This case is part of a dispute over whether it is lawful for the United States Environmental Protection Agency ("EPA") to regulate internal waters which are an integral part of the waste treatment system of the Texas Municipal Power Agency ("TMPA"). The EPA has been doing so for almost nine years. We hold that it may continue for at least two more years.

TMPA, an electric utility, operates a lignite-fired steam generating plant next to the Gibbons Creek Reservoir in Grimes County, Texas. As part of the waste treatment system for the plant, waste water flows through a series of settling ponds and, at the end of this series, discharges into the Gibbons Creek Reservoir. The purpose of the ponds is to allow suspended solids to settle out of the water before it flows into the reservoir. Most of the flow into the settling ponds consists of ash-laden water from TMPA's generating plant. But in addition, at a point known as outfall 301, treated wastewater flows into the ponds from a sewage treatment facility serving the 300 employees at the plant.

Under the Clean Water Act ("CWA"), the EPA regulates TMPA's waste treatment through the terms it includes in the National Pollution Discharge and Elimination System ("NPDES") permit TMPA requires for discharging waste into the Gibbons Creek Reservoir.1 Each NPDES permit runs for a fixed term not exceeding five years.2

Since November 1977, several NPDES permits have regulated TMPA's discharges into the Gibbons Creek Reservoir as well as its discharges into the settling ponds at outfall 301. The EPA last revised its regulations on outfall 301 in March 1984, issuing the current permit, set to expire in April 1989. All this passed without objection from TMPA.

In July 1985, however, over a year after the EPA renewed the relevant permit, TMPA petitioned this court for review of the restrictions for outfall 301, contending that the EPA should have eliminated them. In Texas Municipal Power Agency v. EPA ("TMPA I "), we dismissed this petition because it was not timely filed.3

TMPA then applied to the EPA under CWA section 402,4 asking the agency to modify the permit to delete the restrictions on outfall 301. The EPA denied the modification in November 1986. Here, TMPA timely petitions for review of that denial and for the second time brings the dispute before this court.5

In TMPA I, we held that TMPA has lost the right to a full review of the permit. In this case, unlike TMPA I, we reach the merits of TMPA's complaint against the EPA. But because this petition is for review of a modification denied by the EPA, our review is more restricted than review of EPA renewal or issuance of an NPDES permit.

Section 701 of the Administrative Procedure Act provides that the action of "authority of the Government of the United States" is subject to judicial review except where there is a statutory prohibition on review or where "agency action is committed to agency discretion by law".6 In the often-cited case of Abbott Laboratories v. Gardner, the Supreme Court said that "judicial review of a final agency action ... will not be cut off unless there is persuasive reason to believe that such was the purpose of Congress".7 In TMPA I, however, we found just such a Congressional purpose in section 509 of the CWA. Our reasoning in that case provides the starting point for our analysis in this one:

Section 509(b)(1) of the Clean Water Act states:

Review of the Administrator's action ... in issuing or denying any permit under section 1342 of this title [NPDES], may be had by any interested person in the Circuit Court of Appeals of the United States for the Federal judicial district in which such person resides or transacts such business upon application by such person. Any such application shall be made within days from the date of such determination, approval, promulgation, issuance or denial, or after such date only if such application is based solely on grounds which arose after such [120th] day." 33 U.S.C. Sec. 1369(b)(1) (emphasis added).

Statutory time limits on petitions for review of agency actions are jurisdictional in nature such that if the challenge is brought after the statutory time limit, we are powerless to review the agency's action.... These time limitations impart finality to the administrative process, thus conserving administrative resources.... The requirements show a congressional decision to impose statutory finality on agency actions that we, as a court, may not second-guess.... The statutory time limitations have been strictly enforced.8

We held that Congress intended that anyone wishing to challenge the terms of an NPDES permit must do so within the period prescribed by section 509(b)(1) or " 'lose forever the right to do so, even though that action might eventually result in the imposition of severe civil or criminal penalties' ".9 9] The rule is "now or never".10 If no timely challenge is filed, the permit stands until its expiration--at most, five years.

It follows from this scheme that Congress did not intend petitions for modification to provide a second chance for full review of an NPDES permit after the statute of limitation has run from the issue or renewal of the permit. This conclusion is buttressed by the laws and legislative history concerning petitions for modification. CWA section 402 gives the EPA discretion to modify permits "for cause".11 Section 402 gives three examples of causes for modification; all are reasons to tighten limitations on effluents.12 EPA regulations appearing at 40 C.F.R. Sec. 122.62 list in greater detail the permissible grounds for modification "for cause". These grounds are narrower than the permissible grounds for a challenge to the issuance or renewal of an NPDES permit.13

Thus, our review here is narrow in scope. In particular, TMPA petitioned the EPA for modification on the basis of 40 C.F.R. Sec. 122.62(a)(16): "[t]o correct technical mistakes, such as errors in calculation, or mistaken interpretations of law made in determining permit conditions". Accordingly, we review the permit regulating outfall 301 for "technical mistakes" or "mistaken interpretations of law".

Further, our review of a denial of a modification is more deferential to EPA than review of an issuance of an NPDES permit. Although the denial of modification is reviewable agency action,14 it is action at one remove from the issuance of the permit and entrusted to broad EPA discretion. Congress clearly did not view modification as a vehicle which dischargers could freely use to relax or eliminate effluent limitations.15 Rather, it intended modification to give the EPA flexibility to alter permits in response to changed circumstances or to correct routine errors. The language of both CWA section 402 and 40 C.F.R. Sec. 122.62 make it clear that the EPA is not required to modify any NPDES permit.16 Rather, the EPA may do so at its discretion. In keeping with the goal of limiting challenges to the permit after the period for full review has passed, the regulations twice state that only actual modification of an NPDES permit reopens the matter.17 A mere request for modification does not. The EPA's broad discretion to accept or deny a proposed modification places a heavy burden on TMPA to show that the denial requires reversal.18

We now address whether the NPDES permit regulating outfall 301 is based upon "mistaken interpretations of law". The EPA is charged with regulating pollution of the "waters of the United States".19 To do so, it is usually necessary only that the EPA regulate direct discharges into those waters; the EPA does not generally regulate waters internal to a waste processing system. But the EPA's "internal waste stream rule", first promulgated in 1979, provides:(1) When permit effluent limitations or standards imposed at the point of discharge are impractical or infeasible, effluent limitations or standards for discharges of pollutants may be imposed on internal waste streams or cooling water streams.

(2) Limits on internal waste streams will be imposed only when [EPA notice to the polluter] sets forth the exceptional circumstances which make such limitations necessary, such as when the final discharge point is inaccessible ..., the wastes at the point of discharge are so diluted as to make monitoring impracticable, or the interferences among pollutants at the point of discharge would make detection analysis impossible.20

This rule is the legal basis for EPA regulation of outfall 301.

TMPA's principal contention is that the internal waste stream rule exceeds EPA authority. TMPA argues that internal waste streams are not "waters of the United States" and that therefore the EPA has no right to regulate discharges into them. Thus, TMPA concludes, the regulation of outfall 301 is rooted in a mistaken interpretation of law and must be modified.

Even if modification to correct mistakes of law were mandatory, we find no such limit on EPA powers. The CWA does not define "waters of the United States". Rather, the EPA promulgated 40 C.F.R. Sec. 122.2 to define the term:

Waters of the United States or waters of the U.S. means:

(a) All waters which are currently used, were used in the past, or may be susceptible to use in interstate or foreign commerce, including all waters which are subject to the ebb and flow of the tide;

(b) All interstate waters, including interstate "wetlands;"

(c) All other waters such as intrastate lakes, rivers, streams (including intermittent streams), mudflats, sandflats, "wetlands," sloughs, prairie potholes, wet meadows, playa lakes, or natural ponds the use, degredation, or destruction of which would affect or could affect interstate or foreign commerce ...;

(d) All impoundments of waters otherwise defined as waters of the United States under this definition;

(e) Tributaries of waters identified in paragraphs (a) through (d) of this definition;

(f) The territorial sea; and

(g) "Wetlands" adjacent to waters (other than waters that are themselves wetlands) identified in paragraphs (a) through (f) of this definition.21

This definition is expansive and, in keeping with the intent of Congress, the courts construe it liberally to give the broadest possible reach to EPA regulation.22 We have no doubt that its language embraces TMPA's treatment ponds.

But this definition also contains one, very relevant exception:

Waste treatment systems, including treatment ponds or lagoons designed to meet the requirements of [the] CWA ... are not waters of the United States.23

This treatment pond exception appears to conflict with the internal waste stream rule.

As it must to present grounds for modification, TMPA contends that the language of the treatment pond exception renders ultra vires the internal waste stream rule. TMPA seems to argue that the EPA promulgated the treatment pond exception to bind itself, like Ulysses before the sirens, lest the agency succumb to the temptation to regulate internal waste streams. The EPA, on the other hand, contends that the treatment pond exception is simply notice to polluters that it will not generally regulate internal waste streams, not a promise that it never will.24 Accordingly, EPA understands the internal waste stream rule as an exception to the treatment pond exception.

Notwithstanding the apparent conflict between the two regulations, there is no support for TMPA's position. We are required to defer to any reasonable EPA construction of its enabling statutes.25 When resolving an apparent conflict among EPA regulations, even greater deference is in order. As the Supreme Court stated in Udall v. Tallman:

When the construction of an administrative regulation rather than a statute is in issue, deference is even more clearly in order.... "[T]he ultimate criterion is the administrative interpretation, which becomes controlling weight unless it is plainly erroneous or inconsistent with the regulation."26

Here, we find the EPA's reading of the internal waste stream rule and the treatment pond exception regulations not erroneous and consistent with the regulations.

The holding TMPA urges upon us would unreasonably restrict the EPA's power to regulate pollution in important circumstances and, in so doing, frustrate Congress's purpose in the CWA. As the EPA points out, it is sometimes necessary to regulate discharges within the treatment process to control discharges at the end. Regardless of whether internal waste streams are "waters of the United States", the internal waste stream rule falls within EPA power to promulgate regulations necessary to control discharges into "waters of the United States".27 This position has support in the language of the CWA, its legislative history, and common sense.

The CWA is strong medicine. Section 301(a) prohibits the discharge by any person of any pollutant into the nation's waters except that which the EPA expressly allows in an NPDES permit.28 Section 101(a)(1) states that Congress enacted the CWA to "restore and maintain the chemical, physical, and biological integrity of the Nation's waters" and adopts "the national goal that the discharge of pollutants into the navigable waters be eliminated by 1985".29 Congress explicitly recognized that reduction of the amount of effluents--not merely their dilution or dispersion--is the goal of the CWA.30 This and other courts long ago adopted this view of the EPA's enabling legislation.31

That the CWA's ambitious goal has not been achieved even in 1987 does not vitiate Congress's intent that it be achieved as soon as possible.32 As section 308(a) makes clear, EPA policing of effluent limitations is instrumental to its achievement.33 And, under section 308(a), the EPA has authority to monitor waste streams "at such locations" necessary to "determin[e] whether any person is in violation of [an] effluent limitation".34 Under section 402, the EPA has broad powers to impose NPDES permit conditions "to assure compliance with" effluent limitations required by the CWA.35

In this light, we see two broad justifications for the internal waste stream rule. The first is EPA responsibility to measure and control the discharge of particular effluents.36 In most circumstances, EPA need only focus on discharges at the end of any waste treatment system. But "end-of-pipe" monitoring a particular effluent is sometimes impractical. The final discharge may be inaccessible and the confluence of several effluents in the final discharge may complicate monitoring of each.37 Also, dilution during treatment may leave an effluent present in immeasurably small, but still harmful, concentrations.38

The second justification for the internal waste stream rule is EPA responsibility to monitor the effectiveness of each polluter's treatment technology--here, TMPA's sewage treatment plant. Congress explicitly recognized that the EPA should have power to monitor and regulate the technology used in any waste treatment system.39 This allows the EPA to assess the effectiveness and cost of each waste treatment system--a step necessary to the CWA's broad goal of requiring the maximum feasible elimination of pollutants.40 Also, the CWA specifically requires that in some circumstances the EPA enforce treatment technology standards. Here, for example, section 301 imposes technology standards on the secondary treatment of TMPA's sanitary wastes.41

In sum, we find ourselves in agreement with the opinion of the Seventh Circuit Court of Appeals in Mobil Oil Corp. v. EPA:

[T]he only interest Mobil could possibly have in preventing EPA officials from sampling its untreated waste water is that Mobil might want to keep the EPA in the dark as much as possible about what pollutants are present in the water it dumps into the Des Plaines River and about how efficient its treatment processes are at cleaning its waste water of pollutants. Treatment of waste water may mask the presence of a pollutant.... And if EPA is to assess with any reasonable degree of accuracy how efficient Mobil's treatment processes are, it needs to know what pollutants are present in waste water before it is treated as well as after it has been treated.

Any interest Mobil may have in frustrating the EPA's efforts to assess the efficiency of its treatment processes and to detect trace amounts of toxic pollutants in waste water it dumps into the Des Plaines River is not entitled to protection.42

We hold that there is no mistake of law in EPA regulation of outfall 301 or any other internal waste stream.

We now turn to whether the EPA's decision to apply the internal waste stream rule to outfall 301 is based upon "technical mistakes". TMPA challenges--under the rubric of "technical mistake"--the EPA's determination that, because it is impractical to measure certain pollutants effectively at the discharge into Gibbons Creek Reservoir, it is necessary to measure them at outfall 301.

The restrictions the EPA imposed on outfall 301 apply to two measures of pollution, Biological Oxygen Demand ["BOD"] and Total Suspended Solids ["TSS"]. CWA section 304(a)(4) specifically requires that the EPA formulate limits on BOD and TSS,43 and both are significant concerns in the secondary treatment of sewage. According to the EPA's description,

BOD includes any compound which consumes dissolved oxygen in water. High levels of BOD in streams can kill fish. TSS is a measure of the amount of suspended or particulate matter in water. As TSS increases in a stream, light is blocked from aquatic plants. Also, fish and other aquatic organisms are affected by the cloudy conditions.44

The EPA decided to regulate outfall 301 because it determined that it was impractical to assure that TMPA's sewage treatment plant complies with EPA effluent and technology standards by monitoring only the discharges into Gibbons Creek Reservoir.

TMPA attacks this by contending that BOD and TSS are effectively treated by dilution. TMPA argues that, therefore, the EPA has reason to be concerned with BOD and TSS concentration at the discharges into the Gibbons Creek Reservoir, but not the concentrations at outfall 301. At bottom, TMPA challenges the EPA's finding that it is "impractical" to enforce CWA compliance by measuring TSS and BOD only at the final discharges.

Although TMPA may call into question EPA's fact-findings regarding BOD and TSS, its arguments have nothing to do with the term "technical mistake" in 40 C.F.R. Sec. 122.62(a)(16). Apparently, TMPA believes that "technical mistake" refers to mistakes in findings of "technical"--as opposed to unsophisticated--fact. This is simply wrong.

40 C.F.R. Sec. 122.62(a)(16) does not refer to "technical" fact-findings; rather, it says "technical mistakes, such as errors in calculation ".45 This makes clear that the regulation uses the term "technical mistake" narrowly, to mean only errors in mathematical calculations, computer errors, clerical mistakes, and the like. TMPA's reading would open the door to full judicial review of EPA fact-finding after the section 509 statute of limitations had run on the NPDES permit, contrary to our holding in TMPA I. Moreover, it would do so for precisely the type of agency decision that is normally subject to limited judicial review: complex, scientific determinations.46

Thus, we have no jurisdiction in this case to review the findings of fact behind the application of the internal waste stream rule to outfall 301; they are not relevant to TMPA's petition for modification.47

TMPA also challenges the EPA's refusal to modify the permit in response to alleged procedural flaws. TMPA notes that the internal waste stream rule requires the EPA to give an explanation of the circumstances that justify each application of the rule.48 TMPA alleges that the EPA's explanation of its reasons for regulating outfall 301, contained in the "fact sheet" the EPA sent to TMPA along with the draft permit in 1984,49 is inadequate.

Even if we saw merit in these allegations, procedural deficiencies are not a basis for permit modification under 40 C.F.R. Sec. 122.62(a)(16). The failings TMPA allege are neither "technical mistakes" nor "mistaken interpretations of law". Moreover, they were not "made in determining permit conditions".50 Whether the EPA gives adequate notice of its reasons to polluters will not affect directly the permit conditions it imposes. TMPA could have raised these alleged deficiencies during the course of the permit renewal in 1984,51 but did not. It is three years too late to do so now. Because these allegations have nothing to do with TMPA's petition for modification, we have no jurisdiction to address them.52

TMPA has failed to show any reason to modify the permit under 40 C.F.R. Sec. 122.62. There is no mistake of law in the EPA's regulation of any internal waste stream. While TMPA may be able to show that the EPA applied the internal waste stream rule to outfall 301 inappropriately, it is irrelevant here. The place for judicial review of findings of fact behind the application of the internal waste stream rule in TMPA's permit is in a timely challenge after the permit is issued, not in a petition to modify the permit. Similarly, the place for judicial review of procedural failings in the issuance of TMPA's permit is in a timely challenge after the permit is issued, not in a petition to modify the permit.

We see no reason to allow TMPA to evade the time limit on review of EPA action by expanding the scope of the petition for modification. At some point, agency decisions must become final; at some point, challenges to those decisions must be foreclosed. In this instance, CWA section 509(b)(1) sets the time limit for review of the terms of TMPA's permit. As we ruled in TMPA I, TMPA lost its right to challenge this permit because it failed to make the challenge timely. TMPA's failure may be regretable, but that is entirely its own fault. Moreover, the NPDES permit governing outfall 301 will be up for renewal only two years from now and, at that time, will be open to a full challenge by TMPA.

Accordingly, the petition for review of the EPA's refusal to modify the permit is DENIED.

Judge Wisdom's opinion for the court is correct and I must therefore concur in it, although I do so with some reluctance. My concurrence rests upon his clear demonstration that the actions of which TMPA complains are neither "mistaken interpretations of the law" nor "technical mistakes," and hence that at this stage of proceedings our review cannot reach them. My reluctance stems from a conviction that the administrator has conducted itself in an unreasonable fashion, treating BOD and TSS "pollution"--pollution which is produced by concentration and is destroyed by dilution--as though it involved toxic substances. Even in the service of the environment, it appears that bureaucracy remains bureaucracy. We hold no general roving commission to right wrongs, however; and the majority is correct in holding that these must be endured for two more years.