. However, the fundamental problems of electrode fouling must be
addressed. Perhaps a combination of membranes and potentiometric detection
would produce a promising system for ozone determinations. The system appears
to have modest potential for development.
Ultraviolet Measurements.
Ultraviolet absorption measurements also can be used for residual aqueous
ozone at 258-260 nra. There is uncertainty with respect to the -olar
absorptivity for aqueous ozone. In che literature, values ranging frora 2?CO :D
38
-------
3600 M"'ca"1 are reported. This ur.cercai.ncy in che molar aosorpriviry ;s
critical co che future use and calibration uses of che UV aecnods. Clear 1-
further work co verify this value is strongly recommended.
If che solar absorptivity for ozone is known unarabigiously, 'JV absorption
in principle an absolute method for che determination of ozone, which is r.<
dependent upon calibration or standardisation against other analytical mec.-.cc:
Therefore, it can be used for calibration of other analytical methods for ocor.i
It is specific to che determination of ozone, and is applicable to =easure?.ei
in gaseous and aqueous phases.
Physical Methods.
The calorimetric aethod is based on the enthalpy of the cacalvz
decomposition of ozone (AH - 144.41 KJ/mole). The calorimetric deterair.aci
of ozone is calibration-independent. The technique is specific to c
determination of molecular ozone, but is applicable to measurement only in c
gas phase. However, the higher che concencracion of ozone in che gas phase, z
more accurate the method appears to be, since a greater temperature cifferer.
is observed. Potential interferents have not been reported.
The method has been shown to agree with iodometric and UV absorption pr
cedures. particularly for the measurement of ozone in the gases exiting 0=3
generacors. Therefore, che procedure can be used co monicor applied ozo
dosages. Additional decailed incerlaboratory comparisons need co oe cam
out.
The isochernal differencial pressure procedure' is based on che generation
an increased nuaber of gas molecules during the UV destruction of osor.e
constant temperature. '-"hen this reaccion is carried out isocherrallv :r.
closed vessel, che increase in pressure of the concained gas is proport.3r.al
che ozone concencracion. In principle, chis procedure achieves a rocal
physical ozone measurement without requiring calibracion using a chesuc,
method. Various automated instrumental checks such as che scored sol,
absorptivity, che age of che UV light source, che zero poinc readir.;
aeasuremenc of che flow of che cesc gas and che flushing gas, and che reading c
che diagnostic display are possible.
No specific comparisons are reported. However, in principle it appears ch,
chis physical method is the best candidate for calibrating che gas phase ozor
instruments currently being used for ozonation control. As long as pure oxygi
is used for ozone generation this method would be free of interferences a:
would be subject only to strict temperature control of the measurement: eel?
Furcher study of chis system would be necessary before it could be recomnier.st
for further consideration.
General Summary and Recommendations for Ozone.
In comparing all che methods co che "Ideal Method" we find chac rone co-
close co our ideal scandard. Concinued development of che various selecciv
mechods will, however, come closer and closer co che ideal.
39
-------
In terns of gas phase measurements, none of the existing r.ethods can be
recommended for accurace determinations of ozone. If a relative value of tr.e
ozone concentration is needed for control purposes, aost of the aethocs reported
could be applicable.
The accurace determination of ozone in the aqueous phase is corr.olicatea ;v
the decomposition of ozone, its reactivity to the other species present, ar.d tr.e
by-produces of the ozonation reactions. Most current methods were developed
without a clear knowledge of the associated ozone chemistry. Therefor: aost of
the methods are unacceptable or cannoc be recoesended. In particular, -o
io dome trie based chemistry is acceptable for the determination of aqueous ozc-r.e.
Indigo trisulfonace and arsenic(III) direct oxidation are acceptable -etr.ccs.
Aaperometery continues co improve •- especially as an automated control zet.-.oa.
The stripping techniques have some merit in terms of improved o::r.e
selectivity. However, automated chemical systems such as flow ir.jectisn
analysis offer considerably more promise. The current GD-rlA ir.sigo prscec-re
is superior for residual ozone measurements due to its selectivity for ozone.
The most important aspect of any potential new or improved ozone sr.alvtical
method will be speed of analysis and selectivity of the aetect-.on svstea for
only ozone. As a point of comparison, we strongly recommend that ail future ar.a
existing methods be compared against the 'Ideal Method".
LITERATURE
1. Symons, J.M.; et al "Ozone, Chlorine Dioxide and Chloraair.es is
Alternatives to Chlorine for Disinfection of Drinking Vacer" .n
Water Chlorination: Environmental Irrpaet and Health Effects , v ?". .
2., Jolley, R.L.; Gorchev, H. and Hamilton, D.H., Jr., Editors, vAr.n
Arbor,. MI: Ann Arbor Science Publishers, Inc., 1979) pp. 535-560
and Complete Report entitled "State of the Art ..." (Cincinnati.
OH: U.S. EPA, November, 1977), 84 pp.
2. Proceedings of Seminar on "The Design and Operation of Drinking '."ater
Facilities Using Ozone or Chlorine Dioxide", Rice. R.C., Editor,
(Dedhaa, MA: New England Water Works Assoc., 1979).
3. Miller. C.tf, ; Rice, R.C. ; Robson, C.M.; Scullin, R.L. ; Kuhn. '-. and
Uolf, H., "An Assessment of Ozone and Chlorine Dioxide
Technologies for Treatment of Municipal Water Supplies". '.' S.
Environmental Protection Agency, EPA Project Report, EPA-600/2-
78/018. 1978, 571 pp.
4. Miltner, R.J. "Measurement of Chlorine Dioxide and Related Products',
in Proceedings of The 'Jarer Quality Teehr.olee;gal '".'-fe!••-?•;.
(Denver, CO: American Water Works ASSOC.. 1976). pp. 1-il.
-------
5. Gordon. C. "Improved Methods of Analysis for Chlorate, Chlorite,
and Hypochlorite Ions at the Sub-rag/L Level", U.S. Environmental
Protection Agency, EPA Technical Report. EPA-600/4-85/079, October.
1985, 35 p. and Presented at AWWA WQTC, in ?roe. AVTJA Vater Quali-v
Technology Conference. December. Nashville, IN, 1982. pp. 175-139.
6. Aieta, E.M.; Roberts. P.V. "Chlorine Dioxide Chemistry: Generation
and Residual Analysis" in Chemistry in Vater Reuse. Vo^. ^,
Cooper, W.J., Editor (Ann Arbor, MI: Ann Arsor Science Publishers,
Inc., 1981). pp. 429-452.
7. Hoigne. J.; Bader, H. "Sestimmung von Ozon und Chlordioxid in Vasser
mit der Indigo-Methode" ("Determination of Ozone and Chlorine
Dioxide in Water With the Indigo Method"), Voo Uasser, 1980. H,
261-280.
8. Gilbert, E. ; Hoigne, J. "Messung von Ozon in Wasserwerken; Vergleich
der DPD- und Indigo-Methode" ("Ozone Measurement in Water Treatser.t
Plants: Comparison of the DPD and Indigo Methods"), GFV-
Wasser/Abwasser, 1983. 124. 527-531.
9. Schalekaap, M. "European Alternatives and Experience" in Proceedings
of the National (Canadian) Conference on Critical Issues in
Drinking Vater Qualify. (Ottawa, Ontario, Canada: federation of
Associations on Canadian Environment, 1984), pp. 140-159.
10. Ikeda. V.; Tang. T-F.; Gordon, G. "lodometric Method of Determination
of Trace Chlorate Ion", Anal. Chem.. 1984, i£. 71-73.
11. Ernnenegger F. ; Gordon, G. "The Rapid Interaction between Sodiua
Chlorite and Dissolved Chlorine", Inorg. Chem., 1967, 6, 633-635.
12. Aieta. E.M.; Berg, J.D. "A Review of Chlorine Dioxide in Drinking
Water Treatment", J. Am. Water Works Assoc.. 1986, 7$, 62-72.
13. Standard Methods for The Examination of Water and tfastevacer. 16th
Editioc. Creenberg, A.E.; Trussell. R.R.; Clesceri, L.S.; Franson,
M.A.H., Editors (Washington, D.C.: American Public Health ASSOC.,
1985), 1268 pp. and 15th Edition. Greenberg, A.E.; Connors. J.J.;
Jenkins, D.; Franson. M.A.H.. Editors (Washington, DC: American
Public Health Assoc., 1980), 1134 pp.
14. Whittle, G.P.: Lapteff. A.. Jr. "New Analytical Techniques for the
Study of Water Disinfection" in ghe,mistrv of Water Supply.
Treatment, and Distribution. Rubin, A.J.. Editor, (Ann Arbor, MI:
Ann Arbor Sci. Pub., Inc.. 1974), pp. 63-88.
15. Tomiyasu, H.; Fukutorai, H.; Cordon, G. "Kinetics and Mechanism of
Ozone Decomposition in Basic Aqueous Solution". Inorg. Chem., 1985,
24, 2962-2966.
41
-------
16. Grunwell, J.; Senga, J.; Cohen, H., Cordon, G. "A Derailed Comparison
of Analytical Methods for Residual Ozone Measurement", Ozone Sci.
Eng., 1983. I, 203-223.
17. Flanm. D.L.; Anderson. S.A. "lodace Formation and Decomposition in
lodometric Analysis of Ozone", Environ. Sex. Technol., 1975, 3..
• 660-663.
18. Rehme. K.A.; Purak, J.C.; Beard. M.E.; Smith. C.F.; ?aur, R.J.
"Evaluation of Ozone Calibration Procedures", U.S. E?.viror-T.er.:al
Protection Agency, EPA Project Summary, EPA-600/S4-80-C30.
February, 1980, 277 pp.
19. Liebennann, J.. Jr.; Roscher, tf.M.; Meier, E.P.; Cooper. W.J. Develop-
ment of the FACTS Procedure for Combined Forms of Chlorine ar.d
Ozone in Aqueous Solutions". Environ. Sci. Technol., 1980. 1^,
1395-1400.
20. Palin. A.T.; Derreumaux, A. "Determination de 1'Ozone Residuel dar.s
1'eau" ("Determination of Ozone Residual in Water"), L'Zau e:
1'lndustrie. 1977, 12, 57-60.
21. Sader, H.; Hoigne, J. "Colorimetric Method for the Measurement
of Aqueous Ozone Based on the Decolorization of Indigo
Derivatives", in Ozenization y.ar.ual for Water and '-'asrevarer
Treatment. Masschelein. W.J.. Editor, (Sew York. N"Y: Jonn
Wiley & Sons. 1982). pp. 169-172.
22. Sader. H.; Hoigne. J. "Determination of Ozone in Water by the
Indigo Method", Water Research 1981. 15_. 449-456.
23. Bader, H.; Hoigne, J. "Determination of Ozone in Water by rhe Indigo
Method; A Submitted Standard Method", Ozone: Science and Er.g ,
1982. 4, 169-176.
24. Scraka, M.R.; Cordon, G.; Pacey, G.E. "Residual Aqueous Ozone Deter-
mination by Gas Diffusion Flow Injection Analysis", Anal. Chen.,
1985, H, 1799-1803.
25. Masschelein, W.J. "Continuous Amperometric Residual Ozone Analysis
in the Tailfer (Brussels, Belgium) Plant", in Ozonyzation "arual
for Water and Wastewater Treatment. Masschelein. W.J., Editor, (.v.'ew
York, NY: John Uiley & Sons, 1982). pp. 187-188.
26. Stanley, J.H.; Johnson, J.D. "Amperometric Membrane Electrode for
Measurement of Ozone in Water". Anal. Chem. . 1979. 5_1, 2144-2147.
27. Stanley, J.W.; Johnson, J.D. "Analysis of Ozone in Aqueous Solution",
in Handbook of Ozone Technology and Applications. Vol. 1, Rice,
R.C. and Netzer, A., Editors (Ann Arbor, MI: Ann Arbor Sci. Pub.,
Inc., 1982), pp. 255-276.
42
-------
A GUIDE FOR EFFICIENT USE OF THIS REPORT (AND A BRIEF GLOSSARY OF TERMS)
This Report contains a very detailed review of all disinfectant residual
aeasuremenc mtchods. The Executive Summary is intended to give readers a brie:
overview of the advantages and disadvantages of each method. To that end, Table
I (Characteristics and Comparisons of Analytical Methods) has been included :o
summarize each of our findings and to recommend possible directions for future
research. In addition, Table II (Equivalent Weights for Calculating
Concentracions on the Basis of Mass) describes the equivalent weights of each of
che disinfection species in terms of the actual reactions involved in :.-.e
disinfection process.
Each chapter contains individual recommendation* following the discussion of
the method. A summary of all of the recommendations is also given at rhe end of
each chapter. Additional help is given by means of an alphabetical Ir.cex
containing more than 2500 individual terms. Specific cross referencing for ail
recoanendations can be found in the Index either under the "recommendation", or,
in terms of the subject of the numbered recommendation itself.
The term Referee Method is used to describe appropriate comparisons with
existing methods and Standard Methods refers to a specifically recocsenced
method. The Index should be an additional aid to finding rhe details of
specific methods.
In this context, it should be noted that the individual literature ciratior.s
are specific to each individual chapter -- and are either numbered individuallv
within chapters 2 and 3, or alphabetically sequenced within chapters -* ar.c :
Chapter 4 (Indexed Reference Citations) has been included in this report :-
order to assist readers in locating particular papers of interest. The -5
categories for chlorine, chloramines, and the oxy-chlorine species, along vim
the additional 60 categories for ozone, should make the task of finding in-
dividual papers of interest considerably less cumbersome. Papers which describe
several methods have been included in each of the appropriate categories. All
together, the 1.400 references cited in Chapters 1*3 number more than 2.CCO
individual citation* when distributed in the indexed form of Chapter 4.
Chapter S is an alphabetical listing of the individual references citations.
finally, a detailed Index has been included in order to assist readers in
locating subjects of specific interest. We hope the readers will find these
additional chapters as useful as have we in preparing this reporc.
A brief Glossary follows on the next page in order to assist readers in :i-.e
various specialized terms and abbreviations used in this reporc. For additional
terms, the reader is referred to the Index.
-------
GLOSSARY
Accuracy -- the ability co determine the correct concentration
3AKI •• boric acid buffered potassium iodide method for ozone
Breakpoint •• the inorganic reaction of chlorine with asaonia r.irrijen
CDFV -• chlorine demand free water
Combined Chlorine •- inorganic and organic chloraaines
Detection Limit •- a signal that is 3 times the noise level of the system
DOC •- dissolved organic carbon
DPD -- (N.N-diethyl-p-phenylenediamine)
FACTS •- free available chlorine tesc with syrir.galdazine
FIA •- flow injection analysis, an automated analysis procedure
Free Chlorine -- the species, C17 +• HOC1 •«• OC1"
KI •• potassium iodide method for ozone
LCV -- leuco crystal violet
aL -- ailliliter(s). standard unit of volume
Molar Absorptivity (O reported in units of M~'cm"1
N5KI -- neutral buffered potassiua iodide method for ozone
Precision -- how wall the method reproducibly measures the sane
concentration
Reactive Intelnaediate -• species such as 02~, HO.," , H02, OH, 03~, etc.
Referee Method •- the method aqainsc which a working method is compared
Sensitivity -- th« change in signal per unit concentration [i.e. Aaps/taol]
Standard Methods -• the book, Standard Methods for the Exar!i-3tion._
tfater_.and Uastewater published by APHA. AW"»A, and U'PCF
THM's -- trih*lom«thanes
Total Chlorine -- Che combination of Free Chlorine and Combined Chlorine
TOC •- total organic carbon
TOX •- total organic halogen
-------
-------
APPENDIX E
INACTIVATIONS ACHIEVED
BY VARIOUS DISINFECTANTS
-------
z
M
u
— 3 iT 9
a z * Q
a — » ~*
•J * £ a
<2feS
2 > *<
3 ^. O
O
= 1
o. — w>
r
i/^
• CM
vO «
M
HLORI
CONCENTRATION
<-R/L)
S
§
OO
« — (nvo«o —
S — — — — (N
SfM^r~
— — —
» J 5 S 5 8
P P P
II
V
(StMCSCN
CM
9 I o
4i3
o
S"
2|5
CRINE
CONCENTRAT
(«*/D
SSh ^^
MK
— «CMCMCM«MCMfMNtM^{MCMO ^* ^» to ^ ^ P p
CMOt«M(S«Me>if«){M
-------
a
z
z *
2 S
- * M ^
SSS*
UJ CJ
Zj <
> °
i- <
o 3
!o
m a <^
r- .2
- I
•s.^
I-
41
S
.
= 1
^. - *>
- 9
> N
— "^
1"
o
HLORINE
CONCENTRAT
(,nR/L)
M «M (S
— m >e •
o. — min«o
§
oeo — r»>nf^«eOfs''^'nf~9»~
r-««o«o«e«e«8*»<>o\a*o
CM R R~-r5 S
OOO —
^«» > '
n
*m "^ *^
'1-
da
1
sia
l
CHLORINE
CONCENTRATIO
(«"J!/U
O*Ownmo«o>O — n » in
tStS«NtSJMtS(NP4
— »>o«!«Ncs«
-------
UJ
Z
Z *
2 S
< 0
— a
SI
z >
U (j
_; <
> S
u.
O
.
1
— «•>
I"
il
I
= 1
a. — <«
5"
PM
E
RATION
HLORI
CONCEN
(niR/L)
S — «vn>«r»o\ —
O> —
?3 S N r>i 01 «M
RO — — M «
ft m ^ f*l ^i
m >n wi >e >o t^irt
r%rtrtft*^r^r»*^r
w e* «s es ts ts ts
*.'*.*'.'" p***t"
Si
I
z
-------
aj
z
z i
;;
>|
^ UB
Z >
2 <
< 5
^ *v
o 3
o
pH=7.5
Log IfUKlivatiun
1.5 2.0
i°
s <•*
is. *>
>
1 1
o
• f<
iris
p
Log In
0 1.5
CHLORINE
CONCENTRATION
(mR/L)
*«rt/iiAvt«^r<»«««oe>
OiWRtSNtMCMPXCSSMpSfMN
O O O — — — — — (N(MP*{SeS
O O O —
II
V
— •- «srxes«M
5 -
-
I
U
.0
pH
Log laa
I.S
E
I
et
i^
ae a
3^
S8
in«'«««*r~r»r»r-i--r-
es w (S
{S (N (N CM
6d6
II
V
•»CM«r^f*CM(S«>0**
,2
t
ii
3
-------
LU
Z
ll
M
M
U ae
< u.
z >
~ a
O 52
"• K
s§
< a
> at
i_ <
e. — ir>
I"
2 IS
iJ.
i
25
=
iv
2.0
pll<=6
Log liwctivation
.5
CRINE
CONCENTRATION
(raR/L)
~£i'*9n'$P~ii9&O'~mt*
mmfnmm«>t<)(mnKK9«<*
m»»i«»n«SR?5»o>«r-r-r^«««o*o>.
»«»»»» O
Sp*mfn«»in«r»f"-«o«
««nfnr>i*ienM«^«^««t
ni»»m»»>«o^r-r«r»p-r-(^r-r-«»«««««»
O O O — •— — — WCMfSN
II
V
9.0
vatiooi
2.0 2.5
Log
5
vrti
2.0
pH
Log liM
I.S
vatKMM
2.0 2.S 3.0
Log
I
.1
UJ ^
j 8 1
win^r*te
fnr^n«»v5r<»^'»*fr«rv^
csrtm^««minin««
-------
TABLE E 6
CT VALUES 1 OR INACTIVATION
Or GIARDIA CYSTS BY FREE CHLORINE
AT 25 C (I)
9
i o
>/•> 1 is
5 2
» ~ VI
*fi M
i/i
d
e
M
• X O
9 a N
i- .5
U,
r
0
o
p
1/1
•» is
*•> 1 °
III
0
W1
O
p
l\"
'i-
P
vr
O
CHLORINE
CONCENTRATION
(•"R'M
";;^^»r.!!"
„„„„„„„,,.,«„„
•a»aBa«saB»»»«
O O O •.-.-.— is is is is
II
V
p
«^
m
(S
H
p | 9
p
d
pit =8 5
Log liMctivalioni
05 1.0 15 20 25 30
p
• (S
iK
sl
6
ErilLORINE
XJNCENTRATION
mB/L)
9 R R S A ft S S 3 S S S S 5
OOO —••••-•«• tscsr-ip*
II
V
N.rtc*
(1) CT = CT for 3 log iiwciivation
?;
-------
TABLE E-7
CT VALUES FOR
INACTIVATIQN OF VIRUSES BY FREE CHLQRINE(1)
Loa Inactivation
Temperature (C)
0.5
5
10
15
20
25
2.0
pH
^
6
4
3
2
1
1
14
45
30
22
15
11
7
3.0
oH
fiz*
9
6
4
3
2
1
14
66
44
33
22
16
11
4.0
DH
6=2
12
8
6
4
3
2
14
90
60
45
30
22
15
Notes:
1. Basis for values given in Appendix F.
-------
TABLE E-8
CT VALUES FOR
INACTIVATION OF GIARDIA CYSTS
BY CHLORINE DIOXIDEU)
Temoerature (C^
Inactivation
0.5-log
1-log
1.5-log
2-1og
2.5-log
3-log
-------
Notes:
TABLE E-9
CT VALUES FOR
INACTIVATION OF VIRUSES
BY CHLORINE DIOXIDE oH 6-9("
Temoerature (C)
Removal
2-log
3-log
4-log
£•1 5_ ifl
8.4 5.6 4.2
25.6 17.1 12.8
50.1 33.4 25.1
15
2.8
8.6
16.7
22
2.1
6.4
12.5
25
1.4
4.3
8.4
1. Basis for values given in Appendix F.
-------
TABLE E-10
CT VALUES FOR
INACTIVATION OF GIARDIA CYSTS
BY QZQNE(1)
Temoerature 1C]
Inactivation <»1
0.5-log
1-log
1.5-log
2-log
2.5-log
3-log
0.48
0.97
1.5
1.9
2.4
2.9
-5.
0.32
0.63
0.95
1.3
1.6
1.9
UL
0.23
0.48
0.72
0.95
1.2
1.43
Ifi.
0.16
0.32
0.48
0.63
0.79
0.95
2Q_
0.12
0.24
0.36
0.48
0.60
0.72
25_
0.08
0.16
0.24
0.32
0.40
0.48
Note;
1. Basis for values given in Appendix F.
-------
TABLE E-ll
CT VALUES FOR
INACTIVATION OF VIRUSES BY OZONE
(i)
Temperature (C)
Inactivation <=1 5 }Q 15
2-log 0.9 0.6 0.5 0.3
3-log 1.4 0.9 0.8 0.5
4-1og 1.8 1.2 1.0 0.6
Note;
1. Basis for values given in Appendix F.
2P_ £5_
0.25 0.15
0.4 0.25
0.5 0.3
-------
TABLE E-12
CT VALUES FOR
INACTIVATION OF GIARDIA CYSTS
BY CHLORAMINE DH 6-9(1)
Temoerature (C]
Inactivation <*1 5 10
0.5-log
1-log
1.5-log
2-log
2.5-log
3-1og
Note;
1.
635
1,270
1,900
2,535
3,170
3,800
Basis for
365
735
1,100
1,470
1,830
2,200
values giver
310
615
930
1,230
1,540
1,850
i in Appendix F.
15
250
500
750
1,000
1,250
1,500
20
185
370
550
735
915
1,100
25
125
250
375
500
625
750
-------
TABLE E-13
CT VALUES FOR
INACTIVATION OF VIRUSES BY CHLORAMINE
O)
Temoerature (C)
Inactivation
-------
TABLE E-14
CT VALUES FOR
INACTIVATIQN OF VIRUSES BY UV(1)
Log Inactivation
2.0 3.0
21 36
Note:
1. Basis for values given in Appendix F,
-------
APPENDIX F
BASIS FOR CT VALUES
-------
APPENDIX F
BASIS OF CT VALUES
F.I Inactivation of Giardia Cvsts
F.I.I Free Chlorine
The CT values for free chlorine in Tables E-l through E-6 are based
on a statistical analysis (Clark et al., 1988; attached to this appendix),
which considered both animal infectivity studies (Hibler et al., 1987) and
excystation studies (Jarroll et al., 1981; Rice et al., 1982; Rubin et
al., 1988). A multiplicative model was selected to best represent the
chemical reactions during the inactivation process. This model was
applied to each of the data sets, listed above, and in various combina-
tions. The animal infectivity data were included in all combinations
studied". The animal infectivity data was considered essential for
inclusion in all the analysis of combined data sets because it included
many more data points than the other data sets, all of which represented
inactivation levels at 99.99 percent. Because of limitations with the
excystation methodology, only data for achieving less than 99.9 percent
inactivation was available from such studies.
Statistical analysis supported the choice of combining the Hibler et
al. and the Jarroll et al. data (and excluding the Rice et al. (1981) and
Rubin et al. (1987) data), to form the best fit model for predicting CT
values for different levels of inactivation. As a conservative regulatory
strategy, Clark and Regli (1990) (attached at the end of thsi appendix),
recommended that CT values for different levels of inactivation be
determined by applying first order kinetics to the 99 percent upper
confidence interval of the CT99 99 values predicted by the model.
The model was applied using the above strategy as a safety factor,
to determine the CT values ranging from 0.5-log to 3-log inactivation at
0.5 and 5 C. CT values for temperatures above 5 C were estimated assuming
a twofold decrease for every 10 C. CT values for temperatures at 0.5 C
were estimated assuming a 1.5 times increase to CT values at 5 C. This
general principle is supported by Hoff (1986). It is important to note
that the CT values for free chlorine are sensitive to the residual
F-l
-------
concentration, C. For example, at a pH of 7 and a temperature of 10 C, a
3-log Giardia cyst inactivation results from a CT of 107 mg/L-min with a
free residual of 0.6 g/L and a CT of 124 mg/L-min with a free residual of
2.0 mg/L.
Application of the model to pHs above 8, up to 9, was considered
reasonable because the model is substantially sensitive to pH (e.g., CTs
at pH 9 are over three times greater than CTs at pH 6 and over two times
greater than CTs at pH 7). At a pH of 9, approximately four percent of
the hypochlorous acid fraction of free chlorine is still present. Recent
data indicate that in terms of HOC1 residuals (versus total free chlorine
residuals including HOC1 and OC1") the CT products required for inactiva-
tion of Giardia muris and Giardia lamblia cysts decrease with increasing
pH from 7 to 9 (Leahy et al., 1987; Rubin et al., 1988b). However, with
increasing pH, the fraction of free chlorine existing as the weaker
oxidant species (OCT) increases. In terms of total free chlorine
residuals (i.e., HOC1 and OCV) the-CT products required for inactivation
of Giardia muris cysts increase with increasing pH from 7 to 9 by less
than a factor of 2 at concentrations of less than 5.0 mg/L (see
Table F-l). Also, the significance of pH on the value of CT products
achieving 99 percent inactivation appears to decrease with decreasing
temperature and free chlorine concentration. The relative effects of pH,
temperature, and chlorine concentration, on inactivation of Giardia muris
cysts appears to be the same for Giardia lamblia cysts (Rubin et al.,
1988b), although not as much data for Giardia lamblia cysts for high pH
and temperature values as for Giardia muris cysts is yet available.
F.I.2 Ozone and Chlorine Dioxide
The CT values for ozone in Table E-10 are based on disinfection
studies using in vitro excystation of Giardia lamblia
(Wickramanayake, G. B., et al., 1985). CT99 values at 5 C and pH 7 for
ozone ranged from 0.46 to 0.64 (disinfectant concentrations ranging from
0.11 to 0.48 mg/L). No CT values were available for other pHs. The
highest CT99 value, 0.64, was used as a basis for extrapolation to obtain
the CT values at 5 C, assuming first order kinetics and applying a safety
factor of 2, e.g., (0.64 X 3/2 X 2 = 1.9). CT values for temperatures
F-2
-------
7
8
9
TABLE F-l
CT VALUES TO ACHIEVE 99 PERCENT
INACTIVATION OF GIARDIA HURIS CYSTS BY FREE CHLORINE
Temperature
1
15
1
15
1
15
(Source: Rubin,
0.2-0.5
500
200
510
440
310
et al., 1988b)
Concentration
0.5-1.0
760
290
820
220
1,100
420
1.0-2.0
1,460
360
1,580
1,300
620
2.0-5.C
1,200
290
1,300
320
2,200
760
-------
above 5 C were estimated assuming a twofold decrease for every 10 C. CT
values for temperatures at 0.5 C were estimated assuming a 1.5 times
increase to CT values at 5 C.
The CT values for chlorine dioxide in Table E-8 are based on
disinfection studies using in vitro excystation of Giardia muris CT99
values at pH 7 and 1 C, 5 C, 15 C and 25 C (Leahy, 1985 and Rubin, 1988b).
The average CT99 value at each temperature (27.9 at 1 C, 11.8 at 5 C, 8.5
at 15 C, and 4.7 at 25 C) was extrapolated using first order kinetics and
multiplied by a safety factor of 1.5 to obtain the CT99 9 values, e.g.,
at 1 C, C999 « 27.9 x 1.5 x 1.5 » 63.
Because of the limited data available at pHs other than pH 7, the same CT
values are specified for all pHs. Although most of the CT99 data were
determined at pH 7, it is known that chlorine dioxide is more effective at
pH 9. Thus, the CT values in the rule are more conservative for higher
pHs than for lower pHs.
A lower safety factor is used for chlorine dioxide than for ozone,
because the data was generated using Giardia muris cysts which are more
resistant than Giardia lamblia cysts. CT values at other temperatures
were estimated, based on the same rule of thumb multipliers assumed for
ozone.
A larger safety factor was applied to the ozone and chlorine dioxide
data than to the chlorine data because:
a. Less data were available for ozone and chlorine dioxide than
for chlorine;
b. Data available for ozone and chlorine dioxide, because of the
limitations of the excystation procedure, only reflected up to
or slightly beyond 99 percent inactivation. Data for chlo-
rine, based on animal infectivity studies rather than excysta-
tion procedures, reflected inactivation of 99.99 percent.
Extrapolation of data to achieve CT values for 99.9 percent
inactivation with ozone and chlorine dioxide, involved greater
uncertainty than the direct determination of CT values for
99.9 percent inactivation using chlorine.
c. The CT values for ozone and chlorine dioxide to achieve 99.9
percent inactivation are feasible to achieve; and
d. Use of ozone and chlorine dioxide is likely to occur within
the plant rather than in the distribution system (versus
chlorine and chloramines which are the likely disinfectants
F-3
-------
for use in the distribution system). Contact time measure-
ments within the plant will involve greater uncertainty than
measurement of contact time in pipelines.
EPA recognizes that the CT values for ozone and chlorine dioxide are
based on limited data. Therefore, EPA encourages the generation of
additional data in accordance with the protocols provided in Appendix G to
determine conditions other than the specified CT values, for providing
effective disinfection at a particular system.
F.I.3 Chloramines
- >*
The CT values for chloramines in Table E-12 are based on disinfec-
tion studies using preformed chloramines and in vitro excystation of
Giardia muris (Rubin, 1988). Table F-2 summarizes CT values for achieving
99 percent inactivation of Giardia muris cysts. The highest CT values for
achieving 99 percent inactivation at 1 C (2,500) and 5 C (1,430) were each
multiplied by 1.5 (i.e., first order kinetics were assumed) to estimate
the CT99 9 values at 0.5 C and 5 C, respectively, in Table E-12. The CT99
value of 970 at 15 C was multiplied by 1.5 to estimate the CT99 9 value.
The highest CT99 value of 1,500 at 15 C and pH 6 was not used because it
appeared anomalous to the other data. Interesting to note is that among
the data in Table F-2 the CT values in the lower residual concentration
range (<2 mg/L) are higher than those in the higher residual concentration
range (2-10 mg/L). This is opposite to the relationship between these
variables for free chlorine. For chloramines, residual concentration may
have greater influence than contact time on the inactivation of Giardia
cysts within the range of chloramine residual concentrations practiced by
water utilities (less than 10 mg/L). No safety factor was applied to
these data since chloramination, conducted in the field, is more effective
than using preformed chloramines. Also, Giardia muris appears to be more
resistant than Giardia lamblia to chloramines (Rubin, 1988b).
The protocol in Appendix G can be used to demonstrate if less
stringent disinfection conditions than those cited in Table E-12 can
achieve comparable levels of inactivation for specific system characteris-
tics.
F-4
-------
TABLE F-2
CT VALUES FOR 99 PERCENT
INACTIVATION OF GIARDIA MURIS CYSTS BY MONOCHLORAMINE*
EH
6
8
(Source:
Temperature
(C)
15
5
.1
15
5
1
15
5
1
15
5
1
Rubin, 1988)
Monochloramine
<0_^
1,500
>1,500
>1 , 500
>970
>970
2,500
1,000
>1,000
>1,000
890
>890
>890
Concentration fmp/L)
2.0-10.0
880
>880
>880
970
1,400
>1,400
530
1,430
1,880
560
>560
>560
*CT values with ">" signs are extrapolated from the known data.
-------
F.2 Inactivation of Viruses
F.2.1 Free Chlorine
CT values for free chlorine are based on data by Sobsey (1988) for
inactivation of Hepatitus A virus (HAV), Strain HM175, at pH 6,7,8,9 and
10, chlorine concentrations of 0.5 to 0.2, and a temperature of 5 C, as
contained in Table F-3. The highest CT value for the pH range 6-9 for
achieving 2, 3, and 4-log inactivation of HAV were multiplied by a safety
factor of 3 to obtain the CT values listed in Table E-7. (e.g., the CT
value for achieving 4-log inactivation at pHs 6-9 was determined by
multiplying 2.55 X 3 = 7.6 = 8). The CT values at pH 10 were significant-
ly higher than those for pHs 6-9 and are considered separately. The CT
values in Table E-7 for pH 10 also include a safety factor of 3. CT
values at temperatures other than 5 C were determined assuming a two fold
decrease for every 10 C increase. CT values for inactivating viruses in
general are based on HAV data since they give higher CT values than those
for inactivation of polio and rotaviruses under similar conditions of pH
and temperature (Hoff, 1986).
F.2.2 Chlorine Dioxide
Data by Sobsey (1988) for inactivation of Hepatitus A virus, strain
HM 175, by a chlorine dioxide concentration of 0.5 mg/1 at pH 6 and 5 C is
shown in Table F-4. The CT values in Table E-9 for pHs 6-9 and tempera-
ture = 5 C were determined by applying a safety factor of 2 to the average
CT values presented in Table F-4 at pH 6. This safety factor is lower
than that used to determine CT values for chlorine because chlorine
dioxide appears to be significantly more effective at higher pHs and most
waters are assumed to have a higher pH than 6.
CT values at temperatures other than 5 C in Table E-9 were
determined by applying a twofold decrease for every 10 C increase. The
data for pH 9 was not considered because it is very limited and other
viruses are more resistant to chlorine dioxide than Hepatitus A is at pH
9. According to Hoff (1986) at a pH of 9 and a temperature of 21 C, a CT
of 0.35 provides a 4-log inactivation of poliovirus 1. Applying the same
safety factor and rule of thumb multipliers to this data results in a CT
F-5
-------
. of 2.8 for a 4-log virus inactivation at 0.58C, in contrast to a CT of
50.1 resulting from the Hepatitus A data at pH 6. Therefore, in order to
assure inactivation of Hepatitus A, the higher CT values are needed.
Systems with high pHs may wish to demonstrate the effectiveness of
chlorine dioxide at lower CT values based on the protocol in Appendix G.
Chlorine dioxide is much more effective for inactivating rotavirus and
polio virus than it is for inactivating HAV (Hoff 1986).
F.2.3 Chloramines
The CT values in Table E-13 at 5 C were based directly on data by
Sobsey (1988) using preformed chloramines at pH 8. No safety factor was
applied to the laboratory data since chloramination in the field, where
some transient presence of free chlorine would occur, is assumed more
effective than preformed chloramines.
HAV is less resistant to preformed chloramines than are other
viruses. For example, CTs of 3,800-6,500 were needed for 2-log inactiva-
tion of simian rotavirus at pH » 8.0 and temperature « 5 C (Berman and
Hoff, 1984). However, these same viruses are very sensitive to free
chlorine. CT values ranging from less than 0.025 to 2.16 were required to
achieve 99 percent inactivation of rotavirus by free chlorine at pH = 6-10
and temperature « 4-5 C (Hoff, 1986). HAV is more resistant to free
chlorine than are rotaviruses.
The CT values in Table E-13 apply for systems using combined
chlorine where-chlorine is added prior to ammonia in the treatment
sequence. This should provide sufficient contact with free chlorine to
assure inactivation of rotaviruses. CT values Table E-13 should not be
used for estimating the adequacy of disinfection in systems applying
preformed chloramines or ammonia ahead of chlorine, since CT values based
on HAV inactivation with preformed chloramines may not be adequate for
destroying rotaviruses. In systems applying preformed chloraraines, it is
recommended that inactivation studies as outlined in Appendix G be
performed with Bacteriophage MS2 as the indicator virus to determine
sufficient CT values. Also, the protocol in Appendix G can be used by
systems applying chlorine ahead of ammonia to demonstrate less stringent
disinfection conditions than those indicated in Table E-13.
F-6
-------
TABLE F-3
CT VALUES FOR INACTIVATION OF HEPATITUS A VIRUS
BY FREE CHLORINE
(Source: Sobsey 1988)
LOG INACTIVATION DH
& z a a
2 1.18 0.70 1.00 1.25
3 1.75 1.07 1.51 1.9
4 2.33 U43 2.03 2.55
IP.
19.3
14.6
9.8
-------
TABLE F-4
CT VALUES FOR INACTIVATION OF HEPATITUS A VIRUS
BY CHLORINE DIOXIDE (SOBSEY 1988)
pH6
pH9
Experiment
No.
1
2
3
4
1
2
_£1Q,
' t
Concentration (ma/L^
Initial Average
0.49
0.50
0.51
0.51
0.5
0.5
0.32
0.33
0.36
0.37
0.5
0.5
Inactivation Time
Log
Inactivation
pH6
pH9
>2.5
>3.6
Experiment No.
1
12
30
55
9
29
59
0.33 --
0.33 --
3
5
22
43
7
20
39
Exoeriment
1
3
9
17
<0
<0
.8
.4
.17
.17
2
3
9
20
.0
.6
3
1
7
16
.8
.9
No
4
2
7
14
.6
.4
Average
CT
2.8
8.6
16.7
<0.17
<0.17
Note;
1. CT values were obtained by multiplying inactivation time by the average
concentration shown above for each experiment.
-------
F.2.4 Ozone
No laboratory CT values based on inactivation of HAV virus are yet
available for ozone. Based on data from Roy (1982), a mean CT value of
0.2 achieved 2-log inactivation of poliovirus 1 at 5 C and pH 7.2. Much
lower CT values are needed to achieve a 2-log inactivation of rotavirus
(Vaughn, 1987). No CT values were available for achieving greater than a
2-log inactivation. The CT values in Table E-ll for achieving 2-log
inactivation at 5 C were determined by applying a safety factor of 3 to
the data from Roy (1982). CT values for 3 and 4-log inactivation were
determined by applying first order kinetics and assuming the same safety
factor of 3. CT values were adjusted for temperatures other than 5 C by
applying a twofold decrease for every 10 C increase. Based on the
available data, CT values for ozone disinfection are not strongly
dependent on pH. Therefore, data obtained at pH = 7.2 is assumed to apply
for pHs in the range of 6.0 to 9.0. However, it should be noted that the
maintenance of an ozone residual is affected by pH.
F.2.5 Ultraviolet Light (UV)
The CT values for inactivation of viruses by UV are based on studies
by Sobsey (1988) on inactivation of Hepatitis A virus (HAV) by UV. These
data were used because HAV has been established as an important cause of
waterborne disease. The CT values were derived by applying a safety
factor of 3 to the HAV inactivation data. The CT values in Table E-14 are
higher than the CT values for UV inactivation of poliovirus 1 and simian
rotavirus from previous studies (Chang et al., 1985).
F.2.6 Potassium Permanganate
Potassium permanganate is a commonly used oxidant in water
treatment. Preliminary testing by Yahya, et al 1988, indicates that
potassium permanganate may contribute to virus inactivation. The test
data included in Table F-5 indicates the inactivation of bacteriophage
MS-2 using potassium permanganate with a pure water-buffer solution.
These data do not include safety factors. It is likely that CT values for
actual water treatment processes will differ from these values. This data
has only been provided here as an indication of the potential of potassium
F-7
-------
TABLE F-5
CT VALUES FOR 2-LOG INACTIVATION
OF MS-2 BACTERIOPHAGE WITH POTASSIUM PERMANGANATE
KMn04
(mg/L)
0.5
1.5
2.0
5.0
Notes;
1.
2.
oH 6.0
27.4 a(1)
32.0 a
ND(I)
63.8 a
Letters indicate different experimental
Not determined.
pH 8.0
26.1 a
50.9 b
53.5 c
35.5 c
conditions.
-------
THE BASIS FOR GIARDIA C T VALUES IN THE SURFACE WATER
TREATMENT RULE: INACTIVATION BY CHLORINE
by
Robert M. Clark, Director
Drinking Water Research Division
Risk Reduction Engineering Laboratory
Cincinnati, OH 45268
and
Stig Regli
Office of Drinking Water
U.S. Environmental Protection Agency
Washington, DC 20460
RISK REDUCTION ENGINEERING LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OH 45268
February 1991
-------
THE BASIS FOR GIARDIA C T VALUES IN THE SURFACE WATER
TREATMENT RULE: INACTIVAT10N BY CHLORINE
by
Robert M. Clark,' and Stig Reglib
INTRODUCTION
The 1986 amendments to the Safe Drinking Water Act (SDWA) require EPA to
promulgate primary drinking water regulations (a) specifying criteria under
which filtration would be required, (b) requiring disinfection as a treatment
technique for all public water systems, and (c) establishing maximum
contaminant levels (MCLs) or treatment requirements for control of Giardla
1amblia. viruses, Leqionella. heterotrophic plate count bacteria, and
turbidity. EPA has promulaged treatment technique requirements to fulfill the
SDWA requirement for systems using surface waters and ground waters under the
direct influence of surface water.1 Additional regulations specifying
disinfection requirements for systems using ground water sources not under the
direct influence of surface water will be proposed and promulgated at a later
date. This paper presents a model that relates pH, temperature, chlorine
concentration, and inactivation level to Giardia inactivation by free
chlorine. Because Giardia lamblia is known to be one of the most resistant
organisms to disinfection by chlorine found in water, much interest and effort
Director, Drinking Water Research Division, Risk Reduction Engineering
Laboratory, 26 W. Martin Luther King Drive, Cincinnati, Ohio 45268
bUSEPA, Office of Drinking Water, 401 M Street, S.W., Washington, DC 20460
-------
has been devoted to determination of C't values for Giardia lamblia. The
model has been used to predict *C t" values that have been included as part of
EPA's Surface Water Treatment Rule (SWTR).
BACKGROUND
Under the SWTR all community and non-community public water systems using
surface water, or ground water under the direct influence of surface water,
are required to provide minimum disinfection to control Giardia lamblia.
enteric viruses and bacteria.1 In addition, unless the source water is well
protected and meets certain water quality criteria (total or fecal coliforms
and turbidity limits), treatment must also include filtration. The treatment
provided, in any case, is required to achieve at least 99.9 percent removal
and/or inactivation of Giardia 1ambli a cysts and at least 99.99 percent
removal and/or inactivation of viruses (i.e., virus of fecal origin and
infectious to humans). Unfiltered systems are required to demonstrate that
disinfection alone achieves the minimum performance requirements by monitoring
disinfectant residual(s), disinfectant contact time(s), pH (if chlorine is
used), and water temperature. These data must be applied to determine if their
"C t" value [the product of disinfectant concentration (mg/L) and disinfectant
contact (minutes)] equals or exceeds the C't values for Giardia 1ambli a
specified in the SWTR.1 With the exception of chloramines, where ammonia is
added prior to chlorine, these C't values are also adequate to achieve greater
than 99.99 percent inactivation of viruses. For filtered systems, states are
required to specify the level of disinfection for each system to ensure that
their overall treatment achieves at least 99.9 and 99.99 percent removal
and/or inactivation of Giardia 1ambli a cysts and viruses, respectively.1
In the Guidance Manual to the SWTR, EPA recommends C't values for
-------
different disinfectants to achieve levels of inactivation for unfiltered
systems. Filtered systems will be required to achieve less inactivation then
required for unfiltered systems. The percent inactivation that filtered
systems should achieve as a function of the filtration technology in place and
source water quality conditions is also recommended.2
PROBLEM
The destruction of pathogens by chlorination is dependent on a number of
factors, including water temperature, pH, disinfectant contact time, degree of
mixing, turbidity, presence of interfering substances, and concentration of
chlorine available. The pH has a significant effect on inactivation
efficiency because it determines the species of chlorine found in solution,
each of which has a different inactivation efficiency.
The impact of temperature on disinfection efficiency is also significant.
For Example, Clarke's work in virus destruction by chlorine indicates that
contact time must be increased two to three times when the temperature is
lowered 10°C.3 Disinfection by chlorination can inactivate Giardia cysts, but
only under rigorous conditions. Most recently, Hoff et al. concluded that (1)
these cysts are among the most resistant pathogens known, (2) disinfection at
low temperatures is especially difficult, and (3) treatment processes prior to
disinfection are important.4
Typical C t values for 99 percent inactivation of Giardia lamblia by free
chlorine at different temperatures and pH values are shown in Table 1.
-------
TABLE 1. C T VALUES FOR 99% INACTIVATION OF GIARDIA4
LAMBLIA CYSTS BY FREE CHLORINE
Ranae
Temp
(°c
5
15
25
PH
6
7
8
6
7
8
6
7
8
Disinfectant
Concentration
(mg/L)
1.0-8.0
2.0-8.0
2.0-8.0
2.5-3.0
2.5-3.0
2.5-3.0
1.5
1.5
1.5
Time
(min)
6-47
7-42
72-164
7
6-18
7-21
< 6
< 7
< 8
Ct
47-84
56-152
72-164
18-21
18-45
21-52
< 9
<10
<12
Mean
C t
65
97
110
20
32
37
< 9
<10
<12
No. of
Experiments
4
3
3
2
2
2
1
1
1
Jarroll et al., using in vitro excystation to determine cyst viability,
showed that greater than 99.8 percent of Giardia lamblia cysts can be killed
by exposure to 2.5 mg/L of chlorine for 10 minutes at 15°C and pH 6, or after
60 minutes at pH 7 or 8. At 5°C, exposure to 2 mg/L of chlorine killed at
least 99.8 percent of all cysts at pH 6 and 7 after 60 minutes.5 While it
required 8 mg/L to kill the same percentage of cysts at pH 6 and 7 after 10
minutes, it required 8 mg/L to inactivate cysts to the same level at pH 8
after 30 minutes. Inactivation rates decreased at lower temperatures and at
higher pH values as indicated by the higher C t values.
Because of the obvious interactions among these variables it is essential
that a model be developed for predicting C t values under the various
conditions that may exist in drinking water systems.
-------
OBJECTIVE
As Indicated, many factors influence Giardia lamblia reaction kinetics.
The objective of the study described in this paper therefore is to develop an
equation that will relate f/t values for Giardia inactivated by chlorine to
such factors as pH, temperature, level of inactivation and chlorine
concentration. As mentioned previously, this equation ultimately provided the
values presented in the SWTR and associated Guidance Manual for disinfection
of Giardia 1amblia by free chlorine.
SIGNIFICANCE
The significance of these efforts relates to the fact that EPA's Office
of Drinking Water has adopted the C't concept to quantify the inactivation of
Giardia 1ambli a by disinfection with free chlorine. Whether or not a utility
is forced to install surface water treatment will depend on its ability to
meet the C't values specified by the SWTR. Even if the utility is not
required to install filtration a utility may have to make significant
investments in holding basins and disinfection capacity in order to meet these
requirements. Therefore C't values established under the SWTR will be
extremely important to the drinking water industry and the authors believe it
»-k
is important that the industry understand the basis for the procedures used to
estimate these values. This paper describes the way in which C t values were
calculated for the SWTR. It is unlikely that utilities can directly use the
models developed in this paper, although it is important that they understand
the mechanism by which C't values have been derived. Tables generated from
the model will be useful as they provide the C't values for Giardia inacti-
vation by chlorine that utilities must achieve. These tables are presented at
the end of the paper.
-------
THEORY
Current disinfection theory is based on the Chick or Chick-Watson model.
Chick's law expresses the rate of destruction of microorganisms based on a
first-order chemical reaction.6
dN/dt - -kt (1)
which when integrated yields
In (Nt/No) - -kt (2)
where
Nt - number of organisms present at time t (minutes)
NQ « number of organisms present at time 0
k » rate constant characteristic of type of disinfectant,
microorganism, and water quality aspects of system (minutes'1)
«
t - time (minutes)
Watson, using Chick's data, refined this equation to produce an empirical
relation that included changes in the disinfectant concentration:7
In (N/No) - r Cnt (3)
where
C = concentration of disinfectant [(milligrams/liter)17"]
r - coefficient of specific lethality (liters/milligram ' minutes)
n « coefficient of dilution (liters/milligrams ' minutes)
or
(1/r) In (Nt/N0) - Cnt (4)
For a given level of survival such as Nt/No « 0.001 (3 log reduction) the left
hand side of equation 4 is a constant K, or
K - Cnt (5)
The value K will vary depending on the level of inactivation.
-------
EFFECT OF OTHER VARIABLES
As indicated previously C't values have been found to be a function of
pH, temperature, disinfectant concentration and level of inactivation.8
Therefore in this study equation 5 was reformulated as follows:
C t « C'(n'n K (6)
where
K - f (pH, temp, I)
I » ratio of organisms at time t to the organisms at time 0 (Nt/NQ)
temp « temperature at which experiment was conducted in °C
pH - pH at which experiment was conducted in pH units
Equation 6 can be rewritten in the form:
C t - R I'CbpHctempd (7)
where
R,a,b,c, and d are coefficient to be determined.
A more convenient form for coefficient estimation and the one used in this
paper is as follows:
t » R I'Cb'1pHctempd (8)
As will be discussed in the following sections these coefficients will be
•»-.
determined by a statistical analysis using appropriate data bases.
COEFFICIENT ESTIMATES
Several data sets are available for estimating the coefficients in
equation 8. Data sets have been developed by Jarroll, Hibler, Rice and
Rubin.5'9'11U1
Much of the available Giardia inactivation data is based on excystation
rather than animal infectivity since it is an easier measure of cyst
viability.11 Hoff et al. compared mouse infectivity and excystation for
-------
determining the viability of £. murls cysts exposed to chlorine and reported
that both methods yielded similar results.12 Hibler et al. used Mongolian
gerbils to determine the effects of chlorine on £. 1amblia cysts.9 In a
series of experiments, cysts were exposed for various time periods to free
chlorine concentrations ranging from 0.4 to 4.2 mg/L at 0.5, 2.5, and 5.0°C
and pH 6, 7, and 8. Each of 5 gerbils was fed 5 x 104 of the chlorine exposed
cysts and subsequently examined for evidence of Infection. Since the test
animals had each received a dose of 5 x 10* of the chlorine exposed cysts and
subsequently examined for evidence of infection and since infectivity studies
with unchlorinated cysts showed that approximately 5 cysts usually constituted
an infective dose, the following assumptions were made depending on the
infectivity patterns occurring in the animals. If all five animals were
infected, it was assumed that C t had produced less than 99.99 percent
inactivation and if no animals were infected, that it had produced greater
than 99.99 percent inactivation.9 If, however, 1-4 animals were infected it
was assumed that the level of viable cysts were 5 per animal and that 99.99
percent of the original cyst population had been inactivated. Hibler
interpolated from the results and provided comprehensive tables showing C t
values at 0.5°C temperature intervals.9 Because of observations indicating
that C t values increased as chlorine concentration increased within the range
of chlorine concentrations used, Hibler et al. advised against use of the C t
values for chlorine concentrations above 2.5 mg/L.
Table 2 summarizes Hibler's data for the different experimental
conditions examined. Column 3 shows the range of chlorine concentrations in
mg/L to which cysts were exposed before being fed to the gerbils, and Column 7
shows the number of experiments which yields 1-4 infected gerbils out of 5.
8
-------
Column 4 shows the range of cyst exposure times and Column 5 contains the
range of C t values that are the product of the chlorine concentration and
cyst exposure time.
TABLE 2. C T VALUES FOR 99.99 PERCENT INACTIVATION BASED
ON ANIMAL INFECTIVITY DATA
Range of
Range of Cyst Exposure Range of Range of Number of
Temp Cone. Time C't values from Predicted Observa-
°C (mg/L) (min) Data C't Values tions
6
6
6
7
7
7
8
8
8
0.5
2.5
5
0.5
2.5
5
0.5
2.5
5
0.56-3.96
0.53-3.80
0.44-3.47
0.51-4.05
0.64-4.23
0.73-4.08
0.49-3.25
0.50-3.24
0.84-3.67
39-300
18-222
25-287
76-600
55-350
47-227
132-593
54-431
95-417
113-263
65-212
50-180
156-306
124-347
144-222
159-526
175-371
200-386
136-192
107-151
93-134
205-295
169-235
156-211
294-410
233-324
209-299
25
15
26
14
14
15
22
21
15
Hibler's data set, based on animal infectivity, is appealing because it
is a more direct indicator of cyst viability than data based on excystation.
However the C t values in this data set are based solely on 99.99 percent
inactivation. The o£her three data sets, based on excystation, have values
calculated for all four parameters in equation 8. Table 3 contains a summary
characterization of the studies on which these data sets were based. Because
no one individual experiment provided the exact characteristics required for
this study an attempt was made to find the "most consistent* set of data for
parameter estimation, which might include several of the data sets discussed.
-------
TABLE 3. CHARACTERIZATION OF fi. LAHBLIA FREE CHLORINE*
INACTIVATION STUDIES USED IN PREDICTIVE MODELS
Reference
No.
5
Cyst
Source
Symptomatic
human
Viability
Assay
excy station
Comments
Conventional
survival curves
8
Gerbils, adapted
from infected
humans. (CDC
isolate)
Symptomatic and
nonsymptomatic
humans
Gerbils adapted
from infected
humans. (Several
isolates used)
based on multiple
samples. End
point - 0.1%
survival
gerbil infec- No survival curves.
tivity (5 Endpoint sought
animals/sample) - 0.01% survival
excystation
excystation
Conventional
survival curves
based on multiple
samples. End
point - 0.1%
survival
Conventional
survival curves
based on multiple
samples. End
*Data provided by DrT John Hoff formerly of USEPA
The Hibler data set was included in all combinations considered because
it was the largest data set, the data set was based on animal infectivity, and
the data reflected higher percent inactivation than required under the SWTR.
Since the data based on excystation, with the exception of a few data points,
only reflected percent inactivation up to 1 log or less than that required
under the SWTR, inclusion of the Hibler data was considered essential for
developing a model that could predict disinfection conditions for achieving
10
-------
99.9 percent inactivation with minimum uncertainty. Filtered systems will
need to know disinfection conditions for achieving less than 99.9 percent
inactivation. Therefore data from at least one of the excystation studies was-
considered essential since the C't values in the SWTR may be used for
calculating partial inactivation levels (i.e., less than 99.9 percent).
A fundamental question that needed to be addressed was the statistical
compatibility of the data sets. Initial regression estimates for each of the
data sets were made using equation 8.13 High V* were obtained for these
fits but significant differences were found for the *R* coefficient or slope.
This indicated that the basic model was adequate but that there were
differences in the coefficients as defined by the individual estimates using
equation 8. It was decided to "anchor* all of the data sets to the Hibler
data set. The approach used was to construct an indicator random variable to
move the regression intercept or slope to compensate for data set
differences.13 The significance of the indicator random variable would
support the hypothesis of different regression surfaces, i.e., incompatibility
of the data sets chosen. The indicator random variable was created in such a
way as to always differentiate between the Hibler data set and other data sets
considered and to move the regression intercept not the slope. The indicator
random variable was defined as follows:
o if Hibler data
Z - [ ] - (9)
1 if other data
Therefore equation 8 was modified as follows:
t » R I*Cb'lpHeteropd10-l (10)
where t, I, C, pH, temp are defined as in equation 8, and R,a,b,c,d,e are
constants determined from regression.
11
-------
Equation 10 can be transformed as follows:
log t • log R + a log I + (b-1) log C + c log pH + d temp + ez (11)
In equation 11 when z « 0 equation 10 is defined over the Hibler data set,
and
t - R I* C6"1 pHc tempd (12)
When z • 1 equation 10 is defined over the remaining data and
t - (R ' 10e) I* Cb'1 pHc tempd (13)
Table 4 displays the data set combinations and regression diagnostics. Note
that z is the indicator random variable.
In Table 4, the first column shows the various data sets considered in
/ the analysis. Column two contains the "r2" values based on equation 13 for
each of the data combinations. Column three indicates major results of the
analysis. For example it was found, for the first data set combination, that
the intercept, and temperature variable were not significant. Column 4 shows
the test that was used to determine whether or not the equation yields biased
results.
As indicated in Column 4 of Table 4 residual plots were used to determine
constant variance and normality. Fortunately a strict assumption of normality
is not required. As stated in Neter, Wasserman and Whitmore "Small departures
from normality do not create any serious problems.14 Major departures, on the
other hand, should be of concern". Further they write, "Unless the departures
from normality are serious, particularly with respect to skewness, the actual
confidence coefficients and risks of error will be close to the levels for
exact normality". In addition because of the large sample size one would
expect the central limit theories would apply and symmetry would not be an
issue.
12
-------
It was found that 90% of the data fell within plus/or/minus 1.64 standard
deviations of the mean. In addition 75% of the data fell within plus or minus
1 minus standard deviation which gives support for the normality assumption.
[For a perfect normal distribution we would expect 68% of the data to lie
within plus or minus 1 standard deviation. Similarly, we would expect 90% to
lie within plus or minus 1.64 standard deviation of the mean].
The indicator random variable for the intercept variable using the
Hibler, Jarroll data base was not significant (p-value » 0.3372). All other
data bases considered had a significant indicator random variable at the 0.095
level of significance. A formal test for differences of intercept and/or
slope between the Hibler and Jarroll data sets was conducted and no difference
was detected.
As mentioned previously the Hibler data set does possesses some desirable
characteristics and it is the largest data set among all data sets available.
However one might argue that by forcing the Hibler data set into the analysis
the possibility has been ignored that the other data sets may be mutually
consistent, and the Hibler data set may represent an "outlier*. In addition,
one might hypothesize that data from different experimental situations
prohibits us from making a reasonable comparison among these excystation
studies. Table 4 shows that the Hibler and Jarroll data sets are compatible.
Since Table 4 also shows that Hibler-Rice and Hibler-Rubin is not consistent,
then it is reasonable to assume that the Jarroll date is not consistent with
the Rice and Rubin data so that the Hibler data is not alone in being
inconsistent with the other data sets. It seems reasonable therefore to start
with the Hibler data set, the largest one, then incorporate other smaller data
sets into the modeling process. Thus logic supports the use of the Hibler,
13
-------
Jarroll data base for extending the model development and the coefficients in
equation 8 were estimated using these data as shown in Table 5 in the log
transformed form.13
TABLE 4. DIAGNOSTIC RESULTS FROM
Data sets considered
Hibler,
Hibler,
Hibler,
Hibler,
Hibler,
Hibler,
Hibler,
Hibler,
Hibler,
Hibler,
Hibler,
Hibler,
Hibler,
Hibler,
Rice, Jarroll, Rubin
Rice, Jarroll, Rubin, z
Rice, Rubin
Rice, Rubin, z
, Jarroll, Rubin
Jarroll, Rubin, z
Rice, Jarroll
Rice, Jarroll, z
Rubin
Rubin, z
Rice
Rice, z
Jarroll
Jarroll, z
R-Square
0.6801
0.7316
0.6649
0.7899
0.6424
0.6879
0.8619
0.865
0.6483
0.7593
0.8548
0.8578
0.8452
0.8459
DATA SET COMBINATION ANALYSIS
Variables
intercept, temp
not-significant
Intercept, temp
not-slanificant
intercept, temp
not-significant
intercept
not-significant
intercept, temp
not-significant
intercept, temp
not-significant
all variables
significant
all variables
significant
temp
not-significant
intercept
not-significant
all variables
significant
all variables
significant
all variables
significant
z not significant
Plots
non-normal data
non-constant var
non-normal data
non-constant var
non-normal data
non-constant var
non-normal data
non-constant var
non-normal data
non-constant var
non-normal data
non-constant var
non-normal data
non-constant var
non-normal data
non-constant var
non-normal data
non-constant var
non-normal data
constant var
non-normal data
constant var
non-normal data
constant var
non-normal data
constant var
non-normal data
constant var
14
-------
TABLE 5. COEFFICIENT ESTIMATES FOR EQUATION 8.
Statistical Analysis
Standard T for HO:
Variable
DF
Coefficient
Error
Parameter«0
PROB > 0 1
Vari
Infl
ance
ation
Factor
INTERCEP
LOGI
LOGCHLOR
LOGPH
LOGTEHP
1
1
1
1
1
-0
-0
-0
2
-0
.902
.268
.812
.544
.146
0
0
0
0
0
.200
.014
.042
.221
.028
-4
-19
-19
11
-5
.518
.420
.136
.535
.117
0
0
0
0
0
.0001
.0001
.0001
.0001
.0001
0.
1.
1.
1.
1.
000
183
033
032
179
In Table 5 column 7 entitled the 'Variance Inflation Factor (VIF)* is
defined as (l-R^) where Rk2 is the coefficient of multiple determination when
Xk is regressed on the other variables in the model. The minimum value of VIF
is 1 if there is no multicollinearity. As shown in column 7 all of the
variance inflation factors are close to one.
DISCUSSION OF MODEL
As discussed in the previous sections the coefficients for equation 8
were determined by a combination of log transformation and linear regression.
An issue to consider is the probability that there is measurement error in the
model's independent variables and the effect that this could have on estimates
of the parameters.~
Regression is intended to fulfill the dual purposes of prediction and
explanation. The purpose of equation 8 is primarily to predict by providing
water utilities guidance as to what C't values will be needed for a desired
level of inactivation. The purpose of this model is to predict C't values and
will not be hampered by measurement error as long as consistency is
maintained.15 Since any measurement is subject to some type of error, the
approach taken to deal with this issue was to provide safe or "conservative
estimates* of C t values.
15
-------
As one of the diagnostic procedures applied to the analysis equation 13
was evaluated for multicolinearity. As can be see from Table 5 all of the
coefficients are highly significant and there is no multicolinearity.
TABLE 6. COLLINEARITY DIAGNOSTICS
Condition
Number
1.000
2.495
2.801
10.662
45.636
VAR PROP
Intercep
0.0002
0.0001
0.0003
0.0147
0.9847
VAR PROP
LOG I
0.0031
0.0063
0.0067
0.9266
0.0574
VAR PROP
LOGCHLOR
0.0214
0.0138
0.9285
0.0029
0.0334
VAR PROP
LOGPH
0.0003
0.0001
0.0004
0.0253
0.9739
VAR PROP
LOGTEMP
0.0174
0.7833
0.0005
0.1918
0.0071
In Table 6 VAR PROP is the variance-decomposition proportion (VDP) and
has a maximum value of 1. A high condition number coupled with high VDP
values for two or more coefficients is an indication of multicollinearity
between those variables. A condition of 45.636 in conjunction with an
intercept VDP of 0.9847 and Log(pH) VDP of 0.9739 indicated a dependency
between the intercept and Log(pH) variable, however, multicollinearity among
the other coefficients were nonexistent.
The final equation used for predicting C t values in the SWTR was based
on equation 8 as follows:
C t - RI'CbpHctempd (14)
Confidence intervals of the coefficients estimate for equation 14 based
on the Bonferroni method at a 99% confidence interval are:14'16
R: ( 0.384, 0.4096)
a: (-0.2321, -0.3031)
b: ( 0.0792, 0.2977)
c: ( 1.9756, 3.1117)
d: (-0.2192, -0.0724)
16
-------
RESULTS
There are many uncertainties regarding the various data sets that might
be considered for calculating C't values. The random variable analysis shows '
the statistical incompatibility among most of these data sets. More work
needs to be done to define the impact of strain variation and in vivo versus
in vitro techniques on C't values. In order to provide conservative estimates
for C t values in the SWTR and the guidance document the authors used the
approach illustrated in Figure 1.
In Figure 1 the 99% confidence interval of the 4 log inactivation level
is calculated. First order kinetics are then assumed so that the inactivation
"line* goes through 1 at C't - 0 and a C't value equal to the upper 99% con-
fidence interval at 4 logs of inactivation. As can be seen the inactivation
line consists of higher C't values than all of the mean predicted C't values
from equation 14, most of the Jarroll et al., and most of the Hibler data
points. Conservative C't values, for a specified level of inactivation, can
be obtained from the inactivation line prescribed by the disinfection condi-
tions. For the example indicated in Figure 1, the appropriate C't for
achieving 99.9% inactivation would be 105. This approach (assumption of first
order kinetics) also provides the basis for establishing credits for sequen-
tial disinfection steps allowed under the SWTR. It should be noted that this
approach provides very conservative estimates at mid range levels of C t.
Note in Figure 1 that some of the individual data points fall outside the
99% confidence interval estimated at the four logs of inactivation. This is
to be expected since the confidence intervals constructed were for mean C't
values, but also indicated the high variability of the Hibler data.
Equation 14 was applied using the above strategy, as a safety factor, to
determine the C't "alues for 99.9 percent inactivation at 0.5°C and 5°C in the
17
-------
•
o
UJ
cc
Q.
O
1
5
-i
D
h-
O
<
•
1
mmm
$
^^
CC
UJ
H-
Z
•
U.
z
o
o
*
0)
O)
<
1 VATION
»-
O
Z
u.
0
CO
0
o
_J
»»
L_
k
o
^f
^f
CM
V ^i
0
04
04
O
O
04
O
00
*~
0
(O
T"
2 w
2 u
3
0 -1
04 <
§ 5
r"
O
CO
O
tn
vw
0
CO
CO
J
LU
>
LU
-J
LU
0
Z
G
LL
O
O
^a
S^
0)
O)
• »
~
O)
e
CNJ
u
LLJ
Z
f^
LL
O
-J
X
o
a:
o
LL
^
g
<
^.
G
01
0
o
to
u
Hi
GC
D
£
^v
rr
k^
LU
^k
Q.
2
LU
1-
• »
(D
u
I
Q.
o
04
O
o
o
o
•
o
o
o
T-
6
0
o
f~
o
d
o
*••
o
o
d
_
o
o
o
d
o
o
o
o
d
LU
CC
D
g
ti-
-Z
-------
final SWTR.1 C t values for temperatures above 5°C were estimated assuming a
twofold decrease for every 10°C increase in temperature since all the Hibler
data was generated at 5°C or less. This general principle is supported by
Hoff.
Application of equation 14 to pHs above 8, up to 9, was considered
reasonable because the model is substantially sensitive to pH (e.g., C ts at
pH 9 are about three times greater than C ts at pH 6 and about two times
greater than C ts at pH 7). At a pH of 9, approximately four percent of the
hypochlorous acid fraction of free chlorine is still present. Other data
indicate that in terms of HOC1 residuals (versus total free chlorine residuals
including HOC1 and OC1~) the C t values required for inactivation of Giardia
muris and Giardia 1amblia cysts decrease with increasing pH from 7 to 9.10
However, with increasing pH, the fraction of free chlorine existing as the
weaker oxidant species (OCT) increases. In terms of total free chlorine
residuals (i.e., HOC1 and OC1") the C't values required for inactivation of
Giardia muris and Giardia lamblia cysts increase with increasing pH from 7 to
9 but generally less than by a factor of 2 at concentrations of less than 5.0
mg/L.'° Table 7 compares the C t values in the proposed SWTR to those given
in the SWTR. The C't values in the proposed SWTR were based only on the
Hibler data and included different safety factors.2'8
19
-------
TABLE 7. COMPARISON BETWEEN MODIFIED APPROACH (MEANS) AND RULE C TS
AT 99.9% INACTIVATION AND 5°C IN THE PROPOSED AND FINAL SWTR
pjj
Concentration 6 789
mg/l Proposed Final Proposed Final Proposed Final Proposed Final
1
2
105
116
108
122
149
165
165
186
216
243
238
269
329
371
312
353
The C't values in the final SWTR are 0-10 percent lower than in the
proposed SWTR. Table 8 presents representative C't values determined by
application of the above described approach.
TABLE 8. CALCULATED C'T VALUES FOR GIARDIA INACTIVATION
USING USING EQUATION 14 AT 0.5°C and 5°
Values for Inactivation of Giardia Cysts
by Free Chlorine at 0.5°C
hlorine
centration
pH » 6
Log Inactivation
0.5 1.0 2.0 3.0
pH « 7
Log Inactivation
0.5 1.0 2.0 3.0
pH - 8
Log Inactivation
0.5 1.0 2.0 3.0
pH = 9
Log Inactivation
0.5 1.0 2.0 3.8
0.4
1
2
3
23
25
28
30
46
49
55
60
91
99
110
121
137
148
165
181
33
35
39
44
65
70
79
87
130
140
157
174
195
210
236
261
46
51
58
64
92
101
115
127
185
203
231
255
277
304
346
382
65
73
83
92
130
146
167
184
260
291
333
368
390
437
500
552
Values for Inactivation of Giardia Cysts
by Free Chlorine at 5°C
:hlorine pH - 6 pH - 7 pH - 8 pH « 9
icentration Log Inactivation Log Inactivation Log Inactivation Log Inactivation
mg/L 0.5 1.0 2.0 3.0 C.5 1.0 2.0 3.0 0.5 1.0 2.0 3.0 0.5 1.0 2.0 3.8
0.4
1
2
3
16
18
19
21
32
35
39
42
65 97
70 105
77 116
84 126
23
25
28
30
46 93
50 99
55 110
61 121
139
149
165
182
33
36
41
45
66 137
72 144
81 162
89 179
198
216
243
268
47 93
52 104
59 118
65 130
186 279
208 312
235 353
259 389
20
-------
Because calculations for the SWTR C t values are the upper limit on the
error bounds associated with equation 14 (Table 8), an equation was developed
to estimate these C't values for 0.5 and 5°C directly. C t values above 5°C
can be estimated by using the method given below to estimate C't values at
5°C, then the assumption that there is a twofold decrease in C t values for
every 10°C increase in temperature can be applied. The equation for the
estimated C't values at 0.5 and 5°C is as follows:
C't - 0.36 pH2'69temp'°'15C° 15(-log I)1'00 (R2 - 0.998} (15)
where the variables in equation 15 are as defined previously.
Table 9 compares the values estimated by equation 15 and the SWTR values
shown in Table 8.
TABLE 9. CALCULATED C T VALUES FOR GIARDIA INACTIVATION
USING EQUATION 15 AT 0.5 AND 5°C
Chi orine
Concentration
mg/L
0.4
1
2 •
3
Chlorine
Concentration
mg/L
0.4
1
2
3
Log
0.5
22
25
27
29
Log
0.5
15
17
19
20
Values
pH - 6
Inactivation
1.0 2.0 -3.0
43
49
55
58
86 129
99 148
109 164
116 174
Values
pH - 6
Inactivation
1.0 2.0 3.0
31
35
39
41
61 91
70 104
77 116
82 123
for Inactivation of
bv Free Chlorine at
pH - 7
Log Inactivation
0.5 1.0 2.0 3.0
33 65 131 196
37 75 149 224
41 83 165 248
44 88 175 263
for Inactivation of
bv Free Chlorine at
pH « 7
Log Inactivation
0.5 1.0 2.0 3.0
23 46 92 138
26 53 106 158
29 58 117 175
31 62 124 186
Giardia
0.5°C
Cysts
pH - 8
Log Inactivation
0.5 1.0 2.0 3.0
47 94
54 107
59 118
63 126
Giardia
5°C
187
214
137
251
Cysts
281
321
355
377
pH « 8
Log Inactivation
0.5 1.0 2.0 3.0
33 66
38 76
42 84
44 89
132
151
167
178
198
227
251
266
Loq
0.5
64
74
81
86
LOQ
0.5
46
52
58
61
pH = 9
Inactivat
1.0
129
147
163
173
2.0
257
294
325
345
pH = 9
Inactivat
1.0
91
104
115
122
2.0
182
208
320
244
ior
3.
3S
4-
4J
5;
ior
3.
27
31
3'
36
21
-------
FUTURE WORK
Because of the importance from an economic and a public health viewpoint
of the calculation of C't values for the inactivation of Giardia 1amblia by
free chlorine, much effort has been expended in developing models that
interrelate the important variables effecting these values.8 The work
reported in this paper reflects the authors attempts to develop such a
relationship for inclusion in the SWTR. However, it also raises a very
interesting point regarding the application of statistical methodology to
public policy decision problems. There is no perfect "regulatory* experiment
that answers all of the textbook questions that could be raised regarding
regulatory decision making. One has to use available data and incorporate the
best judgment that can be brought to bear on a given issue to insure that
public health and welfare is protected. 'The need to combine data sets from
different investigations and then develop a decision rule based on the data,
as shown in this paper, as an example of the this process.
There is no doubt in the authors' mind that other better models may be
developed. For example, Haas' work in applying the Horn model to inactivation
data and incorporating the method of Maximum Likelihood for estimating
parameters is promising.17 The authors believe that the public is best served
by examining problems from many different points of view and encourage others
to pursue these difficult, frustrating but extremely challenging problems.
SUMMARY AND CONCLUSIONS
Amendments to the Safe Drinking Water Act clearly require that all
surface water suppliers in the U.S. filter and/or disinfect to protect the
health of their customers. G. lamblia has been identified as one of the
leading causes of waterborne disease outbreaks in the U.S. G. lamblia cysts
are also one of the most resistant organisms to disinfection by free chlorine.
22
-------
EPA's Office of Drinking Water has adopted the C t concept to quantify the
inactivation of £. lamblia cysts by disinfection. If a utility can assure
that a large enough C't can be maintained to ensure adequate disinfection
then, depending upon site specific factors, it may not be required to install
filtration. Similarly, the C't concept can be applied to filtered systems for
determining appropriate levels of protection.
In this paper, an equation has been developed that can be used to predict
C't values for the inactivation of fi. lamblia by free chlorine based on the
interaction of disinfectant concentration, temperature, pH, and inactivation
level. The parameters for this equation have been derived from a set of
animal infectivity and excystation data. The equation can be used to predict
C't values for achieving 0.5 to 4 logs of inactivation, within temperature
ranges of 0.5 to 5°C, chlorine concentration ranges up to 4 mg/L, and pH
levels of 6 to 8. While the model was not based on pH values above 8, the
model is still considered applicable up to pH level of 9. The equation shows
the effect of disproportionate increases of C't versus inactivation levels.
Using 99% confidence intervals at the 4 log inactivation levels and applying
first order kinetics to these end points a conservative regulatory strategy
for defining C't at various levels of inactivation has been developed. This
approach represents an alternative to the regulatory strategy previously
proposed.
23
-------
ACKNOWLEDGMENTS
The authors would like to acknowledge Ms. Patricia Pierson and Ms. Diane
Routledge for their assistance in preparing this manuscript. The authors are
grateful to Dr. John Hoff, formerly of USEPA, Ms. Shirley Pien and Ms. Eleanor
Read of the Computer Sciences Corporation, Mr. Dennis Black of the University
of Nevada, Las Vegas, and Dr. Charles Haas of the Illinois Institute of
Technology for their review and suggestions to improve the manuscript. The
authors would like to extend a special acknowledgement to Ms. Dianne Wild for
her assistance in the preparation of this manuscript.
24
-------
REFERENCES
1. National Primary Drinking Water Regulations: Filtration, Disinfection,
Turbidity, Giardia lamblia. Viruses, Leaionella. and Heterotrophic
Bacteria. Final Rule, 40 CFR parts 141 and 142. Fed. Reg. 54:124:27486
(June 29, 1989).
2. U.S. Environmental Protection Agency, Office of Drinking Water, Criteria
and Standards Division. Guidance Manual for Compliance with the Filtra-
tion and Disinfection Requirements for Public Water Systems Using Surface
Water Sources, October 1987.
3. Clarke, N. A., Berg, C., Kabler, P. W., and Chang, S. I., "Human Enteric
Viruses in Water: Source, Survival, and Removability*. International
Conference on Water Pollutions Research, Landar, September, 1962.
4. Hoff, J. C., Rice, E. W., and Schaefer III, F. W. "Disinfection and the
Control of Waterborne Giardiasis", In Proceedings of the 1984 Specialty
Conference, Environmental Engineering Division, ASCE, June 1984.
5. Jarroll, E. L., Bingham, A. K., and Meyer, E. A. "Effect of Chlorine on
Giardia Lamblia Cyst Viability". Applied and Environmental Microbiology.
Vol. 41, pp. 483-48, February, 1981.
6. Chick, H., "an Investigation of the Laws of Disinfection:, J. Hygiene, 8,
92 (1908).
7. Watson, W. E., "A Note on the Variation of the Rate of Disinfection with
Change in the Concentration of the Disinfectant", J. Hygiene, 8, 536
(1908).
8. Clark, R. M., Read, E. J., and Hoff, J. C. "Analysis of Inactivation of
Giardia Lamblia by Chlorine". Journal of Environmental Engineering, ASCE,
Vol. 115, No. 1, February, 1989, pp. 80-90.
9. Hibler, C. P., Hancock, C. M., Perger, L. M., Wegrzyn, J. G. and Swabby,
K. D. "Inactivation of Giardia Cysts with Chlorine at 0.5°C to 5.0°C.
american Water Works Association Research Foundation, 6666 West Quincy
Avenue, Denver, Colorado 80235, 1987.
10. Rice, E. W., Hoff, J. C. and Schaefer III, F. W. "Inactivation of Giardia
Cysts by Chlorine", Applied and Environmental Microbiology. Jan. 1982,
Vol. 43, No. 1, pp. 250-251.
11. Rubin, A. J., Evers, D. P., Eyman, C. M., and Jarroll, E. L.,
"Inactivation of Gerbil-Cultured Giardia Lamblia Cysts by Free Chlorine",
Applied and Environmental Microbiology. Oct. 1989, Vol. 55, No. 10, p.
2592-2594.
12. Hoff, J. C., "Inactivation of Microbial Agents by Chemical Disinfectants"
EPA/600/2-86-067.
13. Draper, N. and Smith, H. (1981) Second Edition, Applied Regression
Analysis. Wiley: New York.
25
-------
14. Neter, J. and Wasserman, W. (1974), Applied Linear Statistical Models.
Irwin: Homewood, IL.
15. Fuller, Wayne, Measurement Error Models. John Wiley & Sons, 1987.
16. Belsley, D. A., Kuh, e. and Welsch, R. E. (1980), Regression Diagnostics.
Wiley: New York.
17. Haas, Charles W., and Hillar, B. Statistical Analysis of Data on
Chlorine Inactivation of Giardia Lamblia. Final Report prepared for U.S.
EPA Office of Drinking Water, January 6, 1988.
26
-------
GLOSSARY
d Nt/dt « rate of change of organisms with respect to time
k - inactivation rate in minutes"1
t * time in minutes
Nt « number of organisms at time t
N0 « number of organisms at time 0
r - coefficient of specific lethality (liters/milligram - minutes)
C « concentration of disinfectant [milligrams/liter]17"
n - coefficient of dilution
K « constant at given level temperature, pH and inactivation level
pH - pH in water phase
temp « temperature in °C
I * level of inactivation
C t - concentration in mg/l times time in minutes
R - coefficient to be determined
a » coefficient to be determined
c * coefficient to be determined
d - coefficient to be determined
e - coefficient to be determined
z - coefficient to be determined
VIF » variance inflation factor. If VIF is 1 there is no multicolinearity
VDP - variance decomposition number. If VDP is high for two or more
variables there is an induction of multicolinearity between
variables
Bonferroni technique - a conservative method of estimating confidence
intervals
27
-------
permanganate as a disinfectant. It is not meant to be used as a basis for
establishing CT requirements.
References
Berman, D.; Hoff, J. Inactivation of Simian Rotavirus SA 11 by Chlorine,
Chlorine Dioxide and Monochloramine. Appl. Environ. Microbiol.,
48:317-323, 1984.
Chang, J.C.H.; Ossoff, S.F.; Lobe, D.C.; Dorfman, M.H.; Dumais, C.M.;
Quails, R.G.; Johnson, J.D. Inactivation of Pathogenic and Indicator
Microorganisms. Applied Environ. Micro., June 1985, pp. 1361-1365.
Clark, R.M.; Read, E.J.; Hoff, J.C. Inactivation of Giardia Iambi la by
Chlorine: A Mathematical and Statistical Analysis. Unpublished Report,
EPA/600/X-87/149, DWRD, Cincinnati, OH, 1987.
Clark, R.; Regli, S.; Black, D. Inactivation of Giardia lamblia by Free
Chlorine: A Mathematical Model. Presented at AWWA Water Quality
Technology Conference. St. Louis, Mo., November 1988.
Clark, M.R.; Regli, S. The Basis for Giardia CT Values in the Surface
Water Treatment Rule: Inactivation by Chlorine. Submitted to Journal of
Water Supply Research and Technology-Aqua, August 1990.
Hibler, C.P.; C.M. Hancock; L.M. Perger; J.G. Wegrzn; K.D. Swabby Inacti-
vation of Giardia cysts with Chlorine at 0.5 C to 5.0 C American Water
Works Association Research Foundation, In press, 1987.
Hoff, J.C. Inactivation of Microbial Agents bv Chemical Disinfectants.
EPA/600/52-86/067, U.S. Environmental Protection Agency, Water Engineering
Research Laboratory, Cincinnati, Ohio, September, 1986.
Jarroll, E.L.; A.K. Binham; E.A. Meyer Effect of Chlorine on Giardia
1amb1ia Cyst Viability. Appl. Environ. Microbiol., 41:483-487, 1981.
Leahy, J.G.; Rubin, A.J.; Sproul, O.J. Inactivation of Giardia muris.
Cysts by Free Chlorine. Appl, Environ. Microbiol., July 1987.
Rice, E.; Hoff, J.; Schaefer, F. Inactivation of Giardia Cysts by
Chlorine. Appl. and Environ. Microbiology, 43:250-251, January 1982.
Roy, D., R.S. Engelbrecht, and E.S.K. Chian. Comparative Inactivation of
Six Enteroviruses by Ozone. J. AWWA, 74(12):660, 1982.
Rubin, A. Factors Affecting the Inactivation of Giardia Cysts by Mono-
chloramine and Comparison with other Disinfectants. Water Engineering Re-
search Laboratory, Cincinnati, OH March 1988a.
F-8
-------
Rubin, A. "CT Products for the Inactivation of Giardia Cysts by Chlorine,
Chloramine, Iodine, Ozone and Chlorine Dioxide" submitted for publication
in J. AWWA, December, 1988b.
Sobsey, M. Detection and Chlorine Disinfection of Hepatitus A in Water.
CR-813-024. EPA Quarterly Report. December 1988.
Vaughn, J.; Chen, Y.; Lindburg, K.; Morales, D. Inactivation of Human and
Simian Rotaviruses by Ozone. Appl. Environ. Microbiol., 53(9):2218-2221,
September 1987.
Wickramanayake, G.; Rubin, A.; Sproul, 0. Effects of Ozone and Storage
Temperature on Giardia Cysts. J.AWWA, 77(8):74-77, 1985.
Yahya, M.T., Landeen, L.K., Forsthoefel, N.R., Kujawa, K., and Gerba, C.P.
Evaluation of Potassium Permanganate for Inactivation of Bacteriophage
MS-2 in Water Systems. Copyright 1988, Carus Chemical Company, Ottawa,
Illinois.
F-9
-------
APPENDIX S-l
DETERMINING CHLORAMINE INACTIVATION OF 3IAP.DIA
FOR THE SURFACE WATER TREATMENT RULE
Microbiological Treatment Branch
Risk 'eduction Engineering Laboratory
and
Parasitology and Immunology Branch
invironmental Monitoring Systems Laboratory
U.S. Environmental Protection Agency
26 West Martin Luther King Drive
Cincinnati, Ohio 45263
-------
TABLE OF CONTENTS
I. Materials 3
II. Reagents 4
III. Giardia muris Assay 7
IV. Disinfection Procedures for Giardia 10
/. Procedure for Determining Inactivation ;....12
VI. Bibliography 13
VII. Technical Contacts 14
Appendix
A. Use of the Hemocytometer 15
3. Preparation and Loading of Chamber Slides 20
-------
The Surface Water Treatment Rule requires 39.9% or greater removal/
inactivati'on of Giardia. The following protocol may be used to determine
the percentage of Giardia inactivation obtained by a treatment plant
using ciloramine disinfection.
I. yATEDIALS
A. Materials for Disinfection
1. Stock chlorine solution
2. Stock ammonia solution
3, Stirring'device
4. Incubator or water bath for temperatures below ambient
5, Water fro.n treatment plant
6. Giardia muris cysts
7. Assorted glassware
8. Assorted pipettes
9. Reagents and instruments for determining disinfectant residual
10. Sterile sodium thiosulfate solution
11. Vacuum filter device, for 47mm diameter filters
12. 1.0 jm pore size polycarbonate filters, 47 mm diameter
13. Vacuum source
14. Crushed ice and ice bucket
15. Timer
3. M.aterials for Excystation
1. Exposed and control Giardia muris cysts
2. Reducing solution
3. 0.1 M sodium bicarbonate
4. Trypsin-Tryode's solution
5. 15 ml conical screw cap centrifuge tubes
6. Water bath, 37°C
7. Warm air incubator or slide warming tray, 37'C
8. Aspirator flask
9. Vacuum" source
10. Assorted pipettes
11. Vortex mixer
12. Centrifuge with swinging bucket rotor
13. Chamber si ides
14. Phase contrast microscope
15. Differential cell counter
16. Timer
-------
II. REAGENTS
A. Reducing Solution
Ingredient Amount
glJtathione (reduced form)0.2 g
L-cysteine-HCl 0.2 g
IX Hanks' balanced salt solution 20.0 ml
Dissolve the dry ingredients in the IX Hanks' balanced salt
solution and warm to 378C before use in the experiment.
Prepare fresh, within 1 hour of use.
3. Sodiuoi Bicarbonate Solution, 0.1 H
Ingredient Amount
Sodium bicarbonate0.42 g
Dissolve the salt in 10 to 15 ml distilled water. Adjust
the volume to 50 ml with additional distilled water and
warm to 37'C before use in the experiment. Prepare fresh,
within 1 hour of use.
C. Sodium Bicarbonate Solution, 7.5*
Ingredient Amount
Sodium bicarbonate7.5 g
Dissolve the sodium bicarbonate in 50 ml distilled water.
Adjust the volume to 100 ml with additional distilled
water. Store at room temperature.
D. Sodium Thiosulfate Solution, 10%
Ingredient Amount
Sodium thiosulfate10.0 g
Dissolve the sodium thiosulfate in 50 ml distilled water.
Adjust the volume to 100 ml with additional distilled
water. Filter sterilize the sol ution - through a 0.22 urn
porosity membrane or autoclave for 15 minutes at 1213C.
Store at room temperature.
-------
E. Tyrode's Solution, 20X
Ingredient
Nad
KC1
CaCl2
MgCV6H20
M a L! ^ f\ tU A
• 1 Q i o r U /i n o j
Glucose
Amount
160.0 g
4.0 g
4.0 g
2.0 g
1.0 g
20.0 g
Dissolve the dry ingredients in the order listed in 750 ml
distilled water. Adjust the volume to 1.0 liter with addi-
tional distilled water. If long term storage (up to 1
year) is desired, filter sterilize the solution through a
0.22 u;n porosity membrane.
F. Tyrode's Solution, IX
Ingredient
20X Tyrode1 s solution
Dilute 5 ml of the 20X Tyrode's
of 100 ml with distilled water.
Amount
5.0 ml
solution to a final
volume
Trypsin-Tyrode's Solution
Ingredient
Trypsin,
Na'HC03
IX Tyrode
i :
's
100,
sol
U
ut
.S. Biochemical Co.
ion
Amount
0.
0.
100.
50
15
00
g
g
ml
'-.'ith continuous mixing on a stirplate, gradually add 100 ~1
IX Tyrode's solution to the dry ingredients. Continue
stirring until the dry ingredients are completely dissolved.
Adjust the pH of the solution to 8.0 with 7.5" NaHC03.
ChiVl the trypsin Tyrode's solution to 4°C. NOTE: Trypsii
lots must be tested for their excystation efficiency.
Prepare fresh, within 1 hour of use.
H.
Polyoxyethylene Sorbitan Monolaurate (Tween 20)
(v/v)
Ingredient
Solution, 0.01'
Amount
Tween 20
Add the
well.
O.TmT
Tween 20 to 1.0 liter of distilled water. Mix
-------
I. Vaspar
Ingredient Amount
Paraffin 1 part
Petroleum jelly 1 part
Heat the two ingredients in a boiling water bath until melt-
ing and mixing is complete.
-------
III. GIARDIA MURIS AS5AY
A. Cysts
Giardia muris cysts may be available from commercial sources.
The cysts may be produced in Mongolian gerbils (Meriones unguicu-
1atus) or in mice. Mus musculus, the laboratory mouse, CF-1,
BAL3c, and C3H/he strains have been used to produce j2. muris
cysts. The method is labor intensive and requires a good animal
facility.
In order for the disinfection procedure to work properly, the _G.
muris cysts used must be of high quality. Evaluation of a cyst
suspension is a subjective procedure involving aspects of morpho-
logy and microbial contamination as well as excystment. A good
quality G_. muris cyst preparation should exhibit the following:
1. Examine cyst stock suspension microscopically for the presence
of empty cyst walls (ECW). Cyst suspensions containing equal
to or greater than 1% ECW should not be used for determining
inactivation at any required level. However, if a 99.9»
level of disinfection inactivation is required, the stock
cyst suspension must contain <0.1* ECW.
2. Excystation should be 90* or greater.
3. The cyst suspension should contain little or no detectable
microbial contamination.
4. Good Jj. muris cysts are phase bright with a defined cyst wall ,
peritrophic space, and agranular cytoplasm. Cysts which are
phase dark, have no detectable p.eritrophic space, and have a
granular cytoplasm may be non-viable. There generally should
be no more than 4 to 5% phase dark cysts in the cyst prepara-
tion. -
3ood G. muris cyst 'preparations result when the following
guideTines are followed during cyst purification from feces:
a. Use feces collected over a period of 24 hours or less.
b. The isolation of the cysts from the feces should be done
immediately after the fecal material is collected.
c. Initially, G. muris cysts should be purified from the
fecal materfal by flotation using 1.0 M sucrose.
d. If the £. muris cyst suspension contains an undesirable
density of contaminants after the first sucrose float,
further purification is necessary. Two methods for
further purification are suggested.
1) Cysts may be reconcentrated over a layer of 0.85 M
sucrose in a 50 ml conical centrifuge tube. If this
-------
second exposure to sucrose is not done quickly, high
cyst losses can occur due to their increased bcuyant
density in the hyperosmotic sucrose medium. The
cysts must be thoroughly washed free of the sucrose
immediately after collection of the interface.
2) Cysts can be separated from dissimilar sized contami-
nants by sedimentation at unit gravity, which will
not adversely affect cyst bouyant density, morphology,
or viabil ity.
B. Maintenance of Cysts
1. Preparation of stock suspension
Determine the suspension density of the £. muris cyst prepara-
tion using a henocytometer (see Appendix A"H Adjust the cyst
suspension density with distilled water to approximately 3-5
x 106 cysts/ml.
2. Storage
Store cysts in distilled water in a refrigerator at 4°C.
Cysts should not be used for disinfection experiments if they
are .nore than 2 weeks old (from time of feces deposition).
C. Excystation Assay
A nunber of 3. muris excystation procedures have been described in
the scienti fie 1 iterature (see Bibliography, Section VI). Any of
these procedures nay be used provided 90S or greater excystation
of control, undisinfected _G. muris cysts is obtained. The
following protocol is used to evaluate the suitability of cysts in
the stocL suspension , and to determine excystation in control and
disinfected cysts.
1. For evaluating a cyst suspension or for running an unexposed
control , transfer 5 x 10^ G. muris cysts fron the stock
preparation to a 15 ml conical screw cap centrifuge tube. An
unexposed control should be processed at the same time as the
disinfectant exposed cysts.
2. Reduce the volume of G. muris cyst suspension in each 15 ml
centrifuge tube to 0.5~ml or less by centrifugation at 400 x
g for 2 minutes. Aspirate and discard the supernatant to no
less than 0.2 ml above the pellet.
3. Add 5 ml reducing solution, prewarmed to 37°C, to each tube.
4. Add 5 ml 0.1 M NaHCOs, prewarmed to 37°C, to each tube. NOTE:
Tightly close the caps to prevent the loss of C02. If the
escapes, excystation will not occur.
5. Mix the contents of each tube by vortexing and place in
a 37°C water bath for 30 minutes.
-------
6. Remove the tubes from the water bath and centrifuge each for
2 minutes at 400 x g.
7. Aspirate and discard the supernatant to no less than 0.2 ml
above the pellet and resuspend the pellet in each tube in 10
ml trypsin-Tyrode1s solution chilled to 48C.
8. Centrifuge the tubes for 2 minutes at 400 x g.
9. Aspirate and discard the supernatant to no less than 0.2 ml
above the pellet.
10. Add 0.3 ml trypsin-Tyrode1 s solution, prewarmed *o 37°C, to
each tube. Resuspend the _G. muris cysts by low speed vortex-
ing.
11. Prepare a chamber slide for each tube (see Appendix S).
12. Seal the coverslip on each chamber slide with melted vaspar
and incubate at 37'C for 30 minutes in an incubator or on a
si ide warmer.
13. After incubation, place a chamber slide on the stage of an
upright phase contrast microscope. Focus on the slide with a
low power objective. Use a total magnification of 490X or
more for the actual quantitation. NOTE: Be careful to keep
the objectives out of the vaspar.
13. lihile scanning the slide and using a differential cell coun-
ter, enumerate the number of empty cyst walls (ECVJ), partial-
ly excysted trophozoites (PET), and intact cysts (1C) observed
(see Section V for a further description of these forms and
the method for calculating percentage excystation). If the
percentage excystation in the stock suspension is not 90* or
greater, do not continue with the disinfection experiment.
-------
IV. DISINFECTION PROCEDURES FOR GIARDIA
A. The treatment plant water to be used should be the water influent
into the chloramine disinfection unit process used in the plant.
If chloramine disinfection is performed at more than one point in
the treat-lent process, e.g., prefiltration and postfiltration,
the procedure should simulate as closely as possible actual
treatment practice.
B. Prepare stock ammonia and chlorine solutions to be added to the
treatment plant water to achieve the same stoichiometric relation-
ship between chlorine and ammonia that is used in the water
treatment plant. These solutions should be concentrated enough
so that no more than 2 ml of each solution will be added to the
treatment plant water being disinfected.
C. Determine the contact tine by the methods described in the Surface
'..'ater Treatment Rule and/or the associated Guidance Manual.
D. Rinse a 600 ml beaker with treatment plant water to remove any
extraneous material that may cause disinfectant demand. Then
add 400 ml treatment plant water to the beaker.
E. Mix the contents of the beaker short of producing a vortex in the
center and continue until the conclusion of the experiment.
F. Equilibrate the 600ml beaker and its contents as well as the dis-
infectant reagents to the desired experimental temperature.
3. Adjust the stock _G. muris cyst suspension with distilled water so
that the concentration is 2-5 x 10" cysts/ml.
H. Add 0.5 ml of the adjusted cyst suspension to the contents of the
600 ml beaker.
I. Add the disinfectant reagents to the beaker using the same rea-
gents, the same sequence of addition of reagents, and the sane
time interval between addition of reagents that is used in the
disinfection procedure in the treatment plant.
J. Just prior to the end of the exposure time, remove a sample ade-
quate for determination of the disinfectant residual concentra-
tion. Use methods prescribed in the Surface Water Treatnent ^ule
for the determination of combined chlorine. This residual should
be the sane (±20*) as residual present in the treatment plant
operation.
K. At the end of the exposure time, add 1.0 ml 10% sodium thiosulfate
soljtion to the contents of the 600 ml beaker.
L. Concentrate the _G. muris cysts in the beaker by filtering the
entire contents through a 1.0 urn porosity 47 mm diameter polycar-
bonate filter.
-------
M. Place the filter, cyst side up, on the side of a 150 ml beaker.
Add 10 ill 0.01* Tween 20 solution to the beaker. Using a Pasteur
pipette, wash the _G. myri s cysts from the surface of the filter
by aspirating and Fxpel 1 ing the 0.01% Tween 20 solution over the
surface of the filter.
N. Transfer the contents of the 150 ml beaker to an appropriately
labeled 15 ml screw cap conical centrifuge tube.
0. Keep the tube on crushed ice until the excystation assay is
performed (see Section III, C) on the disinfectant exposed cysts
and on an unexposed control preparation obtained from the stock
cyst suspension.
-------
V. PROCEDURE FOR DETERMINING INACTIVATION
A- Giardia muris Excystation Quantitation Procedure
The percentage excystation is calculated using the following for-
mula:
• excystation = . ECV].!_PET.. x 100,
ECW + PET + 1C
where ECW is.the number of empty cyst walls,
PET is the number of partially excysted trophozoites, and
1C is the number of intact cysts.
An ECW is defined as a cyst wall which is open at one end and is
completely devoid of a trophozoite. A PET is a cyst which has
started the excystation process and progressed to the point where
the trophozoite has either started to emerge or has completely
emerged and is still attached to the cyst wall. An 1C is a
trophozoite rfhich is completely surrounded with a cyst wall
showing no evidence of emergence. For the control, generally 100
forms are counted to determine the percent excystation.
The number of cysts that must be observed and classified (ECW,
PET, 1C) in the disinfected sample is dependent on the level of
inactivation desired and on the excystation percentage obtained
in the control.
ror 3.5, 1 and 2 login, reductions, (68*, 99* and 99* inacti-
vation, respectively], the minimum number of cysts to be
observed and classified is determined by dividing 100 by the
percentage excystation (expressed as a decimal) obtained in
the control.
For a 3 logio reduction (99.9% inactivation) the minimum
number of cysts to be observed and classified is determined
by dividing 1,000 by the percentage excystation (expressed
as a decimal) obtained in the control.
B. Determining Inactivation
The amount of inactivation is determined by comparing the percent-
age excystation of the exposed cyst preparation to the percentage
excystation in the control preparation using the following for-
mula :
% inactivation = 100% - [(exposed % excysted/control % excysted) x 100]
If the percentage excystation in the exposed preparation is zero,
i.e., only 1C (no ECW or ?ET) are observed and counted, use <1 as
the value for "exposed % excysted" in the formula for calculating
I inactivation.
-------
13
VI. BIBLIOGRAPHY
Anerican Public Health Association; American Water Works Association;
Water Potation Control Federation. Standard Methods for the Examina-
tion of Uater and Wastewater, 16th ed. (1985).
Belosevic, M. & G.M. Faubert. Giardia muris: correlation between
oral dosage, course of infection, and trophozoite distribution in the
mouse small intestine. Exp. Parasitol., 56:93 (1983).
Erlandsen, L.S. and E.A. Meyer. Giardia and Giardiasis. Plenum
Press, new York, (1984).
Faubert, G.M. et al. Comparative studies on the pattern of infec-
tion with Giardia spp. in Mongolian gerbils. J. Parasitol., 69:802
(1983).
Feely, D.E. A simplifed method for in vitro excystation of G i ardia
muris. J. Parasitol., 72:474-475 (198T).
Feely, D.E. Induction of excystation of Giardia muris by CO?. 62nd
Annual Meeting of the American Society of Parasitologists, Lincoln,
Nebraska, Abstract No. 91 (1987).
Gonzalez-Castro, J., Bermejo-Vicedo, M.T. and Palacios-Gonzalez, F.
Desenquistaniento y cultivo de Giardia muris. Rev. Iber. Parasitol.,
45:21-25 (1986).
Melvin, C.M. and M.M. Brooke. Laboratory Procedures for the Diagnosis
of Intestinal Parasites. 3rd ed., HHS Publication No. (CDC) 32-8282
(1982).
Miale, J.8. Laboratory Medicine Hematology, 3rd ed. C. V. MosDy
Company, St. Louis, Missouri (1967).
Roberts-Thomson, I.e. et al.. Giardiasis in the mouse: an ani'ma1
model. Gastroenterol., 71:57 (1976).
Sauch, J.F. Purification of Giardia muris cysts by velocity sedi-
mentation. Appl. Environ. Microbiol., 48:454 (1984).
Sauch, J.F. A new method for excystation of Giardia. Advances in
Giardia Research. University of Calgary, Calgary, Canada (In Press).
Schaefer, III, F.W., Rice, E.W., & Hoff, J.C. Factors promoting
In vitro excystation of Giardia muris cysts. Trans. Roy. Soc. Trop.
Med. Hyg., 78:795 (1984).
-------
VII. TECHNICAL CONTACTS:
Eugene W. Rice
Microbiological Treatment Branch
Risk Reduction Engineering Laboratory
U.S. Environmental °rotection Agency
26 West Martin Luther King Drive
Cincinnati, Ohio 45258
Phone: (513) 569-7233
Frank W. Schaefer, III
Parasitology and Immunology Sranch
Environmental Monitoring Systems Laboratory
U.S. Environmental Protection Agency
26 West Martin Luther King Drive
Cincinnati, Ohio 45268
'hone: (513) 559-7222
-------
15
Appendix A: Use of the Hemocytometer
Suspension Density Determination Using the Improved Neubauer (Bright-line)
HemocytO'ieter
The hemocytometer consists of two chambers separated by a transverse
trench and bordered bilaterally by longitudinal trenches. Each chamber
is ruled and consists of nine squares, each 1 x 1 x 0.1 mm with a
volume of 0.1 mrn^. Each square mm is bordered by a triple line. The
center line of the three is the boundary line of the square. (See
Figure 1).
According to the U. S. Bureau of Standards' requirements, the cover
glass must be free of visible defects and must be optically plane on
both sides within plus or ninus 0.002 mm. ONLY HEMOCYTOMETER COVER
GLASSES MAY BE USED. ORDINARY COVER GLASSES AND SCRATCHED HEMOCYTOMETERS
ARE UNACCEPTABLE, as they introduce errors into the volume relationships.
The suspension to be counted must be evenly distributed and free of
large debris, so that the chamber floods properly. The suspension to be
counted should contain 0.011 Tween 20 solution to prevent Giardia cysts
from sticking and causing improper hemocytometer chamber flooding. Cyst
suspensions should be adjusted so that there are a total of 60 to 100 cysts
in the four corner counting squares. Counts are statistically accurate
in this range. If the suspension is too numerous to'be counted, then it
must be diluted sufficiently to bring it into this range. In some cases,
the suspension will be too dilute after concentration to give a statisti-
cally reliable count in the 60-130 cyst range. There is nothing that can
be done about this situation other than to record the result as question-
able.
To use the hemocytometer:
1. Dilute or^concentrate the suspension as required.
2. Apply a clean cover glass to the hemocytometer and load the
hemocytometer chamber with 8-10 yl of vortexed suspension per
chamber. If this operation has been properly executed, the
liquid should amply fill the entire chamber without bubbles or
overflowing Into the surrounding moats. Repeat this step with a
clean, dry hemocytometer and cover glass, 1f loading has been
incorrectly done. See step (12) below for the hemocytometer
cleaning procedure.
3. Do not attempt to adjust the cover glass, apply clips, or in any
way disturb the chamber after it has been filled. Allow the
Giardia cysts to settle 30 to 60 seconds before starting the
count.
4. The Giardia cysts may be counted using a magnification 200-600X.
5. Move the chamber so the ruled area is centered underneath it.
6. Then, locate the objective close to the cover glass "^Uln^Se"
ing it from the side of rather than through the microscope.
-------
16
7. Focus up from the covers!1p until the hemocytometer ruling
appears.
8. At each of the four corners of the chanber is a 1 mm? divided
into 15 squares in which Giardia cysts are to be counted (see
Figure 1). Beginning with the top row of four squares,.count
with a hand tally counter in the directions indicated in Figure
2. Avoid counting Giardia cysts twice by counting only those
touching the top and left boundary lines and none of those touch-
ing the lower and right boundary lines. Count each square mm in
this fashion.
9. The formula for determining the number of Giardia cysts per ml
suspension 1s:
ft of cysts counted x 10 x dilution factor x 1,000 mnr _
# of sq. mm counted 1 mm 1 1 ml
if cysts/ml
10. Record the result on a data sheet similar to that shown in
Figure 3.
11. A total of six different hemocytometer chambers -nust be loaded,
counted, and then averaged for each Giardia cyst suspension to
achieve optimal counting accuracy.
12. After each use, the hemocytometer and coverslip must be cleaned
immediately to prevent the cysts and debris from drying on it.
Since this apparatus is precisely machined, abrasives cannot be
used to clean it as they will disturb the flooding andvTTune
relationships.
a. Rinse the hemocytometer and cover glass first with tap
water-, then 70% ethanol , and finally with acetone.
b. 5ry and polish the hemocytometer chamber and cover glass
with lens paper. Store it in a secure place.
13. A number of factors are known to introduce errors into hemocyto-
meter counts. These include:
a. Inadequate suspension mixing before flooding the chamber.
b. Irregular filling of the chamber, trapped air bubbles,
dust, or oil on the chamber or coverslip.
c. Chamber coverslip not flat.
d. Inaccurately ruled chamber.
e. The enumeration procedure. Too many or too few Giardia
cysts per square, skipping or recounting some Giardia cysts.
-------
17
f. Total number of Giardia cysts counted is too low to
give statistical confidence in result.
g. Error in recording tally.
h. Calculation error; failure to consider dilution factor,
or area counted.
i. Inadequate cleaning and removal of cysts froin the previous
count.
j. Allowing filled chamber to sit too long so that chamber sus-
pension dries and concentrates.
-------
18
I mm.
1/5 mm.
of Chomfev * 0.1 mm.
Figure 1. Hemocytometer platform ruling. Squares 1, 2, 3, and 4 are
used to count Giardia cysts. (From Miale, 1967)
Figure 2. Manner of counting Giardia cysts in 1 square mm. Dark cysts
are counted and light cysts are omitted. (After Miale, 1967)
rtt
-------
-------
-------
Date
Person
Counting
!
Count
1
1
2
3
4
5
6
7
e
9
10
11
12
13
14
15
~16
17
18
19
20
* Cells
Counted
* IW2
Counted
!
Dilution
Fictor
1 Cysts*
ml
Rourfcs
1
cysts/ml . # of cysts counted x 10 x dilution factor x 1.000 mm3
# of sq. mm counted 1 mm I
TniT
Figure 3. Hemocytometer Data Sheet for Glardla Cysts
-------
y
i *
Appendix B. Reparation and Loading of Excystation Chamber Slides
'Jsing tape which is sticky on both sides, cut strips approximately 12
x 3 mm.
2. Apply a strip of the tape to one side of a 22 x 22 mm covers! i?.
3. Apply a second strip of tape to the opposite edge but same side of
the covers! ip.
4. Handling the covers!ip by the edges only, attach the covers!ip to the
center of a 3x1 inch glass slide by placing the taped sides of the
covers! ip down along the long edge of the glass slide.
5. Make sure the covers! ip is securely attached to the slide by lightly
pressing down on the edges of the coverslip with your fingers. Care
should be taken to keep finger prints off the center of the coverslip.
5. To load the chamber slide, place a Pasteur or micro!itar pipette
containing at least 0.2 ml of the Giardia cyst suspension about 2 ^m
from an untaped edge of the covers!ip.Slowly allow the cyst suspen-
sion to flow toward the coverslip. As it touches the coverslip it
will be wicked or drawn rapidly under the coverslip by adhesive forces.
Only expel! enough of the cyst suspension to completely fill the
chanber formed by the tape, slide, and coverslip.
?. '.-Jipe away any excess cyst suspension which is not under the covers!'c
with an absorbant paper towe! , but be careful not to pull cyst
suspension from under the covers!ip.
3. Seal all sides of the coverslip with vaspar to prevent the slide f^on
drying out djring the incubation.
Figure 1. Excystation Chamber Slide
NOTE: Prepared excystation chamber slides may be commercially avail-
able from Spiral Systems, Inc., 5740 Clough Pike, Cincinnati,
Ohio 45244, (513) 231-1211 or 232-3122, or from other sources.
-------
APPErnix 3-2
DETERMINING CHLORAMIfJE INACTIVATION OF VIRUS
FOR THE SURFACE WATER TREATMENT RULE
Microbiological Treatment Branch
Risk Reduction Engineering Laboratory
and
Parasitology and Immunology Branch
Environmental Monitoring Systems Laboratory
U.S. Environmental Protection Agency
26 '/Jest Martin Luther King Drive
Cincinnati , Ohio 45268
-------
TABLE OF CONTENTS
I. Materials 3
II. Reagents and Media .....4
III. MS2 Sacteriophage Assay 6
IV. Disinfection Procedure 8
V. Procedure for Determining Inactivation 9
VI. Bibliography 10
VII. Technical Contacts 11
-------
The Surface Water Treat-ient Rule requires 99.99% or greater removal/
inactivation of viruses. The following protocol may be used to determine
the percentage of virus inactivation obtained by a treatment plant using
chloramine disinfection.
I. MATERIALS
A. Materials for Disinfection
1. Stock chlorine solution
2. Stock ammonia solution
3. Stirring device
4. Incubator .or water bath for less than ambient temperature
5. Water from treatment plant
6. MS2 bacteriophage
7. Assorted glassware
8. Assorted pipettes
9. Aqueous, sterile sodium thiosulfate solution
19. Refrigerator
11. Vortex mixer
12. Timer
B. Materials for MS2 Assay
1. MS2 bacteriophage and its Escherichia coli host
2. Assorted glassware
3. Assorted pipettes
4. Incubator, 37°C
5. Refrigerator
6. Petri dishes, 100 x 15 mm, sterile
7. Vortex mixer
3. Water bath, 453C
9. Sterile rubber spatula
10. EDTA, disodium salt
11. Lyso^yme, crystallized from egg white
12. Centrifuge with swinging bucket rotor
-------
II. PEAGENTS AND MEDIA
A. Tryptone-Yeast Extract (TYE) Broth
Ingredient
Bacto tryptone
Yeast extract
Glucose
NaCl
1.0 M CaC12
Amount
10.0 g
1.0 g
1.0 g
8.0 g
2.0 ml
Dissolve in distilled water to a total volume of 1.0 liter,
then add 0.3 ml of 6.0 M NaOH. This medium should be steri-
lized either by autoclaving for 15 minutes at 121'C or
filtration through a 0.22 jm porosity membrane and then
stored at approximately 4°C. It is used in preparing
bacterial host suspensions for viral assays.
B. Tryptone-Yeast Extract (TYE) Agar
Ingredient Amount
TYE broth 1.0 liter
Agar 15.0 g
The agar should be added to the broth prior to steriliza-
tion. The medium should be sterilized by autoclaving for
15 minutes at 121°C. This medium is used to prepare slant
tubes for maintenance of bacterial stock cultures. The
prepared slant tubes should be stored at approximately 4C
Bottom Agar for Sacteriophage Assay
I 0 f
Ingredient
Bacto tryptone
Agar -
NaCl
KC1
1.0 M CaCl2
Amount
10.0 g
15.0 g
2.5 g
2.5 g
1.0 ml
Dissolve the ingredients in distilled water to a total
volume of 1 liter. The medium should be sterilized by
autoclaving for 15 minutes at 1219C. After autoclaving and
cooling, store at 4°C. Immediately prior to use, liquefy
the medium by heating. Add approximately 15 ml of lique-
fied agar into each Petri dish. This bottom layer serves
as an anchoring substrate for the top agar layer.
-------
D. Top Agar for Bacteriophage Assay
Ingredient
Bacto tryptone
Agar
MaCl
veast extract
Glucose
1.0 M CaCl2
Amount
10.0 g
8.0 g
8.0 g
1.0 g
1.0 g
1.0 ml
Dissolve the ingredients in distilled water to a total
volume of 1 liter. This medium should be sterilized by
autoclaving 15 minutes at 1218C. After cooling, store at
4°C until needed in bacteriophage assays. Immediately
prior to use in assays, liquefy the medium by heating and
then cool to and maintain at a temperature of 45'C.
Salt Diluent for Bacteriophage Assay
Ingredient Amount
NaCl 8.5 g
1.0 M CaCl2 2.0 ml
Dissolve in distilled water to a total volume of 1 liter.
This diluent should be sterilized either by autoclaving
for 15 minutes at 12TC or filtration through a 0.22 .m
porosity membrane. Store at room temperature.
CaCl?, 1.0 M
Ingredient Amount
Ingt
"cTn
Dissolve in distilled water to a total volume of 100 ml.
Autoclave 15 minutes at 1218C or filter sterilize the
solution through a 0.22 urn porosity membrane. Store at
room temperature.
G. Sodium Thiosulfate, 1%
Ingredient Amount
Sodium thiosulfate1.0 g
Dissolve the sodium thiosulfate in 50 ml distilled water.
Adjust the volume to 100 ml with additional distilled
water. Filter sterilize the solution through a 0.22 jm
porosity membrane or autoclave 15 minutes at 121°C. Store
at room temperature.
-------
III.- MS2 SACTERIOPHAGE ASSAY
A. Microorganisms
1. MS2 bacteriophage: catalog number 15597-81, American Type
Culture Collection, 12301 Parklawn Drive, Rockville, MD 20352
2. Bacterial host: Escherichia coli. catalog number 15597,
American Type Culture Collection.
B. Growth and Maintenance of Microorganisms
1. Preparation of bacterial host stock cultures
Inoculate host bacteria onto TYE agar slant tubes, incubate
24 hours at 37°C to allow bacterial growth, and then refriger-
ate at 4°C. At monthly intervals the cultured bacterial
hosts should be transferred to a new TYE agar slant.
2. Preparation of bacteriophage stock suspension
Melt top agar and maintain at 45°C. Add 3 ml of the agar to
a 13 x 100 mm test tube contained in a rack in a 458C water
bath. Add 0.5 to 1.0 ml of the bacteriophage suspension
diluted so that the host bacterial "lawn" will show nearly
complete lysis after overnight incubation. Add 0.1 to 0.2 ml
of a TYE broth culture of the host bacteria that has been
incubated overnight. Mix gently and pour the contents on the
surface of bottom agar contained in a Petri dish that has
been prepared previously. Rock the Petri dish to spread the
added material evenly over the agar surface. After the top
agar solidifies (about 15 minutes), invert the Petri dish
and incubate overnight at 376C. Repeat the above procedure
so that a minimum of 5 but no more than 10 Petri dishes are
prepared^.
Following this incubation and using a sterile rubber spatula,
gently scrape the top and bottom agar layers into a large
beaker. Add to this pool of agar layers an amount of TYE
broth sufficient to yield a total volume of 80 ml. To this
mixture add 0.4 g of EDTA (disodium salt) and 0.052 g of
lysozyme (crystallized from egg white). Incubate this mixture
at room temperature for 2 hours with continuous mixing. Then
centrifuge the mixture for 15 minutes at 3,000 x g. Carefully
remove the upper fluid layer. This fluid layer constitutes a
viral stock suspension for use in subsequent testing and
assays. The viral stock suspension may be divided into
aliquots and stored either frozen or at 4°C.
C. Performance of Sacteriophage Assay
A two-week supply of Petri dishes may be poured with bottom agar
ahead of time and refrigerated inverted at 4°C. If stored in a
refrigerator, allow agar plates to equilibrate to room temperature
-------
before use. Eighteen hours prior to beginning a bacteriophage
assay, prepare a bacterial host suspension by inoculating 5 nl of
7YE broth with a snail amount of bacteria taken directly from a
slant tube culture. Incubate the broth containing this bacterial
inoculum overnight (approximately 13 hours) at 37°C immediately
prior to use in bacteriophage assays as described below. This
type of broth culture should be prepared freshly for each 'day's
bacteriophage assays. If necessary, a volume greater than 5 inl
can be prepared in a similar manner.
On the day of assay, melt a sufficient amount of top agar and
maintain at 45°C in a water bath. Place test tubes (13 x 100 mn)
in a rack in the same water bath and allow to warm, then add 3 ml
of top agar to each tube. Inoculate the test tubes containing
top agar with the bacteriophage samples (0.5 to 1.0 ml of the
sample/tube) plus 0.1 to 0.2 ml of the overnight bacterial host
suspension. Dilute the bacteriophage samples from 10*^ to 10'4
in salt diluent prior to inoculation and assay each dilution in
triplicate. In addition, assay the uninoculated salt diluent as
a negative control. Agitate the test tubes containing top agar,
bacteriophage inoculum, and bacterial host suspension gently on a
vortex mixer, and pour the contents of each onto a hardened
bottom agar layer contained in an appropriately numbered dish.
Quickly rock the Petri dishes to spread the added material evenly,
and place on a flat surface at room temperature while the agar
present in the added material solidifies (approximately 15 min-
utes). Invert and incubate the dishes at 37°C overnight (approxi-
mately 13 hours). The focal areas of viral infection which
develop during this incubation are referred to as "plaques" and,
if possible, should be enumerated immediatly after the incubation.
If necessary, the incubated Petri dishes can be refrigerated at
4°C overnight prior to plaque enumeration. As a general rule,
count only those plates that contain between 20 and 200 plaques.
-------
IV. DISINFECTION PROCEDURE
A. The treatment plant water to be used should be the water influent
into the chloramine disinfection unit process used in the plant.
If chloramine disinfection is performed at more than one point in
the treatment process, e.g. prefiltration and postfiltration, the
procedure should simulate as closely as possible actual treatment
practice.
3. Prepare stock ammonia and chlorine solutions to be added to the
treatment plant water to achieve the same stoichiometric relation-
ship between chlorine and ammonia that is used in the water
treatment plant; These solutions should be concentrated enough
so that no more than 2 ml of each solution will be added to the
treatment plant water being disinfected.
C. Determine the contact time by the methods described in the Surface
Water Treatment Rule and/or the associated Guidance Manual.
D. Rinse two 600 ml beakers with treatment plant water to remove any
extraneous material that may cause disinfectant demand. Then add
400 ml treatment plant water to the beaker. The first beaker
will be seeded with MS2 before the contents are chloraninated.
The second beaker will be an indigenous virus control and will
be chloraninated without addition of extraneous phage.
E. Mix the contents of the beaker short of producing a vortex in the
center and continue until the conclusion of the experiment.
F. Equilibrate the 600 ml beakers and their contents as well as the
disinfectant reagents to the desired experimental temperatjre.
G. Dilute the stock MS2 bacteriophage so that the bacteriophage con-
centration is 1 to 5 x 10s PF'J/ml.
H. Add 1.0 ml of the diluted MS2 bacteriophage to the contents of the
first 600 ml beaker.
I. Remove a 10 ml sample from the contents of the first beaker after
2 minutes of mixing. Assay the MS2 bacteriophage concentration
in this sample within 4 hours and record the results as PFU/ml.
This value is the initial MS2 concentration.
J. Remove a 10 ml sample from the contents of the second beaker
after 2 minutes of mixing. Assay the indigenous bacteriophage
concentration in this sample within 4 hours (at the same time as
you assay the sample from the first beaker) and record the
results as PFU/ml. This value is the initial unseeded concentra-
tion.
K. Add the disinfectant reagents to the contents of both beakers
using the same sequence, time, and concentrations as are used in
the actual treatment plant operations.
-------
L. Just prior to the end of the contact time, remove a volume of sam-
ple adequate for determination of the disinfectant residual con-
centration from both beakers. Use methods prescribed in the
Surface Jater Treatment Rule for the determination of combined
chlorine. This residual should be the same (t20%) as the
residual present in the treatment plant operation.
M. At the end of the exposure time, remove a 10 ml sample from the
first 500 ml beaker and neutralize with 0.25 ml of 1.0* aqueous,
sterile sodium thiosulfate. Assay for the MS2 bacteriophage
survivors and record the results as PFU/ml . This value is the
exposed MS2 concentration.
N. At the end of the exposure time, remove a 10 ml sample from the
second 500 ml beaker and neutralize with 0.25 ml of 1.0% aqueous,
sterile sodiu/n thiosulfate. Assay for the indigenous bacterio-
phage survivors and record the results as PFU/ml . This value is
the exposed unseeded concentration.
V. PROCEDURE FOR
DETERMINING
INACTIVATION
A. Calculation of Percentage Inactivation
Use the following formula to calculate the percent inactivation
of MS2:
1. % inactivation = 130% - [(exposed MS2/initial MS2) x 100]
'Jsing values from Section IV steps I, J, M and N calculate initial
MS2 and exposed MS2 as follows:
2. Initial MS2 (PFU/ml) = I - J.
3. Exposed MS2 (PF'J/ml) = M - N.
If the number of PFU/ml in exposed MS2 is zero, i.e., no plaques
are produced after assay of undiluted and diluted samples, use <1
PFU/ml as the value in the above formula.
5. Comparison of Percentage Inactivation to Log^Q of Inactivation
68% inactivation is equivalent to 0.5 login, inactivation
90% inactivation is equivalent to 1 log^n. inactivation
99% inactivation is equivalent to 2 log^n. inactivation
99.9% inactivation is equivalent to 3 log^n inactivation
-------
VI. BIBLIOGRAPHY
Adams, M.H. Bacteriophages. Interscience Pub!ishers, New York (1959).
American Public Health Association; American Water Works Association;
'Jater Pollution Control Federation. Standard Methods for the Examina-
tion of Water and Wastewater. 16th ed. (1985).
Grabow, U.O.K. et al. Inactivation of hepatitus A virus, other enter-
ic viruses and indicator organises in water by chlorination. Water
Sci. Techno!., 17:657 (1985)
Jacangelo, J.D.; Qlivieri, V.P.; & Kawata, K. Mechanism of inactiva-
tion of microorganisms by combined chlorine. AWWARF Rept., Denver,
CO (1987).
Safe Drinking Water Committee. The disinfection of drinking water.
In: Drinking Water and Health, National Academy Press, Washington,
D.C., 2:5 (1980).
Shah, P. & Me Garnish, J. Relative resistance of poliovirus 1 and coli-
phages *2 and T2 in water. Appl. Microbiol. 24:658 (1972).
'J.S. Environmental Protection Agency. Guidance Manual for Compliance
with the Filtration and Disinfection Requirements for Public Water
Systems Using Surface Water Sources. Appendix'G. U.S. EPA, Office
of Water, Criteria and Standards Division, Washington, D.C. (1983).
Ward, fl.R.; Wolfe, R.L.; * Olson, S.H. Effect of pH, application
technique, and chlorine-to-nitrogen ratio on disinfectant activity of
inorganic chloranines with pure culture bacteria. Appl. Environ.
Microbiol ., 48:508 (1984).
-------
VII. TECHNICAL CONTACTS;
Donald Bernan
Microbiological Treatment Branch
Risk Reduction Engineering Laboratory
U.S. Environmental Drotection Agency
25 West Martin Luther King Drive
Cincinnati, Ohio 45268
Phone: (513) 569-7235
Christon J. Hurst
Microbiological Treatment Branch
Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency
26 West Martin Luther King Drive
Cincinnati , Ohio 45263
"hone: (513) 569-7331
-------
G.3 DETERMINING CHLORINE DIOXIDE INACTIVATION
OF GIARDIA CYSTS AND VIRUS
Giardia Cvsts
The basis for the chlorine dioxide CT values for Giardia cysts in
the Guidance Manual is given in Appendix F.I.2. The CT values are based
on data collected mainly at pH 7. Very little data was available at other
pHs. A review of data from Hoff (1986) indicates that the disinfection
efficiency of chlorine dioxide for bacteria and viruses increases
approximately 2 to 3 fold as pH increases from 7 to 9. Data on which the
CT values in the SWTR are based indicate that at 25 C, £. muris cyst
inactivation CTs were approximately 2 fold higher at pH 7 than at pH 9
(Leahy, 1985). In addition, the data also indicate that chlorine dioxide
efficiency increases as disinfectant concentration increases within the
range studied.
Because the data on effects of chlorine dioxide concentration and
water pH on Giardia cyst inactivation efficiency were very limited, they
were not considered in calculating the Giardia cyst CT values in Appendix
E. However, the data suggest that site specific conditions, i.e. water
pH and disinfectant concentration, can have significant effects on
chlorine dioxide effectiveness. Therefore, the option of allowing the
Primacy Agency to consider the use of lower CT values by individual
systems has been provided.
This approval should be based on acceptable experimental data
provided by the system. The data should be collected using the protocol
provided in Appendix G-l for determining Giardia cyst inactivation by
chloramine with appropriate changes in Section IV A, B, I and J to reflect
the use of chlorine dioxide rather than chloramines. This procedure can
be used for any disinfectant which can be prepared in an aqueous solution
and is stable over the course of the testing. To do this, chloramine
should be replaced with the test disinfectant in the above noted sections.
G.3 - 1
-------
Virus
The basis for the chlorine dioxide CT values for virus in Appendix
F.2.2 consists of limited data from Sobsey (1988). Because the pH 9 data
available were very limited, the CT values are based on the pH 6 data with
a safety factor of 2 applied. As indicated previously, review of data
from a number of studies (Hoff, 1986) shows that chlorine dioxide
efficiency increases 2 to 3 fold as pH increases from 7 to 9.
Because the virus CT values for chlorine dioxide are very conserva-
tive and most systems operate at water pHs higher than those on which the
CT values are based, the option of allowing the Primacy Agency to consider
the use of lower CT values has been provided.
Th: approval should be based on acceptable experimental data
provided by the system. The data should be collected using the protocol
provided in Appendix G.2 with appropriate changes in Sections I A,l and
2 and IV A, B, D, K, and L to reflect the use of chlorine dioxide rather
«
than chloramines. This procedure can be used for any disinfectant which
can be prepared in an aqueous solution and is stable over the course of
the testing. To do this, chloramine should be replaced with the test
disinfectant in the above noted sections.
REFERENCES
Hoff, J.C. Inactivation of Microbial Agents By Chemical Disinfectant's,
EPA/600/52-86/067, U.S. Environmental Protection Agency, Water Engineering
Research Laboratory, Cincinnati, Ohio, September, 1986.
Leahy, J.G. Inactivation of Giardia Mun's Cysts by Chlorine and Chlorine
Dioxide. Thesis, Department of Civil Engineering, Ohio State University,
1985.
Sobsey, M.D. Detection and Chlorine Disinfection of Hepatitis A Virus in
Water. CR813024, EPA Quarterly Report, Dec. 1988.
G.3 - 2
-------
G.4 DETERMING OZONE INACTIVATION OF GIARDIA CYSTS AND VIRUS
G.4.1 'BACKGROUND
The basis for the ozone CT values are given in Appendices F.I.2
fGiardia cysts) and F.2.4 (Virus). As indicated, both sets of CT values
are based on limited data and because of this, the values established are
conservative and employ large safety factors. In addition, the difference
between the way the laboratory experiments used to develop the CT values
and how ozone is used in water treatment presents a problem with
translating the data for field use. The laboratory studies were conducted
using steady state ozone concentrations with ozone continually added
during the contact period. In contrast, steady state ozone concentrations
are not maintained in field use. Also, the effectiveness of ozone
contactors used in field applications may vary from each other and from
the mixing efficiencies applied in the laboratory experiments used to
establish the CT values.
The net effect of all of these differences is to limit the appli-
cability of the CT values in the SWTR and Guidance Manual to individual
systems. Therefore, the option of allowing the Primacy Agency to consider
the use of lower CT values by individual systems has been provided.
This approval should be based on acceptable experimental data
provided by the^system. In general, the procedures provided in Appendix
G.I for determining Giardia cyst inactivation and Appendix G.2 for
determining virus inactivation can be used. However, unlike chloramines
ozone is not a stable disinfectant. Because of ozone's rapid dissipation,
a pilot study must be used in lieu of the batch system to demonstrate the
disinfection efficiency. General considerations for conducting pilot
studies to demonstrate the disinfection ability of ozone or any other
unstable disinfectant are enumerated below.
G.4.2 GENERAL CONSIDERATIONS FOR PILOT TEST
A. All microorganisms, reagents and media are prepared as in-
dicated in sections G.I for Giardia and G.2 for virus.
G.4-1
-------
B. The disinfectant should be prepared, measured and added to the
test water as it would be added to the water at the water
treatment plant.
C. Specific reactor design should be a function of the disinfec-
tant and reflect how the disinfectant is added at the water
treatment plant. Provisions should be made to determine
concentration of disinfectant and microbial survival to be
measured with contact time.
An example of conducting a pilot test for a plug flow reactor using
ozone or another unstable disinfectant is provided below.
Example - Plug Flow Reactor Protocol
The size of the plug flow reactor can be approximated from the table
below. Glass, stainless steel, copper, plastic tubing or other material
compatible with the disinfectant can be used to construct the plug flow
reactor. Table 1 shows the approximate length of pipe for a plug flow
reactor to yield 10 minutes contact at flow rates between 50 and 500
ml/nrin. Depending on pipe size and material an economical reactor can be
constructed.
TABLE 1 APPROXIMATE LENGTH AND DIAMETER OF PIPE
BASED ON FLOW
LINEAR
PIPE. LENGTH. METERS
NOMINAL PIPE DIAMETER, CM
FLOW
ml/mi n
50
100
200
300
400
500
TIME
MIN
10
10
10
10
10
10
VOLUME
L 1 TE£S
0 5
1
2
3
4
5
cc
500
1000
2000
3000
4000
sooo
0 6
0 28
17 7
35 4
70 7
206 1
141. 5
176 I
1 2
1 31
4 4
8 t
17 7
26. 5
3S.4
44 2
1 8
2.54
2 0
3 9
7 9
11. (
15.7
19 6
2 54
5 07
1 0
2 0
3 9
5 9
7.9
9 9
3 81
11 40
0 4
0 9
1 1
2.6
3 5
4. 4
5 08
20 27
0 2
0 5
1 0
1.5
2 0
2.5
Additional information on the design of specific pilot studies can
be found in the following references by Thompson (1982), Montgomery (1985),
and Al-Ani (1985).
Additional Materials to those in G.I and/or G.2
plug flow reactor
cyst suspension, 2xl07 cysts/trial
G.4-2
-------
cyst quantity - cysts are prepared as indicated in G.I.
103 cysts/ml X 20,000 ml = 2xl07 cysts required/trial
MS2 stock, 2xl010/trial
2-20 liter (5 gal) carboy
test water pump, mid range 200 ml/min
disinfectant generator
disinfectant pump, mid range 10-20 ml/min
disinfectant residual reagents and equipment
Test Procedure
A. Reactor conditions
1. Test Water Flow rate* 200 ml/min (this may vary from 50 to 500
ml/min with 20 1 reservoir total experimental time* 100 min)
2. Disinfectant flow
gas-requires specific contactor designed for disinfectant
Liquid*10 to 20 ml/min
3. Temperature
controlled
4. Prepare 20 liter reservoir (5 gal) of test water at the pH and
temperature of the CT trial. Do not add microorganisms
5. Prepare 20 liter reservoir (5 gal) of test water and equi-
librate to the temperature of the CT trial. Add Giardia muris
cysts at an initial density of 103 cysts/ml and/or MS2 bacter-
ial virus at an initial density of 10s PFU/ml. Mix thoroughly
and adjust pH to the pH of the CT trial. Continuous mixing of
the test water feed stock should be carried out over the
course of the CT trial to prevent the Giardia cysts from
settling.
B. Disinfection Procedure - Prior to Disinfection Trial
1. Determine contact time for the sample ports in the plug flow
reactor under conditions of the CT trial by methods described
in the SWTR.
2. Determine disinfectant concentration with no microorganisms in
the feed test water.
G.4-3
-------
C. CT Trial Procedure
1. Start test water feed without cysts and or virus (approx. 200
ml/min), start disinfectant feed (gas or liquid).
Allow system to equilibrate.
Monitor disinfectant residual by appropriate method during
this time. Samples for disinfectant residual should be taken
directly into tubes or bottles containing reagents to fix the
disinfectant at the time the sample is collected. Keep a plot
of disinfectant residual vs running time to evaluate steady
state conditions.
2. After the disinfectant residual has stabilized, switch to the
reservoir containing the test microorganism(s).
3. Allow system to equilibrate for a time « 3 X final contact
time.
example
final contact time =10 min, allow 30 min.
4. Monitor disinfectant residual by appropriate method during
this time. If the disinfectant residual is stable begin
chemical and biological sampling for calculation of CT.
5. Sampling
a. Chemical
A sufficient volume (about 250 ml should be collected
from the sampling tap prior to the biological composite
to determine:
pH
Residual disinfectant - Samples
should be collected directly
into tubes or bottles containing
reagents to fix the disinfectant
at the time the sample is collected.
b. Biological
Samples for microbial analysis are collected as short
time composite samples over a 10 to 20 minute time
period. Several trials may run for a given 20 liter
test water preparation as long as sufficient equilibra-
G.4-4
-------
tion and flow recovery times are allowed between
trials.
- Zero time samples should be collected as 250 ml
composite samples either directly from the test
water feed reservoir or in line prior to the addi-
tion of the disinfectant.
- Four 250 ml samples are collected separately into
a 2 1 sterile bottle containing a neutralizing
agent for the particular disinfectant. Each
sample is thoroughly mixed upon collection and
stored at 4 C. If multiple sample ports are used,
the order of collection should be from longest to
shortest contact time to minimize flow changes due
to sampling.
6. Giardia cyst recovery and assay.
Concentrate the 1000 ml composite sample by filtration
according to the method given in section G.I. Record and
report the data as described in section G.I. The expected
cysts/sample is given below:
Cysts/sample = 4 x 250 ml X 103 cyst/ml * Ixl08cyst/samp1e.
7. Virus Assay
Before filtration for Giarcjia. remove 10.0 ml from the
biological composite sample to a sterile screw cap culture
tube containing 2 to 3 drops chloroform. Assay for MS2,
record and report the virus data according to the methods and
procedures described in G.2. Be sure to correct the Giardia
sample volume to 990 ml.
8. Calculation of CT
Calculate CT in a manner described in Section G.I for Giardia
and Section G.2 for virus. The residual disinfectant should
be the average of the four residual determinations performed
prior to the individual samples collected for the biological
composite and the time should be the time determined for the
sample port under similar flow conditions.
G.4-5
-------
REFERENCES
Al-Ani, C.S.U., Filtration of Giardia Cysts and other substances:
Volume 3. Rapid Rate Filtration (EPA/600/2-85/027) 1985.
Montgomery, James M. Consulting Engineers Inc., Water Treatment Principles
and Design. John Wiley and Sons, May 1982.
Wallis, P.M., Davies, J.S., Nuthonn, R.,Bichanin-Mappin, J.M., Roach, P.O.,
and Van Roodeloon, A. Removal and Inactivation of Giardia Cysts in a
Mobile Water Treatment Plant Under Field Conditions: Preliminary Results.
In Advances in Giardia Research. P.M. Wallis and B.R. Hammand, eds, Union
of Calgary Press, p. 137-144, 1989.
Wolfe, R.L., Stewart, M.H., Liange, S.L., and McGuire, M.J., Disinfection
of Model Indicator Organisms in a Drinking Water Pilot Plant by Using
PEROXONE, Applied Environmental Microbiology, Vol 55, 1989, pp 2230-2241.
Olivieri, V.P. and Sykora, J.L., Field and Evaluation of CT for Determining
the Adequacy of Disinfection. American Water Works Association Water
Quality Technology Conference. In press, 1989.
G.4-6
-------
APPENDIX H
SAMPLING FREQUENCY FOR TOTAL COLIFORMS
IN THE DISTRIBUTION SYSTEM
-------
TABLE H-l
TOTAL COLIFORM SAMPLING REQUIREMENTS
BASED UPON POPULATION
Population
Served
1,
2,
3,
4,
4,
5,
6,
7,
8,
12,
17,
21,
25,
33,
41,
50,
25
001
501
301
101
901
801
701
601
501
901
201
501
001
001
001
001
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
Notes:
1.
1
2
3
4
4
5
6
7
8
12
17
21
25
33
41
50
59
,000
,500
,300
,100
,900
,800
,700
,600
,500
,900
,200
,500
,000
,000
,000
,000
,000
Minimum
Number
of Samples
Per Month11"*'-1'
1
2
3
4
5
6
7
8
9
10
15
20
25
30
40
50
60
Non-community systems ui
59
70
83
96
130
220
320
450
600
780
970
1,230
1,520
1,850
2,270
3,020
3,960
sing all or j
Population
Served
,001
,001
,001
,001
,001
,001
,001
,001
,001
,001
,001
,001
,001
,001
,001
,001
,001
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
or
70
83
96
130
220
320
450
600
780
970
1,230
1,520
1,850
2,270
3,020
3,960
more
,000
,000
,000
,000
,000
,000
,000
,000
,000
,000
,000
,000
,000
,000
,000
,000
?art surface water
Minimum
Number
of Samples
Per Month
70
80
90
100
120
150
180
210
240
270
300
330
360
390
420
450
480
and community
systems'" oust monitor total colifora at this frequency. A r.on-
community water system using ground water and serving 1,000 persons
or fewer must monitor quarterly, beginning 5 years after the rule's
promulgation, although this can be reduced to yearly if a sanitary
survey shows no defects. A non-community water system serving mere
than 1,000 persons during any month, or a non-community water system
using surface water, must monitor at the same frequency as a like-
sized community public water system for each month the system
provides water to the public.
Unfiltered surface water systems must analyze one coliform sample
each day the turbidity exceeds 1NTU.
-------
TABLE H-l
TOTAL COLIFORM SAMPLING REQUIREMENTS
BASED UPON POPULATION (Continued)
3. Systems collecting fewer than 5 samples per month on a regular basis
must conduct sanitary surveys. Community and non-community systems
must conduct the initial sanitary surveys within 5 and 10 years of
promulgation, respectively. Subsequent surveys must be conducted
every 5 years, except for non-community systems using protected and
disinfected ground water, which have up to 10 years to conduct
subsequent surveys.
-------
TABLE H-2
MONITORING AND REPEAT SAMPLE FREQUENCY
System Size
NCWS(1)
25 - 1,000
1,001 - 2,500
2,501 - 3,300
3,301 - 4,100
4,101 - 4,900
>4,900
# Routine
Samples
Quarterly (2)
Monthly (2)
2/mo
3 /mo
4/mo
5 /mo
Table 1
* Repeats
4
4
3
3
3
3
3
More Monitoring For
5 /mo for 1 additional
5/mo for 1 additional
5/mo for 1 additional
5/mo for 1 additional
5/mo for 1 additional
None
None
r.o
mo
mo
mo
mo
Notes;
1. Non-community Water systems.
2. For exceptions, see Table 1.
-------
APPENDIX I
MAINTAINING REDUNDANT
DISINFECTION CAPABILITY
-------
APPENDIX I
REDUNDANT DISINFECTION CAPABILITY
The SWTR requires that unfiltered water systems provide redundant
disinfection components to ensure the continuous application of a
disinfectant to the water entering the distribution system. In many
systems, both filtered and unfiltered, a primary disinfectant is used to
provide the overall, inactivation/removal and a secondary residual is
applied to maintain a residual in the distribution system. As outlined in
Sections 3.2.4 and 5.5.4, redundancy of the disinfection system(s) is
recommended to ensure that the overall treatment requirement of 3-1 og
Giardia cyst and 4-log virus removal/inactivation is achieved, and a
residual is maintained entering the distribution system. This is
particularly important for unfiltered supplies where the only treatment
barrier is disinfection. Redundancy of components is necessary to allow
for disinfection during routine repairs, maintenance and inspection and
possible failures.
In reviewing water disinfection facilities for compliance with
redundancy requirements, the following items should be checked:
I. General
A. Are the capacities of all components of both the primary
system and the backup system equal to or greater than the
required capacities?
Some systems may have two or more units that provide the
required dosage rates when all units are operating. In these
cases, an additional unit is needed as backup, during the
downtime of any of the operating units. The backup must have
a capacity equal to or greater than that of the largest
on-line unit.
B. Are adequate safety precautions being followed, relative to
the type of disinfectant being used?
C. Are redundant components being exercised or alternated with
the primary components?
D. Are all components being properly maintained?
E. Are critical spare parts on hand to repair disinfection
equipment?
1-1
-------
F. Are spare parts available for components that are indispens-
ible for disinfecting the water?
II. Disinfectant Storage
A minimum of two storage units capable of being used alternately
should be provided. The total combined capacity of the storage units
should provide as a minimum the system design capacity.
A. Chlorine
Storage for gaseous chlorine will normally be in 150-lb cylinders,
2,000-lb containers, or larger on-site storage vessels.
1. Is there automatic switchover equipment if one cylinder
or container empties or becomes inoperable?
2. Is the switching equipment in good working order,
(manually tested on a regularly scheduled basis), and
are spare parts on hand?
3. Are the scales adequate for at least two cylinders or
containers.
B. Hvoochlorite
Storage of calcium hypochlorite or sodium hypochlorite is normally
provided in drums or other suitable containers. Redundancy requirements
are not applicable to these by themselves, as long as the required minimum
storage quantity is on hand at all times.
C. Ammonia
An hydrous ammonia is usually stored in cylinders as a pressurized
liquid. Aqua ammonia is usually stored as a solution of ammonia and water
in a horizontal pressure vessel.
1. Is the available storage volume divided into two or more
usable units?
2. Is automatic switching equipment in operation to change
over from one unit to another when one is empty or
inoperable?
3. Are there spare parts for the switching equipment?
1-2
-------
III. Generation
Ozone and chlorine dioxide are not stored on-site. Rather, because
of their reactivity, they are generated and used immediately.
To satisfy the redundancy requirements for these disinfectants it is
recommended that two generating units, or two sets of units, capable of
supplying the required feed rate be provided. In systems where there is
more than one generation system, a standby unit should be available for
times the on-line units need repair. The backup unit should have a
capacity equal to or greater than the unit(s) it may replace.
A. Chlorine Dioxide
Chlorine, sodium chlorite, or sodium hypochlorite should be stored
in accordance with storage guidelines previously described.
B. Ozone
Are all generation components present and in working order for both
the primary and the redundant units (whether using air or oxygen)?
C. Common
Is switchover and automatic start-up equipment installed and
operable to change from the primary generating unit(s) to the redundant
unit(s)?
IV. Feed Systems
Redundancy in feed systems requires two separate units, or systems,
each capable of supplying the required dosage of disinfectant. If more
than one unit is needed to apply the required feed rate, a spare unit
should be available to replace any of the operating units during times of
malfunction. The replacement unit should, therefore, have a capacity
equal to or greater than that of the largest unit which it may replace.
This requirement applies to all disinfection methods, and is best
implemented by housing the on-line and redundant components in separate
rooms, enclosures, or areas, as appropriate.
In reviewing these systems for redundancy, the following components
should be checked:
A. Chlorine
1. Evaporators
2. Chlorinators
3. Injectors
1-3
-------
B. Hvpochlorite
1. Mixing tanks and mixers
2. Chemical feed pumps and controls
3. Injectors
C. Ozone
1. Dissolution equipment, including compressor and delivery
piping systems
D. Chlorine Dioxidg
1. Chlorine feed equipment
2. Sodium chlorite mixing and metering equipment
3. Day tank and mixer
4. Metering pumps
5. If a package C102 unit is used, two must be provided
E. Chloramination
1. Chlorine feed equipment
2. Ammonia feed equipment, including applicable equipment
for either:
a. Anhydrous ammonia (gas)
b. Aqua ammonia (solution)
V. Residual Monitoring
The best method of monitoring a disinfection facility for continuous
operation is by continuous recording equipment. To improve reliability,
it is suggested that duplicate continuous monitors are present for backup
in the event of monitor failure. However, if there is a failure in the
monitoring system for indicating that a continuous residual is being
maintained, the SWTR allows systems to take grab samples every four hours
for up to five days during monitor repair. For systems without 24 hour
staffing it will not be practical to take grab samples and redundant
monitoring equipment is recommended. Failure of continuous monitoring
would be a violation of a monitoring requirement, not a treatment
requirement.
A. Chlorine
1. Does the facility have a continuous monitor for chlorine
residual at the disinfection system site with an alarm
1-4
-------
or indicator to show when the monitor is not function-
ing? For added assurance, the provision of a backup
monitoring unit is also recommended.
2. Is there instrumentation in place to automatically
switch from one monitor to the other if the first one
fails?
B. Hvpochlorite
Same as for chlorine system.
C. Ozone
1. Does the facility have a continuous ozone monitor with
automatic switchover capability and alarms?
2. Does the facility have a continuous ozone residual
monitor with automatic switchover capability and alarms?
D. Chlorine Dioxide
1. Does the facility have a continuous chlorine dioxide
monitor with automatic switchover capability and alarms?
2. Does the facility have a continuous chlorine dioxide
residual monitor with automatic switchover capability
and alarms?
E. Chloramination
1. Does the facility have a continuous ammonia monitor with
_ automatic switchover capability and alarm?
2. Does the facility also have a continuous chlorine
residual monitor on-site with automatic switchover
capability and alarms?
VI. Power Supply
A permanently installed standby generator, capable of running all
electrical equipment at the disinfection station, and equipped for
automatic start-up on power failure, should be on-site and functional.
Alternatives to a standby generator, such as a feed line from a
different power source, are acceptable if they can be shown to have equal
reliability.
1-5
-------
VII. Alarms
Indicators and alarms, both local and remote, should be capable of
promptly alerting operating and supervisory personnel of problem
conditions.
A. Local
Lights, buzzers, and horns should be installed and functioning to
alert on-site personnel to problem conditions.
B. Remote
Alarm signals should be relayed to a central control panel which is
manned 24 hours per day and whose operators can notify response personnel
immediately.
C. Problem Conditions
A minimum list of problem conditions which should have indicators
and alarms, both locally and at a 24-hour per day switchboard, are as
follows:
1. Disinfectant leak
2. Feeder pump failure
3. Power outage
4. Generator or alternate power source on
5. Disinfectant residual less than setpoint value
VIII. Facility Layout
Maximum reliability is ensured when redundant units are separated
from primary unfts. The type of separation should be appropriate to the
type of potential malfunction. For example, any area within a building
subject to a chlorine leak should have primary components separated from
redundant components by an airtight enclosure, i.e., separate rooms of
varying sizes.
IX. Separate Facility
Under certain conditions, such as location of a disinfection
facility in an area of high earthquake potential, the most reliable means
of providing redundant facilities may be to house them in a completely
separate structures at a different site.
1-6
-------
APPENDIX J
WATERSHED CONTROL PROGRAM
-------
APPENDIX J
WATERSHED CONTROL PROGRAM
The following is a guideline for documenting a watershed control
program. The SWTR only requires a watershed control program for
unfiltered supplies. A watershed control program can also benefit a
filtered system by providing protection for maintaining the source water
quality, minimizing the level of disinfection to be provided. It is
therefore recommended that all systems conduct the basic elements of a
watershed control program. However, the scope of the program should
increase as the complexity and size of the watershed/ system increases.
The program could be more or less comprehensive than this outline, and
will be determined on a case-by-case basis by the utility and the Primacy
Agency. In addition to the guidelines below, a wellhead protection
program could be the basis of a watershed control program in many states.
All of the elements found below would also be part of a local wellhead
protection program.
A. Watershed Description
1. Geographical location and physical features of the
watershed.
2. Location of major components of the water system in
relationship to the watershed.
3.~ Hydrology: Annual precipitation patterns, stream flow
characteristics, etc.
4. Agreements and delineation of land use/ownership.
B. Identification of the Watershed Characteristics
and Activities Detrimental to Water Quality
1. Naturally Occurring:
a. Effect of precipitation, terrain, soil types and
land cover
*
b. Animal populations (describe) -- include a dis-
cussion of the Giardia contamination potential,
any other microbial contamination transmitted by
animals
J-l
-------
Other - any other activity which can adversely
affect water quality
2. Man-Made:
a. Point sources of contamination such as wastewater
treatment plant, industrial discharges, barnyard,
feedlots, or private septic systems
The impact of these sources on the microbiological
quality of the water source should be evaluated. In
cases resulting in identifiable degradation, the
discharges should be eliminated in order to minimize the
treatment of the water needed.
b. Nonpoint Source of Contamination:
1) Road construction - major highways, rail-
roads
2) Pesticide usage
3) Logging
4) Grazing animals
5) Discharge to ground water which recharges
the surface source
6) Recreation activities
7) Potential for unauthorized activity in the
watershed
8) Describe any other human activity in the
watershed and its potential impact on water
quality
It should be noted that grazing animals in the watershed
may lead to the presence of Crvptosporidium in the
water. Crvptosporidium is a pathogen which may result
in a disease outbreak upon ingestion. No information is
available on its resistance to various disinfectants,
therefore it is recommended that grazing should not be
permitted on watersheds of non-filtering systems.
Sewage discharges will introduce viruses into the water
source which may be occluded in solids and protected
from inactivation through disinfection. It is, there-
fore, recommended that sewage discharges should not be
permitted within watersheds of non-filtering supplies.
Although it is preferable to not have grazing or sewage
discharges within the watershed, Primacy Agencies will
J-2
-------
need to evaluate the impact of these activities on a
case-by-case basis. In cases where there is a long
detention time and a high degree of dilution between the
point of the activity and the water intake, these activ-
ities may be permissible for unfiltered supplies. The
utility should set priorities to address the impacts in
B.I. and 2., considering their health significance and
the ability to control them.
C. Control of Detrimental Activities/Events
Depending on the activities occurring within the watershed,
various techniques could be used to eliminate or minimize
their effect. Describe what techniques are being used to
control the effect of activities/events identified in B.I. and
2. in its yearly report.
Example:
Activity; Logging in the watershed.
Management Decision; Develop program to minimize impact
of logging.
Procedure; Establish agreements with logging companies
to maintain practices which will minimize adverse
impacts on water quality. These practices should
include:
limiting access to logging sites
ensuring cleanup of sites
controlling erosion from site.
Monitoring; Periodically review logging practices to
ensure they are consistent with the agreement between
the utility and the logging companies.
Example:
Activity; Point sources of discharge within the
watershed.
Management Decision; Eliminate those discharges or
minimize their impact.
Procedures; Actively participate in the review of
discharge permits to alert the reviewing agency of the
potential (actual) impacts of the discharge and lobby
for its elimination or strict control.
J-3
-------
Monitoring: Conduct special monitoring to ensure
conditions of the permit are met and to document adverse
effects on water quality.
D. Monitoring
1. Routine: Minimum specifications for monitoring several
raw water quality parameters are listed in Section 3.1.
Describe when, where and how these samples will be
collected. These results will be used to evaluate
whether the source may continue to be used without
filtration.
2. Specific: Routine monitoring may not provide informa-
tion about all parameters of interest. For example, it
may be valuable to conduct special studies to measure
contaminants suspected of being present (Giardia. pesti-
cides, fuel products, enteric viruses, etc.). Frequent
presence of either Giardia or enteric viruses in raw
water samples prior to disinfection would indicate an
inadequate watershed control program. Monitoring may
also be useful to assess the effectiveness of specific
control techniques, and to audit procedures or opera-
tional requirements instituted within the watershed.
Utilities are encouraged to conduct additional monitor-
ing as necessary to aid them in controlling the quality
of the source water.
E. Management/Operations
1. Management
., a. Organizational structure
b. Personnel and education/certification requirements
2. Operations
a. Describe system operations and design flexibi 1 ity.
b. The utility should conduct some form of ongoing
review or survey in the watershed to identify and
react to potential impacts on water quality. The
scope of this review should be documented and
agreed upon by the utility and Primacy Agency on
a case-by-case basis.
c. Specifically describe operational changes which
can be made to adjust for changes in water quali-
ty. Example: Switching to alternate sources;
increasing the level of disinfection; using
0-4
-------
settling basins. Discuss what triggers, and who
decides to make, those changes.
3. Annual Report: As part of the watershed program, an
annual report should be submitted to the Primacy Agency.
The contents of the report should:
a. Identify special concerns that occurred in the
watershed and how they were handled (example:
herbicide usage, new construction, etc.).
b. Summarize other activities in the watershed such
as logging, hunting, water quality monitoring,
etc.
c. Project what adverse activities are expected to
occur in the future and describe how the utility
expects to address them.
F. Agreements/Land Ownership
The goal of a watershed management program is to achieve the
highest level of raw water quality practicable. This is
particularly critical to an unfiltered surface supply.
1. The utility will have maximum opportunity to realize
this goal if they have complete ownership of the
watershed. Describe efforts to obtain ownership, such
as any special programs or budget. When complete
ownership of the watershed is not practical, efforts
should be taken to gain ownership of critical elements,
such as, reservoir or stream shoreline, highly erodable
land, and access areas to water system facilities.
2. Where ownership of land is not possible, written
agreements should be obtained recognizing the watershed
as part of a public water supply. Maximum flexibility
should be given to the utility to control land uses
which could have adverse effect on the water,quality.
Describe such agreements.
3. Describe how the utility ensures that the landowner
complies with these agreements.
J-5
-------
APPENDIX K
SANITARY SURVEY
-------
APPENDIX K
SANITARY SURVEY
The SWTR requires that an on-site inspection be conducted each year
as outlined in Section 3. It is recommended that at the onset of
determining the classification of a source water that a detailed sanitary
survey be conducted. In addition, it is recommended that a sanitary
survey such as contained in this appendix be conducted every 3 to 5 years
by both filtered and unfiltered systems to ensure that the quality of the
water and service is, maintained. This time period is suggested since the
time and effort needed to conduct the comprehensive survey makes it
impractical for it to be conducted annually. A periodic sanitary survey
is also required under the Total Coliform Rule for systems collecting
fewer than 5 samples/month. The survey must be conducted every 5 years
for all systems except for protected ground water systems which disinfect.
These systems must conduct the survey every 10 years.
The sanitary survey involves three phases, including planning the
survey, conducting the survey and compiling the final report of the
survey, as will be presented in the following pages.
1. Planning the Survey
Prior to conducting or scheduling a sanitary survey, there
should be a detailed review of the water system's file to
prepare for the survey. The review should pay particular
attention to past sanitary survey reports and correspondence
describing previously identified problems and their solutions.
These should be noted, and action/inaction regarding these
problems should be specifically verified in the field. Other
information to review includes: any other correspondence,
water system plans, chemical and microbiological sampling
results, operating reports, and engineering studies. This
review will aid in the familiarization with the system's past
history and present conditions, and the agency's past interac-
tions with the system.
The initial phase of the water quality review will be carried
out prior to conducting the survey as well, and will consist
of reviewing the water system's monitoring records. Records
should be reviewed for compliance with all applicable microbi-
ological, inorganic chemical, organic chemical, and radiologi-
cal contaminant MCLs, and also for compliance with the
monitoring requirements for those contaminants. The survey
K-l
-------
will provide an opportunity to review these records with the
utility, and to discuss solutions to any MCL or monitoring
violations. The survey will also provide an opportunity to
review how and where samples are collected, and how field
measurements (turbidity, chlorine residual, fluoride, etc.)
are made. Points to cover include:
a. Is the system in compliance with all applicable MCLs
(organic chemical, inorganic chemical, microbiological,
and radiological)?
b. Is the system in compliance with all monitoring require-
ments?
The pre-survey file review should generate a list of items to
check in the field, and a list of questions about the system.
It will also help to plan the format of the survey and to
estimate how much time it may take. The next step is to make
the initial contact with the system management to establish
the survey date(s) and time. Any records, files, or people
that will be referenced during the survey should be mentioned
at the outset. Clearly laying out the intent of the survey up
front will greatly help in managing the system, and will
ensure that the survey goes smoothly without a need for repeat
trips.
2. Conducting the Survey
The on-site portion of the survey is the most important and
will involve interviewing those in charge of managing the
water system as well as the operators and other technical
people. The survey will also review all major system compo-
nents from the source(s) to the distribution system. A
standard form is frequently used to ensure that all major
components and aspects of each system are consistently
reviewed. However, when in the field, it is best to have an
open-jnind and focus most attention on the specifics of the
water system, using the form only as a guide. The surveyor
should be certain to be on time when beginning the survey.
This consideration will help get the survey started smoothly
with the operator and/or manager.
As the survey progresses, any deficiencies that are observed
should be brought to the attention of the water system
personnel, and the problem and the corrective measures should
be discussed. It is far better to clarify technical details
and solutions while standing next to the problem than it is to
do so over the telephone. Points to cover include:
a. Is the operator competent in performing the necessary
field testing for operational control?
K-2
-------
b. Are testing facilities and equipment adequate, and do
reagents used have an unexpired shelf life?
c. Are field and other analytical instruments properly and
regularly calibrated?
d. Are records of field test results and water quality
compliance monitoring results being maintained?
e. Conduct any sampling which will be part of the survey.
Also, detailed notes of the findings and conversations should
be taken so that the report of the survey will be an accurate
reconstruction of the survey.
Specific components/features of the system to review and some
pertinent questions to ask are:
Source Evaluation
All of the elements for a source evaluation enumerated below
may also be part of a Wellhead Protection Program.
1. Description: based on field observations and
discussion with the operator, a general charac-
terization of the watershed should be made.
Features which could be included in the descrip-
tion are:
a. Area of watershed or recharge area.
b. Stream flow.
c. Land usage (wilderness, farmland, rural
housing, recreational, commercial, indus-
trial, etc.).
d. Degree of access by the public to watershed.
e. Terrain and soil type.
f. Vegetation.
g. Other.
2. Sources of contamination in the watershed or
sensitive areas surrounding wells or well fields
should be identified. Not only should this be
determined by physically touring and observing
the watershed and its daily uses, but the survey-
or should also actively question the water system
manager about adverse and potentially adverse
K-3
-------
activities in the watershed. An example of types
of contamination includes:
a. Man Made.
1. Point discharges of sewage, storm-
water, and other wastewater.
2. On-site sewage disposal systems.
3. Recreational activities (swimming,
boating, fishing, etc.).
4. Human habitation.
5. Pesticide usage.
6. Logging.
7. Highways or other roads from which
there might be spills.
8. Commercial or industrial activity.
9. Solid waste or other disposal facili-
ties.
10. Barnyards, feed lots, turkey and
chicken farms and other concentrated
domestic animal activity.
11. Agricultural activities such as graz-
ing, tillage, etc., which affects
soil erosion, fertilizer usage, etc.
12. Other.
b. Naturally Occurring.
1. Animal populations, both domestic and
wild.
2. Turbidity fluctuations (from precipi-
tation, landslides, etc.).
3. Fires.
4. Inorganic contaminants from parent
materials (e.g., asbestos fibers).
5. Algae blooms.
K-4
-------
6. Other.
This list is by no means all inclusive.
The surveyor should rely principally on his
observations and thorough questioning
regarding the unique properties of each
watershed to completely describe what may
contaminate the source water.
3. Source Construction.
a. Surface Intakes.
1. Is the source adequate in quantity?
2. Is the best quality source or loca-
tion in that source being used?
3. Is the intake protected from icing
problems if appropriate?
4. Is the intake screened to prevent
entry of debris, and are screens
maintained?
5. Is animal activity controlled within
the immediate vicinity of the intake?
6. Is there a raw water sampling tap?
b. Infiltration Galleries.
1. Is the source adequate in quantity?
2. Is the best quality source being used?
3. Is the lid over the gallery water-
tight and locked?
4. Is the collector in sound condition
and maintained as necessary?
5. Is there a raw water sampling tap?
c. Springs.
1. Is the source adequate in quantity?
2. Is there adequate protection around
the spring such as fencing to control
the area within 200 feet?
K-5
-------
3. Is the spring constructed to best
capture the spring flow and exclude
surface water infiltration?
4. Are there drains to divert surface
water from the vicinity of the
spring?
5. Is the collection structure of sound
construction with no leaks or cracks?
6. Is there a screened overflow and
drain pipe?
7. Is the supply intake located above
the floor and screened?
8. Is there a raw water sampling tap?
Catchment and Cistern.
1. Is source adequate in quantity?
2. Is the cistern of adequate size?
3. Is the catchment area protected from
potential contamination?
4. Is the catchment drain properly
screened?
5. Is the catchment area and cistern of
sound construction and in good condi-
tion?
6. Is catchment constructed of approved
non-toxic, non-leaching material?
7. Is the cistern protected from contam-
ination -- manholes, vents, etc?
8. Is there a raw water tap?
Other Surface Sources.
1. Is the source adequate in quantity?
2. Is the best possible source being
used?
3. Is the immediate vicinity of the
source protected from contamination?
K-6
-------
4. Is the structure in good condition
and properly constructed?
5. Is there a raw water sampling tap?
4. Pumps, Pumphouses, and Controls.
a. Are all intake pumps, booster pumps, and
other pumps of sufficient capacity?
b. Are all pumps and controls operational and
maintained properly?
c. Are check valves, blow off valves, water
meters and other appurtenances operated and
maintained properly?
d. Is emergency power backup with automatic
start-up provided and does it work (try
it)?
e. Are underground compartments and suction
wells waterproof?
f. Is the interior and exterior of the pump-
house in good structural condition and
properly maintained?
g. Are there any safety hazards (electrical or
mechanical) in the pumphouse?
h. Is the pumphouse locked and otherwise
protected against vandalism?
i. Are water production records maintained at
the pumphouse?
5. Watershed Management (controlling contaminant
sources): The goal of the watershed management
program is to identify and control contaminant
sources in the watershed (see Section 3.3.1 of
this document, "Watershed Control Program").
Under ideal conditions each source of contamina-
tion identified in 2 will already have been
identified by the utility, and some means of
control instituted, or a factual determination
made that its impact on water quality is insig-
nificant. To assess the degree to which the
watershed management program is achieving its
goal, the following types of inquiries could be
made:
K-7
-------
a. If the watershed is not entirely owned by
the utility, have written agreements been
made with other land owners to control land
usage to the satisfaction of the utility?
Are appropriate regulations under the
contract of state/local department of
health in effect?
b. Is the utility making efforts to obtain as
complete ownership of the watershed as
possible? Is effort directed to control
critical elements?
c. Are there means by which the watershed is
regularly inspected for new sources of
contamination or trespassers where access
is limited?
d. Are there adequately qualified personnel
employed by the utility for identifying
watershed and water quality problems and
who are given the responsibility to correct
these problems?
e. Are raw water quality records kept to
assess trends and to assess the impact of
different activities and contaminant con-
trol techniques in the watershed?
f. Has the system responded adequately to
concerns expressed about the source or
watershed in past sanitary surveys?
g. Has the utility identified problems in its
yearly watershed control reports, and if
so, have these problems been adequately
addressed?
h. Identify what other agencies have control
or jurisdiction in the watershed. Does the
utility actively interact with these agen-
cies to see that their policies or activi-
ties are consistent with the utility's goal
of maintaining high raw water quality?
B. Treatment Evaluation
1. Disinfection.
a. Is the disinfection equipment and disinfec-
tant appropriate for the application
(chloramines, chlorine, ozone, and chlorine
K-8
-------
dioxide are generally accepted disinfec-
tants)?
b. Are there back-up disinfection units on
line in case of failure, and are they
operational?
c. Is there auxiliary power with automatic
start up in case of power outage? Is it
tested and operated on a regular basis,
both with and without load?
d. Is there an adequate quantity of disinfec-
tant on hand and is it properly stored
(e.g., are chlorine cylinders properly
labeled and chained)?
e. In the case of gaseous chlorine, is there
automatic switch over equipment when cylin-
ders expire?
f. Are critical spare parts on hand to repair
disinfection equipment?
g. Is disinfectant feed proportional to water
flow?
h. Are daily records kept of disinfectant
residual near the first customer from which
to calculate CTs?
i. Are production records kept from which to
determine CTs?
j. Are CTs acceptable based on the level of
treatment provided (see Surface Water
Treatment Rule for CT values, and
Sections 3 and 5 of this guidance manual
for calculation of CT).
k. Is a disinfectant residual maintained in
the distribution system, and are records
kept of daily measurements?
1. If gas chlorine is used, are adequate
safety precautions being followed (e.g.,
exhaust fan with intake within six inches
of the floor, self-contained breathing
apparatus that is regularly tested, regular
safety training for employees, ammonia
bottles and/or automatic chlorine detec-
tors)? Is the system adequate to ensure
K-9
-------
the safety of both the public and the
employees in the event of a chlorine leak?
2. Other.
a. Are other treatment processes appropriate
and are they operated to produce consis-
tently high water quality?
b. Are pumps, chemical feeders, and other
mechanical equipment in good condition and
properly maintained?
c. Are controls and instrumentation adequate
for the process, operational, well main-
tained and calibrated?
d. Are accurate records maintained (volume of
water treated, amount of chemical used,
etc.)?
e. Are adequate supplies of chemical on hand
and properly stored?
f. Are adequate safety devices available and
precautions observed?
Sections of a sanitary survey pertaining to
systems containing filtration facilities have
been omitted, as this section of the guidance
document pertains to non-filtering systems.
C. Distribution System Evaluation
After water has been treated, water quality must be
protected and maintained as it flows through the
^ distribution system to the customer's tap. The follow-
ing questions pertain to the water purveyor's ability to
maintain high water quality during storage and distribu-
tion.
1. Storage.
a. Gravity.
1. Are storage reservoirs covered and
otherwise constructed to prevent
contamination?
2. Are all overflow lines, vents, drain-
lines, or cleanout pipes turned down-
ward and screened?
K-10
-------
3. Are all reservoirs inspected regular-
ly?
4. Is the storage capacity adequate for
the system?
5. Does the reservoir (or reservoirs)
provide sufficient pressure through-
out the system?
6. Are surface coatings within the res-
ervoir in good repair and acceptable
for potable water contact?
7. Is the hatchcover for the tank water-
tight and locked?
8. Can the reservoir be isolated from
the system?
9. Is adequate safety equipment (caged
ladder, OSHA approved safety belts,
etc.) in place for climbing the tank?
10. Is the site fenced, locked, or other-
wise protected against vandalism?
11. Is the storage reservoir disinfected
after repairs are made?
12. Is there a scheduled program for
cleaning storage reservoir sediments,
slime on floor and side walls.
Hydropneumatic.
1, Is the storage capacity adequate for
the system?
2. Are instruments, controls, and equip-
ment adequate, operational, and main-
tained?
3. Are the interior and exterior surfac-
es of the pressure tank in good con-
dition?
4. Are tank supports structurally sound?
5. Does the low pressure cut in provide
adequate pressure throughout the
entire system?
K-ll
-------
6. Is the pump cycle rate acceptable
(not more than 15 cycles/hour)?
2. Cross Connections.
a. Is the system free of known uncontrolled
cross connections?
b. Does the utility have a cross connection
prevention program, including annual test-
ing of backflow prevention devices?
c. Are backflow prevention devices installed
at all appropriate locations (wastewater
; treatment plant, industrial locations,
hospitals, etc.)?
3. Other.
a. Are proper pressures and flows maintained
at all times of the year?
b. Do all construction materials meet AWWA or
equivalent standards?
c. Are all services metered and are meters
read?
d. Are plans for the system available and
current?
e. Does the system have an adequate mainte-
nance program?
Is there evidence of leakage in the
system?
Is there a pressure testing program?
Is there a regular flushing program?
Are valves and hydrants regularly
exercised and maintained?
Are AWWA standards for disinfection
followed after all repairs?
Are there specific bacteriological
criteria and limits prescribed for
new line acceptance or following line
repairs?
K-12
-------
Describe the corrosion control pro-
gram.
Is the system interconnected with
other systems?
D. Management/Operation
1. Is there an organization that is responsible for
providing the operation, maintenance, and manage-
ment of the water system?
2. Does the utility regularly summarize both current
and long-term problems identified in their water-
shed, or other parts of the system, and define
how they intend to solve the problems i.e., is
their planning mechanism effective; do they
follow through with plans?
3. Are customers charged user fees and are collec-
tions satisfactory?
4. Are there sufficient personnel to operate and
manage the system?
5. Are personnel (including management) adequately
trained, educated, and/or certified?
6. Are operation and maintenance manuals and manu-
facturers technical specifications readily avail-
able for the system?
7. Are routine preventative maintenance schedules
established and adhered to for all components of
the water system?
8. Are sufficient tools, supplies, and maintenance
parts on hand?
9. Are sufficient operation and maintenance records
kept and readily available?
10. Is an emergency plan available and usable, and
are employees aware of it?
11. Are all facilities free from safety defects?
When the survey is completed, it is always preferable to
briefly summarize the survey with the operator(s) and
management. The main findings of the survey should be
reviewed so it is clear that there are not misunder-
standings about findings/conclusions. It is also good
K-13
-------
to thank the utility for taking part in the survey,
arranging interviews with employees, gathering and
explaining their records, etc. The information and help
which the utility can provide an invaluable to a
successful survey, and every attempt should be made to
continue a positive relationship with the system.
3. Reporting the Survey
A final report of the survey should be completed as soon as
possible to formally notify the system and other agencies of
the findings. There is no set or necessarily best format for
doing so, and the length of the report will depend on the
findings of the survey and size of the system. Since the
report may be used for future compliance actions and inspec-
tions, it should include as a minimum: 1) the date of the
survey; 2) who was present during the survey; 3) the findings
of the survey; 4) the recommended improvements to identified
problems; and 5) the dates for completion of any improvements.
Any differences between the findings discussed at the conclu-
sion of the survey and what's included in the final report
should be discussed and clarified with the utility prior to
sending out the final report. In other words, the utility
should be fully aware of the contents of the final report
before receiving it.
K-14
-------
APPENDIX L
SMALL SYSTEM CONSIDERATIONS
-------
APPENDIX L
SMALL SYSTEM CONSIDERATIONS
Introduction
Under the provisions of the SWTR, systems with fewer than 500
service connections may be eligible for an exemption. Guidance on the
requirements for an exemption is provided in Section 9. For systems which
are not eligible for an exemption, compliance with the SWTR is mandatory.
It is recognized that the majority (approximately 75 percent) of people in
the United States are served by a relatively small number of large
systems. However, most water systems in the United States are small. For
small systems, compliance with the various provisions of the SDWA has
traditionally been a problem. Records show small systems have a
disproportionately higher incidence of drinking water quality and
monitoring difficulties. The reasons for these difficulties can generally
be broken down into the following three categories:
Economics
Treatment Technologies
Operations (lack of qualified personnel)
Small water systems typically face severe economic constraints.
Their lack of operating revenues results in significant limitations on
their ability to respond to the requirements of the SDWA. These systems
cannot benefit from the economies of scale which are available to larger
systems.
The second difficulty facing the small systems has been the lack of
appropriate treatment technologies. Although methods for removing most of
the contaminants known to occur in drinking water are available, many of
these technologies have only recently been scaled down for the smaller
systems.
The third problem which has traditionally plagued small systems is
the lack of well trained operators. This deficiency is the result of many
combined factors. First of all, many of these operators are employed only
on a part-time basis or if they are employed on a full-time basis they
have a myriad of additional duties. In addition, the operator's technical
L-l
-------
background may be limited as well. This results from the low salary of
the position, which is uninviting to qualified operators. Also, in spite
of the requirement of retaining certified operators upheld in many states,
it seems to be difficult to enforce this requirement in small systems.
The purpose of this appendix is to provide assistance to the Primacy
Agency in defining the problems and potential solutions typically
associated with small systems. It is beyond the scope of this document to
provide an indepth dicussion of the needs of small systems. However, over
the past several years the needs of the small water systems have been
recognized to be of primary concern and numerous workshops, seminars and
committees have been attempting to more clearly define workable solutions.
A partial listing of the papers, reports and proceedings which discuss
problems and solutions pertaining to small systems beyond that which is
possible in this manual is presented in the reference list of this
appendix.
Economics
One of the most severe constraints of small systems is the small
economic base from which to draw funds. Certain treatment and services
must be provided for a community regardless of how few people are served.
Thus, as the number of connections to the system decrease, the cost per
connection increases. The economic limitations of small utilities makes
it difficult to_ provide needed upgrading of existing facilities or an
adequate salary to maintain the employment of a qualified operator to
monitor and maintain the system. Adding to the severity of the economic
hardships of small systems is the fact that many of the small water
systems are privately owned, with private ownership increasing as system
size decreases. The ownership of the plant presents difficulties since
privately owned systems are subject to rate controls by the local public
utility commission, are not eligible for public grants and loans, and may
find commercial loans hard to obtain.
Financing options for small systems include; federal and state loan
and grant programs, federal revenue sharing and revenue bonds (for
municipal systems) and loans through the United States Small Business
Administration (SBA) and use of industrial development bonds or privatiza-
L-2
-------
tion (for private utilities). These options are explained in greater
detail in the "Guidance Manual - Institutional Alternatives for Small
Water Systems" (AWWA, 1986). The following paragraphs will explain some
existing options which may ease the hardship of financing small water
treatment facilities.
The major cause of small system difficulties arises from the lack of
funds and resources. It is therefore in the best interest of small
utilities to expand their economic base and the resources available to
them, to achieve the economies of scale available to larger systems.
Regionalization is the physical or operational union of small systems to
effect this goal. This union can be accomplished through the physical
interconnection of two or more small systems or the connection of a
smaller system to a pre-existing larger system. Water supply systems can
also join together for the purchase of supplies, materials, engineering
services, billing and maintenance. The union of the small systems
increases the population served, thereby dispersing the operational costs
and decreasing the cost per consumer.
The creation of utility satellites is another form of regionaliza-
tion. A satellite utility is one which taps into the resources of an
existing larger facility without being physically connected to, or owned
by, the larger facility. The larger system may provide any of the
following for the smaller system:
1. Varying levels of technical operational, or managerial
assistance on a contract basis.
2. Wholesale treated water with or without additional services.
3. Assuming ownership, operation and maintenance responsibility
when the small system is physically separate with a separate
source.
The formation of a satellite offers many advantages for both the
satellite and the parent utility. These advantages include: an improved
economy of scale for satellites, an expanded revenue base for the parent
utility, provisions of needed resources to satellites, the retention of
the satellites' local autonomy, improved water quality management of the
satellite, improved use of public funds for publicly owned satellites.
L-3
-------
In order to create a more definite structure for the union of
resources of water treatment facilities, water districts may be created.
Water districts are formed by county officials and provide for the public
ownership of the utilities. The utilities in any given district would
combine resources and/or physically connect systems so that one or two
facilities would provide water for the entire district. The creation of
water districts creates eligibility for public monies, has the potential
for economies of size, facilitates the takeover or contract services with
publicly owned non-community systems and small privately owned systems,
and offers a tax advantage. Drawbacks include subjection to politics, a
strong local planning effort is needed for success, and competition with
private enterprises.
The centralization of utilities can be taken one step further
through the creation of county utilities or even state utilities. The
government will create a board which may then act to acquire, construct,
maintain and operate any public water supply within its district, the
system may provide water on its own or purchase water from any municipal
corporation. The board may adopt and administer rules for the construc-
tion, maintenance, protection and use of public water supplies and the
fixation of reasonable rates for water supplies. The cost of construction
and/or upgrading of facilities may be defrayed through the issuance of
bonds and/or property assessment. As with all the alternatives, the
creation of government control of the utilities has its advantages and
disadvantages. The advantages include: the creation of central
management, creation of economy of scale for utilities, eligibility for
public grants and loans, savings through centralized purchasing, manage-
ment, consultation, planning and technical assistance, and possible
provision for pool of trained operators. The disadvantages include the
subjectivity to politics, the slow response caused by bureaucracy, and
competition to private contractors.
Treatment Technologies
The high cost of available treatment technologies has limited their
use in small water supply systems. Recently prefabricated package plants
and individual treatment units have been developed to lessen these costs.
L-4
-------
At the present time, the treatment technologies which are available to
enable systems to comply with the Safe Drinking Water Act are identified
to be the following:
Package plants
Slow-sand filters
Diatomaceous earth filters
Cartridge filtration
A brief discussion of each treatment method is provided below.
Package Plants
Clarification and filtration units which require minimal assembly in
the field can now be manufactured. To minimize required operator skill
level and operational attention, the equipment should be automated.
Continuous effluent turbidity and disinfectant residual monitoring systems
with alarms and emergency shutdown provisions are features that safeguard
water quality and should be provided for unattended plants.
Slow-Sand Filters
Slow-sand filters are applicable to small' water supply systems.
Their proven record of effective removal of turbidity and Giardia cysts
makes them suitable for application where operational attention is
minimal. Since no chemicals other than a disinfectant are needed, and no
mechanical equipment is involved, the required operator skill level is the
lowest of the filtration alternatives available to small systems.
Diatomaceous Earth Filters
"*•>•
Diatomaceous earth (DE) pressure and vacuum filters can be used on
relatively low turbidity surface waters (less than 1 to 2 NTU) for removal
of turbidity and Giardia cysts. DE filters can effectively remove
particles as small as 1 micron, but would require coagulating chemicals
and special filter aids to provide significant virus removal.
Cartridge Filters
Cartridge filters using microporous ceramic filter elements with
pore sizes as small as 0.2 urn may be suitable for producing potable water,
in combination with disinfection, from raw water supplies containing
moderate levels of turbidity, algae, protozoa and bacteria. The advantage
to a small system, is, with the exception of chlorination, that no other
chemicals are required. The process is one of strictly physical removal
L-5
-------
of small particles by straining as the water passes through the porous
membranes. Other than occasional cleaning or membrane replacement,
operational requirements are not complex and do not require skilled
personnel.
Selection of a Filtration Technology
The criteria for selection of a filtration technology for a small
community are essentially the same as those for a larger community. That
is, the utility must first screen the complete list of available
alternatives to eliminate those which are either not technically suited to
the existing conditions (Table 4-1) or not affordable by the utility.
Remaining alternatives should then be evaluated based on both cost
(capital, annual, and life-cycle) and non-cost bases (operation and
maintenance, technical requirements versus personnel available; flexibili-
ty regarding future needs; etc.). In these evaluations it should be noted
that even though automated package plants are cost-competitive with slow
sand filters, their operation requirements to achieve optimum performance
could be complicated. Also, the maintenance requirements for package
plants would be mechanically and electrically oriented and might require
a maintenance agreement with the manufacturer.
During the process of installing the treatment system, interim
measures should be taken to ensure the delivery of a reasonably safe water
to the consumers. In addition to the available interim measures listed in
Section 9.3, temporary installation of mobile filtration plants may be
possible. These trailer-mounted units are sometimes available from state
agencies for emergencies, but more often may be rented or leased from an
equipment manufacturer.
Modification of Existing Filtration Systems
Small treatment systems that are already in existence should comply
with the performance criteria of the SWTR. If the systems are not found
to be performing satisfactorily, modifications to the existing process may
be required. Improvement in treatment efficiency depends on the type of
filtration system in use. Operation of slow sand filters could be checked
for bed depth, short-circuiting, excessive hydraulic loading, and for the
need to pretreat the raw water. Infiltration galleries, or sometimes,
roughing filters ahead of a slow sand filter may provide for better
1-6
-------
performance by reducing the solids load on the filters. However, the
design criteria and costs for this alternative have not yet been defined.
Site specific studies may be required before roughing filters could be
used to achieve compliance with the regula- tions. Diatomaceous earth
(DE) filters should be checked for appropriate precoat and body feed
application, hydraulic loading, grade (size) of DE being used, and
possible need for chemical pretreatment. Package plants would have to be
checked process-by-process, similar to the system used for a conventional
plant. Other filtration processes would have to be checked for hydraulic
loading rate, appropriateness of the filter material (pore size), and
possible need for additional pretreatment.
Disinfection
Disinfection (CT) requirements for small systems can be met in
several different ways. The most obvious method of maintaining a
disinfectant residual in the distribution system is to add disinfectant at
one or more additional locations. An alternate method is to increase the
disinfectant dose at the existing application point(s). The latter
alternative, however, may increase disinfectant byproducts, including
THMs, in the system.
If it is a relatively short distance between the treatment system
and the first customer, additional contact time can be provided so that
the disinfectant dose does not have to be increased beyond desirable
residuals. Two specific methods of increasing contact time for small
systems are 1) installing a pressure vessel or closed storage vessel,
baffled to provide adequate contact time, or 2) constructing a looped
pipeline, on the finished water line between the filtration-disinfection
system and the first customer. The feasibility of either of these methods
would depend on system specifics that include size, physical conditions,
and cost.
If it is not practical to provide additional storage time to achieve
the desired CT, an alternate, more effective disinfectant may be used. An
alternate disinfectant may provide a sufficient CT without altering the
system configuration.
L-7
-------
Operations
Water treatment facilities need to be operated properly in order to
achieve maximum treatment efficiencies. There is currently a lack of well
trained operators at many small treatment plants. The main cause is lack
of awareness of the importance of correct plant operation, lack of
training programs, lack of enforcement of the requirement for employment
of a certified operator and lack of funds to employ such an operator.
Small systems may wish to implement a circuit rider/operator
program. In this program a qualified, certified, experienced operator
works for several water supply systems. The rider can either directly
operate the plants, or provide technical assistance to individual plant
operators, by acting as a trainer through on-the-job supervision. The
latter would be preferable since it could create a pool of well trained
operators.
The main cause of inadequately trained operators is the lack of well
established training programs. Until such training programs are begun,
systems must depend on other training means, such as seminars and books.
One resource which may be helpful in running the plant is "Basic
Management Principles for Small Water Systems - An AWWA Small-Systems
Resource Book", 1982.
Most package plant manufacturers' equipment manuals include at least
brief sections on operating principles, methods for establishing proper
chemical dosages, instructions for operating the equipment, and trouble-
shooting guides. An individual who studies these basic instructions and
receives comprehensive start-up training should be able to operate the
equipment satisfactorily. These services are vital to the successful
performance of a package water treatment plant and should be a requirement
of the package plant manufacturer. The engineer designing a package plant
facility should specify that start-up and training services be provided by
the manufacturer, and also should consider requiring the manufacturer to
visit the plant at 6-month and 1-year intervals after start-up to adjust
the equipment, review operations, and retrain operating personnel.
Further, this program should be ongoing and funds should be budgeted every
year for at least one revisit by the package plant manufacturer.
L-8
-------
Another way for small systems to obtain qualified plant operation
*ould be to contract the services of administrative, operations, and/or
maintenance personnel from a larger neighboring utility, government
agencies, service companies or consulting firms. These organizations
could supply assistance in financial and legal planning, engineering,
purchasing accounting and collection services, laboratory support,
licensed operators or operator training, treatment and water quality
assurance, regulatory liaison, and/or emergency assistance. Through the
contracting of these services the utility provides for the resources
needed, improves water quality management and retains its autonomy.
However, if and when the contract is terminated, the utility returns to
its original status.
References
American Water Works Association. Basic Management Principles for Small
Water Systems. 1982.
American Water Works Association. Design and Construction of Small Water
Systems. 1984.
Kelly, Gidley, Blair and Wolfe, Inc. Guidance Manual - Institutional
Alternatives for Small Water Systems. AWWA Research Foundation Contract
79-84, 1986.
L-9
-------
APPENDIX M
PROTOCOL FOR DEMONSTRATION
OF EFFECTIVE TREATMENT
-------
APPENDIX M
PROTOCOL FOR DEMONSTRATION
OF EFFECTIVE TREATMENT
This appendix presents approaches which can be taken to demonstrate
overall effective removal and/or inactivation of Giardia cysts.
M.I Demonstration for Alternate Technology
Systems using a filtration technology other than those enumerated
in the SWTR may demonstrate the effectiveness of the treatment process
through pilot or full scale testing. As a minimum, testing should be
conducted when the source exhibits its worst case annual conditions. Some
systems may have two periods of "worst case" water quality including the
cold water in winter or algae blooms during the summer.
Pilot units should include the following:
filtration rate of the pilot system equal to filtration
rate on full scale unit
pilot filter diameter greater than or equal to 50 times
the media diameter, (Robeck, et al 1959)
media diameter, depth, and size gradation should be
identical to full scale,
coagulant dosing identical to full scale
any mixing and settling occurring before filtration in
the full scale plant should be reproduced as closely as
possible in the pilot. Mixing should be of the same G
value(s), and the detention time for settling should be
close to the average flow detention time for the
projected full scale plant.
According to the SWTR, alternate technologies must be capable of
meeting the same turbidity performance criteria of slow sand filtration
systems. Thus the filtered water from the process should be monitored
continuously or with grab samples every four hours for turbidity. The
requirement for meeting turbidity performance has been established to
M-l
-------
ensure that there will be no interference of turbidity with virus
inactivation through disinfection.
Following the demonstration of meeting the turbidity requirements,
the level of GJardia cyst removal achieved must be determined. The
protocol in M.2 may be followed for this demonstration.
M.2 Particle Size Analysis Demonstration for Giardia Cyst Removal Credit
Particle size analysis may be used to demonstrate the level of
actual Giardia cyst removal provided by the system. This demonstration
can be done using samples from the full scale plant or a pilot unit.
In the case of either a full scale or pilot scale demonstration,
removal of particles in the range of 5 to 15 urn in diameter should be
determined using an electronic particle counter that has been calibrated
with latex spheres. If a light blockage device is used (e.g. HIAC) this
calibration should have been done during installation of the device. The
calibration should be checked before taking measurements for the purposes
of this demonstration. Samples should be diluted appropriately to ensure
that measurements do not reflect coincident error. Coincident error
results when more than one particle passes the detector at one time,
causing an inaccurate particle count and diameter measurement. An
electrical sensing zone device (e.g. Coulter Counter or Elzone) may also
be used. Appropriate dilutions, electrolyte strength, and calibration
procedures should be followed (these are scheduled to be outlined in the
17th edition of Standard Methods). When using an electrical sensing zone
instrument, an orifice no larger than 125 urn and no smaller than 40 urn
should be used since only particles between 2% and 40% of the orifice
diameter are accurately sized and counted (Karuhn et al 1975).
Samples of the filter influent and effluent should be taken 5
minutes after the backwashed filter is placed in operation, and every 30
minutes thereafter for the first 3 hours of operation, followed by hourly
samples up until backwash (Wiesner et al 1987). All samples should show
at least a 2-log removal. The SWTR establishes an overall treatment
requirement of 3-log Giardia cyst removal/inactivation. Thus, disinfec-
M-2
-------
tion must be provided to supplement the particulate removal and meet this
requirement.
Samples from repeated filter runs may be averaged at each sampling
time, but samples should not be averaged within one filter run.
Additional suggestions on particle counting technique (Wiesner
1985):
1) If particle counts are not determined immediately upon
sampling (within 10 minutes) samples should be diluted.
2) For an electrical sensing zone measurement, samples
should be diluted 1:5 to 1:20 with a "particle-free1
electrolyte solution (approximately 1% NaCl) containing
100 particles per ml or fewer.
3) For a light blockage measurement, particle free w-ater
should be used to dilute samples.
4) Dilutions should be done to produce particle concentra-
tions as close to the tolerance for coincident error as
possible to minimize background counts.
5) Particle counts should be determined within 8 hours of
sampling.
6) All sampling vessels should be washed with laboratory
detergent, double rinsed in particle free water, and
rinsed twice with the water being sampled at the time
of sampling.
The log reduction of particles in the size range of 5 to 15 urn in
size can be assumed to correspond to the log reduction of Giardia cysts
which would be achieved.
M.3 Demonstration for Increased Turbidity Allowance
Based upon the requirements of the SWTR, the minimum turbidity
performance criteria for systems using conventional treatment or direct
filtration is filtered water turbidity less than or equal to 0.5 NTU in
95 percent of the measurements taken each month. However, at the dis-
cretion of the Primacy Agency, filtered water turbidity levels of less
than or equal to 1 NTU in 95 percent of the measurements taken every month
M-3
-------
may be permitted on a case-by-case basis depending on the capability of
the total system to remove and/or inactivate at least 99.9 percent of
Giardia 1amb1i$ cysts.
Treatment plants that use settling followed by filtration, or direct
filtration are generally capable of producing a filtered water with a
turbidity of 0.2 NTU or less. The most likely cause of high turbidities
in the filtered water is incorrect coagulant dosing (O'Melia, 1974).
Regardless of the turbidity of the raw or finished water, coagulant
addition at some point prior to filtration is required to destabilize
particles for removal in the filter. Only plants documenting continuous
coagulant feed prior to filtration should be eligible for being allowed
higher filtered water turbidities than the 0.5 NTU requirement. At plants
that continuously feed coagulant and do not meet the 0.5 NTU requirement,
a series of jar tests, and perhaps sand column filtration tests (in batch)
should be performed to evaluate the optimum coagulant dose for turbidity
removal.
In the event that plants can document continuous coagulant feed,
and, after running the plant under conditions determined in batch testing
to be optimal for turbidity removal, still do not meet the 0.5 NTU
requirement, effective filtration status may still be appropriate. This
would further be supported if it can be shown that the full scale plant
is capable of achieving at least a 2-log reduction in the concentration
of particles between 5 and 15 urn in size through particle size analysis
as outlined in Section M.2. Where a full scale plant does not yet exist,
appropriately ~~ scaled-down pilot filters might be used for such a
demonstration.
Disinfection
The level of disinfection could also be considered for determining
when to allow a higher turbidity performance criterion for a system. For
example, if a system achieves 3-log Giardia cyst inactivation through
disinfection, as determined by CT values, it may be appropriate to allow
higher filtered water turbidities (i.e., greater than 0.5 NTU but less
than 1 NTU in 95 percent of the measurements and never exceeding 5 NTU).
M-4
-------
The expected level of fecal contamination and Giardia cyst
concentrations in the source water should be considered in the above
analysis. High levels of disinfection (e.g., 2 to 3-log inactivation of
Giardia cysts), in addition to filtration which achieves less than 0.5 NTU
in 95 percent of the measurements may be appropriate, depending upon
source water quality. Further guidance on the level of disinfection to be
provided for various source water conditions is provided in Section 4.4.2.
In all cases the minimum disinfection to be provided must supplement the
particulate removal to ensure at least a 3-log Giardia cyst remov-
al/inactivation.
References
American Public Health Association; American Water Works Association;
Water Pollution Control Federation. Standard Methods for the Examination
of Water and Wastewater. 17th ed. (supplement), September 1989.
Coulter Electronics 600 W. 20th Street, Hialeah, FL 33010-2428
Karuhn, R.; Davies, R.; Kaye, B.; Clinch, M. Studies on the Coulter
Counter Part I. Powder Technology Volume II, pp. 157-171, 1975.
O'Melia, C. The Role of Polvelectrolytes in Filtration Processes. EPA -
67012-74-032, 1974.
Robeck, G.; Woodword, R. L. Pilot Plants for Water Treatment Research,
Journal of Sanitary Engineering ASCE Vol. 85;SA4; 1, August 1959.
Wiesner, M. "Optimum Water Treatment Plant Configuration Effects of Raw
Water Characteristics," dissertation John Hopkins University, Baltimore,
MD, 1985.
Wiesner, M.; Rook, J. J.; Fiessinger, F. Optimizing the Placement of GAC
Filters, J. AWWA VOL 79, pp. 39-49, Dec 1987.
M-5
-------
APPENDIX N
PROTOCOLS FOR POINT-OF-USE
DEVICES
-------
ITITED STATES
ENVIRONMENTAL PROTECTION AGENCY
Registration Division
Office of Pesticide Programs
Criteria and Standards Division
Office of Drinking Water
GUIDE STANDARD AND PROTOCOL TOR
rESTING MICROS 1C LOGICAL WATER PL'RIFIERS
Report of Task Force
Submitted April, 1986
Revised April, 1987
-------
CONTENTS
Page
PREFACE
1. GENERAL r.-l
2. PERFORMANCE REQUIREMENTS N-6
3. MICROBIOLOGICAL WATER PURIFIER TEST PROCEDURES N-8
APPENDIX N-l SUMMARY FOP. BASIS OF STANDARDS AND N-21
TEST WATER PARAMETERS
APPENDIX N-2 LIST OF PARTICIPANTS IN TASK FORCE N-29
APPENDIX N-3 RESPONSE BY REVIEW SUBCOMMITTEE TO N-21
PUBLIC COMMENTS
-------
Preface
The protocol presented in this paper can be applied to demonstrate the
effectiveness of new technologies as well as point-of-use devices. The
evaluation presented here deals with the removal of particulates and
disinfection. In areas which pertain to disinfection, the guidelines
contained in Appendix G take precedence.
-------
1. GENEPAL
1.1 Introduction
The subject of microbiological purification for waters of unknown micro-
biological quality repeatedly presents itself to a variety of governmental =r.d
non-governmental agencies, consumer groups, manufacturers and others. Exam-
ples of possible application of such purification capabilities include:
- Backpackers and campers
Non-standard military requirements
Floods and other natural disasters
- Foreign travel and stations (however, not for extreme contamination
situations outside of the U.S.)
Contaminated individual sources, wells and springs (however, net for
the conversion of waste water to microbiologically potable water)
Motorhomes and trailers
Hatch methods of water purification based on chlorine and iodine disin-
fection or boiling are well known, but many situations and personal choice
call for the consideration of water treatment equipment. Federal agencies
specifically involved in responding to questions and problems relating to
microbiological purifier equipment include:
- Registration Division, Office of Pesticide Programs (OPP), Environ-
mental Protection Agency (EPA): registration of microbiological
purifiers (using chemicals);
- Compliance Monitoring Staff, EPA: control of microbiological
purifier device claims (non-registerable products such as ultra-
violet units, ozonators, chlorine generators, others);
- U.S. Army Medical Bioengineering Research and Development Laboratory
(USAMBRDL), U.S. Army Natick Research and Development Center and
other Army and military agencies: research and development for
possible field applications;
- Criteria and Standards Division, Office of Drinking water (CCW),
EPA: Consideration of point-of-use technology as acceptable tech-
nology under the Primary Drinking Water Regulations; consumer
information and service;
N-l
-------
Drinking Water Research, Water Engineering Research Laboratory
(WERL), EPA; responsible for water treatment technology research;
- Microbiology Branch, Health Effects Research Laboratory (HERD, EPA;
responsible for study of health effects related to drinking water
filters.
A number of representatives of the above mentioned agencies provided
excellent participation in the task force to develop microbiological testing
protocols for water purifiers. Major participation was also provided by the
following:
- A technical representative from the Water Quality Association;
- A technical representative from the Environmental Health Center,
Department of Health and Welfare of Canada; and
- An associate professor (microbiology) from the University of
Arizona.
1.2 Basic Principles
1.2.1 Definition
As set forth in EPA Enforcement Strategy and as supported by a Federal
Trade Commission (FTC) decision (FTC v. Sibco Products Co., Inc., et al.,
Nov. 22, 1965) , a unit, in order to be called a microbiological water
purifier, must remove, kill or inactivate all types of disease-causing micro-
organisms from the water, including bacteria, viruses and protozoan cysts so
as to render the processed water safe for drinking. Therefore, to qualify, a
microbiological water purifier must treat or remove all types of challenge
organisms to meet specified standards.
1.2.2 General Guide
The standard and protocol will be a general guide and, in some cases, may
present only the minimum features and framework for testing. While basic
features of the standard and protocol have been tested, it was not feasible to
conduct full-fledged testing for all possible types of units. Consequently,
protocol users should include pre-testing of their units in a testing rig,
including the sampling techniques to be used. Where users of the protocol
find good reason to alter or add to the guide in order to meet specific
operational problems, to use an alternate organism or laboratory procedure, or
to respond to innovative treatment units without decreasing the level of
N-2
-------
testing or altering the intent of the protocol, they should feel free to do
so. For example, the OPP Registration Division might find it necessary to
amend the guide somewhat for different types of treatment units. Another
example would be ultraviolet (U.V.) units, which may have specific require-
ments in addition to the guide protocol.
1.2.3 Performance-Based
The standard will be performance-based, utilizing realistic worst case
challenges and test conditions and use of the standard shall result in water
quality equivalent to that of a public water supply meeting the
microbiological requirements and intent of the National Primary Drinking Water
Regulations.
1.2.4 Exceptions
A microbiological water purifier must remove, kill or inactivate all
types of pathogenic organisms if claims are made for any organism. However,
an exception for limited claims may be allowed for units removing specific
organisms to serve a definable environmental need (i.e., cyst reduction units
which can be used on otherwise disinfected and microbiologically safe drir.kir.g
water, such as a disinfected but unfiltered surface water containing cysts.
Such units are not to be called microbiological water purifiers and should r.ct
be used as sole treatment for an untreated raw water.)
1-2-5 Not to Cover Non-Microbiological Reduction Claims
The treatment of water to achieve removal of a specific chemical or ether
non-microbiological- substances from water will not be a part of this standard.
National Sanitation Foundation (NSF) Standards 42 (Aesthetic Effects) and 53
(Health Effects) provide partial guides for chemical removal and other claims
testing.
1.2.6 Construction and Information Exclusions
while the standard recommends safe, responsible construction of units
with non-toxic materials for optimum operation, all such items and associated
operational considerations are excluded as being beyond the scope of the
standard. Included in the exclusion are materials of construction, electrical
and safety aspects, design and construction details, operational instructions
and information, and mechanical performance testing.
1.2.7 Research Needs Excluded
N-3
-------
The guide standard and protocol must represent a practical testing
program and not include research recommendations. For example, consideration
cf mutant organisms or differentiation between injured and dead organisms
would be research items at this time and not appropriate for inclusion in the
standard.
1.2.8 Not to Consider Sabotage
Esoteric problems which could be presented by a variety of hypothetical
terrorist (or wartime) situations, would provide an unnecessary complication,
and are not appropriate for inclusion in the standard.
1.2.9 Continuity
The guide standard and protocol will be a living document, subject to
revision and updating with the onset of new technology and knowledge. It is
recommended that the responsible authorities for registration and drinking
water quality review potential needs every two to three years and reconvene
the task force upon need or upon request from the water quality industry, to
review and update the standard and testing protocol.
1.3 Treatment Units Coverage
1.3.1 Universe of Possible Treatment Units
A review of treatment units that might be considered as microbiological
purifiers discloses a number of different types covering treatment principles
ranging from filtration and chemical disinfection to ultraviolet light ra-
diation.
1.3.2 Coverage of This Standard
In view of the limited technical data available and in order to expedite
the work of the task force, the initial coverage is limited, on a priority
basis, to three basic types of microbiological water purifiers or active
components with their principal means of action as follows:
1.3.2.1 Ceramic Filtration Candles or Units (may or
may not contain a chemical bacteriostatic agent)
Filtration, and adsorption, and chemical anti-microbial activity if a
chemical is included.
1.3.2.2 Halooenated Resins and Units
N-4
-------
Chemical disinfection and possibly filtration. (Note: While not
included in this guide standard, halogen products for disinfection or systems
using halogen addition and fine filtration may be tested using many of its
elements, i.e., test water parameters, microbiological challenge and reduction
requirements, analytical techniques and other pertinent elements.)
1.3.2.3 Ultraviolet (UV) Units
UV irradiation with possible add-on treatment for adsorption and filtra-
tion (not applicable to UV units for treating potable water from public water
supply systems).
1.3.3 Application of Principles to Other Units
while only three types of units are covered in this standard, the princi-
ples and approaches outlined should provide an initial guide for the testing
of any of a number of other types of units and/or systems for the microbiolog-
ical purification of contaminated water.
N-5
-------
-------
-------
2. PERFORMANCE REQUIREMENTS
2.1 Microbiological Water Purifier
In order to make the claim of "microbiological water purifier," units
must be tested and demonstrated to meet the microbiological reduction require-
ments of Table 1 according to the test procedures described in Section 3
(Appendix N-l) for the specific type of unit involved.
2.2 Chemical Health Limits
Where silver or some other pesticidal chemical is used in a unit, that
chemical concentration in the effluent water must meet any National Primary
Drinking Water Maximum Contaminant Level (MCL) , additional Federal guidelir.es
or otherwise be demonstrated not to constitute a threat to health from con-
sumption or contact where no MCI* exists.
2.3 Stability of Pesticidal Chemical
Where a pesticidal chemical is used in the treatment unit, the stability
of the chemical for disinfectant effectiveness should be sufficient for the
potential shelf life and the projected use life of the unit based on manufac-
turer's data. Where stability cannot be assured from historical data and
information, additional tests will be required.
2.4 Performance Limitations
2.4.1 Effective Lifetime
The manufacturer must provide an explicit indication or assurance of the
unit's effective use lifetime to warn the consumer of potential diminished
treatment capability either through:
a. Having the unit terminate discharge of treated water, or
b. Sounding an alarm, or
c. Providing simple, explicit instruction for servicing or replacing
units within the recommended use life (measurable in terms of volume
throughput, specific time frame or other appropriate method).
2.4.2 Limitation on Use of Iodine
N-6
-------
EPA policy initially developed in 1973 and reaffirmed in 1982 (memo of
March 3, 1982 from J. A. Cotruvo to G. A. Jones, subject: "Policy on Iodine
Disinfection") is that iodine disinfection is acceptable for short-term or
limited or emergency use but that it is not recommended for long-term or
routine community water supply application where iodine-containing species may
remain in the drinking water.
N-7
-------
3. MICROBIOLOGICAL WATER PURIFIER TEST PROCEDURES
3.1 Purpose
These tests are performed on ceramic filtration candles or units, halo-
genated resins and units and ultraviolet (UV) units in order to substantiate
their microbiological removal capabilities over the effective use life of the
purifier as defined in Table 1 and, where a pesticidal chemical is used, to
determine that said chemical is not present in the effluent at excessive
levels (see Section 3.5.3.4, Appendix N).
3.2 Apparatus
Three production units of a type are to be tested, simultaneously, if
feasible; otherwise, in a manner as similar to that as possible.
Design of the testing rig must parallel and simulate projected field use
conditions. For plumbed-in units a guide for design of the test rig may be
taken from "Figure 1: Test Apparatus-Schematic" (p. A-2 of Standard Musiber 53
"Drinking Water Treatment Units — Health Effects," National Sanitation
Foundation). Otherwise, the test rig must be designed to simulate field use
conditions (worst case) for the unit to be tested.
3.3 Test waters — Non-Microbiological Parameters
In addition to the microbiological influent challenges, the various test
waters will be constituted with chemical and physical characteristics as
follows:
3.3.1 Test Water »1 (General Test Water)
This water is intended for the normal non-stressed (non-challenge) phase
of testing for all units and shall have specific characteristics which nay
easily be obtained by the adjustment of many public system tap waters, as
follows:
a. It shall be free of any chlorine or other disinfectant residual;
b. pH — 6.5 - 8.5;
c. Total Organic Carbon (TOC) 0.1 - 5.0 mg/L;
d. Turbidity 0.1-5 NTU;
N-8
-------
e. Temperature 20 C * 5 C; and
f. Total Dissolved Solids (TDS) 50 - 500 mg/L.
3.3.2 Test Water *2 (Challenge Test Water/Halogen Disinfection)
This water is intended for the stressed challenge phase of testing where
units involve halogen disinfectants (halogen resins or other units) and shall
have the following specific characteristics:
a. Free of chlorine or other disinfectant residual;
b. (1) pH 9.0 * .2, and
(2) for iodine-based units a pH of 5.0 * .2 (current information
indicates that the low pH will be the most severe test for virus
reduction by iodine disinfection);
c. Total Organic Carbon (TOO not less than 10 mg/L;
d. Turbidity not less than 30 NTU;
e. Temperature 4 C " 1 C; and
f. Total Dissolved Solids (TDS) 1,500 mg/L * 150 mg/L.
3.3.3 Test Water *3 (Challenge Test Water/Ceramic Candle
or Units With or without Silver Impregnation)
This water is intended for the stressed challenge phase of testing for
the indicated units but not for such units when impregnated with a halogen
disinfectant (for"the latter, use Test Water #2). It shall have the following
specific characteristics:
a. It shall be free of any chlorine or other disinfectant residual;
b. pH 9.0 * .2;
c. Total Organic Carbon (TOO — not less than 10 mg/L;
d. Turbidity — not less than 30 NTU;
e. Temperature 4 C * 1 C; and
f. Total Dissolved Solids (TDS) — 1,500 mg/L * 150 mg/L.
3.3.4 Test Water »4 (Challenge Test Water for Ultraviolet Units)
This water is intended for the stressed phase of testing for UV units and
shall have the following specific characteristics:
N-9
-------
a. Free of chlorine or other disinfectant residual;
b. pH 6.5 - 8.5;
c. Total Organic Carbon (TOO — not less than 10 mg/L;
d. Turbidity — not less than 30 NTU;
e. Temperature 4 C * 1 C;
f. Total Dissolved Solids (TDS) ~ 1,500 mg/1 * 150 mg/L;
g. Color U.V. absorption (absorption at 254 run) — Sufficient para-
hydroxybenzoic acid (PHBH) to be just below the trigger point of the
warning alarm on the U.V. unit. (Note that Section 3.5.1.1 provides
an alternative of adjusting the U.V. lamp electronically, especially
when the U.V. lamp is preceded by activated carbon treatment.)
3.3.5 Test Water *5 (Leaching Test Water for Units Containing Silver)
This water is intended for stressed leaching tests of units containing
silver to assure that excess levels of silver will not be leached into the
drinking water. It shall have the following specific characteristics:
a. Free of chlorine or other disinfectant residual;
b. pH — 5.0 * 0.2;
c. Total Organic Carbon (TOO — approximately 1.0 mg/L;
d. Turbidity — 0.1 - 5 NTU;
i*..
e. Temperature —• 20 C " 5 C; and
f. Total Dissolved Solids (TDS) — 25 - 100 mg/L.
3.3.6 Recommended Materials for Adjusting Test Water Characteristics
a. pH: inorganic acids or bases (i.e., HC1, NaOH);
b. Total Organic Carbon (TOO: hmnic acids;
c, Turbidity: A.C. Fine Test Dust (Park No. 1543094)
from: A.C. Spark Plug Division
General Motors Corporation
1300 North Dort Highway
Flint, Michigan 48556;
d. Total Dissolved Solids (TDS): sea salts, Sigma Chemical Co., S9683
(St. Louis,MO) or another equivalent source of TDS;
N-10
-------
e. Color U.V. Absorption: p-hydroxybenzoic acid (grade: general
purpose reagent).
3.4 Analytical Methods
3.4.1 Microbiological Methods
Methods in this section are considered "state-of-the-art" at the time of
its preparation and subsequent improvements should be expected. Methods used
for microbiological analyses should be compatible with and equal to or better
than those given below. .
3.4.1.1 Bacterial Tests
a. Chosen Organism: Klebsiella terrigena (ATCC-33257).
b. Method of Production: Test organism will be prepared by overnight
growth in nutrient broth or equivalent to obtain the organism in the
stationary growth phase (Reference: Asburg, E.D., Methods of
Testing Sanitizers and Bacteriostatic Substances In: Disinfection,
Sterilization and Preservation, Seymour S. Block, ed., pp. 964-930,
1983) . The organism will be collected by centrifugation ar.d washed
three times in phosphate buffered saline before use. Alternatively,
the organisms may be grown overnight on nutrient agar slants or
equivalent and washed from the slants with phosphate buffered
saline. The suspensions should be filtered through sterile what.ran
Number 2 filter paper (or equivalent) to remove any bacterial
clumps. New batches of organisms must be prepared daily for use in
challenge testing.
c. State of Organism: Organisms in the stationary growth phase and
suspended in phosphate buffered saline will be used.
d. Assay Techniques: Assay may be by the spread plate, pour plate or
membrane filter technique on nutrient agar, M.F.C. or m-Endo medium
(Standard Methods for the Examination of Water and Wastewater, 16th
edition, 1985, APHA). Each sample dilution will be assayed in
triplicate.
3.4.1.2 Virus Tests
a. Chosen Organisms: Poliovirus type 1 (LSc) (ATCC-VR-59) , and Rota-
virus Strain SA-11 (ATCC-VR-899) or WA (ATCC-VR-2018).
b. Method of Production: All stocks should be grown by a method
described by Smith and Gerba (in Methods in Environmental Virology,
pp. 15-47, 1982) and purified by the procedure of Sharp, et al.
(Appl. Microbiol., 29:94-101, 1975), or similar procedure (Berman
and Hoff, Appl. Environ. Microbiol., 48:317-323, 1984), as these
methods will produce largely monodispersed virion particles.
N-ll
-------
c. State of the Organism: Preparation procedure will largely produce
monodispersed particles.
d. Assay Techniques: Poliovirus type 1 may be grown in the 3GM, MA-104
or other cell line which will support the growth of this virus. The
rotaviruses are best grown in the MA-104 cell line. Since both
viruses can be assayed on the MA-104 cell line, a challenge test may
consist of equal amounts of both viruses as a mixture (i.e., the
mixture must contain at least 1.0 x 10 /mL of each virus). Assays
may be as plaque forming units (PFU) or as immunofluorescence foci
(IF) (Smith and Gerba, In: Methods in Environmental Virology,
pp. 15-47, 1982). Each dilution will be assayed in triplicate.
3.4.1.3 Cyst Tests
a. Chosen Organism
1. Giardia lamblia or the related organism, Giardia muris, may be
used as the challenge cyst.
2. Where filtration.is involved, tests with 4-6 micron spheres or
particles have been found to be satisfactory and may be used as
a substitute for tests of occlusion using live organisms (see
Table 1). Spheres or particles may only be used to evaluate
filtration efficacy. Disinfection efficacy can only be evalu-
ated with the use of viable Giardia cysts.
b. Method of Production: Giardia muris may be produced in laboratory
mice and Giardia lamblia may be produced in Mongolian gerbiis;
ir.activation results based on excystation measurements correlate
well with animal infectivity results.
c. State gj£ the Organism: Organisms may be separated from fecal
material by the procedure described by Sauch (Appl, Environ.
Microbiol., 48:454-455, 1984) or by the procedure described by
Bingham, et al. (Exp. Parasitol., 47:284-281, 1979).
d. Assay Techniques: Cysts are first reconcentrated (500 ml., minimum
sample size) according to the method of Rice, Hoff and Schaefer
(Appl. Environ. Microbiol., 43:250-251, 1982). The excystation
method described by Schaefer, et al. (Trans., Royal Soc. of Trop.
Med. & Hyg. 78:795-800, 1984) shall be used to evaluate Giardia
muris cyst viability. For Giardia lamblia cysts, the excystation
method described by Bingham and Meyer (Nature, 277:301-302, 1979) or
Rice and Schaefer (J. Clin. Microbiol., 14:709-710, 1981) shall be
used. Cyst viability may also be determined by an assay method
involving the counting of trophozoites as well .as intact cysts
(Bingham, et al., Exp. Parasitol., 47:284-291, 1979).
3.4.2 Chemical and Physical Methods
N-12
-------
All physical and chemical analyses shall be conducted in accordance with
procedures in Standard Methods for the Examination of Water and Wastewater,
16th Edition, American Public Health Association, or equivalent.
3.5 Test Procedures
3.5.1 Procedure - Plumbed-in Units
a. 1. Install three production units of a type as shown in Figure 1
and condition each unit prior to the start of the test in
accordance with the manufacturer's instructions with the test
water without the addition of the test contaminant. Measure
the flow rate through each unit. The unit shall be tested at
the maximum system pressure of 60 psig static and flow rate
will not be artificially controlled.
2. Test waters shall have the defined characteristics continuously
except for test waters 2, 3 and 4 with respect to turbidity.
The background non-sampling turbidity level will be maintained
at 0.1-5 NTU but the turbidity shall be increased to the
challenge level of not less than 30 NTU in the following
manner:
In the "on" period(s) prior to the sampling "on" period.
In the sampling "on" period when the sample actually will
be taken. (Note: at least 10 unit void volumes of the 30
NTU water shall pass through the unit prior to actual
sampling so as to provide adequate seasoning and uni-
formity before sample collection.)
b. 1. Use appropriate techniques of dilution and insure continual
mixing to prepare a challenge solution containing the bacterial
contaminant. Then spike test water continuously with the
influent concentration specified in Table 1.
2. Use appropriate techniques to prepare concentrated virus and
Giardia suspensions. Feed these suspensions into the influent
stream so as to achieve the influent concentrations specified
in Table 1 in the following manner:
- In the "on" period(s) prior to the sampling "on" period.
- In the sampling "on" period when the sample actually will
be taken. (Note: at least 10 unit void volumes of seeded
water shall pass through the unit prior to sampling so as
to provide adequate seasoning and uniformity before sample
collection.)
N-13
-------
c. Purge the system of the uncontaminated water with a sufficient flow
of contaminated test water. Start an operating cycle of 10 percent
on, 90 percent off with a 15 to 40 minute cycle (Example: 3 minutes
on, 27 minutes off) with the contaminated test water. This cycle
shall be continued for not more than 16 hours per day (minimum daily
rest period of 8 hours). The total program shall extend to 100% of
estimated volume capacity for halogenated resins or units and for
10-1/2 days for ceramic candles or units and U.V. units.
d. Sampling: Samples of influent and effluent water at the specified
sampling points shall be collected as shown below for the various
units; these are minimum sampling plans which may be increased in
number by the investigator. All samples shall be collected in
duplicate from the flowing water during the sampling "on" portion of
the cycle and they shall be one "unit void volume" in quantity (or
of appropriate quantity for analysis) and represent worst case
challenge conditions. Effluent samples shall usually be collected
near the middle of the sampling "on" period (or the whole volume
during one "on" period) except for samples following the specified
"stagnation" periods, for which sampling shall be conducted on the
first water volume out of the unit. Each sample will be taken in
duplicate and shall be retained and appropriately preserved, if
required, for chemical or microbiological analysis in the event
verification is required. (For units where the volume of a single
"on" period is insufficient for the required analysis, samples frcn
successive "on" periods nay be accumulated until a sufficient volume
has been collected.)
l(a). Sampling Plan: Halogenated Resins or Units (Non-iodine Based)
Tests
Test Point
(% of Estimated
Capacity)
Start
25%
50%
After 48 hours
stagnation
Influent
Background
General
Active
Agent/
Residual k
X
X
X
Microbiological
X
X
X
60%
75%
After 48 hours
stagnation
100%
Chal-
lenge
pH -
9.0 " 0.2
X
X
X
X
X
X
X
X
Kb). Sampling Plan: lodinated Resins or Units
N-14
-------
Tests
Test Point
(% of Estimated
Capacity)
Start
25%
50%
After 48 hours
stagnation
General
Influent
Background
Active
Agent/
Residual
X
X
X
Microbiological
X
X
X
60%
75%
After 48 hours
stagnation
Chal-
lenge
PH -
9.0 ' 0.2
X
X
X
X
90%
100%
After 48 hours
stagnation
Chal-
lenge
PH -
5.0 * 0.2
X
X
X
X
2. Sampling Plan: Ceramic Candles or Units and U.V. Units
Tests
Test Point
Start
Day 3 (middle)
Day 6 (middle)
After 48 hours
stagnation
Influent
Background
General
Microbiological
X
X
X
Day 7 (middle)
Day 8 (near end)
After 48 hours
stagnation
Day 10-1/2
Chal-
lenge
X
X
X
X
(Note: All days are "running days" and exclude stagnation periods, when
the units contain silver, a leaching test shall be conducted as shown in
Section 3.5.1.e and silver residual will be measured at each microbiological
sampling point.)
N-15
-------
e. Leaching Tests for Silverized Units: Where the unit ccntair.
silver, additional tests utilizing Test Water #5 will be conducted
as follows:
Tests
Influent
Test Point Background Silver/Residual
Start X X
Day 2 • X
After 48 hours
stagnation X
f. Alternate Sampling Plans:
1. Since some laboratories may find it inconvenient to test soir.e
units on a 16 hour on/8 hour off cycle, two alternates are
recognized:
Go to a shorter operational day but lengthen the days of
test proportionally
- Use up to 20 percent "on"/80 percent "off" for a propor-
tionally shorter operational day
2. Sampling points must be appropriately adjusted in any alternate
sampling plan.
g. Application of Test Waters: The application of test waters is
designed to provide information on performance under both norrai and
stressed conditions; it should be the same or equivalent to the
following-:
1. a. Halogenated Resins or Units (Non-iodine based) —
First 50% of test period: Test Water 1 (General)
Last 50% of test period: Test Water 2 (Challenge)
(pH - 9.0 * 0.2)
b. lodinated Resins or Units —
First 50% of test period: Test Water 1 (General)
Next 25% of test period: Test Water 2 (Challenge)
(pH - 9.0 ' 0.2)
Last 25% of test period: Test Water 2 (Challe ge)
(but with pH - 5.0 * C.2)
2. Ceramic Candles or Units —
First 6 days of testing: Test Water 1 (General!
N-16
-------
Last 4-1/2 days of testing:
3. Ultraviolet (U.V.) Units —
First 6 days of testing:
Last 4-1/2 days of testing:
h. Analyses and monitoring:
Test Water 3 (Challenge)
Test Water 1 (General)
Test Water 4 (Challenge)
1. Microbiological sampling and analysis shall be conducted of the
specified influent and effluent sampling points during each
indicated sampling period.
2. Test Water Monitoring: The specified parameters of the various
test waters (see Section 3.3) will be measured and recorded at
each microbiological sampling point; the specified parameters
will be measured at least once on non-sampling days when the
units are being operated.
3. Background chemical analyses of influent watsr shall be con-
ducted at least once at the start of each test period to
determine the concentration of the U.S. EPA primary inorganic
contaminants, secondary contaminants and routine water para-
meters, not otherwise covered in the described test waters.
4. In addition, quality assurance testing shall be conducted for
the seed bacteria under environmental conditions on the first
and last days of testing to make sure that there is no signifi-
cant change over the test day. Populations will be measured
(for example, as dispersed in the supply tank) at the beginning
and end of the test day to detect possible incidental effects
such as proliferation, die-off, adsorption to surfaces, etc.
Relatively stable bacterial seed populations are essential to
an acceptable test program.
5. When a .unit contains a halogen or silver, the active agent
residual will be measured in the effluent at each microbiologi-
cal test (sampling) point.
6. Silver will additionally be measured three times in the efflu-
ent as specified in Section 3.5.I.e.
Neutralization of Disinfection Activity: Immediately after col-
lection, each test sample must be treated to neutralize residual
disinfectant. For halogen- and silver-based disinfectants this may
be done by addition of thioglycollate-thiosulfate neutralizer
solution (Chambers, et al., J. Amer. Water Works Assoc., 54:208-216,
1962). This solution should be prepared daily. All results are
invalid unless samples are neutralized immediately upon collection.
Special Provisions for Ceramic Candles or Units:
N-17
-------
1. Provisions for slow flow: Ceramic units may be subject to
clogging and greatly reduced flow over the test period. An
attempt should be made to maintain manufacturer rated or
claimed flow rates, but even at reduced flows the sampling
program set forth in Section 3.5.1.d.2 shall be maintained.
2. Cleaning of ceramic units: Units should be cleaned according
to manufacturer's directions. Two cleanings should occur
during the period of test (in order to prove the unit's
durability through the cleaning procedure). However, near the
time of microbiological sampling, the units should not be
cleaned until after the sampling. Further, no anti-microbial
chemical (for cleaning or sanitizing) may be applied to the
units during the test period unless the manufacturer specifies
the same as part of routine maintenance.
k. Halogenated units or U.V. units with mechanical filtration processes
separate from the microbiological disinfection components shall have
the mechanical filtration components replaced or serviced when
significant flow reduction (clogging) occurs in accordance with the
manufacturer's instructions in order to maintain the test flow rate.
Units with non-removable mechanical filtration components will be
run until flow is below that considered acceptable for consumer
convenience. (If premature clogging presents a problem, some
specialized units may require a customized test plan.)
1. Special Provisions for Ultraviolet (U.V.) Units:
1. The units will be adequately challenged by the prescribed test
waters; consequently they will be operated at normal intensity.
However, where the U.V. treatment component is preceded by
activated carbon treatment, the output of the U.V. lamp shall
be adjusted electronically, such as by reducing the current to
the lamp or other appropriate means, to be just above the alarm
point. This option shall be available for use under other U.V.
configurations, at the choice of the persons responsible for
testing, as an alternative to the use of the U.V. absorbent,
p-hydroxybenzoic acid.
2. Fail/safe: Units will provide and will be tested for fail/safe
warnings in the event of water quality changes or equipment
failures which may interfere with its microbiological purifica-
tion function.
3. Cleaning: Manufacturer's guidance with respect to cleaning
will be followed.
3.5.2 Procedure; Non-Plumbed Units
a. General: The basic procedures given in Section 3.5.1 shall be used
with necessary adaptations to allow for the specific design of the
N-18
-------
unit. In any event, the testing procedures shall provide a test
challenge equivalent to those for plumbed-in units.
b. Test conditions and apparatus should be adapted to reflect proposed
or actual use conditions in consultation with the manufacturer,
including flow rate and number of people to be served per day. In
some cases variable flow or other non-standard conditions may be
necessary to reflect a worst-case test.
3.5.3 Acceptance and Records
3.5.3.1
To qualify as a microbiological water purifier, all three production
units of a type must continuously meet or exceed the reduction requirements of
Table 1, within allowable measurement tolerances for not more than ten percent
of influent/effluent sample pairs, defined as follows:
Virus: one order of magnitude
Bacteria: one order of magnitude
Cysts: one/half order of magnitude
The geometric mean of all microbiological reductions must meet or exceed
the requirements of Table 1. An example is given as follows:
Unit: iodinated resin.
- Number of sample pairs over the completed test program:
10 per unit — 3 units » 30.
- Number of allowable sample pairs where log reduction is insuffi-
cient: 10% of 30 « 3 sample pairs.
- Allowable minimum log reductions in these 3 pairs:
Bacteria - 5 log
Virus - 3 log
Cyst - 2-1/2 log
- Conclusion: If the geometric mean of all reductions meets or
exceeds the requirements of Table 1, the indicated insufficient
sample pairs will be allowed.
3.5.3.2 Records
N-19
-------
All pertinent procedures and data shall be recorded in a standard format
and retained for possible review until the report of results has been com-
pletely accepted by review authorities, in no case for less than a year.
3.5.3.3 Scaling Up or Down
Where a manufacturer has several similar units using the same basic
technology and parallel construction and operation, it may sometimes be
appropriate to allow the test of one unit to be considered representative of
others. Where any serious doubt exists, all units of various sizes may
require testing. A "rule of three" is suggested as a matter of judgment.
Scaling up to three times larger or on-third, based on the size of either the
test unit or of its operative element, may be allowed. However, for UV units,
any size scale-up must be accompanied by a parallel increase in radiation
dose.
3.5.3.4
Where silver or some other chemical is used in the unit, concentrations
in the effluent water must meet any National Primary Drinking Water Maximum
Contaminant Level (MCL), additional Federal guidelines, or otherwise must rot
constitute a threat to health where no MCL exists.
N-20
-------
APPENDIX N-l
SUMMARY FOR BASIS OF STANDARDS AND TEST WATER PARAMETERS
A. Microbiological Reduction Requirements
1. Bacteria
Current standards for the microbiological safety of drinking
water are based on the presence of coliform bacteria of which
Klebsiella is a member. Members of the genus Klebsiella are also
potential pathogens of man (Vlassof, 1977). Klebsiella terrigena is
designated as the test organism since it is commonly found in
surface waters (Izard, et al., 1981).
Experience with the use of coliform bacteria to estimate the
presence of enteric bacterial pathogens in drinking water as per-
formed over the last 75 years indicates a high degree of reliabil-
ity. Required testing of more than one bacterial pathogen appears
unjustified since viral and Giardia testing will be required.
Enteric viruses and Giardia are known to be more resistant to common
disinfectants than enteric bacterial pathogens and viruses are more
resistant to removal by treatments such as filtration. Thus, any
treatment which would give a good removal of both virus and Giardia
pathogens would most likely reduce enteric bacteria below levels
considered infectious (Jarroll, et al., 1981; Liu, et al., 1971).
The concentration of coliform bacteria in raw sewage is approx-
imately 10 /100 ml. Concentrations in polluted stream waters have
been found to exceed 10 per 100 ml (Culp, et al., 1978, Table 10).
Based on the over 10 /100 ml concentrations observed in highly
polluted stream water and a target effluent concentration of less
than 1/100 ml, a 6 log reduction is recommended.
2. Virus
In the United States concentrations of enteroviruses are esti-
mated to range from 10 -10 /liter in raw sewage (Farrah and Schaub,
1971). Based on this observation it is estimated that, natural
waters contaminated with raw sewage may contain from 10 to 10
enteric viruses per liter.
There are currently no standards for viruses in drinking water
in the United States. However, EPA has proposed a non-enforceable
health-based recommended maximum contaminant level (RMCL) of zero
for viruses (EPA, 1985) . Several individuals and organizations have
developed guidelines for the presence of viruses in drinking water
and various experts have proposed standards (WHO, 1979, 1984; Berg,
1971; Melnick, 1976). It has generally been felt that drinking
N-21
-------
water should be free of infectious virus since even one virus i.-
potentially infectious and .suggested standards are largely based on
technological limits of our detection methodology. Guidelines
suggested by the World Health Organization (1984) and others
recommend that volumes to be tested be in the order of 100-1,000
liters and that viruses be absent in these volumes.
Assuming a target effluent level of less than one virus in 100
liters of water and a concentration of 10 enteric viruses in 100
liters of sewage-contaminated waters, the water purifier units
should achieve at least 4 logs of virus removal.
The relative resistance of enteric viruses to different dis-
infectants varies greatly among the enteric viruses and even among
members of the same group (i.e., enteroviruses). For example, while
f2 coliphage is one of the most resistant viruses to inactivation by
chlorine it is one of the most susceptible to inactivation by ozone
(Harakeh and Butler, 1984). Ionic conditions and pH can also affect
the relative resistance of different viruses to a disinfectant
(Engelbrecht, et al., 1980). On this basis it is felt that more
than one enteric virus should be tested to ensure the efficacy of
any disinfection system. Poliovirus typ« 1 (Strain LSc) was chosen
as one of the test viruses because it has been extensively used in
disinfection and environmental studies as representative of the
enterovirus family. It is recognized that it is not the most
resistant virus to inactivation by chlorine, but is still resistar.t
enough to serve as a useful indicator. Rotavirus is selected as the
second test enteric virus since it represents another group of
enteric viruses in nucleic acid composition and size. It is also a
major cause of viral gastroenteritis and has been documented as a
cause of water borne gastroenteritis (Gerba, et al. , 1985). The
human rotavirus or the similar Simian rotavirus may be used in the
test procedure. A net 4-log reduction for a joint challenge of
1 x 10 /L each for poliovirus and rotavirus is recommended.
3. Cysts (Protozoan)
Over the past several years, giardiasis has consistently been
one of the most frequently reported waterborne diseases transmitted
by drinking water in the United States (Craun, 1984). EPA has
proposed a RMCL of zero for Giardia (EPA, 1985). Its occurrence has
generally been associated with treatment deficiencies including
either inadequate or no filtration. Giardiasis has not been known
to occur from drinking water produced by well-operated filtration
treatment plants. De Walle, et al. (1984), in a study of filtration
treatment plant efficiencies, cited percent removals for Giardia in
pilot plant tests as follows:
- Rapid filtration with coagulation-sedimentation: 96.6-99.9%;
- Direct filtration with coagulation: 95.9-99.9%.
N-22
-------
From this research and from the lack of Giardia cases in
systems where adequate filtration exists, a 3-log (99.9%) reduction
requirement is considered to be conservative and to provide a
comparable level of protection for water purifiers to a
well-operated filtration treatment plant.
Data on environmental levels for cysts in natural waters is
limited because of the difficulties of sampling and analysis.
Unpublished data indicate very low levels from less than 1/L to less
than 10/L. Here a 3-log reduction would provide an effluent of less
than 1/100 L, comparable to the recommended virus reduction require-
ments.
Either Giardia lamblia or the related organism, Giardia muris,
which is reported to be a satisfactory test organism (Hoff, et al.,
1985) , may be used as the challenge organism. Tests will be con-
ducted with a challenge of 10 organisms per liter for a 3-log
reduction.
Where the treatment unit or component for cysts is based on the
principle of occlusion filtration alone, testing for a 3-log reduc-
tion of 4-6 micron particles or spheres (National Sanitation Founda-
tion Standard 53, as an example) is acceptable. Difficulties in the
cyst production and measurement technologies by lesser-equipped
laboratories may require the use of such alternative tests where
applicable.
B. Microbiological Purifier Test Procedures
1. Test Waters
a. The ^general test water (test water *1) is designed for the
normal, non-stressed phase of testing with characteristics that
may easily be obtained by the adjustment of many public system
tap waters.
b. Test water #2 is intended for the stressed phase of testing
where units involve halogen disinfectants.
1. Since the disinfection activity "of some halogens falls
with a rising pH, it is important to stress test at an
elevated pH. The recommended level of 9.0 * 0.2, while
exceeding the recommended secondary level (Environmental
Protection Agency, 1984) is still within a range seen in
some natural waters (Environmental Protection Agency,
1976). However, for iodine-based units, a second stress-
ful condition is provided — a pH of. 5.0 * 0.2 since
current information indicates that the disinfection
activity of iodine falls with a low pH (National Research
Council, 1980). While beneath the recommended secondary
level (Environmental Protection Agency, 1984) a pH of 5.0
K-23
-------
is not unusual in natural waters (Environmental Protection
Agency, 1976).
2. Organic matter as total organic carbon (TOO is known to
interfere with halogen disinfection. While this TOC is
higher than levels in many natural waters, the designated
concentration of 10 mg/L is cited as typical in stream
waters (Culp/Wesner/Culp, 1978).
3. High concentrations of turbidity can shield microorganisms
and interfere with disinfection. While the recommended
level of not less than 30 NTU is in the range of turbidi-
ties seen in secondary wastewater effluents, this level is
also found in many surface waters, especially durir.g
periods of heavy rainfall and snow melt (Culp/Wesner/Culp,
1978).
4. Studies with Giardia cysts have shown decreasing halogen
disinfection activity with lower temperatures (Jarroll,
et al., 1980); 4 C, a common low temperature in many
natural waters, is recommended for the stress test.
5. The amount of dissolved solids (TDS) may impact the
disinfection effectiveness of units that rely on displace-
afale or exchange elements by displacement of halogens or
resins, or it may interfere with adsorptive processes.
While TDS levels of 10,000 mg/L are considered unusable
for drinking, many supplies with over 2,000 mg/L are used
for potable purposes (Environmental Protection Agency,
1984). The recommended level of 1,500 mg/L represents a
realistic stress challenge.
c. Test water #3 is intended for the stressed phase of testing of
ceramic filtration candles or units with or without silver
impregnation.
1. Since viruses are typically eluted from adsorbing media at
high pHs (Environmental Protection Agency, 1978) it may be
concluded that a high pH will provide the most stressful
testing for a ceramic-type unit; consequently, the high
natural water pH of 9.0 is recommended.
2. Expert opinion also holds that organic material will
interfere with adsorption of viruses. Thus, a high total
organic carbon level of not less than 10 mg/L is recom-
mended .
3. Turbidity may enhance the entrapment and removal of
microorganisms but it also may stimulate "short-
circuiting" through some units. A turbidity level of
30 NTU will provide stress at time of sampling but the
-------
non-sampling level of 0.1-5 NTU will allow routine opera-
tion of units.
4. Expert opinion holds that low water temperatures and high
TDS would most likely interfere with virus reduction by
adsorption; consequently, a 4 C temperature and 1,500 mg/L
TDS are recommended.
d. Test water #4 is intended for the stressed phase of testing for
ultraviolet (UV) units.
1. In general, high TOC, turbidity and TDS and low tempera-
ture are considered most stressful for UV, and the in-
dicated challenge levels are the same as for test
water #2.
2. The pH is not critical and may range from 6.5 to 8.5.
3. In order to test the UV units at their most vulnerable
stage of operation, a color challenge (light absorption at
254 run) is to be maintained at a level where UV light
intensity is just above the unit's low intensity warning
alarm point. However, an alternate to the absorption
challenge is provided through adjusting the light intensi-
ty output of the UV lamp electronically by red-jcir.g
current to the lamp, or other appropriate means, to be
just above the alarm point; this approach would be
particularly necessary where the UV lamp is preceded by
activated carbon treatment.
e. Test water #5 is intended for the stressed leaching tests of
units containing silver. Low pH, TOC, turbidity, and TDS ar.d
higher temperature are felt to be the characteristics associ-
ated with increased leachability. The recommended pH of
5.0 * .2, while being beneath the recommended secondary range
of 6.5-S.5 (Environmental Protection Agency, 1984) is still
found in some natural waters.
2. Test Procedures
The plan for testing and sampling is designed to reveal unit
performance under both "normal" and "stressed" operating conditions.
The stressed phase would utilize a set of water quality and opera-
tions conditions to give the units a realistic worst case challenge.
Testing plans for a specific model might involve modifications to
the recommended plan; more samples could be taken and analyzed; more
units could be studied. The principle of demonstrating adequate
performance even under realistic worst case conditions should be
maintained and the final selected test procedures should be agreed
as between investigators and reviewers or regulators.
N-25
-------
While some aspects of the testing procedures have been utilized
in actual experiments, the proposed protocol has not been verified
or utilized for the various units that may be considered.. Conse-
quently, investigators and users of this protocol may find reasons
to alter some aspects through their practical experience; needed
changes should be discussed and cleared with involved reviewers/-
regulators.
N-26
-------
REFERENCES;
Berg, G. Integrated approach to the problem of viruses in water, j. ASCE,
Sanit. Eng. Div. 97:867-882, 1971.
Culp/Wesner/Culp. Guidance for planning the location of water supply intakes
downstream from municipal wastewater treatment facilities. EPA Report, Office
of Drinking Water. Washington, DC, 1978.
Craun, G. F. 1984. Waterborne outbreaks of giardiasis: Current status. In:
Giardia and giardiasis. D. L. Erlandsen and E. A. Meyer Eds., Plenum Press,
New York, pp. 243-261, 1984.
DeWalle, F. B.; j. Engeset; Lawrence, W. Removal of Giardia lamblia cyst by
drinking water treatment plants. Report No. EPA-600/52-84-069, Office of
Research and Development, Cincinnati, OH, 1984.
Engelbrecht, R. S. , et al. Comparative inactivation of viruses by chlorine.
Appl. Environ. Microbiol. 40:249-256, 1980.
Environmental Protection Agency. Quality criteria for water. Washington, DC,
1976.
Environmental Protection Agency. National secondary drinking water
regulations. EPA-570/9-76-000, Washington, DC, 1984.
Environmental Protection Agency. National primary drinking water regulations;
synthetic organic chemicals, inorganic chemicals and microorganisms; Proposed
rule. Federal Register, Nov. 13, 1985.
Farrah, S. R. , and S. A. Schaub. Viruses in wastewater sludges. In: Viral
Pollution of the Environment, G. Berg, Ed. CRC Press, Boca Raton, Florida.
pp. 161-163, 1983. ~"
Gerba, C. P.; Rose, J. B.; Singh, S. N. Waterborne gastroenteritis and viral
hepatitis. CRC Critical Rev. Environ. Contr. 15:213-236, 1985.
Harakeh, M.; Butler, M. Inactivation of human rotavirus, SA-11 and other
enteric viruses in effluent by disinfectants. J. Hyg. Camb. 93:157-163, 1984.
Hoff, J. C.; Rice, E. W.; Schaefer, F. W. Comparison of animal infectivity
and excystation as measures of Giardia muris cyst inactivation by chlorine.
Appl. Environ. Microbiol. 50:1115-1117, 1985.
Izard, D.; Farragut, C.;Gavini, F.; Kersters, K.; DeLey, J.; Ledere, H.
Klebsiella terrigena, a new species from water and soil. Intl. J. Systematic
Bacteriol. 31:116-127, 1981.
Jakubowski, w. Detection of Giardia cysts in drinking water. In: Giardia
and Giardiasis, Erlandsen, S. L.; Meyer, E. A. Eds., Plenum Press, NY.
pp. 263-286, 1984.
N-27
-------
Jarroll, E. L.; Bingham, A. K.; Meyer, E. A. Giardia cyst destruction:
Effectiveness of six small-quantity water disinfection methods. Am. j. Trop.
Med. 29:8-11, 1980
Jarroll, E. L.; Bingham, A. K.; Meyer, E. A. Effect of chlorine on Giardia
cyst viability. Appl. Environ. Microbiol. 43:483-487, 1981.
Liu, 0. C., et al. Relative resistance of 20 human enteric viruses to free
chlorine in Potomac River water. Proceedings of 13th Water Quality Conference
Snoeyink, V.; Griffin, V. Eds., pp. 171-195, 1971.
Melnick, J. L. Viruses in water. In: Viruses in Water Berg, G.;
Bodily, H. L.; Lennette, E. H.; Melnick, J. L.; Metclaf T. G., Eds. Amer.
Public Hlth. Assoc., Washington, DE, pp. 3-11, 1976.
National Research Council. The disinfection of drinking water, In: Drinking
Water and Health, Volume 2. Washington, DC, pp. 5-137, 1980.
National Sanitation Foundation. Drinking water treatment units: Health
effects. Standard 53. Ann Arbor, MI, 1982.
Vlassoff, L. T. Klebsiella. In: Bacterial Indicators/Health Hazards
Associated with Water Hoadley, A. W.; Dutka, B. J., Eds. American Society for
Testing and Materials, Philadelphia, PA. pp. 275-288, 1977.
World Health Organization. Human Viruses in Water, Technical Support
Series 639, World Health Organization, Geneva, 1979.
world Health Organization. Guidelines for Drinking Water Quality. Volume 1.
Recommendations. World Health Organization, Geneva, 1984.
N-28
-------
APPENDIX N-2
LIST OF PARTICIPANTS; TASK FORCE ON GUIDE STANDARD AND PROTOCOL FOR
TESTING^MICROBIOLOGICAL WATER PURIFIERS
Stephen A. Schaub, Chairman — U.S. Army Medical Bioengineering Research and
Development Laboratory (USAMBRDL), Fort Detrick, Maryland 21701, FTS:
8/935-7207 — Comm: 301/663-7207.
Frank A. Bell; Jr., Secretary — Criteria and Standards Division, Office of
Drinking Water (WH-550), Environmental Protection Agency, Washington,
D.C. 20460, Phone: 202/382-3037.
Paul Berger, Ph.D. — Criteria and Standards Division, Office of Drinking
Water (WH-550), Environmental Protection Agency, Washington, D.C. 20460,
Phone: 202/382-3039.
Art Castillo — Disinfectants Branch, Office of Pesticide Programs (TS-767CO,
Environmental Protection Agency, Washington, D.C. 20460, Phone: 703/557-
3695.
Ruth Douglas — Disinfectants Branch, Office of Pesticide Programs (TS-767O,
Environmental Protection Agency, Washington, D.C. 20460, Phone: 703/357-
3675.
Al Dufour — Microbiology Branch, Health Effects Research Laboratory,
Environmental Protection Agency, 26 W. St. Clair Street, Cincinnati, Ohio
45268, Phone: FTS: 8/684-7870 — Comm: 513/569-7870.
Ed Geldreich — Chief, Microbiological Treatment Branch, Water Engineering
Research Laboratory, Environmental Protection Agency, 26 W. St. Clair
Street, Cincinnati, Ohio 45268, Phone: FTS: 8/684-7232 — Conm:
513/569-7232.
Charles Gerba — Associate Professor, Department of Microbiology and
Immunology, University of Arizona, Tucson, Arizona 85721, Phone:
602/621-6906.
John Hoff — Microbiological Treatment Branch, Water Engineering Research
Laboratory, Environmental Protection Agency, 26 W. St. Clair Street,
Cincinnati, Ohio 45268, Phone: FTS: 8/684-7331 -- Comm: 513/569-7331.
Art Kaplan — Office of Research and Development (RD-681) Environmental
Protection Agency, Washington, D.C. 20460, Phone: 202/382-2583.
Bala Krishnan — Office of Research and Development (RD-681) Environmental
Protection Agency, Washington D.C. 20460, Phone: 202/382-2583.
John Lee — Disinfectants Branch, Office of Pesticide Programs (TS-767C)
N-2 9
-------
Environmental Protection Agency, Washington, D.C. 20460, Phone:
703/557-3663.
Dorothy Portner — Disinfectants Branch, Office of Pesticide Programs
(TS-767-C), Environmental Protection Agency, Washington, D.C. 20460,
Phone: 703/557-0484.
Don Reasoner — Microbiological Treatment Branch, Water Engineering Research
Laboratory, Environmental Protection Agency, 26 w. St. Clair Street,
Cincinnati, Ohio 45268, Phone: 312/654-4000.
P. Reguanthan (Regu) — EVerpure, Inc., 660 N. Blackhawk Drive, Westmont,
Illinois 60559, Phone: 312/654-4000.
David Stangel — Policy and Analysis Branch, Office of Compliance Monitoring,
Environmental Protection Agency, Washington, D.C., Phone: 202/382-7845.
Richard Tobin — Monitoring and Criteria Division, Environmental Health
Center, Department of Health and Welfare of Canada, Tunney's Pasture,
Ottawa, Ontario, K1A OL2, Canada, Phone: 613/990-8982.
N-30
-------
APPENDIX N-3
PZSPONSE BY REVIEW SUBCOMMITTEEfl) TO PUBLIC COMMENTS ON GUIDE STANDARD
AND PROTOCOL FOR TESTING MICROBIOLOGICAL WATER PURIFIERS
A. Recommendation for the use of Giardia lamblia cysts as a replacement for
Giardia muris cysts as the protozoan cyst test organisms.
Recommendation;
The subcommittee concurs with the recommendation and further endorses the
use of Giardia lamblia as the preferred cyst test for evaluation of all
treatment units and devices. Obviously the use of the protozoan orga-
nisms of actual health concern in testing is most desirable. Anyone
finding the Giardia lamblia strain feasible for testing and cost-
effective to work with is encouraged to use same instead of Giardia
nuris.
B. Substitution of 4-6 micron bead or particle tests as an alternate option
instead of the Giardia cysts for evaluating devices that rely strictly on
occlusion filtration for microbiological removal: Several commenters
criticized the use of beads or particles (e.g., A.C. fine dust) and
recommended only use of live Giardia cysts for performance tests.
discussion;
The subcommittee recognizes and favors the use of the natural human
parasite, Giardia lamblia, but was not aware of any convincing scientific
data which would disallow the optional use of testing with beads or
particles for units or devices using only occlusion filtration to remove
microorganisms. Previous development of the National Sanitation Standard
(NSF) 53 (1982) requirement for cyst reduction (using 4-6 micron parti-
cles as cyst models) was based on engineering and scientific opinion and
experimental evidence at that time. Specifically, Logsdon used
radioactive cyst . models in the initial phase of a study of removal
efficiencies for diatomaceous earth filters; subsequent experiments with
Giardia muris cysts confirmed the efficacy of the diatomaceous earth
filters. Further studies by Hendricks and DeWalle with Giardia
lanblia cysts also showed comparable reduction efficiencies for
diatomaceous earth filters.
l.S.A. Schaub; F.A. Bell, Jr.; P. Berger; C. Gerba; J. Hoff;
P. Regunathan; and R. Tobin. [Includes additional revision pursuant to
Scientific Advisory Panel review (Federal Insecticide, Fungicide, and
Rodenticide Act).]
N-31
-------
Subsequently confirmatory parallel testing results have been developed
vis-a-vis 4-6 micron particles as compared to Giardia lamblia cysts.
Specifically, two units listed by NSF for cyst reduction (using 4-6
micron particles) have also been tested and lifted for 100% efficiency
reduction (using Giardia lamblia cysts) by Hibler :
1. Everpure Model QC4-SC
2. Royal Boulton Model F303.
Again we prefer the use of the human pathogen, Giardia lamblia; however,
no experimental data has been provided regarding the lack of validity or
of failure in previous tests utilizing beads or particles of 4-6 microns.
In most cases the bacterial or viral challenges to occlusion filters ill
represent a greater problem in terms of microbiological reduction
requirements than will cysts. Therefore, without substantiation of
deficiencies, the use of 4-6 micron beads or particles is considered to
be as feasible as the use of live cysts for routine performance testing
of water filtration (occlusion) devices.
Seeonmendation;
Recommend retaining the optional use of 4-6 micron particles or beads for
cyst reduction testing in occlusion filtration devices only.
References;
Logsdon, G. S. , et al. Alternative Filtration Methods for Removal
of Giardia Cysts and Cyst Models, JAWWA, 73(2)111-118, 1981.
Logsdon, G. S.; Hendricks, D. W., et al. Control of Giardia Cysts
by Filtration.- The Laboratory's Ro'se. Presented at the AWWA Water
Quality Technology Conference, December, 1983.
DeWalle, et al. Removal of Giardia lamblia Cysts by Drinking Water
Treatment Plants, Grant No. R806127, Report to Drinking Water
Research Division, U.S. EPA (ORD/MERL), Cincinnati, Ohio.
(4)
National Sanitation Foundation, Listing of Drinking Water Treatment
Units, Standard 53. May, 1986.
Hibler, C. P. An Evaluation of Filters in the Removal of Giardia
lamblia. Water Technology, pp. 34-36. July, 1984.
C. Alternate assay techniques for cyst tests (Jensen): Proposed alterations
in cyst tests include a different method for separating cysts from fecal
material and an assay method involving the counting of trophozoites as
well as intact cysts. Both alterations have been used by- Bingham, et -al.
(Exp. Parasitol., 47:284-291, 1979).
Recommendation:
N-32
-------
These alterations appear to be reasonable laboratory procedures, support-
ed by a peer-reviewed article and will be included in the Report as
options for possible development and use by interested laboratories.
D. The use of pour plate techniques as an option for Klebsiella terrigena
bacteria analyses.
Recommendation;
The pour plate technique adds a heat stress factor to the bacteria which
constitutes a possible deficiency. However, it is a recognized standard
method and probably will not adversely affect the Klebsiella terrigena.
Consequently, it will be added to the Report as one of the acceptable
techniques.
E. Option of using Escherichia coli in lieu of Klebsiella terrigena for the
bacterial tests.
Discussion;
Appendix N-l, Section A.I. of the Guide Standard and Protocol sets forth
the basis for selection of K. terrigena as the test bacteria. The
selection was made along pragmatic line emphasizing the occurrence of K^
terrigena in surface waters and that it would represent the enteric
bacteria. It was also pointed out that the tests with virus and Giardia
were expected to be more severe than the bacterial tests. For comprehen-
siveness, bacterial tests were included in the protocol but were not felt
to be as crucial as the virus and Giardia tests.
E. coli, or any number of other generally accepted indicator bacteria,
could be used for the test program if they were shown to have good
testing and survival characteristics (equivalent to K. terrigena) by the
interested research laboratory.
Recommendation;
The intent of the Guide Standard and Protocol is to provide a baseline
program subject to modification when properly supported by an interested
laboratory. Consequently, any laboratory could propose and with proper
support (demonstrating challenge and test equivalency to K. terrigena)
use Escherichia coli or one of the other enteric bacteria. This idea
will be included in revised working in Section 1.2.2, "General Guide."
F. Performance requirements for Giardia cysts and virus in relation to the
EPA-Recommended Maximum Contamination Levels (RMCLs) of zero.
Discussion;
The RMCLs of zero for Giardia and viruses which have been proposed by EPA
are health goals. They are no enforceable standards since to assure the
presence of "no organisms" would require an infinite sample. The
N-33
-------
rationale for the recommended performance requirements for Giardia cyst i
and virus is set forth in Sections A.2 and A.3 of Appendix A. We feel
that these requirements together with the application of realistic worst
case test conditions will provide a conservative test for units resulting
in treated effluent water equivalent to that of a public water supply
meeting the microbiological requirements and intent of the National
Primary Drinking Water Regulations.
Recommendation;
Retain recommended performance (log reduction) requirements for cyst and
virus reduction.
G. Rotavirus and its proposed assay: One commenter states that the rota-
virus tests are impractical because Amirtharajah (J. AWWA, 78 (3) :34-49,
1976) cites "no satisfactory culture procedures available for analysis of
these pathogens and, therefore, monitoring would not be feasible."
Discussion;
Section 3.4.1.2, "Virus Tests" of the Report, presents means for cul-
turing and assaying rotaviruses. This means for doing the rotavirus
tests are available and are practical for application in the laboratory.
Dr. Amirtharajah was referring to the field collection, identification in
the presence of a wide variety of microorganisms, and quantification as
not being "satisfactory." Laboratory analysis of rotaviruses is practi-
cal but their field monitoring may not yet be feasible.
Further, the selection of both poliovirus and rotavirus as test viruses
was necessitated by the fact that the surface adsorptive properties and
disinfection resistance of the various enteric viruses have been shown to
differ significantly by virus group and by strains of a specific virus.
while all enteric viruses and their strains could not be economically
tested, it was determined by the task force that at least two distinctly
different virus types should be tested to achieve some idea of the
diversity of removal by the various types of water purifiers. Polio and
rota viruses have distinctly different physical and chemical charac-
teristics representative of the viruses of concern. Polioviruses are
small singl« stranded RNA viruses with generally good adsorptive proper-
ties to surfaces and filter media while rotaviruses are over twice as
large, are double stranded RNA and in some studies have been found to
possess less potential for adsorption onto surfaces or filter media.
These two viruses also have been demonstrated to have somewhat different
disinfection kinetics.
Recommendation;
Retain the rotavirus test requirements.
H. Definition of microbiological water purifier: One general comment
requested redefinition based on "lack of any virus removal "requirement
N-34
-------
in the EPA primary drinking water regulations, so that no virus reduction
requirement should be included. Also, it was claimed that the separation
of purifiers from non-purifiers would be a "disservice to consumers and
other users."
Discussion;
viruses are recognized in the EPA regulations vis-a-vis a proposed recom-
mended maximum contaminant level of zero. Since virus monitoring for
compliance with a possible MCL is not yet feasible, a treatment require-
ment is necessary. Virus control will be considered in the Safe Drinking
Water Act filtration and disinfection treatment regulations. The reduc-
tion of viruses by treatment is discussed by Amirtharajah (J. AWWA,
78:3:34-49, 1986).
With respect to consumers and other users, we feel that the current
definition is appropriate and necessary. The average consumer cannot be
expected to know the difference between viruses, bacteria and cysts, or
when a raw water will or will not contain any of these organisms. In
order to protect the average consumer, the subject units either alone or
with supplementary treatment, should be able to cope with all of the
specified organisms.
Recommendation;
Retain the current definition for microbiological water purifier.
I. Coverage of units: Several comments related to the coverage of units.
These questions are addressed individually as follows:
1. Ultraviolet units that are used for supplemental treatment of water
from public water system taps would not be covered. We agree that
such units are not covered and parenthetical language has been
included in Section 1.3.2.3 to clarify this point.
2. A special status should be given to units which remove Giardia and
bacteria but not virus. Specifically, the meaning of Section 1.2.4,
"Exceptions," was addressed. The "Exceptions" section was specif-
ically developed to relate to the problem of public water syster.s
having disinfection but no filtration on a surface supply. Cysts
alone have been found to survive disinfection treatment and could be
present in such treated waters. In this case an effective cyst
filter serves an independent, beneficial purpose and should not be
required to be a microbiological water purifier. However, such a
unit should not be used as sole treatment for untreated raw water.
Additional parenthetical language has been added to Section 1.2.4.
3. The entire treatment unit or system should be tested, not :ust a
single component. We agree but believe that it is sufficiently
clear without providing additional language.
N-35
-------
4. The protocol should be expanded to cover units for the reduction of
TCE, EDB and other chemical pollutants. We felt that the introduc-
tion of non-microbiological claims to the standard would make it
large, unwieldy and duplicateve of an existing third-party standards
and testing program (see Section 1.2.5).
J. Alleged preference of National Sanitation Foundation (NSF) over other
laboratories for conducting the microbiological water purifier testing
protocol. The comment indicated that we were giving NSF preferential
treatment "to the detriment of other laboratories well qualified to
perform the required protocol."
Discussion;
we have made appropriate references to existing standards (#42 and *53)
developed by the NSF standards development process. Standard 53, the
health effects standard, was developed by a broadly based Drinking Water
Treatment Units Committee, including representatives from local, State
and Federal health and environmental agencies, universities, professional
and technical associations, as well as water quality industry
representatives. It was adopted in 1982 and the only test from it
utilized in our Report has been substantiated as described in Part E of
this "Response."
Nowhere in our report have we advocated NSF (or any other laboratory) as
the prime or only laboratory for implementing "the required protocol."
Recommendation;
No action needed.
K. Instruction concerning effective lifetime. One comment described an
alternate mea.ns for determining lifetime where a ceramic unit is
"brushed" to renew its utility and is gradually reduced in diameter. A
gauge is provided to measure diameter and to determine when replacement
is needed.
Recommendation;
Where a manufacturer provides a satisfactory "other" means of determining
lifetime, this should be accepted. Appropriate words have been added to
Section 2.4.I.C.
L. Ceramic candles should not be cleaned during testing because some
consumers would not clean them and this would provide the "worst case
test." One comment asserted this point.
Discussion;
There is some truth to this proposition. However, the other approach nay
also have validity. Frequent brushing may reduce filtration efficiency.
N-36
-------
In any event, where a manufacturer prescribes filter cleaning and how to
do it, and provides a gauge to determine lifetime, we feel the testing
program is bound to follow the manufacturer's directions.
Recommendation;
No change needed.
M. Scaling up or down. One comment points out that one or more manufac-
turers may vary size of treatment units by increasing or decreasing the
number of operative units rather than the size of the operative unit.
The comment suggests allowing scaling based on size of operative unit.
Recommendation;
We agree with the comment and have added clarifying words to Sec-
tion 3.5.3.3.
N. Turbidity level of "not less that 30 MTU" for ceramic candles or units.
One comment states that "Such levels are impossible to utilize in testing
• mechanical filtration devices which will clog entirely or require such
frequent brushing as to render the test impossible as a practical
matter."
Discussion;
We recognized the potential "clogging problems" in Section 3.5.1.a(2)
where the 30 NTU water is only to be applied immediately before ar.d
during each sampling event; the non-sampling turbidity level, which will
be applied over 90% of the "on" time, is currently set at no less than
10 NTU.
Turbidity levels of 30 NTU are commonly found in surface waters during
heavy rainfall or snow melt. Treatment units may be used under these
circumstances, so this challenge level should be retained. However, most
usage will occur under background conditions so the non-sampling
turbidity levels should be 0.1-5 NTU.
Recommendations;
1. Retain sampling turbidity level of not less than 30 NTU, and
2. Change non-sampling turbidity to 0.1-5 NTU. Appropriate wording
changes have been introduced in Section 3.5.1.a(2) and in Appen-
dix N-l, Section B.
0. Chlorine in test water #5. One comment asserts that chlorine "tends to
increase silver ion leaching activity" and that a high chlorine level
should be included in the silver leaching test; but no reference or
evidence, however, is provided to back this assertion.
N-37
-------
Discussion;
We have no compelling evidence or reason to expect that chlorine will
enhance the leaching of silver. However, the prescribed.low pH and 7ES
levels will provide a clearly severe test for silver leaching.
Recommendation;
No change needed.
P. Unnecessary difficulty and expense of test protocols. Several comments
were made under this general heading. These comments are outlined and
discussed as follows:
1. Too many sampling events are required; sampling of a few units at
start, middle and finish should be satisfactory: The committee has
carefully laid out the standard and protocol and we feel the minimum
sampling plan must be maintained for the consumers' health pro-
tection.
2. Three units are too many to study; parallel testing of two units
should be satisfactory: For consumer protection, the Disinfectants
Branch, Office of Pesticide Programs, has traditionally required the
testing of three units. The committee recognizes the additional
cost involved in testing a third unit but feels that this will
provide a minimum level of assurance to prevent infectious disease
and recommends retention of the 3-unit requirement.
3. The protocol requires large tanks and microbiological reseeding or. a
daily basis: We feel that the tank size requirements are not
extreme and can be met by an interested laboratory. With respect to
reseeding, it should be pointed out that virus and cyst seeding r.eed
only be conducted immediately before and during the sampling "on"
period (s~ee Section 3.5.1.b(2)), equivalent to less that 10% of the
"on" time. This "spot" seeding for viruses and cysts recognized the
expense and .difficulty of maintaining large populations of these
organisms. Continuous seeding was provided for bacteria because
they are easier to grow and maintain and might have the capacity to
grow through some units, given enough time and opportunity.
4. Challenge levels of contaminants are too high compared to known
environmental conditions and the required log reductions exceed Safe
Drinking Water Act requirements: As explained in a footnote to
Table 1, Section 2, the influent challenges may constitute greater
concentrations than would be anticipated in source waters. These
levels are necessary to test properly for the required log reduc-
tions without having to utilize sample concentration procedures
which are time/labor intensive and which may, on their own, intro-
duce quantitative errors to the microbiological assays. As men-
tioned in Part I of this paper, the log reductions for bacteria,
virus and Giardia have been suggested for public water system
N-38
-------
treatment in a paper by Amirtharajah (1986, JAWWA, 78:3:34-49). The
reductions in the microbiological purifier standard are entirely
compatible with the reductions cited for public water supply
treatment.
N-J9
-------
APPENDIX 0
GUIDELINES TO EVALUATE OZONE DISINFECTION
Principal Technical Author: Dr. Ovadia Lev
Division of Environmental Sciences
Hebrew University of Jerusalem
Jerusalem, Israel
-------
APPENDIX 0
GUIDELINES TO EVALUATE OZONE DISINFECTION
TABLE OF CONTENTS
Page
0.1 INTRODUCTION 0.1-1
0.1.1 Background 0.1.1
0.1.2 Objectives of the Recommended Guidelines 0.1-2
0.1.3 EPA's Approach in Setting the Recommended
Guidelines 0.1-3
0.1.4 Typical Ozone Disinfection Units 0.1-4
0.2 DETERMINATION OF CONTACT TIME (T) 0.2-1
0.2.1 Background 0.2-1
0.2.2 T10 Analysis 0.2-1
0.2.3 Additional Considerations for T10:
Multiple Chamber Contactors 0.2-3
0.2-4 Alternative Analysis of Disinfection Kinetics 0.2-7
0.2.5 Continuously Stirred Tank Reactor (CSTR) Approach 0.2-9
0.2.6 Segregated Flow Analysis (SFA) 0.2-10
0.2.7 Relative Inactivation of Giardia Cysts and Viruses 0.2-12
0.2.8 Examples of Determining Contact Time (T) 0.2-13
0.2.8.1 Evaluation Using T.0 0.2-13
0.2.8.2 Evaluations Using CSTR Calculations 0.2-16
0.2.8.3 Evaluations Using SFA 0.2-18
0.2.9 Estimating T 0.2-21
0.3 DETERMINATION OF OZONE CONCENTRATION (C) 0.3-1
0.3.1 Introduction 0.3-1
0.3.2 Direct Measurement of C 0.3-2
0.3.3 Estimating C Based on Residual Measurements
at the Outlet 0.3-6
0.3.3.1 First Chambers 0.3-6
0.3.3.2 Subsequent Chambers 0.3-7
0.3.4 Estimating C 0.3.-10
0.4 SITE SPECIFIC EVALUATION OF OZONE CONTACTORS 0.4-1
0.4.1 Introduction 0.4-1
0.4.2 Site Specific Correlations of C with an
Observable Variable 0.4-2
0.4.2.1 Utilizing Off-Gas Measurements 0.4-4
0.4.3 Modeling the Performance of Full Scale Operations 0.4-7
0.4.4 Microbial Indicator Studies to Model Inactivation
Contactors 0.4-9
REFERENCES
-------
TABLE OF CONTENTS (Continued)
LIST OF FIGURES
Figure Following
No. Title Page
0-1 Turbine Contactor, Haworth Water Treatment Plant . 0.1-5
Hackensack, NJ
0-2 Multiple-Chamber Ozone System, Los Angeles
Aqueduct Filtration Plant 0.1-5
0-3 Multiple-Chamber Ozone System, City of Tucson, Arizona 0.1-5
0-4 Multiple-Chamber Ozone System. East Bay Municipal
Utility District, Oakland, California 0.1-5
0-5 Schematic of Deep U-Tube Ozone Contactor 0.1-6
0-6 Schematic of In-Line Static Mixer 0.1-6
0-7 Principal of Segregated Flow Analysis 0.2-11
0-8 Segregated Flow Analysis of an Ozone Contactor
Tracer Study 0.2-14
0-9 Tracer Study of Sturgeon Bay Ozone Contactor 0.2-15
0-10 Segregated Flow Analysis of Ozone Contactor
- Integration of Survival Efficiency. 0.2-20
0-11 Segregated Flow Analysis of Ozone Contact Chamber 0.2-20
0-12 Decision Tree for Estimating T 0.2-21
0-13 Flow Configurations in Ozone Contact Chambers. 0.3-1
0-14 Direct Measurements for Determining T 0.3-5
0-15 Ozone Concentration Profiles 0.3-8
0-16 Decision Tree for Estimating T 0.3-10
0-17 Example of Empirical Correlation of Residual Ozone
and Off-Gas 0.4-6
-------
TABLE OF CONTENTS (Continued)
LIST OF TABLES
Table Following
No. Title Page
0-1 Recommended Procedures to Calculate Contact Time (T) 0.2-2
0-2 CT Values for Inactivation by Ozone 0.2-7
0-3 k Values for Inactivation by Ozone 0.2-7
0-4 Spread Sheet Notations for SFA 0.2-18
0-5 Segregated Flow Analysis of an Ozone Disinfection
Contactor at Hackensack 0.2-20
0-6 Correlations to Predict C Based on Outlet Ozone
Concentrations 0.3-6
0-7 Henry's Constants for Ozone 0.4-5
0-8 Empirical Correlation Between Cout and yout 0.4-6
-------
0.1 INTRODUCTION
0.1.1 Background
The Surface Water Treatment Rule (SWTR) specifies overall minimal
removal/inactivation efficiencies by filtration and disinfection for
Giardia cysts and viruses. The SWTR uses the "CT" concept to predict
inactivation efficiencies of microorganisms by disinfection. "CT"
represents the product of contact or exposure time ("T") and the
concentration of disinfectant ("C") during disinfection. The Guidance
Manual suggests design, operating and performance criteria for specific
surface water quality conditions to provide compliance with the SWTR.
Appendix C of the Guidance Manual recommended guidelines for the
determination of contact time (T10) for the disinfection of drinking water.
T10 is the time defined to assure that 90 percent of the water that enters
the disinfection chamber will remain for at least T10 minutes. This
appendix recommends additional procedures which may be used for consistent
determination of the C and T for systems using disinfection by ozone.
Ozone has unique characteristics and warrants special consideration
for estimating inactivation efficiencies. In developing these recommended
procedures, EPA addressed the following complications that are specific to
ozone disinfection and distinguish it from other typical disinfection
processes.
Despite the long operational experience with ozone disinfec-
tion, the data regarding performance of ozone as a disinfectant
are rather limited. Most of the available inactivation rate
data are derived from laboratory conditions which are sub-
stantially different than full scale continuous operation,
generally more so than for other disinfectants.
From a technical point of view, disinfection of drinking water
by ozone is more complicated than disinfection by other common
disinfectants because of ozone's unique gas-liquid mass
transfer characteristics. Ozone requires sophisticated mass
transfer equipment to introduce it into water, because of the
relatively low ozone concentration in the feed gas. Ozone is
a powerful oxidant, that reacts rapidly with organic and
inorganic substances present in the water and undergoes auto-
decomposition. Therefore, it's residual is much less stable
than that of other disinfectants and dissipates rapidly.
0.1-1
-------
Ozone contactors exhibit more diversified types of flow
configurations relative to the flow pattern in contactors for
the other disinfectants. The flow configuration often ranges
from an almost continuously stirred tank reactor (CSTR) to an
almost ideal plug flow configuration, making the determination
of contact time for ozonation more complex than for other
disinfectants.
Ozone contactors are closed vessels because of ozone's toxici-
ty. The contactors have limited access for measurement of the
ozone concentration profile within the contactor. Gas bubbles
also may interfere with the determination of the dissolved
ozone concentration, if the bubbles are entrapped during
samp!ing.
Ozone technology is still evolving and new types of ozone
contactors are being developed. These guidelines should not
set unnecessary obstacles that will inhibit engineering
progress and prevent innovative designs of disinfection
systems.
EPA's procedures for determining C and T for disinfection with ozone
differ from those recommended for systems using chlorine, chloramines or
chlorine dioxide as disinfectants. The CT evaluation procedures
presented in previous chapters of the Guidance Manual are not appropriate
for ozone disinfection because they would result in excessive ozone
dosages. Excessive ozone doses result in high energy requirements and
costs and may 1-ead to unnecessary production of ozonation by-products
which may have associated health risks. Additionally, excess dissolved or
entrained ozone should be destroyed or removed before reaching the first
drinking water consumer or plant personnel, in order to prevent health
risks. Therefore, excessive dosage of ozone may require an additional
unit operation to destroy the remaining residual ozone. This process is
expensive and may not be necessary if guidelines such as those presented
in this section are used for compliance with the SWTR.
0.1.2 Objectives of the Recommended Guidelines
The recommended guidelines were developed to assure compliance with
the SWTR for the whole range of flow rates, flow configurations and water
quality conditions that may be encountered with ozone disinfection of
drinking water. The primary goal of these guidelines is to assure
0.1-2
-------
compliance with the SWTR even under "worst case" conditions. Without
compromising this primary goal, these guidelines were developed to meet
the following criteria:
1. Simplicity: The guidelines for selecting contact time (T) and
concentration.(C) have to be easily understood by practitio-
ners, even by those who do not have an engineering background.
2. Implementation: The procedure to estimate concentration and
time should be easily implemented, even by water utilities that
have only limited engineering and technological means.
3. Economics: The guidelines should be designed to minimize
capital and operating costs and to minimize ozone consumption.
The guidelines should be flexible enough to allow systems to
take advantage of site specific characteristics of the treated
water and the various designs of ozone contactors.
0.1.3 EPA's Approach in. Setting the Recommended Guidelines
EPA is aware that the current technological knowledge is insufficient
to formulate a consistent and efficient single set-of general rules that
will achieve these conflicting goals and still guarantee compliance with
the SWTR. Therefore, EPA developed two alternative sets of guidelines
that systems may use depending on their technological resources:
Alternative 1: General guidelines which assure compliance with
the SWTR regardless of the site specific conditions,
Alternative 2: A sophisticated evaluation procedure that water
utilities may use to take advantage of their site specific
conditions.
These guidelines are considered to be state-of-the-art. As more
information becomes available, more accurate approaches and models may be
developed. A brief description of the current alternatives follows.
Alternative 1 - General Guidelines
This alternative consists of a simple set of general guidelines that
assure compliance with the SWTR even under worst case conditions. These
guidelines were developed to emphasize generality and simplicity.
However, they may not result in the lowest cost alternative(s).
0.1-3
-------
The second and third sections of this Appendix contain detailed
descriptions of the general guidelines. Section 0.2 contains procedures
to estimate the contact time (T) and Section 0.3 contains procedures to
calculate the concentration (C) in ozone contactors based on simple
measurements of some oarameters. The basis for these general guidelines
is discussed in two papers (Lev and Regli, 1990a,b).
Alternative 2 - Site Specific Evaluation Procedures
This alternative consists of a more sophisticated set of evaluation
procedures to characterize the performance of ozone contactors and thereby
take advantage of site specific conditions. EPA recommends that systems
be given opportunity to prove by further experimental and analytical data
that the performance of their ozone contactors are better than the
performance predicted by the first alternative, thereby allowing a system
to minimize costs while providing adequate treatment.
Section 0.4 outlines recommended procedures for demonstrating that
ozone contactors achieve better performance than that predicted by the
first alternative.
0.1.4 Topical Ozone Disinfection Units
Several types of ozone contactors are currently in use for disinfec-
tion of drinking-water in the United States. Other types of contactors
are being designed or are being used for disinfection of treated sewage
effluents. The following characteristics illustrate the diversity of
ozone contactors:
The capacity of ozonation systems ranges from less than 1
million gallons per day (mgd) up to 600 mgd.
The volume of ozone contactors ranges from less than 35 cubic
feet up to more than 35,000 cubic feet for a single chambers.
The ozone gas stream may be introduced into the water by
several ways including porous diffusers, submerged turbines and
gas injectors.
Ozone contactors include single or multiple gas/liquid contact
chambers.
0.1-4
-------
Four typical ozone contactors currently in use or in design in the
United States are shown on Figures 0-1 through 0-4. Figure 0-1 presents
a schematic of an aspirating turbine contactor, operating in countercur-
rent flow. A turbine agitator is used to introduce the ozone into the
contactor and to mix the liquid phase. This unit may serve as the first
ozone chamber in a series of chambers or as a single chamber. The unit
shown in this figure is from the Hackensack Water Company's Haworth Plant
at Haworth, New Jersey. The turbine chamber is followed by a reactive
chamber to provide additional contact time. Studies conducted in the full
scale turbine agitated contactor demonstrated that even when the ozone
demand was high, the dissolved ozone concentration was almost constant
throughout the contactor as a result of the vigorous action of the turbine
(Schwartz et al, 1990).
The 600 mgd ozone system of the city of Los Angeles is comprised of
four parallel contactors each consisting of six chambers. A schematic of
one of these contactors is presented on Figure 0-2. (Stolarik and
Christie, 1990) As indicated on this figure:
An oxygen stream containing a few percent by weight of ozone
is compressed through bubble diffusers into the first and third
chambers of the contactor.
The second and fourth chambers are used to provide contact
time, without supplying additional gas to the liquid stream.
The size of the first three gas/liquid contact chambers is
20,400 cubic feet each.
The fifth and sixth chambers are the ozonated water channel and
the rapid mixer basins.
The liquid and gas streams in the first and third chambers flow
in a counter-current pattern; the gas stream flows upward and
the water stream flows downward.
As illustrated on Figure 0-3, a similar design approach was taken by
the City of Tucson, Arizona. This contactor is comprised of five
chambers, all of which are equipped with gas diffusers. The sixth chamber
has no diffusers. The flow in all six chambers is counter-current flow.
These counter-current chambers are separated by narrower co-current liquid
channels in which the water flows upward to the inlet of the next chamber.
0.1-5
-------
The East Bay Municipal Utility District Oakland, California is
currently designing two 60 mgd ozone contactors, the first of which is to
be operational in 1991. As illustrated on Figure 0-4, the contactor
includes three ozone gas/liquid chambers followed by three more reactive
chambers to provide additional contact time. The first and third chambers
are counter-current and the second chamber is co-current. In the latter,
the water and the gas bubbles flow in the same direction. Hydrogen
peroxide can be added at the outlet of the contactor to dissipate any
residual dissolved ozone.
The following types of contactors are already used in other parts of
the world, but have not yet been installed in the United States:
The Deep U Tube contactor shown on Figure 0-5, is comprised of
two concentric flow tubes. Water and gas streams are intro-
duced at the top of the inner tube and the mixture is pumped
10 to 30 meters downwards at a velocity greater than the rise
rate of the gas. After reaching the very bottom of the
contactor the mixture flows up in the outer section of the
contactor. The Deep U-tube is basically a co-current operation
taking advantage of the increased mass transfer at high
pressures.
The Static Mixer (shown on Figure 0-6) consists of a flow tube
equipped with baffles to produce efficient contact between the
liquid and the gas streams. This installation is gaining
popularity in Europe particularly for small and medium size
disinfection units. Here the flow is basically co-current, the
liquid and gas flow is in the same direction, through a tube
equipped with baffles that create turbulence and thus increases
the rate of gas-liquid mass transfer. The ozone is applied to
the water prior to the mixer either through an eductor or a
diffuser. Following dissolution through the mixer, the water
flows through a pipeline in plug flow.
Some contactors, particularly for disinfection of wastewater
effluents, use packed beds to increase mass transfer. Co-
current or counter-current flow configuration may be used.
The guidelines were developed to represent four different flow
conditions in ozone contactors. However, other types of contactors or
flow conditions may still use the same guidelines if the features of the
gas-liquid flow configuration as presented in Section 0.4 of this appendix
are taken into account.
0.1-6
-------
•
B
Ll
3
J
L
L
J
—
•
J
J
•
c
J
c
t
t_
(
c
\
^>
« «
— h»
(
r
^
a
f
I
/
$
. i . .
c
\
• :«-.'
. • . \ • i
*
-
1
•
Li'.*zn
^*-
* ? ; .
ua
> UJ
K«
ii?
Ui Z
CO
\
L. • . « ,v- •
Ji \
\
•^ % •
• •
• » %.r
t jg.
./Ji
*7#>-
/ "J.
o
z
K UJ
••*
H
Z
j
Q.
^
oc z
0 u,
(J H
< <
KLU
Zf^
cc
O H-
oc •>
UJ 01 Z
o i -
g ^ <
0 z 5
UJ ^ Z
5 n w
£ ^ 0
3 < <
H Z X
»»
1
O
UJ
QC
O
u»
-------
Z
<7i
03
'sJ
Z
O
o
UJ
O
INI
O
•4-
0
0
0
0
0
0
0
0
0
0
0
0
CN
UJ
u.
UJ
o
ISJ
o
Q
UJ
D
O
<
CO
UJ
_J
UJ
O
Z
5 _:
UJ f
S:
(A *
U C
2 *
Sc
CC
UJ
CD
UJ
i
(M
O
UJ
cc
cr
UJ
-------
TO
RAP'D
^ r
o
i
FflCM
RAW WATER
1
1
'
•
-
2
i
i
i
r
3
1 1
i
«i
i
4
1 ,
^
1
5
"V
•**
6
1
x 02CNE DIFRJS£=S
iMPCUNCMefT
FIGURE 0-3
MULTIPLE-CHAMBER OZONE SYSTEM, CITY
OF TUCSON, ARIZONA, JOOST tt al. (1969)
-------
§?~i
iiSs
- sea
3.'.'
sr-r-
v ;
r
«• 3
u -
=*)
1
i
— £ *•
* i*i
a S^
i "3
-i j
3 I
i
2
k
I
>
J
x^ i ;
=i . !
*~ !|! M
^
_!ls = H
»2= 3 f|
^ '
/V,' 1 1 '
i.
.-i ri
'^1
^ *»
C S
33
e
UD
-------
INLET OF WATER
INLET OF GAS
DESCENDING
TUBES
EXTERNAL-
TUBING CUVE
OUTLET
OF GAS
OUTLET OF WATER
1
k*.
n
i
i
i
i
t
J .
«£
l_
FIGURE 0-5 - SCHEMATIC OF THE DEEP U-TUBE OZONE
CONTACTOR, ROUSTAN ft al. (1987)
-------
oc
ILI
X
H
UJ
U.
O
U
2
ui
Z
O
I
(0
I
e
iu
cc
3
O
(Z
-------
0.2 DETERMINATION OF CONTACT TIME (T)
0.2.1 Background
The hydraulic characteristics in ozone contactors range from an
almost Continuous Stirred-Tank Reactor (CSTR) to an ideal plug flow
configuration. Because the T10 approach may not be adequate for determin-
ing the inactivation.provided for systems resembling a CSTR, and because
the T10 approach is overly conservative in other cases, EPA recommends the
following three numerical methods to predict the contact time (T) in ozone
contactors:
T10: The T10 method discussed in Appendix C (and in Section 0.2.2) is
a good measure to characterize the contact time in most cases.
However, this method reduces the possibility of complying with the
SWTR for systems that have relatively high back-mixing and require
high inactivation levels.
Segregated Flow Analysis (SFA): (See Section 0.2.6) This is an alter-
native procedure to calculate the disinfection contact time. This
procedure is applicable only to systems that have good data from
tracer studies of high resolution as explained in Section 0.2.6.
CSTR: The Continuously Stirred-Tank Reactor (CSTR) method described
in Section 0.2.5, assumes the ozone contactor behaves as a CSTR.
This procedure is extremely conservative. However, no apparent
simplified analysis is currently available to make it less conserva-
tive. The-CSTR approach should be used only when:
Other predicting techniques are not recommended,
The required inactivation level is very low, or
Systems cannot afford to get good tracer study data for other
methods.
Systems may choose the optimal method for their situation based on
the available data to perform the calculations. A discussion of each is
presented below.
0.2.2 T10 Analysis
The simplest method of calculating the contact time, T, of microor-
ganisms in a contactor is by the T10 approach. T10 is defined as the
detention time to assure that 90 percent of the liquid that enters the
0.2-1
-------
contactor will remain at least T,0 minutes in the contactor. A system
achieving a CT10 corresponding to X percent inactivation, will assure that
90 percent of the water passing through the contactor is receiving at
least X percent inactivation, while 10 percent of the water will receive
less than X percent inactivation.
When conducting a step-input tracer study, T,0 is the 'time interval
required for the outlet tracer concentration to achieve 10 percent of its
ultimate response, following an inlet step addition. Appendix C of this
manual contains procedures to conduct and evaluate tracer studies for the
determination of T10. Appendix C also contains procedures to estimate the
T10 of contactors based on their baffling conditions and flow configura-
tion.
The results of tracer studies conducted on several ozone contactors
(Stolarik and Christie, 1990, Schwartz et al, 1990, Rosenbeck et al, 1989)
indicate that high quality tracer data on ozone contactors can be obtained
and that T10 can be estimated with high precision, but to a lesser degree
when T10 is less than one minute.
T,0 is a good measure of the contact time in most contactors and the
safety margin provided by using T10 compensates for the inferior perfor-
mance of contactors with a high degree of short-circuiting and backmixing
relative to contactors that approach plug flow conditions, (see Lev and
Regli, 1990a, for further detail.) However, for contactors with a high
degree of short-circuiting and a need to provide a high level of inactiva-
tion, this safety margin fails to compensate for the effect of backmixing.
In such cases, approximately 10 percent of the water passing through the
contactor receives significantly less than the inactivation indicated by
CT10. In these cases, either the SFA or the CSTR approach should be used
for determining the contact time.
The recommended alternatives for determining the contact time (T) for
various conditions of T10 versus hydraulic detention time (HOT) are
presented in Table 0-1. HOT is determined by dividing the liquid volume
of the contactor by the rate of flow through the contactor. As illustrat-
ed in this table:
0.2-2
-------
TABLE 0-1
Recomended Procedures to Calculate the
Disinfection Contact Time (T)
Condition: T10<(HDT)/3 T,0<(HDT)/3 T10>(HDT)/3
)(1) < 2.5 -Log(I/Ift)<1) >• 2.5
Notes:
Recommended
Methods: T«T10 T«T10
SFA(2) SFA(2) SFAC2)
CSTRC3) CSTR(3) CSTR<3>
1. Required level of inactivation in logs of either Giardja
lamblia cysts or viruses whichever value is greater;
I - # live organisms in outlet of ozone contactor and
I0 » # live organisms in inlet to ozone contactor
2. High resolution tracer characterization of the ozone contactor
be available.
3. The-CSTR method is extremely conservative and should be avoided
when alternative approaches are possible.
-------
The T10 method is applicable for systems that are required to
achieve less than a 2.5-log inactivation of Giardij cysts even
if the flow configuration in their ozone contactor approaches
that of a CSTR, such as disinfection in contactors using
turbine agitators.
Likewise, the T10 approach is appropriate for systems demon-
strating T.Q/HDT greater than 1/3 regardless of the required
level of disinfection.
Systems for which the T10 approach is appropriate to have the
option of applying either the SFA or CSTR analysis. The method
resulting in the highest T value, or thereby the lowest C value
may then be followed.
The SFA or CSTR should be used in lieu of T10 when the:
Level of inactivation required for Giardia cysts and/or viruses
is 2.5-log or higher
T10/HDT is less than 1/3.
Systems should be aware that the 2.5-log inactivation guideline refers to
the inactivation provided by the ozone system alone regardless of
inactivation provided by other disinfectants. For example, if a system
requires an overall inactivation of 3-log and provides 1-log inactivation
by chlorine, then a 2-log inactivation is required by ozone and the T10
approach can be used.
Examples for applying the different methods of calculation for T are
included in Section 0.2.8.
0.2.3 Additional Considerations for T10: Multiple Chamber Contactors
This section provides guidelines for computing T10 for several
contactors in series. The main shortcoming of the T10 approach is the
inherent non-linearity of this measure. In contrast to the HOT, which is
a linear measure, T10's of individual subunits do not sum up to give the
T10 of the overall unit. For example:
The HOT of two equal CSTRs in series is exactly twice the HOT
of each CSTR.
The T.p for the same two CSTRs in series is more than twice the
sum of the individual T10's.
0.2-3
-------
This raises some practical questions:
How should the T10 of a multiple-chamber contactor be deter-
mined using tracer studies?
Is it necessary to conduct individual tracer studies for each
chamber or is it sufficient to conduct an overall study of the
whole contactor?
How can the contact time of one chamber be determined based on
the T10 of the overall system?
Conducting tracer studies of individual chambers in a multiple
chamber ozone contactor is likely to be difficult. In addition, an
analysis conducted by Lev and Regli (1990a) indicates that the computation
of the contact time (T) based on tracer studies of the individual chambers
is likely to lead to over design. The excess volume of a system designed
by summing the T10s of the separate chambers may be up to 9.5 times higher
than one designed by the overall T10 approach. Therefore, EPA.recommends
the use of an overall tracer study of the whole contactor, in order to
lower operation costs and to avoid overly complex tracer studies.
Disinfection credits for a multiple chamber contactor should be based
only on the active chambers, those which have a detectable ozone residual.
Based on the recommendation to use overall tracer studies, guidelines are
needed for determining the disinfection credit for the active part of a
system based on overall tracer studies. The average concentration in the
individual chambers of a multiple-chamber system may deviate considerably
from one another. Therefore, systems must be able to assign contact
times for each chamber.
Lev and Regli, (1990a) evaluated the consequences of using a linear
approximation based on relative contact chamber volumes and overall T10 of
the contactor to determine the contact time of individual chambers in an
ozone contactor:
0.2-4
-------
Tv
Where:
TIO,eh«(*«r * An approximation for the contact time of one
chamber.
^10.total " T10 °f tne entire multi-chamber ozone contactor as
determined by tracer studies
VCh»iTto«r * Volume of the individual chamber
vtotai " Overall volume of the multi-chamber ozone conta-
ctor
They demonstrated that such linear extrapolation may lead to an underesti-
mate of the required T. This underestimate can be significant when the
concentration in the different chambers deviate considerably from each
other. This would be the case when the residual ozone concentration in
one chamber is zero.
Considering the various safety margins that are included in the T,0
approach, and considering the practical complexity involved in conducting
separate tracer studies, EPA recommends the use of the linear approxima-
tion described in Equation 1 provided that the volume of the portion of
the contactor that has zero residual ozone is less than half of the
overall volume o^ the ozone contactor:
V /V < 0 5
'inactivt chtwbtK 'total , u ' J
Where:
Vir»ctivt chwtwr * Tne volume °f the chambers in the contactor
where the ozone concentration is zero
V. . , - The volume of the chambers with a residual
totii
The following examples illustrate the computation of the overall
inactivation performance of multiple-chamber systems using the linear
approximation of Equation 1:
Example 0.2-1 Linear approximation to predict T
10
An ozone contactor has three chambers in series. Each chamber
has a volume of 353 cubic feet.
0.2-5
-------
The average ozone concentration in each chamber is:
First chamber: C,«0 mg/L ozone.
Second chamber: C2»l mg/L ozone.
Third chamber: C3 -0.5 mg/L ozone.
Cv C? and C3 are the average concentrations, determined as
described in Section 0.3.
The utility measured T10 - 5 min for the entire ozone con-
tactor.
The volumetric fraction of the chamber which has no ozone
residual is V1/(V,+V2+VJ)) - 0.33 which is less than the 0.5
guideline. Therefore it is permissible to use Equation 1 in
order to estimate the CT achieved in the ozone contactor.
The total CT achieved by the ozone contactor is:
CT - (C2)[(T10>tot-l)(V,)/(VMtil)] + (Cs)[(T10ftotil)(V,)/(Vtotil)]
CT - (1)[(5)(10)/(30)] + (0.5)[(5)(10)/(30)] • 2.5
The CT achieved by the ozone contactor is 2.5 mg-min/L.
Example 0.2-2 Linear approximation not applicable
An ozone contactor consists of:
A chamber with a volume of 70 cubic feet and equipped
with a turbine agitator
Followed by a second chamber with a volume of 200 cubic
feet.
The first chamber has an ozone residual of 0.5 mg/L and the
second chamber has an ozone residual of zero
The T10ftotil - 8 min for both chambers at the peak flow rate
The volumetric fraction of the chamber with no ozone residual
is 200/270 • 0.74 which is greater then 0.5 of the total
volume. Therefore, the use of Equation 1 to approximate the
T10 of the chamber that contains an ozone residual is not
recommended.
The system may estimate its performance by either the CSTR
approach taking into account only the detention time of the
first chamber or conduct tracer studies of the first chamber.
0.2-6
-------
0.2.4 Alternative Analysis of Disinfection Kinetics
The CSTR and the SFA approaches utilize the Chick-Watson inactivation
rule directly rather than relying on the CT approach. The following
section describes this alternative approach to represent the disinfection
kinetics.
The Guidance Manual recommends that systems should calculate the
inactivation level in their disinfection contactors by the CT approach.
Table 0-2 presents CT data corresponding to specified inactivation levels
of Gjardia cysts and viruses by ozone. An alternative way to present the
same information is by tables of the kinetic coefficients used to
calculate the CT values.
The CT values presented in Table 0-2 were calculated based on batch-
reactor experimental information that was fitted into a logarithmic
correlation according to a first order Chick-Watson's rule (Chick, 1907;
Watson 1908; and Hoff, 1987):
log(I/I0) - - k CT (2)
Where:
I/I0 - Survival ratio of the Giardia cysts or viruses
C - Residual concentration of ozone in mg/L
T » Exposure time in min.
k - A kinetic coefficient which characterizes the
specific rate of inactivation of the microorgan-
isms at the appropriate temperature and pH.
Solving Equation 2 for k yields:
k - -loo (I/IJ (3)
CT
Equation 3 can be used to calculate k values corresponding to the CT
values in Table 0-2. Table 0-3 summarizes these k values. Equation 3 may
also be used to transform inactivation levels (I/I0) to CT values and vice
versa.
The following example illustrates the use of the values presented in
Table 0-3 to calculate the performance of multiple-chamber ozone
contactors:
0.2-7
-------
Example 0.2-3 Multiple-chamber Ozone Contactor
An ozone contactor consists of three chambers in series
Temperature is 5 C.
The first chamber has a 10 percent survival ratio for Giardia
cysts, or (I/I0) - 0.1, which also corresponds to 90 percent
inactivation
The second chamber has an I/I0 » 0.07
The third chamber has an I/I0 « 0.03
The total inactivation may be calculated by either summing CT's
or summing logs of inactivation, as presented below.
Summing CT's:
At 5 C the k for Giardia cysts -1.58
<
The survival fractions are:
First Chamber - 0.1
Second Chamber - 0.07
Third Chamber - 0.03
Therefore, the CT values in each of the chambers are:
First chamber:
CT - -log(I/I0)/k - -log (0.1) /(I.58) • 0.63
Second chamber:
CT - -log (I/I0)/k - log (0.07)/1.58 - 0.73
Third chamber:
CT - -log(I/I0)/k - -log (0.03J/1.58 - 0.96
- Total CT is : 0.63 + 0.73 + 0.96 - 2.32
As indicated in Table 0-2, a CT of 2.32 is sufficient to
achieve a 3-log inactivation of Giardia cysts.
Summing logs of inactivation:
- First chamber: -log (I/I0) - -log(O.l) - 1
- Second chamber: -log(I/I0) * -log (0.07) - 1.15
0.2-8
-------
TABLE 0-2
CT VALUES FOR
INACTIVATION BY OZONE
Giardia
Tefficerature (C)
Inactlvation
0.5 log
1 log
1.5 log
2 log
2.5 log
3 log
Virus
Inactivation
2 log
3 log
4 log
0.48
0.97
1.5
1.9
2.4
2.9
0.9
1.4
1.8
0.32
0.63
0.95
1.3
1.6
1.9
0.6
0.9
1.2
0.23
0.48
0.72
0.95
1.2
1.4
0.5
0.8
1.0
0.16
0.32
0.48
0.63
0.79
0.95
0.3
0.5
0.6
0.12
0.24
0.36
0.48
0.60
0.72
0.25
0.4
0.5
0.08
0.16
0.24
0.32
0.40
0.48
0.15
0.25
0.3
-------
TABLE 0-3
k Values for Ozone Inactivation05
TEMPERATURE (C) 0.5 5 10 15 .20 25
Inactivation 1.03 1.58 2.08 3.12 4.17 6.25
of Gia-dia cysts
Inactivation 2.22 3.33 4.00 6.67 8.00 13.3
of Viruses
(1) k « -log(I/I0)/(CT) in L/mg-min. When Chick's rule is repre-
sented by the formula ln(I/L) - -K CT (In stands for the
natural logarithm) then k should be calculated by k * 2.303(K)
-------
Third chamber: -log(I/I0) - -log (0.03) « 1.52
The total logs of inactivation is:
-log (I/I0) - 1 + 1.15 + 1.52 - 3.67,
The 3.67-log inactivation of Giardia cysts is higher than
the required 3-log inactivation
0.2.5 Continuously Stirred-Tank Reactor (CSTR1 Approach
The CSTR method assumes that the flow configuration in the ozone
contactor approaches that of completely stirred reactor. In most cases,
this calculation method is the most conservative approach. Studies by
Schwartz et al (1990) suggest that well-operated turbine contactors
approach ideal CSTR characteristics and the CSTR calculation is appropri-
ate. In some cases, CSTR calculations offer the only apparent method to
evaluate the performance of the ozone contactors. CSTR calculations
should be used under the following conditions if systems have no other
means for demonstrating the inactivation efficiency.
Tracer data are not available,
The required inactivation level is greater than 2.5-log, and
ozone disinfection is applied in a single chamber contactor
with T10/HDT < 1/3.
If ejther the required inactivation level is less than 2.5-log
g_r T10/HDT > 1/3 then the inactivation predicted by CT1p is
appropriate provided that tracer data are available. If nigh
resolution tracer data are available then the SFA method can
be applied regardless of the level of inactivation required or
the ratio of T10/HDT.
In some cases, systems may actually receive more credit by using the
CSTR approach then by using the T10 approach. Higher credits result when
a low level of ozone disinfection such as 0.5-log is required and mixed
contactors are used.
When using the CSTR approach, the inactivation performance should be
evaluated for viruses and Giardia cysts, regardless of which required CT
is higher. This recommendation results from the influence of flow
characteristics on contactor performance, as discussed in Section 0.2.7.
The performance equation for a CSTR is based on two important
assumptions:
0.2-9
-------
1. The concentration of disinfectant and microorganisms is
homogeneously distributed in the contactor.
2. First order Chick-Watson's law applies. That is, the rate of
inactivation of the microorganisms is approximately proportion-
al to the concentration of the microorganisms and the concen-
tration of disinfectant.
The performance of a CSTR contact chamber is given by:
(I/I0) - 1/[1 t 2.303(k)C(HDT)] (4)
Where:
k - kinetic coefficient for microorganism inactivation
[k values are listed in Table 0-4 (L/mg-min)]
(I/I0) - Survival ratio of organisms
C » Average concentration of disinfectant (mg/L)
HOT « Hydraulic detention time (min)
Equation 4 may also be used to calculate the ozone concentration that
is required to achieve a specified level of inactivation for a given HOT
or to compute the HOT required to achieve a desired inactivation level for
a given ozone concentration. Equation 5 restates Equation 4 for use in
determining C or HOT
C(HDT) - [1-(I/I0)]/[2.303 k (I/I0)] (5)
The effects of mixing on improving disinfection effectiveness may be
very significant in CSTR contactors, and are not accounted for in this
model.
Examples demonstrating how to calculate the operating conditions
necessary to meet the required inactivation levels by the CSTR approach
are included in Section 0.2.8.2.
0.2-.6 Segregated Flow Analysis (SFA1
SFA is a method that is often used to characterize chemical
reactions. Better approximations may be determined through analysis and
modelling of the specific details of the flow pattern in the ozone
0.2-10
-------
contactor, but such modelling cannot be done based on tracer studies
alone, as the SFA can. Comprehensive descriptions of the SFA can be found
in several references including Levenspiel (1972) and Seinfeld and Lapidus
(1984). The SFA assumes that the inactivation in a contactor'can be
determined by the product of the probabilities of two events: the
probability distribution for water to remain in the contactor; and the
probability distribution for organisms to survive as they pass through the
contactor.
The first probability function describes the chances of a microor-
ganism remaining in the contactor for a specified time period. The water
passing through the contactor has a probability distribution, determined
by tracer studies which indicate the detention time for each fraction of
the flow through the contactor.
The second probability function describes the chances of a microbio-
logical species surviving following exposure to a disinfectant for a
certain amount of time. This probability function is given by the
modified Chick's equation: (I/I0) » 10""ct. Each fraction of the flow would
have a different "t" for which this equation would apply. For example, a
virus that is exposed for 1 minute to C-l mg/L ozone when k-1 L/mg-min has
0.1 (10 percent) chances to survive.
The following illustrates the intuitive origin of the SFA approach:
The flow in an imaginary contactor may be viewed by flow lines.
A microorganism that is introduced at time t«0 will follow one
of these flow lines.
For simplicity, consider that only four flow lines exist as
represented on Figure 0-7.
A microorganism that is introduced in the feed to the contactor
has some probability (PI) of following any one of these four
lines.
The microorganism will then remain for a specific detention
time, characteristic of each flow line, in the contactor.
This concept is presented schematically on Figure 0-7, where
the flow lines are represented by four different tubes whose
lengths (or detention times) correspond to the lengths of the
flow lines on Figure 0-7.
0.2-11
-------
Microorganisms that are introduced into various tubes have
different probabilities of survival (P2), because of different
susceptibilities to disinfection in each of these tubes.
The product of the probability that a microorganism will be
carried into a specific tube (PI) times the probability of
survival after being exposed to the hostile environment for the
appropriate time (P2) is the probability that a microorganism
introduced into the feed inlet will get out alive from a
specific tube (P1)(P2).
For example, if 20 percent of the flow is directed into the first
flowline and I/I0 for this fraction of the flow equals 0.25, a microorgan-
ism has (0.2)(0.25) or a 5 percent chance of emerging alive from this
specific flowline. The total survival of microorganisms that are
introduced into the inlet to the entire contactor can be computed by
summing up all four survival probabilities (P1)(P2).
Complete examples for the application of the SFA are included in
Section 0.2.8.3. For SFA to be applied, a high resolution tracer study
must be available. The requirements for a high resolution tracer study
are:
Appropriate time distribution of sampling points.
Limited degree of scatter in sample points.
The first requirement is to have several sample points prior to the
occurrence of T10 and less frequent sampling points thereafter. Several
sampling points~pnor to T10 are essential to get in accurate representa-
tion of what is occurring in the early flow through the contactor, when
organisms are most likely to exit the contactor while still viable. The
second requirement if for a limited degree of scatter between the sample
points. The plotted curve should ideally be continuous to allow for more
accurate integration to predict the survival of microorganisms.
0.2.7 Relative Inactivation of Giardia Cvsts and Viruses
In most cases, when the CT required for the inactivation of Giardia
cysts is greater than the CT required for the inactivation of viruses,
compliance with the inactivation requirements for Giardia cysts will
0.2-12
-------
A. FLOW LINES IN A CONTACTOR
INPUT
OUTPUT
B. SCHEMATIC REPRESENTATION OF THE FLOW LINES
C. SURVIVAL PROBABILITY FOR AN ORGANISM
Flow
Line
1
2
3
4
p,
Fraction of Flow
into the Flow! ine
2/10
4/10
1/10
3/10
d/50)
Survival Ratios
1/4
1/8
1/16
1/32
p p
12
Overall
Survival Ratios
16/320
16/320
2/320
2/320
SUM 1
37/320
FIGURE 0-7 - PRINCIPLE OF SEGREGATED FLOW ANALYSIS
-------
assure compliance with the virus inactivation requirements. Specifically,
this is true when:
kv,rus/kcyst > 1 °9( I/ I0) virus/ l<>g( I/I0)cygt (6)
The SWTR, however, requires a higher level of inactivation of viruses
than Giardia cysts. Therefore, ozone contactors that are characterized by
a high degree of turbulence will find that, as the flow configuration
approaches that of a CSTR (T10/HDT < 1/3), compliance with the virus
inactivation requirements may be a more demanding task than meeting the
inactivation requirements for Giardia cysts. Consequently, an ozone
contactor that has a T10/HDT < 1/3 and a low (I/I0) should be checked for
compliance with the inactivation of viruses as well as for cysts. Another
way to understand this is that as the inactivation indicated by CT10
increases, the 10 percent of the water passing through the contactor with
less contact time than T10 becomes more significant for lowering the
overall inactivation efficiency for all the water passing through the
contactor.
0.2.8 Examples of Determining Contact Time (T)
This section presents examples for 'the application of the three
general approaches - T10, SFA, and CSTR - for determining contact time.
0.2.8.1 Evaluation Using T,0
The following four examples illustrate when the T10 approach should
be used and when alternate approaches are appropriate. Procedures for
calculating the required ozone residual based on the T10 approach are
outlined in the examples.
Example 0.2-4 Inactivation Required >2.5-1oq
The Haworth Water Treatment Plant, Hackensack, New Jersey, uses a
turbine ozone chamber followed by a contact chamber to provide additional
contact time. A schematic of the contactor is shown on Figure 0-1. The
treatment plant provides filtration after the ozone contactor. For the
purposes of this example, although it is not the case for Hackensack, the
0.2-13
-------
ozone system must provide disinfection for 2-log Giardia and 3-log virus
inactivation to supplement filtration. The following conditions apply:
Water Temperature • 0.5 C
CT for 2-log Giardia • 1.9 mg-min/L
CT for 3-log virus - 1.4 mg-min/L
A tracer study was conducted on one of the four ozone contact-
ors. Figure 0-8 depicts the chart recorder of the raw data
that were collected during the tracer study.
The HOT at the flow rate of the study was 20 minutes, and the
T10 occurs at 11 min.
The T10/HDT of 0.55, is greater than 1/3, making the T10
approach valid for this system.
The CT for Giardia inactivation is the controlling CT because
it is greater than the CT for virus inactivation.
Using the T10 of 11 min, the residual needed to meet the CT
requirement of 1.9 mg/L-min is determined as follows:
C " 1.9 ^q-min/L - 0.17 mg/L
11 min
As a result of using the T.0 approach, the system must maintain
an ozone concentration of 6.17 mg/L in the contactor to provide
the necessary disinfection.
The application of the SFA method for this contactor is presented in
Section 0.2.8.3.^
Example 0.2-5 Low Detention Time. Inactivation Required <2.5-1og
A system using slow sand filtration must provide disinfection for 1-
log Giardia cyst and 2-log virus inactivation. The system has an ozone
contactor equipped with a turbine mixer. The following conditions apply:
Water Temperature « 25 C
CT for 1-log Giardia • 0.16 mg-min/L
CT for 2-log virus « 0.15 mg-min/L
The CT for Giardia cyst inactivation is greater than the CT for
virus inactivation and is therefore the controlling CT.
A tracer study was conducted for the ozone contactor and
resulted in a T10 of 30 seconds.
C.2-14
-------
The HOT of the contactor at the flow rate of the study was 150
seconds.
Thus T10/HDT - 30/150 - 0.2, is less than 1/3, however, because
the required inactivation is less than 2.5-log, the T10
evaluation for this system is appropriate.
Based on the T10 evaluation, the residual needed to meet the CT
requirement is determined as follows:
CT « 0.16 mg-min/L
C - 0.16 mq-min/L - 0.32 mg/L
0.5 min
Thus, according to this approach, the system must provide an
ozone concentration of 0.32 mg/L to meet the inactivation
requirements.
Because of the low T10/HDT value for this system, the CSTR
approach is an alternative for determining C. This example is
presented in Section 0.2.8.2.
Example 0.2-6 Low Detention Time. Inactivation Required >2.5-1oq
An unfiltered water system must provide disinfection for a 4-log
inactivation of viruses and a 3-log inactivation of Giardia cysts. The
ozone system uses a single chamber turbine contactor for disinfection:
The hydraulic detention time measured at peak flow rate is 30
minutes and T10 determined by a tracer study is 9 minutes.
The Lp approach is not recommended for this system because
T10/HDT of 0.3 is less than 1/3 and the required level of 4-log
virus inactivation is higher than the 2.5-log level.
SFA or the more conservative CSTR calculations may be used to
determine the required ozone concentration for this system.
Examples of the CSTR and SFA calculations are presented in
Sections 0.2.8.2 and 0.2.8.3, respectively.
Example 0.2-7 High Detention Time. Inactivation Required <2.5-1og
The Sturgeon Bay Water Treatment system (Rosenbeck, 1989) uses a
series of two submerged turbine ozone contactors followed by a reactive
chamber to disinfect ground water:
The results of a tracer study conducted on one of the mixed
contactors is shown on Figure 0-9.
The T,, from this study is approximately 30 seconds while the
hydraulic detention time is 62 seconds.
0.2-15 / .
i
-------
30/62 - 0.48 which is greater than 1/3. Therefore,
e T10 approach is appropriate for this system.
In this case, the SFA method is not recommended as an alternative to
the T10 approach because of the minimal detention times in the contactor.
With such a short period for the collection of samples, the data are
insufficient for the SFA method. The resolution of the tracer studies,
apparent in Figure 0-9, will lead to an overly conservative estimate of
the inactivation if differentiation is conducted by a forward algorithm.
0.2.8.2 Evaluations Using CSTR Calculations
The following two examples demonstrate the CSTR approach. One illus-
trates the benefit of the CSTR analysis over the T10 analysis. The other
identifies conditions for which the CSTR approach is not practical.
Example 0.2-8 Low Detention Time. Inactivation Required <2.5-log
The system identified in Example 0.2-5 is a slow sand filtration
plant, using ozone to provide for a 1-log Giardia cyst inactivation.
Chlorine provides the 2-log virus inactivation. Because the level of
inactivation required from ozone disinfection is less than <2.5-log, the
system may choose any method for the determination of the contact time.
A tracer study conducted on the ozone contactor resulted in a
T10 of 30 sec for a HOT of 150 sec.
The fraction of T10/HDT is 0.2, which is less than 1/3,
indicating that the CSTR approach may be appropriate.
Chlorine provides disinfection for the viruses, therefore the
CSTR calculation for the ozone disinfection requirements will
be based on Giardia cyst inactivation.
The following conditions apply:
Water Temperature - 25 C
CT for 1-log Giardia cyst - 0.16 mg-min/L
- Equation 5 from Section 0.2.5 applies for the CSTR calculation:
C(HDT) - [1 - (I/I0)]/[(2.303)k
0.2-16
-------
-I O
-------
UJ
O
N -^
O o>
to
52
CD *-
2°
UJ m
3 UJ
H W
WQ
U.CC
°
so
UJ<
OH
I
O
UJ
E
(%) n/ow Nouv«iN30NOO soiaonid
-------
The parameters are determined as follows:
1. From Table 0-3, keym - 6.25 for CT - 0.16 min/L at 25c
2. For 1-log inactivation, I/I0 » 0.1
3. HOT - 150 sec or 2.5 min
C is determined as follows:
C(HDT) - (0.9]/[(2.303)(6.25) (0.1)] - 0.625 mg-min/l
C» 0.625/2.5 » 0.25 mg/L
Thus, according to the CSTR approach, the system must provide an
ozone concentration of 0.25 mg/L to meet the inactivation requirements.
For this case, the system would prefer to use the CSTR approach rather
than the T10 approach since the T10 approach would require a 0.32 mg/L
ozone residual, as shown in Example 0.2-5.
"Example 0.2-9 Low Detention Time. Inactivation Required >2.5-log
An unfiltered water system must provide disinfection for 4-log
inactivation of viruses and 3-log inactivation of Giardia cysts. The
system uses a single chamber turbine ozone contactor. Hydraulic detention
time measured at peak flow rate is 30 minutes and T10 determined by tracer
studies is 9 minutes. T10/HDT is less than 1/3 and greater than 2.5-log
inactivation is-required, therefore the T10 approach should not be used.
The CSTR or SFA methods 'are appropriate.
The CSTR calculation must be conducted for both Giardia cysts
and viruses to determine the controlling parameter
Compute the C required for inactivation of Giardia cysts:
- k cysts - 6.25 (Table 0-3).
For 3-log inactivation, I/I0 • 0.001
Using the CSTR equation:
C(HDT) - [1-0.001]/[2.303(6.25)(0.001)] • 69.5 mg-min/L
C - (69.5 mg-min/L)/(30 min) • 2.3 mg/L
0.2-17
-------
Compute the required C for inactivation of viruses:
- *v,vu. • 13-3 (Table 0-3)
For 4-log inactivation I/I0 - 0.001
Applying the CSTR Equation:
C(HDT) - [1 - 0.0001]/[(2.303) (6.25) (0.0001)] - 326 mg-min/L
C - (326 mg-min/l)/(30 min) - 10.8 mg/L
As indicated, virus inactivation is the controlling parameter,
requiring a C of 10.8 mg/L. Because of the higher ozone residual needed
for the virus inactivation, this example illustrates why systems should
verify compliance with the inactivation requirements for viruses as well
as for the inactivation requirements for Giardia cysts. Since obtaining
an ozone residual of 10.8 mg/L is unrealistic, this example illustrates
how stringent disinfection conditions can become assuming CSTR character-
istics. Consequently, the SFA would result in a more feasible residual
requirement for this system.
0.2.8.3 Evaluations Using SFA
The SFA method can be conducted on spread sheets. Table 0-4 presents
the calculation procedure in spread sheet notations for a step tracer
input:
The first column of Table 0-4 represents the sequential
numbering of consecutive tracer study measurements or digital
measurement points fed into the computer.
The second column represents the time interval that elapsed
between the step change in tracer concentration and the
sampling of the specific tracer point.
The third column represents the tracer effluent concentration
at a point in time determined by the analyzer (spectropho-
tometer conductivity meter, etc.) reading.
The fourth column represents the tracer response on a scale of
0-1, where 0 corresponds to background reading of the analyzer
and 1 to ultimate response after a long time interval. In
other words, it is C /Cjn where C^, is the tracer concentra-
tion in the outlet of the contactor and Cin is the basal me
tracer concentration in the inlet.
0.2-18
-------
The fifth column represents the forward derivative of the
F(t) response. It is the slope of the tracer curve at a
specific time interval, or the rate at which C^/C, changes
with respect to time at different intervals in time. Note that
by forward evaluation of the derivative: E(t) « [F(t+dt)-
F(t)]/dt the E(t) curve Is shifted by half a dt toward the
origin.
This method of differentiation introduces an inherent safety
margin to the calculation. Systems can reduce this safety
margin by collecting more tracer points at the initial period
of the tracer response, when the response is starting to
increase.
This period has the largest effect on the accuracy of the
tracer analysis because most of the contribution to the total
survival of microorganisms comes from the organisms that remain
only for short time interval in the contactor.
The sixth column represents Chick's inactivation rule, computed
at the concentration and the appropriate 10"kct.
The seventh column represents the survival expectancy function
(Es(t) - E(t)(10"kct) which is the product of columns 5 and 6.
The eighth column represents the organism survival in each
segment passing through the contactor. It is also known as the
integral of the survival expectancy function (Es presented in
the 7th column).
The survival ratio (I/I0) is the sum of column 8. This
represents the sum of organism survival in all the water
segments passing through the contactor.
Table 0-4 Illustrates only one form of performing the integra-
tion (i.e., quadratic integration). Other integration methods
can also be used.
The corresponding log inactivation and the corresponding
calculated CT may be computed by the procedures outlined in
Section 0.2.4.
The following examples illustrate the use of the SFA method to
calculate conditions in ozone contactors, and a situation where SFA cannot
be used.
Example 0.2-10 Turbine Contactor
As noted in Example 0.2-4, the ozone system at Haworth Water
Treatment Plant, uses a turbine ozone chamber followed by a reactive
0.2-19
-------
§
vll -f -f
o o
3 ?::
o
_J
u.
O
01 o o
. . o
3 1- > .- J>
a.
UJ
O
X
U.
UJ
>
QC
3
U
s
UJ
- a
i
00 —
y *
U
< u
»» 0.
OS
"-co
<5
l!
u
K
O
o
UJ
cc
o
UJ
U. ff
^J
o g
2 <
O (A
< >
UJ ^
-------
chamber to provide additional contact time. A tracer study was conducted
on one of the contactors resulting in a T10 value of 11 minutes for a HOT
of 20 minutes. Using the same conditions as the above cited example, the
SFA will be conducted on the tracer data. The following illustrates a
step by step procedure for conducting a SFA:
The digitized tracer response (F(i)) is depicted in Figure 0-10
as a function of t(i) where:
i stands for the consecutive numbering of randomly chosen
points from the tracer study chart, and
t(i) is the corresponding time coordinate.
The slope of the tracer curve, also known as the density of the
expectancy function, E(t) approximated by the following
equation is depicted in Figure 0-10.
The digitized points were not translated into a smooth curve
in order to avoid numeric compromises.
The survival expectancy (Es(t)) was then calculated by Es(i)=Et
(i)(10'kct0)) and summed to give the survival ratio (I/I0) as
shown in Table 0-5.
Figure 0-11 depicts the integration for conditions where the
ozone residual is C - 0.15 mg/l.
The cumulative survival ratio is 0.00982 which is below 0.01
assuring compliance with the 2-log or 99 percent inactivation
requirement for Giardia cysts. A survival ratio of <0.01
corresponds to an inactivation of greater than 99 percent or
2-log.
The residual value determined from this method .is lower than O0.17
mg/l predicted by the T10 approach presented in Example 0.2-4. Although
this example only shows a small difference in C values needed, other cases
may result in a greater reduction of C compared to the C resulting from
the T1Q approach.
0.2-20
-------
Z iu
-) \u
u. a:
°s
5<
Ss
UJ A
i
i
(I)J '01 (*1)3
-------
TABLE 0-5
Segregated Flow Analysis
of an Ozone Disinfection Contactor at Hackensack
time
(min)
0
1
2
3
4
5
6
6.5
7
8
9
10
11
12
13
14
15
16
17
18
19
21
23
25
27
34
36
41
45
height
(mm)
0.0
0.0
0.0
0.0
0.0
0.0
0.3
0.5
2.0
3.0
5.0
8.0
12.0
19.0
26.0
31.0
36.0
42.0
51.0
59.0
69.0
80.0
90.0
94.0
98.0
113.0
114.0
124.0
124.0
F(t)
0.000 •
0.000
0.000
0.000
0.000
0.000
0.002
0.004
0.016
0.024
0.040
0.065
0.097
0.153
0.210
0.250
0.290
0.339
0.411
0.476
0.556
0.645
0,726
0.758
0.790
0.911
0.919
1.000
1.0000
E(t)
0.000
0.000
0.000
0.000
0.000
0.002
0.014
0.013
0.008
0.016
0.024
0.032
0.056
0.056
0.040
0.040
0.048
0.073
0.065
0.081
0.044
0.040
0.016
0.016
0.017
0.004
0.016
0.000
0.000
10'kct
(C-0.16
k-1.03)
1.000
0.708
0.502
0.355
0.252
0.178
0.126
0.106
0.089
0.063
0.045
0.032
0.022
0.016
0.011
0.008
0.006
0.004
0.003
0.002
0.001
0.001
0.000
0.000
0.000
0.000
0.000
0.000
0.000
Es-
-E(t)10'kct
0.00000
0.00000
0.00000
0.00000
0.00000
0.00043
0.00173
0.00143
0.00072
0.00102
0.00108
0.00102
0.00127
0.00090
0.00045
0.00032
0.00029
0.00027
0.00018
0.00016
0.00006
0.00003
0.00001
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
(Es)At
0.00000
0.00000
0.00000
0.00000
0.00000
0.00013
0.00035
0.00215
0.00072
0.00204
0.00324
0.00408
0.00889
0.00630
0.00225
0.00160
0.00174
0.00243
0.00144
0.00160
0.00066
0.00030
0.00004
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00982
-------
a
>
_J
<
Cfl
r
i
UJ
(3
!u
QC
o
at
10
! a
a
M
49
UJ
X
o
o
r ?! -
I
(A
O
a
a
o:
C
q
ci
001 x ip U)S3
-------
0.2.9 ESTIMATING T
The results of this section are summarized in Figure 0-12. The
decision tree shows the applicable methods of estimating T for each
approach, and provides a quick means to compare alternatives and make a
selection.
0.2-21
-------
C3
Z
(ft
UJ
u
LU
CE
O
UJ
U
CE
H-
Z
UJ
Q
CM
i
O
UJ
CC
D
o
II
-------
0.3 DETERMINATION OF OZONE CONCENTRATION (C)
0.3.1 Introduction
This section presents ways to measure or estimate the ozone
concentration, C, for the calculation of CT. An alternative, more
elaborate concept, requiring better characterization of the hydrodynamics
of the ozone contactor is presented in Section 0.4 of this appendix.
EPA recommends use of the average dissolved ozone concentration in
the water for C for all types of ozone contactors. The average concentra-
tion may be determined by one of following methods:
1. Direct measurement of the concentration profile of dissolved
ozone in each contact chamber
2. Indirect prediction of the average concentration by assuming
a set of conservative correlations between an observed variable
such as the concentration of ozone in the outlet from the ozone
chamber and the average concentration within the ozone chamber.
The application of these methods to estimate the average concentra-
tion should take into account the gas/liquid flow configuration in the
ozone contactor. The next section presents a short discussion of the
types of liquid/gas contact in ozone chambers, followed by two sections
that describe the methods to estimate the average concentration in the
chamber based on simple measurements.
Classification of Ozone Chambers
Ozone contactors currently in use or in design stage in the US may
be classified into four types of flow configurations as illustrated on
Figure 0-13. This, of course, does not preclude the use of other types of
contactors. The four configurations are as follows:
1. Continuously Stirred-Tank Reactor (CSTR):
Ozone contactors using turbine agitators, where the water may
be considered uniformly mixed as shown on Figure 0-13, diagram
1. Studies conducted in a full scale turbine contact chamber
indicate that turbine contactors may be considered uniformly
0.3-1
-------
mixed (Schwartz et, al., 1990). This study was conducted in the
first contact chamber under conditions of high ozone demand.
Therefore, it is assumed that under less stringent kinetic
conditions, turbine contactors can still be considered uniform-
ly mixed.
2. Counter-Current Flow Chambers
In these chambers, the water flows opposite the direction of
the gas bubbles. For example, the first and third chambers in
the Los-Angeles ozone treatment system, as shown on Figure 0-2.
3. Co-Current plow Chambers
In these chambers, the gas bubbles and the water flow in the
same direction. For example, the Deep U-Tube contactor shown
in Figure 0-5 and the Static Mixer contactor. This is the case
also for the conventional gas/liquid contact chambers such as
the second contact chamber in the configuration designed for
the East Bay MUD water disinfection system, as shown on Figure
0-4.
4. Reactive Flow Chambers
In these chambers, no gas (and ozone) is being introduced into
the chamber or conduit. The second and fourth chambers of the
Los Angeles water disinfection system are reactive chambers
(Figure 0-2).
0.3.2 Direct Measurement of C
Direct measurement of the dissolved ozone concentration is the
preferred method to determine the ozone concentration in ozone contact
chambers. However, very little full scale experience is currently
available with this type of measurement. Some guidelines were developed
based on the limited studies conducted at the Haworth, NJ (Schwartz et al.
1990) and Los Angeles water treatment systems (Stolarik and Christie,
1990). The guidelines developed for direct measurement of ozone
concentration in the liquid phase are detailed in the following sections.
Analyze Each Chamber Separately
Every chamber of a multiple-chamber unit should be analyzed separate-
ly. Different chambers in series exhibit different ozone consumption rates
and reactivities and, therefore, are likely to have different dissolved
ozone profiles.
0.3-2
-------
1 TURBINE CHAMBER
2 COUNTER-CURRENT 4-
CHAMBER
J L
3 CO-CURRENT CHAMBER
4. REACTIVE FLOW
CHAMBER
L: Liquid
G: Ga*
FIGURE 0 13- FLOW CONFIGURATIONS IN OZONE CONTACTOR CHAMBERS
-------
Avoid Interference From Gas Bubbles
Gas bubbles may strongly interfere with the measurement of ozone
concentration, particularly if some bubbles are carried into the sampling
taps. This interference may be reduced by directing the sampling port
opposite to the direction of the bubble flow in order to prevent gas from
entering the sampling tube. Additionally, the operator should verify, by
visual inspection, that the sample water does not contain gas bubbles.
Systems using in-situ ozone analyzers should be careful to prevent
direct contact of gas bubbles with the measuring probe which is usually a
gas permeable membrane. Such contact may bias the measurements and give
high results.
Minimize Distance to Ozone Analyzers
Minimize the distance from the sampling ports to the ozone analyzer
to limit ozone consumption by reducing agents in the water. This
consideration is particularly important when evaluating the concentration
profile in chambers with high ozone demand such as the first chamber in
multiple-chamber units.
Provide Proper Soacial Distribution
The vertical profile of the ozone concentration in ozone contact
chambers should be measured in at least five vertical locations and at
least two different horizontal locations for each vertical sampling point
within the contact chamber. Each sample should represent the time
averaged concentration at the specific location. This may be achieved by
sampling a large volume of water into a container and analyzing the water
by the indigo trisulfonate method (Bader and Hoigne, 1982). In-situ
measurement of ozone should be carried out over a sufficient time interval
to suppress temporal fluctuations. Such instruments should be initially
calibrated by the indigo trisulfonate method. Facilities that have more
than 25 percent deviation between the average concentration at two
horizontal locations should collect additional measurements at a third
location. The average of all measurements may be taken as the average
concentration of dissolved ozone in the ozone contact chamber. For
systems with a symmetrical vertical distribution of ozone concentration
0.3-3
-------
the vertical sampling points should be equidistant. Systems with an
asymmetrical distribution of available sampling points can perform an
integration of the data to estimate the average concentration in the
chamber. An example of this is given at the end of this section.
Some contact chambers, such as the Deep U-Tube chambers, static
mixers and reactive flow chambers have a high length to width ratio, where
the length of the chamber in the direction of fluid flow is greater than
four times the cross section length. These chambers have more uniform
radial concentration profiles, eliminating the need to measure the
concentration at various vertical or horizontal locations. Therefore,
measuring the concentration profile at several points along the flow path
should be sufficient to accurately determine the average concentration.
Select Representative Locations
All sampling positions should be placed in representative locations,
avoiding stagnant zones and zones near the wall. Measurements in stagnant
locations will lead to low values of the average residual concentrations.
While measurements at the wall may result in either an underestimate or
overestimate of the residual depending on the ozone flow pattern.
Systems having two or more identical parallel ozone contact chambers
may determine the average ozone concentration by measuring the concentra-
tion profile at one horizontal location in each contact chamber. These
systems should, however, show by dual or triple horizontal measurements in
at least one of-the parallel chambers that the measurement in the
particular horizontal location adequately represents the concentration
profile in the contact chamber.
Example 0.3-1
A system with a co-current chamber with dimensions of 10' X 10' x 20'
was sampled to determine the average concentration in the chamber. In
accordance with the recommended guidelines, the following samples were
taken:
0.3-4
-------
Water Ozone Residual (mg/t)
Depth (ft) H. H,
2 0.1 0.12
6 0.15 0.17
10 0.15 0.14
14 0.3 0.25
18 0.6 0.65
The horizontal sampling point measurements are within 25
percent of each other indicating that no additional horizontal
sampling' is needed. Figure 0-14a shows the sampling locations
and the resulting ozone profile.
Average the H, and H, sampling points to determine C :
Cava " (0-1 + 0-15 + 0.15 + 0.3 + 0.6 + 0.12 + 0.17 + 1).14 +
O.Z5 + 0.65)/10 = 0.26. Cavg equals 0.26 mg/L, which is C for
the chamber.
Example 0.3-2
A system with a co-current chamber and the same dimensions of the
system in Example 0.3-1 has sampling results as follows:
Water Ozone Residual (mq/L)
Depth (ft)
2
8
14
16
18
H.
0.1
0.16
0.27
0.70
0.52
H,
0.12
0.14
0.3
0.73
0.61
Averaae10
0.11
0.15
0.285
0.715
0.615
(1)
Average ~ (H, + H2)/2
The sampling points are not vertically equidistant so the
system will plot the average ozone concentration of the
horizontal sampling points versus depth to calculate the area
under the curve. This approach should only be used if the
sampling points cover the range of the water depth.
As shown on Figure 0-14b, the area under the curve is deter-
mined for the range of depths sampled from 2 to 18 ft.
Several methods can be used for calculating the area includ-
ing:
Measurement with a planimeter
Mathematical methods such as:
Simson's Rule
Runge Kutta
0.3-5
-------
UJ
O
B
0.
Uj
z
O
N
O
I z
o O
*• P
UJ
o
z
O
O
UJ
z
O
N
O
UJ
Z
O
N
O
u
O
DC
ID
O
QC
O
H31VM
(II) HldSQ U31VM
UJ
S
UJ
ff
^ cv
s s
2 CH
3 z
a.
z
_j
0.
1
<
en
en
O
5
en
O
U
o
(9
Z
3
Ok
<
(U) HJldBO b31V*
Hid30 U3J.VM
O
Z
to
O
UJ
UJ
-------
The area under the curve is in units of mg/L-ft. C is deter-
mined as:
area (mq/L-ft)
range of depth sampled (ft)
For this data, use of a planimeter results in an area of 5.44
mg/L-ft, with the concentration determined as follows:
5.44 mq/L - ft * 0.34 mg/L
18 ft - 2 ft
0.3.3 Estimating C Based on Residual Measurements at the Outlet
For many systems, measuring ozone profiles in their ozone chamber may
be impractical because of physical constraints. These systems may
estimate C in the chamber based on measurements of the ozone residual at
the outlet from the chamber. EPA has established correlations for
different types of gas-liquid contact configurations currently in use in
ozone contactors. These relationships were derived based on conservative
assumptions regarding the type of flow configuration in the contactor.
Due to the highly reactive nature of ozone the values for C vary slightly
between first chambers and subsequent chambers. The recommended
concentrations for first and subsequent chambers are summarized in Table
0-6.
0.3.3.1 First Chambers
A first chamber is the chamber in which ozone is initially intro-
duced. In establishing guidelines for determining C values for the first
ozone contact chamber, the following items were considered:
1. The relationship between C and the outlet concentration in the
first chamber of a multiple-chamber system (or single chamber)
may be very sensitive to the reaction order of the ozone
consumption kinetics.
The average concentration in the contactor may be less than 10
percent of the outlet concentration. This was demonstrated in
pilot plant studies conducted in a multiple chamber system by
Stolarik and Christie, 1990. Therefore, general relationships
between the residual ozone concentration at the outlet from a
first (or single) ozone contact chamber and the average
concentration in this chamber cannot be developed.
0.3-6
-------
TABLE 0-6
CORRELATIONS TO PREDICT C BASED
ON OUTLET OZONE CONCENTRATIONS'1-3)
FLOW CONFIGURATION
CO-CURRENT
TURBINE FLOW
First Chamber
C PARTIAL'2'
CREDIT
Subseauent Chambers
C • Cout C • Cout
or
COUNTER-CURRENT
FLOW
PARTIAL(2)
CREDIT
C - C^/2
REACTIVE
FLOW
NOT
APPLICABLE
c - cout
Cfn)/2
1. Definitions:
C Characteristic Concentration (mg/L)
CQLJt Dissolved ozone concentration at the outlet from the chamber (mg/L)
C
Concentration of ozone at the inlet to t » chamber (mg/L)
2. 1-log of virus inactivation providing that Cout > 1 mg/L and 1/2-log Giardia
cysts inactivation providing that CM > 0.3 mg/L.
3. Alternatively, C may equal the average concentration as evaluated by the direct
measurement method (Section 0.3.2).
-------
2. The rate of disinfection of viruses (coliphage) by ozone often
decreases with respect to contact time whereby the initial
inactivation rate is very fast and deteriorates afterwards.
3. Pilot plant experiments reported by Wolfe et al, (1989)
suggest that the inactivation of organisms including MS2
bacteriophages, Giardia muris cysts, R2A bacteria and E. Coli,
in the first chamber of a multiple-chamber reactor is very
rapid even when high ozone demand waters are used.
Considering these, items, EPA recommends a general guideline of
crediting the first ozone chamber with CT credits equivalent to 1-log
virus inactivation and 0.5-log Giardia cyst inactivation, provided that
the residual concentration measured at the outlet from the first contact
chamber exceeds 0.1 mg/L and 0.3 mg/L, respectively, regardless of the
contactor configuration. However, this guideline does assume that the
volume of the first chamber is equal to the volume of subsequent chambers.
The credit for 1-log virus inactivation at an outlet residual of 0.1 mg/L
may appear conservative with respect to MS2 bacteriophage data, however.
only limited data for ozone inactivation of the animal viruses of concern
is currently available. Preliminary test results indicate that bacterio-
phage may not be an appropriate indicator for virus inactivation by ozone
(Finch, 1990).
Systems may prove higher performance of their first contact chambers
by measuring the concentration profiles in the first chamber, as outlined
in Section 0.3.2t>r by applying the more sophisticated methods that are
presented in Section 0.4.
0.3.3.2 Subsequent Chambers
The correlations in Table 0-6 are based on analysis of the dissolved
concentration profile in liquid/gas contacting chambers. All correlations
rely on the accurate measurement of ozone concentration outside of the
gas/liquid contacting regime. Concentrations at the outlet from the ozone
contact chambers can be measured accurately without interferences from the
ozone bubbles. The correlations represent the highest possible estimate
of C that can be supported without site-specific test data. These
estimates are conservative and systems may choose to determine C based on
0.3-7
-------
direct measurement of the concentration profile in the contact chamber, or
use one of the procedures recommended in Section 0.4.
Correlations were developed for the four types of flow configura-
tions:
Turbine
Counter Current Flow
Co-current Flow
Reactive Flow
Turbine
For turbine chambers or rigorously mixed chambers, the flow
characteristics in the chamber approach that of a CSTR and, therefore, the
concentration at the outlet from the contactor (Cout) is assumed to be
representative of the dissolved concentration of ozone in the liquid phase
(C). Currently, contactors using turbine agitators appear to approximate
CSTR characteristics (Schwartz et al, 1990). Other systems with T10/HOT
values less than 0.33 may use the same correlations. This correlation is
applicable to every chamber, including turbine contactors used for first
chambers or as a single chamber contactor.
The measurement of ozone concentration in the gas phase is a possible
alternative for determining C although such correlations will be highly
site specific. A procedure to develop site specific correlations between
the average ozone concentration and the off-gas concentration is presented
in Section 0.4.2.1.
*-^
Counter-Current Flow
In counter-current flow, the water flows opposite to the direction
of bubble rise. Measurement of the concentration profile in such systems
revealed that the concentration in the liquid phase uniformly increased
with depth in the ozone chamber as shown in Figure 0-15. The maximum
concentration in the chamber is achieved near the water outlet from the
ozone chamber.
Measurement of the ozone concentration in an ideal plug flow chamber
reveals that the average concentration is only 25 to 50 percent of the
outlet concentration for these chambers under typical operating condi-
tions. Additional contributions to the average concentration that are not
accounted for by the plug flow analysis, include the contribution of
0.3-8
-------
TOP
Q.
LU
o
QC
UJ
BOTTOM
INCREASES
DISSOLVED OZONE RESIDUAL
A COUNTER-CURRENT FLOW PROFILE
TOP
z
K-
a.
UJ
o
QC
UJ
BOTTOM
INCREASES
DISSOLVED OZONE RESIDUAL
B. CO-CURRENT FLOW PROFILE
FIGURE 0-15-020NE CONCENTRATION PROFILES
-------
turbulence and the contribution of the inlet concentration. Based on these
considerations, EPA recommends the use of one-half the outlet concentra-
tion of ozone as an estimate for C.
The measurement of ozone concentration in the off gas is a possible
alternative for determining the average ozone concentration although the
correlations will be highly site specific. A procedure to develop site
specific correlations between the average ozone concentration and the off-
gas concentration is presented in Section 0.4.2.1.
Co-Current Flow
In co-current flow, both the water and gas flow in the same
direction. The ozone concentration profile in co-current operation
increases until it reaches a maximum and then decreases along the contact
chamber as shown on Figure 0-15. The dissolved ozone concentration
increases at the beginning of the column due to dominant mass transfer
from the ozone rich bubbles. Then the gas phase becomes depleted of ozone
and the impact of ozone consumption in the liquid phase dominates the
ozone profile. C can be estimated as the concentration of dissolved ozone
at the outlet or by the average of the inlet and outlet concentrations of
dissolved ozone, whichever is higher. This estimate should still be
conservative, particularly for systems exhibiting high transfer efficien-
cies.
The measurement of ozone concentration in the off gas is a possible
alternative for determining the average ozone concentration although the
correlations will be highly site specific. A procedure to develop site
specific correlations between the average ozone concentration and the off-
gas concentration is presented in Section 0.4.2.1.
Reactive Flow
In ozone chambers operated in a reactive flow configuration, the
water contains dissolved ozone residual from previous chambers but no
additional ozone is being introduced. Reactive flow chambers are used:
for other disinfectants, such as chlorine, chlorine dioxide and chloramin-
es; for the decay of ozone following a contactor or a static mixer; and
for combining ozone with hydrogen peroxide.
0.3-9
-------
For static mixers, the mixer acts as a turbine chamber with the
pipeline following the mixer acting as the reactive chamber. The pipeline
is in effect the second chamber and the guidelines in Table 0-6 apply for
the determination of C. The contact time in the pipeline can be
calculated by assuming plug flow.
In order to be consistent with the recommendations for monitoring
other disinfectants in reactive flow chambers, and in order to assure
compliance under worst case conditions, the use of the residual outlet
from the chamber (C^) is recommended as a conservative measure of C. The
CT for reactive flow chambers may be estimated by dividing the chamber
into subunits, measuring the concentration at the end of each subunit, and
adding the CT credits.
Estimates of C based on the outlet concentration were conservatively
developed based on available test data. EPA's recommended values for C
<
are summarized in Table 0-6. A system may choose to perform additional
testing for direct measurement of ozone residuals to support a higher
value, if appropriate. In addition, a reactive flow chamber may be
subdivided into smaller units with ozone measurements at the end of each
unit to improve CT credit.
0.3.4 Estimating C
The results of this section are summarized in Figure 0-16. The
decision tree shows the applicable methods of estimating C for each flow
configuration, and provides a quick means to compare alternatives and make
a selection.
0.3-10
-------
O
z
H
<
I
K
V)
UJ
aj
S
u
II
o
REACTIVE
FLOW
CHAMBER
ij
5
§
u
II
u
e
Z
O
£
U
UJ
z
z
U
O
CO
UJ
tr
o
u.
UJ
UJ
cc
z
g
CO
u
u
Q
UJ
I
O
s
CO
-------
0.4 SITE-SPECIFIC EVALUATION OF OZONE CONTACTORS
0.4.1 Introduction
The second set of guidelines is designed to prevent systems from
costly over-design and use of overdoses of ozone, by performing site
specific characterization of their ozone contactors. This approach was
partially utilized in the previous two sections by recommending a direct
measurement of the ozone concentration profile and by allowing systems to
use the SFA or CSTR approaches. In this section the site specific
evaluation procedure will be further developed by presenting additional
options to improve disinfection credits or simplify monitoring procedures.
EPA recommends the following three alternatives for site specific
evaluations:
Estimating C by measurement of another variable
Modeling performance of field scale operation
Use of microbial indicator studies
C may be estimated by measuring an easily monitored (observable)
variable. Systems should develop site specific correlations between C and
another observable parameter such as the gas or liquid concentration
exiting (Cout) the chamber and monitor this observable parameter instead
of C. Guidelines to develop such site specific correlations are presented
in Section 0.4.2
Modelling the. performance of full scale operations is an alternative
to the separate C and T approach. The first procedure separated the
analysis into two separate issues related to determining C and T.
Extensive modelling of the system may predict higher inactivation levels,
even for the same C and T. EPA recommends that systems construct
mathematical models of their ozone contactors to predict the disinfection
performance, provided that the models are confirmed by experimental
observation of the actual ozone concentration profile in the contact
chambers, as discussed in Section 0.4.3.
Microbial indicator studies may be used to determine the inactivation
of viruses and Giardia cysts in ozone contactors. EPA recommends that
systems be allowed to evaluate the performance of their disinfection
systems by spiking a pilot of the contactor with an indicator microorgan-
ism and predicting the actual inactivation of Giardia cysts and viruses
0.4-1
-------
based on the inactivation of the indicator microorganisms. Guidelines to
conduct such pilot scale performance evaluations are presented in Section
0.4.4.
0.4.2 Site Specific Correlation of C with an Observable Variable
Section 0.3 recommends determining the concentration of ozone in
contactors by one of the following ways:
1. Measure the concentration profile in the chambers and determine
the average dissolved ozone concentration for C.
2. Measure the dissolved concentration of ozone in the water
outlet from each chamber (C ) and estimate C by the correla-
tions presented in Tables O-o.
This section presents an alternative method to determine C.
The SWTR requires unfiltered systems to report a daily CT for their
disinfection systems. Similar requirements may be specified by the
Primacy Agency for filtered systems. Measuring the concentration in the
ozone chambers each day may be difficult. Determining the ozone
concentration in a chamber by continuous or daily measurements of other
variables is probably preferable. Likewise, many systems may prefer to
monitor the ozone concentration in the off gas (Y^) or via the applied
ozone dose rather than monitor C^. However, based on available data, a
non-site specific correlation between the average ozone concentration in
the chamber and an observable variable other than taut could not be
developed.
EPA encourages systems to develop such site specific correlations and
use them instead of the general procedures. These correlations may be
developed in one of the following ways:
1. Determine site specific correlations between C^. and another
variable that can be easily monitored. Measure the variable.
estimate C0 t and then use the correlations presented in Tables
0-6 to predict C.
2. Determine site specific correlations directly between C and
another variable such as the ozone concentration in the off gas
(Y .} or C t. Measure that variable and estimate C.
* OUt' OUt
0.4-2
-------
Correlations between C or Cout and a measurable parameter may vary in
complexity from a simple empirical linear correlation to a highly
sophisticated mathematical model accounting for the ozone concentration
profile in the contact chamber. Development of appropriate correlations
depends on the engineering capabilities of the utility. Therefore, EPA
does not recommend any particular mathematical relationships. However,
the following sections present guidelines to assist systems in developing
appropriate correlations.
Correlations for Specific Chamber^
The correlations should refer to a specific contact chamber and
should be verified to fit the performance of this chamber. For example, a
correlation for the first chamber should not be used to predict C in the
second chamber of a multiple-chamber system.
Developing the Correlation
When fitting the correlation with experimental data, a record of the
following variables should be kept:
a. Water flow rate
b. Gas flow rate
c. Ozone concentration in the gas feed
d. "bzone transfer efficiency
d. Water temperature and pH
e. Concentrations of all major inorganic reducing agents,
if they constitute a substantial proportion of the total
ozone demand, such as iron(II) and manganese, TOC,
alkalinity and turbidity.
f. C^ or whatever is being correlated
g. The measurable variables such as ozone dosage or C^
The system should also record the dependent {C or C^) and indepen-
dent measurable variables.
0.4-3
-------
Application of the Correlation
The correlation should be evaluated with at least a 90 percent
confidence level. Since confidence margins are very sensitive to the
number of observations used to develop the relationship, this requirement
will prevent the use of correlations that are based on a limited amount of
observations. On the other hand, because systems usually make daily
records of most of the parameters needed to develop a correlation, the
number of observations will usually be very high, thereby, providing a
high confidence level for the correlation. Simple procedures to determine
confidence intervals are presented in statistical textbooks.
The correlation nust be checked periodically, such as monthly, as an
additional precaution against unexpected shifts in water conditions.
The correlation should be applied only to conditions that are within
the parametric range for which the correlation was developed, as noted in
the second guideline. Interpolation is permitted but extrapolation is
not. Correlations developed during the winter time should not be used to
evaluate performance in the summer.
EPA believes that by permitting such correlations, systems will be
encouraged to apply sophisticated mathematical models in order to decrease
the confidence interval and administer smaller doses of ozone. EPA also
expects that systems will develop correlations between C in the contactors
and measurable parameters to simplify their operations. Small or lesser
"***
equipped systems will then be able to use these relationships to estimate
the performance of their ozone contactors. EPA intends to follow advances
in this field and issue updated examples and guidelines regarding the
selection of efficient site specific correlations. ..
0.4.2.1 Utilizing Off-Gas Measurements
In ozone contactors, the gas and liquid streams equilibrate when the
contact between the gas and liquid is intimate enough and for sufficient
time, otherwise the concentration in the water phase will be much lower
than the equilibrium concentration. It can be assumed that close to
equilibrium conditions are reached, when the transfer efficiency in the
contactors is greater than 85 percent ((Y^-Y^J/Y,,, > 0.85). When the
transfer efficiency is greater than 85 percent, systems may use solubility
0.4-4
-------
constant data to calculate Cout from the contactor, based on the ozone
concentration in the off gas. This may lead to a slight over estimate of
the concentration in the liquid phase but this over estimate is justified,
in view of the better reliability of gas phase measurements.
Henry's constants for ozone at various temperatures are presented in
«
Table 0-7. The residual concentration of ozone may be estimated by:
Where:
Yc^t = The concentration of ozone in the gas phase (ppm -
volume or partial pressure-atm)
C = The concentration of ozone in the liquid phase
H = Henry'^s constant (atm/mg/L)
When applying off-gas modelling, liquid phase measurements must be
made periodically to check the correlation, as the ozone transfer
efficiency has a high impact on the results of this correlation.
Systems must be cautioned against the use of off-gas measurements
for multiple chamber contactors with a common headspace. As noted
previously, modelling must be specific to individual chambers. Thus, if
a contactor has a common head space between chambers, no distinction can
be made as to the concentration in each chamber. Therefore, off-gas
measurements for modelling are recommended for use with single chamber
contactors.
Example 0.4-1
The Metropolitan Water District of Southern California conducted
off-gas monitoring on a single chamber co-current flow pilot contactor to
determine the dissolved ozone concentration:
Operating conditions were as follows:
source water: Colorado River
feed gas ozone concentration * 2 percent by weight
off gas ozone concentration - 0.185 percent by weight
(or 0.123 percent by volume)
0.4-5
-------
transfer efficiency - 90.8 percent
temperature » 16.5 C
observed ozone residual « 1.04 mg/L
Henry's constant 16.5°C - 0.001179 atm/mg/L
The ozone residual estimated from the off gas concentration
Cout-Yout/H « 0.00123/0.001179
, - 1.04 mg/L
The measured residual is the same as that predicted by the
off-gas measurement indicating that this approach is appropri-
ate for this system.
Example 0.4-2 Empirical Correlation between Cout and Yout
A system using two counter-current contact chambers in series wants
to predict Cout in the second chamber by the concentration of ozone in the
off-gas (Yout). Daily observations of the pertinent parameters during tne
first month of operation are presented in Table 0-8.
The system chose to correlate £M and Yout by linear empirical
correlation.
The daily observations, and the best linear fit are presented
in Figure 0-17.
The 90 percent confidence interval is presented by the lower
1 ine"~in Figure 0-17.
The system may use the 90 percent confidence level line to
estimate Cout based on measurements of Y^.
For example when Yout - 0.4 percent then the system may use Cout
- 0.36 mg/L.
Although the best estimate is C^ « 0.4 mg/L, the system
should predict Cout - 0.36 mg/L.
Now, according to Table 0-6, the system may predict C using
the recommended guideline of C - C^/2 - (0.36J/2 » 0.18 mg/L.
The system measures the ozone concentration at the chamber
outlet monthly, to check the model correlation.
0.4-6
-------
TABLE 0-7
HENRY'S CONSTANTS FOR OZONE<1)
Water
Temperature Henry's Constant Henry's Constant
(°C) atm/Mole Fraction (a.tm/mq/1 ozone)
0 1,940 0.00073
5 2,180 0.00082
10 2,480 0.00093
15 2,880 0.00108
20 3,760 0.00141
25 4,570 0.00171
30 5,980 0.00224
NOTE: {1) EPA, 1986
-------
TABLE 0-8
Empirical Correlation
Between C and Y
Temp. 8C YjnS
0.5 0.5 2.0 20 20
0.47 0.43 2.8 15 2.0
0.38 0.41-- 2.5 17 2.0
0.39 0.4 2.3 18 2.0
0.28 0.32 2.4 18 2.0
0.2 0.17 2.6 20 2.0
0 25 0.23 2.0 20 2.0
0.32 0.27 2.0 21 1.9
0.29 0.27 2.0 18 1.9
0.2 0.18 2.0 17 2.0
0.22 0.2 1.9 18 2.1
0.30 0.33 1.8 20 2.0
0.32 0.34 1.9 17 2.0
0.28 0.27 1.9 18 1.8
0.29 0.32 2.5 18 1.9
0.4 0.42 2.4 19 1.9
0.47 0.45 2.3 19 1.8
0.35 0.37 2.4 21 1.9
0.30 0.29 1.9 19 1.8
0.20 0.17 1.9 19 1.8
0.15 1.19 2.0 19 2.0
0.12 0.20 1.9 17 2.0
0.17 0.17 1.9 19 1.9
0.14 0.16 2.0 19 2.0
0.13 J3.12 1.9 18 2.0
0.25 0.27 1.9 17 2.0
0.29 0.32 1.9 18 2.1
0.30 0.29 1.8 17 2.0
0.22 0.20 1.9 17 2.0
0.22 0.20 1.9 18 1.9
-------
I-* s
z
o
LU CO
CC <
CZ (5
8s
-i O
-3°
f i<
5 CL m
^ 2z
•: LU o
1 S3
ct
\ 03
s, \
\ \
X LU
LU CC
LU
CC
-------
If this system had the means to monitor the concentration
profile in the contactor and determine C directly it could
develop a correlation between C and Y^, instead of using Table
0*6.
0.4.3 Modeling the Performance of Full Scale Operations
More extensive site specific mathematical modelling of the actual
performance of the ozone contactor may determine higher inactivation
levels than those determined by the separate C and T approach. Therefore,
systems should be allowed to use such advanced modelling, provided that
these models are confirmed by direct measurement of the dissolved ozone
profile in the contactor. Only after the model is confirmed to correctly
estimate the concentration profile in the contactor can it be used to
estimate the inactivation performance of the contactor. Systems with
multiple chamber contactors must develop models for each of the chambers.
Various types of mathematical models for reaction-diffusion systems
were reported (Danckwerts, 1976) and some were shown to be applicable for
ozone contactors (Gurol and Singer, 1982). This section deliberately
avoids giving preference to any type of mathematical modelling in order to
encourage engineering innovations. The guidelines presented below may
help systems to select appropriate modelling that will be consistent with
the requirements of the SWTR.
The model should account for the ozone demand of the water being
treated in the -contactor. The rate of ozone reaction and decomposition
should be based on batch experiments, on-site pilot plant columns, or
full-scale measurements.
The model should represent the actual flow distribution in the ozone
contact chambers by incorporating a dispersion term and/or a three
dimensional velocity distribution term in the contactor.
The modelled profile of the concentration of dissolved ozone in the
contactor should fit the actual distribution of dissolved ozone, as
verified by direct measurements, with a variation of less than 10 to 20
percent. This difference between the model and measured residual allows
for the inherent inaccuracies in measuring the actual ozone residual. The
mathematically modelled concentration profile should not be used without
comparing it with actual measurements. Even elaborate mathematical models
0.4-7
-------
are not considered reliable enough to estimate the concentration
distributions of dissolved gasses in complex gas/liquid operations,
without additional verification of the actual concentration profile in the
contactor.
In addition to the above guidelines, the model may also account for
other phenomena that may affect the performance of the ozone contact
chambers, such as: the effects of varying bubble diameter during its
movement through the.contactor, the effect of stagnant regions in the
contactor and the variation of the hydrostatic pressure.
For example, a system may use the two film theory coupled with
reaction kinetics to estimate the performance of an ozone contact chamber.
Using the two film theory the relevant differential equations are:
L dC/dz - Mt + Mr + Md
G dy/dz - Mt
L dl/dz - Md - KCI
Where:
C * Concentration of dissolved ozone (mg/L)
G « Gas flow rate per cross section of the contactor (m2.Kg
gas/min)
I - Concentration of the target microorganism (Giardia or viruses)
*••*
L « Water flow rate per cross section area of the reactor (Kg wat-
er/min.m )
y - Concentration of ozone in the gas phase (mg/L)
z « Length coordinate of the contactor
Mt » An expression for ozone transfer from the bubble phase to the
water phase. For example, kta(Ci-C) where kta stands for the
volumetric mass transfer coefficient, C, represents the
interfacial concentration of ozone, given by solubility data
(Table 0-10).
Mr * An expression for the rate of ozone consumption in the water
due to auto-decomposition and the ozone demand of the treated
water. For example, Mr- k,C -k2(C)(R). Where k, and k2 are
kinetic coefficients, and R represents the variable ozone
demand, such as TOC. An additional equation may be required to
represent the variation of R along the contactor.
0.4-8
-------
Md - An expression for the dispersion by turbulence and bubble flow
of dissolved ozone in the specific contactor. For example,
OdzC/dzz the dispersion coefficient (D) may be evaluated by
analysis of tracer study data. The third equation describing
the microorganism concentration (dl/dz) should incorporate the
same dispersion coefficient (0).
KCI- Chick's inactivation term (K«2.303k, where k » Chick-Watson's
inactivation coefficient presented in Table 0-4, C represents
the local concentration of ozone and I represents the concen-
tration of microorganisms).
The validity of these equations is subject to the appropriate
boundary conditions at the bottom and top of the contactor. The signs of
the various terms depend on the definition of coordinates and the type of
flow configuration (co-current or counter-current flow configuration).
0.4.4 Microbial Indicator Studies to Model Inactivation Contactors
According to the recommendations in Appendix G, systems may
demonstrate the actual performance of a disinfection system rather than
rely on the CT approach. The procedures outlined in Appendix G recommend
the use of Giardia muris cysts as indicators of Giardia inactivation and
bacteriophage (MS2) as indicators for virus inactivation by disinfection
in general. However, recent data indicate that MS2 phages may be
substantially more sensitive to ozone disinfection than pathogenic
viruses, and therefore are not a good indicator for determining adequate
ozonation conditions for inactivating pathogenic viruses (Finch, 1990).
Additional research is needed to determine which coliphage species, if
any, can be used as an appropriate indicator for virus inactivation by
ozone. Pilot scale inactivation experiments using appropriate indicator
microorganisms can serve as powerful tools to indicate the performance of
the ozone contactors. This section contains guidelines for conducting
indicator studies. At this time, full-scale testing with indicator
organisms is not feasible because of the high volume of organisms needed
and the concern for introducing organisms into the finished water.
However, with the development of naturally occurring indicators such as
resistant species of coliphage, demonstration on the full-scale level may
be feasible in the future.
0.4-9
-------
Systems may determine the performance of their disinfection basins
by demonstrating levels of inactivation of indicator microorganisms such
as Giartji? muris cysts, or other indicator microorganisms provided that
such demonstrations are based on solid engineering principles. The
following steps can be used for conducting indicator studies:
1. Batch Experiments
On-site batch disinfection experiments are recommended with treated
water spiked with indicator microorganisms to determine the inactivation
kinetics of the indicator used in the pilot scale experiments. Microor-
ganisms should be used as indicators preferably in the range where the
inactivation kinetics approximate Chick's law. This protocol assumes that
within the desired inactivation range, the inactivation kinetics will
approximate Chick's law. It is important to note that other disinfection
kinetic models, not yet apparent, may be developed to more accurately
predict ozone inactivation efficiency than the Chick-Watson model.
Evidence that other models may be more appropriate is shown with data
generated by several researchers for different organisms (Wolfe, R.L. et
al, 1989; Finch G., et al 1988; Finch G. and Smith, D.W. 1989).
2. Pilot Scale Indicator Experiments
Pilot-scale experiments should then be conducted using identical
strains of biological indicators to those used in the batch experiments.
The pilot-scale experiments should be repeated under identical gas and
water flow conditions with and without introducing ozone into the gas
stream. The actual performance may then be calculated by subtracting the
inactivation achieved in the control experiment (without ozone) from the
inactivation achieved in the ozone disinfection experiments.
3. Evaluation of Inactivation Performances
Systems may choose direct or indirect methods to interpret the
inactivation performance of ozone contactors based on indicator studies.
The direct method is more conservative and simple while the indirect
method is more accurate but requires mathematical modelling of the
contactors. The two procedures are outlined below:
0.4-10
-------
1. Direct prediction of inactivation performance
a. Determine k< (where k, is Chick-Watson's inactivation
coefficient of the indicator microorganism) from batch
test data with the expression:
where:
indicator ' Survival ratio of indicator microorganism
as . determined by batch experiments.
C - Dissolved ozone concentration in the batch
experiment (mg/L)
t - time (minutes) elapsed from the beginning
of the batch experiment
Note: This assumes that the inactivation data will provide a
reasonable fit for this equation. If this is not true,
then the following is not applicable and other relation-
ships should be developed.
b. Determine the disinfection performance of the pilot
scale disinfection system on the indicator microorganism
( I/Jo' indicator'
c. Calculate the inactivation of Giardia cysts or viruses
(I/I0) using the appropriate k' values from Table 0-3:
log(I/I0) - log(I/I0)indic.tor (k'/k,) I
^ log (I/IO) « log (I/I0)indieitor (k1 < k) (8)
This equation still represents an approximation because it neglects
dispersion effects. The laws used in deriving the above equations are
based on conservative similarity approaches. When the indicator
microorganism is less resistant to ozone disinfection than the target
organism (k, > k), then the plug flow"deration represents the more
conservative prediction approach. Equation J is based on the assumption
that the flow configuration irv the chamber approaches plug flow. When the
indicator microorganism is'more vulnerable then the target microorganism
(k, > k) then the CSTR approach provides a more conservative estimate.
Equation 4 represents a conservative approximation to the CSTR similarity
0.4-11
-------
rule. A more accurate determination of the inactivation performance of
the contactor may be calculated by the following approach:
2. Indirect determination of the disinfection performance
a. Determine k, (where k. is Chick-Watson's inactivation
coefficient of the indicator microorganism) from batch
test data with the expression:
^9 U/10) indicator ' •*««
where:
(I/Io).nd.c«top • Survival ratio of indicator micro-organisms
as determined by batch experiments.
C - Dissolved ozone concentration in the batch
experiment (mg/L)
t - time (minutes) elapsed from the beginning
of the batch experiment
b. Determine the disinfection performance of the pilot
scale inactivation level of the indicator microorganism
(I/Ifl' indicator1
c. Determine the actual concentration profile in the
disinfection chamber (see Section 0.3.2).
d. Construct a mathematical model that estimates the
concentration profile in the contactor as discussed in
Section 0.4.3
e. Confirm the mathematical model by fitting its parameters
such as dispersion or kinetic coefficients to describe
accurately the concentration profile of ozone in the
contactor and the overall Inactivation of the indicator
microorganism. A model that predicts within 10-20
percent the inactivation of the indicator microorganism
and the concentration profile of dissolved ozone in the
contacyfe#$?*4£ be considered to be valid and can be
used bf^ncoffr&GM&M k values from Table 0-3 to
estimate the inactivaJTon of Giardia cysts or viruses in
the contactor.
0.4-12
-------
REFERENCES
Bader, H. and J. Hoigne, Determination of Ozone in Water by the Indigo
Method, A Submitted Standard Method, Ozone: Science and Engineering, pp
449-456, 1982, 4
Cams, K, Design of East Bay Municipal Utility District ozone disinfection
plants in Oakland Ca., presented at EPA workshop on Ozone CT, Cincinnati,
OH., Feb. 1990
Chick, H., An investigation of the laws of disinfection, J. Hygiene, 8, 92
(1908)
Danckwerts, P.V. Gas Liquid Reactions, McGraw Hill Inc. (1976)
Finch, G.R., Smith, D.W., Stiles, M.E., Dose Response of Escherichia Coli
in Ozone Demand-free Phosphate Buffer, Water Research Vol. 22, No.12, pp.
1563-1570, 1988.
Finch, G., University of Alberta, Canada, Private Communication (1990).
Hoff, J. C., Strengths and weaknesses of using CT values to evaluate
disinfection practice, in assurance of adequate disinfection, or C-T or
not C-T, AWWA seminar proceedings, Annual AWWA Conference, Kansas City.
pp. 9 (1987)
Gurol M.D. and P.C. Singer, Kinetics of ozone decomposition, A dynamic
approach, Env. Sci. &Tech., 16, 377 (1982)
Joost, R.D., 1.A. Jackson and L.J. Bollyky, Optimization of ozone
contactors for drinking water disinfection, Proceedings of the 9th Ozone
World Congress", L.J. Bollyky, Editor, New York, NY, vol 1, pp. 551 (1989).
Lev, 0. and S. Regli, Compliance of ozone disinfection systems with the
Surface Water Treatment Rule, Selection of characteristic exposure time
(T), submitted to J. of Env. Eng. ASCE (1990a)
Lev, 0. and S. Regli, Compliance ofozon| disinfection systems with the
Surface Water Treatment Rule, SMflBBf characteristic concentration
(C), submitted to J. of Env.
Levenspiel, 0., Chemical ti^^H^^^^BRg* 2nd ed. Wiley, New York, NY
(1972).
Perry R.H., Perry's CtiWrffaTTngineering Handbook, McGraw Hill Book
Company, New York, NY (1963)
Richards, D. A. and Fleischman, M., Ozone transfer to aqueous systems in
a static mixer, 2nd Symposium on ozone technology, Montreal, Canada, R. G.
Rice, P. Pichet and M. A. Vincent, eds, p. 57 (1975).
-1-
-------
/p
Rice, R. EPA, in preparation, Source Document: Ozone Treatment of
Drinking Water, EPA/625/1-86/021.
Rosenbeck, G.L, McMahon Associates, Inc., Private communication (1989)
Schwartz, B., Hackensack Water Company, Private communication (1989)
Schwartz, B., L.C. Fung and C. N. Weng, Determination of ozone concentra-
tion in turbine contactor, presented at EPA Workshop on Ozone CT,
Cincinnati, OH., Feb. 1990
Seinfeld J. H. and Lapidus L., Mathematical methods in Chemical Engineer-
ing, Vol. 3, Process Modeling Estimation and Identification, Prentice Hall
Inc. (1984).
Stolarik, G.L. and J. D. Christie, Projection of ozone C-T values, Los
Angeles Aqueduct Filtration Plant, presented at Int. Ozone Assoc. Conf.
Monroe, Mich., April 1987.
Stolarik, G.L. and J. D. Christie, Projection of ozone C-T values, Los
Angeles Aqueduct Filtration Plant, presented at EPA workshop on Ozone CT,
Cincinnati, OH., Feb. 1990
Treybal, R.E., Mass-transfer operations, McGraw-Hill Book Company, 3rd ed.
(1980)
Watson, H.E., A note on the variation of the rate of disinfection with
change in the concentration of the disinfectant, J. Hygiene, 8, 536 (1908)
Wolfe, R.L, M.H. Stewart, K.N. Scott and M. J. McGuire, Inactivation of
Giardia Muris and indicator organisms seeded in surface water supplies by
peroxone and ozone, Env. Sci. and Tech., 23, 744 (1989)
Yuteri C. and Gurol M.D. Evaluation of kinetic parameters for the
ozonation of organic micropollutants, Wat. Sci. & Tech. 21, 465 (1989)
DATE DUE
(I
-2-
------- |