Plaintiffs, Natural Resources Defense Council, Inc. and Delaware Audubon Society (collectively "NRDC"), brought this citizen suit pursuant to section 505 of the Federal Water Pollution Control Act, 33 U.S.C. Sec. 1365 ("Clean Water Act" or "Act"). The complaint alleges that Texaco Refining and Marketing, Inc. ("Texaco") repeatedly violated National Pollution Discharge Elimination System ("NPDES") permits limiting effluent discharges from its Delaware City oil refinery. Following a bench trial, the district court held Texaco liable for 365 permit violations corresponding to 3,360 days of violation, and imposed civil penalties totalling $1,680,000. The district court also issued a permanent injunction prohibiting further permit violations and ordering Texaco to comply with the permit's investigatory, reporting and monitoring provisions. Texaco claims that the district court erred by: (1) exercising jurisdiction over certain wholly past permit violations; (2) failing to dismiss some of the NRDC's claims for penalties as moot; (3) failing to dismiss the NRDC's complaint for lack of standing; (4) ordering injunctive relief; and (5) overcounting the days of violation. The NRDC cross-appeals, claiming that the district court erred by exercising jurisdiction over too few permit violations.

We will affirm in part and reverse in part, and remand for further proceedings consistent with this opinion.

The NRDC brought this action in May 1988. At that time, Texaco was discharging pollutants into the Delaware River pursuant to an NPDES permit issued by the Delaware Department of Natural Resources and Environmental Control ("DNREC"). The permit was originally issued in 1977 to a prior owner of the refinery. Texaco acquired the refinery in 1984.

Texaco's permit established seven monitoring points, or outfalls, through which pollutants were discharged. The outfalls were labelled 001, 002, 101, 201, 301, 401, and 501. The permit also designated parameters for each outfall. A parameter is a particular attribute of a discharge. Parameters under the permit included specific types of pollutants, as well as other discharge characteristics, such as temperature and flow rate. Most parameters were subject to strict effluent limits, of which there were two types. Maximum daily limits specified the maximum amount of a pollutant that could be discharged from an outfall during any calendar day. Daily average limits, which were lower than maximum daily limits, specified the maximum average amount that could be discharged over the course of the days in a calendar month that the facility operated.

The permit provided that "[t]he discharge of any pollutant identified in this permit ... at a level in excess of that authorized shall constitute a violation of the permit." App. at 157. To assure compliance, the permit required Texaco to sample every parameter at every outfall at prescribed intervals, using specified methods, and to report the test results in monthly Discharge Monitoring Reports ("DMRs") to DNREC. The permit also required Texaco to "take all reasonable steps to minimize any adverse impact to waters of the State resulting from this permit, including such accelerated or additional monitoring as necessary to determine the nature and impact of the noncomplying discharge." Id.

On December 31, 1988 Texaco transferred ownership of the refinery to Star Enterprise ("Star"), a partnership between a wholly-owned subsidiary of Texaco and a subsidiary of the Saudi Arabian Oil Company. In January 1989, DNREC issued a new NPDES permit to Star. The 1989 permit modified the 1977 permit by, among other things, eliminating outfall 501 and altering numerous effluent limits.

Prior to trial, Texaco filed two motions for summary judgment. Relying on the Supreme Court's holding in Gwaltney of Smithfield, Ltd. v. Chesapeake Bay Foundation, 484 U.S. 49, 108 S. Ct. 376, 98 L. Ed. 2d 306 (1987) ("Gwaltney I"), Texaco claimed that the district court lacked jurisdiction to hear the NRDC's claims relating to wholly past violations. Texaco also argued, among other things, that the relaxation or elimination of certain effluent limitations in the 1989 permit rendered several other of the NRDC's claims moot. The NRDC filed a cross-motion for summary judgment on the issue of liability, arguing that Texaco's DMRs contained dispositive proof of the alleged violations. The district court denied Texaco's motions, rejecting Texaco's contention that the court lacked jurisdiction over certain wholly past violations. Natural Resources Defense Council v. Texaco Refining and Marketing, Inc., 719 F. Supp. 281, 287 (D.Del.1989) (Texaco I ). The court granted the NRDC's motion and found Texaco liable for all alleged permit violations. Id. at 287, 289. Determining that the NRDC had shown probable cause that the Clean Water Act was being violated, the district court also enjoined Texaco from future violations of the 1989 permit. Id. at 292. After Texaco filed an interlocutory appeal, we vacated the injunction, holding that the district court erred by failing to apply traditional equitable principles in determining whether injunctive relief was appropriate. Natural Resources Defense Council, Inc. v. Texaco Refining and Marketing, Inc., 906 F.2d 934, 937 (3d Cir.1990) ("Texaco II").

Following remand, the district court conducted a three-week bench trial in February 1991. In June 1991, the court reopened the trial and allowed the NRDC to introduce additional evidence of their standing. After all the evidence in the case had been presented, the district court held that the NRDC had shown sufficient evidence of standing and concluded that Texaco had exceeded NPDES permit effluent limits 414 times between March 1983 and February 1991. Natural Resources Defense Council v. Texaco Refining and Marketing, Inc., 800 F. Supp. 1 (D.Del.1992). While the case before the district court was pending on remand, the Court of Appeals for the Fourth Circuit decided Chesapeake Bay Foundation v. Gwaltney of Smithfield, Ltd., 890 F.2d 690 (4th Cir.1989) ("Gwaltney III"). Relying in part on this decision, the district court reconsidered its earlier holding in Texaco I and concluded that it had jurisdiction only over continuous or intermittent violations, and not over wholly past violations. Of Texaco's 414 violations, the district court found that 365 were continuous or intermittent at the time the complaint was filed, while 49 were wholly past.

To calculate the number of days of violations, the district court counted each excess of a daily maximum limitation as the equivalent of one day of violation and each excess of a daily average limitation as a violation of each day of the month during which the excess occurred. The court thereby arrived at a total of 3,360 days of violation. After considering the relevant statutory factors, see 33 U.S.C. Sec. 1319(d), the court imposed a civil penalty of $500 per day of violation, for a total penalty of $1,680,000. The district court also enjoined further violations of the 1989 permit issued to Star and directed Texaco to comply with the permit's investigatory, reporting, and monitoring provisions.

In its cross-appeal, the NRDC claims that the district court erred by failing to exercise jurisdiction over forty-nine violations that occurred prior to the filing of the complaint. We will address the NRDC's cross-appeal first because it raises an important threshold issue as to the proper standard for determining the scope of the district court's subject matter jurisdiction.

The district court exercised subject matter jurisdiction pursuant to section 505(a) of the Clean Water Act, which provides that

any citizen may commence a civil action on his own behalf--

(1) against any person ... who is alleged to be in violation of (A) an effluent standard or limitation....

33 U.S.C. Sec. 1365(a) (emphasis added). In Gwaltney I, the Supreme Court interpreted the phrase "alleged to be in violation" in section 505 as imposing a jurisdictional requirement "that citizen-plaintiffs allege a state of either continuous or intermittent violation--that is, a reasonable likelihood that a past polluter will continue to pollute in the future." 484 U.S. at 57, 108 S.Ct. at 381. Consistent with this requirement, the Court held that jurisdiction will not lie where a plaintiff alleges claims for "wholly past" violations. Id., at 57-58, 108 S.Ct. at 381-82.

The Court explained that the jurisdictional rule announced in Gwaltney I is suggested by the repeated use of the present tense throughout section 505. For example, in addition to requiring a citizen to allege that a polluter is "in violation" of an effluent standard or limitation, the Act states that a citizen suit may be brought for violation of a permit limitation "which is in effect." 33 U.S.C. Sec. 1365(f); Gwaltney I, 484 U.S. at 59, 108 S.Ct. at 382. Citizen-plaintiffs also must provide notice of an action to, among others, the State in which the alleged violation "occurs." 33 U.S.C. Sec. 1365(b)(1)(A). According to the Court, the most salient use of the present tense is in the Act's definition of "citizen" as "a person ... having an interest which is or may be adversely affected" by the alleged violations. Gwaltney I, 484 U.S. at 59, 108 S.Ct. at 382; 33 U.S.C. Sec. 1365(g). "This definition makes plain what the undeviating use of the present tense strongly suggests: the harm sought to be addressed by the citizen suit lies in the present or the future, not in the past." Gwaltney I, 484 U.S. at 59, 108 S.Ct. at 382.

At issue in this cross-appeal is the proper standard for determining the extent of a court's subject matter jurisdiction over cases in which a plaintiff alleges a series of permit violations involving multiple parameters. The NRDC argues that courts should determine the scope of their jurisdiction using a "permit-based" standard. Under a permit-based standard, jurisdiction attaches to entire cases, not to individual violations alleged within a case. See Sierra Club v. Port Townsend Paper Corp., 19 Envtl.L.Rep. (Envtl.L.Inst.) 20,532, 20,533, 1988 WL 160580 (W.D.Wash.1988). If the plaintiff alleges in good faith that violations of the permit as a whole or of any particular parameter within the permit are occurring on a continuous or intermittent basis, the court may automatically exercise jurisdiction over all alleged past, present, and potential future violations. See Texaco I, 719 F. Supp. at 287. Texaco, on the other hand, advocates a "parameter-based," or "by-parameter" standard. Under the by-parameter standard, a court looks separately at each violation or set of violations of a parameter and exercises jurisdiction only over those which the plaintiff alleges to be continuous or intermittent as of the time of the complaint. See id. at 286.

Although the district court initially adopted the permit-based standard advocated by the NRDC, see id. at 287, the court later reconsidered its position after the standard was rejected by the Court of Appeals for the Fourth Circuit in Gwaltney III. In that case, the district court imposed over $1,000,000 in penalties for violations of the defendant's total Kjeldahl nitrogen ("TKN") and chlorine NPDES permit limits. The court of appeals reversed the judgment as it related to the chlorine violations, finding that although the TKN violations were ongoing at the time the suit was filed, it was " 'absolutely clear' that there were no ongoing chlorine violations at the time this suit was brought." Gwaltney III, 890 F.2d at 697. The court expressly rejected the argument that, having found "an ongoing violation of the permit," it could order penalties "for any past violation of any permit parameter."

If Congress wished to permit citizen suits only when a discharger fails to abate a problem and the government fails to take enforcement action ... then it would make little sense to permit penalties for wholly past violations of one parameter simply because there are ongoing violations of another parameter.

Gwaltney III, 890 F.2d at 698 (emphasis in original).

While agreeing that the permit-based approach is overly broad, the district court in the present case found the Gwaltney III by-parameter approach too narrow insofar as it implies that for each violation or set of violations of a parameter, the plaintiff must eventually be able to prove a continuing likelihood that the same exact parameter will be violated. The district court found that one flaw in the treatment process can often lead to violations of several parameters. The court also stated that the Clean Water Act's policy of full compliance demands that polluters identify and implement effective measures for remedying the underlying causes of violations. When appropriate measures are not taken, the district court reasoned, "the fact that the failure to solve the problem is manifested from time to time in different parameters should not preclude a citizen's suit directed at the unresolved source of trouble." 800 F. Supp. at 14.

These concerns led the district court to adopt a modified by-parameter standard that expressly accounts for the interrelationship between parameter violations and their underlying causes. Like a strict by-parameter standard, the district court's standard requires a separate determination of jurisdiction for each violation or set of violations of a parameter and allows a plaintiff to establish jurisdiction by alleging in good faith that the violation or violations are likely to recur on a continuous or intermittent basis. Under the modified by-parameter approach, however, a plaintiff can establish at trial that violations are continuous or intermittent in either of two ways: first, by proving a likelihood of recurring violations of the same parameter; or second, by proving a likelihood that the same inadequately corrected source of trouble will cause recurring violations of one or more different parameters.1

We agree that a modified by-parameter approach is most consistent with the language of the Act and its underlying policies and find the NRDC's arguments in favor of a permit-based approach unpersuasive.2 The NRDC contends, for example, that the plain language of section 505 mandates a permit-based rather than a by-parameter jurisdictional standard.3 A citizen may "commence a civil action" under section 505(a) against any person "alleged to be in violation of ... an effluent standard or limitation." 33 U.S.C. Sec. 1365(a). An "effluent standard or limitation" is defined to include an NPDES "permit or condition thereof." Id. Sec. 1365(f)(6). Plugging in this statutory definition, the NRDC concludes that section 505 assures jurisdiction over a citizen suit in which a discharger is "alleged to be in violation" of an NPDES "permit."

On its face, however, section 505 addresses only the circumstances in which a citizen "may commence" a civil action, not the scope of such an action. Although the plain language of the statute authorizes a citizen to commence a civil action against a person who is in violation of a "permit," it does not follow that such a suit, once commenced, can extend to cover all past as well as present permit violations, including those that are not reasonably likely to recur. Of course, neither does the plain language of section 505 preclude such a result. Rather, the plain language is not controlling.

Still, the statutory language is instructive. As the Court emphasized in Gwaltney I, the language in section 505(a) that delimits the scope of a court's jurisdiction over citizen suits are the words "to be in violation." See 484 U.S. at 57, 108 S. Ct. at 381. This phrase in conjunction with the language and structure of section 505 as a whole "make plain that the interest of the citizen-plaintiff is forward-looking." Id. at 59, 108 S. Ct. at 382. The prospective orientation of citizen suits is particularly reflected in, among other places, the citizen enforcement provision's primary emphasis on injunctive relief. While section 505 authorizes civil penalties, citizens may seek such penalties only in a suit brought to enjoin or otherwise abate a continuous or intermittent violation. Id.

Compared with private citizens, the EPA Administrator is given a broader range of enforcement powers and greater discretion in choosing among them. The Administrator, unlike private citizens, may bring enforcement actions for "wholly past violations." Id. at 58, 108 S.Ct. at 382. The Administrator also may pursue civil penalties without simultaneously seeking injunctive relief. See id. at 58-59, 108 S. Ct. at 382. Thus, while the Act unquestionably provides the public with "the right to seek vigorous enforcement action," S.Rep. No. 92-414, 92nd Cong., 1st Sess. 64 (1971), reprinted in 1972 U.S.C.C.A.N. 3668, 3730, it does not provide private citizens and the Administrator with coequal enforcement roles. Rather, the broader enforcement powers conferred upon the government reflect a conscious decision to accord citizens a "supplementary," or "interstitial," role. Gwaltney I, 484 U.S. at 60-61, 108 S.Ct. at 383.

Under the permit-based jurisdictional approach advocated by the NRDC, the enforcement powers enjoyed by private citizens would closely approximate the powers that the Act confers on the Administrator alone and would allow citizens to pursue the type of retrospective actions that the Court in Gwaltney I held to be strictly prohibited. The Court in Gwaltney I feared that allowing citizens to pursue wholly past violations "could undermine the supplementary role envisioned for the citizen suit." Id. at 60, 108 S.Ct. at 383.

This danger is best illustrated by an example. Suppose that the Administrator identified a violator of the Act and issued a compliance order under Sec. 309(a). Suppose further that the Administrator agreed not to assess or otherwise seek civil penalties on the condition that the violator take some extreme corrective action, such as to install particularly effective but expensive machinery, that it otherwise would not be obliged to take. If citizens could file suit, months or years later, in order to seek the civil penalties that the Administrator chose to forgo, then the Administrator's discretion to enforce the Act in the public interest would be curtailed considerably.

Id. at 60-61, 108 S. Ct. at 383. The scenario depicted above illustrates the shortcomings of a permit-based approach, which allows citizens who allege an ongoing violation of a single parameter to litigate past violations of different and unrelated parameters, even if corrective measures were taken to eliminate any reasonable likelihood that such past violations would recur. By allowing what are in essence wholly past and unrelated parameter violations to be brought within the purview of citizen suits, a permit-based standard would in essence "change the nature of the citizens' role from interstitial to potentially intrusive." See id., at 61, 108 S.Ct. at 383.

The NRDC also argues that a by-parameter standard is untenable because it would transform citizen enforcement suits from the simple actions Congress envisioned to complex proceedings frontloaded with jurisdictional fact-finding. The same criticism was recently voiced by the United States District Court for the District of New Jersey in Public Interest Research Group v. ELF Atochem North America, Inc., 817 F. Supp. 1164, 1175-76 (D.N.J.1993), a case decided while this appeal was pending. Citing legislative history, the district court in ELF Atochem noted that Congress intended to streamline enforcement actions under the Act by avoiding the necessity of lengthy fact-findings and investigations. 817 F. Supp. at 1175; see S.Rep. No. 92-414, supra, at 64, reprinted in 1972 U.S.C.C.A.N. 3668, 3730. The district court adopted a permit-based rather than a parameter-based standard, explaining:

The parameter-by-parameter approach--whether conducted at the outset of the suit or at trial--requires a particularized and fact-intensive inquiry, especially if we attempt to account for underlying causes in determining which violations are "ongoing".... This cumbersome analysis is antithetical to th[e] statutory purpose.

Elf Atochem, 817 F. Supp. at 1175.

A closer look at the legislative history reveals that the section of the Senate Report relied upon by the court in ELF Atochem applies specifically to section 309 of the Act, the federal enforcement provision. S.Rep. No. 92-414, supra, at 63-64, reprinted in 1972 U.S.C.C.A.N. 3668, 3729-30. While it is fair to assume that the goal of streamlining enforcement actions applies in some degree to citizen suits also, we do not think it applies with equal force. If Congress intended to authorize the Administrator but not private citizens to redress wholly past violations, then we must assume Congress also contemplated that some extra analysis would be necessary in citizen suits in order to distinguish wholly past violations from ongoing ones.

Moreover, we do not think the added analysis required under a by-parameter approach will greatly complicate citizen suits. To begin with, a district court need not engage in any fact-finding whatsoever in order to determine the scope of its subject matter jurisdiction in citizen suits. As the Court in Gwaltney I made exceedingly clear, good faith allegations "suffice for jurisdictional purposes." 484 U.S. at 65, 108 S.Ct. at 385. Nor is it particularly burdensome for plaintiffs to make jurisdictional allegations on a by-parameter basis, since federal regulations already require plaintiffs to provide a notice of intent to file suit containing "sufficient information to permit the recipient to identify the specific standard, limitation, or order alleged to have been violated, the activity alleged to constitute a violation ... [and] the date or dates of such violation." 40 C.F.R. Sec. 135.3; see 33 U.S.C. Sec. 1365(c)(3). In addition, citizen-plaintiffs should have ready access to the relevant information, which is contained in publicly available DMRs. See 40 C.F.R. Secs. 2.101; 122.7; 122.41(j), (l ); 122.44(i); see also 33 U.S.C. Sec. 1314(i).

Although subject matter jurisdiction depends entirely on the sufficiency of a plaintiff's good faith allegations, a plaintiff nevertheless is required at trial to prove that the alleged violations are continuous or intermittent in order to satisfy the requirement of standing. See Gwaltney I, 484 U.S. at 66, 108 S.Ct. at 386.4 We therefore acknowledge that under a modified by-parameter approach, a court may have to consider evidence concerning the nature and causes of some violations in order to determine the likelihood that the same or related violations will continue to recur. As the NRDC concedes, however, much of the same evidence will often be considered by the court anyway when deciding whether to issue injunctive relief and how much to impose in civil penalties. Consequently, while a permit-based approach could be expected to facilitate a court's decision as to standing, such an approach would not necessarily lessen the complexity of citizen suits overall.

Texaco's first argument on appeal is that the district court erred by treating post-complaint violations as dispositive proof that certain pre-complaint violations involving the same parameters were continuous or intermittent. The district court found that 42 violations of various parameters occurred after the filing of the complaint. The court also found that 323 violations of the same parameters occurred prior to the filing of the complaint. Based almost entirely on the correspondence between the pre-complaint and post-complaint violations, the district court concluded that all 323 of the pre-complaint violations were continuous or intermittent when the complaint was filed.

In cases decided since the Court's decision in Gwaltney I, courts of appeals have consistently held, and we agree, that plaintiffs can establish that violations are continuous or intermittent "either (1) by proving violations that continue on or after the date the complaint is filed, or (2) by adducing evidence from which a reasonable trier of fact could find a continuing likelihood of a recurrence in intermittent or sporadic violations." Chesapeake Bay Foundation v. Gwaltney of Smithfield, Ltd., 844 F.2d 170, 171-72 (4th Cir.1988) ("Gwaltney II"); see Carr v. Alta Verde Indus., Inc., 931 F.2d 1055, 1062 (5th Cir.1991); Sierra Club v. Union Oil Co. of California, 853 F.2d 667, 671 (9th Cir.1988).5 As the use of the disjunctive makes clear, a plaintiff need not prove both that a post-complaint violation has occurred and that independent evidence proves a continuing likelihood of recurring violations. Either method of proof will suffice. See, e.g., Gwaltney II, 844 F.2d at 171-72. Because proof of one or more post-complaint violations is itself conclusive, see Carr, 931 F.2d at 1065 n. 12, the district court did not err by relying solely on such violations to determine that corresponding pre-complaint violations were continuous or intermittent.6

Subject matter jurisdiction is determined with respect to circumstances at the time the complaint is filed. See Lujan v. Defenders of Wildlife, --- U.S. ----, ---- n. 4, 112 S. Ct. 2130, 2141 n. 4, 119 L. Ed. 2d 351 (1992). Post-filing events are thus incapable of stripping a court of its properly conferred jurisdiction. Principles of mootness, however, preclude a court from adjudicating claims that no longer present a concrete and substantial controversy capable of being redressed through specific relief. See United States v. Michigan Nat'l Corp., 419 U.S. 1, 4, 95 S. Ct. 10, 11, 42 L. Ed. 2d 1 (1974). Thus, the doctrine of mootness controls whether a citizen-plaintiff may maintain a suit if the allegations of ongoing permit violations cannot be established because "the allegedly wrongful behavior [of the defendant] could not reasonably be expected to recur." United States v. Concentrated Phosphate Export Ass'n, 393 U.S. 199, 203, 89 S. Ct. 361, 364, 21 L. Ed. 2d 344 (1968). The Gwaltney I Court, in dicta, observed:

Longstanding principles of mootness ... prevent the maintenance of suit when "there is no reasonable expectation that the wrong will be repeated." In seeking to have a case dismissed as moot, however, the defendant's burden "is a heavy one." The defendant must demonstrate that it is "absolutely clear that the allegedly wrongful behavior could not reasonably be expected to recur." Mootness doctrine thus protects defendants from the maintenance of suit under the Clean Water Act based solely on violations wholly unconnected to any present or future wrongdoing, while it also protects plaintiffs from defendants who seek to evade sanction by predictable "protestations of repentance and reform."

Gwaltney I, 484 U.S. at 66-67, 108 S.Ct. at 386 (citations omitted) (emphasis in original).

Texaco argues that the district court should have dismissed many of the NRDC's claims for penalties on mootness grounds because the 1989 permit relaxed several parameters sufficiently to eliminate any reasonable expectation that the refinery would violate the new standards. Specifically, the flow limit at outfall 201 increased from 334 to 398 gallons per day and outfall 501 was completely eliminated in the 1989 permit. The refinery's average flow from outfall 201 has not exceeded 398 gallons per day since 1983, and the NRDC does not dispute Texaco's contention that future violations of this limit are not reasonably expected. Clearly no outfall 501 violation could reoccur because that outfall is no longer subject to regulation. Injunctive relief with respect to these effluent standards therefore is moot. See id.

Texaco argues that because the NRDC's claims for injunctive relief are moot, the corresponding penalty claims also must be dismissed as moot. Otherwise, Texaco asserts, the NRDC would maintain a suit for wholly past violations in contravention of Gwaltney I. The NRDC urges us to follow the three courts of appeals that have addressed this issue and have held that claims for penalties survive even though the corresponding claims for injunctive relief become moot. See Atlantic States Legal Foundation, Inc. v. Pan American Tanning Corporation, 993 F.2d 1017, 1021 (2d Cir.1993) ("We hold ... that a defendant's ability to show, after suit is filed but before judgment is entered, that it has come into compliance with limits on the discharge of pollutants will not render a citizen suit for civil penalties moot."); Atlantic States Legal Foundation, Inc. v. Tyson Foods, Inc., 897 F.2d 1128, 1135-37 (11th Cir.1990) ("[I]f the parties are able to make a valid request for injunctive relief at the time the complaint is filed, then they may continue to maintain a suit for civil penalties, even when injunctive relief is no longer appropriate."); Gwaltney III, 890 F.2d at 696 ("[T]he penalty factor keeps the controversy alive between plaintiffs and defendants in a citizen suit ... even though the ultimate judicial remedy is the imposition of civil penalties assessed for past acts of pollution.").

The Supreme Court in Gwaltney I failed to address whether mootness bars only claims for injunctive relief or also damage claims.7 We find the reasoning in Pan American Tanning, Tyson Foods and Gwaltney III persuasive, and hold that claims for damages are not moot because an intervening NPDES permit eliminates any reasonable possibility that Texaco will continue to violate specified parameters.

The language and structure of the Clean Water Act support our holding. The Act states that any person who violates a permit condition "shall be subject to a civil penalty." 33 U.S.C. Sec. 1319(d) (emphasis added). This mandatory language demonstrates that once a citizen plaintiff establishes an ongoing violation of a parameter at the time the complaint is filed, the court is obliged to assess penalties for all proven violations of that parameter. See Pan American Tanning, 993 F.2d at 1020-21; Gwaltney III, 890 F.2d at 697. Allowing a polluter to escape all liability through post-complaint compliance is at odds with the mandatory language of section 1319(d).

The policies underlying the existence of citizen suits compel the same result. Citizens may initiate suits only if no government action to require compliance with the Act is pending. See 33 U.S.C. Sec. 1365(b)(1)(B). As private attorneys general, citizens constitute a special category of plaintiffs who ensure that companies comply with the Act even when the government's limited resources prevent it from bringing an enforcement action. See Tyson Foods, 897 F.2d at 1136. Citizen plaintiffs may seek penalties in addition to injunctive relief. See 33 U.S.C. Sec. 1365(a). The potential imposition of penalties is an important mechanism to deter companies from violating their NPDES permits.

A citizen suit would lose much of its effectiveness if a defendant could avoid paying any penalties by post-complaint compliance. See Pan American Tanning, 993 F.2d at 1020-21; Tyson Foods, 897 F.2d at 1136-37.8 If penalty claims could be mooted, polluters would be encouraged to "delay litigation as long as possible, knowing that they will thereby escape liability even for post-complaint violations, so long as violations have ceased at the time the suit comes to trial." Tyson Foods, 897 F.2d at 1137. Moreover, whether or not damage claims are mooted would depend on the vagaries of when the district court happens to set the case for trial. See id. We cannot embrace a rule that would weaken the deterrent effect of the Act by diminishing incentives for citizens to sue9 and encourage dilatory tactics by defendants. See Pan American Tanning, 993 F.2d at 1020-21; Tyson Foods, 897 F.2d at 1137.

At the pretrial conference, Texaco argued that the NRDC had to prove standing at trial even though the district court held in its opinion granting summary judgment that the NRDC had standing. In a long colloquy with the court, Texaco's counsel stated that there was no proof of standing because "nobody has come in and said that particular damage to the river affected me" and because "the parameter-by-parameter approach applies to both standing and jurisdiction." App. at 128. The district court stated that it had ruled on the issue of standing in its summary judgment motion, but then told Texaco's counsel that after trial it could submit a proposed conclusion of law that plaintiffs lack standing if such a conclusion was warranted.

Following the court's timetable, at the conclusion of the trial, Texaco moved to dismiss the case because the NRDC had failed to prove that a member of their organization suffered sufficient injury in fact. In response, the NRDC filed a motion to reopen the trial so that members of their organization could testify concerning the adverse consequences they experienced from Texaco's polluting. The district court granted this motion without explanation, and allowed the NRDC to take two hours of additional testimony. We review this decision for abuse of discretion. See Skehan v. Board of Trustees of Bloomsburg State College, 590 F.2d 470, 478 (3d Cir.1978), cert. denied, 444 U.S. 832, 100 S. Ct. 61, 62 L. Ed. 2d 41 (1979). Because the district court "has a feel for the case that an appellate court can seldom have, the trial court's ruling is subject to reversal only in a rare case where abuse is clearly shown." Id. (quoting 6A Moore's Federal Practice, p 59.04, at 36-37 (2d ed. 1974)).

Texaco argues that no legitimate reason existed for allowing the NRDC to introduce additional testimony. We disagree. Failure to adduce evidence adequate to sustain a judgment is generally fatal. See Rochez Bros., Inc. v. Rhoades, 527 F.2d 891, 894 (3d Cir.1975) (citing Youngstown Sheet & Tube Co. v. Lucey, 403 F.2d 135, 139 (5th Cir.1968)), cert. denied, 425 U.S. 993, 96 S. Ct. 2205, 48 L. Ed. 2d 817 (1976). A trial may be reopened, however, when the failure is due to a reasonable misunderstanding among the parties and the trial court. See id. at 894-95.

The NRDC filed an affidavit stating that they believed standing would be revisited only to allow Texaco to raise its by-parameter standing theory. Plaintiffs apparently believed that the summary judgment disposition finally resolved the issue of injury in fact. At the pretrial conference, the district court had not clarified what aspect or aspects of standing Texaco could raise after trial. Because the district court reasonably could have found that there was a legitimate misunderstanding concerning whether Texaco would be permitted to challenge plaintiff's proof of injury in fact, it did not abuse its discretion by reopening the case for limited additional testimony.

Texaco further argues that even considering the additional testimony, the NRDC still has failed to prove their standing to bring a citizen suit. Whether sufficient evidence supported the district court's conclusion that the NRDC had standing is a legal question subject to plenary review. See Public Interest Research Group of New Jersey, Inc. v. Powell Duffryn Terminals, Inc., 913 F.2d 64, 71 (3d Cir.1990), cert. denied, 498 U.S. 1109, 111 S. Ct. 1018, 112 L. Ed. 2d 1100 (1991).

Standing analysis focuses on whether a party has a sufficient stake in the controversy. See Sierra Club v. Morton, 405 U.S. 727, 731-32, 92 S. Ct. 1361, 1364, 31 L. Ed. 2d 636 (1972). Representational standing, when an organization represents the interests of its members, is established when 1) the members would have individual standing to sue; 2) the interests the organization seeks to protect are related to its goals; and 3) the claims do not require individual participation by the members. See International Union, United Automobile, Aerospace & Agricultural Implement Workers v. Brock, 477 U.S. 274, 282, 106 S. Ct. 2523, 2529, 91 L. Ed. 2d 228 (1986) (quoting Hunt v. Washington State Apple Advertising Comm'n, 432 U.S. 333, 343, 97 S. Ct. 2434, 2441, 53 L. Ed. 2d 383 (1977)). To fulfill the first prong of this test, demonstrating individual standing for the organization's members, plaintiffs must show that 1) the individual has personally suffered an actual or threatened injury from the defendant's allegedly illegal conduct; 2) the injury may be fairly traced to this conduct; and 3) a favorable decision will redress the harm. See Valley Forge Christian College v. Americans United for Separation of Church and State, Inc., 454 U.S. 464, 472, 102 S. Ct. 752, 758, 70 L. Ed. 2d 700 (1982).

Texaco argues that the NRDC has not shown that their injuries are fairly traceable to Texaco's permit violations. In Powell Duffryn, this court expounded on the fairly traceable requirement in the context of a citizen suit to enforce the Clean Water Act. We held that plaintiffs need not prove to a scientific certainty that defendant's effluent caused the precise harm suffered by plaintiffs. Powell Duffryn, 913 F.2d at 72. Plaintiffs simply must show a substantial likelihood that defendant's conduct caused their harm. Id. We elaborated:

In a Clean Water Act case, this likelihood may be established by showing that a defendant has 1) discharged some pollutant in concentrations greater than allowed by his permit 2) into a waterway in which the plaintiffs have an interest that is or may be adversely affected by the pollutant and that 3) this pollutant causes or contributes to the kinds of injuries alleged by the plaintiffs.

Id. Texaco does not dispute that the NRDC has satisfied the first two prongs of the Powell Duffryn test.

Texaco argues that the NRDC did not show that the effluents it discharged in unlawful quantities cause the precise types of injuries its members suffered. Members of the NRDC testified that the sight and unpleasant smell of pollution in the vicinity of the refinery interfered with their enjoyment of bird watching, swimming, boating and fishing in Delaware City waterfront parks, and walking along the river. The presence of toxic chemicals in the waterway also inhibited them from eating local fish. Plaintiffs' expert, Dr. Livingston, stated that the effluents Texaco placed into the Delaware River may poison fish, render areas unsafe for people to swim, and generally limit the river as a fertile ground for recreation. See App. at 1004-06. Although Texaco objected to this testimony as speculation, the district court found it persuasive. This testimony linked Texaco's unlawful discharges to the exact injuries asserted by the NRDC.10

Several NRDC members also stated that the River had an oily sheen in certain areas. Dr. Livingston testified that excessive oil discharges, such as Texaco's violation of its oil and grease limit, contribute to this type of harm. Dr. Livingston explained that some compounds in oil can float to the surface and create an oil slick. Texaco complains that Dr. Livingston did not state that the compounds that cause oil slicks were compounds found in the oil Texaco discharged. This type of specificity is unnecessary. Indeed, the NRDC made precisely the showing that this court deemed sufficient to prove standing in Powell Duffryn. There, we found that:

[S]everal affiants stated that the water had an oily or greasy sheen they found offensive.... PDT's reports to the EPA indicate that PDT has discharged oil and grease in excess of [its NPDES permit] limits. Thus the aesthetic injury suffered by the plaintiffs may fairly be traced to PDT's effluent.

Powell Duffryn, 913 F.2d at 73. The NRDC adequately has shown that their injuries are fairly traceable to Texaco's permit violations.VI. INJUNCTIVE RELIEF

Texaco next argues that the permanent injunction, which prohibited it from violating any aspect of the 1989 NPDES permit, was inappropriate for several reasons. Texaco first asserts that the district court may not enjoin it because Star, not Texaco, now owns the refinery and is the named permittee. At the time the NRDC filed this suit, Texaco owned and operated the refinery and held the permit at issue. Shortly thereafter, on December 31, 1988, Texaco transferred ownership of the refinery to Star. The NRDC moved to join Star as a party, but the district court denied this motion because Star was not necessary to the litigation and would be bound by any injunction as a successor to Texaco.

Federal Rule of Civil Procedure 25(c) reveals the lack of merit in Texaco's assertion. Rule 25(c) clearly states that "[i]n case of any transfer of interest, the action may be continued by or against the original party, unless the court upon motion directs the person to whom the interest is transferred to be substituted...." One commentator has noted that "[t]he most significant feature of Rule 25(c) is that it does not require that anything be done after an interest has been transferred." 7C C. Wright, A. Miller & M. Kane, Federal Practice and Procedure Sec. 1958, at 555 (2d ed. 1986). Joinder of the transferee is unnecessary unless the trial court makes a discretionary determination that the transferee's presence would facilitate the case. Id. at 555-57. Moreover, an injunction against the named party will bind all successors in interest and assigns. See id. ("the judgment will be binding on [a named party's] successor in interest even though he is not named"); Kloster Speedsteel AB v. Crucible Inc., 793 F.2d 1565, 1581-82 (Fed.Cir.1986), cert. denied, 479 U.S. 1034, 107 S. Ct. 882, 93 L. Ed. 2d 836 (1987); see also Regal Knitwear Co. v. National Labor Relations Board, 324 U.S. 9, 13-15, 65 S. Ct. 478, 481, 89 L. Ed. 661 (1945) (NLRB injunction binds successors if they continue operation of previous employer).

Texaco next challenges the propriety of injunctive relief on the grounds that the factual predicates supporting the imposition of penalties do not also demonstrate irreparable injury. Not every violation of the Clean Water Act justifies the issuance of an injunction. See Weinberger v. Romero-Barcelo, 456 U.S. 305, 320, 102 S. Ct. 1798, 1807, 72 L. Ed. 2d 91 (1982). Injunctive relief is appropriate only when the district court deems it "necessary to secure prompt compliance with the Act." Id. Although the district court exercises considerable discretion in making this determination, proof of irreparable injury and the inadequacy of legal remedies are prerequisites to the granting of injunctive relief in a Clean Water Act case. Id., 456 U.S. at 311-13, 102 S. Ct. at 1803; Texaco II, 906 F.2d at 937-41.

The NRDC demonstrated irreparable injury arising from Texaco's unlawful conduct. The NRDC established a decade-long history of substantial permit noncompliance at Texaco's refinery. Texaco even concedes that "the court reasonably concluded that there may be future permit violations." Texaco Opening Brief at 44. Plaintiffs' expert, whom the district court credited, testified to the long-term and cumulative effects of excessive discharges of pollutants. Such discharges may impact a river's capacity to serve as a source of food, a sanctuary for wildlife and a fertile ground for recreation. The Supreme Court has stated: "Environmental injury, by its nature, can seldom be adequately remedied by money damages and is often permanent or at least of long duration, i.e., irreparable." Amoco Prod. Co. v. Village of Gambell, Alaska, 480 U.S. 531, 545, 107 S. Ct. 1396, 1404, 94 L. Ed. 2d 542 (1987).

The district court found that Texaco's investigations of the causes of the violations were haphazard and incomplete. Often, possible causes of violations were not investigated or reported, but simply chalked up to sampling and testing errors. These shoddy investigative techniques almost guaranteed that permit violations would not cease. Although the number of Texaco's permit violations declined in recent years, the district court properly considered the entire record and reasonably found that the NRDC had demonstrated irreparable injury.

Although conceding that it has never been and does not expect to be in full compliance with the Act, Texaco points to the EPA Training Manual for NPDES Permit Writers, which states that permissible effluent levels are set so that a facility is statistically expected to discharge in excess of its permit one to five percent of the time. App. at 391. Other EPA documents contradict this statement. See, e.g., EPA, Development Document for Effluent Limitations Guidelines 146 (April 1974), App. at 310 ("short-term limits (30 day average or daily) ... should never be exceeded" by properly operating plants). More importantly, however, we are interpreting Texaco's permit and the Clean Water Act, not reviewing the EPA's methodology for setting permit limits. The permit unequivocally requires complete compliance, see App. at 184, as does the Act, see 33 U.S.C. Sec. 1311(a). Therefore, injunctive relief is appropriate to "secure prompt compliance with the Act." Romero-Barcelo, 456 U.S. at 320, 102 S.Ct. at 1807. Texaco even undermines the reasonableness of its reliance on the EPA Manual by conceding that "[t]he fact that some violations are expected by the agency does not excuse such violations as a matter of law." Texaco Opening Brief at 44.

Finally, Texaco challenges the broad scope of the injunction. The district court enjoined Texaco from violating the 1989 permit in any respect. Since the district court properly determined that forty-nine violations were wholly pre-complaint violations, an injunction that proscribes violations of these parameters is overbroad. In the context of a citizen suit, intended to be supplemental in nature, a district court may not grant any relief, monetary or injunctive, with respect to parameters over which it lacks Article III or subject matter jurisdiction.11 Moreover, the court could not grant equitable relief with respect to parameters that have never been violated or claims that are moot. On remand, therefore, the district court must narrow the scope of the injunction to order compliance only with the parameters that were being continuously or intermittently violated, and which have not been rendered moot by virtue of the relaxed standards in the superseding 1989 permit.12

Texaco argues that the district court overcounted the days of violation in several respects. Initially, we note that the NRDC withdrew all claims relating to alleged violations prior to March 10, 1983. The district court, however, found four violations that occurred on March 2, 1983. Both parties agree that the four days of penalties imposed for these violations must be reversed.

Texaco's principal overcounting argument challenges the district court's method for calculating the penalties for average daily limit violations at outfall 002. The permit defines daily average discharge as "the total discharge by weight during a calendar month divided by the number of days in the month that the production or commercial facility was operating." App. at 155. If the daily average discharge for a given month exceeded the permissible level, the district court assessed penalties for each day of that month even if the facility only operated a few days each month.

Although the permit clearly delineates how to calculate the daily average limit, neither the permit nor the Clean Water Act states how many days of penalties should be assessed for a daily average limit violation. We therefore are left without guidance to predict how DNREC intended penalties to be imposed.

Since the daily average limit is computed by averaging effluent levels only for days on which the facility operated, it seems only logical and fair to assess penalties for a violation of this limit based on the number of days the facility was in operation. Otherwise, the amount of the penalty would be disproportionate to the environmental harm caused by the pollution. For example, if Texaco exceeded its daily average limit by ten percent because it discharged in excess of its permissible average for a parameter on a single day in a month, this violation certainly results in less environmental harm than if Texaco exceeded its daily average limit by ten percent every day in a given month. The district court's formula, however, would assess equal damages under these two scenarios.13 We therefore hold that violations of the daily average limit result in penalties only for the number of days within the month that the facility operated.14

For the foregoing reasons, we will affirm in part and reverse in part the judgment of the district court, and remand for further proceedings consistent with this opinion.