# National Primary Drinking Water Regulations: Monitoring Requirements for Public Drinking Water Supplies: Cryptosporidium, Giardia, Viruses, Disinfection Byproducts, Water Treatment Plant Data and Other Information Requirements

> Briefs, arguments, decisions, and more.

URL: https://www.frixlaw.com/law-library/documents/fr%3A94-2587

## Record

- **Collection:** Federal Register
- **Document type:** Uncategorized Document
- **Published:** February 10, 1994

## Text

SUMMARY: EPA is proposing to require public water systems which serve
10,000 people or greater to generate and provide the Agency with
specific monitoring data and other information characterizing their
water systems. Systems which use surface water, or ground water under
the influence of surface water, and serve between 10,000-100,000 people
would be required to (a) monitor their source water at the intake of
each plant for two disease-causing protozoa, Giardia and
Cryptosporidium; fecal coliforms or Escherichia coli; and total
coliforms; and (b) provide specific engineering data as it pertains to
removal of disease-causing microorganisms. Systems which use surface
water, or ground water under the influence of surface water, and serve
more than 100,000 people would be required to monitor their source
water at the intake of each plant for the microorganisms indicated
above, plus viruses, and, when pathogen levels exceed one pathogen/
liter in the source water, finished water for these microorganisms;
monitor for certain disinfection byproducts (DBPs) as well as other
water quality indicators; and provide specific engineering data as they
pertain to removal of disease causing organisms and control of DBPs.
All ground water systems that serve more than 100,000 people would be
required to monitor for certain DBP, other water quality indicators,
and to provide specific physical and engineering data. Systems which
use surface water and serve more than 100,000 people and systems which
use ground water and serve more than 50,000 people would be required to
conduct bench or pilot scale studies to evaluate treatment performance
for the removal of precursors to DBPs unless they have met certain
source water or treated water quality criteria. This information will
be used to consider possible changes to the current Surface Water
Treatment Rule (SWTR) and to develop drinking water regulations for
disinfectants and DBPs. If the SWTR is amended, information collected
under this monitoring rule would assist utilities in complying with
such amendments.

DATES: Comments should be postmarked or delivered by hand on or before
March 14, 1994. Comments received after this date may not be considered
because of time constraints.

ADDRESSES: Send written comments to ESWTR/DBPR Monitoring Docket Clerk,
Water Docket (MC-4101); U.S. Environmental Protection Agency; 401 M
Street, SW; Washington, DC 20460. Please submit any references cited in
your comments. EPA would appreciate an original and three copies of
your comments and enclosures (including references). Commenters who
want EPA to acknowledge receipt of their comments should include a
self-addressed, stamped envelope. No facsimiles (faxes) will be
accepted because EPA cannot ensure that they will be submitted to the
Water Docket.
The proposed rule with supporting documents and all comments
received are available for review at the Water Docket at the address
above. For access to Docket materials, call (202) 260-3027 between 9
a.m. and 3:30 p.m. for an appointment.

FOR FURTHER INFORMATION CONTACT: The Safe Drinking Water Hotline,
Telephone (800) 426-4791. The Safe Drinking Water Hotline is open
Monday through Friday, excluding Federal holidays, from 9 a.m. to 5:30
p.m. Eastern Time. For technical inquiries, contact Stig Regli or Paul
S. Berger, Ph.D., Office of Ground Water and Drinking Water (WH-550D),
U.S. Environmental Protection Agency, 401 M Street SW., Washington DC
20460, telephone (202) 260-7379 (Regli) or (202) 260-3039 (Berger).

SUPPLEMENTARY INFORMATION:

Table of Contents

I. Statutory Authority
II. Regulatory Background
III. Discussion of Proposed Rule
A. Enhanced Surface Water Treatment Requirements (ESWTR)
1. Need for Enhanced SWTR
2. Monitoring and reporting requirements and rationale
3. Reasons for monitoring listed pathogens and indicators
4. Rationale for frequency of microbial monitoring
5. Rationale for reporting physical data and engineering
information
6. Analytical methods
7. Laboratory approval
8. Quality assurance
B. Disinfection Byproducts Rule (Stage 2)
1. Need for additional data
2. Monitoring and reporting requirements and rationale
3. Treatment process information collection
4. Database development
5. Analytical methods
6. Quality assurance
7. Bench/pilot scale testing
C. Dates
D. Reporting Requirements
E. List of Systems Required to Submit Data
IV. State Implementation
V. Cost of Rule
VI. Other Statutory Requirements
A. Executive Order 12866
B. Regulatory Flexibility Act
C. Paperwork Reduction Act
D. Science Advisory Board, National Drinking Water Advisory
Council, and Secretary of Health and Human Services
VII. Request for Public Comments
VIII. References

I. Statutory Authority

The Safe Drinking Water Act (SDWA or the Act), as amended in 1986,
requires EPA to promulgate National Primary Drinking Water Regulations
(NPDWRs) which specify maximum contaminant levels (MCLs) or treatment
techniques for drinking water contaminants (42 U.S.C. 300g-1). NPDWRs
apply to public water systems (42 U.S.C. 300f(1)(A). Section 1412(b)(3)
of the Act requires EPA to publish regulations for at least 25
contaminants at three year intervals. Section 1412(b)(9) of the Act
requires EPA to review existing national primary drinking water
regulations at least once every 3 years.
According to section 1445(a)(1) of the Act, every public water
system ``shall establish and maintain such records, make such reports,
conduct such monitoring, and provide such information as the
Administrator may reasonably require by regulation to assist him in
establishing regulations, [or] * * * in evaluating the health risks of
unregulated contaminants''. This section authorizes EPA to require
systems to monitor and provide the Agency with these data as well as
other data characterizing the system, including source and treated
water quality.
In addition, section 1401(1)(d) of the Act defines NPDWRs to
include ``criteria and procedures to assure a supply of drinking water
which dependably complies with such maximum contaminant levels;
including quality control and testing procedures * * * ''. This section
authorizes EPA to require systems and laboratories to use Agency-
approved methods and quality assurance criteria for collecting and
analyzing water samples.

II. Regulatory Background

Two regulations attempt to control disease-causing microorganisms
(pathogens) in public water supplies--the Total Coliform Rule (54 FR
27544; June 29, 1989) and the Surface Water Treatment Requirements
(SWTR) (54 FR 27486; June 29, 1989). A third regulation, the
Groundwater Disinfection Rule, which is currently under development,
will add further protection for systems using ground water. The Agency
is considering revising the SWTR in conjunction with the development of
other new regulations.
Another rule EPA is currently developing will address chemical
byproducts that form when disinfectants used for microbial control in
drinking water react with various organic chemicals in the source
water. Some of these disinfection byproducts are toxic or are probable
human carcinogens. As such, they were included on the 1991 Drinking
Water Priority List (56 FR 1470; January 14, 1991) as candidates for
future regulations. They are among the candidate contaminants for which
EPA must meet a Court-ordered deadline that is currently being
negotiated.
To develop the Disinfectant/Disinfection Byproducts (D/DBP) Rule,
EPA instituted a formal regulation negotiation process in 1992 (57 FR
53866; Nov 13, 1992) including representatives from water utilities,
State and local agencies, environmental groups, consumer groups, and
EPA. The Negotiating Committee agreed to propose three rules: a) an
information collection rule (ICR), which is proposed herein, b) an
``interim'' enhanced surface water treatment rule (ESWTR), to be
proposed within the next few months, and c) D/DBP regulations, to be
proposed concurrently with the interim ESWTR.
During the development of the D/DBP Rule, a number of members of
the Negotiating Committee did not believe that there were adequate data
available to address some of the DBPs on EPA's priority list (56 FR
1473; January 14, 1991). They believed that insufficient data were
available on many aspects of DBPs necessary to make appropriate
regulatory decisions including health effects and health effect related
issues, occurrence of and exposure to contaminants, and the
capabilities of treatment technologies. Also of concern were the
limited data on microbial contaminants for making regulatory decisions.
The Negotiating Committee's development of the three proposed rules
mentioned above was based on the premise of (1) taking prudent
immediate steps by proposing a two staged D/DBP rule and an interim
ESWTR, and (2) developing additional data through monitoring and
research for future regulatory decisions that would support refinements
to the proposed interim ESWTR and the Stage 2 D/DBP rule. For example,
decisions on the direction of an ESWTR will be limited without more
data on the occurrence of microorganisms, the effectiveness of current
and advanced treatment schemes, potential consumer exposure, dose
response relationships for certain pathogens, pathogen strain
differences, and cyst/oocyst viability measures. Likewise, important
decisions on the Stage 2 D/DBP rule would benefit from additional data
on occurrence of DBPs, effects of current and advanced treatment
approaches on DBP formation, potential consumer exposures, acute short-
term health effects, chronic health effects, and the use of surrogates
as tools for defining adequacy of treatment for specific contaminants
and reduced monitoring.
The ICR was developed to obtain both microbial and DBP occurrence,
exposure, and treatment data for input to the ESWTR and Stage 2, as
outlined below, and would require the expenditure of an estimated $130
million over three and a half years by a segment of public water
suppliers. The commitment by the public water supply community to
support the collection of additional data was linked to EPA's
commitment to provide (1) adequate quality control procedures for
collecting and managing the information obtained under the ICR and (2)
additional funding, especially on health effects, for properly
interpreting the data collected under the ICR. As evidence of this
linkage, non-EPA members of the Negotiating Committee sought to assist
the Agency in obtaining funding for the health effects and other
research equally critical to the future decisions. On May 20, 1993,
these committee members sent letters to the Administration and members
of Congress requesting support for a federal commitment of $4 million
per year for five years to support the needed research. The letters
noted that the American Water Works Association Research Foundation
had, independent of the negotiations, presented a public-private
partnership research plan under which they committed to provide up to
$2 million per year for the research under a one for two match.
On a related effort, non-EPA Negotiating Committee members
requested on July 14, 1993, in a letter to EPA's Administrator,
consideration of reallocation of Agency research funds to support the
research needs described above. The July 14, 1993 letter also spoke of
the need for the Agency to commit funds necessary to adequately
collect, manage, and analyze data collected under the ICR. A number of
Negotiating Committee members believed that, without additional federal
research and data management funding, the ICR data generated by systems
would not be particularly useful in developing the ESWTR or Stage 2 D/
DBP Rule.
The Negotiating Committee agreed that more data, especially
monitoring data, should be collected under the ICR to assess possible
shortcomings of the SWTR and develop appropriate remedies, if needed,
to prevent increased risk from microbial disease when systems began
complying with the new D/DBP Rule. It was also agreed that EPA would
propose an interim ESWTR for systems serving greater than 10,000 people
that included a wide range of regulatory alternatives. Data gathered
under the ICR would form the basis for developing the most appropriate
criteria among the options presented in the proposed interim ESWTR, and
eventually for a long-term ESWTR that would include possible
refinements to the interim ESWTR and be applicable to all system sizes.
Both of these ESWTR rules would become effective concurrently with the
requirements of the Stage 1 D/DBP rule for the respective different
system sizes.
The Negotiating Committee also agreed that additional data on the
occurrence of disinfectants, DBPs, potential surrogates for DBPs,
source water and within-treatment conditions affecting the formation of
DBPs, and bench-pilot scale information on the treatability for removal
of DBP precursors would be useful for developing Stage 2 D/DBP
regulatory criteria beyond those currently being considered for
proposal in Stage 1. To this end, today's proposed ICR rule, which
would require this additional information, was accepted as necessary
and reasonable by the Negotiating Committee.

III. Discussion of Proposed Rule

A. Enhanced Surface Water Treatment Requirements

1. Need for Enhanced SWTR
The SWTR, which became effective on December 31, 1990, requires all
systems using surface water, or ground water under the direct influence
of surface water, to disinfect. It also requires all such systems to
filter their water unless they can demonstrate that they have an
effective watershed protection program and can meet other EPA-specified
requirements. The SWTR also specifies that systems using surface water
must treat water to remove/inactivate at least 99.9% (3 logs10) of
the Giardia lamblia cysts (a protozoan) and at least 99.99% (4
logs10) of the viruses. The SWTR does not require a system to
monitor its source water or drinking water for these pathogens.
During the development of the SWTR, the United States experienced
its first large recognized waterborne disease outbreak of
cryptosporidiosis, caused by the protozoan, Cryptosporidium (Hayes et
al., 1989). Other outbreaks caused by this pathogen have since been
reported both in the United States and other countries. Because of the
lack of data before 1989 on Cryptosporidium oocyst occurrence and
susceptibility to treatment, EPA decided to regulate this pathogen in a
future rulemaking, rather than to delay publication of the SWTR until
these data were available. EPA and others are now performing research
to understand the health risks posed by Cryptosporidium. Although some
occurrence and treatment data are now available, EPA believes that much
more is needed before EPA can promulgate a suitable regulation for
Cryptosporidium. EPA is planning to propose an MCLG and treatment
technique requirement for Cryptosporidium in the ESWTR, and use the
data from this rule to determine the need for, and specifics of, that
regulation.
Another shortcoming of the SWTR is that a 3-log removal/
inactivation of Giardia and a 4-log removal/inactivation of enteric
viruses may be inadequate when a system is supplied by a poor quality
source water. In developing the SWTR, EPA assumed on the basis of data
available at that time, that this level of treatment was adequate for
most systems. The Agency published associated guidance recommending
greater treatment for systems supplied by poor quality source waters
(EPA, 1991).
Subsequent data on Giardia densities in source water and drinking
water (LeChevallier et al., 1991a,b), however, bring into question the
assumption that the treatment specified in the SWTR was adequate for
most systems. These new data suggest that Giardia cyst concentrations
in the source waters of many systems may be too great for the specified
minimum level of treatment to adequately control waterborne giardiasis
(to be discussed in the preamble of the forthcoming proposed interim
ESWTR).
As a result of this uncertainty, EPA needs much more data on the
concentration of Giardia cysts and viruses for various qualities of
source waters, with variation over time and seasonal influences, to
determine the need for additional treatment to provide adequate Giardia
and virus control. In addition, EPA needs more field data on the
effectiveness of different types of water treatment for controlling
these pathogens.
If these new data indicate that EPA's original assumption was
correct, i.e., that only a small percentage of systems have source
water Giardia and virus concentrations that are too great for adequate
control under the SWTR, then guidance (EPA, 1991) may suffice and no
revision of the SWTR would be needed. In contrast, if a high percentage
of systems have elevated concentrations of Giardia, then EPA believes
that the SWTR may need to be revised to require additional treatment
for such systems.
If the data indicate that a revision of the SWTR is needed, then
one regulatory option would be to tailor required treatment levels to
Giardia concentrations in the source water. For example, the Agency
might require a system to achieve at least a 99.9 percent (3-log)
reduction if the source water(s) contained less than 1 cyst/100 liters,
a 99.99 percent (4-log) reduction if the source water(s) contained 1 to
9 cysts/100 liters, a 99.999 percent (5-log) reduction if the source
water(s) contained 10 to 99 cysts/100 liters, and a 99.9999 percent (6-
log) reduction if the source water(s) contained more than 99 cysts/100
liters. These suggested level of treatment requirements are consistent
with existing EPA Guidance (USEPA 1991). Based on the dose response
curve developed by Rose et al (1991) these levels of treatment have
been predicted to ensure a risk of less than 1 infection per 10,000
people per year. The concept of utilities providing higher levels of
treatment to meet a desired acceptable risk level will be one of the
options discussed in the preamble of the forthcoming proposed ESWTR.
The data collected under today's monitoring rule, if promulgated, could
be used as the basis for the treatment level prescribed.
If EPA decides to revise the SWTR according to the above or similar
approach, then the monitoring data would assist the Agency in
determining the most appropriate manner for calculating source water
pathogen densities. For example, options include the arithmetic means,
geometric means, highest value, or a 90th percentile value (e.g., for
ten data points, the system would select the second highest, or for 18
data points, the system would select the third highest). These options
will be discussed in greater detail in the forthcoming proposed interim
ESWTR. These proposed revisions would be modified or withdrawn based on
monitoring data collected under the present rule.
In summary, today's proposed rule, if promulgated, would provide
the Agency with much needed field data to determine the need for
amending the SWTR to control microorganisms in an appropriate manner.
Data collected under this proposed rule could also form the basis by
which systems could establish levels of treatment, perhaps beyond those
minimally required under the SWTR, that are appropriate for controlling
microbial risk while complying with new D/DBP regulations. EPA
understands that the water industry may voluntarily provide additional
useful data for these purposes. The data collected under today's
proposed rule, if promulgated, would also support the long-term ESWTR
rule.
2. Monitoring and Reporting Requirements and Rationale
The rule would require systems using surface water that serve a
population greater than 100,000 (about 233 systems nationally) to
monitor their influent to each plant for Giardia cysts, Cryptosporidium
oocysts, ``total culturable viruses'' (hereafter referred to as
``viruses'', unless otherwise indicated), fecal coliforms or
Escherichia coli, and total coliforms. Monitoring would be monthly for
18 months. If a plant has several sources of water, the system must
sample the blended water from all sources or, if this is not possible,
sample the source with the expected highest pathogen concentration. If,
during the first twelve months of monitoring, any pathogen were to
exceed a density of one/liter, or if the detection limit for any
pathogen exceeds one/liter, the system would be required to monitor
their finished water for the entire set of pathogens and indicators at
the same frequency as source water sampling for the remaining months.
Under this rule, systems would not be required to continue
monitoring for viruses if: (1) viruses are not detected in the source
water at the intake (for each plant) during the first twelve months of
monitoring, or (2) the system has tested the source water at the intake
(for each plant) for either total coliforms or fecal coliforms at least
five times per week between [insert first day of month, 4 months prior
to the promulgation date of this rule] and [insert first day of month,
2 months after the promulgation date of this rule], and the density of
total coliforms or fecal coliforms is less than 100 colonies/100 ml or
20 colonies/100 ml, respectively, for at least 90% of the samples.
For surface water systems that serve between 10,000 and 100,000
people, the rule would require source water monitoring at the intake of
each plant for the organisms listed above, except that they would not
have to monitor for viruses. Monitoring for this category of systems
would be every two months for 12 months. The rule would require all
systems serving more than 10,000 people to provide the above monitoring
data and other, system-specific information to EPA. The rule would not
apply to systems that purchase all of their water from other systems.
The rationale for requiring this information is to provide EPA with
much needed data on the concentrations and variations with time of
viral and protozoan pathogens in various types of source waters. It
would also help EPA evaluate whether current assumptions on water
treatment removal efficiencies for pathogenic protozoa and viruses are
appropriate. Together, these data and the data on source water
concentrations would provide EPA and the system a better understanding
of pathogen concentrations following treatment, which would allow for a
more accurate assessment of the pathogen levels and the associated
health risk to which the public may be exposed. These data, along with
possible additional data on dose-response patterns, pathogen strain
differences, and cyst/oocyst viability measures, would allow EPA to
determine the circumstances under which the SWTR is not adequate and to
revise this rule accordingly to overcome any shortcomings.
The data would also help EPA characterize occurrence relationships
among Giardia cysts, Cryptosporidium oocysts, and viruses. For example,
these data would help the Agency evaluate the merits of using Giardia
as the primary target to define treatment requirements, as it did in
the SWTR. In addition, the data may help EPA identify and prevent
treatment changes that systems might inappropriately consider to meet
the forthcoming D/DBP rule.
The source water data collected under this rule might also be used
for determining appropriate levels of treatment for particular systems
serving more than 10,000 people, if minimum treatment requirements were
specified as a function of source water quality conditions under the
interim ESWTR.
EPA would not require systems serving between 10,000 and 100,000
people to monitor treated water because the Agency believes that
sufficient data for microorganisms would be provided by the larger
systems, which are generally better able to fund the collection of the
needed data. EPA would also not require these sized systems to monitor
viruses in source waters because the Agency believes that the larger
systems would provide sufficient data to establish any relationship
between the viruses and the two protozoan pathogens being monitored,
regarding source water densities and treatment effectiveness. The
Agency, in the absence of data suggesting otherwise, would continue to
use Giardia, possibly including Cryptosporidium, as the primary target
organism(s) for regulation, given their greater disinfection resistance
compared to most other organisms, and consequently less data would be
needed for the viruses.
The data from these larger systems would also be useful for
estimating pathogen concentrations in many source waters serving
systems with fewer than 10,000 people, which EPA believes typically do
not have the financial resources or technical expertise to collect and
process the samples as part of the above monitoring requirements. The
Agency would use the large system data to define the relationship
between the pathogen concentrations in the source water and the
concentrations of potential/existing microbial indicators of water
quality. If such a relationship were found, then small systems could
use one or more of these easily-measured indicators to estimate
pathogen concentrations in their source waters.
In addition, small systems that use the same source water and are
in the same vicinity as a large system may be able to use the same
pathogen concentrations measured by the large system as a basis for
determining the minimum level of treatment required. Finally, EPA may
be able to use these data to develop national occurrence patterns that
would allow the Agency to establish more appropriate treatment criteria
for small systems. By characterizing source water quality using any one
or a combination of these three approaches, a small system could
evaluate the effectiveness of treatment in place for pathogen control
and determine the need for additional treatment steps.
The Agency requests suggestions for assessing pathogen exposure in
small systems in addition to the three approaches provided above.
Following the full compilation of data under the ICR and other research
developments, EPA is considering proposing a long-term ESWTR that would
include criteria by which systems serving less than 10,000 people could
determine appropriate levels of treatment for different source water
qualities.
As stated above, under this proposed ICR, systems using surface
water and serving more than 100,000 people would be required to monitor
their finished water for the entire set of pathogens and indicators if
any pathogen density in the source water were to exceed one/liter.
Since pathogen occurrence in a particular source water can vary by
several orders of magnitude, a pathogen density of slightly greater
than one/liter during one month might be followed by considerably
greater densities in subsequent months. Requiring a system to monitor
its raw and filtered water concurrently in the months following a
source water pathogen concentration of greater than one/liter would be
more likely to result in pathogen detection in the filtered water
compared to a situation where source water pathogen densities are less
than one/liter. EPA believes that, at Giardia occurrence levels above
one/liter or virus occurrence levels above 10/liter, a 3-log Giardia
reduction or 4-log virus reduction, depending upon the efficacy of
treatment, should still be countable in the treated water. At a density
less than one/liter in source water, the sample volume needed to detect
pathogens in treated waters would be unreasonably high and technically
difficult to achieve.
To avoid virus monitoring that is likely to be uninformative
because of exceptionally good source water quality, EPA would allow two
circumstances under which a system that serves more than 100,000 people
could forgo all or part of the virus monitoring requirement. In one
case, a system that does not detect any viruses during the first twelve
months of monitoring would not be required to monitor viruses during
the last six months of monitoring. In the other case, if a system has
monitored for total coliforms or fecal coliforms in the source water
for at least five days/week every week for six months before the
effective date of this rule, and 90 percent of all samples are no
greater than 100 total coliform/100 ml or 20 fecal coliforms/100 ml,
the system may forgo the virus monitoring requirement, per approval by
EPA upon submission of this data. EPA believes that systems that do not
detect viruses during a full year of monitoring, or where the densities
of total coliforms or fecal coliforms do not exceed the values
specified in the SWTR above which a system is required to filter, could
assume that treatment that removes/inactivates Giardia satisfactorily
would also reduce viruses to a safe level.
One issue raised during rule development is whether a system could
submit previously collected monitoring data in lieu of part or all of
the data required by this rulemaking. EPA believes such data would be
useful only if (1) the laboratory used the same analytical methods
approved under this rulemaking, (2) the Agency has some assurance that
the laboratory used adequate quality assurance procedures in analyzing
the samples, (3) the system provides all data, rather than selected
data, and that these data include seasonal information, and (4) the
laboratory analyzed the full set of pathogens and bacterial indicators
required by this rule so that microbial interrelationships can be
evaluated. The Agency solicits comment on whether to allow systems to
submit previously collected data in lieu of the requirements of this
rulemaking and, if so, the appropriateness of the criteria outlined
above regarding the admissibility of such data.
Another issue is whether EPA should require systems to submit some
percentage of their processed microbiological samples to the Agency or
some other repository for archiving. Such a repository would allow EPA,
States, systems, and research centers to study the samples in the
future for any newly identified pathogens or any additional
relationships. Also, a repository could allow for very efficient
research since particular samples of interest could be selected from
the same sites based on previous ICR monitoring results. The previous
data could, in part, be validated using new analytical methods that
become available in the future. An examination of archived data may
allow EPA to require monitoring of an easily measured indicator rather
than pathogens in any future rulemaking.
If the Agency determines that archiving is appropriate, based on
public comments received, EPA would facilitate its implementation by
making any requirement as simple as possible for systems and
laboratories. For this purpose, EPA intends to serve as the repository
for all archived samples under this rule. For Giardia/Cryptosporidium
samples, systems/laboratories would collect a total volume of at least
140L and 1400L for raw and treated waters, respectively, and send
approximately one-fourth of the sample concentrate (\1/4\ of the
pellet), i.e., about 5 ml of sediment in 5 ml of formalin, to EPA for
archiving under refrigeration. For viruses, systems/laboratories would
collect a total volume of at least 200L and 1400L for raw and treated
waters, respectively, and ship a 100-ml filter eluant (pH neutralized)
on dry ice to EPA for each sample.
EPA solicits comment on the feasibility and utility of archiving
samples.
EPA also requests comment on the option for requiring systems to
collect particle size count data within the treatment plant in lieu of,
or in addition to, finished water monitoring for Giardia and
Cryptosporidium. The intent of the finished water monitoring is to
provide data on removal efficiencies throughout the treatment process,
and applicability of pathogen removal credits for various treatment
processes. However, because suspended solids in some source waters may
clog the filters and thus limit the sample volume collected, systems
may only be able to determine an upper limit for pathogen
concentration, i.e., less than the detection limit, rather than an
actual concentration. This problem would preclude a system from
calculating pathogen reduction efficiencies by treatment. Additionally,
the analytical method currently specified does not clearly
differentiate between live or dead cysts/oocysts of Giardia and
Cryptosporidium. Potential public misunderstandings of cysts/oocysts
detected in plant effluent is another reason to allow particle count
data.
Removal efficiencies indicated by particle count data may
approximate removal efficiencies of Giardia cysts and Cryptosporidium
oocysts. Particle size counting may be used as a tool for evaluating
removal efficiencies of physical removal processes. Ongoing research
may provide enough information to establish a quantitative relationship
between reductions by treatment of particle counts of specific size and
reductions of Giardia cysts and Cryptosporidium oocysts. Due to
recovery problems of Giardia and Cryptosporidium by the methodology and
the inability to quantitate removal efficiencies in many waters, the
use of particle counts in the same or smaller size range as Giardia and
Cryptosporidium may be a better method for the evaluation of removal
efficiencies by treatment.
The intent of the option for allowing particle size measurements in
lieu of finished water monitoring for Giardia and Cryptosporidium is to
obtain data on the use of particle count data as a surrogate for
Giardia and Cryptosporidium removal. Under this option particle counts
would be taken on the plant influent, settled water, filter effluent,
and plant effluent. The particle count data would be taken on the same
day as the plant influent data for Giardia and Cryptosporidium.
The particle count data would be recorded on a form similar to that
shown in Appendix A of this preamble. The data would be recorded as
particle size counts for each treatment step between the plant influent
and effluent. By requiring particle size counts in increments of
``greater than'' values for some specified volume of flow, removal
efficiency for a specified particle size range (e.g., 5-10 m),
could be calculated for a particular treatment process. This would be
done by subtracting the count in the higher size range (e.g., >10
m) from the count in the lower size range (e.g., >5
m) for the effluent of one treatment process (or the raw
water) and comparing this value, ``a'', to a similarly calculated
value, ``b'', for a subsequent treatment process (i.e., [``a'' -
``b'']/``a'' x 100). Removal efficiencies calculated based upon
particle size counts in the ranges of 2-5 m and 5-10
m, as indicated in Appendix A of this preamble, may be
conservative indicators for estimating the removal efficiency of
Giardia or Cryptosporidium which are generally in the respective size
ranges of 3-7 microns and 8-12 microns, respectively.
EPA solicits comment on the following issues pertaining to
monitoring of particle size counts: Under what circumstances, if any,
should monitoring of particle size counts be allowed in lieu of
monitoring finished water for Giardia and Cryptosporidium? What
particle size ranges and sample volumes should be monitored? What
analytical method(s), including instrumentation, should be used for
such monitoring? What criteria should be specified to ensure that
particle size data collected from different systems could be
appropriately compared? What criteria should be specified to ensure
that the particle size measurements would be most representative of
removal of Giardia and Cryptosporidium? Should methods in addition to,
or in lieu of, particle size counting, such as Microscopic Particulate
Analysis (MPA), be included as a condition for avoiding finished water
monitoring of Giardia and Cryptosporidium?
3. Reasons for Monitoring Listed Pathogens and Indicators
EPA would require monitoring of Giardia concentrations because this
pathogen causes more reported waterborne disease outbreaks than any
other single known pathogen and is more resistant to environmental
stresses and disinfection than almost all other known waterborne
pathogens. The Agency would require monitoring of Cryptosporidium
because this pathogen has caused major waterborne disease outbreaks in
the United States, England, and elsewhere, and is even more resistant
to disinfection than is Giardia. Cryptosporidium may also not be as
readily removed by filtration as Giardia, given its smaller size.
A number of enteric viruses have caused waterborne disease and they
may be responsible for many, if not most, of the outbreaks where a
causative agent was not specifically identified (about half of all
reported outbreaks). EPA believes, however, that it would be
prohibitively expensive to monitor for all of them, using current
technology. Moreover, adequate analytical methodology is not yet
available for routine analysis for many of them. For this reason, the
Agency would require systems to monitor total culturable viruses (as
determined by BGM (Buffalo Green Monkey) tissue cultures), a group of
enteric viruses that are commonly found in fecally polluted waters and
which EPA believes are at least somewhat representative of other
pathogenic enteric viruses. Total culturable viruses contain some
strains that are capable of causing waterborne disease, have been
widely studied for many years, and analytical methods are far better
defined for them than is the case for many specific enteric viruses.
EPA believes that monitoring for total culturable viruses is useful
both because this group of viruses contains pathogens and is a
potential indicator for other viral pathogens.
Some individuals believe that systems which satisfactorily control
for Giardia cysts will adequately control for pathogenic viruses, since
viruses generally are much less resistant to disinfection than are
Giardia cysts, and thus virus monitoring is not warranted under this
rulemaking. They point out that, based on the Guidance Manual to the
Surface Water Treatment Requirements (EPA, 1991), the disinfection CT
values (disinfection concentration in mg/l x disinfection contact time
in minutes) for achieving the SWTR compliance level inactivation of
viruses, which is based on hepatitis A inactivation data, is about one
to two orders of magnitude below that for achieving the SWTR compliance
level of inactivation of Giardia.
EPA, however, does not believe that sufficient data are yet
available to conclude that the Giardia density in source waters is an
adequate gauge to define the necessary treatment for viruses in all
types of source waters. The Agency is not aware of data on relative
densities between Giardia and viruses in source water. If the virus
concentration in some source waters greatly exceeds that of Giardia,
and some pathogenic viruses are significantly more resistant to
disinfection than is hepatitis A, an adequate treatment for Giardia may
not result in adequate control of viruses. Moreover, the Agency notes
that viruses have often been detected in fully treated waters (i.e.,
coagulation, sedimentation, filtration, and disinfection) (Gerba and
Rose, 1990; Payment et al., 1985; Hurst, 1991), and it is not aware of
any data demonstrating that viruses in raw water or treated water are
usually or always accompanied by Giardia cysts. The Agency also notes
that the CT values for viruses in the Guidance Manual to the SWTR (EPA,
1991) were based upon laboratory studies on free (i.e., non-aggregated)
viruses; in environmental waters, viruses are usually aggregated or
associated with cell debris, some of which may not be removed entirely
by filtration processes. Such cell-associated aggregates are
considerably more resistant to disinfection than free viruses
(Williams, 1985; Sobsey et al., 1991). Moreover, some pathogenic
enteric viruses may be substantially more resistant to disinfection
than others (Keswick et al., 1985).
Because of these uncertainties, it may not be appropriate to assume
that by controlling Giardia densities, systems will adequately control
viral pathogens. EPA needs monitoring data from many systems nationwide
to determine the level of treatment needed to control viruses.
Specifically, the Agency needs to determine the extent to which Giardia
are present in source waters when viruses are present. The Agency also
needs to determine what minimum level of disinfection inactivation is
necessary for surface water supplies to ensure adequate virus control,
regardless of Giardia densities. These data will allow the Agency to
determine whether a system that consistently provides an overall
Giardia reduction of 3-logs (of which at least 0.5-log is due to
disinfection alone) or any greater reduction level for Giardia, will
also consistently provide an adequate control for viruses, especially
in cases where virus densities in source waters are much higher than
those for Giardia. Information collected under this rule would provide
part of these data. The Agency believes that these data, along with a
more intensive voluntary monitoring effort among a small number of
systems, should clarify this situation sufficiently to allow it to
develop suitable revisions to the SWTR.
With regard to bacterial pathogens, EPA believes that pathogenic
protozoa and many waterborne viruses are more resistant to
environmental stress and disinfection than most enteric bacteria that
cause waterborne disease. Thus a system that protects the public from
pathogenic protozoa and viruses will concurrently protect them from
most pathogenic bacteria (except possibly for those bacteria that can
proliferate within the distribution system or which have special
protective factors). For this reason, EPA would not require these
systems to monitor pathogenic bacteria in the source water or in
treated water.
While EPA would not require systems to monitor pathogenic bacteria,
the Agency would require them to monitor potential bacterial indicators
for waterborne pathogens in source water and treated water. Under this
rule, EPA is proposing to require systems to monitor for total
coliforms and either fecal coliforms or E. coli. Total coliforms and
fecal coliforms have been used widely for decades to assess source
water quality, testing for these two groups of bacteria is very simple
and inexpensive, and systems are familiar with these tests. Total
coliforms are usually much more numerous in water than fecal coliforms,
and therefore enumeration in source waters and treated water is more
sensitive than with fecal coliforms. However, fecal coliforms are a
better indicator of fresh fecal contamination than are total coliforms.
Because the bacterium E. coli is more closely related to fresh fecal
pollution and to gastrointestinal illness among bathers than are fecal
coliforms, EPA would allow a system to analyze for E. coli in lieu of
fecal coliforms.
EPA solicits comment on the requirement to monitor the specific
pathogens and bacterial indicators mentioned above. The Agency
specifically seeks comment on whether to require systems to monitor
both fecal coliforms and E. coli, rather than one or the other. In
addition, the Agency may include a requirement to monitor for two other
potential indicators--Clostridium perfringens (C. perfringens) and
coliphage which are discussed below.
Clostridium perfringens. C. perfringens is a bacterium that is
common in the intestinal tract of warm-blooded animals. This organism
forms an endospore in the environment that is extremely resistant to
environmental stresses and disinfection. Of the more than 60 species of
Clostridium, C. perfringens is the one most consistently associated
with human fecal wastes (Cabelli, 1977). It is consistently present in
human feces at a relatively high density (Bisson and Cabelli, 1980) and
appears to be excreted in greater numbers than are fecal pathogens
(NATO, 1984). There is controversy over whether other important animal
hosts exist, since C. perfringens spores are widely found in
terrestrial and aquatic environments (Cabelli, 1977). The survivability
of C. perfringens spores in water and their resistance to treatment
compared to the pathogens is much greater than other indicators (Bonde,
1977), except possibly for Giardia and Cryptosporidium. Analysis is
relatively easy and inexpensive. The European Community has a
supplementary standard for the endospores of sulfite-reducing
Clostridium for drinking waters.
Recently, Payment and Franco (1993) published a paper that showed
that C. perfringens may be a suitable indicator for viral and protozoan
pathogens in both raw water and filtered water. In this study, the
investigators collected large-volume samples from three water treatment
plants and analyzed them for Giardia cysts, Cryptosporidium oocysts,
cultivable human enteric viruses, Clostridium, and somatic and male-
specific coliphage. They found that Clostridium densities were
significantly correlated with the densities of viruses, cysts, and
oocysts in river water and with viruses and oocysts (but not Giardia
cysts) in filtered water.
For the above reasons, EPA is considering a requirement that
systems monitor their source and filtered water for C. perfringens at
the same frequency as is being proposed for the other organisms. C.
perfringens may be appropriate as a low cost monitoring indicator for
estimating pathogen densities in the source water and/or for defining
treatment effectiveness. If feasible, such an indicator could greatly
reduce monitoring costs for determining appropriate levels of treatment
to address microbial concerns. This would be of special benefit for
smaller systems under the long-term ESWTR. EPA solicits comment on this
issue.
Coliphage. The Agency also seeks comment on the utility of
coliphage as an indicator of pathogen presence. Coliphages, which are
viruses that infect the bacterium E. coli, are far simpler to analyze
than other viruses and are, like E. coli, generally associated with
fecal contamination. They have often been discussed as a possible
indicator of treatment effectiveness for enteric viruses. Coliphages
are commonly categorized into two groups: the somatic phage and the
male-specific (or F-specific) phage. The somatic phage gain entry into
E. coli cells via the cell wall, while the male-specific phage gain
entry only through the sex-pili of those E. coli cells that have them
(referred to as male cells).
Because coliphages are so much simpler to analyze than human
viruses, EPA wants to determine whether systems can use coliphages to
indicate the presence of the human viruses in source waters and
filtered water. Data on relative densities in natural waters are
sparse. Somatic phages are common in the feces of humans and other
animals but, unlike human viruses, some of them apparently can multiply
in natural water, probably in species other than E. coli. Male-specific
phages are not common in humans and other animals, but are common in
sewage, suggesting they can multiply in the sewerage system (IAWPRC,
1991). Data on the relative resistance and removal of coliphages and
human viruses during the water treatment process is also scarce, and
the data which exist are inconsistent, especially for the somatic
phages (IAWPRC, 1991). Some of the male-specific phages (e.g., MS2),
however, appear to be more resistant to chemical disinfection than most
waterborne pathogens (Sobsey, 1989).
One recent study suggests that coliphages are suitable as an
indicator for viruses, at least in filtered water. In the Payment and
Franco (1993) study indicated above, the densities of somatic
coliphages (E. coli CN13 host) were statistically correlated with human
enteric viruses and Cryptosporidium oocysts (but not Giardia cysts) in
filtered water, and not in river water. Male-specific coliphages
(Salmonella typhimurium WG49 host) were correlated with human enteric
viruses in filtered water, but not in river water. The male-specific
coliphages were also correlated with Giardia cysts, but not
Cryptosporidium oocysts, in river water.
In another study, Havelaar et al. (1993) compared the
concentrations of culturable viruses (BGM cell line) with those of
thermotolerant coliforms, fecal streptococci, and male-specific RNA
phages (Salmonella typhimurium WG49 host) for a variety of water types.
The investigators found that the male- specific phages were
significantly correlated (significant at P g/ml of 4-
methylumbelliferyl-beta-D-glucuronide (MUG), as specified in
Sec. 141.21(f)(6)(i). In this method, each total coliform-positive
broth culture from the Multiple Tube Fermentation (MTF) Technique
(Sec. 141.74(a)(2)) or each total coliform-positive colony from the
Membrane Filter Technique (Sec. 141.74(a)(2)) is transferred to 10 ml
of EC + MUG. After incubation, the inoculated medium is examined with
an ultraviolet light. If fluorescence is observed, the medium contains
E. coli.
(2) Nutrient agar supplemented with 100 g/ml of MUG, as
specified in Sec. 141.21(f)(6)(ii), with the additional requirement
that E. coli colonies be counted.
(3) Minimal Medium ONPG-MUG Test, often referred to as the Colilert
Test, as specified in Sec. 141.74(a)(2), with the additional
requirement that total coliform-positive tubes be examined with an
ultraviolet light. If fluorescence is observed, the medium contains E.
coli.
Giardia, cryptosporidium, and total culturable viruses. In August
1993, EPA sponsored a workshop of invited experts in Giardia,
Cryptosporidium, and virus analysis and quality assurance procedures to
help the Agency develop standardized methods for these organisms for
use with the ICR. Workshop participants included representatives from
academia; water industry; commercial laboratories; and federal, State
and local governments. As the basis for the discussion, the workshop
used the Giardia/Cryptosporidium method published by ASTM (1992) and
the method to be published shortly in the 18th edition Supplement to
Standard Methods for the Examination of Water and Wastewater. Two virus
methods in the 18th edition of Standard Methods (Method 9510C for virus
collection and elution; Method 9510G for virus assay) (APHA, 1992) were
used. The methods in ASTM (1992) and Standard Methods were used as the
basis for this discussion because these texts are highly respected and
widely used references that have been peer-reviewed throughout the
scientific community. The workshop generally recommended use of the
methods above, but, because these methods allow many sub-options,
decided to refine and standardize them to achieve more precise
comparisons among systems under the ICR (USEPA, 1993a).
The method for Giardia/Cryptosporidium, as revised, is in Appendix
C of the proposed rule. This method includes sample collection,
purification, and microscopic assay, and allows the density of Giardia
and Cryptosporidium to be determined simultaneously on the same sample.
The microscopic assay includes the use of epifluorescence along with
differential-interference- (or Hoffman Modulation) contrast optics to
identify morphological characteristics.
One issue with regard to the Giardia/Cryptosporidium method
concerns how to express the results. The total number of cysts and
oocysts are counted, based on immunofluorescence, size, shape, and
presence of internal structures. Then the total number of cysts with
internal structures is tallied. The issue is what terminology to use
for these two steps. One procedure is to categorize the first step as a
``presumptive'' test and the second step as the ``confirmed'' test. The
terminology ``confirmed'' could be used if at least two internal
structures are identified as being Giardia/Cryptosporidium cysts/
oocysts. The second procedure is to categorize the first step as the
``total number of cysts and/or oocysts per 100L'' (which would be
equivalent to ``presumptive'') and the second step as the ``total
number of cysts and/or oocysts with internal structures.'' The
terminology ``with internal structures'' could be used if at least one
internal structure is identified as being Giardia/Cryptosporidium
cysts/oocysts.
The rationale for considering the two steps as presumptive and
confirmed is: (1) Some algal and yeast cells recovered with this
procedure cross-react with the protozoan monoclonal antibodies used,
(2) many algae and other particles autofluoresce and thereby confuse
the analyst, and (3) depending upon the criteria that will be used for
defining level of treatment requirements in the interim ESWTR, use of
the terminology ``confirmed'' may reduce the number of false positives
and thereby not lead to excessive levels of treatment to achieve the
desired health risk goal. However, the use of these terms is somewhat
inaccurate in that it diminishes the importance of the total count
(i.e., the presumptive test). The confirmed test only reflects those
particles where internal structures can be specifically observed, which
may represent only a small fraction of the cysts/oocysts on the slide.
EPA requests comment on which terminology is most suitable for
referring to the two steps.
Other methods for the assay of Giardia and Cryptosporidium are
currently being developed. One of these assays (the electrotation
assay) is based on the observation that particles in a rotating
electric field also rotate if the frequency is right. In addition to
this assay, other potential assays for the protozoa include polymerase
chain reaction and flow cytometry. The Agency requests comment about
the most appropriate means for incorporating new and easier analytical
methods for Giardia and Cryptosporidium into the ICR.
The method for viruses, as revised, is in Appendix D of the
proposed rule. This method relies on a most probable number technique
using BGM tissue culture monolayers, with cytopathic effect (CPE) as
the sole enumeration endpoint. Attendees at the workshop considered
plaque-forming units (PFU) as an endpoint, but rejected it. Although
the PFU endpoint can be determined without the use of a microscope,
unlike the CPE endpoint, it may not be as sensitive as CPE, i.e., use
of CPE should result in greater virus densities. The workshop members
determined that sensitivity was more important than precision in
quantitation for comparing virus and protozoan data to determine the
appropriateness of using Giardia and possibly Cryptosporidium as the
primary target organism(s) for defining adequacy of treatment.
Clostridium perfringens. If EPA decides to require systems to
monitor Clostridium perfringens, as was discussed in Section IIIA3
above, the Agency would also specify a method for this bacterium. The
Agency believes that the most appropriate method is a membrane filter
procedure using M-CP medium (Bisson and Cabelli, 1979), possibly as
modified by Armon and Payment (1988). The Agency solicits comment on
whether this method is most suitable for monitoring Clostridium
perfringens. The Agency notes that this organism must be grown under
strict anaerobic conditions (i.e., without oxygen).
Coliphage. If EPA decides to require the monitoring of somatic
coliphages and/or male-specific coliphage, as was discussed in Section
IIIA3, the Agency believes that the most appropriate method is a simple
agar overlay procedure. For somatic phage testing, the Agency believes
that the most suitable host is E. coli C. The Agency solicits comment
on whether this procedure and host are most suitable for monitoring the
somatic coliphage. The Agency also seeks comment, with data, on what
bacterial host is most suitable for monitoring male-specific
coliphages. The method for sample collection, sample processing, and
assay for somatic and male-specific coliphage is presented in Appendix
D of the proposed rule.
EPA requests comment on the appropriateness of the above methods.
7. Laboratory Approval
General. EPA is developing a program for approving laboratories to
analyze the pathogens that would be monitored under this rule. This
program would ensure that these laboratories are competent to perform
the analyses. Analytical skill is especially important for the
difficult and sophisticated processing and analyses specified for the
total culturable viruses and Giardia and Cryptosporidium. Another
prominent reason for approving laboratories is to ensure that
laboratory procedures are as standardized as possible for uniform data
comparison among systems.
Currently, EPA has a laboratory certification program for drinking
water analyses. All laboratories that analyze drinking water samples to
determine compliance with MCLs must be certified by EPA or the State,
as specified by 40 CFR 142.10(b)(4) and 141.28. Under this program, EPA
certifies the principal State laboratory and, with certain exceptions
(see 40 CFR 142.10), each State certifies all drinking water
laboratories within the State. Laboratories certified to perform
analysis for coliforms under the Total Coliform Rule would be approved
to analyze for total coliforms, fecal coliforms, and E. coli under the
ICR without further action. The current program does not address
pathogens.
Rather than broaden the present laboratory certification program to
include Giardia, Cryptosporidium, and the viruses, EPA believes that it
would be more appropriate to develop a separate program and to
differentiate the two programs by using the term laboratory
``approval'' instead of ``certification'' to refer to laboratories
performing pathogen analyses required by the ICR. The rationale for
this approach is that (1) EPA expects that only a small number of
laboratories will be qualified to perform analyses for the protozoa and
viruses because of the complexity of the methods, (2) few States and
EPA Regions are currently able to certify laboratories for the
pathogens of interest, and (3) the short time constraints for
implementing this rule and the short-term nature of the sampling (up to
18 months) do not provide time for a full certification program.
Nevertheless, EPA is proposing to use several major elements of the
current certification program in its program to ``approve''
laboratories for pathogen analysis, including performance evaluation
(PE) samples, training, and on-site evaluations. If an interim or long-
term ESWTR were to require some systems to monitor the same pathogens
as those specified by the ICR, then the laboratory approval criteria
would probably be incorporated into the drinking water laboratory
certification program.
Performance evaluation samples. Under the laboratory approval
program proposed herein, a laboratory would need to analyze
satisfactorily a set of PE samples to become approved and subsequent
sets of PE samples (e.g., 6, 12, 18 months) to maintain approval.
Workshop participants recommended that a set of PE samples for Giardia/
Cryptosporidium consist of (1) a mixture of Giardia cysts and
Cryptosporidium oocysts, (2) a mixture of Giardia cysts and
Cryptosporidium oocysts plus algal cells, and (3) algal cells only
(negative control). According to workshop recommendations, a set of PE
samples for viruses should include virus samples of varying titers
(concentrations) that the laboratory would process as if they were
filter eluates. Currently, EPA is developing a PE sample program
intended to satisfy these recommendations.
Training. In addition to PE samples, at least one principal analyst
in each laboratory would need to complete an EPA-specified training
course or meet the requirements of equivalent training, as defined by
the Agency. Although EPA has not yet defined ``equivalent training'',
the Agency is considering an approach involving a training video or an
apprenticeship with an expert. EPA is developing two training courses--
one in Giardia/Cryptosporidium analysis, and the other in environmental
virus analysis. Each of these courses would also include training in
sample collection.
On-site evaluation. EPA is also proposing to require a laboratory
to pass an on-site evaluation before receiving approval. The EPA
Regional Administrator would be the ultimate approval authority. The
Agency would develop criteria for determining whether an individual has
the necessary expertise to conduct the intended tests.
The Agency has drafted a laboratory approval manual that lists the
specific criteria that an on-site evaluator would examine. These
criteria are based on workshop recommendations. This manual, which is
available in the Water Docket, includes a number of certification
criteria from Chapters III and V of EPA's laboratory certification
manual (USEPA, 1990). For example, as part of the on-site evaluation,
the certification officer would ensure that the laboratory has prepared
and is using a written laboratory Quality Assurance Plan. This plan is
described in EPA's laboratory certification manual (Chapter III). Some
draft criteria pertaining to the qualifications of laboratory personnel
are indicated below.
For Giardia and Cryptosporidium analysis:
Technician: This person performs at the bench level and
is actively involved in collecting samples, extracting filters, and/or
processing the filter eluent for Giardia/Cryptosporidium analysis. The
technician must have two years of college (full time) in life sciences
or a related field.
Analyst: This person must have two years of college (full
time) in the life sciences or a related field and have at least three
months experience in examining indirect fluorescent antibody stains
under the microscope.
Principal Analyst/Supervisor: This person is a qualified,
experienced microbiologist with a minimum of a B.A./B.S. degree in
microbiology or a closely related field. The principal analyst must
have completed the ICR protozoan training course (mentioned above) or
have equivalent experience, as approved by EPA.
For virus analysis:
Technician: This person extracts the filter and processes
the sample, but does not perform tissue culture work. The technician
must have at least three months experience in filter extraction of
virus samples and sample processing.
Analyst: This person performs at the bench level and is
involved in all aspects of the analysis, including sample collection,
filter extraction, sample processing, and assay. The analyst must have
two years of college (full time) in the life sciences or at least six
months of bench experience in cell culturing and animal virus analyses.
Principal Analyst/Supervisor: This person is a qualified,
experienced microbiologist who oversees the entire analysis. The
individual must have a B.A./B.S. degree in the life sciences with three
years experience in cell culture and animal virus analyses. This
individual must have completed the ICR environmental virology training
course or have equivalent experience, as approved by EPA.
Because of the tight time constraints and the limited number of
national experts capable of participating in on-site evaluations, EPA
proposes to give highest priority in evaluating those laboratories
(e.g., commercial, academic, utility, State) that (1) have been
analyzing Giardia and Cryptosporidium or virus samples for at least one
year, (2) have nationally recognized experts in protozoan or virus
analyses, or (3) have the technical capability, capacity, and
willingness to analyze at least four samples/month under the ICR
requirements for Giardia and Cryptosporidium or viruses.
Laboratory capacity. If, following the beginning effective date of
this rule, a system cannot locate an approved laboratory to analyze its
water samples for the indicated pathogens, the system would be required
to notify EPA in writing (see Section III.C). EPA will inform the
system which laboratories are available for performing the requisite
analysis, or when new approved laboratories become available to do such
analysis.
EPA solicits comment on the approach above for approving
laboratories and, more broadly, on the most appropriate means for
ensuring that laboratories performing the pathogen analyses are
competent. Laboratories wishing to become approved for doing these
analyses should contact ICR Laboratory Coordinator, USEPA, Office of
Ground Water and Drinking Water, Technical Support Division, 26 West
Martin Luther King Drive, Cincinnati, Ohio 45268, for an application
form to initiate the approval process.
8. Quality Assurance
Sample collection. For the collection of samples for pathogens, the
laboratory would document that each sample collector, either from the
laboratory or the system, is properly trained. Without such
documentation, the laboratory would not proceed with analyzing the
system's samples. EPA encourages approved laboratories to provide
adequate training, if needed, not only to laboratory sample collectors,
but to individuals at client water systems who collect their own
samples for pathogens. Other criteria for sampling are included in the
draft laboratory approval manual mentioned in Section 7, above.
Data reporting. EPA proposes to require a laboratory to submit data
results to both the Agency and the client system for the pathogens. The
water system would also be required to submit the same data results to
the Agency. By receiving and comparing both data submissions, EPA can
reduce reporting errors. EPA would require that systems report data in
a computer-readable form; in addition, systems serving at least 100,000
people would be required to report data in an EPA-specified electronic
format (see Section III.B6 for more discussion). EPA encourages systems
serving 10-100,000 people to also submit data using the electronic
format.
EPA also proposes to require a laboratory, when the laboratory
submits pathogen data to the Agency, to include its results on the most
recent set of PE samples for that pathogen. This quality assurance
criterion would allow EPA to assess the quality of that data,
especially if the data appear to be atypical or equivocal.

B. Stage 2 Disinfection By-Products Rule

1. Need for Additional Data
When drinking water is disinfected, the organic material and
bromide that are naturally present in the water react with the
disinfectant to form hundreds of DBPs. Only a small subset of these
chemicals have been identified due to the complexities of measuring
them. Many of them are not stable, so they decompose during the
sampling or analytical process. Others are polar and so are not easily
extracted from the water for further analysis.
Most of the DBPs that can be measured in drinking water (i.e.,
there are analytical techniques available to detect them) are
byproducts from the use of chlorine. However, there is limited
occurrence information on even these DBPs, so the extent of exposure
cannot be estimated. Only a subset of them have been studied to
determine whether exposure to them presents a risk to health.
Several DBPs were included on the 1991 Drinking Water Priority List
(56 FR 1470; January 14, 1991), as candidates for future regulations.
During development of the proposed Stage 1 D/DBP Rule, the Negotiating
Committee did not believe there were adequate data available to address
most of the DBPs on the Priority List, so MCLs were recommended for a
subset of the Priority List DBPs (trihalomethanes [THMs], haloacetic
acids [HAAs], chlorite and bromate). The Stage 1 D/DBP Rule would
address the ``other'' DBPs in two ways: 1) EPA would assume that
control of other Priority List DBPs would occur if systems could meet
the MCLs for THMs and HAAs; and 2) EPA would require some surface water
systems using conventional treatment to implement optimized coagulation
to remove as much organic material as possible before disinfection,
thereby minimizing the formation of all DBPs. Total organic carbon
(TOC) was designated as the surrogate for the organic precursor
material removed during optimized coagulation.
Many members of the Negotiating Committee expressed concern on the
adequacy of data to support the use of surrogate limits such as TOC for
inclusion in the Stage 1 regulatory criteria. The lack of field data
led the Negotiating Committee to base its decisions on the Stage 1 D/
DBP Rule using a water treatment plant model to predict DBP
concentrations resulting from various changes in treatment practices.
The THM and HAA compliance monitoring requirements being considered
for proposal in the Stage 1 D/DBP Rule were modeled after the
requirements of the current Total Trihalomethane (TTHM) Rule (44 FR
68624, November 1979). Some members of the Negotiating Committee were
concerned that quarterly monitoring for THMs and HAAs would not
accurately reflect consumer exposure to DBPs. An under-prediction of
consumer exposure would be especially serious if research indicated
there were short-term adverse health effects from exposure to DBPs.
Field data were not available to assess the spatial and seasonal
variability of DBP concentrations within distribution systems. Data
were also lacking concerning the usefulness of surrogates, such as
total organic halide (TOX), as tools for reducing compliance monitoring
costs.
As a result of the above uncertainties, the Negotiating Committee
strongly recommended that additional information be collected and
analyzed to assist in the development of a Stage 2 D/DBP Rule. Field
data are needed to: (1) Characterize source water parameters that
influence DBP formation, (2) determine the concentrations of DBPs in
drinking water, (3) refine models for predicting DBP formation based on
treatment and water quality parameters, and (4) establish cost-
effective monitoring requirements that are protective of the public
health. Today's proposed rule would provide EPA with the data necessary
to accomplish the above tasks.
2. Monitoring and Reporting Requirements and Rationale
The rule would require all community and nontransient, noncommunity
systems serving at least 100,000 persons to: (1) Perform the monitoring
summarized in Table III.1-.2 and (2) report treatment plant operational
data specified in Table III.6. Treatment plants that use alternate
disinfectants (chloramines, ozone, or chlorine dioxide) or hypochlorite
solutions would also be required to perform monitoring for DBPs that
are of particular concern for the disinfectant being used. Community
and nontransient, noncommunity systems that use groundwater not under
the direct influence of surface water and serve between 50,000 and
99,999 persons would be required to conduct monthly monitoring for
total organic carbon (TOC) in water entering the distribution system.

Table III.1.--Sampling Points for All Systems
------------------------------------------------------------------------
Sampling point Analyses\1\ Frequency
------------------------------------------------------------------------
Treatment plant influent.. pH, alkalinity, turbidity, Monthly.
temperature, calcium and
total hardness, TOC,
UV254, bromide, and
ammonia.
Treatment plant influent Optional oxidant demand Monthly.
(optional for waters with test.
high oxidant demand due
to the presence of
inorganics).
Treatment plant influent.. TOX....................... Quarterly.
After air stripping....... Ammonia................... Monthly.
Before and after pH, alkalinity, turbidity, Monthly.
filtration. temperature, calcium and
total hardness, TOC, and
UV254.
At each point of pH, alkalinity, turbidity, Monthly.
disinfection\2\. temperature, calcium and
total hardness, TOC, and
UV254.
At end of each process in Disinfectant residual\3\.. Monthly.
which chlorine is applied.
After filtration (if THMs, HAAs(6), HANs, CP, Quarterly.
chlorine is applied prior HK, CH, and TOX.
to filtration).
Entry point to pH, alkalinity, turbidity, Monthly.
distribution system. temperature, calcium and
total hardness, TOC,
UV254, and disinfectant
residual\3\.
Entry point to THMs, HAAs(6), HANs, CP, Quarterly.
distribution system. HK, CH, TOX, and SDS\4\.
4 THM Compliance THMs, HAAs (6), HANs, CP, Quarterly.
Monitoring Points in HK, CH, TOX, pH,
Distribution System (1 Temperature, Alkalinity,
sample point will be Total Hardness and
chosen to correspond to Disinfectant Residual\3\.
the SDS sample,\4\ 1 will
be chosen at a maximum
detention time, and the
remaining 2 will be
representative of the
distribution system).
------------------------------------------------------------------------
\1\TOC is total organic carbon. UV254 is absorbance of ultraviolet light
at 254 nanometers. THMs are chloroform, bromodichloromethane,
dibromochloromethane, and bromoform. HAAs(6) is mono-, di-, and
trichloroacetic acid; mono- and di- bromoacetic acid; and
bromochloroacetic acid. HANs are dichloro-, trichloro-, bromochloro-,
and dibromo- acetonitrile. CP is chloropicrin. HK is 1,1-
dichloropropanone and 1,1,1- trichloropropanone. CH is chloral
hydrate. TOX is total organic halide. SDS is the simulated
distribution system test.
\2\For utilities using ozone or chlorine dioxide, Tables III.4 and
III.5, respectively, show additional monitoring requirements at this
sampling point.
\3\Free chlorine residual will be measured in systems using free
chlorine as the residual disinfectant; total chlorine residual will be
measured in systems using chloramines as the residual disinfectant.
\4\The SDS (simulated distribution system test) sample will be stored in
such a manner that it can be compared to the results from one of the
distribution system sampling points. This distribution system sampling
point will be selected using the following criteria: 1) No additional
disinfectant added between the treatment plant and this point, 2)
Approximate detention time of water is available, and 3) No blending
with water from other sources. The SDS sample will be analyzed for
THMs, HAAs(6), HANs, CP, HK, CH, TOX, pH and disinfectant residual.
\5\Five THM samples.

Monitoring of source water quality. EPA would require all community
and nontransient noncommunity water systems serving at least 100,000
persons to conduct monthly monitoring of the raw water entering each
treatment plant for pH, alkalinity, turbidity, temperature, calcium and
total hardness, total organic carbon (TOC), ultraviolet absorbance at
254 nm (UV254), bromide ion, and ammonia. If the raw water were to
contain a sufficiently high concentration of inorganic chemicals (i.e.,
hydrogen sulfide, iron, manganese) to cause a high oxidant demand, then
the system would be encouraged to monitor for this inorganic oxidant
demand at the same frequency. Systems would collect samples from the
plant influent after water from multiple sources is blended. The
sampling point would be before the first treatment step to characterize
the chemical quality of the water being treated. A system that uses
ground water not under the direct influence of surface water and with
multiple wells in the same aquifer would only be required to collect
raw water samples from representative wells in the two aquifers serving
the largest portion of the system's population.
The above parameters were selected because they influence the
quantity and chemical character of the DBPs formed when the
disinfectant is added to the water. High oxidant demand water should be
characterized because the availability of the disinfectant for reaction
with organic material to form DBPs will depend on the amount of
disinfectant that is consumed by inorganic chemicals. EPA solicits
comments on the definition of high oxidant demand water and the type(s)
of measurements necessary to characterize it.
Monthly sampling at the treatment plant influent would provide an
estimate of the variability in raw water quality. EPA would use data
from this portion of the rule to characterize source water parameters
that influence DBP formation.
Monitoring within the treatment plant. EPA would require systems
serving at least 100,000 people to monitor for most of the same
parameters at several points within the treatment plant. These
requirements are summarized in Table III.1. Samples from representative
points before and after the filters collected on a monthly basis would
be measured for pH, alkalinity, turbidity, temperature, calcium and
total hardness, TOC, and UV254. These measurements would provide
data on changes in water quality between the plant influent and the
last filtration step. Of particular importance are data on how the
organic precursor material (as represented by TOC and UV254) is
removed prior to and through filtration.
Monthly monitoring of the same parameters (pH, alkalinity,
turbidity, temperature, calcium and total hardness, TOC, and
UV254) would be required at each point of disinfection. These data
are critical, because most data now available for comparing these
variables with DBP concentrations are based on source water data. Most
utilities do some treatment of the water prior to the addition of
disinfectant, so source water measurements do not accurately reflect
the quality of the water when the disinfectant is added. These data
would provide a more accurate determination of how these parameters
influence DBP formation.
Disinfectant residuals would be measured monthly at the end of each
treatment process in which chlorine is applied. Free and total chlorine
residual would be reported if free chlorine is used as the
disinfectant; total chlorine residual would be reported if ammonia is
added in combination with chlorine or when sufficient ammonia is
present in the source water that breakpoint chlorination is not
achieved. These data combined with information on the applied
disinfectant dosages and contact times (from the plant operational data
discussed in the next section) would give a more accurate picture on
DBP formation, because the chlorine or chloramine demand of the water
can be estimated. Part of this demand is reflected in the formation of
DBPs.
If a water plant practices air stripping to remove volatile organic
compounds (VOCs) from the raw water prior to the addition of a
disinfectant and the raw water contains ammonia, then a monthly sample
collected immediately following the air stripper and analyzed for
ammonia would be required. Air stripping might change the concentration
of ammonia, and an accurate concentration of ammonia is necessary to
determine DBP formation.
EPA would also require systems serving at least 100,000 people to
analyze samples from the entry point to the distribution system
monthly. The monitoring would consist of pH, alkalinity, turbidity,
temperature, calcium and total hardness, TOC, UV254, and
disinfectant residual.
Systems are already monitoring for many of the parameters listed
above, either to comply with other drinking water regulations or for
operational considerations. Therefore, the additional costs of
providing monthly data would not be excessive for these parameters.
The monthly data from the treatment plants would provide EPA with
the necessary information to conduct two analyses essential for the
development of the Stage 2 D/DBP Rule: (1) The variability in source
water quality and treatment operation and its impacts on the parameters
that influence the formation of DBPs, and (2) when the data are
combined with the DBP data described below, EPA will have a better
understanding of how water quality and treatment practices influence
DBP formation. This understanding would allow EPA to refine models for
predicting DBP formation based on treatment and water quality
parameters and thus to further clarify the interrelationships between
disinfectant concentrations and DBPs under field conditions.
EPA would require community and nontransient, noncommunity water
systems that use only ground water not under the direct influence of
surface water and serve between 50,000 and 99,999 people to analyze TOC
samples monthly from the entry points to the distribution system.
Additional monitoring for chlorination by-products. EPA would
require monitoring for specific chlorination by-products quarterly to
fulfill three objectives: (1) To relate water quality and treatment
practices to DBP formation, (2) to determine the concentration of DBPs
in drinking water, and (3) to establish cost effective monitoring
requirements that are protective of public health. The Agency would
require analysis for the following chlorination by-products:
chloroform, bromodichloromethane, dibromochloromethane, bromoform,
monochloroacetic acid, dichloroacetic acid, trichloroacetic acid,
monobromoacetic acid, dibromoacetic acid, bromochloroacetic acid,
trichloroacetonitrile, dichloroacetonitrile, bromochloroacetonitrile,
dibromoacetonitrile, 1,1- dichloropropanone, 1,1,1-trichloropropanone,
chloropicrin, and chloral hydrate. Each time a DBP sample is collected,
the system would also be required to measure and report pH,
temperature, alkalinity, and disinfectant residual. Free chlorine
residual would be measured in systems using free chlorine as the
disinfectant. Total chlorine residual would be measured at sampling
points after the addition of ammonia, because the residual disinfectant
would be chloramines.
To relate DBP formation to water quality and treatment practices,
EPA would require systems to monitor the above DBPs at the following
locations: (1) At a representative point immediately after the last
filtration step (if chlorine is applied prior to the filters), (2) at
the entry point to the distribution system, and (3) at a TTHM
compliance monitoring sampling point in the distribution system which
can be related to a simulated distribution system (SDS) sample. This
distribution system sampling point would be selected using the
following criteria: (1) No additional disinfectant is added to the
water between entry to the distribution system and the sampling point,
(2) the approximate detention time of the water is available, and (3)
there is no blending with water from other treatment plants. A sample
would also be collected at the entry point to the distribution system
and incubated at a time and temperature corresponding to the
distribution system sample. This SDS sample would be analyzed for the
same DBPs as the distribution system sample and it would provide a
measure of DBP formation under controlled conditions. Data from SDS
samples would also be evaluated as a cost-effective alternative to
distribution system compliance monitoring.
The concentration of chlorination by-products would be determined
by requiring the utilities to conduct quarterly monitoring at four
points in the distribution system using the same criteria for sampling
point selection as specified in the THM Rule. One sample would be taken
from a point representing a maximum detention time in the system. The
sample point with the highest THM concentrations would meet this
criterion. The second sample would correspond to the SDS sampling point
described above. The remaining two points would be representative of
the distribution system. All four sampling points would be routine
sampling points for TTHM compliance monitoring. This regimen minimizes
the sampling costs, since additional sampling points are not required.
It also provides a link between the measurements made for this rule and
the historical TTHM compliance monitoring data for each system. Systems
that have two or more treatment plants serving the same distribution
system would only be required to collect four DBP samples in the
distribution system.
Six quarters of DBP monitoring would provide EPA with information
concerning the spatial and seasonal variability of DBPs within
distribution systems. In an effort to evaluate lower cost monitoring
options, EPA would also require systems to monitor total organic halide
(TOX) concentrations at the same sampling points and at the same time
DBP concentrations are measured. Total organic halide (TOX) is an
indicator of the total quantity of dissolved halogenated organic
material present in water. Essentially all of the TOX present in
chlorinated drinking water in the United States is the result of
reactions between chlorine and the organic material and bromide ion
present in the source water. The eighteen chlorination by-products
listed above typically account for less than 50% of the TOX that is
measured in chlorinated drinking water. Since TOX also includes the
halogenated by-products not routinely measured, it might be a better
surrogate of chlorination by-product concentrations than are TTHMs and
THAAs. The TOX analysis of treatment plant influent would also be
required quarterly, because the source water could contain background
concentrations of halogenated organic compounds as a result of chemical
contamination or upstream discharges of chlorinated water. The DBP,
TOX, and surrogate precursor (i.e., TOC and UV254) data will be
evaluated to determine the most cost-effective monitoring requirements
that are protective of public health.
All the samples for the above-named parameters would be collected
as close together in time as feasible (during the same working day if
possible). Samples would be collected during normal plant operating
conditions, when there were no obvious changes in source water quality
due to storm events, chemical spills, etc. The quarterly sampling for
DBPs would be conducted at the same time as the sampling from the
treatment plant(s). The quarterly samples would be collected at a time
when the source water quality and plant operations had been stable for
several days, so that the distribution system sample can be related to
the SDS sample that is collected at the same time.
Additional monitoring required for systems using chloramines. EPA
would require systems serving at least 100,000 people and using
chloramines to analyze for one additional DBP, cyanogen chloride. This
by-product is formed when chlorine reacts with organic material in the
presence of the ammonium ion (Ohya and Kanno, 1985). There are little
data available to assess the occurrence of this compound and the
factors influencing its formation are poorly understood. Therefore,
these data are necessary to determine how the distribution of by-
products would change if utilities switched from free chlorine to
chloramines as the residual disinfectant to meet the MCLs for TTHM and
THAA.
Monitoring for cyanogen chloride would be required quarterly, as
summarized in Table III.2. Only one sample would be required from the
distribution system, because of the analytical complexities of
measuring the compound. By sampling at the entry point to the
distribution system and at a point of maximum detention time, EPA would
be able to assess the concentration range at which this compound
occurs. Cyanogen chloride is very reactive, and would be expected both
to decompose and be produced within the distribution system.

Table III.2.--Additional Sampling Required of Systems Using Chloramines
------------------------------------------------------------------------
Sampling point Analyses Frequency
------------------------------------------------------------------------
Entry point to Cyanogen chloride......... Quarterly.
distribution system.
One THM compliance Cyanogen chloride......... Quarterly.
monitoring sample point
representing a maximum
detention time in
distribution system.
------------------------------------------------------------------------

Additional monitoring required of systems using hypochlorite
solutions. EPA would require systems serving at least 100,000 people
and using hypochlorite solutions for chlorination to perform the
additional monitoring presented in Table III.3. The monitoring would
include quarterly measurements for chlorate in the treatment plant
influent, hypochlorite feedstock solution, and water at the entry point
to the distribution system. Chlorate is a decomposition product found
in hypochlorite feedstock (Lister, 1956; Bolyard, et al., 1992; and
Gordon et al., 1993). Its concentration in the drinking water would not
be expected to change in the distribution system unless additional
hypochlorite solution was added, because it is not a DBP from chlorine
reactions under drinking water conditions. Quarterly monitoring of the
hypochlorite stock solution to assess the factors that influence
chlorate formation (pH, storage temperature, and hypochlorite ion
concentration) would also be required. These data would allow EPA to
assess the significance of chlorate ion resulting from the use of
hypochlorite solutions. EPA anticipates chlorate would be regulated as
part of the Stage 2 DBP Rule.

Table III. 3.--Additional Sampling Required of Systems Using
Hypochlorite Solutions
------------------------------------------------------------------------
Sampling point Analyses Frequency
------------------------------------------------------------------------
Treatment plant influent.. Chlorate.................. Quarterly.
Hypochlorite stock pH, temperature, free Quarterly.
solution. residual chlorine, and
chlorate.
Entry point to Chlorate.................. Quarterly.
distribution system.
------------------------------------------------------------------------

Additional monitoring required of systems using ozone. EPA would
require systems serving at least 100,000 people and using ozone in
their treatment process to perform the additional monitoring listed in
Table III.4. The ozone contactor influent would be monitored monthly
for parameters that influence formation of by-products: pH, alkalinity,
turbidity, temperature, calcium and total hardness, TOC, UV254,
bromide, and ammonia. The ozone residual would be measured in the
contactor effluent and immediately prior to filtration. These data
would be combined with the operational data and the DBP data to better
understand and predict DBP formation.

Table III.4.--Additional Sampling Required of Systems Using Ozone
------------------------------------------------------------------------
Sampling point Analyses Frequency
------------------------------------------------------------------------
Ozone contactor influent.. pH, alkalinity, turbidity, Monthly.
temperature, calcium and
total hardness, TOC,
UV254, bromide, and
ammonia.
Ozone contactor influent.. Aldehydes\1\ and AOC/ Quarterly.
BDOC\2\.
Ozone contactor effluent.. Ozone residual............ Monthly.
Ozone contactor effluent.. Aldehydes\1\ and AOC/ Quarterly.
BDOC\2\.
Before filtration......... Ozone residual............ Monthly.
Entry point to Bromate................... Monthly.
distribution system.
Entry point to Aldehydes\1\ and AOC/ Quarterly.
distribution system. BDOC\2\.
------------------------------------------------------------------------
\1\The aldehydes to be included in this analysis are: formaldehyde,
acetaldehyde, butanal, propanal, pentanal, glyoxal, and methyl
glyoxal. Measurement of other aldehydes is optional.
\2\Submission of data for assimilable organic carbon (AOC) or
biodegradeable organic carbon (BDOC) is optional.

Water systems using ozone would also be required to monitor for
specific DBPs that are known to be formed as the result of oxidation
reactions. The contactor influent, contactor effluent and water from
the entry point to the distribution system would be monitored on a
quarterly basis for aldehydes. Utilities would also be encouraged to
voluntarily measure assimilable organic carbon (AOC) or biodegradeable
dissolved organic carbon (BDOC) at the same sampling points and at the
same frequency and voluntarily submit the data. The concentration of
bromate would be monitored on a monthly basis at the entry point to the
distribution system. The concentration of bromate is not expected to
increase in the water after it leaves the treatment plant.
Additional monitoring required of systems using chlorine dioxide.
EPA would require systems serving 100,000 people and using chlorine
dioxide in their treatment process to conduct the additional monitoring
listed in Table III.5. Parameters that influence the formation of by-
products would be measured on a monthly basis at sampling point(s)
prior to each application of chlorine dioxide. The analyses would
include: pH, alkalinity, turbidity, temperature, calcium and total
hardness, TOC, UV254, and bromide.

Table III.5--Additional Sampling Required of Systems Using Chlorine
Dioxide
------------------------------------------------------------------------
Sampling point Analyses Frequency
------------------------------------------------------------------------
Treatment plant influent.. Chlorate.................. Quarterly.
Before each chlorine pH, alkalinity, turbidity, Monthly.
dioxide application. temperature, calcium and
total hardness, TOC,
UV254, and bromide.
Before first chlorine Aldehydes\1\ and AOC/ Quarterly.
dioxide application. BDOC\2\.
Before application of pH, chlorine dioxide Monthly.
ferrous salts, sulfur residual, chlorite,
reducing agents, or GAC. chlorate.
Before downstream chlorine/ Aldehydes\1\ and AOC/ Quarterly.
chloramine application. BDOC\2\.
Entry point to Chlorite, chlorate, Monthly.
distribution system. chlorine dioxide
residual, bromate.
Entry point to Aldehydes\1\ and AOC/ Quarterly.
distribution system. BDOC\2\.
3 distribution system chlorite, chlorate, Monthly.
sampling points (1 near chlorine dioxide
first customer, 1 in residual, pH, and
middle of distribution temperature.
system, and 1 at a
maximum detention time in
the system).
------------------------------------------------------------------------
\1\The aldehydes to be included in this analysis are: formaldehyde,
acetaldehyde, butanal, propanal, pentanal, glyoxal, and methyl
glyoxal. Measurement of other aldehydes is optional.
\2\Submission of data for AOC or BDOC is optional.

The by-products of particular concern from the use of chlorine
dioxide are chlorite and chlorate. Since the application of ferrous
salts or sulfur reducing agents changes the concentrations of these by-
products, utilities would be required to monitor for chlorite and
chlorate prior to and following each of these treatment processes.
Monitoring would also be required before and after granular activated
carbon (GAC) filtration. These data would provide a better
understanding of the formation and control of these two by-products and
would allow the development of predictive models for use in development
of the Stage 2 D/DBP Rule.
Very little data are available concerning the chlorite and chlorate
concentrations generally present in drinking water as a result of
chlorine dioxide use. Therefore, utilities would be required to monitor
for these by-products at the entry point to the distribution system and
at three sites within the distribution system. The concentrations of
chlorite and chlorate are expected to change as the water is
distributed through the system, so distribution system samples are
needed to assess the magnitude of the changes. One sample would be
collected near the first customer; another sample would be collected at
a point representing the maximum detention time in the distribution
system and the last sample would be collected at a point representative
of the average consumer.
These water systems would also be required to monitor the chlorine
dioxide residual concentrations, pH and temperature at the above
sampling points. Of particular concern is the possible re-formation of
chlorine dioxide in the distribution system as a result of reactions
between chlorite and chlorine. Since chlorine dioxide and its by-
products may pose acute health risks, monitoring for them would be
required on a monthly basis. The proposed Stage 1 D/DBP Rule may
require daily monitoring for chlorine dioxide at the point of entry
into the distribution system and monthly monitoring for chlorite at
three points in the distribution system.
Because low levels of chlorate have been reported in source water
(Bolyard, et al., 1993; and Gordon, et al., 1993), EPA would also
require systems using chlorine dioxide to monitor the treatment plant
influent monthly for chlorate. This monitoring would provide data to
assess the relative amounts of chlorate from source water versus the
amount produced as the result of chlorine dioxide use.
EPA would also require systems using chlorine dioxide to perform
quarterly monitoring for several oxidation by-products, because there
is a small amount of data indicating their presence as the result of
chlorine dioxide use. Quarterly monitoring for aldehydes would be
required: (1) Before the first chlorine dioxide application in order to
determine background levels from the source waters; (2) before
application of the secondary disinfectant to determine what was
produced by chlorine dioxide; and (3) at the entry point to the
distribution system to evaluate the total level delivered to the
consumers based upon all the treatment processes and disinfectants. EPA
would also encourage systems to voluntarily measure AOC or BDOC at the
same sampling points and at the same frequency and voluntarily submit
the data. The Agency would require systems to report the bromate
concentration present in the sample analyzed for chlorite and chlorate
from the entry point to the distribution system, because there are
limited data indicating that bromate may be formed as a result of
sunlight catalyzed reactions between chlorine dioxide and bromide ion
present in the source water (Zika et al., 1985). This would be an
additional sample, because the measurement of low levels of bromate
(g/L) in the presence of much higher levels of chlorite
(100-1000 g/L) would require special treatment of the sample.
3. Treatment Process Information Collection
Background/justification. EPA proposes collecting treatment process
information as part of this rule to characterize the various forms of
treatment currently being used by treatment plants serving more than
100,000 persons. The treatment process information will be used to
evaluate options available to large water utilities to monitor and
reduce DBP formation. The Water Treatment Plant (WTP) Model
(Harrington, et al., 1992) was used to predict THM and HAA levels in
the development of the Stage 1 D/DBP Rule. The model is available from
the Safe Drinking Water Act Hotline (1-800-426-4791). It uses raw water
quality and treatment process data to predict THM and HAA formation.
The WTP model is calibrated on fewer than 100 bench-, pilot-, and full-
scale studies. This rule would provide a sufficiently large database to
upgrade the model to include additional processes, predict other DBPs,
and better calibrate the model based on hundreds of plant experiences.
The process data would be coupled with the water quality data
described in Tables III.1 through III.5 to assess how treatment impacts
precursor removal; how treatment affects the formation of THMs, HAAs
and other DBPs; and how parameters like TOX and SDS compare to
distribution system compliance parameters. Relationships between the
process data and water quality data collected under this rule would be
evaluated to help define Stage 2 requirements of the D/DBP Rule and to
better evaluate and refine prediction models that will be used for the
Stage 2 D/DBP Rule development.
Specific Process Information. The treatment plant information and
unit processes listed in Table III.6 and the water quality data
described in previous sections will provide the information necessary
to develop predictions between raw water quality, treatment conditions,
precursor removal, and DBP formation. EPA selected the parameters
listed to characterize the unit process for use in developing the
predictions and Stage 2 D/DBP Rule development. For example,
coagulation parameters are needed for evaluation of efficiencies to
better define the impact of enhanced coagulation for precursor (TOC)
control. The depth of the filter is needed to evaluate the feasibility
of adding GAC to the filter for precursor removal. The complete process
train details are needed to evaluate the feasibility and costs of
treatment changes being considered for DBP control. The list does not
include every possible water treatment process parameter, but does
include the parameters that would be used to characterize the treatment
practices for the purpose of this monitoring rule.

Table III.6.--Treatment Plant Information

Utility information:
Utility Name
Mailing Address
Contact Person & Phone Number
Public Water Supply Identification Number FRDS (PWSID)
Population Served
Plant information:
Name of plant
Design flow (MGD)
Annual minimum water temperature (C)
Annual maximum water temperature (C)
Hours of operation (hours per day)
Source water information:
Name of source
Type of source (One of the following)
1River
2Stream
3Reservoir
4Lake
5Ground water under the direct influence of surface water
6Ground water
7Spring
8Purchased from Utility Name, FRDS PWSID
9Other
Surface water as defined by SWTR (TRUE/FALSE)
Monthly Average Flow of this Source (MGD)
Upstream sources of microbiological contamination
Wastewater plant discharge in watershed (yes/no)
Distance from intake (miles)
Monthly average flow of plant discharge (MGD)
Point source feedlots in watershed (yes/no)
Distance of nearest feedlot discharge to intake (miles)
Non-point sources in watershed
Grazing of animals (yes/no)
Nearest distance of grazing to intake (miles)
Plant influent (ICR influent sampling point):
Monthly average flow (MGD)
Monthly peak hourly flow (MGD)
Flow at time of sampling (MGD)
Plant effluent (ICR effluent sampling point):
Monthly average flow (MGD)
Monthly peak hourly flow (MGD)
Flow at time of sampling (MGD)
Sludge treatment:
Monthly average solids production (lb/day)
Installed design sludge handling capacity (lb/day)
General process parameters:
The following data will be required for all unit processes:
Number of identical parallel units installed
Number of identical parallel units in service at time of sampling
The following parameters will be required for all unit processes except
chemical feeders:
Design Flow per unit (MGD)
Liquid volume per unit (gallons)
Tracer study flow (MGD)
T50 (minutes)
T10 (minutes)
Presedimentation basin:
Surface loading at design flow (gpm/ft2)
Chemical feeder:
Type of feeder (one of the following)
1Liquid
2Gas
3Dry
Capacity of each unit (lb/day)
Purpose (one or more of the following)
1Coagulation
2Coagulation aid
3Corrosion control
4Dechlorination
5Disinfection
6Filter aid
7Fluoridation
8Oxidation
9pH adjustment
10Sequestration
11Softening
12Stabilization
13Taste and odor control
14Other
Chemical feeder chemicals (one of the following):
Alum
Anhydrous ammonia
Ammonium hydroxide
Ammonium sulfate
Calcium hydroxide
Calcium hypochlorite
Calcium oxide
Carbon dioxide
Chlorine dioxide--acid chlorite
Chlorine dioxide--chlorine/chlorite
Chlorine gas
Ferric chloride
Ferric sulfate
Ferrous sulfate
Ozone
Polyaluminum chloride
Sodium carbonate
Sodium chloride
Sodium fluoride
Sodium hydroxide
Sodium hypochlorite
Sodium hexametaphosphate
Sodium silicate
Sulfuric acid
Zinc orthophosphate
Other
Notes:
1. The above list is intended to be a comprehensive list of chemicals
used at water treatment plants. If the name of a chemical does not
appear in the list then ``Other Chemical'' information will be
requested.
2. Formulas and feed rate units will be included in data reporting
software.
Monthly average feed rate based on inventory (mg/L) Feed rate at time
of sampling (mg/L)
Other chemical:
Note:
In addition to Chemical Feeder information the following will be
required for any chemical not included in the Chemical Feeder list of
chemicals.
Trade name of chemical
Formula
Manufacturer
Rapid mix:
Type of mixer (one of the following)
1Mechanical
2Hydraulic jump
3Static
4Other
If mechanical: horsepower of motor
If hydraulic: head loss (ft)
If static: head loss (ft)
Flocculation basin:
Type of mixer (one of the following)
1Mechanical
2Hydraulic
3Other
If mechanical: Mixing power (HP)
If hydraulic: head loss (ft)
Sedimentation basin:
Loading at Design Flow (gpm/ft2)
Depth (ft)
Filtration:
Loading at Design Flow (gpm/ft2)
Media Type (one or more of the following)
1Anthracite
2GAC
3Garnet
4Sand
5Other
Depth of top media (in)
If more than 1 media: Depth of second media (in)
If more than 2 media: Depth of third media (in)
If more than 3 media: Depth of fourth media (in)
If GAC media: Carbon replacement frequency (months):
Water depth to top of media (ft)
Depth from top of media to bottom of backwash trough (ft)
Backwash Frequency (hours)
Backwash volume (gallons)
Contact basin (Stable liquid level):
Baffling Type (one of the following as defined in SWTR guidance
manual)
1Unbaffled (mixed tank)
2Poor (inlet/outlet only)
3Average (Inlet/Outlet and intermediate)
4Superior (Serpentine)
5Perfect (Plug flow)
Clearwell (Variable liquid level):
Baffling Type (one of the following as defined in SWTR guidance
manual)
1Unbaffled (mixed tank)
2Poor (inlet/outlet only)
3Average (Inlet/Outlet and intermediate)
4Superior (Serpentine)
5Perfect (Plug flow)
Minimum liquid volume (gallons)
Liquid volume at time of tracer study (gallons)
Ozone contact basin:
Basin Type
1Over/Under (Diffused O3)
2Mixed (Turbine O3)
Number of Stages
CT (min mg/L)
EPA requests comments on the design and operating parameters to be
reported for ozone contact basins.
Tube settler:
Surface loading at design flow (gpm/ft2)
Tube angle from horizontal (degrees)
Upflow clarifier:
Design horse power of turbine mixer (HP)
Surface loading at design flow (gpm/ft2)
Special Equipment (none, one, or more of the following)
1Lamella plates
2Tubes
Plate settler:
Surface loading at design flow (gpm/ft2)
DE filter:
Surface loading at design flow (gpm/ft2)
Precoat (lb/ft3)
Bodyfeed (mg/L)
Run length (hours)
Granular activated carbon:
Empty bed contact time at design flow (minutes)
Design regeneration frequency (days)
Actual regeneration frequency (days)
Membranes:
Type (one of the following)
1Reverse osmosis
2Nanofiltration
3Ultrafiltration
4Microfiltration
5Electrodialysis
6Other
Name of Other type
Membrane type (one of the following)
1Cellulose acetate and derivatives
2Polyamides
3Thin-film composite
4Other
Name of other membrane type
Molecular weight cutoff (gm/mole)
Configuration (one of the following)
1Spiral wound
2Hollow fiber
3Tube
4Plate and frame
5Other
Name of other configuration
Design flux (gpd/ft2)
Design pressure (psi)
Purpose of membrane unit (one or more of the following)
1Softening
2Desalination
3Organic removal
4Other
5Contaminant removal--name of contaminant
Percent recovery (%)
Operating pressure (psi)
Air stripping:
Packing height (ft)
Design liquid loading (gpm/ft2)
Design air to water ratio
Type of packing (Name)
Nominal size of packing (inch)
Operating air flow (SCFM)
Adsorption clarifier:
Surface loading at design flow (gpm/ft2)
Dissolved air flotation:
Surface loading at design flow (gpm/ft2)
Slow sand filtration:
Surface loading at design flow (gpd/ft2)
Ion exchange:
Purpose (one or more of the following)
1Softening
2Contaminant removal
Contaminant name
Media type (Name)
Design exchange capacity (equ/ft3)
Surface loading at design flow (gpm/ft2)
Bed depth (ft)
Regenerant Name (one of the following)
1Sodium Chloride (NaCl)
2Sulfuric Acid (H2SO4)
3Sodium Hydroxide (NaOH)
4Other
If other: Name and formula
Operating regeneration frequency (hr)
Regenerant concentration (%)
Regenerant Used (lb/day)
Other treatment:
Name
Purpose
Design Parameters

EPA will be working with the industry to develop the software to
collect this process information as described in the following section.
Utilities would use the data collection software to input the process
data once at the beginning of the monitoring period with monthly
updates of the operating data and any treatment modifications.
EPA requests comments on the completeness of Table III.6 to
describe treatment plant configurations and the specific design
parameters for the unit processes that would be relevant to Stage 2 D/
DBP rule development and future model development for predicting DBPs.
Is all the requested information essential? Are more efficient
mechanisms available than those proposed herein for obtaining the
desired information? Will the treatment plant information requested be
adequate for developing models by which to predict the ability of
utilities to achieve various potential regulatory criteria under Stage
2 (e.g., DBP and TOX occurrence levels in the distribution system)?
Will the treatment plant information required for systems serving
100,000 or more people be adequate for developing predictive models of
DBP formation for systems serving less than 100,000 people? What
additional information, if any, would be important to obtain to predict
the formation of DBPs in systems serving less than 10,000 people? If
additional information is needed, what mechanisms should be used for
obtaining it? For example, would any survey techniques of
representative systems be useful for obtaining this information?
Data collection software design. Since the collection of DBP
occurrence data and source water quality data must be combined with
information about the treatment processes, EPA proposes using data
collection software as a mechanism for obtaining the monitoring data
and treatment plant process information necessary for developing the
Stage 2 D/DBP Rule. The software would capture information about source
water quality, treatment plant design, unit processes, chemical
dosages, and the monitoring results listed in Tables III.1-III.6. EPA
would provide technical assistance for use of the data collection
software.
To capture both water quality data and process information from
each plant, the data collection software and database would be designed
to handle various treatment configurations including split flow,
process parameters relevant to each configuration, and water quality
monitoring data described in earlier sections.
EPA would provide each utility a diskette containing the data
collection software. The software would generate screen driven data
entry forms that are customized for the water utility depending on the
treatment process configuration entered by the utility. The water
quality parameters listed in Tables III.1 through III.5 and the results
of the microbiological monitoring would also be entered by the utility.
The water utility would only enter monitoring results pertinent to its
system. Table III.6 lists the unit process choices that would be used
to develop the process train for a given water treatment plant. The
computer program would be designed to prompt the user for the process
parameters based on the process choices selected. For example, a plant
using only chlorine for disinfection would not see prompts for chlorine
dioxide residual, bromate, or chlorite on its data entry screen.
The software will determine such details as where sampling points
should generally be located and which water quality parameters should
be measured. The user would have the option of printing a series of
data forms to be used as a guide in identifying sample point locations,
requesting laboratory analysis, and gathering design and operation
parameters. The software will be designed in data segments and will
save data to a monthly data file on a hard drive or diskette. The
utility will send data to EPA as described in the following section.
4. Database Development
The proposed procedure would entail each PWS collecting the data on
a computer diskette provided by EPA using the data collection software,
sending the data via modem or by diskette to a database coordinator,
having the data reviewed for correctness by an engineer or scientist
familiar with water treatment, loading the data into a master database,
having the data analyzed periodically throughout the monitoring period,
generating interim reports, and having the database in final usable
form for Stage 2 D/DBP Rule development shortly after the conclusion of
the sixth quarter monitoring period. Any interested party would have
access to the data at various points in time during the collection
period. EPA would provide technical assistance throughout the data
collection and reporting process.
EPA proposes that a personal computer with an MS-DOS operating
system be used for data entry. EPA would provide the ICR data
collection software to the utilities for data collection. The utilities
would provide the personal computer. The software will have many built
in features to guide the user through the process train configuration
and data input. In addition, EPA intends to make technical assistance
available, if needed, to help assure the quality of information
provided.
The output from the data collection software would be monthly data
files in ASCII format. Data files on diskette would be mailed to EPA
and transferred to the master data base. Data files transferred via
modem would be sent using telecommunication software supplied by the
utility. EPA requests comment on the use of diskettes, modem or other
means for data reporting.
Design of the database, its input/output mechanisms, and its output
formats would be considered before start-up of the monitoring effort.
The output would target the requirements being considered for the Stage
2 D/DBP Rule and the Enhanced SWTR. Examples of the many questions the
output would address are: (1) What is the national distribution of
bromide, TOC, etc., i.e., the factors that affect DBP formation? (2)
What is the distribution of HAAs, chloral hydrate, etc. in distribution
system waters? (3) What treatment processes and operating conditions
are associated with minimum DBP levels? (4) What levels of bromate form
in ozonation plants under different conditions?
Testing data collection and transfer. Before monitoring begins, EPA
would need to beta test the ICR data collection software for
transferring data from the utility to a master database to identify
unforeseen problems with the data collection procedure. Therefore, the
Agency's schedule for beta testing must have enough lead time to modify
the process, if needed, before monitoring begins. EPA intends to
conduct the data collection software beta testing with the cooperation
of a small number of utilities with diverse characteristics. The master
database and its data manipulation and output procedures would also be
beta tested to identify unforeseen problems with the data handling
procedures after the data are reported to EPA.
Frequency of reporting. EPA would require systems to submit data to
the Agency two months after monitoring begins and thereafter monthly.
Periodic reporting would allow EPA to review the data and resolve
problems associated with data collection and submission, and also to
quicken the pace of regulatory development of the interim and long-term
ESWTRs.
Data availability. EPA would make raw (unanalyzed) data available
to interested organizations and individuals periodically throughout the
monitoring period via electronic transfer. EPA proposes that the data
be made available after the first two quarters' raw data have been
verified, and for every 6 months of data thereafter following the
verification of that data until the conclusion of the monitoring
period. This access would be a ``read only'' mode.
EPA would make analyzed data available in summary form. The
analyzed data would be grouped by source water type, utility size, type
of treatment, distribution of DBPs, distribution of TOC, treatment
effectiveness, etc. These data would be used in developing the interim
and long-term ESWTR and the Stage 2 D/DBP rule.
5. Analytical Methods
Approved methods. Analytical methods that are currently approved
for monitoring purposes under other drinking water regulations would be
approved for use under this rule. These include the parameters: (1) pH;
(2) alkalinity; (3) turbidity; (4) temperature; (5) calcium hardness;
(6) free residual chlorine; (7) total residual chlorine; (8) chlorine
dioxide residual; (9) ozone residual; (10) chloroform; (11)
bromodichloromethane; (12) dibromochloromethane; and (13) bromoform.
Analytical methods for several of the above named parameters have
also been updated in the 18th edition of Standard Methods for the
Examination of Water and Wastewater for the Examination of Water and
Wastewater. These include: (1) pH; (2) alkalinity; (3) turbidity; (4)
temperature; (5) calcium hardness; (6) free residual chlorine; (7)
total residual chlorine; (8) chlorine dioxide residual; and (9) ozone
residual. The updated versions of these methods would also be approved
for compliance monitoring under this rule.
In addition to the methods currently approved for monitoring
purposes under other drinking water regulations and their most recent
versions, approved methods for the remainder of the parameters that
must be measured for this rule are listed in Table III.7. The methods
are published and contain descriptions of the methodology and
information on the precision and accuracy of the methods.
EPA is proposing one new method (EPA Method 551) for trihalomethane
(chloroform, bromodichloromethane, dibromochloromethane, and bromoform)
monitoring under this rule. EPA is also soliciting comment on whether
use of this method should also be approved for compliance with the
monitoring requirements under the Trihalomethane rule [44 FR 68264,
November 29, 1979].
Monitoring for the six haloacetic acids (HAAs) would be done using
EPA Method 552.1 or an expanded version of Method 6233 B which is
published in the 18th edition of Standard Methods. Bromochloroacetic
acid is not listed as an analyte in the published version of Method
6233 B, because an analytical standard was not commercially available
when the method was first developed. The feasibility of including it in
Method 6233 B has been demonstrated (Barth and Fair, 1992), and it will
be added to the method during the next revision.
Method 6233 B is undergoing revision for the 19th edition of
Standard Methods, so EPA proposes that a draft version be made
available to laboratories performing HAA analyses for this monitoring
rule.
EPA would require laboratories to use EPA Method 551 for measuring
trichloroacetonitrile, dichloroacetonitrile, bromochloroacetonitrile,
dibromoacetonitrile, 1,1-dichloropropanone, 1,1,1-trichloropropanone,
and chloropicrin. The use of pentane instead of methyl-tertiary-butyl
ether (MTBE), the solvent described in the method, would be permissible
when analyzing for these analytes and for the THMs.
Chloral hydrate (CH) would also be measured using EPA Method 551,
but its concentration would be determined by analyzing a separate
sample from the one collected for the other 551 analytes. CH requires a
different dechlorinating agent than the other DBPs included in the
method. The THMs can also be measured in the chloral hydrate sample.
MTBE must be used as the extracting solvent when measuring CH.
EPA Method 551 specifies that the pH of the sample be adjusted to
between 4.5 and 5.0 when the sample is collected, to prevent base-
catalyzed hydrolysis of several of the analytes. Sample stability has
been demonstrated for 14 days when this technique is used in the
laboratory. However, field application of this preservation technique
(i.e., titration) has not been tested and may not be practical. EPA
proposes that the samples be collected without adjusting the pH and
that the laboratories be required to extract the samples within 24-48
hours of sample collection. This requirement would result in a negative
bias in the data for several of the analytes, with the bias increasing
as the pH of the samples increases. EPA solicits comments on this
approach or suggestions on alternative approaches.
Chlorate, chlorite, bromide, and bromate would be measured using
EPA Method 300.0. Laboratories would be permitted to use alternate
eluents (e.g., borate eluent) or sample cleanup or concentration
techniques in order to lower the detection limit for bromate, as long
as the quality assurance criteria specified in the method are met.
EPA is aware that the above method may not be sensitive enough to
provide quantitative data for bromate at concentrations g/
L. Some laboratories may be able to detect bromate in samples at
concentrations as low as 5 g/L, but the data will not be
precise enough to be used for making decisions on how treatment
practices and source water characteristics influence bromate formation.
Since the Stage 1 D/DBP Rule may propose a maximum contaminant level
goal (M

[Text truncated at 120,000 characters. The full text is on the page linked above.]

---

Source: Frix Law Library, https://www.frixlaw.com/law-library/documents/fr%3A94-2587. Public record. Not legal advice.