Sie sind auf Seite 1von 107


January 4, 2006

Part II

Protection Agency
40 CFR Parts 9, 141, and 142
National Primary Drinking Water
Regulations: Stage 2 Disinfectants and
Disinfection Byproducts Rule; Final Rule
wwhite on PROD1PC65 with RULES2

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00001 Fmt 4717 Sfmt 4717 E:\FR\FM\04JAR2.SGM 04JAR2
388 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations

ENVIRONMENTAL PROTECTION reducing peak and average levels of Publicly available docket materials
AGENCY DBPs in drinking water supplies. are available either electronically
The Stage 2 DBPR applies to public through or
40 CFR Parts 9, 141, and 142 water systems (PWSs) that are in hard copy at the Water Docket, EPA/
community water systems (CWSs) or DC, EPA West, Room B102, 1301
[EPA–HQ–OW–2002–0043; FRL–8012–1]
nontransient noncommunity water Constitution Ave., NW., Washington,
RIN 2040–AD38 systems (NTNCWs) that add a primary DC. The Public Reading Room is open
or residual disinfectant other than from 10 a.m. to 4 p.m., Monday through
National Primary Drinking Water ultraviolet light or deliver water that has Friday, excluding legal holidays. The
Regulations: Stage 2 Disinfectants and been treated with a primary or residual telephone number for the Public
Disinfection Byproducts Rule disinfectant other than ultraviolet light. Reading Room is (202) 566–1744, and
AGENCY: Environmental Protection This rule also makes minor the telephone number for the Water
Agency (EPA). corrections to drinking water Docket is (202) 566–2426.
regulations, specifically the Public
Notification tables. New endnotes were
SUMMARY: The Environmental Protection added to these tables in recent technical inquiries, contact Tom
Agency (EPA) is promulgating today’s rulemakings; however, the Grubbs, Standards and Risk
final rule, the Stage 2 Disinfectants and corresponding footnote numbering in Management Division, Office of Ground
Disinfection Byproducts Rule (DBPR), to the tables was not changed. In addition, Water and Drinking Water (MC 4607M),
provide for increased protection against this rule makes a minor correction to the Environmental Protection Agency, 1200
the potential risks for cancer and Stage 1 Disinfectants and Disinfection Pennsylvania Ave., NW., Washington,
reproductive and developmental health Byproducts Rule by replacing a sentence DC 20460; telephone number: (202)
effects associated with disinfection that was inadvertently removed. 564–5262; fax number: (202) 564–3767;
byproducts (DBPs). The final Stage 2 DATES: This final rule is effective on e-mail address:
DBPR contains maximum contaminant March 6, 2006. For judicial review For general information, contact the
level goals for chloroform, purposes, this final rule is promulgated Safe Drinking Water Hotline, Telephone
monochloroacetic acid and as January 4, 2006. The incorporation by (800) 426–4791. The Safe Drinking
trichloroacetic acid; National Primary reference of certain publications listed Water Hotline is open Monday through
Drinking Water Regulations, which in the rule is approved by the Director Friday, excluding legal holidays, from
consist of maximum contaminant levels of the Federal Register as of March 6, 10 a.m. to 4 p.m. Eastern Time.
(MCLs) and monitoring, reporting, and 2006. SUPPLEMENTARY INFORMATION:
public notification requirements for ADDRESSES: EPA has established a
total trihalomethanes (TTHM) and docket for this action under Docket ID I. General Information
haloacetic acids (HAA5); and revisions No. EPA–HQ–OW–2002–0043. All A. Does This Action Apply to Me?
to the reduced monitoring requirements documents in the docket are listed on
for bromate. This document also the Web Entities potentially regulated by the
specifies the best available technologies site. Stage 2 DBPR are community and
for the final MCLs. EPA is also Although listed in the index, some nontransient noncommunity water
approving additional analytical methods information is not publicly available, systems that add a primary or residual
for the determination of disinfectants e.g., CBI or other information whose disinfectant other than ultraviolet light
and DBPs in drinking water. EPA disclosure is restricted by statute. or deliver water that has been treated
believes the Stage 2 DBPR will reduce Certain other material, such as with a primary or residual disinfectant
the potential risks of cancer and copyrighted material, is not placed on other than ultraviolet light. Regulated
reproductive and developmental health the Internet and will be publicly categories and entities are identified in
effects associated with DBPs by available only in hard copy form. the following chart.

Category Examples of regulated entities

Industry ............................................................... Community and nontransient noncommunity water systems that use a primary or residual dis-
infectant other than ultraviolet light or deliver water that has been treated with a primary or
residual disinfectant other than ultraviolet light.
State, Local, Tribal, or Federal Governments .... Community and nontransient noncommunity water systems that use a primary or residual dis-
infectant other than ultraviolet light or deliver water that has been treated with a primary or
residual disinfectant other than ultraviolet light.

This table is not intended to be the section entitled ‘‘coverage’’ (§ 141.3) B. How Can I Get Copies of This
exhaustive, but rather provides a guide in Title 40 of the Code of Federal Document and Other Related
for readers regarding entities likely to be Regulations and applicability criteria in Information?
regulated by this action. This table lists § 141.600 and 141.620 of today’s See the ADDRESSES section for
the types of entities that EPA is now proposal. If you have questions information on how to receive a copy of
aware could potentially be regulated by regarding the applicability of this action this document and related information.
this action. Other types of entities not to a particular entity, contact the person
Regional contacts:
listed in the table could also be listed in the preceding FOR FURTHER
wwhite on PROD1PC65 with RULES2

I. Kevin Reilly, Water Supply Section,

regulated. To determine whether your INFORMATION CONTACT section.
JFK Federal Bldg., Room 203,
facility is regulated by this action, you Boston, MA 02203, (617) 565–3616.
should carefully examine the definition
II. Michael Lowy, Water Supply Section,
of ‘‘public water system’’ in § 141.2 and 290 Broadway, 24th Floor, New

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00002 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 389

York, NY 10007–1866, (212) 637– acid, trichloroacetic acid, SBAR Small Business Advisory
3830. monobromoacetic acid, and Review
III. Jason Gambatese, Drinking Water dibromoacetic acid) SBREFA Small Business Regulatory
Section (3WM41), 1650 Arch Street, HAN Haloacetonitriles Enforcement Fairness Act
Philadelphia, PA 19103–2029, (215) (trichloroacetonitrile, SDWA Safe Drinking Water Act, or the
814–5759. dichloroacetonitrile, ‘‘Act,’’ as amended in 1996
IV. Robert Burns, Drinking Water bromochloroacetonitrile, and SER Small Entity Representative
Section, 61 Forsyth Street SW., dibromoacetonitrile) SGA Small for gestational age
Atlanta, GA 30303, (404) 562–9456. IC Ion chromatograph SUVA Specific ultraviolet absorbance
V. Miguel Del Toral, Water Supply IC/ICP–MS Ion chromatograph SWAT Surface Water Analytical Tool
Section, 77 W. Jackson Blvd., coupled to an inductively coupled SWTR Surface Water Treatment Rule
Chicago, IL 60604, (312) 886–5253. plasma mass spectrometer TC Total coliforms
VI. Blake L. Atkins, Drinking Water IDSE Initial distribution system TCAA Trichloroacetic acid
Section, 1445 Ross Avenue, Dallas, evaluation TCR Total Coliform Rule
TX 75202, (214) 665–2297. ILSI International Life Sciences THM Trihalomethane
VII. Douglas J. Brune, Drinking Water Institute TOC Total organic carbon
Management Branch, 901 North 5th IESWTR Interim Enhanced Surface TTHM Total trihalomethanes (sum of
Street, Kansas City, KS 66101, (800) Water Treatment Rule four THMs: chloroform,
233–0425. IPCS International Programme on bromodichloromethane,
VIII. Bob Clement, Public Water Supply Chemical Safety dibromochloromethane, and
Section (8P2-W-MS), 999 18th IRIS Integrated Risk Information bromoform)
Street, Suite 500, Denver, CO System (EPA) TWG Technical work group
80202–2466, (303) 312–6653. LOAEL Lowest observed adverse effect UMRA Unfunded Mandates Reform
IX. Bruce Macler, Water Supply Section, level Act
75 Hawthorne Street, San LRAA Locational running annual UV 254 Ultraviolet absorption at 254
Francisco, CA 94105, (415) 972– average nm
3569. LT1ESTWR Long Term 1 Enhanced VSL Value of Statistical Life
X. Wendy Marshall, Drinking Water Surface Water Treatment Rule WTP Willingness To Pay
Unit, 1200 Sixth Avenue (OW–136), LT2ESTWR Long Term 2 Enhanced Table of Contents
Seattle, WA 98101, (206) 553–1890. Surface Water Treatment Rule
MBAA Monobromoacetic acid I. General Information
Abbreviations Used in This Document MCAA Monochloroacetic acid A. Does This Action Apply to Me?
B. How Can I Get Copies of This Document
ASDWA Association of State Drinking MCL Maximum contaminant level and Other Related Information?
Water Administrators MCLG Maximum contaminant level II. Summary of the Final Rule
ASTM American Society for Testing goal A. Why is EPA Promulgating the Stage 2
and Materials M–DBP Microbial and disinfection DBPR?
AWWA American Water Works byproducts mg/L Milligram per liter B. What Does the Stage 2 DBPR Require?
Association MRL Minimum reporting level 1. Initial Distribution System Evaluation
AwwaRF American Water Works MRDL Maximum residual disinfectant 2. Compliance and monitoring
Association Research Foundation level requirements
BAT Best available technology MRDLG Maximum residual 3. Operational Evaluation Levels
BCAA Bromochloroacetic acid 4. Consecutive systems
disinfectant level goal
C. Correction of § 141.132
BDCM Bromodichloromethane NDMA N-nitrosodimethylamine III. Background
CDBG Community Development Block NDWAC National Drinking Water A. Statutory Requirements and Legal
Grant Advisory Council Authority
CWS Community water system NF Nanofiltration B. What is the Regulatory History of the
DBAA Dibromoacetic acid NOAEL No Observed Adverse Effect Stage 2 DBPR and How Were
DBCM Dibromochloromethane Level Stakeholders Involved?
DBP Disinfection byproduct NODA Notice of data availability 1. Total Trihalomethanes Rule
DBPR Disinfectants and Disinfection NPDWR National primary drinking 2. Stage 1 Disinfectants and Disinfection
Byproducts Rule water regulation Byproducts Rule
DCAA Dichloroacetic acid NRWA National Rural Water 3. Stakeholder involvement
a. Federal Advisory Committee process
EA Economic analysis Association b. Other outreach processes
EC Enhanced coagulation NTNCWS Nontransient C. Public Health Concerns to be Addressed
EDA Ethylenediamine noncommunity water system 1. What are DBPs?
EPA United States Environmental NTP National Toxicology Program 2. DBP Health Effects
Protection Agency NTTAA National Technology Transfer a. Cancer health effects
ESWTR Enhanced Surface Water and Advancement Act i. Epidemiology
Treatment Rule OMB Office of Management and ii. Toxicology
FACA Federal Advisory Committee Budget b. Reproductive and developmental health
Act PAR Population attributable risk effects
GAC Granular activated carbon PE Performance evaluation i. Epidemiology
GC/ECD Gas chromatography using ii. Toxicology
PWS Public water system
c. Conclusions
electron capture detection RAA Running annual average D. DBP Occurrence and DBP Control
wwhite on PROD1PC65 with RULES2

GWR Ground Water Rule RFA Regulatory Flexibility Act 1. Occurrence

GWUDI Ground water under the direct RfD Reference dose 2. Treatment
influence of surface water RSC Relative source contribution E. Conclusions for Regulatory Action
HAA5 Haloacetic acids (five) (sum of RUS Rural Utility Service IV. Explanation of Today’s Action
monochloroacetic acid, dichloroacetic SAB Science Advisory Board A. MCLGs

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00003 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
390 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations

1. Chloroform MCLG M. System Reporting and Recordkeeping B. Paperwork Reduction Act

a. Today’s rule Requirements C. Regulatory Flexibility Act
b. Background and analysis 1. Today’s rule D. Unfunded Mandates Reform Act
c. Summary of major comments 2. Summary of major comments E. Executive Order 13132: Federalism
2. HAA MCLGs: TCAA and MCAA N. Approval of Additional Analytical F. Executive Order 13175: Consultation
a. Today’s rule Methods and Coordination With Indian Tribal
b. Background and analysis 1. Today’s Rule Governments
c. Summary of major comments 2. Background and Analysis G. Executive Order 13045: Protection of
B. Consecutive Systems O. Laboratory Certification and Approval Children from Environmental Health
1. Today’s Rule 1. PE acceptance criteria Risks and Safety Risks
2. Background and analysis a. Today’s rule H. Executive Order 13211: Actions
3. Summary of major comments b. Background and analysis Concerning Regulations That
C. LRAA MCLs for TTHM and HAA5 c. Summary of major comments Significantly Affect Energy Supply,
1. Today’s rule 2. Minimum reporting limits Distribution, or Use
2. Background and analysis a. Today’s rule I. National Technology Transfer and
3. Summary of major comments b. Background and analysis Advancement Act
D. BAT for TTHM and HAA5 c. Summary of major comments J. Executive Order 12898: Federal Actions
1. Today’s rule P. Other regulatory changes to Address Environmental Justice in
2. Background and analysis V. State Implementation Minority Populations or Low-Income
3. Summary of major comments A. Today’s rule Populations
E. Compliance Schedules 1. State Primacy Requirements for K. Consultations with the Science
1. Today’s rule Implementation Flexibility
Advisory Board, National Drinking
2. Background and analysis 2. State recordkeeping requirements
Water Advisory Council, and the
3. Summary of major comments 3. State reporting requirements
Secretary of Health and Human Services
F. Initial Distribution System Evaluation 4. Interim primacy
L. Plain Language
(IDSE) 5. IDSE implementation
B. Background and Analysis M. Analysis of the Likely Effect of
1. Today’s rule Compliance With the Stage 2 DBPR on
a. Applicability C. Summary of Major Comments
VI. Economic Analysis the Technical, Managerial, and Financial
b. Data collection Capacity of Public Water Systems
i. Standard monitoring A. Regulatory Alternatives Considered
B. Analyses that Support Today’s Final N. Congressional Review Act
ii. System specific study VIII. References
iii. 40/30 certification
1. Predicting water quality and treatment II. Summary of the Final Rule
c. Implementation
2. Background and analysis
2. Estimating benefits A. Why is EPA Promulgating the Stage
a. Standard monitoring
3. Estimating costs 2 DBPR?
b. Very small system waivers
4. Comparing regulatory alternatives
c. 40/30 certifications C. Benefits of the Stage 2 DBPR The Environmental Protection Agency
d. System specific studies 1. Nonquantified benefits is finalizing the Stage 2 Disinfectants
e. Distribution System Schematics 2. Quantified benefits and Disinfection Byproduct Rule
3. Summary of major comments 3. Timing of benefits accrual (DBPR) to reduce potential cancer risks
G. Monitoring Requirements and D. Costs of the Stage 2 DBPR
Compliance Determination for TTHM and address concerns with potential
1. Total annualized present value costs reproductive and developmental risks
and HAA5 MCLs 2. PWS costs
1. Today’s Rule from DBPs. The Agency is committed to
a. IDSE costs
a. IDSE Monitoring b. PWS treatment costs ensuring that all public water systems
b. Routine Stage 2 Compliance Monitoring c. Monitoring costs provide clean and safe drinking water.
i. Reduced monitoring 3. State/Primacy agency costs Disinfectants are an essential element of
ii. Compliance determination 4. Non-quantified costs drinking water treatment because of the
2. Background and Analysis E. Household Costs of the Stage 2 DBPR barrier they provide against harmful
3. Summary of Major Comments F. Incremental Costs and Benefits of the waterborne microbial pathogens.
H. Operational Evaluation Requirements Stage 2 DBPR
initiated by TTHM and HAA5 Levels However, disinfectants react with
G. Benefits From the Reduction of Co-
1. Today’s rule naturally occurring organic and
occurring Contaminants
2. Background and analysis H. Potential Risks From Other inorganic matter in source water and
3. Summary of major comments Contaminants distribution systems to form
I. MCL, BAT, and Monitoring for Bromate 1. Emerging DBPs disinfection byproducts (DBPs) that may
1. Today’s rule 2. N-nitrosamines pose health risks. The Stage 2 DBPR is
2. Background and analysis 3. Other DBPs designed to reduce the level of exposure
a. Bromate MCL I. Effects of the Contaminant on the from DBPs without undermining the
b. Criterion for reduced bromate General Population and Groups within control of microbial pathogens. The
monitoring the General Population that are
3. Summary of major comments Long Term 2 Enhanced Surface Water
Identified as Likely To Be at Greater Risk
J. Public Notice Requirements of Adverse Health Effects Treatment Rule (LT2ESWTR) is being
1. Today’s rule J. Uncertainties in the Risk, Benefit, and finalized and implemented
2. Background and analysis Cost Estimates for the Stage 2 DBPR simultaneously with the Stage 2 DBPR
3. Summary of major comments K. Benefit/Cost Determination for the Stage to ensure that drinking water is
K. Variances and Exemptions 2 DBPR microbiologically safe at the limits set
1. Today’s Rule L. Summary of Major Comments for DBPs.
2. Background and Analysis 1. Interpretation of health effects studies Congress required EPA to promulgate
a. Variances 2. Derivation of benefits the Stage 2 DBPR as part of the 1996
b. Affordable Treatment Technologies for 3. Use of SWAT
wwhite on PROD1PC65 with RULES2

Small Systems 5. Unanticipated risk issues

Safe Drinking Water Act (SDWA)
c. Exemptions 6. Valuation of cancer cases avoided Amendments (section 1412(b)(2)(C)).
3. Summary of major comments VII. Statutory and Executive Order Reviews The Stage 2 DBPR augments the Stage
L. Requirements for Systems to Use A. Executive Order 12866: Regulatory 1 DBPR that was finalized in 1998 (63
Qualified Operators Planning and Review FR 69390, December 16, 1998) (USEPA

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00004 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 391

1998a). The goal of the Stage 2 DBPR is reductions of DBPs in distribution The IDSE is designed to offer
to target the highest risk systems for systems. flexibility to public water systems. The
changes beyond those required for Stage The Stage 2 DBPR presents a risk- IDSE requires TTHM and HAA5
1 DBPR. Today’s rule reflects consensus targeting approach to reduce risks from monitoring for one year on a regular
recommendations from the Stage 2 DBPs. The new requirements provide schedule that is determined by source
Microbial/Disinfection Byproducts (M– for more consistent, equitable protection water type and system size.
DBP) Federal Advisory Committee (the from DBPs across the entire distribution Alternatively, systems have the option
Advisory Committee) as well as public system and the reduction of DBP peaks. of performing a site-specific study based
comments. New risk-targeting provisions require on historical data, water distribution
New information on health effects, systems to first identify their risk level; system models, or other data; and
occurrence, and treatment has become then, only those systems with the waivers are available under certain
available since the Stage 1 DBPR that greatest risk will need to make circumstances. The IDSE requirements
supports the need for the Stage 2 DBPR. operational or treatment changes. The are discussed in Sections IV.E, IV.F.,
EPA has completed a more extensive Stage 2 DBPR, in conjunction with the and IV.G of this preamble and in
analysis of health effects, particularly LT2ESWTR, will help public water subpart U of the rule language.
reproductive and developmental systems deliver safer water to
endpoints, associated with DBPs since Americans with the benefits of 2. Compliance and Monitoring
the Stage 1 DBPR. Some recent studies disinfection to control pathogens and Requirements
on both human epidemiology and with fewer risks from DBPs. As in Stage 1, the Stage 2 DBPR
animal toxicology have shown possible focuses on monitoring for and reducing
B. What Does the Stage 2 DBPR Require? concentrations of two classes of DBPs:
associations between chlorinated
drinking water and reproductive and The risk-targeting components of the total trihalomethanes (TTHM) and
developmental endpoints such as Stage 2 DBPR focus the greatest amount haloacetic acids (HAA5). These two
spontaneous abortion, stillbirth, neural of change where the greatest amount of groups of DBPs act as indicators for the
tube and other birth defects, intrauterine risk may exist. Therefore, the provisions various byproducts that are present in
growth retardation, and low birth of the Stage 2 DBPR focus first on water disinfected with chlorine or
weight. While results of these studies identifying the higher risks through the chloramine. This means that
have been mixed, EPA believes they Initial Distribution System Evaluation concentrations of TTHM and HAA5 are
support a potential hazard concern. (IDSE). The rule then addresses monitored for compliance, but their
New epidemiology and toxicology reducing exposure and lowering DBP presence in drinking water is
studies evaluating bladder, colon, and peaks in distribution systems by using representative of many other
rectal cancers have increased the weight a new method to determine MCL chlorination DBPs that may also occur
of evidence linking these health effects compliance (locational running annual in the water; thus, a reduction in TTHM
to DBP exposure. The large number of average (LRAA)), defining operational and HAA5 generally indicates an overall
people (more than 260 million evaluation levels, and regulating reduction of DBPs.
Americans) exposed to DBPs and the consecutive systems. This section The second provision of the Stage 2
potential cancer, reproductive, and briefly describes the requirements of DBPR is designed to address spatial
developmental risks have played a this final rule. More detailed variations in DBP exposure through a
significant role in EPA’s decision to information on the regulatory new compliance calculation (referred to
move forward with regulatory changes requirements for this rule can be found as locational running annual average)
that target lowering DBP exposures in Section IV. for TTHM and HAA5 MCLs. The MCL
beyond the requirements of the Stage 1 values remain the same as in the Stage
1. Initial Distribution System Evaluation
DBPR. 1. The Stage 1 DBPR running annual
While the Stage 1 DBPR is predicted The first provision, designed to average (RAA) calculation allowed some
to provide a major reduction in DBP identify higher risk systems, is the locations within a distribution system to
exposure, national survey data suggest Initial Distribution System Evaluation have higher DBP annual averages than
that some customers may receive (IDSE). The purpose of the IDSE is to others as long as the system-wide
drinking water with elevated, or peak, identify Stage 2 DBPR compliance average was below the MCL. The Stage
DBP concentrations even when their monitoring sites that represent each 2 DBPR bases compliance on a
distribution system is in compliance system’s highest levels of DBPs. Because locational running annual average
with the Stage 1 DBPR. Some of these Stage 2 DBPR compliance will be (LRAA) calculation, where the annual
peak concentrations are substantially determined at these new monitoring average at each sampling location in the
greater than the Stage 1 DBPR maximum sites, only those systems that identify distribution system will be used to
contaminant levels (MCLs) and some elevated concentrations of TTHM and determine compliance with the MCLs of
customers receive these elevated levels HAA5 will need to make treatment or 0.080 mg/L and 0.060 mg/L for TTHM
of DBPs on a consistent basis. The new process changes to bring the system into and HAA5, respectively. The LRAA will
survey results also show that Stage 1 compliance with the Stage 2 DBPR. By reduce exposures to high DBP
DBPR monitoring sites may not be identifying compliance monitoring sites concentrations by ensuring that each
representative of higher DBP with the highest concentrations of monitoring site is in compliance with
concentrations that occur in distribution TTHM and HAA5 in each system’s the MCLs as an annual average, while
systems. In addition, new studies distribution system, the IDSE will offer providing all customers drinking water
indicate that cost-effective technologies increased assurance that MCLs are being that more consistently meets the MCLs.
including ultraviolet light (UV) and met across the distribution system and A more detailed discussion of Stage 2
granular activated carbon (GAC) may be that customers are receiving more DBPR MCL requirements can be found
wwhite on PROD1PC65 with RULES2

very effective at lowering DBP levels. equitable public health protection. Both in Sections IV.C, IV.E, and IV.G of this
EPA’s analysis of this new occurrence treatment changes and awareness of preamble and in § 141.64(b)(2) and (3)
and treatment information indicates that TTHM and HAA5 levels resulting from and subpart V of the rule language.
significant public health benefits may be the IDSE will allow systems to better The number of compliance
achieved through further, cost-effective control for distribution system peaks. monitoring sites is based on the

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00005 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
392 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations

population served and the source water the Stage 1 Disinfection Byproducts on the health of persons,’’ is ‘‘known to
type. EPA believes that population- Rule. This rule corrects a technical error occur or there is a substantial likelihood
based monitoring provides better risk- made in the January 16, 2001, Federal that the contaminant will occur in
targeting and is easier to implement. Register Notice (66 FR 3769) (see page public water systems with a frequency
Section IV.G describes population-based 3770). This rule restores the following and at levels of public health concern,’’
monitoring and how it affects systems sentence that was inadvertently and for which ‘‘in the sole judgement of
complying with this rule. removed from § 141.132 (b)(1)(iii), the Administrator, regulation of such
The Stage 2 DBPR includes new ‘‘Systems on a reduced monitoring contaminant presents a meaningful
MCLGs for chloroform, schedule may remain on that reduced opportunity for health risk reduction for
monochloroacetic acid, and schedule as long as the average of all persons served by public water
trichloroacetic acid, but these new samples taken in the year (for systems systems’’ (SDWA section 1412(b)(1)(A)).
MCLGs do not affect the MCLs for which must monitor quarterly) or the MCLGs are non-enforceable health goals
TTHM or HAA5. result of the sample (for systems which set at a level at which ‘‘no known or
3. Operational Evaluation Levels must monitor no more frequently than anticipated adverse effects on the health
annually) is no more than 0.060 mg/L of persons occur and which allows an
The IDSE and LRAA calculation will and 0.045 mg/L for TTHMs and HAA5, adequate margin of safety.’’ These
lead to lower DBP concentrations respectively.’’ This text had been part of health goals are published at the same
overall and reduce short term exposures the original regulation when it was time as the NPDWR (SDWA sections
to high DBP concentrations in certain codified in the CFR on December 16, 1412(b)(4) and 1412(a)(3)).
areas, but this strengthened approach to 1998. However, as a result of a SDWA also requires each NPDWR for
regulating DBPs will still allow subsequent amendment to that which an MCLG is established to
individual DBP samples above the MCL regulatory text, the text discussed today specify an MCL that is as close to the
even when systems are in compliance was removed. EPA recognized the error MCLG as is feasible (sections 1412(b)(4)
with the Stage 2 DBPR. Today’s rule only after publication of the new and 1401(1)(C)). The Agency may also
requires systems that exceed operational amendment, and is now correcting the consider additional health risks from
evaluation levels (referred to as error. EPA is merely restoring to the other contaminants and establish an
significant excursions in the proposed CFR language that EPA had MCL ‘‘at a level other than the feasible
rule) to evaluate system operational promulgated on December 16, 1998. level, if the technology, treatment
practices and identify opportunities to EPA is not creating any new rights or techniques, and other means used to
reduce DBP concentrations in the obligations by this technical correction. determine the feasible level would
distribution system. This provision will Thus, additional notice and public result in an increase in the health risk
curtail peaks by providing systems with comment is not necessary. EPA finds from drinking water by—(i) increasing
a proactive approach to remain in that this constitutes ‘‘good cause’’ under the concentration of other contaminants
compliance. Operational evaluation 5 U.S.C. 553(b)(B). in drinking water; or (ii) interfering with
requirements are discussed in greater the efficacy of drinking water treatment
III. Background techniques or processes that are used to
detail in Section IV.H.
A combination of factors influenced comply with other national primary
4. Consecutive Systems the development of the Stage 2 DBPR. drinking water regulations’’ (section
The Stage 2 DBPR also contains These include the initial 1992–1994 1412(b)(5)(A)). When establishing an
provisions for regulating consecutive Microbial and Disinfection Byproduct MCL or treatment technique under this
systems, defined in the Stage 2 DBPR as (M–DBP) stakeholder deliberations and authority, ‘‘the level or levels or
public water systems that buy or EPA’s Stage 1 DBPR proposal (USEPA treatment techniques shall minimize the
otherwise receive some or all of their 1994); the 1996 Safe Drinking Water Act overall risk of adverse health effects by
finished water from another public (SDWA) Amendments; the 1996 balancing the risk from the contaminant
water system. Uniform regulation of Information Collection Rule; the 1998 and the risk from other contaminants
consecutive systems provided by the Stage 1 DBPR; new data, research, and the concentrations of which may be
Stage 2 DBPR will ensure that analysis on disinfection byproduct affected by the use of a treatment
consecutive systems deliver drinking (DBP) occurrence, treatment, and health technique or process that would be
water that meets applicable DBP effects since the Stage 1 DBPR; and the employed to attain the maximum
standards, thereby providing better, Stage 2 DBPR Microbial and contaminant level or levels’’ (section
more equitable public health protection. Disinfection Byproducts Federal 1412(b)(5)(B)). In today’s rule, the
More information on regulation of Advisory Committee. The following Agency is establishing MCLGs and
consecutive systems can be found in sections provide summary background MCLs for certain DBPs, as described in
Sections IV.B, IV.E, and IV.G. information on these subjects. For Section IV.
additional information, see the Finally, section 1412(b)(2)(C) of the
C. Correction of § 141.132
proposed Stage 2 DBPR and supporting Act requires EPA to promulgate a Stage
Section 553 of the Administrative technical material where cited (68 FR 2 DBPR. Consistent with statutory
Procedure Act, 5 U.S.C. 553(b)(B), 49548, August 18, 2003) (USEPA provisions for risk balancing (section
provides that, when an agency for good 2003a). 1412(b)(5)(B)), EPA is finalizing the
cause finds that notice and public LT2ESWTR concurrently with the Stage
procedure are impracticable, A. Statutory Requirements and Legal
2 DBPR to ensure simultaneous
unnecessary, or contrary to the public Authority
protection from microbial and DBP
interest, the agency may issue a rule The SDWA, as amended in 1996, risks.
without providing prior notice and an authorizes EPA to promulgate a national
B. What is the Regulatory History of the
wwhite on PROD1PC65 with RULES2

opportunity for public comment. In primary drinking water regulation

addition to promulgating the Stage 2 (NPDWR) and publish a maximum Stage 2 DBPR and How Were
regulations, this rule also makes a minor contaminant level goal (MCLG) for any Stakeholders Involved?
correction to the National Primary contaminant the Administrator This section first summarizes the
Drinking Water Regulations, specifically determines ‘‘may have an adverse effect existing regulations aimed at controlling

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00006 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 393

levels of DBPs in drinking water. The EPA finalized the Interim Enhanced calculation to a locational running
Stage 2 DBPR establishes regulatory Surface Water Treatment Rule (63 FR annual average (LRAA) calculation. In
requirements beyond these rules that 69477, December 16, 1998) (USEPA the first phase, systems would continue
target high risk systems and provide for 1998b) at the same time as the Stage 1 to comply with the Stage 1 DBPR MCLs
more equitable protection from DBPs DBPR to ensure simultaneous as RAAs and, at the same time, comply
across the entire distribution system. compliance and address risk tradeoff with MCLs of 0.120 mg/L for TTHM and
Next, this section summarizes the issues. Both rules were products of 0.100 mg/L for HAA5 calculated as
extensive stakeholder involvement in extensive Federal Advisory Committee LRAAs. RAA calculations average all
the development of the Stage 2 DBPR. deliberations and final consensus samples collected within a distribution
recommendations in 1997. system over a one-year period, but
1. Total Trihalomethanes Rule
3. Stakeholder Involvement LRAA calculations average all samples
The first rule to regulate DBPs was
taken at each individual sampling
promulgated on November 29, 1979. a. Federal Advisory Committee
location in a distribution system during
The Total Trihalomethanes Rule (44 FR process. EPA reconvened the M-DBP
68624, November 29, 1979) (USEPA Advisory Committee in March 1999 to a one-year period. Systems would also
1979) set an MCL of 0.10 mg/L for total develop recommendations on issues carry out an Initial Distribution System
trihalomethanes (TTHM). Compliance pertaining to the Stage 2 DBPR and Evaluation (IDSE) to select compliance
was based on the running annual LT2ESWTR. The Stage 2 M-DBP monitoring sites that reflect higher
average (RAA) of quarterly averages of Advisory Committee consisted of 21 TTHM and HAA5 levels occurring in
all samples collected throughout the organizational members representing the distribution system. The second
distribution system. This TTHM EPA, State and local public health and phase of compliance would require
standard applied only to community regulatory agencies, local elected MCLs of 0.080 mg/L for TTHM and
water systems using surface water and/ officials, Native American Tribes, large 0.060 mg/L for HAA5, calculated as
or ground water that served at least and small drinking water suppliers, LRAAs at individual monitoring sites
10,000 people and added a disinfectant chemical and equipment manufacturers, identified through the IDSE. The first
to the drinking water during any part of environmental groups, and other phase has been dropped in the final
the treatment process. stakeholders. Technical support for the rule, as discussed in section IV.C.
Advisory Committee’s discussions was The Agreement in Principle also
2. Stage 1 Disinfectants and Disinfection
provided by a technical working group provided recommendations for
Byproducts Rule
established by the Advisory Committee. simultaneous compliance with the
The Stage 1 DBPR, finalized in 1998 The Advisory Committee held ten LT2ESWTR so that the reduction of
(USEPA 1998a), applies to all meetings from September 1999 to July
community and nontransient DBPs does not compromise microbial
2000, which were open to the public, protection. The complete text of the
noncommunity water systems that add with an opportunity for public comment
a chemical disinfectant to water. The Agreement in Principle (USEPA 2000a)
at each meeting.
rule established maximum residual can be found online at
The Advisory Committee carefully
disinfectant level goals (MRDLGs) and considered extensive new data on the
enforceable maximum residual occurrence and health effects of DBPs, b. Other outreach processes. EPA
disinfectant level (MRDL) standards for as well as costs and potential impacts worked with stakeholders to develop
three chemical disinfectants—chlorine, on public water systems. In addition, the Stage 2 DBPR through various
chloramine, and chlorine dioxide; they considered risk tradeoffs associated outreach activities other than the M-
maximum contaminant level goals with treatment changes. Based upon this DBP Federal Advisory Committee
(MCLGs) for three trihalomethanes detailed technical evaluation, the process. The Agency consulted with
(THMs), two haloacetic acids (HAAs), committee concluded that a targeted State, local, and Tribal governments; the
bromate, and chlorite; and enforceable protective public health approach National Drinking Water Advisory
maximum contaminant level (MCL) should be taken to address exposure to Committee (NDWAC); the Science
standards for TTHM, five haloacetic DBPs beyond the requirements of the Advisory Board (SAB); and Small Entity
acids (HAA5), bromate (calculated as Stage 1 DBPR. While there had been Representatives (SERs) and small
running annual averages (RAAs)), and substantial research to date, the system operators (as part of an Agency
chlorite (based on daily and monthly Advisory Committee also concluded outreach initiative under the Regulatory
sampling). The Stage 1 DBPR uses that significant uncertainty remained Flexibility Act). Section VII includes a
TTHM and HAA5 as indicators of the regarding the risk associated with DBPs complete description of the many
various DBPs that are present in in drinking water. After reaching these stakeholder activities which contributed
disinfected water. Under the Stage 1 conclusions, the Advisory Committee to the development of the Stage 2 DBPR.
DBPR, water systems that use surface developed an Agreement in Principle
water or ground water under the direct (65 FR 83015, December 29, 2000) Additionally, EPA posted a pre-
influence of surface water and use (USEPA 2000a) that laid out their proposal draft of the Stage 2 DBPR
conventional filtration treatment are consensus recommendations on how to preamble and regulatory language on an
required to remove specified further control DBPs in public water EPA Internet site on October 17, 2001.
percentages of organic materials, systems, which are reflected in today’s This public review period allowed
measured as total organic carbon (TOC), final rule. readers to comment on the Stage 2
that may react with disinfectants to form In the Agreement in Principle, the DBPR’s consistency with the Agreement
DBPs. Removal is achieved through Advisory Committee recommended in Principle of the Stage 2 M-DBP
enhanced coagulation or enhanced maintaining the MCLs for TTHM and Advisory Committee. EPA received
wwhite on PROD1PC65 with RULES2

softening, unless a system meets one or HAA5 at 0.080 mg/L and 0.060 mg/L, important suggestions on this pre-
more alternative compliance criteria. respectively, but changing the proposal draft from 14 commenters,
The Stage 1 DBPR was one of the first compliance calculation in two phases to which included public water systems,
rules to be promulgated under the 1996 facilitate systems moving from the State governments, laboratories, and
SDWA Amendments (USEPA 1998a). running annual average (RAA) other stakeholders.

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00007 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
394 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations

C. Public Health Concerns to be formation within and between water weight of evidence involves
Addressed systems. considerations of the quality and
THMs and HAAs are widely occurring adequacy of data and consistency of
EPA is promulgating the Stage 2 rule
classes of DBPs formed during responses. These factors are not scored
to reduce the potential risks of cancer
disinfection with chlorine and mechanically by adding pluses and
and reproductive and developmental
chloramine. The four THMs (TTHM) minuses; they are judged in
health effects from DBPs. In addition,
and five HAAs (HAA5) measured and combination.
the provisions of the Stage 2 DBPR
regulated in the Stage 2 DBPR act as • Criteria for determining ‘causality’
provide for more equitable public health
indicators for DBP occurrence. There are include consistency, strength, and
protection. Sections C and D describe
other known DBPs in addition to a specificity of association, a temporal
the general basis for this public health
variety of unidentified DBPs present in relationship, a biological gradient (dose-
concern through reviewing information disinfected water. THMs and HAAs response relationship), biological
in the following areas: the health effects typically occur at higher levels than plausibility, coherence with multiple
associated with DBPs, DBP occurrence, other known and unidentified DBPs lines of evidence, evidence from human
and the control of DBPs. (McGuire et al. 2002; Weinberg et al. populations, and information on agent’s
1. What Are DBPs? 2002). The presence of TTHM and structural analogues (USEPA 2005i).
HAA5 is representative of the Additional considerations for individual
Chlorine has been widely used to kill
occurrence of many other chlorination study findings include reliable exposure
disease-causing microbes in drinking
DBPs; thus, a reduction in the TTHM data, statistical power and significance,
water. The addition of chlorine in PWSs
and HAA5 generally indicates an overall and freedom from bias and
across the U.S. to kill microbial
reduction of DBPs. confounding.
pathogens in the water supply has been • The term ‘hazard’ describes not a
cited as one of the greatest public health 2. DBP Health Effects definitive conclusion, but the possibility
advances of the twentieth century Since the mid 1980’s, epidemiological that a health effect may be attributed to
(Okun 2003). For example, during the studies have supported a potential a certain exposure, in this case
decade 1880–1890, American cities association between bladder cancer and chlorinated water. Analyses done for the
experienced an average mortality rate of chlorinated water and possibly also Stage 2 DBPR follow the 1999 EPA
58 per 100,000 from typhoid, which was with colon and rectal cancers. In Proposed Guidelines for Carcinogenic
commonly transmitted through addition, more recent health studies Risk Assessment (USEPA 1999a). In
contaminated water. By 1938, this rate have reported potential associations March 2005, EPA updated and finalized
had fallen to 0.67 deaths per 100,000, between chlorinated drinking water and the Cancer Guidelines and a
largely due to improved treatment of reproductive and developmental health Supplementary Children’s Guidance,
drinking water (Blake 1956). effects. which include new considerations on
During the disinfection process, Based on a collective evaluation of mode of action for cancer risk
organic and inorganic material in source both the human epidemiology and determination and additional potential
waters can combine with chlorine and animal toxicology data on cancer and risks due to early childhood exposure
certain other chemical disinfectants to reproductive and developmental health (USEPA 2005i; USEPA 2005j).
form DBPs. More than 260 million effects discussed below and in Conducting the cancer evaluation using
people in the U.S. are exposed to consideration of the large number of the 2005 Cancer Guidelines would not
disinfected water and DBPs (USEPA people exposed to chlorinated result in any change from the existing
2005a). Although chlorine is the most byproducts in drinking water (more analysis. With the exception of
commonly applied disinfectant, other than 260 million), EPA concludes that chloroform, no mode of action has been
disinfectants, including ozone, chlorine (1) new cancer data since Stage 1 established for other specific regulated
dioxide, chloramine, and ultraviolet strengthen the evidence of a potential DBPs. Although some of the DBPs have
radiation, are in use. In combination association of chlorinated water with given mixed mutagenicity and
with these, all surface water systems bladder cancer and suggests an genotoxicity results, having a positive
must also use either chlorine or association for colon and rectal cancers, mutagenicity study does not necessarily
chloramine to maintain a disinfectant (2) current reproductive and mean that a chemical has a mutagenic
residual in their distribution system. developmental health effects data do not mode of action. The extra factor of
The kind of disinfectant used can support a conclusion at this time as to safety for children’s health protection
produce different types and levels of whether exposure to chlorinated does not apply because the new
disinfectant byproducts in the drinking drinking water or disinfection Supplementary Children’s Guidance
water. byproducts causes adverse requires application of the children’s
Many factors affect the amount and developmental or reproductive health factor only when a mutagenic mode of
kinds of DBPs in drinking water. Areas effects, but do support a potential health action has been identified.
in the distribution system that have had concern, and (3) the combined health a. Cancer health effects. The following
longer contact time with chemical data indicate a need for public health section briefly discusses cancer
disinfectants tend to have higher levels protection beyond that provided by the epidemiology and toxicology
of DBPs, such as sites farther from the Stage 1 DBPR. information EPA analyzed and some
treatment plant, dead ends in the This section summarizes the key conclusions of these studies and reports.
system, and small diameter pipes. The information in the areas of cancer, Further discussion of these studies and
makeup and source of the water also reproductive, and developmental health EPA’s conclusions can be found in the
affect DBP formation. Different types of studies that EPA used to arrive at these proposed Stage 2 DBPR (USEPA 2003a)
organic and inorganic material will form conclusions. Throughout this writeup, and the Economic Analysis for the Final
wwhite on PROD1PC65 with RULES2

different types and levels of DBPs. Other EPA uses ‘weight of evidence,’ Stage 2 Disinfectants and Disinfection
factors, such as water temperature, ‘causality,’ and ‘hazard’ as follows: Byproducts Rule (Economic Analysis
season, pH, and location within the • A ‘weight of evidence’ evaluation is (EA)) (USEPA 2005a).
water purification process where a collective evaluation of all pertinent Human epidemiology studies and
disinfectants are added, can affect DBP information. Judgement about the animal toxicology studies have

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00008 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 395

examined associations between ascertainment of cause of the cancer, surface waters. The database of studies
chlorinated drinking water or DBPs and and reduction of potential selection bias on colon and rectal cancers continues to
cancer. While EPA cannot conclude in case-control and cohort studies (by support a possible association, but
there is a causal link between exposure having comparable cases and controls evidence remains mixed. For colon
to chlorinated surface water and cancer, and by limiting loss to follow-up). cancer, one newer study supports the
EPA believes that the available research Epidemiology studies provide extremely evidence of an association (King et al.
indicates a potential association useful information on human exposure 2000a) while others showed
between bladder cancer and exposure to to chlorinated water, which inconsistent findings (Hildesheim et al.
chlorinated drinking water or DBPs. complement single chemical, high dose 1998; Yang et al. 1998). Rectal cancer
EPA also believes the available research animal data. studies are also mixed. Hildesheim et al.
suggests a possible association between In the Stage 1 DBPR, EPA concluded
(1998) and Yang et al. (1998) support an
rectal and colon cancers and exposure that the epidemiological evidence
association with rectal cancer while
to chlorinated drinking water or DBPs. suggested a potential increased risk for
bladder cancer. Some key studies EPA King et al. (2000a) did not. A review of
This is based on EPA’s evaluation of all
considered for Stage 1 include Cantor et colon and rectal cancer concluded
available cancer studies. The next two
sections focus on studies published al. (1998), Doyle et al. (1997), Freedman evidence was inconclusive but that
since the Stage 1 DBPR. Conclusions are et al. (1997), King and Marrett (1996), there was a stronger association for
based on the research as a whole. McGeehin et al. (1993), Cantor et al. rectal cancer and chlorination DBPs
i. Epidemiology. A number of (1987), and Cantor et al. (1985). Several than for colon cancer (Mills et al. 1998).
epidemiological studies have been studies published since the Stage 1 The WHO (2000) review reported that
conducted to investigate the DBPR continue to support an studies showed weak to moderate
relationship between exposure to association between increased risk of associations with colon and rectal
chlorinated drinking water and various bladder cancer and exposure to cancers and chlorinated surface water or
cancers. These studies contribute to the chlorinated surface water (Chevrier et THMs but that evidence is inadequate to
overall evidence on potential human al. 2004; Koivusalo et al. 1998; Yang et evaluate these associations.
health hazards from exposure to al. 1998). One study found no effects on Recent studies on kidney, brain, and
chlorinated drinking water. a biomarker of genotoxicity in urinary lung cancers and DBP exposure support
Epidemiology studies provide useful bladder cells from TTHM exposure a possible association (kidney: Yang et
health effects information because they (Ranmuthugala et al. 2003). al. 1998, Koivusalo et al. 1998; brain:
reflect human exposure to a drinking Epidemiological reviews and meta-
water DBP mixture through multiple Cantor et al. 1999; lung: Yang et al.
analyses generally support the
routes of intake such as ingestion, 1998). However, so few studies have
possibility of an association between
inhalation and dermal absorption. The chlorinated water or THMs and bladder examined these endpoints that
greatest difficulty with conducting cancer (Villanueva et al. 2004; definitive conclusions cannot be made.
cancer epidemiology studies is the Villanueva et al. 2003; Villanueva et al. Studies on leukemia found little or no
length of time between exposure and 2001; Mills et al. 1998). The World association with DBPs (Infante-Rivard et
effect. Higher quality studies have Health Organization (WHO 2000) found al. 2002; Infante-Rivard et al. 2001). A
adequately controlled for confounding data inconclusive or insufficient to recent study did not find an association
and have limited the potential for determine causality between between pancreatic cancer and DBPs
exposure misclassification, for example, chlorinated water and any health (Do et al. 2005). A study researching
using DBP levels in drinking water as endpoint, although they concluded that multiple cancer endpoints found an
the exposure metric as opposed to type the evidence is better for bladder cancer association between THM exposure and
of source water. Study design than for other cancers. all cancers when grouped together
considerations for interpreting cancer In the Stage 1 DBPR, EPA concluded (Vinceti et al. 2004). More details on the
epidemiology data include sufficient that early studies suggested a small cancer epidemiology studies since the
follow-up time to detect disease possible increase in rectal and colon Stage 1 DBPR are outlined in Table II.D–
occurrence, adequate sample size, valid cancers from exposure to chlorinated 1.


Study type Exposure(s) studied Findings

Do et al. 2005 Case-control Estimated chlorinated DBPs, Pancreatic can- No association was found between pancreatic cancer and
study in chloroform, BDCM con- cer. exposure to chlorinated DBPs, chloroform, or BDCM.
Canada, centrations.
Chevrier et al. Case-control Compared THM levels, dura- Bladder cancer. A statistically significant decreased risk of bladder cancer
2004.. study in tion of exposure, and 3 was found as duration of exposure to ozonated water in-
France, types of water treatment creased. This was evident with and without adjustment
1985–1987. (ozonation, chlorination, for other exposure measures. A small association was
ozonation/chlorination). detected for increased bladder cancer risk and duration
of exposure to chlorinated surface water and with the es-
timated THM content of the water, achieving statistical
wwhite on PROD1PC65 with RULES2

significance only when adjusted for duration of ozonated

water exposures. Effect modification by gender was
noted in the adjusted analyses.

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00009 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
396 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations


Study type Exposure(s) studied Findings

Vinceti et al. Retrospective Standardized mortality ratios 15 cancers in-

Mortality ratio from all cancers showed a statistically signifi-
2004. cohort study from all causes vs. cancer cluding colon,
cant small increase for males consuming drinking water
in Italy, for consumers drinking rectum, and with high THMs. For females, an increased mortality ratio
1987–1999. water with high THMs. bladder. for all cancers was seen but was not statistically signifi-
cant. Stomach cancer in men was the only individual
cancer in which a statistically significant excess in mor-
tality was detected for consumption of drinking water with
high THMs.
Ranmuthugala Cohort study Estimated dose of TTHM, Frequency of Relative risk estimates for DNA damage to bladder cells for
et al. 2003. in 3 Aus- chloroform, and bromoform micronuclei in THM dose metrics were near 1.0. The study provides no
tralian com- from routinely-collected urinary blad- evidence that THMs are associated with DNA damage to
munities, THM measurements and der epithelial bladder epithelial cells, and dose-response patterns were
1997. fluid intake diary. cells. not detected.
Infante-Rivard Population- Estimated prenatal and post- Acute Data are suggestive, but imprecise, linking DNA variants
et al. 2002. based case- natal exposure to THMs lymphoblastic with risk of acute lymphoblastic leukemia associated with
control study and polymorphisms in two leukemia. drinking water DBPs. The number of genotyped subjects
in Quebec, genes. for GSTT1 and CYP2E1 genes was too small to be con-
1980–1993. clusive.
Infante-Rivard Population- Compared water chlorination Acute No increased risk for lymphoblastic leukemia was observed
et al. 2001. based case- (never, sometimes, always) lymphoblastic for prenatal exposure at average levels of TTHMs, met-
control study and exposure to TTHMs, leukemia. als or nitrates. However, a non-statistically significant,
in Quebec, metals, and nitrates. small increased risk was seen for postnatal cumulative
1980–1993. exposure to TTHMs and chloroform (both at above the
95th exposure percentile of the distribution for cases and
controls), for zinc, cadmium, and arsenic, but not other
metals or nitrates.
King et al. Population- Compared source of drinking Colon and rec- Colon cancer risk was statistically associated with cumu-
2000a. based case- water and chlorination sta- tal cancer. lative long term exposure to THMs, chlorinated surface
control study tus. Estimated TTHM lev- water, and tap water consumption metrics among males
in southern els, duration of exposure, only. Exposure-response relationships were evident for
Ontario, and tap water consumption. exposure measures combining duration and THM levels.
1992–1994. Associations between the exposure measures and rectal
cancer were not observed for either gender.
Cantor et al. Population- Compared level and duration Brain cancer .... Among males, a statistically significant increased risk of
1999. based case- of THM exposure (cumu- brain cancer was detected for duration of chlorinated
control study lative and average), source versus non-chlorinated source water, especially among
in Iowa, of water, chlorination, and high-level consumers of tap water. An increased risk of
1984–1987. water consumption. brain cancer for high water intake level was found in
men. No associations were found for women for any of
the exposure metrics examined.
Cantor et al. Population- Compared level and duration Bladder cancer A statistically significant positive association between risk
1998. based case- of THM exposure (cumu- of bladder cancer and exposure to chlorinated ground-
control study lative and average), source water or surface water reported for men and for smokers,
in Iowa, of water, chlorination, and but no association found for male/female non-smokers,
1986–1989. water consumption. or for women overall. Limited evidence was found for an
association between tapwater consumption and bladder
cancer risk. Suggestive evidence existed for exposure-re-
sponse effects of chlorinated water and lifetime THM
measures on bladder cancer risk.
Hildesheim et Population- Compared level and duration Colon and rec- Increased risks of rectal cancer was associated with dura-
al. 1998. based case- of THM exposure (cumu- tal cancer. tion of exposure to chlorinated surface water and any
control study lative and average), source chlorinated water, with evidence of an exposure-re-
in Iowa, of water, chlorination, and sponse relationship. Risk of rectal cancer is statistically
1986–1989. water consumption. significant increased with >60 years lifetime exposure to
THMs in drinking water, and risk increased for individuals
with low dietary fiber intake. Risks were similar for men
and women and no effects were observed for tapwater
measures. No associations were detected for water ex-
posure measures and risk of colon cancer.
Koivusalo et Population- Estimated residential duration Bladder and Drinking water mutagenicity was associated with a small,
al. 1998. based case- of exposure and level of kidney cancer. statistically significant, exposure-related excess risk for
control study drinking water mutagenicity. kidney and bladder cancers among men; weaker asso-
in Finland, ciations were detected for mutagenic water and bladder
1991–1992. or kidney cancer among women. The effect of mutage-
wwhite on PROD1PC65 with RULES2

nicity on bladder cancer was modified by smoking status,

with an increased risk among non-smokers.

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00010 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 397


Study type Exposure(s) studied Findings

Yang et al. Cross-sec- Examined residence in Cancer of rec- Residence in chlorinating municipalities (vs. non-
1998. tional study chlorinated (mainly surface tum, lung, chlorinating) was statistically significantly associated with
in Taiwan, water sources) relative to bladder, kid- the following types of cancer in both males and females:
1982–1991. non-chlorinated (mainly pri- ney, colon, rectal, lung, bladder, and kidney cancer. Liver cancer
vate well) water. and 11 others. and all cancers were also statistically significantly ele-
vated in chlorinated towns for males only. Mortality rates
for cancers of the esophagus, stomach, colon, pancreas,
prostate, brain, breast, cervix uteri and uterus, and ovary
were comparable for chlorinated and non-chlorinated res-
Doyle et al. Prospective Examined chloroform levels Colon, rectum, Statistically significant increased risk of colon cancer,
1997. cohort study and source of drinking bladder, and breast cancer and all cancers combined was observed
in Iowa, water. 8 other can- for women exposed to chloroform in drinking water, with
1987–1993. cers in evidence of exposure-response effects. No associations
women. were detected between chloroform and bladder, rectum,
kidney, upper digestive organs, lung, ovary, endo-
metrium, or breast cancers, or for melanomas or non-
Hodgkin’s lymphoma. Surface water exposure (compared
to ground water users) was also a significant predictor of
colon and breast cancer risk.
Freedman et Population- Estimated duration of expo- Bladder cancer There was a weak association between bladder cancer risk
al. 1997. based case- sure to chlorinated water. and duration of exposure to municipal water for male cig-
control study Compared exposure to arette smokers, as well as an exposure-response rela-
in Maryland, chlorinated municipal water tionship. No association was seen for those with no his-
1975–1992. (yes/no). tory of smoking, suggesting that smoking may modify a
possible effect of chlorinated surface water on the risk of
bladder cancer.
King and Case-control Compared source of drinking Bladder cancer Statistically significant associations were detected for blad-
Marrett 1996. study in On- water and chlorination sta- der cancer and chlorinated surface water, duration or
tario, Can- tus. Estimated TTHM lev- concentration of THM levels and tap water consumption
ada, 1992– els, duration of exposure, metrics. Population attributable risks were estimated at
1994. and tap water consumption. 14 to 16 percent. An exposure-response relationship was
observed for estimated duration of high THM exposures
and risk of bladder cancer.
McGeehin et Population- Compared source of drinking Bladder cancer Statistically significant associations were detected for blad-
al. 1993. based case- water, water treatment, and der cancer and duration of exposure to chlorinated sur-
control study tap water versus bottled face water. The risk was similar for males and females
in Colorado, water. Estimated duration and among nonsmokers and smokers. The attributable
1990–1991. of exposure to TTHMs and risk was estimated at 14.9 percent. High tap water intake
levels of TTHMs, nitrates, was associated with risk of bladder cancer in a expo-
and residual chlorine. sure-response fashion. No associations were detected
between bladder cancer and levels of TTHMs, nitrates,
and residual chlorine.
Cantor et al. Population- Compared source of drinking Bladder cancer Bladder cancer was statistically associated with duration of
1987 (and based case- water. Estimated total bev- exposure to chlorinated surface water for women and
Cantor et al. control study erage and tap water con- nonsmokers of both sexes. The largest risks were seen
1985). in 10 areas sumption and duration of when both exposure duration and level of tap water in-
of the U.S., exposure. gestion were combined. No association was seen for
1977–1978. total beverage consumption.
Villanueva et Review and Individual-based exposure Bladder cancer The meta-analysis suggests that risk of bladder cancer in
al. 2004. meta-anal- estimates to THMs and men increases with long-term exposure to TTHMs. An
ysis of 6 water consumption over a exposure-response pattern was observed among men
case-control 40-year period. exposed to TTHMs, with statistically significant risk seen
studies. at exposures higher than 50 ug/L. No association be-
tween TTHMs and bladder cancer was seen for women.
Villanueva et Review and Compared source of water Bladder cancer The meta-analysis findings showed a moderate excess risk
al. 2003 meta-anal- and estimated duration of of bladder cancer attributable to long-term consumption
(and Goebell ysis of 6 exposure to chlorinated of chlorinated drinking water for both genders, particu-
et al. 2004). case-control drinking water. larly in men. Statistically significance seen with men and
studies and combined both sexes. The risk was higher when expo-
2 cohort sure exceeded 40 years.
wwhite on PROD1PC65 with RULES2

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00011 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
398 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations


Study type Exposure(s) studied Findings

Villanueva et Qualitative re- Compared exposure to TTHM Cancer of blad- Review found that although results for cancer studies var-
al. 2001. view of 31 levels, mutagenic drinking der, colon, ied and were not always statistically significant, evidence
cancer stud- water, water consumption, rectum, and 5 for bladder cancer is strongest, and all 10 of the bladder
ies. source water, types of dis- other can- cancer studies showed increased cancer risks with in-
infection (chlorination and cers.. gestion of chlorinated water. The authors felt associa-
chloramination), and resi- tions with chlorinated water and cancer of the colon, rec-
dence times. tum, pancreas, esophagus, brain, and other cancers
were inconsistent.
WHO 2000 ..... Qualitative re- Various exposures to THMs. Various cancers Studies reviewed reported weak to moderate increased rel-
views of var- ative risks of bladder, colon, rectal, pancreatic, breast,
ious studies brain or lung cancer associated with long-term exposure
in Finland, to chlorinated drinking water. The authors felt evidence is
U.S., and inconclusive for an association between colon cancer
Canada. and long-term exposure to THMs; that evidence is insuffi-
cient to evaluate a causal relationship between THMs
and rectal, bladder, and other cancers. They found no
association between THMs and increased risk of cardio-
vascular disease.
Mills et al. Qualitative re- Examined TTHM levels and Cancer of Review suggests possible increases in risks of bladder
1998. view of 22 water consumption. Com- colon, rec- cancer with exposure to chlorinated drinking water. The
studies. pared source of water and tum, and authors felt evidence for increased risk of colon and rec-
2 types of water treatment bladder. tal cancers is inconclusive, though evidence is stronger
(chlorination and for rectal cancer.

Overall, bladder cancer data provide cancer toxicology and mode of action (USEPA 2001a), chlorine dioxide and
the strongest basis for quantifying studies completed since the Stage 1 chlorite (USEPA 2000c), and bromate
cancer risks from DBPs. EPA has chosen DBPR was finalized in December 1998. (USEPA 2001b), and is currently
this endpoint to estimate the primary In support of this rule, EPA has reassessing TCAA.
benefits of the Stage 2 DBPR (see developed health criteria documents
which summarize the available Slope factors and risk concentrations
Section VI).
ii. Toxicology. Cancer toxicology toxicology data for brominated THMs for BDCM, bromoform, DBCM and
studies provide additional support that (USEPA 2005b), brominated HAAs DCAA have been developed and are
chlorinated water is associated with (USEPA 2005c), MX (USEPA 2000b), listed in Table II.D–2. For BDCM,
cancer. In general, EPA uses long term MCAA (USEPA 2005d), and TCAA bromoform, and DBCM, table values are
toxicology studies that show a dose (USEPA 2005e). The 2003 IRIS derived from the brominated THM
response to derive MCLGs and cancer assessment of DCAA (USEPA 2003b) criteria document (USEPA 2005b),
potency factors. Short term studies are and an addendum (USEPA 2005k) also which uses IRIS numbers that have been
used for hazard identification and to provides analysis released after Stage 1. updated using the 1999 EPA Proposed
design long term studies. Much of the It summarizes information on exposure Guidelines for Carcinogenic Risk
available cancer toxicology information from drinking water and develops a Assessment (USEPA 1999a). For DCAA,
was available for the Stage 1 DBPR, but slope factor for DCAA. IRIS also has the values are derived directly from
there have also been a number of new toxicological reviews for chloroform IRIS.


LED 10a ED 10a

Disinfection byproduct Slope 10 ¥6 Risk Slope 10 ¥6 Risk

factor concentration factor concentration
(mg/kg/day)¥1 (mg/L) (mg/kg/day) ¥1 (mg/L)

Bromodichloromethane .................................................................... 0.034 0.001 0.022 0.002

Bromoform ....................................................................................... 0.0045 0.008 0.0034 0.01
Dibromochloromethane .................................................................... 0.04 0.0009 0.017 0.002
Dichloroacetic Acid .......................................................................... 0.048 0.0007 0.015 b 0.0023 b
10 is the lower 95% confidence bound on the (effective dose) ED10 value. ED10 is the estimated dose producing effects in 10% of ani-
b The ED
10 risk factors for DCAA have been changed from those given in the comparable table in the proposed Stage 2 DBPR to correct for
transcriptional errors.
wwhite on PROD1PC65 with RULES2

More research on DBPs is underway been completed on BDCM and chlorate.
at EPA and other research institutions. intro.html). Two-year bioassays by the The draft abstract on BDCM reported no
Summaries of on-going studies may be National Toxicology Program (NTP) evidence of carcinogenicity when
found on EPA’s DRINK Web site (http:// released in abstract form have recently BDCM was administered via drinking

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00012 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 399

water (NTP 2005a). Another recent ability to detect statistically significant endpoints (spontaneous abortion or
study, a modified two-year bioassay on associations for small to moderate stillbirth) remains inconsistent as a
BDCM in the drinking water, reported relative risk estimates. Small sample whole, there is suggestive evidence of
little evidence of carcinogenicity sizes also result in imprecision around an association between fetal loss and
(George et al. 2002). In a previous NTP risk estimates reflected by wide chlorinated water or DBP exposure.
study, tumors were observed, including confidence intervals. In addition to the Various studies support the possibility
an increased incidence of kidney, liver, limitations of individual studies, that exposure to chlorinated water or
and colon tumors, when BDCM was evaluating reproductive and DBPs is associated with decreased fetal
administered at higher doses by gavage developmental epidemiology studies viability (Toledano et al. 2005; Dodds et
in corn oil (NTP 1987). EPA will collectively is difficult because of the al. 2004; King et al. 2000b; Dodds et al.
examine new information on BDCM as methodological differences between 1999; Waller et al. 1998; Aschengrau et
it becomes available. In the chlorate studies and the wide variety of al. 1993; Aschengrau et al. 1989). Other
draft abstract, NTP found some evidence endpoints examined. These factors may studies did not support an association
that it may be a carcinogen (NTP 2004). contribute to inconsistencies in the (Bove et al. 1995) or reported
Chlorate is a byproduct of hypochlorite scientific body of literature as noted inconclusive results (Savitz et al. 2005;
and chlorine dioxide systems. A long- below. Swan et al. 1998; Savitz et al. 1995)
term, two-year bioassay NTP study on More recent studies tend to be of between fetal viability and exposure to
DBA is also complete but has not yet higher quality because of improved THMs or tapwater. A recent study by
undergone peer review (NTP 2005b). exposure assessments and other King et al. (2005) found little evidence
b. Reproductive and developmental methodological advancements. For of an association between stillbirths and
health effects. Both human example, studies that use THM levels to haloacetic acids after controlling for
epidemiology studies and animal estimate exposure tend to be higher trihalomethane exposures, though non-
toxicology studies have examined quality than studies that define statistically significant increases in
associations between chlorinated exposure by source or treatment. These stillbirths were seen across various
drinking water or DBPs and factors were taken into account by EPA exposure levels.
reproductive and developmental health when comparing and making Fetal malformations. A number of
effects. Based on an evaluation of the conclusions on the reproductive and epidemiology studies have examined
available science, EPA believes the data developmental epidemiology literature. the relationship between fetal
suggest that exposure to DBPs is a What follows is a summary of available malformations (such as neural tube, oral
potential reproductive and epidemiology literature on reproductive cleft, cardiac, or urinary defects, and
developmental health hazard. and developmental endpoints such as chromosomal abnormalities) and
The following section briefly spontaneous abortion, stillbirth, neural chlorinated water or DBPs. It is difficult
discusses the reproductive and tube and other birth defects, low birth to assess fetal malformations in
developmental epidemiology and weight, and intrauterine growth aggregate due to inconsistent findings
toxicology information available to EPA. retardation. Information is grouped, and disparate endpoints being examined
Further discussion of these studies and where appropriate, into three categories in the available studies. Some studies
EPA’s conclusions can be found in the on fetal growth, viability, and support the possibility that exposure to
proposed Stage 2 DBPR (USEPA 2003a) malformations, and reviews are chlorinated water or DBPs is associated
and the Economic Analysis (USEPA described separately afterward. Table with various fetal malformations
2005a). II.D–3 provides a more detailed (Cedergren et al. 2002; Hwang et al.
i. Epidemiology. As discussed description of each study or review. 2002; Dodds and King 2001; Klotz and
previously, epidemiology studies have Fetal growth. Many studies looked for Pyrch 1999; Bove et al. 1995;
the strength of relating human exposure an association between fetal growth Aschengrau et al. 1993). Other studies
to DBP mixtures through multiple (mainly small for gestational age, low found little evidence (Shaw et al. 2003;
intake routes. Although the critical birth weight, and pre-term delivery) and Källén and Robert 2000; Dodds et al.
exposure window for reproductive and chlorinated water or DBPs. The results 1999; Shaw et al. 1991) or inconclusive
developmental effects is much smaller from the collection of studies as a whole results (Magnus et al. 1999) between
than that for cancer (generally weeks are inconsistent. A number of studies chlorinated water or DBP exposure and
versus years), exposure assessment is support the possibility that exposure to fetal malformations. Birth defects most
also a main limitation of reproductive chlorinated water or DBPs are consistently identified as being
and developmental epidemiology associated with adverse fetal growth associated with DBPs include neural
studies. Exposure assessment effects (Infante-Rivard 2004; Wright et tube defects and urinary tract
uncertainties arise from limited data on al. 2004; Wright et al. 2003; Källén and malformations.
DBP concentrations and maternal water Robert 2000; Gallagher et al. 1998; Other endpoints have also been
usage and source over the course of the Kanitz et al. 1996; Bove et al. 1995; examined in recent epidemiology
pregnancy. However, classification Kramer et al. 1992). Other studies studies. One study suggests an
errors typically push the true risk showed mixed results (Porter et al. association between DBPs and
estimate towards the null value (Vineis 2005; Savitz et al. 2005; Yang 2004) or decreased menstrual cycle length
2004). According to Bove et al. (2002), did not provide evidence of an (Windham et al. 2003), which, if
‘‘Difficulties in assessing exposure may association (Toledano et al. 2005; corroborated, could be linked to the
result in exposure misclassification Jaakkola et al. 2001; Dodds et al. 1999; biological basis of other reproductive
biases that would most likely produce Savitz et al. 1995) between DBP endpoints observed. No association
substantial underestimates of risk as exposure and fetal growth. EPA notes between THM exposure and semen
well as distorted or attenuated that recent, higher quality studies quality was found (Fenster et al. 2003).
wwhite on PROD1PC65 with RULES2

exposure-response trends.’’ Studies of provide some evidence of an increased More work is needed in both areas to
rare outcomes (e.g., individual birth risk of small for gestational age and low support these results.
defects) often have limited statistical birth weight. Reviews. An early review supported
power because of the small number of Fetal viability. While the database of an association between measures of fetal
cases being examined. This limits the epidemiology studies for fetal loss viability and tap water (Swan et al.

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00013 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
400 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations

1992). Three other reviews found data et al. (2000) concluded that the weight reproductive effects and drinking
inadequate to support an association of evidence from epidemiology studies chlorinated water. Graves et al. (2001)
between reproductive and suggests that ‘‘DBPs are likely to be supports a possible association for fetal
developmental health effects and THM reproductive toxicants in humans under growth but not fetal viability or
exposure (Reif et al. 1996; Craun 1998; appropriate exposure conditions,’’ but malformations. More recently, Bove et
WHO 2000). Mills et al. (1998) from a risk assessment perspective, data al. (2002) examined and supported an
examined data on and found support for are primarily at the hazard association between small for
an association between fetal viability identification stage. Nieuwenhuijsen et gestational age, neural tube defects and
and malformations and THMs. Another al. (2000) found some evidence for an spontaneous abortion endpoints and
review presented to the Stage 2 MDBP association between fetal growth and DBPs. Following a meta-analysis on five
FACA found some evidence for an THM exposure and concluded evidence malformation studies, Hwang and
association with fetal viability and some for associations with other fetal Jaakkola (2003) concluded that there
fetal malformations and exposure to endpoints is weak but gaining weight. A was evidence which supported
DBPs but reported that the evidence was qualitative review by Villanueva et al. associations between DBPs and risk of
inconsistent for these endpoints as well (2001) found evidence generally birth defects, especially neural tube
as for fetal growth (Reif et al. 2000). Reif supports a possible association between defects and urinary tract defects.


Author(s) Study type Exposure(s) studied Outcome(s) measured Findings

Porter et al. Cross-sectional study in Estimated THM and Intrauterine growth re- No consistent association or dose-response rela-
2005. Maryland, 1998–2002. HAA exposure during tardation. tionship was found between exposure to either
pregnancy. TTHM or HAA5 and intrauterine growth retar-
dation. Results suggest an increased risk of
intrauterine growth retardation associated with
TTHM and HAA5 exposure in the third tri-
mester, although only HAA5 results were sta-
tistically significant.
Savitz et al. Population-based pro- Estimated TTHM, HAA9, Early and late preg- No association with pregnancy loss was seen
2005. spective cohort study and TOC exposures nancy loss, preterm when looking at high exposure of TTHM com-
in three communities during pregnancy. In- birth, small for gesta- pared to low exposure of TTHM. When exam-
around the U.S., dices examined in- tional age, and term ining individual THMs, a statistically significant
2000–2004. cluded concentration, birth weight. association was found between
ingested amount, ex- bromodichloromethane (BDCM) and preg-
posure from show- nancy loss. A similar, non-statistically signifi-
ering and bathing, cant association was seen between
and an integration of dibromochloromethane (DBCM) and preg-
all exposures com- nancy loss. Some increased risk was seen for
bined. losses at greater than 12 weeks’ gestation for
TTHM, BDCM, and TOX (total organic halide),
but most results generally did not provide sup-
port for an association. Preterm birth showed
a small inverse relationship with DBP expo-
sure (i.e. higher exposures showed less
preterm births), but this association was weak.
TTHM exposure of 80 ug/L was associated
with twice the risk for small for gestational age
during the third trimester and was statistically
Toledano et Large cross-sectional Linked mother’s resi- Stillbirth, low birth A significant association between TTHM and risk
al. 2005. study in England, dence at time of deliv- weight. of stillbirth, low birth weight, and very low birth
1992–1998. ery to modeled esti- weight was observed in one of the three re-
mates of TTHM levels gions. When all three regions were combined,
in water zones. small, but non-significant, excess risks were
found between all three outcomes and TTHM
and chloroform. No associations were ob-
served between reproductive risks and BDCM
or total brominated THMs.
Dodds et al. Population-based case- Estimated THM and Stillbirth ......................... A statistically significant association was ob-
2004 (and control study in Nova HAA exposure at resi- served between stillbirths and exposure to
King et al. Scotia and Eastern dence during preg- total THM, BDCM, and chloroform. Associa-
2005). Ontario, 1999–2001. nancy. Linked water tions were also detected for metrics, which in-
consumption and corporated water consumption, showering and
showering/bathing to bathing habits. Elevated relative risks were ob-
THM exposure. served for intermediate exposures for total
HAA and DCAA measures; TCAA and
brominated HAA exposures showed no asso-
wwhite on PROD1PC65 with RULES2

ciation. No statistically significant associations

or dose-response relationships between any
HAAs and stillbirth were detected after control-
ling for THM exposure.

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00014 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 401


Author(s) Study type Exposure(s) studied Outcome(s) measured Findings

Infante- Case-control study of Estimated THM levels Intrauterine growth re- No associations were found between exposure
Rivard newborns in Montreal, and water consump- tardation. to THMs and intrauterine growth retardation.
2004. 1998–2000. tion during pregnancy. However, a significant effect was observed be-
Exposure from show- tween THM exposure and intrauterine growth
ering and presence of retardation for newborns with the CYP2E1
two genetic gene variant. Findings suggest that exposure
polymorphisms. to THMs at the highest levels can affect fetal
growth but only in genetically susceptible
Wright et al. Large cross-sectional Estimated maternal Birth weight, small for Statistically significant reductions in mean birth
2004. study: Massachusetts, third-trimester expo- gestational age, weight were observed for BDCM, chloroform,
1995–1998. sures to TTHMs, chlo- preterm delivery, ges- and mutagenic activity. An exposure-response
roform, BDCM, total tational age. relationship was found between THM expo-
HAAs, DCA, TCA, MX sure and reductions in mean birth weight and
and mutagenicity in risk of small for gestational age. There was no
drinking water. association between preterm delivery and ele-
vated levels of HAAs, MX, or mutagenicity. A
reduced risk of preterm delivery was observed
with high THM exposures. Gestational age
was associated with exposure to THMs and
Yang et al. Large cross-sectional Compared maternal Low birth weight, Residence in area supplied with chlorinated
2004 (and studies in Taiwan, consumption of preterm delivery. drinking water showed a statistically significant
Yang et 1994–1996. chlorinated drinking association with preterm delivery. No associa-
al. 2000). water (yes/no). tion was seen between chlorinated drinking
water and low birth weight.
Fenster et Small prospective study Examined TTHM levels Sperm motility, sperm No association between TTHM level and sperm
al. 2003. in California, 1990– within the 90 days morphology. mobility or morphology. BDCM was inversely
1991. preceding semen col- associated with linearity of sperm motion.
lection. There was some suggestion that water con-
sumption and other ingestion metrics may be
associated with different indicators of semen
Shaw et al. 2 case-control maternal Estimated THM levels Neural tube defects, oral No associations or exposure-response relation
2003. interview studies: CA, for mothers’ resi- clefts, selected heart were observed between malformations and
1987–1991. dences from before defects. TTHMs in either study.
conception through
early pregnancy.
Windham et Prospective study: CA, Estimated exposure to Menstrual cycle, fol- Findings suggest that THM exposure may affect
al. 2003. 1990–1991. THMs through show- licular phase length ovarian function. All brominated THM com-
ering and ingestion (in days). pounds were associated with significantly
over average of 5.6 shorter menstrual cycles with the strongest
menstrual cycles per finding for chlorodibromomethane. There was
woman. little association between TTHM exposure and
luteal phase length, menses length, or cycle
Wright et al. Cross-sectional study: Estimated TTHM expo- Birth weight, small for Statistically significant associations between 2nd
2003. Massachusetts, 1990. sure in women during gestational age, trimester and pregnancy average TTHM expo-
pregnancy (average preterm delivery, ges- sure and small for gestational age and fetal
for pregnancy and tational age. birth weight were detected. Small, statistically
during each trimester). significant increases in gestational duration/
age were observed at increased TTHM levels,
but there was little evidence of an association
between TTHM and preterm delivery or low
birth weight.
Cedergren Retrospective case-con- Examined maternal Cardiac defects ............. Exposure to chlorine dioxide in drinking water
et al. 2002. trol study: Sweden, periconceptional DBP showed statistical significance for cardiac de-
1982–1997. levels and used GIS fects. THM concentrations of 10 ug/L and
to assign water sup- higher were significantly associated with car-
plies. diac defects. No excess risk for cardiac defect
and nitrate were seen.
Hwang et al. Large cross-sectional Compared exposure to Birth defects (neural Risk of any birth defect, cardiac, respiratory sys-
2002. study in Norway, chlorination (yes/no) tube defects, cardiac, tem, and urinary tract defects were signifi-
1993–1998. and water color levels respiratory system, cantly associated with water chlorination. Ex-
for mother’s residence oral cleft, urinary posure to chlorinated drinking water was sta-
wwhite on PROD1PC65 with RULES2

during pregnancy. tract). tistically significantly associated with risk of

ventricular septal defects, and an exposure-re-
sponse pattern was seen. No other specific
defects were associated with the exposures
that were examined.

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00015 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
402 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations


Author(s) Study type Exposure(s) studied Outcome(s) measured Findings

Dodds and Population-based retro- Estimated THM, chloro- Neural tube defects, Exposure to BDCM was associated with in-
King 2001. spective cohort in form, and cardiovascular de- creased risk of neural tube defects, cardio-
Nova Scotia, 1988– bromodichloromethan- fects, cleft defects, vascular anomalies. Chloroform was not asso-
1995. e (BDCM) exposure. chromosomal abnor- ciated with neural tube defects, but was asso-
malities. ciated with chromosomal abnormalities. No as-
sociation between THM and cleft defects were
Jaakkola et Large cross-sectional Compared chlorination Low birth weight, small No evidence found for association between pre-
al. 2001. study in Norway, (yes/no) and water for gestational age, natal exposure to chlorinated drinking water
1993–1995. color (high/low) for preterm delivery. and low birth weight or small for gestational
mother during preg- age. A reduced risk of preterm delivery was
nancy. noted for exposure to chlorinated water with
high color content.
Källén and Large cross-sectional Linked prenatal expo- Gestational duration, A statistically significant difference was found for
Robert cohort study in Swe- sure to drinking water birth weight, intra- short gestational duration and low birth weight
2000. den, 1985–1994. disinfected with var- uterine growth, mor- among infants whose mother resided in areas
ious methods (no tality, congenital mal- using sodium hypochlorite, but not for chlorine
chlorine, chlorine di- formations, and other dioxide. Sodium hypochlorite was also associ-
oxide only, sodium birth outcomes. ated with other indices of fetal development
hypochlorite only). but not with congenital defects. No other ef-
fects were observed for intrauterine growth,
childhood cancer, infant mortality, low Apgar
score, neonatal jaundice, or neonatal
hypothyroidism in relation to either disinfection
Dodds et al. Population-based retro- Estimated TTHM level Low birth weight, A statistically significant increased risk for still-
1999 (and spective cohort study for women during preterm birth, small births and high total THMs and specific THMs
King et al. in Nova Scotia, 1988– pregnancy. for gestational age, during pregnancy was detected, with higher
2000b). 1995. stillbirth, chromosomal risks observed among asphyxia-related still-
abnormalities, neural births. Bromodichloromethane had the strong-
tube defects, cleft de- est association and exhibited an exposure-re-
fects, major cardiac sponse pattern. There was limited evidence of
defects. an association between THM level and other
reproductive outcomes. No congenital anoma-
lies were associated with THM exposure, ex-
cept for a non-statistically significant associa-
tion with chromosomal abnormalities.
Klotz and Population-based case- Estimated exposure of Neural tube defects ...... A significant association was seen between ex-
Pyrch control study in New pregnant mothers to posure to THMs and neural tube defects. No
1999 (and Jersey, 1993–1994. TTHMs and HAAs, associations were observed for neural tube
Klotz and and compared source defects and haloacetic acids or
Pyrch of water. haloacetonitriles.
Magnus et Large cross-sectional Compared chlorination Birth defects (neural Statistically significant associations were seen
al. 1999. study in Norway, (yes/no) and water tube defects, major between urinary tract defects and chlorination
1993–1995. color (high/low) at cardiac, respiratory, and high water color (high content of organic
mothers’ residences urinary, oral cleft). compounds). No associations were detected
at time of birth. for other outcomes or all birth defects com-
bined. A non-statistically significant, overall ex-
cess risk of birth defects was seen within mu-
nicipalities with chlorination and high water
color compared to municipalities with no
chlorination and low color.
Gallagher et Retrospective cohort Estimated THM levels in Low birth weight, term Weak, non-statistically significant association
al. 1998. study of newborns in drinking water during low birthweight, and with low birth weight and TTHM exposure dur-
Colorado, 1990–1993. third trimester of preg- preterm delivery. ing the third trimester. Large statistically sig-
nancy. nificant increase for term low birthweight at
highest THM exposure levels. No association
between preterm delivery and THM exposure.
Swan et al. Prospective study in Compared consumption Spontaneous abortion ... Pregnant women who drank cold tap water com-
1998. California, 1990–1991. of cold tap water to pared to those who consumed no cold tap
bottled water during water showed a significant finding for sponta-
early pregnancy. neous abortion at one of three sites.
wwhite on PROD1PC65 with RULES2

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00016 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 403


Author(s) Study type Exposure(s) studied Outcome(s) measured Findings

Waller et al. Prospective cohort in Estimated TTHM levels Spontaneous abortion ... Statistically significant increased risk between
1998 (and California, 1989–1991. during first trimester high intake of TTHMs and spontaneous abor-
Waller et of pregnancy via in- tion compared to low intake. BDCM statis-
al. 2001). gestion and show- tically associated with increased spontaneous
ering. abortion; other THMs not. Reanalysis of expo-
sure yielded less exposure misclassification
and relative risks similar in magnitude to ear-
lier study. An exposure-response relationship
was seen between spontaneous abortion and
ingestion exposure to TTHMs.
Kanitz et al. Cross-sectional study in Compared 3 types of Low birth weight, body Smaller body length and small cranial circum-
1996. Italy, 1988–1989. water treatment (chlo- length, cranial circum- ference showed statistical significant associa-
rine dioxide, sodium ference, preterm de- tion with maternal exposure to chlorinated
hypochlorite, and livery, and other ef- drinking water. Neonatal jaundice linked statis-
chlorine dioxide/so- fects. tically to prenatal exposure to drinking water
dium hypochlorite). treated with chlorine dioxide. Length of preg-
nancy, type of delivery, and birthweight
showed no association.
Bove et al. Large cohort cross-sec- Examined maternal ex- Low birth weight, fetal Weak, statistically significant increased risk
1995 (and tional study in New posure to TTHM and deaths, small for ges- found for higher TTHM levels with small for
Bove et Jersey, 1985–1988. various other contami- tational age, birth de- gestational age, neural tube defects, central
al. 1992a nants. fects (neural tube de- nervous system defects, oral cleft defects, and
& 1992b). fects, oral cleft, cen- major cardiac defects. Some association with
tral nervous system, higher TTHM exposure and low birth weight.
major cardiac). No effect seen for preterm birth, very low birth
weight, or fetal deaths.
Savitz et al. Population-based case- Examined TTHM con- Spontaneous abortion, There was a statistically significant increased
1995. control study: North centration at resi- preterm delivery, low miscarriage risk with high THM concentration,
Carolina, 1988–1991. dences and water birth weight. but THM intake (based on concentration times
consumption (during consumption level) was not related to preg-
first and third tri- nancy outcome. No associations were seen for
mesters). preterm delivery or low birth weight. Water
source was not related to pregnancy outcome
either, with the exception of a non-significant,
increased risk of spontaneous abortion for bot-
tled water users. There was a non-statistically
significant pattern of reduced risk with in-
creased consumption of water for all three out-
Aschengrau Case-control study in Source of water and 2 Neonatal death, still- There was a non-significant, increased associa-
et al. 1993. Massachusetts, 1977– types of water treat- birth, congenital tion between frequency of stillbirths and mater-
1980. ment (chlorination, anomalies. nal exposure to chlorinated versus
chloramination). chloraminated surface water. An increased risk
of urinary track and respiratory track defects
and chlorinated water was detected. Neonatal
death and other major malformations showed
no association. No increased risk seen for any
adverse pregnancy outcomes for surface
water versus ground and mixed water use.
Kramer et Population-based case- Examined chloroform, Low birth weight, pre- Statistically significant increased risk for intra-
al. 1992. control study in Iowa, DCBM, DBCM, and maturity, intrauterine uterine growth retardation effects from chloro-
1989–1990. bromoform levels and growth retardation. form exposure were observed. Non-significant
compared type of increased risks were observed for low birth
water source (surface, weight and chloroform and for intrauterine
shallow well, deep growth retardation and DCBM. No intrauterine
well). growth retardation or low birth weight effects
were seen for the other THMs, and no effects
on prematurity were observed for any of the
Shaw et al. Small case-control Estimated chlorinated Congenital cardiac Following reanalysis, no association between
1991 (and study: Santa Clara tap water consump- anomalies. cardiac anomalies and TTHM level were ob-
Shaw et County, CA, 1981– tion, mean maternal served.
al. 1990). 1983. TTHM level, show-
ering/bathing expo-
sure at residence dur-
ing first trimester.
wwhite on PROD1PC65 with RULES2

Aschengrau Case-control study in Source of water and ex- Spontaneous abortion ... A statistically significantly association was de-
et al. 1989. Massachusetts, 1976– posure to metals and tected between surface water source and fre-
1978. other contaminants. quency of spontaneous abortion.

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00017 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
404 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations


Author(s) Study type Exposure(s) studied Outcome(s) measured Findings

Hwang and Review and meta-anal- Compared DBP levels, Birth defects (respiratory The meta-analysis supports an association be-
Jakkola ysis of 5 studies. source of water, chlo- system, urinary sys- tween exposure to chlorination by-products
2003. rine residual, color tem, neural tube de- and the risk of any birth defect, particularly the
(high/low), and 2 fects, cardiac, oral risk of neural tube defects and urinary system
types of disinfection: cleft). defects.
chlorination and
Bove et al. Qualitative review of 14 Examined THM levels. Birth defects, small for Review found the studies of THMs and adverse
2002. studies. Compared drinking gestational age, low birth outcomes provide moderate evidence for
water source and type birth weight, preterm associations with small for gestational age,
of water treatment. delivery, spontaneous neural tube defects, and spontaneous abor-
abortion, fetal death. tions. Authors felt risks may have been under-
estimated and exposure-response relation-
ships distorted due to exposure
Graves et al. Review of toxicological Examined water con- Low birth weight, Weight of evidence suggested positive associa-
2001. and epidemiological sumption, duration of preterm delivery, tion with DBP exposure for growth retardation
studies using a weight exposure, THM levels, small for gestational such as small for gestational age or intra-
of evidence approach. HAA levels, and other age, intrauterine uterine growth retardation and urinary tract de-
contaminants. Com- growth retardation, fects. Review found no support for DBP expo-
pared source of specific birth defects, sure and low birth weight, preterm delivery,
water, water treat- neonatal death, de- some specific birth defects, and neonatal
ment, water color creased fertility, fetal death, and inconsistent findings for all birth de-
(high/low), etc. resorption, and other fects, all central nervous system defects, neu-
effects. ral tube defects, spontaneous abortion, and
Villanueva et Qualitative review of 14 Compared exposure to Spontaneous abortion, Review found positive associations between in-
al. 2001. reproductive and de- TTHM levels, muta- low birth weight, small creased spontaneous abortion, low birth
velopmental health ef- genic drinking water, for gestational age, weight, small for gestational age, and neural
fect studies. water consumption, neural tube defects, tube defects and drinking chlorinated water in
source water, types of other reproductive most studies, although not always with statis-
disinfection and developmental tical significance.
(chlorination and outcomes.
chloramination), and
residence times.
Nieuwenhuij- Qualitative review of nu- Examined levels of var- Low birth weight, The review supports some evidence of associa-
sen et al. merous toxicological ious DBPs, water con- preterm delivery, tion between THMs and low birth weight, but
2000. and epidemiological sumption, and dura- spontaneous abor- inconclusive. Review found no evidence of as-
studies. tion of exposure. tions, stillbirth, birth sociation between THMs and preterm delivery,
Compared water defects, etc. and that associations for other outcomes
color, water treatment, (spontaneous abortions, stillbirth, and birth de-
source of water, etc. fects) were weak but gaining weight.
Reif et al. Qualitative reviews of Compared source of Birth weight, low birth Weight of evidence suggested DBPs are repro-
2000. numerous epidemio- water supply and weight, intrauterine ductive toxicants in humans under appropriate
logical studies. methods of disinfec- growth retardation, exposure conditions. The review reports find-
tion. Estimated TTHM small for gestational ings between TTHMs and effects on fetal
levels. age, preterm deliver, growth, fetal viability, and congenital anoma-
somatic parameters, lies as inconsistent. Reviewers felt data are at
neonatal jaundice, the stage of hazard identification and did not
spontaneous abortion, suggest a dose-response pattern of increasing
stillbirth, develop- risk with increasing TTHM concentration.
mental anomalies.
WHO 2000 Qualitative reviews of Various exposures to Various reproductive Review found some support for an association
various studies in Fin- THMs. and developmental ef- between increased risks of neural tube defects
land, U.S., and Can- fects. and miscarriage and THM exposure. Other as-
ada. sociations have been observed, but the au-
thors believed insufficient data exist to assess
any of these associations.
Craun, ed. Qualitative review of 10 Examined THM levels Stillbirth, neonatal Associations between DBPs and various repro-
1998. studies, focus on Cali- and water consump- death, spontaneous ductive effects were seen in some epidemio-
fornia cohort study. tion, and compared abortion, low birth logical studies, but the authors felt these re-
source of water and weight, preterm deliv- sults do not provide convincing evidence for a
wwhite on PROD1PC65 with RULES2

water treatment (chlo- ery, intrauterine causal relationship between DBPs and repro-
rine, chloramines, growth retardation, ductive effects.
chlorine dioxide). neonatal jaundice,
birth defects.

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00018 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 405


Author(s) Study type Exposure(s) studied Outcome(s) measured Findings

Mills et al. Qualitative review of 22 Examined TTHM levels Various reproductive Review found studies suggest possible increases
1998. studies. and water consump- and developmental ef- in adverse reproductive and developmental ef-
tion. Compared fects. fects, such as increased spontaneous abortion
source of water and 2 rates, small for gestational age, and fetal
types of water treat- anomalies, but that insufficient evidence exists
ment (chlorination and to establish a causal relationship.
Reif et al. Review of 3 case-con- Examined THM levels at Birth defects (central Studies reviewed suggest that exposure to DBPs
1996. trol studies and 1 residences, dose con- nervous system, neu- may increase intrauterine growth retardation,
cross-sectional study. sumption, chloroform. ral tube defects, car- neural tube defects, major heart defects, and
Compared source of diac, oral cleft, res- oral cleft defects. Review found epidemiologic
waters and 2 types of piratory, urinary tract), evidence supporting associations between ex-
water treatment spontaneous abortion, posure to DBPs and adverse pregnancy out-
(chlorination and low birth weight, comes to be sparse and to provide an inad-
chloramination). growth retardation, equate basis to identify DBPs as a reproduc-
preterm delivery, tive or developmental hazard.
intrauterine growth re-
tardation, stillbirth,
neonatal death.
Swan et al. Qualitative review of 5 Compared maternal Spontaneous abortion ... Four of the studies reviewed suggest that
1992. studies in Santa Clara consumption of resi- women drinking bottled water during the first
County, CA (Deane et dence tap water to trimester of pregnancy may have reduced risk
al. 1992, Wrensch et bottled water. of spontaneous abortion relative to drinking
al. 1992, Hertz- tap water. No association seen in the fifth
Picciotto et al. 1992, study. Review concluded that if findings are
Windham et al. 1992, causal and not due to chance or bias, data
Fenster et al. 1992). suggest a 10–50% increase in spontaneous
abortion risk for pregnant women drinking tap
water over bottled water.

ii. Toxicology. To date, the majority of endpoints. The authors identified a possible association between adverse
reproductive and developmental NOAEL and LOAEL of 50 ppm and 150 reproductive and developmental health
toxicology studies have been short term ppm, respectively, based on delayed effects and exposure to chlorinated
and higher dose. Many of these studies sexual maturation for BDCM and a surface water.
are summarized in a review by Tyl NOAEL and LOAEL of 50 ppm and 250 c. Conclusions. EPA’s weight of
(2000). A summary of this review and of ppm based on abnormal evidence evaluation of the best available
additional studies is provided in the spermatogenesis for DBAA. The authors science on carcinogenicity and
proposed Stage 2 DBPR (USEPA 2003a). concluded that similar effects in reproductive and developmental effects,
Individual DBP supporting documents humans would only be seen at levels in conjunction with the widespread
evaluate and assess additional studies as many orders of magnitude higher than exposure to DBPs, supports the
well (USEPA 2000b; USEPA 2000c; that of current drinking water levels. As incremental regulatory changes in
USEPA 2001a; USEPA 2001b; USEPA discussed in more detail in the today’s rule that target lowering DBPs
2003b; USEPA 2005b; USEPA 2005c; proposal, EPA believes that because of and providing equitable public health
USEPA 2005d; USEPA 2005e; USEPA key methodological differences protection.
2005k). A number of recent studies have indicated as being important in other EPA believes that the cancer
been published that include in vivo and studies (Bielmeier et al. 2001; Bielmeier epidemiology and toxicology literature
in vitro assays to address mechanism of et al. 2004; Kaydos et al. 2004; provide important information that
action. Overall, reproductive and Klinefelter et al. 2001; Klinefelter et al. contributes to the weight of evidence for
developmental toxicology studies 2004), definitive conclusions regarding potential health risks from exposure to
indicate a possible reproductive/ BDCM and DBAA cannot be drawn. chlorinated drinking water. At this time,
developmental health hazard although Other multi-generation research the cancer epidemiology studies support
they are preliminary in nature for the underway includes a study on BCAA, a potential association between
majority of DBPs, and the dose-response but this research is not yet published. exposure to chlorinated drinking water
characteristics of most DBPs have not Biological plausibility for the effects and cancer, but evidence is insufficient
been quantified. Some of the observed in reproductive and to establish a causal relationship. The
reproductive effects of DCAA were developmental epidemiological studies epidemiological evidence for an
quantified as part of the RfD has been demonstrated through various association between DBP exposure and
development process, and impacts of toxicological studies on some individual colon and rectal cancers is not as
DCAA on testicular structure are one of DBPs (e.g., Bielmeier et al. 2001; consistent as it is for bladder cancer,
the critical effects in the study that is Bielmeier et al. 2004; Narotsky et al. although similarity of effects reported in
the basis of the RfD (USEPA 2003b). 1992; Chen et al. 2003; Chen et al. animal toxicity and human
wwhite on PROD1PC65 with RULES2

A few long term, lower dose studies 2004). Some of these studies were epidemiology studies strengthens the
have been completed. Christian et al. conducted at high doses, but similarity evidence for an association with colon
(2002a and 2002b) looked for an of effects observed between toxicology and rectal cancers. EPA believes that the
association between BDCM and DBAA studies and epidemiology studies overall cancer epidemiology and
and reproductive and developmental strengthens the weight of evidence for a toxicology data support the decision to

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00019 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
406 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations

pursue additional DBP control measures the ICR was supplemented by a survey addressing high DBP concentrations that
as reflected in the Stage 2 DBPR. conducted by the National Rural Water occur at particular locations or in single
Based on the weight of evidence Association, data provided by various samples within systems in compliance.
evaluation of the reproductive and States, the Water Utility Database
developmental epidemiology data, EPA 2. Treatment
(which contains data collected by the
concludes that a causal link between American Water Works Association), The analysis of the new treatment
adverse reproductive or developmental and ICR Supplemental Surveys for small study data confirmed that certain
health effects and exposure to and medium water systems. technologies are effective at reducing
chlorinated drinking water or DBPs has After analyzing the DBP occurrence DBP concentrations. Bench- and pilot-
not been established, but that there is a data, EPA and the Advisory Committee scale studies for granular activated
potential association. Despite reached three significant conclusions carbon (GAC) and membrane
inconsistent findings across studies, that in part led the Advisory Committee technologies required by the
some recent studies continue to suggest to recommend further control of DBPs Information Collection Rule provided
associations between DBP exposure and in public water systems. First, the data information on the effectiveness of the
various adverse reproductive and from the Information Collection Rule two technologies. Other studies found
developmental effects. In addition, data showed that the RAA compliance UV light to be highly effective for
from a number of toxicology studies, calculation under the Stage 1 DBPR inactivating Cryptosporidium and
although the majority of them were allows elevated TTHM or HAA5 levels Giardia at low doses without promoting
conducted using high doses, to regularly occur at some locations in the formation of DBPs (Malley et al.
demonstrate biological plausibility for the distribution system while the overall 1996; Zheng et al. 1999). This new
some of the effects observed in average of TTHM or HAA5 levels at all treatment information adds to the
epidemiology studies. EPA concludes DBP monitoring locations is below the treatment options available to utilities
that no dose-response relationship or MCLs of the Stage 1 DBPR. Customers for controlling DBPs beyond the
causal link has been established served at those sampling locations with requirements of the Stage 1 DBPR.
between exposure to chlorinated DBP levels that are regularly above
drinking water or disinfection 0.080 mg/L TTHM and 0.060 mg/L E. Conclusions for Regulatory Action
byproducts and adverse developmental HAA5 experience higher exposure After extensive analysis of available
or reproductive health effects. EPA’s compared to customers served at data and rule options considered by the
evaluation of the best available studies, locations where these levels are Advisory Committee and review of
particularly epidemiology studies is that consistently met. public comments on the proposed Stage
they do not support a conclusion at this Second, the new data demonstrated 2 DBPR (USEPA, 2003a), EPA is
time as to whether exposure to that DBP levels in single samples can be finalizing a Stage 2 DBPR control
chlorinated drinking water or substantially above 0.080 mg/L TTHM strategy consistent with the key
disinfection byproducts causes adverse and 0.060 mg/L HAA5. Some customers elements of the Agreement in Principle
developmental and reproductive health receive drinking water with signed in September 2000 by the
effects, but do provide an indication of concentrations of TTHM and HAA5 up participants in the Stage 2 M–DBP
a potential health concern that warrants to 75% above 0.080 mg/L and 0.060 mg/ Advisory Committee. EPA believes that
incremental regulatory action beyond L, respectively, even when their water exposure to chlorinated drinking water
the Stage 1 DBPR. system is in compliance with the Stage may be associated with cancer,
1 DBPR. Some studies support an reproductive, and developmental health
D. DBP Occurrence and DBP Control association between acute exposure to risks. EPA determined that the risk-
New information on the occurrence of DBPs and potential adverse targeting measures recommended in the
DBPs in distribution systems raises reproductive and developmental health Agreement in Principle will require
issues about the protection provided by effects (see Section III.C for more detail). only those systems with the greatest risk
the Stage 1 DBPR. This section presents Third, the data from the Information to make treatment and operational
new occurrence and treatment Collection Rule revealed that the highest
changes and will maintain simultaneous
information used to identify key issues TTHM and HAA5 levels can occur at
protection from potential health
and to support the development of the any monitoring site in the distribution
concerns from DBPs and microbial
Stage 2 DBPR. For a more detailed system. In fact, the highest
contaminants. EPA has carefully
discussion see the proposed Stage 2 concentrations did not occur at the
evaluated and expanded upon the
DBPR (USEPA 2003a). For additional maximum residence time locations in
recommendations of the Advisory
information on occurrence of regulated more than 50% of all ICR samples. The
Committee and public comments to
and nonregulated DBPs, see the fact that the locations with the highest
develop today’s rule. EPA also made
Occurrence Assessment for the Final DBP levels vary in different public
simplifications where possible to
Stage 2 Disinfectants and Disinfection water systems indicates that the Stage 1
minimize complications for public
Byproducts Rule (USEPA 2005f). DBPR monitoring may not accurately
water systems as they transition to
represent the high DBP concentrations
1. Occurrence compliance with the Stage 2 DBPR
that actually exist in distribution
EPA, along with the M-DBP Advisory while expanding public health
systems, and that additional monitoring
Committee, collected, developed, and protection. The requirements of the
is needed to identify distribution system
evaluated new information that became Stage 2 DBPR are described in detail in
locations with elevated DBP levels.
available after the Stage 1 DBPR was These data showed that efforts beyond Section IV of this preamble.
published. The Information Collection the Stage 1 DBPR are needed to provide IV. Explanation of Today’s Action
Rule (ICR) (USEPA 1996) provided new more equitable protection from DBP
wwhite on PROD1PC65 with RULES2

field data on DBP exposure for large exposure across the entire distribution
water systems and new study data on system. The incremental regulatory MCLGs are set at concentration levels
the effectiveness of several DBP control changes made under the Stage 2 DBPR at which no known or anticipated
technologies. The unprecedented meet this need by reevaluating the adverse health effects occur, allowing
amount of information collected under locations of DBP monitoring sites and for an adequate margin of safety.

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00020 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 407

Establishment of an MCLG for each 1. Chloroform MCLG of 0.01 mg/kg/day and an adult tap
specific contaminant is based on the a. Today’s rule. The final MCLG for water consumption of 2 L per day for a
available evidence of carcinogenicity or chloroform is 0.07 mg/L. The MCLG was 70 kg adult. A relative source
noncancer adverse health effects from calculated using toxicological evidence contribution (RSC) of 20% was used in
drinking water exposure using EPA’s that the carcinogenic effects of accordance with Office of Water’s
guidelines for risk assessment. MCLGs chloroform are due to sustained tissue current approach for deriving RSC
are developed to ensure they are toxicity. EPA is not changing the other through consideration of data that
protective of the entire population. THM MCLGs finalized in the Stage 1 indicate that other routes and sources of
Today’s rule provides MCLGs for DBPR. exposure may potentially contribute
b. Background and analysis. The substantially to the overall exposure to
chloroform and two haloacetic acids,
MCLG for chloroform is unchanged chloroform. See the proposed Stage 2
monochloroacetic acid (MCAA) and
from the proposal. The MCLG is DBPR (USEPA 2003a) for a detailed
trichloroacetic acid (TCAA).
calculated using a reference dose (RfD) discussion of the chloroform MCLG.

(0.01 mg/kg /day)(70 kg)(0.2)

MCLG for Chloroform = = 0.07 mg/L (rounded)
2 L/day

Based on an analysis of the available carcinogenic to humans under MCLGs for TCAA and MCAA of 0.02
scientific data on chloroform, EPA conditions that do not cause mg/L and 0.03 mg/L, respectively, in the
believes that the chloroform dose- cytotoxicity and cell regeneration Stage 2 proposal (USEPA 2003a). The
response is nonlinear and that (USEPA 2001a). Therefore, the dose- proposed TCAA MCLG and its
chloroform is likely to be carcinogenic response is nonlinear, and the MCLG is supporting analysis is being finalized
only under high exposure conditions set at 0.07 mg/L. This conclusion has unchanged in today’s final rule. The
(USEPA 2001a). This assessment is been reviewed by the SAB (USEPA MCLG calculation for MCAA is revised
supported by the principles of the 1999 2000d), who agree that nonlinear in this final rule, based on a new
EPA Proposed Guidelines for approach is most appropriate for the reference dose, as discussed later. See
Carcinogen Risk Assessment (USEPA risk assessment of chloroform; it also the proposed Stage 2 DBPR (USEPA
1999a) and reconfirmed by the 2005 remains consistent with the principles 2003a) for a detailed discussion of the
final Cancer Guidelines (USEPA 2005i). of the 1999 EPA Proposed Guidelines calculation of the MCLGs.
The science in support of a nonlinear for Carcinogenic Risk Assessment
approach for estimating the TCAA. The MCLG for TCAA was
(USEPA 1999a) and the final Cancer calculated based on the RfD of 0.03 mg/
carcinogenicity of chloroform was Guidelines ( USEPA 2005i), which
affirmed by the Chloroform Risk kg/day using a 70 kg adult body weight,
allow for nonlinear extrapolation.
Assessment Review Subcommittee of a 2 L/day drinking water intake, and a
EPA also received some comments
the EPA SAB Executive Committee requesting a combined MCLG for THMs relative source contribution of 20%. An
(USEPA 2000d). Since the nonzero or HAAs. This is not appropriate additional tenfold risk management
MCLG is based on a mode of action because these different chemicals have factor has been applied to account for
consideration specific to chloroform, it different health effects. the possible carcinogenicity of TCAA.
does not affect the MCLGs of other This approach is consistent with EPA
trihalomethanes. 2. HAA MCLGs: TCAA and MCAA policy. TCAA induces liver tumors in
c. Summary of major comments. EPA a. Today’s rule. Today’s rule finalizes mice (Ferreira-Gonzalez et al. 1995;
received many comments in support of the proposed Stage 2 MCLG for TCAA Pereira 1996; Pereira and Phelps 1996;
the proposed MCLG calculation for of 0.02 mg/L (USEPA 2003a) and sets an Tao et al. 1996; Latendresse and Pereira
chloroform, although some commenters MCLG for MCAA of 0.07 mg/L. EPA is 1997; Pereira et al. 1997) but not in rats
disagreed with a non-zero MCLG. not changing the other HAA MCLGs (DeAngelo et al. 1997). Much of the
At this time, based on an analysis of finalized in the Stage 1 DBPR (USEPA recent data on the carcinogenicity of
all the available scientific data on 1998a). TCAA have focused on examining the
chloroform, EPA concludes that b. Background and analysis. The Stage carcinogenic mode(s) of action.
chloroform is likely to be carcinogenic 1 DBPR included an MCLG for TCAA of However, at this time, neither the
to humans only under high exposure 0.03 mg/L and did not include an MCLG bioassay nor the mechanistic data are
conditions that lead to cytotoxicity and for MCAA (USEPA 1998a). Based on sufficient to support the development of
regenerative hyperplasia and that toxicological data published after the a slope factor from which to quantify
chloroform is not likely to be Stage 1 DBPR, EPA proposed new the cancer risk.

(0.03 mg/kg/day)(70 kg)(0.2)

MCLG for TCAA = = 0.02 mg/L (rounded)
(2 L/day)(10)

The chronic bioassay for TCAA by and NOAEL were determined. The data NOAEL of 32.5 mg/kg/day for liver
wwhite on PROD1PC65 with RULES2

DeAngelo et al. (1997) was selected as are consistent with the findings in both histopathological changes in rats
the critical study for the development of the Pereira (1996) chronic drinking (DeAngelo et al. 1997). A composite
the RfD. In this chronic drinking water water study and the Mather et al. (1990) uncertainty factor of 1000 was applied
study, a dose-response was noted for subchronic drinking water study. The in the RfD determination. A default

several endpoints and both a LOAEL RfD of 0.03 mg/kg/day is based on the uncertainty factor of 10 was applied to

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00021 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
408 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations

the RfD to account for extrapolation suggesting some presence in the change as a NOAEL. Increased spleen
from an animal study because data to atmosphere (Reimann et al. 1996); weights in the absence of
quantify rat-to-human differences in however, due to the low volatility (0.5— histopathological effects are not
toxicokinetics or toxicodynamics are not 0.7 mm Hg at 25 °C) of TCAA, exposure necessarily adverse. In addition, spleen
available. The default uncertainty factor from ambient air is expected to be weights were decreased, rather than
of 10 was used to account for human minimal. Dermal exposure to increased in the mid- and high-dose
variability in the absence of data on disinfected water is also unlikely to be groups in the DeAngelo et al. (1997)
differences in human susceptibility. significant. A study by Xu et al. (2002) study and were accompanied by a
Although subchronic and chronic reports that dermal exposure from significant decrease in body weight,
studies of TCAA have been reported for bathing and showering is only 0.01% of decreased relative and absolute liver
multiple species, many studies have that from oral exposure. In addition, the weights, decreased absolute kidney
focused on liver lesions and a full solvents trichloroethylene, weight, and an increase in relative testes
evaluation of a wide range of potential tetrachlorethylene, 1,1,1-trichloroethane weight. Accordingly, the mid-dose in
target organs has not been conducted in (often found in ambient air and drinking this same study (26.1 mg/kg/day) has
two different species. In addition, there water), and the disinfection byproduct been categorized as the LOAEL with the
has been no multi-generation study of chloral hydrate all contribute to the lower 3.5 mg/kg/day dose as a NOAEL.
reproductive toxicity and the data from body’s TCAA load since each of these Based on a NOAEL of 3.5 mg/kg/day
teratology studies in rats provide compounds is metabolized to TCAA (DeAngelo et al. 1997), the revised RfD
LOAEL values but no NOAEL for (ATSDR 2004; ATSDR 1997a; ATSDR was calculated as shown below, with a
developmental toxicity. Thus, an 1997b; USEPA 2000e). Due to the composite uncertainty factor of 300.
additional uncertainty factor of 10 was limitations primarily in the dietary data EPA used a default uncertainty factor of
used to account for database and a clear indication of exposure from 10 to account for extrapolation from an
insufficiencies. other sources, EPA applied a relative animal study, since no data on rat-to-
The MCLG calculation also includes a source contribution of 20%. human differences in toxicokinetics or
relative source contribution (RSC) of MCAA. The MCLG for MCAA uses toxicodynamics were identified. A
20%. The RSC was derived consistent the following calculations: An RfD of default uncertainty factor of 10 was
with Office of Water’s current approach 0.01 mg/kg/day, a 70 kg adult used to account for human variability in
for deriving RSC. In addition to consuming 2 L/day of tap water, and a the absence of data on the variability in
disinfected water, foods are expected to relative source contribution of 20%. the toxicokinetics of MCAA in humans
contribute to daily exposure to TCAA The RfD included in the proposal was or in human susceptibility to MCAA.
(Raymer et al. 2001, 2004; Reimann et based on a chronic drinking water study An additional uncertainty factor of three
al. 1996). Some of the TCAA in foods in rats conducted by DeAngelo et al. was used to account for database
comes from cleaning and cooking foods (1997). In the assessment presented for insufficiencies. Although there is no
in chlorinated water. Additional TCAA the proposed rule, the LOAEL from this multi-generation reproduction study,
is found in some foods because of the study was identified as 3.5 mg/kg/day the available studies of reproductive
widespread use of chlorine as a based on increased absolute and relative and developmental processes suggest
sanitizing agent in the food industry spleen weight in the absence of that developmental toxicity is unlikely
(USFDA 1994). EPA was not able to histopathologic changes. After to be the most sensitive endpoint. This
identify any dietary surveys or duplicate reviewing comments and further led to the following calculation of the
diet studies of TCAA in the diet. TCAA analysis of the data, EPA concludes that Reference Dose (RfD) and MCLG for
also has been identified in rain water, it is more appropriate to identify this MCAA:

(3.5 mg /kg/day)
RfD = = 0.012 mg /kg/day rounded to 0.01 mg /kg/day

Where: exposed to MCA for 104 weeks in extrapolation, inter-individual

drinking water (DeAngelo et al. variability in humans, and
3.5 mg/kg/day = NOAEL for decreased
1997). deficiencies in the database.
body weight plus decreased liver, 300 = composite uncertainty factor
kidney and spleen weights in rats chosen to account for inter species

(0.01 mg/kg/day)(70 kg)(0.2)

MCLG for MCAA = = 0.07 mg/L
2 L/day

The RSC for MCAA was selected did not increase the MCAA content of (USFDA 1994). As with TCAA,
using comparable data to that discussed foods to the same extent as was inhalation and dermal exposures are
for TCAA. MCAA, like TCAA, has been observed for TCAA (Raymer et al. 2004). unlikely to be significant. Dermal
found in foods and is taken up by foods MCAA was found to be completely exposure from bathing and showering
wwhite on PROD1PC65 with RULES2

during cooking (15% in chicken to 62% stable in water boiled for 60 minutes was estimated to contribute only 0.03%
in pinto beans) and cleaning (2.5% for and is likely to be found in the diet due of that from oral exposure (Xu et al.
lettuce) with water containing 500 ppb to the use of chlorinated water in food 2002). As with TCAA, due to the
MCAA (Reimann et al.1996; Raymer et preparation and the use of chlorine as limitations in dietary data and a clear

al. 2001, 2004). Rinsing of cooked foods a sanitizing agent by the food industry indication of exposure from other

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00022 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 409

sources, EPA applied a relative source 1. Today’s Rule pumping, treatment, equipment, and
contribution of 20%. As public water systems, consecutive personnel; assuring an adequate supply
c. Summary of major comments. EPA systems must provide water that meets during peak demand periods; acquiring
received few comments on MCAA and the MCLs for TTHM and HAA5 under emergency supplies; selling surplus
TCAA. The majority of comments about the Stage 2 DBPR, use specified supplies; and delivering a better product
the MCLGs for TCAA and MCAA were analytical methods, and carry out to consumers. EPA estimates that there
general MCLG questions, including RSC associated monitoring, reporting, are more than 10,000 consecutive
derivation. Some commenters recordkeeping, public notification, and systems nationally.
questioned why MCAA, TCAA, and other requirements. The following Consecutive systems face particular
chloroform were calculated using an discusses a series of definitions needed challenges in providing water that meets
RSC of 20%. In particular, some for addressing consecutive system regulatory standards for contaminants
commenters compared these requirements in today’s rule. Later that can increase in the distribution
calculations to that for DBCM in the sections of this preamble provide system. Examples of such contaminants
Stage 1 DBPR, which uses 80%. Each of further details on how rule requirements include coliforms, which can grow if
the MCLGs set for chloroform, TCAA, (e.g., schedule and monitoring) apply to favorable conditions exist, and some
and MCAA under this rule is calculated consecutive systems. DBPs, including THMs and HAAs,
using the best available science and EPA A consecutive system is a public which can increase when a disinfectant
Office of Water’s current approach for water system that receives some or all and DBP precursors continue to react in
deriving the RSC. EPA chose an RSC of of its finished water from one or more the distribution system.
20%, not 80%, because of clear wholesale systems. EPA included requirements
indications of exposure from other Finished water is water that has been
specifically for consecutive systems
sources; data limitations preclude the introduced into the distribution system
because States have taken widely
derivation of a specific RSC. of a public water system and is intended
varying approaches to regulating DBPs
for distribution and consumption
The RSC for DBCM was 80% in the in consecutive systems in previous
without further treatment, except as
Stage 1 DBPR. The DBCM MCLG is not rules. For example, some States have
necessary to maintain water quality in
part of today’s rulemaking. Any possible not regulated DBP levels in consecutive
the distribution system (e.g., booster
future revision to the DBCM MCLG as systems that deliver disinfected water
disinfection, addition of corrosion
a result of an RSC change would not but do not add a disinfectant. Other
control chemicals).
affect the MCL for TTHM finalized in States have determined compliance
A wholesale system is a public water
today’s rule. with DBP standards based on the
system that treats source water as
combined distribution system that
In response to comments received on necessary to produce finished water and
includes both the wholesaler and
the RfD for MCAA, EPA has reviewed then delivers finished water to another
consecutive systems. In this case, sites
the critical study regarding the public water system. Delivery may be
in consecutive systems are treated as
appropriateness of an increase in spleen through a direct connection or through
monitoring sites within the combined
weight in the absence of histopathology the distribution system of one or more
distribution system. Neither of these
as a LOAEL. EPA has determined that consecutive systems.
The combined distribution system is approaches provide the same level of
the dose associated with this endpoint public health protection as non-
is more appropriately categorized as a defined as the interconnected
distribution system consisting of the consecutive systems receive under the
NOAEL rather than a LOAEL and has Stage 1 DBPR. Once fully implemented,
revised the RfD and MCLG for MCAA. distribution systems of wholesale
systems and of the consecutive systems today’s rule will ensure similar
B. Consecutive Systems that receive finished water from those protection for consumers in consecutive
wholesale system(s). systems.
Today’s rule includes provisions for EPA is allowing States some In developing its recommendations,
consecutive systems, which are public flexibility in defining what systems are the Stage 2 M-DBP Advisory Committee
water systems that receive some or all a part of a combined distribution recognized two principles related to
of their finished water from another system. This provision determines consecutive systems: (1) consumers in
water system (a wholesale system). effective dates for requirements in consecutive systems should be just as
Consecutive systems face particular today’s rule; see Section IV.E well protected as customers of all
challenges in providing water that meets (Compliance Schedules) for further systems, and (2) monitoring provisions
regulatory standards for DBPs and other discussion. EPA has consulted with should be tailored to meet the first
contaminants whose concentration can States and deferred to their expertise principle. Accordingly, the Advisory
increase in the distribution system. regarding the nature of the connection Committee recommended that all
Moreover, previous regulation of DBP in making combined distribution system wholesale and consecutive systems
levels in consecutive systems varies determinations. In the absence of input comply with provisions of the Stage 2
widely among States. In consideration from the State, EPA will determine that DBPR on the same schedule required of
of these factors, EPA is finalizing combined distribution systems include the wholesale or consecutive system
monitoring, compliance schedule, and all interconnected systems for the serving the largest population in the
other requirements specifically for purpose of determining compliance combined distribution system. In
consecutive systems. These schedules for implementation of this addition, the Advisory Committee
requirements are intended to facilitate rule. recommended that EPA solicit
compliance by consecutive systems comments on issues related to
2. Background and Analysis
wwhite on PROD1PC65 with RULES2

with MCLs for TTHM and HAA5 under consecutive systems that the Advisory
the Stage 2 DBPR and help to ensure The practice of public water systems Committee had not fully explored
that consumers in consecutive systems buying and selling water to each other (USEPA 2000a). EPA agreed with these
receive equivalent public health has been commonplace for many years. recommendations and they are reflected
protection. Reasons include saving money on in today’s rule.

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00023 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
410 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations

3. Summary of Major Comments systems have only a marginal under the Initial Distribution System
Commenters generally supported the association (such as an infrequently Evaluation (IDSE) or using existing
proposed definitions. However, used emergency connection) with other Stage 1 DBPR compliance monitoring
commenters did express some concerns, systems in the combined distribution locations (as discussed in Section IV.F).
especially with including a time period system. To prepare for the IDSE and EPA has dropped the proposed phased
of water delivery that defined whether subsequent Stage 2 implementation, approach for LRAA implementation
a system was a consecutive system EPA has worked with States in (Stage 2A and Stage 2B) by removing
(proposed to trigger plant-based identifying all systems that are part of Stage 2A and redesignating Stage 2B as
monitoring requirements) or wholesale each combined distribution system. Stage 2.
system (proposed to allow Finally, several commenters requested
that the wholesale system definition Details of monitoring requirements
determination that a combined and compliance schedules are discussed
distribution system existed). EPA has replace ‘‘public water system’’ with
‘‘water system’’ so that wholesale in preamble Sections IV.G and IV.E,
dropped this requirement from the final respectively, and may be found in
rule; population-based monitoring systems serving fewer than 25 people
would not be considered public water subpart V of today’s rule.
requirements in the final rule do not
need to define how long a plant must systems. EPA did not change the 2. Background and Analysis
operate in order to be considered a definition in today’s rule; EPA considers
plant, and EPA has provided some any water system to be a public water The MCLs for TTHM and HAA5 are
flexibility for States to determine which system (PWS) if it serves 25 or more the same as those proposed, 0.080 mg/
systems comprise a combined people either directly (retail) or L TTHM and 0.060 mg/L HAA5 as an
distribution system (without presenting indirectly (by providing finished water LRAA. See the proposed rule (68 FR
a time criterion). to a consecutive system) or through a 49584, August 18, 2003) (USEPA 2003a)
Other commenters expressed concern combination of retail and consecutive for a more detailed discussion of the
that the proposed definition of system customers. If a PWS receives analysis supporting the MCLs. The
consecutive system was inconsistent water from an unregulated entity, that primary objective of the LRAA is to
with use of the term prior to the PWS must meet all compliance reduce exposure to high DBP levels. For
rulemaking. EPA acknowledges that the requirements (including monitoring and an LRAA, an annual average must be
Agency has not previously formally treatment techniques) that any other computed at each monitoring location.
defined the term, but believes that the public water system that uses source The RAA compliance basis of the 1979
definition in today’s rule best considers water of unknown quality must meet. TTHM rule and the Stage 1 DBPR allows
all commenters’ concerns, while also C. LRAA MCLs for TTHM and HAA5 a system-wide annual average under
providing for accountability and public which high DBP concentrations in one
health protection in as simple a manner 1. Today’s Rule or more locations are averaged with, and
as is possible given the many This rule requires the use of dampened by, lower concentrations
consecutive system scenarios that locational running annual averages elsewhere in the distribution system.
currently exist. (LRAAs) to determine compliance with Figure IV.C–1 illustrates the difference
Several States requested flexibility to the Stage 2 MCLs of 0.080 mg/L TTHM in calculating compliance with the
determine which systems comprised a and 0.060 mg/L HAA5. All systems, MCLs for TTHM between a Stage 1
combined distribution system under including consecutive systems, must DBPR RAA, and the Stage 2 DBPR
this rule; EPA has included that comply with the MCLs for TTHM and LRAA.
flexibility for situations in which HAA5 using sampling sites identified BILLING CODE 6560–50–P
wwhite on PROD1PC65 with RULES2

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00024 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 411


the expected impact on the water requirements. In the process of
EPA and the Stage 2 M–DBP Advisory industry and its customers. Strategies evaluating alternatives, EPA and the
Committee considered an array of considered included across the board Advisory Committee reviewed vast
wwhite on PROD1PC65 with RULES2

alternative MCL strategies. The requirements, such as significantly quantities of data and many analyses
Advisory Committee discussions decreasing the MCLs (e.g., 40/30) or that addressed health effects, DBP
primarily focused on the relative single hit MCLs (e.g., all samples must occurrence, predicted reductions in DBP
magnitude of exposure reduction versus be below 80/60); and risk targeting levels, predicted technology changes,

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00025 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
412 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations

and capital, annual, and household systems are able to comply with an RAA level suggested by the consecutive
costs. The Advisory Committee MCL even if they have a plant with a systems, and the time and money it
recommended and EPA proposed the poor quality water source (that thus could take to work out differences.
risk targeting approach of 80/60 as an produces high concentrations of DBPs) Although setting up a contract is a
LRAA preceded by an IDSE. Today’s because they have another plant that has prudent business action, commenters
rule finalizes these requirements. a better quality water source (and thus noted that small consecutive water
EPA has chosen compliance based on lower concentrations of DBPs). systems have few resources to sue for
an LRAA due to concerns about levels Individuals served by the plant with the damages should the wholesaler provide
of DBPs above the MCL in some poor quality source will usually have water exceeding the MCL.
portions of the distribution system. The higher DBP exposure than individuals The purpose of DBPRs is to protect
LRAA standard will eliminate system- served by the other plant. public health from exposure to high
wide averaging of monitoring results In part, both the TTHM and HAA5 DBP levels. Not requiring violations
from different monitoring locations. The classes are regulated because they occur when distributed water exceeds MCLs
individuals served in areas of the at high levels and represent chlorination undermines the intent of the rule. While
distribution system with above average byproducts that are produced from EPA recognizes consecutive systems do
DBP occurrence levels masked by source waters with a wide range of not have full control over the water they
averaging under an RAA are not water quality. The combination of receive, agreements between wholesale
receiving the same level of health TTHM and HAA5 represent a wide and consecutive systems may specify
protection. Although an LRAA standard variety of compounds resulting from water quality and actions required of the
still allows averaging at a single location bromine substitution and chlorine wholesaler if those water quality
over an annual period, EPA concluded substitution reactions (e.g., bromoform standards are not met.
that changing the basis of compliance has three bromines, TCAA has three
from an RAA to an LRAA will result in chlorines, BDCM has one bromine and Finally, commenters recommended
decreased exposure to higher DBP levels two chlorines). EPA believes that the that the Stage 2A provisions in the
(see Section VI for predictions of DBP TTHM and HAA5 classes serve as an proposed rule be removed. These
reductions under the LRAA MCLs). This indicator for unidentified and provisions (compliance with locational
conclusion is based on three unregulated DBPs. EPA believes that running annual average MCLs of 0.120
considerations: controlling the occurrence levels of mg/L for TTHM and 0.100 mg/L for
(1) There is considerable evidence TTHM and HAA5 will help control the HAA5) required systems to comply with
that under the current RAA MCL overall levels of chlorination DBPs. the Stage 1 MCLs (as running annual
compliance monitoring requirements, a averages) and the Stage 2A MCLs (as
small but significant proportion of 3. Summary of Major Comments LRAAs) concurrently until systems were
monitoring locations experience high Commenters supported the proposed, required to comply with Stage 2B MCLs.
DBP levels at least some of the time. Of risk-targeted MCL strategy over the Commenters noted that having two
systems that collected data under the alternative MCL strategies that were separate MCLs for an individual system
Information Collection Rule that met the considered by the Advisory Committee to comply with at the same time was
Stage 1 DBPR RAA MCLs, 14 percent as the preferred regulatory strategy. confusing to the system and its
had TTHM single sample concentrations Commenters concurred with EPA’s customers. In addition, State resources
greater than the Stage 1 MCL, and 21 analysis that such an approach will needed for compliance determinations
percent had HAA5 single sample reduce peak and average DBP levels. and data management for this short-term
concentrations above the MCL. Commenters supported the Stage 2 long- requirement would be resource-
Although most TTHM and HAA5 term MCLs of 0.080 mg/L TTHM and intensive. Finally, resources spent to
samples were below 100 µg/L, some 0.060 mg/L HAA5 as LRAAs. comply with Stage 2A would be better
ranged up to 140 µg/L and 130 µg/L, EPA received many comments on spent in complying with Stage 2B,
respectively. today’s MCLs specific to consecutive especially given that some of the
(2) In some situations, the populations systems. While commenters supported changes for Stage 2A compliance might
served by certain portions of the consecutive system compliance with the not provide any benefit for Stage 2B.
distribution system consistently receive Stage 2 DBPR in order to provide Since EPA agrees with commenters’
water that exceeds 0.080 mg/L for comparable levels of public health concerns, the Stage 2A requirements
TTHM or 0.060 mg/L for HAA5 (both as protection, they noted that it would be have been removed from the final rule.
LRAAs) even though the system is in difficult for many consecutive systems
D. BAT for TTHM and HAA5
compliance with Stage 1 MCLs). Of to meet Stage 2 requirements because
Information Collection Rule systems they have not had to meet the full scope 1. Today’s Rule
meeting the Stage 1 DBPR MCLs as of DBP requirements under previous
RAAs, five percent had monitoring rules. EPA has developed a training and Today, EPA is identifying the best
locations that exceeded 0.080 mg/L outreach program to assist these systems available technology (BAT) for the
TTHM and three percent exceeded and encourages States, wholesale TTHM and HAA5 LRAA MCLs (0.080
0.060 mg/L HAA5 as an annual average systems, and professional associations mg/L and 0.060 mg/L respectively) for
(i.e., as LRAAs) by up to 25% to also provide assistance. systems that treat their own source
(calculated as indicated in Figure IV.C– Some commenters expressed concern water as one of the three following
1). Customers served at these locations about holding consecutive systems technologies:
consistently received water with TTHM responsible for water quality over which (1) GAC10 (granular activated carbon
and/or HAA5 concentrations higher they have no control. Several filter beds with an empty-bed contact
than the system-wide average and commenters were concerned about the time of 10 minutes based on average
wwhite on PROD1PC65 with RULES2

higher than the MCL. establishment of contracts between daily flow and a carbon reactivation
(3) Compliance based on an LRAA wholesale and consecutive systems, frequency of every 120 days)
will remove the opportunity for systems including concern about a strain on (2) GAC20 (granular activated carbon
to average out samples from high and their relationship, wholesale system filter beds with an empty-bed contact
low quality water sources. Some reluctance to commit to keep DBPs at a time of 20 minutes based on average

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00026 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 413

daily flow and a carbon reactivation conducted an Information Collection membrane process. Also, nanofiltration
frequency of every 240 days) Rule GAC treatment study, is an accepted technology for treatment
(3) Nanofiltration (NF) using a approximately 70 percent of the surface of high TOC ground waters in Florida
membrane with a molecular weight water plants studied could meet the and parts of the Southwest, areas of the
cutoff of 1000 Daltons or less. 0.080 mg/L TTHM and 0.060 mg/L country with elevated TOC levels in
EPA is specifying a different BAT for HAA5 MCLs, with a 20 percent safety ground waters.
consecutive systems than for systems factor (i.e., 0.064 mg/L and 0.048 mg/L, The second method that EPA used to
that treat their own source water to meet respectively) using GAC with 10 examine alternatives for BAT was the
the TTHM and HAA5 LRAA MCLs. The minutes of empty bed contact time and Surface Water Analytical Tool model
consecutive system BAT is a 120 day reactivation frequency, and 78 that was developed to compare
chloramination with management of percent of the plants could meet the alternative regulatory strategies as part
hydraulic flow and storage to minimize MCLs with a 20 percent safety factor of the Stage 1 and Stage 2 M–DBP
residence time in the distribution using GAC with 20 minutes of empty Advisory Committee deliberations. EPA
system for systems that serve at least bed contact time and a 240 day modeled a number of BAT options. In
10,000 people and management of reactivation frequency. Because the the model, GAC10 was defined as
hydraulic flow and storage to minimize treatment studies were conducted at granular activated carbon with an empty
residence time in the distribution plants with much poorer water quality bed contact time of 10 minutes and a
system for systems that serve fewer than than the national average, EPA believes reactivation or replacement interval of
10,000 people. that much higher percentages of plants 90 days or longer. GAC20 was defined
2. Background and Analysis nationwide could meet the MCLs with as granular activated carbon with an
the proposed GAC BATs. empty bed contact time of 20 minutes
The BATs are the same as was
proposed, except that consecutive Among plants using GAC, larger and a reactivation or replacement
systems serving fewer than 10,000 systems would likely realize an interval of 90 days or longer.
people do not have chloramination as economic benefit from on-site The compliance percentages
part of the consecutive system BAT. See reactivation, which could allow them to forecasted by the SWAT model are
the proposal (68 FR 49588, August 18, use smaller, 10-minute empty bed indicated in Table IV.D–1. EPA
2003) (USEPA 2003a) for more detail on contact time contactors with more estimates that more than 97 percent of
the analysis supporting these frequent reactivation (i.e., 120 days or large systems will be able to achieve the
requirements. The Safe Drinking Water less). Most small systems would not Stage 2 MCLs with the GAC BAT,
Act directs EPA to specify BAT for use find it economically advantageous to regardless of post-disinfection choice
in achieving compliance with the MCL. install on-site carbon reactivation (Seidel Memo, 2001). Because the
Systems unable to meet the MCL after facilities, and thus would opt for larger, source water quality (e.g., DBP
application of BAT can get a variance 20-minute empty bed contact time precursor levels) in medium and small
(see Section IV.K for a discussion of contactors, with less frequent carbon systems is expected to be comparable to
variances). Systems are not required to replacement (i.e., 240 days or less). or better than that for the large system
use BAT in order to comply with the The Information Collection Rule (USEPA 2005f), EPA believes it is
MCL. PWSs may use any State-approved treatment study results also conservative to assume that at least 90
technologies as long as they meet all demonstrated that nanofiltration was percent of medium and small systems
drinking water standards. the better DBP control technology for will be able to achieve the Stage 2 MCLs
EPA examined BAT options first by ground water sources with high TOC if they were to apply one of the
analyzing data from the Information concentrations (i.e., above proposed GAC BATs. EPA assumes that
Collection Rule treatment studies approximately 6 mg/L). The results of small systems may adopt GAC20 in a
designed to evaluate the ability of GAC the membrane treatment studies showed replacement mode (with replacement
and NF to remove DBP precursors. that all ground water plants could meet every 240 days) over GAC10 because it
Based on the treatment study results, the 0.080 mg/L TTHM and 0.060 mg/L may not be economically feasible for
GAC is effective for controlling DBP HAA5 MCLs, with a 20% safety factor some small systems to install and
formation for waters with influent TOC (i.e., 0.064 mg/L and 0.048 mg/L, operate an on-site GAC reactivation
concentrations below approximately 6 respectively) at the system average facility. Moreover, some small systems
mg/L (based on the Information distribution system residence time using may find nanofiltration cheaper than the
Collection Rule and NRWA data, over nanofiltration. Nanofiltration would be GAC20 in a replacement mode if their
90 percent of plants have average less expensive than GAC for high TOC specific geographic locations cause a
influent TOC levels below 6 mg/L ground waters, which generally require relatively high cost for routine GAC
(USEPA 2003c)). Of the plants that minimal pretreatment prior to the shipment.

Compliance with 0.080 mg/L TTHM and 0.060 Compliance with 0.064 mg/L TTHM and 0.048
mg/L HAA5 LRAAs mg/L HAA5 LRAAs (MCLs with 20% Safety fac-
Residual disinfectant
Technology Residual
All systems disinfectant All systems
Chlorine (per- Chloramine (percent) (percent)
cent) (percent) Chlorine (per- Chloramine
wwhite on PROD1PC65 with RULES2

cent) (percent)

Enhanced Coagulation (EC) .................... 73.5 76.9 74.8 57.2 65.4 60.4
EC (no pre-disinfection) ........................... 73.4 88.0 78.4 44.1 62.7 50.5
EC & GAC10 ............................................ 100 97.1 99.1 100 95.7 98.6

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00027 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
414 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations

Compliance with 0.080 mg/L TTHM and 0.060 Compliance with 0.064 mg/L TTHM and 0.048
mg/L HAA5 LRAAs mg/L HAA5 LRAAs (MCLs with 20% Safety fac-
Residual disinfectant
Technology Residual
All systems disinfectant All systems
Chlorine (per- Chloramine (percent) (percent)
cent) (percent) Chlorine (per- Chloramine
cent) (percent)

EC & GAC20 ............................................ 100 100 100 100 100 100

EC & All Chloramines .............................. NA 83.9 NA NA 73.6 NA
Note: Enhanced coagulation/softening is required under the Stage 1 DBPR for conventional plants.
Source: Seidel (2001).

The BAT requirements for large EPA believes that the BATs for contradicts the premise of the Stage 1
consecutive systems are the same as nonconsecutive systems are not DBPR that DBPs are best controlled
proposed, but the requirements have appropriate for consecutive systems through TOC removal and optimizing
changed for small consecutive systems. because their efficacy in controlling disinfection processes, the SDWA
EPA believes that the best compliance DBPs is based on precursor removal. requires EPA to identify a BAT for all
strategy for consecutive systems is to Consecutive systems face the unique systems required to meet an MCL. No
collaborate with wholesalers on the challenge of receiving waters in which commenter recommended an alternative
water quality they need. For consecutive DBPs are already present if the BAT. EPA still believes that precursor
systems that are having difficulty wholesale system has used a residual removal remains a highly effective
meeting the MCLs, EPA is specifying a disinfectant, which the BATs for non- strategy to reduce DBPs. Thus, EPA
BAT of chloramination with consecutive systems do not effectively encourages States to work with
management of hydraulic flow and remove. GAC is not cost-effective for wholesale systems and consecutive
storage to minimize residence time in removing DBPs. Nanofiltration is only systems to identify strategies to ensure
the distribution system for systems moderately effective at removing THMs compliance, especially those systems
serving at least 10,000 and management or HAAs if membranes with a very low with DBP levels close to the MCL.
of hydraulic flow and storage to molecular weight cutoff (and very high
minimize residence time in the E. Compliance Schedules
cost of operation are employed).
distribution system for systems serving Therefore, GAC and nanofiltration are 1. Today’s Rule
fewer than 10,000. EPA believes that not appropriate BATs for consecutive
small consecutive systems can use this This section specifies compliance
BAT to comply with the Stage 2 DBPR, dates for the IDSE and MCL compliance
but if they cannot, then they can apply 3. Summary of Major Comments requirements in today’s rule. As
to the State for a variance. Commenters concurred with EPA’s described elsewhere in Section IV of
Chloramination has been used for identification of BATs for non- this preamble, today’s rule requires
residual disinfection for many years to consecutive systems but expressed PWSs to carry out the following
minimize the formation of chlorination concern about the BAT for consecutive activities:
DBPs, including TTHM and HAA5 systems. Many commenters agreed that • Conduct initial distribution system
(USEPA 2003d). EPA estimates that over Stage 2 compliance for consecutive evaluations (IDSEs) on a required
50 percent of large subpart H systems systems would usually best be achieved schedule. Systems may comply by using
serving at least 10,000 use by improved treatment by the wholesale any of four approaches for which they
chloramination for Stage 1. The BAT system. However, they noted that the qualify (standard monitoring, system
provision to manage hydraulic flow and proposed BAT may not be practical for specific study, 40/30 certification, or
minimize residence time in the compliance if water delivered to the very small system waiver).
distribution system is to facilitate the consecutive system is at or near DBP • Determine Stage 2 monitoring
maintenance of the chloramine residual MCLs. In addition, chloramination locations based on the IDSE.
and minimize the likelihood for requires operator supervision and • Comply with Stage 2 MCLs on a
nitrification. EPA has not included adjustment and many consecutive required schedule.
chloramination for consecutive systems systems that buy water may be reluctant Compliance dates for these activities
as part of the BAT for systems serving to operate chemical feed systems. vary by PWS size. Table IV.E–1 and
fewer than 10,000 due to concerns about Therefore, EPA included chloramines as Figure IV.E–1 specify IDSE and Stage 2
their ability to properly control the part of the BAT in today’s rule only for compliance dates. Consecutive systems
process, given that many have no systems serving at least 10,000 because of any size must comply with the
treatment capability or expertise and the of the operator attention it requires and requirements of the Stage 2 DBPR on the
Agency’s concern about such systems concerns with safety and nitrification. same schedule as required for the largest
having operational difficulties such as While some commenters believed that system in the combined distribution
distribution system nitrification. having a BAT for consecutive systems system.
wwhite on PROD1PC65 with RULES2

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00028 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 415


Compliance dates by PWS size (retail population served) 1

Requirement CWSs and CWSs and CWSs and CWSs serving NTNCWSs serving
NTNCWSs serving NTNCWSs serving NTNCWSs serving <10,000 <10,000
at least 100,000 50,000–99,999 10,000–49,999

Submit IDSE monitoring plan OR October 1, 2006 ..... April 1, 2007 ........... October 1, 2007 ..... April 1, 2008 ....... Not applicable.
Submit IDSE system specific
study plan OR.
Submit 40/30 certification OR .....
Receive very small system waiv-
er from State.
Complete standard monitoring or September 30, 2008 March 31, 2009 ...... September 30, 2009 March 31, 2010 .. Not applicable.
system specific study.
Submit IDSE Report ................... January 1, 2009 ..... July 1, 2009 ............ January 1, 2010 ..... July 1, 2010 ....... Not applicable.
Begin subpart V (Stage 2) com- April 1, 2012 ........... October 1, 2012 ..... October 1, 2013 ..... October 1, 2013
pliance monitoring 2. (October 1,
2014 if Crypto-
sporidium mon-
itoring is re-
quired under
Subpart W)..
1 Wholesale and consecutive systems that are part of a combined distribution system must comply based on the schedule required of the larg-
est system in the combined distribution system.
2 States may grant up to an additional 2 years for systems making capital improvements.


wwhite on PROD1PC65 with RULES2

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00029 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
416 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations
wwhite on PROD1PC65 with RULES2



VerDate Aug<31>2005 21:06 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00030 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 417

2. Background and Analysis locations and schedules to the State or requiring IDSE plan review prior to
The compliance schedule in today’s primacy agency. Where required, PWSs conducting the IDSE.
final rule stems from the risk-targeted must provide the necessary level of • Provides additional time to develop
approach of the rule, wherein PWSs treatment to comply with the Stage 2 budgets and establish contracts with
conduct initial monitoring to determine MCLs within three years of the laboratories.
completion of State or primacy agency
locations and concentrations of high
review of the IDSE report, though States • Spreads out the workload for
DBPs. A primary objective of this technical assistance and guidance. The
may allow an additional two years for
schedule is to ensure that PWSs identify staggered schedule will allow States and
PWSs making capital improvements.
locations with high DBP concentrations EPA has modified the proposed EPA to provide more support to
and provide appropriate additional compliance schedule to stagger individual PWSs as needed.
treatment in a timely manner for high monitoring start dates for PWSs serving • Provides time for DBP analytical
risk areas, while not requiring low risk 10,000 to 99,999 people and to allow laboratories to build capacity as needed
systems to add additional treatment. more time for development and review to accommodate the sample analysis
The compliance schedule balances the of IDSE monitoring plans prior to the needs of PWSs and extends and
objective of early risk-targeted start of monitoring. The following smooths the demand for laboratory
monitoring with adequate time for discussion addresses these changes from services.
PWSs and the State or primacy agency the proposal.
to assure full implementation and • Maintains simultaneous rule
The proposed rule required all PWSs compliance with the LT2ESWTR as
compliance. EPA is establishing serving at least 10,000 people (plus
concurrent compliance schedules under recommended by the Stage 2 M-DBP
smaller systems that are part of a Advisory Committee and as mandated
the Stage 2 DBPR for all systems (both combined distribution system with a
wholesale systems and consecutive by the 1996 SDWA Amendments, which
PWS that serves at least 10,000 people) require that EPA ‘‘minimize the overall
systems) in a particular combined to complete IDSE monitoring and
distribution system because this will risk of adverse health effects by
submit IDSE reports (including balancing the risk from the contaminant
assure comparable risk-based targeting recommended Stage 2 compliance
information being available at the same and the risk from other contaminants
monitoring locations) two years after the concentrations of which may be
time for all PWSs that are part of a rule promulgation, followed by one year
combined distribution system and affected by the use of a treatment
for review of IDSE reports, after which
thereby allow for more cost-effective technique or process that would be
systems had three years to come into
compliance with TTHM and HAA5 employed to attain the maximum
compliance with Stage 2B MCLs.
MCLs. Under today’s final rule, PWSs contaminant level’’ (Sec.
SDWA section 1412(b)(10) states that serving at least 100,000 people (plus 1412(b)(5)(B)(i)).
a drinking water regulation shall take smaller systems that are part of the The Advisory Committee
effect 3 years from the promulgation combined distribution system) will meet recommended the Initial Distribution
date unless the Administrator the same Stage 2 compliance deadlines
System Evaluation, as discussed in
determines that an earlier date is as proposed. However, the timing of the
Section IV.F, and EPA is finalizing an
practicable. Today’s rule requires PWSs IDSE has been changed to allow for a
IDSE schedule generally consistent with
to begin monitoring prior to 3 years more even workload and a greater
the Advisory Committee timeframe
from the promulgation date. Based on opportunity for primacy agency
recommendation, but modified to
EPA’s assessment and recommendations involvement (e.g., through monitoring
stagger the schedule for systems serving
of the Advisory Committee, as described plan review and approval). The IDSE
more than 10,000 but less than 100,000,
in this section, EPA has determined that plan submission dates for PWSs serving
and to address public comments on the
these monitoring start dates are 50,000 to 99,999 people (plus smaller
IDSE requirements.
practicable and appropriate. systems that are part of the combined
Systems must submit their IDSE plans distribution system) will be 12 months For all systems, the IDSE schedule has
(monitoring plans for standard after the effective date; for PWSs serving been revised to allow systems to submit
monitoring, study plans for system 10,000 to 49,999 (plus smaller systems and States or primacy agencies to
specific studies) to the primacy agency that are part of the combined review (and revise, if necessary)
for review and approval. The State or distribution system), the IDSE plan systems’ recommendations for IDSE and
primacy agency will then have 12 submission dates will be 18 months Stage 2 monitoring locations, while still
months to review, and, as necessary, after the effective date. The Stage 2 allowing systems three years after
consult with the system. A number of compliance schedule for systems completion of the State or primacy
PWSs will then conduct one year of serving fewer than 10,000 people agency review of Stage 2 compliance
distribution system monitoring for remains the same as proposed. Stage 2 monitoring locations to make necessary
TTHM and HAA5 at locations other MCL compliance dates are modified treatment and operational changes to
than those currently used for Stage 1 accordingly. comply with Stage 2 MCLs.
DBPR compliance monitoring. At the This staggering of IDSE start dates for Figure IV.E–2 illustrates compliance
conclusion of this monitoring, these PWSs serving 10,000 to 99,999 people is schedules for examples of three
PWSs have three months to evaluate advantageous in several respects: combined distribution systems, with the
analysis and monitoring results and • Provides PWSs greater assurance schedule dictated by the retail
submit Stage 2 compliance monitoring that IDSEs are properly conducted by population served by the largest system.


wwhite on PROD1PC65 with RULES2

—Wholesale system (pop. 64,000) with three consecutive systems (pops. 21,000; 15,000; 5,000):
—IDSE monitoring plan due for all systems April 1, 2007 since wholesale system serves 50,000–99,999
—Stage 2 compliance beginning October 1, 2012 for all systems
—Wholesale system (pop. 4,000) with three consecutive systems (pops. 21,000; 5,000; 5,000):

VerDate Aug<31>2005 21:06 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00031 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
418 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations


—IDSE monitoring plan due for all systems October 1, 2007 since the largest system in combined distribution system serves 10,000–
—Stage 2 compliance beginning October 1, 2013 for all systems
—Wholesale system (pop. 4,000) with three consecutive systems (pops. 8,000; 5,000; 5,000):
—IDSE monitoring plan due for all systems April 1, 2008 since no individual system in combined distribution system exceeds 10,000 (even
though total population exceeds 10,000)
—Stage 2 compliance beginning October 1, 2013 if no Cryptosporidium monitoring under the LT2ESWTR is required or beginning October
1, 2014 if Cryptosporidium monitoring under the LT2ESWTR is required

This schedule requires wholesale the August 18, 2003 proposal. Major a larger system in their combined
systems and consecutive systems that issues raised by commenters include distribution system, do not begin
are part of a combined distribution providing more time for PWSs to monitoring until more than 36 months
system with at least one system with an prepare for monitoring, giving States or after the effective date.
earlier compliance deadline to conduct primacy agencies more time to oversee EPA believes that the final
their IDSE simultaneously so that the monitoring, and establishing consistent compliance schedule allows PWSs
wholesale system will be aware of schedules for consecutive PWSs. A sufficient time to develop IDSE plans
compliance challenges facing the summary of these comments and EPA’s with these compliance dates. The
consecutive systems and will be able to responses follows. schedule also allows 12 months for
implement treatment plant, capital, and Standard monitoring plan and system State or primacy agency review of IDSE
operational improvements as necessary specific study plan preparation. Many plans, which allows additional time for
to ensure compliance of both the commenters were concerned about the review and for coordination with
wholesale and consecutive systems. The proposed requirement to develop and systems and provides more time to
Advisory Committee and EPA both execute an IDSE monitoring plan address deficiencies in IDSE plans. This
recognized that DBPs, once formed, are without any primacy agency review. is especially important for smaller
difficult to remove and are generally PWSs specifically expressed concern PWSs, which are likely to need the most
best addressed by treatment plant about the financial commitment without assistance from States. By staggering
improvements, typically through prior State approval and noted that monitoring start dates, today’s rule also
precursor removal or use of alternative some PWSs would need more than the eases implementation by reducing the
disinfectants. For a wholesale system to time allowed under the proposed rule to number of PWSs that will submit plans
make the best decisions concerning the develop and implement an IDSE at any one time, when the most
treatment steps necessary to meet monitoring plan, especially without an assistance from regulatory agencies will
TTHM and HAA5 LRAAs under the opportunity for State or primacy agency be required.
Stage 2 DBPR, both in its own review and approval. Smaller PWSs In summary, today’s schedule has
distribution system and in the may require substantial time and been modified so that systems are
distribution systems of consecutive planning to budget for IDSE expenses, required to submit IDSE plans for
systems it serves, the wholesale system especially for systems that have not primacy agency review and approval
must know the DBP levels throughout previously complied with DBP MCLs. prior to conducting their IDSE. Systems
the combined distribution system. EPA recognizes these concerns and can consider that their plan has been
Without this information, the wholesale today’s final rule provides time for approved if they have not heard back
system may design treatment changes PWSs to submit IDSE plans (monitoring from the State by the end of the State
that allow the wholesale system to plans, study plans, or 40/30 review period. Systems are also required
achieve compliance, but leave the certifications) for State or primacy to conduct the approved monitoring and
consecutive system out of compliance. agency review and more time before submit their IDSE report (including the
In summary, the compliance schedule having to begin monitoring. system’s recommended Stage 2
for today’s rule maintains the earliest Specifically, PWSs serving 50,000 to compliance monitoring) for State or
compliance dates recommended by the 99,999 people and those serving 10,000 primacy agency review on a schedule
Advisory Committee for PWSs serving to 49,999 people must submit IDSE that allows for systems to still have a
at least 100,000 people (plus smaller plans about 12 months and 18 months minimum of full three years to comply
systems that are part of the combined after the effective date, respectively, and with Stage 2 following State or primacy
distribution system). These PWSs serve complete standard monitoring or a agency review of the system’s Stage 2
the majority of people. The schedule system specific study within two years recommended monitoring. As with the
also maintains the latest compliance after submitting their IDSE plan. This is review of plans, systems can consider
dates the Advisory Committee significantly more time than was that their IDSE report has been
recommended, which apply to PWSs specified under the proposal, where approved if they have not heard back
serving fewer than 10,000 people. EPA these systems would have had to from the State by the end of the State
has staggered compliance schedules for conduct their IDSE and submit their review period.
PWSs between these two size categories IDSE report 24 months after the effective State/primacy agency oversight. EPA
in order to facilitate implementation of date. PWSs serving at least 100,000 is preparing to support implementation
the rule. This staggered schedule is people must submit IDSE plans about of IDSE requirements that must be
consistent with the schedule required six months after the effective date and completed prior to States achieving
under the LT2ESWTR promulgated complete standard monitoring or a primacy. Several States have expressed
wwhite on PROD1PC65 with RULES2

elsewhere in today’s Federal Register. system specific study about 30 months concern about EPA providing guidance
after the effective date, which also and reviewing reports from systems that
3. Summary of Major Comments provides more time than was specified the State has permitted, inspected, and
EPA received significant public under the proposal. PWSs serving fewer worked with for a long time. These
comment on the compliance schedule in than 10,000 people, not associated with States believe that their familiarity with

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00032 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 419

the systems enables them to make the system (the interconnected distribution F. Initial Distribution System Evaluation
best decisions to implement the rule system consisting of the distribution (IDSE)
and protect public health and that the systems of wholesale systems and of the 1. Today’s Rule
rule requirement should be delayed consecutive systems that receive
until States receive primacy. finished water) on the same Stage 2 Today’s rule establishes requirements
Commenters were concerned that some compliance schedule. Extending the for systems to perform an Initial
States will not participate in early Stage 2 compliance dates would Distribution System Evaluation (IDSE).
implementation activities and indicated The IDSE is intended to identify sample
unnecessarily delay the public health
that States would prefer monitoring to locations for Stage 2 compliance
protection afforded by this rule.
begin 24 months after rule monitoring that represent distribution
Consecutive systems must be able to
promulgation. Commenters also noted system sites with high DBP
evaluate whether wholesale system concentrations. Systems will develop an
that States need sufficient time to changes are sufficient to ensure
become familiar with the rule, train IDSE plan, collect data on DBP levels
compliance and, if they are not, to make throughout their distribution system,
their staff, prepare primacy packages,
cost-effective changes to ensure evaluate these data to determine which
and train PWSs.
EPA agrees that State familiarity is an compliance where wholesale system sampling locations are most
important component of the review and efforts address some, but not all, of the representative of high DBP levels, and
approval process, looks forward to concerns with compliance. Public compile this information into a report
working closely with the State drinking health protection through compliance for submission to the State or primacy
water program representatives during with Stage 2 MCLs will occur on the agency. Systems must complete one
IDSE implementation, and welcomes schedule of the largest system for all IDSE to meet the requirements of
proactive State involvement. However, systems in the combined distribution today’s rule.
the Agency believes that delaying system (regardless of size). If a a. Applicability. This requirement
implementation of risk-based IDSE consecutive system must make capital applies to all community water systems,
targeting activities until States receive improvements to comply with this rule, and to large nontransient
primacy is an unacceptable delay in the State may use its existing authority noncommunity water systems (those
public health protection and also to grant up to an additional 24 months serving at least 10,000 people) that use
inconsistent with the Advisory to that system. In addition, a primary or residual disinfectant other
Committee’s recommendations. EPA implementation and data tracking will than ultraviolet light, or that deliver
remains committed to working with be simplified because all systems in a water that has been treated with a
States to the greatest extent feasible to combined distribution system will be on primary or residual disinfectant other
implement today’s rule, consistent with than ultraviolet light. Systems serving
the same IDSE and Stage 2 compliance
the schedule promulgated today. For fewer than 500 people are covered by
schedule. EPA believes that this is a
States unable to actively participate in the very small system waiver provisions
better approach from both a public
IDSE implementation, however, EPA of today’s rule and are not required to
health standpoint and an complete an IDSE if they have TTHM
believes it has an obligation to provide implementation standpoint.
support and guidance to PWSs who are and HAA5 data collected under Subpart
covered and independently responsible EPA agrees with many commenters L. Consecutive systems are subject to
for complying with the IDSE that a high level of coordination among the IDSE requirements of today’s rule.
requirements of today’s rule and is wholesaler, consecutive system, and Consecutive systems must comply with
prepared to oversee implementation. States will be necessary to ensure IDSE requirements on the same
Moreover, EPA believes that the compliance. The schedule in today’s schedule as the system serving the
staggered compliance schedule in rule provides more time for planning, largest population in the combined
today’s final rule will enhance States’ reviewing, and conducting the IDSE distribution system, as described in
ability to help implement the rule. than the schedule in the proposed rule, section IV.E.
Consecutive systems. Most which will allow more time for b. Data collection. For those systems
commenters supported consecutive necessary coordination, including small not receiving a very small system
systems being on the same IDSE consecutive systems that need help in waiver, there are three possible
schedule as wholesale systems, negotiations with their wholesale approaches by which a system can meet
recognizing the benefits of treatment system. EPA will work with ASDWA the IDSE requirement.
plant capital and operational and States to develop guidance to i. Standard monitoring. Standard
improvements by the wholesale system facilitate wholesale/consecutive system monitoring requires one year of DBP
as the preferred method of DBP cooperation. This additional time and monitoring throughout the distribution
compliance, with the timely collection the staggered schedule discussed in this system on a specified schedule. Prior to
of DBP data throughout the combined section also lessens the laboratory commencing standard monitoring,
distribution system a key component. systems must prepare a monitoring plan
burden associated with IDSE
Several commenters preferred that and submit it to the primacy agency for
consecutive systems have a later Stage review. The frequency and number of
2 compliance date to allow for The staggered schedule also helps samples required under standard
evaluation of whether wholesale system address commenter concerns about monitoring is determined by source
treatment changes are adequate to evaluating combined distribution water type and system size. The number
ensure compliance and to consider systems. Other commenters’ concerns of samples does not depend on the
changes to water delivery specifications. about time needed for developing number of plants per system. Section
EPA disagrees with those commenters contracts between systems and for IV.G provides a detailed discussion of
wwhite on PROD1PC65 with RULES2

recommending a different Stage 2 planning, funding, and implementing the specific population-based
compliance date and thus has treatment changes are addressed by not monitoring requirements for IDSE
maintained the approach in the requiring Stage 2 compliance until at standard monitoring. Although standard
proposal, which keeps all systems that least six years following rule monitoring results are not to be used for
are part of a combined distribution promulgation. determining compliance with MCLs,

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00033 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
420 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations

systems are required to include a wide range of sample sites HAA5 monitoring violations during the
individual sample results for the IDSE representative of the entire distribution same period. The State or primacy
results when determining the range of system, including those judged to agency may require systems to submit
TTHM and HAA5 levels to be reported represent high TTHM and HAA5 compliance monitoring results,
in their Consumer Confidence Report concentrations, and (2) extended period distribution system schematics, or
(see section IV.J). simulation hydraulic models that recommend subpart V compliance
ii. System specific study. Under this simulate water age in the distribution monitoring locations as part of the
approach, systems may choose to system, in conjunction with one round certification. This certification must be
perform a system specific study based of TTHM and HAA5 sampling. kept on file and submitted to the State
on earlier monitoring studies or iii. 40/30 certification. Under this or primacy agency for review. Systems
distribution system hydraulic models in approach, systems must certify to their that qualify for reduced monitoring for
lieu of standard monitoring. Prior to State or primacy agency that every the Stage 1 DBPR during the two years
commencing a system specific study, individual compliance sample taken prior to the start of the IDSE may use
systems must prepare a study plan and under subpart L during the period results of reduced Stage 1 DBPR
submit it to the primacy agency for specified in Table IV.F–2 were less than monitoring to prepare the 40/30
approval. The two options for system or equal to 0.040 mg/L for TTHM and certification. The requirements for the
specific studies are: (1) TTHM and less than or equal to 0.030 mg/L for 40/30 certification are listed in Table
HAA5 monitoring data that encompass HAA5, and that there were no TTHM or IV.F–1.


40/30 Certification Requirements ... • A certification that every individual compliance sample taken under subpart L during the period specified
in Table IV.F–2 were less than or equal to 0.040 mg/L for TTHM and less than or equal to 0.030 mg/L
for HAA5, and that there were no TTHM or HAA5 monitoring violations during the same period.
• Compliance monitoring results, distribution system schematics, and/or recommended subpart V compli-
ance monitoring locations as required by the State or primacy agency.


Then your eligibility for 40/30 certification is based on eight consecutive
If your 40/30 Certification Is Due calendar quarters of subpart L compliance monitoring results beginning
no earlier than1

(1) October 1, 2006 .................................................................................. January 2004.

(2) April 1, 2007 ........................................................................................ January 2004.
(3) October 1, 2007 .................................................................................. January 2005.
(4) April 1, 2008 ........................................................................................ January 2005.
1 Unless you are on reduced monitoring under subpart L and were not required to monitor during the specified period. If you did not monitor
during the specified period, you must base your eligibility on compliance samples taken during the 12 months preceding the specified period.

c. Implementation. All systems monitoring, study plan for system requirements for the IDSE plan depend
subject to the IDSE requirement under specific study) or 40/30 certification to on the IDSE approach that the system
this final rule (except those covered by the State or primacy agency. IDSE plans selects and are listed in Tables IV.F–1
the very small system waiver) must and 40/30 certifications must be and IV.F–3.
prepare and submit an IDSE plan submitted according to the schedule
(monitoring plan for standard described in section IV.E and IV.M. The


IDSE data collection alternative IDSE plan requirements

Standard Monitoring ........................ • Schematic of the distribution system (including distribution system entry points and their sources, and
storage facilities), with notes indicating locations and dates of all projected standard monitoring, and all
projected subpart L compliance monitoring.
• Justification for all standard monitoring locations selected and a summary of data relied on to select
those locations.
• Population served and system type (subpart H or ground water).
System Specific Study:
Hydraulic Model .............................. Hydraulic models must meet the following criteria:
• Extended period simulation hydraulic model.
• Simulate 24 hour variation in demand and show a consistently repeating 24 hour pattern of residence
• Represent 75% of pipe volume; 50% of pipe length; all pressure zones; all 12-inch diameter and larger
pipes; all 8-inch and larger pipes that connect pressure zones, influence zones from different sources,
storage facilities, major demand areas, pumps, and control valves, or are known or expected to be sig-
wwhite on PROD1PC65 with RULES2

nificant conveyors of water; all pipes 6 inches and larger that connect remote areas of a distribution sys-
tem to the main portion of the system; all storage facilities with standard operations represented in the
model; all active pump stations with controls represented in the model; and all active control valves.

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00034 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 421


IDSE data collection alternative IDSE plan requirements

• The model must be calibrated, or have calibration plans, for the current configuration of the distribution
system during the period of high TTHM formation potential. All storage facilities must be evaluated as
part of the calibration process.
• All required calibration must be completed no later than 12 months after plan submission.
Submission must include:
• Tabular or spreadsheet data demonstrating percent of total pipe volume and pipe length represented in
the model, broken out by pipe diameter, and all required model elements.
• A description of all calibration activities undertaken, and if calibration is complete, a graph of predicted
tank levels versus measured tank levels for the storage facility with the highest residence time in each
pressure zone, and a time series graph of the residence time at the longest residence time storage facil-
ity in the distribution system showing the predictions for the entire simulation period (i.e., from time zero
until the time it takes for the model to reach a consistently repeating pattern of residence time).
• Model output showing preliminary 24 hour average residence time predictions throughout the distribution
• Timing and number of samples planned for at least one round of TTHM and HAA5 monitoring at a num-
ber of locations no less than would be required for the system under standard monitoring in § 141.601
during the historical month of high TTHM. These samples must be taken at locations other than existing
subpart L compliance monitoring locations.
• Description of how all requirements will be completed no later than 12 months after submission of the
system specific study plan.
• Schematic of the distribution system (including distribution system entry points and their sources, and
storage facilities), with notes indicating the locations and dates of all completed system specific study
monitoring (if calibration is complete) and all subpart L compliance monitoring.
• Population served and system type (subpart H or ground water).
• If the model submitted does not fully meet the requirements, the system must correct the deficiencies
and respond to State inquiries on a schedule the State approves, or conduct standard monitoring.
System Specific Study:
Existing Monitoring Results ............ Existing monitoring results must meet the following criteria:
• TTHM and HAA5 results must be based on samples collected and analyzed in accordance with
§ 141.131. Samples must be collected within five years of the study plan submission date.
• The sampling locations and frequency must meet the requirements identified in Table IV.F–4. Each loca-
tion must be sampled once during the peak historical month for TTHM levels or HAA5 levels or the
month of warmest water temperature for every 12 months of data submitted for that location. Monitoring
results must include all subpart L compliance monitoring results plus additional monitoring results as
necessary to meet minimum sample requirements.
Submission must include:
• Previously collected monitoring results
• Certification that the reported monitoring results include all compliance and non-compliance results gen-
erated during the time period beginning with the first reported result and ending with the most recent
subpart L results.
• Certification that the samples were representative of the entire distribution system and that treatment
and distribution system have not changed significantly since the samples were collected.
• Schematic of the distribution system (including distribution system entry points and their sources, and
storage facilities), with notes indicating the locations and dates of all completed or planned system spe-
cific study monitoring.
• Population served and system type (subpart H or ground water).
• If a system submits previously collected data that fully meet the number of samples required for IDSE
monitoring in Table IV.F–4 and some of the data are rejected due to not meeting the additional require-
ments, the system must either conduct additional monitoring to replace rejected data on a schedule the
State approves, or conduct standard monitoring.


Number of Number of samples
System type Population size category monitoring lo-
cations TTHM HAA5

Subpart H:

<500 3 3 3

500–3,300 3 9 9

3,301–9,999 6 36 36
wwhite on PROD1PC65 with RULES2

10,000–49,999 12 72 72

50,000–249,999 24 144 144

250,000–999,999 36 216 216

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00035 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
422 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations


Number of Number of samples
System type Population size category monitoring lo-
cations TTHM HAA5

1,000,000–4,999,999 48 288 288

≥ 5,000,000 60 360 360

Ground Water: <500 3 3 3

500–9,999 3 9 9

10,000–99,999 12 48 48

100,000–499,999 18 72 72

≥ 500,000 24 96 96

The State or primacy agency will the IDSE plan approved by the State or monitoring plan submission may submit
approve the IDSE plan or 40/30 primacy agency according to the a combined monitoring plan and report
certification, or request modifications. If schedule described in section IV.E. on the required schedule for IDSE plan
the State or primacy agency has not All systems completing standard submissions. The requirements for the
taken action by the date specified in monitoring or a system specific study IDSE report are listed in Table IV.F–5.
section IV.E or has not notified the must submit a report to the State or Some of these reporting requirements
system that review is not yet complete, primacy agency according to the have changed from the proposal to
systems may consider their submissions schedule described in section IV.E. reduce reporting and paperwork burden
to be approved. Systems must Systems that have completed their on systems.
implement the IDSE option described in system specific study at the time of


IDSE data collection alternative IDSE report requirements

Standard Monitoring ........................ • All subpart L compliance monitoring and standard monitoring TTHM and HAA5 analytical results in a
tabular format acceptable to the State.
• If changed from the monitoring plan, a schematic of the distribution system, population served, and sys-
tem type.
• An explanation of any deviations from the approved monitoring plan.
• Recommendations and justifications for subpart V compliance monitoring locations and timing.
System Specific Study .................... • All subpart L compliance monitoring and all system specific study monitoring TTHM and HAA5 analytical
results conducted during the period of the system specific study in a tabular or spreadsheet form accept-
able to the State.
• If changed from the study plan, a schematic of the distribution system, population served, and system
• If using the modeling provision, include final information for required plan submissions and a 24-hour
time series graph of residence time for each subpart V compliance monitoring location selected.
• An explanation of any deviations from the original study plan.
• All analytical and modeling results used to select subpart V compliance monitoring locations that show
that the system specific study characterized TTHM and HAA5 levels throughout the entire distribution
• Recommendations and justifications for subpart V compliance monitoring locations and timing.

All systems must prepare Stage 2 today’s final rule, and in section IV.G. agency has not taken action by the date
compliance monitoring Generally, a system must recommend specified in section IV.E or has not
recommendations. All IDSE reports locations with the highest LRAAs unless notified the system that review is not
must include recommendations for it provides a rationale (such as ensuring yet complete, systems may consider
Stage 2 compliance monitoring geographical coverage of the their submission to be approved and
locations and sampling schedule. distribution system instead of clustering prepare to begin Stage 2 compliance
Systems submitting a 40/30 certification all sites in a particular section of the monitoring.
must include their Stage 2 compliance distribution system) for selecting other EPA has developed the Initial
monitoring recommendations in their locations. In evaluating possible Stage 2 Distribution System Evaluation
Stage 2 (Subpart V) monitoring plan compliance monitoring locations, Guidance Manual for the Final Stage 2
wwhite on PROD1PC65 with RULES2

unless the State requests Subpart V site systems must consider both Stage 1 Disinfectants and Disinfection
recommendations as part of the 40/30 DBPR compliance data and IDSE data. Byproducts Rule (USEPA 2006) to assist
certification. The number of sampling The State or primacy agency will systems with implementing each of
locations and the criteria for their approve the IDSE report or request these requirements. This guidance may
selection are described in § 141.605 of modifications. If the State or primacy be requested from EPA’s Safe Drinking

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00036 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 423

Water Hotline, which may be contacted compliance data, identify high DBP small systems (serving <500) at both
as described under FOR FURTHER locations. Monitoring at additional sites average residence time and maximum
INFORMATION CONTACT in the beginning increases the chance of finding sites residence time locations are lower than
of this notice. This guidance manual is with high DBP levels and targets both levels at both of those locations in larger
also available on the EPA Web site at DBPs that degrade and DBPs that form systems, and the change in residual as residence time increases in the concentration between those two
index.html. distribution system. EPA believes that locations is smaller in very small
the required number of standard systems compared to larger sized
2. Background and Analysis
monitoring locations plus Stage 1 systems. The magnitude of the
In the Stage 2 DBPR proposal monitoring results will provide an reduction in residual concentration
(USEPA, 2003a), EPA proposed adequate characterization of DBP levels gives an indication of how much
requirements for systems to complete an throughout the distribution system at a disinfectant has reacted to form DBPs,
IDSE. The Agency based its proposal reasonable cost. By revising Stage 2 including TTHM and HAA5. The
upon the Stage 2 M–DBP Advisory compliance monitoring plans to target smaller reduction in disinfectant
Committee recommendations in the locations with high DBPs, systems will concentration between average
Agreement in Principle. The Advisory be required to take steps to address high residence time and maximum residence
Committee believed and EPA concurs DBP levels at locations that might time in very small systems compared to
that maintaining Stage 1 DBPR otherwise have gone undetected. larger systems indicates that DBP
monitoring sites for the Stage 2 DBPR The Advisory Committee formation potential is probably lower in
would not accomplish the risk-targeting recommended that an IDSE be very small systems compared to larger
objective of minimizing high DBP levels performed by all community water systems, and the likelihood for
and providing consistent and equitable systems, unless the system had significant DBP variation within the
protection across the distribution sufficiently low DBP levels or is a very distribution system of very small
system. Most of these requirements have small system with a simple distribution systems is low if the distribution system
not changed from the proposed rule. system. EPA believes that large is small and not complex. However,
The data collection requirements of nontransient noncommunity water there may be some small systems with
the IDSE are designed to find both high systems (NTNCWS) (those serving at extended or complex distribution
TTHM and high HAA5 sites (see section least 10,000 people) also have systems that should be studied further
IV.G for IDSE monitoring requirements). distribution systems that require further to determine new sampling locations.
High TTHM and HAA5 concentrations evaluation to determine the locations For this reason, States or primacy
often occur at different locations in the most representative of high DBP levels agencies can require any particular very
distribution system. The Stage 1 DBPR and proposed that they be required to
monitoring sites identified as the small system to conduct an IDSE. Very
conduct an IDSE. Therefore, large small systems subject to the Stage 2
maximum location are selected NTNCWS and all community water
according to residence time. HAAs can DBPR that do not have a Stage 1
systems are required to comply with compliance monitoring location may
degrade in the distribution system in the IDSE requirements under today’s final
absence of sufficient disinfectant monitor in accordance with the Stage 1
rule, unless they submit a 40/30 DBPR provisions to be eligible for this
residual (Baribeau et al. 2000). certification or they are covered by the
Consequently, residence time is not an waiver.
very small system waiver provisions.
ideal criterion for identifying high b. Very small system waivers. Systems c. 40/30 certifications. Systems that
HAA5 sites. In addition, maximum serving fewer than 500 people that have certify to their State or primacy agency
residence time locations that are taken samples under the Stage 1 DBPR that all compliance samples taken
associated with high TTHM levels may will receive a very small system waiver. during eight consecutive calendar
not be constant due to daily or seasonal EPA proposed and the Advisory quarters prior to the start of the IDSE
changes in demand. The analysis of Committee recommended a very small were ≤0.040 mg/L TTHM and ≤0.030
maximum residence time completed for system waiver following a State mg/L HAA5 are not required to collect
the selection of Stage 1 monitoring sites determination that the existing Stage 1 additional DBP monitoring data under
may not have been capable of detecting compliance monitoring location the IDSE requirements as long as the
these variations. The Information adequately characterizes both high system has no TTHM or HAA5
Collection Rule data show that over 60 TTHM and high HAA5 for the monitoring violations. These criteria
percent of the highest HAA5 LRAAs and distribution system because many very were developed because both EPA and
50 percent of the highest TTHM LRAAs small systems have small or simple the AdvisoryCommittee determined that
were found at sampling locations in the distribution systems. The final rule these systems most likely would not
distribution system other than the grants the very small system waiver to have DBP levels that exceed the MCLs.
maximum residence time compliance all systems serving fewer than 500 that Systems must have qualifying TTHM
monitoring location (USEPA 2003a). have Stage 1 DBPR data. This provision and HAA5 data for eight consecutive
Therefore, the method and assumptions was changed from the proposal to reflect calendar quarters according to the
used to select the Information Collection that most very small systems that schedule in Table IV.F–2 to be eligible
Rule monitoring sites and the Stage 1 sample under the Stage 1 DBPR have for this option. Systems on reduced
DBPR compliance monitoring sites may sampling locations that are monitoring that did not monitor during
not reliably capture high DBP levels for representative of both high TTHM and the specified time period may use data
Stage 2 DBPR compliance monitoring high HAA5 because most very small from the prior year to meet the 40/30
sites. systems have small and simple certification criteria. Systems that have
a. Standard monitoring. The Advisory distribution systems. In addition, many not previously conducted Stage 1 DBPR
wwhite on PROD1PC65 with RULES2

Committee recommended that systems very small systems are ground water compliance monitoring may begin such
sample throughout the distribution systems that typically have stable DBP monitoring to collect the data necessary
system at twice the number of locations levels that tend to be lower than surface to qualify for 40/30 certification. The
as required under Stage 1 and, using water DBP levels. NRWA survey data certification and data supporting it must
these results in addition to Stage 1 show that free chlorine residual in very be available to the public upon request.

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00037 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
424 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations

The qualifying time period for the 40/ compliance monitoring as part of the requirements reflect the fact that the
30 certification has changed from the 40/30 certification. This provision was purpose of the model is to predict water
proposed rule. included to facilitate primacy agency age. ICR data show that HAA5 data do
Under the proposed rule, the rule review of 40/30 certifications; the not necessarily correlate well with water
language identified a specific two year additional information is only required age (USEPA 2003a). Because the
window with start and end dates. In if requested by the primacy agency. purpose of the IDSE is to locate
today’s final rule, the qualifying time d. System specific studies. Advisory representative high locations for both
period has been changed to ‘‘eight Committee members recognized that TTHM and HAA5, one round of
consecutive calendar quarters of subpart some systems have detailed knowledge monitoring must be completed at
L compliance monitoring results of their distribution systems by way of potential Stage 2 compliance monitoring
beginning no earlier than * * *’’ (see ongoing hydraulic modeling and/or locations to determine appropriate
Table IV.F–2). This change was made so existing widespread monitoring plans HAA5 monitoring locations during the
that systems that have made a treatment (beyond that required for compliance historical high month of TTHM
change within the two years prior to monitoring) that would provide concentrations. The number of locations
rule promulgation and have collected equivalent or superior monitoring site must be no less than would be required
initial data that meet the 40/30 criteria selection information compared to under standard monitoring.
might have the opportunity to collect standard monitoring. Therefore, the Preliminary average residence time
eight consecutive quarters of qualifying Advisory Committee recommended that data are required as a part of the study
data and apply for a 40/30 certification. such systems be allowed to determine plan for systems to demonstrate that
This schedule change also allows new monitoring sites using system- their distribution system hydraulic
systems that have not previously specific data such as hydraulic model model is able to produce results for
monitored under Stage 1 an opportunity results or existing monitoring data; this water age throughout the distribution
to qualify for a 40/30 certification. provision remains in the final rule. In system, even though calibration may not
Under the proposed Stage 2 DBPR, the proposed rule, the only specification be complete. Systems also need to
systems that missed the deadline for for SSSs was to identify monitoring sites describe their plans to complete the
submitting a 40/30 certification would that would be equivalent or superior to modeling requirements within 12
be required to conduct either standard those identified under Standard months of submitting the study plan.
monitoring or a system specific study Monitoring. The final rule includes These last two requirements were
even if the system otherwise qualified more specific requirements on how developed so that States can be assured
for the 40/30 certification. Under these studies should be completed. The that systems have the technical capacity
today’s final rule, systems that do not requirements in the final rule were to complete their modeling
make any submission by the IDSE plan developed to be consistent with the requirements by the IDSE report
submission deadline will still receive a proposal, yet more specific to help deadline. If systems cannot demonstrate
violation, but may submit a late 40/30 systems better understand expectations that they are in a position to complete
certification if their data meet the under this provision and lessen the the modeling requirements according to
requirements. This change was made so chances of a study plan not being the required schedule, they will be
that systems and primacy agencies do approved. required to complete standard
not spend time preparing and reviewing The new modeling requirements were monitoring.
standard monitoring plans and IDSE developed to reflect that hydraulic All new modeling requirements were
reports for systems with a low models can identify representative high added to help systems demonstrate how
likelihood of finding high TTHM and TTHM monitoring locations by their model will fulfill the purpose and
HAA5 levels. predicting hydraulic residence time in requirements of the IDSE and to assist
The reporting requirements for this the distribution system. Water age has primacy agencies with approval
provision have been reduced from the been found to correlate with TTHM determinations. The associated
requirements in the proposed formation in the distribution system. reporting requirements were developed
rulemaking. In the proposal, systems Consequently, for this system specific to balance the needs of systems to
qualifying for the 40/30 certification study approach, hydraulic residence demonstrate that they have fulfilled the
were required to submit all qualifying time predicted by the model is used as requirements and the needs of primacy
data and provide recommendations for a surrogate for TTHM formation to agency reviewers to be able to
Stage 2 compliance monitoring locate appropriate Stage 2 compliance understand the work completed by the
locations. The final rule requires monitoring locations. To predict system.
systems to submit a certification that hydraulic residence time in the EPA has specified new requirements
their data meet all the requirements of distribution system, the model must for systems complete an SSS using
the 40/30 certification and to include represent most of the distribution existing monitoring data to help systems
their Stage 2 compliance monitoring system and must have been calibrated understand the extent of historical data
recommendations in their Stage 2 recently and appropriately to reflect that would meet the requirements of the
monitoring plan. These changes were water age in the distribution system. IDSE. The number of required sample
made to reduce the reporting burden on Requirements to reflect this are in locations and samples are consistent
systems that qualify for the 40/30 today’s rule. All storage facilities must with sampling requirements under
certification and to maintain be evaluated for the calibration, and standard monitoring and the
consistency with monitoring plan systems using this option must submit recommendations made by the Advisory
requirements under the Stage 1 DBPR. a graph of predicted tank levels versus Committee. The Advisory Committee
This approach also gives systems more measured tank levels for the storage recommended that systems complete an
time to select appropriate monitoring facility with the highest residence time IDSE sample at twice the number of
wwhite on PROD1PC65 with RULES2

sites for Stage 2 compliance monitoring. in each pressure zone. These calibration sites required by the Stage 1 DBPR in
The State or primacy agency may requirements are focused on storage addition to Stage 1 DBPR sampling.
request systems to submit the data, a facilities because they are the largest Because the number of required Stage 1
distribution system schematic, and/or controlling factor for water age in the DBPR monitoring locations varies
recommendations for Stage 2 distribution system. The calibration within each population category under

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00038 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 425

the Stage 1 plant-based monitoring existing monitoring results were taken In addition to addressing the very
approach (since systems have different during the earliest eligible dates. Again, small system waivers, commenters
numbers of plants), EPA used the these clarifications were made so that suggested that different criteria should
number of required Standard systems could better understand the be used for the 40/30 certification, such
Monitoring locations plus the number of extent of data necessary for a monitoring as higher minimum DBP levels, cut-offs
Stage 2 compliance monitoring plan to be deemed acceptable and be of 40/30 as LRAAs or RAAs rather than
locations to develop minimum confident that efforts to complete an single sample maximums, or State or
requirements for the use of existing SSS would be found acceptable to the primacy agency discretion on which
monitoring data for the SSS. The State or primacy agency. systems should qualify for 40/30
number of required locations and e. Distribution System Schematics. certification. There was no consensus
samples are shown in Table IV.F–4. EPA has considered security concerns among the commenters on what changes
Systems will use their Stage 1 that may result from the requirement for should be made to the proposal for the
monitoring results plus additional non- systems to submit a distribution system 40/30 certification requirements. EPA
compliance or operational samples to schematic as part of their IDSE plan. did not change the requirements for the
fulfill these requirements. Small EPA believes that the final rule strikes 40/30 certification eligibility because
systems with many plants may have an appropriate balance between security the recommended alternatives were not
been collecting a disproportionate concerns and the need for States and technically superior to the requirements
number of samples under the Stage 1 primacy agencies to be able to review of the proposed rule. Implementation of
DBPR compared to the population based IDSE plans. EPA has developed 40/30 criteria using an LRAA or RAA
monitoring requirements presented in guidance for systems on how to submit would result in reduced public health
today’s rule and may have sufficient a distribution system schematic that protection from the rule by allowing
historical data to characterize the entire does not include sensitive information. higher DBP levels to go undetected. EPA
distribution system. These requirements 3. Summary of Major Comments did change the eligibility dates and
allow those systems to submit an SSS reporting requirements for the 40/30
based on existing Stage 1 monitoring The Agency received significant certification to reduce the burden on the
results, and they also accommodate comments on the following issues system. Under today’s final rule, States
systems that have been completing related to the proposed IDSE or primacy agencies can request TTHM
additional monitoring throughout the requirements: Waiver limitations, and and HAA5 data as desired for a more in-
distribution system. State or primacy agency review of IDSE depth review of a system’s
The requirement to sample during the plans. qualifications.
historical month of high TTHM, high In the proposed rule, EPA requested Many commenters expressed concern
HAA5, or warmest water temperature comment on what the appropriate over the implementation schedule for
during each year for which data were criteria should be for States or primacy the IDSE. Commenters were especially
collected was added to maintain agencies to grant very small system concerned that IDSE plans would be
consistency with the standard waivers. Commenters responded with a developed and implemented prior to
monitoring requirements where each wide range of suggestions including State primacy, and once States receive
location must be sampled one time support for the proposal as written, primacy, they might not support the
during the peak historical month. different population cut-offs, State or IDSE plan and would reject the results
Samples that qualify for this SSS must primacy agency discretion on what of the completed IDSE. To address this
have been collected within five years of system size should qualify for the issue, commenters requested the
the study plan submission date and waiver, and alternative waiver criteria opportunity for States to review the
must reflect the current configuration of such as pipe length or number of IDSE plans prior to systems completing
treatment and the distribution system. booster stations. There was no their IDSEs. In today’s rule EPA has
Five years was selected as a cut off for consensus among the commenters on modified the compliance schedule for
eligible data so that all data submitted what changes should be made to the the Stage 2 DBPR so that systems have
would be reasonably representative of proposal for the very small system the opportunity to complete their IDSE
current source water conditions and waiver requirements. EPA did not plan and have it reviewed by the
DBP formation within the distribution change the population cutoff for the primacy agency prior to completing the
system. Data that are older may not very small system waiver because IDSE to address the concern that States
reflect current DBP formation potential analysis of NRWA survey data also or primacy agencies may reject the
in the distribution system. Five years showed that systems serving fewer than results of the completed IDSE. The
prior to the submission of the study 500 had different residence times and changes to the compliance schedule are
plan also correlates with the signing of lower free chlorine residual discussed further in section IV.E.
the Agreement in Principle where the concentrations compared to other
population categories, indicating that G. Monitoring Requirements and
Advisory Committee made the
larger systems have different DBP Compliance Determination for TTHM
recommendation for this provision.
formation characteristics compared to and HAA5 MCLs
Systems interested in using this
provision would have started eligible very small systems. Some of the EPA is finalizing monitoring
monitoring after the agreement was suggested changes for very small system requirements under a population-based
signed. waiver criteria may require data that are approach described in this section. EPA
Systems that submit existing not readily available to systems (such as believes the population-based approach
monitoring data must submit all Stage 1 pipe length in service) and for which will provide more representative high
sample results from the beginning of the there were no specific criteria proposed DBP concentrations throughout
SSS to the time when the SSS plan is or recommended by the commenters. distribution systems than would plant-
wwhite on PROD1PC65 with RULES2

submitted. The purpose of this Implementation of subjective very small based monitoring, is equitable, and will
requirement is to demonstrate that there system waiver criteria would result in simplify implementation for both States
have been no significant changes in reduced public health protection from and systems. For these reasons, EPA
source water quality since the first the rule by allowing higher DBP levels believes this approach is more
samples were collected, especially if all to go undetected. appropriate than the proposed plant-

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00039 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
426 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations

based monitoring. Detailed discussion based approach, monitoring must collect samples during the peak
of the two approaches is presented in requirements are based solely on the historical month for DBP levels or water
the preamble of the proposed rule retail population served and the type of temperature; this will determine their
(USEPA 2003a) and EA for today’s rule source water used and not influenced by monitoring schedule. Table IV.G–1
(USEPA 2005a). the number of treatment plants or entry contains the IDSE monitoring
1. Today’s Rule points in the distribution system as in frequencies and locations for all source
previous rules (i.e., TTHM Rule (USEPA water and size category systems. Section
Today’s rule establishes TTHM and
1979) and Stage 1 DBPR (USEPA IV.F identifies other approaches by
HAA5 monitoring requirements for all
systems based on a population-based 1998a)). which systems can meet IDSE
monitoring approach instead of a plant- a. IDSE Monitoring. All systems requirements.
based approach. Under the population- conducting IDSE standard monitoring


Distribution system monitoring locations 1
Source water Monitoring periods and
Population size category Total per Average
type frequency of sampling monitoring Near entry residence High TTHM High HAA5
points locations locations
period time

Subpart H
<500 consecutive sys- one (during peak histor- 2 1 .................... 1
tems. ical month) 2.
<500 non-consecutive .......................................... 2 .................... .................... 1 1

500–3,300 non-consecu- four (every 90 days) ........ 2 1 .................... 1 ....................

tive systems.
500–3,300 consecutive .......................................... 2 .................... .................... 1 1

3,301–9,999 ..................... .......................................... 4 .................... 1 2 1

10,000–49,999 ................. six (every 60 days) .......... 8 1 2 3 2
50,000–249,999 ............... .......................................... 16 3 4 5 4
250,000–999,999 ............. .......................................... 24 4 6 8 6
1,000,000–4,999,999 ....... .......................................... 32 6 8 10 8
≥5,000,000 ....................... .......................................... 40 8 10 12 10

<500 consecutive sys- one (during peak histor- 2 1 .................... 1 ....................
tems. ical month) 2.

<500 non-consecutive .......................................... 2 .................... .................... 1 1

500–9,999 ........................ four (every 90 days) ........ 2 .................... .................... 1 1
10,000–99,999 ................. .......................................... 6 1 1 2 2
100,000–499,999 ............. .......................................... 8 1 1 3 3
≥500,000 .......................... .......................................... 12 2 2 4 4
1A dual sample set (i.e., a TTHM and an HAA5 sample) must be taken at each monitoring location during each monitoring period.
2 The peak historical month is the month with the highest TTHM or HAA5 levels or the warmest water temperature.

b. Routine Stage 2 Compliance additional monitoring locations, if levels. Systems may also recommend
Monitoring. For all systems conducting required by the transition from plant- locations with lower levels of DBPs that
either standard monitoring or a system based monitoring to population-based would not be picked up by the protocol
specific study, initial Stage 2 monitoring. if they provide a rationale for the
compliance monitoring locations are Systems recommend Stage 2 recommendation. Examples of
based on the system’s IDSE results, monitoring locations generally by rationales include ensuring better
together with an analysis of a system’s arraying results of IDSE standard distribution system or population
Stage 1 DBPR compliance monitoring monitoring (or system specific study coverage (not having all locations in the
results. Systems receiving 40/30 results) and Stage 1 compliance same area) or maintaining existing
certification or a very small system monitoring by monitoring location (from locations with DBP levels that are nearly
waiver, and nontransient highest to lowest LRAA for both TTHM as high as those that would otherwise be
noncommunity water systems serving and HAA5). Using the protocol in selected. The State or primacy agency
<10,000 not required to conduct an § 141.605(c) of today’s rule, systems will review these recommendations as
IDSE, base Stage 2 initial compliance then select the required number of part of the review of the IDSE report
wwhite on PROD1PC65 with RULES2

monitoring locations on the system’s locations. Larger systems include submitted by systems that conducted
Stage 1 DBPR compliance monitoring existing Stage 1 monitoring locations in standard monitoring or a system specific
results. Some of these systems may also order to be able to have historical study.
need an evaluation of distribution continuity for evaluating how changes Table IV.G–2 contains the routine
system characteristics to identify in operations or treatment affect DBP Stage 2 TTHM and HAA5 compliance

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00040 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 427

monitoring requirements for all systems report (those receiving a 40/30 period’’ column at current Stage 1
(both non-consecutive and consecutive certification or very small system waiver compliance monitoring locations, unless
systems), as well as the protocol for and nontransient noncommunity water the State or primacy agency specifically
Stage 2 compliance monitoring location systems serving <10,000) must conduct directs otherwise. All systems are then
selection in the IDSE report. Systems Stage 2 compliance monitoring as required to maintain and follow a Stage
that do not have to submit an IDSE indicated in the ‘‘Total per monitoring 2 compliance monitoring plan.


Distribution system monitoring location
Source water type Population size category Monitoring frequency1 Total per Highest Highest Subpart L
monitoring TTHM loca- HAA5 loca- compliance
period2 tions tions locations

Subpart H:
<500 .................................. per year ............................. 2 1 1 ....................
500–3,300 ......................... per quarter ........................ 2 1 1 ....................
3,301–9,999 ...................... per quarter ........................ 2 1 1 ....................
10,000–49,999 .................. per quarter ........................ 4 2 1 1
50,000–249,999 ................ per quarter ........................ 8 3 3 2
250,000–999,999 .............. per quarter ........................ 12 5 4 3
1,000,000–4,999,999 ........ per quarter ........................ 16 6 6 4
≥ 5,000,000 ....................... per quarter ........................ 20 8 7 5
Ground water:
<500 .................................. per year ............................. 2 1 1 ....................
500–9,999 ......................... per year ............................. 2 1 1 ....................
10,000–99,999 .................. per quarter ........................ 4 2 1 1
100,000–499,999 .............. per quarter ........................ 6 3 2 1
≥ 500,000 .......................... per quarter ........................ 8 3 3 2
1 All
systems must monitor during month of highest DBP concentrations.
2 Systems on quarterly monitoring must take dual sample sets every 90 days at each monitoring location, except for subpart H systems serving
500–3,300. Systems on annual monitoring and subpart H systems serving 500–3,300 are required to take individual TTHM and HAA5 samples
(instead of a dual sample set) at the locations with the highest TTHM and HAA5 concentrations, respectively. Only one location with a dual sam-
ple set per monitoring period is needed if highest TTHM and HAA5 concentrations occur at the same location, and month, if monitored annually).

Today’s rule provides States the and chloramines. States may use the 0.040 mg/L for TTHM or 0.030 mg/L for
flexibility to specify alternative Stage 2 provisions of § 141.134(c) to modify HAA5 or if the source water annual
compliance monitoring requirements reporting requirements. For example, average TOC level, before any treatment,
(but not alternative IDSE monitoring the State may require that only the exceeds 4.0 mg/L at any of the system’s
requirements) for multiple consecutive consecutive system distribution system treatment plants treating surface water
systems in a combined distribution point-of-entry disinfectant or ground water under the direct
system. As a minimum under such an concentration be reported to influence of surface water, the system
approach, each consecutive system must demonstrate MRDL compliance, must resume routine monitoring. For
collect at least one sample among the although monitoring requirements may systems with annual or less frequent
total number of samples required for the not be reduced.
reduced monitoring, systems may
combined distribution system and will i. Reduced monitoring. Systems can
qualify for reduced monitoring, as remain on reduced monitoring as long
base compliance on samples collected
within its distribution system. The specified in Table IV.G–3, if the LRAA as each TTHM sample is ≤0.060 mg/L
consecutive system is responsible for at each location is ≤0.040 mg/L for and each HAA5 sample is ≤0.045 mg/L.
ensuring that required monitoring is TTHM and ≤0.030 mg/L for HAA5 based If the annual (or less frequent) sample
completed and the system is in on at least one year of monitoring at at any location exceeds either 0.060 mg/
compliance. It also must document its routine compliance monitoring L for TTHM or 0.045 mg/L for HAA5,
monitoring strategy as part of its subpart locations. Systems may remain on or if the source water annual average
V monitoring plan. reduced monitoring as long as the TOC level, before any treatment,
Consecutive systems not already TTHM LRAA is ≤0.040 mg/L and the exceeds 4.0 mg/L at any treatment plant
conducting disinfectant residual HAA5 LRAA is ≤0.030 mg/L at each treating surface water or ground water
monitoring under the Stage 1 DBPR monitoring location for systems with under the direct influence of surface
must comply with the monitoring quarterly reduced monitoring. If the water, the system must resume routine
requirements and MRDLs for chlorine LRAA at any location exceeds either monitoring.


Population size cat- Monitoring fre-
Source water type Distribution system monitoring location per monitoring period
egory quency 1
wwhite on PROD1PC65 with RULES2

Subpart H:
<500 ...................... ........................... Monitoring may not be reduced.

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00041 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
428 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations


Population size cat- Monitoring fre-
Source water type Distribution system monitoring location per monitoring period
egory quency 1

500–3,300 ............. per year ................. 1 TTHM and 1 HAA5 sample: one at the location and during the quarter with
the highest TTHM single measurement, one at the location and during the
quarter with the highest HAA5 single measurement; 1 dual sample set per
year if the highest TTHM and HAA5 measurements occurred at the same
location and quarter.
3,301–9,999 .......... per year ................. 2 dual sample sets: one at the location and during the quarter with the highest
TTHM single measurement, one at the location and during the quarter with
the highest HAA5 single measurement.
10,000–49,999 ...... per quarter ............ 2 dual sample sets at the locations with the highest TTHM and highest HAA5
50,000–249,999 .... per quarter ............ 4 dual sample sets—at the locations with the two highest TTHM and two high-
est HAA5 LRAAs.
250,000–999,999 .. per quarter ............ 6 dual sample sets—at the locations with the three highest TTHM and three
highest HAA5 LRAAs
1,000,000– per quarter ............ 8 dual sample sets—at the locations with the four highest TTHM and four
4,999,999. highest HAA5 LRAAs.
≥5,000,000 ............ per quarter ............ 10 dual sample sets—at the locations with the five highest TTHM and five
highest HAA5 LRAAs.
Ground Water:
<500 ...................... every third year ..... 1 TTHM and 1 HAA5 sample: one at the location and during the quarter with
the highest TTHM single measurement, one at the location and during the
quarter with the highest HAA5 single measurement; 1 dual sample set per
year if the highest TTHM and HAA5 measurements occurred at the same
location and quarter.
500–9,999 ............. per year ................. 1 TTHM and 1 HAA5 sample: one at the location and during the quarter with
the highest TTHM single measurement, one at the location and during the
quarter with the highest HAA5 single measurement; 1 dual sample set per
year if the highest TTHM and HAA5 measurements occurred at the same
location and quarter.
10,000–99,999 ...... per year ................. 2 dual sample sets: one at the location and during the quarter with the highest
TTHM single measurement, one at the location and during the quarter with
the highest HAA5 single measurement.
100,000–499,999 .. per quarter ............ 2 dual sample sets; at the locations with the highest TTHM and highest HAA5
≥500,000 ............... per quarter ............ 4 dual sample sets at the locations with the two highest TTHM and two high-
est HAA5 LRAAs.
1 Systems on quarterly monitoring must take dual sample sets every 90 days.

ii. Compliance determination. A PWS the wholesale system fails to monitor, the impacts of using the population-
is in compliance when the annual the consecutive system is in violation based approach.
sample or LRAA of quarterly samples is because it has the legal responsibility The plant-based approach was
less than or equal to the MCLs. If an for monitoring under State/EPA adopted from the 1979 TTHM rule and
annual sample exceeds the MCL, the regulations. the Stage 1 DBPR and was derived from
system must conduct increased • If a wholesale system has a the generally valid assumption that, as
(quarterly) monitoring but is not violation and provides that water to a systems increase in size, they tend to
immediately in violation of the MCL. consecutive system, the wholesale have more plants and increased
The system is out of compliance if the system is in violation. Whether the complexity. During the development of
LRAA of the quarterly samples for the consecutive system is in violation will the Stage 2 proposal, EPA identified a
past four quarters exceeds the MCL. depend on the situation. The number of issues associated with the
Monitoring and MCL violations are consecutive system will also be in use of the plant-based monitoring
assigned to the PWS where the violation violation unless it conducted approach. These included: (1) Plant-
occurred. Several examples are as monitoring that showed that the based monitoring is not as effective as
follows: violation was not present in the population-based monitoring in
• If monitoring results in a consecutive system. targeting locations with the highest risk;
consecutive system indicate an MCL (2) a plant-based approach can result in
violation, the consecutive system is in 2. Background and Analysis disproportionate monitoring
violation because it has the legal EPA proposed the plant-based requirements for systems serving the
responsibility for complying with the approach for all systems that produce same number of people (due to widely
MCL under State/EPA regulations. The some or all of their finished water and varying numbers of plants per system);
consecutive system may set up a the population-based monitoring (3) it cannot be adequately applied to
contract with its wholesale system that approach for systems purchasing all of plants or consecutive system entry
wwhite on PROD1PC65 with RULES2

details water quality delivery their finished water year-round. As part points that are operated seasonally or
specifications. of the proposal, EPA presented a intermittently if an LRAA is used for
• If a consecutive system has hired its monitoring cost analysis for applying compliance due to complex
wholesale system under contract to this approach to all systems in the implementation and a need for repeated
monitor in the consecutive system and Economic Analysis to better understand transactions between the State and

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00042 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 429

system to determine whether and how Compared to the 1995 CWSS, the 2000 (serving fewer than 10,000 people) and
compliance monitoring requirements CWSS contained questions more the ranges have been modified to be
may need to be changed; (4) State relevant for determining the number of consistent with those for other existing
determinations of monitoring plants in each system. Based on 2000 rules (such as the Lead and Copper
requirements for consecutive systems CWSS data, EPA has modified the Rule). This change will reduce
would be complicated, especially in number of monitoring sites per system implementation transactional costs. For
large combined distribution systems for several categories (particularly for medium and large subpart H systems
with many connections between the larger subpart H systems) to align (serving at least 10,000 people), EPA has
systems; and (5) systems with multiple the median population-based gone from seven categories in the
disinfecting wells would have to monitoring requirements with the proposal to five categories in final rule.
conduct evaluation of common aquifers median monitoring requirements under The population groups are sized so that
in order to avoid taking unnecessary plant-based monitoring, as was the ratio of maximum population to
samples for compliance (if they did not proposed. minimum population for each of the
conduct such evaluations under Stage EPA also believes that more samples categories is consistent. EPA believes
1). EPA requested comment on two are necessary to characterize larger that this will allow most systems to
approaches to address these issues: (1) systems (as defined by population) than remain in one population size category
keep the plant-based monitoring for smaller systems. This progressive and maintain the same monitoring
approach and add new provisions to approach is included in Table IV.G–4. requirements within a reasonable range
address specific concerns; and (2) base As system size increases, the number of of population variation over time. In
monitoring requirements on source samples increases to better reflect the addition, it assures that systems within
water type and population served, in hydraulic complexity of these systems. a size category will not have disparate
lieu of plant-based monitoring. While the national monitoring burden monitoring burdens as could occur if
under the population-based approach is there were too few categories. Overall,
The final rule’s requirements of slightly less than under a plant-based
population-based monitoring for all EPA believes that the population-based
approach, some larger systems with few monitoring approach allows systems to
systems are based on improved public plants relative to system population will
health protection, flexibility, and have more flexibility to designate their
take more samples per system than they
simplified implementation. For monitoring sites within the distribution
had under plant-based monitoring.
determining monitoring requirements, system to better target high DBP levels
However, EPA believes that many of
EPA’s objective was to maintain and is more equitable.
these large systems with few plants have
monitoring loads consistent with Stage traditionally been undermonitored (as To derive the number of monitoring
1 and similar to monitoring loads noted in the proposal). Systems with sites for IDSE standard monitoring, EPA
proposed for Stage 2 under a plant- more plants will see a reduction in doubled the number of routine
based approach, using a population- monitoring (e.g., small ground water compliance monitoring sites per system
based approach to facilitate systems with multiple wells). for each size category. This is consistent
implementation, better target high DBP While population-based monitoring with the advice and recommendations
levels, and protect human health. This requirements for ground water systems of the M-DBP Advisory Committee for
leads to a more cost-effective in today’s rule remain the same as those the IDSE. EPA has developed the Initial
characterization of where high levels in the proposed rule, the final rule Distribution System Evaluation
occur. For the proposed rule, EPA used consolidates ten population categories Guidance Manual for the Final Stage 2
1995 CWSS data to derive the number for subpart H systems into eight Disinfectants and Disinfection
of plants per system for calculating the categories for ease of implementation. Byproducts Rule (USEPA 2006) to assist
number of proposed monitoring sites As indicated in Table IV.G–4, EPA has systems in choosing IDSE monitoring
per system. During the comment period, gone from four to three population size locations, including criteria for selecting
2000 CWSS data became available. categories for smaller subpart H systems monitoring.
Plant-based Number of plants per sys- Calculated number of sites
approach* tem (Based on 2000 per system for plant-based Number of
Ratio of Number of CWSS data) approach monitoring
maximum sampling sites per
Population category population periods per Based on Based on system for
to minimum # Sites per
year median # mean # pop-based
population plant Median Mean plants per plants per approach
system system


<500 .................................................................. .................... 1 **1 1 1.21 1 1.2 **1

500–3,300 ......................................................... 6.6 4 **1 1 1.22 1 1.2 **1
3,301–9,999 ...................................................... 3 4 2 1 1.56 2 3.1 2
10,000–49,999 .................................................. 5 4 4 1 1.37 4 5.5 4
50,000–249,999 ................................................ 5 4 4 1 1.83 4 7.3 8
250,000–<1 million ............................................ 4 4 4 2 2.53 8 10.1 12
1 million–<5 million ........................................... 5 4 4 4 3.62 16 14.5 16
≥5 million ........................................................... .................... 4 4 4 4.33 16 17.3 20
wwhite on PROD1PC65 with RULES2

* As in the proposal.
** System is required to take individual TTHM and HAA5 samples at the locations with the highest TTHM and HAA5 concentrations, respectively, if highest TTHM
and HAA5 concentrations do not occur at the same location.

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00043 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
430 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations

Note: To determine the number of routine compliance monitoring sites per population category, EPA took these steps: (1) Maintaining about the same sampling
loads in the nation as required under the plant-based approach, but basing on population rather than number of plants to better target high DBP levels in distribution
systems and facilitate implementation; (2) The number of monitoring sites per plant under the plant-based approach (Column B) were multiplied by the number of
plants per system (Columns C and D) to calculate the number of monitoring sites per system under the plant-based approach (Columns E and F in terms of median
and mean, respectively); and (3) The number of monitoring sites per system under the population-based approach were derived with adjustments to keep categories
consistent and to maintain an even incremental trend as the population size category increases (Column G).

3. Summary of Major Comments for complicated distribution system eases the complexity by specifying
EPA received significant support for relationships, such as where minimum system-level requirements;
applying the population-based approach neighboring systems buy from and sell simplicity is essential for meeting the
to all systems. EPA also received to each other regularly throughout the implementation schedule in today’s
comments concerning the specific year. In this case, water may pass rule. If monitoring requirements were
requirements in a population-based through multiple consecutive systems determined by the combined
approach. before it reaches a user. Another distribution system population, many
Excessive Sampling Requirements. example would be a large group of implementation problems would occur.
Several commenters believed that the interconnected systems that have a Some of these problems would have the
proposed sampling requirements were complicated combined distribution potential to impact public health
excessive (especially in the larger system. This approach also allows the protection. For example, States or
population categories for subpart H combined distribution system to primacy agencies would have to decide
systems) and that some individual concentrate IDSE and Stage 2 how to allocate IDSE distribution
systems would be required to sample monitoring sites in the system with the system samples (where and how much
more under the population-based highest known DBP concentrations, to monitor in individual PWSs) in a
approach than the plant-based while assigning fewer sample sites to complicated combined distribution
approach. EPA recognizes that a small systems with low DBP concentrations. system with many systems, multiple
fraction of systems in some categories Population Size Categories. Some sources, multiple treatment plants, and
will have to take more samples under commenters recommended fewer varying water demand and with limited
the population-based approach than the population categories for subpart H understanding of DBP levels throughout
plant-based approach because their systems (those using surface water or the combined distribution system. This
number of plants is substantially less ground water under the direct influence would have to happen shortly after rule
than the national median or mean. of surface water as a source) than promulgation in order to meet the
However, the number of samples proposed while others recommended schedule. For example, some
required under the Stage 1 DBPR for more. Today’s rule has fewer categories consecutive systems buy water
these systems may not have been than proposed. However, EPA believes seasonally (in times of high water
sufficient to determine the that further reduction of the number of demand) or buy from more than one
concentrations of DBPs throughout the population size categories will not wholesale system (with the volume
distribution system of these systems. On reflect the fact that the number of plants purchased based on many factors). The
the other hand, systems with many and complexity of distribution systems State or primacy agency would find it
plants may have taken excessive (and DBP exposure) tend to increase as difficult to properly assign a limited
samples under the Stage 1 DBPR that the population served increases. As a number of IDSE monitoring locations
were not necessary to appropriately result, the population served by a large (especially since there are States where
determine DBP levels throughout the system in one particular category would many consecutive systems have no DBP
distribution system. Consequently, the receive much less protection from the data) to adequately reflect DBP levels in
total number of samples taken DBP risks than a smaller system in the such a system, as well as throughout the
nationally will be comparable to the same size category. On the other hand, combined distribution system.
Stage 1 DBPR, but will better target DBP too many categories with smaller EPA believes that assigning
risks in individual distribution systems. population ranges would result in compliance monitoring requirements
Consecutive systems. Some frequent category and requirement shifts appropriately throughout the combined
commenters noted that a consecutive as population fluctuates. Much greater distribution system requires a case-by-
system may need to take more samples implementation effort would be needed case determination based on factors
than its associated wholesale system. for those systems without much benefit such as amount and percentage of
Under today’s rule, all systems, in DBP exposure knowledge. finished water provided; whether
including consecutive systems, must Population Definition. Some finished water is provided seasonally,
monitor based on retail population commenters supported use of the intermittently, or full-time; and
served. Thus, large consecutive systems population of a combined distribution improved DBP occurrence information.
will take more samples than a smaller system (i.e., the wholesale and Since the IDSE will provide improved
wholesale system. The population-based consecutive systems should be DBP occurrence information throughout
monitoring approach will allow the considered a single system for the combined distribution system,
samples to better represent the DBP monitoring purposes) while others States may consider modifications to
concentrations consumed by the preferred use of the retail population for Stage 2 compliance monitoring
population associated with the sampling each individual system (i.e., wholesale requirements for consecutive systems on
locations and to understand the DBP systems and consecutive systems are a case-by-case basis as allowed by
concentrations reaching consumers. considered separately). Today’s final § 141.29 or under the special primacy
There is also a provision that allows rule uses the retail population for each condition at § 142.16(m)(3) by taking all
States to specify alternative monitoring individual system. EPA chose this these factors into consideration. In
wwhite on PROD1PC65 with RULES2

requirements for a consecutive system approach for today’s rule because of the making these case-by-case
in a combined distribution system (40 complexity involved in making determinations, the State will be able to
CFR 142.16(m)(3)). This special primacy implementation decisions for use its system-specific knowledge, along
condition allows the State to establish consecutive systems. Using the retail with the IDSE results, to develop an
monitoring requirements that account population to determine requirements appropriate monitoring plan for each

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00044 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 431

system within the combined 1. Today’s Rule operational evaluation does not extend
distribution system. Today’s rule defines the Stage 2 DBP the schedule (90 days after notification
Changes to monitoring plans. operational evaluation levels that of the analytical result) for submitting
Commenters requested more specific require systems to conduct operational the operational evaluation report.
language regarding how IDSE and Stage evaluations. The Stage 2 DBP 2. Background and Analysis
2 monitoring plans should be updated operational evaluation levels are
as a result of treatment or population The Stage 2 DBPR proposal outlined
identified using the system’s Stage 2 three components of the requirements
changes in the distribution system. DBPR compliance monitoring results.
Changes to IDSE plans should not be for significant excursions (definition,
The operational evaluation levels for system evaluation and excursion
necessary since the State or primacy each monitoring location are
agency will have reviewed those plans report). In response to public comments,
determined by the sum of the two the term ‘‘significant excursion’’ has
shortly before the system must conduct previous quarters’ TTHM results plus
the IDSE and the reviewed plan should been replaced by the term ‘‘operational
twice the current quarter’s TTHM result, evaluation level’’ in today’s rule. The
identify such issues. EPA provided a at that location, divided by 4 to
process in the Stage 2 DBPR proposal evaluation and report components
determine an average and the sum of the remain the same as those outlined in the
for updating monitoring plans for two previous quarters’ HAA5 results
systems that have significant changes to proposed rule for significant excursions.
plus twice the current quarter’s HAA5 However, the scope of the evaluation
treatment or in the distribution system result, at that location, divided by 4 to
after they complete their IDSE. This and report components of the
determine an average. If the average operational evaluation has also been
process remains in today’s rule, with an TTHM exceeds 0.080 mg/L at any
added requirement that systems must modified from the proposed significant
monitoring location or the average excursion evaluation components based
consult with the State or primacy HAA5 exceeds 0.060 mg/L at any on public comments.
agency to determine whether the monitoring location, the system must In the Stage 2 DBPR proposal, States
changes are necessary and appropriate conduct an operational evaluation and were to define criteria to identify
prior to implementing changes to their submit a written report of the significant excursions rather than using
Stage 2 monitoring plan. operational evaluation to the State. criteria defined by EPA. Concurrent
In addition, the State or primacy Operational evaluation levels with the Stage 2 DBPR proposal, EPA
agency may require a system to revise (calculated at each monitoring location) issued draft guidance (USEPA 2003e)
its IDSE plan, IDSE report, or Stage 2 IF (Q1 + Q2 + 2Q3)/4> MCL, then the for systems and States that described
monitoring plan at any time. This system must conduct an operational how to determine whether a significant
change was made so that systems could evaluation excursion has occurred, using several
receive system-specific guidance from where: different options. The rule proposal
the State or primacy agency on the specifically requested public comment
appropriate revisions to the Stage 2 Q3 = current quarter measurement
Q2 = previoius quarter measurement on the definition of a significant
monitoring plan. Regulatory language excursion, whether it should be defined
regarding changes that might occur is Q1 = quarter before previous quarter
measurement by the State or nationally, and the scope
not appropriate because any of the evaluation.
modifications would be system-specific MCL = Stage 2 MCL for TTHM (0.080 After reviewing comments on the
and a national requirement is not mg/l) or Stage 2 MCL for HAA5 (0.060 Stage 2 DBPR proposal, EPA determined
capable of addressing these system- mg/L) that DBP levels initiating an operational
specific issues. The operational evaluation includes evaluation should be defined in the
an examination of system treatment and regulation to ensure national
H. Operational Evaluation
distribution operational practices, consistency. Systems were concerned
Requirements Initiated by TTHM and
including changes in sources or source with the evaluation requirements being
HAA5 Levels
water quality, storage tank operations, initiated based on criteria that might not
A system that is in full compliance and excess storage capacity, that may be consistent nationally. Also, many
with the Stage 2 DBPR LRAA MCL may contribute to high TTHM and HAA5 States believed the requirement for
still have individual DBP measurements formation. Systems must also identify States to define criteria to initiate an
that exceed the Stage 2 DBPR MCLs, what steps could be considered to evaluation would be difficult for States
since compliance is based on individual minimize future operational evaluation to implement.
DBP measurements at a location level exceedences. In cases where the Under today’s rule, EPA is defining
averaged over a four-quarter period. system can identify the cause of DBP operational evaluation levels with an
EPA and the Advisory Committee were levels that resulted in the operational algorithm based on Stage 2 DBPR
concerned about these higher levels of evaluation, based on factors such as compliance monitoring results. These
DBPs. This concern was clearly water quality data, plant performance operational evaluation levels will act as
reflected in the Agreement in Principle, data, and distribution system an early warning for a possible MCL
which states, ‘‘. . . significant configuration the system may request violation in the following quarter. This
excursions of DBP levels will sometimes and the State may allow limiting the early warning is accomplished because
occur, even when systems are in full evaluation to the identified cause. The the operational evaluation requirement
compliance with the enforceable State must issue a written determination is initiated when the system assumes
MCL. . .’’. approving limiting the scope of the that the current quarter’s result is
Today’s final rule addresses this operational evaluation. The system must repeated and this will result in an MCL
concern by requiring systems to conduct submit their operational evaluation violation. This early identification
wwhite on PROD1PC65 with RULES2

operational evaluations that are initiated report to the State for review within 90 allows the system to act to prevent the
by operational evaluation levels days after being notified of the violation.
identified in Stage 2 DBPR compliance analytical result that initiates the Today’s rule also modifies the scope
monitoring and to submit an operational operational evaluation. Requesting of an operational evaluation. EPA has
evaluation report to the State. approval to limit the scope of the concluded that the source of DBP levels

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00045 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
432 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations

that would initiate an operational DBP levels that might not be warranted distribution systems, as in the proposal.
evaluation can potentially be linked to rather than on system operational issues Others felt that the treatment processes
a number of factors that extend beyond and compliance with Stage 2 DBPR should be included in the evaluation,
distribution system operations. MCLs. noting that these can be significant in
Therefore, EPA believes that evaluations Basis. The proposed requirements for the formation of DBPs.
must include a consideration of significant excursion evaluations were The Agency agrees with commenters
treatment plant and other system not based upon health effects, but rather that treatment processes can be a
operations rather than limiting the were intended to be an indicator of significant factor in DBP levels initiating
operational evaluation to only the operational performance. To address an operational evaluation and that a
distribution system, as proposed. commenter’s concerns and to emphasize comprehensive operational evaluation
Because the source of the problem could what EPA believes should initiate a should address treatment processes. In
be associated with operations in any of comprehensive evaluation of system cases where the system can clearly
these system components (or more than operations that may result in elevated identify the cause of the DBP levels
one), an evaluation that provides DBP levels and provide a proactive initiating an operational evaluation
systems with valuable information to procedure to address compliance with (based on factors such as water quality
evaluate possible modifications to Stage 2 DBP LRAA MCLs , EPA has data, plant performance data,
current operational practices (e.g. water replaced the term ‘‘significant distribution system configuration, and
age management, source blending) or in excursion’’ used in the Stage 2 DBPR previous evaluations) the State may
planning system modifications or proposal with the term ‘‘operational allow the system to limit the scope of
improvements (e.g. disinfection evaluation level’’ in today’s rule. the evaluation to the identified cause. In
practices, tank modifications, Definition of the operational other cases, it is appropriate to evaluate
distribution looping) will reduce DBP evaluation levels. The majority of the entire system, from source through
levels initiating an operational commenters stated that EPA should treatment to distribution system
evaluation. EPA also believes that State define the DBP levels initiating an configuration and operational practices.
review of operational evaluation reports operational evaluation (‘‘significant Timing for completion and review of
is valuable for both States and systems excursion’’ in the proposal) in the the evaluation report. While some
in their interactions, particularly when regulation to ensure national commenters agreed that the evaluation
systems may be in discussions with or consistency rather than requiring States report should be reviewed as part of the
requesting approvals from the State for to develop their own criteria (as was sanitary survey process (as proposed),
system improvements. Timely reviews proposed). Commenters suggested many commenters felt that the time
of operational evaluation reports will be several definitions, including a single between sanitary surveys (up to five
valuable for States in reviewing other numerical limit and calculations years) minimized the value of the
compliance submittals and will be comparing previous quarterly DBP evaluation report in identifying both the
particularly valuable in reviewing and results to the current quarter’s result. causes of DBP levels initiating an
approving any proposed source, Commenters that recommended a single operational evaluation and in possible
treatment or distribution system numerical limit felt that such an changes to prevent recurrence.
modifications for a water system. Under approach was justified by the available Moreover, a number of commenters felt
today’s rule, systems must submit a health effects information, while other that the evaluation report was important
written report of the operational commenters felt available heath effects enough to warrant a separate submittal
evaluation to the State no later than 90 information did not support a single and State review rather than have the
days after being notified of the DBP numerical limit. Commenters evaluation report compete with other
analytical result initiating an recommended that any definition be priorities during a sanitary survey.
operational evaluation. The written easy to understand and implement. The Agency agrees that completion
operational evaluation report must also EPA agrees with commenter and State review of evaluation reports
be made available to the public upon preference for national criteria to on a three or five year sanitary survey
request. initiate an operational evaluation. The cycle, when the focus of the evaluation
DBP levels initiating an operational is on what may happen in the next
3. Summary of Major Comments evaluation in today’s rule consider quarter, would allow for an
EPA received comments both in favor routine operational variations in unreasonable period of time to pass
of and opposed to the proposed distribution systems, are simple for between the event initiating the
evaluation requirements. While some water systems to calculate, and operational evaluation and completion
commenters felt that the evaluation minimize the implementation burden and State review of the report. This
requirements should not be a part of the on States. They also provide an early would diminish the value of the
Stage 2 DBPR until there was more warning to help identify possible future evaluation report for both systems and
information regarding potential health MCL violations and allow the system to States, particularly when systems may
effects correlated to specific DBP levels, take proactive steps to remain in be in discussions with or requesting
other commenters felt that the existing compliance. EPA emphasizes, as it did approval for treatment changes from
health effects data were sufficient to in the proposal and elsewhere in this States, and as noted above, the focus of
warrant strengthening the proposed notice, that health effects research is the report is on what may occur in the
requirements for an evaluation. Today’s insufficient to identify a level at which next quarter. EPA believes that timely
final rule requirements are consistent health effects occur, and thus today’s reviews of evaluation reports by States
with the Agreement in Principle methodology for initiating operational is important, would be essential for
recommendations. evaluation is not based upon health States in understanding system
Some commenters noted that health effects, but rather is intended as an operations and reviewing other
wwhite on PROD1PC65 with RULES2

effects research on DBPs is insufficient indicator of operational performance. compliance submittals, and would be
to identify a level at which health Scope of an evaluation. Some extremely valuable in reviewing and
effects occur and were concerned that commenters felt that the scope of an approving any proposed source,
the proposed significant excursion evaluation initiated by locational DBP treatment or distribution system
requirements placed an emphasis on levels should be limited to the modifications for a water system.

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00046 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 433

Having the evaluation information on an utilization of alternative technologies, 2. Background and Analysis
ongoing basis rather than a delayed such as UV, and whether the risk/risk EPA requested comment on including
basis would also allow States to concerns reflected in today’s rule, as language in the proposed rule
prioritize their resources in scheduling well as in the LT2ESWTR, remain valid. concerning potential reproductive and
and reviewing particular water system b. Criterion for reduced bromate developmental health effects. EPA
operations and conditions as part of any monitoring. Because more sensitive believes this is an important issue
on-site system review or oversight. bromate methods are now available, because of the large population exposed
Therefore, today’s rule requires that EPA is requiring a new criterion for (58 million women of child-bearing age;
systems complete the operational reduced bromate monitoring. In the USEPA 2005a) and the number of
evaluation and submit the evaluation Stage 1 DBPR, EPA required ozone studies that, while not conclusive, point
report to the State within 90 days of the systems to demonstrate that source towards a potential risk concern. While
occurrence. water bromide levels, as a running EPA is not including information about
annual average, did not exceed 0.05 mg/ reproductive and developmental health
I. MCL, BAT, and Monitoring for
L. EPA elected to use bromide as a effects in public notices at this time, the
surrogate for bromate in determining Agency plans to reconsider whether to
1. Today’s Rule eligibility for reduced monitoring include this information in the future.
Today EPA is confirming that the because the available analytical method As part of this effort, EPA intends to
MCL for bromate for systems using for bromate was not sensitive enough to support research to assess
ozone remains at 0.010 mg/L as an RAA quantify levels well below the bromate communication strategies on how to
for samples taken at the entrance to the MCL of 0.010 mg/L. best provide this information.
distribution system as established by the EPA approved several new analytical The responsibilities for public
Stage 1 DBPR. Because the MCL remains methods for bromate that are far more notification and consumer confidence
the same, EPA is not modifying the sensitive than the existing method as reports rest with the individual system.
existing bromate BAT. EPA is changing part of today’s rule. Since these methods Under the Public Notice Rule (Part 141
the criterion for a system using ozone to can measure bromate to levels of 0.001 subpart Q) and Consumer Confidence
qualify for reduced bromate monitoring mg/L or lower, EPA is replacing the Report Rule (Part 141 subpart O), the
from demonstrating low levels of criterion for reduced bromate wholesale system is responsible for
bromide to demonstrating low levels of monitoring (source water bromide notifying the consecutive system of
bromate. running annual average not to exceed analytical results and violations related
0.05 mg/L) with a bromate running to monitoring conducted by the
2. Background and Analysis annual average not to exceed 0.0025 mg/ wholesale system. Consecutive systems
a. Bromate MCL. Bromate is a L. are required to conduct appropriate
principal byproduct from ozonation of In the past, EPA has often set the public notification after a violation
bromide-containing source waters. As criterion for reduced monitoring (whether in the wholesale system or the
described in more detail in the Stage 2 eligibility at 50% of the MCL, which consecutive system). In their consumer
DBPR proposal (USEPA 2003a), more would be 0.005 mg/L. However, the confidence report, consecutive systems
stringent bromate MCL has the potential MCL for bromate will remain at 0.010 must include results of the testing
to decrease current levels of microbial mg/L, representing a risk level of 2×10/ conducted by the wholesale system
protection, impair the ability of systems b 2×10¥4, 10¥4 and 10¥6 (higher than unless the consecutive system
to control resistant pathogens like EPA’s usual excess cancer risk range of conducted equivalent testing (as
Cryptosporidium, and increase levels of 10¥4 to 10¥6) because of risk tradeoff required in today’s rule) that indicated
DBPs from other disinfectants that may considerations) (USEPA 2003a). the consecutive system was in
be used instead of ozone. EPA EPA believes that the decision for compliance, in which case the
considered reducing the bromate MCL reduced monitoring is separate from consecutive system reports its own
from 0.010 mg/L to 0.005 mg/L as an these risk tradeoff considerations. Risk compliance monitoring results.
annual average but concluded that many tradeoff considerations influence the
systems using ozone to inactivate selection of the MCL, while reduced 3. Summary of Major Comments
microbial pathogens would have monitoring requirements are designed to EPA requested and received many
significant difficulty maintaining ensure that the MCL, once established, comments on the topic of including
bromate levels at or below 0.005 mg/L. is reliably and consistently achieved. public notification language regarding
In addition, because of the high doses Requiring a running annual average of potential reproductive and
required, the ability of systems to use 0.0025 mg/L for the reduced monitoring developmental effects. A number of
ozone to meet Cryptosporidium criterion allows greater confidence that comments called for including
treatment requirements under the the system is achieving the MCL and reproductive and developmental health
LT2ESWTR would be diminished if the thus ensuring public health protection. effects language to address the potential
bromate MCL was decreased from 0.010 3. Summary of Major Comments health concerns that research has
to 0.005 mg/L; higher doses will shown. Numerous comments also
generally lead to greater bromate Commenters supported both the opposed such language due to
formation. After evaluation under the retention of the existing bromate MCL uncertainties in the underlying science
risk-balancing provisions of section and the modified reduced monitoring and the implications such language
1412(b)(5) of the SDWA, EPA concluded criterion. could have on public trust of utilities.
that the existing MCL was justified. EPA J. Public Notice Requirements EPA agrees on the importance of
will review the bromate MCL as part of addressing possible reproductive and
1. Today’s Rule
wwhite on PROD1PC65 with RULES2

the six-year review process and developmental health risks. However,

determine whether the MCL should Today’s rule does not alter existing given the uncertainties in the science
remain at 0.010 mg/L or be reduced to public notification language for TTHM, and our lack of knowledge of how to
a lower level. As a part of that review, HAA5 or TOC, which are listed under best communicate undefined risks, a
EPA will consider the increased 40 CFR 141.201–141.210 (Subpart Q). general statement about reproductive

VerDate Aug<31>2005 21:46 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00047 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
434 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations

and developmental health effects is exemptions in accordance with section EPA has determined that affordable
premature at this time. The Agency 1416(a) of the SDWA and EPA’s compliance technologies are available.
needs to understand how best to regulations. The 1996 Amendments to the SDWA
characterize and communicate these identify three categories of small public
risks and what to do to follow up any 2. Background and Analysis water systems that need to be addressed:
such communication. The public a. Variances. The SDWA provides for (1) Those serving a population of 3301–
deserves accurate, timely, relevant, and two types of variances—general 10,000; (2) those serving a population of
understandable communication. The variances and small system variances. 500–3300; and (3) those serving a
Agency will continue to follow up on Under section 1415(a)(1)(A) of the population of 25–499. The SDWA
this issue with additional research, SDWA, a State that has primary requires EPA to make determinations of
possibly including a project to work enforcement responsibility (primacy), or available compliance technologies for
with stakeholders to assess risk EPA as the primacy agency, may grant each size category. A compliance
communication strategies. general variances from MCLs to those technology is a technology that is
Some comments also suggested affordable and that achieves compliance
public water systems of any size that
leaving the choice of language up to the with the MCL and/or treatment
cannot comply with the MCLs because
water server. EPA believes that this technique. Compliance technologies can
of characteristics of the raw water
strategy would cause undue confusion include point-of-entry or point-of-use
sources. The primacy agency may grant
to both the PWS and the public. treatment units. Variance technologies
Commenters generally agreed that general variances to a system on
condition that the system install the best are only specified for those system size/
both wholesale and consecutive systems source water quality combinations for
that conduct monitoring be required to technology, treatment techniques, or
which there are no listed affordable
report their own analytical results as other means that EPA finds available
compliance technologies.
part of their CCRs. One commenter and based upon an evaluation
Using its current National
requested clarification of consecutive satisfactory to the State that indicates
Affordability Criteria, EPA has
system public notification requirements that alternative sources of water are not
determined that multiple affordable
when there is a violation in the reasonably available to the system. At
compliance technologies are available
wholesale system but the consecutive the time this type of variance is granted,
for each of the three system sizes
system data indicate that it meets DBP the State must prescribe a compliance
(USEPA 2005a), and therefore did not
MCLs. schedule and may require the system to
identify any variance treatment
Although EPA requires consecutive implement additional control measures.
technologies. The analysis was
systems to conduct appropriate public Furthermore, before EPA or the State
consistent with the current methodology
notification of violations (whether in the may grant a general variance, it must
used in the document ‘‘National-Level
wholesale or consecutive system), there find that the variance will not result in
Affordability Criteria Under the 1996
may be cases where the violation may an unreasonable risk to health (URTH) Amendments to the Safe Drinking Water
only affect an isolated portion of the to the public served by the public water Act’’ (USEPA 1998d) and the ‘‘Variance
distribution system. Under the public system. In today’s final rule, EPA is Technology Findings for Contaminants
notification rule, the State may allow specifying BATs for general variances Regulated Before 1996’’ (USEPA 1998e).
systems to limit distribution of the under section 1415(a) (see section IV.D). However, EPA is currently reevaluating
notice to the area that is out of Section 1415(e) authorizes the its national-level affordability criteria
compliance if the system can primacy agency to issue variances to and has solicited recommendations
demonstrate that the violation occurred small public water systems (those from both the NDWAC and the SAB as
in a part of the distribution system that serving fewer than 10,000 people) where part of this review. EPA intends to
is ‘‘physically or hydraulically isolated the primacy agent determines (1) that apply the revised criteria to the Stage 2
from other parts of the distribution the system cannot afford to comply with DBPR once they have been finalized for
system.’’ This provision remains in an MCL or treatment technique and (2) the purpose of determining whether to
place. As for a consecutive system that the terms of the variances will enable States to give variances. Thus,
whose wholesale system is in violation, ensure adequate protection of human while the analysis of Stage 2 household
the consecutive system is not required health (63 FR 43833, August 14, 1998) costs will not change, EPA’s
to conduct public notification if DBP (USEPA 1998c). These variances may determination regarding the availability
levels in the consecutive system are in only be granted where EPA has of affordable compliance technologies
compliance. determined that there is no affordable for the different categories of small
K. Variances and Exemptions compliance technology and has systems may.
identified a small system variance b. Affordable Treatment Technologies
1. Today’s Rule technology under section 1412(b)(15) for for Small Systems. The treatment trains
States may grant variances in the contaminant, system size and source considered and predicted to be used in
accordance with sections 1415(a) and water quality in question. As discussed EPA’s compliance forecast for systems
1415(e) of the SDWA and EPA’s below, small system variances under serving under 10,000 people, are listed
regulations. States may grant section 1415(e) are not available because in Table IV.K–1.

SW Water Plants GW Water Plants
wwhite on PROD1PC65 with RULES2

• Switching to chloramines as a residual disinfectant ............................. • Switching to chloramines as a residual disinfectant

• Chlorine dioxide (not for systems serving fewer than 100 people) ...... • UV
• UV ......................................................................................................... • Ozone (not for systems serving fewer than 100 people)
• Ozone (not for systems serving fewer than 100 people) ..................... • GAC20
• Micro-filtration/Ultra-filtration ................................................................. • Nanofiltration

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00048 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 435

SW Water Plants GW Water Plants

• GAC20.
• GAC20 + Advanced disinfectants.
• Integrated Membranes.
Note: Italicized technologies are those predicted to be used in the compliance forecast.
Source: Exhibits 5.11b and 5.14b, USEPA 2005a.

The household costs for these systems under the Stage 2 DBPR will be schedule or use excessive capacity to
technologies were compared against the the same as under the Stage 1 DBPR avoid installing a costly technology to
EPA’s current national-level (because many systems serving fewer comply with the Stage 2 DBPR. The
affordability criteria to determine the than 500 people will have the same system also may identify another water
affordable treatment technologies. The single sampling site under both rules); source that has lower TTHM and HAA5
Agency’s national level affordability these systems will have already precursor levels. Systems that can
criteria were published in the August 6, installed the necessary compliance identify such an alternate water source
1998 Federal Register (USEPA 1998d). technology to comply with the Stage 1 may not have to treat that new source
A complete description of how this DBPR. It is also possible that less costly water as intensely as their current
analysis was applied to Stage 2 DBPR is technologies such as those for which source, resulting in lower treatment
given in Section 8.3 of the Economic percentage use caps were set in the costs. Systems may elect to connect to
Analysis (USEPA 2005a). decision tree may actually be used to a neighboring water system. While
Of the technologies listed in Table achieve compliance (e.g., chloramines,
IV.K–1, integrated membranes with connecting to another system may not
UV). Thus, EPA believes that be feasible for some remote systems,
chloramines, GAC20 with advanced compliance by these systems will be
oxidants, and ozone are above the EPA estimates that more than 22 percent
affordability threshold in the 0 to 500 As shown in Table IV.K–2, the cost of all small water systems are located
category. No treatment technologies are model predicts that some households within metropolitan regions (USEPA
above the affordability threshold in the served by very small systems will 2000f) where distances between
500 to 3,300 category or the 3,300 to experience household cost increases neighboring systems will not present a
10,000 category. As shown in the greater than the available expenditure prohibitive barrier. Low-cost
Economic Analysis for systems serving margins as a result of adding advanced alternatives to reduce total
fewer than 500 people, 14 systems are technology for the Stage 2 DBPR trihalomethanes (TTHM) and haloacetic
predicted to use GAC20 with advanced (USEPA 2005a). This prediction may be acid (HAA5) levels also include
disinfectants, one system is predicted to overestimated because small systems distribution system modifications such
use integrated membranes, and no may have other compliance alternatives as flushing distribution mains more
systems are predicted to use ozone to available to them besides adding frequently, looping to prevent dead
comply with the Stage 2 DBPR (USEPA treatment, which were not considered in ends, and optimizing storage to
2005a). However, several alternate the model. For example, some of these minimize retention time. More
technologies are affordable and likely systems currently may be operated on a discussion of household cost increases
available to these systems. In some part-time basis; therefore, they may be is presented in Section VI.E and the
cases, the compliance data for these able to modify the current operational Economic Analysis (USEPA 2005a).
Number of Number of
Number of Number of Total num-
households surface
households groundwater ber of plants
served by water plants
with annual plants with with annual
plants add- 90th Per- 95th Per- with annual
Mean an- Median an- Available cost in- annual cost cost in-
Systems size ing treat- centile an- centile an- cost in-
nual house- nual house- expenditure creases increases creases
(population ment (Per- nual house- nual house- creases
hold cost in- hold cost in- margin ($/ greater than greater than greater than
seved) cent of all hold cost in- hold cost in- greater than
crease crease hh/yr) the avail- the avail- the avail-
households crease crease the avail-
able ex- able ex- able ex-
subject to able ex-
penditure penditure penditure
the Stage 2 penditure
margin margin margin
DBPR) margin
0–500 ................ 43045(3) $201.55 $299.01 $299.01 $414.74 $733 964 15 0 15
501–3,300 ......... 205842 (4) $58.41 $29.96 $75.09 $366.53 $724 0 9 0 0
3,301–10,000 .... 342525 (5) $37.05 $14.59 $55.25 $200.05 $750 0 0 0 0
Notes: Household unit costs represent treatment costs only. All values in year 2003 dollars.
Source: Exhibit 8.4c, USEPA 2005a.

c. Exemptions. Under section 1416(a), such as qualification of the PWS as effective date of the NPDWR, or for a
EPA or a State that has primary serving a disadvantaged community), system that was not in operation by that
wwhite on PROD1PC65 with RULES2

enforcement responsibility (primacy) the PWS is unable to comply with the date, only if no reasonable alternative
may exempt a public water system from requirement or implement measures to source of drinking water is available to
any requirements related to an MCL or develop an alternative source of water the new system; and (4) management or
treatment technique of an NPDWR, if it supply; (2) the exemption will not result restructuring changes (or both) cannot
finds that (1) due to compelling factors in an unreasonable risk to health; and; reasonably result in compliance with
(which may include economic factors (3) the PWS was in operation on the the Act or improve the quality of

VerDate Aug<31>2005 21:46 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00049 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
436 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations

drinking water. If EPA or the State to recognize and respond to problems. recommended Stage 2 compliance
grants an exemption to a public water Subpart H systems were required to be monitoring sites as part of the IDSE
system, it must at the same time operated by qualified operators under report.
prescribe a schedule for compliance the SWTR (§ 141.70). The Stage 1 DBPR Systems must report compliance with
(including increments of progress or added requirements for all disinfected Stage 2 TTHM and HAA5 MCLs (0.080
measures to develop an alternative systems to be operated by qualified mg/LTTHM and 0.060 mg/L HAA5, as
source of water supply) and personnel who meet the requirements LRAAs) according to the schedules
implementation of appropriate control specified by the State, which may differ specified in §§ 141.620 and 141.629 and
measures that the State requires the based on system size and type. The rule discussed in section IV.E of today’s
system to meet while the exemption is also requires that States maintain a preamble. Reporting for DBP
in effect. Under section 1416(b)(2)(A), register of qualified operators (40 CFR monitoring, as described previously,
the schedule prescribed shall require 141.130(c)). While the Stage 2 DBPR will remain generally consistent with
compliance as expeditiously as requirements do not supercede or current public water system reporting
practicable (to be determined by the modify the requirement that disinfected requirements (§ 141.31 and § 141.134);
State), but no later than 3 years after the systems be operated by qualified systems will be required to calculate
effective date for the regulations operators, such personnel play an and report each LRAA (instead of the
established pursuant to section important role in delivering drinking system’s RAA) and each individual
1412(b)(10). For public water systems water that meets Stage 2 MCLs to the monitoring result (as required under the
which do not serve more than a public. States should also review and Stage 1 DBPR). Systems will also be
population of 3,300 and which need modify, as required, their qualification required to provide a report to the State
financial assistance for the necessary standards to take into account new about each operational evaluation
improvements, EPA or the State may technologies (e.g., ultraviolet (UV) within 90 days, as discussed in section
renew an exemption for one or more disinfection) and new compliance IV.H. Reports and evaluations must be
additional two-year periods, but not to requirements (including simultaneous kept for 10 years and may prove
exceed a total of 6 years, if the system compliance and consecutive system valuable in identifying trends and
establishes that it is taking all requirements). EPA received only one recurring issues.
practicable steps to meet the comment on this topic; the commenter
requirements above. A public water 2. Summary of Major Comments
supported the need for a qualified
system shall not be granted an operator. EPA requested comment on all system
exemption unless it can establish that reporting and recordkeeping
either: (1) the system cannot meet the M. System Reporting and Recordkeeping requirements. Commenters generally
standard without capital improvements Requirements supported EPA’s proposed
that cannot be completed prior to the 1. Today’s Rule requirements, but expressed concern
date established pursuant to section about two specific issues. The first issue
1412(b)(10); (2) in the case of a system Today’s Stage 2 DBPR, consistent
with the existing system reporting and was the data management and tracking
that needs financial assistance for the difficulties that States would face if EPA
necessary implementation, the system recordkeeping regulations under 40 CFR
141.134 (Stage 1 DBPR), requires public finalized a monitoring approach which
has entered into an agreement to obtain had both plant-based and population-
financial assistance pursuant to section water systems (including consecutive
systems) to report monitoring data to based requirements, as was proposed.
1452 or any other Federal or state Since today’s rule contains only
program; or (3) the system has entered States within ten days after the end of
the compliance period. In addition, population-based monitoring
into an enforceable agreement to requirements, this concern is no longer
become part of a regional public water systems are required to submit the data
required in § 141.134. These data are an issue. See section IV.G in today’s
system. preamble for further discussion.
required to be submitted quarterly for
3. Summary of Major Comments any monitoring conducted quarterly or The second concern related to
more frequently, and within ten days of reporting associated with the IDSE.
Several commenters agreed with the Commenters who supported an
proposal not to list variances the end of the monitoring period for less
frequent monitoring. As with other approach other than the IDSE for
technologies for the Stage 2 DBPR. One determining Stage 2 compliance
commenter requested that EPA modify chemical analysis data, the system must
keep the results for 10 years. monitoring locations did not support
the methodology used to assess IDSE-related reporting. The IDSE
affordability. As mentioned earlier, EPA In addition to the existing Stage 1
reporting requirements, today’s rule remains a key component of the final
is currently reevaluating its national- rule; thus, EPA has retained IDSE-
level affordability criteria and has requires systems to perform specific
IDSE-related reporting to the primacy related reporting. However, the Agency
solicited recommendations from both has modified both the content and the
the NDWAC and the SAB as part of this agency, except for systems serving fewer
than 500 for which the State or primacy timing of the reporting to reduce the
review. EPA intends to apply the burden. See sections IV.F and IV.E,
revised criteria to the Stage 2 DBPR for agency has waived this requirement.
Required reporting includes submission respectively, of today’s preamble for
the purpose of determining whether to further discussion.
enable States to give variances. of IDSE monitoring plans, 40/30
certification, and IDSE reports. This N. Approval of Additional Analytical
L. Requirements for Systems to Use reporting must be accomplished on the Methods
Qualified Operators schedule specified in the rule (see
EPA believes that systems that must § 141.600(c)) and discussed in section 1. Today’s Rule
wwhite on PROD1PC65 with RULES2

make treatment changes to comply with IV.E of today’s preamble. System EPA is taking final action to:
requirements to reduce microbiological submissions must include the elements (1) allow the use of 13 methods
risks and risks from disinfectants and identified in subpart U and discussed published by the Standard Methods
disinfection byproducts should be further in section IV.F of today’s Committee in Standard Methods for the
operated by personnel who are qualified preamble. These elements include Examination of Water and Wastewater,

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00050 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 437

20th edition, 1998 (APHA 1998) and 12 sample preservation; and require that changes in the same sections of the CFR.
methods in Standard Methods Online. TOC samples be preserved at the time EPA decided to make all the changes to
(2) approve three methods published of collection. § 141.131 as part of the Stage 2 DBPR
by American Society for Testing and (7) clarify which methods are and the remainder of the methods that
Materials International. approved for magnesium hardness were proposed with the Stage 2 DBPR
(3) approve EPA Method 327.0 determinations (40 CFR 141.131 and will be considered as part of the
Revision 1.1 (USEPA 2005h) for daily 141.135). Methods Update Rule, which will be
monitoring of chlorine dioxide and finalized at a later date. Two ASTM
2. Background and Analysis
chlorite, EPA Method 552.3 (USEPA methods, D 1253–86(96) and D 1253–03,
2003f) for haloacetic acids (five) The Stage 1 Disinfectants and that were proposed in the Methods
(HAA5), EPA Methods 317.0 Revision 2 Disinfection Byproducts Rule (Stage 1 Update Rule, are being approved for
(USEPA 2001c) and 326.0 (USEPA 2002) DBPR) was promulgated on December measuring chlorine residual as part of
for bromate, chlorite, and bromide, EPA 16, 1998 (USEPA 1998a) and it included today’s action.
Method 321.8 (USEPA 2000g) for approved analytical methods for DBPs, Minor corrections have been made in
bromate only, and EPA Method 415.3 disinfectants, and DBP precursors. two of the methods that were proposed
Revision 1.1 (USEPA 2005l) for total Additional analytical methods became in the Stage 2 DBPR. In today’s rule, the
organic carbon (TOC) and specific available subsequent to the rule and Agency is approving EPA Method 327.0
ultraviolet absorbance (SUVA). were proposed in the Stage 2 (Revision 1.1, 2005) which corrects
(4) update the citation for EPA Disinfectants and Disinfection three typographical errors in the
Method 300.1 (USEPA 2000h) for Byproducts Rule (Stage 2 DBPR) proposed method.
bromate, chlorite, and bromide. (USEPA 2003a). These methods are EPA is also approving EPA Method
(5) standardize the HAA5 sample applicable to monitoring that is required 415.3 (Revision 1.1, 2005), which does
holding times and the bromate sample under the Stage 1 DBPR. After the Stage not contain the requirement that
preservation procedure and holding 2 DBPR proposal, analytical methods for samples for the analysis of TOC must be
time. additional drinking water contaminants received within 48 hours of sample
(6) add the requirement to remove were proposed for approval in a collection.
inorganic carbon prior to determining Methods Update Rule proposal (USEPA A summary of the methods that are
TOC or DOC, remove the specification 2004). The Stage 2 DBPR and Methods included in today’s rule is presented in
of type of acid used for TOC/DOC Update Rule proposals both included Table IV.N–1.


Standard methods 20th
Analyte EPA method Standard methods online Other

§ 141.131—Disinfection Byproducts

HAA5 .................................. 552.3 ................................. 6251 B .............................. 6251 B–94 ........................

Bromate ............................. 317.0, Revision 2.0 ........... ........................................... ........................................... ASTM D 6581–00
321.8 .................................
326.0 .................................
Chlorite (monthly or daily) 317.0, Revision 2.0 ........... ........................................... ........................................... ASTM D 6581–00
326.0 .................................
Chlorite (daily) .................... 327.0, Revision 1.1 ........... 4500–ClO2 E ..................... 4500–ClO2 E–00 ...............

§ 141.131—Disinfectants

Chlorine (free, combined, ........................................... 4500–Cl D ......................... 4500–Cl D–00 ................... ASTM D 1253–86(96)
total). 4500–Cl F ......................... 4500–Cl F–00 ................... ASTM D 1253–03
4500–Cl G ......................... 4500–Cl G–00 ...................
Chlorine (total) ................... ........................................... 4500–Cl E ......................... 4500–Cl E–00 ...................
4500–Cl I .......................... 4500–Cl I–00 ....................
Chlorine (free) .................... ........................................... 4500–Cl H ......................... 4500–Cl H–00 ...................
Chlorine Dioxide ................ 327.0, Revision 1.1 ........... 4500–ClO2 D ..................... 4500–ClO2 E–00 ...............
4500–ClO2 E .....................

§ 141.131—Other parameters

Bromide .............................. 317.0, Revision 2.0 ........... ........................................... ........................................... ASTM D 6581–00
326.0 .................................
TOC/DOC .......................... 415.3, Revision 1.1 ........... 5310 B .............................. 5310 B–00 ........................
5310 C .............................. 5310 C–00 ........................
5310 D .............................. 5310 D–00 ........................
UV254 .................................. 415.3, Revision 1.1 ........... 5910 B .............................. 5910 B–00 ........................
SUVA ................................. 415.3, Revision 1.1 ........... ........................................... ...........................................
wwhite on PROD1PC65 with RULES2

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00051 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
438 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations

O. Laboratory Certification and analyses to be conducted by approved PE acceptance criteria will be subject to
Approval parties. It revises the acceptance criteria the new criteria when it is time for them
1. PE Acceptance Criteria for performance evaluation (PE) studies to analyze their annual DBP PE
which laboratories must pass as part of sample(s). Today’s rule also requires
a. Today’s rule. Today’s rule the certification program. The new that TTHM and HAA5 analyses that are
maintains the requirements of acceptance limits are effective 60 days performed for the IDSE or system-
laboratory certification for laboratories
after promulgation. Laboratories that specific study be conducted by
performing analyses to demonstrate
were certified under the Stage 1 DBPR laboratories certified for those analyses.
compliance with MCLs and all other


limits (per-
DBP Comments
cent of true

Chloroform ............................................................................. ±20 Laboratory must meet all 4 individual THM acceptance limits in
order to successfully pass a PE sample for TTHM
Bromodichloromethane ......................................................... ±20
Dibromochloromethane ......................................................... ±20
Bromoform ............................................................................. ±20
Monochloroacetic Acid .......................................................... ±40 Laboratory must meet the acceptance limits for 4 out of 5 of the
HAA5 compounds in order to successfully pass a PE sample
for HAA5
Dichloroacetic Acid ................................................................ ±40
Trichloroacetic Acid ............................................................... ±40
Monobromoacetic Acid .......................................................... ±40
Dibromoacetic Acid ............................................................... ±40
Chlorite ......................................................................................... ±30
Bromate ........................................................................................ ±30

b. Background and analysis. The Stage DBP in the PE studies, so that confusion, EPA has modified the rule
1 DBPR (USEPA 1998a) specified that in laboratories would not be required to language to allow laboratories one year
order to be certified the laboratory must meet tighter criteria in the PE study than from today’s date to meet the new PE
pass an annual performance evaluation they are required to meet with the criteria.
(PE) sample approved by EPA or the minimum reporting level (MRL) check
State using each method for which the standard. EPA has addressed this 2. Minimum Reporting Limits
laboratory wishes to maintain concern by directing the PE sample a. Today’s rule. EPA is establishing
certification. The acceptance criteria for suppliers to use concentrations no less regulatory minimum reporting limits
the DBP PE samples were set as than 10 µg/L for the individual THM
(MRLs) for compliance reporting of
statistical limits based on the and HAAs, 100 µg/L for chlorite, and 7
DBPs by Public Water Systems. These
performance of the laboratories in each µg/L for bromate in PE studies used for
regulatory MRLs (Table IV.O–2) also
study. This was done because EPA did certifying drinking water laboratories.
not have enough data to specify fixed Two commenters requested that the define the minimum concentrations that
acceptance limits. effective date for the new PE acceptance must be reported as part of the
Subsequent to promulgation of the criteria be extended from 60 days to 180 Consumer Confidence Reports (40 CFR
Stage 1 DBPR, EPA was able to evaluate days, because they felt that 60 days was § 141.151(d)). EPA is incorporating
data from PE studies conducted during not enough time for laboratories to meet MRLs into the laboratory certification
the Information Collection Rule (USEPA the new criteria. EPA realized from program for DBPs by requiring
1996) and during the last five general those comments that the original intent laboratories to include a standard near
Water Supply PE studies. Based on the of the proposal was not clearly the MRL concentration as part of the
evaluation process as described in the explained; the 60 days was to be the calibration curve for each DBP and to
proposed Stage 2 DBPR (USEPA 2003a), deadline for when the PE providers verify the accuracy of the calibration
EPA determined that fixed acceptance must change the acceptance criteria that curve at the MRL concentration by
limits could be established for the DBPs. are used when the studies are analyzing an MRL check standard with
Today’s action replaces the statistical PE conducted. Laboratories would have to a concentration less than or equal to
acceptance limits with fixed limits meet the criteria when it is time for 110% of the MRL with each batch of
effective one year after promulgation. them to analyze their annual PE samples samples. The measured DBP
c. Summary of major comments. Four in order to maintain certification. concentration for the MRL check
commenters supported the fixed Depending upon when the last PE standard must be ±50% of the expected
acceptance criteria as presented in the sample was analyzed, laboratories could value, if any field sample in the batch
proposed rule. One requested that a have up to one year to meet the new has a concentration less than 5 times the
wwhite on PROD1PC65 with RULES2

minimum concentration be set for each criteria. In order to eliminate this regulatory MRL.

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00052 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 439


Minimum reporting level
DBP Comments
(mg/L) 1

Chloroform ...................................................... 0.0010
Bromodichloromethane .................................. 0.0010
Dibromochloromethane .................................. 0.0010
Bromoform ...................................................... 0.0010
HAA5 2
Monochloroacetic Acid ................................... 0.0020
Dichloroacetic Acid ......................................... 0.0010
Trichloroacetic Acid ........................................ 0.0010
Monobromoacetic Acid ................................... 0.0010
Dibromoacetic Acid ........................................ 0.0010
Chlorite .................................................................. 0.020 Applicable to monitoring as prescribed in § 141.132(b)(2)(i)(B)
and (b)(2)(ii).
Bromate ................................................................. 0.0050 or 0.0010 Laboratories that use EPA Methods 317.0 Revision 2.0, 326.0 or
321.8 must meet a 0.0010 mg/L MRL for bromate.
1 The calibration curve must encompass the regulatory minimum reporting level (MRL) concentration. Data may be reported for concentrations
lower than the regulatory MRL as long as the precision and accuracy criteria are met by analyzing an MRL check standard at the lowest report-
ing limit chosen by the laboratory. The laboratory must verify the accuracy of the calibration curve at the MRL concentration by analyzing an
MRL check standard with a concentration less than or equal to 110% of the MRL with each batch of samples. The measured concentration for
the MRL check standard must be ±50% of the expected value, if any field sample in the batch has a concentration less than 5 times the regu-
latory MRL. Method requirements to analyze higher concentration check standards and meet tighter acceptance criteria for them must be met in
addition to the MRL check standard requirement.
2 When adding the individual trihalomethane or haloacetic acid concentrations to calculate the TTHM or HAA5 concentrations, respectively, a
zero is used for any analytical result that is less than the MRL concentration for that DBP, unless otherwise specified by the State.

b. Background and analysis. EPA analyze the MRL check standard, but that are no longer effective. These
proposed to establish regulatory MRLs the laboratory is only required to meet sections have been superceded by new
for DBPs in order to define expectations the accuracy criteria (±50%) if a field requirements elsewhere in Part 141.
for reporting compliance monitoring sample has a concentration less than Sections 141.12 (Maximum
data to the Primacy Agencies and in the five times the regulatory MRL contaminant levels for total
Consumer Confidence Reports. The concentration. trihalomethanes) and 141.30 (Total
proposed MRLs were generally based on EPA proposed a regulatory MRL of trihalomethanes sampling, analytical
those used during the Information 0.200 mg/L for chlorite, because data and other requirements) were
Collection Rule (USEPA 1996), because from the Information Collection Rule promulgated as part of the 1979 TTHM
an analysis of the quality control data indicated that most samples would Rule. These sections have been
set from the Information Collection Rule contain concentrations greater than superceded in their entirety by § 141.64
(Fair et al. 2002) indicated that 0.200 mg/L (USEPA 2003c). EPA also (Maximum contaminant levels for
laboratories are able to provide took comment on a lower MRL of 0.020 disinfection byproducts) and subpart L
quantitative data down to those mg/L. Commenters were evenly divided (Disinfectant Residuals, Disinfection
concentrations. concerning which regulatory MRL Byproducts, and Disinfection Byproduct
concentration should be adopted in the Precursors), respectively, as of
EPA also proposed that laboratories final rule. EPA has decided to set the December 31, 2003. Also, § 141.32
be required to demonstrate ability to chlorite regulatory MRL at 0.020 mg/L (Public notification) has been
quantitate at the MRL concentrations by in today’s rule. This decision was based superceded by subpart Q (Public
analyzing an MRL check standard and on two factors. First, the approved Notification of Drinking Water
meeting accuracy criteria on each day analytical methods for determining Violations), which is now fully in effect.
that compliance samples are analyzed. compliance with the chlorite MCL can Section 553 of the Administrative
Three public commenters noted that easily support an MRL of 0.020 mg/L. Procedure Act, 5 U.S.C. 553(b)(B),
meeting the accuracy requirement for More importantly, since the proposal, provides that, when an agency for good
the MRL check standard did not EPA has learned that water systems that cause finds that notice and public
contribute to the quality of the data in have low chlorite concentrations in procedure are impracticable,
cases in which the concentration of a their water have been obtaining data on unnecessary, or contrary to the public
DBP in the samples was much higher these low concentrations from their interest, the agency may issue a rule
than the MRL. For example, if laboratories and have been using these without providing prior notice and an
chloroform concentrations are always data in their Consumer Confidence opportunity for public comment. In
greater than 0.040 mg/L in a water Reports. Setting the MRL at 0.020 mg/ addition to updating methods, this rule
system’s samples, then verifying L is reflective of current practices in also makes minor corrections to the
accurate quantitation at 0.0010 mg/L is laboratories and current data National Primary Drinking Water
unnecessary and may require the expectations by water systems. Regulations, specifically the Public
laboratory to dilute samples or maintain c. Summary of major comments. Notification tables (Subpart Q,
two calibration curves in order to There were no major comments. Appendices A and B). Two final
wwhite on PROD1PC65 with RULES2

comply with the requirement. EPA has drinking water rules (66 FR 6976 and 65
taken this into consideration in today’s P. Other Regulatory Changes FR 76708) inadvertently added new
rule and has adjusted the requirement As part of today’s action, EPA has endnotes to two existing tables using the
accordingly. EPA is maintaining the included several ‘‘housekeeping’’ same endnote numbers. This rule
requirement for all laboratories to actions to remove sections of Part 141 corrects this technical drafting error by

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00053 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
440 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations

renumbering the endnote citations in case basis, if a State will use the enforcement; (3) keeping records and
these two tables. Thus, additional notice authority to modify monitoring making reports available on activities
and public comment is not necessary. requirements under this special primacy that EPA requires by regulation; (4)
EPA finds that this constitutes ‘‘good condition. issuing variances and exemptions (if
cause’’ under 5 U.S.C. 553(b)(B). For the allowed by the State), under conditions
2. State Recordkeeping Requirements no less stringent than allowed under
same reasons, EPA is making this rule
change effective upon publication. 5 Today’s rule requires States to keep SDWA; and (5) adopting and being
U.S.C. 553(d)(3). additional records of the following, capable of implementing an adequate
including all supporting information plan for the provisions of safe drinking
V. State Implementation and an explanation of the technical water under emergency situations.
A. Today’s Rule basis for each decision: 40 CFR part 142 sets out the specific
—very small system waivers. program implementation requirements
This section describes the regulations for States to obtain primacy for the
—IDSE monitoring plans.
and other procedures and policies States public water supply supervision
—IDSE reports and 40/30
must adopt to implement today’s rule. program as authorized under SDWA
certifications, plus any
States must continue to meet all other section 1413. In addition to adopting
modifications required by the State.
conditions of primacy in 40 CFR Part —operational evaluations conducted basic primacy requirements specified in
142. To implement the Stage 2 DBPR, by the system. 40 CFR Part 142, States may be required
States must adopt revisions to the to adopt special primacy provisions
following: 3. State Reporting Requirements pertaining to specific regulations where
—§ 141.2—Definitions Today’s rule has no new State implementation of the rule involves
—§ 141.33—Record maintenance; reporting requirements. activities beyond general primacy
—§ 141.64—Maximum contaminant provisions. States must include these
levels for disinfection byproducts; 4. Interim Primacy
regulation specific provisions in an
—subpart L—Disinfectant Residuals, States that have primacy for every application for approval of their
Disinfection Byproducts, and existing NPDWR already in effect may program revision.
Disinfection Byproduct Precursors; obtain interim primacy for this rule, The current regulations in 40 CFR
—subpart O, Consumer Confidence beginning on the date that the State 142.14 require States with primacy to
Reports; submits the application for this rule to keep various records, including the
—subpart Q, Public Notification of USEPA, or the effective date of its following: analytical results to
Drinking Water Violations; revised regulations, whichever is later. determine compliance with MCLs,
—new subpart U, Initial Distribution A State that wishes to obtain interim MRDLs, and treatment technique
System Evaluation; and primacy for future NPDWRs must obtain requirements; PWS inventories; State
—new subpart V, Stage 2 Disinfection primacy for today’s rule. As described approvals; enforcement actions; and the
Byproducts Requirements. in Section IV.F, EPA expects to work issuance of variances and exemptions.
1. State Primacy Requirements for with States to oversee the individual Today’s final rule requires States to
Implementation Flexibility distribution system evaluation process keep additional records, including all
that begins shortly after rule supporting information and an
In addition to adopting basic primacy promulgation. explanation of the technical basis for
requirements specified in 40 CFR part decisions made by the State regarding
142, States are required to address 5. IDSE Implementation
today’s rule requirements. The State
applicable special primacy conditions. As discussed in section IV.E, many may use these records to identify trends
Special primacy conditions pertain to systems will be performing certain IDSE and determine whether to limit the
specific regulations where activities prior to their State receiving scope of operational evaluations. EPA
implementation of the rule involves primacy. During that period, EPA will currently requires in 40 CFR 142.15 that
activities beyond general primacy act as the primacy agency, but will States report to EPA information such as
provisions. The purpose of these special consult and coordinate with individual violations, variance and exemption
primacy requirements in today’s rule is States to the extent practicable and to status, and enforcement actions; today’s
to ensure State flexibility in the extent that States are willing and rule does not add additional reporting
implementing a regulation that (1) able to do so. In addition, prior to requirements or modify existing
applies to specific system configurations primacy, States may be asked to assist requirements.
within the particular State and (2) can EPA in identifying and confirming On April 28, 1998, EPA amended its
be integrated with a State’s existing systems that are required to comply State primacy regulations at 40 CFR
Public Water Supply Supervision with certain IDSE activities. Once the 142.12 to incorporate the new process
Program. States must include these rule- State has received primacy, it will identified in the 1996 SDWA
distinct provisions in an application for become responsible for IDSE Amendments for granting primary
approval or revision of their program. implementation activities. enforcement authority to States while
These primacy requirements for their applications to modify their
implementation flexibility are discussed B. Background and Analysis primacy programs are under review (63
in this section. SDWA establishes requirements that a FR 23362, April 28, 1998) (USEPA
To ensure that a State program State or eligible Indian Tribe must meet 1998f). The new process grants interim
includes all the elements necessary for to assume and maintain primary primary enforcement authority for a
an effective and enforceable program enforcement responsibility (primacy) for new or revised regulation during the
under today’s rule, a State primacy its PWSs. These requirements include period in which EPA is making a
wwhite on PROD1PC65 with RULES2

application must include a description the following activities: (1) Adopting determination with regard to primacy
of how the State will implement a drinking water regulations that are no for that new or revised regulation. This
procedure for modifying consecutive less stringent than Federal drinking interim enforcement authority begins on
system and wholesale system water regulations; (2) adopting and the date of the primacy application
monitoring requirements on a case-by- implementing adequate procedures for submission or the effective date of the

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00054 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 441

new or revised State regulation, the staggered rule schedule will analyses summarized in this section and
whichever is later, and ends when EPA facilitate State involvement in pre- additional analytical results.
makes a final determination. However, primacy implementation.
A. Regulatory Alternatives Considered
this interim primacy authority is only Many commenters also requested that
available to a State that has primacy the State have more flexibility to grant The Stage 2 DBPR is the second in a
(including interim primacy) for every sampling waivers and exemptions. EPA set of rules that address public health
existing NPDWR in effect when the new believes that it has struck a reasonable risks from DBPs. EPA promulgated the
regulation is promulgated. States that balance among competing objectives in Stage 1 DBPR to decrease average
have primacy for every existing NPDWR granting State flexibility. State exposure to DBPs and mitigate
already in effect may obtain interim flexibility comes at a resource cost and associated health risks—compliance
primacy for this rule and a State that excessive system-by-system flexibility with TTHM and HAA5 MCLs is based
wishes to obtain interim primacy for could overwhelm State resources. Also, on averaging concentrations across the
future NPDWRs must obtain primacy for EPA believes that much of the distribution system. In developing the
this rule. monitoring and water quality Stage 2 DBPR, EPA sought to identify
EPA is aware that due to the information a State would need to and further reduce remaining risks from
complicated wholesale system- properly consider whether a waiver is exposure to chlorinated DBPs.
consecutive system relationships that appropriate is generally not available The regulatory options EPA
exist nationally, there will be cases and, if available, difficult to evaluate. considered for the Stage 2 DBPR are the
where the standard monitoring direct result of a consensus rulemaking
framework will be difficult to VI. Economic Analysis process (Federal Advisory Committee
implement. Therefore, States may This section summarizes the Act (FACA) process) that involved
develop, as a special primacy condition, Economic Analysis for the Final Stage 2 various drinking water stakeholders (see
a program under which the State can Disinfectants and Disinfection Section III for a description of the FACA
modify monitoring requirements for Byproducts Rule (Economic Analysis process). The Advisory Committee
consecutive systems. These (EA)) (USEPA 2005a). The EA is an considered the following key questions
modifications must not undermine evaluation of the benefits and costs of during the negotiation process for the
public health protection and all today’s final rule and other regulatory Stage 2 DBPR:
systems, including consecutive systems, • What are the remaining health risks
alternatives the Agency considered.
must comply with the TTHM and HAA5 after implementation of the Stage 1
Specifically, this evaluation addresses
MCLs based on the LRAA at each DBPR?
both quantified and non-quantified
compliance monitoring location. Each • What are approaches to addressing
benefits to PWS consumers, including
consecutive system must have at least these risks?
the general population and sensitive • What are the risk tradeoffs that need
one compliance monitoring location. subpopulations. Costs are presented for
However, such a program allows the to be considered in evaluating these
PWSs, States, and consumer approaches?
State to establish monitoring
households. Also included is a • How do the estimated costs of an
requirements that account for
discussion of potential risks from other approach compare to reductions in peak
complicated distribution system
contaminants, uncertainties in benefit DBP occurrences and overall DBP
relationships, such as where
and cost estimates, and a summary of exposure for that approach?
neighboring systems buy from and sell
major comments on the EA for the The Advisory Committee considered
to each other regularly throughout the
proposed Stage 2 DBPR. DBP occurrence estimates to be
year, water passes through multiple
consecutive systems before it reaches a EPA relied on data from several important in understanding the nature
user, or a large group of interconnected epidemiologic and toxicologic studies, of public health risks. Although the ICR
systems have a complicated combined the Information Collection Rule (ICR), data were collected prior to
distribution system. EPA has developed and other sources, along with analytical promulgation of the Stage 1 DBPR, they
a guidance manual to address these and models and input from technical were collected under a similar sampling
other consecutive system issues. experts, to understand DBP risk, strategy. The data support the concept
occurrence, and PWS treatment changes that a system could be in compliance
C. Summary of Major Comments that will result from today’s rule. with the RAA Stage 1 DBPR MCLs of
Public comment generally supported Benefits and costs are presented as 0.080 mg/L and 0.060 mg/L for TTHM
the special primacy requirements in the annualized values using social discount and HAA5, respectively, and yet have
August 11, 2003 proposal, and many rates of three and seven percent. The points in the distribution system with
commenters expressed appreciation for time frame used for benefit and cost either periodically or consistently
the flexibility the special primacy comparisons is 25 years—approximately higher DBP levels.
requirements provided to States. five years account for rule Based on these findings, the Advisory
Many commenters expressed concern implementation and 20 years for the Committee discussed an array of
about EPA as the implementer instead average useful life of treatment alternatives to address disproportionate
of the State, given the existing technologies. risk within distribution systems.
relationship between the State and EPA has prepared this EA to comply Alternative options included lowering
system. EPA agrees that States perform with the requirements of SDWA, DBP MCLs, revising the method for
an essential role in rule implementation including the Health Risk Reduction MCL compliance determination e.g.,
and intends to work with States to the and Cost Analysis required by SDWA requiring individual sampling locations
greatest extent possible, consistent with section 1412(b)(3)(C), and Executive to meet the MCL as an LRAA or
the rule schedule promulgated today. Order 12866, Regulatory Planning and requiring that no samples exceed the
wwhite on PROD1PC65 with RULES2

EPA believes that pre-promulgation Review. The full EA is available in the MCL), and combinations of both. The
coordination with States, changes in the docket for today’s rule, which is Advisory Committee also considered the
final rule strongly supported by States available online as described in the associated technology changes and costs
(e.g., population-based monitoring ADDRESSES section. The full document for these alternatives. After narrowing
instead of plant-based monitoring), and provides detailed explanations of the down options, the Advisory Committee

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00055 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
442 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations

primarily focused on four types of —MCLs of 0.080 mg/L for TTHM and TTHM/HAA5 alternatives described and
alternative MCL scenarios. These are the 0.060 mg/L for HAA5 as absolute the Stage 1 DBPR for a hypothetical
alternatives EPA evaluated in the EA, as maximums for individual large surface water system. This
follows: measurements hypothetical system has one treatment
Preferred Alternative —Bromate MCL remaining at 0.010 plant and measures TTHM in the
—MCLs of 0.080 mg/L for TTHM and mg/L distribution system in four locations per
0.060 mg/L for HAA5 as LRAAs Alternative 3 quarter (the calculation methodology
—Bromate MCL remaining at 0.010 —MCLs of 0.040 mg/L for TTHM and shown would be the same for HAA5).
mg/L 0.030 mg/L for HAA5 as RAAs Ultimately, the Advisory Committee
Alternative 1
—MCLs of 0.080 mg/L for TTHM and —Bromate MCL remaining at 0.010 recommended the Preferred Alternative
0.060 mg/L for HAA5 as LRAAs mg/L. in combination with an IDSE
—Bromate MCL of 0.005 mg/L Figure VI.A–1 shows how compliance requirement (discussed in Section IV.F).
Alternative 2 would be determined under each of the
wwhite on PROD1PC65 with RULES2

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00056 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 443
wwhite on PROD1PC65 with RULES2



VerDate Aug<31>2005 21:06 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00057 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
444 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations

B. Analyses That Support Today’s Final predicted pre-Stage 1 baseline, the ICR Epidemiology, and End Results program
Rule Matrix Method uses unadjusted ICR in conjunction with data from the 2000
EPA’s goals in designing the Stage 2 TTHM and HAA5 pre-Stage 1 data to U.S. Census to estimate the number of
DBPR were to protect public health by estimate the percent of plants changing new bladder cancer cases per year
reducing peak DBP levels in the technology to comply with the Stage 2 (USEPA 2005a). Three approaches were
distribution system while maintaining DBPR. EPA gives equal weight to SWAT then used to gauge the percentage of
microbial protection. As described and ICR Matrix Method predictions in cases attributable to DBP exposure (i.e.,
estimating Stage 2 compliance forecasts population attributable risk (PAR)).
earlier, the Stage 1 DBPR reduces
and resultant reductions in DBP Taken together, the three approaches
overall average DBP levels, but specific
exposure. The ICR Matrix Method is provide a reasonable estimate of the
locations within distribution systems
also used to estimate reductions in the range of potential risks. EPA notes that
can still experience relatively high DBP
occurrence of peak TTHM and HAA5 the existing epidemiological evidence
concentrations. EPA believes that high
concentrations because SWAT- has not conclusively established
DBP concentrations should be reduced
predicted TTHM and HAA5 causality between DBP exposure and
due to the potential association of DBPs
concentrations are valid only when any health risk endpoints, so the lower
with cancer, as well as reproductive and
considering national averages, not at the bound of potential risks may be as low
developmental health effects.
plant level. as zero.
EPA analyzed the benefits and costs When evaluating compliance with a
of the four regulatory alternatives The first approach used the range of
DBP MCL, EPA assumed that systems
presented in the previous section. PAR values derived from consideration
would maintain DBP levels at least 20
Consistent with the recommendations of of five individual epidemiology studies.
percent below the MCL. This safety
the Advisory Committee, EPA is This range was used at the basis for the
margin represents the level at which
establishing the preferred alternative to Stage 1 and the proposed Stage 2
systems typically take action to ensure
achieve the Agency’s goals for the Stage economic analyses (i.e., 2 percent to 17
they meet a drinking water standard and
2 DBPR. The following discussion percent) (USEPA 2003a).
reflects industry practice. In addition,
summarizes EPA’s analyses that support the safety margin accounts for year-to- The second approach used results
today’s final rule. This discussion year fluctuations in DBP levels. To from the Villanueva et al. (2003) meta-
explains how EPA predicted water address the impact of the IDSE, EPA analysis. This study develops a
quality and treatment changes, also analyzed compliance using a safety combined Odds Ratio (OR) of 1.2 that
estimated benefits and costs, and margin of 25 percent based on an reflects the ever-exposed category for
assessed the regulatory alternatives. analysis of spatial variability in TTHM both sexes from all studies considered
1. Predicting Water Quality and and HAA5 occurrence. EPA assigned in the meta-analysis and yields a PAR
Treatment Changes equal probability to the 20 and 25 value of approximately 16 percent.
percent safety margin for large and The third approach used the
Water quality and treatment data from medium surface water systems for the Villanueva et al. (2004) pooled data
the ICR were used in predicting final analysis because both alternatives analysis to develop a dose-response
treatment plant technology changes (i.e. are considered equally plausible. EPA relationship for OR as a function of
compliance forecasts) and reductions in assumes the 20 percent operational average TTHM exposure. Using the
DBP exposure resulting from the Stage safety margin accounts for variability in results from this approach, EPA
2 DBPR. Because ICR data were gathered small surface water systems and all estimates a PAR value of approximately
prior to Stage 1 DBPR compliance groundwater systems. 17 percent.
deadlines, EPA first accounted for
treatment changes resulting from the 2. Estimating Benefits EPA used the PAR values from all
Stage 1 DBPR. Benefit and cost Quantified benefits estimates for the three approaches to estimate the number
estimates for the Stage 2 DBPR reflect Stage 2 DBPR are based on potential of bladder cancer cases ultimately
changes following compliance with the reductions in fatal and non-fatal bladder avoided annually as a result of the Stage
Stage 1 DBPR. cancer cases. In the EA, EPA included 2 DBPR. To quantify the reduction in
The primary model used to predict a sensitivity analysis for benefits from cases, EPA assumed a linear
changes in treatment and reductions in avoiding colon and rectal cancers. EPA relationship between average DBP
DBP levels was the Surface Water believes additional benefits from this concentration and relative risk of
Analytical Tool (SWAT), which EPA rule could come from reducing potential bladder cancer. Because of this, EPA
developed using results from the ICR. reproductive and developmental risks. considers these estimates to be an upper
SWAT results were applied directly for EPA has not included these potential bound on the annual reduction in
large and medium surface water systems risks in the primary benefit analysis bladder cancer cases due to the rule.
and were adjusted for small surface because of the associated uncertainty. A lag period (i.e., cessation lag) exists
water systems to account for differences The major steps in deriving and between when reduction in exposure to
in source water DBP precursor levels characterizing potential cancer cases a carcinogen occurs and when the full
and operational constraints in small avoided include the following: (1) risk reduction benefit of that exposure
systems. EPA used ICR data and a estimate the current and future annual reduction is realized by exposed
Delphi poll process (a group of drinking cases of illness from all causes; (2) individuals. No data are available that
water experts who provided best estimate how many cases can be address the rate of achieving bladder
professional judgment in a structured attributed to DBP occurrence and cancer benefits resulting from DBP
format) to project technologies selected exposure; and (3) estimate the reduction reductions. Consequently, EPA used
by ground water systems. in future cases corresponding to data from epidemiological studies that
wwhite on PROD1PC65 with RULES2

To address uncertainty in SWAT anticipated reductions in DBP address exposure reduction to cigarette
predictions, EPA also predicted occurrence and exposure due to the smoke and arsenic to generate three
treatment changes using a second Stage 2 DBPR. possible cessation lag functions for
methodology, called the ‘‘ICR Matrix EPA used results from the National bladder cancer and DBPs. The cessation
Method.’’ Rather than a SWAT- Cancer Institute’s Surveillance, lag functions are used in conjunction

VerDate Aug<31>2005 21:06 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00058 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 445

with the rule implementation schedule performing these activities and on broad spectrum of microbial pathogens,
to project the number of bladder cancer laboratory costs. including microorganisms like
cases avoided each year as a result of While systems vary with respect to Cryptosporidium that are resistant to
the Stage 2 DBPR. many of the input parameters to the chlorine.
Although EPA used three approaches Stage 2 DBPR cost analysis (e.g., plants Alternative 2 would have prohibited
for estimating PAR, for simplicity’s per system, population served, flow per any single sample from exceeding the
sake, EPA used the Villanueva et al. population, labor rates), EPA believes TTHM or HAA5 MCL. This is
(2003) study to calculate the annual that mean values for the various input significantly more stringent than the
benefits of the Stage 2 DBPR. The parameters are appropriate to generate preferred alternative and would likely
benefits estimates derived from the best estimate of national costs for require a large fraction of surface water
Villanueva et al. (2003) capture a the rule. Uncertainty in the national systems to switch from their current
substantial portion of the overall range average unit capital and O&M costs for treatment practices to more expensive
of results, reflecting the uncertainty in the various technologies has been advanced technologies. Consistent with
both the underlying OR and PAR values, incorporated into the cost analysis the Advisory Committee, EPA does not
as well as the uncertainty in DBP (using Monte Carlo simulation believe such a drastic shift is warranted
reductions for Stage 2. procedures). Costs of the Stage 2 DBPR at this time.
To assign a monetary value to avoided are estimated at both mean and 90 Similarly, Alternative 3, which would
bladder cancer cases, EPA used the percent confidence bound values. decrease TTHM and HAA5 MCLs to
value of a statistical life (VSL) for fatal EPA assumes that systems will, to the 0.040 mg/L and 0.030 mg/L,
extent possible, pass cost increases on to respectively, and would require a
cases and used two alternate estimates
their customers through increases in significant portion of surface water
of willingness-to-pay to avoid non-fatal
water rates. Consequently, EPA has also systems to implement expensive
cases (one based on curable lymphoma
estimated annual household cost advanced technologies in place of their
and the other based on chronic
increases for the Stage 2 DBPR. This existing treatment. Further, compliance
bronchitis). EPA believes additional
analysis includes costs for all with TTHM and HAA5 MCLs under this
benefits from this rule could come from
households served by systems subject to alternative would be based on the RAA,
a reduction in potential reproductive
the rule, costs just for those households which does not specifically address DBP
and developmental risks. See Chapter 6
served by systems actually changing peaks in the distribution system as the
of the EA for more information on
treatment technologies to comply with LRAA, in conjunction with the IDSE,
estimating benefits (USEPA 2005a). the rule, costs for households served by are designed to do. Based on these
3. Estimating Costs small systems, and costs for systems considerations, EPA and the Advisory
served by surface water and ground Committee did not favor this alternative.
Analyzing costs for systems to comply
water sources.
with the Stage 2 DBPR included C. Benefits of the Stage 2 DBPR
identifying and costing treatment 4. Comparing Regulatory Alternatives
The benefits analysis for the Stage 2
process improvements that systems will Through the analyses summarized in DBPR includes a description of non-
make, as well as estimating the costs to this section, EPA assessed the benefits quantified benefits, calculations of
implement the rule, conduct IDSEs, and costs of the four regulatory quantified benefits, and a discussion of
prepare monitoring plans, perform alternatives described previously. when benefits will occur after today’s
additional routine monitoring, and Succeeding sections of this preamble final rule is implemented. An overview
evaluate significant DBP excursion present the results of these analyses. As of the methods used to determine
events. The cost analysis for States/ recommended by the Advisory benefits is provided in Section VI.B.
Primacy Agencies included estimates of Committee, EPA is establishing the More detail can be found in the final
the labor burdens for training employees preferred regulatory alternative for EA. A summary of benefits for the Stage
on the requirements of the Stage 2 today’s Stage 2 DBPR. This regulation 2 DBPR is given in this section.
DBPR, responding to PWS reports, and will reduce peak DBP concentrations in
record keeping. distribution systems through requiring 1. Nonquantified Benefits
All treatment costs are based on mean compliance determinations with Non-quantified benefits of the Stage 2
unit cost estimates for advanced existing TTHM and HAA5 MCLs using DBPR include potential benefits from
technologies and chloramines. the LRAA. Further, the IDSE will ensure reduced reproductive and
Derivation of unit costs are described in that systems identify compliance developmental risks, reduced risks of
detail in Technologies and Costs for the monitoring sites that reflect high DBP cancers other than bladder cancer, and
Final Long Term 2 Enhanced Surface levels. EPA believes that these provision improved water quality. EPA believes
Water Treatment Rule and Final Stage 2 are appropriate given the association of that additional benefits from this rule
Disinfectants and Disinfection DBPs with cancer, as well as potential could come from a reduction in
Byproducts Rule (USEPA 2005g). Unit reproductive and developmental health potential reproductive and
costs (capital and O&M) for each of nine effects. developmental risks. However, EPA
system size categories are calculated Alternative 1 would have established does not believe the available evidence
using mean design and average daily the same DBP regulations as the provides an adequate basis for
flows values. The unit costs are then preferred alternative, and would have quantifying these potential risks in the
combined with the predicted number of lowered the bromate MCL from 0.010 to primary analysis.
plants selecting each technology to 0.005 mg/L. The Advisory Committee Both toxicology and epidemiology
produce national treatment cost did not recommend and EPA did not studies indicate that other cancers may
estimates. establish this alternative because it be associated with DBP exposure but
wwhite on PROD1PC65 with RULES2

Non-treatment costs for could have an adverse effect on currently there is not enough data to
implementation, the IDSE, monitoring microbial protection. The lower bromate include them in the primary analysis.
plans, additional routine monitoring, MCL could cause many systems to However, EPA believes that the
and operational evaluations are based reduce or eliminate the use of ozone, association between exposure to DBPs
on estimates of labor hours for which is an effective disinfectant for a and colon and rectal cancer is possibly

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00059 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
446 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations

significant, so an analysis of benefits is from exposure to chemicals associated al. (2003), as described in Section VI.B.
presented as a sensitivity analysis. with accidental spills or environmental Table VI.C–1 summarizes the benefits
To the extent that the Stage 2 DBPR runoff. for the Preferred Regulatory Alternative
changes perceptions of the health risks for the Stage 2 DBPR. Benefits estimates
2. Quantified Benefits
associated with drinking water and for the other regulatory alternatives
improves taste and odor, it may reduce EPA has quantified the benefits were derived using the same methods as
actions such as buying bottled water or associated with the expected reductions for the Preferred Regulatory Alternative
installing filtration devices. Any in the incidence of bladder cancer. As and are presented in the EA.
resulting cost savings would be a discussed in Section VI.B, EPA used the The confidence bounds of the results
regulatory benefit. Also, as PWSs move PAR values from all three approaches to in Table VI.C–1 reflect uncertainty in
away from conventional treatment to estimate the number of bladder cancer PAR, uncertainty in the compliance
more advanced technologies, other non- cases ultimately avoided annually as a forecast and resulting reduction in DBP
health benefits are anticipated besides result of the Stage 2 DBPR, shown in concentrations, and cessation lag.
better tasting and smelling water. For Figure VI.C–1. Confidence bounds of the monetized
example, GAC lowers nutrient Table VI.C–1 summarizes the benefits also reflect uncertainty in
availability for bacterial growth, estimated number of bladder cancer valuation parameters. An estimated 26
produces a biologically more stable cases avoided as a result of the Stage 2 percent of bladder cancer cases avoided
finished water, and facilitates DBPR, accounting for cessation lag and are fatal, and 74 percent are non-fatal
management of water quality in the the rule implementation schedule, and (USEPA 1999b). The monetized benefits
distribution system. Since GAC also the monetized value of those cases. The therefore reflect the estimate of avoiding
removes synthetic organic chemicals benefits in Table VI.C–1 were developed both fatal and non-fatal cancers in those
(SOCs), it provides additional protection using the PAR value from Villanueva et proportions.
wwhite on PROD1PC65 with RULES2

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00060 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 447
wwhite on PROD1PC65 with RULES2



VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00061 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
448 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations

Annual average cases avoided Annualized benefits of cases avoided
Discount rate, WTP for non- Cessation lag model
fatal cases
Mean 5th 95th Mean 5th 95th

279 103 541 3%, Lymphoma ................... $1,531 $233 $3,536 Smoking/Lung Cancer
7% Lymphoma .................... 1,246 190 2,878
3% Bronchitis ...................... 763 165 1,692
7% Bronchitis ...................... 621 135 1,376
188 61 399 3%, Lymphoma ................... 1,032 157 2,384 Smoking/Bladder Cancer
7% Lymphoma .................... 845 129 1,950
3% Bronchitis ...................... 514 111 1,141
7% Bronchitis ...................... 420 91 932
333 138 610 3%, Lymphoma ................... 1,852 282 4,276 Arsenic/Bladder Cancer
7% Lymphoma .................... 1,545 235 3,566
3% Bronchitis ...................... 922 200 2,045
7% Bronchitis ...................... 769 167 1,704
Notes: Values are discounted and annualized in 2003$. The 90 percent confidence interval for cases incorporates uncertainty in PAR, reduc-
tion in average TTHM and HAA5 concentrations, and cessation lag. The 90 percent confidence bounds for monetized benefits reflect uncertainty
in monetization inputs relative to mean cases. Based on TTHM as an indicator, benefits were calculated using the Villanueva et al. (2003) PAR.
EPA recognizes that benefits may be as low as zero since causality has not yet been established between exposure to chlorinated water and
bladder cancer. Assumes 26 percent of cases are fatal, 74 percent are non-fatal (USEPA 1999b).
Source: Exhibit 6.1, USEPA 2005a.

3. Timing of Benefits Accrual exposures. EPA developed cessation lag years after the exposure reduction has
models for DBPs from literature to occurred, the annual cases avoided will
EPA recognizes that it is unlikely that describe the delayed benefits, in be 489 for the smoking/lung cancer
all cancer reduction benefits would be keeping with the recommendations of cessation lag model, 329 for the
realized immediately upon exposure the SAB (USEPA 2001d). Figure VI.C–2 smoking/bladder cancer cessation lag
reduction. Rather, it is expected that illustrates the effects of the cessation lag model, and 534 cases for the arsenic/
there will likely be some transition models. The results from the cessation bladder cancer cessation lag model.
period as individual risks reflective of lag models show that the majority of the These represent approximately 84%,
higher past exposures at the time of rule potential cases avoided occur within the 57%, and 92%, respectively, of the
implementation become, over time, first fifteen years after initial reduced estimated 581 annual cases ultimately
more reflective of the new lower exposure to DBPs. For example, fifteen avoidable by the Stage 2 DBPR.

In addition to the delay in reaching a six years from rule promulgation to treatment by 2016. The delay in
steady-state level of risk reduction as a meet the new Stage 2 MCLs, with up to exposure reduction resulting from the
result of cessation lag, there is a delay a two-year extension possible for capital rule implementation schedule is
in attaining maximum exposure improvements. In general, EPA assumes incorporated into the benefits model by
reduction across the entire affected that a fairly constant increment of adjusting the cases avoided for the given
population that results from the Stage 2 systems will complete installation of year and is illustrated in Table VI.C–2.
wwhite on PROD1PC65 with RULES2

DBPR implementation schedule. For new treatment technologies each year,

example, large surface water PWSs have with the last systems installing

VerDate Aug<31>2005 21:46 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00062 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 449

Smoking/lung cancer Smoking/bladder can- Arsenic/bladder can-
cessation lag model cer cessation lag cer cessation lag
Year model model
Total Percent Total Percent Total Percent

1 ............................................................................................................... 0 0 0 0 0 0
2 ............................................................................................................... 0 0 0 0 0 0
3 ............................................................................................................... 0 0 0 0 0 0
4 ............................................................................................................... 0 0 0 0 0 0
5 ............................................................................................................... 0 0 0 0 0 0
6 ............................................................................................................... 24 4 23 4 45 8
7 ............................................................................................................... 62 11 54 9 110 19
8 ............................................................................................................... 111 19 90 16 187 32
9 ............................................................................................................... 170 29 132 23 275 48
10 ............................................................................................................. 220 38 161 28 334 58
11 ............................................................................................................. 265 46 184 32 379 65
12 ............................................................................................................. 305 53 204 35 412 71
13 ............................................................................................................. 341 59 221 38 438 76
14 ............................................................................................................. 371 64 237 41 458 79
15 ............................................................................................................. 396 68 251 43 475 82
16 ............................................................................................................. 416 72 265 46 488 84
17 ............................................................................................................. 433 75 278 48 499 86
18 ............................................................................................................. 448 77 289 50 509 88
19 ............................................................................................................. 460 79 301 52 516 89
20 ............................................................................................................. 471 81 311 54 523 90
21 ............................................................................................................. 481 83 321 55 528 91
22 ............................................................................................................. 489 84 330 57 533 92
23 ............................................................................................................. 496 86 339 59 537 93
24 ............................................................................................................. 503 87 347 60 541 93
25 ............................................................................................................. 509 88 355 61 544 94
Notes: Percent of annual cases ultimately avoidable achieved during each of the first 25 years. The benefits model estimates 581 (90% CB =
229–1,079) annual cases ultimately avoidable using the Villanueva et al. (2003) PAR inputs and including uncertainty in these and DBP reduc-
tions. EPA recognizes that benefits may be as low as zero since causality has not yet been established between exposure to chlorinated water
and bladder cancer.
Source: Summarized from detailed results presented in Exhibits E.38a, E.38e and E.38i, USEPA 2005a.

D. Costs of the Stage 2 DBPR present value costs, PWS costs, State/ $77 million per year. The mean and
National costs include those of Primacy agency costs, and non- range of annualized costs are similar at
treatment changes to comply with the quantified costs. a 7 percent discount rate. State costs are
rule as well as non-treatment costs such estimated to be between $1.70 and $1.71
1. Total Annualized Present Value Costs
as for Initial Distribution System million per year depending on the
Evaluations (IDSEs), additional routine Tables VI.D–1 and VI.D–2 summarize discount rate. These estimates are
monitoring, and operational the average annualized costs for the annualized starting with the year of
evaluations. The methodology used to Stage 2 DBPR Preferred Regulatory promulgation. Actual dollar costs
estimate costs is described in Section Alternative at 3 and 7 percent discount during years when most treatment
VI.B. More detail is provided in the EA rates, respectively. System costs range changes are expected to occur would be
(USEPA 2005a). The remainder of this from approximately $55 to $101 million somewhat higher (the same is true for
section presents summarized results of annually at a 3 percent discount rate, benefits that occur in the future).
EPA’s cost analysis for total annualized with a mean estimate of approximately BILLING CODE 6560–50–P
wwhite on PROD1PC65 with RULES2

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00063 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
450 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations
wwhite on PROD1PC65 with RULES2


VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00064 Fmt 4701 Sfmt 4725 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 451
wwhite on PROD1PC65 with RULES2


VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00065 Fmt 4701 Sfmt 4725 E:\FR\FM\04JAR2.SGM 04JAR2
452 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations

2. PWS costs installation as well as operation and incur some costs. EPA’s analysis
maintenance. Significant PWS costs for allocated systems into five categories to
PWS costs for the Stage 2 DBPR IDSEs, treatment, and monitoring are determine the costs of the IDSE—those
include non-treatment costs of rule described in this section, along with a conducting standard monitoring, SSS,
implementation, Initial Distribution sensitivity analysis. VSS, 40/30, and NTNCWS not required
System Evaluations (IDSEs), Stage 2 a. IDSE costs. Costs and burden to do an IDSE. EPA then developed cost
DBPR monitoring plans, additional associated with IDSE activities differ estimates for each option. Tables VI.D–
routine monitoring, and operational depending on whether or not the system 3, VI.D–4, and VI.D–5 illustrate PWS
evaluations. Systems required to install performs the IDSE and, if so, which
costs for IDSE for systems conducting an
treatment to comply with the MCLs will option a system chooses. All systems
SMP, SSS, and 40/30, respectively.
accrue the additional costs of treatment performing the IDSE are expected to
wwhite on PROD1PC65 with RULES2


VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00066 Fmt 4701 Sfmt 4725 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 453
wwhite on PROD1PC65 with RULES2


VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00067 Fmt 4701 Sfmt 4725 E:\FR\FM\04JAR2.SGM 04JAR2
454 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations

b. PWS treatment costs. The number forecast. The percent of systems VI.D–6. The cost model includes
of plants changing treatment as a result predicted to make treatment technology estimates for the cost of each
of the Stage 2 DBPR and which changes and the technologies predicted technology; the results of the cost model
technology various systems will install to be in place after implementation of for PWS treatment costs are summarized
are determined from the compliance the Stage 2 DBPR are shown in Table in Table VI.D–7.
wwhite on PROD1PC65 with RULES2


VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00068 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 455
wwhite on PROD1PC65 with RULES2


VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00069 Fmt 4701 Sfmt 4725 E:\FR\FM\04JAR2.SGM 04JAR2
456 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations
wwhite on PROD1PC65 with RULES2

c. Monitoring costs. Because systems costs for additional routine monitoring number of samples to be collected from
already sample for the Stage 1 DBPR, are determined by the change in the the Stage 1 to the Stage 2 DBPR. The

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00070 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 457

Stage 2 DBPR monitoring requirements number of treatment plants. With this increase for some systems but actually
for systems are based only on modification in monitoring scheme, the decrease from the Stage 1 to the Stage
population served and source water average system will have no change in 2 DBPR for many systems. Table VI.D–
type, while the Stage 1 DBPR monitoring costs. The number of 8 summarizes the estimated additional
requirements are also based on the samples required is estimated to routine monitoring costs for systems.
wwhite on PROD1PC65 with RULES2

BILLING CODE 6560–50–c


VerDate Aug<31>2005 21:06 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00071 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
458 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations

3. State/Primacy Agency Costs is multiplied by the number of labor activity based on experience
hours per FTE, the State/Primacy implementing previous rules, such as
To estimate State/Primacy Agency Agency hourly wage, and the number of the Stage 1 DBPR. State/Primacy Agency
costs, the estimated number of full-time States/Primacy Agencies. EPA estimated costs are summarized in Table VI.D–9.
equivalents (FTEs) required per activity the number of FTEs required per

4. Non-quantified Costs such as changing storage tank operation E. Household Costs of the Stage 2 DBPR
All significant costs that EPA has were also not considered as alternatives EPA estimates that, as a whole,
identified have been quantified. In some to treatment. These might be options for households subject to the Stage 2 DBPR
instances, EPA did not include a systems with a single problem area with face minimal increases in their annual
potential cost element because its effects a long residence time. In the absence of costs. Approximately 86 percent of the
are relatively minor and difficult to detailed information needed to evaluate households potentially subject to the
estimate. For example, it may be less situations such as these, EPA has rule are served by systems serving at
costly for a small system to merge with included a discussion of possible effects least 10,000 people; these systems
neighboring systems than to add where appropriate. In general, however, experience the lowest increases in costs
advanced treatment. Such changes have the expected net effect of such due to significant economies of scale.
both costs (legal fees and connecting situations is lower costs to PWSs. Thus, Households served by small systems
infrastructure) and benefits (economies the EA tends to present conservatively that add treatment will face the greatest
of scale). Likewise, procuring a new high estimates of costs in relation to increases in annual costs. Table VI.E–1
source of water would have costs for non-quantified costs. summarizes annual household cost
new infrastructure, but could result in increases for all system sizes.
lower treatment costs. Operational costs
wwhite on PROD1PC65 with RULES2


VerDate Aug<31>2005 21:06 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00072 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 459


Percentage Percentage
90th per- 95th per- of annual of annual
Mean an- Median an-
Total number centile an- centile an- household household
nual house- nual house-
of households nual house- nual house- cost in- cost in-
hold cost in- hold cost in-
served hold cost in- hold cost in- crease < crease <
crease crease crease crease $12 (per- $120 (per-
cent) cent)

Households Served by All Plants

All Systems .......................................... 101,553,868 $0.62 $0.03 $0.36 $0.98 99 100

All Small Systems ................................ 14,261,241 2.20 0.10 0.79 2.57 97 100
SW < 10,000 ........................................ 3,251,893 4.58 0.79 2.69 7.24 95 99
SW ≥ 10,000 ........................................ 62,137,350 0.46 0.02 0.35 1.81 99 100
GW < 10,000 ....................................... 11,009,348 1.49 0.02 0.39 0.99 98 100
GW ≥ 10,000 ........................................ 25,155,277 0.13 0.00 0.03 0.08 100 100

Households Served by Plants Adding Treatment

All Systems .......................................... 10,161,304 $5.53 $0.80 $10.04 $22.40 92 99

All Small Systems ................................ 591,623 46.48 18.47 168.85 197.62 38 89
SW < 10,000 ........................................ 285,911 43.05 13.79 173.53 177.93 47 85
SW ≥ 10,000 ........................................ 9,060,119 2.83 0.80 6.98 11.31 96 100
GW < 10,000 ....................................... 305,712 49.69 16.65 109.86 197.62 31 92
GW ≥ 10,000 ........................................ 509,562 5.97 1.37 26.82 33.84 79 100
Notes: Detail may not add to total due to independent rounding. Number of households served by systems adding treatment will be higher
than households served by plants adding treatment because an entire system will incur costs even if only some of the plants for that system add
treatment (this would result in lower household costs, however).
Source: Exhibit 7.15, USEPA 2005a.

F. Incremental Costs and Benefits of the benefit and cost comparisons may be are greater due to the additional control
Stage 2 DBPR unrepresentative of the true net benefits of bromate. However, the benefits of
Incremental costs and benefits are of the rule because a significant portion Alternative 1 are less than the Preferred
those that are incurred or realized in of the rule’s potential benefits are not Alternative because the Agency is not
reducing DBP exposures from one quantified, particularly potential able to estimate the additional benefits
alternative to the next more stringent reproductive and developmental health of reducing the bromate MCL.
alternative. Estimates of incremental effects (see Section VI.C). Table VI.F–1 Alternative 1 was determined to be
costs and benefits are useful in shows the incremental monetized costs unacceptable due to the potential for
considering the economic efficiency of and benefits for each regulatory increased risk of microbial exposure.
different regulatory options considered alternative. Evaluation of this table Both benefits and costs are greater for
by the Agency. Generally, the goal of an shows that incremental costs generally Alternative 2 and Alternative 3 as
incremental analysis is to identify the fall within the range of incremental compared to the Preferred Alternative.
regulatory option where net social benefits for each more stringent However, these regulatory alternatives
benefits are maximized. However, the alternative. Equally important, the do not have the risk-targeted design of
usefulness of this analysis is addition of any benefits attributable to the Preferred Alternative. Rather,
constrained when major benefits and/or the non-quantified categories would add implementation of these stringent
costs are not quantified or not to the benefits without any increase in standards would require a large number
monetized. Also, as pointed out by the costs. of systems to change treatment
Environmental Economics Advisory Table VI.F–1 shows that the Preferred technology. The high costs of these
Committee of the Science Advisory Alternative is the least-cost alternative. regulatory alternatives and the drastic
Board, efficiency is not the only A comparison of Alternative 1 with the shift in the nation’s drinking water
appropriate criterion for social decision Preferred Alternative shows that practices were considered unwarranted
making (USEPA 2000i). Alternative 1 would have approximately at this time. (See Section VI.A of this
For the proposed Stage 2 DBPR, the same benefits as the Preferred preamble for a description of regulatory
presentation of incremental quantitative Alternative. The costs of Alternative 1 alternatives.)


Annual Annual ben- Incremental costs Incremental benefits Incremental net bene-
WTP for non-fatal costs efits fits
Rule alternative
bladder cancer cases C D

3 Percent Discount Rate

wwhite on PROD1PC65 with RULES2

Lymphoma ............... Preferred ................. $79 $1,531 $79 .......................... $1,531 ..................... $1,452
Alternative 1 1 .......... 254 1,377 (1) ............................ (1) ............................ (1)
Alternative 2 ............ 422 5,167 343 .......................... 3,637 ....................... 3,294
Alternative 3 ............ 634 7,130 212 .......................... 1,962 ....................... 1,750
Bronchitis ................. Preferred ................. 79 763 79 ............................ 763 .......................... 684

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00073 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
460 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations


Annual Annual ben- Incremental costs Incremental benefits Incremental net bene-
WTP for non-fatal costs efits fits
Rule alternative
bladder cancer cases C D

Alternative 1 1 .......... 254 686 (1) ............................ (1) ............................ (1)

Alternative 2 ............ 422 2,575 343 .......................... 1,812 ....................... 1,469
Alternative 3 ............ 634 3,552 212 .......................... 978 .......................... 765

7 Percent Discount Rate

Lymphoma ............... Preferred ................. $77 $1,246 $77 .......................... $1,246 ..................... $1,170
Alternative 1 1 .......... 242 1,126 (1) ............................ (1) ............................ (1)
Alternative 2 ............ 406 4,227 330 .......................... 2,981 ....................... 2,651
Alternative 3 ............ 613 5,832 207 .......................... 1,605 ....................... 1,399
Bronchitis ................. Preferred ................. 77 621 77 ............................ 621 .......................... 544
Alternative 1 1 .......... 242 561 (1) ............................ (1) ............................ (1)
Alternative 2 ............ 406 2,105 330 .......................... 1,484 ....................... 1,154
Alternative 3 ............ 613 2,904 207 .......................... 799 .......................... 593
Notes: Estimates are discounted to 2003 and given in 2003 dollars. Based on TTHM as an indicator, Villanueva et al. (2003) for baseline risk,
and smoking/lung cancer cessation lag model. Assumes 26 percent of cases are fatal, 74 percent are non-fatal (USEPA 1999b). EPA recognizes
that benefits may be as low as zero since causality has not yet been established between exposure to chlorinated water and bladder cancer.
1 Alternative 1 appears to have fewer benefits than the Preferred Alternative because it does not incorporate the IDSE, as explained in Chapter
4. Furthermore, this EA does not quantify the benefits of reducing the MCL for bromate (and potentially associated cancer cases), a requirement
that is included only in Alternative 1. This means that Alternative 1 is dominated by the Preferred Alternative in this analysis (having higher costs
than the Preferred Alternative but lower benefits), and so it is not included in the incremental comparison of alternatives (Columns C–E). OMB
states this in terms of comparing cost effectiveness ratios, but the same rule applies to an incremental cost, benefits, or net benefits comparison:
‘‘When constructing and comparing incremental cost-effectiveness ratios, [analysts] * * * should make sure that inferior alternatives identified by
the principles of strong and weak dominance are eliminated from consideration.’’ (OMB Circular A–4, p. 10)
Source: Exhibit 9.13, USEPA 2005a.

G. Benefits From the Reduction of Co- TTHM and HAA5 as indicators for Iodoacetic acid was found to be
occurring Contaminants chlorination DBP occurrence and cytotoxic and genotoxic in Salmonella
Installing certain advanced believes that operational and treatment and mammalian cells (Plewa et al.
technologies to control DBPs has the changes made because of the Stage 2 2004a) as were some of the
added benefit of controlling other DBPR will result in an overall decrease halonitromethanes (Kundu et al. 2004;
drinking water contaminants in addition in risk. Plewa et al. 2004b). Although potent in
to those specifically targeted by the these in vitro screening studies, further
1. Emerging DBPs
Stage 2 DBPR. For example, membrane research is needed to determine if these
technology installed to reduce DBP Iodo-DBPs and nitrogenous DBPs DBPs are active in living systems. No
precursors can also reduce or eliminate including halonitromethanes are DBPs conclusions on human health risk can
many other drinking water that have recently been reported be drawn from such preliminary
contaminants (depending on pore size), (Richardson et al. 2002, Richardson studies.
including those that EPA may regulate 2003). One recent occurrence study
sampled quarterly at twelve surface 2. N-Nitrosamines
in the future. Removal of any
contaminants that may face regulation water plants using different Another group of nitrogenous DBPs
could result in future cost savings to a disinfectants across the U.S. for several are the N-nitrosamines. A number of N-
water system. Because of the difficulties iodo-THMs and halonitromethane nitrosamines exist, and N-
in establishing which systems would be species (Weinberg et al. 2002). The nitrosodimethylamine (NDMA), a
affected by other current or future rules, concentrations of iodo-THMs and probable human carcinogen (USEPA
no estimate was made of the potential halonitromethane in the majority of 1993), has been identified as a potential
cost savings from addressing more than samples in this study were less than the health risk in drinking water. NDMA is
one contaminant simultaneously. analytical minimum reporting levels; a contaminant from industrial sources
plant-average concentrations of iodo- and a potential disinfection byproduct
H. Potential Risks From Other THM and halonitromethane species from reactions of chlorine or chloramine
Contaminants were typically less than 0.002 mg/L, with nitrogen containing organic matter
Along with the reduction in DBPs which is an order of magnitude lower and from some polymers used as
from chlorination such as TTHM and than the corresponding average coagulant aids. Studies have produced
HAA5 as a resultof the Stage 2 DBPR, concentrations of TTHM and HAA5 at new information on the mechanism of
there may be increases in other DBPs as those same plants. Chloropicrin, a formation of NDMA, but there is not
systems switch from chlorine to halonitromethane species, was also enough information at this time to draw
alternative disinfectants. For all measured in the ICR with a median conclusions regarding a potential
disinfectants, many DBPs are not concentration of 0.00019 mg/L across all increase in NDMA occurrence as
regulated and many others have not yet surface water samples. No occurrence systems change treatment. Although
wwhite on PROD1PC65 with RULES2

been identified. EPA will continue to data exist for the iodoacids due to the there are studies that examined the
review new studies on DBPs and their lack of a quantitative method and occurrence of NDMA in some water
occurrence levels to determine if they standards. Further work on chemical systems, there are no systematic
pose possible health risks. EPA formation of iodo-DBPs and evaluations of the occurrence of NDMA
continues to support regulation of halonitromethanes is needed. and other nitrosamines in U.S. waters.

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00074 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 461

Recent studies have provided new I. Effects of the Contaminant on the J. Uncertainties in the Risk, Benefit, and
occurrence information that shows General Population and Groups Within Cost Estimates for the Stage 2 DBPR
NDMA forms in both chlorinated and the General Population That Are For today’s final rule, EPA has
chloraminated systems. Barrett et al. Identified As Likely To Be at Greater estimated the current baseline risk from
(2003) reported median concentrations Risk of Adverse Health Effects exposure to DBPs in drinking water and
of less than 2ng/L for the seven chlorine projected the risk reduction and cost for
systems studied and less than 3 ng/L for EPA’s Office of Water has historically
various rule alternatives. There is
considered risks to sensitive
13 chloramine systems. Another study uncertainty in the risk calculation, the
subpopulations (including fetuses,
demonstrated that factors other than benefit estimates, the cost estimates, and
infants, and children) when establishing
disinfectant type may play an important the interaction with other regulations.
drinking water assessments, advisories The EA has an extensive discussion of
role in the formation of NDMA
and other guidance, and standards relevant uncertainties (USEPA 2005a).
(Schreiber and Mitch 2005). More
(USEPA 1989) (56 FR 3526, January 30, This section briefly summarizes the
research is underway to determine the
1991) (USEPA 1991). In the case of major uncertainties. Table VI.J–1
extent of NDMA occurrence in drinking Stage 2 DBPR, maximizing health
water systems. EPA has proposed presents a summary of uncertainty in
protection for sensitive subpopulations the cost and benefit estimates, refers to
monitoring for NDMA under requires balancing risks to achieve the
Unregulated Contaminant Monitoring the section or appendix of the EA where
recognized benefits of controlling the information is introduced, and
Rule 2 (70 FR 49094, at 49103, August waterborne pathogens while minimizing
22, 2005) (USEPA 2005m). estimates the potential effects that each
risk of potential DBP toxicity. may have on national cost and benefit
Risk assessments have estimated that Experience shows that waterborne estimates.
the 10¥6 lifetime cancer risk level is 7 disease from pathogens in drinking EPA believes that uncertainty in the
ng/L based on induction of tumors at water is a major concern for children compliance forecast has a potentially
multiple sites. NDMA is also present in and other subgroups (e.g., the elderly, large influence on cost and benefit
food, tobacco smoke, and industrial immunocompromised, and pregnant estimates for today’s rule. Thus, the
emissions, and additional research is women) because of their greater Agency has attempted to quantify the
underway to determine the relative vulnerabilities (Gerba et al. 1996). EPA uncertainty by giving equal weight to
exposure of NDMA in drinking water to believes DBPs may also potentially pose two different compliance forecast
these other sources. risks to fetuses and pregnant women approaches. One compliance forecast
(USEPA 1998a). In addition, because the approach is based on the SWAT
3. Other DBPs elderly population (age 65 and above) is predictions, and the other is based on
Some systems, depending on bromide naturally at a higher risk of developing the ‘‘ICR Matrix Method.’’ The ICR
and organic precursor levels in the bladder cancer, their health risks may Matrix Method uses the same basic
further increase as a result of long-term approach as SWAT, but uses TTHM and
source water and technology selection,
DBP exposure (National Cancer Institute HAA5 data from the ICR directly to
may experience a shift to higher ratios,
2002). estimate the percent of plants changing
or concentrations, of brominated DBPs technology to comply with the Stage 2
while the overall TTHM or HAA5 In developing this rule, risks to
DBPR and the resulting DBP reduction.
concentration may decrease. In some sensitive subpopulations, including
To characterize the uncertainty of the
instances where alternative children, were taken into account in the
compliance forecast results, EPA
disinfectants are used, levels of chlorite assessments of disinfectants and DBPs.
assumes a uniform distribution between
and bromate may increase as a result of More details on sensitive
SWAT and ICR Matrix Method results
systems switching to chlorine dioxide or subpopulations can be found in the
(USEPA 2005a). That is, the cost and
ozone, respectively. However, EPA Economic Analysis (USEPA 2005a). For benefit estimates presented in the
anticipates that changes in chlorite and each of the DBPs included in the Stage preamble represent the midpoint
bromate concentration as a result of the 2 DBPR, the maximum contaminant between costs and benefits estimated
Stage 2 DBPR will be minimal (USEPA level goals (MCLG) are derived using the using the SWAT model, and those
2005a). For most systems, overall levels most sensitive endpoint among all estimated using the ICR Matrix Method.
of DBPs, as well as brominated DBP available data and an intraspecies Cost estimates using the SWAT model
species, should decrease as a result of uncertainty factor of 10 which accounts are about 25% lower than the midpoint
this rule. EPA continues to believe that for human variability including estimates, while those using the ICR
precursor removal is a highly effective sensitive subpopulations, like children. Matrix Method are about 25% higher.
The Agency has evaluated several Benefits estimated using the SWAT
strategy to reduce levels of DBPs.
alternative regulatory options and model are about 30% lower than the
EPA also considered the impact this selected the one that balances cost with midpoint estimates, while those using
rule may have on microbial significant benefits, including those for the ICR Matrix Method are about 30%
contamination that may result from sensitive subpopulations. The Stage 2 higher.
altering disinfection practices. To DBPR will result in a potential EPA believes the compliance forecast
address this concern, the Agency reduction in cancer risk and a potential may be overstated because the
developed this rule jointly with the reduction in reproductive and technology decision tree does not
Long Term 2 Enhanced Surface Water developmental risk to fetuses and consider low-cost, non-treatment system
Treatment Rule (LT2ESWTR). EPA pregnant women. It should be noted that improvements that could be used to
expects that the LT2ESWTR provisions the LT2ESWTR, which accompanies comply with the Stage 2 DBPR. These
wwhite on PROD1PC65 with RULES2

will prevent increases in microbial risk this rule, reduces pathogens in drinking improvements, including things like
resulting from the Stage 2 DBPR. water and further protects sensitive flushing more frequently and managing
subpopulations. See Section VII.G for a storage facilities to reduce water age,
discussion of EPA’s requirements under could be used by systems to reduce
Executive Order 13045. TTHM and HAA5 levels for specific

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00075 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
462 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations

locations in their distribution system to produce the national average are In a number of different contexts over
meet Stage 2 DBPR MCLs. Thus, the uncertain. the past few years, the Agency has
standard compliance forecast method as For the cost estimates, uncertainty considered the relative merits and
developed during the M/DBP FACA also exists in baseline data inputs, such assumptions encountered when
(with a 20 percent safety margin) is a as the total number of disinfecting employing meta-analyses. Cessation lag
reasonable estimation. However, SWAT plants and their typical average and modeling is a relatively recent analysis
does not explicitly consider the IDSE. design flow rates. Other cost model that the Agency has incorporated into
To address uncertainty in the impact of inputs such as labor rates and laboratory its risk analyses to more appropriately
the IDSE on the compliance forecast, fees also contain uncertainties. In these model the timing of health benefits. The
EPA revised the compliance forecast cases, EPA has evaluated available data specific papers upon which the Stage 2
methodology, assigning equal and estimated a cost input value to
analysis is based have been peer
probability to 20 and 25 percent represent the average of all water
reviewed. However, the Agency believes
operational safety margins. EPA believes systems nationally. EPA recognizes that
there is uncertainty in this average and that it is time to consider these Agency-
the 25 percent safety margin is a wide science issues in a broader sense
reasonable high-end estimate of system variability in the characteristics of
individual systems. The influence of with outside experts to better inform the
response to account for the influences of Agency’s future analyses.
the IDSE. EPA used a spatial variability these uncertainties on national cost
analysis to determine the appropriate estimates is expected to be fairly minor. For monetization of benefits, EPA
safety margin to use to estimate the For the benefits estimates, uncertainty uses two alternatives for valuing non-
impact of the IDSE on the compliance exists in model inputs such as the fatal bladder cancer. Other
forecast. estimated PAR values and the cessation uncertainties, such as the linear
lag models. EPA considered three relationship between DBP reductions
These alternative approaches for the approaches to estimate attributable risk:
compliance forecast estimate are used to and reductions in bladder cancer cases
(1) a range of risk derived from avoided, are discussed qualitatively.
represent a range of possible results and individual studies, (2) a risk estimate
are incorporated into the cost and from a meta-analysis, and (3) a risk In addition to the uncertainties
benefit models using Monte Carlo estimate from a pooled analysis. To quantified as part of the benefits
probability functions. EPA believes this quantify uncertainty in cessation lag, evaluation, other uncertainties that have
approach helps inform the reader of the three independent cessation lag models not been quantified could result in
likely magnitude of the impact of the derived from three different either an over-or under-estimation of the
uncertainties. epidemiological studies are used. Also, benefits. Two of the greatest
In addition to quantifying some two functional forms are used for each uncertainties affecting the benefits of
uncertainties in the compliance of these data sets and uncertainty in the the Stage 2 DBPR, benefits from
forecasts, EPA has explicitly accounted parameters of those functions is potential reductions of cancers other
for uncertainty in estimated treatment included in the analysis. As noted than bladder and benefits from possible
technology costs. Treatment costs are previously, causality has not been reductions in potential reproductive and
modeled using a triangular distribution established between DBP levels and developmental health effects, are
of ± 30 percent for Capital, and ± 15 cancer endpoints, so the lower bound of unquantified. Both of these factors
percent for O&M costs to recognize that potential risk reductions may be as low could result in an underestimation of
the assumptions for cost analysis to as zero. quantified Stage 2 DBPR benefits.


Potential effect on benefit estimate Potential effect on cost estimates
Section with
Assumptions for which there full discussion
is uncertainty Under-esti- Unknown im- Under-esti- Unknown im-
of uncertainty Over-estimate Over-estimate
mate pact mate pact

Uncertainty in the industry 3.4 ................ .................. .................. X ................... .................. .................. X
baseline (SDWIS and
1995 CWSS data).
Uncertainty in observed data 3.7 ................ .................. .................. X ................... .................. .................. X
and predictive tools used
to characterize DBP oc-
currence for the pre-Stage
1 baseline.
Uncertainty in predictive Chapter 5, Quantified in primary analysis (addresses po- Quantified in primary analysis (addresses po-
tools used to develop the Appendix A. tential underestimate or overestimate) tential underestimate or overestimate)
compliance forecast for
surface water systems
(SWAT and ICR Matrix
Uncertainty in ground water Chapter 5, A .................. .................. X ................... .................. .................. X
compliance forecast meth- and B.
Operational safety margin of 5.2 ................ .................. .................. X ................... .................. .................. X
wwhite on PROD1PC65 with RULES2

Impacts of the IDSE on the 5.3 ................ Quantified in the primary analysis (addresses Quantified in the primary analysis (addresses
compliance forecast for potential underestimate) potential underestimate)
the Preferred Regulatory

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00076 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations 463


Potential effect on benefit estimate Potential effect on cost estimates
Section with
Assumptions for which there full discussion
is uncertainty Under-esti- Unknown im- Under-esti- Unknown im-
of uncertainty Over-estimate Over-estimate
mate pact mate pact

Uncertainty in the PAR 6.1.1 Appen- Quantified in the primary analysis (addresses
value. dix E. range of potential effects, but true values
could lie outside range)
Reduction in TTHM and 6.3.3 ............. .................. .................. X.
HAA5 used as proxies for
all chlorination DBPs.
DBPs have a linear no- 6.2.1 ............. .................. X.
threshold dose-response
relationship for bladder
cancer effects.
Uncertainty in benefits valu- 6.5.2 ............. Quantified in the primary analysis (addresses
ation inputs. potential underestimate or overestimate)
Benefits of reduced cancers 6.7 ................ Quantified in a sensitivity analysis (addresses
other than bladder cancer potential underestimate)
are not included in the
quantitative analysis.
Value of potential reproduc- 6.8 ................ X.
tive and developmental
health effects avoided is
not quantified in the pri-
mary analysis.
Treatment costs do not in- 7.4.1 ............. .................. .................. .................. X.
clude costs for minor
operational changes pre-
dicted by SWAT.
Median operational and 7.4.1 ............. .................. .................. .................. .................. .................. X
water quality parameters
considered for technology
unit costs.
Economies of scale for com- 7.4.1 ............. .................. .................. .................. .................. X.
bination treatment tech-
nologies not considered.
Possible UV-chloramine 7.4.1 ............. .................. .................. .................. .................. X.
synergy not taken into ac-
Potential low-cost alter- 7.4.2 ............. .................. .................. .................. .................. X.
natives to treatment not
Uncertainties in unit costs ... 7.4.3 ............. .................. .................. .................. Quantified in primary analysis (addresses po-
tential overestimate or underestimate)

K. Benefit/Cost Determination for the drinking water quality and may allow Table VI.K–1 shows that the Preferred
Stage 2 DBPR some systems to avoid installing Alternative is the least cost alternative.
additional technology to meet future The Preferred Alternative has higher
The Agency has determined that the drinking water regulations. mean net benefits than Alternative 1.
benefits of the Stage 2 DBPR justify the Table VI.K–1 presents net benefits for Alternatives 2 and 3 have higher
costs. As discussed previously, the main the four regulatory alternatives benefits than the Preferred Alternative
concern for the Agency and the evaluated by EPA. This table shows that but also much greater costs. These
Advisory Committee involved in the net benefits are positive for all four regulatory alternatives do not have the
Stage 2 rulemaking process was to regulatory alternatives. Generally,
provide more equitable protection from risk-targeted design of the Preferred
analysis of net benefits is used to
DBPs across the entire distribution Alternative. Rather, a large number of
identify alternatives where benefits
system and reduce high DBP levels. The exceed costs, as well as the alternative systems would be required to make
final rule achieves this objective using that maximizes net benefits. However, treatment technology changes to meet
the least cost alternative by targeting analyses of net benefits should consider the stringent standards under these
sampling locations with high DBP levels both quantified and non-quantified regulatory alternatives. Also, because
and modifying how the annual average (where possible) benefits and costs. As causality has not been established
DBP level is calculated. This will reduce discussed previously with incremental between DBP exposure and bladder
both average DBP levels associated with net benefits, the usefulness of this cancer, actual benefits may be as low as
bladder cancer (and possibly other analysis in evaluating regulatory zero. EPA is promulgating the preferred
wwhite on PROD1PC65 with RULES2

cancers) and peak DBP levels which are alternatives for the Stage 2 DBPR is regulatory alternative because the
potentially associated with reproductive somewhat limited because many Agency believes that such a drastic shift
and developmental effects. In addition, benefits from this rule are non- in the nation’s drinking water practices
this rule may reduce uncertainty about quantified and non-monetized. is not warranted at this time.

VerDate Aug<31>2005 20:53 Jan 03, 2006 Jkt 208001 PO 00000 Frm 00077 Fmt 4701 Sfmt 4700 E:\FR\FM\04JAR2.SGM 04JAR2
464 Federal Register / Vol. 71, No. 2 / Wednesday, January 4, 2006 / Rules and Regulations


Mean annual Mean annual Mean net
Rule alternative WTP for non-fatal bladder cancer cases costs benefits benefits

3 Percent Discount Rate, 25 Years

Preferred ......................................................... Lymphoma ...................................................... $78.8 $1,530.8 $1,452

A1 ..............