Arsenic in Drinking Water: Regulatory Developments and Issues

Arsenic in Drinking Water:
Regulatory Developments and Issues
Mary Tiemann
Specialist in Environmental Policy
Resources, Science, and Industry Division
The Safe Drinking Water Act Amendments of 1996 directed the Environmental
Protection Agency (EPA) to update the standard for arsenic in drinking water. In 2001,
EPA issued a new arsenic rule that set the legal limit for arsenic in tap water at 10 parts
per billion (ppb), replacing a 50 ppb standard set in 1975, before arsenic was classified
as a carcinogen. The arsenic rule was to enter into effect on March 23, 2001, and water
systems were given until January 2006 to comply. EPA concluded that the rule would
provide health benefits, but projected that compliance would be costly for some small
systems. Many water utilities and communities expressed concern that EPA had
underestimated the rule’s costs significantly. Consequently, EPA postponed the rule’s
effective date to February 22, 2002, to review the science and cost and benefit analyses
supporting the rule. After completing the review in October 2001, EPA affirmed the 10
ppb standard. The new standard became enforceable for water systems in January 2006.
Since the rule was completed, Congress and EPA have focused on how to help
communities comply with the new standard. In the past several Congresses, numerous
bills have been offered to provide more financial and technical assistance and/or
compliance flexibility to small systems; however, none of the bills has been enacted.
Sources of arsenic in water include natural sources, particularly rocks and soils, and
also releases from its use as a wood preservative, in semi-conductors and paints, and from
mining and agricultural operations. Elevated levels of arsenic are found more frequently
in ground water than in surface water. Because small communities typically rely on wells
for drinking water, while larger cities often use surface-water sources, arsenic tends to
occur in higher concentrations more frequently in water used by small communities.
In the United States, the average arsenic level measured in ground-water samples is
less than or equal to 1 part per billion (ppb, or micrograms per liter [µg/L]); however,
higher levels are not uncommon. Compared with the rest of the United States, Western
states have more water systems with levels exceeding 10 ppb; levels in some locations in

the West exceed 50 ppb. Parts of the Midwest and New England also have some water
systems with arsenic levels exceeding 10 ppb, but most systems meet the new standard.
When issuing the rule, EPA estimated that roughly 4,000 (5.5%) of regulated water
systems, serving a total of 13 million people, were likely to exceed the 10 ppb standard.
The previous drinking water standard for arsenic, 50 ppb, was set by the U.S. Public
Health Service in 1942. EPA adopted that level and issued an interim drinking water
regulation for arsenic in 1975. This standard was based on estimated total dietary intake1
and non-cancer health effects. In 1986, Congress amended the Safe Drinking Water Act
(SDWA), converted all interim standards to National Primary Drinking Water
Regulations, and included arsenic on a list of 83 contaminants for which EPA was
required to issue new standards by 1989. EPA’s extensive review of arsenic risk
assessment issues caused the agency to miss the 1989 deadline. As a result of a citizen
suit, EPA entered into a consent decree with a new deadline for the rule of November

1995. EPA continued work on risk assessment, water treatment, analytical methods,

implementation, and occurrence issues, but in 1995 decided to delay the rule in order to
better characterize health effects and assess cost-effective removal technologies for small
Arsenic and the 1996 SDWA Amendments
In the 1996 SDWA Amendments (P.L. 104-182), Congress directed EPA to propose
a new drinking water standard for arsenic by January 1, 2000, and to promulgate a final
standard by January 1, 2001. Congress also directed EPA to develop a research plan for
arsenic to support the rulemaking effort and to reduce the uncertainty in assessing health
risks associated with low-level exposures to arsenic. EPA was required to conduct the
study in consultation with the National Academy of Sciences. In 1996, EPA requested
the National Research Council (NRC) to review the available arsenic toxicity data base
and to evaluate the scientific validity of EPA’s risk assessments for arsenic.
The NRC issued its report in 1999 and recommended that the standard be reduced,
but it did not recommend a particular level. The NRC affirmed that the available data
provided ample evidence for EPA’s classification of inorganic arsenic as a human
carcinogen, but that EPA’s dose-response assessment, which was based on a Taiwan
study, deserved greater scrutiny. The NRC explained that the data in the study lacked the
level of detail needed for use in dose-response assessment. The Council also reported that
research suggested that arsenic intake in food is higher in Taiwan than in the United
States, further complicating efforts to use the data for arsenic risk assessment. Based on
findings from three countries where individuals were exposed to very high levels of
arsenic (several hundreds of parts per billion or more), the NRC concluded that the data
were sufficient to add lung and bladder cancers to the types of cancers caused by ingestion
of inorganic arsenic; however, the NRC noted that few data addressed the risk of ingested
arsenic at lower concentrations, which would be more representative of levels found in

1 Food is a significant source of arsenic. The National Research Council estimates that, in the
United States, arsenic intake from food is comparable to drinking water containing 5 ppb arsenic.

the United States. The NRC concluded that key studies for improving the scientific
validity of risk assessment were needed, and recommended specific studies to EPA.2
EPA’s Final Arsenic Rule
In June 2000, EPA published its proposal to revise the arsenic standard from 50 ppb
to 5 ppb and requested comment on options of 3 ppb, 10 ppb, and 20 ppb. EPA stated that
the proposal relied primarily on the NRC analysis and some recently published research,
and that it would further assess arsenic’s cancer risks before issuing the final rule. As
proposed, the standard would have applied only to community water systems. Non-
transient, non-community water systems (e.g., schools with their own wells) would have
been required only to monitor and then report if arsenic levels exceeded the standard. In
the final rule, published on January 22, 2001 (66 FR 6976), EPA set the standard at 10
ppb and applied it to non-transient, non-community water systems, as well as community
water systems. The agency gave the water utilities five years to comply (the maximum
amount of time allowed under SDWA). EPA estimated that 3,000 (5.5%) of the 54,000
community water systems, and 1,100 (5.5%) of the 20,000 non-transient, non-community
water systems, would need to take measures to meet the standard.3
Standard-Setting Process. In developing standards under the Safe Drinking
Water Act, EPA is required to set a maximum contaminant level goal (MCLG) at a level
at which no known or anticipated adverse health effects occur and that allows an adequate
margin of safety. (EPA sets the MCLG at zero for carcinogens [as it did for arsenic],
unless a level exists below which no adverse health effects occur.) EPA must then set an
enforceable standard, the MCL, as close to the MCLG as is “feasible” using the best4
technology, treatment, or other means available (taking costs into consideration). EPA’s
determination of whether a standard is feasible typically has been based on costs to large
water systems (serving more than 50,000 people). Less than 2% of community water
systems (roughly 750 of 54,000 systems) are this large, but they serve roughly 56% of all5
people served by community systems.
Variances and Exemptions. Congress has long recognized that the technical
and cost considerations associated with technologies selected for large cities often are not
applicable to small systems. In the 1996 amendments, Congress expanded SDWA
variance and exemption provisions to address small system compliance concerns.

2 National Research Council, Arsenic in Drinking Water, National Academy of Sciences,
National Academy Press, Washington, DC, 1999, pp. 7, 22.
3 See EPA’s Technical Fact Sheet: Final Rule for Arsenic in Drinking Water, available online
at [].
4 For a more detailed discussion of these and other SDWA provisions, see CRS Report RL31243,
Safe Drinking Water Act: A Summary of the Act and Its Major Requirements, by Mary Tiemann.
5 SDWA does not discuss how EPA should consider cost in determining feasibility; thus, EPA
has relied on legislative history for guidance. The Senate report for the 1996 amendments states
that “[f]easible means the level that can be reached by large regional drinking water systems
applying best available treatment technology.... This approach to standard setting is used because
80% of the population receives its drinking water from large systems and safe water can be
provided to this portion of the population at very affordable costs.” (U.S. Senate, Safe Drinking
Water Amendments Act of 1995, S.Rept. 104-169, November 7, 1995, p. 14.) Systems serving

10,000 or more people serve about 80% of the population served by community water systems.

The small system variance provisions require that for each rule establishing an MCL,
EPA must list technologies that comply with the MCL and are affordable for three size
categories of small systems. If EPA does not list affordable compliance technologies for
small systems, then it must list variance technologies. A variance technology need not
meet the MCL, but must be protective of public health. If EPA lists a variance
technology, a state then may grant a variance to a small system, allowing the system to
use a variance technology to comply with a regulation. EPA has not identified variance
technologies for arsenic or any other standards because, based on its current affordability
criteria, EPA has determined that affordable compliance technologies are available for all
standards. Thus, small system variances are not available.
Congress took issue with EPA’s assessment that small system variance technologies
were not merited for the arsenic standard, and in 2002, directed EPA to review the criteria
it uses to determine whether a compliance treatment technology is affordable for small
systems. In March 2006, EPA proposed three options for revising its affordability criteria
(71 FR 10671). Under the current affordability criteria, EPA considers a treatment
technology affordable unless the average compliance cost exceeds 2.5% of the area’s
median household income. Based on this measure, EPA determined that affordable
technologies are available for all SDWA standards. The proposed options under
consideration are well below the current level: 0.25%, 0.50%, and 0.75% of an area’s
median household income. EPA also stated that it expects to address in the revised
criteria the issue of how to ensure that a variance technology would be protective of
public health. According to EPA, the final criteria would apply only to the new Stage 2
Disinfectants/Disinfection Byproducts Rule and future rules, and not to the arsenic rule.
Exemptions potentially offer a source of compliance flexibility for small systems.
States may grant temporary exemptions from a standard if, for certain reasons (including
cost), a system cannot comply on time. The arsenic rule gives systems five years to
comply with the new standard; an exemption allows another three years for qualified
systems. Systems serving 3,300 or fewer persons may receive up to three additional
two-year extensions, for a total exemption duration of nine years (a total of 14 years to
achieve compliance). In the final rule, EPA noted that exemptions will be an important
tool to help states address the number of systems needing financial assistance to comply
with this rule and other SDWA rules (66 FR 6988). However, to grant an exemption, the
law requires a state to hold a public hearing and make a finding that the extension will not
result in an “unreasonable risk to health.” Because the exemption process is complex,
states have seldom granted them. State officials have noted that “unreasonable risk to
health” has never been defined, and that states must make a separate finding for each
system. Many states have granted few or no exemptions for the arsenic rule.
Balancing Costs and Benefits. When proposing a rule under SDWA, EPA
must publish a determination as to whether or not the benefits of the standard justify the
costs. If EPA determines that costs are not justified, then it may set the standard at the
level that maximizes health risk reduction benefits at a cost that is justified by the
benefits. EPA determined that the “feasible” arsenic level (for large systems) was 3 ppb,
but that the benefits of that level did not justify the costs. Thus, EPA proposed a standard
of 5 ppb. Also, EPA proposed to require non-transient, non-community water systems
(e.g., schools) only to monitor and report (as opposed to treating), largely because of cost-
benefit considerations. In setting the standard at 10 ppb, EPA cited SDWA, stating that
this level “maximizes health risk reduction benefits at a cost that is justified by the
benefits.” The final rule applies to schools and similar non-community water systems.

Anticipated Benefits and Costs. In the final rule, EPA estimated that reducing
the standard to 10 ppb could prevent roughly 19 to 31 bladder cancer cases and 5 to 8
bladder cancer deaths each year. The agency further estimated that the new standard
could prevent 19 to 25 lung cancer cases and 16 to 22 lung cancer deaths each year, and
provide other cancer and non-cancer health benefits that were not quantifiable.
Regarding the cost of meeting the 10 ppb standard, EPA estimated that for systems
that serve fewer than 10,000 people, the average cost per household could range from $38
to $327 per year. Roughly 97% of the systems that were expected to exceed the standard
are in this category, and most of these systems serve fewer than 500 people. For larger
systems, projected water cost increases range from $0.86 to $32 per household. The
estimated national, annualized cost of the rule is approximately $181 million.
EPA’s Science Advisory Board (SAB) had raised concerns about the rule’s economic
and engineering assessment, and concluded that several cost assumptions were likely to
be unrealistic and other costs seemed to be excluded. The SAB also suggested that EPA6
give further thought to the concept of affordability as applied to this standard. Many
municipalities and water system representatives also disagreed with the agency’s cost
estimates. The American Water Works Association (AWWA), while supporting a stricter
standard, estimated that the new rule would cost $600 million annually and require $5
billion in capital outlays. The AWWA attributed differences in cost estimates partly to
the costs of handling arsenic-contaminated treatment residuals and the estimated number
of wells affected. The AWWA projected that the rule could cost individual households
in the Southwest, Midwest, and New England as much as $2,000 per year.7
Arsenic Rule Review. EPA issued the final rule on January 22, 2001. In March
2001, the Administrator delayed the rule for 60 days, citing concerns about the science
supporting the rule and its estimated cost to communities. On May 22, 2001, EPA
delayed the rule’s effective date until February 22, 2002, but did not change the 2006
compliance date for water systems (66 FR 28342). At EPA’s request, the NRC undertook
an expedited review of EPA’s arsenic risk analysis and subsequent health effects research,
the National Drinking Water Advisory Council (NDWAC) reassessed the rule’s cost, and
the SAB reviewed its benefits. EPA also requested public comment on whether the data
and analyses for the rule support setting the standard at 3, 5, 10, or 20 ppb (66 FR 37617).
The NRC determined that “recent studies and analyses enhance the confidence in risk
estimates that suggest chronic arsenic exposure is associated with an increased incidence
of bladder and lung cancer at arsenic levels in drinking water below the current MCL of

50 µg/L.”8 The NDWAC concluded that EPA had produced a credible cost estimate,

given constraints and uncertainties, and suggested ways to improve estimates. The SAB
offered ways to improve the benefits analysis. In October 2001, EPA affirmed that 10
ppb was the appropriate standard and announced plans to provide $20 million for research
on affordable arsenic removal technologies to help small systems comply.

6 Science Advisory Board, Arsenic Proposed Drinking Water Regulation: A Science Advisory
Board Review of Certain Elements of the Proposal, EPA-SAB-DWC-01-001, December 2000,
p. 4.
7 AWWA, January 17, 2001. See [].
8 National Research Council, Arsenic in Drinking Water: 2001 Update, NAS, p. 14.

Legislative Action
Since the arsenic standard was revised, Congress repeatedly has expressed concern
over the cost of this regulation, especially to small, rural communities. The 107th
Congress directed EPA to review its affordability criteria and how the small system
variance and exemption programs should be implemented for arsenic (P.L. 107-73,
H.Rept. 107-272, p. 175). The conferees directed the agency to report on its affordability
criteria, administrative actions, funding mechanisms for small system compliance, and
possible legislative actions. In 2002, EPA submitted its report to Congress, Small Systems
Arsenic Implementation Issues, on actions EPA was taking to address these directives.
Major activities included developing and implementing a small community assistance
plan to improve access to financial and technical assistance, improve compliance capacity,
and simplify the use of exemptions. EPA also has sponsored research on low-cost arsenic
treatment technologies and has issued guidance to help states grant exemptions.
The 108th Congress again expressed concern over the economic impact that the
revised standard could have in many communities. In the conference report for the
omnibus appropriations act for FY2005 (P.L. 108-447), Congress provided $8.2 million
for arsenic removal research. The conferees expressed concern that the new requirements
could pose a “huge financial hardship” for many rural communities. Congress directed
EPA to report on the extent to which communities were being affected by the rule and to
propose compliance alternatives and make recommendations to minimize costs. This
report is pending.
In the 110th Congress (as in the 109th Congress), legislative efforts focused on
helping economically struggling communities comply with the arsenic rule and other
drinking water standards. Various bills were offered to promote small system compliance
by providing technical assistance, financial assistance, and/or compliance flexibility. The
Senate Environment and Public Works Committee reported several bills that would have
authorized new funding for drinking water infrastructure. The Water Infrastructure
Financing Act (S. 3617, S.Rept. 110-509), which paralleled the committee bill from the
109th Congress, proposed to increase funding authority for the drinking water (and clean
water) state revolving fund (DWSRF) program and to create a grant program at EPA for
small or economically disadvantaged communities for critical drinking water and water
quality projects. S. 1933 (S.Rept. 110-475) would have authorized a new grant program
for small water systems, and S. 1429 (S.Rept. 110-242) and H.R. 6313 would have
reauthorized funding authority for small system technical assistance under SDWA. Other
bills included S. 2509, which proposed to require EPA to promote the use of affordable
technologies (e.g., point-of-use technologies and bottled water), revise its affordability
criteria, and provide more compliance assistance for high-priority rules including the
arsenic rule. S. 2509 also would have required EPA or a state to ensure that funds have
been made available to small systems before taking enforcement actions. H.R. 2141
would have required states to grant exemptions to eligible small systems for rules
covering naturally occurring contaminants (such as arsenic and radium). None of these
bills was enacted.
SDWA compliance and, more broadly, drinking water infrastructure issues are likely
to remain on the congressional agenda. However, severe competition for federal resources
and uncertainty in the policy agenda make the prospects for new funding legislation
unclear. (For more information on this and other issues, see CRS Report RL34201, Safe
Drinking Water Act: Selected Regulatory and Legislative Issues, by Mary Tiemann.)