[Federal Register Volume 61, Number 239 (Wednesday, December 11, 1996)]
[Rules and Regulations]
[Pages 65183-65185]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 96-31429]
=======================================================================
-----------------------------------------------------------------------
ENVIRONMENTAL PROTECTION AGENCY
40 CFR Part 131
[FRL-5663-5]
National Toxics Rule: Remand of Water Quality Criteria for Dioxin
and Pentachlorophenol to EPA for Response to Comments
AGENCY: U.S. Environmental Protection Agency.
ACTION: Notice of availability of US EPA response to comments.
-----------------------------------------------------------------------
SUMMARY: In this document, the U.S. Environmental Protection Agency
(``EPA'') is publishing a document entitled ``Response to Comments from
American Forest and Paper Association (``AFPA'') on Two of the Exposure
Assumptions Used by EPA in Developing the Human Health Water Quality
Criteria for Dioxin and Pentachlorophenol''. AFPA challenged EPA's
promulgation of human health water quality criteria for dioxin and
pentachlorophenol. The District Court remanded these criteria to EPA
for an adequate response to AFPA's comments regarding two exposure
assumptions used by EPA in developing those criteria: an assumption
that daily water consumption is 2 liters, and an assumption that all
consumed fish are contaminated at criteria levels. EPA has prepared a
response in accordance with the court's order, and is publishing that
response in this document.
FOR FURTHER INFORMATION CONTACT: Denis R. Borum, Office of Science and
Technology, Office of Water (4304), USEPA, 401 M Street, SW.,
Washington, D.C. 20460, (202) 260-8996.
SUPPLEMENTARY INFORMATION: In November 1991, EPA proposed chemical-
specific, numeric criteria for priority toxic pollutants, including
dioxin and pentachlorophenol, necessary to bring all States into
compliance with the requirements of section 303(c)(2)(B) of the Clean
Water Act. (The ``National Toxics Rule'' or ``NTR'', 56 FR 58420;
codified at 40 CFR 131.36.) AFPA commented on a number of aspects of
the proposal, including the exposure assumptions used in EPA's water
quality criteria methodology. The NTR was promulgated in December 1992
(57 FR 60848; codified at 40 CFR 131.36). AFPA challenged the rule as
arbitrary and capricious in violation of the Administrative Procedure
Act, 5 U.S.C. 551 et seq. (Civil Action No. 93-CV-0694 (RMU), DCDC.) On
September 4, 1996, the court issued an order remanding the human health
criteria for dioxin and pentachlorophenol to EPA for ``an adequate
response to AFPA's comments'' regarding two of the exposure assumptions
used by EPA in developing the criteria. These assumptions are that
daily water consumption is 2 liters, and that all consumed fish are
contaminated at the criteria levels.
The court directed EPA to respond to AFPA's comments on these two
issues by December 13, 1996, or the human health criteria for dioxin
and pentachlorophenol will be vacated automatically. This notice
publishes EPA's response to AFPA's comments. Under the order, AFPA has
60 days from the publication of EPA's response to re-open the
litigation; upon expiration of the 60 days, the action will stand
dismissed with prejudice.
In accordance with section 553 of the Administrative Procedure Act,
EPA has determined that there is good cause not to solicit public
comment on this notice. In this notice, the Agency is simply responding
to comments on the proposed NTR and such responses are not subject to
further public comment. Moreover, the public has had ample opportunity
to comment on the exposure assumptions addressed in this notice since
the assumptions have been reflected in a number of Agency regulatory
actions. For these reasons, EPA finds further public comment to be
unnecessary.
Dated: December 5, 1996.
Robert Perciasepe,
Assistant Administrator for Water.
Response to Comments From the American Forest and Paper Association on
Two Exposure Assumptions Used by EPA To Develop Human Health Water
Quality Criteria for Dioxin and Pentachlorophenol
Background
The purpose of the Clean Water Act (``CWA'') is to protect the
nations waters, on which public health and the environment depend.
Toward this end, the CWA requires those discharging into surface waters
of the United States to have permits that limit the amount of
pollutants discharged. To set such limits, ``criteria'' are established
for each pollutant at a level necessary to preserve or achieve the uses
designated for particular waterbodies by the States. In other words,
for waterbodies designated as drinking water supplies, the criteria
should assure that people can safely drink the water. Where waterbodies
are to be used for fishing, swimming or recreation, the criteria should
assure that people can safely eat fish that are taken from those
waters, and safely use
[[Page 65184]]
the waters for other designated purposes. These criteria, intended to
protect public health, are referred to as ``human health criteria''.
Human health criteria are derived to establish quantitative
estimates of chemicals which, if not exceeded, will protect the general
population from adverse health impacts from exposure to contaminated
surface water. There are two routes of human exposure: water
consumption and fish consumption. In order to develop the criteria, EPA
needed to determine appropriate exposure assumptions for these
pathways. In 1980, EPA announced its methodology for establishing human
health criteria. 45 FR 79318 (Nov. 28, 1980). To predict the effects of
low doses of the pollutant on a hypothetical person over a 70-year
lifetime, EPA assumed the exposed individual is a male who weighs 70-
kilograms and who on a daily basis consumes an average of 6.5 grams of
fish and shellfish and 2 liters of water. Id, at 79323-24. EPA also
assumed for purposes of the methodology that the consumed water and
fish are contaminated at the criteria levels. Id., at 79323.
Issue 1: EPA's Estimate of Water Intake as 2 Liters per Day
As noted above, in order to derive human health criteria, EPA
needed to make assumptions concerning daily exposure to pollutants in
surface water from two primary routes: water consumption and fish
consumption. EPA has assumed an average daily water consumption of 2
liters. The Agency recognizes that a number of other drinking water
consumption rates have been suggested. Having reviewed those studies,
EPA's policy judgment continues to be that an assumed daily consumption
of 2 liters is reasonable to provide the margin of safety needed to
protect most people and thereby meet the objectives of the CWA. EPA is
not required, by the CWA or regulation, to base its assumed water
consumption on ``average ingestion'' in statistical terms. Rather, as
EPA explained in the proposed NTR, the assumed water consumption rate
is based on an ``approximate'' national average. (56 FR 58436), i.e.,
the approximate national average may be a starting point not an end
point. Also, both the Agency and the National Academy of Sciences
(``NAS'') have indicated that policy reasons are appropriate
considerations in adopting ``average'' drinking water consumption
rates. Since 1980, EPA has on several occasions reviewed and publicly
addressed the rationale for its water consumption value, but to the
extent that questions remain as to the basis for the assumption, EPA
here further explains that rationale.
The Agency's 1980 methodology for deriving human health criteria
assumed a water consumption of 2 liters per day. EPA cited a study done
by the NAS in support of this assumption. The NAS study was undertaken
to meet the needs expressed in the 1974 Safe Drinking Water Act
(``SDWA''). Under the SDWA, EPA was required to establish federal
standards for protection from harmful contaminants in the drinking
water supplies of the nation. Congress directed EPA to arrange with the
NAS to study the adverse effects on health attributable to contaminants
in drinking water. In 1977, NAS produced a multi-volume study entitled
Drinking Water and Health, National Academy of Sciences, Washington,
D.C. 1977. In this study, NAS considered 2 liters to be the average
amount of water consumed per day. While noting that the average per
capita water consumption of the U.S. population, as calculated from a
survey of nine different literature sources, was 1.63 liters per day,
NAS adopted 2 liters per day as representing the ``intake of the
majority of water consumers''. Id. at 11. EPA adopted 2 liters per day
as the drinking water exposure for its human health criteria
methodology, understanding that it included a margin of safety that
would ensure that most of the population would be protected.
In its comments on the proposed NTR, AFPA argued that the assumed 2
liters per day water consumption rate was overly conservative:
In a paper recently accepted for publication in Risk Analysis
(Exhibit 9) * * * (the) analysis demonstrated that the 50th
percentile intake of ``tap water'' * * * was slightly less than one
liter per day. * * * ChemRisk recently analyzed similar water
consumption data and came up with a similar figure for ``tap water''
consumption--1.2 liters per day. (Exhibit 2) Since an individual
exposed to contaminated surface water would at most only be exposed
to that contamination in the ``tap water'' he consumes, and not in
the moisture content inherent in foods that he purchases. * * * the
two liter per day assumption EPA has used overstates by a factor of
2 the potentially contaminated water that an average individual
might consume. AFPA Comments on Proposed Rule, Dec. 19, 1991, pp.
59-60.
The ChemRisk analysis states that EPA's 2.0 liters per day value is
based on the daily ration of water required by US Army field personnel;
ChemRisk questions whether this value is appropriate for a general
population with access to other beverages and that does not engage in
as much physical exertion and is not as exposed to the outdoors.
ChemRisk reviewed several studies that show the average adult
consumption rate for liquids ranges from 0.4 to 2.2 L/day. Based on a
study showing that approximately 60% of the total dietary fluid intake
is water, ChemRisk concludes that if a total fluid consumption rate of
2 liters per day is reasonable, then 60% of that consumption rate or
1.2 liters per day is water. (pp. 5-1 to 5-2)
EPA is familiar with the studies, including those cited by AFPA,
that estimate average consumption of water to be less than 2 liters per
day. Indeed, in 1990, EPA conducted its own analysis of data that
suggested that the average water consumption rate across the U.S. adult
population is 1.4 liters per day. ``Exposure Factors Handbook'', EPA
600/8-89/043, at 2-6 (AR VA-103). However, while noting that the
scientific literature suggests a daily rate of 1.4 liters, EPA made
clear that ``[p]olicy or precedent reasons may support the continued
use of the 2.0 L/day [figure] as the average adult drinking water
consumption rate.'' This analysis further indicates that consumption of
2 liters per day covers about 90 percent of the population; the
remaining 10 percent of the population consumes more than a daily
average of 2 liters. In this analysis, 2 liters per day is
characterized as a reasonable worst-case water consumption rate for
adults. Since EPA's purpose in selecting 2 liters as an average daily
water consumption rate was to provide a margin of safety sufficient to
protect most people--to the extent that 2 liters per day is protective
of approximately 90 percent of the population--using 2 liters per day
as the assumed water consumption rate for the NTR is consistent with
EPA's approach in setting human health criteria.
In a 1992 SDWA rulemaking that established health-based contaminant
levels for numerous pollutants in drinking water (57 FR 31,776), the
issue of water consumption estimates was re-examined yet again. In the
SDWA rulemaking, the Chemical Manufacturers Association (``CMA'')
submitted comments (which mirror those submitted by AFPA in the
contemporaneous NTR rulemaking) objecting to EPA's use of 2 liters per
day to set drinking water standards. CMA recommended instead the 1.4
liters per day estimate in EPA's Exposure Factors Handbook. In response
to CMA's comments, EPA acknowledged that the 1.4 liters per day
estimate is ``an overall average of a number of studies'' but rejected
using that value since some of the studies did not necessarily consider
indirect water consumption (such as use in cooking) and therefore may
not account for all exposures related to the
[[Page 65185]]
occurrence of contaminants in drinking water. EPA reiterated that the 2
liters per day assumption was a more appropriate value ``in order to be
conservative and allow for an adequate margin of safety.'' Id. at
31787. EPA further noted that the Exposure Factors Handbook considered
2 liters per day a reasonable worst case estimate.
The Agency's rationale and conclusion in the drinking water
regulation is equally applicable to the NTR. Therefore, EPA included
the Federal Register notice (Id. 31787-31788) containing EPA's response
to CMA's comments on the 2 liters per day figure in the record for the
NTR rulemaking. In the NTR, an assumption of water consumption of 2
liters per day provides a sufficient margin of safety to ensure that
most people can safely drink from waterbodies designated as drinking
water sources.
In sum, AFPA disagrees with EPA's choice of methodology and desired
level of health protection in deriving an estimate of assumed water
consumption. EPA is not required under the CWA to base its water
consumption estimate on ``average ingestion'' in statistical terms. In
order to meet the objectives of the CWA, EPA believes that its assumed
water consumption must include a margin of safety so that the general
population is protected. The NAS adopted a water consumption figure of
2 liters per day in its study of drinking water and public health as
representing the consumption of the majority of water consumers. EPA
has reviewed the subsequent studies of water consumption, but continues
to believe that 2 liters per day is appropriate for ensuring protection
of public health under the CWA.
Issue 2. EPA's Assumption That All of the Fish Consumed Is Contaminated
at the Criteria Level
In developing a methodology for deriving human health criteria, EPA
made assumptions about exposure to contamination from eating fish taken
from surface waters. The purpose of the assumptions was to ensure that
if the criteria were met in a waterbody designated for fishing, most
people could safely eat fish from that waterbody. In addition to the
assumption in the methodology that the hypothetical man has an average
daily consumption of 6.5 grams of fish, EPA assumes that all of that
fish is taken from water with pollutants present at the criteria level.
It is EPA's view that to ensure that people can safely eat fish
from waters designated for fishing, it is necessary to assume that all
of the consumed fish is taken from waterbodies at the criteria level.
EPA recognizes that there are differences in fishing patterns and the
degree to which fish bioaccumulate contaminants from the water.
However, it is EPA's judgment that this assumption regarding fish
contamination is necessary to derive criteria that are sufficiently
protective to meet the objectives of the CWA.
AFPA commented that this assumption overstates the actual expected
exposure to a contaminant:
Another source of overestimation of exposure comes from the
implicit assumption that each portion of freshwater fish consumed by
an individual will have the maximum concentration of the subject
contaminant * * * This assumption is obviously an overstatement,
since not all fish (presumably very few of them, in fact) will have
been exposed to ambient water which is just barely achieving the
water quality standard. Likewise, if the water quality standards are
being met, it would only be on rare occasions that the water
consumed will have a concentration as high as the water quality
standard allows. By definition, if the water quality standard is
implemented, ambient concentrations of the pollutant will normally
be less. In addition, depending on the dilution calculations (if
any) used in implementing the water quality standard, there may be
little or no portion of the stream where the concentration of the
pollutant is ever as high as the water quality standard allows (due
to dilution and the use of low stream flows * * * EPA has very
recently made this point forcefully in briefs and argument in the
Eastern District of Virginia in NRDC, et al. v. U.S. EPA, No.
3:91CV0058. [cite omitted]. EPA has noted that FDA's analysis of
risk from eating dioxin-contaminated fish in the Great Lakes assumed
that * * * 90 percent of the fish an individual consumed would show
no measurable contamination or would be taken from uncontaminated
areas. (cite omitted). AFPA Comments on Proposed National Toxics
Rule, December 19, 1991, pp. 60-61.
Two exhibits to AFPA's comments were prepared for the National
Council of the Paper Industry for Air and Stream Improvement. Exhibit 2
discusses studies of fish consumption of anglers in New York and Maine,
and Exhibit 4 addresses exposure to dioxin from the consumption of fish
caught in fresh waters impacted by certain pulp mills. Both reports
conclude that it is unlikely that all of the fish consumed by sport
anglers come from only one waterbody or from impacted waters. The
dioxin report notes, however, that no data are available on the number
of waterbodies fished by members of the general population or sport
fishermen over a course of time.
In its methodology, EPA assumes that all fish consumed by the
hypothetical exposed individual are contaminated at the maximum
concentration level that is ``safe'' (i.e., the criteria level). This
is the same assumption that EPA makes as to water consumption, and the
Agency's rationale supporting that assumption is equally applicable to
fish consumption.
AFPA offers examples of situations which, it contends, make it
unlikely that individuals will be exposed at the criteria level. EPA is
aware that levels of actual exposure to contamination from consuming
fish will vary depending on a number of factors. Daily fish consumption
may be both greater than and less than 6.5 grams. As EPA noted in the
proposed NTR, the exposure assumptions are based on approximate
national averages, but ``considerably understate the exposure that
would occur for certain segments of the population that have high fish
consumption or depend on fish consumption for subsistence.'' Id. at
58,436.
AFPA's exhibits note that sport fishing patterns may differ among
communities. Fishermen with access to a number of different waterbodies
may very well fish in several places and the levels of contamination
may differ among those waterbodies. Further, different species of fish
bioaccumulate pollutants at different rates. There are many
circumstances that may be relevant to fish consumption in different
communities and the level of contamination of those fish. However,
whether people fish from a number of locations, or whether some
waterbodies are not as contaminated as others does not demonstrate that
EPA's assumption is invalid. EPA must develop national criteria (that
States may modify) that must be protective of the general population.
Neither AFPA nor other commenters provided EPA with evidence sufficient
to allow the Agency to use a less conservative assumption.
It continues to be EPA's view that in order to develop criteria
that are sufficiently protective, it is necessary to assume that all
consumed fish are taken from waters at the criteria level. By deriving
criteria based on that assumption, EPA is better able to ensure that
people can safely eat fish from waters designated for fishing.
The local circumstances that AFPA reports are best addressed by the
States which have chief responsibility for implementing the CWA. States
can modify or adapt EPA's recommended human health criteria to reflect
just such local environmental conditions, and EPA encourages them to do
so. (See 57 FR 60888, Dec. 22, l992).
[FR Doc. 96-31429 Filed 12-10-96; 8:45 am]
BILLING CODE 6560-50-P
