A group of public agencies and private companies based in Indiana and Ohio have petitioned this court for a review of a final decision issued by the United States Environmental Protection Agency (EPA) regarding state regulatory schemes governing toxic discharges into the Great Lakes. For the reasons set forth below, we deny the petitions for review.

This litigation arises under the Clean Water Act, the purpose of which is to "restore and maintain the chemical, physical, and biological integrity of the Nation's waters." 33 U.S.C. § 1251(a). Section 402 of the Clean Water Act established the National Pollutant Discharge Elimination System (NPDES), which obligates polluters to obtain permits for their discharges. Each permit issued under NPDES must include restrictions on pollution discharge and, if necessary, water effluent toxicity (WET) limitations. Although the regulations under NPDES allow a permitting authority to impose a WET limitation whenever a discharge has at least a "reasonable potential" to lead to a water quality violation, they do not establish a specific procedure for the permitting authority to follow. The permitting authority is instead required to take into account certain enumerated factors.

In the late 1980s, the governors of the eight states surrounding the Great Lakes entered into an agreement to protect and preserve the environmental integrity of the Great Lakes waters. Congress followed up by enacting the Great Lakes Critical Programs Act of 1990, which amended Section 118 of the Clean Water Act. 33 U.S.C. § 1268. In this amendment, Congress instructed the EPA to promulgate regulations to ensure that water quality procedures in the states around the Great Lakes would be "no less restrictive" than the already-existing water quality criteria. 33 U.S.C. § 1268(c)(2)(A). Congress also provided that "the Great Lakes States shall adopt water quality standards, antidegredation policies, and implementation procedures for waters within the Great Lakes System which are consistent with such guidance." 33 U.S.C. § 1268(c)(2)(C).

The EPA set forth these rules in a regulation titled the "Final Water Quality Guidance for the Great Lakes System" (the Guidance), found at 60 Fed.Reg. 15,366 (Mar. 23, 1995). Specific provisions of the Guidance govern individual discharges of toxic pollutants into Great Lakes waters. Following the issuance of the Guidance, the eight Great Lakes states were required to adopt and submit to the EPA implementation procedures of their own. The states have the flexibility to create and modify their own regulatory schemes, so long as the net level of protection offered by their provisions match or exceed those promulgated under the Guidance. See 40 C.F.R. § 132.5(g)(3). If the states fail to establish regulations consistent with the Guidance, Congress has mandated that the EPA impose its own standards on the states. 33 U.S.C. § 1268(c)(2)(C).

Under the Guidance, the procedures for establishing WET limits are roughly as follows: Samples are taken from a facility's discharge. Marine organisms are then exposed to these samples as well as to control samples, and tests are done to determine what percentage of these organisms die upon exposure. A "toxic unit," which is a quantitative value, is based upon the percentage of deaths that occur. The highest daily "toxic unit" value measured in the discharge (or, in some cases, the highest weekly value) is then multiplied by a predetermined statistical variable, which depends upon the number of the data points and the variation in results. If the resulting value is greater than the EPA's predetermined criterion, the permitting authority is required to impose a WET limitation.

Indiana submitted its rules to the EPA in late 1997. The EPA subsequently criticized Indiana's proposed procedure because, unlike the Guidance, which used the maximum value from all of the WET tests during the relevant time period, Indiana's procedure would use the geometric mean of all values sampled. In addition, the EPA disapproved of the fact that Indiana's procedure did not employ a statistically based multiplier in its analysis. The EPA thus concluded that Indiana's proposed procedures were inconsistent with the Guidance and issued a final order to that effect on August 4, 2000.

Ohio also submitted its rules to the EPA in late 1997 and, like Indiana, part of its proposal was rejected as being inconsistent with the Guidance. Unlike the Guidance's rigid statistical test, the Ohio test adopted a "weight-of-the-evidence" approach that would require the permitting authority to look to a variety of factors in deciding whether to impose a WET limit. The EPA found that this approach granted too much flexibility to the permitting authority, noting that it would "devalue observed WET test results and would not require a [limit] even where WET test results show observed levels of unacceptable toxicity." Approved and Disapproved Elements of the Great Lakes Guidance Submissions, 65 Fed.Reg. 47864, 47867 (Aug. 4, 2000). It also faulted the fact that Ohio did not employ a statistically based multiplier. The EPA concluded that the procedures were inconsistent with the Guidance in the same final order that it issued regarding the Indiana procedures.

On December 1, 2000, approximately four months after the final EPA order was issued, a group of Ohio entities (hereinafter the Ohio petitioners) filed suit in this court challenging the EPA's conclusion that the proposed Ohio rules were inconsistent with the Guidance. A similar group of Indiana petitioners, united under the moniker Indiana Water Quality Coalition (IWQC), filed suit on December 18, 2000 in the United States Court of Appeals for the Seventh Circuit, challenging the EPA's final decision that the Indiana proposed rules were inconsistent with the Guidance. Pursuant to 28 U.S.C. § 2112(a)(1) and (5), the EPA moved to transfer the case to this court, where the Ohio petitioners had already filed a similar petition. The Seventh Circuit granted the EPA's petition on April 16, 2001. This court accepted the transfer on May 15, 2001, consolidating the two actions into the present case.

This case is governed by the Federal Administrative Procedure Act (APA), 5 U.S.C. § 706. Under the APA, courts will not set aside a final agency action unless it is found to be "arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law." 5 U.S.C. § 706(a).

This standard is deferential. "Even when an agency explains its decision with less than ideal clarity, a reviewing court will not upset the decision on that account if the agency's path may reasonably be discerned." Alaska Dep't of Envtl. Conservation v. EPA, 540 U.S. 461, 124 S. Ct. 983, 1006, 157 L. Ed. 2d 967 (2004) (quotation marks omitted). In considering whether an agency rule is "arbitrary and capricious," we may consider whether

the agency has relied on factors which Congress has not intended it to consider, entirely failed to consider an important aspect of the problem, offered an explanation for its decision that runs counter to the evidence before the agency, or is so implausible that it could not be ascribed to a difference in view or the product of agency expertise. The reviewing court should not attempt itself to make up for such deficiencies; we may not supply a reasoned basis for the agency's action that the agency itself has not given.

Motor Vehicle Mfrs. Ass'n v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43, 103 S. Ct. 2856, 77 L. Ed. 2d 443 (1982) (quotation marks omitted). This court has also noted that it will "defer in large part to EPA's scientific findings." BP Exploration & Oil v. EPA, 66 F.3d 784, 792 (6th Cir.1995). Although the court's review is to be "searching and careful, the ultimate standard of review is a narrow one. The court is not empowered to substitute its judgment for that of the agency." Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U.S. 402, 416, 91 S. Ct. 814, 28 L. Ed. 2d 136 (1971), questioned on other grounds by Califano v. Sanders, 430 U.S. 99, 105, 97 S. Ct. 980, 51 L. Ed. 2d 192 (1977).

The EPA asserts at the outset that this consolidated case is nothing more than an attempt by the petitioners to have this court review the validity of the Guidance itself, as opposed to a review of the Indiana and Ohio regulatory schemes. Under the Clean Water Act, any petitions for judicial review of the Guidance, which was enacted in 1995, had to have been filed within 120 days of its promulgation. 33 U.S.C. § 1369(b)(1). If the EPA is correct, therefore, then we would have to dismiss the present case as a result of it being filed nine years too late.

We find this argument unconvincing. The petitioners have repeatedly conceded that their challenge is limited to the reasonableness of the EPA's finding that the Indiana and Ohio implementation procedures are inconsistent with the Guidance. Although the EPA is correct in noting that many of the petitioners' arguments originate in the language of the Guidance, this is because the Guidance provides the frame of reference to which the Indiana and Ohio procedures must be compared. That the Guidance is held to some degree of scrutiny does not necessarily make this action a challenge to its validity. To the contrary, petitioners' briefs nowhere suggest that they contest the legitimacy of the Guidance. Dismissing this case on the basis that it is an overbroad challenge to the Guidance itself would thus be inappropriate.

C. Res judicata and collateral estoppel arguments

The EPA further argues that the petitioners are barred by the doctrines of res judicata and collateral estoppel from challenging the Guidance on account of American Iron & Steel Institute v. EPA, 115 F.3d 979 (D.C.Cir.1997), in which the Court of Appeals for the District of Columbia rejected the claims of various petitioners, including the petitioners in the present case, that the Guidance was an abuse of the EPA's authority. It argues that the petitioners should have raised in that case any challenges they had about the "validity and level of protection afforded" by the Guidance, and that "issue preclusion" bars the relitigation of any claims in that case. The EPA concludes that "[o]nly the reasonableness of EPA's finding that Indiana and Ohio's implementation procedures were not as protective to the environment as the Guidance's corresponding procedure falls within the scope of this proceeding."

The petitioners, however, do not challenge the EPA's position on the scope of this case. They admit that the present suit is, indeed, nothing more than a challenge of the narrow administrative ruling regarding the Indiana and Ohio implementation procedures. The Ohio petitioners even observe that "[n]ot only do the Ohio Petitioners not seek to `revisit' the holding of [American Iron], they actually rely upon that case and cite to it in their Opening Brief for one of the very principles that EPA erroneously claims to be under attack." Furthermore, as discussed above, the present suit cannot be appropriately deemed to be a suit against the validity of the Guidance. Nothing in the briefs of the petitioners suggests otherwise. Because the EPA has little support for its conclusion that the petitioners are, in fact, challenging the legitimacy of the Guidance as was done in American Iron, we decline to dismiss the case under the theories of either res judicata or collateral estoppel.

IWQC and the Ohio petitioners argue that the EPA erred when it concluded that their regulatory schemes were inconsistent with the Guidance. All of the parties agree that to be "consistent with" the Guidance, the regulatory scheme in question must be at least as protective of the environment as the Guidance. See 40 C.F.R. 132.5(g)(3) ("The Great Lakes State[s]... may adopt provisions that are more protective than those contained in this part."); American Iron, 115 F.3d at 988 (deferring to the EPA's assessment that to be "consistent with" the Guidance, the regulatory regime in question must be "as protective as" the Guidance itself). Accordingly, both IWQC and the Ohio petitioners assert that the proposals submitted by their respective states are at least as protective of the environment as the Guidance, if not more so. Because of variations between the Indiana and Ohio regulations, each will be examined in turn.

Indiana's proposed scheme differs from the Guidance in two key ways. First, under Indiana's proposal, "toxic unit levels" required to assess water toxicity are measured using a geometric mean of tested values. The EPA's Guidance, by contrast, uses a maximum tested value to ascertain toxicity, a value that is, by mathematical definition, greater than the geometric mean. Under Indiana's system, then, a facility's effluent would have to be consistently higher than the mean before a permitting authority would impose a WET limit. Second, the Indiana scheme does not employ the statistical multiplier called for by the Guidance, using instead its own independently derived factor. IWQC asserts that Indiana's scheme is more protective of the environment than the Guidance.

The EPA, however, observed in its final decision that Indiana's procedure, which calls for using a mean value of toxicity, "lessens the impact of observed toxicity on the [limitation] calculation and fails to account for the reasonable possibility that effluent may exceed the level observed in the test because sampling did not coincide with periods of maximum toxicity." Approved and Disapproved Elements of the Great Lakes Guidance Submissions, 65 Fed.Reg. 47864, 47868 (Aug. 4, 2000). A maximum value taken from a sample will necessarily be greater than the geometric mean of the sample.

IWQC also argues that the EPA erred because its "analysis was based on a review of only one small part of Indiana's WET procedures: the use of a geometric mean of the measured effluent values." In reviewing the EPA's analysis, IWQC "see[s] no indication that EPA has conducted a holistic review of Indiana's WET procedures and compared them to EPA's procedure, as the EPA Guidance contemplates." IWQC then points to numerous provisions in the proposed Indiana rules that, according to its brief, make the scheme more protective than the Guidance.

But there is no reason to believe that the EPA acted in an arbitrary or capricious manner in evaluating the Indiana regulatory scheme. The EPA points to a long list of correspondence in which it considered and commented on various aspects of Indiana's proposal. It noted, for example, the fact that Indiana would "require multiple failures" of a toxicity test before a WET limit could be imposed. EPA's Analysis of Steps Taken By Indiana in Response to EPA's 90-Day Letter 9 (July 31, 2000). The EPA also considered, and rejected, the argument that Indiana's catch-all rule, which would allow a permitting authority to "exercise best professional judgment" to enact WET limits, would correct what the EPA viewed as the deficiencies in the rest of the scheme. Id. at 10. This catch-all provision, the EPA noted, "is not a substitute for having a reasonable potential procedure that is as protective" as the Guidance. Id. IWQC's charge that the EPA failed to view the proposed regulatory scheme in its entirety is therefore unpersuasive and does not demonstrate that the EPA acted arbitrarily or capriciously.

In sum, the EPA acted rationally and on the basis of considerable evidence when it rejected Indiana's regulatory scheme. Given that Indiana's averaging of toxicity will call for fewer WET limits than a system using maximum values, the EPA's conclusion that the state's scheme would be less protective of the environment was far from being arbitrary or capricious. Because the "agency's path may reasonably be discerned," Alaska Dep't of Envtl. Conservation, 124 S.Ct. at 1006, there is no basis for us to grant IWQC's petition for review.

Under Ohio's proposed regulations, the permitting authority must take into account a variety of data on the toxicity of the discharge before issuing a permit or establishing a WET limit. The factors to be considered under this "weight of the evidence" standard include the "magnitude of discharge," the "degree and type of effects," the "quality and quantity of each type of data," and "[o]ther relevant factors." Only if the discharge in question meets a variety of criteria will WET limits be imposed. The Ohio petitioners argue that this system, with its comprehensive approach to water quality, is more protective of the environment than the imposition of the statistical multiplier required by the Guidance.

Adopting Ohio's "weight of the evidence" approach, however, necessarily grants the permitting authority wide discretion not afforded by the Guidance. The EPA can rationally read Ohio's rule as failing to provide the permitting authority with guidance on how to weigh the various factors. This discretion arguably makes the Ohio regulatory scheme less protective than the Guidance, in which a statistically determined multiplier is used to determine toxicity.

Ohio's regulatory scheme further limits the permitting authority from imposing a WET limit in the absence of "biological data," except where "[1] the maximum observed toxicity value is at least three times greater than the expected toxicity limit, [2] the average toxicity exceeds one third the expected effluent limit, [3] and more than 30 percent of the test results exceed a projected wasteload allocation." Approved and Disapproved Elements of the Great Lakes Guidance Submissions, 65 Fed.Reg. 47864, 47867 (Aug. 4, 2000). The Ohio petitioners concede that this provision may produce results that are "less restrictive than the Guidance." But this fact, they argue, is balanced out because "if the data set is large, ... Ohio's procedure will require a WET limit where the [Guidance] would not." They further point out that "the [Guidance] will not require a limit even if the average values are much higher than those that Ohio considers to be significant." Ultimately, they argue, the EPA's rejection of this scheme was arbitrary and capricious because the "[a]doption of a more protective element in one provision may be used to offset a less protective element in the same provision." 40 C.F.R § 132.5(g)(3).

The Ohio petitioners concede, however, that their proposed scheme would allow certain toxic discharges to go unregulated in instances where no "biological data" was available. "Ohio's procedure," the EPA concluded, "would not require a reasonable potential finding even where testing has shown actual, observed toxicity. This is clearly inconsistent with [the Guidance.]" Approved and Disapproved Elements of the Great Lakes Guidance Submissions, 65 Fed.Reg. 47864, 47867 (Aug. 4, 2000). Because this provision allows pollution where the Guidance would not, the EPA's final action disapproving the regulatory scheme can hardly be considered arbitrary or capricious.

Moreover, the Ohio petitioners' argument that their "weight of the evidence" approach might be more protective than the Guidance in certain hypothetical situations is unpersuasive. The fact remains that Ohio's scheme "fail[s] to require a limit even in cases of observed toxicity." Id. In its final ruling, the EPA identified this weakness as a significant and determinative flaw in the regulatory scheme. It therefore had little need to address the hypothetical situations where the Ohio scheme might be marginally more protective than the Guidance.

The Ohio petitioners are correct in their observation that there is no discussion of the Ohio scheme's more protective elements in the EPA's final disapproval. But this fact alone does not make the EPA's final agency decision arbitrary or capricious. To the contrary, the tone of the EPA's final decision indicates that the Ohio procedure allowing toxic discharges is so antithetical to the very goal of the Guidance that the more protective provisions would have been insufficient to neutralize its effects.

In sum, we find nothing in the record indicating that the EPA acted in an arbitrary or capricious manner in rejecting Ohio's proposed scheme as inconsistent with the Guidance. The Ohio petitioners have simply not met their burden of establishing sufficient grounds to set aside the EPA's final action.

E. The "scientifically indefensible" exception

[f]or any pollutant ... for which the State or Tribe demonstrates that a methodology or procedure in this part is not scientifically defensible, the Great Lakes States and Tribes shall: (1) Apply an alternative methodology or procedure acceptable under 40 CFR part 131 when developing water quality criteria; or (2) Apply an alternative implementation procedure that is consistent with all applicable Federal, State, and Tribal laws.

40 C.F.R. § 132.4(h). IWQC and the Ohio petitioners contend that the system of WET testing espoused by the Guidance is "not scientifically defensible" because of the huge variation introduced by taking only maximum tested values and by utilizing an independent statistical multiplier. They argue that their schemes are appropriate alternatives sanctioned by the Guidance.

The EPA responds by submitting that the petitioners are in effect making a wholesale challenge to the Guidance, a position that is barred both by the statute of limitations and by American Iron & Steel Institute v. EPA, 115 F.3d 979 (D.C.Cir.1997). We again find the EPA's argument unpersuasive. The petitioners do not challenge the applicability of the Guidance generally; they instead argue that § 132.4(h) allows them to adopt an alternative scheme should they be able to prove that the WET-testing provision provided by the Guidance is scientifically indefensible.

But the EPA makes a compelling argument when it points out that the provision allowing alternative schemes was developed because "there may be pollutants identified in the future for which some of the methodologies or procedures [in the Guidance] may not be technically appropriate. Under these circumstances, EPA wishes to provide sufficient flexibility for permitting to address these pollutants on a case-by-case basis." Proposed Water Quality Guidance, 58 Fed.Reg. 20802, 20843 (Apr. 16, 1993) (emphasis added); see also EPA, Water Quality Guidance for the Great Lakes System: Supplementary Information Document, EPA-820-B-95-001 58 (Mar.1995) ("The reason for this exclusion is that there may be pollutants identified in the future for which some of the methodologies or procedures in the final Guidance may not be technically appropriate.") Moreover, the EPA notes that this exception for scientific indefensibility was intended to be "applied to a specific situation," id., in one or more sites, not to a sweeping alternative regulatory scheme.

WET discharges, which have concerned the EPA for decades, do not fall under the category of "pollutants identified in the future." The "scientific indefensible" exception was not designed to apply to such run-of-the-mill discharges, but rather only on a case-by-case basis if, for whatever reason, applying the Guidance would be inappropriate. This exception to the Guidance is therefore inapplicable to the petitioners' regulatory schemes.

For all of the reasons set forth above, we deny the petitions for review.