United States Court of Appeals
FOR THE DISTRICT OF COLUMBIA CIRCUIT
Argued October 14, 2004 Decided January 11, 2005
No. 03-1455
CARUS CHEMICAL COMPANY,
PETITIONER
V.
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY,
RESPONDENT
On Petition for Review of an Order of the
Environmental Protection Agency
Thomas W. Dimond argued the cause for petitioner. With
him on the brief were John C. Berghoff, Jr. and Jennifer R.
Hagan.
Ammie Roseman-Orr, Attorney, U.S. Department of Justice,
argued the cause for respondent. With her on the brief were
John C. Cruden, Deputy Assistant Attorney General, and Sheila
M. Igoe, Attorney, U.S. Environmental Protection Agency.
Before: GINSBURG, Chief Judge, and HENDERSON and
ROBERTS, Circuit Judges.
Opinion for the Court filed by Chief Judge GINSBURG.
2
GINSBURG, Chief Judge: The Environmental Protection
Agency placed a site owned in part by Carus Chemical
Company on the National Priorities List (NPL) of hazardous
waste sites, pursuant to the Comprehensive Environmental
Response, Compensation, and Liability Act of 1980 (CERCLA),
94 Stat. 2767, 42 U.S.C. § 9601 et seq., and its implementing
regulations. Carus argues the EPA’s action was arbitrary and
capricious because the agency misinterpreted, and hence
misapplied, its Hazard Ranking System, and because it
disregarded more recent data that contradicted those upon which
the agency relied.
We hold that neither the EPA’s interpretation nor its
application of the disputed regulation was unreasonable.
Further, the agency reviewed the data Carus submitted and
correctly concluded they did not cast doubt upon the decision to
list the site. Accordingly, we deny the petition for review.
I. Background
Carus operates a manufacturing plant east of La Salle,
Illinois on a parcel of land, once part of a larger property to
which the EPA refers as the Matthiessen & Hegeler Zinc
Company Site, where the latter company operated a smelter and
a rolling mill for more than 100 years. During that period, there
accumulated at the site two large slag piles, one of which, six
acres in extent, is located adjacent to (and partly in) the Little
Vermilion River and partly on Carus’s property. The EPA
decided that hazardous substances in the slag piles posed a threat
to human health and to the environment sufficient to warrant
putting the entire Matthiessen & Hegeler site on the NPL.
Before recounting the specifics of the EPA’s decision, an
explanation of the statute and regulations underlying the NPL is
3
in order. The CERCLA directs the President, who delegated the
responsibility to the EPA, to compile a list of cleanup priorities
among hazardous waste sites around the country. The EPA’s
listing a site on the NPL, however, does not necessarily mean it
will order remedial action at that site, see Honeywell Int’l, Inc.
v. EPA, 372 F.3d 441, 443 (D.C. Cir. 2004); rather, it guarantees
only more detailed study, see Eagle-Picher Indus., Inc. v. EPA,
759 F.2d 905, 919-20 (D.C. Cir. 1985). Nonetheless, listing can
have significant adverse consequences for the owner of a listed
property. See Mead Corp. v. Browner, 100 F.3d 152, 155 (D.C.
Cir. 1996) (costs in business reputation, property value, and
increased probability of remediation).
In order to identify candidates for the NPL, the EPA
promulgated the Hazard Ranking System (HRS), see 40 C.F.R.
pt. 300, App. A, a comprehensive methodology and
mathematical model the agency uses to “evaluate[] the observed
or potential release of hazardous substances” and to “quantif[y]
the environmental risks a site poses.” Tex Tin Corp. v. EPA, 992
F.2d 353, 353 (D.C. Cir. 1993).
In order to evaluate a waste site using the HRS, the EPA
first identifies the “sources” of contamination, the “[h]azardous
substances associated with these sources,” and the “[p]athways
potentially threatened by these hazardous substances.” HRS §
2.2. The HRS lists four possible pathways: soil exposure, air
migration, ground water migration, and the one relevant to this
case, surface water migration. Id. at § 2.1. For each pathway
deemed potentially affected in light of conditions at the site, the
agency calculates a score based upon particular “threats.” The
surface water migration pathway is scored based upon threats to
drinking water, to the human food chain, and to the
environment. See id. at § 4.0.1. With respect to each pathway
and threat to be scored, the HRS calls for the EPA to measure
4
three so-called factor categories: “likelihood of release (or
likelihood of exposure)”; “waste characteristics”; and “targets,”
which may include an individual, a human population,
resources, and sensitive environments. Id. at § 2.1.3. The
agency’s measurements of the first two categories are relevant
to this case.
The “[l]ikelihood of release is a measure of the likelihood
that a waste has been or will be released to the environment.”
Id. at § 2.3. When, as in this case, the EPA determines there has
already been a release, it assigns a fixed number for this
component of the overall score of the pathway, regardless of the
level of that release. Id.
With respect to waste characteristics, the HRS first requires
the EPA to “select the hazardous substance potentially posing
the greatest hazard for the pathway.” Id. at § 2.4.1. The agency
is then to evaluate persistence, bioaccumulation, and toxicity
factors pertaining to that substance, id. at § 2.4 (only the last of
which features in this case).
For each substance being scored, the agency uses a toxicity
factor value between 0 and 10,000, reflecting the potential of
that substance to cause adverse health effects. For a single
substance there may be multiple toxicity factor values, each
corresponding to a route of exposure (e.g., inhalation, ingestion)
through which that substance may come into contact with
humans. If there are, and if the agency has “usable toxicity
data” for more than one such exposure route, then it should
“consider all exposure routes and use the highest assigned value,
regardless of exposure route, as the toxicity factor value.” Id. at
§ 2.4.1.1.
Based upon the considerations recounted above (as well as
5
others not relevant here), the EPA assigns each site a score from
0 to 100. A site with a score greater than 28.50 is eligible for
the NPL. See 68 Fed. Reg. 55,875, 55,876 (Sept. 29, 2003).
The EPA’s study of the Matthiessen & Hegeler site began
with an aerial photograph of the site taken in 1988 and with data
compiled by the Environmental Protection Agency of Illinois in
1991 and 1993 from sediment, groundwater, and soil samples
taken around the slag piles. Upon the basis of this evidence, the
EPA determined that hazardous substances were being released
into the Little Vermilion River. Because the observed release
was into a river, the agency scored the surface water migration
pathway, and because Illinois classified that river as a fishery,
the agency scored that pathway for the threat it posed to the
human food chain.
Following the method set forth in the HRS, the EPA then
assessed the “waste characteristics” of the hazardous substances
found at the site, namely, cadmium, copper, lead, nickel, and
zinc. Carus’s principal dispute is over the EPA’s choice,
purportedly in compliance with HRS § 2.4.1.1, to use the
toxicity factor value for cadmium corresponding to the
inhalation route of exposure. Plugging this value into the model,
the EPA calculated a score of 100 for the pathway and a total
score of 50 for the Matthiessen & Hegeler site.
Because the total score exceeded 28.50, the EPA proposed
to list the site on the NPL. See 66 Fed. Reg. 32,287, 32,290,
32,293 (June 14, 2001). Carus submitted comments in
opposition to the listing, which comments included technical
documents prepared for it by GeoSyntec Consulting. The EPA
found the objections raised in those comments unpersuasive and,
accordingly, published a final rule adding the Matthiessen &
Hegeler site to the NPL. 68 Fed. Reg. at 55,878. Carus now
6
petitions for review of that decision.
II. Analysis
Carus raises two separate challenges to the EPA’s decision
to place the Matthiessen & Hegeler site on the NPL. First,
Carus argues the agency wrongly interpreted HRS § 2.4.1.1 as
requiring it to use a toxicity factor value for cadmium
corresponding to a route of exposure (inhalation) unlikely to
occur in light of the conditions at the site. Had the agency
instead applied the toxicity factor value corresponding to the
ingestion route of exposure, Carus maintains, the HRS score for
the site would have been below 28.50. Second, Carus contends
the sampling data and the documents it submitted during the
comment period rendered unreasonable the EPA’s reliance
solely upon data collected earlier by the Illinois EPA.
A. The EPA’s Interpretation of HRS § 2.4.1.1
As the EPA understands HRS § 2.4.1.1, and as it applied
that regulation in the rulemaking under review, 68 Fed. Reg. at
55,875-55,882, the agency was required to use the toxicity factor
value for the inhalation of cadmium even though it was scoring
the surface water migration pathway. Carus takes exception to
that interpretation, arguing it is nonsensical to read the rule as
mandating the use of a toxicity factor value corresponding to an
exposure route (inhalation) unlikely to present a threat
considering the pathway being scored.
An agency is, of course, entitled to “substantial deference”
in its “interpretation of its own regulations.” Thomas Jefferson
Univ. v. Shalala, 512 U.S. 504, 512 (1994). Carus does not
quarrel with this principle but nonetheless argues deference to
the EPA’s interpretation of § 2.4.1.1 is not appropriate because
7
(Carus nakedly asserts) that interpretation is simply a “litigation
position.” See Bowen v. Georgetown Univ. Hosp., 488 U.S. 204,
212 (1988) (court owes no “deference to an agency counsel’s
interpretation of a statute where the agency itself has articulated
no position on the question”). But see Auer v. Robbins, 519 U.S.
452, 462 (1997) (agency interpretation of regulation first
appearing in legal brief not categorically unworthy of
deference).
We need not, however, concern ourselves today with the
degree of deference owed to an agency’s regulatory
interpretation first articulated in the course of litigation. See
generally Drake v. FAA, 291 F.3d 59, 68 (D.C. Cir. 2002)
(listing three conditions for deference in that circumstance). The
EPA has consistently interpreted § 2.4.1.1 as requiring it to use
the highest toxicity factor value for a substance regardless of the
most likely route of exposure. Indeed, the protest Carus now
advances – that there is no scientific justification for using the
toxicity factor value for a route of exposure that is improbable
in light of the pathway being scored – was made and rejected
when the EPA first issued the regulation. At that time, the
agency defended its interpretation of the rule as follows:
EPA recognizes that toxicity values for substances are
route-specific. However, the three pathways ... receiving a
toxicity factor value are substance migration pathways, not
human exposure routes. Multiple human exposure routes
are possible for each substance migration pathway (e.g.,
volatile substances in ground water or surface water used
for drinking can be inhaled during showering), and,
therefore, use of a single route-specific toxicity value is not
necessarily appropriate .... For this reason, and to avoid the
added complexity of route-specific toxicity evaluations,
EPA decided to base the toxicity factor on the highest
8
route-specific value (if more than one is available) ....
[T]his may, in a few site situations, overstate toxicity ....
Responses to Comments on the Revisions to the Hazard Ranking
System (HRS), at 3-G1 (1990).
The EPA’s original understanding of HRS § 2.4.1.1 is
reflected in its response to Carus’s comments in the proceeding
now under review, and Carus points to not one instance in the
intervening dozen years in which the agency applied the
regulation differently. We therefore have no basis for treating
the EPA’s interpretation of its regulation as though it were a
mere litigation position. Consequently, we must defer to that
interpretation unless “an alternative reading is compelled by the
regulation’s plain language” or by other indicia of the agency’s
“intent at the time of the regulation’s promulgation.” Thomas
Jefferson Univ., 512 U.S. at 512; see also Castlewood Products,
L.L.C. v. Norton, 365 F.3d 1076, 1082 (D.C. Cir. 2004).
Under that standard of review, we must uphold the EPA’s
interpretation because it accords both with the text of the
regulation and with what we know of the agency’s
understanding of the rule when it was issued.
As for the former, the regulation reads in relevant part as
follows:
Toxicity Factor. Evaluate toxicity for those hazardous
substances at the site that are available to the pathway being
scored. For all pathways and threats, except the surface
water environmental threat, evaluate human toxicity as
specified below.
....
9
For hazardous substances having usable toxicity data for
multiple exposure routes (for example, inhalation and
ingestion), consider all exposure routes and use the highest
assigned value, regardless of exposure route, as the toxicity
factor value.
HRS § 2.4.1.1. The EPA’s interpretation clearly hews close to
the text. The rule first directs the agency to “[e]valuate [the]
toxicity” of each hazardous substance “available to the pathway
being scored,” and then unequivocally requires that, when the
substance being scored has “usable toxicity data for multiple
exposure routes” – as does cadmium – the EPA “consider all
exposure routes and use the highest assigned value, regardless
of exposure route.”
Carus does not claim the EPA failed to follow its own long-
standing interpretation of the rule; rather, Carus complains that
the agency has misread the rule – presumably ever since its
promulgation. Read properly, per Carus, the rule allows the use
of only those toxicity factor values corresponding to routes of
exposure the EPA finds are a threat at the site. In support of this
interpretation, Carus points to the reference, in the opening
sentence of the rule, to the “pathway being scored.”* According
*
In its brief, Carus also claims support from HRS § 1 which,
it says, provides the agency is to assign “numerical values to factors,
that relate to or indicate risk, based on conditions at the site.” Neither
the italicized words nor any variation thereof appear in § 1 nor, as far
as we can tell, anywhere in the HRS. Instead, that phrase appears only
in the document issued by the EPA along with the rule that added the
Matthiessen & Hegeler site to the NPL. See Support Document for the
Revised National Priorities List Final Rule – September 2003, at ix.
Because the quoted words are not part of the HRS they are of no
assistance to Carus in its effort to show the agency misinterpreted the
10
to Carus, this phrase modifies not just the preceding clause of
the same sentence but the entire rule, thus requiring the EPA to
decide whether exposure via a particular route is likely,
considering the pathway being scored, before it uses a toxicity
factor value corresponding to that exposure route.
We need not tarry long over this claim. “A challenge to an
agency’s interpretation of its own regulation ... turns not on
whether the challenger has articulated a rationale to support its
interpretation, but on whether the agency has offered an
explanation that is reasonable and consistent with the
regulation’s language and history.” Trinity Broad. of Fla., Inc.,
v. FCC, 211 F.3d 618, 627 (D.C. Cir. 2000). Even if we owed
no deference to the agency, however, we would not be
persuaded by Carus’s alternative reading of the rule. First, the
phrase “available to the pathway” plainly limits only the
universe of hazardous substances the EPA may use in scoring a
pathway; the phrase cannot reasonably be read to limit, as Carus
suggests, the universe of toxicity factor values upon which the
agency may draw. Second, Carus would read out of the
regulation the direction to use the highest toxicity factor value
“regardless of exposure route,” in contravention of the principle
that every word of a legal text should be given effect. See
United States v. Menasche, 348 U.S. 528, 520 (1955)
(interpretation must “give effect, if possible, to every clause and
word of a statute”).
Carus nonetheless seeks to bolster the plausibility of its
interpretation by arguing the EPA’s reading lacks a “rational
HRS. In any event, in the same support document the EPA explained
that HRS § 2.4.1.1 required it to use the highest toxicity factor value
for a substance regardless of exposure route. Id. at 1.1-21.
11
basis” in science and in common sense. The rationality of
§ 2.4.1.1, however, is not before us; the Congress, in § 113(a) of
the CERCLA, provided that a substantive challenge to a
regulation promulgated under that Act must be made within 90
days of its promulgation. See 42 U.S.C. § 9613(a). The time to
challenge the reasonableness of § 2.4.1.1 therefore passed in
1991. Accord RSR Corp. v. EPA, 102 F.3d 1266, 1269 (D.C.
Cir. 1997) (Section 113(a) bars untimely claim that HRS states
improper toxicity factor for lead). It is some comfort, therefore,
that EPA’s rendering of § 2.4.1.1 seems consistent with the
purpose of the NPL to provide a quick and an inexpensive
means of identifying high priority sites for further study, see
Eagle-Picher, 759 F.2d at 911.
B. Carus’s Submissions to the EPA
Carus also claims the EPA relied upon insufficient and out-
dated sampling data obtained by its Illinois counterpart in 1991
and 1993. Carus apparently believes more recent data from
samples collected by GeoSyntec Consulting shows the EPA
erred in listing the Matthiessen & Hegeler site on the NPL.
Before considering this argument we must address two anterior
matters.
First, because the CERCLA does not specify a standard of
review applicable to this issue, we shall proceed under the
standard prescribed by the Administrative Procedure Act and
will set aside the agency’s action only if it is “arbitrary,
capricious, an abuse of discretion, or otherwise not in
accordance with law.” 5 U.S.C. § 706(2)(A); see Eagle-Picher,
759 F.2d at 921 n.80. In applying that standard, we remain
mindful of the “significant deference” we owe to the EPA’s
decision to list a site on the NPL because of the “highly
technical issues involved” and because the NPL serves merely
12
as a “rough list of priorities, assembled quickly and
inexpensively.” Bradley Mining Co. v. EPA, 972 F.2d 1356,
1359 (1992). Even so, we must assure ourselves the EPA has
“examined relevant data and has articulated a rational
explanation for its actions.” Eagle-Picher, 759 F.2d at 921.
Second, the EPA contends Carus has forfeited any claim the
agency failed adequately to consider the GeoSyntec data
because Carus failed to say with “reasonable specificity during
the period for public comment,” Mossville Envtl. Action Now v.
EPA, 370 F.3d 1213, 1238 (D.C. Cir. 2004), how those data
would have affected the HRS score for the Matthiessen &
Hegeler site. Although we do not doubt the need for specificity,
we see that Carus made three comments before the EPA with
sufficient specificity that we may consider them on review.
Carus first argued before the EPA that the documents it
submitted showed the slag piles were “highly resistant to further
leaching.” Neither before the agency nor here, however, does
Carus describe the significance of this purported fact for the
HRS scoring of the site. As far as we are told – and that is by
the EPA – the only arguable connection between resistance to
leaching and the HRS score is whether cadmium, the substance
scored for the surface water migration pathway, was “available
to the pathway.” Yet Carus does not challenge the EPA’s
contention that both direct observation (based upon the aerial
photograph taken in 1988) and chemical analyses (performed in
1991 and 1993), established cadmium from the slag piles had
been released into the Little Vermilion River, and hence was
“available to the pathway.” The toxicity factor value for a
substance available to the pathway pursuant to HRS § 2.4.1.1,
notwithstanding any putative resistance to “further leaching,”
may be used in scoring that pathway.
13
Next, Carus asserts the documents it submitted to the EPA
“establish[ed] risk-based clean-up levels” – whatever that may
mean – and “evaluate[d] possible remedial alternatives,” both
of which contributions the agency apparently failed to consider.
Perhaps so, but Carus did and still does nothing to explain how
either consideration could have affected the HRS score for the
Matthiessen & Hegeler site, see Northside Sanitary Landfill, Inc.
v. Thomas, 849 F.2d 1516, 1519 (D.C. Cir. 1988) (comment to
agency must show alleged mistake “was of possible significance
in the results”). Nor does Carus refute the EPA’s response that
it had no obligation to consider such matters because they are
irrelevant to the decision to list a site on the NPL, see Honeywell
Int’l, 372 F.3d at 445 (EPA not required to explore remedial
options when placing site on NPL).
Finally, Carus maintains the sampling data that GeoSyntec
prepared and submitted to the EPA shows the agency overstated
the environmental and health risks posed by conditions at the
site. Here Carus cites Linemaster Switch Corp. v. EPA, 938
F.2d 1299 (D.C. Cir. 1991), for the proposition the EPA must
consider more recent data that have been included in the
administrative record. Carus remains fatally silent, however, in
the face of the EPA’s indisputable observation that it did review
the data Carus submitted but found them consistent with its
scoring of the site. See Support Document for the Revised
National Priorities List Final Rule – September 2003, at 1.1-10
(cited in 68 Fed. Reg. at 55,877).
In the EPA’s response to public comments it explained that
the data compiled by GeoSyntec, far from contradicting the data
upon which the agency relied, “confirmed the presence of ...
cadmium and lead in soil samples” taken from around the slag
piles. Id. The EPA also responded to Carus’s claim the
GeoSyntec data showed the levels of hazardous substances
14
present in the Little Vermilion River were below applicable
limits; the agency explained that an “observed release” of a
hazardous substance may be established “[e]ven though levels
may be lower than regulatory limits ... if the measured levels are
significantly higher than background levels.” Id. In sum, the
EPA concluded Carus did not present “any specific comments
[showing] that the data used in the HRS scoring [were] incorrect
or why [Geosyntec’s] data would suggest that the site score is
incorrect.” Id. Just so.
III. Conclusion
For the foregoing reasons Carus’s petition for review is
Denied.