Petition for Writ of Certiorari — PPG Industries, Inc. v. United States Environmental Protection Agency

Cr, aes as ae Supreme Court, U.S,

Ua J Ge. FILED

No. __ JAN 8 1990

SESEEPPEEPANIOL, JR,

IN THE nto

Supreme Court of the United States

OCTOBER TERM, 1989

PPG INDUSTRIES, INC.,

Petitioner,

Vv.

UNITED STATES ENVIRONMENTAL PROTECTION AGENCY,

Respondent.

PETITION FOR A WRIT OF CERTIORARI

TO THE UNITED STATES COURT OF APPEALS

FOR THE FIFTH CIRCUIT

*ROBERT BRAGER

SHELLEY V. LUCAS

BEVERIDGE & DIAMOND, P.C.

1350 I Street, N.W., Suite 700

Washington, D.C. 20005

(202) 789-6000

MADELYN A. REILLY

PPG INDUSTRIES, INC.

One PPG Place

Pittsburgh, Pennsylvania 15272

(412) 434-2430

Counsel for Petitioner

PPG Industries, Inc.

*Counsel of Record

ee em

PRESS OF BYRON 8S. ADAMS, WASHINGTON, D.C. (202) 347-8203

B8 pj”

QUESTIONS PRESENTED

1. Are effluent limitations ‘‘achievable’”’ and ‘‘at-

tainable”’ within the meaning of Sections 301(b\2\A)

and 304(bX2\A) of the Clean Water Act (‘‘Act’’) when

no plant, not even the two plants identified by the

Environmental Protection Agency (‘‘EPA’”’ or

“‘Agency’’) as the best plants in the regulated indus-

try, can comply with all of the limitations?

2. Are effluent limitations ‘‘achievable’”’ and “‘at-

tainable’’ within the meaning of the Act when they

fail to account for all technological variability, thus

assuring that unavoidable exceedances will occur at

even the best plants that properly operate and main-

tain EPA-designated pollution control technology?

ii —

LIST OF PARTIES

Petitioner is PPG Industries, Inc.* Parties sup-

porting this petition are:

The Dow Chemical Company;

Chemical Manufacturers Association;

National Paint and Coatings Association;

Synthetic Organic Chemical Manufacturers

Association, Inc.;

* PPG Industries, Inc.’s subsidiaries and affiliates are: Ace

Insurance Company, Ltd.; Ampaspace S.r.].; Arkansas Chemi-

cals, Inc.; Asahi-Penn Chemical Company, Ltd.; AZDEL, Inc.;

Beiletech Corp.; Deutsche Pittsburgh Corning GmbH; Disvica,

C.A.; Dongju Industrial Co., Ltd.; Exel, Ltd.; Glaskontoret A/

S; Glass Plaza Associates; Glasunit A/S; Guangdong Float Glass

Co. Ltd.; Industrie Vernici Italiane S.p.A.; Inveca-Pittsburgh

C.A.; J.M. Eltzroth & Associates, Inc.; Lenhardt Machinenbau

GmbH; Nippon Pittsburgh Corning, K.K.; PPG Fiber Glass Cor-

poration; PPG Crystal A.G. (PPG Crystal S.A.); PPG Glasexco

S.A.; PPG Glass (UK) Ltd.; PPG Iberica, S.A.; PPG Industries

Foundation; PPG Industries (France) S.A.; PPG Industries, Inc.

(Alaska); PPG Industries Taiwan Ltd.; PPG-Nanchang Chemical

Technology Development Corporation, Ltd.; PPG Scandinavia A/

S; PPG-Siam Silica Company, Limited; PPG Sweden AB; PPG

Vernate Pennitalia S.p.A.; P.T. Asahimas Subentra Chemical

Co., Ltd.; Pennvasia Limited; Pittsburgh Corning Corporation;

Pittsburgh Corning Europe N.V.; Pittsburgh Corning France

S.A.R.L.; Pittsburgh Corning Gesellschaft mbH; Pittsburgh

Corning International Sales Company Limited; Pittsburgh Corn-

ing Nederland B.V.; Pittsburgh Corning Scandinavia AB; Pitts-

burgh Corning (Schweiz) AG; Pittsburgh Corning (United

Kingdom) Limited; Pittsburgh Plate Glass Company (Alabama);

Pittsburgh Plate Glass Company (Delaware); Pittsburgh Plate

Glass Company (Maine); Quantum Technologies, Inc.; Separations

Technology, Inc.; Silenka B.V.; Silenka Deutschland G.M.B.H.;

Silenka France; Silenka U.K. Ltd.; Societe Industrielle, Com-

merciale et Financiere de Peintures; Taiwan Chlorine Industries

Ltd.; Tatung Coatings Co.; VFG-Sudamtex, C.A.; and Vidrio

C.A.

ill

Air Products Manufacturing Corporation and

Air Products and Chemical, Inc.;

Akzo Chemicals, Inc.;

Allied-Signal, Inc.;

Dixie Chemical Company, Inc.;

E.I. du Pont de Nemours & Co.;

Ethyl Corporation;

FMC Corporation;

Goodyear Tire & Rubber Company;

Hoechst Celanese Corporation and Hoechst

Celanese Chemical Group, Inc.;

Koppers Company, Inc.;

LaRoche Chemicals, Inc.;

M & T Chemicals, Inc.;

Monsanto Company;

Rubicon, Inc.:;

Sherex Chemical Company, Inc.;

Sterling Chemicals, Inc.;

Texas Eastman Company, a division of

Eastman Kodak Company;

Union Carbide Corporation; and

W.R. Grace & Company.

Other parties to the case below were:

United States Environmental Protection

Agency;

Borg-Warner Specialty Chemicals, Inc.;

Courtoulds Fibers, Inc.;

Gulf Coast Waste Disposal Authority;

The Lubrizol Corporation; and

Natural Resources Defense Council.

iv

TABLE OF CONTENTS

QUESTIONS PRESENTED. ............ccccscccccssesscceseseeees i

BI Ee IEE bteiieniicietanindcesitibintegeinidn cate ii

A. Statutory Background ...........c..cceeeeeeee.

B. Regulatory Background ........0..ccceeeeeeeeee

C. The Court’s Decision ................ccccscccosees

REASONS FOR GRANTING WRIT. ...........cccceeseeeeeeee 1]

I. THE COURT’S FAILURE TO REQUIRE EPA

TO DEMONTRATE THAT THE BEST PLANTS

IN THE INDUSTRY CAN COMPLY WITH THE

BAT LIMITATIONS CREATES A CONFLICT

WITH FOURTH, NINTH, EIGHTH, AND D.C.

CIRCUIT PRECEDENT ON AN ISSUE OF

FUNDAMENTAL IMPORTANCE TO THIS AND

1

2

2

STATEMENT OF THE CASE. .....sccccsecccscsssesssesscsseens 2

3

4

7

OTHER RULEMAKING PROCEEDINGS ........ ll

A. The Decision of the Court Below Creates

a Conflict Among the Circuit Courts .... 11

B. This Issue Is Fundamental to This and

Future Rulemakings Under the Clean

Wet 20D naiiiedeia ee. 15

II. THE COURT’S HOLDING THAT EPA MAY

PROMULGATE LIMITATIONS THAT PLANTS

CAN BE EXPECTED TO EXCEED BECAUSE

OF UNAVOIDABLE VARIABILITY IS AT

ODDS WITH FOURTH, FIFTH, NINTH, AND

D.C. CIRCUIT PRECEDENT ON AN ISSUE OF

FUNDAMENTAL IMPORTANCE TO THIS AND

OTHER RULEMAKING PROCEEDINGS ........ 16

CONE ATII ccncennsnvnsccsimnndanshncntaapanaiiaidipmilishinisianiis 20

TABLE OF AUTHORITIES

Page

Cases:

American Petroleum Institute v. EPA, 661 F.2d

I TIED ecthstcsnaswanecncsrnansnersssenenananss 10,16,19

Association of Pacific Fisheries v. EPA, 615 F.2d

Be Ne le HED wecesivnnscscnsesepnevncsnsenseesesenanses passim

Chemical Manufacturers Ass’n v. EPA, 870 F.2d

Ee te IED cctssinstncesnenasnssesnessnsinsainninas passim

Chemical Manufacturers Ass’n v. EPA, 885 F.2d

EIR IID cetnenscecessssarsnceceresuvietensiinians passim

CPC International, Inc. v. Train, 540 F.2d 1329

(8th Cir. 1976), cert. denied, 430 U.S. 966

TIP sssiiilphessdeicibinistsaacdonseanamibinchoesonnilaasacnivensun 3,12,13

FMC Corp. v. Train, 539 F.2d 973 (4th Cir

REI bahia niuncersnrtenanwsnanpcnensnanseintatnssacenmsansnsnasnsas passim

Marathon Oil Co. v. EPA, 564 F.2d 1253

SE MD, ~ BIO ED sicinranetsinanencncociesesesasitidaniunaniansvess passim

National Lime Ass’n v. EPA, 627 F.2d 416

Se, SUED, SINE sanceniesesanconssatnanessnneniansessansonsecs passim

Natural Resources Defense Council v. EPA, 859 F.2d

ARTs BEPUIED snnvenciccsnssansnsisbsdvasasussnsseeces 20

Securities Industry Ass’n v. Board of Governors of

the Federal Reserve System, 468 U.S. 137

RT Titi titsciiccchadlinnaineinndencinanencwiendepnaiaansonnsotexese 11-12

Southeastern Community College v. Davis, 442 U.S.

TE il cncstindathiaidinaininneaaninsiiininnstnaranennensees 12

Tanners’ Council of America v. Train, 540 F.2d

I: SNE scensherensscaninesedcmnienvannsnonnses passim

Weyerhaeuser Co. v. Costle, 590 F.2d 1011 (D.C.

iia caaiclsainsnamatininssnssannnniwennxsien 10,16

Statutes:

I in ies inanincanninnnievinndientennsecsanenion 2

SE MIA, AE TIUIIED ccncecsecccocnegssccoscccsssescvercosccsscensss 2

vi

Table of Authorities Continued

abe Page

Oe eat, © TUR ID sccnnsiscissictnsidencmeenaae passim

ee Wee Th I tintstcrctawsinanrnstoensrediieieenead passim

Se eis II charcesccsiccshccncchaannshbabhecunicala ate 16

Se SEA: BD TRIE IIIOD ancosscisncsnnsnrensrhanaeeadsntciananes 5

i NIE EE PINE Andi constuinttnasicedasdercin acd ens 14

Regulations:

A I Se i ee 17

Oe ee © I ii vinirsiniccmnscc 17

QD CPR. § UBRGUAWENT) occcsecccccnsecscscecceassscaszensere 17

lar ey ET pints ASnceacidsccsaceth eas 2

me UR, SE CIID a svenersesenssccsarieccaccsncntanaces 4,5,7

55 Fed. Reg. 80, 81 (1990) ........... By Jeera eater ore 16

IN THE

Supreme Court of the Gnited States

OCTOBER TERM, 1989

No. _—

PPG INDUSTRIES, INC.,

Petitioner,

We

UNITED STATES ENVIRONMENTAL PROTECTION AGENCY,

Respondent.

PETITION FOR A WRIT OF: CERTIORARI

TO THE UNITED STATES COURT OF APPEALS

FOR THE FIFTH CIRCUIT

Petitioner PPG Industries, Inc. prays that a writ of

certiorari issue to review the judgment and opinions of

the United States Court of Appeals for the Fifth Circuit

entered on March 30, 1989 and October 10, 1989.

OPINIONS BELOW “

The initial opinion of the United States Court of

Appeals for the Fifth Circuit is reported in Chemical

Manufacturers Association v. EPA, 870 F.2d 177 (5th

Cir. 1989) (hereinafter “CMA I’), and is reprinted in

the Appendix at p. la. The opinion of the United

States Court of Appeals for the Fifth Circuit denying

ing is reported in Chemical Manufacturers As-

sociation v. EPA, 885 F.2d 253 (5th Cir. 1989) (here-

2

inafter “CMA II’’), and is reprinted in the Appendix

at p. 178a.

JURISDICTIONAL STATEMENT

The -eriginal judgment in this case was entered on

March 30, 1989. Petitioner’s motion for panel re-

hearing was denied by the Fifth Circuit on October

10, 1989. This petition for a writ of certiorari has

been timely filed pursuant to 28 U.S.C. § 2101(c). This

Court’s jurisdiction to review the judgment and opin-

ions of the Fifth Circuit is invoked pursuant to 28

U.S.C. § 1254(1).

STATUTE AND REGULATION INVOLVED

This case involves an EPA rulemaking under the

Clean Water Act. 33 U.S.C. § 1251 et seg. The sta-

tutory provisions relevant to this petition are Sections

301(a),(b) and 304(b) of the Act. 38 U.S.C.

§§ 1311(a),(0) and 1314(b). These provisions are re-

printed in the Appendix at p. 196a. The effluent lim-

itations at issue here are codified at 40 C.F.R. Part

414, and are reprinted in the Appendix at p. 202a.

STATEMENT OF THE CASE

This petition for a writ of certiorari is before the

Court because the court below found that a set of

effluent limitations is ‘‘achievable” and “‘attainable’”’

within the meaning of Sections 301(bX2XA) and

304(bX2XA) of the Clean Water Act, 33 U.S.C.

§§ 1311(bX2\A) and 1314(bX2XA), even though those

limitations cannot be achieved by any plant in the

industry, not even by the two “best” plants on whose

performance the limitations were based. The Fifth

Circuit conceded that its opinion conflicts with the

Fourth Circuit’s decision in Tanners’ Council of Amer-

ica v. Train, 540 F.2d 1188, 1192-94 (4th Cir. 1976),

but expressly declined to follow the Fourth Circuit's

decision. CMA II, 885 F.2d at 264; App. at 19la. The

decision of the court below is also at odds with As

sociation of Pacific Fisheries v. EPA, 615 F.2d 794

(9th Cir. 1980); CPC International, Inc. v. Train, 540

F.2d 1329 (8th Cir. 1976), cert. denied, 430 U.S. 966

(1977); and National Lime Ass’n v. EPA, 627 F.2d

416, 430-51 (D.C. Cir. 1980).

Additionally, this petition is before the Court be-

cause the court below held that effluent limitations

are achievable even though they are set at a level

that can be expected to result in exceedances for each

pollutant 1% of the time (for daily maximum limita-

tions) and 5% of the time (for monthly average lim-

itations), and the limitations provide no affirmative

defense even though the exceedances are unavoidable.

In analogous circumstances, other circuits have

deemed EPA regulations to be unattainable. National

Lime, 627 F.2d at 430-44; Marathon Oil Co. v. EPA,

564 F.2d 1253, 1271-74 (9th Cir. 1977); FMC Corp.

v. Train, 5389 F.2d 978, 985-86 (4th Cir. 1976).

This petition should be granted because the decision

below creates a conflict among the circuits and raises

important and recurring issues concerning develop-

ment of regulations under the Clean Water Act.

A. Statutory Background

Sections 301(bX2XA) and 304(bX2XA) of the Clean

Water Act require EPA to promulgate regulations,

known as “effluent limitations guidelines,’ which are

sets of numerical limitatior.s restricting the amounts

of pollutants that may be discharged from various

categories or classes of industrial facilities. 33 U.S.C.

§§ 1311(bKX2KA) and 1314(bX2XA). These limitations

reflect the reduction of pollutants that is achievable

using the “best available technology” (“BAT’’). Jd.

The Act requires BAT effluent limitations to be

“achievable” and “attainable” by the class or cate-

gory of regulated plants. Jd.

As a matter of established practice, when devel-

oping BAT limitations for a category or class of fa-

cilities, EPA identifies the best technology available

to treat a set of pollutants and determines whether

any plants in the regulated industry effectively use

that technology. If no plant in the regulated industry

does so, the Agency looks to other industries and,

through modeling and technical analyses, transposes

that technology to the industry being regulated.

Where, as here, some plants in the industry already

effectively use the EPA-designated technology, the

~~~Agency collects and analyzes data from those plants

to determine what pollutant reductions are possible.

Data representing BAT-level performance are re-

tained, while data not representing BAT-level per-

formance are excluded from EPA’s data base. Based

on the data retained in EPA’s data base, the Agency

then promulgates two limitations (a daily maximum

limitation and a monthly average limitation) for each

regulated pollutant.

B. Regulatory Background

On November 5, 1987, EPA promulgated effluent

limitations guidelines applicable to the more than 1000

plants in the organic chemicals, plastics, and synthetic

fibers (“OCPSF’’) industry. 52 Fed. Reg. 42,522

(1987). Those limitations were challenged in the Fifth

Circuit which had jurisdiction pursuant to 33 U.S.C.

§ 1369(b\1XE).

The OCPSF Guidelines impose daily maximum and

monthly average limitations for 66 pollutants. Twenty-

eight of those pollutants are volatile compounds for

which EPA properly designated a single technology—

steam stripping—as BAT. 52 Fed. Reg. at 42,538-41.

Steam stripping is a process in which wastewater is

preheated to temperatures near boiling and then

pumped into the top of a distillation column (usually

20 to 30 feet high). Superheated steam is continuously

injected into the bottom of the column. When the

steam contacts the wastewater, the volatile com-

pounds evaporate out of the wastewater into the

steam and exit through the top of the distillation

column into a condenser. The compounds then are

removed from the condenser and either recycled or

incinerated. The cleansed wastewater exits at the bot-

tom of the distillation column.

BAT-level steam strippers typically remove more

than 99% of all of the volatile compounds in the

wastestream. A BAT-level steam stripper’s effective-

ness in treating any single volatile compound, how-

ever, depends on the characteristics of the

wastestream (i.e., the nature, amounts, and interac-

tions of the pollutants in the wastestream) and the

stripper’s operating parameters. For example, in-

creasing the temperature of the wastestream in the

preheater or increasing the volume and temperature

of the steam injected into the distillation column can

enhance the stripper’s ability to remove certain pol-

lutants but, depending on wastestream characteris-

tics, can also impair the stripper’s ability to remove

other pollutants. Consequently, steam strippers are

operated to optimize overall removal of all of the

volatile compounds in a particular plant’s wastes-

tream.

To develop the BAT steam stripping limitations,

EPA reviewed performance data from many plants

and then compiled a data base using data from the

few plants that effectively use BAT-level steam strip-

pers. As EPA has stated, the Agency then deleted

certain data from its data base, retaining “‘only those

data that are consistent with good or sound opera-

tion.... In the case of all pollutants, EPA has se-

leeted-only-a-very-smalt subset of the industry using

the best available technology appropriately, and EPA

has only used data that represented good perform-

ance.’’ See EPA’s Response to Comments, reprinted

in the Appendix at p. 259a.

There were two plants in EPA’s data base whose

steam strippers outperformed all of the others: PPG

Plant 913 and Dow Plant 415.' Recognizing the su-

perior performance of the steam strippers at PPG

Plant 913 and Dow Plant 415, EPA relied almost

exclusively on data from those two plants to establish

the BAT steam stripping limitations for the 28 vol-

atile compounds. Indeed, no data from either plant

were excluded from the data base during EPA’s rig-

orous data editing process. Both the data themselves

‘The performance of the steam strippers at Plants 913 and

415 was exemplary: Many of the volatile pollutants found in

those plants’ wastestreams prior to treatment could not even be

detected after treatment, and at no time at either plant did the

total of all volatile compounds in the wastestream after treat-

ment exceed 1 part per million (“‘ppm’”’). Indeed, at both plants

the total of all volatile compounds averaged less than a quarter

of a part per million after treatment.

|

and EPA’s method for editing those data demonstrate

| that PPG Plant 913 and Dow Plant 415 have BAT-

level steam strippers; as EPA stated in the preamble,

“only data representing BAT-level design and oper-

ation were retained for purposes of developing limi-

tations.”’ 52 Fed. Reg. at 42,540.

In light of the statutory command that limitations

be achievable and attainable by the category or class

of plants being regulated, 33 U.S.C. §§ 1311(b\(2\A),

1314(bX2\A), one would have expected EPA to es-

tablish effluent limitations that could be attained by

the two best steam strippers in the industry. How-

ever, neither plant can achieve the limitations for all

pollutants. As the court below noted, a comparison

of the BAT-level performance data with the limita-

tions demonstrates that “‘PPG plant 913 exceeded the

monthly average limit for chloroform, and Dow plant

415 exceeded the daily maximum limitation for tri-

chloroethylene ....’’ CMA I, 870 F.2d at 238; App.

at 112a.

C. The Court’s Decision

The court below acknowledged that neither PPG

Plant 913, Dow Plant 415, nor any other plant in

EPA’s data base can comply with the full set of steam

stripping limitations, but deemed this ‘‘irrelevant’’ as

a matter of law. CMA I, 870 F.2d at 239; App. at

i13a. According to the court, PPG Plant 913’s ina-

bility to achieve the chloroform limitation was “‘ir-

relevant” because Dow Plant 415 could do so, while

Dow Plant 415’s inability to achieve the trichloroe-

thylene (“TCE”) limitation was ‘‘irrelevant’’ because

PPG Plant 913 could do so. Jd. The court’s conclusion

was based on its holding that, as a matter of statutory

construction, EPA could demonstrate that an entire

™

set of limitations is achievable by selecting perform-

ance data from different plants for different pollu-

tants, and can ignore as “‘irrelevant’’ data

demonstrating that no plant can attain the full set of

limitations. Jd.

The Fifth Circuit did not find that PPG Plant 913

can alter its performance to achieve the chloroform

limitation, and there is no basis in the record that

would support such a conclusion. Regarding Dow

Plant 415, the Fifth Circuit erroneously hypothesized

that “EPA could reasonably conclude’ that Plant

415’s inability to achieve the TCE limitation was

either due to an upset or a quality control problem.

CMA I, 870 F.2d at 238-39; App. at 112a. The court

did not and could not base such a conclusion on the

record, however, because EPA did not reach this con-

clusion.? Indeed, because EPA excluded from the data

base all data reflecting upsets and quality control

problems, but did not edit any TCE data from Dow

Plant 415, EPA must have concluded that the TCE

exceedance did not result from an upset or quality

2The absence of such a conclusion is especially noteworthy

because Dow Plant 415 was the subject of a comprehensive,

month-long EPA study. The Agency was at the plant monitoring

operations, sampling steam stripper influent and effluent, and

analyzing the data at all times during the study, including the

day on which the elevated TCE reading occurred. The study

resulted in a 350-page Agency report on Plant 415’s perform-

ance, yet the report nowhere mentions a steam stripper upset

or quality control problem. At other plants studied by EPA, the

Agency identified upsets and quality control problems and edited

the cata base accordingly. Had an upset or quality control prob-

lem occurred at Dow Plant 415, EPA would have noted that

fact in the record and edited the resulting data from the Agen-

cy’s data base.

ee ey enna it

control problem. See 52 Fed. Reg. at 42,540; infra

pp. 17-19.

Despite EPA’s determination that all of the data

from PPG Plant 913 and Dow Plant 415 “‘repre-

sent[ed] BAT-level design and operation,” 52 Fed.

Reg. at 42,540, and despite the fact that this data

demonstrate that neither plant can achieve all of the

OCPSF limitations, the court below upheld the limi-

tations. This decision presents a direct, acknowledged

conflict with the Fourth Circuit’s decision in Tanners’

Council and is at odds with Justice (then Judge) Ken-

nedy’s opinion in Association of Pacific Fisheries, the

Eighth Circuit’s decision in CPC International, and

the D.C. Circuit’s decision in National Lime.

Petitioner prays that the Court will grant this petition

and resolve this conflict.

Additionally, all of the OCPSF limitations (i.e. those

based on other technologies as well as those based

on steam stripping) are unattainable because of the

methodology used by EPA to develop the OCPSF

limitations. In developing the limitations, EPA rec-

ognized all technology has its limits, and even the

best designed and operated pollution control technol-

ogy will experience fluctuations (or ‘‘variability’’) in

performance. As the court below stated,

[t]he same plant using the same treatment

method to remove the same toxic does not

always achieve the same result. Tests con-

ducted one day may show a different con-

centration of the same toxic than are shown

by the same test the next day. This varia-

bility may be due the inherent inaccuracy of

analytical testing, i.e., ‘‘analytical variabil-

10

ity,’ or to routine fluctuations in a plant’s

treatment performance.

CMA I, 870 F.2d at 228; App. at 89a. To prevent

the occurrence of unavoidable exceedances, EPA must

allow for all of the variability inherent in use of the

EPA-designated technology.

Rather than do so, however, EPA used a statistical

methodology that allows for only 99% of the inherent

variability for the daily maximum limitation for each

pollutant and only 95% of the inherent variability for

the monthly average limitation for each pollutant. As

all courts addressing the issue have held, when EPA

does not allow for 100% variability, unavoidable ex-

ceedances will result. See American Petroleum Insti-

tute v. EPA, 661 F.2d 340, 350-53 (5th Cir. 1981);

Weyerhaeuser Co. v. Costle, 590 F.2d 1011, 1056-58

(D.C. Cir. 1978); Marathon Oil, 564 F.2d at 1266;

FMC Corp., 539 F.2d at 985-86. Consequently, other

courts have required EPA to provide an affirmative

defense or otherwise account for the unavoidable ex-

ceedances that result from EPA’s use of less than

100% variability factors. See National Lime, 627 F.2d

at 430-51; Marathon Oil, 564 F.2d at 1271-74; FMC

Corp., 539 F.2d at 985-86. However, the court below

upheld the limitations even though EPA did not pro-

vide an affirmative defense or otherwise account for

the expected unavoidable exceedances. Petitioner

prays that the Court will grant this petition and re-

solve this additional conflict.

11

REASONS FOR GRANTING WRIT

I. THE COURT’S FAILURE TO REQUIRE EPA TO DEM-

ONSTRATE THAT THE BEST PLANTS IN THE INDUS-

TRY CAN COMPLY WITH THE BAT LIMITATIONS

CREATES A CONFLICT WITH FOURTH, NINTH,

EIGHTH, AND D.C. CIRCUIT PRECEDENT ON AN IS-

SUE OF FUNDAMENTAL IMPORTANCE TO THIS AND

OTHER RULEMAKING PROCEEDINGS.

A. The Decision of the Court Below Creates a Conflict

Among the Circuit Courts.

Prior to this case, every court addressing the issue

has held that limitations are not ‘‘achievable’’ within

the meaning of the Act, 33 U.S.C. § 1311(b\2 A),

when the record indicated that no plant in the reg-

ulated industry was able to achieve all of those lim-

itations using the EPA-designated technology. To

justify its unprecedented opinion to the contrary, the

court below noted that BAT limitations may be based

on the performance of ‘“‘the single best performing

plant in an industrial field.” CMA I, 870 F.2d at 239;

App. at 113a. Instead of requiring the Agency to

demonstrate that the single best performing plant or

plants can achieve all of the limitations, however, the

court below held that achievability can be based on

the performance of identical technology at different

“best” plants ‘‘on a pollutant-by-pollutant basis.’’® Jd.

3’ The court provided no statutory interpretation to support

this conclusion, but relied solely on “‘deference’”’ to EPA’s alleged

construction of the Act. CMA I, 870 F.2d at 239; App.

at 113a. This alleged interpretation of the Act was neither ar-

ticulated in the record nor in the Agency’s brief, but was raised

for the first time by the Department of Justice at oral argument.

As such, the interpretation is a post hoc rationalization that ‘is

entitled to no deference. See Securities Industry Ass'n v. Board

12

The court unjustifiably concluded that as long as a

hypothetical ‘‘best’’ plant—created by taking perform-

ance data for different pollutants from different

plants—theoretically could achieve the full set of lim-

itations, then it is irrelevant that no actual plant can

do so. This holding ignores the technological limits

inherent in simultaneously steam stripping numerous

pollutants from multi-pollutant wastestreams. See su-

pra pp. 5-6.

Moreover, the decision of the court below is con-

trary to all applicable precedent; it directly conflicts

with the Fourth Circuit’s decision in Tanners’ Coun-

cil, 540 F.2d at 1192, and is at odds with the Ninth

Circuit’s decision in Association of Pacific Fisheries,

615 F.2d at 816-20; the Eighth Circuit’s decision in

CPC International, 540 F.2d at 1338-40; and the D.C.

Circuit’s decision in National Lime, 627 F.2d at 430-

51. In Tanners’ Council, 540 F.2d at 1190-92, the

court found that by using the EPA-designated tech-

nology, some plants could meet the limitations for one

pollutant (TSS) and other plants could meet the lim-

itations for another pollutant (BOD), but no plant

could meet the limitations for both pollutants.‘ Had

the court in Tanners’ Council determined the achiev-

ability of the tannery limitations by selecting per-

of Governors of the Federal Reserve System, 468 U.S. 137, 143-

44 (1984). Moreover, this deference was improper because any

such EPA “‘interpretation” would be contrary to law. South-

eastern Community College v. Davis, 442 U.S. 397, 411-12 (1979).

‘ Similarly, Dow Plant 415 can meet the limitations for chlo-

roform, PPG Plant 913 can meet the limitations for TCE, but

neither plant can meet the limitations for both pollutants, even

though both plants use the identical EPA-designated technology.

CMA I, 870 F.2d at 238-39; App. at 113a.

13

formance data for different pollutants from different

plants, those limitations would have been upheld.

However, the Fourth Circuit did not assess the lim-

itations’ achievability on a pollutant-by-pollutant basis,

but instead remanded the tannery limitations because

the entire set of limitations must be ‘‘achievable by

the affected plants.’’ Tanners’ Council, 540 F.2d at

1192.

In Association of Pacific Fisheries, 615 F.2d at 817,

Justice (then Judge) Kennedy stated that the ‘“‘best”’

plants in the industry must be able to achieve all of

the limitatiuns with EPA-designated technology. The

technologies at issue in Association of Pacific Fish-

eries were dissolved air flotation (‘“‘DAF’’) and aerated

lagoons; EPA had designated each technology as BAT

for different types of plants. When discussing DAF

plants, the court stated that ‘EPA is ... charged

with the burden of showing that .. . the best existing

DAF units can meet the limitations.’ Id. (emphasis

added). Because all the limitations were met by ‘“‘the

best existing DAF units,” those limitations were up-

held. Jd. However, because EPA had no data showing

that two of the limitations could be achieved with

aerated lagoons, the limitations for those two pollu-

tants were remanded to EPA. /d. at 819. According

to the Ninth Circuit, all of the limitations must be

achievable by the best plant or plants.

In CPC International, the Eighth Circuit found that

the model plant could meet the limitations for one

pollutant (BOD), but not the other (TSS), and no plant

could meet both limitations. Rejecting EPA’s reliance

on other evidence that allegedly implied that the TSS

limitations could be achieved, the court struck down

the TSS limitations and suggested that EPA set those

14

limitations at levels achievable by the model plant.

540 F.2d at 1338-40.

In National Lime, 627 F.2d at 480-51, the D.C.

Circuit struck down standards of performance for lime

manufacturing plants that had been promulgated pur-

suant to Section 111 of the Clean Air Act, 42 U.S.C.

§ 7411. The court held that EPA had not demon-

strated that those standards were achievable when

one out of the three best plants using the EPA-des-

ignated technology could not achieve the standards.

The court stated:

[i]f, for unexplained reasons, one-third of the

tested plants initially chosen by EPA for their

well-controlled systems fail to meet the

standard, the conclusion is just as plausible

that the standard is not achievable as that

the plants chosen did not have well controlled

systems. It is up to EPA to dispel such

doubts, and they have not done so here.

Id. at 444. Rather than require EPA to “‘dispel such

doubts” here, the court below simply deemed “‘irrel-

evant” the fact that neither “‘best”’ plant can achieve

all of the limitations.®

5 The court below tacitly admitted that PPG Plant 913 cannot

achieve the limitations, but stated that ‘‘the industrial petitioners

have failed to demonstrate that the Dow Plant is incapable of

meeting the BAT, limitations.’”” CMA I, 870 F.2d at 239; App.

at 112a. The court’s decision thus impermissibly shifted the bur-

den of proof from EPA to petitioner. The Act requires that

effluent limitations be achievable, 33 U.S.C. § 1311(bX2XA), and,

as this Court has held, “‘[i]t is the proponent of a rule or order

who has the burden of proof in administrative proceedings.”

Industrial Union AFL-CIO v. American Petroleum Institute, 448

U.S. 607, 653 (1980). Accordingly, EPA must demonstrate, based

15

B. This Issue Is Fundamental to This and Future Rule-

makings Under the Clean Water Act.

Not only does the decision below create an unac-

ceptable conflict among the circuits, but the issue

raised is one of fundamental importance to this and

to future EPA rulemaking proceedings under the

Clean Water Act. The ultimate premise underlying

technology-based limitations is that they can be

achieved through use of the EPA-designated tech-

nology by the category or class of plants subject to

regulation. See 33 U.S.C. § 1311(b)2XA). Where, as

here and in most industries, plants have multi-pollu-

tant wastestreams, the EPA-designated technology

must be able to attain simultaneously all of the lim-

itations. Permitting EPA to demonstrate the achiev-

ability of limitations that are based on a single

technology with performance data from different

plants for different pollutants negates this achieva-

bility requirement. The correct view, adopted by all

of the other circuits that have addressed the issue,

is to require EPA to demonstrate, with performance

data from the best plants using the EPA-designated

technology, that all of the limitations are achievable

by the best plant or plants.

This conflict should be resolved now because the

issue posed will recur in numerous future rulemaking

proceedings under the Clean Water Act. Promulga-

tion of effluent limitations under the Clean Water Act

is an ongoing process; the Act requires EPA to re-

view and, where appropriate, revise, effluent limita-

tions for the more than 50 existing industrial

categories on an annual basis, and requires EPA to

on evidence in the record, that the OCPSF industry can achieve

the limitations.

16

promulgate limitations for new categories of sources

as well. 33 U.S.C. § 1814(m). For example, EPA has

recently announced its intention to develop limitations

for five new industrial categories; to revise limitations

for three existing industrial categories; to study eight

industries to determine whether limitations are war-

ranted; and to review limitations for three existing

categories to determine whether they should be re-

vised. 55 Fed. Reg. 80, 81 (1990). Unless the conflict

among the circuits is resolved, EPA can be expected

to follow the inappropriate course of action the

Agency pursued here when it develops future effluent

limitations. Accordingly, petitioner prays that the

Court will grant this petition.

II. THE COURT’S HOLDING THAT EPA MAY PROMUL-

GATE LIMITATIONS THAT PLANTS CAN BE EX-

PECTED TO EXCEED BECAUSE OF UNAVOIDABLE

VARIABILITY IS AT ODDS WITH FOURTH, FIFTH,

NINTH, AND D.C. CIRCUIT PRECEDENT ON AN IS-

SUE OF FUNDAMENTAL IMPORTANCE TO THIS AND

OTHER RULEMAKING PROCEEDINGS.

EPA has admitted that it set all of the OCPSF

limitations using variability factors of 99% for the

daily maximum limitations and 95% for the monthly

average limitations. Consequently, EPA admittedly

failed to allow for 1% and 5% of the variability in-

herent in the use of EPA-designated technology. As

every circuit addressing the issue has found, this fail-

ure will result in corresponding (1% and 5%) instances

of noncompliance. See American Petroleum Institute,

661 F.2a at 351; Weyerhaeuser, 590 F.2d at 1056-58;

Marathon Oil, 564 F.2d at 1266; FMC Corp., 539 F.2d

at 986. Indeed, EPA has itself acknowledged this fact,

stating: ‘‘a discharger running a properly operated

and maintained treatment facility has a 95-99% chance

17

of complying with its permit limits in any single mon-

itoring observation.” See EPA’s “Training Manual for

NPDES Permit Writers” at p.17, reprinted in the

Appendix at p. 26la. Thus, EPA’s use of 99% and

95% variability factors in the OCPSF limitations can

be expected to result in unavoidable exceedances for

each pollutant 1% and 5% of the time, despite ap-

propriate use of the EPA-designated technology.

Plants are required to comply with the OCPSF lim-

itations 100% of the time, are strictly liable if they

fail to do so, and must report all exceedances to EPA.

33 U.S.C. § 1311(a); 40 C.F.R. § 122.41(a) and (1X6),(7).

Consequently, when EPA has failed to account for all

technological variability, other circuit courts have re-

quired EPA to provide an affirmative defense or oth-

erwise account for the inevitable and unavoidable

noncompliance. See National Lime, 627 F.2d at 430-

51; Marathon Oil, 564 F.2d at 1272-74; FMC Corp.,

539 F.2d at 986. Nevertheless, the court below upheld

EPA’s use of 99% and 95% variability factors without

a corresponding affirmative defense, stating incor-

rectly that the resulting 1% and 5% instances of non-

compliance represent either upsets for which the upset

defense is available,* or quality control problems for

which no defense is required. CMA I, 870 F.2d at

229-30; App. at 112a.

* An upset is ‘‘an exceptional incident in which there is un-

intentional and temporary noncompliance with technology based

permit effluent limitations because of factors beyond the rea-

sonable control of the permittee.” 40 C.F.R. § 122.41(nX1). A

permittee who has violated technology-based permit limitations

as the result of an upset can use the upset defense as an af-

firmative defense to liability. See id. at § 122.41(n).

18

The Fifth Circuit’s erroneous conclusion that the

1% and 5% exceedances resulted from upsets or qual-

ity control problems was based on the court’s incor-

rect assumption that the limitations were derived from

all of the data collected from the data base plants

(i.e., both the retained and the edited data), and that

unreasonably high readings were excluded by setting

the variability factors at 99% and 95% instead of at

100%. In fact, however, as the Agency has acknowl-

edged, EPA first edited the data from the model

plants to expunge all data “‘not representative of BAT

technology performance,” including ‘‘treatment sys-

tem upsets[,] equipment malfunctions[, and] perform-

ance not up to design specifications.”” CMA II, 885

F.2d at 263; App. at 189a. Only after data repre-

senting upsets and quality control problems had been

deleted from the data base were the 99% and 95%

variability factors applied to formulate the limitations.

Consequently, the data points that fell above the 99

and 95 percentiles of the edited data could not have

resulted from upsets or quality control problems, but

represented normal, expected variability in perform-

ance that occurs relatively infrequently, 7.e., 1% and

5% of the time.

On rehearing, the court apparently recognized the

correctness of this reasoning; its only response was

to assert, incorrectly, that EPA edited only whole sets

of data from improperly operated or maintained

plants, but ‘‘did not use its editing criteria to exclude

individual data points.’ Jd. This assertion is contra-

dicted by the administrative record, which shows that

EPA in fact edited individual data points representing

inferior treatment from model plants whose data was

otherwise used to set the limitations. For example,

19

EPA edited two days’ worth of data from a third

company’s plant (Plant 725) ‘“‘because steam stripper

performance for those two days was considered in-

adequate.’’ However, the remaining data from that

third plant were used to set the limitations. See De-

velopment Document for Effluent Limitations Guide-

lines and Standards for the OCPSF Industry, at p.

VII-190, reprinted in the Appendix at p. 263a.

The excluded data from this third plant may well

have reflected an upset or quality control problem;

the discharges were in excess of 3359 ppm and 1558

ppm. In contrast, Dow Plant 415’s daily maximum

TCE exceedance was .085 ppm and PPG Plant 913’s

monthly average chloroform exceedance was .129

ppm. These data were not edited, but were retained

in the data base as data representing good perform-

ance. Significantly, because the exceedances experi-

enced by Plants 913 and 415 did not result from

upsets, the upset defense is inapplicable.

Because data representing inadequate performance

were excluded from the data base prior to application

of the variability factors, the 1% and 5% exceedances

can be expected despite BAT-level performance. Since

the limitations do not contain an applicable affirma-

tive defense or otherwise account for these unavoid-

able exceedances, the OCPSF limitations are not

‘achievable’ and ‘“‘attainable’’ within the meaning of

the Act, and the decision of the court below sustain-

ing them is at odds with decisions of the Fourth,

Fifth, Ninth, and D.C. Circuits in FMC Corp., 539

F.2d at 936; American Petroleum Institute, 661 F.2d

at 350-52; Marathon Oil, 564 F.2d at 1272; and Na-

tional Lime, 627 F.2d at 430-51.

20

This issue is one of fundamental importance to this

and future rulemaking proceedings under the Clean

Water Act. As the D.C. Circuit recently stated, ‘‘[a]

technology-based standard discards its fundamental

premise when it ignores the limits inherent in the

technology.”” Natural Resources Defense Council v.

EPA, 859 F.2d 156, 208 (D.C. Cir. 1988). Moreover,

as noted supra pp. 15-16, the Agency is both devel-

oping effluent limitations for new categories of

sources and reevaluating the limitations imposed on

the more than 50 existing industrial categories al-

ready subject to regulation. Unless the conflict among

the circuits is resolved, EPA can be expected to fol-

low, in the development of those limitations, the in-

appropriate course of action the Agency followed here.

Accordingly, petitioner prays that the Court will grant

this petition.

CONCLUSION

This petition should be granted both to resolve the

conflict among the circuits and to correct errors of

fundamental importance to this and future rulemak-

ings under the Clean Water Act. As a result of the

decision below, the most important check on the

Agency’s rulemaking authority under the Clean Water

Act has been vitiated. Allowing that decision to stand

will permit EPA to promulgate technology-based lim-

itations without making any meaningful demonstra-

tion that those limitations are achievable and

attainable. Petitioner prays that the Court will grant

this petition.

21

Respectfully submitted,

ROBERT BRAGER

SHELLEY V. LUCAS

BEVERIDGE & DIAMOND, P.C.

1350 I Street, N.W.

Washington, D.C. 20005

(202) 789-6000

MADELYN A. REILLY

PPG INDUSTRIES, INC.

One PPG Place

Pittsburgh, Pennsylvania 15272

(412) 434-2451

Counsel for PPG Industries, Inc.