Reply Brief — Environmental Defense v. Duke Energy Corporation

Qo G4) , a

No. 05-848 OcT 19 2006

In The

Supreme Court of the United States

ENVIRONMENTAL DEFENSE, et al.,

Petitioners,

Vv.

DUKE ENERGY CORPORATION, et al.,

Respondents.

.

On Writ Of Certiorari To The

United States Court Of Appeals

For The Fourth Circuit

S

REPLY BRIEF OF PETITIONERS

+

JEFFREY M. GLEASON SEAN H. DONAHUE

J. BLANDING HOLMAN, [IV Counsel of Record

CALEB JAFFE 2000 L St., NW

SOUTHERN ENVIRONMENTAL Suite 808

LAW CENTER Washington, D.C. 20036

200 W. Franklin (202) 466-2234

Suite 330

- DAVID T. GOLDBERG

Chapel Hill, N.C. 27516

New York, N.Y. 10011

VICKIE PATTON (212) 334-8813

ENVIRONMENTAL DEFENSE

2334 North Broadway

Boulder, Colorado 80304

(303) 447-7215

Counsel for Petitioners

TABLE OF CONTENTS

TABLE OF AUTHORITIES ..............0ccscccsosssossssceeseserssesnerse li

UTI BIIIS crccccscccesessseseseseresesesoencensensecsvesenscceseuesavarense l

I. DUKE COULD HAVE OBTAINED D.C.

CIRCUIT REVIEW OF ANY CLAIM THAT

EPA HAD TO USE AN HOURLY RATE

FERS PRI BI cecececcennnsecensscnssenconassnanetmnnanenenssesessoonsnnes 2

II. DUKE’S NEW READING OF THE

REGULATIONS IS OBVIOUSLY WRONG............... 9

Ill. THE ACT DOES NOT REQUIRE USE OF AN

HOURLY RATE MEASURE FOR PSD

EMISSIONS “INCREASES” ........0..-...cccercssescccseseeeses 14

CIITA AIIDS cxcccnssesccsensesecnssqenvnensessnonssesosesnseonssnneenssonsenant 20

TABLE OF AUTHORITIES

CASES

Aetna Health Inc. v. Davila, 542 U.S. 200 (2004).........0...... 9

Alabama Power v. EPA, 636 F.2d 323 (1979).........000+ passim

Baldwin v. Reese, 541 U.S. 27 (2004) ........cceccseseeesesseerenenees 9

Bragdon v. Abbott, 524 U.S. 624 (1998) .......ccccccceseeseeeeeees 15

Brown v. Gardner, 513 U.S. 115 (1994) ......cccccescesesseseeseeeees 15

General Dynamics Land Sys., Inc. v. Cline, 540 U.S.

Se a isecsniisnsieciatiignaithinhiinnieriinaiiaiaantintinnpatapatinanvent 18

Harrison v. PPG Industries, Inc., 446 U.S. 578 (1980) ......... 8

Hawaiian Elec. Co. v. EPA, 723 F.2d 1440 (9th Cir.

SE cvcinisonsraneneistnntnmnbncnionntaisapitinaineneninre 8

Helvering v. Wilshire Oil Co., 308 U.S. 90 (1939)........00:000 16

IBP v. Alvarez, 126 S.Ct. 514 (2005)...........cscccsecceseeeeseeeseeess 17

Johnson v. United States, 529 U.S. 694 (2000) .............:0000- 15

Lorillard v. Pons, 43 U.S. 575 (1978) .....:.cccccsseesseeeees 6,7, 15

National Muffler Dealers Ass'n, Inc. v. United States,

BEB, Sie Ci i wintrcercenspesisinenseiimnimenimmneiiamitatiin 16

New York v. EPA, 413 F.3d 3 (D.C. Cir. 2005)............ passim

il

New York v. EPA, 431.F.3d 301 (D.C. Cir. 2005)............ 7, 15

New York v. EPA, 443 F.3d 880 (D.C. Cir. 2006)

SOR TRIN AE Wis rshcsindateciniicinsdanitiasteehsiiesbiaildenabaananialpealai 7, 15

Office of Personnel Management v. Richmond, 496

U.S. 414 (1990)... eticcitachabinanlaitonaitnabiadienaiiee 8

Potomac Elec. Power Co. v. EPA, 650 F.2d 509 (4th

ne TNT iesiicinciiiactinincndaedcedebeibhatbeiattipihctaadbinaodaasedectiaiiae 8

Puerto Rican Cement v. EPA, 889 F.2d 292 (1st Cir.

RIES EES ees ee Re CSE passim

Robinson v. Shell Oil Co., 519 U.S. 337 (1997) ......ccccseeseeeee 18

Steiner v. Mitchell, 350 U.S. 247 (1956) ..........ccccsceeseeseeneeeee 17

United States v. Calamaro, 354 U.S. 351 (1957) ...........:00006 15

United States v. Cinergy, 458 F.3d 705 (2006) ............ passim

Whitman v. Am. Trucking Ass'n, 531 U.S. 457 (2001)........... 3

Wis. Elec. Power Co. v. Reilly, 893 F.2d 901 (7th Cir.

ea ditinretincsncnesiiuchiteiionaibbcshnsiionietbiniciicimahipteipiemiiinediitaaeia passim

STATUTES

NUNN ah MN ciscstaicsessiceihiceiesciileitaatinchadidibiahddindedbaati 17

eas Al scastinspcivaiannttchinietsicaushdeiiaiaitabadiieniennpbbuanati 17

Clean Air Act 42 U.S.C. 7401 ef S09. .0......cccccscscseseeees passim

Sa ine CU cbncstieiciiteinincnendicicibaeiiibesinateibiineaes 12, 15, 18

iii

I el 18

Oe Sle PO COE OR inritcescitinctepcnnscsicnssansccininnavtnis 16

ee ees Pe icicaniatenininsnisintndnsbpsusiaesiichdinittiniimippidiniges 16

42 U.S.C. 7419(d)(5) (superseded)...........:ccccccceseeseeeeeees 16

ae FUN icisiesnteenssnidigstisneentetotesiteimcitieadicieiiitundi 18

ns Se csichiniincssinthniinctcipisntnininaiinnaisieipsicistinninigetdl 16

Sa ad SUIT ieciainrtnichcighticiptibgteithtsindenssnisinnenintndgadein 16

A anti IUD teiscintncnestpisesiisiniannancinitetlibeainnienpescies 18, 19

ne SITE cccnrsecichthlintetytavitehiinnnescinnnnisectieninianea 18

Fa ay SET intenetncitianininiocamainnnnisisneninandindinanegele 18

Oe ise i chipertensiciccncicntenintandcitanastnmncnicnnensesivcdsentttietens 16

as Se Ricsccictnsninintetniieninnapstniiediabubanadseiinnats 18

ie es SE ceaccccnanientiiemsensensininenitntinibrsiasedicinastitaita 18

Fe a HERE vcicesstincicteninnscsaniniananmnineneptesecnebennsneininien 18

eae Se ccierncesiccstnechichunsctsntnsincsiscinnticinonatansianstduiniti 18

ee FR sicdncinicsinsirnsntnnncssndsesetentuinne 16

PF aes ST ictinsccohseniiinininsdentanttinneedvonnsidaninnninennibeiita 19

Say PE hacresciincniectniectenninptinigticcncimiminissencvensnvutipiatens 2

ee Pe iienticedcntintagiicnedaletiiniiarecvaticamenteasin passim

I Reais PN ciiceccntitacecnccsnteintiansnnnenoseimnnenonenstianei 8

ns: HEE cinsinincansciinpincbintinsciciesniesdinsnnneenpsnensincti 3, 20

es PED ctiteransciincidcinnnitisictavecnernainitanttninenntante 19

is CIE ncccticcitaictentantspnrisiinadcsinnnnimantieveinisecstant 19

Pies NII iccissitiscinscnonpincabpininsiniimnieinssoeishasniinisenisieinies 15

iV

REGULATIONS

40 C.F.R. (1987):

OD Fe Be UID ccnccccvenciccesnenisianerastisiicitsasinaiiniaan 4

I PRD scccicnvtivsintsscsininaiesibsiapnitaiaiboaiinsattialitiied passim

EE FEU sccinsccininntovepiniepasntinicciineniditcnteianimiaiaaae 12

I Fa ee acivscssntcnnesssintitineiiibieiteaiiiiisinianbaaiel passim

FE ND icsnitencivicssiiniaieriiiatsicinainidaebiaiiieanidl 9

ONTO FE DET vcccinscsccscsessstnssetnnssntalinisncteiio passim

PR Fo DED escnccencersnsnenscicccsnttamnsetmnanninnsiaiaeigl 4

Section 31.166(O 2G NG) ........ersccccoccceressvseresssorseesoseees 20

Section 51.166(b)(2)(iii)(D)-(€).......ccsecseseesesesesesseneereneeees 19

Section 51.166(b)(2)(i1i)(f) ..........cecscceseereseeeseseeeeeeees 13, 19

ORIN Be te chcnteisssscepcsntmenenttinanieatiumsnniaiai passim

AIR FERED sivisintciiscsiintnmcitasiniltiarsineimiiiaman ial 4

OSE Sh. TOG accnccccstensccncnscmvssenassiniuetiiiia 11,14

I FS BR ccncessiesinstatitiiitaistidiabiiniaiaiainid 1]

AD Fee ciccicecnisnincinistsitascaiideiinidaliiheieedieniaan 12

REI FE. A accicsestcncsnssslsssseniitiiteastiiecibeageiliaiaaadldl 4

ama SE SOG ccscnccsncnvtsiiitaiitleuseianiaiasipendiidideal =

Ef, TIT NAS 4

ED FE Be scisnstensccnititinitstbineibiiniassediatin! 3,4, 12

nea FE AGRI occocisiticieentisninicialassainastiaats passim

ERE RMT RE BM nS ar 19

Sec cncirscecsssnnssnnscnonecnennsenensnentstannete 13

EE Ser ee eRe 10

TET ciintisictininsninitiitsininicibinpniennienteiniiataeneinend 4

hierniana inndindnpanmentinsretitnatatin passim

EES ee 9

| ES ENS ae ae eT 4, 15,17

NT ieeicierireistetihahantneitaieceiaiapetnacinttagemesnemnante 15

EE 16

LEGISLATIVE MATERIALS

Pub. L. No. 95-95, § 109(f), 91 Stat. 685 (1977)... 16

Pub. L. No. 95-95, § 129(a)(1), 91 Stat. 685 (1977)............. 16

Pub. L. No. 95-95, § 223, 91 Stat. at 764 ooo... ccccececeeseeeeenes 16

nL re ia csiciicineneenlindenibiencennninnsioonanianl 16

H.R. Rep. No. 95-297 (1977)..........ccccceeseee: sinilietencteiiiitcnni 17

ADMINISTRATIVE MATERIALS

38 Fed. Reg. 18986 (July 16, 1973) ...........cccccccceseceeseeeeeeeeees 13

39 Fed. Reg. 36946 (Oct. 15, 1974) .......cccccccecceeseeeeenenes 13, 16

5D Fed. Raw, 42510 (Das. S, 1974)......e.ecceccersscssssessessessececee 13

41 Fed. Reg. $5524 (Dec. 21, 1976) ..............ccscssesceseeseeeeees 15

42 Fed. Reg. 57459 (Nov. 3, 1977) ..n...ccccecsecseeeeceneenees 10, 16

44 Fed. Reg. 51924 (Sept. 5, 1979) ......ccccccccccesseseeeeeennees 9,10

45 Fed. Reg. 52676 (Aug. 7, 1980) .........ccccccceescesseeeees passim

48 Fed. Reg. 38742 (August 23, 1983) ...........ccccccscseceseeeeenens 11

57 Fed. Reg. 32314 (July 21, 1992)... aia 7,14

61 Fed. Reg. 38249 (July 23, 1996)..........cccccceeseeseeseeeeeeeenes 5

67 Fed. Reg. 80186 (Dec. 31, 2002)..............rcccecccsessesscsssoress 6

70 Fed. Reg. 61081 (Oct. 20, 2005)..........ccccccceseeeeeeeeees 6, 19

vil

INTRODUCTION

Making no effort to defend the Fourth Circuit’s reasoning,

Duke seeks to defend its result by urging that EPA’s 1980

PSD regulations “compel the conclusion that a PSD ‘major

modification’ first requires an NSPS ‘modification’ — a

change that causes an increase in a unit’s hourly emissions

rate.” Br. 33. Duke advances this theory both as a purported

basis to avoid Section 307(b), and a substantive argument that

the agency’s reading of its rules is impermissible.

This makes for painful reading: In fact, the 1980

regulations and their preamble make no mention whatsoever

of the NSPS maximum hourly rate test Duke now asserts to

be a prerequisite for every PSD “major modification.”

Instead, they set forth in considerable detail a PSD

applicability test triggered by increases in “actual” emissions

measured in “tons per year,” regardless of whether the same

project would satisfy the NSPS regulatory test. Duke not

only “could have” challenged these patent features of the PSD

rules, it did: Duke was party to D.C. Circuit proceedings that

began with Duke and others (in 1982) importuning EPA to

add an hourly-rate increase requirement to the PSD

regulations, and ended with the court rejecting Duke’s

arguments that the Act mandates an hourly rate test for PSD.

New York v. EPA, 413 F.3d 3 (D.C. Cir. 2005). Duke’s

identical statutory arguments here are “not subject to review”

in this “enforcement action,” 42 U.S.C. 7607(b). (They are

also meritless.)

Duke’s merits brief unveils a deus ex machina — two

“general” regulatory provisions now said to import an hourly

rate increase requirement into PSD. See Br. 28 (1980 “rules

as promulgated clearly provided that a PSD ‘major

modification’ first requires an NSPS ‘modification.””) (citing

40 C.F.R. 51.100, 52.01(d)). As with many 13th-hour

arguments, it was for good reason Duke did not cite either

provision in its 10]-page brief below (or its opposition to

certiorari). Neither even arguably “provides” — “clearly” or

otherwise — that an NSPS hourly rate increase is a prerequisite

for PSD applicability.

I. DUKE COULD HAVE OBTAINED D.C. CIRCUIT

REVIEW OF ANY CLAIM THAT EPA HAD TO

USE AN HOURLY RATE TEST FOR PSD

As our opening brief explained, the Fourth Circuit’s

decision represents a flagrant violation of Section 307(b), the

provision allocating judicial authority over CAA rulemakings.

Although Duke protests (Br. 29) that this provision does not

disable a court in an enforcement proceeding from rejecting

EPA’s interpretation of a regulation, that is not what

happened below: the Fourth Circuit did not interpret the

regulations, deeming their text “largely irrelevant.” Pet. App.

llan.3. Its decision rested on a statutory ruling that vitiated

nationally applicable rules. As we explained, the vice of that

conclusion (beyond being wrong about the CAA’s substance)

is not simply that the question was statutory, but that it was a

statutory objection that could have been — and was, in fact —

presented to the D.C. Circuit. Making no effort to defend the

Fourth Circuit’s overstep, Duke even denigrates critiques of

the court’s reasoning as attacks on a “straw man” that does

not reflect “Duke’s position.” Br. 31.

Duke calls EPA’s view that the 1980 PSD regulations do

not require an increase in hourly rates an “enforcement

interpretation” (Br. 25) that Duke “could not have

challenged” (Br. 28) in the D.C. Circuit.’ But a glance at the

' It is not true that the United States raised Section 307(b) “[o]nly after the

Fourth Circuit signaled in a supplemental briefing order that it was not

accepting EPA's position{.]” Duke Br. 26-27. Its opening brief (US CA4

Br. 23) pointed to the D.C. Circuit's “exclusive jurisdiction,” and when, at

oral argument, the panel questioned the regulations’ consonance with the

CAA, counse! for the United States observed that “reject{ing] portions of

the PSD regulations” was “something only the D.C. Circuit has

jurisdiction to do” under “Section 7607” of the CAA. CA4 Arg. Tr. 15.

When one judge responded, “[t}his would not be that. This would be

statutory — this is Congress telling us that the definition must be the

same,” counsel explained that “the proper vehicle for making that claim

plain language of the regulations reveals the utter

insubstantiality of this plea. The 1980 PSD regulations

contain no requirement that there be an increase in a source’s

hourly rate for PSD to apply. To the contrary, they direct, in

specific and technical detail, that emissions be measured in

“tons per year” of “actual” emissions, 40 C.F.R. 51.166(b)(3),

(21). EPA’s adoption of this “actual, annual” emissions

increase test in 1980 was “action” “of which review could

have been obtained” in the D.C. Circuit. See 42 U.S.C.

7607(b), (d). As the Seventh Circuit recently explained in

rejecting a challenge to the 1980 PSD regulations advanced

by Duke’s affiliate:

Cinergy’s principal argument * * * is that Congress

required that the regulation define “modification” as a

change in the hourly emission rate. Since the [1980 PSD]

regulation does not define it so, this seems an attack on the

validity of the regulation rather than an argument about its

meaning, and [pursuant to CAA Section 307(b)] issues of

validity * * * are beyond the jurisdiction of a regional

circuit to resolve.

United States v. Cinergy, 458 F.3d 705, 709 (2006). That it

is necessary here, as in Cinergy, to look preliminarily at the

regulations to see whether a party’s claim “could have been

obtained” in the D.C. Circuit, is not, as Duke seeks to argue

(Br. 30-31), reason to ignore Section 307(b) and proceed to

the merits of challenges that could have been (and here,

were) presented to that court.

Here, Duke could not have been lulled by the regulatory

text. A requirement that no project can trigger PSD unless it

first triggers the NSPS regulatory test would have been an

enormously important feature of the 1980 rules. Yet the

comprehensive regulatory text and preamble contain not even

would be in a challenge to the regulations which supposedly violated that

statutory mandate.” /d. The court ordered supplemental briefing on the

statutory issue three months later. See JA 9; Pet. App. 21a.

a “mousehole” for such a brontosaurian proviso. Whitman vy.

Am. Trucking Ass'n, 531 U.S. 457, 468 (2001). The PSD

rules provide elaborate and specific instructions on how to

determine whether a physical or operational change has

caused a PSD-triggering emissions increase. 40 C.F.R.

51.166(bX(3), (21) (1987). See Petr. Br. 10-11. When, in

those regulations, EPA intended to reference NSPS standards,

it did so in clear terms.? The PSD rules make no mention of

the detailed regulatory test for NSPS modifications in 40

C.F.R. 60.14(b). Unsurprisingly, the four appellate panels to

have examined the 1980 regulations’ text, cf. Pet. App. Ila

n.3, have readily concluded that the test for PSD applicability

— turning on increases in actual, annual emissions — differs

“fundamentally” from the NSPS hourly rate test.’

The 1980 preamble highlighted that EPA had

“change[d]” the PSD applicability standards set out in the

1979 proposed rule, 45 Fed. Reg. 52676, 52680 (Aug. 7,

1980), based on the D.C. Circuit’s reading of the statute in

Alabama Power v. EPA, 636 F.2d 323, 399-400 (1979), and

EPA’s own scrutiny of the “language of the statutory

definition,” which embraces “any” physical or operational

“change” that “increases the amount of any air pollutant

emitted by such source.” 45 Fed. Reg. at 52700 (emphasis

EPA’s). The 1980 PSD rules, in turn, define a “major

modification” as a (1) “physical change in or change in

? See, e.g., 40 C.F.R. 51.166(b)(12) (BACT under PSD must comply with

“any applicable standard under 40 C.F.R. 60 and 61”); 40 C.F.R.

51.166(b)(16)(i) (“allowable emissions” for PSD must be at least as

stringent as “standards as set forth in 40 CFR Parts 60 and 61”); see also

id. 51.166(b)(17), 51.166()\(1), 51.18G)(1 Xxvii).

> Wis. Elec. Power Co. v. Reilly, 893 F.2d 901, 913, 915 (7th Cir. 1990)

(“WEPCo.”) (NSPS and PSD rules measure emissions increases in

“fundamentally distinct manner,” namely, hourly rates versus actual,

annual emissions). See Cinergy, 458 F.3d at 708 (“the natural reading of

the regulation is that any physical change or change in operating methods

that increases annua! emissions is covered”); New York, 413 F.3d at 18;

Puerto Rican Cement v. EPA, 889 F.2d 292, 297 (1st Cir. 1989).

method of operation” at a source (2) that “results in a

significant net emissions increase,” 40 C.F.R. 51. i66(b)(2)(1),

‘ and define those constituent phrases in actual annual terms,

id. at (b)(3)(i) (net emissions increase defined as “actual

emissions”), (b)(21)(i) (“actual emissions” from unit specified

as “rate” in “tons per year”).

The 1980 preamble is likewise bereft of any support for

Duke’s assertion that “a PSD ‘major modification’ first

requires an NSPS ‘modification.”” Br. 28 (citing 40 C.F.R.

51.100, 52.01(d), neither of which, 2s explained below,

“provided” anything of the kind). To the contrary, the

preamble explains that, once a physical or operational change

is identified, “[t]he first step in determining whether a ‘net

emissions increase’ would occur is to determine whether the

physical or operational change in question would itself result

in an increase in ‘actual emissions.’”” 45 Fed. Reg. at 52698;

see id. at 52705 (EPA examples discussed in U.S. Br. 23).

The actual emissions “rate,” as explained, is expressed in

“tons per year.” 40 C.F.R. 51.166(0)(21)(i).

In 1982, a group of D.C. Circuit petitioners, including

Duke, see Petr. Br. 12-13; U.S. Br. 8, 32-33, entered into a

settlement whereby EPA agreed to propose for comment a

change in the 1980 regulations’ definition of “major

modification,” removing the references to “actual” emissions

“wherever it occurs in paragraph (b)(3)” (the definition of

“net emissions increase,” now 40 C.F.R. 51.166(b)(3)), and to

add new regulatory language expressly providing that PSD

would not apply absent an increase in hourly emissions rates.

See Petr. Br. 13. Settlement Agreement, D.C. Cir. No. 79-

1112, Exh. B, Sec. A, § 1 (1982). This change would have

made an increase in hourly emissions rates a sine qua non for

PSD applicability - just what Duke now argues, without

colorable textual basis, the 1980 regulations themsclves

already did. See, e.g., Duke Br. 33.‘

* See 61 Fed. Reg. 38249, 38269 (July 23, 1996) (proposed language

would have “eliminate[d] a source's level of operations as a factor when

Duke actually obtained D.C. Circuit review of the

“divergence” between the 1980 PSD regulations from the

NSPS regulatory test, and the D.C. Circuit rejected its

arguments. See New York, 413 F.3d at 20.° Attempting to

downplay that review, Duke asserts that New York

“principally involved a challenge to the 2002 rules,” and that

the D.C. Circuit merely disagreed “that Congress adopted

NSPS rules wholesale into PSD.” Br. 32 (citing 413 F.3d at

19-20) (emphasis added). But the court rejected Duke’s

challenge to the “1980 and 2002 rules,” specifically, the

claim “that modification must have the same regulatory

determining whether a proposed change will result in an increase” and to

require regulators to “completely disregard{]” post-change utilization).

See also 67 Fed. Reg. 80186, 80205 (Dec. 31, 2902); 70 Fed. Reg. 61081,

61098 (Oct. 20, 2005) (referring to settlement test as “NSPS-like”). Duke

ignores the settlement's terms, failing even to note (Br. 11) it was a party.

Duke does, however, try to explain away — as having “nothing to do with

Duke's position here” (Br. 11) - the D.C. Circuit brief filed in February

1981 by General Motors er al. (Petr. Br. 12), which complained that,

under EPA's 1980 rules, PSD could apply “even though the source's net

capacity to emit remains constant or declines.” Br. Ind. Pet’rs On Actual

Emissions 5-6 (No. 79-1112; filed Feb. 12, 1981) (“GM Brief”) (emphasis

added). Yet Duke's central position is that PSD should only apply where

there is an increase in “basic emissions capacity” (Duke Br. 2), or a

facility's “design emitting capacity” (Duke Br. Opp’n 4). Seeking to blur

the incompatibility between this position and the D.C. Circuit's rulings

that PSD turns on increases in “actual” emissions, e.g., New York, 413

F.3d at 38-40, Duke unveils an oxymoronic new phrase, “actual emissions

capacity.” Br. 22, 32 n.13; see also Br. 5 (“actual emissions capabilities”).

We have requested leave under Rule 32.3 to lodge the settlement

agreement and GM Brief with the Clerk.

* Duke’s arguments here are eerily similar to those made before the D.C.

Circuit. Compare, e.g., Joint Reply Br. Ind. Pet'rs 6 (No. 02-1387; filed

Sept. 20, 2004) (“in adopting the NSPS and 1974 PSD meaning and usage

of modification into NSR, Congress selectively modified certain of the

rules while adopting others (including the well-established definition of

‘modification’) without change. Such selectivity is strong evidence of

Congress's intent to incorporate in the statute those regulatory provisions

that it did not specifically change. Cf. Lorillard v. Pons, 43 U.S. 575, 582

(1978)") with Duke Br. 44.

meaning for NSR as prevailed for NSPS in 1977.” 413 F.3d

at 19-20.° As the D.C. Circuit itself has explained, New York

“rejected industry's eontention that Congress ratified the

[NSPS] regulations on ‘modification’ in the 1977

amendments * * * * [and] their position that ‘modifications’

require an increase in maximum emissions rates[.]" New

York v. EPA, 443 F.3d 880, 889 (D.C. Cir. 2006) (New York

Il) (citing 413 F.3d at 19-20, 40 and 431 F.3d 801, 802-863

(2005) (Williams, J., conc. den. reh’g. en banc)). No party

petitioned for certiorari in New York, though the time to file

elapsed months after the petition in this case was filed.

Nor can Duke escape from New York by pointing to (Br.

32-33) the D.C. Circuit’s refusal to address, as unmpe,

industry's challenge to an “allegedly new interpretation of the

1980 rule contained in the preamble to the 2002 rule,” 413

F.3d at 20. See also id. at 21 (noting that preamble language

“appears to be — as EPA claims ~- no more than a short-hand

reference to the 1980 rule, not a formal interpretation”).’ The

* Duke and other petitioners also urged that the 1980 final rule should be

vacated as an impermissible departure from the 1979 proposal, which they

read to restrict PSD to changes that increased units’ capacity to emit.

Joint Br. Ind. Pet'rs 35-37 (D.C. Cir. No. 02-1387; filed May 11, 2004).

” Duke's repeated assertions before this Court that EPA's position is an

“enforcement interpretation” (¢.g., Br. 22, 24, 25, 27, 33, 41, 46) — ansing

from an “abrupt reversal of position” by EPA occurring “in November

1999” (Br. 16) — are inconsistent with its own D.C. Circuit submissions:

“Beginning in the late 1980s, EPA, without even acknowledging its

contemporaneous interpretation of the 1980 NSR rules, attempted to

rewrite the rules by reinterpreting them as having repealed the first step of

the emissions increase test-—i.e., the requirement that there be NSPS

modification activity.” Joint Br. Ind. Pet’rs 16 (D.C. Cir. No. 02-1387;

filed May 11, 2004) (emphasis added); see New York, 413 F.3d at 15

(discussing WEPCo., Puerto Rican Cement, and 1992 preamble, 57 Fed.

Reg. 32314, 32328 (July 21, 1992)). Similarly, the record reflects that

since at least 1990, Duke and the Unlity Air Resources Group (of which

Duke is a member and with whom it shares counsel) have known that

increased hours of operation, or “utilization,” could trigger PSD even if

hourly rates do not increase. JA 243, 272, 276, 380, 496.

D.C. Circuit did not regard as unripe challenges to the 1980

regulations themselves — including their “divergence” (id. at

20) from the NSPS hourly rate regulation; it rejected those

challenges on the merits, id. at 18-20. If there were others the

D.C. Circuit did “not resolve” (Duke Br. 32), it was because

Duke did not present them, see 413 F.3d at 20. Any CAA or

APA-style challenge to the 1980 regulations’ patent failure to

impose an NSPS-like hourly rate “trigger” for PSD would

have been cognizable by petition for review immediately

upon promulgation, and is barred in an enforcement action.

42 U.S.C. 7607(b)(2). If “fundamental fairness” (Duke Br.

29) speaks to these circumstances, it calls for enforcement of

that limitation, which governs, and protects, the large range of

interests affected by CAA rulemakings. See, e.g., Br. of

Amici Curiae New Jersey et al. 5-20; Br. of Amici Curiae

STAPPA et al. 15-18; Br. of Former EPA Admin'rs 1-13.*

* Duke’s narrative of regulator acquiescence rings decidedly hollow. That

a letter from Duke claiming its units would be in “extended cold

shutdown” with “minimal expenditures” (JA 187), prompted agreement

from state environmental officials that PSD did not apply (Duke Br. 18), is

meaningless given the projects actually Duke undertook — work more

accurately explained to utility regulators from whom Duke sought

increased rates. See Petr. Br. 19-20 (quoting Duke's statements of need

for “total rehabilitation” of “geriatric” units “too dangerous to operate”

and otherwise due to be “retired and scrapped”). Also unavailing are

routine field inspection reports (Duke Br. 18): A state inspector explained

that PSD “determinations are not made in the field” (JA 363) and require

testing and analysis such that “just from the inspection and looking at the

boiler, you couldn’t say it is subject to PSD or not.” JA 347. See also JA

313, 323, 333. Despite the scale of Duke’s Plant Modernization Program,

it never sought an applicability determination from EPA. Cf. Harrison v.

PPG Industries, Inc., 446 U.S. 578 (1980); WEPCo, 893 F.2d at 913;

Puerto Rican Cement, 889 F.2d at 297-298; Potomac Elec. Power Co. v.

EPA, 650 F.2d 509 (4th Cir. 1981); Hawaiian Elec. Co. v. EPA, 723 F.2d

1440 (9th Cir. 1984). To the extent claims of agency acquiescence or

prejudicial delay in initiating enforcement could have any legal relevance

here, they would not bear on the validity of the EPA’s regulations or

interpretation, but on affirmative defenses such as estoppel or laches—

defenses Duke asserted below. But see Office of Personnel Management

v. Richmond, 496 U.S. 414, 419-24 (1990).

Il. DUKE’S NEW READING OF THE REGULATIONS

IS OBVIOUSLY WRONG

Duke contends “that an NSPS ‘modification’ is required

before a [PSD] ‘major modification’ analysis can occur,” so

that PSD cannot apply unless “an increase in a unit’s

maximum achievable hourly emissions rate occurs.” Br. 32

(citing 40 C.F.R. 60.14, the NSPS regulation). Although the

PSD regulations’ highly specific instructions on how to

measure emissions “increases” for PSD purposes say nothing

of an hourly rate “trigger,” Duke claims it materializes by

operation of two “general” regulatory provisions, which are

said to make satisfaction of the NSPS hourly rate increase test

a prerequisite for any PSD “major modification.” See Duke

Br. 2, 10, 28, 32, 34 (citing 40 C.F.R. 51.100, 52.01(d)).

Duke did not present this account of the regulations in its

briefs in the Fourth Circuit or its opposition to certioran. In

neither place did Duke cite either of the regulatory provisions

now assigned the monumental role of importing the NSPS

hourly rate standard into the PSD regime.” To the contrary,

its opposition (correctly) referred to “major modification,” 40

C.F.R. 51.166(b\(2), as “the definition of ‘modification’ for

PSD.” Duke Cert. Opp. 7-8. See Baldwin v. Reese, 541 U.S.

27, 34 (2004) (waiver under Sup. Ct. Rule 15.2); Aetma

Health Inc. v. Davila, 542 U.S. 200, 212 n.2 (2004) (same).

This late-breaking regulatory argument is, in any event,

completely disconnected from the 1980 PSD regulations,

which nowhere incorporate — as the “tigger” (see Duke Br.

32) for a PSD modification, or for any other purpose — the

NSPS regulations’ hourly rate provisions in Part 60.14.

Rather, the definition of “major modification” in 40 C.F.R.

51.166(b) sets forth a free-standing and complete standard for

judging whether activity is subject to PSD, and the

? In Cinergy, unlike this case, the appellant did argue that a PSD “major

modification” must, by operation of 40 C.F.R. 52.01(d), first constitute a

NSPS “modification” — this was among the arguments the Seventh Circuit

dismissed as “makeweights.” Cinergy, 458 F.3d at 711. See Duke Br. 19.

regulations carefully define each of the terms used therein.

The definition prescribes what a PSD “‘[m]ajor modification’

means,” id. 51.166(b)(2)(i) (emphasis added); activities that

satisfy the criteria therein are subject to PSD, id. 51.166(i)(1).

The term “major modification” was introduced in the

1977 PSD regulations to correspond to the PSD regulatory

term “major source,” and EPA’s initial regulations required

that emissions increases from a change be large enough

independently to reach the statutory thresholds for “major

emitting sources.” See 42 Fed. Reg. 57459, 57480 (Nov. 3,

1977). After Alabama Power disapproved this “regulatory

definition of modification,” 636 F.2d at 399, EPA retained

“major modification” as the PSD implementation of the CAA

term “modification” in the revised rules. See, e.g., 45 Fed.

Rey. at 52704 (providing examples of “the way in which the

definition of modification works’) (emphasis added); id. at

52705 (examples of changes that “qualify as

modifications”). '°

The 1980 regulations define “major modification” in

high-resolution detail, leaving no room for Duke’s subliminal

NSPS trigger. According to Duke, before a “major

modification” can occur, “first” a “unit's” hourly rate must

increase. Br. 33. But the 1980 regulations explicitly define

“actual emissions” at the “unit” level in “tons per year,” 40

C.F.R. 51.166(b)(21)(i); see 45 Fed. Reg. at 52698 (“the first

step” is to “determine whether the physical or operational

change in-question would itself” increase annual emissions).

At the next stage, the regulator must sum “any increase in

actual emissions from a particular physical change” with “any

other increases and decreases in actual emissions at the

source.” 40 C.F.R. 51.166(b)(3)i)(a), (b) (emphases added).

'° EPA did not use Duke’s allegedly NSPS-referring definitional structure

for PSD prior to 1980, either. See, e.g., 44 Fed. Reg. 51924, 51948 (Sept.

5, 1979) (proposed regulations, describing “major modification” as “the

definition of ‘modification’ * * * when used in the Act in reference to a

major stationary source”).

10

Thus, a change that causes a significant increase in a unit's

actual, annual emissions qualifies as a “major modification”

even if there is no decrease to be “netted” against it, and even

if there is no increase in hourly rates. An NSPS

“modification” is nowhere made an interceding “first step”

for PSD applicability.''

Duke’s account is also inconsistent with the 1980

preamble’s explanation of why the “actual emissions”

standard supplanted the “potential emissions rate” test

proposed in 1979. Both the Alabama Power court, and EPA

itself, emphasized the broad statutory definition of

“modification” — and its interplay with the express statutory

requirements for PSD - in concluding that PSD must capture

“actual” emissions increases. See 636 F.2d at 400; 45 Fed.

Reg. at 52700; see also id. at 52718 (“Use of actual emissions

for increment consumption is consistent with using an actual

emissions baseline for defining a major modification.”); id.

(linking choice of “actual emissions” to effort to “reflect

actual air quality” in area).

Against the precision of the 1980 regulations, the two

“general” provisions Duke cites as the basis for its argument

that the PSD regulations incorporate the NSPS hourly rate

test, 40 C.F.R. 51.100 and 52.01(d), do very little. Section

51.100 provides that “[a)]s used in this Part, all terms not

defined herein will have the meaning given them in the Act.”

This provision applies only to undefined terms, and the

specific and detailed “major modification” definition in

51.166(b)(2)(i) is the regulatory definition of the statutory

'' The January 1981 Reich memorandum (cited in Duke Br. 11-12, 28)

makes this very point. Reich explained that the “first” step in the PSD

inquiry is “to determine if there will be a significant net emissions

increase from the modification itself” with the emissions “rates” expressed

as tons per year. JA 36. Reich noted that this understanding would be

published in the Federal Register (JA 36), and it was: 48 Fed. Reg. 38742,

38746 (Aug. 23, 1983) (“first step” is to sum emissions from change at

unit); see id. n.14 (“tpy” emissions at all steps, including for units).

1]

term “modification” for PSD purposes — as has been

recognized by EPA, supra, p. 10; by courts both before and

after the 1980 rulemaking, see Alabama Power, 636 F.2d at

399: New York, 413 F.3d at 20; and (until its merits brief) by

Duke itself, see Duke Cert. Opp. 7-8. Ever since the term

“major modification” was introduced for PSD in 1977, EPA

has invariably treated it as the operative NSR regulatory

definition of the statutory term “modification,” and there 1s no

sign, anywhere, that EPA understood or intended that detailed

regulatory definition to be dramatically qualified by Section

51.100’s bare reference to the statute. '”

Equally unavailing is Duke’s reliance on Section

§2.01(d), which, uncited in the Fourth Circuit or in Duke’s

opposition to our petition, now rates a “passim,” Duke Br. x.

Promulgated for the 1974 administrative PSD program,

Section 52.01(d) provides that “the phrases modification or

modified source mean any physical change in, or change in

the method of operation of, a stationary source which

increases the emission rate of any pollutant * * * *” Section

§2.01(d) applies only to “this part” - Part 52 - and, as Duke

has recognized, this case is governed by Part 51, establishing

the PSD requirements for state plans. See Duke Cert. Opp. 7

n.15 (state rules “are identical to the federal 1980 PSD rules,

40 C.F.R. § 51.166 (1987)”). The Part 51 regulations instruct

that “All state plans shall use the following definitions for

purposes of this section,” and that states may use different

definitions only if they are “at least as stringent, in all respects

as the corresponding definitions below.” 40 C.F.R. 51.166(b)

2 Even if it applied, 40 C.F.R. 51.100 would get Duke nowhere. Inserting

the CAA 111(a)(4) language into the regulatory definition of “major

modification,” would still predicate PSD applicability (albeit in oddly

prolix fashion) on a “net emission increase,” measured in “tons per year”

of “actual emissions.” See 40 C.F.R. 51.166(bX3), (6)(21). Nor

(obviously) does the inclusive definition of “construction” in the PSD

regulations, id. 51.166(b\8) (cited in Duke Br. 23, 33, 35, but also not

cited below) establish an hourly rate-increase precondition for PSD.

12

(emphasis added). "”

In any event, by its terms, Section 52.01(d) does not

purport to mandate the use of the NSPS (or any other) “hourly

rate” test. An “emission rate” can be measured in hourly,

daily, annual, or other terms ~ and the 1980 regulations

explicitly measure emissions “rates” in “tons per year.” See

40 C.F.R. 51.166(b)(21)(i); (b(23)(i); see also 38 Fed. Reg.

18986, 18996 (July 16, 1973) (proposed PSD rules describing

“rate” in “annual * * * tons” of pollution) (emphasis

added).'* The reference to “rate” in Section 52.01(d) does

not vary the specific and detailed methodology for identifying

PSD emissions increases set forth in 40 C.F.R. 51.166(b).

Duke also invokes the limited exclusion for “[a]n increase

in the hours of operation or in the production rate.” 40 C.F.R.

51.166(b)(2)(iiif). But, as its plain language makes clear,

this provision addresses what constitutes a “physical change”

or “change in the method of operation,” not how emissions

from such a change should be measured. See, e.g., Petr. Br.

36-37; 45 Fed. Reg. at 52704; WEPCo., 893 F.2d at 916 n.11;

Puerto Rican Cement, 889 F.2d at 298; 45 Fed. Reg. at 52698

(under regulations, physical and operational “changes” “do

not encompass certain specific types of events”). When such

a “change” is present — in an extensive plant “modernization,”

for example — the regulations require that “[a}ny increase in

> Alabama Power specifically described Section 52.01(d) as having been

“superseded,” 636 F.2d at 348 n.24, and EPA never referenced it in the

1980 preamble. See 45 Fed. Reg. 52686 (1980 regulations would in time

“displace the old” Part 52 regulations “entirely”). It is quite implausible

that EPA - crafting new regulations in direct response to Alabama Power

~ would have silently assigned this provision an enormously important

role in determining PSD coverage.

'* When EPA first promulgated Section 52.01(d) in 1974, the NSPS

regulations did not provide for an hourly rate test. See 39 Fed. Reg.

42,510, 42,513 (Dec. 5, 1974) (explaining in the PSD rule preamble that

the Administrator was stil] evaluating how to define “modification” for

NSPS purposes); 39 Fed. Reg. 36946, 36947 (Oct. 15, 1974) (soliciting

comment on the unit of measurement for emissions increases under NSPS,

and including “year{ly]” increases as an option).

13

actual emissions from a particular physical change or change

in the method of operation” be counted, in “tons pér year.”

40 C.F.R. 51.166(b)(3)(i)(@), (621d).

Duke pleads that “changes in hours operation, standing

alone, are not changes in the ‘method of operation,”” so “the

‘hours of operation’ exclusion must do more.” Br. 39-40.

But this overlooks the breadth of the statutory definition,

which could be read to include stand-alone increases in

utilization, as EPA as expressly noted in 1980, 45 Fed. Reg.

at 52705 (“The increase in hours of operation is a change in

the method of operation.”). See 57 Fed. Reg. at 32316

(because statutory definition could “encompass the most

mundane activities,” agency has defined “modification”. to

“include common-sense exclusions from the ‘physical or

operational change’ component of the definition”). The

exemption — also part of 1992 and 2002 PSD rules that even

Duke does not claim to employ an hourly rate test —- confines

coverage in an important way, but by its plain terms does not

preclude consideration of emissions increases resulting from

physical or operational changes.

Il. THE ACT DOES NOT REQUIRE USE OF

AN HOURLY RATE MEASURE FOR PSD

EMISSIONS “INCREASES”

Duke’s brief (Br. 42-47) mingles two separate statutory

arguments: that Congress ratified some specific pre-existing

NSPS or administrative PSD regulations in the 1977 PSD

enactment (the argument rejected on its merits in New York),

or that it commanded that the statutory term “modification”

be implemented by identical regulations in both NSPS and

PSD (the argument accepted by the Fourth Circuit in this

case, but found to have been waived in New York). The two

arguments, inconsistent with one another (see Pet. App. 18a),

are alike in two respects: they are both jurisdictionally

improper challenges to the regulations EPA actually adopted,

and they are both, in any event, meritless.

The statute does not support Duke's argument (Br. 42-44)

14

that the 1977 PSD enactment’s use of the Section 111(a)

statutory definition of “modification” constituted a command

that EPA must use a regulatory standard from the 1975 NSPS

regulation. Nowhere does the 1977 PSD enactment purport to

incorporate any prior regulation concerning how to measure

emissions increases. See New York, 413 F.3d at 18-20;

Cinergy, 458 F.3d at 710.

The 1975 NSPS regulations contained “different” and

“possibly inconsistent” definitions of “modification,” only

one of which contained the hourly rate standard. New York,

413 F.3d at 19 (citing 40 C.F.R. 60.2(h), 60.14(b)); see id. at

12 (“{iJn its various permutations, this regulatory framework

had not been long in place when” Congress enacted the 1977

legislation). The regulations in place under the 1974

administrative PSD program contained other definitions,

lacking any reference to hourly rates, 40 C.F.R. 52.01(d),

while the non-attainment NSR program contained yet a

different test. See New York, 431 F.3d 801, 802 (Williams, J.,

conc. den. reh’g en banc) (discussing “potential allowable

emissions” test at 41 Fed. Reg. 55524, 55528 (Dec. 21,

1976)). There is no evidence Congress was aware of any of

these regulatory definitions (none of them judicially

reviewed), let alone that it intended to entrench any of them

as statutory requirements. See Brown v. Gardner, 513 U.S.

115, 121 (1994) (citing United States v. Calamaro, 354 U.S.

351, 359 (1957)); Johnson v. United States, 529 U.S. 694, 726

(2000) (Scalia, J., dissenting); cf. Bragdon v. Abbott, 524 U.S.

624, 631-32 (1998) (finding regulatory ratification where

Congress not only was aware of but expressly referred to and

incorporated regulations) (citing 42 U.S.C. 12201(a)).

Congress, in 1977, did adopt other pre-existing

regulations as statutory requirements. It amended the

definition of “modification” in CAA Section 111(a) by adding

a new legislative exemption for coal-conversion projects

approved under Section 119(d) of the Act, one that tracked a

15

pre-existing 1975 NSPS regulatory exemption.’ And

“elsewhere” in the amendments, Congress “incorporate[d]

regulatory provisions expressly by reference.” New York, 413

F.3d at 19 (citing Pub. L. No. 95-95, § 129(a)(1), 91 Stat. 685,

745 (1977); 42 U.S.C. § 7502 note); see also Pub. L. No. 95-

95, § 223, 91 Stat. at 764 (adding since-superseded CAA §

211(g)(3), 42 U.S.C. 7545(g)(3) (1978), referring to specific

EPA regulations concerning fuel additives). When Congress

wished to incorporate existing EPA regulations in the 1977

legislation, it knew how to do so.

Section 168 of the Act (cited in Duke Br. 44) provides

that pre-1977 PSD regulations “shall remain in effect” “[u)ntil

such time as an applicable implementation plan is in effect for

any area” that complies with new statutory PSD provisions,

and would continue to govern projects commenced between

July 1, 1975 and August 7, 1977. 42 U.S.C. 7478(a), (b). It

manifestly expresses an intent to phase out the preexisting

regulations, not capture them in statutory amber. See

Alabama Power, 636 F.2d at 363; 42 Fed. Reg. at 57459,

$7471 (discussing “comprehensive changes” to regulations

required by 1977 statute’s “comprehensive new

requirements”). There is no evidence that Congress, outside

the narrow provisions just discussed, intended for pre-1977

regulations to constrain EPA’s rulemaking authority. See

National Muffler Dealers Ass'n, Inc. v. United States, 440

U.S. 472, 485-86 (1979) (citing Helvering v. Wilshire Oil Co.,

308 U.S. 90, 101 (1939))."*

1S See Pub. L. No. 95-95, § 109(f), 91 Stat. 685, 703 (1977), codified at 42

U.S.C. 7411(aX8) (a “conversion to coal,” under CAA § 119(d)(5) “shall

not be deemed to be a modification for purposes of paragraphs (2) and (4)

of this subsection.”). See 40 C.F.R. 60.14(e)(4) (1975) (NSPS regulation

exempting “conversion to coal” under § 119(d)(5)); 39 Fed. Reg. 36946,

36948 (Oct. 15, 1974).

'* Duke quotes (Br. 46) the House Report for the proposition that “existing

sources * * * and their emissions’ [sic] capacity are “grandfathered.”

H.R. Rep. No. 95-294, at 144 (1977), but fails to mention that Congress

rejected the House bill’s approach, which premised PSD “baseline

16

Nor is there any merit to Duke’s other account of what

the Act is supposed to have unambiguously provided —

Duke’s passing nod (Br. 43) to the Fourth Circuit’s Rowan-

based theory that Congress intended that regulations

implementing the term “modification” for NSPS and NSR

must be the same. Duke’s own commitment to this theory is

weak indeed, for it advocates an NSR “modification” test

substantially different from the NSPS test. Duke fails to

explain why Congress would impose as an unyielding

statutory requirement one particular feature of the regulatory

test for NSPS modifications - the hourly rate test enunciated

at 40 C.F.R. 60.14(b) — while simultaneously allowing large

departures with respect to others, including minimum

significance thresholds and netting, or why, if the mandate is

regulatory sameness, the NSPS test should be preferred over

an NSR test that is far more consonant with the statutory text,

see New York, 413 F.3d at 38-40; 45 Fed. Reg. at 52700.

Contrary to Duke’s argument (Br. 43), JBP v. Alvarez,

126 S.Ct. 514 (2005), does not make the “presumption of

uniform usage irrebut[t]able” here. That decision construed

the term “principal activity or activities” as interpreted by

Steiner v. Mitchell, 350 U.S. 247 (1956), and as employed in

two contiguous subsections of the Portal-to-Portal Act,

covering worker travel “to and from * * * the principal

activity or activities,” and activities before and after “said

principal activity or activities,” 29 U.S.C. 254(a)(1), (2)

(emphasis added). Because the second reference to “said”

activities was “an explicit reference to the use of the identical

term” earlier in this statute, the Court found “no plausible

argument” for diverging interpretations. 126 S.Ct. at 524-25.

Here, rather than simultaneously-enacted statutory

subsections sharing a single “said” term, the Court encounters

separate statutory programs enacted years apart using a

general term with component phrases (“source,” “increase,”

concentration” on “plant capacity in existence.” H.R. Rep. No. 95-297

(1977) (§ 160(c)2EXi)). See Alabama Power, 636 F.2d at 380-81.

17

“amount,” “emitted”) that must themselves be defined to

mesh with the distinct regulatory programs. See Robinson v.

Shell Oil Co., 519 U.S. 337, 343 (1997); Cinergy, 458 F.3d at

710 (noting that NSPS and NSR definitions were enacted “by

different Congresses for different purposes”). "”

Finally, the language of the statutory definition -

relegated to a cameo role in Duke’s script — is inconsistent

with the regulatory approach Duke claims it to command.

See New York, 413 F.3d at 38-40.'* Congress did not speak

of increases in “rate” let alone “hourly” rate), “designed

emitting capacity” (Duke CA4 Br. 8), “fundamental ability to

emit pollutants” (Duke Br. 24), or “actual capacity” (Duke Br.

22, 32 n.13, 36). Congress, instead, targeted increases in the

“amount” of pollutants “emitted.” 42 U.S.C. 7479(2\C),

7411(a). See id. 7475(b), 7479(1) (denominating emissions

for the PSD program in “tons per year’’).

Duke dismisses arguments based on the text and structure

of the Act by treating the term “purposes” as an epithet (Br.

47-50). As Robinson illustrates, however, no rule of statutory

construction prevents a court or an agency from construing a

term in light of the role it (and the provision in which it

appears) plays in the overall statutory scheme. Moreover,

neither our argument nor EPA's longstanding interpretation

'’ Duke's assertion (Br. 43) that Congress explicitly defined PSD

modification as “used” in NSPS is incorrect. The term appears in the PSD

provision as follows: “construction when used in connection with any

source or facility includes modification,” 42 U.S.C. 7479(2\C). Far from

mandating identity, that usage invokes the fundamentally distinct

“sources” to which the programs apply. See also Petr. Br. 45-48

(summarizing substantive differences between NSPS and NSR programs).

See General Dynamics Land Sys., Inc. v. Cline, 540 U.S. 581, 595 (2004)

(meaning of same term may vary depending on the “‘connection in which

the words are used’”) (citation omitted).

'S The 1990 Amendments to the NNSR program, which also employs the

Section 111(a)(4) definition, 42 U.S.C. 7501(4), built upon EPA's NSR

regulatory definition, using annual pollutant thresholds, id. 751 1a(c), (d),

751 1a(e), and “actual emissions,” id. 7503(c)(1). Petr. Br. 15-16, 40.

18

of the statute and regulations privileges the CAA’s “remedial

purpose” over all others. On the contrary, it also serves the

different purpose of enabling economic growth: Because the

increases in actual emissions that Duke claims must be

ignored at the permitting stage indisputably do consume

available “increment,” allowing projects like Duke's to

proceed without emissions controls crowds out construction

of new, more efficient projects. See, e.g., Petr. Br. 40-41; Br.

of States of New York, et al. 26-28.

Seeking to rescue arguments unmoored to statutory or

regulatory text, Duke and its amici suggest that transplanting

the NSPS hourly rate test into PSD is necessary to avoid

undue burdens for industry, and request a cropping of NSR in

light of later legislative developments (but see, e.g., Br. of

Amicus Curiae STAPPA ef al., at 8-12). At best, such

contentions would be grist for EPA rulemaking procedures,

not an enforcement action, See 42 U.S.C. 7607(b), (d)(1),

(e).'* But they also disregard the numerous exemptions built

into the Act and rules. The statute exempts sources emitting

less than “major” pollution thresholds entirely, id. 7479(1),

and contains a partial exemption for modifications involving

less that 50 tons per year of pollution, id. 7574(b). EPA's

regulations, in turn, exempt individual projects not only

through source-wide netting and de minimis thresholds, but

also through exemptions for multiple alternative fuel and

repowering projects, id. 51.166(b)(2)(iii)(b)-(e), and stand-

alone increases in hours of operation or production rate, id.

51.166(b)(2)(i1i)(/). And, contrary to assertions that stretch

beyond hyperbole, the rules exempt routine maintenance,

repair, and replacement. /d. 51.166(b)(2)iii(a).-” See 45

'* Petitioners oppose EPA's October 20, 2005 rulemaking proposal, 70

Fed. Reg. 61081. But Duke's efforts to capitalize on that ininative (Br.

21-22) weaken its position in this case, since EPA’s notice recogmzes that

the proposal is a significant departure, requiring changes to regulatory text

in effect since 1980, see, e.g., 70 Fed. Reg. at 61088-89.

*® See, e.g., Br. of Alabama, ef al. 17 (“nearly every physical change made

19

Fed. Reg. at 52726 (“adequate exemptions have been

provided in today’s regulations and no further ones are

authorized under the Act”). But neither Act nor regulations

provide for the exemption Duke seeks here — “perpetual

immunity,” Alabama Power, 636 F.2d at 400, for old sources

undergoing renovations that increase the amount of pollution

actually emitted by thousands of tons per year.

As Judge Posner explained in Cinergy, the rule sought by

Duke here, “besides not conforming well to the language of

the regulation,” would encourage investment choices made

only to “elude the permit requirement,” “distort the choice

between rebuilding an old plant and replacing it with a new

one,” and open “a loophole that would allow pollution to soar

unregulated.” 458 F.3d at 709, 711. See Br. of Amici Curiae

STAPPA et al. 7-16; Br. of States of New York, et al. 22-29;

Br. of Former Administrators Browner and Train 21-24. It

would undermine an entire statutory structure predicated in

every particular (see Petr. Br. 39-42) on a realistic accounting

of increases in the “amount * * * emitted” by major sources

of pollution. Duke’s policy arguments are misguided, and

provide no basis to ignore the statutory allocation of judicial

authority in CAA implementation, 42 U.S.C. 7607(b), (d), (e).

CONCLUSION

The judgment of the court of appeals should be reversed.

Respectfully submitted.

to every component of every sub-part unit will tngger application of new

NSR/PSD requirements”) (emphases original). Duke's massive

renovation of “geriatric” units (JA 201), see Pet. Br. at 19-20 & nn.12-14,

serves as a real-world corrective. Despite the scope and cost of its

projects, below Duke invoked the regulatory exemption for “routine

maintenance,” an issue that would be taken up on remand. See Pet. App.

7a-8a n.2.

JEFFREY M. GLEASON

J. BLANDING HOLMAN, IV

CALEB JAFFE

Southern Environmental

Law Center

200 W. Franklin Suite 330

Chapel Hill, N.C. 27516

(434) 967-1450

VICKIE PATTON

Environmental Defense

2334 North Broadway

Boulder, Colorado 80304

(303) 447-7215

SEAN H. DONAHUE

Counsel of Record

2000 L St., NW

Suite 808

Washington, D.C. 20036

(202) 466-2234

DAVID T. GOLDBERG

99 Hudson St., 8th Floor

New York, N.Y. 10013

(212) 334-8813

Counsel for Petitioners

OCTOBER 2006