Perchlorate Contamination of Drinking Water: Regulatory Issues and Legislative Actions
Perchlorate Contamination of Drinking Water:
Regulatory Issues and Legislative Actions
Mary Tiemann
Specialist in Environmental Policy
Resources, Science, and Industry Division
Summary
Perchlorate is the explosive component of solid rocket fuel, fireworks, road flares,
and other products. Used heavily by the Department of Defense (DOD) and related
industries, perchlorate also occurs naturally and is present in some organic fertilizer.
This soluble, persistent compound has been detected in drinking water supplies,
especially in California. It also has been found in milk and many foods. Because of this
widespread occurrence, concern over the potential health risks of perchlorate exposure
has increased, and some states, water utilities, and Members of Congress have urged the
Environmental Protection Agency (EPA) to set a federal drinking water standard for this
chemical. Regulatory issues involve the relative health benefits and costs of federal
regulation, including environmental cleanup and water treatment costs, both of which
are driven by federal and state standards. (California and Massachusetts have set
standards.) EPA has spent years assessing perchlorate’s health effects and occurrence
(including its occurrence in food) to determine whether a national standard is warranted.
Interagency disagreements over the risks of perchlorate exposure led several federal
agencies to ask the National Research Council (NRC) to evaluate perchlorate’s health
effects and EPA’s risk analyses. In 2005, the NRC issued its report, and EPA adopted
the NRC’s recommended reference dose (i.e., the expected safe dose) for perchlorate
exposure. Subsequent studies raised more concerns about potential effects of low-level
exposures, particularly for infants in certain cases. In October 2008, EPA announced
its preliminary determination not to regulate perchlorate in drinking water. This report
reviews perchlorate contamination issues and related developments.
Background
Ammonium perchlorate is the key ingredient in solid fuel for rockets and missiles;
other perchlorate salts are used to manufacture products such as fireworks, air bags, and
road flares. Uncertainty about the health effects of perchlorate exposure has slowed efforts
to establish drinking water and environmental cleanup standards. However, because of
perchlorate’s persistence in water and ability to affect thyroid function, concern has
escalated with the detection of perchlorate in water in at least 33 states. In the absence
of a federal standard, states have begun to adopt their own measures. Massachusetts set
a drinking water standard of 2 parts per billion (ppb, or micrograms per liter [µg/L]) in
2006, and California adopted a 6 ppb standard in October 2007. Several states have issued
health goals or advisory levels ranging from 1 ppb in Maryland (advisory level) and New
Mexico (drinking water screening level) to 51 ppb in Texas (industrial cleanup level).
Occurrence. Perchlorate has been used heavily by the DOD and its contractors,
and perchlorate contamination has been found near weapons and rocket fuel
manufacturing facilities and disposal sites, research facilities, and military bases.
Fireworks, road flares, construction sites, and other manufacturing activities and facilities
also have been sources of contamination. Moreover, perchlorate occurs naturally (in West
Texas, for example), is present in organic fertilizer imported from Chile, and can occur
as a breakdown product of other products. It has been detected in drinking water sources,
primarily in the Southwest and in scattered locations across the country. Contamination
has been found most often in ground water, including some large aquifers in California.
In 1999, EPA required public water systems to monitor for perchlorate under the
Unregulated Contaminant Monitoring Rule (UCMR) to determine the frequency and
levels at which it is present in public water supplies nationwide. The UCMR required
monitoring by all systems serving more than 10,000 persons and by a sample of smaller
systems. Of some 3,700 water systems tested, perchlorate was detected in 153 systems
in 26 states and two commonwealths, including 58 systems in California. In 14 systems,
perchlorate levels exceeded EPA’s preliminary remediation goal of 24.5 ppb. California
has required more comprehensive monitoring, and perchlorate has been detected at least
twice in 241 active or standby sources of drinking water in that state since 2002. In 2005,
EPA reported perchlorate contamination at 65 DOD facilities, 7 other federal facilities,1
and 37 private sites. All sampling results combined (i.e., soil, public and private drinking
water wells, groundwater monitoring wells, and surface water), the Government2
Accountability Office reported that perchlorate has been found at 395 sites.
Monitoring also has been conducted to assess the presence of perchlorate in foods.
In 2004, the Food and Drug Administration (FDA) tested 500 samples of foods, including
vegetables, milk, and bottled water for perchlorate. Samples were taken in areas where
water was thought to be contaminated. The FDA found perchlorate in roughly 90% of
lettuce samples (average levels ranged from 11.9 ppb to 7.7 ppb for lettuces), and in 101
of 104 bottled milk samples (with an average level of 5.7 ppb). To assess the presence of
perchlorate in a wider range of foods, the FDA began testing all samples in its Total Diet
Study in 2005.3 Perchlorate was detected in 625 of 1065 (50%) of samples, and in 2114
of the 285 (74%) foods tested. In most cases, perchlorate levels were in the low single
1 EPA, Federal Facilities Restoration and Reuse, Known Perchlorate Releases in the U.S., March
2 U.S. Government Accountability Office, Perchlorate: A System to Track Sampling and Cleanup
Results is Needed, GAO-05-462, May 2005, pp. 29-44.
3 The FDA test results are available online at [http://www.cfsan.fda.gov/~dms/clo4dat2.html].
4 Clarence W. Murray et al., “U.S. Food and Drug Administration’s Total Diet Study: Dietary
Intake of Perchlorate and Iodine,” Journal of Exposure Science and Environmental Epidemiology
(continued...)
digits; however, levels were higher in some foods (e.g., shrimp, tomatoes, spinach, and
bacon). The study found that 2-year-olds have the highest total perchlorate intake per
kilogram body weight per day, followed by infants (6 to 12 months of age) and children
6 to 10 years of age. The widespread detection of perchlorate in food is relevant to EPA’s
standard-setting efforts, because EPA considers non-water exposures when determining
whether to establish a standard for a contaminant, and at what level to set a standard.
Health Effects. Perchlorate is not known to cause cancer. It is known to disrupt
the uptake of iodine in the thyroid, and health effects associated with perchlorate exposure
are expected to parallel those caused by iodine deficiency. Iodine deficiency decreases
the production of thyroid hormones, which help regulate the body’s metabolism and
growth. A key concern is that impairment of thyroid function in pregnant women can
affect fetuses and nursing infants and can result in delayed development and decreased
learning capacity. Several human studies have indicated that thyroid changes occur in
humans at significantly higher levels of perchlorate than the amounts typically observed
in water supplies.5 However, a 2006 study by the Centers for Disease Control and
Prevention (CDC) of a representative sample of the U.S. population found that
environmental exposures to perchlorate have an effect on thyroid hormone levels in
women with iodine deficiency. (No effect was found in men.) Fully 36% of the 1,111
women in this study were found to be iodine deficient, and the median level of urinary6
perchlorate measured in the women was 2.9 ppb.
EPA Assessment of Perchlorate for Regulation
Over the past decade, EPA has evaluated perchlorate to determine whether a federal
drinking water standard is needed. Under the Safe Drinking Water Act (SDWA,
§1412(b)(1)), EPA must regulate a contaminant if the Administrator determines that the
contaminant (1) may have an adverse health effect, (2) occurs in public water systems at
a frequency and level of public health concern, and (3), in the sole judgment of the
Administrator, regulation of the contaminant presents a meaningful opportunity for
reducing health risks. In 1997, when a sensitive detection method became available for
perchlorate and detections increased, scientific information was limited. In 1998, EPA
placed perchlorate on the list of contaminants that were candidates for regulation, but
concluded that information was insufficient to determine whether perchlorate should be
regulated under the SDWA. EPA listed perchlorate as a priority for further research on
health effects and treatment technologies and for collecting occurrence data. In 1999, EPA
required water systems to monitor for perchlorate under the Unregulated Contaminant
Monitoring Rule to determine the frequency and levels at which it is present in public
water supplies nationwide. In January 2007, EPA reported that it had collected sufficient
4 (...continued)
(2008), 1-10.
5 Michael A. Kelsh et al., “Primary Congenital Hypothyroidism, Newborn Thyroid Function,
and Environmental Perchlorate Exposure Among Residents of a Southern California
Community,” Journal of Occupational Environmental Medicine, 2003, p. 1117.
6 Benjamin C. Blount, James L. Pirkle, et al., “Urinary Perchlorate and Thyroid Hormone Levels
in Adolescent and Adult Men and Women Living in the United States,” Centers for Disease
Control and Prevention, in Environmental Health Perspectives, December 2006, p. 1865.
occurrence data, and that further monitoring was not needed for the agency to make a
regulatory determination (72 Fed. Reg. 367, January 4, 2007).
Perchlorate Risk Assessment. In 1992, and again in 1995, EPA issued draft
reference doses (RfDs) for perchlorate exposure. An RfD is an estimate (with uncertainty
spanning perhaps an order of magnitude) of a daily oral exposure that is not expected to
cause any adverse, non-cancer health effects during a lifetime. In developing an RfD,
EPA incorporates factors to account for sensitive subpopulations, study duration, inter-
and intraspecies variability, and data gaps. The draft RfDs range of 0.0001 to 0.0005
milligrams per kilogram (mg/kg) body weight per day translated to a drinking water
equivalent level of 4 ppb-18 ppb. EPA takes the RfD into account when setting a
drinking water standard; it also considers costs, the capabilities of monitoring and
treatment technologies, and other sources of perchlorate exposure, such as food.
EPA’s 1999 draft risk characterization resulted in a human risk benchmark of 0.0009
mg/kg per day (with a 100-fold uncertainty factor), which converted to a drinking water
equivalent level of 32 ppb. However, EPA determined that the available health effects and
toxicity database was inadequate for risk assessment. In 1999, EPA issued an Interim
Assessment Guidance for Perchlorate, which recommended that EPA risk managers use
the earlier reference dose range and drinking water equivalent level (DWEL) of 4-18 ppb
for perchlorate-related assessment activities at hazardous waste sites.
In 2002, EPA prepared a draft risk assessment that concluded that the potential
human health risks of perchlorate exposures include effects on the developing nervous
system and thyroid tumors, based on rat studies that observed benign tumors and adverse
effects in fetal brain development. The document included a draft RfD of 0.00003 mg/kg
per day, which translated to a drinking water equivalent level of 1 ppb. This document
was controversial, both for its implications for cleanup costs and for science policy
reasons. (For example, some peer reviewers expressed concern over EPA’s risk
assessment methodology and reliance on rat studies.) DOD, water suppliers, and other
commentors expressed concern that the draft RfD could lead to unnecessarily stringent
and costly cleanups of perchlorate releases at federal facilities and in water supplies. In
2002, a federal interagency perchlorate working group convened to discuss perchlorate
risk assessment, research and regulatory issues, and related agency concerns. Working
group members included DOD, EPA, the Department of Energy, the National Aeronautics
and Space Administration, the Office of Science and Technology Policy, the Council on
Environmental Quality, and the Office of Management and Budget.
NRC Perchlorate Study. To resolve some of the uncertainty and debate over
perchlorate’s health effects and the 2002 draft risk assessment, the interagency working
group asked the National Research Council (NRC) to review the available science for
perchlorate and EPA’s draft assessment. The NRC was asked to comment and make
recommendations. The NRC Committee to Assess the Health Implications of Perchlorate
Ingestion issued its review in January 2005 and suggested several changes to EPA’s draft
risk assessment. The committee concluded that because of key differences between rats
and humans, studies in rats were of limited use for quantitatively assessing human health
risk associated with perchlorate exposure. Although the committee agreed that thyroid
tumors found in a few rats were likely perchlorate treatment-related, it concluded that
perchlorate exposure is unlikely to lead to thyroid tumors in humans. The committee
noted that, unlike rats, humans have multiple mechanisms to compensate for iodide
deficiency and thyroid disorders. Also, the NRC found flaws in the design and methods
used in the rat studies. The committee concluded that the animal data selected by EPA
should not be used as the basis of the risk assessment.
The committee also reviewed EPA’s risk assessment model. It agreed that EPA’s
model for perchlorate toxicity represented a possible early sequence of events after
exposure, but it did not think that the model accurately represented possible outcomes
after changes in thyroid hormone production. Further, the committee disagreed with
EPA’s definition of a change in thyroid hormone level as an adverse effect. Rather, the
NRC defined transient changes in serum thyroid hormone as biochemical events that
might precede adverse effects, and identified hypothyroidism as the first adverse effect.
Because of research gaps regarding perchlorate’s potential effects following changes
in thyroid hormone production, the committee made the recommendation that EPA use
a nonadverse effect (i.e., the inhibition of iodide uptake by the thyroid in humans) rather
than an adverse effect as the basis for the risk assessment. The committee explained that
“[i]nhibition of iodide uptake is a more reliable and valid measure, it has been
unequivocally demonstrated in humans exposed to perchlorate, and it is the key event that
precedes all thyroid-mediated effects of perchlorate exposure.”7 Based on the use of this
point of departure, the reliance on human studies, and the use of an uncertainty factor of
10 (for intraspecies differences), the NRC’s recommendations led to an RfD of 0.0007
mg/kg per day. The committee concluded that this RfD should protect the most sensitive
population (i.e., the fetuses of pregnant women who might have hypothyroidism or iodide
deficiency) and noted that the RfD was supported by clinical studies, occupational and
environmental epidemiologic studies, and studies of long-term perchlorate administration
to patients with hyperthyroidism. In addition, the NRC identified data gaps and research
needs. The committee has received some criticism for the extent to which it relied on a
small, short-term human study, and debate over perchlorate’s health risks has continued.
EPA’s Response. In 2005, EPA adopted the NRC recommended reference dose
of 0.0007 mg/kg per day, which translates to a drinking water equivalent level of 24.5
ppb. The DWEL is the concentration of a contaminant in water that is expected to have
no adverse effects; it is intended to include a margin of safety to protect the fetuses of
pregnant women who might have a preexisting thyroid condition or insufficient iodide
intake. EPA based the DWEL on the assumption that all exposure would come from
drinking water. If EPA were to develop a drinking water standard for perchlorate, it
would lower the DWEL to account for other sources of exposure, particularly food.
In January 2006, EPA’s Superfund office issued guidance adopting the NRC
reference dose and the DWEL of 24.5 ppb as the recommended value to be considered as
the preliminary remediation goal (PRG) to guide perchlorate assessment and cleanup at
Superfund sites. In March, EPA’s Children’s Health Protection Advisory Committee
(CHPAC) wrote to the EPA Administrator that the PRG does not protect infants, who are
highly susceptible to neurodevelopmental toxicity and may be more exposed than fetuses
to perchlorate. The CHPAC noted that perchlorate is concentrated in breast milk and that
nursing infants could receive daily doses greater than the RfD if the mother is exposed to
7 National Research Council, Health Implications of Perchlorate Ingestion, Committee to Assess
the Health Implications of Perchlorate Ingestion, National Academy of Sciences, 2005, p. 9.
24.5 ppb perchlorate in tap water. The committee recommended that the Superfund office
lower the PRG and that the Office of Water develop a standard for perchlorate and, in the
interim, issue a drinking water health advisory that takes into account early life exposures.
In October 2008, EPA announced a preliminary determination not to regulate
perchlorate, noting that less than 1% of water systems have perchlorate levels above the
health reference level. EPA concluded that perchlorate failed to meet two of SDWA’s
regulatory criteria (i.e., that a contaminant occur frequently at levels of health concern,
and that establishing a national drinking water standard would provide a “meaningful
opportunity for health risk reduction”). In response, EPA’s Science Advisory Board’s
(SAB’s) Drinking Water Committee argued that, given perchlorate’s occurrence and well-
documented toxicity, EPA must have a compelling basis to support a determination not
to regulate. The SAB requested more time to review the new model EPA relied on, and
to comment on the preliminary determination. If the final determination is not to regulate,
EPA will develop a nonenforceable health advisory to guide state and other officials on
health effects, monitoring and treatment technologies, and expected safe exposure levels.
Department of Defense Perchlorate Actions
DOD is responsible for some large releases of perchlorate contamination and has
allotted significant resources to address this problem. DOD has spent more than $114
million on research activities regarding perchlorate treatment technologies, detection
methods, toxicity studies, and substitutes. Additional funds have been spent on cleanup.
Although remediation has proceeded at some sites, cleanups typically are driven by
drinking water standards or other established cleanup standards. With no federal
standard, cleanup goals and responsibilities have been ambiguous outside of California
and Massachusetts where standards have been set. In 2006, after EPA established a
DWEL for perchlorate, DOD adopted a policy setting 24 ppb as the level of concern to
be used in managing perchlorate releases, unless a more stringent federal or state standard
exists. The policy applies broadly to DOD installations and former military lands, and
directs the services to test for perchlorate when it is reasonably expected that a release has
occurred. If perchlorate levels exceed 24 ppb, a site-specific risk assessment must be
conducted; if the assessment indicates that the perchlorate could result in adverse health
effects, then the site must be prioritized for risk management. DOD uses a relative risk
site evaluation framework to help prioritize environmental restoration work and to
allocate resources among sites. (For more information, see DOD website:
[https://www.denix .osd.mil/portal/page/ portal/denix /environment/MERIT/EC/ECAL].)
Congressional Actions
Several perchlorate bills were considered, but none were enacted during the 110th
Congress. Responding to EPA’s 2007 decision not to require further monitoring for
perchlorate as an unregulated contaminant, S. 24 was introduced to require community
water systems to test for perchlorate and disclose its presence in annual consumer reports.
S. 150 and H.R. 1747 would have required EPA to set a standard for perchlorate. The
Senate Environment and Public Works Committee reported S. 24 (S.Rept. 110-483) and
S. 150 (S.Rept. 110-484). Additionally, H.Con.Res. 347 expressed the sense of Congress
that the CDC and FDA should take action to educate the public on the importance of
adequate iodine intake. (Iodine is protective against perchlorate exposure.)