On January 3, 2013, the District Court for the Eastern District of Virginia ruled that EPA lacks the statutory authority to set a Clean Water Act (“CWA”) total maximum daily load (“TMDL”) for “stormwater flow rates” as a surrogate for sediment deposition. Virginia Dep’t of Transportation et al v EPA et al.  EPA has decided not to appeal.  The case has received national attention because of its implications for other TMDLs that use surrogates. This article will discuss the decision and its significance for the TMDL and water quality regulatory regime.

The relevant statutory framework is CWA Section 303, under which each state establishes water quality standards for waters within its boundaries.  These consist of a designated use (trout fishing, contact recreation, etc.) and numerical or narrative “water quality criteria” necessary to support that use.  For “impaired waters” where the criteria are not being met, the state must set a TMDL (think “pollution budget”) for each pollutant for which the criteria are exceeded, and implement a “planning process” leading to achievement.  Where the state fails to act, or sets a TMDL which EPA regards as insufficient, CWA Section 303(d)(2) directs EPA to set the TMDL.

Accotink Creek is a 25 mile tributary to the Potomac River in Virginia, in which the benthic organisms were impaired, primarily because of sediment deposited by stormwater running off impervious urban and suburban areas.  In April 2011, after Virginia failed to set a TMDL, EPA set one which limited the flow rate of stormwater into Accotink Creek to 681.8 cu ft/ acre-day.  The court said that the parties agreed that “sediment is a pollutant, and that stormwater is not” (Slip op. 3). While EPA’s brief contains a fallback argument that stormwater can be viewed as a “pollutant”, it did not dispute that stormwater flow was being used as a surrogate for sediment.  Thus the question addressed by the court was whether EPA has the statutory authority to set a TMDL for a “surrogate” which is not itself a “pollutant”. 

EPA has used surrogates in a number of circumstances where, in its view, the surrogate would provide appropriate reduction of pollutants, and would be either easier to measure or provide other benefits (such as, in this case, reduction of stream bank scouring caused by heavy stormwater discharges), or both. The court rejected EPA’s argument that since the CWA does not expressly address the use of surrogates, EPA’s use of them should be upheld as reasonable “gap-filling”, consistent with the broad remedial objectives of the CWA, and entitled to substantial Chevron step 2 deference. The court held instead that because the CWA instructed EPA to set TMDLs for “pollutants”, not “surrogates”, the statute was clear.  The court distinguished EPA’s use of surrogates in this case from other instances  in which surrogates have been used under other CWA provisions (notably Sections 301, 304 and 402) where EPA appears to have greater latitude.

EPA and states have used stormwater surrogates in TMDLs in Connecticut, Missouri and North Carolina. They have also used other types of surrogates, such as impervious surface area limits and secchi disc readings.  Some of those have been challenged, and this decision will no doubt provide ammunition for those who oppose their use.  Nationally, however, this amounts to a very small percentage of the TMDLs that are in place, even if one focuses only on sediment (for which, the court noted, EPA has issued approximately 3700 TMDLs).

In addition, this ruling will have no effect whatever on EPA’s permitting of  industrial and municipal stormwater  discharges, including municipal separate storm sewer systems (“MS4s”), or its ongoing development of stormwater regulations, because these activities are expressly authorized under CWA Section 402(p).  This is especially important, because EPA and many states now recognize stormwater as a major source of contamination and water quality impairment.  For a thoughtful article on this subject and emerging approaches, see Dave Owen, Urbanization, Water Quality, and the Regulated Landscape,  82 U. of Colo. L. Rev.  431 (April 2011).

After decades of sparring over nutrient loading in the Illinois River, and following several short term extensions of a previous truce, Arkansas and Oklahoma recently executed an agreement, the “Second Statement of Joint Principles and Actions”, that establishes a procedural framework for attempting to resolve their long running trans-boundary water quality dispute.

The Illinois River heads up in a rapidly developing section of Northwest Arkansas and flows west into a comparatively undeveloped portion of Northeast Oklahoma, where the river is designated by state statute as a scenic river.  For more than two decades Oklahoma has worked to reduce the amount of nutrients, and particularly phosphorus, discharged into the Illinois River watershed.  In 2002 Oklahoma adopted a numeric water quality criterion for Total Phosphorus that many considered impossible to attain in a developed watershed.  In an effort to avoid litigation over the validity of the numeric criterion, Arkansas and Oklahoma entered into an agreement in 2003 known as the Statement of Joint Principles and Actions.  This agreement provided, among other things, that:  (i) Oklahoma would postpone for 10 years the date on which the numeric criterion would be fully effective; (ii) Arkansas sources would take a number of steps to reduce phosphorus discharges; and (iii) Oklahoma would review the existing numeric criterion, with an opportunity for Arkansas representatives to participate, before the end of the ten year period to determine whether the numeric criterion should be changed.

The ten year truce created by the Statement of Joint Principles and Actions was originally scheduled to expire in July 2012.  During the ten year period Arkansas sources made significant progress in reducing the amount of phosphorus they discharged in the watershed.  As a result, phosphorus levels in the Illinois River began to decline and most observers agreed that conditions in the river were significantly improved.  Towards the end of the ten year period Oklahoma undertook a review, with full participation by representatives of Arkansas, EPA, and the Cherokee Nation.  The review ended in a sharply divided report, with Oklahoma representatives stating that no change in the numeric phosphorus criterion was warranted and Arkansas representative stating that significant change was necessary.

As the end of the ten year truce approached, officials from Arkansas and Oklahoma began negotiations once again on how to avoid litigation.  Focus on the potential for costly litigation was sharpened by the fact that EPA had publicly commenced work on a Phosphorus TMDL for the entire Illinois River watershed.  After several agreements on short term extensions of the July 2012 deadline, Arkansas and Oklahoma reached agreement in February 2013 on a Second Statement of Joint Principles and Actions.  This new agreement provides, among other things, that Arkansas and Oklahoma will fund a joint three year water quality study using EPA protocols to determine the threshold Total Phosphorus levels at which shifts in algal species or biomass production occur that result in undesirable aesthetic or water quality conditions.  Oklahoma and Arkansas agree in the Second Statement to be bound by the findings of the joint study, and Oklahoma agrees to adopt a new numeric criterion for Total Phosphorus in the Illinois River if the results of the joint study are significantly different from the existing criterion (i.e., more than -0.010 mg/l or +0.010 mg/l than the existing .037 mg/l criterion).  During the term of the Second Statement of Joint Principles and Actions, both states agree not to initiate or maintain litigation contrary to the terms of the agreement, and the statute of limitations on all claims is extended.  Oklahoma agrees that it will postpone for the duration of the new agreement the date on which its existing, hotly disputed numeric criterion is to be fully effective.

EPA was not a party to the negotiation of the new agreement and it has not announced any formal position on its effect.  It is not clear what impact the new agreement will have on EPA’s work to develop a TMDL for Phosphorus in the Illinois River watershed or on the various NPDES permits for POTWs on the Illinois River that are currently pending review in EPA Region 6.

On January 3, 2013, an EPA-set TMDL for Accotink Creek (a Fairfax County, Virginia tributary of the Potomac River) was invalidated on the grounds EPA exceeded its statutory authority when it attempted to regulate, via the TMDL, a Clean Water Act pollutant – sediment – by instead regulating a surrogate non-pollutant – stormwater flow.  The opinion granted plaintiff’s motion for judgment on the pleadings; in a separate order Judge O’Grady vacated the Accotink Creek TMDL and remanded the matter to EPA for further consideration.

Utilizing the two-step Chevron statutory interpretation analysis, the court found the first of the two criteria had been met:  Congress had addressed in unambiguous language the precise question at issue “… and its answer is that EPA’s authority does not extend to establishing TMDLs for nonpollutants as surrogates for pollutants.”  While directly acknowledging he did not need to reach the second Chevron criterion (whether the agency’s reading of an ambiguous statute is “permissible”), Judge O’Grady nonetheless noted in some detail that “there is substantial reason to believe EPA’s motives go beyond ‘permissible gap-filling.’”

Based upon EPA’s own pleadings, the opinion notes it appears the Agency could have set a TMDL based upon the underlying problem – sediment in the creek – rather than pursuing a surrogate approach.
In recent years, EPA has been pursuing so-called “innovative” TMDLs in an attempt to address stormwater’s contribution to impaired watersheds.  In particular, those TMDLs purport to set load limits based upon various surrogate approaches, including such things as the percentage of impervious cover (“IC”) allowed in the impacted watershed, or (as with Accotink Creek) stormwater flow rates.  While the Accotink Creek TMDL was in EPA Region 3, New England’s Region 1 seems to be in the forefront of this approach and currently has at least three stormwater-source TMDLs in place (including two so-called IC-TMDLs), as well as others in development. 

Given the extensive resources EPA has invested in trying to manage stormwater impacts to impaired streams (see e.g., TMDLs to Stormwater Permits Handbook (Draft Nov 2008)), the Virginia Dept. of Transportation case is clearly a significant setback for the surrogate approach propounded by the Agency.  Whether the United States appeals the decision or retreats and re-evaluates its initiatives in this area is yet to be determined.  Whichever way EPA decides to go, communities dealing with impaired watersheds certainly will need to pay close attention.

EPA recently issued a new draft vision statement for the Clean Water Act 303(d) program, the program under which impaired water bodies are identified and TMDLs performed.ⁱ   It is fitting that this draft vision statement coincides with the 60th anniversary of the classic episode of I Love Lucy in which Lucy and Ethel struggle mightily to wrap chocolate bonbons as they proceed down a conveyor belt.ⁱⁱ  At first all goes well.  Then the conveyor belt speeds up.  Lucy and Ethel are soon frantically pulling candies off the conveyor belt, stuffing them into their hats, their mouths, and even down the front of their uniforms in order to keep pace with the conveyor belt.

The CWA 303(d) program started off quietly.  States occasionally sent EPA lists of impaired water bodies (the so-called 303(d) lists), and EPA dutifully filed most of them away.  Even a few TMDLs – the studies intended to describe how impaired water bodies should be restored – got performed.  Then came the TMDL litigation.  Dozens of lawsuits in the 1990’s led EPA and the states to begin serious periodic assessments of state water bodies for compliance with water quality standards; and quotas were set by judicial decree or regulatory fiat for the issuance of TMDLs for all of the water bodies found to be impaired.  Soon TMDLs were flying down the conveyor belt at a speed never previously experienced.  EPA reports that we are very close to reaching the milestone of 50,000 TMDLs.

With two decades of experience issuing thousands of TMDLs per year it is reasonable to step back and consider the water quality results obtained.  The simple truth is that restoration of impaired water bodies has not come anywhere close to keeping up with the pace at which TMDLs have been rolling off the bureaucratic assembly line.  There are a variety of reasons for this paucity of results in restoring impaired water bodies.  Frequently the principal causes of impairment identified in a TMDL are non-point source contributions that are largely beyond the direct reach of regulatory programs.  Load reductions prescribed by TMDLs usually require significant expenditures, but funding is rarely available.  Public buy-in for the changes prescribed by TMDLs is often lacking, frequently because the TMDLs are viewed as products of obscure regulatory processes that are not responsive to stakeholder sentiment.  Although the problems addressed by TMDLs are commonly quite complex, the quality of the analysis in some TMDLs is indefensibly poor.  Sometimes the water quality standards that triggered the TMDL are unrealistically ambitious to begin with.  Even when the relevant standards are appropriate, the process of identifying impaired water bodies is commonly mired in bureaucratic minutiae and delay.  States are frequently late in sending in their biennial 303(d) lists to EPA; and EPA is equally delinquent in meeting the 30-day time limit set in its own regulations for completing federal review and approval of state 303(d) submissions.

Given the litany of issues that have limited the success in achieving most TMDL load reductions, it is interesting to read EPA’s new draft vision statement:

“The Clean Water Act Section 303(d) Program provides effective integration for implementation of activities to restore and protect the nation’s aquatic resources, where the nation’s waters have been assessed, restoration and protection objectives have been systematically prioritized, and Total Maximum Daily Load and alternative approaches are being adaptively implemented to achieve water quality goals with the collaboration of States, federal agencies, tribes, stakeholders, and the public.”

In fairness to EPA’s draft, almost all vision statements are insufferably stuffy and meaningless to everyone except those few who were in the room and argued vigorously for the addition or deletion of a particular word or phrase when the inscrutable language was crafted.  Moreover, EPA’s new draft vision statement is accompanied by “goals statements” that offer somewhat more concrete prospects for program improvements at the margin, such as protection of unimpaired waters, increased focus on prioritization, and more flexible approaches to TMDL implementation.  If we want to achieve major improvements in water quality, however, we need visionary leaders not a new vision statement.  We need to spend a lot of money, and we need to spend it in a smart way.  We need new programs that fix the regulatory limitations of the current laws, not repackaged versions of the status quo.  And we need a broad public commitment to achieve change.  In the early 1970s political leaders managed to do these things on a bipartisan basis for the sake of improving the nation’s water quality.  Unfortunately, the current state of partisan gridlock offers little hope for major change in the foreseeable future.

Maybe a new vision statement is the best we can hope for given the strong anti-regulatory sentiment in the current public discourse.  Against this background, one cannot help but reflect on the end of the iconic scene in the chocolate factory.  The supervisor comes into the room, pleased that Lucy and Ethel have not allowed a single bonbon to make it past their work station unwrapped.  She announces “Why you girls are doing splendidly,” and then yells to the conveyor belt operator, “Speed it up a little!”

ⁱ A Long-Term Vision for Assessment, Restoration, and Protection under the Clean Water Act Section 303(d) Program, FINAL DRAFT FOR STATE/EPA REVIEW (1 August 2012).  An earlier version of the same document, a Stakeholder Review Draft dated June 2012, is accessible here.
ⁱⁱ I Love Lucy, Season 2, Episode 4, broadcast on September 15, 1952, accessible here. Copies of the chocolate factory scene are accessible on YouTube.

There is nothing new about transboundary water quality disputes under the Clean Water Act. Introductory classes on environmental law commonly trace the history of Supreme Court decisions arising from Milwaukee’s battles with Illinois over sewer discharges into Lake Michigan, the challenge Vermonters’ raised against New York paper mill discharges into Lake Champlain, and Oklahoma’s objections to a permit EPA issued to a sewage treatment plant in Arkansas. Given the length of time that the Clean Water Act program has been in place and the large number of instances in which upstream discharges drain into and therefore arguably affect water quality in downstream states, one would expect that most of the relevant legal questions would be well-settled. Recently, however, transboundary water quality disputes have arisen with increasing frequency.

Coal bed methane development in Wyoming has given rise to disputes with Montana over salinity impacts in the Powder and Tongue Rivers. Efforts by the state of Washington to protect dissolved oxygen levels in Spokane Lake have prompted a dispute over Washington’s attempt to impose wasteload allocations that would limit nutrient discharges by upstream sources in Idaho. Oklahoma’s efforts to restore the Illinois River to pristine scenic river conditions have resulted in recurrent and steadily intensifying disputes with agricultural interests on both sides of the border and point sources located predominantly in the headwaters on the Arkansas side. EPA’s imposition of nutrient water quality standards in Florida could have direct effects on discharges originating in Georgia; and the agency’s showcase multi-state TMDL for the Chesapeake Bay has recently precipitated challenges by state and national agricultural interests. In what undoubtedly is the most dramatic transboundary claim under the Clean Water Act, environmental groups have filed a petition with EPA asking the Agency to impose nutrient water quality standards and adopt TMDLs for the main stem of the Mississippi River, all of its tributaries, and certain related coastal waters in the Gulf of Mexico.
 

These recent disputes have recurring themes that arise out of weaknesses or unresolved questions regarding the Clean Water Act program. The statute empowers each state to exercise sovereign independence in adopting water quality standards that apply within the state’s own borders (so long as minimum federal standards are met), but the statute does little to address or even give consideration to the interests of other states that may be directly affected by those standards. Transboundary disputes frequently involve situations in which the regulatory burdens fall disproportionately on interests in one state while the resulting environmental benefits are realized largely or entirely in another state, but the Clean Water Act does nothing to address questions of transboundary fairness. Transboundary disputes frequently involve regulatory decisions that have enormous financial and long term planning consequences, but the decisions are often based upon limited factual data, imperfect scientific analysis, and less than comprehensive computer modeling. The Clean Water Act offers no process for seeking to assure that the quality of the decision making will be commensurate with the gravity of the consequences at stake. Indeed, the program largely makes the magnitude of financial consequences simply irrelevant. Disparities between the magnitude of the consequences and the limited quality of analysis and data supporting the regulatory decision are particularly problematic when the regulatory decision is being made by one jurisdiction that has no political accountability to the other.

It is perhaps no surprise that most of the recent transboundary water quality disputes are arising out of efforts to regulate the discharge of nutrients. Nutrient pollution is the largest unresolved water quality issue nationally; and the adverse effects of excess nutrients frequently occur at locations far downstream from the original source. The fact that most of the current transboundary water quality disputes involve nutrient pollution probably makes the disputes even more difficult than normal to resolve. Nutrient pollution has no simple, universally accepted means of measurement. It is costly and time consuming to establish a clear causal link between a given discharge of nutrients and an observed adverse effect; and many of the most important sources of nutrient pollution are non-point sources which are beyond direct control under the Clean Water Act. Unfortunately, this appears to be a recipe for increased frustration and controversy.
Transboundary water quality disputes may not be a new phenomenon, but it does not appear that we are any closer to finding a good way to resolve them.