Jump to main content.


National Primary Drinking Water Regulations; Arsenic and Clarifications to Compliance and New Source Contaminants Monitoring

 [Federal Register: June 22, 2000 (Volume 65, Number 121)]
[Proposed Rules]
[Page 38887-38983]
From the Federal Register Online via GPO Access [wais.access.gpo.gov]
[DOCID:fr22jn00-30]

[[Page 38887]]

-----------------------------------------------------------------------

Part II

Environmental Protection Agency

-----------------------------------------------------------------------

40 CFR Parts 141 and 142

National Primary Drinking Water Regulations; Arsenic and Clarifications
to Compliance and New Source Contaminants Monitoring; Proposed Rule

[[Page 38888]]

-----------------------------------------------------------------------

ENVIRONMENTAL PROTECTION AGENCY

40 CFR Parts 141 and 142

[WH-FRL-6707-2]
RIN 2040-AB75


National Primary Drinking Water Regulations; Arsenic and
Clarifications to Compliance and New Source Contaminants Monitoring

AGENCY: Environmental Protection Agency (EPA).

ACTION: Notice of proposed rulemaking.

-----------------------------------------------------------------------

SUMMARY: The Environmental Protection Agency (EPA) is proposing a
drinking water regulation for arsenic, as required by the 1996
amendments to the Safe Drinking Water Act (SDWA). The proposed health-
based, non-enforceable goal, or Maximum Contaminant Level Goal (MCLG),
for arsenic is zero, and the proposed enforceable standard, or maximum
contaminant level (MCL), for arsenic is 0.005 mg/L. EPA is also
requesting comment on 0.003 mg/L, 0.010 mg/L and 0.020 mg/L for the
MCL. EPA is listing technologies that will meet the MCL, including
affordable compliance technologies for three categories of small
systems serving less than 10,000 people. This proposal also includes
monitoring, reporting, public notification, and consumer confidence
report requirements and State primacy revisions for public drinking
water programs affected by the arsenic regulation.
    In addition, in this proposal the Agency is clarifying compliance
for State-determined monitoring after exceedances for inorganic,
volatile organic, and synthetic organic contaminants. Finally, EPA is
proposing that States will specify the time period and sampling
frequency for new public water systems and systems using a new source
of water to demonstrate compliance with the MCLs. The requirement for
new systems and new source monitoring will be effective for inorganic,
volatile organic, and synthetic organic contaminants.

DATES: EPA must receive public comments, in writing, on the proposed
regulations by September 20, 2000. EPA will hold a public meeting on
this proposed regulation this summer. EPA will publish a notice of the
meeting, providing date and location, in the Federal Register, as well
as post it on EPA's Office of Ground Water and Drinking Water web site
at http://www.epa.gov/safewater.

ADDRESSES: You may send written comments to the W-99-16 Arsenic
Comments Clerk, Water Docket (MC-4101); U.S. Environmental Protection
Agency; 1200 Pennsylvania Ave., NW, Washington, DC 20460. Comments may
be hand-delivered to the Water Docket, U.S. Environmental Protection
Agency; 401 M Street, SW; EB-57; Washington, DC 20460; (202) 260-3027
between 9 a.m. and 3:30 p.m. Eastern Time, Monday through Friday.
Comments may be submitted electronically to ow-docket@epamail.epa.gov.
See SUPPLEMENTARY INFORMATION for file formats and other information
about electronic filing and docket review. The proposed rule and
supporting documents, including public comments, are available for
review in the Water Docket at the above address.

FOR FURTHER INFORMATION CONTACT: Regulatory information: Irene Dooley,
(202) 260-9531, email: dooley.irene@epa.gov. Benefits: Dr. John B.
Bennett, (202) 260-0446, email: bennett.johnb@epa.gov General
information about the regulation: Safe Drinking Water Hotline, phone:
(800) 426-4791, or (703) 285-1093, email: hotline.sdwa@epa.gov.

SUPPLEMENTARY INFORMATION:

Regulated Entities

    A public water system, as defined in 40 CFR 141.2, provides water
to the public for human consumption through pipes or other constructed
conveyances, if such system has ``at least fifteen service connections
or regularly serves an average of at least twenty-five individuals
daily at least 60 days out of the year.'' A public water system is
either a community water system (CWS) or a non-community water system
(NCWS). A community water system, as defined in Sec. 141.2, is ``a
public water system which serves at least fifteen service connections
used by year-round residents or regularly serves at least twenty-five
year-round residents.'' The definition in Sec. 141.2 for a non-
transient, non-community water system [NTNCWS] is ``a public water
system that is not a [CWS] and that regularly serves at least 25 of the
same persons over 6 months per year.'' EPA has an inventory totaling
over 54,000 community water systems and approximately 20,000 non-
transient, non-community water systems nationwide. Entities potentially
regulated by this action are community water systems and non-transient,
non-community water systems. The following table provides examples of
the regulated entities under this rule.

                       Table of Regulated Entities
------------------------------------------------------------------------
                                   Examples of potentially regulated
           Category                             entities
------------------------------------------------------------------------
Industry.....................  Privately owned/operated community water
                                supply systems using ground water or
                                mixed ground water and surface water.
State, Tribal, and Local       State, Tribal, or local government-owned/
 Government.                    operated water supply systems using
                                ground water or mixed ground water and
                                surface water.
Federal Government...........  Federally owned/operated community water
                                supply systems using ground water or
                                mixed ground water and surface water.
------------------------------------------------------------------------

    The table is not intended to be exhaustive, but rather provides a
guide for readers regarding entities likely to be regulated by this
action. This table lists the types of entities that EPA is now aware
could potentially be regulated by this action. Other types of entities
not listed in this table could also be regulated. To determine whether
your facility is regulated by this action, you should carefully examine
the applicability criteria in Secs. 141.11 and 141.62 of the rule. If
you have any questions regarding the applicability of this action to a
particular entity, consult Irene Dooley, the regulatory information
person listed in the FOR FURTHER INFORMATION CONTACT section.

Additional Information for Commenters

    Please submit an original and three copies of your comments and
enclosures (including references). To ensure that EPA can read,
understand, and therefore properly respond to comments, the Agency
would prefer that comments cite, where possible, the paragraph(s) or
sections in the notice or supporting documents to which each comment
refers. Commenters should use a separate paragraph for each issue
discussed. Electronic comments must be submitted as a WordPerfect 5.1,
WP6.1

[[Page 38889]]

or WP8 file or as an ASCII file avoiding the use of special characters.
Comments and data will also be accepted on disks in WP5.1, WP6.1 or
WP8, or ASCII file format. Electronic comments on this Notice may be
filed online at many Federal Depository Libraries. Commenters who want
EPA to acknowledge receipt of their comments should include a self-
addressed, stamped envelope. No facsimiles (faxes) will be accepted.

Availability of Docket

    The docket for this rulemaking has been established under number W-
99-16, and includes supporting documentation as well as printed, paper
versions of electronic comments. The docket is available for inspection
from 9 a.m. to 4 p.m., Monday through Friday, excluding legal holidays,
at the Water Docket; EB 57; U.S. EPA; 401 M Street, SW; Washington,
D.C. For access to docket materials, please call (202) 260-3027 to
schedule an appointment.

Abbreviations Used in This Proposed Rule

>--greater than
³--greater than or equal to
--less than
£--less than or equal to
Sec. --Section
ACWA--Association of California Water Agencies
AA--activated alumina
As (III)--trivalent arsenic. Common inorganic form in water is arsenite
As (V)--pentavalent arsenic. Common inorganic form in water is arsenate
ATSDR--Agency for Toxic Substances and Disease Registry, U.S.
Department of Health & Human Services
ASTM--American Society for Testing and Materials
ASV--anodic stripping voltammetry
AWQC--Ambient Water Quality Criterion
AWWA--American Water Works Association
BAT--best available technology
BFD--Blackfoot disease
BOD--biochemical oxygen demand
BOSC--Board of Scientific Counselors, ORD
CASRN--Chemical Abstracts Service registration number
CCA--chromated copper arsenate
CCR--consumer confidence report
CDC--Centers for Disease Control and Prevention
CFR--Code of Federal Regulations
CPI--Consumer Price Index
CSFII--Continuing Survey of Food Intakes by Individuals
CV--coefficient of variation=standard deviation divided by the mean  x
100
CWS--community water system
CWSS--Community Water System Survey
DBPs--disinfection byproducts
DBPR--Disinfectants/Disinfection By-products Rule
DMA--Di-methyl arsinic acid, cacodylic acid,
(CH3)2HAsO2
DSMA--Disodium methanearsonate
DWSRF--Drinking Water State Revolving Fund
DNA--Deoxyribonucleic acid
EB--East Tower Basement
EDL--Estimated Detection Limit
EDR--Electrodialysis Reversal
e.g.--such as
EJ--Environmental Justice
EO--Executive Order
EPA--U.S. Environmental Protection Agency
FDA--Food and Drug Administration
FR--Federal Register
FTE--full-time equivalents (employees)
GDP--Gross Domestic Product
GFAA--Graphite Furnace Atomic Absorption
GHAA--Gaseous Hydride Atomic Absorption
GI--gastrointestinal
gw--ground water
HRRCA--Health Risk Reduction and Cost Analysis
IARC--International Agency for Research on Cancer
ICP-MS--Inductively Coupled Plasma Mass Spectroscopy
i.e.--that is
ICP-AES--Inductively Coupled Plasma-Atomic Emission Spectroscopy
IESWTR--Interim Enhanced Surface Water Treatment Rule
IOCs--inorganic contaminants
IRFA--Initial Regulatory Flexibility Analysis
IRIS--Integrated Risk Information System
IX--Ion exchange
K--thousands
kg--kilogram, which is one thousand grams
L--Liter, also referred to as lower case ``l'' in older citations
LC50--The concentration of a chemical in air or water which
is expected to cause death in 50% of test animals living in that air or
water
LCP--laboratory certification program
LD50--The dose of a chemical taken by mouth or absorbed by
the skin which is expected to cause death in 50% of the test animals so
treated
LOAEL--Lowest-observed-adverse-effect level
LS--lime softening
LT2ESWTR--Long-Term 2 Enhanced Surface Water Treatment Rule
M--millions
m3--Cubic meters
MCL--maximum contaminant level
MCLG--maximum contaminant level goal
MDL--method detection limit
Metro--Metropolitan Water District of Southern California
mg--Milligrams--one thousandth of gram, 1 milligram = 1,000 micrograms
mg/kg--milligrams per kilogram
mg/m3--Milligrams per cubic meter
microgram (µg)--One-millionth of gram (3.5  x  10-8
oz., 0.000000035 oz.)
µg/L--micrograms per liter
M/DBP--Microbial/Disinfection By-product
MMA--Mono-methyl arsenic, arsonic acid,
CH3H2AsO3
MOS--margin of safety
MSMA--Monosodium methanearsonate
NAOS--National Arsenic Occurrence Survey
NAS--National Academy of Sciences
NAWQA--National Ambient Water Quality Assessment, USGS
NCI--National Cancer Institute
NCWS--non-community water system
NDWAC--National Drinking Water Advisory Council
NELAC--National Environmental Laboratory Accreditation Council
NIRS--National Inorganic and Radionuclide Survey
NIST--National Institute of Standards and Technology
NOAEL--No-observed-adverse-effect level
NODA--notice of data availability
NOEL--No-observed-effect level
NPDWR--National Primary Drinking Water Regulation, OGWDW
NRC--National Research Council, the operating arm of NAS
NTNCWS--non-transient non-community water system
NTTAA--National Technology Transfer and Advancement Act of 1995
NWIS--National Water Information System
O&M--operational and maintenance
OGWDW--Office of Ground Water and Drinking Water
PBMS--Performance-Based Measurement System
PE--performance evaluation, studies to certify laboratories for EPA
drinking water testing
P.L.--Public Law
PNR--Public notification rule
POD--point of departure
POE--Point-of-entry treatment devices
POU--Point-of-use treatment devices
ppb--Parts per billion. Also, µg/L or micrograms per liter
ppm--Parts per million. Also, mg/L or milligrams per liter
PQL--Practical quantitation level
PRA--Paperwork Reduction Act
PT--performance testing
PWS--Public water systems
PWSS--Public Water Systems Supervision
RCRA--Resource Conservation and Recovery Act

[[Page 38890]]

REFs--relative exposure factors
RFA--Regulatory Flexibility Act
RfD--Reference dose
RIA--Regulatory Impact Analysis
RMCL--Recommended Maximum Contaminant Level
RO--reverse osmosis
RWS--Rural Water Survey
SAB--Science Advisory Board
SBA--Small Business Administration
SBREFA--Small Business Regulatory and Enforcement Flexibility Act, SBA
SDWA--Safe Drinking Water Act of 1974, as amended
SDWIS--Safe Drinking Water Information System
SER--Small Entity Representative for SBREFA
SISNOSE--Substantial impact on a significant number of small entities,
SBREFA
SM--Standard Methods for the Examination of Water and Wastewater
SMRs--Standardized mortality ratios, comparing deaths in test areas to
deaths in unexposed areas
SSCTs--Small System Compliance Technologies
STP-GFAA--Stabilized Temperature Platform Graphite Furnace Atomic
Absorption
SW--Office of Solid Waste publication or test method
SW-846--Solid Waste publication #846, Test Methods for Solid and
Hazardous Waste
TC--toxicity characteristic
TDS--total dissolved solids
TNC--transient, non-community
TOC--total organic carbon
µg--Microgram, 1000 micrograms = 1 milligram
UMRA--Unfunded Mandates Reform Act
U.S.--United States
USDA--U.S. Department of Agriculture
USGS--U.S. Geological Survey
USPHS--U.S. Public Health Service
VSL--Value of Statistical Life
WESTCAS--Western Coalition of Arid States
WHO--World Health Organization
WITAF--Water Industry Technical Action Fund
WS--water supply
WTP--Willingness to pay

Table of Contents

I. Summary of Regulation
II. Background
    A. What is the Statutory Authority for the Arsenic Drinking
Water Regulation?
    B. What is arsenic?
    C. What are the sources of arsenic exposure?
    1. Natural Sources of Arsenic
    2. Industrial Sources of Arsenic
    3. Dietary Sources
    4. Environmental Sources
    D. What is the regulatory history for arsenic?
    1. Earliest U.S. Arsenic Drinking Water Standards
    2. EPA's 1980 Guidelines
    3. Research and Regulatory Work
    E. EPA's Arsenic Research Plan
III. Toxic Forms and Health Effects of Arsenic
    A. What are the toxic forms of arsenic?
    B. What are the effects of acute toxicity?
    C. What cancers are associated with arsenic?
    1. Skin Cancer
    2. Internal Cancers
    D. What non-cancer effects are associated with arsenic?
    E. What are the recent developments in health effects research?
    1. Funding of Health Effects Research
    2. Expert Panel on Arsenic Carcinogenicity
    3. NAS Review of EPA's Risk Assessment
    4. May 1999 Utah Mortality Study
    5. 1999 Review of health effects
    6. Study of Bladder and Kidney Cancer in Finland
    F. What did the National Academy of Sciences/National Research
Council report?
    1. The National Research Council and its Charge
    2. Exposure
    3. Essentiality
    4. Metabolism and Disposition
    5. Human Health Effects and Variations in Sensitivity
    6. Modes of Action
    7. Risk Considerations
    8. Risk Characterization
IV. Setting the MCLG
    A. How did EPA approach it?
    B. What is the MCLG?
    C. How will a health advisory protect potentially sensitive
subpopulations?
    D. How will the Clean Water Act criterion be affected by this
regulation?
V. EPA's Estimates of Arsenic Occurrence
    A. What data did EPA evaluate?
    B. What databases did EPA use?
    C. How did EPA estimate national occurrence of arsenic in
drinking water?
    D. What are the national occurrence estimates of arsenic in
drinking water for community water systems?
    E. How do EPA's estimates compare with other recent national
occurrence estimates?
    F. What are the national occurrence estimates of arsenic in
drinking water for non-transient, non-community water systems?
    G. How do arsenic levels vary from source to source and over
time?
    H. How did EPA evaluate co-occurrence?
    1. Data
    2. Results of the Co-occurrence Analysis (US EPA, 1999f)
VI. Analytical Methods
    A. What section of SDWA requires the Agency to specify
analytical methods?
    B. What factors does the Agency consider in approving analytical
methods?
    C. What analytical methods and method updates are currently
approved for the analysis of arsenic in drinking water?
    D. Will any of the approved methods for arsenic analysis be
withdrawn?
    E. Will EPA propose any new analytical methods for arsenic
analysis?
    F. Other Method-Related Items
    1. The Use of Ultrasonic Nebulization with ICP-MS
    2. Performance-Based Measurement System
    G. What are the estimated costs of analysis?
    H. What is the practical quantitation limit?
    1. PQL determination
    2. PQL for arsenic
    I. What are the sample collection, handling and preservation
requirements for arsenic?
    J. Laboratory Certification
    1. Background
    2. What Are the Performance Testing criteria for arsenic?
    3. How often is a laboratory required to demonstrate acceptable
PT performance?
    4. Externalization of the PT Program (formerly known as the PE
Program)
VII. Monitoring and Reporting Requirements
    A. What are the existing monitoring and compliance requirements?
    B. How does the Agency plan to revise the monitoring
requirements?
    C. Can States grant monitoring waivers?
    D. How can I determine if I have an MCL violation?
    E. When will systems have to complete initial monitoring?
    F. Can I use grandfathered data to satisfy the initial
monitoring requirement?
    G. What are the monitoring requirements for new systems and
sources?
    H. How does the Consumer Confidence Report change?
    I. How will public notification change?
VIII. Treatment Technologies
    A. What are the Best Available Technologies (BATs) for arsenic?
What are the issues associated with these technologies?
    B. What are the likely treatment trains? How much will they
cost?
    C. How are variance and compliance technologies identified for
small systems?
    D. When are exemptions available?
    E. What are the small systems compliance technologies?
    F. How does the Arsenic Regulation overlap with other
regulations?
IX. Costs
    A. Why does EPA analyze the regulatory burden?
    B. How did EPA prepare the baseline study?
    1. Use of baseline data
    2. Key data sources used in the baseline analysis for the RIA?
    C. How were very large system cost derived?
    D. How did EPA develop cost estimates?
    E. What are the national treatment costs of different MCL
options?
    1. Assumptions affecting the development of the decision tree
    2. Assumptions affecting unit cost curves
X. Benefits of Arsenic Reduction
    A. Monetized Benefits of Avoiding Bladder Cancer
    1. Risk reductions: The Analytic Approach
    2. Water Consumption
    3. Monte Carlo Analysis
    4. Relative Exposure Factors
    5. NRC Risk Distributions

[[Page 38891]]

    6. Estimated Risk Reductions
    B. ``What if?'' scenario for lung cancer risks
    C. Evaluation of Benefits
    1. Fatal Risks and Value of a Statistical Life (VSL)
    2. Nonfatal Risks and Willingness to Pay (WTP)
    D. Estimates of Quantifiable Benefits of Arsenic Reduction
    F. NDWAC Working Group (NDWAC, 1988) on Benefits
XI. Risk Management Decisions: MCL and NTNCWSs
    A. What is the Proposed MCL?
    1. Feasible MCL
    2. Principal Considerations in Analysis of MCL Options
    3. Findings of NRC and Consideration of Risk Levels
    4. Non-monetized Health Effects
    5. Sources of Uncertainty
    6. Comparison of Benefits and Costs
    7. Conclusion and Request for Comment
    B. Why is EPA proposing a total arsenic MCL?
    C. Why is EPA proposing to require only monitoring and
notification for NTNCWSs?
    1. Methodology for analyzing NTNCWS risks
    2. Results
XII. State Programs
    A. How does arsenic affect a State's primacy program?
    B. When does a State have to apply?
    C. How are Tribes affected?
XIII. HRRCA
    A. What are the requirements for the HRRCA?
    B. What are the quantifiable and non-quantifiable health risk
reduction benefits?
    C. What are the Quantifiable and Non-Quantifiable Costs?
    D. What are the Incremental Benefits and Costs?
    E. What are the Risks of Arsenic Exposure to the General
Population and Sensitive Subpopulations?
    F. What are the Risks Associated with Co-Occurring Contaminants?
    G. What are the Uncertainties in the Analysis?
XIV. Administrative Requirements
    A. Executive Order 12866: Regulatory Planning and Review
    B. Regulatory Flexibility Act (RFA), as amended by the Small
Business Regulatory Enforcement Fairness Act of 1996 (SBREFA), 5
U.S.C. 601 et seq.
    1. Overview
    2. Use of Alternative Small Entity Definition
    3. Initial Regulatory Flexibility Analysis
    a. Number of Small Entities Affected
    b. Reporting, Recordkeeping and Other Requirements for Small
Systems
    4. Small Business Advocacy Review (SBAR) Panel Recommendations
C. Unfunded Mandates Reform Act (UMRA)
    1. Summary of UMRA Requirements
    a. Authorizing legislation
    b. Cost-benefit analysis
    c. Financial Assistance
    d. Estimates of future compliance costs and disproportionate
budgetary effects
    e. Macroeconomic effects
    f. Summary of EPA's consultation with State, local, and tribal
governments and their concerns
    g. Nature of State, local, and Tribal government concerns and
how EPA addressed these concerns
    h. Regulatory Alternatives Considered
    2. Impacts on Small Governments
    D. Paperwork Reduction Act (PRA)
    E. National Technology Transfer and Advancement Act (NTTAA)
    F. Executive Order 12898: Environmental Justice
    G. Executive Order 13045: Protection of Children from
Environmental Health Risks and Safety Risks
    H. Executive Order 13132: Federalism
    I. Executive Order 13084: Consultation and Coordination with
Indian Tribal Governments
    J. Request for Comments on Use of Plain Language
XV. References
List of Tables
Table V-1. Summary of Arsenic Data Sources
Table V-2. Regional Exceedance Probability Distribution Estimates
Table V-3. Statistical Estimates of Number of Ground Water CWSs with
Average Arsenic Concentrations in Specified Ranges
Table V-4. Statistical Estimates of Number of Surface Water CWSs
with Average Arsenic Concentrations in Specified Ranges
Table V-5. Comparison of CWSs from EPA, NAOS, and USGS Estimates
Exceeding Arsenic Concentrations
Table V-6. Statistical Estimates of Number of Ground Water NTNCWSs
with Average Arsenic Concentrations in Specified Ranges
Table V-7. Statistical Estimates of Number of Surface Water NTNCWSs
with Average Arsenic Concentrations in Specified Ranges
Table V-8. Correlation of Arsenic with Sulfate and Iron (surface and
ground waters)
Table V-9. Correlation of Arsenic with Radon (ground water)
Table VI-1. Approved Analytical Methods (and Method Updates) for
Arsenic (CFR 141.23)
Table VI-2. Estimated Costs for the Analysis of Arsenic in Drinking
Water
Table VI-3. Acceptance Limits and PQLs for Other Metals (in order of
decreasing PQL)
Table VII-1. Comparison of Sampling, Monitoring, and Reporting
Requirements
Table VII-2. Treatment in-place at small water systems (US EPA,
1999e and US EPA, 1999m)
Table VII-3. Table Identifying Regulatory Changes
Table VII-4. Table Listing Deleted Sections
Table VIII-1. Best Available Technologies and Removal Rates
Table VIII-2. Treatment Technology Trains
Table VIII-3. Annual Costs of Treatment Trains (Per household)
Table VIII-4. Affordable Compliance Technology Trains for Small
Systems
Table VIII-5. Affordable Compliance Technology Trains for Small
Systems
Table IX-1. Summary of General Baseline Categories of Affected
Entities
Table IX-2. List of Large Water Systems that Serve More Than 1
Million People
Table IX-3. Total Annual Costs for Large Systems for (serving more
than 1 million people)
Table IX-4. Systems Needing to Add Pre-Oxidation
Table IX-5. Percent of Systems with Coagulation-Filtration and Lime-
Softening in Place
Table IX-6. Waste Disposal Options
Table IX-7. Ground Water: Arsenic and Sulfate
Table IX-8. Surface Water: Arsenic and Sulfate
Table IX-9. Ground Water: Arsenic and Iron
Table IX-10. Surface Water: Arsenic and Iron
Table IX-11. National Annual Treatment Costs (Dollars in Millions)
Table IX-12. Total Annual Costs per Household (Dollars)
Table IX-13. Incremental National Annual Costs (Dollars in Millions)
Table IX-14. Incremental Annual Costs per Household (Dollars)
Table X-1. Source of Water Consumed
Table X-2a. Bladder Cancer Incidence Risks \1\ for High Percentile
U.S. Populations Exposed At or Above MCL Options, After Treatment
\2\ (Community Water Consumption Data \3\)
Table X-2b. Bladder Cancer Incidence Risks1 for High Percentile U.S.
Populations Exposed At or Above MCL Options, After Treatment \2\
(Total Water Consumption Data \3\)
Table X-3a. Percent of Exposed Population At 10\-4\ Risk or Higher
for Bladder Cancer Incidence\1\ After Treatment \2\ (Community Water
Consumption Data \3\)
Table X-3b. Percent of Exposed Population At 10\-4\ Risk or Higher
for Bladder Cancer Incidence\1\ After Treatment \2\ (Total Water
Consumption Data \3\)
Table X-4a. Mean Bladder Cancer Incidence Risks \1\ for U.S.
Populations Exposed At or Above MCL Options, after Treatment \2\
(Community Water Consumption Data \3\)
Table X-4b. Mean Bladder Cancer Incidence Risks \1\ for U.S.
Populations Exposed At or Above MCL Options, after Treatment \2\
(Total Water Consumption Data \3\)
Table X-5. Lifetime Avoided Medical Costs For Survivors (preliminary
estimates, 1996 dollars \1\)
Table X-6. Mean Bladder Cancer Incidence Risks \1\ for U.S.
Populations Exposed At or Above MCL Options, after Treatment \2\
(Composite of Tables X-5a and X-5b)
Table X-7. Estimated Costs and Benefits from Reducing Arsenic in
Drinking Water ($millions, 1999)
Table XI-1. Estimated Costs and Benefits from Reducing Arsenic in
Drinking Water (In 1999 $ millions)
Table XI-2. Exposure Factors Used in the NTNC Risk Assessment
Table XI-3. Composition of Non-Transient, Non-Community Water
Systems (Percentage of Total NTNC Population Served by Sector)

[[Page 38892]]

Table XI-4. Upper Bound School Children Risk Associated with Current
Arsenic Exposure in NTNC Water Systems
Table XI-5. Non-Transient Non-Community Benefit Cost Analysis
Table XI-6. Sensitive Group Evaluation Lifetime Risks
Table XIII-1. Risk Reduction from Reducing Arsenic in Drinking Water
Table XIII-2. Mean Bladder Cancer Risks and Exposed Population
Table XIII-3. Estimated Costs and Benefits from Reducing Arsenic in
Drinking Water (in 1999 $ millions)
Table XIII-4. Estimated Annualized National Costs of Reducing
Arsenic Exposures (in 1999 $ millions)
Table XIII-5. Estimated Annual Costs per Household \1\ (in 1999 $)
Table XIII-6. Summary of the Total Annual National Costs of
Compliance with the Proposed Arsenic Rule Across MCL Options (in
1997 $ millions)
Table XIII-7. Estimates of the Annual Incremental Risk Reduction,
Benefits, and Costs of Reducing Arsenic in Drinking Water
($millions, 1999)
Table XIV-1. Profile of the Universe of Small Water Systems
Regulated Under the Arsenic Rule
Table XIV-2. Average Annual Cost per CWS by Ownership
Table XIV-3. Average Compliance Costs per Household for CWSs
Exceeding MCLs
Table XIV-4. Average Compliance Costs per Household for CWSs
Exceeding MCLs as a Percent of Median Household Income
Table XIV-5. Hour Burden per Activity for Public Water Systems
Table XIV-6. Hour Burden per Activity for States and Tribes

I. Summary of Regulation

    EPA is proposing an arsenic regulation for community water systems,
which are systems that provide piped water to at least fifteen service
connections used by year-round residents or regularly serves at least
twenty-five year-round residents. This proposal will require non-
transient, non-community water systems (NTNCWS) to monitor for arsenic
and report exceedances of the MCL. The proposed health-based, non-
enforceable goal, or Maximum Contaminant Level Goal (MCLG), is zero,
based on EPA's revised risk characterization.
    EPA evaluated the analytical capability and laboratory capacity,
likelihood of water systems choosing treatment technologies for several
sizes of systems based on source water properties, and the national
occurrence of arsenic in water supplies to determine the proposed
Maximum Contaminant Level (MCL). Furthermore, the Agency analyzed the
quantifiable and nonquantifiable costs and health risk reduction
benefits likely to occur at the treatment levels considered, and the
effects on sensitive subpopulations. Based on the determination that
the costs for the feasible MCL do not justify the benefits, EPA is
proposing an MCL of 0.005 mg/L and requesting comment on 0.003 mg/L,
0.010 mg/L, and 0.020 mg/L. The treatment technologies for large
systems are primarily coagulation/filtration and lime softening, while
EPA expects that small systems (serving less than 10,000 people) will
be able to use ion exchange, activated alumina, reverse osmosis,
nanofiltration, and electrodialysis reversal. The effective date will
be five years after the final rule comes out for community water
systems serving 10,000 people or less, and three years after
promulgation for all other community water systems. EPA is proposing
that States applying to adopt the revised arsenic MCL may use their
most recently approved monitoring and waiver plans or note in their
primacy application any revisions to those plans.
    The Agency is clarifying the procedure used for determining
compliance after exceedances for inorganic, volatile organic, and
synthetic organic contaminants in this proposal. Finally, EPA is
proposing in this proposal that States will specify the time frame
which new systems and systems using a new source of water have to
demonstrate compliance with the MCL's including initial sampling
frequencies and compliance periods for new systems and systems that use
a new source of water for inorganic, volatile organic, and synthetic
organic contaminants.

II. Background

A. What Is the Statutory Authority for the Arsenic Drinking Water
Regulation?

    Section 1401 of the Safe Drinking Water Act (SDWA) requires a
``primary drinking water regulation'' to specify a maximum contaminant
level (MCL) if it is economically and technically feasible to measure
the contaminant and include testing procedures to insure compliance
with the MCL and proper operation and maintenance. In addition, section
1401(1)(D)(i) requires EPA to establish the minimum quality of
untreated, or raw, water taken into a public water system. A national
primary drinking water regulation (NPDWR) that establishes an MCL also
lists the technologies that are feasible to meet the MCL, but systems
are not required to use the listed technologies (section
1412(b)(3)(E)(i)). As a result of the 1996 amendments to SDWA, when
issuing a NPDWR, EPA will also list affordable technologies for small
systems serving 10,000 to 3301, 3300 to 501, and 500 to 25 that achieve
compliance with the MCL or treatment technique. EPA can list modular
(packaged) and point-of-entry and point-of-use treatment units for the
three small system sizes, as long as the units are maintained by the
public water system or its contractors. Home units must contain
mechanical warnings to notify customers of problems (section
1412(b)(4)(E)(ii)). In section 1412(b)(12)(A) of SDWA, as amended
August 6, 1996, Congress directed EPA to propose a national primary
drinking water regulation for arsenic by January 1, 2000 and issue the
final regulation by January 1, 2001. At the same time, Congress
directed EPA to develop a research plan by February 2, 1997 to reduce
the uncertainty in assessing health risks from low levels of arsenic
and conduct the research in consultation with the National Academy of
Sciences, other Federal agencies, and interested public and private
entities. The amendments allowed EPA to enter into cooperative
agreements for research.
    Section 1412(a)(3) requires EPA to propose a maximum contaminant
level goal (MCLG) simultaneously with the national primary drinking
water regulation. The MCLG is defined in section1412(b)(4)(A) as ``the
level at which no known or anticipated adverse effects on the health of
persons occur and which allows an adequate margin of safety.'' Section
1412(b)(4)(B) specifies that each national primary drinking water
regulation will specify a maximum contaminant level (MCL) as close to
the MCLG as is feasible, with two exceptions added in the 1996
amendments. First, the Administrator may establish an MCL at a level
other than the feasible level if the treatment to meet the feasible MCL
would increase the risk from other contaminants or the technology would
interfere with the treatment of other contaminants (section1412(b)(5)).
Second, if benefits at the feasible level would not justify the costs,
EPA may propose and promulgate an MCL ``that maximizes health risk
reduction benefits at a cost that is justified by the benefits (section
1412(b)(6)).''
    When proposing an MCL, EPA must publish, and seek public comment
on, the health risk reduction and cost analyses (HRRCA) of each
alternative maximum contaminant level considered (section
1412(b)(3)(C)(i)). This includes the quantifiable and nonquantifiable
benefits from reductions in health risk, including those from removing
co-occurring contaminants (not counting benefits resulting from
compliance with other proposed or final regulations), costs of
compliance (not counting costs resulting from other regulations), any
increased health risks (including those from co-occurring contaminants)
that

[[Page 38893]]

may result from compliance, incremental costs and benefits of each
alternative MCL considered, and the effects on sensitive subpopulations
(e.g., infants, children, pregnant women, elderly, seriously ill, or
other groups at greater risk). EPA must analyze the quality and extent
of the information, the uncertainties in the analysis, and the degree
and nature of the risk.
    The 1996 amendments also require EPA to base its action on the best
available, peer-reviewed science and supporting studies and to present
health effects information to the public in an understandable fashion.
To meet the latter obligation, EPA must specify, among other things,
the methodology used to reconcile inconsistencies in the scientific
data for the final regulation (section 1412(b)(3)(B)(v)).
    Section 1451(a) allows EPA to delegate primary enforcement
responsibility to federally recognized Indian Tribes, providing grant
and contract assistance, using the procedures applied to States.
Section 1413(a)(1) allows EPA to grant States primary enforcement
responsibility for NPDWRs when EPA has determined that the State has
adopted regulations that are no less stringent than EPA's. States must
adopt comparable regulations within two years of EPA's promulgation of
the final rule, unless a two-year extension is justified. State primacy
also requires, among other things, adequate enforcement (including
monitoring and inspections) and reporting. EPA must approve or deny
State applications within 90 days of submission (section 1413(b)(2)).
In some cases, a State submitting revisions to adopt a national primary
drinking water regulation has enforcement authority for the new
regulation while EPA action on the revision is pending (section
1413(c)).

B. What Is Arsenic?

    Arsenic is an element that occurs naturally in rocks, soil, water,
air, plants, and animals. Arsenic is a metalloid, which exhibits both
metallic and nonmetallic chemical and physical properties. The primary
valence states for arsenic are 0, -3, +3 and +5. Although arsenic is
found in nature to a small extent in its elemental form (0 valence), it
occurs most often as inorganic and organic compounds in either the As
(III) (+3) or As (V) (+5) valence states. The trivalent forms of
inorganic arsenic [As (III) (e.g., arsenite,
H3AsO3)] and the pentavalent forms [As (V) (e.g.,
arsenate, H2AsO4-,
HAsO42-)] are inorganic species which tend to be
more prevalent in water than the organic arsenic species (Irgolic,
1994; Clifford and Zhang, 1994). The dominant inorganic species present
in water is largely a function of the pH and the oxidizing/reducing
conditions which affects the need for pretreatment and removal effects.
Arsenates are more likely to occur in aerobic surface waters and
arsenites are more likely to occur in anaerobic ground waters.

C. What Are the Sources of Arsenic Exposure?

1. Natural Sources of Arsenic
    There are numerous natural sources as well as human activities that
may introduce arsenic into food and drinking water. The primary natural
sources include geologic formations (e.g., rocks, soil, and sedimentary
deposits), geothermal activity, and volcanic activity. Arsenic and its
compounds comprise 1.5-2% of the earth's crust (Welch, personal
communication). While concentrations of arsenic in the earth's crust
vary, the average concentrations are generally reported to range from
1.5 to 5 mg/kg. Arsenic is a major constituent of many mineral species
in igneous and sedimentary rocks. It is commonly present in the sulfide
ores of metals including copper, lead, silver, and gold. There are over
100 arsenic-containing minerals, including arsenic pyrites (e.g.,
FeAsS), realgar (AsS), lollingite (FeAs2,
Fe2As3, Fe2As5), and
orpiment (As2S3). Geothermal water can be a
source of inorganic arsenic in surface water and ground water. Welch et
al. (1988) identified fourteen areas in the Western United States where
dissolved arsenic concentrations ranged from 80 to 15,000 µg/L.
In addition, natural emissions of arsenic are associated with forest
fires and grass fires. Volcanic activity appears to be the largest
natural source of arsenic emissions to the atmosphere (ATSDR, 1998).
Arsenic compounds, both inorganic and organic, are also found in food.
2. Industrial Sources of Arsenic
    Major present and past sources of arsenic include wood
preservatives, agricultural uses, industrial uses, mining and smelting.
The human impact on arsenic levels in water depends on the level of
human activity, the distance from the pollution sources, and the
dispersion and fate of the arsenic that is released. The production of
chromated copper arsenate (CCA), an inorganic arsenic compound and wood
preservative, accounts for approximately 90% of the arsenic used
annually by industry in the United States (USGS, 1998; USGS, 1999). CCA
is used to pressure treat lumber, which is typically used for the
construction of decks, fences, and other outdoor applications. In
addition to wood preservatives, the other EPA-registered use of
inorganic arsenic is for sealed ant bait. In the past, agricultural
uses of arsenic included pesticides, herbicides, insecticides,
defoliants, and soil sterilants. Inorganic arsenic pesticides are no
longer used for agricultural purposes; the last agricultural
application was voluntarily canceled in 1993 (58 FR 64579, US EPA,
1993b).
    Organic forms of arsenic are constituents of some agricultural
pesticides that are currently used in the U.S. Monosodium
methanearsonate (MSMA) is the most widely applied organoarsenical
pesticide, which is used to control broadleaf weeds and is applied to
cotton (Jordan et al., 1997). Small amounts of disodium methanearsonate
(DSMA, or cacodylic acid) are also applied to cotton fields as
herbicides. The Food and Drug Administration regulates other organic
arsenicals (e.g., roxarsone and arsanilic acid) used as feed additives
for poultry and swine for increased rate of weight gain, improved feed
efficiencies, improved pigmentation, and disease treatment and
prevention. These additives undergo little or no metabolism before
excretion (NAS, 1977; Moody and Williams, 1964; Aschbacher and Feil,
1991).
    Arsenic and arsenic compounds (arsenicals) are used for a variety
of industrial purposes, including: electrophotography, catalysts,
pyrotechnics, antifouling paints, pharmaceutical substances, dye and
soaps, ceramics, alloys (automotive solder and radiators), battery
plates, optoelectronic devices, semiconductors, and light emitting
diodes in digital watches (Azcue and Nriagu, 1994). In addition,
burning of fossil fuels, combustion of wastes, mining and smelting,
pulp and paper production, glass manufacturing, and cement
manufacturing can result in emissions of arsenic to the environment (US
EPA, 1998). Arsenic has been identified as a contaminant of concern at
916 of the 1,467 National Priorities List (Superfund) hazardous waste
sites (ATSDR, 1998).
3. Dietary Sources
    Because arsenic is naturally occurring, the entire population is
exposed to low levels of arsenic through food, water, air, and contact
with soil. The National Research Council report (NRC, 1999) described
in sections III.C. and III.E.3. provides Food and Drug Administration
(FDA) ``market basket'' data for total arsenic intake by age

[[Page 38894]]

group. NRC assumed that, for fish and seafood, inorganic arsenic is 10%
of the total arsenic and that other food contains entirely inorganic
arsenic. These assumptions are probably high and conservative for
public health protection to avoid underestimating the contributions
from food. Table 3-5 in the 1999 NRC report characterizes inorganic
arsenic intake from food in the U.S. as being 1.3 µg/day for
infants under one year old, 4.4 µg/day for 2-year olds, almost
10 µg/day for 25-30 year-old males, with a maximum of 12.5
µg/day for 60-65 year-old males (females had lower arsenic
intake in every age group). MacIntosh et al. (1997) estimated that 785
adults had a mean inorganic arsenic consumption of 10.22 µg/
day, with a standard deviation of 6.54 µg/day and a range of
0.36-123.84 µg/day based on semi-quantitative food surveys.
    Likewise, the 2 L/day assumption of adult drinking water intake
used to develop the MCLG does not represent intake by the average
person; rather it represents intake of a person in the 90th percentile.
(See Section X.B.1.a. for a description of water consumption for the
general population.)
4. Environmental Sources
    Internal exposure after skin contact with water or soil containing
arsenic or inhalation of arsenic from air is believed to be low.
Studies of inorganic arsenic absorption from skin from cadavers
estimated 0.8% uptake from soil and 1.9% uptake from water over a 24-
hour period (Wester et al., 1993). EPA's arsenic health assessment
document for the Clean Air Act (US EPA, 1984) cited respiratory arsenic
as being about 0.12 µg/day from a daily ventilation rate of 20
m\3\ using a 1981 national average arsenic air concentration of 0.006
µg/m\3\. Assuming 30 percent absorption, the daily amount of
arsenic from breathing would be 0.03 µg, so air is a minor
source of arsenic (50 FR 46936 at 46960; US EPA, 1985b). At this time,
EPA is basing health risks on estimates of arsenic exposure from food
and water. The Centers for Disease Control and Prevention (CDC) is
initiating a study of arsenic intake from bathing. EPA requests comment
on whether available data on skin absorption and inhalation indicate
that these are significant exposure routes that should be considered in
the risk assessment.

D. What is the Regulatory History for Arsenic?

    Regulation of arsenic has been the subject of scientific debate
that has lasted for decades despite research and scientific review. The
controversy has affected policy and regulatory decisions for arsenic in
drinking water from low, environmental exposure.
1. Earliest U.S. Arsenic Drinking Water Standards
    In 1942 the U.S. Public Health Service first established an arsenic
drinking water standard for interstate water carriers at 0.05 mg
arsenic per liter (mg/L, or 50 µg/L), as measured with a
colorimetric method. The report did not cite any reason for choosing
that level, but it defined ``safety of water supplies'' as ``the
danger, if any, is so small that it cannot be discovered by available
means of observation (US Public Health Service 1943).'' In 1946, the
Surgeon General of the U.S. Public Health Service noted that the
American Water Works Association had accepted the 1942 drinking water
standards, including the arsenic standard (U.S. Public Health Service
1946). In 1962 (U.S. Public Health Service 1962) the U.S. Public Health
Service issued more stringent drinking water standards for arsenic of
0.01 mg/L (10 µg/L) for a water supply in 42 CFR 72.205(b)(1)
and 0.05 mg/L in 42 CFR 72.205(b)(2) as grounds for rejection of a
water supply, as measured by the current edition of Standard Methods
for the Examination of Water and Wastewater per 42 CFR 72.207(a).
    The Safe Drinking Water Act of 1974 amended the Public Health
Service Act and specified that EPA set primary and secondary drinking
water standards. On December 24, 1975 (40 FR 59566 at 59570; US EPA,
1975), EPA issued a National Interim Primary Drinking Water Regulation
for arsenic in Sec. 141.23(b) of 0.05 mg/L (50 µg/L), effective
18 months later (Sec. 141.6). Commenters recommended an MCL of 100
µg/L, saying there were no observed adverse health effects (40
FR 59566 at 59576; US EPA, 1975). EPA noted long-term chronic effects
at 300-2,750 µg/L, but observed no illnesses in a California
study at 120 µg/L. Drinking 2 liters of water a day containing
arsenic at 50 µg/L would provide approximately 10% of total
ingested arsenic from food and water, estimated to be 900 µg/
day. The section on arsenic noted that arsenic has been believed to be
a carcinogen ``[s]ince the early nineteenth century  * *; however
evidence from animal experiments and human experience has accumulated
to strongly suggest that arsenicals do not produce cancer. One
exception is a report from Taiwan * * *. The text goes on to note
occupational skin and lung cancer from arsenic dust and skin cancer in
England from drinking water with 12 mg/L. (US EPA, 1976 Appendix A).
2. EPA's 1980 Guidelines
    Scientific data at the time the 1980 Ambient Water Quality
Guidelines were formulated did not support a safe or ``threshold''
concentration for carcinogens, so EPA's public health policy was

``that the recommended concentration for maximum protection of human
health is zero. In addition, the Agency presented a range of
concentrations corresponding to incremental cancer risks of 10-\7\
to 10-\5\ (one additional case of cancer in populations ranging from
ten million to 100,000, respectively) * * * [that did not
necessarily represent] an Agency judgement on an `acceptable' risk
level (45 FR 79318 at 79323, US EPA, 1980).''

    In the November 28, 1980 Federal Register document, using its then
current risk assessment approach (assumed toxicity increased as a
natural logarithm linear function across species), EPA set the Clean
Water Act surface water quality criterion for arsenic at 2.2 nanograms
(ng/L) (0.0022 µg/L) at an increased cancer risk of 10-\6\. The
criterion was to prevent skin cancer in humans drinking contaminated
water and eating aquatic organisms from those water bodies (45 FR 79318
at 79326). The 1980 Federal Register notice indicated that drinking
water standards consider a range of factors, including health effects,
technological and economic feasibility of removal, and monitoring
capability. On the other hand the Clean Water Act criteria of section
304(a)(1) ``have no regulatory significance under the SDWA.'' The Clean
Water Act section 304(a)(1) criteria are more similar to the health-
based goals of the recommended maximum contaminant levels (now referred
to as MCLGs), than to MCLs; and differences in mandates ``may result in
differences between the two numbers.'' (45 FR 79318 at 79320; US EPA,
1980). In 1992, the Clean Water Act criterion was recalculated based on
the updated cancer risk assessment in EPA's Integrated Risk Information
System (IRIS) database, to a level of 0.018 µg/L for arsenic at
a 10-\6\ cancer risk (57 FR 60848; US EPA, 1992c).
3. Research and Regulatory Work
    The 1980 National Academy of Science (NAS) Volume III of ``Drinking
Water and Health'' report encouraged EPA to research whether arsenic is
essential for humans, as demonstrated in four studies of mammalian
species. The 1983 NAS Volume V report projected that 0.05 mg/kg of
total arsenic may be a desirable level for people, and 25 to 50
µg a day may be required (as cited in 50 FR 46936 at 46960; US
EPA, 1985b).

[[Page 38895]]

    In 1983, EPA requested comment on whether the arsenic MCL should
consider carcinogenicity, other health effects, and nutritional
requirements, and whether MCLs are necessary for separate valence
states (e.g., arsenite vs. arsenate) (48 FR 45502 at 45512; US EPA,
1983). On November 13, 1985, EPA proposed (50 FR 46936; US EPA, 1985b)
a recommended maximum contaminant level (RMCL), a non-enforceable
health goal now known as an MCLG, of 50 µg/L based on the 1983
NAS conclusion that 50 µg/L balanced toxicity and possible
essentiality and provided ``a sufficient margin of safety'' (50 FR
46936 at 46960). EPA also requested comment on alternate RMCLs of 100
µg/L based on noncarcinogenic effects (calculated from an
animal study and an uncertainty factor of 1000) and 0 µg/L
based on carcinogenicity (50 FR 46936 at 46961). EPA chose not to base
the proposed RMCL on the animal study because each dose group had only
four Rhesus monkeys. Also, at that time, studies had ``not detected
increased risks via drinking water in the USA'' (50 FR 46936 at 46960).
The 1985 proposed drinking water regulation preamble noted the 1980
excess cancer risk values derived from the ambient water quality
criteria were based on skin cancer using the 1968 Tseng et al. study
(50 FR 46936 at 46961).
    The June 19, 1986 amendments to the Safe Drinking Water Act (SDWA;
Public Law 99-339) converted the 1975 interim arsenic standard to a
National Primary Drinking Water Regulation (section 1412(a)(1)),
subject to revision by 1989 (section 1412(b)(1)). Review of the arsenic
risk assessment issues caused the Agency to miss the 1989 deadline for
proposing a revised NPDWR. As a result of a citizen suit to enforce the
deadline, EPA entered into a consent decree providing deadlines for
issuing the arsenic rule.
    In 1988, EPA's Risk Assessment Forum issued the Special Report on
Ingested Inorganic Arsenic: Skin Cancer; Nutritional Essentiality (EPA/
625/3-87/013), in part, to evaluate the validity of applying skin
cancer data from Taiwanese studies (published in 1968 and 1977) in
dose-response assessments in the U.S. As described in the report, the
maximum likelihood estimate of risk ranged from 3  x  10-\5\ to 7  x
10-\5\ for a 70-kilogram person consuming 2 liters of water per day
contaminated with 1 µg of arsenic per liter. Calculated at the
50 µg/L standard, the U.S. lifetime risk of skin cancer ranged
from 1  x  10-\3\ to 3  x  10-\3\, which means one to three skin
cancers would occur in a group of one thousand people drinking water
containing arsenic at 50 µg/L. Existing studies could not
determine whether arsenic was an essential nutrient.
    After reviewing the scientific evidence for carcinogenicity, EPA's
Science Advisory Board (US EPA, 1989a and b) stated in its August 1989
and September 1989 reports that (1) the animal studies suggesting
arsenic is an essential nutrient are not definitive; (2) the skin
changes seen in hyperkeratosis may not always result in skin cancer;
(3) the 1968 Taiwan data demonstrate that high doses of ingested
arsenic can cause skin cancer; (4) the Taiwan study is inconclusive to
determine cancer risk at levels ingested in the United States (U.S.);
and (5) As (III) levels below 200-250 µg per day may be
detoxified. SAB recommended that EPA set the MCL using a non-linear
dose-response (at some low dose, arsenic would not be toxic). The SAB
report recommended that EPA revise the risk assessment based on dose of
arsenic to target tissues (the concentration of arsenic that damages
tissues, rather than the concentration in water) and consider
detoxification.
    The SAB also reviewed EPA's April 12, 1991 Arsenic Research
Recommendations (US EPA, 1991c). The final report provided SAB's
recommendations (US EPA, 1992a) and ``identified research needed to
resolve major uncertainties about inorganic arsenic cancer risk'' to
evaluate if work could be done in three to five years. It noted that
``important work can be done within the time available. Although the
results from this work will not completely resolve any issue, * * * the
results will likely significantly improve the Agency's ability to
evaluate the risk. * * * through improved knowledge of arsenic
metabolism and * * * as a carcinogen.'' The report reflected
uncertainty as to whether or not EPA could obtain enough data to
regulate arsenic using a non-linear model, which needed more
information on how arsenic induces cancer. The group noted that it
would take longer than five years to develop an animal model to help
understand the toxicity of arsenic. SAB recommended four short-term
studies: (a) Investigation of chromosome damage, arsenic metabolites,
and the times cells are most susceptible to arsenic, (b) study of human
liver capacity to add methyl groups to arsenic, (c) identifying the
species in urine in several populations to look for evidence of
saturation of methylation enzymes, and (d) comparing methylated arsenic
excreted in the U.S., Taiwan, Mexico, and Argentina to consider the
effect of nutritional or genetic differences on methylation capacity.
However, if time were not a factor, SAB ranked developing an animal
model of arsenic-induced cancer as the first priority.
    In 1993 SAB reviewed EPA's draft ``Drinking Water Criteria Document
on Inorganic Arsenic (US EPA, 1993a).'' In 1995, SAB reviewed the
analytical methods, occurrence estimate, treatment technologies, and
approach for assigning costs in the regulatory impact analysis (US EPA,
1995). Besides highlighting previous SAB reviews of 1989, 1992, and
1994 on health effects, the 1995 report recommended changes to the
practical quantitation limit approach, use of occurrence data, review
of technologies, and support for the decision tree, with some
reservations.
    EPA held internal workgroup meetings throughout 1994, addressing
risk assessment, treatment, analytical methods, arsenic occurrence,
exposure, costs, implementation issues, and regulatory options. EPA
decided in early 1995 to defer the arsenic regulation in order to
better characterize health effects and assess cost-effective removal
technologies for small utilities.
    The 1996 amendments to SDWA included a new statutory deadline for
the arsenic regulations, as discussed in section II.A.

E. EPA's Arsenic Research Plan

    EPA held a workshop in March 1994 entitled ``Workshop on Developing
an Epidemiology Research Strategy for Arsenic in Drinking Water.'' The
cover letter to the final report (US EPA, 1997b), dated April 14, 1997,
notes that EPA has been using the recommendations to direct its
research directions. The report listed ten projects and seventeen
conclusions on exposure, endpoints, study design and statistical power,
population selection, feasibility of conducting a study in the U.S.,
international studies, importance of developing biomarkers to measure
health effects of arsenic, and animal studies.
    In 1995, the Water Industry Technical Action Fund (WITAF) ( funded
by the American Water Works Association, National Association of Water
Companies, Association of Metropolitan Water Agencies, National Rural
Water Association, and National Water Resources Association), the AWWA
Research Foundation, and the Association of California Water Agencies
(ACWA) sponsored an Expert Workshop on Arsenic Research Needs in
Ellicott City, MD, May 31-June 2, 1995. The final report (AWWA et al.,
1995) identified research projects in mechanisms, epidemiology,
toxicology, and treatment. It identified ten high

[[Page 38896]]

priority projects which would need over $3 million to fund, eleven
medium priority projects needing over $6 million, and ten low priority
projects costing over $9 million, that totaled over $19 million in
research needs.
    Congress recognized the importance of health effects research in
regulating arsenic, as demonstrated by the 1996 statutory requirement
to develop a research plan within 180 days ``in support of drinking
water rulemaking to reduce the uncertainty in assessing health risks
associated with exposure to low levels of arsenic * * * (section
1412(b)(12)(A)(ii)). In the research plan EPA recognized that ``[t]he
research needs are broader than those that EPA can address alone, and
it is anticipated that other entities will be involved in conducting
some of the needed research (US EPA, 1998a).'' (See section III.E.1. on
industry-funded research and the arsenic research plan (at www.epa.gov/
ORD/WebPubs/final/arsenic.pdf) for EPA-funded projects.) In December
1996, EPA submitted its draft research plan for peer review by its
Board of Scientific Counselors' (BOSC) Ad Hoc Committee, and the
committee met in January 1997. The February 1998 Arsenic Research Plan
addressed the June 1997 comments from BOSC.
    Major areas covered in the research plan included studies to:
    
  • Improve our qualitative and quantitative assessment of the human toxicity of arsenic;
  • Understand mechanisms of arsenic toxicity that may aid in extension of the observed human findings when extrapolation is required;
  • Measure exposures of the US population to arsenic from various sources (particularly diet) to allow better definition of cumulative exposures to arsenic;
  • Refine treatment technologies that may better remove arsenic from water supplies;
  • Improve methods for analyzing and monitoring arsenic in drinking water. EPA also set priorities in the plan and identified projects that met the short term and long term criteria: Short Term Criteria 1. Will the research improve the scientific basis for risk assessments needed for proposing a revised arsenic MCL by January 1, 2000? 2. Will the research improve the scientific basis for risk management decisons needed for proposinig a revised arsenic MCL by January 1, 2000? Long Term Criteria 1. Will the research improve the scientific basis for risk assessment and risk management decisions needed to review and develop future MCLs beyond the year 2000? 2. Is the research essential to improving our scientific understanding of the health risks of arsenic? The research plan included the following priority topics for research under the five major areas of investigation supporting drinking water rulemaking: Exposure Analysis
  • Arsenic speciation and preservation: Improvements in analytical methods to support water treatment decisions.
  • Measurement of background exposures to arsenic in U.S. population, particularly inorganic arsenic intake in the U.S. diet.
  • Development and evaluation of biomarkers (e.g., species of arsenic in urine) of exposures.
  • Development of standard reference material for arsenic in water, food, urine, tissues. Cancer Effects
  • Further study of internal cancers associated with arsenic exposures.
  • Dose response data on hyperkeratosis as a likely precursor to skin cancer.
  • Research on factors influencing human susceptibility including age, genetic characteristics and dietary patterns.
  • Metabolic and pharmacokinetic studies that can identify dose dependent metabolism.
  • Mechanistic studies for arsenic-induced genotoxicity and carcinogenicity. Noncancer Effects
  • Development of human dose-response data for hyperkeratosis, cardiovascular disease, neurotoxicity and developmental effects.
  • Development of additional health effects and hazard identification data on other non-cancer endpoints such as diabetes and hematologic effects. Risk Management Research
  • Identification of limitations of treatment technologies and impacts on water quality.
  • Development of treatment technologies for small water systems.
  • Development of data on cost and performance capabilities of various treatment options.
  • Consideration of residuals management issues, including disposal options and costs. Risk Assessment/Characterization
  • Development of risk characterizations to provide interim support to States and local communities.
  • Development of predictive tools and statistical models for assessing bioavailability, interactions and dose-response as better mass balance data become available.
  • Comprehensive assessment of exposure levels and incorporation of data into risk estimates for better characterization of actual risks associated with arsenic exposure.
  • Comprehensive assessment of arsenic mode of action provide a greater understanding of biological mechanisms and factors that may impact the shape of a dose response curve.
  • Comprehensive assessment of non-cancer risks and consideration of appropriate modeling tools for quantitative estimation of non-cancer risks.
  • Comprehensive assessment of human dose-response data for hyperkeratosis, cardiovascular disease, neurotoxicity and developmental effects. III. Toxic Forms and Health Effects of Arsenic A. What Are the toxic Forms of Arsenic? Arsenic exists in several forms which vary in toxicity and occurrence. Accordingly, for this proposed regulation, it is important to consider those forms that can exert toxic effects and to which people may be exposed. For example, the metallic form of arsenic (0 valence) is not absorbed from the stomach and intestines and does not exert adverse effects. On the other hand, a volatile compound such as arsine (AsH3) is toxic, but is not present in water or food. Moreover, the primary organic forms (arsenobetaine and arsenocholine) found in fish and shellfish seem to have little or no toxicity (Sabbioni et al., 1991). Arsenobetaine quickly passes out of the body in urine without being metabolized to other compounds (Vahter, 1994). Arsenite (+3) and arsenate (+5) are the most prevalent toxic forms of inorganic arsenic that are found in drinking water. However, recovery of identified arsenic species in vegetables, grains and oils has been limited and difficult, so little is known about types of species in these foods (NRC, 1999). In animals and humans, inorganic pentavalent arsenic is converted to trivalent arsenic that can be methylated (i.e., chemically bonded to a methyl group, which is a carbon atom linked to [[Page 38897]] three hydrogen atoms) to mono-methyl arsenic (MMA) and di-methyl arsinic acid (DMA), which are organic arsenicals. The primary route of excretion for arsenic metabolites is in the urine. Studies indicate that the organic arsenicals MMA and DMA were hundreds of times less likely to produce genetic changes in animal cells than inorganic arsenicals. Moreover, many studies reported organic arsenicals to be less reactive in tissues, to kill less cells, and to be more easily excreted in urine (NRC, 1999). B. What Are the Effects of Acute Toxicity? Inorganic arsenic can exert toxic effects after acute (short-term) or chronic (long-term) exposure. From human acute poisoning incidents, the LD50 of arsenic has been estimated to range from 1 to 4 mg/kg (Vallee et al., 1960, Winship, 1984). This dose would correspond to a lethal dose range of 70 to 280 mg for 50% of adults weighing 70 kg. At nonlethal, but high acute doses, inorganic arsenic can cause gastroenterological effects, shock, neuritis (continuous pain) and vascular effects in humans (Buchanan, 1962). Such incidents usually occur after accidental exposures. However, sometimes high dose acute exposures may be self-administered. For example, inorganic arsenic is a component of some herbal medicines and adverse effects have been reported after use. In one report of 74 cases (Tay and Seah, 1975), the primary signs were skin lesions (92%), neurological (i.e., nerve) involvement (51%), and gastroenterological, hematological (i.e., blood) and renal (i.e., kidney) effects (19 to 23%). Although acute or short- term exposures to high doses of inorganic arsenic can cause adverse effects, such exposures do not occur from public water supplies in the U.S. at the current MCL of 50 µg/L. EPA's proposed drinking water regulation addresses the long-term, chronic effects of exposure to low concentrations of inorganic arsenic in drinking water. C. What Cancers Are Associated With Arsenic? Inorganic arsenic is a multi-site human carcinogen by the drinking water route. Asian, Mexican and South American populations with exposures to arsenic in drinking water generally at or above several hundred micrograms per liter are reported to have increased risks of skin, bladder, and lung cancer. The current evidence also suggests that the risks of liver and kidney cancer may also be increased following exposures to inorganic forms of arsenic. The weight of evidence for ingested arsenic as a causal factor of carcinogenicity is much greater now than a decade ago, and the types of cancer occurring as a result of ingesting inorganic arsenic have even greater health implications for U.S. and other populations than the occurrence of skin cancer alone. (Until the late 1980s skin cancer had been the cancer classically associated with arsenic in drinking water.) Epidemiologic studies (e.g., of people) provide direct data on arsenic risks from drinking water at exposure levels much closer to those of regulatory concern than environmental risk assessments based on animal toxicity studies. 1. Skin Cancer Early reports linking inorganic arsenic contamination of drinking water to skin cancer came from Argentina (Neubauer, 1947, reviewing studies published as early as 1925) and Poland (Geyer, 1898, as reported in Tseng et al., 1968). However, the first studies that observed dose-dependent effects of arsenic associated with skin cancer came from Taiwan (Tseng et al., 1968; Tseng, 1977). These studies focused EPA's attention on the health effects of ingested arsenic. Physicians physically examined over 40,000 residents from 37 villages and 7500 residents exposed to 0.017 mg/L arsenic ( reference group). The study population was divided into three groups based on exposure to inorganic arsenic (0 to 0.29, 0.30 to 0.59 and 0.60 mg of inorganic As/Liter) measured at the village level. A dose-and age- related increase of arsenic-induced skin cancer among the villagers was noted. No skin cancers were observed in the low arsenic reference areas. The 1999 NRC report noted that the ``primary limitation of this study * * * was the lack of detail'' reported, such as grouping individuals into ``broad exposure groups'' (rather than grouping into 37 village exposures). This limits the usefulness of these studies. However, these Tseng reports and other corroborating studies such as those by Albores et al. (1979) and Cebrian et al. (1983) on drinking water exposure and exposures to inorganic arsenic in medicines (Cuzick et al., 1982) and in pesticides (Roth, 1956) led the EPA, using skin cancer as the endpoint, to classify inorganic arsenic as a human carcinogen (Group A) by the oral route (US EPA, 1984). 2. Internal Cancers Exposure to inorganic arsenic in drinking water has also been associated with the development of internal cancers. ``No human studies of sufficient statistical power or scope have examined whether consumption of arsenic in drinking water at the current MCL results in an increased incidence of cancer or noncancer effects (NRC, 1999, pg. 7).'' Chen et al. (1985) used standardized mortality ratios (SMRs) to evaluate the association between ingested arsenic and cancer risk in Taiwan. (SMRs, ratios of observed to expected deaths from specific causes, are standardized to adjust for differences in the age distributions of the exposed and reference populations.) The authors found statistically significant increased risks of mortality for bladder, kidney, lung, liver and colon cancers. A subsequent mortality study in the same area of Taiwan found significant dose-response relationships for deaths from bladder, kidney, skin, and lung cancers in both sexes and from liver and prostate cancer for males. They also found increases in peripheral and cardiovascular diseases but not in cerebrovascular accidents (Wu et al., 1989). There are several corroborating reports of the increased risk of cancers of internal organs from ingested arsenic including two from two South American countries. In Argentina, significantly increased risks of death from bladder, lung and kidney cancer were reported (Hopenhayn-Rich et al., 1996; 1998). In a population of approximately 400,000 in northern Chile, Smith et al. (1998) found significantly increased risks of bladder and lung cancer mortality. There have only been a few studies of inorganic arsenic exposure via drinking water in the U.S., and most have not considered cancer as an endpoint. People have written EPA asking that the new MCL be set considering that these U.S. studies have not seen increases in cancers at the low levels of arsenic exposure in U.S. drinking water. Optimally, low-exposure arsenic studies involve long-term residency (20-40 years with known drinking water arsenic exposure), access to health records, populations large enough to detect statistically significant increases in cancers and other health endpoints, and limited use of multiple sources of water (bottled, filtered, beverages, food prepared outside the home). Recently, Lewis et al. (1999) conducted a mortality study of a population in Utah whose drinking water contained relatively low concentrations of arsenic (averaged 18-191 µg/L). They reported no significant increase in bladder or lung mortality. They did report a statistically significant dose-response for an increased risk of prostate cancer mortality. Smoking is an established risk factor for bladder and lung cancer, and inorganic arsenic [[Page 38898]] behaves as a comutagen even though it is not mutagenic alone (NRC, 1999, pg. 200). It is possible that inorganic arsenic potentiates other risk factors for these cancers. This potential role is consistent with the NRC, 1999 view that arsenic's mode of action may be to interfere with cell ``housekeeping'' functions that normally repair genetic damage and ensure that damaged cells die (programmed cell death) rather than reproduce (see section III.D.2. below). D. What Non-Cancer Effects Are Associated With Arsenic? A large number of adverse noncarcinogenic effects have been reported in humans after exposure to drinking water highly contaminated with inorganic arsenic. The earliest and most prominent changes are in the skin, e.g., hyperpigmentation and keratoses (calus-like growths). Other effects that have been reported include alterations in gastrointestinal, cardiovascular, hematological (e.g., anemia), pulmonary, neurological, immunological and reproductive/developmental function (ATSDR, 1998). The most common symptoms of inorganic arsenic exposure appear on the skin and occurr after 5-15 years of exposure equivalent to 700 µg/day for a 70 kg adult, or within 6 months to 3 years at exposures equivalent to 2,800 µg/day for a 70 kg adult (pg. 131 NRC, 1999). They include alterations in pigmentation and the development of keratoses which are localized primarily on the palms of the hands, the soles of the feet and the torso. The presence of hyperpigmentation and keratoses on parts of the body not exposed to the sun is characteristic of arsenic exposure (Yeh, 1973, Tseng, 1977). The same alterations have been reported in patients treated with Fowler's solution (1% potassium arsenite; Cuzick et al., 1982), used for asthma, psoriasis, rheumatic fever, leukemia, fever, pain, and as a tonic (WHO 1981 and NRC 1999). Chronic exposure to inorganic arsenic is often associated with alterations in gastroenterological (GI) function. For example, noncirrhotic hypertension is a relatively specific, but not commonly found manifestation in inorganic arsenic-exposed individuals and may not become a clinical observation until the patient demonstrates GI bleeding (Morris et al., 1974; Nevens et al., 1990). Physical examination may reveal spleen and liver enlargement, and histopathological examination of tissue specimens may demonstrate periportal fibrosis (Morris et al., 1974; Nevens et al., 1990; Guha Mazumder et al., 1997). There have been a few reports of cirrhosis after inorganic arsenic exposure, but the authors of these studies did not determine the subjects' alcohol consumption (NRC 1999). Development of peripheral vascular disease (hardening of the arteries to the arms and legs, that can cause pain, numbness, tingling, infection, gangrene, and clots) after inorganic arsenic exposure has also been reported. In Taiwan, blackfoot disease (BFD, a severe peripheral vascular insufficiency which may result in gangrene of the feet and other extremities) has been the most severe manifestation of this effect. Tseng (1977) reported over 1,000 cases of BFD in the arsenic study areas of Taiwan. Less severe cases of peripheral vascular disease have been described in Chile (Zaldivar et al., 1974) and Mexico (Cebrian, 1987). In a Utah study, increased SMRs for hypertensive heart disease were noted in both males and females after exposure to inorganic arsenic-contaminated drinking water (Lewis et al., 1999). These reports link exposure to inorganic arsenic effects on the cardiovascular system. Studies in Taiwan (Lai et al., 1994) and Bangladesh (Rahman et al., 1998) found an increased risk of diabetes among people consuming arsenic-contaminated water. Two Swedish studies found an increased risk of mortality from diabetes among those occupationally exposed to arsenic (Rahman and Axelson, 1995; Rahman et al., 1998). Although peripheral neuropathy (numbness, muscle weakness, tremors, ATSDR 1998) may be present after exposure to short-term, high doses of inorganic arsenic (Buchanan, 1962; Tay and Seah, 1975), there are no studies that definitely document this effect after exposure to levels of less than levels (50 µg/L) of inorganic arsenic in drinking water. Hindmarsh et al. (1977) and Southwick et al. (1983) have reported limited evidence of peripheral neuropathy in Canada and the U.S., respectively, but it was not reported in studies from Taiwan, Argentina or Chile (Hotta, 1989, as cited by NRC 1999). There have been a few, scattered reports in the literature that inorganic arsenic can affect reproduction and development in humans (Borzysonyi et al., 1992; Desi et al., 1992; Tabacova et al., 1994). After reviewing the available literature on arsenic and reproductive effects, the National Research Council panel (NRC 1999) wrote that ``nothing conclusive can be stated from these studies.'' Based on the studies mentioned in this section, it is evident that inorganic arsenic contamination of drinking water can cause dermal and internal cancers, affect the GI system, alter cardiovascular function, and increase risk of diabetes, based on studies of people exposed to drinking water well above the current arsenic MCL. EPA's MCL is chosen to be protective of the general population within an acceptable risk range, not at levels at which adverse health effects are routinely seen (see section III.F.7. on risk considerations). E. What Are the Recent Developments in Health Effects Research? 1. Funding of Health Effects Research As mentioned earlier in section II.A., Congress recognized that we needed more research to determine the health effects at low levels of arsenic (below the observed health effects and below 50 µg/L). On December 6, 1996, EPA issued a Federal Register notice (61 FR 64739; US EPA, 1996e) asking for public comment on four arsenic health research topics to fund research projects with $2 million from EPA appropriations and $1 million in funds raised by water industry groups (US EPA, 1996d). In addition, the Office of Research and Development's (ORD's) Board of Scientific Counselors (BOSC) peer reviewed the draft research topics and the arsenic research plan. In the fall of 1997, EPA and the industry partners funded their respective choices for arsenic research, after having the applications peer reviewed. EPA issued three grants for the following research: Dose Response of Skin Keratoses and Hyper-Pigmentation, Arsenic Glutathione Interactions and Skin Cancer, and Cellular Redox Status. The water industry groups awarded two contracts, studying Contribution of Arsenic From Dietary Sources and Tumor Studies in Mice. 2. Expert Panel on Arsenic Carcinogenicity As part of the Integrated Risk Information System (IRIS) update effort, EPA sponsored an ``Expert Panel on Arsenic Carcinogenicity: Review and Workshop'' in May 1997 (US EPA, 1997d). The panel evaluated existing data to comment on arsenic's carcinogenic mode of action and the effect on dose-response extrapolations. The panel noted that arsenic compounds have not formed DNA adducts (i.e., bound to DNA) nor caused point mutations. Trivalent inorganic forms inhibit enzymes, but arsenite and arsenate do not affect DNA replication. The panel discussed several modes of action, concluding that arsenic indirectly affects DNA, inducing chromosomal changes. The panel thought that arsenic-induced [[Page 38899]] chromosomal abnormalities could possibly come from errors in DNA repair and replication that affect gene expression; that arsenic may increase DNA hypermethylation and oxidative stress; that arsenic may affect cell proliferation (cell death appears to be nonlinear); and that arsenic may act as a co-carcinogen. Arsenite causes cell transformation but not mutation of cells in culture. It also induces gene amplification (multiple copies of DNA sequences) in a way which suggests interference with DNA repair or cell control instead of direct DNA damage. The panel noted that all identified modes of action support a nonlinear dose- response curve, that few data supports any one mode as most important, and that more than one mode of action may be operating. At low doses the slope of the dose response would decrease, and at very low doses ``might effectively be linear but with a very shallow slope, probably indistinguishable from a threshold.'' In terms of implications for the risk assessment, the panel noted that risk per unit dose estimates from human studies can be biased either way. For the Taiwanese study, the ``* * * biases associated with the use of average doses and with the attribution of all increased risk to arsenic would both lead to an overestimation of risk (US EPA, 1997d, page 31).'' While health effects are most likely observed in people getting high doses, the effects are assigned to the average dose of the exposure group. Thus, risk per unit dose estimated from the average doses would lead to an overestimation of risk (US EPA, 1997d, page 31). 3. NAS Review of EPA's Risk Assessment In 1997, at EPA's request, the National Academy of Sciences' (NAS) Subcommittee on Arsenic of the Committee on Toxicology of the National Research Council (NRC) met. Their charge was to review EPA's assessments of arsenic. The NAS/NRC Subcommittee finished their work in March 1999 (The report can be viewed from the National Academy Press website: www.nap.edu/books/0309063337/html/index.html). The detailed discussion of their work is in section III.F. In general, the NRC report confirms and extends concerns about human carcinogenicity of drinking water containing arsenic and offers perspective on dose- response issues and needed research. For the decisions in this regulation, the EPA has relied upon the NRC report as presenting the best available, peer reviewed science as of its completion and has augmented it with more recently published, peer reviewed information. Further work on the risk assessment will also be done before the final rule is issued to analyze the risks of internal cancers. The NRC provided risk numbers for bladder cancer using the Agency's approach. The NRC report noted that ``some studies have shown that excess lung cancer deaths attributed to arsenic are 2-5 fold greater than the excess bladder cancer deaths. * * * (NRC, 1999, pg. 8).'' The NRC recommended that EPA analyze risks of internal cancers both separately and combined. Peer-reviewed quantitative analysis of lung tumor risk is expected to be available for consideration in the final rulemaking. Meanwhile, this proposal, in a ``what if'' analysis (discussed in section X.B), estimates the potential monetary benefits that would result if the lung cancer and bladder cancer risks were the same, which would be the case if the excess lung cancer deaths actually were 2- to 5-fold greater than the excess bladder cancer deaths. 4. May 1999 Utah Mortality Study EPA scientists conducted an epidemiological study of 4,058 Mormons exposed to arsenic in drinking water in seven communities in Millard County, Utah (Lewis et al., 1999). The 151 samples from their public and private drinking water sources had arsenic concentrations ranging from 4 to 620 µg/L with seven mean (arithmetic average) community exposure concentrations of 18 to 191 µg/L and all seven community exposure medians (mid-point of arsenic values) 200 µg/L. Observed causes of death in the study group (numbering 2,203) were compared to those expected from the same causes based upon death rates for the general white male and female population of Utah. Several factors suggest that the study population may not be representative of the rest of the United States. The Mormon church, the predominant religion in Utah, prohibits smoking and consumption of alcohol and caffeine. Utah had the lowest statewide smoking rates in the U.S. from 1984 to 1996, ranging from 13 to 17%. Mormon men had about half the cancers related to smoking (mouth, larynx, lung, esophagus, and bladder cancers) as the U.S. male population from 1971 to 1985 (Lyon et al., 1994). The Utah study population was relatively small (4,000 persons) and primarily English, Scottish, and Scandinavian in ethnic background. While the study population males had a significantly higher risk of prostate cancer mortality, females had no significantexcess risk of cancer mortality at any site. Millard County subjects had higher mortality from kidney cancer, but this was not statistically significant. Both males and females in the study group had less risk of bladder, digestive system and lung cancer mortality than the general Utah population. The Mormon females had lower death rates from breast and female genital cancers than the State rate. These decreased death rates were not statistically significant. Although deaths due to hypertensive heart disease were roughly twice as high as expected in both sexes, increases in death did not relate to increases in dose, calculated as the years of exposure times the median arsenic concentration. The Utah data indicate that heart disease should be considered in the evaluation of potential benefits of U.S. regulation. Vascular effects have also been reported as an effect of arsenic exposure in studies in the U.S. (Engel et al. 1994), Taiwan (Wu et al., 1989) and Chile (Borgono et al., 1977). The overall evidence indicating an association of various vascular diseases with arsenic exposure supports consideration of this endpoint in evaluation of potential noncancer health benefits of arsenic exposure reduction. 5. 1999 Review of Health Effects Tsai et al. (1999) estimated standardized mortality ratios (SMR's) for 23 cancer and non-cancer causes of death in women and 27 causes of death in men in an area of Taiwan with elevated arsenic exposures (Tsai, et al., 1999). The SMRs in this study are an expression of the ratio between deaths that were observed in an area with elevated arsenic levels and those that were expected to occur, compared to both the mortality of populations in nearby areas without elevated arsenic levels and to the national population. Drinking water (250-1,140 µg/L) and soil (5.3-11.2 mg/kg) in the Tsai (1999) population study had high arsenic content. There are, of course, possible differences between the population and health care in Taiwan and the United States; and arsenic levels in the U.S. are not generally as high as they were in the study area of Taiwan. However, the study gives an indication of the types of health effects that may be associated with arsenic exposure via drinking water. The study reports a high mortality rate (SMR > 3) for both sexes from bladder, kidney, skin, lung, and nasal cavity cancers and for vascular disease. Females also had high mortalities for laryngeal cancer. The SMRs calculated by Tsai (1999) used the single cause of death noted on the death certificates. Many chronic [[Page 38900]] diseases, including some cancers, are not generally fatal. Consequently, the impact indicated by the SMR in this study may underestimate the total impact of these diseases. The causes of death reported in this study are consistent with what is known about the adverse effects of arsenic. Tsai et al. (1999) identified ``bronchitis, liver cirrhosis, nephropathy, intestinal cancer, rectal cancer, laryngeal cancer, and cerebrovascular disease'' as possibly ``related to chronic arsenic exposure via drinking water,'' which had not been reported before. In addition, people in the study area were observed to have nasal cavity and larynx cancers not caused by occupational exposure to inhaled arsenic. 6. Study of Bladder and Kidney Cancer in Finland Kurttio et al. (1999) conducted a case-cohort design study of 61 bladder and 49 kidney cancer cases and 275 controls to evaluate the risk of these diseases with respect to arsenic drinking water concentrations. In this study the median exposure was 0.1 µg/L, the maximum reported was 64 µg/L, and 1% of the exposure was greater than 10 µg/L. The authors reported that very low concentrations of arsenic in drinking water were significantly associated with being a case of bladder cancer when exposure occurred 2-9 years prior to diagnosis. Arsenic exposure occurring greater than 10 years prior to diagnosis was not associated with bladder cancer risk. Arsenic was not associated with kidney cancer risk even after consideration of a latency period. F. What Did the National Academy of Sciences/National Research Council Report? 1. The National Research Council and Its Charge Due to controversy surrounding the risk assessment of inorganic arsenic, EPA asked the National Research Council (NRC) to do the following: (1) Review EPA's characterization of potential human health risks from ingestion of inorganic arsenic in drinking water; (2) review the available data on the carcinogenic and noncarcinogenic effects of inorganic arsenic; (3) review the data on the metabolism, kinetics and mechanism(s)/mode(s) of action of inorganic arsenic; and (4) identify research needs to fill data gaps. To accomplish this task, NRC convened a panel of scientific experts with backgrounds in chemistry, toxicology, genetics, epidemiology, nutrition, medicine, statistics and risk assessment. In addition to the general expertise of the panel members, many had conducted research on inorganic arsenic. NRC identified the thirteen scientists with ``diverse perspectives and technical expertise'' that peer reviewed the draft report. The report noted that ``EPA did not request, nor did the subcommittee endeavor to provide, a formal risk assessment for arsenic in drinking water (NRC, 1999).'' 2. Exposure Arsenic is naturally occurring and ubiquitously distributed in the earth's surface. Because of this, the general population is exposed to low levels of arsenic through the food supply. The NRC report provides FDA market basket data for inorganic arsenic intake by age group which, along with similar data for water intake, will permit communication of total exposure estimates of the general population by age group. The assumption is made in the FDA data that, for fish and seafood, inorganic arsenic is 10% of total arsenic. This 10% assumption is acknowledged to be conservative and has been adopted for public health protection so as not to underestimate the contribution from fish and seafood. Likewise, the 2 L/day assumption of adult drinking water intake does not represent intake by the average person; rather it represents intake of a person in the 90th percentile. 3. Essentiality The NRC report examined the question of essentiality of arsenic in the human diet. It found no information on essentiality in humans and only data in experimental animals suggesting growth promotion (arsenicals are fed to livestock for this reason). Inorganic arsenic has not been found to be essential for human well-being or involved in any required biochemical pathway. Given this and the fact that arsenic occurs naturally in food, consideration of essentiality is not necessary for public health decisions about water. 4. Metabolism and Disposition Data from humans show that inorganic arsenic is readily absorbed and transported through the body. It has a half-life in the body of approximately four days and is primarily excreted in the urine. If a human is exposed to the inorganic arsenate form (+5 valence), the arsenite will be reduced to arsenite (+3). Some of the arsenite will be sequentially methylated to form monomethylarsonic acid (MMA) and dimethylarsinic acid (DMA). This methylation process decreases acute toxicity and facilitates excretion from the body. Individuals and populations vary in their metabolism of arsenic. Such variations may be due to genetic differences, species and dose of inorganic arsenic ingested, nutrition, disease and possibly other factors. Whether these methylated products (MMA and DMA) play a role in the development of cancer and noncancer endpoints is unknown at the present time (NRC, 1999). The NRC report recommended that experiments be conducted on the factors affecting interspecies differences in inorganic arsenic toxicity including use of human tissue when available. 5. Human Health Effects and Variations in Sensitivity The NRC panel concluded that there is sufficient evidence that chronic ingestion of inorganic arsenic causes bladder, lung and skin cancers and adverse noncancer effects on the cardiovascular systems, mainly from studies exposed to ``several hundred micrograms per liter. Few data address the degree of cancer risk at lower concentrations of ingested arsenic (NRC, 1999, pg. 130).'' The Utah study (Lewis et al., 1999), published after the NRC report, indicates that cardiovascular effects can occur at lower exposures than those seen in the studies available for the NRC report. At the present time, the NRC report indicates that there is insufficient evidence to judge whether inorganic arsenic can affect reproduction or development in humans. However, inorganic arsenic can pass through the placenta (Concha et al., 1998), and developmental toxicity needs investigation. In animal studies, intraperitoneal (injection into the abdominal cavity) administration of inorganic arsenic can cause malformations, and oral dosing has been reported to alter fetal growth and viability. The NRC report recommended additional studies to characterize the dose-response curve for inorganic arsenic-induced cancer and noncancer health endpoints. They also stated that the reported beneficial effects of inorganic arsenic in animals should be carefully monitored. In addition, the potential effects of inorganic arsenic on human reproduction should be investigated. There are many factors (genetics, diet, metabolism, health and sex) that may affect a human's response to inorganic arsenic exposure. For example, reduction in methylation of inorganic arsenic methylation can cause humans to retain more arsenic in their tissues. The retention of a greater arsenic load could place a person at a greater risk. The NRC report (1999) recommended that various factors that have the ability [[Page 38901]] to alter a human's response to inorganic arsenic exposure be carefully examined. Specifically, these studies should focus on the extent of human variability with respect to metabolism, tissue deposition and excretion under different environmental conditions. Humans are variable in their metabolic processing of inorganic arsenic, and internal dose will vary from person to person because of this as well as because of diet, nutritional status, lifestyle, and health status. Human variability also exists in response characteristics (susceptibility). The full quantitative extent of this variability is not known. For instance, men are more susceptible than women to bladder cancer throughout the world even though bladder cancer rates vary from region to region. We do not know whether arsenic may have a greater effect at different ages (e.g., infants v.s. adults). 6. Modes of Action Knowledge of a ``mode of action'' means that data are available to describe the key events at the cellular and/or subcellular level that lead to the development of the cancer or noncancer endpoint. A number of potential modes of carcinogenic action have been proposed for arsenic, with varying degrees of supporting data. The key events in the cancer process caused by arsenic exposure are not known. Nevertheless, the data are sufficient to support the conclusion of the NRC report and the EPA 1997 expert panel workshop report that: ``Arsenic exposure induces chromosomal abnormalities without direct reaction with DNA (US EPA, 1997d).'' There is strong evidence against a mode of action for inorganic arsenic involving direct reaction with DNA. One of the hallmarks of direct DNA reactivity is multi-species carcinogenic activity. For arsenic, long-term bioassays for carcinogenic activity in rats, mice, dogs, and monkeys have been uniformly negative (Furst, 1983). The kinds of genetic alterations seen in both in vivo and in vitro studies of arsenic effects are at the level of loss and rearrangement of chromosomes; these are results of errors of ``cellular housekeeping'' either in DNA repair or in chromosome replication. The NRC and EPA expert panel (US EPA, 1997d) reports examined several lines of evidence for various modes of action that might be operative. These included changes in DNA methylation patterns that could change gene expression and repair, oxidative stress, potentiation of effects of mutations caused by other agents, cell proliferative effects, and interference with normal DNA repair processes. Further examination in both of these reports of dose-response shapes associated with these effects led to the conclusion that they involve processes that have either thresholds of dose at which there would be no response or sublinearity of the dose response relationship (response decreasing disproportionately as dose decreases). The NRC report concluded: ``For arsenic carcinogenicity, the mode of action has not been established, but the several modes of action that are considered plausible (namely, indirect mechanisms of mutagenicity) would lead to a sublinear dose-response curve at some point below the point at which a significant increase in tumors is observed. * * * However, because a specific mode (or modes) of action has not yet been identified, it is prudent not to rule out the possibility of a linear response.'' The NRC report noted that in certain in vitro studies of human and animal cells, genotoxic effects have been shown to occur at submicromolar concentrations of arsenite that are similar to concentrations found in urine of humans ingesting water at the current MCL. This emphasizes the potentially low margin of exposure (health effects observed at concentrations eight times above the MCL) for arsenic in water at the current MCL. For noncancer effects, inhibition of cellular respiration in mitochondria by arsenic may be the focal point of its toxicity. In addition, inorganic arsenic causes oxidative stress that could play a role in the development of adverse health effects. The NRC report (1999) recommended that biomarkers of inorganic arsenic exposure and cancer appearance be thoroughly studied. Such data might better characterize the dose-response effects of inorganic arsenic at lower exposure levels. For noncancer effects, a greater understanding of arsenic's effects on cellular respiration and subsequent effects of methylation and oxidative stress are needed (NRC, 1999). NRC recommended several mode of action studies, using biomarkers, to help predict the shape of the dose-response curve for cancer and non-cancer endpoints. NRC concluded that `` * * *Additional epidemiological evaluations are needed to characterize the dose- response relationship for arsenic-associated cancer and non-cancer endpoints, especially at low doses.'' 7. Risk Considerations The NRC study used the results of epidemiological, (i.e., human) studies; research on the mode of action, and information about factors affecting sensitivity to arsenic to project to risks to the U.S. population. The numerical estimation of risk in the NRC report has several features to consider. The range of drinking water levels associated with health endpoints in the available studies is generally hundreds of ppb which is, however, within a factor of 10 of the existing standard of 50 ppb. Because of uncertainty about the shape of the dose-response relationship below this range of observed responses, the NRC report used the approach of the 1996 EPA proposed carcinogen risk assessment guidelines (US EPA, 1996b). For the male bladder cancer deaths which were emphasized in the report, NRC used a lower limit on the dose associated with a 1% (1 in 100) cancer response, and the LED01 is estimated to be 400 ppb. This is a point of departure for extrapolating to exposure levels outside the range of observed data based on inference. Consistent with the proposed revisions to the Guidelines for Cancer Risk Assessment, the report shows both a linear extrapolation and a margin of exposure extrapolation (difference between the point of departure and selected exposure). Because current data on potential modes of action are supportive of sub-linear extrapolations, the linear approach could overestimate risk at low doses. However, EPA believes that within the several-fold range (10x) just below the point of departure, this should make little difference. EPA's scientists note that it makes an increasing difference as dose decreases, and the difference results in an overestimate of risk at lower exposures. With a straight-line extrapolation from the point of departure, the report estimated risk to be 1.0 to 1.5 x 10-\3\ at the current MCL of 50 ppb and the margin of exposure to be less than 8. As described further in section X.A., EPA used parts of NRC's risk analysis and applied U.S. water consumption, weights, and estimate of population exposed to arsenic to model the U.S. population risk. In selecting the proposed MCL, EPA considered the uncertainties of the quantitative dose-response assessment for inorganic arsenic's health effects, particularly the possible nonlinearity of the dose-response and multiple cancer risks. Given the current outstanding questions about human risk at low levels of exposure, decisions about safe levels are public health policy judgments. 8. Risk Characterization In 1983 the National Academy of Sciences (NAS, 1983) defined risk [[Page 38902]] assessment as containing four steps: hazard identification, dose- response assessment, exposure assessment, and risk characterization. Risk characterization is the process of estimating the health effects based on evaluating the available research, extrapolating to estimate health effects at exposure levels, and characterizing uncertainties. In risk management, regulatory agencies such as EPA evaluate alternatives and select the regulatory action. Risk management considers ``political, social, economic, and engineering information'' using value judgments to consider ``the acceptability of risk and the reasonableness of the costs of control (NAS, 1983).'' Unlike most chemicals, there is a large data base on the effects of arsenic on humans. Inorganic arsenic is a human poison, and oral or inhalation exposure to the chemical can induce many adverse health conditions in humans. Specifically oral exposure to inorganic arsenic in drinking water has been reported to cause many different human illnesses, including cancer and noncancer effects, as described in Section III. The NRC panel (1999) reviewed the inorganic arsenic health effects data base. The panel members concluded that the studies from Taiwan provided the current best available data for the risk assessment of inorganic arsenic-induced cancer. (There are corroborating studies from Argentina and Chile.) They obtained more detailed Taiwanese internal cancer data and modeled the data using the multistage Weibull model and a Poisson regression model. Three Poisson data analyses showed a 1% response level of male bladder cancer at approximately 400 µg of inorganic arsenic/L. The 1% level was used as a Point of Departure (POD) for extrapolating to exposure levels outside the range of observed data. For an agent that is either acting by reacting directly with DNA or whose mode of action has not been sufficiently characterized, EPA's public health policy is to assume that dose and response will be proportionate as dose decreases (linearity of the extrapolated dose- response curve). This is a science policy approach to provide a public health conservative assessment of risk. The dose-response relationship is extrapolated by taking a straight line from the POD rather than by attempting to extend the model used for the observed range. This approach was adopted by the NRC report which additionally noted that using this approach for arsenic data provides results with alternative models that are consistent at doses below the observed range whereas extending the alternative models below the observed range gives inconsistent results. Drawing a straight line from the POD to zero gives a risk of 1 to 1.5 per 1,000 at the current MCL of 50 µg/ L. Since some studies show that lung cancer deaths may be 2- to 5-fold higher than bladder cancer deaths, the combined cancer risk could be even greater. The NRC panel also noted that the MCL of 50 µg/L is less than 10-fold lower than the 1% response level for male bladder cancer. Based on its review, the consensus opinion of the NRC panel was that the current MCL of 50 µg/L does not meet the EPA's goal of public-health protection. Their report recommended that EPA lower the MCL as soon as possible. IV. Setting the MCLG A. How Did EPA Approach It? For the decisions in this regulation, the EPA has relied upon the NRC report as presenting the best available, peer reviewed science as of its completion and has augmented it with more recently published, peer reviewed information. EPA used the 1999 NRC report and other published scientific papers to characterize the potential health hazards of ingested inorganic arsenic. As NRC (1999) noted, DMA may enhance the carcinogenicity of other chemicals, but more data are needed. Based on current knowledge, the organic forms of arsenic in fish and shellfish do not appear to present a significant risk to humans. The overall weight of evidence indicates that the inorganic arsenate and arsenite forms found in drinking water are responsible for the adverse health effects of ingested arsenic. EPA focused its risk assessment on the carcinogenic effects of inorganic arsenic (the forms found in drinking water sources). A factor that could modify the degree of individual response to inorganic arsenic is its metabolism. There is ample evidence (NRC, 1999) that the quantitative patterns of inorganic arsenic methylation vary considerably and that the extent of this variation is unknown. It is certainly possible that the metabolic patterns of people affect their response to inorganic arsenic. There are studies underway in humans and experimental animals under the EPA research plan and other sponsorships. Over the next several years these will provide better understanding of the mode(s) of carcinogenic action of arsenic, metabolic processes that are important to its toxicity, human variability in metabolic processes, and the specific contributions of various food and other sources to arsenic exposure in the U.S. These are important issues in projecting risk from the observed data range in the epidemiologic studies to lower environmental exposures experienced from U.S. drinking water. Until further research is completed, questions will remain regarding the dose-response relationship at low environmental levels. The several Taiwan studies have strengths in their long-term observation of exposed persons and coverage of very large populations (>40,000 persons). Additionally, the collection of pathology data was unusually thorough. Moreover, the populations were quite homogeneous in terms of lifestyle. Limitations in exposure information exist that are not unusual in such studies. In ecological epidemiology studies of this kind, the exposure of individuals is difficult to measure because their exposure from water and food is not known. This results in uncertainties in defining a dose-response relationship. The studies in Chile and Argentina are more limited in extent, (e.g., years of coverage, number of persons, or number of arsenic exposure categories analyzed), but provide important findings which corroborate one another and those of the Taiwan studies. These epidemiological studies provide the basis for assessing potential risk from lower concentrations of inorganic arsenic in drinking water, without having to adjust for cross-species toxicity interpretation. Ordinarily, the characteristics of human carcinogens can be explored and experimentally defined in test animals. Dose- response can be measured, and animal studies may identify internal transport, metabolism, elimination, and subcellular events that explain the carcinogenic process. Arsenic presents unique problems for quantitative risk assessment because there is no test animal species in which to study its carcinogenicity. While such studies have been undertaken, it appears that test animals, unlike humans, do not respond to inorganic arsenic exposure by developing cancer. Their metabolism of inorganic arsenic is also quantitatively different than humans. Inorganic arsenic does not react directly with DNA. If it did, it would be expected to cause similar effects across species and to cause response in a proportionate relationship to dose. Moreover, its metabolism, internal disposition, and excretion are different and vary across animal and plant species and humans--in test studies and in nature. [[Page 38903]] Until more is known, EPA will take a traditional, public health conservative approach to considering the potential risks of drinking water containing inorganic arsenic. EPA recognizes that the traditional approach may overestimate risk, as explained in the next section. B. What Is the MCLG? EPA concludes that exposure to inorganic arsenic induces cancer in humans. It also is associated with adverse noncancer effects such as hypertension (NRC, 1999). The NRC report stated that ``Data on the modes of action for carcinogenicity can help to predict the shape of cancer dose-response curves below the level of direct observation of tumors. * * * For arsenic carcinogenicity * * * modes of action that are considered most plausible (namely, indirect mechanisms of mutagenicity) lead to a sublinear dose-response at some point below the level at which a significant increase in tumors is observed. However, because a specific mode (or modes) of action has not been identified at this time, it is prudent not to rule out the possibility of a linear response (NRC 1999, pgs. 213-214).'' The expert panel report (US EPA, 1997d, pg. 31) stated: ``* * * for each of the modes of action regarded as plausible, the dose-response would either show a threshold or would be nonlinear. * * * [H]owever, ``the dose response for arsenic at low doses would likely be truly nonlinear--i.e., with a decreasing slope as the dose decreased. However, at very low doses such a curve might effectively be linear but with a very shallow slope, probably indistinguishable from a threshold.'' In the absence of a known mode of action(s), EPA has no basis for determining the shape of a sublinear dose-response curve for inorganic arsenic. As a result, consistent with EPA public health policy, EPA will continue to use a linear dose- response curve for inorganic arsenic effects. Using a linear type of a dose-response curve, EPA is proposing an MCLG of zero. The Agency welcomes comments on setting a nonzero MCLG and submission of data supporting a nonzero MCLG. C. How Will a Health Advisory Protect Potentially Sensitive Subpopulations? The NRC report was inconclusive about the health risks to pregnant woman, developing fetus, infants, lactating women, and children. When the Agency completes this rulemaking, it intends to issue a health advisory on arsenic in drinking water, in order to decrease risk to sensitive subpopulations prior to the implementation of the new MCL. The effective date of a revised MCL will be three to five years after the final rule is issued (2004-2006). A health advisory is a non-regulatory document that supports water providers in their independent decisions on actions to take regarding water contaminants and their communication with the general public. In the health advisory on arsenic the Agency intends to address a precautionary step to protect infants. This step would be to avoid using water containing high levels of arsenic to make up infant formula. The reason for this precaution is that epidemiologic studies indicate that arsenic in drinking water (Lewis et al., 1999) affects the cardiovascular system. While there are no studies of effects of arsenic on human infants, both the cardiovascular system and brain (and its vascular system) continue to develop after birth (Thompson, P.M et al. 2000); thus, the effects discussed in this notice on the cardiovascular system raise a concern about potential effects of arsenic on infant development. In large part, causes of cerebrovascular incidents (stroke) in children are not understood except for certain, known associations with organic diseases and genetic diseases. Congenital and acquired heart disease are the most common cause of stroke in children. The current toxicity data on arsenic do not contradict this precautionary view. D. How Will the Clean Water Act Criterion Be Affected by This Regulation? EPA is also working to harmonize the human health arsenic criteria for the Clean Water Act (CWA) and the SDWA. The major reason for the present difference (discussed in section II.D.) between the MCL and the Ambient Water Quality Criterion (AWQC) was the result of using separate bases for determining the two standards. The AWQC for arsenic was derived from the risk assessment for arsenic-induced skin cancer, while the current SDWA MCL, adopted in 1975 as a National Interim Primary Drinking Water Regulation, evolved from the U.S. Public Health Service standard dating back to the 1940s. The Agency will use the conclusions of the NRC (1999) report to form the human health basis for both the AWQC and the MCL. However, the CWA and SDWA statutes require that the Agency consider different factors during the derivation of a standard. For example, SDWA requires that the Agency consider: (1) Cost/benefit analyses, including sizes of the public water systems, (2) the level of arsenic that can be analyzed by laboratories on a routine basis, [i.e., the practical quantitation limit (PQL)] and (3) treatment techniques for removing the chemical from the water. On the other hand, the CWA requires the EPA to consider water and fish consumption (including amount of fish eaten, percent lipid in the fish and the bioaccumulation factor for the chemical in the fish), but not cost/benefits, analytical or treatment techniques. Accordingly, developing a AWQC under the CWA may produce a standard that differs from the MCL derived under the SDWA even though both standards are based on the same health endpoint. The Agency will begin work on a new AWQC for arsenic after promulgating the MCL for arsenic. V. EPA's Estimates of Arsenic Occurrence One of the key components in the development of the proposed arsenic rule is the analysis of arsenic occurrence in public water supplies, both community water systems (CWS) and non-transient, non- community, water systems (NTNCWS). EPA's national occurrence assessment of arsenic provides a basis for estimating: (1) The number of systems expected to exceed various arsenic levels; (2) the number of people exposed to the different levels of arsenic; and (3) the variability in arsenic levels in water systems among the wells and/or entry points to the distribution system. EPA uses the estimate of the total number of systems and populations affected in the United States in its cost-benefit analysis. EPA is seeking comment on its analysis of arsenic occurrence in the U.S., as well as requesting additional data. A. What Data Did EPA Evaluate? For previous occurrence analyses EPA used four older national arsenic databases: (1) The National Inorganic and Radionuclide Survey (NIRS), conducted from 1984 to 1986, for ground water CWSs; (2) a 1976- 1977 National Organic Monitoring Survey (NOMS); (3) a 1978-1980 Rural Water Survey (RWS); and (4) the 1978 Community Water System Survey (CWSS) for surface water CWSs. However, these older databases have several limitations. First, the surveys of surface water systems will not reflect changes in raw water sources which occurred in the last twenty years. Second, filtration treatment added to comply with the Surface Water Treatment Rule (110 54 FR 27486, June 29, 1989) would tend to decrease arsenic exposure, through incidental arsenic removal. Finally, most of the [[Page 38904]] data were censored (reported as less than the analytical test method detection level or reporting limit, e.g., ``not detected'' or ``5 µg/L''). NIRS, CWSS, and RWS, respectively, had 93%, 97%, and 90% censored data. This limits the estimation of low level occurrence of arsenic and makes it statistically difficult to extrapolate occurrence with the limited amount of non-censored data. The EPA Science Advisory Board recommended that EPA abandon the older data when sufficient new data become available because of the high percentage of censored data in the older surveys and the difficulty of using highly censored data sets to estimate occurrence (US EPA, 1995). Therefore, with improved analytical techniques for detecting arsenic at lower levels, as low as 0.5 µg/L, and the lower reporting limits in the new data received by EPA, the Agency focused the data evaluation on post-1980 data sources for estimating national occurrence. Since 1992, EPA OGWDW has received arsenic databases from other EPA offices, States, public water utilities, and associations. EPA combined the compliance monitoring data obtained from States into the ``25 States'' database. The Agency evaluated the databases listed in Table V-1. (Note that EPA's database, the Safe Drinking Water Information System (SDWIS), only records violations of the current arsenic MCL, so it is censored at 50 µg/L.) A more detailed description of the databases and evaluations are presented in the EPA document titled ``Arsenic Occurrence in Public Drinking Water Supplies,'' (US EPA, 2000b). Table V-1.--Summary of Arsenic Data Sources ---------------------------------------------------------------------------------------------------------------- Reporting level Data source (µg/L) Number of CWSs Source water Water type ---------------------------------------------------------------------------------------------------------------- 25 States\1\.................. 1 to 10......... >19,000......... Surface, Ground...... finished. Metro \2\..................... 1............... 140............. Surface, Ground...... raw & finished. NAOS \3\...................... 0.5............. 517............. Surface, Ground...... raw & predicted finished. USGS \4\...................... 1............... not available Ground............... raw. (20,000 sites). ACWA \5\...................... 0.1 to 1........ 180 (1,500 Surface, Ground...... finished. samples). WESTCAS \6\................... not available... not available... Ground............... finished. ---------------------------------------------------------------------------------------------------------------- \1\ Arsenic compliance monitoring data from community water systems (CWSs) from Alabama, Alaska, Arizona, Arkansas, California, Illinois, Indiana, Kentucky, Kansas, Maine, Michigan, Minnesota, Missouri, Montana, Nevada, New Hampshire, New Jersey, New Mexico, North Carolina, North Dakota, Ohio, Oklahoma, Oregon, Texas, and Utah. \2\ Metropolitan Water District of Southern California (MWDSC, or Metro) 1992-1993 national survey of 140 CWSs serving more than 10,000 people. \3\ 1996 National Arsenic Occurrence Survey (NAOS) funded by the Water Industry Technical Action Fund (WITAF), which includes the following organizations: American Water Works Association, National Association of Water Companies, Association of Metropolitan Water Agencies, National Rural Water Association, and National Water Resources Association. \4\ U.S. Geological Survey (USGS) ambient (raw water) ground water from approximately 20,000 wells throughout the U.S. used for various purposes, including public supply, research, agriculture, industry and domestic supply. \5\ 1993 survey from 180 water agencies, utilities, and cities in southern California, conducted by the Association of California Water Agencies (ACWA). \6\ 1997 Western Coalition of Arid States (WESTCAS) Research Committee Arsenic Occurrence Study which aggregated arsenic data (e.g., median arsenic value for county, city, or provider) from Arizona, New Mexico, and Nevada. B. What Databases Did EPA Use? EPA evaluated the databases for representativeness, accuracy and coverage of community water systems in the U.S. EPA determined that the compliance monitoring data from the 25 States (``25-States database'') would establish the most accurate and scientifically defensible national occurrence and exposure distributions of arsenic in public ground water and surface water supplies. Figure V.1 shows the coverage of these States in the U.S. The 25-States database provides more finished water arsenic data, from over 19,000 ground and surface water CWSs, than the other national databases. EPA is interested in finished water data, rather than raw water data, because it indicates the current arsenic levels in water systems after treatment and reflects their customers' level of exposure to arsenic. The 25-States database provides system and individual arsenic data for a significant number of CWSs in each State. The arsenic data can be linked directly to specific water systems by their identification code to obtain additional information in SDWIS, such as population served, system type (e.g., CWS, NTNCWS), source type (e.g., ground water, surface water, purchased water, ground water under the influence), and location. For this reason, EPA chose to use the compliance monitoring data from the States of California, Nevada, New Mexico, and Arizona, rather than the data about these States from ACWA and WESTCAS. Most of the 25-States data had reporting limits of less than 2 µg/L. In addition, the database includes multiple samples from the water systems over time and from multiple sources within the systems. The multiple samples provide for a more accurate estimate of the arsenic levels in the systems, than a survey with one sample per system. The arsenic compliance monitoring data provides point-of-entry or well data within systems from eight States, which is used for intrasystem variability analysis (discussed in Section V.G). Intrasystem variability analysis provides an understanding of the variation of arsenic levels within a system, from well to well or entry point to entry point. EPA also received arsenic data from Florida, Idaho, Iowa, Louisiana, Pennsylvania, and South Dakota; however EPA did not include these States in the database. These States either provided data that (1) could not be linked to CWSs; (2) did not indicate if the results were censored or non-censored; (3) were all zero, without providing the analytical/reporting limit; or (4) rounded results to the nearest ten µg/L. EPA used the USGS and NAOS databases and their occurrence estimates for comparison purposes. In addition, EPA used the NAOS approach to partitioning of the U.S. for its analysis. We combined State data sets with different data naming conventions, and the database development and data [[Page 38905]] conditioning process is described in Appendix D-3 of the occurrence support document (US EPA, 2000b). Appendix D-1 identifies who provided the data and data provided for each State in the 25-State database. Appendix D-2 lists the data names we used to develop the national database. We assumed that the data represented compliance sampling, and some States have reportedly provided source water data and compliance data. If you are aware of errors in our data set, please let us know. Also, additional data would reduce the uncertainty of our national occurrence estimate. We encourage commenters to submit arsenic compliance monitoring data sets either from States not already in our data set, more recent data that were not included in the described data sets, or a more official version of compliance data. We will use this information to obtain a more representative national occurrence estimate for the final rule. BILLING CODE 6560-50-P [[Page 38906]] [GRAPHIC] [TIFF OMITTED] TP22JN00.000 BILLING CODE 6560-50-C [[Page 38907]] C. How Did EPA Estimate National Occurrence of Arsenic in Drinking Water? EPA derived the national estimates of arsenic occurrence in three steps: (1) Estimate system means; (2) estimate State distribution of system means; and (3) estimate national distributions of system means. As discussed in section V.B, EPA determined that the 25-States database would be used for estimating national occurrence. EPA calculated a system average for each water system in its database. When the database provided 5 or more detected (greater than the reporting limit) arsenic samples in a system, we used the method of ``regression on order statistics'' (Helsel and Cohn, 1988) to extrapolate values for the non-detected observations, then calculated the arithmetic mean. When there were 1 to 4 detected values, we substituted half the reporting limit for each non-detected value (less than the reporting limit) and calculated an arithmetic average. When there were no detected values (all samples had non-detected values), we set the arsenic system average as a non-detect at the mode (most frequently occurring) of the reporting limits. As a result, each system has a calculated system mean, either a non-detected or detected value. In order to estimate the distribution of systems means in a State, EPA aggregated the system means into a single distribution and derived separate estimates of percentage of systems with average arsenic values greater than 2, 3, 5, 10, 15, 20, 25, 30, 40, and 50 µg/L (referred to as exceedance estimates). We developed separate estimates for ground water and surface water systems. Within each State, EPA fit a lognormal distribution to the population of estimated system means, and used the fitted distribution to estimate exceedance probabilities. However, when fitting the lognormal distribution, EPA excluded system means which were estimated to be less than their reporting limit, since these require more extrapolation below the reporting limit and were judged to be less reliable. EPA also did not make exceedance estimates below the most frequently occurring reporting limit or censoring point in each of the States. To estimate the national distribution of system means, EPA grouped the States into the seven regions developed in the NAOS (Frey and Edwards, 1997). Frey and Edwards derived a natural occurrence factor by weighting detection, number of data points, and higher arsenic values from data in the USGS WATSTORE water quality database and the Metro survey. Then they grouped States into seven regions based on the calculated natural occurrence factors. Figure V.1 is a map of the U.S. with the NAOS regions. With this regional grouping of States, EPA developed separate regional estimates for surface water and ground water systems. In a separate analysis, EPA found the national result from using the NAOS regions to be similar to grouping States into different regions, based on a preliminary examination of generally related exceedance probabilities. EPA derived each regional estimate by using exceedance estimates from the States with compliance monitoring data in the region, weighted by the number of community water systems in those specific States. For example, we used the exceedance estimates from Montana and North Dakota, weighted by the number of community water systems in those States, to derive the North Central region estimate. Within each region, we estimated the percentages of systems with average arsenic values greater than 2, 3, 5, 10, 15, 20, 25, 30, 40, and 50 µg/ L. We then weighted the regional exceedance estimates, by the total number of community water systems in each region (including the number of community water systems in the States without compliance monitoring data) to obtain national estimates of percentages of systems with average arsenic values greater than 2, 3, 5, 10, 15, 20, 25, 30, 40, and 50 µg/L. EPA believes that separate estimates are not justified for different system sizes. A graphical analysis (``box and whisker'' plots) of the occurrence distributions suggests that in some regions, systems in different size categories do have different mean concentrations. However the differences in means are much smaller than the variability of the observed concentrations. Moreover, the differences do not vary with system size in a consistent way. For example, for ground water systems, arsenic concentrations in the New England Region (NAOS Region 1) decrease as system size increases, while in the Mid-Atlantic and South Central regions (NAOS Regions 2 and 5), arsenic concentrations increase as system size increases. In the four remaining regions, no systematic patterns are evident. For these reasons, and because additional stratification decreases the precision of the estimates, EPA has not developed separate estimates for different system sizes. The method of substitution that EPA used for non-detected concentrations (described above) is different from the method that water systems use for determining compliance with the MCL: We substituted positive values for non-detects, while for purposes of compliance, non-detected concentrations are treated as zero. Therefore, our estimates of occurrence will be higher on average than those found by water systems monitoring for compliance with the MCL. As a result we might overestimate both the costs and benefits of the proposed MCL. However we believe that our estimate of occurrence is justified, for two reasons. First, it is more accurate (less biased). Second, as the detection limits of analytical methods continue to improve (i.e., lower than 1 µg/L), the difference between the two substitution methods will be small and will occur in the range below the MCL. D. What Are the National Occurrence Estimates of Arsenic in Drinking Water for Community Water Systems? Arsenic is found in both ground water and surface water sources. Figure V.1 presents the regions of the United States referred to in this discussion. Table V-2 data indicate that higher levels of arsenic tend to be found in ground water sources (e.g., aquifers) than in surface water sources (e.g., lakes, rivers). The 25-States finished water data also indicate that the North Central, Midwest Central, and New England regions of the United States tend to have low to moderate (2-10 µg/L) ground water arsenic levels, while the Western region tends to have higher levels of ground water arsenic (>10 µg/L) than the other regions. Systems in the other regions of the U.S. may have high levels of arsenic (hot spots), while many systems and portions of the States in the listed regions may not have any detected arsenic in their drinking water. [[Page 38908]] Table V-2.--Regional Exceedance Probability Distribution Estimates -------------------------------------------------------------------------------------------------------------------------------------------------------- Percent of systems exceeding arsenic concentrations (µg/L) of: Region ------------------------------------------------------------------------------------- 2 3 5 10 15 20 25 30 40 50 -------------------------------------------------------------------------------------------------------------------------------------------------------- Ground Water Systems -------------------------------------------------------------------------------------------------------------------------------------------------------- New England....................................................... 29 21 21 7 4 3 2 2 1 0.7 Mid Atlantic...................................................... ....... ....... *0.3 *1 0.3 0.1 0.06 0.03 0.009 0.003 South East........................................................ 2 1 0.5 0.2 0.1 0.07 0.05 0.04 0.02 .01 Midwest........................................................... 28 21 14 6 4 2 2 1 .8 0.5 South Central..................................................... 27 19 10 4 2 1 0.8 0.5 0.3 0.2 North Central..................................................... 29 21 13 6 4 2 2 1 0.9 0.6 West.............................................................. 42 31 25 12 7 5 4 3 2 1 -------------------------------------------------------------------------------------------------------------------------------------------------------- Surface Water Systems -------------------------------------------------------------------------------------------------------------------------------------------------------- New England....................................................... 11 *8 *9 1.0 0.6 0.4 0.3 0.3 0.2 0.1 Mid Atlantic...................................................... ....... ....... *0.1 *0.1 0.01 0.001 0 0 0 0 South East........................................................ 0.8 0.2 0.03 0.001 0 0 0 0 0 0 Midwest........................................................... 4 3 1 0.4 0.2 0.1 0.1 0.07 0.05 0.03 South Central..................................................... 9 4 1 0.3 0.1 0.08 0.05 0.03 0.02 0.01 North Central..................................................... 20 10 4 0.8 0.2 0.1 0.05 0.02 0.008 0.003 West.............................................................. 19 13 7 3 2 1 0.8 0.6 0.4 0.3 -------------------------------------------------------------------------------------------------------------------------------------------------------- *Estimates at these regions and levels are inconsistent, in that the estimated % exceedances at lower values are smaller than the estimates at higher values. This inconsistency occurs because fewer States were used to estimate % exceedances at lower levels. EPA did not attempt to resolve the inconsistency, but combined the regional distribution into a national distribution which is consistent. The estimates of the number of CWSs expected to exceed different arsenic levels is based on the distribution of average arsenic concentrations in water systems. Using the data from the 25-States database, EPA estimates that 5.4% of ground water CWSs and 0.7% of surface CWSs have average arsenic levels above 10 µg/L. Similarly, 12.1% and 2.9% of ground water CWSs and surface water CWSs, respectively, have average arsenic levels above 5 µg/L. Tables V-3 and V-4 provide estimates by system size category. The percentage of systems that have average arsenic levels within a specific range does not vary across the system size categories. For example, 2.3% of ground water systems in each of the five system size categories have average arsenic levels in the range of >10 µg/L to 15 µg/L. Therefore, the arsenic exceedance estimates have the same distribution in any system size. These estimates of percent (or probability) of systems that have average arsenic levels within a specific range are multiplied by the number of systems in each size category to derive the number of systems in Table V-3 and V-4. Table V-3.--Statistical Estimates of Number of Ground Water CWSs With Average Arsenic Concentrations in Specified Ranges -------------------------------------------------------------------------------------------------------------------------------------------------------- Number of systems with average arsenic concentrations in specified ranges (µg/L; 43,749 systems total) System size (population served) -------------------------------------------------------------------------------------------------- >2.0 to >3.0 to >5.0 to >10.0 to >15.0 to >20.0 to >30.0 to 2.0 3.0 5.0 10.0 15.0 20.0 30.0 50.0 >50.0 -------------------------------------------------------------------------------------------------------------------------------------------------------- 25 to 500............................................ 21,325 2,158 2,268 1,960 674 314 287 188 129 501 to 3,300......................................... 7,616 771 810 700 241 112 103 67 46 3,301 to 10,000...................................... 1,811 183 193 167 57 27 24 16 11 10,001 to 50,000..................................... 933 94 99 86 29 14 13 8 6 >50,000.............................................. 154 16 16 14 5 2 2 1 1 Total............................................ 31,840 3,221 3,386 2,927 1,006 468 429 280 192 (% of systems)................................... (72.8%) (7.4%) (7.7%) (6.7%) (2.3%) (1.1%) (1.0%) (0.6%) (0.4%) -------------------------------------------------------------------------------------------------------------------------------------------------------- Note: Totals may not add up due to rounding of the number of systems to the nearest whole number. Systems serving fewer than 25 people are not included in this table. The estimates in this table do not take into account most treatment in place; in particular most of the systems in the ``>50.0'' column will have treated for arsenic in order to reduce their concentration below 50 µg/L. See text for more details. In Tables V-3 and V-4, the estimated numbers of systems with mean concentrations above 50 µg/L do not represent the number of systems which are believed to be out of compliance with the current MCL of 50 µg/L; nor do they represent actual systems at all. Rather, they are statistical extrapolations above 50 µg/L, based primarily on data below 50 µg/L. Since most data below 50 µg/L comes from systems which have not treated for arsenic, the ``>50.0'' columns in Tables V-3 and V-4 do not take into account most treatment currently in place. Therefore, the ``>50.0'' columns represent the estimated number of systems which would have mean arsenic concentrations above 50 µg/L if they had not treated for arsenic. By comparison with Tables V-3 and V-4, during the three-year period from September 1994 through August 1997, EPA recorded a total of 14 samples from 10 public water systems in which arsenic concentrations exceeded 50 µg/L. [[Page 38909]] Table V-4.--Statistical Estimates of Number of Surface Water CWSs With Average Arsenic Concentrations in Specified Ranges -------------------------------------------------------------------------------------------------------------------------------------------------------- Number of systems with average arsenic concentrations in specified ranges (µg/L; 10,683 systems total) System size (population served) -------------------------------------------------------------------------------------------------- >2.0 to >3.0 to >5.0 to >10.0 to >15.0 to >20.0 to >30.0 to 2.0 3.0 5.0 10.0 15.0 20.0 30.0 50.0 >50.0 -------------------------------------------------------------------------------------------------------------------------------------------------------- 25 to 500............................................ 2,794 122 94 69 11 4 4 2 2 501 to 3,300......................................... 3,308 144 111 82 13 5 4 3 2 3,301 to 10,000...................................... 1,656 72 56 41 6 3 2 1 1 10,001 to 50,000..................................... 1,384 60 47 34 5 2 2 1 1 > 50,000............................................. 477 21 16 12 2 1 1 0 0 Total............................................ 9,622 419 323 239 37 15 13 8 7 (% of systems)................................... (90.1%) (3.9%) (3.0%) (2.2%) (0.4%) (0.1%) (0.1%) (0.1%) (0.1%) -------------------------------------------------------------------------------------------------------------------------------------------------------- Note: Totals may not add up due to rounding of the number of systems to the nearest whole number. Systems serving fewer than 25 people are not included in this table. The estimates in this table do not take into account most treatment in place; in particular most of the systems in the ``>50.0'' column will have treated for arsenic in order to reduce their concentration below 50 µg/L. See text for more details. E. How Do EPA's Estimates Compare With Other Recent National Occurrence Estimates? In addition to EPA's national occurrence results presented in section V.D., two additional studies recently developed national occurrence estimates for arsenic in drinking water: the NAOS study (Frey and Edwards, 1997), and the USGS study of arsenic occurrence in ground water (USGS, 2000). The databases that supported the NAOS and USGS estimates are briefly described in section V.A., ``What data did EPA evaluate?'' Each of these occurrence estimates was developed in a slightly different manner. Whereas EPA's occurrence estimates are based on compliance monitoring data from more than 19,000 CWSs in 25 states, the NAOS occurrence estimates are based on a stratified random sampling from representative groups defined by source type, system size, and geographic location. The NAOS database contains 435 predicted finished water arsenic data points (derived from raw water arsenic concentrations and treatment information), from more than 400 CWSs. The USGS analysis is based on arsenic ambient (untreated, or raw water) ground water data, providing 17,496 samples for 1,528 counties (with 5 or more data points) in the United States (out of a total of 3,222 counties). USGS derived exceedance estimates for each county by calculating the percentage of data points in each county exceeding specific concentrations, from 1 µg/L to 50 µg/L. Then USGS associated the percentages for each county with the number of CWSs that use ground water in these counties, which was based on data derived from SDWIS. This information was aggregated for all of the appropriate counties to derive the national estimates for ground water CWSs. USGS did not have estimates for surface water CWSs. Table V-5.--Comparison of CWSs From EPA, NAOS, and USGS Estimates Exceeding Arsenic Concentrations ------------------------------------------------------------------------ EPA GW NAOS GW & % CWS exceeding &SW SW EPA GW USGS GW (percent) (percent) (percent) (percent) ------------------------------------------------------------------------ 2 µg/L.............. 24.1 21.7 27.2 25.0 5 µg/L.............. 10.3 11.5 12.1 13.6 10 µ/L.............. 4.5 4.5 5.4 7.6 ------------------------------------------------------------------------ Table V-5 compares the EPA, NAOS, and USGS estimates of the percent of samples exceeding various arsenic concentrations. At a concentration of 2 µg/L, the EPA national exceedance estimate for both surface water and ground water CWSs (24.1 percent) is higher than the NAOS estimate (21.7 percent). At 5 µg/L, the EPA and NAOS predicted exceedance probabilities are relatively similar (10.3 and 11.5 percent, respectively). These two estimates are the same at 10 µg/L (4.5 percent). For ground water CWSs, the USGS and EPA estimates are also relatively similar. At 2 µg/L, the EPA national ground water exceedance estimate (27.2 percent) is slightly higher than the USGS estimate (25.0 percent). At 5 and 10 µg/L, the USGS exceedance estimates (13.6 percent and 7.6 percent, respectively) are slightly higher than the EPA estimates (12.1 percent and 5.4 percent). This comparison of exceedance probabilities suggests that EPA's arsenic occurrence projections based on compliance monitoring data are relatively close to the NAOS and USGS projections through the range of this comparison. In addition, the USGS estimates are expected to be slightly higher than the EPA estimates for ground water, because they are based on raw water arsenic levels (untreated). F. What Are the National Occurrence Estimates of Arsenic in Drinking Water for Non-Transient, Non-Community Water Systems? The 25-States database contains data for non-transient, non- community water systems (NTNCWSs) in 15 States (two additional States only provided data from two systems). NTNCWSs are public water systems that regularly serve at least 25 of the same persons more than 6 months a year. Most NTNCWSs serve less than 3,300 people (99.5%) and use ground water (96%). EPA calculated basic statistics for ground water CWSs and NTNCWSs in each of these States. EPA compared the data and found that arsenic distributions in NTNCWSs are quite [[Page 38910]] similar to arsenic distributions in CWSs. In general, the means, standard deviations, and level of censoring for CWSs in a particular State are very close to the levels observed in NTNCWSs in that State. In some States, mean levels are slightly higher in CWSs than in NTNCWSs, whereas in others, mean levels are slightly lower in CWSs. There is no clear pattern and the differences are relatively minor, suggesting that any differences are due to random variation, rather than systematic underlying differences between NTNCWSs and CWSs. As a result, the occurrence distributions for CWSs were used to derive the occurrence distributions for NTNCWS systems. If the NTNCWSs data from the 15 States were used to derive the estimates, there would have been less spatial coverage of United States, which would have resulted in more uncertainty in the estimate. The NTNCWSs estimates are presented in Tables V-6 and V-7. As in the case of Tables V-3 and V-4, the estimated numbers of systems in Tables V-6 and V-7 with mean concentrations above 50 µg/L do not represent the number of systems which are believed to be out of compliance with the current MCL of 50 µg/L; nor do they represent actual systems at all. Rather they represent the estimated number of systems which would have mean arsenic concentrations above 50 µg/L if they had not treated for arsenic. By comparison with Tables V-6 and V-7, during the three-year period from September 1994 through August 1997, EPA recorded a total of 14 samples from 10 public water systems in which arsenic concentrations exceeded 50 µg/L. Table V-6.--Statistical Estimates of Number of Ground Water NTNCWSs With Average Arsenic Concentrations in Specified Ranges -------------------------------------------------------------------------------------------------------------------------------------------------------- Number of systems with average arsenic concentrations in specified ranges (µg/L; 19,293 systems total) System size (population served) -------------------------------------------------------------------------------------------------- >2.0 to >3.0 to >5.0 to >10.0 to >15.0 to >20.0 to >30.0 to 2.0 3.0 5.0 10.0 15.0 20.0 30.0 50.0 >50.0 -------------------------------------------------------------------------------------------------------------------------------------------------------- 25 to 500............................................ 12,088 1,223 1,285 1,111 382 178 163 106 73 501 to 3,300......................................... 1,902 192 202 175 60 28 26 17 11 3,301 to 10,000...................................... 43 4 5 4 1 1 1 0 0 10,001 to 50,000..................................... 8 1 1 1 0 0 0 0 0 > 50,000............................................. 0 0 0 0 0 0 0 0 0 Total............................................ 14,041 1,421 1,493 1,291 444 206 189 123 85 (% of systems)................................... (72.8%) (7.4%) (7.7%) (6.7%( (2.3%) (1.1%) (1.0%) (0.6%) (0.4%) -------------------------------------------------------------------------------------------------------------------------------------------------------- Note: Totals may not add up due to rounding of the number of systems to the nearest whole number. Systems serving fewer than 25 people are not included in this table. The estimates in this table do not take into account most treatment in place; in particular most of the systems in the ``>50.0'' column will have treated for arsenic in order to reduce their concentration below 50 µg/L. See text for more details. Table V-7.--Statistical Estimates of Number of Surface Water NTNCWSs With Average Arsenic Concentrations in Specified Ranges -------------------------------------------------------------------------------------------------------------------------------------------------------- Number of systems with average arsenic concentrations in specified ranges (``µg/L; 764 systems total) System size (population served) -------------------------------------------------------------------------------------------------- >2.0 to >3.0 to >5.0 to >10.0 to >15.0 to >20.0 to >30.0 to 2.0 3.0 5.0 10.0 15.0 20.0 30.0 50.0 >50.0 -------------------------------------------------------------------------------------------------------------------------------------------------------- 25 to 500............................................ 502 22 17 12 2 1 1 0 0 501 to 3,300......................................... 163 7 5 4 1 0 0 0 0 3,301 to 10,000...................................... 18 1 1 0 0 0 0 0 0 10,001 to 50,000..................................... 4 0 0 0 0 0 0 0 0 50,000............................................... 2 0 0 0 0 0 0 0 0 Total............................................ 688 30 23 17 3 1 1 1 0 (% of systems)................................... (90.1%) (3.9%) (3.0%) (2.2%) (0.4%) (0.1%) (0.1%) (0.1%) (0.1%) -------------------------------------------------------------------------------------------------------------------------------------------------------- Note: Totals may not add up due to rounding of the number of systems to the nearest whole number. Systems serving fewer than 25 people are not included in this table. The estimates in this table do not take into account most treatment in place; in particular most of the systems in the ``>50.0'' column will have treated for arsenic in order to reduce their concentration below 50 µg/L. See text for more details. G. How Do Arsenic Levels Vary From Source To Source and Over Time? EPA analyzed the variability of arsenic concentrations within a system, from well to well or entry point to entry point (sampling point). This analysis allows EPA to estimate the number of sampling points in a system that may be above the proposed MCL and to improve estimation of the treatment costs for systems with multiple sampling points. The result of the intrasystem analysis is a constant coefficient of variation (CV), which is one of the inputs to the cost- benefit computer modeling. EPA analyzed six of the eight States that provided intrasystem data: California, Utah, New Mexico, Oklahoma, Illinois and Indiana. Arkansas and Alabama were not analyzed because these States had very little occurrence of arsenic and almost all of the arsenic values were below the detection limit. After statistical analysis of 127 systems with five or more sampling points, EPA derived an arithmetic average CV of 0.64 or 64%. The EPA document titled ``Arsenic Occurrence in Public Drinking Water Supplies,'' presents this statistical analysis (US EPA, 2000b). USGS examined its raw water arsenic data to assess the variability of arsenic levels over time and to determine whether there are temporal trends (USGS, 2000). Data came from about 350 wells with 10 or more arsenic analyses collected over different time periods. These wells were used for various purposes, such as public supply, research, agriculture, industry, and domestic supply, and encompassed non-potable and potable water quality. USGS conducted a regression analysis [[Page 38911]] of arsenic concentration and time for each well and found that most of the wells had little or no change in concentration over time (low ``r- squared'' values when arsenic concentrations were regressed with time). Arsenic levels for most of the wells probably do not consistently increase or decrease over time. In addition, USGS found that well depth had no relationship to temporal variability. To determine the extent of the temporal variability, EPA analyzed the CVs for the mean arsenic level in the wells. More than 100 wells had a CV and standard deviation of zero. Most of these wells consistently had arsenic concentrations below the detection limit of 1 µg/L. EPA examined the CVs for the other wells in relation to the mean arsenic level and found a relatively constant CV on the lognormal scale. The geometric mean of the CVs, excluding CVs of zero, is 0.39 or 39%. The report (USGS, 2000) listed several factors that may contribute to this variability, including natural variability in geochemistry or source of contamination, sampling technique, and changes in pumping over time. H. How Did EPA Evaluate Co-Occurrence? Sections 1412(b)(3)(C)(i)(II), (III) and (VI) of the SDWA, as amended in 1996, require EPA to take into account activities under preceding rules which may have impacts on each new successive rule. To fulfill this need EPA began the analysis of the co-occurrence of drinking water contaminants. The information on co-occurrence will be used to determine the level of overlap in regulatory requirements. For example, this will include cases where treatment technologies applied for one regulation may resolve monitoring and/or additional treatment needs for another regulation or where water utilities may incur costs for installing multiple treatments to address other co-occurring substances. This information may also be used to show where specific levels of one contaminant may interfere with the treatment technology for another. 1. Data For the co-occurrence analysis, EPA relied on data from the National Water Information System (NWIS), a U.S. Geological Survey (USGS) database. The NWIS database was used for several reasons:
  • It contains both ground and surface water data;
  • It is national in scope, representing raw water samples from approximately 40,000 observation stations across the U.S.; and
  • It provides latitude/longitude coordinates for monitoring stations, which can be used in subsequent analyses to associate with Public Water Supply Systems. NWIS contains a water quality data storage retrieval system developed by the USGS Water Resources Division. NWIS is a distributed water database; data can be processed over a network of computers at USGS offices throughout the U.S. The system comprises the Automated Data Processing System, the Ground Water Site Inventory System, the Water-Quality System, and the Water-Use Data System. NWIS does not represent Public Water Supply Systems directly but can be associated with them because it provides latitude/longitude coordinates for monitoring stations. Using the NWIS data, arsenic was analyzed with 18 other constituents. The other constituents included: Sulfate, radon, radium, uranium, nitrate, antimony, barium, beryllium, cadmium, chromium, cyanide, iron, manganese, mercury, nickel, nitrite, selenium, thallium, hardness, and total dissolved solids. An additional set of ancillary parameters were selected for use as indicators of the hydrogeologic and geochemical conditions that could influence the co-occurrence of specific constituents. These ancillary parameters included: turbidity, conductance, dissolved oxygen, pH, alkalinity, well depth, and depth below land. 2. Results of the Co-occurrence Analysis (US EPA, 1999f) Dissolved arsenic was observed to have 5442 detected counts and total arsenic was observed to have 1273 detected counts in the database at the minimal threshold level of 2 µg/L. The national co- occurrence estimates derived from the USGS NWIS data revealed several correlations between arsenic/sulfate and arsenic/iron at the threshold levels chosen by EPA as likely to affect treatment (see section VIII.). First, a significant correlation was observed between dissolved arsenic and sulfate in surface water and ground water samples at the national level. The analysis of the surface and ground water data from EPA Regions 1, 2, 4, 5, 6, 7, 8, 9 and 10 show 339 co-occurrence frequency counts of the data above the threshold values of dissolved arsenic >5 µg/L and sulfate >250 mg/L (Table V-8). For total arsenic and sulfate there are 143 co-occurrence frequency counts for the same threshold levels. There was no significant co-occurrence of arsenic and sulfate in EPA Region 3. Secondly, a correlation was observed between dissolved arsenic and iron and total arsenic and iron in surface and ground waters from EPA Regions 1, 2, 4, 5, 7, 8 and 9 (Table V-8). There are 562 co-occurrence frequency counts of the data above the threshold levels of dissolved arsenic >5 µg/L and iron >300 µg/L. There are 542 co-occurrence frequency counts of the data above the threshold values of total arsenic >5 µg/L and iron >300 µg/L. There was no significant co-occurrence of arsenic and iron in EPA Regions 3, 6 and 10. Table V-8.--Correlation of Arsenic With Sulfate and Iron (Surface and Ground Waters) ------------------------------------------------------------------------ Correlation Arsenic types elements and Frequency EPA regions (threshold levels their threshold counts >5 µg/L) level ------------------------------------------------------------------------ 1, 2, 4, 5, 6, 7, 8, Dissolved Arsenic Sulfate (>250 mg/ 339 9, 10. L). Total Arsenic.... Sulfate (>250 mg/ 143 L). 1, 2, 4, 5, 7, 8, 9... Dissolved Arsenic Iron (>300 562 µg/L). Total Arsenic.... Iron (>300 542 µg/L). ------------------------------------------------------------------------ The results also show some co-occurring pairs of arsenic with radon. This appears to occur in EPA Regions 5 and 6 for ground water. However, the co-occurrence of arsenic and radon at levels of concern is not significant (Table V-9). At present, the analysis does not show significant co-occurring pairs between arsenic and radon in surface water in any EPA region. The impact from the co-occurrence of arsenic and radon is not a concern on a national level because there was no significant co-occurring pairs in EPA Regions 1, 2, 3, 4, 7, 8, 9, and 10. EPA requests comments on whether the NWIS database and this analysis is appropriate to use to represent co- [[Page 38912]] occurrence of arsenic with other constituents. Table V-9.--Correlation of Arsenic With Radon (Ground Water) ------------------------------------------------------------------------ Radon and Arsenic types and threshold Frequency EPA regions threshold levels (g/L) l) ------------------------------------------------------------------------ 5 and 6............ Dissolved 2££5.. 100£ 58 e300 300£ 140 e1000 Dissolved 5£10. 100£ 124 e300 300£ 101 e1000 Total 2£5...... 0£ 2 100 100£ 2 e300 Total 5£10..... 0£ 1 100 100£ 1 e300 ------------------------------------------------------------------------ VI. Analytical Methods A. What Section of SDWA Requires the Agency To Specify Analytical Methods? Section 1401 of SDWA directs EPA to promulgate national primary drinking water regulations (NPDWRs) which specify either MCLs or treatment techniques for drinking water contaminants (42 U.S.C. 300g- 1). EPA is required to set an MCL ``if, in the judgement of the Administrator, it is economically and technologically feasible to ascertain the level of a contaminant in water in public water systems'' (SDWA section 1401(1)(C)(i)). Alternatively, ``if, in the judgement of the Administrator, it is not economically or technologically feasible to so ascertain the level of such contaminant,'' the Administrator may identify known treatment techniques, which sufficiently reduce the contaminant in drinking water, in lieu of an MCL (SDWA section 1401(1)(C)(ii)). In addition, the NPDWRs are required to include ``criteria and procedures to assure a supply of drinking water which dependably complies with such maximum contaminant levels; including accepted methods for quality control and testing procedures to insure compliance with such levels * * *'' (SDWA section 1401(1)(D)) B. What Factors Does the Agency Consider in Approving Analytical Methods? In deciding whether an analytical method is economically and technologically feasible to determine the level of a contaminant in drinking water, the Agency considers the following factors:
  • Is the method sensitive enough to address the level of concern (i.e., the MCL)?
  • Does the method give reliable analytical results at the MCL? What is the precision (or reproducibility) and the bias (accuracy or recovery)?
  • Is the method specific? Does the method identify the contaminant of concern in the presence of potential interferences?
  • Is the availability of certified laboratories, equipment and trained personnel sufficient to conduct compliance monitoring?
  • Is the method rapid enough to permit routine use in compliance monitoring?
  • What is the cost of the analysis to water supply systems? C. What Analytical Methods and Method Updates Are Currently Approved for the Analysis of Arsenic in Drinking Water? EPA approved analytical methods and method updates for the analysis of arsenic in drinking water in previous rulemakings. EPA took the factors listed in section VI.B into consideration when it approved these methods and updates. The methods and updates, listed in Table VI- 1, are based on atomic absorption, atomic emission and mass spectroscopy methodologies and have been used for compliance monitoring of arsenic at the 0.05 mg/L MCL by State, Federal and private laboratories for many years. In this section on the discussion of analytical methods, and in the sections discussing the consumer confidence rule and public notification, EPA uses the mg/L units of measure, the units used in the regulatory language. Note that EPA's drinking water analytical methods refer to mg/L instead of µg/ L, and milligrams are 1,000 times larger than micrograms. Table VI-1.--Approved Analytical Methods (and Method Updates) for Arsenic (CFR 141.23) ---------------------------------------------------------------------------------------------------------------- MDL 2 or Methodology Reference method 1 EDL 3 (mg/ L) ---------------------------------------------------------------------------------------------------------------- Inductively Coupled Plasma Atomic Emission Spectroscopy 200.7 (EPA) 0.008 (ICP-AES). 3120B (SM) 3 0.050 Inductively Coupled Plasma Mass Spectroscopy (ICP-MS) 200.8 (EPA) 0.0014 ICP-MS with Selective Ion Monitoring. 4 (0.0001) Stabilized Temperature Platform Graphite Furnace Atomic 200.9 (EPA) 0.0005 Absorption (STP-GFAA) STP-GFAA with Multiple 5 (0.0001) Depositions. Graphite Furnace Atomic Absorption (GFAA)............... 3113B (SM) 3 0.001 D-2972-93C (ASTM) 3 0.005 Gaseous Hydride Atomic Absorption (GHAA)................ 3114B (SM) 3 0.0005 D-2972-93B (ASTM) 3 0.001 ---------------------------------------------------------------------------------------------------------------- 1 The reference methods approved for measuring arsenic in drinking water are cited in 40 CFR 141.23. The reference methods include: EPA = ``Methods for the Determination of Metals in Environmental Samples--Supplement I'', EPA/600/R-94-111, US EPA, May 1994. (US EPA, 1994b) [[Page 38913]] SM = Standard Methods for the Examination of Water and Wastewater, 18th and 19th eds., Washington, D.C., 1992 and 1995. (APHA, 1992 and 1995 respectively). The 19th edition of SM was approved in the December 1, 1999 final methods rule (64 FR 67450, US EPA 1999j). ASTM = Annual Book of ASTM Standards: Waster and Environmental Technology,'' Vol. 11.01 and 11.02, American Society for Testing and Materials, 1994 and 1996. (ASTM, 1994 and 1996). The 1996 edition of ASTM was approved in the December 1, 1999 final methods rule (64 FR 67450, US EPA 1999j). 2 MDL = Method Detection Limit = ``the minimum concentration of a substance that can be measured and reported with 99% confidence that the analyte concentration is greater than zero.'' (40 CFR Part 136 Appendix B). 3 EDL = Estimated Detection Limit (EDL) is defined as either the MDL or a concentration of a compound in a sample yielding a peak in the final extract with a signal to noise ratio of 5, whichever value is greater. Although the ASTM GFAA method (D-2972-93C) has a reported EDL of 0.005 mg/L, this method is similar to other GFAA methods. EPA believes D-2972-93C is capable of detection limits similar to other GFAA methods. 4 In 1994 (59 FR 62456; US EPA, 1994c), the Agency approved the use of the updated ``Methods for the Determination of Metals in Environmental Samples--Supplement I,'' (US EPA, EPA/600/R-94/111, 1994). The revised manual allows the use of selective ion monitoring with ICP-MS. The determined MDL for the direct analysis of arsenic in aqueous samples was 0.1 µg/L. 5 In 1994 (59 FR 62456; US EPA, 1994c), the Agency approved the use of the updated ``Methods for the Determination of Metals in Environmental Samples--Supplement I,'' (US EPA, EPA/600/R-94/111, 1994). The revised manual allows the use of multiple depositions with STP-GFAA. The determined MDL for arsenic using multiple deposition with STP-GFAA is 0.1 µg/L. D. Will Any of the Approved Methods for Arsenic Analysis Be Withdrawn? EPA believes all of the analytical methods listed in Table VI-1, with the exception of EPA Method 200.7 and SM 3120B, are technically and economically feasible for compliance monitoring of arsenic in drinking water at the proposed MCL of 0.005 mg/L. EPA is proposing to withdraw approval for EPA Method 200.7 and SM 3120B because the detection limit for the first ICP-AES method, 0.008 mg/L, and the estimated detection limit for the second ICP-AES method, 0.050 mg/L, are inadequate to reliably determine the presence of arsenic at the proposed MCL of 0.005 mg/L. Analysis of the Water Supply (WS) studies used to derive the PQL (Analytical Methods Support Document, US EPA, 1999l) indicates that ICP-AES technology was rarely used for low level arsenic analysis. Therefore, the Agency believes the removal of the methods that use ICP-AES technologies will not have an impact on laboratory capacity. Even at the MCL options of 0.003, 0.010 mg/L, and 0.020 mg/L, the Agency would still withdraw both EPA Method 200.7 and SM 3120B. At these MCL options, these methods are still inadequate for compliance monitoring of arsenic in drinking water. E. Will EPA Propose Any New Analytical Methods for Arsenic Analysis? The Agency conducted a literature search to identify additional analytical methods which are capable of compliance monitoring of arsenic at the proposed MCL of 0.005 mg/L (Analytical Methods Support Document, US EPA, 1999l). A large majority of the analytical techniques identified from the literature search were from EPA's Office of Solid Waste SW-846 methods manual, which can be accessed online at www.epa.gov/epaoswer/hazwaste/test/index.html. Of the Solid Waste methods, the Agency evaluated:
  • SW-846 Method 6020 (ICP-MS, MDL = 0.0004 mg/L; US EPA, 1994d);
  • SW-846 Method 7060A (GFAA, MDL = 0.001 mg/L; US EPA, 1994e);
  • SW-846 Method 7062 (GFAA, MDL = 0.001 mg/L; US EPA, 1994f);
  • SW-846 Method 7063 (Anodic Stripping Voltammetry-ASV, MDL = 0.0001 mg/L; US EPA, 1996d); In addition to the SW-846 method, the Agency also reviewed:
  • EPA Method 1632 (a wastewater GHAA method with an MDL = 0.000002 mg/L or 0.002 µg/L; US EPA 1996a); and
  • EPA Method 200.15 (an ICP-AES with ultrasonic nebulization as part of the written method, MDL = 0.003 mg/L or 0.002 mg/L; US EPA, 1994a). Although the SW-846 methods and the EPA 1632 wastewater method are capable of reaching the detection limits needed at the proposed arsenic MCL, most of these techniques (with the exception of the method using ASV technology) are similar to methods that have already been approved for the analysis of arsenic in drinking water. The Agency does not believe approval of these methods for drinking water would provide additional analytical benefits. Moreover, the addition of the SW-846 methods could complicate the laboratory certification process because SW-846 methods are not mandatory procedures, but rather guidance. At this time, laboratories are certified at different times for different EPA programs. Therefore, laboratories certified for both drinking water methods and Office of Solid Waste methods may need to be certified separately under both programs to use SW-846 methods for drinking water. While SW-846 Method 7063 (using ASV technology) is not similar to any technique approved thus far, this method will not be approved for the measurement of arsenic in drinking water because it only detects dissolved arsenic as opposed to total arsenic. Today's proposal would regulate total arsenic in drinking water not dissolved arsenic. The techniques currently approved for drinking water measure total arsenic (arsenic species in the dissolved and suspended fractions of a water sample). A preliminary total metals digestion would be necessary with the ASV technique in order to determine the total arsenic concentration in a drinking water sample. The Agency also reviewed but does not propose to approve EPA Method 200.15, an ICP-AES method which requires the use of ultrasonic nebulization to introduce the sample into the plasma. To provide uniform signal response using EPA Method 200.15, it is necessary for arsenic to be in the pentavalent state. The addition of hydrogen peroxide to the mixed acid solutions of samples and standards prior to ultrasonic nebulization is necessary to convert all of the arsenic species to the pentavalent state. Although EPA Method 200.15 is capable achieving a MDL of 0.003 mg/L using direct analysis and a MDL of 0.002 mg/L using a total recoverable digestion and a 2-fold concentration, these levels of detection are still insufficient for compliance monitoring at the proposed MCL of 0.005 mg/L. At the MCL options of 0.010 mg/L and 0.020 mg/L, the Agency would approve the use of EPA Method 200.15 but only with the use of a total recoverable digestion and a 2-fold concentration (MDL = 0.002 mg/L). At an MCL option of 0.003 mg/L, EPA method 200.15 would not be approved. F. Other Method-Related Items 1. The Use of Ultrasonic Nebulization with ICP-MS In the September 3, 1998 Analytical Methods for Drinking Water Contaminants Proposed Rule (63 FR 47907; US EPA 1998d), EPA proposed the use of ultrasonic nebulization with EPA Method 200.7 (ICP-AES) and EPA Method 200.8 (ICP-MS). Because EPA Method 200.7 and SM 3120B will be withdraw for the analysis of arsenic in drinking water under the proposed MCL of 0.005 mg/L, ultrasonic nebulization as a modification would not be allowed. [[Page 38914]] Even with the modification of ultrasonic nebulization, the ICP-AES method is not capable of compliance monitoring for arsenic at the proposed MCL of 0.005 mg/L. EPA Method 200.8 (ICP-MS) would still be allowed for compliance monitoring at the proposed MCL of 0.005 mg/L. The use of ultrasonic nebulization can enhance transport efficiency and lower the detection limits for ICP-MS by approximately 5 to 10 fold. The final methods update rule was published in the Federal Register on December 1, 1999 (64 FR 67450; US EPA 1999j). 2. Performance-Based Measurement System On October 6, 1997, EPA published a Notice of the Agency's intent to implement a Performance Based Measurement System (PBMS) in all of its programs to the extent feasible (62 FR 52098; US EPA, 1997e). EPA is currently determining how to adopt PBMS into its drinking water program, but has not yet made final decisions. When PBMS is adopted into the drinking water program, its intended purpose will be to increase flexibility in laboratories in selecting suitable analytical methods for compliance monitoring, significantly reducing the need for prior EPA approval of drinking water analytical methods. Under PBMS, EPA will modify the regulations that require exclusive use of Agency- approved methods for compliance monitoring of regulated contaminants in drinking water regulatory programs. EPA will probably specify ``performance standards'' for methods, which the Agency would derive from the existing approved methods and supporting documentation. A laboratory would be free to use any method or method variant for compliance monitoring that performed acceptably according to these criteria. EPA is currently evaluating which relevant performance characteristics under PBMS should be specified to ensure adequate data quality for drinking water compliance purposes. After PBMS is implemented, EPA may continue to approve and publish compliance methods for laboratories that choose not to use PBMS. After EPA makes final determinations about the implementation of PBMS in programs under the Safe Drinking Water Act, the Agency would then provide specific instruction on the specified performance criteria and how these criteria would be used by laboratories for compliance monitoring of SDWA analytes. G. What Are the Estimated Costs of Analysis? To obtain cost information on the analysis of arsenic in drinking water, the Agency collected price information from a random telephone survey of seven commercial laboratories, which were certified in drinking water analysis, and from price lists posted on the Internet (Analytical Methods Support Document, US EPA, 1999l). Table VI-2 summarizes the results of this survey, including the specific methodology and the associated cost range. The actual costs of performing an analysis may vary with laboratory, the analytical technique selected, and the total number of samples analyzed by a laboratory. The estimated cost range is only for the analysis of arsenic and does not include shipping and handling costs. The Agency solicits comments from the public on the cost estimates listed in Table VI-2. Table VI-2.--Estimated Costs for the Analysis of Arsenic in Drinking Water \1\ ------------------------------------------------------------------------ Estimated cost range Methodology ($) ------------------------------------------------------------------------ Inductively Coupled Plasma Atomic Emission 15 to 25. Spectroscopy (ICP-AES). Inductively Coupled Plasma Mass Spectroscopy 10 to 15. (ICP-MS). Stabilized Temperature Platform Graphite Furnace 15 to 50. Atomic Absorption (STP-GFAA). Graphite Furnace Atomic Absorption (GFAA)....... 15 to 50. Gaseous Hydride Atomic Absorption (GHAA)........ 15 to 50. ------------------------------------------------------------------------ \1\ Analytical Methods Support Document (US EPA, 1999l). H. What Is the Practical Quantitation Limit? Method detection limits (MDLs) and practical quantitation levels (PQLs) are two performance measures used by EPA's drinking water program to estimate the limits of performance of analytic chemistry methods for measuring contaminants in drinking water. As cited in Table VI-1, EPA defines the MDL as ``the minimum concentration of a substance that can be measured and reported with 99% confidence that the analyte concentration is greater than zero (40 CFR part 136, appendix B).'' MDLs can be operator, method, laboratory, and matrix specific. MDLs are not necessarily reproducible within a laboratory or between laboratories on a daily basis due to the day-to-day analytical variability that can occur and the difficulty of measuring an analyte at very low concentrations. In an effort to integrate this analytical chemistry data into regulation development, EPA's OGWDW uses the PQL to estimate or evaluate the minimum, reliable quantitation level that most laboratories can be expected to meet during day-to-day operations. EPA's Drinking Water program defined the PQL as ``the lowest concentration of an analyte that can be reliably measured within specified limits of precision and accuracy during routine laboratory operating conditions (50 FR 46906, November 13, 1985).'' 1. PQL Determination A PQL is determined either through the use of interlaboratory studies or, in absence of sufficient information, through the use of a multiplier of 5 to 10 times the MDL. The inter-laboratory data is obtained from water supply (WS) performance evaluation (PE) studies that are conducted twice a year by EPA to certify drinking water laboratories (now referred to as the Performance Testing or PT program). In addition to certification of drinking water laboratories, WS studies also provide:
  • Large-scale evaluation of analytical methods;
  • A database for method validation;
  • Demonstration of method utilization by a large number of laboratories; and
  • Data for PQL determinations. Using graphical or linear regression analysis of the WS data, the Agency sets a PQL at a concentration where at least 75% of the laboratories (generally EPA and State laboratories) could perform within an acceptable level of precision and accuracy. This method of deriving a PQL was used in the past for inorganics such as antimony, beryllium, [[Page 38915]] cyanide, nickel and thallium (57 FR 31776 at 31800; US EPA, 1992b). 2. PQL for Arsenic In 1994, EPA derived a preliminary PQL for arsenic based on data collected by the Agency from WS studies 20 through 33 (WS 31 was excluded because the spiked samples were mixed incorrectly). In response to concerns from the water utility industry, the results of this derivation and a separate evaluation conducted by the American Water Works Association (AWWA) were reviewed by the EPA Science Advisory Board (SAB) in 1995. The SAB noted that the acceptance limits of + 40% used by EPA to derive the PQL in 1994 were wider than those for other SDWA metal contaminants. The acceptance limits and PQLs for several SDWA metals are shown in Table VI-3. The SAB recommended that EPA set the PQL using acceptance limits similar to those used for other inorganics. Table VI-3.--Acceptance Limits and PQLs for Other Metals (in Order of Decreasing PQL) ------------------------------------------------------------------------ Acceptance Contaminant limit \1\ PQL (mg/L) (percent) \2\ ------------------------------------------------------------------------ Barium....................................... ±15 Chromium..................................... ±15 Selenium..................................... ±20 Antimony..................................... ±30 Thallium..................................... ±30 Cadmium...................................... ±20 Beryllium.................................... ±15 Mercury...................................... ±30 ------------------------------------------------------------------------ \1\ Acceptance limits for the listed inorganics are found at CFR 141.23 (k) (3)(ii). \2\ The PQL for antimony, beryllium and thallium was published in 57 FR 31776 at 31801 (July 17, 1992; US EPA, 1992b). The PQL for barium, cadmium, chromium, mercury and selenium was published in 66 FR 3526 at 3459 (January 30, 1991; US EPA, 1991a). Subsequent to SAB's recommendation, EPA derived a new PQL for arsenic (Analytical Methods Support Document, US EPA, 1999l). The process employed by the Agency to determine the new PQL utilized:
  • Data from six voluntary, low-level (0.006 mg/L of arsenic) WS studies;
  • Acceptance limits similar to other low-level inorganics; and
  • Linear regression analysis to determine the point at which 75% of EPA Regional and State laboratories fell within the chosen acceptance range. The derivation of the PQL for arsenic was consistent with the process used to determine PQLs for other metal contaminants regulated under SDWA and took into consideration the recommendations from the SAB. Using acceptance limits of + 30% and linear regression analysis of WS studies 30 through 36 (excluding 31) yielded a PQL of 0.00258 mg/L. The Agency rounded up to derive a PQL for arsenic of 0.003 mg/L at the 30% acceptance limit. While the PQL represents a stringent target for laboratory performance, the Agency believes most laboratories, using appropriate quality assurance and quality control procedures, will achieve this level on a routine basis. I. What Are the Sample Collection, Handling and Preservation Requirements for Arsenic? The manner in which samples are collected, handled and preserved is critical to obtaining valid data. Specific sample collection, handling and preservation procedures for SDWA analytes are outlined in the ``Manual for the Certification of Laboratories Analyzing Drinking Water'' (US EPA, 1997a). For metals such as arsenic, the certification manual specifies the following:
  • Nitric acid (HNO3 at pH 2) as the preservative;
  • A maximum sample holding time of 6 months;
  • And a sample size of 1 liter, collected in an appropriately cleaned plastic or glass container, is suggested. Currently, arsenic does not have an entry for preservation, collection, and holding time. EPA is proposing in this rule, to revise the table following Sec. 141.23(k)(2) to add ``arsenic, Conc. HNO3 to pH 2, P or G, and 6 months.'' EPA requests comment on the appropriateness of this revision. While 40 CFR 141.23(a)(4) allows compositing of up to 5 samples from the same PWS, the detection limit required for compositing must be \1/5\ of the MCL. Also, compositing for inorganic samples must be done in the laboratory. Samples should only be held if the laboratory detection limit is adequate for the number of samples being composited. In any case, the composite is not to exceed five samples. EPA is adding the test methods and detection limits for the approved arsenic analytical methods to the table following Sec. 141.23(a)(4)(i). J. Laboratory Certification 1. Background The ultimate effectiveness of today's regulation depends upon the ability of laboratories to reliably analyze arsenic at the proposed MCL. The existing drinking water laboratory certification program (LCP), which was established by States with guidance and recommendations from EPA, requires that only certified laboratories analyze compliance samples. External checks of a laboratory's ability to analyze samples of regulated contaminants within specific limits is the one means of judging laboratory performance and determining whether or not to grant certification. Under a performance testing (PT) program (formerly known as the performance evaluation or PE program), laboratories are required to successfully analyze PT samples (contaminant concentrations are unknown to the laboratory being reviewed) that are prepared by appropriate third parties. Successful participation in a PT program is a prerequisite for a laboratory to achieve certification and to remain certified for analyzing drinking water compliance samples. Achieving acceptable performance in these studies of unknown test samples provides some indication that the laboratory is following proper practices. Unacceptable performance may be indicative of problems that could affect the reliability of the compliance monitoring data. 2. What Are the Performance Testing Criteria for Arsenic? The Agency has historically identified acceptable performance using one of two different approaches: (a) Regressions from the performance of preselected laboratories (using 95 percent confidence limits), or [[Page 38916]] (b) Specified accuracy requirements. Acceptance limits based on specified accuracy requirements are developed from past PE study data. EPA has traditionally preferred to use the second (``true value'') approach because it is the better indicator of performance and provides laboratories with a fixed target. Under this approach, each laboratory demonstrates its ability to perform within pre-defined limits. Laboratory performance is evaluated using a constant ``yardstick'' independent of performance achieved by other laboratories participating in the same study. A fixed criterion based on a percent error around the ``true'' value reflects the experience obtained from numerous laboratories and includes relationships of the accuracy and precision of the measurement to the concentration of the analyte. It also assumes little or no bias in the analytical methods that may result in average reporting values different from the reference ``true'' value. In today's rulemaking, the Agency is proposing that the laboratory certification criteria for arsenic be set at an acceptance limit of + 30 % at > 0.003 mg/L in Sec. 141.23(k)(3)(ii). Analysis of water supply data indicate that laboratory capacity at this level should be sufficient for compliance monitoring. At this level, 75 % of EPA Regional and State laboratories and 62 % of non-EPA laboratories were capable of achieving acceptable results. As discussed in the Analytical Methods Support Document, (US EPA, 1999l), setting an acceptance limit of 20% would have decreased laboratory capacity. EPA requests comment on setting the acceptance limit at the upper range of SAB's recommendation. 3. How Often is a Laboratory Required To Demonstrate Acceptable PT Performance? EPA requires that a PT (PE) sample for chemical contaminants be successfully analyzed at least once a year using each method which is used to report compliance monitoring results. For arsenic this would require that the laboratory successfully analyze a PT (PE) sample using the method which is used to report the results for compliance monitoring. Additional guidance on the minimum quality assurance requirements, conditions of laboratory inspections and other elements of laboratory certification requirements for laboratories conducting compliance monitoring measurements are detailed in the Manual for the Certification of Laboratories Analyzing Drinking Water, Criteria and Procedures Quality Assurance (US EPA, 1997a), which can be downloaded via the Internet at ``http://www.epa.gov/ogwdw000/certlab/ labindex.html.'' 4. Externalization of the PT Program (Formerly Known as the PE Program) Due to resource limitations, on July 18, 1996 EPA proposed options for the externalization of the PT studies program (61 FR 37464; US EPA, 1996c). After evaluating public comment, in the June 12, 1997 final notice EPA (62 FR 32112; US EPA, 1997c): ``decided on a program where EPA would issue standards for the operation of the program, the National Institute of Standards and Technology (NIST) would develop standards for private sector PE (PT) suppliers and would evaluate and accredit PE suppliers, and the private sector would develop and manufacture PE (PT) materials and conduct PE (PT) studies. In addition, as part of the program, the PE (PT) providers would report the results of the studies to the study participants and to those organizations that have responsibility for administering programs supported by the studies.'' EPA has addressed this topic in public stakeholders meetings and in some recent publications, including the Federal Register notices mentioned in this paragraph. More information about laboratory certification and PT (PE) externalization can be accessed at the OGWDW laboratory certification website under the drinking water standards heading (www.epa.gov/safewater). VII. Monitoring and Reporting Requirements The currently applicable monitoring requirements for arsenic are different than the other inorganic contaminants (IOCs). First of all, arsenic's MCL and compliance requirements are found in Sec. 141.11, instead of in Sec. 141.62(b). Monitoring, compliance, and reporting requirements for arsenic are also different than the standardized monitoring framework for the grouped IOCs (which does not include radon). EPA is proposing to move arsenic to the standardized monitoring framework for IOCs (antimony, asbestos, barium, beryllium, cadmium, chromium, cyanide, fluoride, mercury, nickel, nitrate, nitrite, selenium, and thallium), including the State reporting and compliance requirements. Table VII-1 presents a comparison of the existing and proposed arsenic requirements, in abbreviated form. For a full picture of the regulations, you must look at the regulatory language. In addition, EPA is proposing to clarify the regulatory language for sampling to determine compliance for inorganics, volatiles and synthetic organic contaminants. Table VII-1.--Comparison of Sampling, Monitoring, and Reporting Requirements [This table is not complete for compliance purposes, but provides an overview for readers.] ------------------------------------------------------------------------ Requirement Current rule Proposed rule ------------------------------------------------------------------------ Compliance with Sec. MCL only applies to Would link 141.11(a). CWS and compliance compliance with 50 is calculated using µg/L with Sec. 141.23. Sec. 141.23(l) and would not add NTNCWS. Compliance with Sec. MCL is 0.05 mg/L.... MCL will remain 50 141.11(b). µg/L for CWS serving 10,000 or less until 5 years after publication of final rule, and be effective for larger systems 3 years after publication of final rule. New lower MCL in Sec. 141.62. NTNCWS will be subject to sampling, monitoring and reporting 3 years after publication of final rule, but not subject to increased monitoring after exceedances, nor to MCL violations. Monitoring frequency........ Groundwater Sec. No change to Sec. 141.23(a)(1) One 141.23(a)(1). sample at each entry point to the distribution system (sampling point). Surface water Sec. No change to Sec. 141.23(a)(2) One 141.23(a)(2). sample at every entry point to the distribution system (sampling point). [[Page 38917]] Compositing inorganics...... Sec. 141.23(a)(4) Adding approved may composite up to arsenic analytical 5 samples in the methods and lab; detection detection limits to limit \1/5\ of the the table following MCL. Sec. 141.23(a)(4)( i). Composite >\1/5\ MCL........ Sec. 141.23(a)(4)(i Same but Sec. ) take follow-up 141.23(a)(4)(i) samples within 14 table will list MCL days of each and detection sampling point in limits for arsenic. the composite. Compositing by system size.. Sec. 141.23(a)(4)(i No change to Sec. i) State may permit 141.23(a)(4)(ii). compositing at sampling points within a system serving >3,300 people. Sec. 141.23(a)(4)(i No change to Sec. i) State may permit 141.23(a)(4)(ii). compositing among different systems, 5-sample limit, systems serving 3,300 people. Resampling composites....... Sec. 141.23(a)(4)(i No change to Sec. ii) Can use 141.23(a)(4)(iii). duplicates of the original sample instead, must be analyzed and reported to State within 14 days of collection. Compliance with Sec. 141.11 Sec. 141.23(l)(1) Sec. 141.23(c)(1) CWSs have same CWS surface water surface water one requirements, but arsenic yearly. sample per monitoring would move from compliance point Sec. 141.23(l) to Sec. annually. 141.23(c). Sec. 141.23(l)(2) Sec. 141.23(c)(1) CWS ground water groundwater one every three years.. sample at each sampling point during each compliance period. Monitoring waivers Sec. None currently Sec. 141.23(c)(2) 141.23(c). available for System may apply to arsenic.. the State. Sec. 141.23(c)(3)Mu st take at least one sample during waiver, which cannot exceed one compliance period (9 years). Minimum data for waivers: .................... Sec. 141.23(c)(4) Surface water Ground water at least 3 years. All results MCL. New water At least 3 rounds source needs three rounds of monitoring. At of monitoring. least one sample must be taken after January 1, 1990. Once MCL exceeded sampling.. Sec. 141.23(m) Sec. 141.23(c)(7) Supplier must exceed MCL as report to State calculated in (i), within 7 days and go to quarterly initiate three monitoring next additional samples quarter. Sec. at the same 141.31(d) within 10 sampling point days of giving within a month. public notice, contact primacy agency. Sec. 141.203(b) Tier 2 public notice no later than 30 days after learning of violation and repeat every 3 months or at least once a year if allowed by primacy agency. Compliance based on less Not currently Sec. 141.23(i)(1) than required number of specified.. for IOCs, Sec. samples. 141.24(f)(15)(i) for VOCs, and Secs. 141.24(h)(11)(i) and (ii) for SOCs will average based on # samples collected. Average that determines Sec. 141.23(n) When Sec. 141.23(i)(5) violation.. the 4 analyses, arsenic will be State notice................ rounded to the same reported to the Public notice............... number of nearest 0.001 mg/L. significant figures Sec. 141.23(i)(1) as the MCL exceeds monitoring > the MCL, supplier annually, running must notify the annual average at State Sec. 141.31 sampling point. If and give notice to less samples taken the public Sec. than required, 141.32. Monitoring compliance is based frequency on average of determined by the samples. Any sample State must continue below method until MCL in two detection limit is consecutive samples assigned zero for or until a calculation. variance, exemption, or enforcement action schedule becomes effective. Sampling frequency after MCL .................. Sec. 141.23(i)(2) compliance monitoring begun. monitoring annually or less often if sampling point > MCL. Confirmation sample......... None currently If State requires a specified for confirmation arsenic. sample, then compliance based on average of the two samples. If State specifies additional monitoring, compliance based on running annual average. If less samples taken than required, compliance is based on average of samples. Increased monitoring .................. Sec. 141.23(f)(1) frequency. State may require one within two weeks. Sec. 141.23(c)(8) State can decrease monitoring after a minimum of 2 quarters for ground water and 4 quarters for surface water MCL. [[Page 38918]] Sec. 141.23(f)(1) If >MCL, State can require a confirmation sample within two weeks. Sec. 141.23(f)(3) Average used to determine compliance with (i). States can delete results with obvious sampling errors. Sec. 141.23(g) State may require more frequent monitoring. New system and new sources.. Only mentions waiver Sec. 141.23(c)(9) eligibility in Sec. IOCs, Sec. 141.23(c)(4). 141.24(f)(22) VOCs, Sec. 141.24(h)(20 SOCs, Compliance demonstrated within State-specified time and sampling frequencies. Subpart O Consumer >50 µg/L Lowers MCL & adds Confidence Reports for CWS. annual report Sec. MCLG to Appendices 141.153(d)(6) A & B to Subpart O- length of effective 30 days violation, after final arsenic potential health rule is published, effects using before compliance Appendix C, actions with lower MCL is taken. 25-50 in place. µg/L informational statement per Sec. 141.154(b). Subpart Q Public >50 µg/L Sec. 141.203(b) Notification for PWS. CWSs Tier 2 annual Tier 2 public report Sec. notice no later 141.203 required than 30 days after October 31, 2000 learning of (if they are in violation and jurisdictions where repeat every 3 the program is months or at least directly once a year if implemented by EPA) allowed by primacy or on the date a agency. primacy State adopts the new requirements (not to exceed May 6, 2002).. Sec. 141.31(d) within 10 days of giving public notice, contact primacy agency. >5 µg/L CWSs & add NTNCWS to Table 1 of Sec. 141.203 to require Tier 2 annual report Sec. 141.203 after effective date of arsenic MCL (3-5 yrs). ------------------------------------------------------------------------ A. What Are the Existing Monitoring and Compliance Requirements? The arsenic monitoring requirements appear in 40 CFR 141.23(a). Surface water systems must collect routine samples annually and ground water systems must collect a routine sample every three years. However, Sec. 141.11(a) currently only requires community water systems (CWS) to monitor for arsenic. EPA understands that some States also require their non-transient non-community water systems (NTNCWS) to collect samples for the analysis of arsenic as well. Under the proposal, CWSs would continue to be allowed to composite samples as specified in Sec. 141.23(a)(3); however, the one-fifth arsenic MCL will no longer be 10 µg/L (It will be 1 µg/L). Sections 141.23(l) through (q) are currently used to determine compliance for arsenic. That is, if arsenic is detected at a concentration greater than the maximum contaminant level (MCL), the community water system must collect 3 additional samples within one month at the entry point to the distribution system that exceeded the MCL (Sec. 141.23(n)). If the average of the four analyses performed, rounded to one significant figure, exceeds the MCL, the system must notify the State; and the system must provide public notice (Sec. 141.23(n)). After public notification, the monitoring continues at the frequency designated by the State until the MCL ``has not been exceeded in two successive samples or until [the State establishes] a monitoring schedule as a condition to a variance, exemption or enforcement action (Sec. 141.23(n)).'' Monitoring waivers are not permitted to exclude a system from the sampling requirements under Sec. 141.23(l)-(q) which currently apply to arsenic. B. How Does the Agency Plan To Revise the Monitoring Requirements? The Agency is proposing to require CWS and NTNCWSs to monitor for arsenic using Sec. 141.23(c). This will make the arsenic monitoring requirements consistent with the inorganic contaminants (IOC's) regulated under the standardized monitoring framework. EPA is proposing that NTNCWSs monitor and report arsenic results to the State and public, as a Tier 2 notice in subpart Q, Public Notification. However, the Agency is proposing that NTNCWSs not be required to meet the MCL, unlike the other inorganics listed in Sec. 141.62(b). EPA's analysis for not requiring NTNCWSs to comply with the MCL is based on the cost- benefit analysis discussed later in section XI.C. of this preamble. If arsenic exceeds the MCL, the CWS will be triggered into quarterly monitoring for that sampling point ``in the next quarter after the violation occurred (Sec. 141.23(c)(7).'' The State may allow the system to return to the routine monitoring frequency when the State determines that the system is reliably and consistently below the MCL. However, the State cannot make a determination that the system is reliably and consistently below the MCL until a minimum of 2 consecutive ground water or 4 consecutive surface water samples have been collected (Sec. 141.23(c)(8)). All systems must comply with the sampling requirements, unless a waiver has been granted in writing by the State (Sec. 141.23(c)(6)). As shown in Table VI-1, the approved methods can measure to 0.001 mg/L or below. In order to use the analytical power of the methods, EPA is proposing that arsenic data be reported to the nearest 0.001 mg/L. Therefore, a result of 0.0055 mg/L would be rounded to 0.006 mg/L, and 0.0145 mg/L would be rounded to 0.014 mg/L (Figures ending in ``5'' rounded down to end on an even digit and up to an even digit.). During the writing of this regulation, some people had asked whether data above 0.01 mg/L could be rounded to one significant figure because the MCL is being proposed with one significant figure. EPA is issuing a clarification to arsenic reporting in Sec. 141.23(i) to indicate that arsenic results will be reported to the nearest 0.001 mg/L. The significance for compliance purposes will be that values between 0.010 mg/ [[Page 38919]] L and 0.014 mg/L will be averaged to the nearest 0.001 mg/L, and the yearly average will more closely reflect the values measured. EPA requests comment on these clarifications to reporting requirements. C. Can States Grant Monitoring Waivers? As proposed, States will be able to grant a 9-year monitoring waiver to a system (Sec. 141.23(c)(3)). Waivers of arsenic sampling requirements must be based on all analytical results from previous sampling and a vulnerability assessment or the assessment from an approved source water assessment program (provided that the assessments were designed to collect all of the necessary information needed to complete a vulnerability assessment for a waiver). States issuing waivers must consider the requirements in 40 CFR 141.23(c)(2)-(6). In order to qualify for a waiver, there must be three previous samples from a sampling point (annual for surface water and three rounds for groundwater) with analytical results reported below the proposed MCL (i.e., the reporting limit must be 0.005 mg/L). The use of grandfathered data collected after January 1, 1990 that is consistent with the analytical methodology and detection limits of the proposed regulation may be used for issuing sampling point waivers. The existing Sec. 141.23(l)-(q) regulations do not permit the use of monitoring waivers. However, a State could now use the analytical results from the three previous compliance periods (1993-1995, 1996-1998, and 1999-2001) to issue ground water sampling point waivers. Surface water systems must collect annual samples so a State could use the previous 3 years sampling data (1999, 2000, and 2001) to issue sampling point waivers. One sample must be collected during the nine-year compliance cycle that the waiver is effective, and the waiver must be renewed every nine years. Vulnerability assessments must be based on a determination that the water system is not susceptible to contamination and arsenic is not a result of human activity (i.e., it is naturally occurring). Although the approved analytical methods can measure to 0.005 mg/L, not all States have required systems to report arsenic results below 50 µg/L. In this case, the States would not have adequate data to grant waivers until enough data are available to make the determinations. EPA has compliance monitoring data from 25 States at 10 µg/L and below. On the other hand, one State submitted data to EPA rounded to tens of µg/L, so some States may not be able to grant waivers until the data are reported below the proposed MCL. EPA believes that some States may have been regulating arsenic under the standardized inorganic framework being proposed today. If so, those States will have to ensure that existing monitoring waivers have been granted using data reported below the new proposed MCL. Otherwise States will have to notify the systems of the new lower reporting requirements that need to be met to qualify for a waiver for the proposed MCL. D. How Can I Determine if I Have an MCL Violation? For this proposal, violations of the arsenic MCL would be determined under Sec. 141.23(f)-(i). If a system samples more frequently than annually (e.g., quarterly), the system would be in violation if the running annual average at any sampling point exceeds the MCL or if any one sample would cause the annual average to be exceeded (Sec. 141.23(i)(1)). If a system conducts sampling at an annual or less frequent basis, the system would be in violation if one sample (or the average of the initial and State-required confirmation sample(s)), at any sampling point exceeds the MCL (Sec. 141.23(i)(2). However, States can require more frequent monitoring per Sec. 141.23(g) for systems sampling annually or less often. Therefore, the Agency is proposing to clarify this section for situations for IOCs in Sec. 141.23(i)(2)) and the corresponding sections for volatile and synthetic organic contaminants (Secs. 141.24(f)(15)(ii) and 141.24(h)(11)(ii), respectively. This proposal clarifies compliance for contaminants subject to Secs. 141.23(i)(2)), 141.24(f)(15)(ii), and 141.24(h)(11)(ii) by pointing out that compliance will be based on the running annual average of the initial MCL exceedance and subsequent State-required confirmation samples. These confirmation samples may be required at State-specified frequencies (e.g., quarterly or some other frequency depending on site-specific conditions). In addition, the clarifications to Secs. 141.23(i)(2)), (141.24(f)(15)(ii) and 141.24(h)(11)(ii) address calculation of compliance when a system fails to collect the required number of samples. Compliance (determined by the average concentration) would be based on the total number of samples collected. The Agency expects systems will conduct all required monitoring. However, some systems have purposely not collected the required number of quarterly samples, and in doing so some avoided reporting an MCL violation. While these systems all incurred monitoring and reporting violations for the uncollected samples, some systems divided the sum of the samples taken by four, which lowered the annual average reported to below the MCL, avoiding an MCL violation. The Agency requests comment on this clarification of exceedances determined under a State-determined monitoring frequency. For purposes of calculating MCL annual averages, Sec. 141.23(i)(1) continues to set all non-detects equal to a value of zero. However, the Agency realizes that some States use the detection limit or a fraction of the detection limit to calculate an average. E. When Will Systems Have To Complete Initial Monitoring? The rule becomes effective 3 years after promulgation (about January 1, 2004) for large PWS (serving over 10,000). This will require all GW and SW systems serving over 10,000 to complete the initial round of monitoring by December 31, 2004. However, States may allow systems, on a case-by-case basis, 2 additional years to comply with the MCL if capital improvements are necessary. The Agency is proposing a national finding that capital improvements are necessary for public water systems serving less than 10,000, on the basis that existing treatments are not expected to be effective in arsenic removal. Table VII-2 shows the percentage of small systems with no treatment in place as well as the percentage of systems which currently have in place technologies that can remove arsenic. The data shows that capital improvements would be necessary for many systems. The rule would be effective 5 years after promulgation (about January 1, 2006) for systems serving under 10,000. This would require these small GW systems to complete the initial round of monitoring by the December 31, 2007 ('05-'07 compliance period), and small SW systems to complete the initial round of monitoring by December 31, 2006. EPA is requesting comment on whether it is appropriate to make a national finding that systems serving less than 10,000 people will need the two additional years to add capital improvements in order to comply with the proposed MCL. The alternative would require States to issue individual two-year extensions for these small systems. [[Page 38920]] Table VII-2.--Treatment In-Place at Small Water Systems (US EPA, 1999e and US EPA, 1999m) -------------------------------------------------------------------------------------------------------------------------------------------------------- Percent of Percent of Percent of Percent of Percent of systems with no systems with ion systems with systems with systems with treatment in exchange in coagulation/ lime softening reverse osmosis System size place place filtration in in place in place ------------------------------------ place ----------------------------------- ------------------ GW SW GW SW GW SW GW SW GW SW -------------------------------------------------------------------------------------------------------------------------------------------------------- 25-100........................................................ 50 7 1.7 0 1.7 21.7 2.6 4.3 0 0 101-500....................................................... 25 6 1.4 0 4.1 53.3 2.7 8.9 0.5 0 501-1K........................................................ 25 0 2.9 0 2.4 73.0 2.4 18.9 0 0 1K-3.3K....................................................... 27 0 1.6 0 2.7 76.4 2.7 16.4 0.4 0 3.3K-10K...................................................... 26 0 2.1 0 8.1 85.3 3.3 7.4 0.6 0 -------------------------------------------------------------------------------------------------------------------------------------------------------- References: Geometries and Characteristics of Public Water Systems, August 1999, (US EPA, 1999e) Drinking Water Baseline Handbook, February 24,1999, (US EPA, 1999m) The regulatory changes affected by the revised arsenic MCL are summarized in Table VII-3. Table VII-3.--Table Identifying Regulatory Changes ------------------------------------------------------------------------ CFR citation Topic or subpart ------------------------------------------------------------------------ Sec. 141.23(a)(4)........... Sample compositing allowed by the State. 141.23(a)(4)(i).............. Detection limit for arsenic. 141.23(a)(5)................. Frequency of monitoring for arsenic determined in Sec. 141.23(c). 141.23(c).................... Standard inorganic monitoring framework, with State waivers possible. 141.23(f)(1)................. Confirmation sampling may be required by the State. 141.23(g).................... More frequent monitoring may be required by the State. 141.23(i)(5)................. Compliance determination reporting. 141.23(k)(1)................. Approved methodology. 141.23(k)(2)................. Container, preservation, and holding time. 141.23(k)(3)(ii)............. Acceptance limit for certified laboratories. 141.62(b)(16)................ MCL for arsenic. 141.62(c).................... BATs for arsenic. 141.26(d).................... Small system compliance technologies (SSCTs). 141.154(b)................... Requires CWS to report exceedances of new MCL in CCR before lower MCL is effective, removing 25-50 µg/L informational statement requirement. Appendix A to Subpart O of Converting lower MCL compliance values 141. for CCRs and listing MCLG. Appendix B to Subpart O of Changes MCLG and MCL values effective 30 141. days after MCL is final. PN, Subpart Q, Table 1 to Add NTNCWS exceeding MCL (not a Sec. 141.203. violation) to Tier 2 reporting. Appendix A to Subpart Q of Public notification regulatory citations 141. revised. Appendix B to Subpart Q of Standard Health Effects Language 141. unchanged; revise MCLG, MCL. ------------------------------------------------------------------------ In order to prevent the arsenic MCL of 5 µg/L from becoming effective immediately, EPA is proposing to delete the reference to Sec. 141.11(a) in Sec. 141.6(c), which provides effective dates. While examining Sec. 141.6(c) for sections that affect arsenic, we found several sections that do not exist. Therefore, EPA is proposing to remove the reference to the following sections in Sec. 141.6(c) listed in Table VII-4: Table VII-4.--Table Listing Deleted Sections ------------------------------------------------------------------------ CFR section Topic or reason ------------------------------------------------------------------------ 141.11(a).............................. New arsenic MCL would be effective immediately. 141.11(e).............................. Section 141.11(e) does not exist 141.14(a)(1)........................... Section 141.14 does not exist. 141.14(b)(1)(i)........................ Section 141.14 does not exist. 141.14(b)(2)(i)........................ Section 141.14 does not exist. 141.14(d).............................. Section 141.14 does not exist. 141.24(a)(3)........................... Section 141.24(a) is reserved. ------------------------------------------------------------------------ The Agency requests comment on whether these deletions to Sec. 141.6(c) are necessary and appropriate. F. Can I use Grandfathered Data To Satisfy the Initial Monitoring Requirement? Ground water systems may use grandfathered data collected after Jan 1, 2002 to satisfy the sampling requirements for the 2002--2004 compliance period. However, the detection limit must be less than the revised MCL. If the grandfathered data is used to comply with the 2002- 2004 compliance period and the analytical result is between the current MCL and the revised MCL, then that system will be in violation of the revised MCL on the effective date of the rule. If the system chooses not to use the grandfathered data, then it must collect another sample by December 31, 2004 to demonstrate compliance with the revised MCL. [[Page 38921]] G. What Are the Monitoring Requirements for New Systems and Sources? The current regulations only address new systems and sources in the waiver provisions of Sec. 141.23(c)(4), so the proposal specifically adds monitoring requirements for these systems for inorganic, volatile organic, and synthetic organics contaminants. All new systems or systems that use a new source of water that begin operation after the effective date of this rule would have to demonstrate compliance with the MCL within a period of time specified by the State. The State would also specify sampling frequencies to ensure a system can demonstrate compliance with the MCL. This requirement would be effective for all inorganic, volatile organic, and synthetic organic contaminants regulated in Sec. 141.23 and Sec. 141.24. The Agency recognizes that many States have established requirements for new systems and new sources, and these are part of the approved State primacy programs. Therefore EPA believes that recognizing State-determined compliance will be the most effective way to regulate new systems and sources. EPA requests comment on this proposed clarification. H. How Does the Consumer Confidence Report Change? On August 19, 1998, EPA issued subpart O, the final rule requiring community water systems to provide annual reports on the quality of water delivered to their customers (63 FR 44512; US EPA, 1998e). The first Consumer Confidence Reports (CCRs) were required by October 19, 1999. The next reports are due by July 1, 2000, for calendar year 1999 data and every July 1 after that (Sec. 141.152(a)). In general, reports must include information on the health effects of contaminants only if there has been a violation of an MCL or a treatment technique. For such violations specific ``health effects language'' in subpart O must be included verbatim in the report. The arsenic health effects language is currently required when arsenic levels exceed 50 µg/L. In addition, the Agency decided to require more information for certain contaminants because of concerns raised by commenters. One of these contaminants was arsenic. As explained in the preamble to the final rule (63 FR 44512 at 44514; US EPA, 1998e) because of concerns about the adequacy of the current MCL, EPA decided that systems that detect arsenic between 0.025mg/L and the current MCL must include some information regarding arsenic (Sec. 141.154(b)). This informational statement is different from the health effects language required for an exceedance of the MCL. EPA noted that the requirement would be deleted upon promulgation of a revised MCL. Another issue which affects handling of arsenic in the CCR is the provision in the statute which authorized the Administrator to require inclusion of language describing health concerns for ``not more than three regulated contaminants'' other than those detected at levels which constitute a violation of an MCL (section 1414(c)(4)(B)(vi)). Based on stakeholder and commenter input, the Agency decided in the final CCR rule that it would use this authority in future rulemaking to require health effects language when certain MCLs are promulgated or revised. The health effects language of Subpart O would have to be included in reports of systems detecting a contaminant above the level of the new or revised MCL, prior to the effective date of the MCL, although technically the systems are not in violation of the regulations. The Agency used this authority in the promulgation of the Disinfectants and Disinfection Byproducts for one contaminant, Total Trihalomethanes on December 16, 1998 (63 FR 69390). The Agency is now proposing to use this same authority to require inclusion of the health effects language in reports of systems which detect arsenic above the level of the revised MCL upon promulgation of these regulations. The Agency believes that it is important to provide this information to customers immediately. The systems have the flexibility to place this information in context and explain to customers that there is no on- going violation. Furthermore, the health advisory EPA is planning to issue in the near future will provide consumers with information about obtaining sources with lower arsenic prior to the effective date of the 5 µg/L arsenic MCL. EPA asks for comment on whether the consumer confidence report should notify customers of arsenic health effects starting with the report issued by July 1, 2002 for calendar year 2001. After the promulgation date of the revised arsenic MCL and before the effective date, community water systems that detect arsenic above 5 µg/L but below 50 µg/L would include the arsenic health effects language. Those systems that detect arsenic above 50 µg/L would include the health effects language and also report violations as required by Sec. 141.153(d)(6). I. How Will Public Notification Change? On May 4, 2000, EPA issued the final Public Notification Rule (PNR) for Subpart Q (US EPA 2000c) to revise the minimum requirements public water systems must meet for public notification of violations of EPA's drinking water standards and other situations that pose a risk to public health from the drinking water. Water systems must begin to comply with the new PNR regulations on October 31, 2000 (if they are in jurisdictions where the program is directly implemented by EPA) or on the date a primacy State adopts the new requirements (not to exceed May 6, 2002). EPA's arsenic drinking water regulation affects public notification requirements and amends the PNR as part of its rulemaking. The PNR divides the public notice requirements into three tiers, based on the seriousness of the violation or situation. Tier 1 is for violations and situations with significant potential to have serious adverse effects on human health as a result of short-term exposure. Notice is required within 24 hours of the violation. Tier 2 is for other violations and situations with potential to have serious adverse effects on human health. Notice is required within 30 days, with extensions up to three months at the discretion of the State or primacy agency. Tier 3 is for all other violations and situations requiring a public notice not included in Tier 1 and Tier 2. Notice is required within 12 months of the violation, and may be included in the consumer confidence report at the option of the water system. Today's proposal will require community water systems (CWS) to provide a Tier 2 public notice for arsenic MCL violations and to provide a tier 3 public notice for violations of the monitoring and testing procedure requirements. Today's proposal would also require NTNCWS to provide a Tier 2 notice for exceedances of the MCL. As later explained in section XI.C., the Agency believes that overall risks from water ingested from NTNCWS cannot justify the costs of treatment. EPA believes that most States will, using their authority as described in Sec. 141.203(b), require NTNCWS to issue repeat notices on a yearly basis rather than every three months. EPA requests comment on the implementation of arsenic public notification requirements by the effective date of the arsenic MCL and on the Tier 2 public notice requirement for quarterly repeat notices for continuing exceedances of the arsenic MCL for NTNCWS. [[Page 38922]] VIII. Treatment Technologies Section 1412(b)(4)(E) of the Safe Drinking Water Act states that each NPDWR which establishes an MCL shall list the technology, treatment techniques, and other means which the Administrator finds to be feasible for purposes of meeting the MCL. Technologies are judged to be a best available technology (BAT) when the following criteria are satisfactorily met:
  • The capability of a high removal efficiency;
  • A history of full scale operation;
  • General geographic applicability;
  • Reasonable cost;
  • Reasonable service life;
  • Compatibility with other water treatment processes; and
  • The ability to bring all of the water in a system into compliance. In order to fulfill this requirement set forth by SDWA, EPA has identified BATs in Section VIII.A. Their removal efficiencies and a brief discussion of the major issues surrounding the usage of each technology are also given in section VIII.A. Likely treatment trains, of which the BAT will be the integral part, are identified in section VIII. B. The costs associated with these treatment trains are also provided. More details about the treatment technologies and costs can be found in ``Technologies and Costs for the Removal of Arsenic From Drinking Water'' (US EPA,1999i). Section 1412(b)(4)(E)(ii) of the Act also states that EPA shall list any affordable small systems compliance technologies that are feasible for the purposes of meeting the MCL. The general process by which EPA identifies compliance, and if necessary, variance technologies is described in section VIII.C. The Agency, for the revised arsenic regulation, is not proposing any variance technologies. Compliance technologies for arsenic are identified in section VIII.E. More details about the technologies and affordability determinations can be found in ``Compliance Technologies for Arsenic'' (US EPA,1999g). Section VIII.F briefly discusses how other rules, presently being developed by the Agency, may impact the arsenic rule, or how the arsenic rule may impact these other regulations. A. What Are the Best Available Technologies (BATs) for Arsenic? What Are the Issues Associated With These Technologies? EPA reviewed several technologies as BAT candidates for arsenic removal: ion exchange, activated alumina, reverse osmosis, nanofiltration, electrodialysis reversal, coagulation assisted microfiltration, modified coagulation/filtration, modified lime softening, greensand filtration, conventional iron and manganese removal, and several emerging technologies. The Agency proposes that, of the technologies capable of removing arsenic from source water, only the technologies in Table VIII-1 fulfill the requirements of the SDWA for BAT determinations for arsenic. The maximum percent removal that can be reasonably obtained from these technologies is also shown in the table. These removal efficiencies are for arsenic (V) removal. Table VIII-1.--Best Available Technologies and Removal Rates ------------------------------------------------------------------------ Maximum Treatment technology percent removal 1 ------------------------------------------------------------------------ Ion Exchange................................................. 95 Activated Alumina............................................ 90 Reverse Osmosis.............................................. >95 Modified Coagulation/Filtration.............................. 95 Modified Lime Softening...................................... 80 Electrodialysis Reversal..................................... 85 ------------------------------------------------------------------------ \1\ The percent removal figures are for arsenic (V) removal. In water, the most common valence states of arsenic are As (V), or arsenate, and As (III), or arsenite. As (V) is more prevalent in aerobic surface waters and As (III) is more likely to occur in anaerobic ground waters. In the pH range of 4 to 10, As (V) species (H2AsO4-and H2AsO42-) are negatively charged, and the predominant As (III) compound (H3AsO3) is neutral in charge. Removal efficiencies for As (V) are much better than removal of As (III) by any of the technologies evaluated, because the arsenate species carry a negative charge and arsenite is neutral under these pH conditions. To increase the removal efficiency when As (III) is present, pre-oxidation to the As (V) species is necessary. Pre-oxidation. As (III) may be converted through pre-oxidation to As (V) using one of several oxidants. Data on oxidants indicate that chlorine, potassium permanganate, and ozone are effective in oxidizing As (III) to As (V). Pre-oxidation with chlorine may create undesirable concentrations of disinfection by-products and membrane fouling of subsequent treatments such as reverse osmosis. EPA has completed research on the chemical oxidants for As (III) conversion, and is presently investigating ultraviolet light disinfection technology (UV) and solid oxidizing media. For point-of-use and point-of-entry (POU/ POE) devices, central chlorination may be required for oxidation of As (III). Coagulation/Filtration (C/F) is an effective treatment process for removal of As (V) according to laboratory and pilot-plant tests. The type of coagulant and dosage used affects the efficiency of the process. Within either high or low pH ranges, the efficiency of C/F is significantly reduced. Below a pH of approximately 7, removals with alum or ferric sulfate/chloride are similar. Above a pH of 7, removals with alum decrease dramatically (at a pH of 7.8, alum removal efficiency is about 40%). Other coagulants are also less effective than ferric sulfate/chloride. Disposal of the arsenic-contaminated coagulation sludge may be a concern especially if nearby landfills are unwilling to accept such a sludge. Lime Softening (LS), operated within the optimum pH range of greater than 10.5 is likely to provide a high percentage of As removal. However, if removals greater than 80% are required, it may be difficult to remove consistently at that level by LS alone. Systems using LS may require secondary treatment to meet that goal (e.g., addition of an ion exchange unit as a polishing step). As with C/F, disposal of arsenic- contaminated sludge from LS may be an issue. Coagulation/Filtration and Lime Softening are technologies primarily used for large systems. Package plants may make it more affordable for small systems to employ these technologies. Package plants are pre-engineered (i.e., the process engineering for the package plants has been done by the manufacturer). What remains for the water system's engineer to design is the specifics of the on-site application of the equipment. However, these technologies still require well trained operators. If it is not possible to keep a trained operator at the plant, an off-site contract operator may be able to monitor the process with a telemetry device. Because of these complexities, these technologies are not likely to be installed solely for arsenic removal. However, if they are already in place, modification of these two technologies to achieve higher arsenic removal efficiencies is a viable option. Activated Alumina (AA) is effective in treating water with high total dissolved solids (TDS). However, the capacity of activated alumina to remove arsenic is very pH sensitive. High removals can be achieved at high pHs, but at shorter run lengths. The use of chemicals for pH adjustment and bed regeneration, storage of sulfuric acid and sodium hydroxide, and process oversight increase operator responsibilities and the need for advanced training. (Decisions on the certification of water operators will be [[Page 38923]] made at the State and local levels). Operators may have to add an acid to lower pH to an optimal range and then afterwards increase the pH to avoid corrosion. Sodium hydroxide and sulfuric acid are required in the regeneration process. Selenium, fluoride, chloride, sulfate, and silica, if present at high levels, may compete for adsorption sites. Suspended solids and precipitated iron can cause clogging of the AA bed. Systems containing high levels of these constituents may require pretreatment or periodic backwashing. AA is highly selective towards As (V), and this strong attraction results in regeneration problems, possibly resulting in 5 to 10 percent loss of adsorptive capacity after each run. As a result, AA may not be efficient in the long term. In addition, activated alumina produces highly concentrated waste streams, which can contain approximately 30,000 mg/L of total dissolved solids (TDS) content. Because of the high content of TDS in the waste stream, disposal of the brine must be taken into consideration. The safety issue of handling corrosive and caustic chemicals associated with this technology may make it inappropriate for small systems. Therefore, in estimating national costs, it was assumed that small systems would not adjust pH and would not regenerate on site. Costs were estimated assuming systems operated a non-optimal pH and operation on a ``throw-away'' basis. Regenerating the media off-site instead of disposing of spent media is another possibility. Ion Exchange (IX) can effectively remove arsenic as well. It is recommended as a BAT primarily for small, ground water systems with low sulfate and TDS, and as a polishing step after filtration. Sulfate, TDS, selenium, fluoride, and nitrate compete with arsenic for binding sites and can affect run length. Column bed regeneration frequency is a key factor in calculating costs. Recent research indicates that ion exchange may be practical up to approximately 120 mg/L of sulfate (Clifford 1994). Passage through a series of columns could improve removal and decrease regeneration frequency. As with AA, suspended solids and precipitated iron can cause clogging of the IX bed. Systems containing high levels of these constituents may require pretreatment. Suspended solids and precipitated iron may also be removed by backwashing. Ion exchange also produces a highly concentrated waste by-product stream, and the disposal of this brine must be considered. Brine recycling can reduce the amount of waste for disposal and lower the cost of operation. Recent research showed that the brine regeneration solution could be reused as many as 20 times with no impact on arsenic removal provided that some salt was added to the solution to provide adequate chloride levels for regeneration (Clifford 1998). Reverse Osmosis (RO) can provide removal efficiencies of greater than 95 percent when operating pressure is ideal (e.g., pounds per square inch, psi). Water rejection (on the order of 20-25%) may be an issue in water-scarce regions. If RO is used by small systems in the western U. S., water recovery will likely need to be optimized due to the scarcity of water resources. Water recovery is the volume of water produced by the process divided by the influent stream (product water/ influent stream). Increased water recovery can lead to increased costs for arsenic removal. Since the ability to blend with an MCL of 5 µg/L would be limited, the entire stream may have to be treated. Therefore, most of the alkalinity and hardness would also be removed. In that case, to avoid corrosion problems and to restore minerals to the water, post-treatment corrosion control may be necessary. Discharge of reject water or brine may also be a concern. Electrodialysis Reversal (EDR) can produce effluent water quality comparable to reverse osmosis. EDR systems are fully automated, require little operator attention, and do not require chemical addition. EDR systems, however, are typically more expensive than nanofiltration and reverse osmosis systems. These systems are often used in treating brackish water to make it suitable for drinking. This technology has also been applied in the industry for wastewater recovery. The technology typically operates at a recovery of 70 to 80 percent. Few studies have been conducted to exclusively evaluate this process for the removal of arsenic, but a removal of approximately 85% can be expected (US EPA, 1999i). Other Technologies Coagulation Assisted Microfiltration. The coagulation process described previously can be linked with microfiltration to remove arsenic. The microfiltration step essentially takes the place of a conventional gravity filter. The University of Houston recently completed pilot studies at Albuquerque, New Mexico on iron coagulation followed by a direct microfiltration system. The results of this study indicated that iron coagulation followed by microfiltration is capable of removing arsenic (V) from water to yield concentrations which are consistently below 2 µg/L. Critical operating parameters are iron dose, mixing energy, detention time, and pH (Clifford, 1997). However, since a full-scale operation history is one of the requirements to list a technology as a BAT, it is not presently being listed as one. It could be designated as such in the future if the technology meets that requirement. Oxidation/Filtration (including greensand filtration) has an advantage in that there is not as much competition with other ions. However, the process has not been used very much for arsenic removal. In addition, similar to activated alumina, greensand filtration may require pH adjustment to optimize removal, which may be difficult for small systems. This technology is not recommended for high removals. The maximum removal percentage was assumed to be 50% when estimating national costs. The presence of iron in the source water is critical for arsenic removal. If the source water does not contain iron, oxidizing and filtering the water will not remove arsenic. In developing national cost estimates, it was assumed that systems would opt for this type of technology only if more than 300 µg/L of iron was present. Oxidation/Filtration is not being listed as a BAT because it does not meet the requirement of a high removal efficiency. However, since it is a relatively inexpensive technology, it may be appropriate for those systems that do not require much arsenic removal and have high iron in their source water. Emerging Technologies There are several emerging technologies for arsenic removal; however, these require more testing before they can be designated as a BAT. Iron-based media products include the following. Iron oxide coated sand removes arsenic using adsorption; the sand also doubles as a filtration media. The technology has only been tested at the bench- scale level and may have a high cost associated with it. Granular ferric hydroxide also employs an adsorption process and is being used in a number of full scale plants in Germany. Costs may be an issue with this technology as well. Iron filings are essentially a filter technology, initially developed for arsenic remediation. Though quite effective at remediation, this technology may have limited use as a drinking water treatment technology; the technology performs well when treating high influent arsenic levels typical of remediation, but needs to be proven in treating lower influent levels expected in raw drinking water to a finished level at the proposed MCL. [[Page 38924]] Sulfur-modified iron appears to remove total organic carbon (TOC) and disinfection byproducts (DBPs) as well as arsenic. However, it has only been tested at the bench scale. ADI Group, Inc.''s proprietary process also has an iron-based media that has been installed in a number of locations. Nanofiltration is of interest because it can be operated at lower pressures than reverse osmosis, which translate into lower operation and maintenance costs. However, when nanofiltration is operated at realistic recoveries, the removal efficiency appears to be low. Electrodialysis Reversal (EDR), although easier to operate than reverse osmosis and nanofiltration, does not appear to be competitive with respect to costs and process efficiency. Waste Disposal Waste disposal will be an important issue for both large and small plants. Costs for waste disposal have been added to the costs of the treatment technologies (in addition to any pre-oxidation and corrosion control costs), and form part of the treatment trains that are listed in Section VIII.B. A sufficient volume of receiving water would be needed in order to directly discharge the contaminated brine stream from membrane technologies. Otherwise, operators may have to pre-treat to meet Clean Water Act permit requirements prior to discharge. If the plant is discharging to a sanitary sewer because of the membranes, there may be a very high salinity in the discharge as well as high levels of arsenic that might, without pretreatment, exceed local sewer use regulations. Ion exchange and activated alumina treatment brines will be even more concentrated (on the order of 30,000 TDS), and more than likely will require pre-treatment prior to discharge to either a receiving body of water or the sanitary sewer. Disposal of solid treatment residuals would be problematic if they fail the toxicity characteristic (TC) of the Resource Conservation and Recovery Act (RCRA). If they fail the TC, the residuals are regulated as hazardous waste because of the concentration of arsenic. For the purposes of the national cost estimate, it was assumed that solid residuals would be disposed of at nonhazardous landfills. B. What Are the Likely Treatment Trains? How Much Will They Cost? Likely treatment trains are shown in Table VIII-2. These trains represent a wide variety of solutions a facility may consider when complying with the proposed arsenic MCL. Not all solutions may be viable for a given system. For example, only those systems with coagulation/filtration in-place will be able to modify their existing treatment system. The treatment trains include BATs, waste disposal, and when necessary, pre-oxidation and corrosion control. Table VIII-2 also contains two ``non-treatment'' options which may be appropriate if the source water is of very poor quality. ``Regionalization'' refers to connecting with another system and purchasing water, and ``alternate source'' refers to finding a new source of water (e.g. drilling a new well). However, since arsenic is a naturally occurring contaminant, it may be ubiquitous at a particular site, so drilling another well may not improve the situation. Table VIII-2.--Treatment Technology Trains ------------------------------------------------------------------------ Train No. Treatment technology trains ------------------------------------------------------------------------ 1............................ Regionalization. 2............................ Alternate Source. 3............................ Add pre-oxidation [if not in-place] and modify in-place Lime Softening. 4............................ Add pre-oxidation [if not in-place] and modify in-place Coagulation/Filtration. 5............................ Add pre-oxidation [if not in-pace] and add Anion Exchange and add POTW waste disposal and add corrosion control [if >90% removal required]. Sulfate level at 25 mg/l. 6............................ Add pre-oxidation [if not in-place] and add Anion Exchange and add POTW waste disposal and add corrosion control [if >90% removal required]. Sulfate level at 150 mg/l. 7............................ Add pre-oxidation [if not in-place] and add Anion Exchange and add evaporation pond/non-hazardous landfill waste disposal and add corrosion control [if >90% removal required]. Sulfate level at 25 mg/l. 8............................ Add pre-oxidation [if not in-place] and add Anion Exchange and evaporation pond/ non-hazardous landfill waste disposal and add corrosion control [if >90% removal required]. Sulfate level at 150 mg/l. 9............................ Add pre-oxidation [if not in-place] and add Activated Alumina and add non- hazardous landfill (for spent media) waste disposal. pH at 7. 10........................... Add pre-oxidation [if not in-pace] and add Reverse Osmosis and add direct discharge waste disposal and add corrosion control [if >90% removal required]. 11........................... Add pre-oxidation [if not in-place] and add Reverse Osmosis and add POTW waste disposal and add corrosion control [if >90% removal required]. 12........................... Add pre-oxidation [if not in-place] and add Reverse Osmosis and add chemical precipitation/non-hazardous landfill and add corrosion control [if >90% removal required]. 13........................... Add pre-oxidation [if not in-place] and add Coagulation Assisted Microfiltration and add mechanical dewatering/non- hazardous landfill waste disposal. 14........................... Add pre-oxidation [if not in-place] and add Coagulation Assisted Microfiltration and add non-mechanical dewatering/non- hazardous landfill waste disposal. 15........................... Add pre-oxidation [if not in-place] and add Oxidation/Filtration (Greensand) and add POTW for backwash stream. 16........................... Add pre-oxidation [if not in-place] and add Anion Exchange and add chemical precipitation/non-hazardous landfill waste disposal and add corrosion control [if >90% removal required]. Sulfate level at 25 mg/l. 17........................... Add pre-oxidation [if not in-place] and add Anion Exchange and add chemical precipitation/non-hazardous landfill waste disposal and add corrosion control [if >90% removal required]. Sulfate level at 150 mg/l. 18........................... Add pre-oxidation [if not in-place] and add Activated Alumina and add POTW/non- hazardous landfill waste disposal. pH at 7. 19........................... Add pre-oxidation [if not in-place] and add POE Activated Alumina. 20........................... Add pre-oxidation [if not in-place] and add POU Reverse Osmosis. 21........................... Add pre-oxidation [if not in-place] and add POU Activated Alumina. ------------------------------------------------------------------------ [[Page 38925]] Costs for each of these treatment trains are given in Table VIII-3. These costs are a function of system size. Some individual systems may experience household costs higher than those estimated in this table. The pre-oxidation costs and corrosion control costs are given separately for each system size category because they will only be incurred by some of the systems. In estimating national costs, it was assumed that only systems without pre-oxidation in-place would add the necessary equipment. It is expected that no surface water systems will need to install pre-oxidation for arsenic removal. Based on Table IX-4, it is expected that fewer than 50% of the ground water systems may need to install pre-oxidation for arsenic removal. Ground water systems without pre-oxidation should determine if pre-oxidation is necessary by determining if the arsenic is present as As (III) or As (V). Groundwater systems with predominantly As (V) will probably not need pre-oxidation to meet the MCL. Similarly, costs for corrosion control were only added to systems that used ion exchange or reverse osmosis to remove more than 90% of the arsenic in the raw water. It is expected that fewer than 1% of the affected systems will need to install corrosion control due to installation of arsenic treatment. For ion exchange, different treatment trains were used for two levels of sulfate. As sulfate affects regeneration frequency, the high sulfate treatment train is more expensive than the low sulfate treatment train. Table VIII-3.--Annual Costs of Treatment Trains (Per Household)* -------------------------------------------------------------------------------------------------------------------------------------------------------- Size --------------------------------------------------------------------------------------- Treatment train 1001-3300 25-100 101-500 501-1000 3301-10K 10K-50K 50K-100K 100K-1M (dollars) (dollars) (dollars) (dollars) (dollars) (dollars) (dollars) (dollars) -------------------------------------------------------------------------------------------------------------------------------------------------------- 1............................................................... $ 1347 $ 202 $ 77 $ 25 $ 8 $ 2 $ 1 $ 0 2............................................................... 96 14 5 2 1 0 0 0 3............................................................... 750 138 70 40 30 26 22 18 4............................................................... 462 82 40 22 49 60 38 18 5............................................................... 519 146 90 106 73 55 44 39 6............................................................... 883 248 160 160 78 60 49 44 7............................................................... 629 226 153 154 108 84 71 58 8............................................................... 1227 469 333 290 197 165 135 88 9............................................................... 384 227 201 182 168 152 144 143 10.............................................................. 2136 800 555 429 300 256 225 206 11.............................................................. 2136 800 555 429 300 256 225 206 12.............................................................. 2819 892 572 409 293 237 204 186 13.............................................................. 1282 293 195 125 72 50 32 18 14.............................................................. 1218 281 187 117 80 54 35 21 15.............................................................. 558 156 102 72 55 42 37 31 16.............................................................. 1008 222 121 128 86 58 46 40 17.............................................................. 1050 246 115 114 96 66 52 45 18.............................................................. 427 243 212 192 177 161 153 152 19.............................................................. 467 427 408 388 367 342 327 298 20.............................................................. 325 289 272 254 236 214 202 178 21.............................................................. 377 334 314 292 271 245 230 202 pre-ox**........................................................ 416 66 26 9 4 2 1 1 corros**........................................................ 63 17 11 6 5 3 3 3 -------------------------------------------------------------------------------------------------------------------------------------------------------- *These costs are based on a discount rate of 7%. **The costs for treatment trains 1-21 do not include pre-oxidation or corrosion control costs. For systems that need to add pre-oxidation or corrosion control, the costs for these additional treatments should be added to those of the trains shown in the table. C. How Are Variance and Compliance Technologies Identified for Small Systems? Section 1415(e)(1) of SDWA allows States to grant variances to small water systems (i.e., systems having fewer than 10,000 customers) in lieu of complying with an MCL if EPA determines that there are no nationally affordable compliance technologies for that system size/ water quality combination. The system must then install an EPA-listed variance treatment technology (section 1412(b)(15)) that makes progress toward the MCL, if not necessarily reaching it. To list variance technologies, three showings must be made: (1) EPA must determine, on a national level, that there are no compliance technologies that are affordable for the given small system size category/source water quality combination. (2) If there is no nationally affordable compliance technology, then EPA must identify a variance technology that may not reach the MCL but that will allow small systems to make progress toward the MCL (it must achieve the maximum reduction affordable). This technology must be listed as a small systems variance technology by EPA in order for small systems to be able to rely on it for regulatory purposes. (3) EPA must make a finding on a national level, that use of the variance technology would be protective of public health and establish. Primacy States must then make a site-specific determination for each system as to whether or not the system can afford to meet the MCL based on State-developed affordability criteria. If the State determines that compliance is not affordable for the system, it may grant a variance, but it must establish terms and conditions, as necessary, to ensure that the variance is adequately protective of human health. In the Agency's draft national-level affordability criteria published in the August 6, 1998 Federal Register (US EPA, 1998h), EPA discussed the affordable treatment technology determinations for the contaminants regulated before 1996. The national-level affordability criteria were derived as follows. First an ``affordability threshold'' (i.e., the total annual household water bill that would be considered affordable) was calculated. In developing this threshold value, EPA considered the percentage of median [[Page 38926]] household income spent by an average household on comparable goods and services such items as housing (28%), transportation (16%), food (12%), energy and fuels (3.3%), telephone (1.9%), water and other public services (0.7%), entertainment (4.4%) and alcohol and tobacco (1.5%). Another of the key factors that EPA used to select an affordability threshold was cost comparisons with other risk reduction activities for drinking water. Section 1412(b)(4)(E)(ii) of the SDWA identifies both Point-of-Entry and Point-of-Use devices as options for compliance technologies. EPA examined the projected costs of these options. EPA also investigated the costs associated with supplying bottled water for drinking and cooking purposes. The median income percentages that were associated with these risk reduction activities were: Point-of-Entry (>2.5%), Point-of-Use (2%) and bottled water (>2.5%). The complete rationale for EPA's selection of 2.5% as the affordability threshold is described in Variance Technology Findings for Contaminants Regulated Before 1996 (US EPA, 1998f). Based on the foregoing analysis, EPA developed an affordability criteria of 2.5% of median household income, or about $750, for the affordability threshold (US EPA 1998f). The median water bill for households in each small system category was then subtracted from this threshold to determine the affordable level of household expenditures for new treatment. This difference is referred to as the ``available expenditure margin.'' Based on EPA's 1995 Community Water System Survey, median water bills were about $250 per year for small system customers. Thus, an average available expenditure margin of up to $500 per year was considered affordable for the contaminants regulated before 1996. However, EPA expects the available expenditure margin may be lower than $500 per household per year for the arsenic rule because EPA believes that water rates are currently increasing faster than median household income. Thus, the ``baseline'' for annual water bills will rise as treatment is installed for compliance with regulations promulgated after 1996, but before the arsenic rule is promulgated. To account for this, EPA intends to adjust its calculation of the baseline for the affordability criteria as follows. The national median annual household water bills for each size category will be adjusted by averaging the total national costs for the size category over all of the systems within the size category. In other words, the costs incurred by these rules at the affected water systems will be averaged over all of the systems in that size category. A revised available expenditure margin will be calculated by subtracting the new baseline from the affordability threshold. The affordable technology determinations will be made by comparing the projected costs of treatment against the lower available expenditure margin. If the projected costs of all treatment technologies for a given system size/ source water quality exceed the revised available expenditure margin, then variance technologies could be considered for those systems. EPA requests comment on this method of accounting for new regulations in its affordability criteria. Applying the affordability criterion to the case of arsenic in drinking water, EPA has determined that affordable technologies exist for all system size categories and has therefore not identified a variance technology for any system size or source water combination at the proposed MCL. (See Table IX-12, Total Annual Costs per Household.) In other words, annual household costs are projected to be below the available affordability threshold for all system size categories for the proposed MCL. EPA solicits comment on its determination in this case as well as its affordability criteria more generally. EPA recognizes that individual water systems may have higher than average treatment costs, fewer than average households to absorb these costs, or lower than average incomes, but believes that the affordability criteria should be based on characteristics of typical systems and should not address situations where costs might be extremely high or low or excessively burdensome. EPA believes that there are other mechanisms that may address these situations to a certain extent, such as rates for disadvantaged communities and grants. For instance, many utilities extend special ``lifeline'' rates to disadvantaged communities. EPA also notes that high water costs are often associated with systems that have already installed treatment to comply with a NPDWR. Such treatment facilities may also facilitate compliance with future standards. EPA's approach to establishing the national-level affordability criteria did not incorporate a baseline for in-place treatment technology. Assuming that systems with high baseline water costs would need to install a new treatment technology to comply with a NPDWR may thus overestimate the actual costs for some systems. To investigate this issue, EPA examined a group of five small surface water systems with annual water bills above $500 per household per year during the derivation of the national-level affordability criteria. All of these systems had installed disinfection and filtration technologies to comply with the surface water treatment rule. If these systems exceeded the revised arsenic standard, modification of the existing processes would be much more cost- effective than adding a new technology to comply with the arsenic rule. These systems have already made the investment in treatment technology and that is reflected in the current annual household water bills. In addition, systems that meet criteria established by the State could be classified as disadvantaged communities under section 1452(d) of the SDWA. They can receive additional subsidization under the Drinking Water State Revolving Fund (DWSRF) program, including forgiveness of principal. Under DWSRF, States must provide a minimum of 15% of the available funds for loans to small communities and have the option of providing up to 30% of the grant to provide additional loan subsidies to the disadvantaged systems, as defined by the State. As previously noted in today's proposal, some technologies can interfere with treatment in-place or require additional treatment to address side effects which will increase costs over the arsenic treatment technology base costs. (An example is corrosion control for lead and copper, which may need to be adjusted to accommodate other treatment). While EPA tries to account for such interferences in its cost estimates for each new compliance technology, it is not possible to anticipate all the site specific issues which may arise. However, EPA has included a discussion of the co-occurrence of radon, sulfate, and iron in this proposal. EPA will also provide guidance identifying cost-effective treatment trains for ground water systems that need to treat for both arsenic and radon after the arsenic rule is finalized. EPA encourages small systems to discuss their infrastructure needs for complying with the arsenic rule with their primacy agency to determine their eligibility for DWSRF loans, and if eligible, to ask for assistance in applying for the loans. D. When Are Exemptions Available? Under section 1416(a), the State may exempt a public water system from any MCL and/or treatment technique requirement if it finds that (1) due to compelling factors (which may include economic factors), the system is unable [[Page 38927]] to comply or develop an alternative supply, (2) the system was in operation on the effective date of the MCL or treatment technique requirement, or, for a newer system, that no reasonable alternative source of drinking water is available to that system, (3) the exemption will not result in an unreasonable risk to health, and (4) management or restructuring changes cannot be made that would result in compliance with this rule. Under section 1416(b), at the same time it grants an exemption the State is to prescribe a compliance schedule and a schedule for implementation of any required control measures. The final date for compliance may not exceed three years after the NPDWR effective date except that the exemption can be renewed for small systems for limited time periods. E. What Are the Small Systems Compliance Technologies? Section 1412(b)(4)(E)(ii) of SDWA, as amended in 1996, requires EPA to issue a list of technologies that achieve compliance with MCLs established under the Act that are affordable and applicable to typical small drinking water systems. These small public water systems categories are: (1) Population of more than 25 but less than 500; (2) Population of more than 500, but less than 3,300; and (3) Population of more than 3,300, but less than 10,000. Owners and operators may choose any technology or technique that best suits their conditions, as long as the MCL is met. Of the treatment trains identified in section VIII.B., the ones identified in Table VIII-4 are deemed to be affordable for systems serving 25-500 people and the ones identified in Table VIII-5 are deemed to be affordable for systems serving 501-3,300 and 3,301-10,000 people, as their annual costs are below the affordability threshold (US EPA, 1999g). Because affordable compliance technologies are available, the Agency does not propose to identify any variance technologies. EPA requests comments on the affordable compliance technology determinations for the three size categories and the determination that there will be no variance technologies. Centralized compliance treatment technologies include ion exchange, activated alumina, modified coagulation/filtration, modified lime softening, and oxidation/filtration (e.g. greensand filtration) for source waters high in iron. In addition, point-of-use (POU) and point-of-entry (POE) devices are also compliance technology options for the smaller systems. EPA is aware that very few water systems have had experience with centrally managed POU or POE options in the past. EPA requests comments on implementation issues associated with a centrally managed POU or POE option for arsenic. The non-treatment alternatives are especially relevant for small systems. EPA is proposing to add the abbreviations ``POU'' and ``POE'' to the definitions in Sec. 141.2 and asks for comment on the utility of adding them. Table VIII-4.--Affordable Compliance Technology Trains for Small Systems With population 25-500 ------------------------------------------------------------------------ Train No. Treatment technology trains ------------------------------------------------------------------------ 3................................. Add pre-oxidation [if not in-place] and modify in-place Lime Softening 4................................. Add pre-oxidation [if not in-place] and modify in-place Coagulation/ Filtration 5................................. Add pre-oxidation [if not in-place] and add Anion Exchange and add POTW waste disposal and add corrosion control [if >90% removal required]. Sulfate level at 25 mg/l. 6................................. Add pre-oxidation [if not in-place] and add Anion Exchange and add POTW waste disposal and add corrosion control [if >90% removal required]. Sulfate level at 150 mg/l. 7................................. Add pre-oxidation [if not in-place] and add Anion Exchange and add evaporation pond/non-hazardous landfill waste disposal and add corrosion control [if >90% removal required]. Sulfate level at 25 mg/ l. 8................................. Add pre-oxidation [if not in-place] and add Anion Exchange and evaporation pond/non-hazardous landfill waste disposal and add corrosion control [if >90% removal required]. Sulfate level at 150 mg/ l. 9................................. Add pre-oxidation [if not in-place] and add Activated Alumina and add non-hazardous landfill (for spent media) waste disposal. pH at 7. 15................................ Add pre-oxidation [if not in-place] and add Oxidation/Filtration (Greensand) and add POTW for backwash stream. 16................................ Add pre-oxidation [if not in-place] and add Anion Exchange and add chemical precipitation/non- hazardous landfill waste disposal and add corrosion control [if >90% removal required]. Sulfate level at 25 mg/l. 17................................ Add pre-oxidation [if not in-place] and add Anion Exchange and add chemical precipitation/non- hazardous landfill waste disposal and add corrosion control [if >90% removal required]. Sulfate level at 150 mg/l. 18................................ Add pre-oxidation [if not in-place] and add Activated Alumina and add POTW/non-hazardous landfill waste disposal. pH at 7. 19................................ Add pre-oxidation [if not in-place] and add POE Activated Alumina. 20................................ Add pre-oxidation [if not in-place] and add POU Reverse Osmosis. 21................................ Add pre-oxidation [if not in-place] and add POU Activated Alumina. ------------------------------------------------------------------------ Table VIII-5.--Affordable Compliance Technology Trains for Small Systems With populations 501-3,300 and 3,301 to 10,000 ------------------------------------------------------------------------ Train No. Treatment technology trains ------------------------------------------------------------------------ 3................................. Add pre-oxidation [if not in-place] and modify in-place Lime Softening 4................................. Add pre-oxidation [if not in-place] and modify in-place Coagulation/ Filtration 5................................. Add pre-oxidation [if not in-place] and add Anion Exchange and add POTW waste disposal and add corrosion control [if >90% removal required]. Sulfate level at 25 mg/l. 6................................. Add pre-oxidation [if not in-place] and add Anion Exchange and add POTW waste disposal and add corrosion control [if >90% removal required]. Sulfate level at 150 mg/l. 7................................. Add pre-oxidation [if not in-place] and add Anion Exchange and add evaporation pond/non-hazardous landfill waste disposal and add corrosion control [if >90% removal required]. Sulfate level at 25 mg/ l. 8................................. Add pre-oxidation [if not in-place] and add Anion Exchange and evaporation pond/non-hazardous landfill waste disposal and add corrosion control [if >90% removal required]. Sulfate level at 150 mg/ l. [[Page 38928]] 9................................. Add pre-oxidation [if not in-place] and add Activated Alumina and add non-hazardous landfill (for spent media) waste disposal. pH at 7. 10................................ Add pre-oxidation [if not in-place] and add Reverse Osmosis and add direct discharge waste disposal and add corrosion control [if >90% removal required]. 11................................ Add pre-oxidation [if not in-place] and add Reverse Osmosis and add POTW waste disposal and add corrosion control [if >90% removal required]. 12................................ Add pre-oxidation [if not in-place] and add Reverse Osmosis and add chemical precipitation/non- hazardous landfill and add corrosion control [if >90% removal required]. 13................................ Add pre-oxidation [if not in-place] and add Coagulation Assisted Microfiltration and add mechanical dewatering/non-hazardous landfill waste disposal. 14................................ Add pre-oxidation [if not in-place] and add Coagulation Assisted Microfiltration and add non- mechanical dewatering/non-hazardous landfill waste disposal. 15................................ Add pre-oxidation [if not in-place] and add Oxidation/Filtration (Greensand) and add POTW for backwash stream. 16................................ Add pre-oxidation [if not in-place] and add Anion Exchange and add chemical precipitation/non- hazardous landfill waste disposal and add corrosion control [if >90% removal required]. Sulfate level at 25 mg/l. 17................................ Add pre-oxidation [if not in-place] and add Anion Exchange and add chemical precipitation/non- hazardous landfill waste disposal and add corrosion control [if >90% removal required]. Sulfate level at 150 mg/l. 18................................ Add pre-oxidation [if not in-place] and add Activated Alumina and add POTW/non-hazardous landfill waste disposal. pH at 7. 19................................ Add pre-oxidation [if not in-place] and add POE Activated Alumina. 20................................ Add pre-oxidation [if not in-place] and add POU Reverse Osmosis. 21................................ Add pre-oxidation [if not in-place] and add POU Activated Alumina. ------------------------------------------------------------------------ Centralized treatment is not always a feasible option. When this is the situation, home water treatment devices can be effective and affordable compliance options for small systems in meeting the proposed arsenic MCL. Home water treatment can consist of either whole-house (point-of-entry) or single faucet (point-of-use) treatment. Whole-house, or POE treatment, is necessary when exposure to the contaminant by modes other than consumption is a concern; this is not the case with arsenic. Single faucet, or POU treatment, is preferred when treated water is needed only for drinking and cooking purposes. POU devices are especially applicable for systems that have a large flow and only a minor part of that flow directed for potable use. POE/ POU options include reverse osmosis, activated alumina, and ion exchange processes. POU systems are easily installed and can be easily operated and maintained. In addition, these systems generally offer lower capital costs and may reduce engineering, legal, and other fees associated with centralized treatment options. Allowing the usage of POU devices is one of the new elements of the Safe Drinking Water Act; on June 11, 1998, EPA issued a Federal Register notice (US EPA, 1998i) to withdraw the prohibition on the use of POU devices as compliance technologies. The SDWA stipulates that POU/POE treatment systems ``shall be owned, controlled and maintained by the public water system, or by a person under contract with the public water system to ensure proper operation and compliance with the MCL or treatment technique and equipped with mechanical warnings to ensure that customers are automatically notified of operational problems.'' Using POU/POE devices introduces some new issues. Adopting a POU/ POE treatment system in a small community requires more record-keeping to monitor individual devices than does central treatment. POU/POE systems require special regulations regarding customer responsibilities and water utility responsibilities. Use of POU/POE systems does not reduce the need for a well-maintained water distribution system. On the contrary, increased monitoring may be necessary to ensure that the treatment units are operating properly. Water systems with high influent arsenic concentrations (i.e., greater than 1 mg/L) may have difficulty meeting the proposed MCL when POU/POE devices are used. As a result, influent arsenic concentration and other source water characteristics must be considered when evaluating POU/POE devices for arsenic removal. EPA assumed that systems would more likely opt to use POU AA or RO (and not IX), and POE AA (and not IX nor RO), when developing national cost estimates (refer to Table VIII-4). Activated alumina and ion exchange units face a breakthrough issue. If the media or resin is not replaced and/or regenerated on time, there is a potential for significantly reduced arsenic removal. Activated alumina units have the advantage of longer run lengths and the option to use the media once and throw it away. However, if POE ion exchange units are regenerated on time, they would also be an effective treatment technology. Units with automatic regeneration are thus viable options. POE IX and RO units also have a potential for creating corrosion control problems. With ion exchange POE units, a reduction in pH can be expected initially with new resin, but the pH reduction should subside over time. F. How Does the Arsenic Regulation Overlap With Other Regulations? Several Federal rules are under development regarding treatment requirements that may relate to the treatment of arsenic for this drinking water rule. The following briefly describes each rule, the impact the Arsenic Rule may have on that rule, and/or how each rule may impact the arsenic standard. The Arsenic Rule is expected to be promulgated in a similar time frame as the Ground Water Rule, the Radon Rule, and the Microbial and Disinfection By-Product Rule (Final December, 1998). In addition, the disposal of residuals may be affected by the hazardous waste regulations of the Resource Conservation and Recovery Act (RCRA). Ground Water Rule (GWR). The goals of the GWR are to: (1) Provide a consistent level of public health protection; (2) prevent waterborne [[Page 38929]] microbial disease outbreaks; (3) reduce endemic waterborne disease; and (4) prevent fecal contamination from reaching consumers. EPA has the responsibility to develop a ground water rule which not only specifies the appropriate use of disinfection, but also addresses other components of ground water systems to assure public health protection. This general provision is supplemented with an additional requirement that EPA develop regulations specifying the use of disinfectants for ground water systems as necessary. To meet these requirements, EPA worked with stakeholders to develop a Ground Water Rule proposal (US EPA, 2000d) and plans to issue a final rule by late Fall 2000. The GWR will result in more systems using disinfection. Under the GWR, a system has options other than disinfection (e.g., protecting source water). However, if a system does add a disinfection technology, it may contribute to arsenic pre-oxidation. This largely depends on the type of disinfection technology employed. If a system chooses a technology such as ultraviolet radiation, it may not affect arsenic pre-oxidation. However, if it chooses chlorination, it will contribute to arsenic pre-oxidation. As discussed previously, arsenic pre- oxidation from As (III) to As (V) will enhance the removal efficiencies of the technologies. In addition, systems may use membrane filtration for the GWR. In that case, depending on the size of the membrane, some arsenic removal can be achieved. Thus, the GWR is expected to alleviate some of the burden of the Arsenic Rule. Radon. In the 1996 Amendments to the SDWA, Congress (section 1412(b)(13)) directed EPA to propose an MCLG and NPDWR for radon by August, 1999 (proposed on December 21, 1999, US EPA 1999n) and finalize the regulation by August, 2000 (section 1412(b)(13)). Like the Ground Water Rule, the Radon Rule will also be finalized before the Arsenic Rule. Systems may employ aeration to comply with the radon rule. Aeration alone, however, will not likely be sufficient to oxidize arsenic (III) to arsenic (V). However, if systems do aerate, they may be required by State regulations to also disinfect. The disinfection process may oxidize the arsenic, depending on the type of disinfection employed. Ultraviolet disinfection may not assist in arsenic oxidation (still under investigation by US EPA), whereas chemical disinfection or oxidation is likely to. Thus, the Radon Rule is expected to alleviate some of the burden of the Arsenic Rule. Microbial and Disinfection By-product Regulations. To control disinfection and disinfection byproducts and to strengthen control of microbial pathogens in drinking water, EPA is developing a group of interrelated regulations, as required by the SDWA. These regulations, referred to collectively as the Microbial Disinfection By-product (M/ DBP) Rules, are intended to address risk trade-offs between the two different types of contaminants. EPA proposed a Stage 1 Disinfectants/Disinfection By-products Rule (DBPR) and Interim Enhanced Surface Water Treatment Rule (IESWTR) in July 1994. EPA issued the final Stage 1 DBPR and IESWTR in November, 1998. The Agency has finalized and is currently implementing a third rule, the Information Collection Rule, that will provide data to support development of subsequent M/DBP regulations. These subsequent rules include a Stage 2 DBPR and a companion Long-Term 2 Enhanced Surface Water Treatment Rule (LT2ESWTR). The IESWTR will primarily affect large surface water systems, so EPA does not expect much overlap with small systems treating for arsenic. However, the Stage 1 DBPR will affect both large and small sized systems and may overlap with small systems treating for arsenic. In addition, the Stage 2 DBPR and possibly the LT2ESWTR would have significance as far as arsenic removal is concerned. For systems removing DBP precursors, systems may use nanofiltration. The use of nanofiltration would also be relevant for removing arsenic, and as a result, would ease some burden when systems implement these later rules. Hazardous Waste. The current toxicity characteristic (TC) regulatory level for designating arsenic as a hazardous waste under the Resource Conservation and Recovery Act (RCRA) is 5 mg/L and is listed in 40 CFR 261.24(a). It is important to differentiate between the toxicity characteristic and the toxicity characteristic leaching procedure (TCLP). The TCLP is the method by which a waste is evaluated to determine if it exceeds the toxicity characteristic. It is also important to note that while the toxicity characteristic was based on multiplying the current drinking water MCL by a factor of 100, the TC is not directly linked to the drinking water MCL. Thus, lowering the drinking water MCL does not mean that the toxicity characteristic would be lowered. A separate RCRA rulemaking would be required to lower the toxicity characteristic regulatory level. The drinking water standards for several inorganic contaminants have been lowered without any lowering of the toxicity characteristic. For example, the cadmium MCL was lowered from 10 µg/L to 5 µg/L in 1991, but the TC for cadmium still remains at 1.0 mg/L. The drinking water standard for lead was revised from an MCL of 50 µg/L to an action level of 15 µg/L. Both drinking water standards were lowered in 1991. The TC for lead remains at 5 mg/L. The studies summarized below show that arsenic residuals should be below the current TC of 5 mg/L and could be disposed in a non-hazardous landfill. In one study, sludges from four different water treatment plants were evaluated. (Bartley et al. 1992). There are data from two lime softening plants, one plant with both lime softening and coagulation/ filtration processes, and one arsenic removal plant utilizing coagulation/filtration. The raw water arsenic in the tow lime softening plants and the one plant using both lime softening and coagulation/ filtration were below 0.001 mg/L. The arsenic removal plant was removing arsenic from 1.1 mg/L to 0.42 mg/L using ferric sulfate coagulation. The product water was blended with water from another source to comply with the MCL. The TCLP extracts ranged from 0.007 to 0.039 mg/L, which is considerably below the current criterion for being designated a hazardous waste under RCRA. In another study, TCLP tests were performed using the activated alumina from two activated alumina plants (Wang et al., 2000). Both plants had similar setups (one is referred to as CS, the other is referred to as BES). Both systems consist of four tanks of activated alumina with two parallel sets of two tanks in series. The first set of tanks are used as roughing filters and the second set of tanks are used as polishing filters. The units were not regenerated, but replaced. For the CS system, the influent arsenic concentration ranged from 0.053 to 0.087 mg/L with an average of 0.062 mg/L. The effluent arsenic concentration was consistently below 0.005 mg/L. When the activated alumina media was removed from the roughing filters, three samples were taken. All three samples had arsenic TCLP test results of less than 0.05 mg/L. Again, these results were well below the regulatory limit. The influent arsenic concentration of the activated alumina plant referred to as BES ranged from 0.021 to 0.076 mg/L, with an average of 0.049 mg/L. Effluent levels were also less than 0.005 mg/L. When the media was removed from the two roughing filters, TCLP tests were taken. The results were 0.05 mg/ [[Page 38930]] L and 0.066 mg/L. Again, the results were below the regulatory limit. Another study examined residuals produced by anion exchange and coagulation-microfiltration (Clifford, 1997). Experiments were performed at the University of Houston-US EPA Drinking Water Research Facility, a 10 ft x 40 ft customized trailer containing various unit processes, including ion exchange and coagulation-microfiltration, and a small analytical lab. The mobile research facility was set up at the West Mesa Pump Station in Albuquerque, NM. The mean arsenic concentration in the source water was 0.021 mg/L. Ion exchange was field tested, and the media was regenerated. This initial waste stream was a brine from the regeneration process. The brine in the ion exchange process was reused 15 times. The average arsenic concentration in the product was below 0.002 mg/L during the 15 cycles. The process produced a highly concentrated spent brine, with arsenic concentrations reaching 26.6 mg/L. It should be noted that the arsenic concentration in the brine would be lower if the brine was not used as many times. After 6 months of storage, the arsenic concentration reduced to 11.3 mg/L. The arsenic was then precipitated out of the brine using iron, resulting in a brine with approximately 0.037 mg/L of arsenic. The precipitated sludge was then subjected to the TCLP extraction procedure. The TCLP extract had an average arsenic concentration of 0.270 mg/L. This is below the current threshold for being designated a hazardous waste. Coagulation-microfiltration was also field tested. Arsenic removal to below 0.002 mg/L could be achieved; 12,000 gallons of water were filtered over 3 days. The backwash water, which is the process waste, had less than 0.5% solids. According to the TCLP Method 1311, for a liquid waste containing less than 0.5% solids, the liquid portion of the waste after filtration, is defined as the TCLP extract. About 20 backwash samples were collected, filtered, and analyzed for arsenic. The average concentration in the backwash water after filtration was 0.0026 mg/L and thus could be disposed as a nonhazardous waste. Additionally, the simulated sludge was subjected to the TCLP leaching procedure. The arsenic concentration in the TCLP extract was 0.0218 mg/ L, which is also considerably lower than the regulatory limit. The University of Colorado performed a series of tests of various arsenic treatment solid residuals using the TCLP test (Amy et al, 1999). The arsenic treatment processes included conventional plants utilizing lime softening, alum and ferric chloride coagulation, activated alumina, and membranes. The results of this analysis for the conventional plant residuals are presented in Table VIII-6. The data indicates that all the plants would pass the current TCLP test although the data from the iron coagulation plant do approach the limit. Table VIII-6.--TCLP Results for Conventional Plant Arsenic Residuals ------------------------------------------------------------------------ TCLP extract Utility ID Type of utility Arsenic (mg/L) ------------------------------------------------------------------------ F, coagulation sludge........ Lime softening....... 0.0009 F, softening sludge.......... Lime softening....... 0.0039 F, filter sludge............. Lime softening....... 0.0014 G............................ Lime softening....... 0.002 J............................ Lime softening....... 0.0284 L............................ Alum coagulation..... 0.0093 C............................ Fe/Mn removal........ 0.0444 O............................ Iron coagulation..... 1.5596 ------------------------------------------------------------------------ Table VIII-7 is a summary of TCLP data on liquid residuals prepared by the University of Colorado for activated alumina regenerant and a reverse osmosis reject water precipitated with ferric chloride. The activated alumina regenerant solution was neutralized to a pH of 6, which caused the aluminum to precipitate and adsorb the arsenic. The membrane reject water was treated with ferric chloride to remove the arsenic and the resulting ferric hydroxide residual was tested. The data indicates that solid residuals generated from the alumina regenerant and membrane residuals would pass the TCLP test. Table VIII-7.--TCLP Test Results for Activated Alumina and Membrane Residuals ------------------------------------------------------------------------ TCLP Sample extract as (mg/L) ------------------------------------------------------------------------ Activated Alumina Column Regenerant........................ 0.0242 Membrane Filter Reject Residuals........................... 0.0179 ------------------------------------------------------------------------ All of the previous data is from residuals produced by central treatment. There is no TCLP data on spent activated alumina from POU or POE devices. The TCLP results of spent activated alumina media from POU and POE devices were simulated by assuming a worst-case scenario for 6- month and one year replacement frequencies (Kempic, 2000). To determine the amount of arsenic that could potentially leach into the extraction fluid during the toxicity characteristic leaching procedure, it was assumed that the influent arsenic concentration was 0.050 mg/L and that the activated alumina column adsorbed all of the arsenic. The first assumption represents the upper bound for influent concentrations since it is the current maximum contaminant level (MCL) for arsenic. The second assumption means that there would be no leakage or any breakthrough of arsenic through the column, which is not realistic. To calculate the total adsorbed arsenic mass, it was assumed that the POU unit treated 24 liters per day. This is the upper bound consumption used in the replacement frequency calculations. Two other assumptions were made to simulate the worst-case scenarios. In the TCLP, the solid phase is extracted with an amount of extraction fluid equal to 20 times the weight of the solid phase. The dry media mass was used for the solid phase for this calculation rather the wet media mass. It was also assumed that all of the adsorbed arsenic would leach into the extraction fluid, which is not realistic. The estimates for the worst-case scenarios are provided in Table VIII- 8. [[Page 38931]] Table VIII-8.--TCLP Projections for Activated Alumina Worst-Case Simulations ------------------------------------------------------------------------ Max TCLP Replacement frequency Conc. (mg/ L) ------------------------------------------------------------------------ POU & 6-months............................................. 2.6 POE & 6-months............................................. 0.8 POU & Annual............................................... 10.4 POE & Annual............................................... 3.2 ------------------------------------------------------------------------ The projections for three of the worst-case scenarios were below the TC of 5 mg/L. The worst-case maximum TCLP concentration for annual replacement for a POU activated alumina device was above the TC. However, despite this projection, activated alumina waste should be non-hazardous. The most unrealistic assumption was that all of the arsenic adsorbed onto the alumina would leach into the extraction fluid. The TCLP uses weak acetic acid (0.57%) at pH 5 for the extraction fluid. The optimal pH for arsenic adsorption onto activated alumina is between pH 5.5 and 6.0. Therefore, arsenic should be retained on the activated alumina at this pH. In fact, adsorbed arsenic is extremely difficult to remove under any conditions. A strong base (4% NaOH) is typically used to regenerate activated alumina. Arsenic is so strongly adsorbed to the activated alumina that only 50 to 70% of the arsenic is eluted during regeneration. Therefore, it is extremely unlikely that the spent activated alumina from POU and POE units would be considered hazardous. All of the TCLP data from solid residuals were below the current TC of 5 mg/L. The arsenic concentrations in TCLP extracts from alum coagulation, activated alumina, lime softening, iron/manganese removal, and coagulation-microfiltration residuals were below 0.05 mg/L, which is two orders of magnitude lower than the current TC regulatory level. The TCLP data for iron coagulation was mixed--the residuals from the arsenic removal plant were below 0.05 mg/L, but the residuals from another iron coagulation plant were above 1 mg/L. For anion exchange, the TCLP data on the precipitated brine stream was 0.27 mg/L. As was noted, this was a highly concentrated brine stream which had been used for fifteen regenerations. Arsenic concentrations in the precipitate would be lower if the brine was used for fewer regeneration cycles. Based on this data, EPA does not believe that drinking water treatment plant residuals would be classified as hazardous waste. The TCLP data also indicate that most residuals could meet a much lower TC regulatory level. EPA requests comment on whether it is appropriate to assume that all residuals can be disposed at a non-hazardous landfill. IX. Costs A. Why Does EPA Analyze the Regulatory Burden? EPA is responsible for issuing regulations that improve the quality of the nation's drinking water and reduce the risk of illness from exposure to harmful contaminants via drinking water supplied by public water systems (PWSs). As part of the regulatory development process, the Agency is required to analyze the regulatory cost and burden imposed on all regulated and affected entities and the benefits associated with the regulation. The Regulatory Impact Analysis (RIA) document is the principal summary of these analyses. Assessing the impacts of proposed SDWA regulations is a complex process, involving many analyses specified by various federal mandates. In particular, EPA must conduct analyses for the following mandates:
  • 1996 Safe Water Drinking Act (SDWA) Amendments
  • Paperwork Reduction Act (PRA)
  • Regulatory Flexibility Act (RFA)
  • Small Business Regulatory Enforcement Fairness Act (SBREFA)
  • Unfunded Mandates Reform Act (UMRA)
  • Executive Order (EO) 12866, ``Regulatory Planning and Review''
  • EO 12989, ``Federal Actions to Address Environmental Justice in Minority Populations and Low-Income Populations''
  • EO 13045, ``Protection of Children from Environmental Health Risks and Safety Risks.'' Executive Order 12866 describes the requirements for and content of the national cost-benefit analyses. Section 1412(b)(3)(C) of SDWA, as amended in 1996, directs EPA to seek comment on a health risk reduction and cost analysis (HRRCA) that will be issued with proposed MCLs. The HRRCA must identify quantifiable and nonquantifiable costs and health benefits of each MCL considered, including the incremental costs and benefits of each MCL considered. In addition, the HRRCA must identify benefits resulting from reducing co-occurring contaminants and exclude costs that will result from other proposed or final regulations. The Paperwork Reduction Act (PRA) requires federal agencies to document the cost and labor burden associated with data collection, recordkeeping, and reporting requirements of proposed regulations. The Regulatory Flexibility Act (RFA), as amended by the Small Business Regulatory Enforcement Fairness Act (SBREFA), mandates that federal agencies consider the impact imposed on small businesses, governments, and non- profit organizations. The objective of these mandates is to provide regulatory relief to small entities affected by SDWA regulations by identifying alternative or lower-cost compliance options. Finally, the Unfunded Mandates Reform Act (UMRA) seeks to assess the burden and costs of federal regulations to local and State governments, while Executive Order 12989 on environmental justice instructs federal agencies to evaluate the impact of proposed regulations on minority and low-income populations. Executive Order 13045 requires EPA to state how the regulation addresses risks for children. An RIA attempts to estimate the possible outcomes in terms of costs and benefits of various levels of regulation. At the most basic level, an RIA is built on estimates of the distribution of arsenic occurrence among the various water systems, the costs of treatment technologies, and predictions of responses by systems above the regulatory level under consideration. Because actual compliance monitoring at the proposed MCL has not been required of all systems at the time of proposal development, projections are based on statistical estimates. EPA believes that the current estimates include appropriate conservative assumptions and on average actual costs are not likely to exceed the estimates. One conservative assumption is that equipment useful life is identical to financing life. The Agency has a long term effort in progress to better characterize how much this issue will affect cost estimations. To be complete, accurate, and consistent, these analyses should be based on a single, integrated set of data and information that defines the baseline characteristics or conditions of the regulated community prior to implementation of the regulation. The regulated community is primarily the water supply industry and State, local, and tribal governments. However, it is the customers of public water systems, especially community water systems, that ultimately incur the cost burden and realize the intended health benefits of these regulations. Therefore, the baseline study identifies and, where possible, quantifies the universe (e.g., characteristics of water suppliers, their customers, and governmental entities) to [[Page 38932]] be used in the regulatory impact analysis (RIA). The current RIA applied national occurrence information in the modeling effort as described earlier in section V.G. EPA requests comment on its analyses for developing cost projections, including household costs, as well as additional cost information. Most previous RIAs conducted for the drinking water program assumed that all the water going into a system was the same concentration. Actually, many water systems (especially those serving more than 500 people) have multiple points where water enters the distribution system. Each of these entry points generally will have a different level of arsenic. Consequently, water systems tend to be impacted by regulations in stages that increase with decreasing regulatory level. Because costs are spread across the entire system, individual household expenditures will vary according to regulatory level. Past RIAs were unable to incorporate this information, and for costing purposes, all entry points to the distribution system required treatment. The arsenic RIA is the first drinking water chemical RIA to incorporate monte carlo simulation of intra-system occurrence variability into the cost and benefits estimation. This simulation permits more accurate characterization of the relative household impacts of various alternatives. Several other changes have also been incorporated into the cost and benefit estimates for the arsenic RIA: Very Large Systems--Very large water systems, those serving more than a million people, can contribute a significant portion to estimates of overall costs and benefits at select regulatory levels. On the other hand, because there are so few of these systems and given that they are of complex configuration, statistically based estimates of arsenic occurrence (especially at low levels of arsenic incidence) introduce very large uncertainty into the RIA. EPA addressed this issue by developing individually tailored estimates through the use of generally available occurrence information and Information Collection Rule data. Estimates were provided to the utilities and they were offered the opportunity to correct errors in the Agency assessment. While these estimates are a considerable improvement over past ones, it is important to keep in mind that they are merely projections and that individual compliance costs could actually still vary by a wide margin depending upon rule timing, interactions with other treatment or capital budget priorities, regulatory commission decisions, or actual compliance sampling results. Inventory Based Modeling--Past RIAs have generally developed benefit and cost estimates by estimating impacts for single representative community water systems within a limited number of size- based classes. Such an approach introduces a slight positive bias to total national cost estimates. This RIA has gone beyond the past approach in the modeling of community water system and non-transient non-community water system impacts. This RIA uses a monte carlo approach to simulate application of occurrence information to the actual SDWIS inventory. Through repeated simulations and assignments, the model is able to develop the most robust [statistically defensible] estimates of actual exposure levels and to better characterize the spread in household costs. B. How Did EPA Prepare the Baseline Study? EPA identified baseline characteristics as the first step in standardizing baseline profiles and information for use across all Agency drinking water RIAs and related analyses. The Agency has several efforts underway to develop improved technical approaches for cost and benefit analyses, including developing characteristic engineering unit costs of treatment plants, assessing financial and operational capacity, and considering the low-cost best available treatment (BAT) options for small systems. Then, EPA reviewed the analytical procedures, and data requirements needed to conduct the analyses. Table IX-1 provides an overview of the overall approach for identifying and classifying specific baseline characteristics. This matrix organizes the baseline characteristics according to the various entities likely to be affected by SDWA regulations and the different categories of data analysis inputs. The affected entities include:
  • State and Tribal Governments: Agencies at the State or local level (including certain Tribes and Alaskan Native Villages) responsible for implementing, administering, and enforcing drinking water programs, and other programs potentially affected by Federal drinking water mandates.
  • Public Water Suppliers: Utilities and other entities that provide potable water to 25 or more persons, 15 or more service connections (includes community and transient/non-transient non- community water systems).
  • Customers: All entities that purchase drinking water from public water systems (including residential, commercial, industrial, wholesale, governmental, agricultural, and other users). The corresponding categories of data analysis inputs shown in Table IX-1 include: 1. Technical/Operational: Characteristics relating to capital assets and operational processes, labor skills and training, and other variable inputs. 2. Managerial/Organizational: Characteristics relating to ownership, control and authority, organizational structure and management approach. 3. Financial/Economic: Characteristics relating to monetary factors, opportunity costs, and benefits. 4. Socio-Economic/Demographic: Composition and characteristics of affected entities (who, where, how much) and demographic trends. Data to describe all the baseline conditions shown in Table IX-1 are contained in a comprehensive EPA document designed to be applicable to all drinking water regulatory impact analyses, ``The Baseline Handbook.'' It is data from this document which is used in Chapter 4 of the RIA for Arsenic. 1. Use of Baseline Data Uses of baseline data include the following analyses: National and Sub-National Benefits, Costs, and Economic Impact Analyses:
  • Occurrence Analysis
  • Exposure/Risk Assessment
  • Model Plants/System Configuration
  • Unit Engineering Cost Analysis
  • Compliance Decision Tree Analysis
  • Financial Analysis
  • Government Implementation
  • Reporting, Recordkeeping, and Monitoring Costs
  • Valuation of Health Benefits
  • Non-Health Benefits Assessment
  • Economic Impact Assessment Small Entity Impact Analyses:
  • Small Entity Definition
  • Reporting and Recordkeeping Requirements for Small Entities
  • Financial Analysis for Small Entities
  • Socio-Economic Analysis for Small Entities
  • Regulatory Alternatives Analysis Other Special Analyses:
  • Health Risks to Sensitive Subpopulations
  • Affordability Analyses
  • Government Budgetary Effects These broad analytical requirements reflect the overlapping nature of the required analyses pursuant to the relevant statutory and administrative mandates. For example, various mandates, including EO 12866, SDWA, [[Page 38933]] UMRA, and PRA, require national cost and benefit analyses. 2. Key Data Sources Used in the Baseline Analysis for the RIA? A number of different data sources were employed in the development of the tables included Chapter 4 of the arsenic RIA. The key data sources used included: 1995 Community Water System Survey (CWSS). This database was compiled by EPA from a survey conducted in 1995 to profile the operational and financial characteristics of community water systems of all source, size, and ownership types. WATER STATS, The Water Utility Database. This database was compiled by the American Water Works Association from a 1996 survey of its member utilities. Data on water system operations and finances were collected in two stages. The first stage involved a comprehensive census of the largest water utilities (i.e., those serving 50,000 or more persons). A second stage data collection involved a statistical sample of smaller water utilities. Safe Drinking Water Information System (SDWIS). This database serves as the U.S. EPA's comprehensive database of public water system regulatory compliance and violation information. SDWIS contains the Agency's inventory of all public water supplies, both community and noncommunity systems and the populations they serve. Survey on State Program Staffing/Funding for FY-97. The Association of State Drinking Water Administrators (ASDWA) conducted a survey of State drinking water programs to solicit estimates on the number of staff (i.e., full-time equivalents, FTEs) involved in drinking water regulatory implementation and enforcement activities by program area, as well as estimates of drinking water program revenues/funding and expenditures by major account categories. 1990 Census of Population. Data from the 1990 Census of Population was used in conjunction with water system data to develop estimates for various demographic characteristics of households and communities served by public water systems. Table IX-1.--Summary of General Baseline Categories of Affected Entities ---------------------------------------------------------------------------------------------------------------- Baseline characteristics ------------------------------------------------------------------------------- Affected entity 1: Technical & 2: Managerial & 3: Economic & 4: Socioeconomic & operational organizational financial demographic ---------------------------------------------------------------------------------------------------------------- A: State Government............. A1.1 PWS A2.1 Program A3.1 Program A4.1 State PWS Inspections & Staffing. Expenditures. Profile. Sanitary Surveys. A2.2 Laboratory A3.2 Program Capacity/ Funding/Revenues. Facilities. A2.3 Division of Authority/ Jurisdiction. B: Public Water Suppliers....... B1.1 Water B2.1 Ownership/ B3.1 Operating B4.1 PWS Type. Sources/Intakes. Organizational Expenses. B4.2 PWS Size/ B1.2 Source Structure. B3.2 Operating Customer Base. Contamination/ B2.2 Plant Revenues. B4.3 PWS Source Protection. Operation/ B3.3 Non- Water. B1.3 Physical Operators. Operating B4.4 Geographic Configuration. Expenses. Location. B1.4 Plant B3.4 Assets & Condition. Liabilities. B1.5 Plant Flow/ B3.5 Rate Capacity. Structures/User B1.6 Treatment/ Burden. Waste Processes B3.6 Capital In-Place. Investment B1.7 Storage Expenditure. Capacity. B1.8 Distribution System. B1.9 Residence Time. B1.1 Monitoring/ Laboratory. C: Customers.................... C1.1 POU/POE C2.1 Alternative C3.1 Residential C4.1 Population Systems In Use. Water Use. Income. Profile. C2.2 Public C3.2 Nonresidenti C4.2 Customer Attitudes/ al Income. Water Use. Perceptions. C3.3 Residential Water Costs. C3.4 Nonresidenti al Water Costs. C3.5 Cost of Drinking Water Alternatives. C3.6 Medical Costs. C3.7 Non-Medical Costs. C3.8 Community Financial Information. ---------------------------------------------------------------------------------------------------------------- C. How Were Very Large System Costs Derived? EPA must conduct a thorough cost-benefit analysis, and provide comprehensive, informative, and understandable information to the public about its regulatory efforts. As part of these analyses, EPA evaluated the regulatory costs of compliance for very large systems, who would be subject to the new arsenic drinking water regulation. The nation's 25 largest drinking water systems (i.e., those serving a million people or more) supply approximately 38 million people and generally account for about 15 to 20 percent of all compliance-related costs. Accurately determining these costs for future regulations is critical. As a result, EPA has developed compliance cost estimates for the arsenic and radon regulations for each individual system that serves greater than 1 million persons. These cost estimates help EPA to more accurately assess the cost impacts and benefits of the arsenic regulation. The estimates also help the Agency identify lower cost regulatory options and better understand current water systems' capabilities and constraints. The system costs were calculated for the 24 public water systems that serve a retail population greater than 1 million persons and one public water [[Page 38934]] system that serves a wholesale population of 16 million persons. Table IX-2 lists these 25 public water systems. The distinguishing characteristics of these very large systems include: (1) A large number of entry points from diverse sources; (2) mixed (i.e., ground and surface) sources; (3) Occurrence not conducive to mathematical modeling; (4) Significant levels of wholesaling; (5) Sophisticated in-place treatment; (6) Retrofit costs dramatically influenced by site-specific factors; and (7) Large amounts of waste management and disposal which can contribute substantial costs. Table IX-2.--List of Large Water Systems That Serve More Than 1 Million People ---------------------------------------------------------------------------------------------------------------- PWS ID # Utility name ---------------------------------------------------------------------------------------------------------------- 1................................... AZ0407025 Phoenix Municipal Water System. 2................................... CA0110005 East Bay Municipal Utility District. 3................................... CA1910067 Los Angeles--City Dept. of Water and Power. 4................................... CA1910087 Metropolitan Water District of Southern California. 5................................... CA3710020 San Diego--City of. 6................................... CA3810001 San Francisco Water Department. 7................................... CA4310011 San Jose Water Company. 8................................... CO0116001 Denver Water Board. 9................................... FL4130871 Miami-Dade Water And Sewer Authority--Main System. 10.................................. GA1210001 City of Atlanta. 11.................................. IL0316000 City of Chicago. 12.................................. MA6000000 Massachusetts Water Resource Authority. 13.................................. MD0150005 Washington Suburban Sanitation Commission. 14.................................. MD0300002 Baltimore City. 15.................................. MI0001800 City of Detroit. 16.................................. MO6010716 St. Louis County Water County. 17.................................. NY5110526 Suffolk County Water Authority. 18.................................. NY7003493 New York City Aqueduct System. 19.................................. OH1800311 City of Cleveland. 20.................................. PA1510001 Philadelphia Water Department. 21.................................. PR0002591 San Juan Metropolitano. 22.................................. TX0570004 Dallas Water Utility. 23.................................. TX1010013 City of Houston--Public Works Department. 24.................................. TX150018 San Antonio Water System. 25.................................. WA5377050 Seattle Public Utilities. ---------------------------------------------------------------------------------------------------------------- Generic models cannot incorporate all of these considerations; therefore, in-depth characterizations and cost analyses were developed utilizing several existing databases and surveys. The profile for each system contains information such as design and average daily flows, treatment facility diagrams, chemical feed processes, water quality parameters, system layouts, and intake and aquifer locations. System and treatment data were obtained from the following sources: (1) The Information Collection Rule (1997); (2) The Community Water Supply Survey (1995); (3) The Association of Metropolitan Water Agencies Survey (1998); (4) The Safe Drinking Water Information System (SDWIS); and (5) The American Water Works Association WATERSTATS Survey (1997). While these sources contained much of the information necessary to perform cost analyses, the Agency was still missing some of the detailed arsenic occurrence data in these large water systems. Where major gaps existed, especially in groundwater systems, occurrence data obtained from the States of Texas, California, and Arizona, the Metropolitan Water District of Southern California Arsenic Study (1993), the National Inorganic and Radionuclides Study (EPA, 1984), and utilities were used. Based on data from the studies, detailed costs estimates were derived for each of the very large water systems. Cost estimates were generated for each system at several MCL options. The total capital costs and operational and maintenance (O & M) costs were calculated using the profile information gathered on each system, conceptual designs (i.e., vendor estimates and RS Means), and modified EPA cost models (i.e., Water and WaterCost models). The models were modified based on the general cost assumptions developed in the Phase I Water Treatment Cost Upgrades (EPA, 1998). Preliminary cost estimates were sent to all of the systems for their review. Approximately 30% of the systems responded by submitting revised estimates and/or detailed arsenic occurrence data. Based on the information received, EPA revised the cost estimates for those systems. Based on the results, the majority of the very large systems will not have capital or O&M expenditures for complying with a MCL of 5 µg/L (Table IX-3). More detailed costs estimates for each very large water system can be found in the water docket. Table IX-3.--Total Annual Costs for Large Systems for (Serving More Than 1 Million People) ------------------------------------------------------------------------ Number MCL option (µg/L) systems Cost treating [$millions] ----------------------------------------------------------------\1\----- 3.......................................... 3 $16-18 5.......................................... 3 11-12 10......................................... 3 6.6-7.47 20 3 2.6-2.7 ------------------------------------------------------------------------ \1\ The lower number shows costs annualized at 3%; the higher number shows costs annualized at 7% capital costs. The 7% rate represents the standard discount rate preferred by OMB for benefit-cost analyses of government programs and regulations. D. How Did EPA Develop Cost Estimates? EPA developed national cost estimates by using the occurrence data, unit cost curves, and a decision tree. The occurrence data provides a measure of the number of systems that would need to install treatment in each size [[Page 38935]] category (the occurrence data was described in Section V). The unit cost curves provide a measure of how much a technology will cost to install. Unit cost curves are continuous functions; they are a function of system size and provide an estimated cost for all design and average flows. The costs for a treatment train for the average flow in each size category were given previously in Table VIII-3. The unit cost curves can be found in ``Technologies and Costs for the Removal of Arsenic From Drinking Water'' (US EPA, 1999i). EPA then developed a decision tree, which is a prediction of what treatment technology trains facilities would likely install to comply with options considered for the revised arsenic standard. A brief discussion of this decision tree follows. A copy of the full 300+ page flowchart and supporting documentation can be found in ``Decision Tree for the Arsenic Rulemaking Process'' (US EPA, 1999d). The following figure is a brief representation of this flowchart. As shown in the flowchart, EPA considered the impact of (1) MCL option and influent arsenic concentration; (2) system size; (3) regional effects (water scarcity); (4) source water type (that is, ground water or surface water); (5) existing treatment in-place; (6) waste disposal issues and costs; and (7) co-occurrence of iron and sulfate, to estimate what systems are likely to install. Ultimately, the decision tree was expressed in decision matrices, in which EPA assigned probabilities as to how often each of the treatment trains in Table VIII-2 will likely be used. EPA developed a different decision matrix for the eight system size categories, for three different removal efficiencies (50%, 50-90% and >90%), and for two source waters (ground and surface). In general, to the extent possible (e.g., based on source water quality), EPA assumed that systems would employ the least-cost technology that can meet the MCL option. BILLING CODE 6560-50-P [[Page 38936]] [GRAPHIC] [TIFF OMITTED] TP22JN00.001 BILLING CODE 6560-50-C [[Page 38937]] MCL option. EPA developed a decision tree that accounted for treatment technology limitations, and only assigned non-zero probabilities in the matrices to those technologies capable of reaching each MCL option. The maximum removal percentages are given in Table VIII-1. For instance, since greensand filtration is only assumed capable of removing 50% of the influent arsenic, for an influent level of 20 µg/L, the technology is assumed to be capable of only producing product water with 10 µg/L of arsenic. Therefore, for an MCL option of 5 µg/L, no usage was assumed for greensand filtration at a 20 µg/L level of influent arsenic. System size. The decision tree also depends on system size. For instance, small systems are assumed to operate activated alumina on a throw-away basis, and thus the probability of using a treatment train that employs on-site regeneration is assumed to be zero. The converse is true for large systems; non-zero probabilities are assumed only for those trains that employ regeneration on-site. Water scarcity. Water scarcity was also taken under consideration when developing the decision tree. It was assumed that this issue would adversely affect the selection of reverse osmosis, since the technology rejects a significant portion of the influent water. However, the costs for reverse osmosis treatment trains are much higher than others (refer to Table VIII-3), and systems would likely opt for other, less expensive, treatment options. For the range of MCL options considered, it was assumed that ion exchange would be capable of delivering the required removal efficiencies. Thus, water scarcity, though considered in the decision tree, did not affect percentages assigned to reverse osmosis. Source water type. Source water type is also a factor in the decision tree. It affects the unit cost curves; one set of curves were developed for surface water, and another was developed for ground water. The treatment-in-place data and co-occurrence data (as shown below) are sorted by source water type. Also, certain technologies are considered appropriate for one source water type, but not the other. For instance, greensand filtration is considered relevant only for ground waters. Existing treatment in-place. Treatments that may already exist at facilities were taken into account in the decision tree. It was assumed that systems would need to pre-oxidize, if they weren't doing so already. Table IX-4 shows the number of systems that were assumed to require addition of pre-oxidation (Source: US EPA,1999e). Table IX-4.--Systems Needing To Add Pre-Oxidation ------------------------------------------------------------------------ Percent of Percent of ground surface System size water water systems systems ------------------------------------------------------------------------ 25-100........................................ 54 9 101-500....................................... 30 4 501-1K........................................ 24 0 1,001-3.3K.................................... 24 0 3,301-10K..................................... 27 3 10,001-50K.................................... 13 1 50,001-100K................................... 41 2 100,001-1 M................................... 16 0 ------------------------------------------------------------------------ It was also assumed that those systems that had coagulation/ filtration in place, or lime softening in place, would modify those treatments to optimize for arsenic removal, since it is a relatively inexpensive option. The percent of systems with these treatments in place is given in Table IX-5 (Source: US EPA,1999e). However, for higher removals (>90%), it was assumed that only half of the systems would be able to achieve the desired removal with a modification. For those systems, an additional cost of a polishing step, such as ion exchange, was added. Table IX-5.--Percent of Systems With Coagulation-Filtration and Lime-Softening in Place ---------------------------------------------------------------------------------------------------------------- Percent of Percent of Percent of Percent of ground water surface water ground water surface water System size systems with systems with systems with systems with CF in place CF in place LS in place LS in place ---------------------------------------------------------------------------------------------------------------- 25-100.......................................... 2 22 3 4 101-500......................................... 4 53 3 9 501-1K.......................................... 2 73 2 19 1,001-3.3K...................................... 3 76 3 16 3,301-10K....................................... 8 85 3 7 10,001-50K...................................... 4 92 5 8 50,001-100K..................................... 4 85 3 5 100,001-1 M..................................... 5 94 10 5 ---------------------------------------------------------------------------------------------------------------- Waste disposal issues and costs. Waste disposal of arsenic contaminated sludges and brines was also factored into the decision tree, and waste costs were added to the treatment trains. The waste disposal options for each of the technologies considered are given in Table IX-6. For ion exchange and activated alumina, it was assumed that the waste streams would be too concentrated to discharge directly. For these technologies, it was assumed that some of the smallest systems would be able to take advantage of evaporation ponds, but that this option would be cost prohibitive in medium and large systems. It was assumed that most systems would opt for either chemical precipitation or discharge to a sanitary sewer. EPA also assumed that systems would dispose of spent activated alumina media in non-hazardous landfills. Costs for reverse osmosis are prohibitive (In Table VIII-3, Annual Costs of Treatment Trains, compare lines 11, 12, and 13 against other technologies), but if used, EPA assumed the relatively large amount of reject water would be discharged directly (because it would not be as concentrated as ion exchange and activated alumina waste streams), to a sanitary sewer or by chemical precipitation. For coagulation assisted microfiltration, modified coagulation filtration, and modified lime softening, EPA assumed the waste would be discharged to non-hazardous landfills after the sludge is mechanically or non-mechanically dewatered. For greensand filtration, it was assumed that the spent media would be disposed of in a non-hazardous landfill. [[Page 38938]] Table IX-6.--Waste Disposal Options -------------------------------------------------------------------------------------------------------------------------------------------------------- POTW waste Non-haz Direct Chemical Non-mech Treatment tech disposal Evap pond landfill discharge precip Mech dewater dewater -------------------------------------------------------------------------------------------------------------------------------------------------------- Ion Exchange.......................................... Activated Alumina..................................... Reverse Osmosis....................................... Coag Assisted Micro-filtration........................ Greensand............................................. Modify CF............................................. Modify LS............................................. -------------------------------------------------------------------------------------------------------------------------------------------------------- Co-occurrence of iron and sulfate. EPA also factored into the decision tree co-occurrence data on iron and sulfate (shown in Tables IX-7 to IX-10, Source: US EPA,1999f ). Co-occurrence of sulfate in water adversely affects the performance of ion exchange, and increases operation and maintenance costs. Three sulfate-level treatment trains were costed for ion exchange: one low-level, one mid-level and one high-level. The percentages in Tables IX-7 to IX-8 were used as ceilings in national cost estimates and limited the number of systems that could be placed in the decision matrices in the low-level and mid- level sulfate ranges. For example, the co-occurrence data shows that the maximum number of systems that can be costed at the low-level sulfate treatment train for an influent level of arsenic between 10 and 20 µg/L is 35%. If more systems were to be placed in the decision matrices under ion exchange, no more than 39% were assumed to face a sulfate level between 25 and 120 mg/L. Any more systems assigned to ion exchange in the decision matrices were assumed to face high sulfate levels. Table IX-7.--Ground Water: Arsenic and Sulfate ------------------------------------------------------------------------ Likelihood of sulfate (percent) Influent arsenic -------------------------------------- 25 mg/L 25-120 mg/L >120 mg/L ------------------------------------------------------------------------ 10 µg/L.................. 48 33 19 10-20 µg/L............... 35 39 26 >20 µg/L................. 33 38 30 ------------------------------------------------------------------------ Table IX-8.--Arsenic Water: Arsenic and Sulfate ------------------------------------------------------------------------ Likelihood of sulfate (percent) Influent arsenic -------------------------------------- 25 mg/L 25-120 mg/L >120 mg/L ------------------------------------------------------------------------ 10 µg/L.................. 28 32 40 10-20 µg/L............... 20 30 51 >20 µg/L................. 12 28 60 ------------------------------------------------------------------------ Table IX-9.--Ground Water: Arsenic and Iron ------------------------------------------------------------------------ Likelihood of sulfate (percent) ------------------------- Influent arsenic 300 >300 µg/L µg/L ------------------------------------------------------------------------ 10 µg/L............................... 82 18 10-20 µg/L............................ 81 19 >20 µg/L.............................. 71 29 ------------------------------------------------------------------------ Table IX-10.--Surface Water: Arsenic and Iron ------------------------------------------------------------------------ Likelihood of sulfate (percent) ------------------------- Influent arsenic 300 >300 µg/L µg/L ------------------------------------------------------------------------ 10 µg/L............................... 91 9 10-20 µg/L............................ 92 8 >20 µg/L.............................. 90 10 ------------------------------------------------------------------------ Co-occurrence of iron in water improves the performance of greensand filtration. Greensand is relatively inexpensive for small systems to use, but not as effective as other treatment technologies. It was assumed that systems would opt for greensand filtration only if the level of iron was greater than 300 µg/L. EPA used the co- occurrence data in Tables IX-9 to IX-10 to determine the ceiling on the number of systems that could use greensand filtration in the decision matrices. E. What Are the National Treatment Costs of Different MCL Options? Under the proposed option of 5 µg/L, the Agency estimates that annual treatment costs to community water systems will be $374 million per year. If required to treat at the proposed level, [[Page 38939]] treatment costs to non-community non-transient systems would be $15 million per year. National annual costs for the MCL options considered (3, 5, 10, and 20 µg/L) are provided in Table IX-11. Table IX-11.--National Annual Treatment Costs [Dollars in millions] ---------------------------------------------------------------------------------------------------------------- Non-community Total MCL option (µg/L) Community Non-transient treatment water systems systems costs ---------------------------------------------------------------------------------------------------------------- 3............................................................... $639 $25 $664 5............................................................... 374 15 389 10.............................................................. 160 6 166 20.............................................................. 59 2 61 ---------------------------------------------------------------------------------------------------------------- Total annual costs per household are given in Table IX-12. Due to economies of scale, costs per household are higher in the smaller size categories, and lower in the larger size categories. For the proposed option of 0.005 µg/L, costs are expected to be $364 per household for systems serving 25-100 people, and $254 per household for systems serving 101-500 people. Costs per households in systems larger than those are substantially lower: from $104 to $21 per household. Costs per household do not vary dramatically across MCL options. This is because of the fact that once a system installs a treatment technology to meet an MCL target, costs do not vary significantly based upon the removal efficiency it will be operated under. Costs are, however, somewhat lower at less stringent MCL options. This is because it was assumed that some systems would blend water at these options, and treat only a portion of the flow. Table IX-12.--Total Annual Costs per Household [Dollars] ---------------------------------------------------------------------------------------------------------------- 10 µg/ 20 µg/ System size 3 µg/L 5 µg/L L L ---------------------------------------------------------------------------------------------------------------- 25-100.......................................... $368 $364 $357 $349 101-500......................................... 259 254 246 238 501-1K.......................................... 106 104 98 93 1K-3.3K......................................... 64 60 57 52 3.3K-10K........................................ 44 41 37 33 10K-50K......................................... 36 33 29 25 50K-100K........................................ 30 27 23 19 100K-1M......................................... 23 21 18 15 ---------------------------------------------------------------------------------------------------------------- Incremental costs are given in Tables IX-13 and IX-14. Incremental costs refer to the dollars that must be spent to obtain the next, more stringent, level of control. The national and household costs under 20 µg/L refer to the amount that must be spent to reach 20 µg/L starting from the baseline of 50 µg/L. The dollar value under 10 µg/L represents the cost differential between 20 µg/L and 10 µg/L. The values under 5 µg/L and 3 µg/L were derived similarly. Table IX-13.--Incremental National Annual Costs [Dollars in millions] ---------------------------------------------------------------------------------------------------------------- Non-community MCL option (µg/L) Community non-transient Total water systems water systems ---------------------------------------------------------------------------------------------------------------- 20.............................................................. $59 $2 $61 10.............................................................. 101 4 105 5............................................................... 214 9 223 3............................................................... 265 19 275 ---------------------------------------------------------------------------------------------------------------- Table IX-14.--Incremental Annual Costs per Household [Dollars] ---------------------------------------------------------------------------------------------------------------- 20 µg/ 10 µg/ System size L L 5 µg/L 3 µg/L ---------------------------------------------------------------------------------------------------------------- 25-100.......................................... $349 $8 $7 $4 101-500......................................... 238 8 8 5 501-1K.......................................... 93 5 6 2 1K-3.3K......................................... 52 5 3 4 3.3K-10K........................................ 33 4 4 3 [[Page 38940]] 10K-50K......................................... 25 4 4 3 50K-100K........................................ 19 4 4 3 100K-1M......................................... 15 3 3 2 ---------------------------------------------------------------------------------------------------------------- In the process of analyzing treatment technologies and developing cost estimates, EPA held several meetings with stakeholders to obtain input on assumptions made. Several of the key assumptions agreed to by stakeholders are given below. 1. Assumptions Affecting the Development of the Decision Tree
  • EPA assumed that ion exchange usage would be prohibited above 120 mg/L of sulfate and 500 mg/L of TDS.
  • EPA assumed that greensand filtration would be used only if iron in the raw water was above 300 µg/L.
  • EPA assumed that systems would pre-oxidize, when existing chlorination or other oxidants are not already present.
  • EPA assumed that systems would not likely use POE-RO nor POE-IX because of corrosion control problems. Also, with IX, if the resin is not replaced and/or regenerated on time, there is a potential for arsenic peaking. EPA assumed that systems will most likely use POE- AA.
  • The breakthrough issue also exists with POU-IX. POU-AA has the advantage of a longer run length. EPA assumed that systems would use either POU-AA or POU-RO. 2. Assumptions Affecting Unit Cost Curves
  • There are significant safety and operating efficiency risks to small systems when adjusting downward. This pH adjustment would require much more oversight than most small systems will have. EPA, in calculating unit costs for activate alumina assumed that systems would not adjust pH downward; thus, AA will be operated at a sub-optimal pH.
  • There is a danger of operating technologies such as ion exchange near breakthrough. EPA incorporated a safety factor, and used 80% of the MCL as the target when calculating costs for all technologies.
  • EPA assumed that small systems would not regenerate Activated Alumina on site--AA will likely be operated on a ``throw- away'' basis.
  • For modifying coagulation/filtration, EPA considered the cost of a new chemical feed system when switching to iron. EPA costed out switching coagulants for high removals. For lower removals, EPA costed out optimizing alum usage.
  • EPA assumed 75% for RO recovery.
  • For Activated Alumina, EPA assumed that there will not be any systems with raw water in the optimal range for arsenic removal (pH between 5.5-6.0).
  • For iron-coagulation-micro-filtration EPA assumed systems would apply a stronger iron dose rather than adjusting to optimum pH.
  • For ion exchange, one or more regenerations per day is not problematic. Regeneration in Ion Exchange can be done automatically. EPA examined cost models on regeneration frequency, volume of waste generated and considered computer-automation for regeneration. X. Benefits of Arsenic Reduction The benefits associated with reductions of arsenic in drinking water arise from a reduction in the risk of adverse human health effects, and a corresponding decrease in the number of expected cases and premature deaths of people experiencing these effects. The various adverse health effects associated with arsenic are known with different levels of certainty. Presently some can be quantified and some cannot. The best characterized benefits can be both quantified and monetized (i.e., a dollar value is attached to the expected decrease in number of cases), while other benefits may be only known well enough to describe. The latter are known as qualitative benefits. The Safe Drinking Water Act (SDWA) amendments of 1996 require that EPA fully consider both quantifiable and non-quantifiable benefits that result from drinking water regulations. The first step in the benefits evaluation process is to consider the adverse health effects that may be expected to decrease with a reduction in the concentrations of arsenic in drinking water. Arsenic has many health effects, both cancer and non-cancer. Section III. discusses these health effects. As discussed in section VIII.A., treatment for arsenic removal may add or remove other contaminants. Using chlorine or other oxidants may increase risk from disinfection by-products. On the other hand, treatments put in place for arsenic may incidentally reduce the risk from other co-occurring contaminants. A. Monetized Benefits of Avoiding Bladder Cancer Reducing arsenic levels in tap water will reduce the risks of suffering the adverse health effects described in the previous sections. In 1999 the National Research Council examined several risk distributions for male bladder cancer in 42 villages in Taiwan with arsenic ranging from 10 to 934 µg/L, grouping arsenic exposure by village. Previous scientific studies analyzed risk using less specific exposure categories, which can obscure ``the true shape of the dose response curve (NRC 1999, page 273).'' Risk assessments for other adverse health effects have not been as thoroughly addressed. To monetize bladder cancer benefits, EPA calculated the number of cases potentially avoided based on the NRC bladder cancer risk analyses. The cases are evaluated in terms of the economic benefits associated with avoiding the cancer cases. In addition to the monetized benefits of avoiding bladder cancer, EPA has chosen to monetize the potential benefits of avoided lung cancer, using a ``What If'' analysis based on statements in the NRC report (see section X.B for applying the ``what-if'' scenario to lung cancer). 1. Risk Reductions: The Analytic Approach EPA applied the 1999 NRC bladder cancer risk assessment to U.S. males and females. The following sections explain how we calculated risk reductions for populations exposed to MCL options of 3 µg/ L and above. The approach for this analysis included five components. First, EPA used data from the recent EPA water consumption study. This study is described in section X.A.2. Second, Monte Carlo simulations (section X.A.3) were used to develop relative exposure factors (section X.A.4). Third, arsenic occurrence estimates were used to identify the population exposed to levels above 3 µg/L. Fourth, NRC risk distributions were chosen for [[Page 38941]] the analysis. Fifth, EPA developed estimates of the risks faced by exposed populations using Monte Carlo simulations, using the relative exposure factors, occurrence, and NRC risk distributions mentioned above. These components of the analysis are described in the following sections. 2. Water Consumption EPA recently updated its estimates of personal (per capita) daily average estimates of water consumption (``Estimated per Capita Water Consumption in the United States,'' EPA 2000a). The estimates used data from the combined 1994, 1995, and 1996 Continuing Survey of Food Intakes by Individuals (CSFII), conducted by the U.S. Department of Agriculture (USDA). The CSFII is a complex, multistage area probability sample of the entire U.S. and is conducted to survey the food and beverage intake of the U.S. Estimates of water consumed include direct water, indirect water and total water (Table X-1). ``Direct'' water is tap water consumed directly as a beverage. ``Indirect'' water is defined as water added to foods and beverages during final preparation at home or by food service establishments such as school cafeterias and restaurants. For the purpose of the report, indirect water did not include ``intrinsic'' water which consists of water found naturally in foods (biological water) and water added by commercial food and beverage manufactures (commercial water). ``Total'' water refers to combined direct and indirect water consumption. Table X-1.--Source of Water Consumed ---------------------------------------------------------------------------------------------------------------- Indirect (from Source Direct food and Bottled water (drinking) beverages) ---------------------------------------------------------------------------------------------------------------- Community Tap................................................ X X Well Tap..................................................... X X Total........................................................ X X X ---------------------------------------------------------------------------------------------------------------- Per capita water consumption estimates are reported by source. Sources include community/tap water, bottled water, and water from other sources, including water from household wells and rain cisterns, and household and public springs. For each source, the mean and percentiles of the distribution of average daily per capita consumption are reported. The estimates are based on an average of 2 days of reported consumption by survey respondents. The estimated mean daily average per capita consumption of community/tap water by individuals in the U.S. population is 1 liter/ person/day. For total water, which includes bottled water, the estimated mean daily average per capita consumption is 1.2 liters per/ person/day. These estimates of water consumption are based on a sample of 15,303 individuals in the 50 States and the District of Columbia. The sample was selected to represent the entire population of the U.S. based on 1990 census data. The estimated 90th percentile of the empirical distribution of daily average per capita consumption of community/tap water for the U.S. population is 2.1 liters/person/day; the corresponding number for daily average per capita consumption of total water is 2.3 liters/ person/day. In other words, current consumption data indicate that 90 percent of the U.S. population consumes up to approximately 2 liters/ person/day, which is the amount many federal agencies use as a standard consumption value. Water consumption estimates for selected subpopulations in the U.S. are described in the analysis, including per capita water consumption by source for gender, region, age categories, economic status, race, and residential status and separately for pregnant women, lactating women, and women in childbearing years. The water consumption estimates by age were used in the computation of the relative exposure factors discussed in the section X.A.4. These water consumption numbers differ somewhat from previous estimates reported in earlier studies. The mean per capita daily intake of total tap water, as estimated from the 1977-78 USDA's Nationwide Food Consumption Survey, was 1.193 liters/person/day (reported by Ershow and Cantor in 1989). Based on the 1977-78 study, the estimated percentile corresponding to 2 liters per day consumed is the 88th. 3. Monte Carlo Analysis Monte Carlo analysis is a technique for analyzing problems where there are a large number of combinations of input values that are too large to calculate for every possible result. A random number generator is used to generate numbers that correspond to assumptions about the distribution or likelihood of various input values. For each set of random input values a single outcome is calculated. As the simulation runs, the outcome is recalculated for each new set of input values and continues until a stopping criterion is reached. The accuracy of this technique, like other statistical techniques, depends on the accuracy of the underlying assumptions about the distribution of input values; it does not resolve the uncertainty behind the assumptions. For the risk distributions calculated in this report, the simulations were carried out 2,000 times. For each simulation, a relative exposure factor, occurrence estimate, and individual risk estimate were calculated. These calculations resulted in estimates of the risks faced by populations exposed to arsenic concentrations in their drinking water. The underlying risk distribution are described in the following sections. 4. Relative Exposure Factors EPA used models to integrate the new drinking water consumption study information into the benefits analysis. We used distributions for both community/tap water and total water consumption because the community water/tap water estimates may underestimate actual tap water consumption. In this analysis, we combined the water consumption data with data on population weight from the U.S. Census. The weight data included a mean and a distribution of weight for male and females on a year-to-year basis throughout a lifetime. Monte Carlo analysis generated male and female relative exposure factors (REFs) for each of the broad age categories used in the water consumption study. Lifetime male and female relative exposure factors were then estimated, where the factors show the sensitivity of exposure to an individual weighing 70 kilograms and consuming 2 liters of water per day. These life-long REFs can be directly multiplied by the average drinking water consumption to provide estimates of individual lifetime consumption [[Page 38942]] practices. The REFs provide a means to incorporate information on various age groups, for example children, into the analysis, as weight and water consumption vary among age groups. The means and variances of the REFs derived from this analysis were: for community water consumption (0.60, 0.37 males; 0.64, 0.36 females), for total water consumption (0.73, 0.39 males; 0.79, 0.37 females). 5. NRC Risk Distributions While the NRC's work did not constitute a formal risk analysis, they did examine many statistical issues (e.g., measurement errors, age-specific probabilities, body weight, water consumption rate, comparison populations, mortality rates, choice of model) and provided a starting point for additional EPA analyses. The report noted that ``poor nutrition, low selenium concentrations in Taiwan, genetic and cultural characteristics, and arsenic intake from food'' were not accounted for in their analysis (NRC, 1999, pg. 295). In its 1999 report, ``Arsenic in Drinking Water,'' the NRC analyzed bladder cancer risks using data from Taiwan. In addition, NRC examined evidence from human epidemiological studies in Chile and Argentina, and concluded that risks of bladder and lung cancer were comparable to those ``in Taiwan at comparable levels of exposure (NRC 1999, page 7).'' The NRC also examined the implications of applying different mathematical procedures to the newly available Taiwanese data for the purpose of characterizing bladder cancer risk. These risk distributions are based on bladder cancer mortality data in Taiwan, in a section of Taiwan where arsenic concentrations in the water are very high by comparison to those in the U.S. It is also an area of very low incomes and poor diets, and the availability and quality of medical care is not of high quality, by U.S. standards. In its estimate of bladder cancer risk, the Agency assumed that within the Taiwanese study area, the risk of contracting bladder cancer was relatively close to the risk of dying from bladder cancer (that is, that the bladder cancer incidence rate was equal to the bladder cancer mortality rate). At the time the study data were collected the chances of surviving were probably poor for individuals diagnosed with bladder cancer. We do not have data, however, on the rates of survival for bladder cancer in the Taiwanese villages in the study and at the time of data collection. We do know that the relative survival rates for bladder cancer in developing countries overall ranged from 23.5% to 66.1% in 1982-1992 (``Cancer Survival in Developing Countries,'' International Agency for Research on Cancer, World Health Organization, Publication No. 145, 1998). We also have some information on annual bladder cancer mortality and incidence for the general population of Taiwan in 1996. The age- adjusted annual incidence rates of bladder cancer for males and females, respectively, were 7.36 and 3.09 per 100,000, with corresponding annual mortality rates of 3.21 and 1.44 per 100,000 (correspondence from Chen to Herman Gibb, January 3, 2000). Assuming that the proportion of males and females in the population is equal, these numbers imply that the mortality rate for bladder cancer in the general population of Taiwan, at present, is 45%. Since survival rates have most likely improved over the years since the original Taiwanese study, this number represents a lower bound on the survival rate for the original area under study (that is, one would not expect a higher rate of survival in that area at that time). This has implications for the bladder cancer risk estimates from the Taiwan data. For this estimate we have made the assumption that all bladder tumors in the study area in Taiwan were fatal. If there were any persons with bladder cancer who recovered and died from some other cause, then our estimate underestimated risk; that is, there were more cancer cases than cancer deaths. Based on the above discussion, we think bladder cancer incidence could be no more than 2 fold bladder cancer mortality; and that an 80% mortality rate would be plausible. In the benefits analysis we include estimates using an assumed mortality rate ranging from 80% to 100%. In the U.S. approximately one out of four individuals who is diagnosed with bladder cancer actually dies from bladder cancer. The mortality rate for the U.S. is taken from a cost of illness study recently completed by EPA (US EPA, 1999a). For those diagnosed with bladder cancer at the average age of diagnosis (70 years), the probability for dying of that disease during each year post-diagnosis were summed over a 20-year period to obtain the value of 26 percent. Mortality rates for U.S. bladder cancer patients have decreased overall by 24 percent from 1973 to 1996. In the NRC report, Table 10-11 shows excess risk estimates based on the Taiwanese male bladder cancer, using a Poisson regression model; a risk at the current MCL of 50 µg/L is in the range of 1 to 1.347 per 1,000. Table 10-12 presents excess lifetime risk estimates for bladder cancer in males calculated using EPA's 1996 proposed revisions to the cancer guidelines (US EPA 1996b). EPA selected four of these distributions as representative of the risks and uncertainty involved (selecting relatively high and relatively low estimates). These distributions (mean 1.049, 95% upper confidence limit 1.347; mean 0.731, 95% upper confidence limit 0.807; mean 1.237, 95% upper confidence limit 1.548; and mean 1.129, 95% upper confidence limit 1.229), were used in the EPA Monte Carlo simulations. All of these risk distributions are linear in the mean, and thus may be conservative assumptions, as the NRC report suggested the true relationship may be sublinear. If the true relationship is sublinear, i.e., lower than the straight line from 50 µg/L to zero, the true risks at levels below 50 µg/L are being overestimated. Other factors which might lower the true risk include the use of grouped data, the high Taiwanese dietary intake of arsenic, and the amount of selenium in the Taiwanese diet. NRC concluded that the present MCL in drinking water of 50 µg/L does not achieve EPA's goal for public health and requires downward revision. EPA did not request nor did NRC recommend a specific new MCL level. 6. Estimated Risk Reductions Estimated risk reductions for bladder cancer at various MCL levels were developed using Monte Carlo simulations. The inputs to the simulations were the distributions of relative risk factors (described in section X.A.4.), distributions of occurrence for arsenic levels at 3 µg/L and above, and bladder cancer risk distributions from the National Research Council report. The relative risk factor and occurrence distributions represent primarily population and occurrence variability, while the cancer risk distributions represent primarily uncertainty about the true risk. Thus the combined distributions reflect both variability and uncertainty. These combined distributions provide our best estimate of the actual risks faced by the exposed population, including the percentiles of the population facing various levels of risk. Estimated risk reductions for bladder cancer at various MCL levels are shown in Tables X-2a and X-2b. Table X-2a uses data on community water consumption from the new EPA study; Table X-2b uses data on total water consumption from the study. Populations at or above 10 -\4\ risk levels are shown in Tables X-3a and X-3b. [[Page 38943]] The after treatment occurrence distributions were assumed to reflect treatment to 80% of the MCL level. The latter assumption is made since water systems tend to treat below the MCL level in order to provide a margin of safety. As shown in Table X-2a, bladder cancer risks at the 90th percentile of water intake, for the various MCL options under consideration, range from a multiple of 10-5 at 3 µg/L (4-6 x 10-5) to a multiple of 10-4 at 20 µg/L (1.2-2.4 x 10-4). At 5 µg/L , the 90th percentile level is 6-11 x 10-5; at 10 µg/L the 90th percentile is 1.0-1.7 x 10-4. Table X-2b presents similar information. The risk estimates in Table X-2b are somewhat higher than those in Table X-2a because total water consumption is higher than community water consumption. Since there is uncertainty about these numbers, it is assumed that the range 1-1.5 x 10-4 represents a risk level of essentially 10-4. It is then assumed that risks above 1.5 x 10-4 represent risks greater than 10-4. Table X-3a gives information about percentages of the exposed populations and the number of people exposed at 10-4 risk levels and above, and, using the stated definition for an over 10-4 risk level, above 10-4. The numbers in this table show that at an MCL of 3 µg/L, only a small number (not quantifiable) face a risk level of greater than 10-4. At an MCL of 5 µg/L, about 0.3 to 0.8 million face such risk levels, at an MCL of 10 µg/L, 0.8 to 4 million, and at an MCL of 20 µg/L, about 2.4 to 6.4 million would be at such levels. Table X-3b gives similar information using total water consumption data. The mean bladder cancer risks for the exposed population at the various MCL options, after treatment, are shown in Tables X-4a and X-4b. These mean risks are used in the computation of the number of cases avoided, used later in the benefits evaluation section. Table X-2A.--Bladder Cancer Incidence Risks \1\ for High Percentile U.S. Populations Exposed at or Above MCL Options, After Treatment \2\ (Community Water Consumption Data \3\) ---------------------------------------------------------------------------------------------------------------- MCL (µg/L) 85th 90th 95th ---------------------------------------------------------------------------------------------------------------- 3............................................. 3.2-5.4 x 10-5 4-6 x 10-5 4.3-7.5 x 10-5 5............................................. 5.3-9.3 x 10-5 6-11 x 10-5 7.5-13.0 x 10-5 10............................................ .88-1.49 x 10-4 1.0-1.7 x 10-4 1.26-2.12 x 10-4 20............................................ 1.2-1.96 x 10-4 1.4--2.4 x 10-4 1.9-3.2 x 10-4 ---------------------------------------------------------------------------------------------------------------- \1\ See Sections III.C. and D. for a description of other health effects, and Section X.B. for ``What-if?'' estimates of magnitude for lung cancer risks. \2\ The bladder cancer risks presented in this table provide our ``best'' estimates at this time. Actual risks could be lower, given the various uncertainties discussed, or higher, as these estimates assume a 100% mortality rate. An 80% mortality rate is used in the computation of upper bound benefits. \3\ Discussed in Section X.A.2. Table X-2B.--Bladder Cancer Incidence Risks \1\ for High Percentile U.S. Populations Exposed at or Above MCL Options, After Treatment \2\ (Total Water Consumption Data \3\) ---------------------------------------------------------------------------------------------------------------- MCL (µg/L) 85th 90th 95th ---------------------------------------------------------------------------------------------------------------- 3............................................. 3.8-6.4 x 10-5 4-7 x 10-5 5-8.7 x 10-5 5............................................. 6.3-10.5 x 10-5 7-12 x 10-5 8.5-14.5 x 10-5 10............................................ 1.02-1.8 x 10-4 1.2-2.0 x 10-4 1.39-2.56 x 10-\4\ 20............................................ 1.4-2.34 x 10-4 1.7-2.8 x 10-4 2.17-3.56 x 10-4 ---------------------------------------------------------------------------------------------------------------- \1\ See Sections III.C. and D. for a description of other health effects, and Section X.B. for ``What-if?'' estimates of magnitude for lung cancer risks. \2\ The bladder cancer risks presented in this table provide our ``best'' estimates at this time. Actual risks could be lower, given the various uncertainties discussed, or higher, as these estimates assume a 100% mortality rate. An 80% mortality rate is used in the computation of upper bound benefits. \3\ Discussed in Section X.A.2. Table X-3A.--Percent of Exposed Population at 10-4 Risk or Higher for Bladder Cancer Incidence \1\ After Treatment \2\ (Community Water Consumption Data \3\) ---------------------------------------------------------------------------------------------------------------- Population at Percent at 10- 10-4 risk or Percent over Population MCL (µg/L) 4 risk or higher 10-4 * over 10-4 higher (millions) (millions) ---------------------------------------------------------------------------------------------------------------- 3............................................... 1-2.6 0.3-0.7 1 5............................................... 1.5-12 0.4-3.2 1-3 0.3-0.8 10.............................................. 11-34 2.9-9.1 3-15 0.8-4 20.............................................. 19.5-41 5.2-11 9-24 2.4-6.4 ---------------------------------------------------------------------------------------------------------------- \1\ See Sections III.C. and D. for a description of other health effects, and Section X.B. for ``What-if?'' estimates of magnitude for lung cancer risks. \2\ The percents presented in this table provide our ``best'' estimates at this time. Actual percents could be lower, given the various uncertainties discussed, or higher, as these estimates assume a 100% mortality rate. An 80% mortality rate is used in the computation of upper bound benefits. \3\ Discussed in Section X.A.2. * Where over 10-4 means 1.5 x 10-4 or above. Too low to calculate. [[Page 38944]] Table X-3B.--Percent of Exposed Population at 10-4 Risk or Higher for Bladder Cancer Incidence \1\ After Treatment \2\ (Total Water Consumption Data \3\) ---------------------------------------------------------------------------------------------------------------- Population at Percent at 10- 10-4 risk or Percent over Population MCL (µg/L) 4 risk or higher 10-4 * over 10-4 higher (millions) (millions) ---------------------------------------------------------------------------------------------------------------- 3............................................... 1-3 0.3-0.8 1 5............................................... 3-18 0.8-4.8 1-4 0.3-1.1 10.............................................. 16-50 4.3-13.4 4-23 1.1-6.2 20.............................................. 26-53 7-14.2 13-33 3.5-8.9 ---------------------------------------------------------------------------------------------------------------- \1\ See Sections III.C. and D. for a description of other health effects, and Section X.B. for ``What-if?'' estimates of magnitude for lung cancer risks. \2\ The percents presented in this table provide our ``best'' estimates at this time. Actual percents could be lower, given the various uncertainties discussed, or higher, as these estimates assume a 100% mortality rate. An 80% mortality rate is used in the computation of upper bound benefits. \3\ Discussed in Section X.A.2. * Where over 10-4 means 1.5 x 10-4 or above. Too low to calculate. Table X-4A.--Mean Bladder Cancer Incidence Risks 1 for U.S. Populations Exposed at or Above MCL Options, After Treatment 2 (Community Water Consumption Data 3) ------------------------------------------------------------------------ MCL (µ/L) Mean exposed population risk ------------------------------------------------------------------------ 3................................... 2.1-3.6 x 10-5 5................................... 3.6-6.1 x 10-5 10.................................. 5.5-9.2 x 10-5 20.................................. 6.9-11.6 x 10-5 ------------------------------------------------------------------------ 1 See Sections III.C. and D. for a description of other health effects, and Section X.B. for ``What-if?'' estimates of magnitude for lung cancer risks. 2 The bladder cancer risks presented in this table provide our ``best'' estimates at this time. Actual risks could be lower, given the various uncertainties discussed, or higher, as these estimates assume a 100% mortality rate. An 80% mortality rate is used in the computation of upper bound benefits. 3 Discussed in Section X.A.2. Table X-4B.--Mean Bladder Cancer Incidence Risks 1 for U.S. Populations Exposed at or Above MCL Options, After Treatment 2 (Total Water Consumption Data 3) ------------------------------------------------------------------------ MCL (µ/L) Mean exposed population risk ------------------------------------------------------------------------ 3................................... 2.6-4.5 x 10-5 5................................... 4.4-7.5 x 10-5 10.................................. 6.7-11.4 x 10-5 20.................................. 8.4-13.9 x 10-5 ------------------------------------------------------------------------ 1 See Sections III.C. and D. for a description of other health effects, and Section X.B. for ``What-if?'' estimates of magnitude for lung cancer risks. 2 The bladder cancer risks presented in this table provide our ``best'' estimates at this time. Actual risks could be lower, given the various uncertainties discussed, or higher, as these estimates assume a 100% mortality rate. An 80% mortality rate is used in the computation of upper bound benefits. 3 Discussed in Section X.A.2. B. ``What if?'' Scenario for Lung Cancer Risks The NRC report ``Arsenic in Drinking Water'' states that ``some studies have shown that excess lung cancer deaths attributed to arsenic are 2-5 fold greater than the excess bladder cancer deaths (NRC, 1999, pg. 8).'' Two-to-five fold greater would be 3.5 fold greater on average. Also in the U.S. the mortality rate from bladder cancer is 26% and the mortality rate of lung cancer is 88%. This suggests that if the risk of contracting lung cancer were identical to the risk of contracting bladder cancer, one would expect 3.4 times the number of deaths from lung cancer as from bladder cancer. Since these numbers are essentially the same, it seems reasonable to assume that the risk of contracting lung cancer is essentially the same as the rate of contracting bladder cancer,\1\ in the context of this ``what-if'' scenario. If the risk of contracting lung cancer from arsenic in drinking water is approximately equal to the risk of contracting bladder cancer, then the combined risk estimates of contracting either bladder or lung cancer would be approximately double the risk estimates presented in the previous tables. --------------------------------------------------------------------------- \1\ If ``X'' is the probability of contracting bladder cancer, then 0.26X is the probability of mortality from bladder cancer. If lung cancer deaths are 2 to 5 times as high as bladder cancer, then they are, on average, 3.5 times as high and the average probability of mortality from lung cancer would be 3.5 times 0.26X, or 0.91X. Since we also know that there is a 88% mortality rate from lung cancer, then if the probability of contracting lung cancer is ``Y,'' the probability of mortality from lung cancer can also be represented as 0.88Y. Setting the two ways of deriving the probability of mortality from lung cancer equal, or 0.91X = 0.88Y, one can solve for Y (Y= (0.91/0.88) X). Thus Y is approximately equal to X, and the rate of contracting lung cancer is approximately the same as the rate of contracting bladder cancer. --------------------------------------------------------------------------- EPA anticipates that a peer-reviewed quantification of lung cancer risk from arsenic exposure may be available between the time of proposal and promulgation. If so, EPA will make this information available for public comment through a Notice of Data Availability (NODA) and consider the analysis and public comment for the final rulemaking. C. Evaluation of Benefits The evaluation stage in the analysis of risk reductions involves estimating the value of reducing the risks. Background information on the economic concepts that provide the foundation for benefits valuation, and the methods that are typically used by economists to monetize the value of risk reductions, such as wage-risk, cost of illness, and contingent valuation studies are provided in the RIA. The following sections describe the use of these techniques to estimate the value of the risk reductions attributable to the regulatory options for arsenic in drinking water. Described first is the approach for valuing the reductions in fatal risks; described next is the approach for valuing the reductions in nonfatal risks. The benefits calculated for this proposal are assumed to begin to accrue on the effective date of the rule and are based on a calculation referred to as the ``value of a statistical life'' (VSL), currently estimated at $5.8 million. The VSL is an average estimate derived from a set of 26 studies estimating what people are willing to pay to avoid the risk of premature mortality. Most of these studies examine willingness to pay in the context of voluntary acceptance of higher risks of immediate accidental death in the workplace in exchange for higher wages. This value is sensitive to differences in population characteristics and perception of risks being valued. For the present rulemaking analysis, which evaluates reduction in premature mortality due to carcinogen exposure, some have argued that the Agency [[Page 38945]] should consider an assumed time lag or latency period in these calculations. Latency refers to the difference between the time of initial exposure to environmental carcinogens and the onset of any resulting cancer. Use of such an approach might reduce significantly the present value estimate. EPA is interested in receiving comments on the extent to which the presentation of more detailed information on the timing of cancer risk reductions would be useful in evaluating the benefits of the proposed rule. Latency is one of a number of adjustments or factors that are related to an evaluation of potential benefits associated with this rule, how those benefits are calculated, and when those economic benefits occur. Other factors which may influence the estimate of economic benefits associated with avoided cancer fatalities include (1) a possible ``cancer premium'' (i.e., the additional value or sum that people may be willing to pay to avoid the experiences of dread, pain and suffering, and diminished quality of life associated with cancer- related illness and ultimate fatality); (2) the willingness of people to pay more over time to avoid mortality risk as their income rises; (3) a possible premium for accepting involuntary risks as opposed to voluntary assumed risks; (4) the greater risk aversion of the general population compared to the workers in the wage-risk valuation studies; (5) ``altruism'' or the willingness of people to pay more to reduce risk in other sectors of the population; and (6) a consideration of health status and life years remaining at the time of premature mortality. Use of certain of these factors may significantly increase the present value estimate. EPA therefore believes that adjustments should be considered simultaneously. The Agency also believes that there is currently neither a clear consensus among economists about how to simultaneously analyze each of these adjustments nor is there adequate empirical data to support definitive quantitative estimates for all potentially significant adjustment factors. As a result, the primary estimates of economic benefits presented in the analysis of this proposed rule rely on the unadjusted estimate. However, EPA solicits comment on whether and how to conduct these potential adjustments to economic benefits estimates together with any rationale or supporting data commenters wish to offer. Because of the complexity of these issues, EPA will ask the Science Advisory Board (SAB) to conduct a review of these benefits transfer issues associated with economic valuation of adjustments in mortality risks. Consistent with the recommendations of the SAB, and subject to resolution of any technical problems, EPA will attempt to develop and present an estimate of the latency structure as a part of the analysis of the final rule, with prior solicitation of comment, if appropriate. 1. Fatal Risks and Value of a Statistical Life (VSL) To estimate the monetary value of reduced fatal risks (i.e., risks of premature death from cancer) predicted under different regulatory options, value of a statistical life (VSL) estimates are multiplied by the number of premature fatalities avoided. VSL does not refer to the value of an identifiable life, but instead to the value of small reductions in mortality risks in a population. A ``statistical'' life is thus the sum of small individual risk reductions across an entire exposed population. For example, if 100,000 people would each experience a reduction of 1/100,000 in their risk of premature death as the result of a regulation, the regulation can be said to ``save'' one statistical life (i.e., 100,000 x 1/100,000). If each member of the population of 100,000 were willing to pay $20 for the stated risk reduction, the corresponding value of a statistical life would be $2 million (i.e., $20 x 100,000). VSL estimates are appropriate only for valuing small changes in risk; they are not values for saving a particular individual's life. Of the many VSL studies, the Agency recommends using estimates from 26 specific studies that have been peer reviewed and extensively reviewed within the Agency (US EPA, 1997f). These estimates, which are derived from wage-risk and contingent valuation studies, range from $0.7 million to $16.3 million and approximate a Wiebull distribution with a mean of $5.8 million (in 1997 dollars). To value the changes in fatal risks associated with the arsenic regulation, the mean estimate of $5.8 million is used. Use of these estimates to value the averted risks of premature death associated with the regulatory options for arsenic is an example of the benefit transfer technique, since the subject of most of the studies (i.e., job-related risks) differs from the fatal cancer risks averted by the regulatory options. Applying these studies results in several sources of potential bias (see latency discussion in section X.C.); however, quantitative adjustments to address these biases generally have not been developed or adequately tested and may be counterbalancing.\2\ EPA notes the uncertainties in the cost-benefit analyses, as required by section 1412(b)(3)(C)(i)(VII) of SDWA, and requests comment on alternate approaches. --------------------------------------------------------------------------- \2\ Some of the key sources of bias include the characteristics of the averted risks (whether they are voluntary or involuntary, ordinary or catastrophic, delayed or immediate, natural or man-made, etc.); the demographic characteristics of the group affected (e.g., age, income); the lag between exposure and diagnosis or incidence of the disease (latency) as well as between incidence and death; the baseline health status (i.e., whether a person is currently in good health) of affected individuals; and the presence of altruism (i.e., individual's willingness to pay to reduce risks incurred by others) (US EPA, 1997f). --------------------------------------------------------------------------- 2. Nonfatal Risks and Willingness To Pay (WTP) Estimates of the willingness to pay to avoid treatable, nonfatal cancers are the ideal economic measures used for evaluation of the reduction in nonfatal risks. However this information is not available for bladder cancer. Willingness to pay (WTP) data to avoid chronic bronchitis is available, however, and has been used before by EPA (the microbial/disinfection by-product (MDBP) rulemaking) as a surrogate to estimate the WTP to avoid non-fatal bladder cancer. The use of such WTP estimates is supported in the SDWA, as amended, at section 1412(b)(3)(C)(iii): ``The Administrator may identify valid approaches for the measurement and valuation of benefits under this subparagraph, including approaches to identify consumer willingness to pay for reductions in health risks from drinking water contaminants.'' The WTP central tendency estimate of $536,000, to avoid chronic bronchitis, is used to monetize the benefits of avoiding non-fatal bladder cancers (Viscusi et al., 1991). EPA has also developed cost of illness estimates for bladder cancer, as reported in Table X-5. These estimates of direct medical costs are derived from a study conducted by Baker et al., (as cited in US EPA, 1997f) which uses data from a sample of Medicare records for 1974-1981. These data include the total charges for inpatient hospital stays, skilled nursing facility stays, home health agency charges, physician services, and other outpatient and medical services. EPA combined these data with estimates of survival rates and treatment time periods to determine the average costs of initial treatment and maintenance care for patients who do not die of the disease. [[Page 38946]] Table X-5.--Lifetime Avoided Medical Costs for Survivors (Preliminary Estimates, 1996 Dollars \1\) ---------------------------------------------------------------------------------------------------------------- Date data Number of cases Estimated survival Mean value per Type of cancer collected studied rate nonfatal case \2\ ---------------------------------------------------------------------------------------------------------------- Bladder.......................... 1974-1981 5% of 1974 Medicare 26 percent (after 20 $179,000 (for patients (sample years). typical individual from national diagnosed at age statistics). 70) ---------------------------------------------------------------------------------------------------------------- \1\ These costs increase by 2.8 percent when inflated to 1997 dollars, based on the consumer price index for the costs of medical commodities and services. \2\ Undiscounted costs. Source: US EPA, 1999a. D. Estimates of Quantifiable Benefits of Arsenic Reduction Benefits estimates for avoided cases of bladder cancer were calculated using mean population risk estimates at various MCL levels. Table X-6 gives the mean populations risk estimates used, which are a composite of the mean population risk estimates discussed earlier. Lifetime risk estimates were converted to annual risk factors, and applied to the exposed population to determine the number of cases avoided. These cases were divided into fatalities and non-fatal cases avoided, based on survival information. The avoided premature fatalities were valued based on the VSL estimates discussed earlier, as recommended by EPA current guidance for cost/benefit analysis. The avoided non-fatal cases were valued based on the willingness to pay estimates for the avoidance of chronic bronchitis. The upper bound estimates have been adjusted upwards to reflect an 80% mortality rate, which is a plausible mortality rate for the area of Taiwan during the Chen study. The ``What if?'' scenario for lung cancer benefits (described in section X.B.) was used to estimate potential benefits for avoided cases of lung cancer. This scenario is based on the statement in the NRC report ``Arsenic in Drinking Water'' that ``some studies have shown that excess lung cancer deaths attributed to arsenic are 2-5 fold greater than the excess bladder cancer deaths (NRC, 1999, pg. 8).'' It was shown in section X.D that the statement implies (if it were accurate for the U.S.), that, because of the relative U.S. mortality rates for bladder and lung cancer, the rate of contracting lung cancer could be essentially the same as the rate of contracting bladder cancer. This would double the number of cancer cases avoided, for both low and high estimates. The potential monetized benefits for lung cancer would be several times higher than those for bladder cancer, due to the higher number of fatalities involved with lung cancer. Another way of considering the addition of lung cancer effects would be to estimate the potential benefits from avoided cases of lung cancer using the 2-5 times range for fatalities (that is, taking the expected number of bladder cancer fatalities and multiplying them by 2 and then 5 to obtain a range of lung cancer fatalities, and then factoring in non-fatal cases). Benefits (and costs) are assumed to accrue on the effective date of the rule. Table X-7 displays the results. Table X-6.-Mean Bladder Cancer Incidence Risks \1\ for U.S. Populations Exposed at or Above MCL Options, After Treatment \2\ (Composite of Tables X-5a and X-5b) ------------------------------------------------------------------------ Mean exposed MCL (µg/L) population risk ------------------------------------------------------------------------ 3................................................. 2.1-4.5 x 10- 5 5................................................. 3.6-7.5 x 10- 5 10................................................ 5.5-11.4 x 10- 5 20................................................ 6.9-13.9 x 10- 5 ------------------------------------------------------------------------ \1\ See Sections III.C. and D. for a description of other health effects, and Section X.B. for ``What-if?'' estimates of magnitude for cancer risks. \2\ The bladder cancer risks presented in this table provide our ``best'' estimates at this time. Actual risks could be lower, given the various uncertainties discussed, or higher, as these estimates assume a 100% mortality rate. An 80% mortality rate is used in the computation of upper bound benefits. Table X-7.--Estimated Costs and Benefits From Reducing Arsenic in Drinking Water [Millions, 1999] -------------------------------------------------------------------------------------------------------------------------------------------------------- ``What if'' scenario \4\ and potential non-quantified benefits -------------------------------------------------------------------------- Total national Total national Total bladder ``What If'' Arsenic level (µg/l) costs to CWSs costs to CWSs cancer health lung cancer \1\ and NTNCWSs benefits \3\ health Potential non-quantifiable health benefits \2\ benefits estimates -------------------------------------------------------------------------------------------------------------------------------------------------------- 3............................ $643.1-753 $644.6-756.3 $43.6-104.2 $47.2-448
  • Skin Cancer. \5\(79) \6\(213.4)
  • Kidney Cancer.
  • Cancer of the Nasal Passages. 5............................ 377.3-441.8 378.9-444.9 31.7-89.9 35-384
  • Liver Cancer. \5\(64.3) \6\(173.4)
  • Prostate Cancer. 10........................... 163.3-191.8 164.9-194.8 17.9-52.1 19.6-224
  • Cardiovascular Effects. \5\(37) \6\(100)
  • Pulmonary Effects.
  • Immunological Effects.
  • Neurological Effects. [[Page 38947]] 20........................... 61.6-72.9 63.2-77.1 7.9-29.8 8.8-128
  • Endocrine Effects. \5\(19.8) \6\(53.4)
  • Reproductive and Developmental Effects. -------------------------------------------------------------------------------------------------------------------------------------------------------- \1\ Costs include treatment, monitoring, O&M, and administrative costs to CWSs and State costs for administration of water programs. The lower number shows costs annualized at a consumption rate of interest of 3%, EPA's preferred approach. The higher number shows costs annualized at 7%, which represents the standard discount rate preferred by OMB for benefit-cost analyses of government programs and regulations. \2\ Costs include treatment, monitoring, O&M, administrative costs to CWSs; monitoring and administrative costs to NTNCWSs; and State costs for administration of water programs. \3\ The upper bound estimate includes an adjustment to account for a possible mortality risk of 80%. It is possible that this risk could have been below 80%, which would lead to increased benefits. The actual risk depends on the survival rate for bladder cancer in the area of Taiwan studied by Chen, which is unknown. \4\ These estimates are based on the ``what if'' scenario for lung cancer, where the risks of a fatal lung cancer case associated with arsenic are assumed to be 2-5 times that of a fatal bladder cancer case. \5\The number in parentheses indicates the bladder cancer health benefits assuming an 80% mortality rate for bladder cancer in the area of the Chen study, and starting from the midpoint of the benefits range when mortality and incidence are assumed equivalent. \6\The number in parentheses is the midpoint of the range and corresponds to an assumption that the risk of fatal lung cancer is 3.5 times the risk of fatal bladder cancer. F. NDWAC Working Group (NDWAC, 1998) on Benefits The National Drinking Water Advisory Council (NDWAC) recommends that: (1) EPA should focus its benefits analysis efforts primarily on assessing effects on human health, defining these effects as clearly as possible and using the best available data to value them. It is also recommended that EPA should also consider, where appropriate, taste and odor improvements, reduction of damage to water system materials, commercial water treatment cost reductions, benefits due to source water protection (e.g., ecological benefits and non-use benefits), and benefits derived from the provision of information on drinking water quality (e.g., a household's improved ability to make informed decisions concerning the need to test or filter tap water); (2) EPA should devote substantial efforts to better understanding the health effects of drinking water contaminants, including the types of effects, their severity, and affected sensitive subpopulations. Better information is also needed on exposures and the effects of different exposure levels, particularly for contaminants with threshold effects. These efforts should pay particular attention to obtaining improved information concerning impacts on children and other sensitive populations; (3) EPA should clearly identify and describe the uncertainties in the benefits analysis, including descriptions of factors that may lead the analysis to significantly understate or overstate total benefits. Factors that may have significant but indeterminate effects on the benefits estimates should also be described; (4) EPA should consider both quantified and non-quantified benefits in regulatory decision-making. The information about quantified and non-quantified (qualitative) benefits should be presented together in a format, such as a table, to ensure that decision-makers consider both kinds of information; (5) EPA should consider incremental benefits and costs, total benefits and costs, the distribution of benefits and costs, and cost- effectiveness in regulatory decision-making. This information should be presented together in a format, such as a table, to ensure its consideration by decision-makers; (6) Whenever EPA considers regulation of a drinking water contaminant, it should evaluate and consider, along with water treatment requirements to remove a contaminant, source water protection options to prevent such a contaminant from occurring. The full range of benefits of those options should be considered. XI. Risk Management Decisions: MCL and NTNCWSs A. What Is the Proposed MCL? EPA is proposing an arsenic MCL of 5 µg/L and soliciting comments on options of 3µg/L, 10 µg/L, and 20 µg/L. EPA is also asking that commenters provide their rationale and any supporting data or information for the option they prefer. The SDWA generally requires that EPA set the MCL for each contaminant as close as feasible to the MCLG, based on available technology and taking costs to large systems into account. The 1996 amendments to the SDWA added the requirement that the Administrator determine whether or not the quantifiable and nonquantifiable benefits of an MCL justify the quantifiable and nonquantifiable costs based on the Health Risk Reduction and Cost Analysis (HRRCA) required under section 1412(b)(3)(C). The 1996 SDWA amendments also provided new discretionary authority for the Administrator to set an MCL less stringent than the feasible level if the benefits of an MCL set at the feasible level would not justify the costs (section 1412(b)(6)). This proposal to establish an MCL for arsenic of 5 µg/L is the first time EPA has invoked this new authority. In conducting this analysis, EPA considered all available scientific information concerning the health effects of arsenic, including various uncertainties in the interpretation of the results. As discussed in more detail below, an array of health endpoints of concern were considered in this analysis. For some of these, the risk can currently be quantified (i.e., expressed in numerical terms); and for some, it cannot. Similarly, there are a variety of health and other benefits attributable to reductions in levels of arsenic in drinking water, some of which can be monetized (i.e., expressed in monetary terms) and others that cannot yet be monetized. All were considered in this analysis. The array of factors taken into account in making risk management decisions for arsenic underscore the difficulty of recommending the most appropriate regulatory level. A detailed [[Page 38948]] discussion of each of the principal factors considered follows. 1. Feasible MCL Because arsenic is a carcinogen with no established mode of action, EPA is proposing that the MCLG be set at zero. To establish the MCL, EPA must first determine the level which is as close to this level as feasible. EPA has determined that 3 µg/L is technologically feasible for large systems based on peer-reviewed treatment information and the practical quantitation level achievable with available analytical methods. 2. Principal Considerations in Analysis of MCL Options In addition to the feasible MCL of 3 µg/L, the Agency evaluated MCL options of 5 µg/L, 10 µg/L, and 20 µg/L. EPA considered the health effects associated with arsenic, the risk levels to the population for these health effects that would remain after implementation, and the costs and benefits of the different options (both those that could be monetized and/or quantified now and those that could not). The Agency's assessment centered on the health risk posed by arsenic in drinking water as well as on the benefits and costs imposed by the options evaluated. These options were then analyzed, taking into consideration the uncertainties involved in each of these factors. EPA solicits public comment on all the factors it considered in making this decision. Estimates of risk levels to the population remaining after the regulation is in place provide a perspective on the level of public health protection and benefits. The SDWA clearly places a particular focus on public health protection afforded by MCLs. For instance, where EPA decides to use its discretionary authority after a determination that the benefits of an MCL would not justify the costs, section 1412(b)(6)(A) requires EPA to set the MCL at a level that ``maximizes health risk reduction benefits at a cost that is justified by the benefits.'' The SDWA requires the Agency to consider both quantifiable and nonquantifiable health risk reduction benefits (quantifiable benefits can include both those that are monetizable and those that are not). Non-monetizable benefits range from those about which some quantitative information is known (such as skin cancer), and those which are more qualitative in nature (such as some of the non-cancer health effects associated with arsenic). If additional potential benefits that are presently not monetized (see Table XI-1) could be estimated at some future point, the benefits might increase further. (Important assumptions inherent in EPA's benefits estimates, including the value of a statistical life and willingness to pay are discussed in section X.C.) EPA considered the relationship of the monetized benefits to the monetized costs for each option. While equality of monetized benefits and costs is not a requirement under section 1412(b)(6)(A), this relationship is still a useful tool in comparing costs and benefits. However, EPA believes that reliance on a simple arithmetic analysis of whether monetized benefits outweigh monetized costs is inconsistent with the HRRCA's instruction to consider both quantifiable and non- quantifiable costs and benefits. The Agency therefore believes it is necessary to also examine the qualitative and non-monetized benefits and consider these benefits in establishing the MCL. 3. Findings of NRC and Consideration of Risk Levels The Agency based its evaluation of the risk posed by arsenic at the MCL options of 3 µg/L, 5 µg/L, 10 µg/L and 20 µg/L on national and international research, the bladder cancer risk analysis provided by the National Research Council (NRC) report issued by the National Academy of Sciences (NRC 1999), and the NRC's qualitative statements of overall risk of combined cancers. The Agency is relying heavily on the findings of the NRC for a number of reasons. In carrying out its charge, the NRC assembled an independent body of preeminent scientists from several disciplines. This committee examined and carefully analyzed more information than has been available before, and NRC had the draft report peer reviewed by thirteen other individuals with ``diverse perspectives and technical expertise (NRC 1999b).'' EPA decided, in 1996, to charge the NRC with evaluating EPA's two risk assessments for arsenic and considering the most current national and international research on arsenic. The NRC determined that the current MCL of 50 µg/L is not adequately protective and should be revised downward as soon as possible. The NRC conducted a number of statistical analyses in making this determination. The report also recommended that EPA conduct separate analyses for ``bladder, lung, and other internal cancers,'' as well as consider the combined impact of these various health effects. Given the release date of the NRC report (March 1999) relative to the timing of the proposed rule and the additional analyses needed to definitively quantify all endpoints of concern, EPA chose to use NRC's bladder cancer analysis to quantify and monetize the bladder cancer risk for the proposed rule. NRC provided quantitative risk factors for bladder cancer, that, when combined with key risk characterization scenarios by EPA and qualitative benefits, yield risks and benefits associated with various possible MCL options. The NRC report also noted that lung cancer deaths due to arsenic could be 2 to 5 times higher than bladder cancer deaths, considering the frequency and incidence of cancers projected from international studies. However, the report did not provide a numeric risk-based quantification analysis for this judgment similar to that provided for bladder cancer. As noted in section X.E., EPA approximated the potential benefits of avoiding arsenic-related lung cancer by assuming that the probability of incidence of lung cancer is approximately equal to that of bladder cancer. One can then use the death rate associated with lung cancer (88% for lung cancer as compared to 26% for bladder cancer) to derive benefits and to consider the implications of this health endpoint on risk. The risk factors associated with various MCL options increase under this ``What If'' analysis, with 10 µg/L being on the upper end or just outside of the Agency's 1 x 10-4 risk range and more stringent MCL options being more solidly under this risk ceiling. EPA anticipates that a peer reviewed quantification of lung cancer risk from arsenic exposure may be available between the time of proposal and promulgation. If so, EPA will make this information available for public comment through a Notice of Data Availability (NODA) and consider the analysis and public comment for the final rulemaking. Individual risk varies widely depending on susceptibility, amount of drinking water consumption, dietary levels of arsenic, years of exposure, and other factors. Consequently, any single MCL does not provide the same level of protection to all individuals. While not required by statute, the Agency has historically set protectiveness levels within a risk range of 10-\4\ to 10-\6\. EPA has sought to ensure that drinking water standards were established at levels such that less than 10% of the exposed population faced a risk that exceeded the chosen risk level. This conclusion is based on a recognition of its responsibility to protect public health, together with its obligation to consider a range of risk management factors when establishing regulatory levels. [[Page 38949]] 4. Non-Monetized Health Effects There are a number of important non-monetized benefits that EPA considered in its analysis. Chief among these are certain health impacts known to be caused by arsenic (such as skin cancer). A number of epidemiologic studies conducted in several countries (e.g., Taiwan, Japan, England, Hungary, Mexico, Chile, and Argentina) report an association between arsenic in drinking water and skin cancer in exposed populations. Studies conducted in the U.S. have not demonstrated an association between inorganic arsenic in drinking water and skin cancer. However, these studies may not have included enough people in their design to detect these types of effects. There were also a large number of other health effects associated with arsenic, discussed in section III, and listed in Table XI.1, which are not monetized, due to lack of appropriate quantitative data. These health effects include other cancers such as prostate cancer and cardiovascular, pulmonary, neurological and other non-cancer endpoints. Other benefits not monetized for this proposal include customer peace of mind from knowing drinking water has been treated for arsenic and reduced treatment costs for currently unregulated contaminants that may be co-treated with arsenic. To the extent that reverse osmosis is used for arsenic removal, these benefits could be substantial. Reverse osmosis is the primary point of use treatment, and it is expected that very small systems will use this treatment to a significant extent. 5. Sources of Uncertainty Among the non-quantifiable factors EPA considered in choosing the proposed MCL was Congress' intent that EPA ``reduce * * * [scientific] uncertainty'' in promulgating the arsenic regulation, reflected in the 1412(b)(12) arsenic research plan provisions and the legislative history for the arsenic provision (S. Rep. 104-169, 104th Cong., 1st Sess. at 39-40). All assessments of risk are characterized by an amount of uncertainty. Some of this can be reduced by collecting more data or data of a different sort; for other types of uncertainty, improved data or assessment methods can allow one to define the degree to which an estimate is likely to be above or below the ``true'' risk. For the arsenic risk assessment, there are some definable sources of uncertainty. These include (but aren't limited to) the following: choice of endpoint and population; uncertainty about the exact exposure of individuals in the study population; issues on applying data from rural Taiwanese to the heterogenous population of the U.S.; the inability to know precisely how a chemical causes cancer in humans (the mode of action, which affects judgments as to the shape of the chemical's dose response curve at low doses); choice of mathematical modeling procedures. Congress established a dual path for arsenic in SDWA: on the one hand, EPA is to issue a proposed MCL in 4\1/2\ years; on a parallel track EPA is to develop a long-term research plan, complete the required consultations and peer reviews, complete the research, and fully consider the research results. While the plan has been developed and research is underway, not all research results will be available for the final rule. However, EPA did obtain through the NRC study the most authoritative review of existing scientific information available. This review examined the areas of uncertainty listed above. EPA considered uncertainty about arsenic's mode of action and the shape of the dose response curve below the observable range of data. EPA is proposing an MCLG of zero. This decision is supported by the NRC's findings that the dose-response relationship at low doses is uncertain and that a conservative, default assumption of linearity is advisable. (An assumption of linearity in the dose-response relationship implies that there is no ``safe'' level that can be identified at which no health effects are expected to occur.) However, the Agency also notes the NRC's conclusion that ``* * * a sublinear dose-response curve in the low dose range is predicted, although linearity cannot be ruled out.'' (NRC, 1999, pg. 6). EPA believes the NRC study's articulation of uncertainty about the shape of the dose- response curve below the observed health effect range is an important qualitative consideration and, given Congress' concern about scientific ``uncertainty'' in setting the arsenic level, guides EPA to a default assumption of linearity. The choice of one endpoint for risk assessment is a judgment call. While this choice is guided by the best available science, it introduces uncertainty. Basing the risk assessment on incidence of bladder tumors will underestimate the combined risk of all arsenic- induced health effects. Section XI.A.4. discusses how assessments of other tumor types and health endpoints would result in a higher estimate of arsenic risk. Another source of uncertainty is in the application of data from one human population to another. EPA believes that the differences in dietary contributions of arsenic that NRC identified in the Taiwan study population and the U.S. are important to consider and a source of uncertainty in interpreting the results. NRC estimated that daily inorganic arsenic intake from food in the U.S. ranges from 1.3 µg/day for infants, to 4.5 µg/day for males 14-16 years old and 5.2 µg/day for females 14-16 years old, to a maximum of 12.5 µg/day for 60-65 year-old males and 9.7 µg/day for 60-65 year old females. On the other hand, NRC cited a study (Schoof et al., 1998) that estimated the Taiwanese obtain 31 µg/day of inorganic arsenic from yams and 19 µg/day from rice, ``for a total of 50 µg/day within a range of estimates of 15-211 µg/day (NRC, 1999, pg. 51).'' NRC noted (p. 24) that ``Limited data on dietary arsenic intake in the blackfoot-disease region now available suggest that arsenic intake from food is higher in Taiwan than in the United States.'' NRC noted that EPA previously observed that arsenic intake from sources other than drinking water would overestimate the unit risk calculated from the Taiwan study (US EPA 1988, pg. 86). The report noted that improved quantification of arsenic in Taiwanese food might affect the risk assessment for arsenic in drinking water in the U.S. (NRC 1999, pg. 6). In addition, the NRC report discussed laboratory animal studies that indicated that selenium reduced the toxicity of arsenic. While there is no direct evidence for humans, NRC noted that ``Selenium status there [in Taiwan] should be considered a moderator of arsenic toxicity and taken into account when the Taiwanese data are applied to populations with adequate selenium intakes (NRC, 1999, pg. 240).'' The NRC report cited studies comparing urinary selenium concentrations and blood serum selenium concentrations; these were lower for the Taiwanese by comparison to other study populations including people in the U.S. NRC noted that the ``model choice can have a major impact on estimated low-dose risks when the analysis is based on epidemiological data (NRC 1999, pg. 294).'' NRC noted that EPA's 1988 risk assessment used the multistage Weibull model to estimate a lifetime skin cancer risk of 1 x 10-\3\ for U.S. males exposed to arsenic at 50 µg/L. In their report NRC discussed the implications (both in a general sense and specifically for the Tseng data) of using data from an ecological study, and of using grouped data. They also reported the results of applying both a multistage Weibull and a Poisson model. When they re-assorted data into varying exposure groups, there [[Page 38950]] was a strong effect on the fitted Weibull model. NRC concluded: ``Thus the fact that grouping does have a strong effect provides evidence of additional measurement error in the arsenic concentrations being assigned at the village level (NRC, 1999, pg. 284).'' NRC used median village arsenic concentrations to represent exposure levels. The Expert Panel (US EPA, 1997d) noted that biases from using average doses for groups leads to overestimation of risk. ``* * * [D]espite a distribution of doses in the population, those individuals exhibiting effects would tend also to be those who received the highest doses; because of this, deriving an average dose based on affected individuals would to some extent bias risk estimates upward. Similarly attribution of the total excess risk in the population to arsenic exposure alone could also be expected to inflate the estimate of risk if the population is also characterized by other risk factors such as smoking, excess exposure to sunlight, nutritional status, and so on (US EPA, 1997d, pg. 31).'' The Poisson model with a quadratic term for age and a linear term for exposure fit as well as the multistage Weibull model, and had less variability in risks from regrouping the exposure intervals. Results from the NRC Poisson model estimations were used in the EPA analysis of bladder cancer risks. NRC noted that ``Ecological studies in Chile and Argentina have observed risks of lung and bladder cancer of the same magnitude as those reported in the studies in Taiwan at comparable levels of exposure.'' This observation increases confidence in the risk estimates based on the Tseng data. That these populations are different in terms of ethnic background, dietary patterns, and potential for other exposures also decreases the level of concern about generalized applicability of the Taiwanese data for risk assessment. EPA considered these various uncertainties associated with interpretation of the health effects of arsenic in making risk management decisions and in selecting an appropriate regulatory level. The Agency requests comment on whether we have properly weighed the uncertainties which overestimate and underestimate risk of the proposed MCL. There is also a measure of uncertainty about the costs associated with various possible regulatory levels. EPA has provided its best estimates of the costs, but recognizes that a number of stakeholders have performed independent analyses suggesting that the costs may be higher than those estimated by EPA. EPA requests comment on its cost estimates and any additional information commenters may have on possible costs of the rule. 6. Comparison of Benefits and Costs The monetized costs and monetized benefits of the proposed rule, and the methodologies used to calculate them, are discussed in detail in sections IX, X, and XIII of this preamble and in the HRRCA. Overall estimates of monetized costs and monetized benefits associated with various MCL options are provided in Table XI-1. There are also many health effects which have not been monetized, as is also shown in Table XI-1. Table XI-1.--Estimated Costs and Benefits From Reducing Arsenic in Drinking Water [In 1999 $ millions] -------------------------------------------------------------------------------------------------------------------------------------------------------- ``What if'' scenario \4\ and potential non-quantified benefits -------------------------------------------------------------------------- Total national Total national Total bladder ``What if'' Arsenic level (µg/L) costs to CWSs costs to CWSs cancer health lung cancer \1\ and NTNCWSs benefits \3\ health Potential non-quantifiable health benefits \2\ benefits estimates -------------------------------------------------------------------------------------------------------------------------------------------------------- 3............................ 643.1-753 644.6-756.3 43.6-104.2 47.2-448
  • Skin Cancer. \5\ (79) \6\ (213.4)
  • Kidney Cancer.
  • Cancer of the Nasal Passages. 5............................ 377.3-441.8 378.9-444.9 31.7-89.9 35-384
  • Liver Cancer. \5\ (64.3) \6\ (173.4)
  • Prostate Cancer.
  • Cardiovascular Effects. 10........................... 163.3-191.8 164.9-194.8 17.9-52.1 19.6-224
  • Pulmonary Effects. \5\ (37) \6\ (100)
  • Immunological Effects.
  • Neurological Effects. 20........................... 61.6-72.9 63.2-77.1 7.9-29.8 8.8-128
  • Endocrine Effects. \5\ (19.8) \6\ (53.4)
  • Reproductive and Developmental Effects. -------------------------------------------------------------------------------------------------------------------------------------------------------- \1\ Costs include treatment, monitoring, O&M, and administrative costs to CWSs and State costs for administration of water programs. The lower number shows costs annualized at a consumption rate of interest of 3%, EPA's preferred approach. The higher number shows costs annualized at 7%, which represents the standard discount rate preferred by OMB for benefit-cost analyses of government programs and regulations. \2\ Costs include treatment, monitoring, O&M, administrative costs to CWSs; monitoring and administrative costs to NTNCWSs; and State costs for administration of water programs. \3\ The upper bound estimate includes an adjustment to account for a possible mortality risk of 80%. It is possible that this risk could have been below 80%, which would lead to increased benefits. The actual risk depends on the survival rate for bladder cancer in the area of Taiwan studied by Chen, which is unknown. \4\ These estimates are based on the ``what if'' scenario for lung cancer, where the risks of a fatal lung cancer case associated with arsenic are assumed to be 2-5 times that of a fatal bladder cancer case. \5\ The number in parentheses indicates the bladder cancer health benefits assuming an 80% mortality rate for bladder cancer in the area of the Chen study, and starting from the midpoint of the benefits range when mortality and incidence are assumed equivalent. \6\ The number in parentheses is the midpoint of the range and corresponds to an assumption that the risk of fatal lung cancer is 3.5 times the risk of fatal bladder cancer. 7. Conclusion and Request for Comment In summary, based on the NRC report, EPA agrees that the current MCL of 50 µg/L is too high and must be made more protective of human health. Because EPA is proposing an MCLG for arsenic of 0, the MCL must be set as close as feasible to the MCLG, unless EPA invokes its discretionary authority to set a different MCL at a level where the costs are justified by the benefits. EPA believes that the feasible level for arsenic is 3 µg/L. Today, EPA is proposing that the arsenic MCL be set at 5 µg/L. EPA believes that setting the MCL at 3 µg/L, the feasible level in this case, may not be justified at this time, given the uncertainty regarding the relationship between the monetized benefits and the monetized costs at that level, the current uncertainty of the non-monetized benefits, and the degree of scientific uncertainty regarding the dose-response curve for an MCL at that level (affected by differences in nutrition and arsenic from food). Because there is a substantial possible imbalance between currently estimated monetized costs and benefits at the [[Page 38951]] feasible level of 3 µg/L, and a lack of certainty concerning the non-monetized costs and potential non-monetized benefits, EPA is proposing a standard other than the feasible level, using its discretionary authority in section 1412(b)(6). (See Senate Rep. 104- 169, 104th Cong., 1st Sess. at 33). The statute requires that a level proposed or promulgated using this discretionary authority be one which maximizes health risk reduction at a level where the costs are justified by the benefits. EPA believes that the 5 µg/L MCL best meets this statutory test. EPA solicits comment on this finding, as described in more detail below. As discussed earlier in section XI.A.4., EPA believes that there are a number of not yet quantified adverse health effects that pose a significant risk to public health. While the relationship of actual monetized benefits to monetized costs at 5 µg/L, $31.7-$89.9 million for bladder cancer benefits (plus possible lung cancer benefits of $35-$384 million based on the ``What If'' scenario) vs. $378.9-444.9 million in costs, is uncertain. EPA believes the range of benefits supports that level, especially when there may potentially be substantial non-monetized benefits factored into the analysis. EPA believes that, given the guidance of the NRC report, these potential non-monetized benefits, including a number of non-cancer health effects (see Table XI-1), are substantial enough to strike a reasonable balance between benefits and costs. Strict parity of monetized costs and monetized benefits is not required to find that the benefits of a particular MCL option are justified under the statutory provisions of Sec. 1412(b)(6). In addition, at 5 µg/L, the remaining risks (of bladder cancer) to the exposed population after the rule's implementation are well within the 10-4 range, which is protective of public health. As a result, EPA finds that the actual risk levels (including risks of potential non-monetized health effects) at 5 µg/L are high enough to justify this MCL, and it is therefore the level which maximizes health protection at a level where the costs are justified. As discussed earlier, EPA has, as a matter of policy typically established MCLs for cancer-causing contaminants to ensure that the risks of excess cancer deaths represented by exposure to drinking water at the MCL over the course of a lifetime are within a range of one in 10,000 to one in 1,000,000. EPA believes that this range is reasonably protective of public health consistent with the goals of the Safe Drinking Water Act. In using its statutory discretion under section 1412(b)(6)(A) to set a standard less stringent than the feasible level that maximizes health risk reduction at a cost that is justified by the benefits, EPA is proposing that it should choose a level that falls within the afore-mentioned target risk range. EPA is proposing to stay within this risk range even if the monetized benefits of a standard set at the upper end of the range are below the costs, as may be the case with this rule. EPA believes that important factors in this evaluation are the considerable non-quantifiable benefits that may be attributable to the proposed MCL. EPA also notes, as discussed earlier, that Congress did not direct EPA to ensure strict equality of monetizable costs and benefits in applying its discretionary authorities under section 1412(b)(6)(A). EPA requests comments on its proposed use of the new authority under section 1412(b)(6)(A) of the SDWA. The risk assessment for bladder cancer indicates that a standard set at 10 µg/L would fall at the upper end of the target risk range, with 5 µg/L more solidly within that risk range. However, there are two important sets of considerations when using available health effects information and studies to help determine the appropriate level for a proposed new standard. On the one hand, multiple health endpoints are of concern in ensuring that the standard is adequately protective. As noted earlier, the NRC expresses concern about lung cancer and other health endpoints and indicated that excess lung cancer deaths from arsenic in drinking water could be 2-5 times the level of bladder cancer deaths. If these other risks were fully quantified, the total risk at 10 µg/L might be well above 1 x 10-4 (the upper end of the risk range), given that the quantified risk of bladder cancer alone appears to be at approximately this level. On the other hand, there is uncertainty in the quantification of bladder cancer risk (as well as other health endpoints) and this risk estimate includes a number of conservative assumptions, as discussed previously. These include the assumptions of using a linear dose- response function; the fact that the dose-response data from the Taiwan epidemiologic study are based upon grouped occurrence information from wells used by the study population; and the possibility that the study population was more susceptible to arsenic in drinking water (as compared to the U.S. population) due to the relatively high dietary intake and dietary deficiencies in other elements (e.g., selenium) that might mitigate the results of arsenic. Thus, the risk of bladder cancer alone might be well below current estimates which represent EPA's best estimate at this time using currently available data and standard methodologies. The proposed MCL attempts to balance these countervailing considerations in establishing a level that is protective of public health. Given these competing sources of uncertainty, EPA believes it is appropriate to propose a standard at 5 µg/L, because at this level it is more likely that the total risk would be within the target range than at a higher standard. However, between now and promulgation of the final rule, EPA will work to resolve as much of this uncertainty as possible, both in terms of quantifying risk of additional health endpoints (e.g., lung cancer) and in terms of reexamining conservative assumptions in the risk estimate. EPA requests comment on its proposed level of 5 µg/L and on its rationale for selecting this level. In selecting the final level of the standard, EPA will evaluate, in light of comments received and any new scientific information, its proposed way of using its discretionary authority under section 1412(b)(6)(A) and the total risk, costs, and benefits associated with each of the levels of the standard under consideration. EPA requests comment on other potential MCLs and which of the MCLs and rationales presented here best fits the statutory framework. First, EPA is requesting comment on setting the MCL at 10 µg/L. The monetized costs of $164.9-$194.8 million, and monetized benefits of $17.9-$52.1 million for bladder cancer (plus possible lung cancer benefits of $19.6-$224 based on the ``What If'' scenario) are closer at 10 µg/L. The risk levels (of bladder cancer) to the exposed population are within the 10-4 risk range, and the uncertainties already discussed in Section XI.A.6. may be a basis for inferring lower expected possible non-monetized benefits than assumed for the MCL option of 5 µg/L. EPA is also requesting comment on an MCL option of 20 µg/L. Some stakeholders favor an MCL in this range and cite, as justification for such a level, their belief that if all uncertainties are taken into consideration, risk estimates would be within the Agency's risk range of range of 1 x 10-6 to 1 x 10-4. As can be seen from Table XI-1, costs are considerably reduced at this level, since far fewer CWSs would be impacted (i.e., occurrence of arsenic, without treatment, is already below this level for many systems). Approximately 1,200 CWSs would be projected to incur costs of approximately $63-$77 million to [[Page 38952]] comply with an MCL of 20 µg/L. Benefits would also be considerably lower than for other options, at $7.9-$29.8 million for bladder cancer (plus possibly $8.8-$128 million for lung cancer, based on the ``What If'' scenario). EPA's principal concern with an MCL option in this range is that it may not be sufficiently protective after consideration of all health endpoints of concern. In other words, when the effects of bladder cancer, lung cancer, and skin cancer are considered, together with the various non-quantifiable endpoints such as circulatory system impacts, an MCL option of 20 µg/L could result in an unacceptably high risk, well outside of the risk range of 1 x 10-6 to 1 x 10-4. As noted above, in using its statutory discretion to set a standard above the feasible level, EPA is proposing not to set a standard that exceeds this target risk range. However, EPA solicits comment on an MCL option of 20 µg/L along with any supporting rationale that commenters wish to offer. EPA is also requesting comment on setting the MCL at 3 µg/ L. As explained in section XI.A.1., this is the level as close to the MCLG as is feasible. It is also the level at which the risks are most solidly within the 10-4 risk range of the three MCLs considered. If EPA were to set the MCL at this level, EPA would not use its discretionary authority to set the MCL at a less stringent level based on costs and benefits. The Agency estimates that the likelihood that actual monetized benefits of $43.6-$104.2 million for bladder cancer (plus possible lung cancer benefits of $47.2-$448 million based on the ``What If'' scenario), are close to monetized costs of $644.6- $756.3 million is less certain than at 5 µg/L. (See Table XI- 1.) While EPA believes that benefits may be substantially less than monetized costs for the feasible level, the feasible level would be the most protective of the options presented here and would conservatively account for the uncertainties about the severity of various health effects endpoints and their potential additive impacts. Finally, Congress indicated interest in assuring that EPA considered impacts of an MCL decision on people served by large systems who could afford protective MCLs and an MCL of 3 would respond to this interest. Section 1412(b)(6)(B), however, provides that the interests of people served by large systems are to be considered along with benefits and costs to systems not expected to get small system variances. Because this proposal does not include small system variance technologies (i.e., affordable technologies for small systems at the proposed MCL have been identified), the interests of persons served by large and small systems are being considered together and the provisions of section 1412(b)(6)(B) do not apply in this case. B. Why Is EPA Proposing a Total Arsenic MCL? The previous drinking water standard for arsenic of 0.05 mg/L was based on total arsenic. Total arsenic includes the dissolved and undissolved arsenic species present in drinking water and makes no distinction between inorganic or organic species. Consistent with the previous standard for arsenic, today's proposed regulation of 0.005 mg/ L will be based on total arsenic. From an occurrence and analytical methods standpoint, the Agency believes it is inappropriate to make a regulatory distinction between inorganic and organic arsenic forms in drinking water. According to Irgolic (1994) and as mentioned in section II.B, the inorganic arsenic species (As III and As V) are present in drinking water, and organic arsenic compounds are rarely found in water supplies. Furthermore, inorganic As V (arsenate) is more prevalent in drinking water supplies than inorganic As III (arsenite), which tends to occur in anaerobic waters. If organic species are present in drinking water, methylarsonic acid (MMA) and dimethylarsonic acid (DMA) are the predominant organic forms. These organic species, when present, can result from the leaching of arsenic-containing herbicides or from the conversion of the inorganic forms to the organic forms in the presence of microbial activity. In arsenic-rich ground water wells from Taiwan, methylated compounds were not present above concentrations of 1 µg/L. No DMA or MMA was detected in the ground water samples from six districts in West Bengal, India (Chatterjee et al.,1995). Regarding surface water, Anderson and Bruland (1991) reported that organic species (DMA and MMA) accounted for 1 to 59% of the total arsenic concentration from fourteen lake and river samples taken in California. As Irgolic pointed out in his review of the Anderson and Bruland study, the level of the organic arsenic found in these surface water samples were in the low nanomolar (nM or nm/L) range. After converting the reported units from nm/L to µg/L, analysis of the Anderson and Bruland data indicate that only two of the fourteen water samples exceeded a concentration of 1 µg/L of organic arsenic (DMA and MMA combined). There is currently no EPA approved method for arsenic analysis in drinking water that distinguishes inorganic arsenic species from organic arsenic forms. The method would need to meet the criteria listed in section VI.B. and would require interlaboratory studies for validation. The estimated costs of such an analytical method could range from $150 to $250 per analysis. In addition, laboratory capacity for this type of method would most likely be limited at this time. Few toxicity studies exist for organic arsenicals. The NRC report noted that methylated arsenic has less developmental toxicity than inorganic arsenic. Concentrations of DMA administered that decreased fetal weight produced over 50% maternal mortality in studies with rats and mice (Rogers et al., 1981 as reported in NRC, 1999); hamsters had no developmental toxicity from exposure to MMA nor DMA (Willhite, 1981, as reported in NRC, 1999). NRC noted that EPA has two unpublished studies of rats fed MMA which had some increase in thyroid tumors, but no effect on mice. In addition, MMA and DMA produced mutations in cells at concentrations over one thousand times higher than the concentrations of inorganic arsenite and arsenate (Moore et al., 1997 as reported in NRC, 1999). It takes roughly ten times more DMA than arsenite to cause chromosome changes in a human cell line (Oya-Ohata et al., 1996, as reported in NRC, 1999). Because of the limited occurrence of organic arsenic species in water and the lack of a suitable and widely available analytical method for inorganic arsenic, the Agency believes compliance with the proposed arsenic standard of 0.005 mg/L should be based on total arsenic. EPA requests comments on setting the MCL based on total arsenic and any data or established analytical methods that would support setting an MCL based on inorganic arsenic. C. Why Is EPA Proposing To Require Only Monitoring and Notification for NTNCWSs? In this rulemaking, the Agency is soliciting comment on an approach which would not extend coverage of the rule to Non-Transient Non- Community (NTNC) water systems, but would instead create an intermediate level of control for these systems (monitoring and notification requirements). The suggested approach would recognize the lower level of risk generally posed to individuals by these systems. Simultaneously, it would provide a mechanism for the public to be adequately informed in those situations where unusual concentrations of NTNC systems, customer overlap, and high [[Page 38953]] local arsenic water concentrations caused risk levels to more closely approach community water system levels. There are approximately 20,000 NTNCs water systems regulated under the Safe Drinking Water Act. By definition, these systems do not serve over 25 people as year round residents, as would be the case for a community water system. However, they must serve at least 25 of the same people for over six months out of the year, or they would be classified as Transient Non-Community (TNC) water systems. It is generally an important distinction since the Agency has not applied regulations for contaminants with chronic health effects to TNC water systems, while it often has regulated NTNC systems similar to community water systems when addressing the risks posed by chronic contaminants. In the case of arsenic, the existing regulation does not apply to NTNC systems. While it is feasible to control arsenic in NTNC water systems, extending regulation to these systems needs to be considered in light of the new SDWA requirement to determine whether the benefits extending coverage to this category would justify the costs and whether such regulation would provide a reasonable opportunity for health risk reduction. As discussed elsewhere in the preamble, this analysis requires a balancing of both quantitative and non-quantitative factors. Based on the modeling to be discussed, the ninetieth percentile lifetime risk of contracting bladder cancer posed to an individual consuming water from a NTNC water system, even in their present untreated state, does not exceed one in 100,000.\ 3\ As a consequence, costs per each bladder cancer case avoided at the proposed MCL would approach the fifty million dollar mark if coverage of the rule were extended to NTNCs. This level is well above the range of historical environmental risk management decisions. --------------------------------------------------------------------------- \3\ Throughout this discussion, exposures and risks were only considered for populations potentially addressable by regulation, i.e., systems with present arsenic levels in excess of 3 µg/ L. --------------------------------------------------------------------------- These much lower risk levels result because most individuals served by NTNC systems are expected to receive only a small portion of their lifetime drinking water exposure from such systems. For example, even with twelve years of perfect attendance at schools served by NTNC water systems, the water consumed by an individual student is estimated to represent less than five percent of lifetime consumption. On the other hand, there are some segments of the NTNC water system population where exposure is a more significant portion of the total lifetime exposure. Manufacturing and other workers, although they represent only five percent of the population served by NTNC systems, could receive twenty to forty percent of their lifetime exposure at work. Nevertheless, as manufacturing workers represent a small portion of the NTNC population, overall risks among the NTNC population are small. Another factor of potential concern is the extent to which users of the different NTNC water systems overlap. It is conceivable that some areas in the country exist where individuals are subjected to arsenic exposure at a number of different non-community systems (e.g., day care center plus school plus factory, etc.). In such circumstances, individuals would be exposed to proportionately higher risks if the water systems all had elevated arsenic levels. For some individuals, the exposure could approach levels observed in corresponding community water systems. This concern is alleviated by the fact that NTNC systems generally serve only a very small portion of the total population. For example, over ninety-five percent of all school children are served by community water systems. Only a small percentage are served by NTNC water systems and, of that group, only about twelve percent (or less than one half of one percent of the overall student population) would be expected to have arsenic in their water above the proposed regulatory level). Likewise, less than 0.1 percent of the work force population receive water from an NTNC water system. With such low portions of the total population exposed to any particular type of NTNC system, the overall likelihood of multiple exposure cases in the NTNC population should also be small. The groups have been treated independently for this analysis. Comment and data are solicited to support any alternative treatments of the exposure data. Finally, although the Agency does not believe there is sufficient evidence to support unusual sensitivity on the part of children, they generally do consume more water on a weight adjusted basis. For this reason, NTNC systems which were likely to pose the greatest exposure risk to children were separately examined and their higher relative doses considered in the modeling effort. All of these factors contributed to the Agency's evaluation of whether or not to extend regulation to NTNC water systems for arsenic and are discussed further in the results section. 1. Methodology for Analyzing NTNCWS Risks Determination of system and individual exposure factors--In the past, the Agency has directly used SDWIS population estimates for assessing the risks posed to users of NTNC water systems. In other words, it was assumed that the same person received the exposure on a year round basis. Under this approach it was generally assumed that all NTNC users were exposed for 270 days out of the year and obtained fifty percent of their daily consumption from these systems. TNC users were assumed to use the system for only ten days per year. With the recent completion of ``Geometries and Characteristics of Public Water Systems (US EPA, 1999e),'' however, the Agency has developed a more comprehensive understanding of NTNC water systems. These systems provide water in due course as part of operating another line of business. Many systems are classified as NTNC, rather than TNC, water systems solely because they employ sufficient workers to trigger the ``25 persons served for over six months out of the year'' requirement. Client utilization of these systems is actually much less and more similar to exposure in TNC water systems. For instance, it is fairly implausible that highway rest areas along interstate highways serve the same population on a consistent basis (with the exception of long distance truckers). Nevertheless, there are highway rest areas in both NTNC and TNC system inventories. The ``Geometries'' report suggests that population figures reported in SDWIS which have been used for past risk assessments generally appear to reflect the number of workers in the establishment coupled with peak day customer utilization. Under these conditions use of the SDWIS figures for population greatly overestimates the actual individual exposure risk for most of the exposed population and also significantly underestimates the number of people exposed to NTNC water.\4\ Adequately characterizing individual and [[Page 38954]] population risks necessitates some adjustments to the SDWIS population figures. For chronic contaminants, such as arsenic, health data reflect the consequences of a lifetime of exposure. Consequently, risk assessment requires the estimation of the portion of total lifetime drinking water consumption that any one individual would receive from a particular type of water system. In turn, one needs to estimate the appropriate portions for daily, days per year, and years per lifetime consumption. These estimates need to be prepared for both the workers at the facility and the ``customers'' of the facility. --------------------------------------------------------------------------- \4\ For example, airports constitute only about a hundred of the NTNC water systems. Washington's Reagan National and Dulles, Dallas/ Fort Worth, Seattle/Tacoma, and Pittsburgh airports are the five largest of the airports. SDWIS reports that these five airports serve about 300,000 people. In actuality, Bureau of Transportation Statistics suggest that they serve about eleven million passengers per year. Examination of this information and other BTS statistics suggests that these airports serve closer to seven million unique individuals over the course of a year and that exposure occurs on an average of ten times per year per individual customer, not 270 times. --------------------------------------------------------------------------- This adjustment was accomplished through a comprehensive review of government and trade association statistics on entity utilization by the U.S. Department of Commerce's Standard Industrial Classification (SIC) code. These figures, coupled with SDWIS information relating to the portion of a particular industry served by non-community water systems, made possible the development of two estimates needed for the risk assessment: customer cycles per year and worker per population served per day. These numbers are required to distinguish the more frequent and longer duration exposure of workers from that of system customers.\5\ A more detailed characterization of the derivation of these numbers is contained in the docket. Table XI-2 provides the factors used in the NTNC risk assessment to account for the intermittent nature of exposure. Comment is solicited on the appropriateness of the various factors. --------------------------------------------------------------------------- \5\ For example, travel industry statistics provide information on total numbers of hotel stays, vacancy rates, traveller age ranges, and average duration of stay. These figures can be combined with the SDWIS peak day population estimates to allocate daily population among workers, customers and vacancies. The combination of these factors provides an estimate of the number of independent customer cycles experienced in a year. --------------------------------------------------------------------------- Once the population adjustment factors were derived, it was possible to determine the actual population served by NTNC water systems. Table XI-3 provides a breakout of these figures by type of establishment. Although not included in Table XI-3, there are other equally important characteristics to note about these systems. With notable exceptions (such as the airports in Washington, DC and Seattle), the systems generally serve a fairly small population on any given day. In fact, 99 percent of the systems serve less than 3300 users on a daily basis. This means that water production costs will be relatively high on a per gallon basis. Risk calculation--Calculations of individual risk were prepared for each industrial sector. Even within a given sector, however, risk varies as a function of an individual's relative water consumption, body weight, vulnerability to arsenic exposure, and the water's arsenic concentration. Computationally, risks were estimated by performing Monte Carlo modeling, as was done in the community water system risk estimation, with two exceptions. First, each realization in a given sector was multiplied by the portion of lifetime exposure factor presented in Table XI-2 to reflect the decreased consumption associated with the NTNC system. Secondly, relative exposure factors were limited to age specific ratings where appropriate.\6\ For example, in the case of school children, water consumption rates and weights for six to eighteen year olds were used. --------------------------------------------------------------------------- \6\ For example, school kid water consumption was weighted to reflect consumption between ages 6 and 18, while factory worker consumption was weighted over ages 20 to 64. Table XI-2.--Exposure Factors Used in the NTNC Risk Assessment -------------------------------------------------------------------------------------------------------------------------------------------------------- Number of Worker Worker Customer Customer NTNCWS cycles per Worker/pop/ fraction Worker days/ exposure fraction Days of use/ exposure yr day daily yr years daily yr years -------------------------------------------------------------------------------------------------------------------------------------------------------- Water wholesalers............................. 1.00 0.000 0.25 270 70 Nursing homes................................. 1.00 0.230 0.50 250 40 1.00 365 10 Churches...................................... 1.00 0.010 0.50 250 40 0.50 52 70 Golf/country clubs............................ 4.50 0.110 0.50 250 40 0.50 52 70 Food retailers................................ 2.00 0.070 0.50 250 40 0.25 185 70 Non-food retailers............................ 4.50 0.090 0.50 250 40 0.25 52 70 Restaurants................................... 2.00 0.070 0.50 250 40 0.25 185 70 Hotels/motels................................. 86.00 0.270 0.50 250 40 1.00 3.4 40 Prisons/jails................................. 1.33 0.100 0.50 250 40 1.00 270 3 Service stations.............................. 7.00 0.060 0.50 250 40 0.25 52 54 Agricultural products/services................ 7.00 0.125 0.50 250 40 0.25 52 50 Daycare centers............................... 1.00 0.145 0.50 250 10 0.50 250 5 Schools....................................... 1.00 0.073 0.50 200 40 0.50 200 12 State parks................................... 26.00 0.016 0.50 250 40 0.50 14 70 Medical facilities............................ 16.40 0.022 0.50 250 40 1.00 6.7 10.3 Campgrounds/RV................................ 22.50 0.041 0.50 180 40 1.00 5 50 Federal parks................................. 26.00 0.016 0.50 250 40 0.50 14 70 Highway rest areas............................ 50.70 0.010 0.50 250 40 0.50 7.2 70 Misc. recreation service...................... 26.00 0.016 0.50 250 40 1.00 14 70 Forest Service................................ 26.00 0.016 1.00 250 40 1.00 14 50 Interstate carriers........................... 93.00 0.304 0.50 250 40 0.50 2 70 Amusement parks............................... 90.00 0.180 0.50 250 10 0.50 1 70 Summer camps.................................. 8.50 0.100 1.00 180 10 1.00 7 10 Airports...................................... 36.50 0.308 0.50 250 40 0.25 10 70 Military bases................................ 1.000 0.50 250 40 Non-water utilities........................... 1.000 0.50 250 40 Office parks.................................. 1.000 0.50 250 40 Manufacturing: Food........................... 1.000 0.50 250 40 Manufacturing: Non-food....................... 1.000 0.50 250 40 Landfills..................................... 1.000 1.00 250 40 Fire departments.............................. 1.000 1.00 250 40 [[Page 38955]] Construction.................................. 1.000 1.00 250 40 Mining........................................ 1.000 1.00 250 40 Migrant labor camps........................... 1.000 1.00 250 40 -------------------------------------------------------------------------------------------------------------------------------------------------------- Table XI-3.--Composition of Non-Transient, Non-Community Water Systems [Percentage of total NTNC population served by sector] -------------------------------------------------------------------------------------------------------------------------------------------------------- -------------------------------------------------------------------------------------------------------------------------------------------------------- Schools.............................. 9.7 Medical Facilities...... 8 Interstate Carriers..... 7.1 Campgrounds............. 1.3 Manufacturing........................ 2.7 Restaurants............. 0.9 State Parks............. 8.6 Misc Recreation......... 1.8 Airports............................. 26.1 Non-food Retail......... 1.6 Amusement Parks......... 17.7 Other................... 3.5 Office Parks......................... 0.6 Hotels/Motels........... 9.2 H'way Rest Area......... 1.0 -------------------------------------------------------------------------------------------------------------------------------------------------------- To illustrate the process, it was conservatively assumed that a child would attend only NTNC served schools for all twelve years. Further, it was assumed that a child would get half of their daily water consumption at school (for an average first grader this would correspond to roughly nine ounces of water per school day). Finally, it was assumed that the child would have perfect attendance and attend school for 200 days per year. Table XI-4 provides a sample output for the upper bound individual risk distribution to school children resulting from exposure to the range of untreated arsenic observed in community ground water systems \7\ as well as an estimate based on more moderate assumptions of four ounces per day and 150 days attendance for four years. Upper and lower bound risk distributions were prepared for both workers and ``customers'' at all types of NTNC water systems and are contained in the docket. --------------------------------------------------------------------------- \7\ Community ground water occurrence information was used since NTNC systems are almost exclusively supplied by ground water sources. Further, as there was no depth dependence of arsenic levels observed in the community information, it is believed that the data are an adequate approximation. Table XI-4.--Upper Bound School Children Risk Associated With Current Arsenic Exposure in NTNC Water Systems [Risks are per 10,000 students. i.e., x 10-4] ------------------------------------------------------------------------ Moderate exposure Upper bound scenario scenario ------------------------------------------------------------------------ Mean Lifetime Risk...................... 0.0087 0.079 90th Percentile Lifetime Risk........... 0.019 0.17 Lifetime Bladder Cancers in Student 0.5 4.5 Population............................. ------------------------------------------------------------------------ Note: This table does not include potential non-quantified lung or skin cancers. The distribution of population risks overall was determined as part of the same simulation by developing sector weightings to reflect the total portion of the NTNC population served by each sector. Population weighted proportional sampling of the individual sectors provided an overall distribution of risk among those exposed at NTNC systems. 2. Results It is important to note that the results presented in the discussion of NTNC benefits are based on the currently quantified health endpoint for arsenic related bladder cancer. As noted elsewhere in Section X of today's proposal, there are a number of health end points that have not yet been quantified and which could provide a rationale for extending coverage to NTNCs--in the event that a substantial portion of the consumers of water from such systems fall outside the 1 in 10,000 risk range frequently used by the Agency as a benchmark for such decisions. (Any additional data quantifying such endpoints would made available for public comment in a Notice of Data Availability.) Table XI-5 presents a summary of the Benefit Cost Analysis for all NTNC systems. As can be seen from a review of the Table, regulation of arsenic in NTNC water systems provides only very limited opportunity for national risk reduction. Table XI-6 presents risk figures for three particular sets of individuals: children in daycare centers and schools, and construction workers. Construction and other strenuous activity workers comprise an extremely small portion of the population served by NTNC systems (less than 0.1%), but face the highest relative risks of all NTNC users (90th percentile risks of 0.7 to 1.6 x 10-4 lifetime risk). Nevertheless, there is considerable uncertainty about these exposure numbers. It is quite likely that they overestimate consumption and may be revised downward by subsequent analysis (Any additional data quantifying such endpoints would made available for public comment in a Notice of Data Availability.). The risks for children are much lower with an upper bound, 90th percentile estimate of 1.7 x 10-5 lifetime risk. What is not possible to determine from the analysis of NTNC systems is the extent to which there is overlap of individual exposure between the various sectors. As mentioned earlier, NTNC establishments generally constitute a small portion of their SIC sectors. This fact and the observation that NTNC populations would only serve about one percent of the total population if all of the sectors with significant exposure (greater than five percent of lifetime) if they were [[Page 38956]] mutually exclusive,\ 8\ provide some support for treating the SIC groups independently. However, it is equally plausible that there are communities where one individual might go from an NTNC day care center to a series of NTNC schools and then work in an NTNC factory. --------------------------------------------------------------------------- \8\ This is considerably less than the estimated rural population in the U.S. which is the smallest group among which users of these systems would conceivably be distributed. --------------------------------------------------------------------------- The Agency is concerned about the potential for local issues to arise with respect to combined arsenic exposures. In the rare community where all ground water is contaminated with the highest levels of arsenic, risks could be outside of the Agency's traditionally allowable realm. Further, different levels of protection being provided by schools served by community water systems versus those served by NTNC systems could be seen as posing equity considerations for rural communities. For all of these reasons, the Agency does not believe it is appropriate to completely exempt NTNC systems from arsenic regulation. On the other hand, it does not believe an adequate basis exists to prescribe a standard. The Agency is proposing to take a somewhat different approach with respect to NTNC water systems than previously practiced. We are proposing that NTNC water systems be subject to arsenic monitoring requirements applicable to community water systems. When an individual NTNC system has arsenic present in excess of the MCL for community systems, it would be required to post a notice to customers as described in Section VII.I. of this rule. The Agency believes that this approach will provide localities with high arsenic concentrations the opportunity to limit their consumption of water from these systems. Because the NTNC is not the sole source of water available to these consumers as would be the case with a community water system, they would have the ability to use bottled water, or in the case of schools for instance, to install voluntary treatment to reduce their exposure. The Agency requests comment on this approach for addressing NTNC water systems as well as on two other possible approaches: exempting NTNC systems entirely from coverage under this rule or extending coverage to NTNC systems in the same manner as CWSs. EPA requests an accompanying rationale and any data commenters wish to submit as part of their comments on this topic. The Agency may decide, as part of the final rule, to incorporate any of these three approaches without further opportunity for comment (except where a NODA may be issued to provide the public with additional new information not taken into consideration for today's rulemaking). Table XI-5.--Non-Transient Non-Community Benefit Cost Analysis [All risk values are per 10,000-i.e., 10-\4\] -------------------------------------------------------------------------------------------------------------------------------------------------------- Untreated 10 5 3 --------------------------------------------------------------------------------------- MCL option Lower Upper Lower Upper Lower Upper Lower Upper bound bound bound bound bound bound bound bound -------------------------------------------------------------------------------------------------------------------------------------------------------- Mean Individual Risk............................................ 0.019 0.042 0.012 0.026 0.0077 0.017 0.0046 0.01 90th Percentile Individual...................................... 0.037 0.08 0.027 0.058 0.017 0.037 0.01 0.022 Annual Bladder Cancers.......................................... 0.427 0.95 0.265 0.583 0.16 0.36 0.101 0.215 Cancer Cases Avoided............................................ 0 0 0.162 0.367 0.267 0.59 0.326 0.735 Benefit Million Dollars......................................... 0 0 0.31 0.70 0.51 1.1 0.62 1.4 Cost Million Dollars............................................ ......... 0 ......... 6.121 ......... 14.69 ......... 25.21 -------------------------------------------------------------------------------------------------------------------------------------------------------- Note: This table does not include potential non-quantified lung cancer benefits. Table XI-6.--Sensitive Group Evaluation Lifetime Risks ------------------------------------------------------------------------ 90th percentile Group Mean risk risk ------------------------------------------------------------------------ Forest Service, Construction and 3.2-7 x 10-\5\ 7.2-16 x 10-\5\ Mining Workers................... School Children................... 3.8-7.9 x 10-\6\ 0.84-1.7 x 10-\5\ Day Care Children................. 3.4-6.8 x 10-\6\ 0.74-1.5 x 10-\5\ ------------------------------------------------------------------------ XII. State Programs A. How Does Arsenic Affect a State's Primacy Program? States must revise their programs to adopt any part of today's rule which is more stringent than the approved State program. Primacy revisions must be completed in accordance with 40 CFR 142.12, and 142.16. States must submit their revised primacy application to the Administrator for approval. State requests for final approval must be submitted to the Administrator no later than 2 years after promulgation of a new standard unless the State requests and is granted an additional 2-year extension. For revisions of State programs, Sec. 142.12 requires States to submit, among other things, ``[a]ny additional materials that are listed in Sec. 142.16 of this part for a specific EPA regulation, as appropriate (Sec. 142.12(c)(1)(ii)).'' Based on comments from stakeholders at the arsenic in drinking water regulatory development meetings held prior to proposal, EPA believes that the information required in Sec. 142.16(e) is not required for States revising the MCL for arsenic. Although that section refers to applications that adopt requirements of Secs. 141.11, 141.23, 141.32, and 141.62, EPA believes that existing State programs which contain the standardized monitoring framework for inorganic contaminants (40 CFR 141.23) can ensure all CWSs monitor for arsenic. Therefore, EPA is proposing to clarify that Sec. 141.16(e) applies only to new contaminants, not revisions of existing contaminants regulations. The Agency requests comment on whether this is an appropriate change. EPA believes that the requirements in Sec. 142.12(c) will provide sufficient information for EPA review of the State revision. The side- by-side comparison of requirements required in Sec. 142.12(c)(1)(i) will only consist of sections revised to adopt the changes required for the arsenic regulation and [[Page 38957]] any other revisions requested by the State. In addition, the Attorney General's statement required in Sec. 142.12(c)(1)(iii) will certify that the revised regulations will be effective and enforceable. The Agency requests comment on whether any other documentation is necessary to approve revisions to State programs enforcing the new arsenic regulation. The Agency is proposing to add Sec. 142.16(j) to clarify primacy requirements relating to monitoring plans and waiver procedures for revisions of existing monitoring requirements such as arsenic. Section 142.16(j) clarifies that the State simply needs to inform the Agency in their application of any changes to the monitoring plans and waiver procedures. Alternatively, a State may indicate in the primacy application that they will use the existing monitoring plans and waiver criteria approved for primacy under the National Primary Drinking Water Standards for other contaminants (for example, i.e. the Phase II/V rules). This information may be provided in the primacy application crosswalk which identifies revisions to the State primacy program. B. When Does a State Have To Apply? To maintain primacy for the Public Water Supply (PWS) program and to be eligible for interim primacy enforcement authority for future regulations, States must adopt today's rule, when final. A State must submit a request for approval of program revisions that adopt the revised MCL and implementing regulations within two years of promulgation unless EPA approved an extension per Sec. 142.12(b). Interim primacy enforcement authority allows States to implement and enforce drinking water regulations once State regulations are effective and the State has submitted a complete and final primacy revision application. To obtain interim primacy, a State must have primacy with respect to each existing NPDWR. Under interim primacy enforcement authority, States are effectively considered to have primacy during the period that EPA is reviewing their primacy revision application. C. How Are Tribes Affected? Currently, no federally recognized Indian tribes have primacy to enforce any of the drinking water regulations. EPA Regions implement the rules for all Tribes under section 1451(a)(1) of SDWA. Tribes must submit a primacy application to have oversight for the inorganic contaminants (i.e., the Phase II/V rule) to obtain the authority for the revised arsenic MCL. Tribes with primacy for drinking water programs are eligible for grants and contract assistance (section 1451(a)(3)). Tribes are also eligible for grants under the Drinking Water State Revolving Fund Tribal set aside grant program authorized by section 1452(i) for public water system expenditures. XIII. HRRCA A. What Are the Requirements for the HRRCA? Section 1412(b)(3)(C) of the 1996 Amendments requires EPA to prepare a Health Risk Reduction and Cost Analysis (HRRCA) in support of any NPDWR that includes an MCL. According to these requirements, EPA must analyze each of the following when proposing a NPDWR that includes an MCL: (1) Quantifiable and non-quantifiable health risk reduction benefits for which there is a factual basis in the rulemaking record to conclude that such benefits are likely to occur as the result of treatment to comply with each level; (2) quantifiable and non- quantifiable health risk reduction benefits for which there is a factual basis in the rulemaking record to conclude that such benefits are likely to occur from reductions in co-occurring contaminants that may be attributed solely to compliance with the MCL, excluding benefits resulting from compliance with other proposed or promulgated regulations; (3) quantifiable and non-quantifiable costs for which there is a factual basis in the rulemaking record to conclude that such costs are likely to occur solely as a result of compliance with the MCL, including monitoring, treatment, and other costs, and excluding costs resulting from compliance with other proposed or promulgated regulations; (4) the incremental costs and benefits associated with each alternative MCL considered; (5) the effects of the contaminant on the general population and on groups within the general population, such as infants, children, pregnant women, the elderly, individuals with a history of serious illness, or other subpopulations that are identified as likely to be at greater risk of adverse health effects due to exposure to contaminants in drinking water than the general population; (6) any increased health risk that may occur as the result of compliance, including risks associated with co-occurring contaminants; and (7) other relevant factors, including the quality and extent of the information, the uncertainties in the analysis, and factors with respect to the degree and nature of the risk. This analysis summarizes EPA's estimates of the costs and benefits associated with various arsenic levels. Summary tables are presented that characterize aggregate costs and benefits, impacts on affected entities, and tradeoffs between risk reduction and compliance costs. This analysis also summarizes the effects of arsenic on the general population as well as any sensitive subpopulations and provides a discussion on the uncertainties in the analysis and any other relevant factors. B. What Are the Quantifiable and Non-Quantifiable Health Risk Reduction Benefits? Arsenic ingestion has been linked to a multitude of health effects, both cancerous and non-cancerous. These health effects include cancer of the bladder, lungs, skin, kidney, nasal passages, liver, and prostate. Arsenic ingestion has also been attributed to cardiovascular, pulmonary, immunological, neurological, endocrine, and reproductive and developmental effects. A complete list of the arsenic-related health effects reported in humans is shown in Table X-1. Current research on arsenic exposure has only been able to define scientifically defensible risks for bladder cancer. Because there is currently a lack of strong evidence on the risks of other arsenic-related health effects noted above, the Agency has based its assessment of the quantifiable health risk reduction benefits solely on the risks of arsenic induced bladder cancers. It is important to note that if the Agency were able to quantify additional arsenic-related health effects, the quantified benefits estimates may be significantly higher than the estimates presented in this analysis. The quantifiable health benefits of reducing arsenic exposures in drinking water are attributable to the reduced number of fatal and non- fatal cancers, primarily of the bladder. Table XIII-1 shows the health risk reductions (number of total bladder cancers avoided and the proportions of fatal and non-fatal bladder cancers avoided) at various arsenic levels. [[Page 38958]] Table XIII-1.--Risk Reduction From Reducing Arsenic in Drinking Water \1\ ---------------------------------------------------------------------------------------------------------------- Risk reduction Risk reduction Risk reduction (non-fatal (total bladder (fatal bladder bladder Arsenic level \2\(µg/L) cancers cancers cancers avoided per avoided per avoided per year) year) year) ---------------------------------------------------------------------------------------------------------------- 3............................................................... 22-42 5.7-10.9 16.3-31.1 5............................................................... 16-36 4.2-9.4 11.8-26.6 10.............................................................. 9-21 2.3-5.5 296 6.7-15.5 20.............................................................. 4-12 1-3 3-9 ---------------------------------------------------------------------------------------------------------------- \1\ The number of bladder cancer cases avoided provide our ``best'' estimates at this time. The actual number of cases could be lower, given the various uncertainties discussed, or higher, as these estimates assume a 100% mortality rate. An 80% mortality rate is used in the computation of upper bound benefits. The above ranges of total, fatal, and non-fatal bladder cancer cases are based on a range of mean bladder cancer risks for exposed populations at or above arsenic levels of 3, 5, 10, and 20 µg/L as shown in Table XIII-2. For example, if we multiply the risk range at 3 µg/L (2.1 x 10-\5\ to 4.5 x 10-\5\) by the population exposed at 3 µg/L (26.6 million), we find that the total cancers avoided at this arsenic level range from 22 to 42 bladder cancers per year, when subtracted from the number of bladder cancers per year at the baseline (50 µg/L). Fatal bladder cancer cases are determined through the relationship (EPA, 1999a) that approximately 26 percent of the total bladder cancer cases avoided at each level result in fatalities. Non-fatal bladder cancer cases are calculated by subtracting the total number of cancers from the number of fatal cancer cases. Table XIII-2.--Mean Bladder Cancer Risks and Exposed Population1 ------------------------------------------------------------------------ Total bladder Arsenic level (µg/L) Mean exposed cancer cases population risk 2 avoided per year ---------------------------------------------------------------3-------- Baseline (50 µg/L): 3............................. 2.1-4.5 x 10-5 22-42 5............................. 3.6-7.5 x 10-5 16-36 10............................ 5.5-11.4 x 10-5 9-21 20............................ 6.9-13.9 x 10-5 4-12 ------------------------------------------------------------------------ 1 The population exposed at 3 µg/L or greater is approximately 26.6 million. 2 The bladder cancer risks presented in this table provide our ``best'' estimates at this time. Actual risks could be lower, given the various uncertainties discussed, or higher, as these estimates assume a 100% mortality rate. An 80% mortality rate is used in the computation of upper bound benefits. 3 Total bladder cancer cases avoided could be higher, depending on the survival rate for bladder cancer in the study area of Taiwan for the duration of the study. The Agency has developed monetized estimates of the health benefits associated with the risk reductions from arsenic exposures. The SDWA, as amended, requires that a cost-benefit analysis be conducted for each NPDWR, and places a high priority on better analysis to support rulemaking. The Agency is interested in refining its approach to both the cost and benefit analysis, and in particular recognizes that there are different approaches to monetizing health benefits. The approach used in this analysis for the measurement of health risk reduction benefits is the monetary value of a statistica