MOORE, Circuit Judge:
Three chemical companies, Hooker Chemicals and Plastics Corporation, Stauffer Chemical Company and Monsanto Company (“Petitioners”), in an action against Russell E. Train, Administrator of The Environmental Protection Agency (“EPA”), seek by petition to review and set aside regulations establishing effluent limitation guidelines
for the phosphate manufacturing industry, issued on February 20, 1974, by the EPA pursuant to the Federal Water Pollution Control Act Amendments (“Act”) 33 U.S.C. § 1251,
et seq.
Although Petitioners contend that this Court lacks jurisdiction, asserting that the
regulations must be initially reviewed in the District Court pursuant to the Administrative Procedure Act (“APA”), 5 U.S.C. §§ 701-06, they have filed their petition as a “protective measure” to avoid the risk of losing their right to challenge the regulations for failure to file the petition within 90 days of promulgation as required by § 509(b)
of the Act.
In 1972, Congress, anxious to end pollution of our Nation’s waters, enacted the Act. It was radically different from previous laws on the subject, and drastically changed the national approach to the control of water pollution. Enforcement of predecessor statutes, which had relied on water quality standards as the primary method of pollution control, had been largely unsuccessful. It was too difficult to establish the necessary correlation between effluent discharges by particular sources
and the quality of the body of water into which the effluent flowed. To solve the dilemma the Act, while retaining water quality standards, predicated pollution control on the application of control technology on the plants themselves rather than on the measurement of water quality.
In the Act, Congress declared its sweeping and praiseworthy goals of which the following are representative: “to restore and maintain the chemical, physical, and biological integrity of the Nation’s waters”; to eliminate discharge of pollutants by 1985; to achieve by July 1,1988, “whereever attainable”, a water quality which would provide for the protection of fish, shellfish and wildlife; to develop technology necessary to eliminate discharge of pollutants, and to protect “the primary responsibilities and rights of States” to eliminate pollution.
The Act distinguishes between existing effluent dischargers and those constructed after the promulgation of the Act’s operative regulations. For the “new” effluent dischargers § 306
of the Act requires the Administrator to promulgate regulations which “reflect the greatest degree of effluent reduction which the Administrator determines to be achievable through application of the best available demonstrated control technology.” When the “new” dis-charger obtains a permit which incorporates as permit conditions the applicable “new” source regulations, compliance with the permit constitutes compliance with the Act.
The same permit mechanism applies to existing effluent discharges, but the Act prescribes two-tiered restrictions.
Section 301(b)(1)(A)
provides:
“[i]n order to carry out the objective of this Act, there shall be achieved . . not later than July 1, 1977, effluent limitations for point sources . . . which shall require the application of the best practicable control technology currently available as defined by the Administrator pursuant to section 1314(b) [§ 304(b)]
Stiffer restrictions are mandated to commence not later than July 1, 1983. Section 301(b)(2)(A)
provides that:
“In order to carry out the objective of this Act, there shall be achieved . not later than July 1, 1983, effluent limitations for categories and classes of point sources . . . which . . . shall require application of the best available technology economically achievable for
such category or class, which will result in reasonable further progress toward the national goal of eliminating the discharge of all pollutants, as determined in accordance with regulations issued by the Administrator pursuant to section 1314(b)(2) [§ 304(b)(2)] . .
Section 304(b)
to which both parts of § 301 refer states,
inter alia:
“For the purpose of adopting or revising effluent limitations under this Act the Administrator shall . . . publish within one year of enactment of this title, regulations, providing guidelines for effluent limitations. . . . Such regulations shall—
(1)(A) identify, in terms of amounts of constituents and chemical, physical, and biological characteristics of pollutants, the degree of effluent reduction attainable . . . [and]
(B) specify factors to be taken into account in determining the control measures and practices to be applicable to point sources . . . within such categories or classes . . . .”
For 1977 limitations, the factors to be considered are:
“the total cost of application of technology in relation to the effluent reduction benefits to be achieved from such application . . . the age of equipment and facilities involved, the process employed, the engineering aspects of the application of various types of control techniques, process changes, non-water quality environmental impact (including energy requirements), and such other factors as the Administrator deems appropriate.”
Comparable factors are specified for 1983 limitations.
JURISDICTION
The first question in this case is whether we have subject matter jurisdiction of the controversy. Under § 509,
the jurisdictional section of the Act, review of the Administrator’s action “in approving or promulgating any effluent limitation or any other limitation under section 1311 [§ 301], 1312 [§ 302], or 1316 [§ 306]” and “in issuing or denying any permit under section 1342 [§ 402]” may be had in the Courts of Appeals. The status of the contested regulations as effluent limitations promulgated under § 301 is the jurisdictional linchpin. The parties’ views on that issue vary with their views of the way the Act is to be administered for existing point sources.
Petitioners argue that we lack jurisdiction because the regulations were promulgated exclusively under § 304, which § 509 omits from the enumeration of sections whose regulations are reviewable initially in the Courts of Appeals. They arrive at this conclusion by a process of elimination. The regulations assertedly were not promulgated pursuant to § 301, which is merely declarative of the Act’s objectives and the time limits for their achievement. Those objectives are effectuated pursuant to the authority conferred by other sections of the Act. The “effluent limitations” mentioned in § 301, they claim, are only established during a case-by-case permit-application procedure. In that process, permit-grantors apply to individual effluent discharges the § 304 guidelines which identify for 1977 “the degree of effluent reduction attainable through the application of the best practicable control technology” and for 1983 “the degree of effluent reduction attainable through the application of the best control measures and practices achievable.” Since the Administrator does not need guidelines to guide himself, the intention of Congress, they argue, was to guide the permit-issuing official or agency who, in issuing permits for specific plants, would thereby create the effluent limitations for that plant.
Thus, since the regulations were not promulgated pursuant to § 301, Petitioners conclude that the regulations are “effluent limitation guidelines” promulgated solely
pursuant to § 304, a section not included in the reviewable sections listed in § 509. The record, upon which the regulations are based, is steeped in technological data and the only section dealing with specific technology and factors to be considered in determining controls and practices applicable to point sources is § 304.
EPA claims that the Administrator has combined in the regulations “efflüent limitations” promulgated pursuant to § 301
and
“effluent limitation guidelines” promulgated pursuant to § 304 and that Congress intended that EPA promulgate § 301 “effluent limitations” by regulation before the permit-granting procedure is initiated. The Administrator promulgated § 304 “effluent limitation guidelines” which assertedly define in concrete terms the abstract norms of § 301 “effluent limitations”. These definitions are then incorporated into § 301 “effluent limitations” and the limitations are then incorporated into permits by permit-grantors. No express statutory authority supports this conclusion except that the effluent limitations of § 301 are to be “as defined by the Administrator pursuant to section 304(b).” On this fragile nexus of § 304 to § 301 to § 509, jurisdiction in the Court of Appeals is asserted to rest.
Seeking guidelines of our own from the other Courts of Appeals which have had to cope with this problem of jurisdiction, a brief review of recent decisions from the Third, Fourth, Seventh and Eighth Circuits is a useful introduction to our discussion.
In
CPC International, Inc. v. Train,
515 F.2d 1032 (8th Cir. 1975), the Eighth Circuit concluded that “the guidelines [§ 304] for existing plants cannot be directly reviewed”
and accordingly dismissed the petitions for review. The Court buttressed its opinion by “legislative history [which] confirms that Congress intended to enforce uniformity of conditions for existing plants, not by authorizing the promulgation of regulations under § 301, but by granting the EPA power to issue permits and to veto state-issued permits which do not comply with guidelines promulgated under § 304(b).”
Next came the Third Circuit with
American Iron and Steel Institute v. Environmental Protection Agency,
526 F.2d 1027 (3rd Cir. 1975) (November 7, 1975). There, a group of steel companies sought to review regulations entitled “Effluent Guidelines and Standards, Iron and Steel Manufacturing Point Source Category”
promulgated by the Administrator as effluent limitations. The Court acknowledged that the “first, and most basic, challenge is to the Administrator’s very power to promulgate nationwide single number effluent limitations for existing point sources”
and found that the answer “which goes to the very heart of the administration of the Act, depends upon our resolution of the interrelationship of three key sections of the Act— 301, 304 and 402.”
The petitioners there had contended (as do Petitioners here) that limitations could only be established by the permit-granting authorities which would apply § 304 guidelines in issuing permits on a plant-by-plant basis under § 402. Although the Court recognized that § 301(b) “does not explicitly authorize the Administrator (or anyone else) to promulgate regulations establishing such limitations”,
it concluded that in light of the statutory scheme and legislative history such power could be “inferred”.
This inference was strengthened, in the Court’s opinion, by the “repeated reference to such limitations” as being “under” or “pursuant” to § 301.
Accordingly, the Third Circuit disagreed with the Eighth’s conclusion in
CPC Inter
national
and held that EPA did have the power to issue limitations under § 301. The Court concluded that “a joint reading of sections 304 and 301 lends support for the Administrator’s position”
and accepted § 509 initial jurisdiction.
A few days later (November 24,1975) the Seventh Circuit in
American Meat Institute v. Environmental Protection Agency,
526 F.2d 442 (7th Cir., 1975), solved the problem of limitations or guidelines by flatly declaring at the outset “This is a review of effluent limitations promulgated by the Administrator . . . ”
The parties agreed that the Court had jurisdiction but
amici
attacked the Administrator’s authority under § 301 to issue limitations. As in
CPC International
and
American Iron and Steel,
it was argued that individual effluent limitations were only to be established by the permit-issuing process of § 402, using § 304(b) guidelines as the criteria.
The Court rejected
CPC International
and held it had authority to review the regulations under § 509
The Court was influenced, as was the Third Circuit, by the many statutory references to § 301 and by EPA’s statement in preambles to both proposed and final regulations that they were promulgated “pursuant to § 301 and § 304(b)”.
The Court believed that EPA’s position was more consistent with Congressional intent and concluded that EPA “had the authority to issue effluent limitations under § 301”
and, by implication, that the regulations were both § 304 guidelines and the § 301 “limitations”.
On December 30,1975, the Fourth Circuit decided
E. I. DuPont de Nemours, Inc. v. Train,
528 F.2d 1136 (4th Cir. 1975),
cert. granted,
- U.S. -, 96 S.Ct. 1662, 48 L.Ed.2d 174, 44 U.S.L.W. 3592 (1976)
(de Nemours I).
The District Court had held that it lacked jurisdiction to hear a petition to review regulations purporting to establish effluent limitations. On an appeal limited to the jurisdictional issue, the Administrator claimed that he had “combined his rulemaking authority under this section [§ 301] with that specifically provided for under § 304(b)”.
The petitioners took the position that the regulations could not have been issued under § 301 or a combination of § 301 and § 304(b), but only under § 304(b). They concluded that since the Act did not confer review jurisdiction of § 304 regulations on the Court of Appeals, the regulations could only be reviewed initially in the District Court. The Fourth Circuit affirmed the District Court’s rejection of petitioners’ arguments. It held that Courts of Appeals have exclusive jurisdiction to review the regulations.
The court was influenced by the fact that in the same section which mentioned § 301 effluent limitations, the Act specifically authorized initial Court of Appeals review of § 306 (new source) regulations.
Therefore, on an
ejusdem generis
basis, it concluded that § 301, concerned with existing sources, should be treated similarly and that review of regulations relating to existing sources should be in the same court as new source regulations. To attain this result, the Court concluded that “any action taken by the Administrator under § 304(b) should properly be considered to be pursuant to the provisions of § 301 and therefore, reviewable by this court under § 509.”
Subsequently, the same Court, although primarily dealing with specific regulations, well summarized the jurisdictional problems presented for resolution in saying: “The conflict among the circuits emphasizes the
confusion caused by this poorly drafted and astonishingly imprecise statute.”
To this criticism may well be added the comments of Judge Adams in his concurring opinion in
American Iron and Steel Institute:
“The failure to provide a clear procedural statutory structure on so basic a matter in the administration of the Act is disquieting. * * * Under these circumstances, neither the states nor the Administrator, nor the industries and municipalities which are to be regulated by the Act, may be confident which agency possesses the legal authority to promulgate effluent limitations.”
Judge Adams’ point is well taken. The jurisdiction question and its subsidiary issues admit of no easy answer. The Act states neither that effluent limitations are to be promulgated in permits nor that they are to be promulgated independently by regulations. Indeed, the Act is “vague, uncertain and inconsistent.”
The legislative history compounds the difficulty. In their briefs, the parties have relied on a two volume, 1766 page work entitled “Legislative History of the Water Pollution Control Act Amendments of 1972” (“Leg. Hist.”), which, as the title implies, sets forth the Congressional debate. From this history may be deduced equally persuasive arguments that on the one hand Congress rejected plant-by-plant proceedings; and that on the other, Congress envisioned limitations adjusted to each specific plant. In short “statements can be found to uphold almost any position which one cares to take.”
No doubt this problem of statutory interpretation was unavoidable under the circumstances. The very magnitude of the task undertaken by Congress and delegated to the EPA for fulfillment probably accounts for the lack of clarity. EPA’s task was not aided by a Federal court decision in a suit brought by NRDC which directed that EPA complete the guidelines by a final deadline.
Natural Resources Defense Council v. Train,
6 ERC 1033 (D.D.C.1973),
aff’d in part,
166 U.S.App.D.C. 312, 510 F.2d 692 (1975). The force of the adage “haste makes waste” is not lessened by that direction. But nonetheless, laws should be written so that those burdened by compliance therewith will know that with which they must comply, and the difficulty of the question posed unfortunately does not relieve the courts of the obligation to endeav- or to answer it.
Against this background of puzzling statutory language, ambiguous legislative history and conflicting court decisions, we must decide whether the contested regulations were promulgated pursuant to §§ 301 and 304 or were promulgated exclusively pursuant to § 304. Since the statutory language is devoid of plain meaning, we turn briefly to the legislative history of the Act and the underlying Congressional purpose to be gleaned therefrom. That history supports the conclusion that effluent limitations were to be promulgated apart from the permit-granting process.
Senator Bentsen, a member of the Committee which reported out the original Senate version of the Act, stated during the Senate debate:
“In phase I, for point sources of pollutants, effluent limits shall be established not later than January 1, 1976 [now July 1, 1977], which comply with specifically defined levels of effluent control and treatment. As defined in section 301(b)(1) of the bill,
and as elaborated in the regulations which we anticipate the Administrator shall issue pursuant to section 301 and 304,
these . . goals shall be at least . . . the ‘best practicable control technology currently available’ for point sources. . . . ”
Leg. Hist,
at 1283. (emphasis added)
Further support is found in the House of Representatives’ Committee on Public
Works’ report. In discussing the original House version of the Act, it states:
“The bill establishes a Federal-State discharge permit program. All permits issued under this program shall be consistent with the specific requirements of the bill,
including effluent limitations or
other limitations, national standards of performance, toxic and pretreatment standards, and ocean discharge guidelines”.
Leg. Hist,
at 761. (emphasis added)
In discussing the Senate version of the Act the Senate Report reiterated:
“It is the Committee’s intention that pursuant to subsection
301(b)(1)(A) and Section 304(b)
the Administrator will interpret the term ‘best practicable’ when applied to various
categories of industries
as a basis for specifying
clear and precise effluent limitations. . .
. ”
Leg. Hist,
at 1468. (emphasis added).
After examining the Conference Report, members of both the House and Senate were satisfied that the Conference version of the Act conformed to this scheme. During Senate consideration of the Conference Report, Senator Muskie, the Act’s principal author, explained:
“It is the intention that pursuant to subsection 301(b)(1)(A)
and
Section 304(b), the Administrator will interpret the term ‘best practicable’ when applied to
various categories of industries as a basis
for specifying clear and precise
effluent limitations
to be implemented by July 1, 1977.”
Leg. Hist,
at 169. (emphasis added)
See, also,
comments of Senator Bayh,
Leg. Hist,
at 216. The Conference Report itself states:
“The conferees intend that the Administrator or the State, as the case may be, will make the determination of the economic impact of an effluent limitation
on the basis of classes and categories
of point sources,
as distinguished from a plant by plant determination.” Leg. Hist,
at 304.
Representative Clausen in commenting upon the utility of the studies mandated by subsection 104(t) observed:
“The Administrator should consider the results of these studies in
promulgating regulations
not only under Section 316 but also under other sections of the act where thermal discharges may be regulated, including section
301 on effluent limitations. . .
. ”
Leg. Hist,
at 264 (emphasis added).
Representative Dingell stated:
“[A] plant-by-plant determination of the economic impact of an effluent limitation is neither expected, nor desired, and, in fact, it should be avoided.”
Leg. Hist,
at 255.
These citations suffice to demonstrate that the draftsmen of the Act intended the promulgation of effluent limitations by regulations apart from the permit-granting process. Furthermore, the Administrator is authorized “to prescribe such regulations as are necessary to carry out his functions under this Act.”
We are persuaded that he did so pursuant to §§ 301 and 304 of the Act. Thus, we believe that we are adhering to Congressional purpose and intent in holding that we have jurisdiction to review these regulations under § 509.
Section 304 structures the procedure and identifies the criteria for consideration whereby the Administrator in the context of industrial categories of existing effluent dischargers gives concrete definitional content to effluent limitations mandated by § 301. As we have previously observed, § 301(b)(1)(A) mandates for 1977 the achievement of effluent limitations “which shall require the application of the best practicable control technology currently available
as defined by the Administrator pursuant to section 304(b).”
Section 304(b)(1)(B) outlines the factors to be taken into account by the Administrator “relating to the assessment of best practicable control technology [1977] currently available to comply with subsection (b)(1) of section 301 of this Act . . . .” It requires that,
inter alia,
the age of the equipment and
facilities involved, the process employed, and “such other factors as
the Administrator
deems appropriate” shall be included in the “consideration”, i. e., determination, of what constitutes the “best practicable technology”. After assessing each of these factors in the context of every particular industrial category, the Administrator translates the standard “best practicable control technology” into objective terms, and determines “the degree of effluent reduction attainable through the application of [that technology].” The ultimate result is synonymous with § 301’s mandated effluent limitations. For 1983 effluent limitations requiring the application of “best available technology economically achievable” the legislative history indicates that an identical relationship and procedure obtains.
Petitioners’ challenges are aimed at the regulations taken as a whole and individually. An appreciation of these claims requires some familiarity with the steps followed in the promulgating process and the resulting administrative record.
GENERAL METHODOLOGY EMPLOYED BY EPA
Prior to the publication of the Advance Notice of Public Review Procedures, the EPA hired General Technologies Corporation, a contractor familiar with the phosphate industry, to survey the technological feasibility and economic cost of eliminating water pollution in the industry. General Technologies prepared a draft which included tentative recommendations of effluent limitation guidelines and “new source” regulations. Copies of the report were mailed to interested institutions for comment. On August 6, 1973, EPA issued its Advance Notice of Public Review Procedures which,
inter alia,
informed the general public that the draft report was available for inspection and that EPA planned independently to examine the industry.
On September 7, 1973, EPA published a notice that it was proposing effluent guidelines for the phosphate manufacturing industry pursuant to several sections of the Act, including §§ 301 and 304(b) and (c). For legal justification of EPA action concerning existing point sources, the notice specifically referred to § 301(b) as requiring achievement of effluent limitations by the respective July 1, 1977 and July 1, 1983 dates “as determined in accordance with regulations issued by the Administrator pursuant to section 304(b) of the Act.” 38 Fed.Reg. 24461-70 (1973). Shortly after the notice of proposed rulemaking, EPA issued a Draft Development Document which disclosed its “general methodology” and recited its proposed findings and conclusions.
After reviewing extensive comments from industry members, including some of the Petitioners in this case, who recommended alterations and amendments, on February 20, 1974, the EPA published its regulations in final form. The notice asserted that §§ 301, 304(b) and (c), 306(b) and (e), and 307(c) constituted the source of statutory authority.
A Final Development Document dated January 1974 was prepared by General Technologies in cooperation with EPA staff and interested chemical companies including some of the Petitioners. The document is a comprehensive 154-page report entitled “Effluent Guidelines and Standards 39 F.R. No. 35 Part 422 Effluent Limitations Guidelines”. The final record supplement is a 50-page report entitled “Economic Analysis of the Proposed Effluent Guidelines — Industrial Phosphate Category” prepared by Arthur D. Little, Inc. A final “Economic Analysis of the Effluent Guidelines” has issued, dated November, 1974.
It is part of the Appendix, but none of the parties rely on it.
COMPLIANCE WITH ADMINISTRATIVE PROCEDURE ACT
Petitioners argue that EPA violated § 4(b) and (c) of the APA since the published notices and regulations assertedly omitted a “reference to the legal authority under which the rule is proposed”, and “a
concise general statement of their basis and purpose.”
We disagree. Although the notice of proposed rulemaking and publication of final regulations might be argued to have self-serving characteristics, they adequately referred to and summarily sketched EPA’s reliance upon § 301 and that section’s interrelationship with § 304.
See
38 F.R. 24470 (1973). Furthermore, the preamble to the regulations succinctly reviews the history of the proceedings, refers to the records of EPA’s methodology and fact findings, cites the regulations’ legal basis, and then outlines both the significant comments received and the EPA’s. response. 39 F.R. 6580 (1974).
Accord, E. I. DuPont de Nemours v. Train,
541 F.2d 1018, at 1025 (4th Cir., 1976)
(“de Nemours II”).
ALLEGED FAILURE TO SPECIFY FACTORS
Petitioners next assert that the regulations generally violate § 304(b)(l)(B)’s purported requirement that the regulations specify factors to be taken into account in determining the control measures and practices to be applied to point sources. In the preamble to each of the individual regulations, the EPA states that in establishing them, it considered information concerning,
inter alia,
age and size of plant, products produced, and available treatment technology. Although Petitioners may quarrel with EPA’s conclusions and rationale for the regulations, the factors considered appear in EPA’s own statement that:
“In identifying such technologies, various factors were considered. These included the total cost of application of technology in relation to the effluent reduction benefits to be achieved from such application, the age of equipment and facilities involved, the process employed, the engineering aspects of the application of various types of control techniques, process changes, non-water quality environmental impact . . . and other factors.
The data upon which the above analysis was performed included EPA permit applications, EPA sampling and inspections, consultant reports, and industry submissions.” 38 F.R. 24471 (1973)
ESTABLISHMENT OF RANGES
The Petitioners next argue that the entire set of regulations is invalid because EPA failed to establish permissible “ranges” of discharges within each category or subcategory of existing point sources, and, instead, improperly promulgated single number maximum discharge levels. We disagree with this argument and believe that whenever Congress spoke of “ranges” in the debates over the Act, it meant only the spectrum comprised of varying discharge levels on a subcategorical, rather than individual, basis,
de Nemours II, supra
at 1029. Although variances are conceivable at the permit-granting stage (see our accompanying opinion
Natural Resources Defense Council v. Environmental Protection Agency, 2
Cir., 537 F.2d 642), Congress intended that the regulations establish a single discharge level for a given subcategory. This is implicit in the Congressional choice of the superlative form in the statutory language requiring achievement of the degree of effluent reduction attainable by application of
“best”
technology-
SPECIFIC REGULATIONS
(a)
The “Substantive” Standard of Review
Under § 10(e) of the Administrative Procedure Act, 5 U.S.C. § 706(2)(A), agency action in rulemaking proceedings is to be sustained unless “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law”. This standard was recently explained in
American Meat Institute,
as follows:
“This standard requires us to determine whether ‘the decision was based on a consideration of the relevant factors and whether there has been a clear error of judgment.’ [citation omitted] We are not to set the effluent limitations ourselves or substitute our judgment for the agency’s, [citations omitted] Rather, we are to determine whether the limitations set by the agency are ‘the result of rea
soned decisionmaking.’ [citations omitted] If the basis stated by the agency for its decision is insufficient, we may not supply another that the agency itself has not chosen to rely on. [citations omitted]”
American Meat Institute, supra
at 452-53.
Mindful of these requirements, we may proceed to an analysis of the individually challenged regulations.
(b)
1977. Effluent Limitations Guidelines for Phosphorus Manufacturers
For phosphorus manufacturers, the EPA promulgated 1977 limitations which restrict the discharge of suspended solids, elemental phosphorus, fluoride and total phosphorus (“phosphates”). 40 C.F.R. § 422.12. Petitioners have confined their challenge of the regulation to the fluoride and phosphates restrictions.
To properly evaluate their contentions, it is necessary to trace the development of these regulations.
When the 1977 regulation for phosphorus manufacturers was first proposed, it completely prohibited discharges of phosphates and fluorides. During the comment period, EPA received a report from the Effluent Standards and Water Quality Information Advisory Committee (“ESWQIAC”)
in which ESWQIAC objected to the zero discharge limitation. It argued that, of the two plants relied upon by EPA, plant no. 181 and TVA, the latter did not achieve zero discharge of phosphate, or fluoride. Consequently, ESWQIAC reasoned that the zero discharge standard was supported by only one plant, Hooker’s no. 181, and that such a limited data base was unreasonable.
As an alternative to the zero discharge standard, the advisory committee suggested a permissible range of discharge for each effluent.
Other parties voiced analogous objections to the zero discharge standard for the effluents. The upshot was a revised standard based upon data from plant no. 181 and two other phosphorus-producing plants, nos. 028 and 159, whose data had not been considered when the standards had first been proposed.
The following table depicts for comparison the discharge rates of Fluorides and Phosphates of the three plants, the average discharge rate of the three plants and the discharge rate as promulgated by the EPA in the final regulation for phosphorus manufacturers for 1977. 40 C.F.R. § 422.12.
Final Reg.
Hooker Monsanto Monsanto Average Standard
Plant 181 Plant 028 Plant 159 Discharge Discharge
Fluorides 0 0.1 0.04 0.05 0.05
Phosphates 0 0.039 0.26 0.10 0.15 (as Phosphorus)
(Data expressed in terms of pounds of pollutants per 1000 pounds of phosphorus)
As is apparent from the table, EPA eschewed ESWQIAC’s recommended range and adhered to its single number effluent limit principle for phosphate and fluoride effluent. In the preamble to the final regulations, the EPA stated:
“The ESWQIAC limits include two
additional
phosphorus plants as exemplary.
EPA has since accepted these plants as exhibiting best practicable control technology and has allowed a discharge based upon the data in the Development Document for the treatment capabilities of these plants. Therefore, although the Agency does not agree with the underlying rationale for establishing the ESWQIAC limits the data in the Development Document does support the specific limits proposed by ESWQIAC.” 39 F.R. 6581 (1974). (emphasis added)
Petitioners raise several objections to the final regulation. First, they assert that in determining the “effluent reduction attainable” and the “best practicable technology applicable”, EPA impermissibly narrowed consideration of production facilities to include only those plants determined to be exemplary. Thus, the Petitioners challenge EPA’s authority to define best practicable as:
“the average of the best existing performance by plants of various sizes, ages, and unit processes within each industrial category or subcategory. This average is not based upon a broad range of plants within an industrial category or subcategory, but is based upon performance levels achieved by exemplary plants.” 39 F.R. 6580 (1974).
Petitioners’ argument is unpersuasive. The Legislative History of the Act clearly substantiates the EPA’s definition of best practicable and its concomitant promulgating methodology. Indeed, the definition adopted by EPA almost mirrors the pertinent portion of the House Report which bases “best practicable” on “the average of the best existing performance” of various plants.
Leg.Hist.
at 1468. This definition is also confirmed by the individual report of Senator Muskie which was appended to the Conference Report.
Leg.Hist.
169-70. A consideration of plants which is confined to those that are characterized as “exemplary” is not a cognizable deviation from the norm of “best existing performance”.
American Meat Institute, supra,
at 453, 462;
American Iron & Steel, supra,
at 1057.
In the specific context of phosphorus manufacturing industries, the record reveals that the EPA made appropriate use of the data of the three “exemplary” plants.
In stating that the Agency relied on two
additional
plants in the preamble to the notice of final regulations, EPA did not suggest that it was excluding consideration of Hooker’s plant no. 181 which it had originally relied on when the regulations were first proposed. Nor do Petitioners contend that these plants should have been classified separately in discrete subcategories. The average of these three plants’ discharge of fluoride and phosphate effluent is less than or equal to the discharge limit set by the EPA for those effluents.
Petitioners complain that the exact mathematical averaging of figures is not in the record and that therefore the record does not adequately corroborate the discharge limit. However, Agency action should be upheld where the path of administrative proceedings may reasonably be discerned even if the demarcations are of less than ideal clarity.
Bowman Transportation, Inc. v. Arkansas-Best Freight System, Inc.,
419 U.S. 281, 286, 95 S.Ct. 438, 442, 42 L.Ed.2d 447, 456 (1974). In the light of the explicit statement in the Final Development Document that the three plants were being used and in light of the definition of “best practicable” which concisely expresses the Agency methodology of averaging, we believe the path of administrative proceedings was adequately set forth.
Petitioners also contend that the regulation is invalid because the limits which the regulation prescribes are not satisfied by the industrial plants on whose data the limits were predicated. It is undeniable that exemplary Monsanto plant no. 028 does not comply with the fluoride discharge limit and exemplary Monsanto plant no. 159 does not comply with the phosphate discharge limit. But this does not invalidate the regulation. The legislatively mandated methodology of averaging the best existing performances of comparable industrial plants
to determine “best practicable technology” of necessity means that those plants which discharge more effluent than the resulting
mean
may need to make some modification of their effluent treatment systems.
American Meat Institute, supra,
at 457. The fallacy of Petitioners’ argument is the assumption that inclusion of an industrial plant within the category “best existing” constitutes an implied determination that that plant’s technology is the “best practicable” and that the plants’ effluent discharge rates are the maximum achievable by 1977. We believe that the figures reached by EPA as to fluorides and phosphates are justified by the information before it.
Petitioners next argue that the regulations are invalid because it is unclear whether they apply on a “net” or “gross” basis. If applied on a “gross” basis, an industrial plant would be liable for all effluents in its process waste regardless of whether they are added during the manufacturing process or already present in the intake water. If applied on a “net” basis, the point source would only be accountable for effluent which it added to the water it uses. The defect, however, appears to have been remedied by promulgation of new regulations following the institution of this appeal. 40 F.R. 29849 (1975). The new regulations, which apply to the issuance of all permits implementing the Act, in effect, provide for application of “net” levels if the permit applicant’s water supply is the same body of water into which effluents are discharged.
Finally, Petitioners argue that the regulation is arbitrary because it makes no provision for excess rainfall or discharge after storms which will assertedly increase the discharge flow and hence the pounds of pollutants discharged. The Agency concedes this point and promises a clarifying amendment. However, we have no way of knowing whether a hypothetical amendment will adequately remedy the defect of an operative regulation, and we decline to resolve the issue solely upon the good faith of respondent.
American Iron & Steel, supra,
at 1056. Accordingly, we vacate the regulation and remand the record for this purpose and further proceedings, if necessary.
(c)
1983 Effluent Limitations Guidelines for Phosphorus Manufacturers
EPA set a 1983 regulation of “no discharge” of process waste water for phosphorus manufacturers. 40 C.F.R. § 422.13. In addition to the confessed failure to make any provision for the concededly unavoidable discharge of rainfall effluent after storms, there are several other defects associated with the regulation which require us to set it aside and remand the record to the EPA for further consideration. Since our judicial review rests in part on the premise that Agency and Court together constitute a “ ‘partnership’ in furtherance of the public interest”,
Kennecott Copper Corp. v. Environmental Protection Agency,
149 U.S.App.D.C. 231, 462 F.2d 846, 848-49 (1972), some discussion of these defects is appropriate. It would be impolitic to require the Agency to grope in the dark for the rationale compelling our disapproval.
The 1983 effluent limitation guideline was predicated on the applicability of the recycle technology employed at Hooker’s plant no. 181 which EPA claims has already achieved zero discharge of fluoride and phosphate effluent except in times of heavy rainfall.
That “technology” is basically a secondary treatment system of settling ponds. Process waste water generated in production is discharged into the first pond and slowly progresses through the other ponds until it is finally ready for reintroduction into the production process.
Petitioners’ principal objection to the regulation is that EPA failed to consider whether such recycle technology could be feasibly applied to plants in cold climates. In such climates the settling ponds would assertedly freeze, preventing recycle. They buttress the argument by referring to the record of EPA’s consideration of the 1977
phosphorus regulation wherein EPA discussed cold weather effects upon recycle technology.
There is no comparable discussion in the section of the record dealing with the 1983 counterpart, even though EPA does not suggest that cold weather will not be a problem in the foreseeable future.
EPA acknowledges that a cold climate may cause recycling difficulties, but it contends that the problems are manageable and that the cold weather analysis in the discussion of the 1977 phosphorus manufacturer regulation adequately addresses Petitioners’ complaints. In that discussion, EPA accepted the factual premise that settling ponds “may” freeze. It concluded that the simple expedient of burying water mains and heating the pumping stations would adequately protect the plumbing in the pond system. In the event that the settling ponds themselves froze, it concluded that a water supply would have to be provided to “uncouple” the process from “climatic perturbations”, that the ponds would have to be sufficiently large to prevent overflow after the introduction of the temporary auxiliary water, and that the ponds would have to provide sufficient evaporative capacity to eliminate the additional auxiliary water.
EPA also claims both that there is “available” technology which could prevent the ponds from freezing altogether and that even if the ponds could not be kept unfrozen, there exists available substitute technology which is not affected by subfreezing temperatures. However, these expedients appear for the first time in EPA’s brief. Courts have repeatedly held that this kind of
post hoc
rationalization is plainly unacceptable.
Burlington Truck Lines v. United States,
371 U.S. 156, 168-69, 83 S.Ct. 239, 245-46, 9 L.Ed.2d 207, 216-17 (1962);
Portland Cement Association v. Ruckelshaus,
158 U.S.App.D.C. 308, 486 F.2d 375, 395 (1973),
cert. denied,
417 U.S. 921, 94 S.Ct. 2628, 41 L.Ed.2d 226 (1974). Moreover, the argument that a substitute technology is “available” to insure compliance with the regulation in instances where land is unavailable, is itself a concession that pond settling, the only technology discussed in the record, is not universally feasible.
Our examination of the record also discloses that the Administrator did not take the costs of the cold weather technology adjustments into consideration. Since the record confirms that at least one of the phosphorus manufacturers is located in a cold climate area, a failure adequately to consider cold weather costs was unreasonable.
American Meat Institute, supra,
at 462-63. Accordingly, the regulation is remanded to the EPA for further proceedings at which time we expect that these issues will be considered.
We are mindful that courts have been reluctant to set aside 1983 effluent limitations which are predicated on technology that is still in the development stage.
See Tanners’ Council of America, Inc. v. Train,
541 F.2d 1018, at 1027 (4th Cir. 1976). Since technological innovations may transform what at present appears to. be fantastic expectations into practical realities for 1983, such trepidation is understandable. No one has the advantage of being able to gaze into a crystal ball which reveals the inventive genius and engineering ability of our scientists. To prophesy the state of the art during the forthcoming years would be pure speculation. The only assumption that can be made with assurance is that the art will not remain static.
For example, in its Economic Analysis report, Arthur D. Little, Inc. hazarded the predication that
“Apart from this factor [possible shutdown of some STPP plants], the costs presented in the effluent guideline development document would not appear to have any significant economic impact in any of the products covered.”
In any event,
“Cost increases of this magnitude will either be absorbed, or, more likely, passed
on to consumers through price increases.”
Arthur D. Little, Inc. concluded that even if costs were “significantly higher”
“we do not expect them to cause directly any plant closings, to lead to unemployment in any of the segments examined, or to have any significant impact to communities where production facilities are located.”
These predictions were accepted by the EPA.
It would be unfair to expect the EPA to pinpoint exactly all conceivable cost ramifications of the 1988 limitations. To precisely identify the costs many years into the future is an almost impossible task.
But the absence of any practical consideration of costs is unjustifiable, and where the record leaves those who are subject to the 1983 limitations without any suggestions or specifications as to how they may attempt to comply, the 1983 limitations must be vacated.
FMC Corporation, Inc. v. Train,
541 F.2d 1018, at 1027 (4th Cir. 1976). The implementation of plans to control 1983 effluent discharge must be begun now. A record which fails to disclose a reasonable basis for belief that a new technology will be available and economically achievable is deficient,
de Nemours II, supra,
at 1032.
(d)
1977 and 1983 Effluent Limitation Guidelines for Phosphorus Pentasulfide
The phosphorus pentasulfide industry in the United States is comprised of five plants. Of these, two were examined by EPA: Hooker Chemical Company’s plants in Niagara Falls, New York, and Columbus, Mississippi. A major effluent source at these and presumably all plants which produce phosphorus pentasulfide is the water used in “wet scrubbers” to reduce noxious fume emissions during the “casting” of phosphorus pentasulfide. Casting is the process whereby molten phosphorus pentasulfide is poured into containers to solidify. The surface of the molten phosphorus pentasulfide mass ignites on contact with the air, producing gaseous phosphorus pentoxide and sulfur dioxide. The investigation of the two plants revealed that little was being done to treat the waste water which resulted from scrubbing these fumes.
(i)
1983 Effluent Limitations Guidelines
EPA concluded that the application of the “best available technology economically achievable” would eliminate effluent discharge, and it announced a standard of no discharge of process waste water for 1983 effluent limitation guidelines applicable to the industry. 40 C.F.R. § 422.22. It determined that either of two alternative technologies are available to totally eliminate the discharge of scrubber water. The first is inert or vacuum casting, during which process molten phosphorus pentasulfide is not exposed to the oxygen in the air, thereby preventing fume production and eliminating the need to use scrubbing water. Inert casting is not presently in use in the industry and its installation would require relatively expensive revisions of the casting process. The second method involves the combined employment of (1) dilute caustic or lime slurry in the scrubber system; (2) partial recycle of scrubber liqúor; (3) lime treatment and sedimentation permitting total recycle of the process effluent.
Petitioners challenge the 1983 phosphorus pentasulfide effluent limitations arguing that there is no basis in the record to conclude that either of these technologies is “available”. They emphasize that neither of the plants inspected use either process and note that there is no citation in the record to extraneous literature which would support the EPA’s conclusion that they could be applied to eliminate process waste water. They also criticize the report for failing to discuss or consider the problem of sediment precipitation endemic in recycle systems which assertedly renders total recycle of effluent water unfeasible. Since neither technology has been shown to be
“available”, petitioners conclude that it was contrary to law for the EPA to predicate the 1983 standard on their application.
That no plant in a given industry has adopted a pollution control device which could be installed does not mean that that device is not “available”. Congress did not intend to permit continuance of pollution by industries which have failed to cope with and attempt to solve the problem of polluted water. Section 304(b)(2)(A) itself explicitly mandates consideration by the Administrator of “treatment techniques, process and procedure
innovations”
(emphasis added) in determining the degree of effluent reduction attainable by the application of the best control measures and practices. The Legislative History supports the plain meaning of this language.
See, e. g., Leg. Hist.,
p. 170.
But even if technology which is not presently in use can be treated as available and achievable, there must be some indication in the administrative record of the reasons for concluding that such technology is feasible and may reasonably be expected to yield the effluent reduction mandated when applied to the particular industry.
American Meat Institute, supra,
at 463;
de Nemours II, supra,
at 30(c). Agencies must present sufficient support for their conclusions in order that the Court can properly review their decisions.
FMC Corp., supra,
at 1027, n. 16. The Court and the parties must not be left with
post hoc
rationalization by counsel as prime authority for agency decision.
Hawaiian Telephone Co. v. Federal Communications Commission,
162 U.S.App.D.C. 229, 498 F.2d 771, 777 (1974). Judged in that light, we conclude that the EPA has failed to adequately identify the facts underlying and explain the reasons for its conclusion that full recycle and inert casting are available technologies which could completely eliminate waste water discharge.
There is nothing in the record which explains how or whether inert casting could be applied in the industrial process. The EPA relies heavily on citations to R. E. Kirk and D. F. Othmer,
Encyclopedia of Chemical Technology Interscience,
N.Y., 2d ed. (1966), as authority that “inert casting” could be successfully implemented. But the record merely contained a generalized reference to the work as a whole and did not specify what sections, if any, of the
Encyclopedia
were being relied upon. This will not suffice.
Portland Cement Association, 'supra,
at 400. Such wholesale incorporation of scientific treatises gives no indication of how the Agency reached its conclusion and encourages too generalized administrative analysis at the expense of reasoned articulation by providing a readily accessible springboard for
post hoc
rationalization. If an agency is going to rely on scientific literature it should specify just what is being employed so that the Court and others may trace its reasoning and proposed application.
As it stands now, the record is devoid of any explanation of how inert or vacuum casting could be adopted to the pentasulfide industry given the distinctions between that industry and the other chemical industries in which that technique is presently employed. The EPA points to the observation that vacuum
distillation
is currently employed in the phosphorus pentasulfide production process, but neglects to say how or whether “casting” is sufficiently similar to distillation to permit technology transfer of vacuum process to casting.
As to “total recycle” of waste water, we again agree with Petitioners that the EPA has failed to adequately explain how the problem of solid salt precipitation in the water recycle system could be avoided without having to discharge periodically the waste water and replace it with fresh water.
Cf. American Iron & Steel, supra,
at 1062. The EPA nowhere responds to the statement by its hired consultant that:
“total recycle probably cannot be carried out or approached in present equipment. The sulfate-sulfite lime system, once concentrations build up, requires special
scrubbing fluid handling techniques to keep the severe scaling under reasonable control.”
Although this statement conflicts with the Agency’s conclusion, no explanation for its rejection is proffered. Nor is there any specification of the available technology which supercedes the performance of “present” equipment. The EPA belatedly argues that dry dust collection technology could be used to trap sulfur dioxide gas in the casting process, thereby preventing scaling in the water recycle system by eliminating from the water flow the crucial ingredient for the formation of sulfite and sulfate precipitates. Since this peculiar use of dry dust technology was never considered at the rulemaking stage
we can only conclude that it, too, is an impermissible
post hoc
rationalization. The administrative record contains nothing which even arguably supports this EPA assertion or begins to explain how technology used to trap dust is “available” to collect gas.
In addition to these fatal defects, the record discloses that the Agency completely neglected the problem of eliminating the discharge of waste water other than that produced in the “casting” part of the production process. The EPA virtually concedes this point when it says that such waste water will be dealt with separately after redefinition of “process waste water”. Although the technology to eliminate the waste water which resülts from the cleaning of storage vessels may already be in existence, it is not the proper function of the Court or adversely affected litigahts to guess what that technology may be in order to substantiate administrative action.
We hold that the 1983 effluent limitation guideline for phosphorus pentasulfide point-sources must be set aside and that further procéedings must be initiated to formulate adequate regulations. This is not to say whether a limitation of “no discharge” for such point source is unattainable, but if the EPA chooses to ¿dhere to that standard, it-must articulate its rationale in such fashion as will permit meaningful judicial review of its action.
(ii)
1977 Effluent Limitation Guideline
EPA adopted a “no discharge” limitation for 1977 based on the application of the same technology which it assertedly relied upon in the formation of the 1983 effluent limitation. 40 C.F.R. § 422.23. Petitioners challenge the validity of the guideline on the ground that whatever its relevance may be to the 1983 guideline, that technology should not have been considered by the Agency in formulating a 1977 guideline. Noting that 1977 guidelines are to be based oh the application of the best practicable control technology, Petitioners cite legislative history for the proposition that technology affecting the manufacturing process itself was not encompassed within the term “best practicable”. Since inert or vacuum casing and total recycle are assertedly “in process” changes, Petitioners conclude that they should not have been considered applicable to point sources for achieving effluent reductions by 1977. EPA candidly concedes the merit of Petitioners’ argument and promises amending regulations. However, in construing such ‘in process’ changes, if in-plant controls, which do not represent changes in the basic manufacturing process and-, which have been carefully considered in light of‘the pertinent factors applicable thereto, may by simple modifications, enable the phosphorus pentasulfide industry to achieve the objectives of thé water pollution legislation with respect to compliance by July 1, 1977, EPA should not be foreclosed from promulgating regulations so long as they are not unattainable by any. technology known • today; but no ‘in process’ changes can be mandated for 1977 unless they may be considered normal practices within the industry.
Accordingly, we set aside the present regulation and remand the record to the EPA • for further proceedings.
See, FMC Corp.,
supra,
at -. But, cf.
de Nemours II, supra,
at --.
(e)
1983 Effluent Limitation Guidelines For Phosphorus Trichloride and Phosphorus Oxychloride
The principal source of fluid effluent in an industrial plant producing phosphorus trichloride or phosphorus oxychloride is produced when water is used to “scrub” phosphorus trichloride or phosphorus oxychloride vapors from reaction and distillation vessels. The 1983 effluent limitation guidelines established by EPA for these plants based on the “best available” technology is “no discharge of process waste water pollutants”. 40 C.F.R. § 422.-23. EPA believed that the installation of refrigerated condensors would reduce vapor volume 90% and that a corresponding reduction of needed scrubber water could be expected. EPA determined that the remaining 10% waste water flow could , be eliminated by evaporation.
Petitioners question EPA’s basis for concluding that evaporation is an “available” technology and also challenges EPA’s conclusion that refrigeration would reduce vapor volume by 90%. It further contends that even if vapor volume could be reduced by 90% there is no basis for EPA’s assumption that a proportionate reduction of waste water flow would follow. Petitioners also argue that EPA either miscalculated or ignored the energy and cost factors which it was obligated to consider in developing the regulations. Finally, it contends that solid wastes produced by the fuel which would have to be consumed to satisfy anticipated technology’s energy requirements would itself constitute a pollution problem.
While contending that a basis for the regulation exists in the record and that the standard of no discharge is achievable, EPA acknowledges the Petitioners’ criticism and the need for revision. Accordingly, we set aside the 1983 effluent limitation guidelines for phosphorus trichloride and phosphorus oxychloride and remand it to the EPA for specific consideration of the Petitioners’ comments.
(f)
1977 and 1983 Effluent Limitation Guidelines For Food Grade Sodium Tripolyphosphate; 1983 Effluent Limitation Guidelines For Food Grade Calcium Phosphate
Our discussion of the regulations for Food Grade Sodium Tripolyphosphate, 40 C.F.R. §§ 422.32; 422.33, and for Food Grade Calcium Phosphate, 40 C.F.R. § 422.-33, will be mercifully brief. In light of the EPA’s candid admission that Petitioners’ arguments expose the lack of a sufficient data base to support the regulations, it is unnecessary to parade the full regalia of Petitioners’ challenges and we simply set aside the regulation. We expect that in subsequent proceedings, the Petitioners will have the opportunity to urge their position and that the EPA will respond accordingly.
(g)
Definition of Process Waste Water
Excepting the 1977 effluent limitation guideline for the Phosphorous Production Subcategory, 40 C.F.R. § 422.12 (1975), each of the effluent limitation guidelines for the phosphate industry governs the discharge of “process waste water” pollutants or discharge of pollutants in “process waste waters”. The term “process waste water” is defined to be:
“any water which, during manufacturing or processing, comes into direct contact with or results from the production or use of any raw material, intermediate product, finished product, by-product or waste product.” 40 C.F.R. §§ 422.11, 422.21 and 422.31, incorporating 40 C.F.R. § 401.-ll(q).
Petitioners challenge this definition as being arbitrary on several grounds. First, they argue that, as it stands now, the definition encompasses leaks and spills which are ineluctable in industrial chemical operations. The inclusion of leaks and spills in the definition of process waste water renders unattainable compliance with permits incorporating effluent guidelines that prohibit all discharge of process waste water. For the same reason, Petitioners claim that the “process waste water” definition is arbitrary because it includes rainwater runoff
which is inevitably contaminated by chemical dust which obviously cannot be controlled by even the most stringent air quality dust control programs.
EPA responds that although the statute may be read as including rainwater runoff and leaks and spills within the definition of “process waste water”, the Agency did not intend to create an overly broad regulation. In light of this concession, we hold that regulation defining process waste water must be set aside and remanded to EPA for further consideration.
See, de Nemours II, supra,
at 1032.
CONCLUSION
Any legislation as innovative as the Act cannot in its initial enactment be expected to cover all possible contingencies or situations either existing or which may arise in the future. Undoubtedly amendments will be required to provide for such situations; and regulations will have to be modified. Much will depend on the reasonableness and the flexibility of those responsible for the administration of the law and regulations. But until those contingencies arise, Cássandra-like prophecies of doom are premature. Despite the fact that we have to sail on unchartered waters (hopefully to become non-polluted in the near future), it is too soon to indulge gloomy fears that industrial plants will have to close; that thousands of employees will be thrown out of work, or that the air will be polluted by fumes of the fúel necessary to drive water pollution control equipment. The spectre of ghost towns which were once prosperous communities is, with proper administrative and court decisions unlikely to become a reality. On remand of these regulations the EPA will have án opportunity to elaborate more fully, if necessary, on such factors, specified in § 304, as may be relevant. For the moment, it would seem that the EPA has endeavored to carry out its assigned functions under difficult circumstances, and there is no reason to believe that it will not continue to do so with the clarifications suggested herein.