Appendix — Central Arizona Water Conservation District v. Environmental Protection Agency

Supreme Court, U.S

ns a)

No.

23 3983

JUN 23 1993 |

OFFICE OF pric i . j

IN THE SUPREME COURT OF THE UNITED STATES

October Term, 1993

CENTRAL ARIZONA WATER CONSERVATION DISTRICT,

CENTRAL ARIZONA IRRIGATION AND DRAINAGE DISTRICT,

MARICOPA-STANFIELD IRRIGATION & DRAINAGE DISTRICT,

and NEW MAGMA IRRIGATION & DRAINAGE DISTRICT, Arizona

Municipal corporations,

Petitioners,

v.

UNITED STATES ENVIRONMENTAL PROTECTION AGENCY,

Respondent.

PETITION FOR A WRIT OF CERTIORARI

TO THE UNITED STATES COURT OF APPEALS

FOR THE NINTH CIRCUIT

APPENDIX TO

CENTRAL ARIZONA WATER

CONSERVATION DISTRICT’S PETITION FOR

WRIT OF CERTIORARI

Marvin S. Cohen, Esq.*

Leslie A. McCarthy, Esq.

SACKS, TIERNEY & KASEN

2929 N. Central, 14th Floor

Phoenix, Arizona 85012-2742

Telephone: (602) 279-4900

Counsel of Record *

Douglas K. Miller, Esq.

CENTRAL ARIZONA WATER

‘CONSERVATION DISTRICT

23636 North 7th Street

Phoenix, Arizona 85024

Telephone: (602) 870-2333

Attorneys for Petitioner Central

Arizona Water Conservation

District

Robert S. Lynch, Esq.

2001 North Third Street, Suite 204

Phoenix, Arizona 85004-1472

Telephone: (602) 254-5908

Attorney for Petitioners Central

Arizona Irrigation and Drainage

District and Maricopa-Stanfield

Irrigation & Drainage District

William D. Baker, Esq.

ELLIS, BAKER, CLARK &

PORTER, P.C.

4444 North 32nd Street, Suite 200

Phoenix, Arizona 85018

Telephone: (602) 956-8878

Attorneys tor Petitioner New Magma

Irrigation & Drainage District

INDEX TO APPENDIX TO

CENTRAL ARIZONA WATER

CONSERVATION DISTRICT’S

PETITION FOR A WRIT OF CERTIORARI TO THE

UNITED STATES COURT OF APPEALS FOR

THE NINTH CIRCUIT

Page

Opinion

Regulations

56 Fed. Reg. 50172 (October 3, 1991) .....-.-.+----

53 Fed. Reg. 35956 (September 15, 1988 .........--

52 Fed. Reg. 7802 (March 12, 1987) .......----+---

45 Fed. Reg. 80084 (December 2, 1980) ......-.----

45 Fed. Reg. 34762 (May 22, 1980) ......-------

40 C.F.R. § 51.300 ef seq. .. 2... ee eee eee eee

tk © ag pS. ee a ere ee eee

40 CPR. BSZAG OG. 2 nn ccc ccc ewncass

Statutes

Te bk wo es ee ee eee eee kee

62 2 ol S| ae aera ia are eee ce

Te Fe ik ON a ERE RS

Miscellaneous

EPA Docket A-79-40, II-A-5

"Protecting Visibility:

An EPA Report to

oe ee

EPA Docket A-79-40, V-A-6 "User's

Manual for the Plume Visibility

Model (PLUVUE), November, 1980 ......------

FOR PUBLICATION

UNITED STATES COURT OF APPEALS

FOR THE NINTH CIRCUIT

Te

CENTRAL ARIZONA WATER

CONSERVATION District, CENTRAL

ARIZONA IRRIGATION AND DRAINAGE

DistTRICT, MARICOPA-STANFIELD

IRRIGATION & DRAINAGE DistrICT,

New MAGMA IRRIGATION &

DRAINAGE DISTRICT, AND

HARQUAHALA VALLEY IRRIGATION

DISTRICT, No. 91-70731

Petitioners,

" OPINION

Untrep STATES ENVIRONMENTAL

PROTECTION AGENCY,

Respondent,

SALT RIVER PROJECT AND POWER

District, GRAND CANYON TRUST,

AND THE WILDERNESS SOCIETY,

Respondents-Intervenors.

f

Appeal from the United States

Environmental Protection Agency

Argued and Submitted

January 13, 1993—San Francisco, California

Filed March 25, 1993

Before: Ruggero J. Aldisert,* Alfred T. Goodwin, and

Betty B. Fletcher, Circuit Judges.

*Honorable Ruggero J. Aldisert, Senior United States Circuit Judge for

the Third Circuit, sitting by designation.

2712 CAWCD v. EPA

Opinion by Judge Goodwin

SUMMARY

Environmental Law/Air

The court of appeals denied a petition for review of a Final

Rule issued by the Environmental Protection Agency. The

court held that, pursuant to Clean Air Act requirements to

reduce visibility impairments around national parks, imposi-

tion by the EPA of emissions reduction requirements on a

power facility in proximity to the Grand Canyon were not

arbitrary and capricious or in excess of regulatory authority,

when the regulations were based on reasonable interpretations

of equivocal evidence.

Petitioner Central Arizona Water Conservation District

obtains electricity to pump its water from the Navajo Generat-

ing Station, which is within twelve miles of the Grand Can-

yon. Respondent United States Environmental Protection

Agency, pursuant to the goal of the Clean Air Act to remedy

visibility impairment in specified areas, including the national

parks, required a 90 percent reduction in certain emissions

from Navajo. Navajo is owned in part by the Bureau of Recla-

mation, with which the District has a contractual relationship

requiring it to repay most of the Bureau’s share of the costs

of the required emission controls.

The Act requires EPA to promulgate regulations to assure

reasonable progress toward the goal. Following an extended

process, which included a tracing study conducted by the

National Park Service, which enabled emissions from Navajo

to be identified in the ambient air; an evaluation of that study

by the National Academy of Sciences, which both acknowl-

edged the difficulty of determining the exact contribution to

Grand Canyon haze reasonably attributable to Navajo, and

CAWCD v. EPA 2713

noted that the EPA was not statutorily required to show such

a precise relationship; and a monitoring study sponsored by

Navajo’s operator, which concluded that emission controls at

Navajo would produce at most a 2 percent improvement in

average visibility, EPA proposed a 70 percent average reduc-

tion in the Navajo emissions. After a public comment period

and consultations with Navajo, environmental groups, includ-

ing the Environmental Defense Fund and the Grand Canyon

Trust, and the State of Arizona, EPA issued a final determina-

tion concluding that Navajo was a dominant contributor to

visibility impairment, and that emissions reduction by 90 per-

cent would more adequately represent reasonable progress

toward the visibility goal than the approach originally sug-

gested by EPA.

The District contended that EPA's Final Rule exceeded the

scope of its regulatory authority, and that its action was arbi-

trary and capricious. In turn, EPA contended that the District

lacked standing to challenge its ruling.

[1] The District had standing to litigate this issue. The Final

Rule would likely cause it some amount of economic harm,

given its contractual relationship with the Bureau of Reclama-

tion. [2] The Final Rule was the direct cause of that harm, not-

withstanding the involvement of a third party and a

contractual relationship, and that harm was likely to be

redressed by a favorable decision. [3] The District's interest

was within the zone protected by the Act, since having to pay

the costs of compliance could not be said to be marginally

related to or inconsistent with the purposes of the Act.

[4] EPA had statutory authority to regulate the contribution

to visibility impairment reasonably attributable to Navajo,

notwithstanding that it might only be a fraction of the total

problem. [5] EPA had broad discretion in determining if

impairment was reasonably attributable to a specific source.

[6] The record reflected adequate technical and scientific sup-

port for EPA‘s conclusion that impairment was reasonably

2714 CAWCD v. EPA

attributable to Navajo. Its choice represented reasonable

accommodation of conflicting policies committed to its care

by the Act, and was therefore entitled to deference from a

reviewing Court.

[7] EPA relied on Congressional direction to assure reason-

able progress toward the national goal as its statutory author-

ity for promulgating its regulations. [8] Reliance on the

reasonable progress provisions was a permissible construction

of the statute, since reasonable progress was the overarching

requirement to be addressed. [9] The Final Rule was the prod-

uct of reasoned decisionmaking. EPA adequately considered

relevant factors in its process. As long as its interpretation

was reasonable, a reviewing Court may not substitute its own

judgment for the agency’s. Deference to EPA’s judgment was

appropriate because the agency was making predictions

within its special expertise at the frontiers of science. [10] The

limits of the Park Service's study were acknowledged, and

EPA’s decision was adequately supported by substantial sci-

entific evidence. Its decision was based on reasonable extrap-

olations from some reliable evidence.

COUNSEL

Marvin S. Cohen, Sacks, Tiemey & Kasen, Phoenix, Arizona,

for the petitioners.

Scott A. Schachter, United States Department of Justice,

Washington, D.C., for the respondent.

Patrick M. Raher, Hogan & Hartson, Washington, D.C., for

the respondents-intervenors.

Beryl I. Dulsky, Assistant Attorney General, Phoenix, An-

zona, for the amicus.

mM

CAWCD v. EPA 2715

OPINION

GOODWIN, Circuit Judge:

Petitioners Central Arizona Water Conservation District

(“CAWCD”) and four irrigation districts’ (collectively

“Petitioners” or the “Districts’) challenge an Environmental

Protection Agency (“EPA”) Final Rule which requires a 90%

reduction in sulfur dioxide (SO,) emissions at the Navajo

Generating Station (“NGS”) in order to improve winter aver-

age visibility in the Grand Canyon National Park (“Grand

Canyon”).

Petitioners argue (1) that, because the Final Rule seeks to

regulate “regional haze” when EPA has yet to promulgate

Phase II implementing regulations addressing regional haze,

EPA exceeded the scope of its regulatory authority by issuing

the Final Rule, and (2) that the Final Rule constitutes arbitrary

and capricious agency action. In addition to responding to

Petitioners’ substantive criticisms, EPA asserts that Petition-

ers lack standing to bring this challenge.

The Final Rule was issued by EPA under 42 U.S.C.

§§ 7410(c)(1), 7491, and is directly appealable to this court

under 42 U.S.C. § 7607(b)(1). Petitioners do have standing to

bring this challenge. We hold, however, that EPA acted

within its statutory and regulatory authority when it promul-

gated the Final Rule, and that EPA has not acted arbitrarily

and capriciously.

I. BACKGROUND

This case involves regulations promulgated by EPA in an

attempt to remedy, at least partially, visibility impairment at

‘Central Arizona Irrigation and Drainage District, Maricopa-Stanfield

Irrigation & Drainage District, New Magma Irrigation & Drainage Dis-

trict, and Harquahala Valley Irrigation District.

2716 CAWCD v. EPA

the Grand Canyon. In a final rule entitled “Approval and Pro-

mulgation of Implementation Plans: Revision of the Visibility

FIP for Arizona,” 56 Fed. Reg. 50,172 (1991) (codified at 40

C.F.R. § 52) (“Final Rule”), EPA required a 90% reduction in

SO, emissions at NGS, a power plant situated approximately

twelve miles from the Grand Canyon, near Page, Arizona.

The Final Rule limits SO, emissions from NGS to 0.10 pound

per million British thermal units (lb/MMBtu), with an esti-

mated 7% winter average visibility improvement in the Grand

Canyon. The estimated cost of the improvement, following an

initial capital cost estimated at $430 million, is $89.6 million

per year.

The Salt River Project (“SRP”), a respondent-intervenor in

this case,” is the operating agent of NGS. NGS is jointly

owned by SRP, the U.S. Department of Interior, Bureau of

Reclamation (“BOR”), Los Angeles Department of Water and

Power, Arizona Public Service Co., Nevada Power Co. and

Tucson Gas & Electric Co.* Petitioners are water districts that

obtain electricity to pump their water primarily from NGS.

They acknowledge that they are not owners of NGS, but

nonetheless claim an economic interest in the Final Rule.

CAWCD claims that it will be required, due to its contractual

relationship with the BOR, to repay the major portion of the

BOR’s 24.3% share of the costs of installing and maintaining

the emission controls required by the Final Rule.

7It is noteworthy that SRP and the other joint-owners are the parties

directly aggrieved by the Final Rule, yet they do not participate in Peti-

tioners’ challenge. In fact, SRP argues in support of the Final Rule as a

respondent-intervenor.

*The joint-owners’ percentages of participation is as follows: BOR,

24.3%; SRP, 21.7%; Los Angeles Department of Water and Power,

21.2%; Arizona Public Service Co., 14%; Nevada Power Co., 11.3%; Tuc-

son Gas & Electric Co., 7.5%.

CAWC(D ~ ERA QT N7

A. Regulatory Framework

1. The Clean Air Act, Visibility Impairment, and the

Grand Canyon

In 1977, Congress substantially amended the Clean Air Act

(the “Act™). Included in the 1977 amendments was section

169A, 42 U.S.C. § 7491, which declared “as a national goal

the prevention of any future, and the remedying of any exist-

ing, impairment of visibility in mandatory class I Federal

areas which impairment results from manmade air pollution.”

42 U.S.C. § 7491(a)(1). Congress required EPA to promulgate

regulations to assure “reasonable progress toward meeting

th[is] national goal.” 42 U.S.C. § 7491(a)(4). EPA was further

directed to require each state with a class I Federal area to

revise its state implementation plan (“SIP”) “to contain such

emission limits, schedules of compliance and other measures

as may be necessary to make reasonable progress toward

meeting the national goal.” 42 U.S.C. § 7491(b)(2). Measures

for achieving “reasonable progress” generally include best

available retrofit technology (“BART™)* and a long-term strat-

egy. 42 U.S.C. §§ 7491(b)(2)(A), (B). If an individual state

fails to fulfill its obligations under the Act, EPA is directed to

take such measures as are required to achieve “reasonable

“The regulations provide the following definition of BART:

Best Available Retrofit Technology (BART) means an emission

limitation based on the degree of reduction achievable through

the application of the best system of continuous emission reduc-

tion for each pollutant which is emitted by an existing stationary

facility. The emission limitation must be established, on a case-

by-case basis, taking into consideration the technology available,

the costs of compliance, the energy and nonair quality environ-

mental impacts of compliance, any pollution control equipment

in use or in existence at the source, the remaining useful life of

the source, and the degree of improvement in visibility which

may reasonably be anticipated to result from the use of such tech-

nology.

40 C.F.R. § 51.301 (c).

2718 CAWCD v. EPA

progress” pursuant to a federal implementation plan (“FIP”)

under section 110(c) of the Act. 42 U.S.C. § 7410(c)(1).

The Act defines class I Federal areas as international parks,

national wilderness areas or memorial parks which exceed

5,000 acres in size, and national parks which exceed 6,000

acres in size. 42 U.S.C. § 7472(a). The Grand Canyon has

been classified as a class I Federal area. See 44 Fed. Reg.

69,122 (1979). Congress recorded its concern with the visibil-

ity impairment at the Grand Canyon caused by NGS. See H.R.

Rep. No. 294, 95th Cong., Ist Sess. 203-04 (1977).

2. EPA's 1980 Regulations

In 1980, EPA promulgated visibility regulations under sec-

tion 169A of the Act. 45 Fed. Reg. 80,084 (1980) (codified

at 40 C.F.R. §§ 51.300-.307). The regulations adopted a

“phased approach to visibility protection.” /d. at 80,085.

Phase I was directed at controlling visibility impairment “that

can be traced to a single existing stationary facility or small

group of existing stationary facilities.” Jd. EPA refers to this

type of impairment as “reasonably attributable” impairment.

45 Fed. Reg. 34,762, 34,779 (1980) (codified at 40 C.F.R.

§ 51). EPA deferred addressing other types of impairment

such as “regional haze” for future phases due to the height-

ened complexity and the scientific and technical limitations

inherent in attempts to identify, measure, and control such

broadscale visibility impairment. See 45 Fed. Reg. at 80,086;

see also id. at 80,085 (defining regional haze as “widespread,

regionally homogeneous haze from a multitude of sources

which impairs visibility in every direction over a large area”).

Generally, EPA’s “Phase I” regulations require affected

states to coordinate the development of SIPs with the appro-

priate Federal land managers, to develop programs to assess

and remedy visibility impairment from new and existing

sources, and to develop a long-term strategy to assure reason-

able progress toward section 169A’s national visibility goal.

CAWCD v. EPA 2719

See 40 C.F.R. §§ 51.300-.307. The regulations specifically

require states to identify those existing sources “which may

reasonably be anticipated to cause or contribute” to any visi-

bility impairment which is “reasonably attributable to that

existing stationary facility.” 40 C.F.R. § 51.302(c)(4)(i). Once

the source is identified, the affected state is required to take

such measures as are required to attain “reasonable progress”;

such measures generally include determination of emissions

limitations for that source under BART and the development

of a long-term strategy. 40 C.F.R. §§ 51.302(c)(1), (2).

The regulations define the term “visibility impairment” as

“any humanly perceptible change iin visibility (visual range,

contrast, coloration) from that which would have existed

under natural conditions.” 40 C.F.R. § 51.301(x). The term

“reasonably attributable” is defined as “attributable by visual

observation or any other technique the State deems

appropriate.” 40 C.F.R. § 51.301(s). The states, or EPA under

§ 7910(c), thus have broad discretion in determining how and

whether impairment may be attributed to an individual source.

See 45 Fed. Reg. at 80,094, 80,085.

B. Prior Proceedings and the Rulemaking History

In its implementation of Phase I, EPA required all states

containing class I Federal areas to submit revised visibility

SIPs within a nine-month period. Arizona was one of thirty-

five states failing to submit a revised SIP to EPA. In 1982, the

Environmental Defense Fund and other environmental groups

brought a citizen suit against EPA to compel performance of

the agency's nondiscretionary duty under 42 U.S.C. § 7410(c)

(1)(A) to promulgate visibility FIPs when states fail to submit

SIPs pursuant to the 1980 regulations. See EDF v. Reilly, No.

C82-6850-RPA (N.D. Cal. Apr. 20, 1984). The parties

reached a settlement agreement which the court approved in

an April 20, 1984 consent decree. This consent decree

required EPA to review existing SIPs for deficiencies and

eT

2720 CAWCD v. EPA

allow states to cure those deficiencies. If states remained defi-

cient, the consent decree required EPA to issue visibility FIPs.

The Department of Interior subsequently certified the exis-

tence of visibility impairment in all class I Federal areas, and

specifically declared NGS as a probable source of impairment

at the Grand Canyon. Following this certification, the

National Park Service (“Park Service”) conducted the Winter

Haze Intensive Tracer Experiment (““WHITEX”’), a winter vis-

ibility attribution study. In part, WHITEX involved the

release from NGS of a unique “tracer” gas, CD; because CD

is not found in the ambient air, its use “fingerprinted” NG

emissions when detected downwind.

In November 1987, EPA disapproved the SIPs of twenty-

nine states, including Arizona, for failing to comply with the

visibility regulations. See 52 Fed. Reg. 45,132, 45,133 (1987)

(codified at 40 C.F.R. §§ 52, 81). Over the next few years,

EPA further investigated visibility impairment at Grand Can-

yon and other class I Federal areas. While acting on many of

the areas, EPA delayed action on the Grand Canyon to allow

the Park Service time to analyze the data obtained from the

WHITEX study. The Park Service issued an April 1989 draft

report which attributed to NGS 70% of the sulfates in the

Grand Canyon during the WHITEX experiment period. See

William Malm et al., National Park Service Report on the

Winter Haze Intensive Tracer Experiment, Draft Final Report

(Apr. 7, 1989).

Relying on the Park Service’s April 1989 draft report, EPA

preliminarily attributed to NGS several episodes of winter-

time visibility impairment at the Grand Canyon. See 54 Fed.

Reg. 36,948, 36,951 (1989) (codified at 40 C.F.R. § 52). EPA

solicited public comment on the merits of its preliminary attri-

bution finding, and began the informal rulemaking process to

determine the appropriate action to be taken. Jd. In December

1989, the Park Service issued its final report on WHITEX.

See William Malm et al., National Park Service Report on the

™ |

CAWCD v. EPA 2721

Winter Haze Intensive Tracer Experiment, Final Report (Dec.

4, 1989).

SRP and others, including Petitioners, submitted comments

severely criticizing the analyses, methodologies, and conclu-

sions of the Park Service's reports. In response to the con-

cerns raised about the Park Service's analysis of the WHITEX

data, the National Research Council of the National Academy

of Sciences (“NAS”) was asked to evaluate the Park Service's

WHITEX report and the other scientific evidence relevant to

EPA’s preliminary attribution. In its October 1990 report, the

NAS confirmed that “at some times during the study period,

NGS contributed significantly to haze™ in the Grand Canyon.

National Research Council, Haze in the Grand Canyon: An

Evaluation of the Winter Haze Intensive Tracer Experiment 3,

37 (1990) [hereinafter Haze in the Grand Canyon}. NAS cau-

tioned, however, that “aspects of the WHITEX data analysis

preclude a quantitative determination of the exact fraction of

the Grand Canyon haze problem that is attributable to NGS.

These aspects are primarily related to problems with imple-

mentation and interpretation of multiple linear regression

models.” Jd. at 37. Still, NAS acknowledged that Congress

did not “require EPA to show a precise relationship between

a source’s emissions and all or a specific fraction of the visi-

bility impairment within a Class I area.” Jd. at 5, 37.

SRP sponsored its own monitoring study in the winter of

1990 — the Navajo Generating Station Visibility Study

(“NGSVS") — to investigate the sources of visibility impair-

ment in the Grand Canyon and to determine what improve-

ment in visibility would result from a reduction of SO,

emissions at NGS. The NGSVS data indicated that controlling

SO, emissions from NGS would at most result in a 2%

improvement in the winter seasonal average standard visual

range. While detailing some episodes of visibility impairment

attributable to NGS, the NGSVS report generally found that

“the NGS was only part of a regional problem and that it was

not the dominant contributor of sulfates.” Sonoma Technol-

2722 CAWCD v. EPA

ogy Inc., Navajo Generating Station Visibility Study 2-11 (L.

Willard Richards et al. eds., 1991).

In February 1991, EPA proposed to revise the FIP for Ari-

zona to include emission limits under BART to address win-

tertime visibility impairment at the Grand Canyon reasonably

attributable to NGS. See 56 Fed. Reg. 5,173 (1991) (codified

at 40 C.F.R. §52). EPA solicited public comment on four

possible regulatory options, and proposed one of those

options — to require as BART for NGS a 70% continuous SO_

emission limitation, determined on a thirty-day rolling aver-

age, to be phased-in between 1995 and 1999. Jd. at 5,183-84.

After a public comment period of over nine weeks, senior

EPA policy officials met with the NGS owners, representa-

tives of the State of Arizona, and certain environmental

groups, including the Grand Canyon Trust and the EDF, to

explore further additional control options which might pro-

duce a pareto optimal solution, more environmentally protec-

tive and economically efficient.* These meetings resulted in a

“memorandum of understanding” which recommended that

EPA adopt a regulatory approach designed to achieve a

greater degree of visibility improvement in the Grand Canyon

at lower cost than the proposal published by EPA in February

1991. EPA published the proposal outlined in the memoran-

dum of understanding and reopened the public comment

period. See 56 Fed. Reg. 38,399, 38,401 (1991) (codified at

40 C.F.R. § 52). EPA received twenty-one comments during

this supplemental period, eighteen of which supported the

new proposal. See 56 Fed. Reg. at 50,177.

In the Final Rule, dated October 3, 1991, EPA issued its

final determination that certain visibility impairment episodes

at the Grand Canyon were “traceable to NGS and that NGS

is a dominant contributor to certain visibility impairment

episodes,” and promulgated revisions to the Arizona visibility

‘Summaries of these meetings were placed in the administrative record.

See 56 Fed. Reg. at 50,177.

CAWCD v. EPA 2723

FIP to address the impairment. /d. The revisions adopted a

regulatory approach consistent with the memorandum of

understanding’s proposal, reducing SO, emissions 90% to a

level of 0.10 Ib/MMBtu. Jd. EPA determined that this

approach would more adequately achieve “reasonable

progress” toward the national visibility goal under section

169A(b)(2) of the Act, 42 U.S.C. § 7491(b)(2), than would

the alternative provided by BART analysis. See 56 Fed. Reg.

at 50,177. As required by section 307(d) of the Act, EPA’s

action was “accompanied by a response to each of the signifi-

cant comments, criticisms, and new data submitted in written

or oral presentations during the comment period.” 42 U.S.C.

§ 7607(d)(6)(B). EPA issued its responses in the regulation

itself and in a supplementary ninety-one page document enti-

tled “Response to Public Comments: Proposed Revisions to

Arizona Visibility FIP for Navajo Generating Station” [here-

inafter Response to Public Comments}.

Il. STANDING

First, as a threshold jurisdictional matter, EPA challenges

whether the Districts have standing to bring this action. We

find that the Districts do have standing.

A. Constitutional Standing

The Supreme Court recently outlined the three elements of

the “irreducible constitutional minimum of standing” in Lujan

v. Defenders of Wildlife, 112 S. Ct. 2130 (1992):

First, the plaintiff must have suffered an “injury in

fact"—an invasion of a legally-protected interest

which is (a) concrete and particularized, and (b)

“actual or imminent, not ‘conjectural’ or

‘hypothetical. “ Second, there must be a causal con-

nection between the injury and the conduct com-

plained of—the injury has to be “fairly

trace[able] to the challenged action of the defendant,

ee

2724 CAWCD v. EPA

and not... th{e] result [of] the independent action of

some third party not before the court.” Third, it must

be “likely,” as opposed to merely “speculative,” that

the injury will be “redressed by a favorable

decision.”

Id. at 2136 (footnote, citations omitted; ellipses in original).

Justice Scalia elaborated that “when the plaintiff is not him-

self the object of the government action or inaction he chal-

lenges, standing is not precluded, but it is ordinarily

‘substanually more difficult’ to establish.” Jd. at 2137 (quot-

ing Allen v. Wright, 468 U.S. 737, 758 (1984)).

1. Injury in Fact

The Districts’ claimed injury is an economic one: CAWCD

Claims it is contractually required to repay much of BOR’s

24.3% share of the costs of installing and maintaining emis-

sion controls at NGS as required by the Final Rule. Pecuniary

injury is Clearly “a sufficient basis for standing.” See Fair v.

EPA, 795 F.2d 851, 853-54 (9th Cir. 1986) (citing cases). But

while pecuniary or economic injury is generally a legally pro-

tected interest, the “injury in fact” test “ ‘requires that the

party seeking review be himself among the injured.’ ” Lujan,

112 S. Ct. at 2137 (quoting Sierra Club v. Morton, 405 U.S.

727, 734-35 (1972)). That party’s injury must be (a) concrete

and particularized, and (b) actual or imminent, and not con-

jectural or hypothetical. /d. at 2136.

EPA argues that the owners of NGS are the ones directly

aggrieved by the Final Rule, and that the indirect economic

harm that the Districts may suffer as a result of their financial

obligation to one of NGS’s owners is too attenuated an inter-

est to provide them with standing to bring this challenge. EPA

claims the Districts’ injury is purely speculative because: (1)

CAWCD concedes it will attempt to pass on increased costs

caused by the Final Rule to its water users and taxpayers

(including the four petitioning irrigation districts); and (2)

CAWCD v. EPA 2725

CAWCD has failed to demonstrate that BOR will in fact have

to pay a portion of any existing compliance costs.° or that

CAWCD will be required under contract to reimburse BOR

for a portion of those costs assuming they exist. Thus, EPA

argues, the Districts’ economic injury is too indirect, specula-

tive, and hypothetical to provide them with standing.

[1] We conclude that CAWCD’s economic injury is suffi-

ciently concrete and imminent to accord it standing to litigate

this action. The Districts’ claimed injury is by no means “a

general or amorphous harm,” National Wildlife Fed'n v. Bur-

ford, 871 F.2d 849, 852 (9th Cir. 1989), nor is it a mere gener-

alized grievance. See Lujan, 112 S. Ct. at 2143. While the

extent of CAWCD's economic harm is not readily determin-

able, the record reveals that the Final Rule will likely cause

Petitioners some amount of pecuniary harm given their obli-

gation to repay BOR’s share of the costs imposed by the Final

Rule. Neither the sale of marketable allowances nor the possi-

bility of passing on increased costs to consumers undermine

the Districts’ showing of the required “actual or threatened

injury.” Valley Forge Christian College v. Americans United

for Separation of Church & State, Inc., 454 U.S. 464, 472

(1982); Competitive Enter. Inst. v. NHTSA, 901 F.2d 107, 112

(D.C. Cir. 1990).

2. Causation and Redressability

[2] In addition to the injury in fact requirement. the Dis-

tricts must prove causation and redressability, i.e., that their

alleged economic injury is fairly traceable to EPA’s chal-

lenged action, and that the relief requested is likely to redress

that injury. Lujan, 112 S. Ct. at 2136. EPA argues that since

*Under section 403(b) of the Act, a regulated party may transfer SO

allowances under an allowance trading program. See 42 USC.

§ 7651b(b). EPA notes that sale of allowances may offset the required

capital investment for control equipment. Thus, EPA argues, capital costs

may never be passed on to the District.

2726 CAWCD v. EPA

the Districts’ alleged economic injury flows from obligations

under the BOR/CAWCD agreement, the injury is not caused

by or fairly traceable to the challenged agency action. This

argument misses the point. While CAWCD’s contractual obli-

gations may provide the basis for its economic liability for the

increased costs imposed by the Final Rule, that hardly means

that the Final Rule itself is not the direct cause of that liabil-

ity. Further, the involvement of an intermediate third-party

here does not undermine the Districts’ causation argument

since “the government’s action [is] substantially likely to

cause the petitioners’ injury despite the presence of intermedi-

ary parties.” Competitive Enter., 901 F.2d at 114 (consumer

organization has standing to challenge NHTSA fuel economy

standards on grounds that organization members would be

hindered in their ability to purchase larger vehicles). Finally,

the Districts’ economic injury is likely to be redressed by a

favorable decision since elimination of the Final Rule would

necessarily eliminate the increased financial burden the rule

causes.

B. Prudential Standing

[3] In addition to the constitutional standing requirements,

Petitioners must also prove that their asserted interest is

“within the zone of interests protected by” the Clean Air Act.

See, e.g., Fair v. EPA, 795 F.2d at 854.

EPA contends that the Districts’ economic injury is not

within the zone of interests of the Act’s visibility provisions,

which are designed to “preserve, protect, and enhance the air

quality in national parks.” 42 U.S.C. § 7470(2). EPA’s argu-

ment ignores the fact that “the zone of interest test is ‘not

meant to be particularly demanding.’” National Wildlife

Fed'n v. Burford, 871 F.2d at 852 (quoting Clarke v. Securi-

ties Indus. Ass'n, 479 U.S. 388, 399 (1987)). As the Supreme

Court clarified in Clarke:

In cases where the plaintiff is not itself the subject of

the contested regulatory action, the test denies a night

a 6

CAWCD v. EPA 2727

of review if the plaintiff’s interests are so marginally

related to or inconsistent with the purposes implicit

in the statute that it cannot reasonably be assumed

that Congress intended to permit the suit.

479 U.S. at 399. Under this permissive standard, the Districts’

economic injury sufficiently falls within the “zone of

interests” protected by the visibility provisions of the Act.

Section 169A of the Act requires the Administrator to con-

sider “the costs of compliance” in setting standards to achieve

reasonable progress towards the national visibility goal. 42

U.S.C. § 7491(g)(1). As entities required to pay those costs of

compliance, the Districts’ interests cannot reasonably be

described as “marginally related to or inconsistent with” the

purposes of the Act.

We therefore conclude that Petitioners have standing to

bring this challenge.

Il. STANDARD OF REVIEW

The appropriate judicial review of the regulations at issue

here is provided for in section 307(d) of the Act, 42 U.S.C.

§ 7607(d). In relevant part, section 307(d)(9) provides that a

reviewing Court may reverse any action of the Administrator

found to be

(A) arbitrary, capricious, an abuse of discretion, or

otherwise not in accordance with law;

(C) in excess of statutory jurisdiction, authority,

or limitations, or short of statutory right; or

(D) without observance of procedure required by

law, if (i) such failure to observe such procedure is

arbitrary or capricious, (ii) the requirement of para-

2728 CAWCD v. EPA

graph (7)(B) has been met, and (iii) the condition of

the last sentence of paragraph (8) is met.

42 U.S.C. §§ 7607(d)(9)(A), (C)-(D).

Paragraph (7)(B) generally limits judicial review to proce-

dural objections “raised with reasonable specificity during the

period for public comment.” /d. at § 7607(d)(7)(B). But para-

graph (7)(B) does provide for mandatory reconsideration of

the rule by the agency if an objecting party can “demonstrate

to the Administrator” that (1) it was “impracticable” to prop-

erly raise a given procedural objection, and (2) the “objection

is of central relevance to the outcome of the rule.” /d.’ The

last sentence of paragraph (8) further provides:

In reviewing alleged procedural errors, the court may

invalidate the rule only if the errors were so serious

and related to matters of such central relevance to

the rule that there is a substantial likelihood that the

tule would have been significantly changed if such

errors had not been made.

42 U.S.C. § 7607(d)(8).

Review of EPA’s actions under section 307(d) of the Act

is guided by the appropriate deference given to an agency s

’The relevant portion of paragraph (7)(B) more fully provides:

If the person raising an objection can demonstrate to the Admin-

istrator that it was impracticable to raise such objection within

such time or if the grounds for such objection arose after the

period for public comment (but within the time specified for judi-

cial review) and if such objection is of central relevance to the

outcome oi the rule, the Administrator shall convene a proceed-

ing for reconsideration of the rule .... If the Administrator

refuses to convene such a proceeding, such person may seek

review of such refusal in the United States court of appeals for

the appropriate circuit ....

42 U.S.C. § 7607(d)( 7B).

CAWCD v. EPA 2729

interpretation of the statute Congress has entrusted it to

administer. Chevron U.S.A. Inc. v. NRDC, Inc., 467 U.S. 837,

843 (1984) (where the Clean Air Act is either “silent or

ambiguous with respect to the specific issue, the question for

the court is whether the agency’s answer is based on a permis-

sible construction of the statute”). “Deference also guides our

review of the Administrator’s interpretation of EPA regula-

tions if the interpretation is not unreasonable.” Citizens for

Clean Air v. EPA, 959 F.2d 839, 844 (9th Cir. 1992); see also

Hawaiian Elec. Co. v. EPA, 723 F.2d 1440, 1447 (9th Cir.

1984) (~[W]hen EPA is interpreting its own regulations, it is

entitled to even more deference.”). Moreover, the Supreme

Court has advised that “a reviewing court must generally be

at its most deferential” when the agency is “making predic-

tions, within its area of special expertise, at the frontiers of

science.” Baltimore Gas & Elec. Co. v. NRDC, Inc., 462 U.S.

87, 103 (1983). In such situations, this court is to “defer to the

agency's interpretation of equivocal evidence, so long as it is

reasonable.” NRDC, Inc. v. EPA, 902 F.2d 962, 968 (D.C. Cir.

1990), vacated, in part, dismissed, 921 F.2d 326 (D.C. Cir.),

certs. dismissed, 111 S. Ct. 806, cert. denied, 111 S. Ct. 952

(1991).

While this court should not defer to the agency where it

“simply has not exercised its expertise,” Public Citizen Health

Research Group v. Tyson, 796 F.2d 1479, 1505 (D.C. Cir.

1986),° courts are “extremely deferential to administrative—

agencies in cases involving technical rulemaking decisions.”

New York v. Reilly, 969 F.2d 1147, 1152 (D.C. Cir. 1992): see

also id. at 1150-51 (“We are particularly deferential when

*For instance, in Motor Vehicle Manufacturers Association of United

States, Inc. v. State Farm Mutual Automobile Insurance Co., 463 U.S. 29

(1983), the Supreme Court found an agency nuling arbitrary and Capricicus

where the agency had “entirely failed to consider an important aspect of

the problem.” /d. at 43. Siate Farm shows that the “expertise model” does

nol necessarily mandate judicial deference, since the Court there took a

“hard look” to ensure that the agency was in fact using its expertise in

arriving at policy decisions.

2730 CAWCD v. EPA

reviewing agency actions involving policy decisions based on

uncertain technical information.”); Tyson, 796 F.2d at 1505

(“[A]s long as Congress delegates power to an agency to reg-

ulate on the borders of the unknown, courts cannot interfere

with reasonable interpretations of equivocal evidence.”).

IV. DISCUSSION

A. EPA’s Statutory Authority to Promulgate the Finai Rule

Petitioners argue that in promulgating the Final Rule, EPA

acted “in excess of statutory jurisdiction, authority, or limita-

tions, or short of statutory right” under 42 U.S.C. § 7607(d)(9)

(c). Petitioners claim that the Final Rule regulates “regional

haze’ when EPA’s own regulations expressly defer regulation

of that type of visibility impairment until future phases. See

45 Fed. Reg. at 80,086. We conclude that the Final Rule was

legitimately promulgated under Phase I regulations directed at

“reasonably attributable” visibility impairment.

Petitioners rely on Vermont v. Thomas, 850 F.2d 99 (2d

Cir. 1988), where EPA reaffirmed that it was without author-

ity to regulate “regional haze and that measures addressing

that type of impairment are outside the scope of EPA's juris-

diction. This reliance is misplaced. In Thomas, the state of

Vermont submitted to EPA an SIP which “proposed a feder-

ally enforceable ‘long-term strategy’ to combat the effects of

regional haze” at the Lye Brook National Wilderness Area.

Id. at 101. The court held that without EPA rulemaking

addressing regional haze, the state’s regional haze measures

were “outside the scope” of EPA’s statutory and regulatory

authority. Jd. at 103-04. Here, EPA properly promulgated its

Final Rule under its Phase I regulations directed at

“reasonably attributable” impairment.’ Thomas is thus distin-

*in the Final Rule itself, EPA made eminently clear that its action

addressed Phase | visibility impairment rather than Phase II “regional

haze.” See 56 Fed. Reg at 50,172. Pursuant to Congress's 1990 amend-

CAWCD v. EPA 2731

guishable since it involved a direct and explicit attempt to reg-

ulate “regional haze.”

[4] EPA has acknowledged that “NGS is not the only

source of visibility impairment” at the Grand Canyon, 56 Fed.

Reg. at 50,177, and that regional haze also adversely affects

visibility there. Response to Public Comments at 23. Nonethe-

less, these mere facts hardly mean that EPA is without statu-

tory authority to remedy the impairment attributable to NGS.

Even if the Final Rule addresses only a small fraction of the

visibility impairment at the Grand Canyon, EPA still has the

Statutory authority to address that portion of the visibility

impairment problem which is, in fact, “reasonably

attributable” to NGS. Congress mandated an extremely low

triggering threshold, requiring the installment of stringent

emission controls when an individual source “emits any air

pollutant which may reasonably be anticipated to cause or

contribute to any impairment of visibility” in a class I Federal

area. 42 U.S.C. § 7491(b)(2)(A). The National Academy of

Sciences correctly noted that Congress has not required iron- .

clad scientific certainty establishing the precise relationship —

between a source’s emission and resulting visibility impair-

ment:

The phrase “may reasonably be anticipated” suggests

that Congress did not intend to require EPA to show

a precise relationship between a source’s emissions

and all or a specific fraction of the visibility impair-

ments to the Act, EPA subsequently established the Grand Canyon Visi-

bility Transport Commission (“GCVTC”), 42 U.S.C. § 7492(f), whose

duty it is to make recommendations on “promulgation of regulations under

section 7491 of this title to address long reuge strategies for addressing

regional haze which impairs visibility in affected class 1 [Federal] areas”

like the Grand Canyon. 42 U.S.C. § 7492(d)(2C). Congress envisioned

GCVTC as a long-term complement to EPA’s then-pending efforts to rem-

edy impairment at the Grand Canyon which was attributable to NGS. See

136 Cong. Rec. $2890 (March 21, 1990). :

2732 CAWCD v. EPA

ment within a Class I area. Rather, EPA is to assess

the risk in light of policy considerations regarding

the respective risks of overprotection and underpro-

tection.

Haze in the Grand Canyon at 5.

[5S] Acting in place of the state of Arizona pursuant to an

FIP under 42 U.S.C. § 7410(c), EPA “stands in the shoes of

the defaulting State, and all of the rights and duties that would

otherwise fall to the State accrue instead to EPA.” 54 Fed.

Reg. at 36,952. EPA is therefore granted broad discretion in

determining whether visibility impairment is “reasonably

attributable” to a given source. See 40 C.F.R. §51.301(s)

(defining “reasonably attributable” as “attributable by visual

Observation or any other technique the State deems

appropriate’) (emphasis added). This broad, all-inclusive def-

inition of the term “reasonably attributable” directly refutes

Petitioners’ argument that EPA is limited to addressing visi-

bility impairment caused by a noticeable plume that is directly

traceable to a given source through the use of visual observa-

tion or simple monitoring techniques.

[6] We conclude that the technical, scientific record more

than adequately supports EPA’s reasonable conclusion that

visibility impairment in the Grand Canyon is “reasonably

attributable” to NGS. We defer to the agency's reasonable

interpretation of its own regulations and statutory mandate. Its

interpretation is most certainly consonant with Congress's

apparent intent. For instance, in his introduction of the Con-

ference Committee report to the House, Representative Henry

Waxman stated:

Protecting the Grand Canyon simply must become a

normal business practice of the American industry

... [T]he Four Corners and Navajo powerplants can

expect to retrofit with additional pollution controls to

ee

CAWCD v. EPA 2733

limit the vast deterioration in visibility which their

plumes have caused.

impairment of visibility is the single most appar-

ent impact air pollution has on the environment. It is

our intent that aggressive steps be taken to reduce

this eyesore which has defaced our grand vistas in

the West.

123 Cong. Rec. 27,076 (1977) (statement of Rep. Waxman).

EPA’s interpretation of its authority to determine whether vis-

ibility impairment is “reasonably attributable” to NGS is enti-

tled to deference from this court since the agency’s * “choice

represents a reasonable accommodatior of conflicting policies

that were committed to the agency’s care by the statute,” ”

which this court “ ‘should not disturb’” since it does not

appear “ ‘from the statute or its legislative history that the

accommodation is not one that Congress would have

sanctioned.’ Chevron, 467 U.S. at 845 (quoting United

States v. Shimer, 367 U.S. 374, 383 (1961)).

B. The Final Rule as the Product of “Reasoned

Decisionmaking”

Petitioners proffer various arguments that the Final Rule is

not the product of “reasoned decisionmaking.” They assert

that EPA has acted arbitrarily and capriciously by overesti-

mating the improvements in visibility expected from the Final

Rule’s emission controls at NGS, by purportedly failing to

address criticisms to the scientific data and analyses on which

it relied, and by allegedly ignoring certain evidence while

placing undue reliance on other evidence. At bottom, how-

ever, Petitioners’ real complaint appears to be that the Final

Rule will most likely lead to minimal visibility improvement

at the Grand Canyon while imposing a substantial financial

burden on them. Nonetheless, we find unsupported Petition-

ers’ legal claim that EPA acted arbitrarily and capriciously in

promulgating the Final Rule. The Final Rule makes

2734 CAWCD v. EPA

“reasonable progress” toward the national goal of remedying

visibility impairment at the Grand Canyon. and is the product

of reasoned decisionmaking.

1. The Final Rule as “Reasonable Progress” toward the

National Goal of Remedying Visibility Impairment at

the Grand Canyon

In reviewing whether the agency's action in promulgating

the Final Rule was arbitrary and capricious, this court “is not

to substitute its judgment for that of the agency.” State Farm,

463 U.S. at 43. Instead, we inquire whether the agency has

“examine[d] the relevant data and articulate[d] a satisfactory

explanation for its action, including a ‘rational connection

between the facts found and the choice made. ~ /d. (quoting

Burlington Truck Lines, Inc. v. United States, 371 U.S. 156,

168 (1962)). Our review is limited to “whether the agency

considered the relevant factors and whether there has been a

clear error of judgment.” NRDC, Inc. v. EPA, 966 F.2d 1292,

1297 (9th Cir. 1992) (citing Citizens to Preserve Overton

Park, Inc. v. Volpe, 401 U.S. 402, 416 (1971)). In this case,

the relevant factors are provided by Congress’s definition of

“reasonable progress” in 42 U.S.C. § 7491(g)(1). Addition-

ally, as the D.C. Circuit recently noted in discussing a similar

provision of the Act, “[bJecause Congress did not assign the

specific weight the Administrator should accord each of these

factors, the Administrator is free to exercise his discretion in

this area.” New York v. Reilly, 969 F.2d at 1150 (discussing

42 U.S.C. § 7411(a)(1)(C)).”°

Petitioners incorrectly suggest that EPA was required to engage in

“cost-benefit” analysis. Congress has not required “cost-benefit” analysis

in the Act. Cf. American Textile Mfrs. Inst., Inc. v. Donovan, 452 U.S.

490, 510 (1981) (“When Congress has intended that an agency engage in

cost-benefit analysis, it has clearly indicated such intent cr the face of the

Statute.”).

CAWCD v. EPA 2735

a. The “Reasonable Progress” Provisions

[7] In the Act, Congress directed EPA to promulgate regu-

lations to assure “reasonable progress toward meeting the

national goal” of preventing future, and remedying existing

visibility impairment in Class I federal areas like the Grand

Canyon. See 42 U.S.C. §§ 7491(a)(4), (a)(1), (b)(2), (b)(2)(B).

Congress chose not to define the term “reasonable progress,”

but instead set forth several factors for the agency to consider:

In determining reasonable progress there shall be

taken into consideration the costs of compliance, the

time necessary for compliance, and the energy and

nonair quality environmental impacts of compliance,

and the remaining useful life of any existing source

subject to such requirements[.]

42 U.S.C. § 7491(g)(1). In promulgating the Final Rule, EPA

relied on the “reasonable progress” provisions as its statutory

authority. See 56 Fed. Reg. at 50,177 (noting that EPA adopts

the “reasonable progress” provisions of section 169A(b)(2),

42 U.S.C. § 7491(b)(2) as the “legal rationale” for the Final

Rule).

[8] Generally, the Act and its regulations require the appli-

cation of BART once it has been determined that visibility

impairment is “reasonably attributable” to an existing source

like NGS. See 42 U.S.C. § 7491(b)(2); 40 C_E.R. § 51.302(c)

(4)(i). Under the unique circumstances of this case, however,

EPA chose not to adopt the emission control limits indicated

by BART analysis, but instead to adopt an emission limita-

tions standard that would produce greater visibility improve-

ment at a lower cost. Congress’s use of the term “including”

in § 7491(b)(2) prior to its listing BART as a method of

attaining “reasonable progress” supports EPA’s position that

it has the discretion to adopt implementation plan provisions

other than those provided by BART anal’ses in situations

where the agency reasonably concludes tha’ more “reasonable

ieee aaa ai aii

2736 CAWCD v. EPA

progress” will thereby be attained. Since the Act itself is

ambiguous on the specific issue, we apply the Supreme

Court’s deferential standard from Chevron and hold that the

agency’s reliance on the “reasonable progress” provisions is

a “permissible construction of the statute,” 467 U.S. at 843,

since “reasonable progress” is the overarching requirement

that implementation plan revisions under 42 U.S.C. § 7491(b)

(2) must address.

b. EPA Reasonably Considered the Relevant Factors

[9] The administrative record reveals that EPA adequately

considered the relevant factors in promulgating the Final Rule."

Petitioners’ essential argument does not claim that EPA failed

to consider the relevant factors, but instead contends that EPA

erred in its consideration of those factors. This court is not to

substitute Petitioners’ judgment, or its own, for that of EPA,

as long as the agency's interpretation is reasonable. In fact,

this is just the type of case in which the Supreme Court has

stated that judicial review should “be at its most deferential,”

because the agency is “making predictions, within its area of

special expertise, at the frontiers of science.” Baltimore Gas

& Elec. Co., 462 U.S. at 103. We therefore find that the agen-

cy’s interpretation of the evidence, and its weighing of the rel-

evant factors, are reasonable, and that the Final Rule is the

product of “reasoned decisionmaking.”

2. Petitioners Have Failed to Demonstrate that EPA

Acted in an Arbitrary and Capricious Manner

Petitioners argue that EPA relied on discredited elements of

a seriously flawed study, failed to refute significant criticisms

which undercut the basis for EPA’s improvement estimate,

and refused to consider highly relevant evidence which con-

‘Actually, EPA not only considered the “reasonable progress” factors,

but also considered expected visibility improvement under the Final Rule,

as required by BART analysis.

CAWCD v. EPA 2737

tradicted its estimate. They conclude that EPA has acted arbi-

trarily and capriciously. We reject these arguments.

[10] The allegedly discredited study on which EPA relied

was the Park Service’s analysis of the WHITEX data. The

record reveals, however, that EPA acknowledged the limita-

tions of the Park Service report (even seeking and obtaining

NAS comment on it), and did not rely solely upon that

report’s conclusions. As discussed supra, EPA’s determina-

tion that visibility impairment at the Grand Canyon is

“reasonably attributable” to emissions at NGS was eminently

reasonable and adequately supported by a substantial body of

scientific and technical evidence in the record. Further, as the

D.C. Circuit noted in NRDC v. EPA:

“The Administrator may apply his expertise to draw

conclusions from suspected, but not completely sub-

stantiated, relationships between facts, from trends

among facts, from theoretical projections from

imperfect data, from probative preliminary data not

yet certifiable as ‘fact,’ and the like.”

902 F.2d at 968 (quoting Ethyl Corp. v. EPA, 541 F.2d 1, 28

(D.C. Cir.) (en banc), cert. denied, 426 U.S. 941 (1976)).

While this court must still ensure that “the agency has made

a reasoned decision based on ‘reasonable extrapolations from

some reliable evidence,’ ” id. (quoting NRDC, Inc. v. Thomas,

805 F.2d 410, 432 (D.C. Cir. 1986)), this standard adequately

has been met here.

Petitioners further challenge as arbitrary and capricious

EPA’s estimate of the quantum of visibility improvement

which the agency expects to result from imposition of the

Final Rule’s emission limits at NGS. In arriving at its estimate

that the Final Rule “should improve the winter seasonal aver-

age visibility above the rim of the [Grand Canyon] approxi-

mately 7 percent,” 56 Fed. Reg at 50,180, EPA discussed the

estimates of three studies: (1) the Park Service’s WHITEX

2738 CAWCD v. EPA

estimate of 14% improvement, (2) the NGSVS estimate of at

most a 2% improvement, and (3) a modelling study, con-

ducted by Douglas Latimer and submitted during the com-

ment period, which estimated a range of 4 to 8%

improvement. Petitioners contend that EPA arrived at its 7%

estimate in an arbitrary and capricious manner because, they

contend, EPA failed to consider and place in the administra-

tive record a new report prepared by Latimer in which he

allegedly reconsiders his previous estimate of visibility

improvement, reducing it to 3 to 4%. Petitioners additionally

claim that failure to put this new report in the administrative

record constitutes an abuse of discretion. We find Petitioners”

challenge both legally untenable and factually misleading.

First, the challenge is legally untenable because it relies on

this court’s consideration of evidence — the new Latimer

report — which is not a part of the administrative record. In

the Act, Congress explicitly and unambiguously provided that

“the record for judicial review shall consist exclusively” of

certain specifically enumerated categories of materials. 42

U.S.C. § 7607(d)(7)(A) (emphasis added). “Nothing in the

statute or its legislative history indicates that a party or the

agency may reopen the record by placing additional materials

(other than those required by the statute and wrongfully omit-

ted by EPA) in the docket after promulgation of the rule.”

Lead Indus. Ass’n v. EPA, 647 F.2d 1130, 1183 (D.C. Cir.),

cert. denied, 449 U.S. 1042 (1980). The new Latimer report

was not included in the administrative record, nor was it

required to be. EPA explains that the report was not directly

related to NGS, but was instead submitted to EPA in the con-

text of the GCVTC regional haze efforts under section 169B

of the Act.

Petitioners’ reliance on Kent County v. EPA, 963 F.2d 391 (D.C. Cir.

1992) is misplaced because, unlike the Clean Air Act here, the statutory

provision at issue in Kent County contained no special procedural provi-

sions displacing general rules of administrative law governing the admin-

istrative record for judicial review. Compare 42 U.S.C. §§ 7607(d)(1), (d)

(7M A) with 42 U.S.C. § 9613(a).

a anaemia

Se

CAWCD v. EPA 2739

Additionally, Petitioners ignore the route for administrative

and judicial review of “new information” and alleged

“procedural errors” in the creation of the administrative

docket that Congress created: 42 U.S.C. § 7607(d)(7)(B)’s

mechanism for reconsideration by the agency. Petitioners

have failed to move for reconsideration and are now trying to

circumvent this congressionally mandated route by petitioning

for direct judicial review. We therefore deny Petitioners”

motion to supplement the administrative record and direct

Petitioners to consult section 7607(d)(7)(B) in order to fash-

ion their claim as a petition for reconsideration to EPA, as

Congress had envisioned.

Second, even if this court were to consider Petitioners’

argument on the merits, their challenge is misleading because

Latimer himself indicates that Petitioners have misrepresented

the statements and conclusions contained in his new report.

We therefore reject Petitioners’ argument that EPA has some-

how acted arbitrarily or capriciously, or abused its discretion,

through the manner in which it arrived at its estimate of visi-

bility improvement.

V. CONCLUSION

In the final analysis, Petitioners simply adhere to a different

interpretation of the rather disparate and equivocal scientific

data in the record. While Petitioners may not be satisfied with

- EPA’s responses, it is not EPA's duty to satisfy all of the con-

cems of potentially affected or aggrieved parties. EPA con-

ducted an extensive and involved notice and comment period,

and adequately met its statutory obligation of responding to

Significant comments and criticisms under 42 U.S.C.

§ 7607(d)(6)(B). Notwithstanding Petitioners’ challenge, the

Final Rule is the result of a site-specific informal rulemaking

process that included virtually unprecedented cooperation

between the governmental agency and the directly affected

parties. See generally D. Michael Rappoport & John F.

Cooney, Visibility at the Grand Canyon: Regulatory Negotia-

2740 CAWCD v. EPA

tions Under the Clean Air Act, 24 Ariz. St. LJ. 627 (1992).

Petitioners’ arguments afford no reason for this court disrup-

tively to interject itself into the picture. Because Congress del-

egated to EPA the power to “regulate on the borders of the

unknown, this court will not interfere with the agency's

“reasonable interpretations of equivocal evidence.” Public

Citizen Health Research Group v. Tyson, 796 F.2d at 150S.

Even if this case highlights how hard it is to engage in

“reasoned decisionmaking” in cases involving scientific

uncertainty, EPA‘s actions in promulgating the Final Rule

were reasonable and within the bounds of its statutory author-

ity, and not arbitrary and capricious.

The Districts” petition for review and motion to supplement

the administrative record are accordingly DENIED.

PRINTED FOR

ADMINISTRATIVE OFFICE—U.S. COURTS

BY BARCLAYS / ELECTROGRAPHIC—SAN FRANCISCO—{415) 588-1155

The summary, which does not constitute a part of the opinion of the court, is copyrighted

© 1993 by Barclays Law Publishers.

50172 Federal Register/Vol. 56, No. 192/ Thursday, October 3,

1991/Rules and Regulations

ENVIRONMENTAL PROTECTION AGENCY

40 CFR Part 52

A.[AD-FRL-4018-5]

Approval and Promulgation of Implementation Plans: Revision

of the Visibility FIP for Arizona

AGENCY: U.S. Environmental Protection Agency (EPA).

ACTION: Final rule.

SUMMARY: This notice revises the Federal implementation plan

(FIP) for the State of Arizona to include a sulfur dioxide (SO,)

emission limit of 42 nanograms per Joule (ng/J) [0.10 pound per

million British thermal units (lb/MMBtu)], heat input for the Navajo

Generating Station (NGS) to remedy visibility impairment in the

Grand Canyon National Park (GCNP). Compliance with this

emission limit will be phased-in by unit in 1997, 1998, and 1999

and determined on a plant-wide annual rolling average basis. In

addition, NGS will be required to reschedule its maintenance such

that 6 unit-weeks of maintenance will be performed during the

winter months.

This action is taken pursuant to sections 169A and 110(c) of the

Clean Air Act (Act), 42 U.S.C. sections 7491 and 7410(c), which

require EPA, upon default by a State, to take appropriate measures

to remedy certain certified visibility impairments in mandatory Class

I areas. The timing of today’s action is in accordance with the

revised settlement agreement between EPA and Environmental

Defense Fund (EDF) in EDF vy. Reilly, No. C82-6850 RPA (N.D.

Cal.).

EFFECTIVE DATE: This action will be effective on November

4, 1991.

ADDRESSES: Docket: Pursuant to section 307(d)(1)(B) of the

Act, 42 U.S.C. 7607(d)(1)(B), this action is Subject to the

procedural requirements of section 307(d). Therefore, EPA

established Docket A-89-02A for this action. Materials related to

the development of this notice have been placed in this docket.

2

Materials related to EPA’s preliminary attribution determination (54

FR 36948 (September 5, 1989)) have been placed in Docket A-89-

02. For background information, materials related to the

development of the visibility protection program (40 CFR 51.300 et

seq.) are available in Docket A-79-40. Also, materials related to

the development of the visibility new source review (NSR) and

visibility monitoring strategies are available in Docket A-84-32.

Materials related to the visibility long-term strategy, implementation

of control strategy, and integral vista program are available in

Docket A-85-26. All dockets are available for public inspection and

copying between 8:30 a.m. to 12 noon and 1:30 p.m. to 3:30 p.m.,

Monday through Friday, at EPA’s Central Docket Section, Office

of the General Counsel, room 1500, 401 M Street, SW.,

Washington, DC 20460. A reasonable fee may be charged for

copies.

FOR FURTHER INFORMATION CONTACT: Mr. David H.

Stonefield, U.S. Environmental Protection Agency, Office of Air

Quality Planning and Standards (MD-15), Research Triangle Park,

North Carolina 27711, (919) 541-5350 or FTS 629-5350.

SUPPLEMENTARY INFORMATION

I. Background

A. Regulatory Requirements

Section 169A of the Act, 42 U.S.C. 7491, sets as a national

goal “the prevention of any future, and the remedying of any

existing, impairment of visibility in mandatory Class I Federal areas

which impairment results from manmade air pollution." Mandatory

Class I Federal areas are certain national parks, wildernesses, and

international parks as described in section 162(a) of the Act, 42

U.S.C. 7472(a). Section 169A requires that EPA promulgate

regulations to assure reasonable progress toward meeting the

national goal for mandatory Class I Federal areas where EPA has

determined that visibility is an important value. On November 30,

1979, EPA identified 156 such areas, including the GCNP in

Arizona, where visibility is an important air quality-related value

(44 FR 69122). Section 169A specifically requires EPA to

a

3

promulgate regulations requiring States to amend their State

implementation plans (SIP’s) to provide reasonable progress toward

meeting the national goal for the 156 areas. On December 2, 1980,

EPA promulgated the required visibility regulations (45 FR 80084,

codified at 40 CFR 51.300 ef seq.). The visibility regulations

require the 36 States listed in section 51.300(b), including Arizona,

to:

1. Coordinate SIP development with the appropriate Federal

land managers (FLM’s).

2. Develop programs to assess and remedy Phase I visibility

impairment from existing sources and to prevent visibility

impairment from new sources.

3. Develop a long-term (10 to 15 years) strategy to address,

among other things, Phase I visibility impairment to assure

reasonable progress toward the national goal.

4. Develop a visibility monitoring strategy to collect

information on visibility conditions.

5. Consider in all aspects of visibility protection any “integral

vistas" (important views of landmarks or panoramas that extend

outside of the boundaries of the Class I area) identified by the

FLM’s as critical to the visitors’ enjoyment of the Class I areas.

The affected States were required to submit revised SIP’s satisfying

these provisions by September 2, 1981 (see 45 FR 80091, codified

at 40 CFR 51.302(a)(1)).'

' The EPA's 1980 regulations implementing the statutory requirements of

section 169A address visibility impairment that is reasonably attributable to a single

source or small group of sources (“Phase |" impairment) (see generally 45 FR

80084 (December 2, 1980)). Recognizing certain scientific and technical

limitations, EPA, in promulgating the 1980 regulations, deferred regulatory action

on more “complex problems such as regional haze and urban plumes.” id. at

80086. Today, in accordance with the 1980 regulations, EPA is taking regulatory

action to remedy visibility impairment that is caused by NGS. The EPA is in the

process of establishing the Grand Canyon Visibility Transport Commission as

required by section 169B of the Act. The duties of that commission include

making recommendations on “promulgation of regulations under section 169A to

address long-range strategies for addressing regional haze which impairs visibility

in * * * the GCNP” (section 169B(d)(2)(C)).

4

The second and third requirements listed above are of particular

relevance to today’s action. Pursuant to 40 CFR 51.302(c)(2), each

affected State is required to include in its SIP such emission

limitations, schedules of compliance, and other measures as may be

necessary to make reasonable progress toward the national visibility

~ goal. Under 40 CFR 51.302(c)(1), an FLM may certify to a State

that there exists impairment of visibility in any mandatory Class I

Federal area. Such impairment must be addressed in accordance

with 40 CFR 51.302(c) which sets forth measures for achieving

reasonable progress, including best available retrofit technology

(BART) and a long-term strategy (see 40 CFR 51.302(c)(1) and (2),

section 169A(b)(2)(A) and (B)). Pursuant to 40 CFR

51.302(c)(4)(i), where impairment is certified at least 6 months

prior to plan submission, an affected State must identify each

existing stationary facility which may “reasonably be anticipated to

cause or contribute" to any such impairment which is “reasonably

attributable to that existing stationary facility," and analyze for

BART any facility so identified. _ "Reasonably attributable"

impairment is impairment “attributable by visual observation or any

other technique the State deems appropriate" (40 CFR 51.301(s)).

Where a State defaults on its obligation sunder the visibility

regulations, EPA may act in place of the State pursuant to . FIP

under section 110(c) of the Act, 42 U.S.C. 7410(c),? and

? Section 110(c) requires EPA to promulgate FIP’s whenever a State fails to

submit an implementation plan (or portion thereof) which meets the requirements

of section 110, whenever the Administrator determines that a plan (or portion) is

not in accordance with the requirements of section 110, or whenever the State fails

to revise its plan within 60 days after notification by the Administrator in

accordance with section 110(a)(2)(H).

During the House of Representatives’ consideration of the Conference Report

on the 1977 Amendments to the Clean Air Act, Congressman Rogers reiterated that

the conferees had agreed that EPA was to act where States failed to carry out their

duties in implementing the requirements of section 169A. The pertinent part of

Congressman Rogers’ statement is as follows:

“The conferees * * * rejected a motion to delete EPA's supervisory role under

section 110 to assure that the required progress towards that goal (the national

visibility goal) will be achieved by the revised State plan. If a State visibility

protection plan is not adequate to assure such progress, the Administrator must

PEL ott Sha Vin att Na

5

promulgate such limitation and measures as are required to achieve

reasonable progress. In such cases, all of the rights and duties that

would otherwise fall to the State accrue instead to EPA.

The visibility regulations promulgated at 40 CFR 51 .302(c)(4)(i)

require that once a Phase I impairment has been certified, a State

(or EPA if the State’s visibility protection program addressing

BART has not been approved, and EPA is acting in its place) must

analyze for BART any specific existing stationary facility it

identifies as a "reasonably attributable" source of the impairment.

Pursuant to section 169A(b) of the Act (42 U.S.C. 7491(b)) and 40

CFR 51.302(C)(4)(iii), the emission limitation representing BART

for fossil fuel-fired power plants with a generating capacity in

excess of 750 megawatts (MW) must be determined pursuant to

guidelines promulgated by the Administrator. This statutorily-

required procedure for conducting a BART analysis is found in

"Guidelines for Determining Best Available Retrofit Technology

Analysis for Coal-Fired Power Plants and Other Stationary

Facilities" (EPA-450/3-80-009b BART Guidelines). A copy of this

document may be found in Docket A-89-02A.

In December 1982, environmental groups, including EDF, filed

a citizens suit in the United States District Court for the Northern

District of California alleging that EPA had failed to perform a

nondiscretionary duty under section 110(c) of the ACT, 42 U.S.C.

7410(c), to promulgate visibility FIP’s for the 35 States? that, at

that time, had failed to submit SIP’s to EPA as called for by the

1980 visibility regulations, EDF v.. Reilly, No. C82-6850 RPA

disapprove that portion of the SIP and promulgate a visibility protection plan under

section 110(c). Thus, visibility protection in mandatory federal class I areas

remains a national commitment, which is nationally enforceable."

See Senate Comm. on Environment and Public Works, 95th Cong. 2d.Sess.

“A Legislative History of the Clean Air Act Amendment of 1977" No. 16, vol. 3

at 311, 320-21 (Comm. Print 1978).

* The State of Alaska had submitted a SIP which was approved on July 5,

1983 at 48 FR 30623.

ee

6

(N.D. Cal.). The State of Arizona was one of the 35 States that

failed to submit a revised SIP to EPA.

The EPA and the plaintiffs negotiated a settlement agreement

for the remaining States which the court approved by order on April

20, 1984. For more information on details of the provisions of the

original settlement, including a schedule of actions by EPA, see

EPA’s announcement of the agreement at 49 FR 20647 (May 16,

1984).

B. Settlement Agreement

To remedy the States’ failure to submit the necessary SIP

revisions during the time specified by the regulations, the settlement

agreement replaced the original regulatory deadlines for visibility

SIP provisions with a rulemaking schedule agreed to by the parties

and approved by the court. This schedule required EPA to review

the existing SIP’s to determine any deficiencies, allow the States to

cure those deficiencies, and to promulgate FIP’s on a specified

schedule for those States that still did not submit visibility SIP

revisions to EPA. Specifically, the first part of the agreement

required EPA to promulgate FIP’s which cover the monitoring and

NSR provisions of 40 CFR 51.305 and 51.307. The EPA

promulgated its monitoring strategy for 23 States and its NSR

provisions for 21 States, including Arizona, at 59 FR 28544 (July

12, 1985), 51 FR 5504 (February 1, 1986), and 51 FR 22937 (June

24, 1986). In separate notices, EPA approved the SIP’s of the other

States with respect to monitoring and NSR.

The second part of the settlement agreement required EPA to

determine the adequacy of the SIP’s to meet the remaining

provisions of the visibility regulations and gave the States until

December 1986 to submit additional measures that would avoid the

need for a FIP. These provisions are the general plan provisions

for achieving reasonable progress toward the national visibility goal

including BART and other implementation control strategies

(§ 51.302), integral vista protection (§ 51-302-307), and long-term

Strategies (§ 51.306). The settlement agreement required EPA to

SIS et abtiee Ne OF

7

promulgate FIP’s to remedy any deficiencies on a specified

schedule.

Pursuant to 40 CFR 51.302(c)(1), on November 14, 1985, the

Department of the Interior (DOI) certified the existence of visibility

impairment in all Class I areas within its jurisdiction in the lower 48

States.

On January 23, 1986, EPA determined that the SIP’s of 32

States (including Arizona) were deficient with respect to the

remaining visibility provisions (51 FR 3046) and offered the States

an Opportunity to submit corrective SIP revisions. Thereafter, EPA

and the plaintiffs negotiated revisions to the settlement agreement

which extended the deadlines for State action or, if the States failed

to respond, Federal action proposing FIP’s to remedy these

deficiencies. The court approved these revisions by its order of

September 9, 1986.‘

On March 24, 1986, the DOI sent a letter to EPA which

supplemented its earlier certification of visibility impairment. The

letter addressed the GCNP and identified the NGS, a coal-fired

power plant located near Page, Arizona, as a probable source of

impairment in this Class I area. A copy of this letter may be found

in Docket A-89-02.

Thirty-two affected States failed to submit visibility SIP

revisions in response to the notice of deficiency. Consequently, in

accordance with the revised settlement agreement, on March 12,

1987 (52 FR 7802), EPA proposed to disapprove the SIP’s of 32

States, including Arizona, for failing to meet the remaining

provisions of the visibility regulations, including general plan

requirements to achieve reasonable progress toward the national

visibility goal (which in turn includes BART, long-term strategies,

and other control strategies). Also in accordance with the

agreement, on November 24, 1987 (52 FR 45132), EPA took final

action disapproving the affected SIP’s, again including Arizona. In

"h copy of the settlement agreement and revisions is available in Docket A-

85-26 at the address given at the beginning of this notice.

8

that action, EPA also promulgated, as FIP measures under section

110(c), general plan requirements for these States. The EPA also

determined that BART was unnecessary in 26 States because it could

not reasonably attribute the visibility impairment of mandatory Class

I Federal areas in these States to specific sources or small groups of

sources. In addition, under the revised agreement, EPA deferred

until August 31, 1988 a decision regarding the need for BART or

other control measures in the FIP’s for the States of Arizona,

Maine, Minnesota, and Utah to address certified visibility

impairments in seven Class I areas in these States which potentially

could be reasonably attributed to a specific source pending

acquisition and evaluation of additional technical information

regarding the potential sources of impairment.

Because all of the additional information needed by EPA was

forthcoming but still not available by August 1988, the Agency

sought and received a second extension, until August 31, 1989, of

the deadline for issuing a rulemaking proposal regarding the need

for BART or other control measures to remedy visibility

impairments in three of the remaining Class I areas (Moosehorn

Wilderness, GCNP, and Canyonlands National Park). With regard

to the impairment in the GCNP, EPA delayed action in order to

allow the National Park Service (NPS) time to analyze the data from

a 1987 winter visibility attribution study, termed the Winter Haze

Intensive Tracer Experiment (WHITEX), conducted in the Colorado

Plateau where the GCNP is located. In the meantime, on May 19,

1989 (54 FR 21904), in accordance with the second revision to the

settlement agreement, EPA promulgated final decisions concerning

certified visibility impairments in four of the seven Class I areas.

B n_ monitoring con in e PA f at

visibility impairments were r l i l

ifi rc Th r PA i it_w

n revi ’s for f Mai inn

Arizon incl ART or n rategi m

impairments in Roosevel Ilo Int i k

AS LE A ARE my a Ng

9

Voyageurs National Park (Minnesota). Saguaro Wilderness

(Arizona), and Petrified Forest National Park (Arizona.

In April 1989, EPA received a draft report on WHITEX from

the NPS.° Because of the delay in receiving this report, EPA

believed that it lacked sufficient time to complete its analyses and

issue a proposed rule by the August 31, 1989 deadline.

Accordingly, EPA and EDF filed a joint motion to revise the

settlement agreement for a third time, which was approved by order

of the court dated July 6, 1989. Under this revision, there was no

change in the deadlines for proposed action regarding Canyonlands

National Park or Moosehorn Wilderness. As to GCNP, the third

revision to the settlement agreement divided EPA’s duty into two

parts. The EPA would proceed, on a preliminary basis, to issue a

finding on reasonably attributable impairment by the August 31,

1989 deadline. However, if EPA did provisionally identify a

specific source of impairment, it would solicit comments on that

finding and would have additional time to conduct a BART analysis.

The EPA was to issue a rulemaking proposal on the need for BART

by February 1, 1990 unless, in response to comments, EPA rejected

its proposed identification and instead determined that BART was

unnecessary.

In accordance with the third revision to the settlement

agreement, EPA published a notice of proposed rulemaking on

September 5, 1989 (54 FR 36948). Regarding Moosehorn

Wilderness, EPA identified a source of the certified impairment but

proposed that BART was unnecessary because the impairment would

be adequately remedied by the retirement of certain _ existing

emission units and the addition of pollution controls on other units

pursuant to a permit issued under the Act’s prevention of significant

deterioration provisions. As to Canyonlands National Park, EPA

could not reasonably attribute the certified impairment to a specific

source at that time and so proposed that BART was unnecessary.

* Malm, et al., “The National Park Service Report on WHITEX Draft Final

Report” (April 7, 1989).

en

10

On June 13, 1990 (55 FR 24060), EPA issued final decisions that

BART was not necessary to address impairment in either the

Moosehorn Wilderness or the Canyonlands National Park.

Regarding the GCNP, in the September 5, 1989 notice, EPA

preliminarily attributed several episodes of wintertime visibility

impairment to emissions form the NGS. In that regard, EPA

reviewed the draft NPS report on WHITEX and concurred with the

findings of the NPS. The EPA solicited comments on the merits of

its preliminary attribution finding.

On November 28, 1989, the Salt River Project Agricultural

Improvement and Power District (SRP), et al.,[sic] and Alabama

Power Company, et al., [sic] intervenors in EDF v. Reilly, filed a

motion requesting a 1-year delay in the rulemaking schedule so that,

among other things, the intervenors could conduct their own studies

of the visibility impairment in the GCNP. By order dated January

9, 1990, the court granted the intervenors’ motion to extend the

deadlines. This order set February 1, 1991, as the new deadline for

EPA to propose whether or not to require BART emission limits for

the NGS to remedy winter visibility impairment. The order also

extended the deadline for final action on any proposal regarding

BART until October 1991 (6 months after the close of the comment

period).°

C. Navajo Generating Station

The NGS is a 2250 MW coal-fired power plant located near

Page, Arizona, approximately 20 kilometers from the northern

boundary of the GCNP. The NGS is a baseload generating station

consisting of three 750 MW units which became operational

between 1974 and 1976. ‘the SRP is the operating agent for NGS

© The EDF has appealed the extension of rulemaking deadlines in this case,

EDF v. Reilly, No. 90-15264 (9th Cir.). That appeal is pending. In a

Memorandum of Understanding (MOU) discussed below, parties to this litigation

have agreed, in light of today’s action, to petition (a) the Ninth Circuit to vacate

the judgment below and remand the matter to the court below with instructions to

dismiss and (b) the District Court for the Northern District of California to dismiss

EDF v. Reilly, No. C82-6850.

11

which is jointly owned by the SRP,’ the Los Angeles Department

of Water and Power, the Arizona Public Service Company, the

Nevada Power Company, and the Tucson Electric Power Company.

Existing pollution control equipment at NGS includes electrostatic

precipitators for particulate matter (PM) removal and specific burner

design for nitrogen oxides (NO,) control.

D. February 8, 1991 Proposal

On February 8, 1991, EPA proposed to revise the FIP for

Arizona to include emission limits to address the visibility

impairment observed in GCNP. The following is a summary of the

main issues discussed in the proposal notice.

1. Emission Limitation

The EPA proposed to adopt a continuous SO, emission

limitation of 0.30 lb/MMBtu. Compliance would be determined on

a 30-day rolling average and would be phased in between 1995 and

1999. Because of the uncertainty in determining the improvement

in visibility expected as a result of reducing emissions at the NGS,

EPA solicited comments on three alternative control Strategies for

NGS:

a. A continuous SO, emission limitation of 0.50 Ib/MMBtu.

Compliance would be determined on a 30-day rolling average and

would be phased in between 1995 and 1999.

b. A continuous SO, emission limitation of 0.10 Ib/MMBtu.

Compliance world be determined on a 30-day rolling average and

would be phased in between 1995 and 1999.

c. AJanuary 10, 1991 proposal submitted by SRP under which

it would test alternative control technologies, and if one of the

technologies met a minimum removal efficiency at a set cost, the

NGS would install that technology and operate it in the wintertime.

’ The SRP owns 21.7 percent of the NGS project for its own use and benefit,

and 24.3 percent for the use and benefit of the United States in accordance with the

NGS project agreements. The term "NGS participants” is used in this notice to

refer to the parties which have rights and responsibilities associated with the

operation of NGS and includes the owners, the operators, and the U.S. Bureau of

Reclamation (for the United States).

|

12

If none of the technologies met the test criteria, NGS would reduce

its emissions by 70 percent (0.30 Ib/MMBtu emission limit)

measured on an annual basis by the year 2000.

In addition, EPA requested comment on whether another

emission limitation or a different averaging period may be more

appropriate.

2. Attribution

As part of the February 8, 1991 proposal, EPA reopened the

comment period on its September 5, 1989 preliminary finding that

several episodes of wintertime impairment in the GCNP was

reasonably attributable to NGS. The EPA noted both in the

September 1989 and the February 1991 notices that the finding was

not based on any single analytical technique in the NPS report on

WHITEX, but rather-on the collection of techniques performed by

the NPS using WHFTEX and other data. In detail, EPA explained

that all of the techniques used by the NPS support the conclusion

that NGS is a source of visibility impairment in the Grand Canyon

during certain wintertime episodes. Also, in some detail, EPA

discussed the National Academy of Science’s (NAS’) review of the

WHITEX study and discussed other visibility studies performed

subsequent to WHITEX.

The National Research Council of the NAS reviewed and

reported on the scientific methods used in the WHITEX report. The

NAS report, entitled "Haze in the Grand Canyon: An Evaluation of

the Winter Haze Intensive Tracer Experiment" (October 1990), 3

contained a qualitative assessment of WHITEX which supported ;

EPA’s finding that, on some days, NGS is a source of visibility :

impairment in the GCNP.* Further, the NAS found that the "rate :

of SO, emissions from NGS is easily large enough to serve as the

ee

* The report found, for example, that the detection at Hopi Point (GCNP) of

the unique tracer released from NGS in the WHITEX study “is an unambiguous

indicator that air parcels containing NGS emissions did impinge on the GCNP on

several occasions.” Executive Summary at p. 3.

13

source of the sulfur measured in the GCNP."° However, the report

also found that the data base and data analyses techniques in the

WHITEX report were not, Standing alone sufficient to ascertain the

quantitative NGS contribution to haze at any given time. The NAS

report has been placed in the rulemaking docket.'°

The SRP has conducted another Study known as the Navajo

Generating Station Visibility Study (NGSVS). That study attempted

to measure the degree of contribution by the NGS to visibility

impairment in the GCNP during the 1989-1990 winter season and

to assess the level of improvement expected as a result of reducing

the SO, emissions at NGS. The SRP Submitted a draft report

describing that study and its results! during the public comment

period on EPA’s February 8, 1991 proposal. The NGSVS

concluded that NGS emissions were present at Hopi Point less

frequently than during the WHITEX period. In addition, the

NGSVS concluded that on days when NGS emissions were present,

visibility impairment at Hopi Point associated with those emission

was substantially less than the amount calculated in the NPS report

on WHITEX. However, results indicate that NGS was responsible

for a significant quantity of sulfate and haze during specific

visibility impairment episodes in GCNP.

3. BART Analysis

In light of its attribution finding, EPA conducted an analysis in

accordance with the BART Guidelines (EPA-450/3-80-009b) and 40

CFR 51.301(c)(4)(iii)."2 That analysis included consideration of

the following: the costs of compliance, the energy and non-air

* The NAS report at p. 33.

'° Additional copies of this report are available from the National Academy

Press, 2101 Constitution Avenue, NW., Washington, DC 20410.

? Sonoma Technology, Inc., "Navajo Generating Station Visibility Study,”

Draft Number 2, April 16, 1991.

'2 U.S. EPA, "Draft Report on Best Available Retrofit Technology (BART)

Analysis for the Navajo Generating Station in Page, Arizona,” January 1990. A

copy of this document has been placed in Docket A-89-02A.

14

quality environmental impacts, any existing pollution control

technology in use at the facility, the remaining useful life of the

source, and the degree of improvement in visibility anticipated to

result from application of controls.

Considering the requirements found in the BART Guidelines,

EPA divided its analysis into two major parts. In the first part,

EPA identified the SO, emission controls that are readily available

to the source, the costs of such controls, and other impacts of

installing and operating the controls. In the second part of the

analysis, EPA attempted to define the source-impairment

relationship which was then used to predict the improvements in

visual air quality that can reasonably be expected to occur as a

result of installing and operating the controls defined in the control

technology analysis. Because EPA was faced with some

uncertainties in each of the parts of the analysis, EPA bounded its

results by giving low and high estimates. A detailed discussion of

this analysis for NGS was provided in EPA’s February 8, 1991

proposal.

To estimate the cost of controlling the SO, emissions at NGS,

EPA used the Integrated Air Pollution Control System (IAPCS) cost

model to predict the capital and operating cost estimates for 0.10,

0.30, and 0.50 Ib/MMBtu control levels.’ The results from the

IAPCS indicated that wet flue gas desulfurization (FGD) achieving

a 0.30 Ib/MMBtu emission limit for all three units, the control level

that was proposed by EPA, was estimated to have total capital cost

requirements of between $245.9 million and $402 million with total

levelized annual costs (including amortized capital, interest,

operating, and maintenance costs) estimated to be between $91.9

million and $128.3 million.'*

In its February 8, 1991 proposal, EPA also estimated the

potential costs to residential electricity customers serviced by the

si Integrated Air Pollution Control System Costing Program, Version 3.0,

copyright PEI and Associates, Inc., 1989.

14 All dollar amounts were measured in 1988 dollars.

Camid SUP aay

;

}

;

j

%

:

15

NGS participants as well as the potential costs to Central Arizona

Project (CAP) customers.

One of the more complicated tasks in the BART analysis was

defining the relationship between the SO, emissions at NGS and the

visibility impairing sulfate in the GCNP. This relationship had to

be addressed in order to estimate the degree of improvement in

visibility that could be anticipated to result from the use of the

alternative SO, control systems. Because of the complex terrain in

and around the GCNP, EPA defined the source/impairment

relationship using the ratio of SO, emissions at NGS to Sulfate in the

GCNP attributable to NGS as found in the final WHITEX report.

Using this ratio, EPA then applied a linear rollback model.

However, in light of the uncertainties Surrounding these analyses

EPA considered a broad range of source-impairment relationships.

In addition, EPA developed and applied 4 nonlinearity factor to the

rollback model to estimate the improvements in visibility that would

result from reductions in SO.,,.

Because no single standard method exists for measuring

improvements in visibility, the EPA employed two of the most

widely used techniques to define the improvements in visibility

expected to occur in the GCNP. One of the methods used to

estimate the visibility improvements in the GCNP was analyzing the

changes in contrast expected to occur as a result of reducing SO,

emissions at NGS. The EPA stated in the proposal that this method

is particularly relevant because human observers use contrast to

make judgments of how atmospheric clarity changes with pollutant

concentration. The contrast is defined as the percent difference

between the brightness of a scenic element end its background. The

second technique was to analyze the seasonal average visual range

change (see 56 FR 5182 Table 1).

Although not specifically addressed in the BART analysis, the

proposal noted that EPA suspected that NGS emissions may

contribute to visibility impairment in the GCNP during other

seasons of the year and in other Class I areas in the region. Thus,

16

EPA sought comment on any other potential visibility impairments

caused by NGS.

The EPA also noted in the proposal that it was not required as

a part of its analysis to estimate monetary benefits associated with

improving visibility in the GCNP. However EPA evaluated the

monetary benefits in developing a part of the preliminary Regulatory

Impact Analysis (RIA) (as required by Executive Order 12291) for

the proposed rule.

4. Control of Particulate Matter and Nitrogen Oxides

In the February 8, 1991 proposal, EPA noted that two other

pollutants emitted by NGS, NO, and PM, are known to contribute

to visibility impairments in some circumstances.’* The current

emissions of NO, and PM from NGS were not identified as

significantly contributing to the visibility impairment in the Grand

Canyon. However, EPA expressed concern based upon its technical

judgment regarding the behavior of the pollutants as potential

contributors to visibility impairment as a general matter, upon the

amount of the pollutants emitted by NGS, upon the proximity of

NGS to the GCNP, and upon the characteristics of the NGS plume,

as shown by WHITEX and other studies, that increased emissions

of these pollutants could result in impairment in the GCNP. The

EPA noted that although the NGS is not subject to any NO,

emission limitations, SRP has stated that NGS currently emits NO,

at a rate of 0.4 to 0.5 Ib/MMBtu. Therefore, EPA proposed an

NO, emission limit consistent with NGS’ existing emission rate, as

well as the rate at other similar plants, of 0.5 Ib/MMBtu.

The EPA noted that PM emissions for the NGS are currently

limited in two ways. The Arizona SIP has both an emission rate

limitation of 0.06 lb/MMBtu and an opacity limit of 40 percent for

NGS. The SRP has stated that NGS currently emits PM at a rate

of 0.03 Ilb/MMBtu with an opacity of between 10 and 15 percent.

Thus, the opacity limit of 40 percent is well above the NGS existing

'S See, e.g., BART Guidelines at p. 12-13.

a a oe

17

emissions. Therefore, EPA proposed a 20 percent Opacity limitation

consistent with current NGS emissions in order to preclude any

visibility deterioration that would be caused by an increase in PM

emissions.

The EPA requested comments on whether NO, and PM

emission limits would be appropriate for NGS.

5. Compliance Schedule

The EPA proposed to require that the emission limitation be

achieved in three phases. Specifically, EPA proposed that the final

emission limitation would need to be met on one unit by January 1,

1995; on two units by January 1, 1997; and on all three units by

January 1, 1999. The EPA solicited comments on whether the

phased-in compliance schedule proposed was appropriate as a matter

of law or policy, and, as noted, requested comment on the January

1991 SRP proposal that a final emission limitation not become

effective before January 1, 2000.

6. Use of Seasonal Controls

Part of the SRP proposal called for the use of seasonal controls.

Therefore, EPA solicited comment on the use of seasonal controls.

In the February 8, 1991 notice, EPA noted that such controls would

have to be both technically and legally justified before they could be

used.

7. Other Issues

In the February 8, 1991 proposal, EPA also specifically

Solicited comments on the following:

a. Use of continuous emission monitoring systems for

compliance determinations.

b. Plant-wide averaging of the emission limitations.

c. Other available technologies which can reduce SO,

emissions at NGS to achieve the proposed or alternative limits and

which may require less capital investment and/or Operating expenses

than those technologies evaluated in the BART analysis.

In addition to soliciting written comments during the comment

period, EPA held a public hearing on March 18 and 19, 1991 in

18

Phoenix, Arizona. A copy of the transcript of the hearing is in

Docket A-89-02A. As a result of the comment period and public

hearing, EPA received over 400 comments on its proposal. All the

comments have been placed in Docket A-89-02A.

Because of the differences in technical opinions expressed and

data analyses submitted to EPA during the comment period, EPA

sponsored a technical review meeting on April 25 and 26, 1991.

The EPA invited the NGS participants and the environmental groups

who submitted technical information to EPA to send technical

representatives to the meeting to discuss the interpretation of the

available data. The meeting was open to the public, and a summary

of that meeting is included in Docket A-89-02A.

E. New information After the Close of the Initial Comment Period

After the comment period closed on April 19, 1991, at the

recommendation of EPA, representatives of SRP, Grand Canyon

Trust (GCT), and EDF met to discuss alternative approaches to

EPA’s February 8, 1991 proposal. In August 1991, the outside

parties reached agreement and together recommended that EPA

adopt an alternative control for NGS consisting of a 0.10 Ib/MMBtu

SO, emission limitation (approximate to a 90 percent control level)

based on a rolling annual average and phased in by unit in

November 1997, November 1998, and August 1999. In addition,

under the agreement, NGS would shift its maintenance schedule

such that 6 unit-weeks of planned maintenance would occur between

November | and March 15 each year. Under specific conditions,

the Administrator of EPA may allow NGS to shift the maintenance

schedule outside of that period or not to conduct scheduled

maintenance in a given year.

Representatives of EPA participated in many of the meetings

with the parties and provided technical assistance. Representatives

of the State of Arizona also attended several of the meetings and

provided additional technical support. New technical materials and

cost information, including adjustments of the potential control

costs, were exchanged between the parties and EPA. Summaries of

the meetings and significant conversations in which EPA was

19

involved and copies of the new material and information which were

submitted to or developed by EPA have been included in Docket A-

89-02A.

The outside parties memorialized their agreement in an MOU

which they submitted to EPA along with recommended regulatory

requirements for EPA’s final rulemaking action. The SRP estimated

(in 1992 dollars) a capital cost of $430 million and a total levelized

annual cost of $89.6 million for the parties’ recommended

alternative. In comparison, SRP estimated a capital cost of $510

million and a total levelized annual cost of $106 million to meet the

alternative EPA proposed in February 1991.

Noting that the alterative incorporated in the MOU would

provide more visibility protection for the GCNP at a lower cost for

NGS and its customers, EPA, on August 8, 1991, reopened the

comment period on its February 8, 1991 proposed action (56 FR

38399, August 13, 1991). In the notice, EPA stated that it was

giving serious consideration to the control option recommended by

the outside parties and incorporated in an appendix to the notice a

memorandum from EPA’s Office of General Counsel concluding

that the outside parties’ legal rationale in support of their

recommended control option was meritorious.

The EPA reopened the commend period until September 9,

1991. Where addresses were available, commenters on the

February 8, 1991 proposal were notified by mail of the reopened

comment period in order to facilitate their ability to comment.

The EPA received 21 comments during the supplemental

comment period. Eighteen of the commenters supported the

alternative incorporated in the MOU, and three opposed it. The

SRP, GCT, EDF, the Wilderness Society, the National Wildlife

Federation, and the Sierra Club conditionally withdrew their prior

comments which were inconsistent with the MOU and associated

documents.

II. Today’s Actions

In today’s action EPA is promulgating its final determination on

the attribution to NGS of certain visibility impairment episodes in

20

GCNP and is promulgating Federal revisions to the visibility

implementation plan for Arizona to address those impairment

episodes. As discussed below, EPA has concluded that certain

visibility impairment episodes in GCNP are traceable to NGS and

that NGS is a dominant contributor to certain visibility impairment

episodes. The EPA finds today that a 0.10 lb/MMBtu SO, emission

limitation (approximately a 90 percent emission reduction) based on

a rolling annual average and phased in by unit in November 1997,

November 1998 and August 1999 in addition to scheduled winter

maintenance at NGS affords greater visibility improvements than the

alternative advanced in EPA’s February 1991 proposal. In

particular, it will reduce by two-thirds the amount of pollution

allowed under the proposed rule. In addition, EPA finds that the

final rule will be significantly less costly than the proposal.

Consequently, EPA further concludes that today’s final action will

provide a greater degree of “reasonable progress" toward the

national goal of remedying such impairment than would be provided

by the February 1991 proposal (see section 169A(b)(2) of the Act,

42 U.S.C. 7491(6)(2)).

In EPA’s August 1991 supplemental notice requesting comment

on the rulemaking alternative largely adopted in final form today,

the Agency noted that SRP, GCT, and EDF had recommended that

the legal rationale in support of this alternative be the requirement

in section 169A(b)(2) of the Act that implementation plan revisions

addressing visibility impairment achieve “reasonable progress"

toward the national visibility goal. The supplemental notice also

indicated that EPA’s Office of General Counsel had reviewed the

matter and concluded, subject to any significant points that may be

raised in the reopened comment period, that EPA could rely on the

reasonable progress provisions as the basis for this alternative. No

significant adverse comments addressing the legal basis were

received. Accordingly, EPA is adopting this rationale as discussed

in the supplemental notice.

Today’s action is based upon the material in the docket

including EPA’s review and consideration of all comments received

ee a Ty

ee ee

21

during the comment periods and at the public hearing. The

agreement between GCT, EDF, and SRP and _ associated

recommendations to EPA, largely adopted in final form today, also

are in the docket. The EPA has responded to all of the significant

comments received. Some responses are provided later in today’s

notice. Additionally, EPA has prepared a document accompanying

today’s action. "Response to Public Comments: Proposed

Revisions to Arizona Visibility FIP for Navajo Generating Station,"

which responds to comments. This document has been placed in

Docket A-89-02A.

A. Impairment Traceable to NGS

The EPA regulations promulgated on December 2, 1980

({r]equire the control of impairment that can be traced to a single

existing stationary facility or small group of existing Stationary

facilities . . ." (45 FR 80085). Accordingly, EPA’s regulations

currently cover existing sources where the impairment is traceable

or “reasonably attributable” to that source by visual observation or

other techniques deemed appropriate by the State (Phase I

impairments). The EPA, acting for the State under section 110(c)

of the Act, deems the techniques in the WHITEX report, the

NGSVS report, and other data and analyses in the docket as

appropriate techniques for determining reasonable attribution in this

case for the following reasons.

1. The unique tracers, deuterated methane used in the

WHITEX study and perfluorocarbons used in the NGSVS, injected

into the emissions of NGS were observed in substantial quantities at

Hopi Point during periods of episodic visibility impairment.

2. The NGS is the predominate source of SO, in the region.

3. SO, released by a power plant converts into sulfates through

chemical transformation in the atmosphere.

4. Meteorological data show that the NGS plume easily can,

and frequently does, travel to GCNP.

5. The data and analyses in the docket show that sulfates are

the major contributor to visibility impairment in GCNP.

22

The EPA recognize that NGS is not the only source of visibility

impairment at GCNP. Under the applicable statutory provisions and

regulations, however, this is not determinative’® (see generally

section 169A of the Act and 40 CFR 51.300 307). Moreover, even

though WHITEX and NGSVS contain some scientific uncertainty,

they are technically adequate for the regulatory purposes at hand, as

they both showed that episodes of visibility impairment in the

GCNP can be reasonably attributed to NGS. Accordingly, EPA

concludes today that certain visibility impairment episodes of the

GCNP are traceable to NGS, and NGS is a dominant source of

those impairments. Notwithstanding that other sources also may

contribute to those impairment episodes at GCNP. EPA concludes

that the addition of emissions controls at NGS alone will result in

a Significant improvement in visibility at GCNP and will make

“reasonable progress" toward meeting the national visibility goal.

B. Control Technology

As part of the BART analysis '’ for the February 1991

proposal, EPA reviewed the available control technologies as well

For example, under section 169A(b)(2) of the Act, EPA is authorized to

require visibility implementation plans containing “emission limits, schedules of

compliance and other measures” necessary to make reasonable progress toward

meeting the national visibility goal (see also 40 CFR 51.302(c)). The national

visibility goal, in turn, calls for the remedying of “any” manmade visibility

impairment in Class I areas (see section 169A(a)(1) and 40 CFR $1.300(a)). As

noted, EPA’s regulations implementing section 169A address visibility impairment

that is reasonably attributable to a single source or small group of sources and

deferred action on complex problems such as regional haze (see generally, 45 FR

80084 (December 2, 1980)). Accordingly, as provided in the 1980 regulations,

EPA may remedy through emission limits, schedules of compliance, or other

measures “any” visibility impairment that is reasonably attributable to an existing

stationary source or small group of sources (see also section 169A(c) of the Act

and so CFR 51,303 (authorizing an exemption from control requirements only

where a plant does not “by itself or in combination with other sources” omit

pollution "which may reasonably be anticipated to cause or contribute to significant

impairment of visibility" in any Class I area and affirms by negative inference,

EPA’s broad rulemaking authority)).

U.S. EPA "Draft Report on Best Available Retrofit Technology (BART)

Analysis for the Navajo Generating Station in Page, Arizona,” January 1990.

iinet

23

as the environmental impact of the use of such technologies and the

estimated costs of installing and operating such equipment. In

today’s action, EPA is not requiring the use of any particular

control technology. Rather, EPA is establishing an emission

limitation and NGS has the discretion to select and install the type

of control system which best meets its needs. The available

information in the docket indicates that a wet FGD system will be

the economically proven control technology for meeting the 0.10

lb/MMBtu emission limitation at NGS.

As a result of new information provided by SRP, EPA has

revised and updated its cost estimates for the installation and

operation of a wet FGD system. The revised estimated Capital cost

and total levelized annual cost to meet the emission limitation EPA

proposed in February 1991, in 1992 dollars, are $510 million and

$106 million, respectively. The estimated capital cost and the total

levelized annual cost, in 1992 dollars, to meet the emission

limitation promulgated today, are $430 million and $89.6 million,

respectively. Although not considered in these cost estimates, EPA

expects that NGS will be able to recoup a portion of the control

costs by the sale of marketable allowances which it will receive as

part of the allowance trading program that EPA is establishing under

Title IV of the Act.

Based on the information submitted during the comment period

by SRP and Nevada Power Company (NPC) and the above cost

figures, EPA has estimated that the average SRP and NPC

customers electric bills will increase approximately $1.72 and $1.57

per month, respectively. The customers of the other utilities will

have smaller increases in their electric bills. | Commenters

representing CAP customers submitted information showing that

water costs would increase between $4.10 and $4.50 per acre foot

based on EPA’s February 1991 proposal. Since the cost figures for

the final action are lower than the figures used for these estimates,

the increase in water costs for the CAP customers is expected to be

less than $4.10 to $4.50 per acre foot.

24

C. Control Strategy

The EPA has adopted the combination of a higher level in

control; compliance phased in by unit in 1997, 1998 and 1999; an

annual averaging period; and scheduled winter maintenance as a

control strategy for NGS -- the alternative recommended to the EPA

by GCT, EDF, and SRP and identified in EPA’s August 1991

supplemental notice - because it better addresses the visibility

effects of concern at a lower cost than the proposal in EPA’s

February 1991 notice. As highlighted here and elsewhere, the

control strategy for NGS must be viewed as a whole and not as

severable parts. The EPA has carefully weighed the relevant

Statutory and regulatory considerations and concluded that, taken in

its entirety, today’s control strategy for NGS will provide a greater

degree of reasonable progress toward the national goal at a lower

cost than EPA’s February proposal.

Essentially, the SO, emissions from NGS can contribute to two

types of visibility impairment. The major impact of NGS is its

dominant single source impact during certain episodic visibility

impairment events, usually during the winter. This impact has been

observed in winter studies (WHITEX and NGSVS) as well as

intense aerosol monitoring within the canyon since 1988.'* The

other concern addressed by the control strategy in today’s action is

the less intense impairments which occur at the GCNP during

seasons other than winter, and occur at other nearby mandatory

Class I Federal areas.’®

The lower level of NGS SO, emissions required by the 0.10

lb/MMBtu emission limitation adopted today will afford GCNP

greater protection from episodic events than an emission limitation

’See Docket A-89-02A, specifically items IV-D-164 and IV-D-375.

See, e.g., Latimer, Douglas, "Haze Impacts on the Golden Circle of

National Parks of Sulfur Dioxide Emission from Navajo Generating Station: Haze

Puff Model Calculations for 1988-90."

25

which allows significantly more SO, emissions but is averaged over

a shorter period, such as the 0.30 Ib/MMBtu emission limitation

proposed by EPA in its February 1991 notice. The probability of

high NGS contribution to severe episodic sulfate impairment events

is limited by the known design reliability of current technologies

which will achieve the required 0.10 Ib/MMBtu emission limitation

as well as the scheduled winter maintenance requirement. Thus,

EPA believes that despite the use of an annual averaging time, the

Significant increase in control level, expected reliability of the

control technology, and winter maintenance, taken together will, for

the most part, prevent continuation of severe winter episodes caused

by NGS. At the same time, the annual average lowers the costs of

compliance.

With respect to the long-term potential visibility impacts on the

Golden Circle area (the other Class I areas in the Colorado Plateau),

there is little doubt that the lower emission rate over a year’s

average will reduce NGS’ contribution to sulfate formation and thus

to sulfate-caused impairment more than the control level proposed

by EPA in February 1991. Given NGS’ central location in relation

to many Class I areas, no matter what meteorological events occur

over a multi-year period, it is likely that the sulfates originating

from NGS and transported to Surrounding Class I areas will be

reduced in proportion to the difference in emission limitations, i.e.,

to one-third of the levels in the February 1991 proposal.

Further, during the comment period on the February 1991

proposal, EPA received technical information Suggesting that NGS

contributes to visibility impairment at GCNP in non-winter

seasons.” For the reasons just discussed, the significant lower

emission rate required today will, over the long-term, curb NGS’

contribution to any visibility impairment in the GCNP during

seasons other than winter to a greater degree than EPA’s proposed

option.

See Docket A-89-02A, item IV-D-171, and season control discussion below.

26

Accordingly, the control scheme adopted today for NGS is

appropriate for addressing the winter episodic visibility impairment

observed at the GCNP and will also reduce any long-term, long-

range impairment that NGS emissions may contribute to during

other seasons at GCNP and at other nearby Class I areas.”

D. Emission Limitation

After reviewing the option recommended by SRP, EDF, and

GCT in their MOU, EPA has determined that such a control

Strategy will provide a substantially greater degree of emissions

reduction, and a correspondingly greater degree of visibility

improvement at a substantially lower cost than the alternative

proposed in February 1991.” For these reasons, EPA also

concludes that this option will make additional reasonable progress

towards attainment of the national visibility goal in comparison to

the February 1991 proposal. Thus, in today’s action EPA is

promulgating an SO, emission limitation of 42 ng/J (0.10

lb/MMBtu) heat input to be phased in by unit in November 1997,

November 1998, and August 1999, with compliance determined on

a rolling annual average basis.

21 As noted, section 169A of the Act establishes a national visibility goal,

not an air quality standard or a specific emission standard. While the objective is

to attain the goal, EPA’s only mandate from Congress is to make reasonable

progress toward the goal. For these reasons, EPA has more flexibility in selecting

a centrol strategy to meet the visibility reasonable progress requirements than it

does when a health-based air quality standard or specific emission standard must

be met. Thus, in determining reasonable progress, EPA, by law, must consider

the costs of compliance, the time necessary for compliance, and the energy and

non-air quality environmental impacts of compliance (see section 169A(g)(1) of the

Act). Consideration of the pertinent statutory scheme and these specific factors has

informed several aspects of today’s action. Thus, for the policy and legal reasons

stated, an annual rolling average emission limitation is acceptable for this case.

Moreover, while this emission limitation is not expected to interfere with

attainment of any other requirement of the Act, its adoption does not relieve NGS

of any responsibility for meeting any air quality standard, emission standard, or

other requirement of the Act.

2 In the August 1991 supplemental notice, EPA published and requested

comment on the parties’ August 8, 1991 MOU and associated documents. During

the public comment period, the parties further clarified their agreement and

submitted a revised version of their MOU and recommended regulatory

requirements to EPA, both documents dated August 22, 1991.

a

27

Under today’s final action, compliance with the annual rolling

average will be determined by computing a weighted plant-wide

average of the SO, emission rate based on the daily SO, emission

rate and the electric energy generated for the previous 365 "boiler

operating days" for each unit.~ The EPA was concerned that the

plant-wide average reflects the actual release of emissions at the

plant. Therefore, after discussion with representatives of SRP,

GCT, and EDF, EPA settled on a methodology that appropriately

weights the emissions from each unit before calculation of the plant-

wide average. The EPA has used the daily electric energy

generated as a weighting factor because that information is readily

available for each unit and in this case is proportional to the heat

input to the boilers. Specifically, compliance will be determined as

follows:

1. For each unit that has accumulated at least 365 boiler

operating days since the passage of the starting date applicable to it,

the plant must measure and record the SO, emission rate and the

electric energy generated on each boiler operating day. The SO,

emission rate will be computed using the date from the required

continuous emission monitoring system for the unit and using

method 19, appendix A, 40 CFR part 60. The electric energy

generated, in megawatt-hours, will be recorded from the megawatt-

hour meter for the unit.

2. For each unit, the previous 365 boiler Operating days will

be identified.

3. For each such day, the product of the SO, emission rate

and the electric energy generated wil! be computed.

3 A “boiler operating day” is specific for each steam-generating unit at

NGS and is defined as a 24-hour calendar day (the period of time between 12:01

a.m. and 12:00 midnight in Page, Arizona) during which coal is combusted at that

NGS unit for the entire 24 hours.” This definition is consistent with the definition

for “boiler operating day” in 40 CFR 60.41a: "a 24-hour period during which

fossil fuel is combusted in a steam-generating unit for the entire 24 hours.” The

EPA notes that the set of previous 365 boiler operating days for each unit will be

unique to that unit and thus the same set of calendar days probably will not be used

for all three units in determining compliance.

28

4. The 365 products for each unit will be added.

5. The electric energy generated for the 365 boiler operating

days for each unit will be added.

6. The sums of the product of the SO, emission rate and the

eiectric energy generated will be divided by the sum of the electric

energy generated to produce a plant-wide weighted annual average

SO, emission rate for comparison with the emission limitation to

determine compliance.

Recording and computation of the daily electric energy

generated and SO, emission rates will commence on:

1. November 19, 1997 for the first unit.

2. November 19, 1998 for the second unit.

3. August 19, 1999 for the third unit.

The NGS has the discretion to determine which of its three units

will be controlled first and second.

In the discussion below, EPA has recognized that certain

difficulties encountered during the first year of operation of a

control system or the catastrophic failure of a control system may,

based on EPA’s judgment, warrant a limited exclusion from

compliance with the emission limitation. However, as discussion

below indicates, the control system, as a general matter, must be

optimally operated consistent with good engineering practices to

keep emissions at, or below, the emission limitation.

For each unit, in determining compliance with the annual

average emission limitation during the first year. of operation of the

control equipment installed to comply with this emission limitation,

periods during which one of the followin; conditions are met will

be excluded:

1. Equipment or systems do not meet designer’s or

manufacturer’s performance expectations.

2. Field installation including engineering or construction

precludes equipment or systems from performing as designed.

The periods during the first year to be excluded will be

determined by the Administrator based on periodic reports of

eee

29

compliance with this emission limitation which must identify the

times proposed for exclusion and provide the reasons for the

exclusion, including the reasons for the Outage of the control

system. The report also must describe the actions taken to avoid the

outage, to minimize its duration, and to reduce SO, emissions at the

plant to the extent practicable while the control System was not fully

operational. Whenever the time to be excluded exceeds a

cumulative total of 30 days for any control system, the NGS owner

or operator must file a report within 15 days addressing the history

of, and prognosis for, the performance of the control equipment.

In addition to the foregoing, the Administrator will exclude

from the compliance determination for a unit any periods of

emissions from a unit for which the Administrator finds that the

control equipment is out of service because of catastrophic failure

of any control system which occurred for reasons beyond the control

of the NGS participants and operators and could not have been

prevented by good engineering practice. The Administrator will not

exclude the period if the equipment failure was a consequence of a

lack of appropriate maintenance; of intentional or negligent conduct

or omissions of the NGS participants or operators or the control

System design, construction, or operating contractors.

The final regulation requires that all equipment needed to

comply with this regulation to be optimally operated consistent with

good engineering practice to keep emissions at or below the

emission limitation. The regulation also requires that following any

control system outages, the system must be returned to full

operation as expeditiously as practicable.

The final regulation provides that continuous emission monitors

must be installed to determine compliance with the emission

limitation. This equipment must meet the specifications listed in

appendix B of 40 CFR part 60, the quality assurance procedures in

appendix F of 40 CFR part 60, and the requirements for estimating

emission rate in ng/J or (Ib/MMBtv) set out in method 19, appendix

30

A, 40 CFR part 60.% The NGS is required to report emissions

and maintain records in accordance with the procedures in 40 CFR

60.7. In addition, EPA is establishing special notification

procedures when an outage of a control system occurs. The NGS

must notify the Administrator by telephone or by writing within one

business day of any outage of the control system needed for

compliance with the emission limitation and must submit a follow-

up written report within 30 days of the repairs stating how the

repairs were accomplished and justifying the amount of time taken

for the repairs.

In the February 1991 notice, EPA proposed new NO, and

opacity limits on NGS emissions consistent with current emission

levels. These were proposed due to a concern that certain SO,

emission control equipment would cause increases in NO, and PM

(which are limited by opacity limitations) emissions at NGS which

could affect visibility in GCNP. However, EPA does not have any

evidence at this time that NO, or PM emissions from NGS

contribute significantly to visibility impairment in the GCNP.

Further, the technology agreed to in the MOU wil! not increase NO,

or PM emissions from NGS. Moreover, the docket does not

contain any evidence that NO, or PM emissions from NGS would

cause any such impairments after compliance with the final

limitations promulgated today. Thus, EPA finds no need to include

additional emission limitations for these pollutants.

” For certain requirements in today’s action (e.g., monitoring, reporting,

recordkeeping, etc.), EPA has relied on its regulations governing standards for

performance for new stationary sources. These regulations have been relied on

because they contain standardized procedures and requirements which apply to

coal-fired steam-generating units. In addition, EPA has provided that the NGS

owners or operators must comply with the requirements relied on in 40 CFR part

60 as in effect today. This approach was agreed to and recommended to EPA by

SRP, GCT, and EDF in their August 22, 1991 MOU and associated documents.

At this time, EPA believes that reliance on the regulations as they currently exist

is reasonable in that they appear to provide adequate technical methods and

procedures for determining and monitoring compliance with the SO2* emission

limitation established today.

31

E. Source/Impairment Relationship

In the February 1991 proposal, EPA estimated seasonal visual

range changes which were based on a simple rollback model applied

to average NGS contribution during the entire WHITEX study

period. Based on this model, a 0.10 lb/MMBtu SO, emission

limitation on NGS should result in an approximate 14 percent

improvement in seasonal average standard visual range at Hopi

Point.

During the comment period on EPA’s February 1991 proposal,

SRP submitted the results and analyses of the NGSVS which

included estimates of NGS’ contribution to visibility impairments in

GCNP and specifically focused on Hopi Point. The NGSVS data

indicate that controlling SO, emissions from NGS would result in at

most 2 percent improvement in the seasonal average standard visual

range.*

Another study™ submitted during the comment period, which

was based on modelling of NGS’ emissions over a long time period,

indicated average wintertime improvements in the range of 4 to 8

percent if NGS emissions are reduced. This study is not based on

any specific monitoring period nor was it based on an EPA-

approved model.

The two intensive studies, WHITEX and NGSVS, included

detailed aerosol, wind flow, and tracer analyses. They represent

snapshots of NGS’ impacts for their respective Study periods. Based

solely on wind-flow analysis during these periods, these studies

varied significantly in concluding how often NGS emissions would

be transported directly to the Hopi Point, the common GCNP

monitoring site for both studies.

The major change in the seasonal average visibility results from

more dramatic improvements expected to occur as a result of

aad Average of the total light scattering to that part apportioned to NGS

over the entire study period as found in appendix C of the NGSVS, Final Report

Draft Number 2. Sonoma Technology, Inc., April 10, 1991.

| © “See Docket A-89-02A, item IV-D-171.

32

reducing emissions from NGS during certain meteoroiogical

conditions. Recent measurements of peak winter sulfate levels in

the canyon approach four times the average winter levels. For

certain periods analyzed in WHITEX and in NGSVS, NGS was

shown to be the dominant contributor (more than 50 percent) to

sulfates measured at Hopi Point. If certain humidity and wind

conditions occur during the period when very high sulfate episodes

are formed, reducing NGS emissions could result in increases in

standard visual range up to 300 percent. Given the variability in

meteorological conditions throughout the winter and from one

winter to the next, it is impossible to exactly quantify the peak

episodic expected improvement for any given period.

Photographic data taken during WHITEX< indicated that airflow

below the rim of the canyon could result in higher visibility

impairment due to trapping of pollution. As a result, a new aerosol

monitoring site at Indian Gardens (4,000 feet below the rim of the

canyon) was established after the WHITEX study. This site has

provided information since 1988 which confirms that the transport

and conversion processes below the rim of the canyon are

sometimes decoupled from the processes above the rim. Both

NGSVS and the NPS report on WHITEX used the Hopi Point site,

which is above the rim of the Canyon, as the basis for determining

the NGS impact on GCNP. Data submitted during the comment

period show that visibility below the rim is impaired more often and

to a greater degree than is the visibility above the rim at Hopi

Point,”’

Taking into account the conclusions from both studies as well

as other monitoring information at GCNP from long-term

monitoring, EPA expects that reducing SO, emissions from NGS to

0.10 Ib/MMBtu should improve the winter seasonal average

visibility above the rim of the canyon approximately 7 percent

principally due to improvements during episodes. Analyses of peak

in-canyon sulfate levels during winter inversion episodes also lead

27“ See Docket A-89-02A, items IV-D-164, IV-D-346, and IV-G-3.

33

EPA to conclude that reductions in NGS SO, emissions may well

contribute to greater episodic visibility improvement resuiting in, as

discussed above, more than a7 percent improvement in the seasonal

average visibility below the rim of the canyon. However, EPA did

not quantify the expected visibility improvement below the rim of

the canyon due to the limited amount of data and understanding of

the air transport mechanisms below the rim of the canyon.

F. Schedule of Compliance

As noted above, compliance with the emission limitation will

be phased-in by unit on the following schedule:

1. One unit by November 19, 1997.

2. Two units by November 19, 1998.

3. All units by August 19, 1999.

The phasing in of the emission limitation will allow NGS to initiate

operation of the control systems one at a time.

To ensure that NGS complies with the emission limitation by

the dates specified, EPA has established a schedule of compliance

containing interim deadlines as follow.

Date of binding contract for architectural/ June 1992

Engineering firm to design and procure the

control system needed for compliance.

Start of onsite construction of the control Jan. 1995

system for the first unit.

Initiation of startup testing:

ks ey a May 1997

Se re May 1998

wes Feb. 1999

The Administrator may extend the interim deadlines if NGS can

demonstrate that compliance with the final deadlines for compliance,

stated above, will not be affected.

G. Maintenance Schedule

The final regulation provides that by March 16, 1993, and

every March 16th thereafter, the NGS owner or operator will

34

prepare a long-term maintenance plan for NGS that maximizes

winter down-time while accommodating the maintenance

requirements for the other generating facilities on the NGS grid.

The plan will cover the period from March 16 to March 15 of the

next year and must provide at least a full 6 unit-weeks of

maintenance for NGS in the November | to March 15 period,

except as provided below, to further reduce SO, emissions during

the winter. The plan will be developed as to be consistent with the

criteria established by the Western States Coordinating Council of

the North American Electric Reliability Council to ensure adequate

reserve margin. The full 6 unit-weeks of winter maintenance need

not occur if any of the following circumstances arise.

1. There is no need for 6 unit-weeks of scheduled periodic

maintenance in the March 16-March 15 year covered by the plan.

2. The reserve margin on any electrical system served by

NGS would fall to an inadequate level, as defined by the criteria

referenced above. In such case the scheduled maintenance may be

moved out of the November 1 to March 15 period.

3. The cost of compliance with this provision would be

excessive. Costs of compliance would be considered excessive

when the economic savings to the participants of moving NGS’

maintenance out of the November | to March 15 period exceeds

$50,000 per unit-day of maintenance moved.

4. A major forced outage at a unit occurs outside the winter

months, and necessary periodic maintenance occurs during the

period of forced outage.

The NGS owner or operator must demonstrate to the

Satisfaction of the Administrator that one or more of the events

listed above render unnecessary or unreasonable a full 6 unit-weeks

of scheduled maintenance during the specified winter period.

Where 6 unit-weeks of scheduled maintenance is unnecessary or

unreasonable, the NGS owner or operator must nevertheless make

best efforts to conduct as much scheduled maintenance as practicable

during the winter period. If NGS does not conduct its scheduled

7 >

|

a eenesiainn eaten

35

winter maintenance (up to 6 unit-weeks), it must report to EPA why

it did not do so.

When maintenance is being conducted on a unit, the unit will

not be in operation and therefore not emitting SO,. Thus, the

shifting of NGS maintenance to the wintertime period will further

reduce NGS SO, emissions during the period when NGS is

suspected to have its greatest impact on the GCNP. Further,

because the scheduled winter maintenance requirement will be

implemented in the winter of 1993-1994, it will have the effect of

making a degree of reasonable progress toward the national

visibility goal before final compliance with the emission limitation

is required. Because NGS must, for normal operation, conduct this

maintenance, rescheduling it in the wintertime period should have

little economic impact on NGS.

H. Air Quality Review

The installation and operation of a wet FGD control system at

NGS (the proven available technology that EPA anticipates will be

used) will lower the exhaust gas temperature from the stacks and

could reduce the plume rise above the NGS stacks. The EPA and

the Arizona Department of Environmental Quality reviewed the

need to reheat the exhaust gases to ensure appropriate plume rise

with the lower gas exit temperatures associated with meeting a 0.10

lb/MMBtu emission limitation using wet FGD and found that

reheating of the exhaust gas will not be necessary to prevent a

violation of any national ambient air quality standards. If the

exhaust gas is not reheated, the control cost will be reduced

significantly.

If NGS were to select a wet FGD SO, control system, the

water content of the plume would increase. Under certain

conditions, a white steam plume will occur as the water vapors

condenses before it dissipates. This steam plume may be visible

near the stack. However, it will not cause any visibility impairment

in GCNP or surrounding mandatory Class I Federal areas and will

not interfere with the attainment of any national ambient air quality

36

standards. Such steam plumes are not subject to any State or

Federal visual emission standards or regulations.

III. Response to Comments

The September 5, 1989, February 8, 1991, and August 13,

1991 notices requested comment on a variety of issues relating to

attribution of impairment, cost of controls, and the benefit expected

from the controls. In addition, EPA held a public hearing on March

18 and 19, 1991 during which EPA received comments on its

February 1991 proposal.

Under section 307(d) of the Act, today’s action must be

accompanied by a response to each of the significant comments,

criticisms, and new data submitted in written or oral presentations

during the comment period. The EPA has carried out this duty.

Moreover, modifications to the proposed rule have been made in

response to public comments. The EPA’s final control strategy for

NGS largely embodies pubic comments received from

representatives of GCT, EDF, and SRP which, in turn, were subject

to public comment (see, e.g., 56 FR 38399 (August 13, 1991)).

Further, in its February 1991 proposal, EPA requested comment on

several control options including, but not limited to, the following

rule elements and specific regulatory alternatives:

1. Emission limitations, including limitations ranging between

0.50 and 0.10 Ib/MMBtu.

2. Averaging times, including a 3-hour, 30-day, and annual

averaging period.

3. Implementation schedules, including one providing for

plant-wide compliance by the year 2000. As noted below, EPA

received many comments in response to the February 1991 proposal

supporting, in part, the control elements adopted by EPA, in total,

today. For example, many citizens and representatives of

environmental organizations submitted comments supporting a 0.10

lb/MMBtu emission limitation, the emission limitation adopted

today. Many representatives of industry and business submitted

comments supporting an annual averaging period and a long

compliance schedule.

37

During the initial public comment period, EPA received over

400 comments including the public hearing transcript. A summary

of the significant comments, criticisms, and new data submitted in

written or verbal presentations during the public comment period

and EPA’s responses are provided below and in supporting

documents which have been placed in Docket A-89-02A.”, ”

During the supplemental comment period, EPA received 21

additional comments. Two commenters objected to requiring any

additional pollution controls on NGS because of the uncertainty in

the impairment attribution. | Eighteen commenters generally

supported the adoption of a control strategy consistent with the

MOU, although one commenter expresssed [sic] some concerns

about the compliance schedule and exemption from winter

maintenance. One commenter objected because of the delayed

compliance and the level of control.

The SRP, GCT, EDF, the Wilderness Society, the National

Wildlife Federation, and the Sierra Club have conditionally

withdrawn prior comments that are inconsistent with the August

MOU entered by representatives of SRP, GCT, and EDF and

associated recommendations to the Agency. In light of this

withdrawal, some of the comments discussed here and in the

supplementary response to comments documents placed in the

rulemaking docket are no longer pending before the Agency and,

clearly, no longer "significant" within the meaning of section 307(d)

of the Act. Nevertheless, EPA has discussed them in order to help

the public better understand today’s action.

A. Attribution

The EPA received a large number of comments on its

attribution of visibility impairment in the GCNP to NGS. These

comments expressed a variety of opinions including statements that

8 JCM Environmental, “Summary of Public Comments, Proposed Revision

to Arizona Visibility FIP for Navajo Generating Station” July 1991.

” See U.S. EPA, “Response to Public Comments, Proposed Revisions to

Anzona Visibility FIP,” September 1991.

38

EPA had already made an attribution decision and that EPA should

not revisit that decision, that more than adequate information was

available to make such a decision and that it is clear from the record

that certain visibility impairment episodes in GCNP are attributable

to NGS, that EPA should not make a decision to attribute the

impairment to NGS until it developed a strong source-impairment

relationship, and that EPA should not attribute the impairment to

NGS since it is not the sole cause of the impairment in the GCNP.

In its February 1991 proposal, EPA reopened its comment period

on and explicitly requested public comment on its preliminary

attribution decision. Therefore, EPA’s proposed attribution decision

was subject to public comment. As explained earlier, EPA has

reviewed the available technical data and analyses and has concluded

that episodic visibility impairment in GCNP is_ reasonably

attributable to NGS.

B. Source-Impairment Relationship

Once EPA attributed impairment to NGS, it then had to

develop a source-impairment relationship in order to estimate the

expected improvements in visibility resulting from the installation

of the controls.

In the February 1991 proposal, EPA provided calculations of

the relationship between reductions in SO, emissions from NGS and

the visibility at the GCNP. The calculations were based on the

overall findings of the WHITEX report which used several

techniques to apportion NGS’ contribution to sulfates and the

resulting haze at GCNP. The EPA calculated two measures of

visibility improvements, one seasonal average change in standard

visual range and the other a distribution of days where control of

NGS would result in several levels of contrast changes. The EPA

received several comments that its calculations used to estimate the

distribution of days when the reduction in SO, emissions from NGS

would result in perceptible, quite noticeable, and very apparent

change in contrast in GCNP were in error because the final contrast

assessment did not take into account natural (Rayleigh) atmospheric

scattering of light. The EPA agrees that these calculations did not

aaaaeneaseammaneneeniaaatanel

39

take into account Rayleigh scattering. However, EPA’s seasonal

visual range estimated in the February 1991 proposal, did not

include error as some commenters had alleged and is relied on in

today’s action.

The EPA has received new information during the comment

period which has improved its understanding of the source-

impairment relationship. Based on this and other available

information, EPA has refined the source-impairment relationship

and concluded that it is as described in the Source-Impairment

discussion in "Today’s Action" above.

C. Emission Limitation

Individuals and groups that commented on the emission

limitation generally requested that EPA adopt a requirement for the

maximum level of control or no controls at all. No commenters

supported the 50 percent control level. Some commenters supported

an SRP proposal, now withdrawn, which called for a 70 percent

level. Most commenters on the emission limitation supported a 90

or 95 percent control level. The commenters that did not support

controls did so because of the cost of the controls and/or the alleged

lack of a strong demonstrated source-impairment relationship. As

discussed, EPA believes that an adequate source-impairment

relationship does exist for EPA to require NGS to control its

emissions. The control strategy that EPA is promulgating today will

provide a 90 percent reduction in the SO, emissions from NGS at

a cost below that which would have been necessary to meet the

control strategy EPA proposed in February 1991. Thus, the control

strategy adopted by EPA satisfies both groups of commenters in that

it will result in a greater emission reduction at a lower cost.

Further, as noted previously, EPA must balance a number of factors

in taking action today under section 169A. Such action must be

balanced in light of EPA’s statutory mandate set out in section 169A

to make “reasonable progress” toward the national goal of "the

prevention of any future, and the remedying of any existing,

impairment of visibility in mandatory class I Federal areas which

impairment results from man-made air pollution." The EPA

40

carefully has weighed, for example, the estimated cost of

compliance with today’s emission limitation and the visibility

benefits expected to be realized and concluded that today’s control

Strategy, taken together, is a reasonable exercise of its delegated

rulemaking authority under section 169A.

D. Averaging Time

Several commenters supported a short (3-hour) averaging time

Stating that it would ensure that maximum visibility protection.

Some of these comments have been withdrawn. Other commenters

supported a longer (30 days to annual) averaging time stating that

it would eliminate the need for expensive backup control equipment.

The EPA’s adoption of the control strategy, which includes an

annual rolling average emission limitation, will eliminate the need

for the backup equipment. However, the control strategy also

includes a stringent emission limitation and winter maintenance

scheduling to ensure maximum visibility protection.

E. Compliance Schedule

Several commenters noted that section 169A of the Act requires

that BART emission limitations be met within 5 years. Some of the

commenters have since withdrawn these comments. Others

commented that it was legally permissible for compliance with a

BART emission limitation to extend beyond 5 years. In EPA’s

August 1991 notice reopening the comment period, EPA published

a memorandum from its Office of General Counsel that described

the legal basis for the 1997, 1998, and 1999 compliance schedule

adopted today. No adverse comments addressing the legal basis

were received during the supplemental comment period. Therefore,

as discussed in the August 1991 notice, EPA is relying principally

on the reasonable progress requirements of section 169A(b)(2). The

requirements of that section, which encompasses both the long-term

strategy and BART provisions, are met by today’s action.

F. BART Guidelines

In the February 1991 proposal (56 FR 5178 col. 1), EPA

requested comments on two potential revisions to the BART

Guidelines. First, EPA requested comment on the continued use of

41

the new source performance standards as a base for the BART

analysis. Second, EPA requested comment on whether it should

include the consideration of the value of marketable allowances

when estimating the cost of control systems. The EPA received

several comments on both issues. However, in making its final

determination on today’s action, neither of these issues were

significant. Therefore, EPA is not making a final decision on

whether to revise the BART Guidelines at this time.

G. Cost of Controls

Several commenters alleged that EPA may have overestimated

the economic impact of the controls. They also noted that costs are

quite small when divided up among the various consumers. Other

commenters alleged that the increased cost will have a significant

impact on the consumers. The EPA has recalculated the cost of the

controls of today’s final action based upon the new information

provided by SRP. Those numbers may be overestimates of the

actual cost of control. Even if the actual control costs are consistent

with SRP and EPA estimates, as noted, EPA has concluded that

today’s action achieves greater visibility benefit at less cost than

EPA’s February 1991 proposed option. Moreover, by balancing the

estimated costs, expected visibility improvement, and other relevant

factors EPA has concluded that the expected improvements in the

visibility in the GCNP justify the estimated costs.

Several commenters noted that the recipients of CAP water face

a greater price increase than electricity customers or will otherwise

bear a significant economic hardship as a result of the control costs

of today’s action. As discussed in the Regulatory Flexibility Act

(RFA) section below, EPA cannot estimate every potential indirect

cost of today’s action. Nevertheless, EPA has carefully considered

the potential impacts of today’s action on recipients of CAP water,

including farmers receiving such water. The Central Arizona Water

Conservation District (CAWCD) submitted an estimate of a 7

percent to 10 percent increase in agricultural water rates, resulting

in an increased water cost of nearly $10,000 per year for a typical

family farmer. The EPA estimates such an increase would be less

42

than a 2 percent increase in the total variable costs of a typical

farm. At the present time, agricultural water in Arizona receives a

substantial Federal subsidy. Current Federal policy calls for

phasing out the water subsidy. When the subsidy is eliminated, the

impact of the increased cost at NGS will be an even smaller

percentage of both total variable cost and total water cost. The EPA

has weighed and balanced the costs of today’s action, the expected

visibility improvement, and other requisite considerations and

concluded that today’s control strategy, taken together, constitutes

a reasonable exercise of EPA’s delegated rulemaking authority

under section 169A of the Act.

H. Impacts on Other Class I Areas

Several commenters noted the impact of NGS on the other class

I areas in the region known as the "Golden Circle." Two of the

commenters provided photographs and modeling to support their

statements. The EPA agrees that NGS could contribute to the

visibility impairment in those other areas, but the data and modeling

that were presented were not adequate for EPA to use in making a

definitive determination on attribution of the impairment in those

areas to NGS. However, the controls that are being required today

will significantly reduce if not eliminate NGS’ contribution to the

visibility impairment in those areas.

I, Seasonal Controls

In the February 1991 proposal, EPA requested comment on

whether, as a policy matter, it should allow seasonal controls at

NGS and whether, if technically justified, seasonal controls would

be legally permissible. Several commenters recommended that EPA

adopt a control strategy for NGS that include seasonal controls

and/or commented that seasonal controls could lawfully constitute

an emission limitation representing BART under section 169A.

Several others commented that the technical record did not support

seasonal controls and/or commented that as a legal matter seasonal

controls could not constitute an emission limitation under section

169A. During the comment period, the GCT and its consultants

presented analyses, which are in the rulemaking docket,

43

demonstrating that NGS’ emissions may significantly impair

visibility year-round at the GCNP as well as other Class I areas in

the region. For example, a simple transport and transformation

model of NGS emissions over a 5-year period indicated impacts at

GCNP in seasons other than winter. While the degree of visibility

impairment during these impacts is less than those documented by

other studies during the winter period, the impacts are nevertheless

strongly suggestive of attributable impairment in non-winter seasons.

Further, the analyses suggest that NGS emissions could impair

visibility in surrounding Class I areas between 60 and 80 percent of

the time year-round. For the foregoing reasons, EPA concludes that

year-round controls are appropriate in this action. The EPA’s

decision also is consistent with the recommendations of the parties.

While EPA’s regulations do not cover regional haze effects from

existing sources, EPA also notes that today’s action will lead to

regional haze benefits as well. Therefore, EPA need not determine

whether such controls would be legally permissible.

J. Monetary Benefits

In the February 1991 proposal, EPA noted that it was not

legally required to estimate the monetary benefits associated with

improving visibility in the GCNP. However, as a check of the

reasonableness of its approach, EPA evaluated and considered the

benefit analysis developed as a part of the draft RIA for the

proposed rule. The EPA requested comment on its monetary

benefits analysis and a monetary benefits analysis conducted by

SRP. Some commenters stated that EPA and SRP had

underestimated the monetary benefits while others indicated that

EPA’s benefits were overstated. Because the benefits analysis forms

no part of legal basis for today’s action, EPA is not responding to

those comments.

IV. EDF v. Reilly

Today’s action regarding NGS completes EPA’s obligations

under the settlement agreement in EDF v. Reilly, No. C82-6850

(N.D. Cal.), and under the first round of SIP/FIP planning called

for in the visibility regulations.

a4

V. Nonprecedential Effect of This Action

Today’s promulgation is limited to the rulemaking requiring an

SO, emission limitation for NGS in order to remedy a single-source

visibility impairment in the GCNP that is traceable to NGS and,

hence, remediable under EPA’s 1980 visibility regulations. As

such, it has no direct precedential effect on any other rulemaking

action EPA might undertake in the future regarding other existing

sources or Class I areas. This is so because the outcome of this

rulemaking has been highly dependent upon facts and circumstances

that are unique to this proceeding and thus does not apply to other

cases. For example, the tracer studies and other analyses contained

in the WHITEX and NGSVS reports that are key to a showing that

impairment at GCNP is traceable to NGS are not being approved

for use in any other rulemaking addressing other potential sources

of visibility impairment. Should EPA or State agencies conduct

visibility rulemakings regarding other Class I areas and other

existing sources of pollution in the future, they will need to rely on

attribution techniques appropriate for those other areas and sources

in order to make reasoned regulatory decisions. They will not be

able to rely solely on WHITEX or NGSVS findings that specifically

address the GCNP and NGS to determine the existence of

impairment that is reasonably attributable to any such other source.

Likewise, neither the methodology used for establishing the

expected improvement in visibility that could be anticipated to result

from use of alternative SO, control systems nor the final emission

limit being promulgated today will have direct precedential effect

elsewhere. The statutory requirement that EPA weigh various

factors before reaching a decision is by nature a case-specific

process.

A. Classification

Executive Order No. 12291 requires each Federal agency to

determine if a regulation is a "major" rule as defined by the order

and "to the extent permitted by law," to prepare and consider an

RIA in connection with every major rule. Major rules are defined

as those likely to result in:

(Ri,

45

1. An annual effect on the economy of $100 million or more.

2. A major increase in costs or prices for consumers or

individual industries; Federal, State, or local government agencies;

or geographic region.

3. Significant adverse effects on competition, employment,

investment, productivity, innovation, or the ability of the United

States-based enterprises to compete with foreign-based enterprises

in domestic or export markets.

The EPA judged the February 1991 proposed regulation for

NGS to be a major rule based on projected annualized costs

potentially in excess of $100 million. The EPA then prepared a

preliminary RIA that included estimates of costs, benefits, and net

benefits for three control options. The preliminary analysis, titled

"Regulatory Impact Analysis of a Revision of the Federal

Implementation Plan for the State of Arizona to Include SO,

Controls for the Navajo Generating Station," and an addendum,

titled "Addendum to the Regulatory Impact Analysis of a Revision

of the Federal Implementation Plan for the State of Arizona to

Include SO, Controls for the Navajo Generating Station" are

available in Docket A-89-02A.

In light of the reduced estimated cost of this rule, the Office of

Management and Budget (OMB) has exempted this action from the

RIA requirements. Therefore, EPA did not finalize the preliminary

RIA.

B. Paperwork Reduction Act

This rule will impose a modest reporting burden on the

participants of the NGS to enable EPA to ensure compliance with

the emission limits. Because the reporting burden affects only a

single source, it is not subject to OMB review under the Paperwork

Reduction Act.

C. Section 317(c) Economic Impact Assessment

This action is taken pursuant to sections 169A and 110(c) of

the Act, and a section 317(C) economic impact assessment may not

be required. However, all of the analytical requirements of section

46

317(C) have been met. The section 317(C)(1) analysis of alternative

regulations, and the impact of varying the effective date, are

essential considerations in promulgating this revision to the FIP for

Arizona. Each of the section 317(C) requirements are included in

the BART analysis prepared by EPA. Portions of the economic

impact assessment are revised in this notice based on additional

information and comments provided to EPA.

D. Regulatory Flexibility Act

Under the RFA, 5 U.S.C. 601-612, EPA must prepare for

rules subject to notice and comment rulemaking initial and final

regulatory flexibility analyses describing the impact on small entities

which includes small businesses, small not-for-profit enterprises,

and governmental entities with jurisdiction over populations of less

than 50,000. The requirement of preparing such is inapplicable,

however, if the Administrator certifies that the rule will not have a

significant economic impact on a substantial number of small entities

(see 5 U.S.C. 605(b)).

In the February 1991 proposal, the Administrator certified that

the proposed rule would not have a significant impact on a

substantial number of small entities because none of the participants

of NGS are classified as small according to the guidelines developed

by the Small Business Administration. Moreover, EPA estimated

that for all of the electric utilities using NGS power, the impact on

electric utility rates would be substantially less than 5 percent.

Similarly, the emission limitation being established today does

not have a direct impact on and will not apply to a substantial

number of small entities because none of the participants of NGS

subject to today’s action are classified as small according to the

guidelines developed by the Small Business Administration.

Further, during the comment period on the February 1991 proposal

and the August 1991 supplemental notice, SRP, NPC, and CAWCD

submitted comments indicating that the potential indirect costs that

may be passed on to customers of the NGS participants were less

than those estimated by EPA.

47

The comments on potential indirect costs incurred by rate

payers submitted by SRP, NPC, CAWCD, and others do not

address the costs incurred directly by the NGS participants, the

entities that are subject to and directly impacted by today’s action

and that do not qualify as "small." Further, except for the general

comments submitted by the Bureau of Reclamation, the comments

do not address EPA’s conclusion under the RFA that the proposed

rule would not have a significant economic impact on a substantial

number of small entities. Before discussing the comments, EPA

notes that it cannot and is not legally required to foresee all of the

indirect impacts associated with today’s action. The EPA cannot

anticipate how the NGS participants will seek to pay for the costs

associated with today’s action, the decisions that public utility

commissions will confront and make, etc. Nevertheless, EPA has

considered and attempted to respond to those comments addressing

indirect costs impacting potential small entities in order to explain

more fully for the public the Administrator’s finding under the RFA

that today’s final rule will not have a significant impact on a

substantial number of small entities.

In the draft RIA, EPA estimated a $92 million per year control

cost would increase SRP’s and NPC’s electric rates by a maximum

of 3.1 percent and 4.4 percent, respectively. The EPA estimated

that of the six utilities receiving power from NGS, these two would

experience the greatest electric rate increases if power generation

costs were passed through completely to consumers because both

SRP and NPC rely on 25 percent of NGS electricity for their total

sales and each has a low average cost/kWh relative to the other four

participants. Further, EPA estimated that these percentage estimates

likely were overestimates due to various dampening effects omitted

from the calculation.

* See generally “Draft Regulatory Impact Analysis of a Revision of the Federal

Implementation Plan for the State of Arizona to Include SO, Controls for the

Navajo Generating Station."

48

The NPC submitted comments on EPA’s February 1991

proposal in which it estimated that the control costs would result in

a 2.27 percent increase for its residential customers. The SRP also

submitted comments on the February 1991 proposal in which it

estimated that a control strategy costing $90 million would result in

a 2.1 percent to 2.5 percent increase to its rate payers. These

estimated increases are less than the 5 percent impact estimated in

EPA’s February 1991 proposal. Further, these costs may be

overestimated in that, for example, today’s control option is cheaper

than that advanced in the February 1991 proposal. In addition, the

sale of SO, allowances under the SO, allowance trading program

established by EPA under title IV of the Act may mitigate costs.

In the draft RIA for the February 1991 proposal, EPA also

estimated the econ

This text is long and has been trimmed here. Open the source document for the complete record.