Natural Resources Defense Council v. Environmental Protection Agency

United States Court of Appeals

FOR THE DISTRICT OF COLUMBIA CIRCUIT

Argued October 24, 2013 Decided April 18, 2014

No. 10-1371

NATURAL RESOURCES DEFENSE COUNCIL,

PETITIONER

v.

ENVIRONMENTAL PROTECTION AGENCY AND GINA

MCCARTHY, IN HER OFFICIAL CAPACITY AS ADMINISTRATOR,

U.S. ENVIRONMENTAL PROTECTION AGENCY,

RESPONDENTS

CEMEX, INC., ET AL.,

INTERVENORS

Consolidated with 10-1378, 13-1112

On Petitions for Review of Final Actions of the

United States Environmental Protection Agency

James S. Pew and Seth L. Johnson argued the causes for

petitioners. With them on the briefs were John Walke,

Meleah Geertsma, and Avinash Kar.

Matthew R. Oakes, Attorney, U.S. Department of Justice,

argued the cause for respondents. With him on the brief were

2

Robert G. Dreher, Acting Assistant Attorney General, and

Steven E. Silverman, Attorney, U.S. Environmental Protection

Agency. Daniel R. Dertke and T. Monique Peoples,

Attorneys, U.S. Department of Justice, entered appearances.

Carter G. Phillips argued the cause for intervenors. With

him on the briefs were Roger R. Martella Jr., Timothy K.

Webster, Ryan C. Morris, William M. Bumpers, Debra J.

Jezouit, Michael B. Schon, Deborah E. Jennings, Chet M.

Thompson, Beth S. Ginsberg, Jason T. Morgan, Ashley C.

Parrish, Cynthia A.M. Stroman, and Richard G. Stoll.

Russell S. Frye and Richard G. Stoll were on the brief for

amici curiae SSM Coalition, et al. in support of respondents.

Before: KAVANAUGH and SRINIVASAN, Circuit Judges,

and EDWARDS, Senior Circuit Judge.

Opinion for the Court filed by Circuit Judge

KAVANAUGH.

KAVANAUGH, Circuit Judge: Portland cement is the key

ingredient in concrete. The basic process for making Portland

cement is much the same today as it was when the material

was first developed nearly 200 years ago. Cement

manufacturers pulverize limestone and minerals, and then

heat those raw materials to several thousand degrees. The

resulting substance, called clinker, is then cooled and ground

into a fine gray powder. This powder – called Portland

cement – is later combined with sand, rocks, and water to

make concrete.

The grinding and heating involved in cement

manufacturing has an unfortunate side effect: It releases into

the air a number of hazardous air pollutants, most notably

3

mercury, hydrochloric acid, hydrocarbons, and particulate

matter. This case concerns EPA’s efforts to develop rules

under the Clean Air Act to limit emissions of those pollutants

from cement plants.

In a previous decision, we considered EPA’s first attempt

to create emission standards for the cement industry, and we

found the agency’s action arbitrary and capricious. See

Portland Cement Association v. EPA, 665 F.3d 177 (D.C. Cir.

2011). Following our ruling, EPA went back to the drawing

board and developed the emission standards at issue here, the

2013 Rule.

Several environmental organizations, including the

Natural Resources Defense Council and the Sierra Club, have

petitioned for review of the 2013 Rule, arguing primarily that

certain aspects of the Rule contravene the Clean Air Act.

They also challenge EPA’s decision to create an affirmative

defense for private civil suits in which plaintiffs sue sources

of pollution and seek penalties for violations of emission

standards. EPA’s affirmative defense would be available to

defendants in cases where an “unavoidable” malfunction had

resulted in impermissible levels of emissions.

We conclude that the emissions-related provisions of

EPA’s 2013 Rule are permissible but that the affirmative

defense for private civil suits exceeds EPA’s statutory

authority. We therefore grant the petitions in part and vacate

the portion of the Rule pertaining to the affirmative defense.

We deny the petitions in all other respects.

I

Section 112 of the Clean Air Act, 42 U.S.C. § 7412,

requires EPA to establish technology-based emission

standards for major sources of certain hazardous air

4

pollutants. Emission standards must reflect “the maximum

degree of reduction in emissions” that EPA determines is

“achievable,” taking into consideration “the cost of achieving

such emission reduction, and any non-air quality health and

environmental impacts and energy requirements.” 42 U.S.C.

§ 7412(d)(2). The resulting standards are commonly known

as the “maximum achievable control technology,” or

“MACT” standards. See National Lime Association v. EPA,

233 F.3d 625, 630 (D.C. Cir. 2000).

EPA uses a two-step process for establishing MACT

standards. The agency begins by setting a minimum

stringency level, or “floor,” based on the results achieved by

the best-performing similar sources. See 42 U.S.C.

§ 7412(d)(3). Once EPA sets the statutory floor, it then

determines, considering cost and the other factors listed in

Section 112(d)(2), whether a more restrictive standard is

“achievable,” and if so then adopts that standard. EPA calls

these stricter requirements “beyond-the-floor” standards.

Sierra Club v. EPA, 479 F.3d 875, 877 (D.C. Cir. 2007).

When EPA sets an emission standard, it also determines a

schedule for compliance with that standard. For existing

sources, EPA must “provide for compliance as expeditiously

as practicable, but in no event later than 3 years after the

effective date” of the emission standard. 42 U.S.C.

§ 7412(i)(3)(A).

In 2010, pursuant to its Section 112 authority, EPA

promulgated National Emission Standards for Hazardous Air

Pollutants from the Portland Cement Manufacturing Industry

and Standards of Performance for Portland Cement Plants, 75

Fed. Reg. 54,970 (Sept. 9, 2010). That 2010 Rule set or

revised emissions limits for mercury, hydrogen chloride, total

hydrocarbons (a surrogate for organic hazardous air pollutants

5

such as benzene and formaldehyde), and particulate matter (a

surrogate for certain non-mercury metals). Cement plants

would be required to comply with the new standards

beginning in September 2013. 75 Fed. Reg. at 54,993.

In addition to establishing emission standards, the 2010

Rule created an affirmative defense in private civil suits when

violations of the standards occurred because of “unavoidable”

malfunctions. See id. at 54,993, 55,053. The affirmative

defense replaced a previous EPA policy creating an

exemption from emissions limitations during malfunction

events. In a prior decision, this Court struck down that

exemption because it was inconsistent with the requirement

that emission standards apply continuously. See Sierra Club

v. EPA, 551 F.3d 1019, 1027-28 (D.C. Cir. 2008).

Industry groups petitioned this Court for review and we

found the 2010 Rule arbitrary and capricious and remanded to

EPA. We ruled specifically that, in calculating the floor for

MACT purposes, EPA had arbitrarily included in its dataset

information from cement kilns properly classified as

commercial incinerators, which are regulated under a separate

provision of the Act. See Portland Cement Association v.

EPA, 665 F.3d 177, 186-89 (D.C. Cir. 2011).

In our 2011 decision, however, we did not vacate the

emission standards set forth in the 2010 Rule or stay its

implementation pending EPA’s reconsideration process,

stating that “it is unlikely that significant changes will be

made to the standards upon reconsideration.” Id. at 189.

On remand, however, EPA made several relevant

changes to the Portland cement emission standards. See

National Emission Standards for Hazardous Air Pollutants for

the Portland Cement Manufacturing Industry and Standards of

6

Performance for Portland Cement Plants, 78 Fed. Reg. 10,006

(Feb. 12, 2013) (the 2013 Rule).

First, consistent with our Portland Cement opinion, EPA

revised its dataset to exclude commercial incinerators. When

recalculated using the revised data, the maximum emissions

level for particulate matter – the floor – was ultimately

revised from 0.04 lb/ton to 0.07 lb/ton of clinker for existing

kilns. See id. at 10,017-19. And EPA declined to re-adopt the

more stringent, 0.04 lb/ton limit of the 2010 Rule as a

beyond-the-floor standard. The agency reasoned that

achieving that additional increment of particulate reduction

would not be cost effective on a cost-per-ton basis. See id. at

10,020-21.

Second, citing additional compliance strategies afforded

cement manufacturers by the revised particulate standard,

EPA established a new compliance date of September 2015

for that standard. See id. at 10,014. EPA further concluded

that although the emissions limits for mercury, hydrochloric

acid, and hydrocarbons remained the same as in the 2010

Rule, the new September 2015 compliance date would also

apply to those emission standards. According to EPA,

coordinating the compliance date for particulate matter,

mercury, hydrochloric acid, and hydrocarbons was essential

because the latter standards “all typically involve some

element of [particulate matter] generation and capture and so

the controls must be integrated with [particulate matter]

control strategies.” Id. at 10,022.

The 2013 Rule also retained the affirmative defense for

private civil suits when the defendant violated emission

standards due to an unavoidable malfunction. EPA explained

that in its view, the affirmative defense was necessary to

resolve a “tension” between the Clean Air Act’s requirement

7

that emission standards apply at all times and the fact that

emission limits may sometimes be exceeded for reasons

beyond the control of the source. See id. at 10,014.

NRDC, the Sierra Club, and other environmental

organizations have petitioned for review of various aspects of

the 2013 Rule. In Part II of this opinion, we address

petitioners’ arguments regarding the emission standards for

particulate matter. In Part III, we address petitioners’

challenge to the compliance schedule implementing some of

the 2013 Rule’s emission standards. In Part IV, we consider

whether EPA’s decision to create the affirmative defense to

civil penalties for certain malfunction-related events exceeds

the agency’s statutory authority.

II

We first consider petitioners’ challenges to the emission

standards for particulate matter.

A

The 2013 Rule ultimately set the emissions level for

particulate matter at 0.07 lb/ton of clinker for existing kilns.

The 2010 Rule had set the level at 0.04 lb/ton of clinker.

Petitioners argue that the 2013 Rule weakens the particulate

matter standard in violation of Section 112(d)(7) of the Clean

Air Act, 42 U.S.C. § 7412(d)(7). That provision, titled “Other

requirements preserved,” states:

No emission standard or other requirement

promulgated under this section shall be interpreted,

construed or applied to diminish or replace the

requirements of a more stringent emission limitation

or other applicable requirement established pursuant to

8

section 7411 of this title, part C or D of this

subchapter, or other authority of this chapter or a

standard issued under State authority.

(emphases added). Petitioners maintain that EPA violated

Section 112(d)(7) because the 2013 Rule’s particulate matter

standards “diminish or replace” the more stringent standards

in the 2010 Rule.

EPA responds that such a reduction does not violate

Section 112(d)(7). In EPA’s view, the most natural reading of

Section 112(d)(7) is that “other authority” refers to authority

other than Section 112 and other than the parts of the Clean

Air Act specifically enumerated in Section 112(d)(7). Stated

another way, EPA suggests that Section 112(d)(7) is simply a

savings clause that makes clear that Section 112 does not

supersede the requirements of other, more restrictive

provisions of the Act.

By contrast, petitioners say that “other authority” of the

Act includes Section 112 itself, as well as other provisions in

the Act. Petitioners read the statute as an anti-backsliding

restriction on EPA’s ability to voluntarily reduce the

stringency of any emission standard issued under Section 112.

As EPA points out, however, when Congress has sought

to include that sort of anti-backsliding provision in the Clean

Air Act, it has done so directly and explicitly. Cf., e.g., 42

U.S.C. § 7410(l) (“The Administrator shall not approve a

revision of a [State Implementation Plan] if the revision

would interfere with any applicable requirement concerning

attainment and reasonable further progress”); id. § 7502(e)

(specifying pollution control requirements if “the

Administrator relaxes a national primary ambient air quality

standard”). Section 112(d)(7) contains no such language.

Furthermore, EPA argues that petitioners’ interpretation of

9

Section 112(d)(7) would mean that any change to a rule

issued under Section 112 – even a necessary change made just

for technical reasons or because there was a calculation error

– would be impermissible if the change made the standard

less stringent. EPA says that it would be extraordinary if the

statute precluded that kind of necessary change.

In wading through this back-and-forth, we ultimately

need not decide whether EPA’s reading is the better or only

reading of this statutory provision, but simply whether it is a

permissible reading. EPA administers the Clean Air Act, and

we must defer to its reasonable interpretation of any

ambiguities in the statute. See Chevron U.S.A. Inc. v. NRDC,

467 U.S. 837, 843-44 (1984). Here, even if the statute does

not compel EPA’s reading, and indeed even if EPA’s reading

is not the better reading, the statute at a minimum is

sufficiently ambiguous on this point to permit EPA’s reading

of “other authority.” Because EPA’s reading is at least

reasonable, we reject petitioners’ argument and rule for EPA

at Chevron step two.

B

Petitioners also contend that EPA should have set a more

restrictive particulate matter standard when considering

whether to set “beyond-the-floor” standards. Petitioners

argue in particular that EPA misinterpreted the statute to

allow it to consider cost-effectiveness when setting beyond-

the-floor standards.

Under Section 112(d), EPA must require “the maximum

degree of reduction in emissions” that EPA determines is

“achievable.” 42 U.S.C. § 7412(d)(2). To determine that

emission level, EPA first establishes a minimum stringency

level, or “floor,” based on the emission results achieved by

10

the best-performing sources in the category at issue. See id.

§ 7412(d)(3). Once EPA sets the statutory floor, it then

determines, based on cost and the other factors listed in

Section 112(d)(2), whether a more restrictive, beyond-the-

floor standard is achievable.1

When it promulgated the 2013 Rule, EPA rejected

petitioners’ argument to set a 0.04 lb/ton limit as a beyond-

the-floor standard. 78 Fed. Reg. at 10,020. EPA estimated

that a beyond-the-floor standard set at the 0.04 lb/ton level

would result in a reduction of 138 tons of particulate matter

per year, at a cost of $37 million. Id. Based on those

estimates, EPA noted that the cost-effectiveness of the

potential beyond-the-floor standard – $268,000 per ton of

particulate matter removed – was substantially lower than the

cost-effectiveness of other emission standards previously

rejected by EPA. See 78 Fed. Reg. at 10,021.

Petitioners take issue with EPA’s consideration of cost-

effectiveness as a component of the Section 112(d)(2) cost

analysis. Petitioners contend that “cost” for purposes of the

statute only concerns whether “the standard is too expensive

for industry to achieve,” in essence, whether the standards

would bankrupt the industry. Pet’rs Br. 34.

1

In relevant part, the statute reads: “Emissions standards

promulgated under this subsection and applicable to new or existing

sources of hazardous air pollutants shall require the maximum

degree of reduction in emissions of the hazardous air pollutants

subject to this section (including a prohibition on such emissions,

where achievable) that the Administrator, taking into consideration

the cost of achieving such emission reduction, and any non-air

quality health and environmental impacts and energy requirements,

determines is achievable for new or existing sources in the category

or subcategory . . . .” 42 U.S.C. § 7412(d)(2) (emphasis added).

11

EPA says that Congress afforded it wide latitude in

deciding how to balance cost and other considerations when

determining that maximum achievable reduction in emissions.

According to EPA, Section 112 does not command EPA to

use a particular form of cost analysis. In taking cost into

account, EPA contends that it may determine whether the

proposed emission levels would be cost-effective. Indeed,

EPA notes that this Court has previously recognized EPA’s

authority to consider cost-effectiveness in setting standards

under nearly identical provisions of the Clean Air Act. See,

e.g., Husqvarna AB v. EPA, 254 F.3d 195, 200 (D.C. Cir.

2001) (“Because section 213 does not mandate a specific

method of cost analysis, we find reasonable the EPA’s choice

to consider costs on the per ton of emissions removed

basis.”); National Association of Clean Water Agencies v.

EPA, 734 F.3d 1115, 1156-57 (D.C. Cir. 2013) (citing

Husqvarna; EPA could consider cost-effectiveness in setting

a beyond-the-floor standard under Section 129(a)(2) of the

Act).

Again on this issue as with the first, we need not decide

whether EPA’s reading is the only reading of this provision.

Even if the statute does not compel EPA’s approach, and even

if EPA’s reading is not the better reading, we conclude that it

is still at least a reasonable reading given the various potential

meanings of “cost” in this context. Therefore, we reject

petitioners’ argument that EPA was required to exclude

consideration of cost-effectiveness and to set a beyond-the-

floor standard of 0.04 lb/ton of clinker.

III

Next, we consider petitioners’ claim that EPA acted

unreasonably in setting a compliance date of September 2015

12

for the emission standards for particulate matter, mercury,

hydrochloric acid, and hydrocarbons.

Under Section 112(i)(3)(A) of the Clean Air Act, EPA

must require compliance with emission standards for existing

sources “as expeditiously as practicable, but in no event later

than 3 years after the effective date of such standard.” 42

U.S.C. § 7412(i)(3)(A). Petitioners here contend that EPA

violated this timing provision by “extending” the compliance

deadline for the 2010 Rule from September 2013 to

September 2015.

As applied to the particulate matter standard, there is a

basic flaw in petitioners’ argument: The 2013 Rule did not

simply “extend” the deadline for complying with the 2010

Rule. Rather, the 2013 Rule established new particulate

matter standards with a new effective date of February 2013,

and a new compliance date of September 2015. EPA

concluded that any earlier date of compliance would not be

practicable because of the multi-year timeline for upgrading

the technology necessary to ensure compliance. See 78 Fed.

Reg. at 10,024. As we have ruled before, EPA may reset the

compliance date for an emission standard when it introduces a

new standard with a new effective date, as was the case for

particulate matter in the 2013 Rule. See NRDC v. EPA, 489

F.3d 1364 (D.C. Cir. 2007).

According to petitioners, even if EPA could change the

compliance date for the particulate matter standard, the

agency could not do the same for the mercury, hydrochloric

acid, and hydrocarbon standards. The 2010 Rule set the

emission levels for those pollutants, and the 2013 Rule did not

alter the emission levels for those pollutants. This situation

does indeed present a bit of a conundrum. On the one hand,

we know under the terms of the statute that EPA has set a

13

compliance date of 2015 as the date that is as expeditious as

practicable for particulate matter. On the other hand, our

prior decision did not vacate the prior 2010 Rule, and the

compliance date for the other pollutants as set forth in that

Rule would otherwise be 2013.

This conundrum is resolved when one realizes that it

would be irrational and even absurd to have different

compliance dates for the different pollutants. EPA explained

that the technology is such that it would be senseless to have

different compliance dates. See 78 Fed. Reg. at 10,024. Even

petitioners do not deny that the compliance date for all the

pollutants should be the same. To be sure, they want 2013 not

2015 as the compliance date. But they recognize the general

undesirability of a compliance date of 2013 for some of the

pollutants and of 2015 for other pollutants. See Tr. of Oral

Argument at 10-11. Finally and perhaps most importantly,

our prior decision in this case also recognized, at least

implicitly, that there must be a single compliance date for all

of the pollutants. See Portland Cement Association, 665 F.3d

at 189. Our decision necessarily relied on an assumption that

if EPA did not alter the level for any of the pollutants, the date

would be 2013. But if EPA changed the level for one of the

pollutants, the compliance date for all the pollutants would

move together. Petitioners’ argument for a 2013 compliance

date would be inconsistent with our prior decision.

In short, we reject petitioners’ argument about the 2015

compliance date.

IV

We next consider petitioners’ challenge to the affirmative

defense that EPA created for cases of “unavoidable”

malfunctions.

14

Section 304(a) of the Clean Air Act, 42 U.S.C. § 7604(a),

allows individuals to file citizen suits in federal district court

against sources that violate emission standards. Under the

law as originally enacted, a court could order only injunctive

relief as a remedy for a violation. But as part of the 1990

amendments to the Act, Congress expanded the citizen suit

provision to give district courts authority to impose “any

appropriate civil penalties,” which may include monetary

penalties. 42 U.S.C. § 7604(a).

In the 2010 Rule, EPA created an affirmative defense to

Section 304(a) for certain emissions violations caused by

“unavoidable” malfunctions. Under the affirmative defense,

the district court may assess penalties only if violators “fail to

meet [their] burden of proving all of the requirements in the

affirmative defense.” 78 Fed. Reg. at 10,039. EPA retained

the affirmative defense when it promulgated the 2013 Rule.

See 40 C.F.R. § 63.1344.

Petitioners now argue that the affirmative defense

exceeds EPA’s statutory authority and that it is for the courts

to decide whether to create an affirmative defense in these

private civil suits, not EPA. We agree.

The threshold question is whether petitioners have

standing to challenge EPA’s adoption of the affirmative

defense. Petitioners are environmental associations with

individual members across the country. EPA’s affirmative

defense would immunize certain emissions that petitioners

contend should be penalized. Some of petitioners’ members

will suffer from those higher emissions, according to their

affidavits. A ruling in their favor would prevent those

emissions and help alleviate that harm. That’s good enough.

Petitioners have shown injury-in-fact, causation, and

redressability, and they thus have standing under Article III.

15

We turn, then, to the substance of petitioners’ challenge

to the affirmative defense.

Section 304(a) grants “any person” the right to

“commence a civil action” against any person “who is alleged

to have violated (if there is evidence that the alleged violation

has been repeated) or to be in violation of” an emission

standard or limitation under the Clean Air Act. 42 U.S.C.

§ 7604(a). The statute further provides that the federal district

courts “shall have jurisdiction, without regard to the amount

in controversy or the citizenship of the parties, to enforce such

an emission standard or limitation . . . and to apply any

appropriate civil penalties.” Id.

When determining whether civil penalties are

appropriate, district courts look to Section 113(e)(1) of the

Act, which directs courts to “take into consideration (in

addition to such other factors as justice may require) the size

of the business, the economic impact of the penalty on the

business, the violator’s full compliance history and good faith

efforts to comply, the duration of the violation as established

by any credible evidence . . . , payment by the violator of

penalties previously assessed for the same violation, the

economic benefit of noncompliance, and the seriousness of

the violation.” 42 U.S.C. § 7413(e)(1).

Section 304(a) creates a private right of action, and as the

Supreme Court has explained, “the Judiciary, not any

executive agency, determines ‘the scope’ – including the

available remedies – ‘of judicial power vested by’ statutes

establishing private rights of action.” City of Arlington v.

FCC, 133 S. Ct. 1863, 1871 n.3 (2013) (emphasis added)

(quoting Adams Fruit Co. v. Barrett, 494 U.S. 638, 650

(1990)). Section 304(a) is in keeping with that principle. By

its terms, Section 304(a) clearly vests authority over private

16

suits in the courts, not EPA. As the language of the statute

makes clear, the courts determine, on a case-by-case basis,

whether civil penalties are “appropriate.” By contrast, EPA’s

ability to determine whether penalties should be assessed for

Clean Air Act violations extends only to administrative

penalties, not to civil penalties imposed by a court. See 42

U.S.C. § 7413(d)(2)(B) (Administrator may “compromise,

modify, or remit, with or without conditions, any

administrative penalty”). To the extent that the Clean Air Act

contemplates a role for EPA in private civil suits, it is only as

an intervenor. See id. § 7604(c)(2). EPA also of course could

seek to participate as an amicus curiae.

EPA argues that its proposed affirmative defense simply

fleshes out the statutory requirement that penalties be applied

only when “appropriate.” But under this statute, deciding

whether penalties are “appropriate” in a given private civil

suit is a job for the courts, not for EPA. When a private suit is

filed, the defendant can argue that penalties should not be

assessed, based on the factors in Section 113(e)(1) such as the

defendant’s “full compliance history and good faith efforts to

comply.” Id. § 7413(e)(1). EPA can support that argument as

intervenor or amicus, to the extent such status is deemed

appropriate by the relevant court. But under the statutory

scheme, the decision whether to accept the defendant’s

argument is for the court in the first instance, not for EPA.

EPA alternatively contends that it is permitted to create

the affirmative defense because of Section 301(a)(1) of the

Clean Air Act. See 78 Fed. Reg. at 10,014. That provision

authorizes EPA’s Administrator to “prescribe such regulations

as are necessary to carry out his functions under” the Act. 42

U.S.C. § 7601(a)(1). But we have consistently held that

EPA’s authority to issue ancillary regulations is not open-

ended, particularly when there is statutory language on point.

17

See, e.g., American Petroleum Institute v. EPA, 52 F.3d 1113,

1119 (D.C. Cir. 1995) (“the general grant of rulemaking

power to EPA cannot trump specific portions of the CAA”);

NRDC v. Reilly, 976 F.2d 36, 41 (D.C. Cir. 1992) (EPA

cannot use its general rulemaking authority as justification for

adding to a statutorily specified list); Sierra Club v. EPA, 719

F.2d 436, 453 (D.C. Cir. 1983) (same); see also Gonzales v.

Oregon, 546 U.S. 243, 264-65 (2006) (“It would

go . . . against the plain language of the text to treat a

delegation for the ‘execution’ of [the Attorney General’s]

functions as a further delegation to define other functions well

beyond the statute’s specific grants of authority.”). Those

precedents establish a simple and sensible rule: EPA cannot

rely on its gap-filling authority to supplement the Clean Air

Act’s provisions when Congress has not left the agency a gap

to fill. So it is here.

On a different tack, EPA notes that Section 304(a)(1)

does not expressly deny EPA the ability to create an

affirmative defense, and EPA emphasizes that this Court has

frequently recognized the need for flexibility in the

administrative process. EPA Br. 46. That’s true. But the

suggestion implicit in EPA’s argument – that we should

“presume a delegation of power absent an express

withholding of such power” – is “plainly out of keeping with

Chevron . . . .” Railway Labor Executives’ Association v.

National Mediation Board, 29 F.3d 655, 671 (D.C. Cir. 1994)

(en banc).

Finally, EPA suggests that an affirmative defense for

malfunctions is necessary to account for the tension between

requirements that emissions limitations be “continuous” and

the practical reality that control technology can fail

unavoidably. See 78 Fed. Reg. at 10,014. That is a good

argument for EPA to make to the courts – and for the courts

18

to then consider – in future civil cases when this issue arises.

But it does not suffice to give EPA authority to create an

affirmative defense.2

***

We grant the petitions for review with regard to EPA’s

affirmative defense and vacate those portions of the 2013

Rule pertaining to the defense. We deny the petitions in all

other respects.

So ordered.

2

The Fifth Circuit recently upheld EPA’s partial approval of

an affirmative defense provision in a State Implementation Plan.

See Luminant Generation Co. v. EPA, 714 F.3d 841 (5th Cir. 2013).

We do not here confront the question whether an affirmative

defense may be appropriate in a State Implementation Plan.