Montana Court's Rememdy for Tardy TMDL Development

Thomas J. Mikula
Timothy G. Lynch
Shea & Gardner
Washington, D.C.

Mr. Mikula is a partner and Mr. Lynch is an associate at Shea & Gardner where they provide environmental counseling and litigation services to companies and trade associations. The views expressed in this article are those of the authors and should not be attributed to the firm or its clients.

Introduction

Section 303(d) of the Clean Water Act (CWA) requires states to identify waterbodies for which technology-based effluent limitations alone are not stringent enough to implement water quality standards, such waters being referred to as "water quality limited segments," or "WQLSs." Once they have identified WQLSs in their borders, states must determine and establish the maximum amount of certain pollutants (from point sources and, in EPA’s view,non-point sources) each WQLS can receive daily and still implement state water quality standards. Those maximum amounts, or "total maximum daily loads" (TMDLs), are to be developed for each pollutant impairing each WQLS. Under federal law, states were to have completed their initial submission of identified WQLSs and appropriate TMDLs on or before June 26, 1979. For a variety of reasons, however, many states failed to meet that deadline and are still in the process of developing TMDLs.

Environmental advocates have lamented the slow pace of TMDL development. They see the TMDL process as essential to ensure that environmental agencies take a big picture approach to achieving the CWA’s goals. As a result, citizen suits have been commenced in several states against EPA, claiming that the agency should not be allowing states to proceed so slowly under § 303(d). In general, these citizen suits have had considerable success in forcing EPA (usually by way of a consent decree after a ruling on liability) to develop a schedule for completing the TMDL process in particular states if the states fail to develop the TMDLs. In fact, the success of these citizen suits was one of the reasons EPA has overhauled its TMDL regulations.

Recently, a federal court in Montana took a giant step further. After holding that EPA had violated the Administrative Procedure Act by approving Montana’s submissions of only 130 out of an estimated 3,000 TMDLs necessary for the State’s WQLSs, the court in Friends of the Wild Swan v. U.S. EPA, No. 97-35-M-DWM (D. Mt.), fashioned an extraordinary and unprecedented remedy: neither the State nor EPA may issue any new permits or increase permitted discharges under the NPDES program until all necessary TMDLs are established for a particular WQLS. Whatever one thinks about who’s to blame over the pace of TMDL development, the court’s decision to link permits and TMDLs in this way is, in our view, contrary to law and sound policy.

Proceedings in Friends of the Wild Swan

In 1992, Montana submitted to EPA a list of 322 WQLS based on a partial assessment of its waterbodies. The State’s 1996 submission identified approximately 858 WQLS and 1 TMDL, and its 1998 submission identified 900 WQLS and 130 TMDLs. EPA has estimated that, with three or four pollutants impairing each of the WQLSs, Montana will have to develop roughly 3,000 TMDLs.

In May 1997, Montana’s legislature set deadlines for the development of TMDLs in the State. Under the State plan, TMDLs must be developed by the year 2007 for all WQLSs identified in 1996. Under Phase 1 of the plan, TMDLs will be developed for all high and medium priority waterbodies identified in the 1996 list, as well as certain low priority waterbodies. Under Phases 2 and 3 of the plan, Montana will develop TMDLs for all remaining waterbodies on the 1996 list. Further, the plan requires the State to develop TMDLs for any new WQLS within ten years of its being identified.

Dissatisfied with the pace of TMDL development in the State, a number of environmental groups brought suit in February 1997 in the United States District Court for the District of Montana to compel EPA to assume Montana’s § 303(d) obligations and identify all WQLSs and develop appropriate TMDLs for the State within three years. Their complaint asserted three claims: one under § 505(a) of the CWA, alleging that EPA breached its duty to identify Montana’s WQLSs and develop appropriate TMDLs; another under § 706(1) of the APA, alleging that EPA’s failure to identify Montana’s WQLSs and develop corresponding TMDLs constituted "agency action unlawfully withheld or unreasonably delayed"; and another under § 706(2) of the APA, alleging that EPA’s approval of Montana’s allegedly deficient WQLSs and TMDLs was arbitrary and capricious in a number of respects. The Montana Department of Environmental Quality and a number of industry groups intervened on EPA’s side.

In November 1999, the court granted summary judgment to EPA on some of the claims and granted summary judgment for the plaintiffs on other claims. On the § 505(a) claim, the court ruled for EPA. That provision authorizes citizen suits against EPA when the agency fails to perform a nondiscretionary duty. Plaintiffs argued that EPA failed to perform its allegedly mandatory duty to identify WQLSs and develop TMDLs under two theories: "constructive submission" and "inadequate state action."

The constructive submission theory was first recognized in Scott v. City of Hammond, 741 F.2d 992 (7th Cir. 1984), which held that a state’s failure to submit lists of WQLSs and TMDLs over an extended period of time constitutes a "constructive submission" by the state that no WQLSs exist and TMDLs are unnecessary – thus triggering a duty on the part of EPA to (1) determine for itself whether there are impaired waterbodies and (2) develop any necessary TMDLs. The court in Friends of the Wild Swan distinguished that theory by following other cases which have held that a state makes such a "constructive submission" only when it has submitted no WQLSs or TMDLs before the suit was commenced. Montana, by contrast, had submitted 858 WQLSs and 1 TMDL before the suit was filed, and, shortly thereafter, submitted 900 WQLSs and 130 TMDLs.

The court also rejected an "inadequate state action" theory. Plaintiffs alleged that Montana’s submission of an insufficient number of WQLSs and TMDLs imposed a duty on EPA to prepare a complete list of WQLSs and TMDLs for the State. The court held that EPA has only two affirmative duties once a state submits any WQLSs or TMDLs: (1) it must review the submission within 30 days, and (2) it must identify appropriate WQLSs and develop corresponding TMDLs within 30 days of disapproving a state’s submission. See 33 U.S.C. § 1313(d)(3). Since there was no affirmative duty to prepare WQLSs and TMDLs with respect to a state, like Montana, that had submitted WQLSs and TMDLs, plaintiffs had no claim under § 505(a) of the Act that EPA had failed to perform a nondiscretionary duty.

The court also granted summary judgment to EPA on plaintiffs’ similar claim that EPA’s failure to identify Montana’s WQLSs and develop appropriate TMDLs constituted "agency action unlawfully withheld or unreasonably delayed," in violation of § 706(1) of the APA. The court held that, because Montana had submitted WQLSs and TMDLs for approval, EPA’s only mandatory duties were to approve the submissions within 30 days or to disapprove the submissions and develop alternatives 30 days thereafter. EPA had fulfilled those duties when it approved Montana’s 1992 submission of WQLSs and its submission of a TMDL in 1996. Thus, plaintiffs had no claim that EPA had unlawfully withheld or unreasonably delayed its fulfillment of a mandatory duty.

The court then granted EPA summary judgment in part and plaintiffs summary judgment in part with respect to plaintiffs’ claim under § 706(2)(A) of the APA that EPA’s approval of Montana’s 1998 submission of WQLS and TMDLs was arbitrary and capricious. The plaintiffs had six theories as to why EPA’s actions were arbitrary, and the court held for EPA on five, and for plaintiffs on one. Plaintiffs’ first four theories involved claims that EPA should not have approved Montana’s WQLSs because the State had (1) failed to identify all of its WQLS, (2) failed to identify 83 particular streams as WQLSs based on existing evidence, (3) failed to consider the effect of non-point source pollution on Montana’s cold water fisheries, and (4) failed to respond adequately to public comments. As to each of those claims, the court held that the State and/or EPA had provided reasonable explanations for their actions and had otherwise not acted arbitrarily, and thus EPA had not acted arbitrarily in approving Montana’s submission.

Plaintiffs’ fifth theory was that EPA had acted arbitrarily in approving technically deficient TMDLs. To that end, plaintiffs alleged that Montana’s 1998 list of TMDLs (1) were not prepared in accordance with the State’s own prioritization of WQLSs, (2) did not contain total maximum daily loads, (3) did not contain explicit loading capacities for point source, non-point source, and background sources of pollutants, and (4) did not account for seasonal variations. The court rejected each of these technical challenges to Montana’s TMDLs and EPA’s approval thereof.

But the court found for plaintiffs’ on their sixth theory, holding that it was arbitrary and capricious for EPA to approve Montana’s 1998 submission of only 130 TMDLs for the 900 WQLSs that were listed. The court reasoned that Congress expressly established a tight deadline for the completion of TMDLs, with states being required to submit initial lists of TMDLs in 1979. The court found that at its current rate of development, Montana would "need over one hundred years to develop the 3,000 TMDLs required for the WQLSs identified in 1998." Having found that Congress placed a firm emphasis on the expeditious development of TMDLs, the court held that EPA’s approval of Montana’s tardy and incomplete submission of TMDLs was arbitrary and capricious.

The court issued its decision and asked the parties to file briefs suggesting a proper remedy. The court’s June 2000 Order specifying the remedy first rejects as too intrusive the plaintiffs’ suggestion that it appoint a special master to oversee a TMDL development schedule that would involve a rate of approximately 15 percent of identified WQLSs per year. It also rejects EPA’s suggestion that the case be remanded to EPA for further action, including, EPA argued, a possible decision that Montana’s TMDL submission was reasonable in light of the State’s current TMDL funding and resources. The court noted that the CWA "features short deadlines" and that "[t]he State has failed to meet * * * deadlines in the past."

The court’s remedy instead requires EPA and Montana jointly to develop a schedule that will result in the completion by 2007 of all TMDLs required for the WQLSs identified in 1998. Significantly, the court’s order purports to give Montana "the opportunity to prioritize among the sites on the 1996 list and to develop TMDLs for EPA approval." But, in an unprecedented move, the court also stated, without citation to any case or other authority:

Until all necessary TMDLs are established for a particular WQLS, the EPA shall not issue any new permits or increase permitted discharge for any permittee under the National Pollutant Discharge Elimination System permitting program.

In September 2000, at plaintiffs’ request, the court amended that paragraph also to restrict Montana’s permitting authority (since Montana, not EPA, is principally responsible for issuing CWA § 402 permits in that State).

EPA and Montana asked the court to amend the paragraph to add the phrase "unless such WQLSs are removed from the list," on the reasonable assumption that Montana should not devote resources to developing TMDLs for waterbodies that are no longer WQLSs. In reasoning that is somewhat unclear, the court rejected defendants’ request, holding that allowing Montana to delete WQLSs would mean that "EPA’s responsibility [to ensure the development of TMDLs] * * * will not be clear until May 4, 2007." Left unanswered was how the deletion of a particular WQLS would make EPA’s responsibilities with respect to the remaining WQLSs unclear. In a subsequent opinion, the court characterized the request to develop TMDLs only for WQLSs that are not removed from the list as "a further request to continue the license to do nothing."

EPA and Montana then moved for a stay of the Order pending appeal. With its motion, the State submitted affidavits from State employees who alleged that the court’s permit bar would prevent the State from issuing discharge permits or authorizations under general stormwater permits necessary for federally funded highway construction projects, as well as other important projects. In fact, on September 29, 2000, the State canceled $28 million in highway construction projects on the ground that it could not issue necessary authorizations under the court’s June 2000 Order. The State furthered asserted in its motion for a stay that another $240 million in future highway construction projects could not be started because of the Order. Plaintiffs opposed the stay, arguing that the court’s remedy did not prohibit authorizations to proceed under a general permit for storm water discharge that covers highway construction projects and that the State never sought guidance from the court as to whether the remedy required the State to terminate highway construction projects.

The court denied defendants’ motion for a stay. It held that defendants did not show a likelihood of success on the merits with respect to their assertion that the court’s remedy inappropriately interfered with EPA’s discretion. As the court noted, "[w]hat TMDLs are ‘necessary,’ what should be the contents of each TMDL, and whether proposed permits will or will not affect a particular WQLS’s impairment, are matters left to the State’s and the EPA’s discretion." Moreover, in the court’s view, "[a]lthough the Clean Water Act does not require completion of TMDLs before permits can be issued," "[e]very new or increased-discharge permit issued for a WQLS site after June 26, 1979, would have and should have been preceded by a TMDL" under the deadlines established by Congress.

The court also declined to issue a stay based on its view that the defendants had not shown a likelihood of irreparable harm absent a stay. The court asserted that, even if new permits or discharges could not be issued until TMDLs were complete, "[t]he State may identify and prioritize those WQLSs where projects that will require permits are contemplated and develop TMDLs for those WQLSs first." Since new permits or requests for increased discharges entail a 180-day waiting period, the court noted, Montana and EPA would have ample time to develop TMDLs for any WQLSs in anticipation of issuing a new permit. Moreover, the court held, its Order did not apply to highway construction projects, because these projects merely required authorizations under an existing general storm water permit, not new permits.

Analysis

It seems clear that the court in Friends of the Wild Swan believed that EPA and Montana needed some significant incentive to complete the TMDLs. Moreover, it is evident from the court’s opinion that Montana did not help its litigation posture by not completing any additional TMDLs between the time the court granted summary judgment to plaintiffs in November 1999 and when it declined EPA’s and the State’s motion for a stay pending appeal (although EPA had just proposed major revisions to the TMDL regulations in August 1999 and so Montana could well have been waiting for guidance in the final regulations before completing any additional TMDLs). It is also possible that the court, after linking the discharge permits to the TMDLs, saw Montana’s decision to cancel contracts for federal highway funds on the grounds that permits necessary for such projects would be blocked by the court’s order as either crying wolf or attempting to bully the judge politically.

But absent from the court’s decisions on the remedy is any recognition of how Montana (and other states) have come to be in a position where TMDL development is so far behind the schedule set in place by Congress. As one commentator has explained, "[i]t is no accident that the TMDL requirements of the Clean Water Act were avoided * * * for so many years. The TMDL program is plagued with numerous difficulties," including "the staggering monetary cost of the development and implementation of TMDLs." Diane K. Conway, Note, TMDL Litigation: So Now What?, 17 Va. Envtl. L.J. 83, 103 (1997). In particular, the technical process of developing TMDLs is quite complex (especially if, as EPA believes, Section 303(d) requires consideration of pollutants from non-point source discharges), and rushing through the process to develop a deficient TMDL is in no one’s interest. Moreover, for many years no one was pressing the states to complete the process. And, if anything, the reaction to EPA’s recently proposed and promulgated revisions to the TMDL regulations suggests that it may take considerable time before any political will develops in favor of strong implementation. Yet the court’s opinion takes no account of how EPA and the states have come to a position in which they are still developing TMDLs and, instead, reads as though the court believed EPA and Montana had the specific intent to flout its TMDL responsibilities for twenty years.

Putting aside the ahistorical nature of the Friends of the Wild Swan decision, its permit remedy is unsupported as a matter of law. The CWA nowhere gives courts the authority to condition the issuance of permits on the completion of TMDLs. Section 303(d) itself merely sets forth the steps EPA and the states must take in the development of TMDLs. It does not imply that permits should be withheld pending their completion. In fact, several provisions of the CWA strongly suggest that a restriction on permitting authority is not an appropriate tool to force EPA or states to comply with § 303(d). For instance, CWA § 402(a)(1) authorizes the issuance of National Pollutant Discharge Elimination System (NPDES) permits, upon condition that the discharge will meet all applicable requirements under certain, specified sections of the CWA. Yet § 303 is not among the sections listed therein. Furthermore, § 301(a) of the CWA prohibits the discharge of any pollutant by any person, except in compliance with certain provisions in the CWA. Again, § 303 is not among the sections listed.

The Friends of the Wild Swan court, in effect, acknowledged the absence of supporting authority when it noted that "the Clean Water Act does not require completion of TMDLs before permits can be issued." Indeed, the Supreme Court has gone so far as to reverse a federal appellate court that had read the CWA to prohibit EPA from issuing a § 402 permit for a source that would discharge pollutants into waters already in violation of existing water quality standards, Arkansas v. Oklahoma, 503 U.S. 91 (1992), a decision that ought to preclude a federal court from barring the issuance of discharge permits merely because all appropriate TMDLs are not yet completed for a WQLS.

Some commentators nonetheless have argued that EPA already has regulatory authority to link the permitting and TMDL development process. E.g., Michael Wenig, How "Total" are "Total Maximum Daily Loads"?, 12 Tul. Envtl. L.J. 87, 120-21 (1998). These advocates point to 40 C.F.R. § 122.4(i) (a regulation that was not cited by the Friends of the Wild Swan court), which provides that no permit shall be issued "[t]o a new source or a new discharger, if the discharge from its construction or operation will cause or contribute to the violation of water quality standards." But EPA appears to have not enforced this provision (although the preamble to the recently promulgated TMDL regulations indicates that EPA may begin to do so) and so its authority under this provision has never been analyzed under Chevron U.S.A. v. Natural Resources Defense Council, 467 U.S. 837 (1984). In any event, regardless of whether EPA has the authority to withhold a permit if the discharge would cause or contribute to the violation of water quality standards, § 122.4(i) does not address the power of a reviewing court that is acting contrary to EPA’s views. Furthermore, the remedy in Friends of the Wild Swan went well beyond what § 122.4(i) would authorize. Instead of mirroring the terms of § 122.4(i), the court barred any new or increased discharge permits until TMDLs for a WQLS were complete, regardless of whether the discharge would actually cause or contribute to excursions of water quality standards.

The remedy is also unlawful as being too severe. It is well established that where injunctive relief is appropriate, it should be narrowly tailored to remedy the specific harm shown. Here, the court did not give Montana and EPA sufficient time to show whether they were moving forward in their efforts to comply with its November 1999 Order. Granted, plaintiffs were able to point out to the court that Montana had not developed a single TMDL since that Order, and that fact certainly did not help Montana’s case. But, far short of banning all new permits or increased discharges until TMDLs were developed, the court. For example, could have required Montana and EPA to report on their progress every two years; and after receiving reports that no progress had been made, it could have considered increasingly severe injunctions (although its present remedy would, in our view, never be justified under the CWA, no matter how slowly a state developed TMDLs). A more modest approach also seems justified, given that the court rejected plaintiffs’ procedural and technical challenges to the TMDLs Montana did submit: not only did the State submit a number of TMDLs, but the ones it did submit were up to standard. The court apparently did not take that record into account. Indeed, the court’s effort to claim that its remedy did not affect highway construction projects has the ring of an ex post rationalization. Although the Order does mention "new permits," thus potentially excluding existing general permits, the Order also prevents any increased discharges. To the extent a highway construction project would potentially entail an increased discharge under an existing general stormwater permit, it’s not hard to see why Montana apparently thought the court’s Order applied.

Friends of the Wild Swan’s injunction against new permits also is unsound as a matter of policy. In rejecting the plaintiff’s proposal to have a special master oversee TMDL development in Montana, the court recognized that it should not interfere with EPA’s and Montana’s discretion in implementing the CWA. Such circumspection is warranted under Alaska Center for the Environment v. Browner, 20 F.3d 981, 986-87 (9th Cir. 1994), which affirmed an order requiring Alaska and EPA to develop a schedule for the development of TMDLs because the lower court showed "great restraint" and "was careful to leave the substance and manner of achieving * * * compliance entirely" to the agencies. Thus, in numerous places in its opinions, the Montana court asserts that it sought to preserve EPA’s and Montana’s discretion. But the court’s goal is completely overwhelmed by its remedy. If Montana must devote resources to developing TMDLs any time a new permit or request for increased discharge is before it, it will have to act to develop those TMDLs regardless of how it would otherwise prioritize among needs. Thus, linking permits to TMDLs directly involves the court in the substance and manner of achieving compliance.

The Friends of the Wild Swan decision also displays what can be described as a certain naivete about the length of time it takes to complete a TMDL. In this regard, the court was not aided by the plaintiffs, who submitted an affidavit attesting that TMDLs could typically be completed in a matter of weeks, and asserted therefore that the typical six-month waiting period for a new permit or increased discharge permit was enough time to complete the necessary TMDLs. But TMDLs that are developed on a rushed basis necessarily favor speed over quality. If the goal is to develop accurate models to be used to help ensure that waterbodies meet state water quality standards, time is just one factor to be considered. Of course, time cannot be an excuse never to develop TMDLs. But a six-month period may simply be insufficient for a WQLS if TMDLs will involve complex estimates about pollutants from non-point sources.

Finally, the court’s apparent refusal to allow Montana to avoid promulgating TMDLs for a WQLS that is no longer in violation of water quality standards is unnecessarily punitive and unproductive. The TMDL process is designed to identify and limit pollutants for streams that cannot, using technology-based effluent limits, meet state water quality standards. If a particular stream is no longer so impaired, it is a waste of resources to dedicate time and effort to developing TMDLs for it, when those resources could be placed elsewhere to achieve the CWA’s goals. Moreover, to the extent that the court was concerned that the State might be tempted to de-list WQLSs to avoid having to promulgate TMDLs, such concerns could be addressed through additional litigation. But that is no reason to require the State to develop TMDLs even for waterbodies that are no longer WQLSs.

Conclusion

The CWA established an ambitious schedule for the establishment of TMDLs, one that has not been met by EPA or most states, including Montana. But the reasons for that delay are practical, not sinister. Against this backdrop, the court in Friends of the Wild Swan took a reasonable position in interpreting the CWA as establishing Congress’s preference for a speedy development of TMDLs. But the court’s remedy linking the issuance of new permits or allowing increases in permitted discharges to development of TMDLs is unsupported by law or policy. The court’s decision probably will not be the last word, however, as EPA and Montana have appealed to the 9th Circuit.