Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 1 of 39 PageID #: 6935 IN THE UNITED STATES DISTRICT COURT FOR THE SOUTHERN DISTRICT OF WEST VIRGINIA HUNTINGTON DIVISION OHIO VALLEY ENVIRONMENTAL COALITION, INC., SIERRA CLUB, WEST VIRGINIA HIGHLANDS CONSERVANCY, INC. and VIRGINIA RIVERS COALITION, Plaintiffs, v. CIVIL ACTION NO. 3:15-0271 GINA MCCARTHY, Administrator, United States Environmental Protection Agency and SHAWN M. GARVIN, Regional Administrator, United States Environmental Protection Agency, Region III, Defendants. MEMORANDUM OPINION AND ORDER I. Introduction Pending in this administrative review action are cross-motions for summary judgment brought by Plaintiffs Ohio Valley Environmental Coalition, Inc., Sierra Club, West Virginia Highlands Conservancy, and Virginia Rivers Coalition (collectively the “Environmental Groups”), ECF No. 30, and by Defendants Gina McCarthy, Administrator of the United States Environmental Protection Agency, and Shawn Garvin, a regional administrator for the same agency (collectively “EPA”), ECF No. 38. In this citizen suit pursuant to the Federal Water Pollution Control Act (Clean Water Act or “CWA”), 33 U.S.C §§ 1251–1388., and under the Administrative Procedure Act (“APA”), 5 U.S.C. §§ 701–706, the Environmental Groups challenge EPA’s failure to Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 2 of 39 PageID #: 6936 disapprove actual or constructive submissions by the West Virginia Department of Environmental Protection (“WVDEP”) that lacked total maximum daily loads (“TMDLs”) for certain West Virginia waterbodies previously identified by WVDEP and EPA as “biologically impaired.” The Environmental Groups seek an order declaring EPA’s alleged failure in violation of the CWA’s process for reviewing state TMDL submissions, 33 U.S.C § 1313, and in violation of the APA’s prohibition on agency action that is arbitrary, capricious, abusive of discretion, and otherwise not in accordance with law, 5 U.S.C. § 706(2). 2d Am. Compl. 35–37, ECF No. 78. The Environmental Groups also request an order declaring EPA’s decision to approve sets of TMDLs for a particular watershed arbitrary and capricious because not all TMDLs for that watershed were submitted to EPA. Id. The Court, in an Order dated September 9, 2016, determined that the Environmental Groups have standing to bring this case. ECF No. 81. On October 20, 2016, the Court heard oral argument on the remaining issues raised in the cross motions for summary judgment. The remaining issues raised by the cross motions for summary judgment require the Court to determine: (1) EPA’s liability on Claims 1 and 2, which allege EPA violated a nondiscretionary duty under the CWA to, first, disapprove WVDEP’s actual or constructive submissions of no biological impairment TMDLs for biologically impaired waterbodies, some of which are impaired specifically by ionic toxicity, and second, to establish those undeveloped TMDLs; and (2) EPA’s liability on Claims 3 through 8, which allege EPA violated the APA by arbitrarily or capriciously approving WVDEP’s TMDL Lists for certain waterbodies, which included no ionic toxicity TMDLs despite those waterbodies’ state of ionic impairment. Having considered the briefing, the administrative record, and the arguments raised at oral argument, the Court GRANTS in part and DENIES in part the Environmental Groups’ Motion -2- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 3 of 39 PageID #: 6937 for Summary Judgment, ECF No. 30, and GRANTS in part and DENIES in part EPA’s Cross Motion for Summary Judgment. ECF No. 38. II. Background A. The structure of the Clean Water Act In 1972 Congress passed amendments to the Federal Water Pollution Control Act, which are commonly known as the Clean Water Act (“CWA”). The goal of the CWA is to “restore and maintain the chemical, physical, and biological integrity of the Nation’s waters” and to attain “water quality which provides for the protection and propagation of fish, shellfish, and wildlife.” 33 U.S.C. § 1251(a). Congress expected the CWA to solve the Nation’s water quality crisis by 1985. Id. The CWA provides for two primary mechanisms to achieve this ambitious goal: point source pollution controls embodied in the National Pollution Discharge Elimination System (“NPDES”) and ambient water quality standards implemented through the Total Maximum Daily Load (“TMDL”) program. See §§ 1311, 1313. NPDES permits address point source pollution outfalls through technology-based controls. § 1311. Any discharge from a point source into waters within the jurisdiction of the CWA is unlawful unless the discharge complies with an NPDES permit. Id. NPDES permits, however, do not address ambient water quality of the waters into which permitted discharges are emitted. TMDLs, the subject of this litigation, were designed to address ambient water quality in bodies of water that do not meet water quality standards even after NPDES permits have been issued to point source discharges. Id.; see also San Francisco BayKeeper v. Whitman, 297 F.3d 877, 880 (9th Cir. 2002). In other words, TMDLs place daily limits on the total load from all sources of a pollutant or pollutants discharged into a body of water. The CWA gave states a key role in developing water quality standards (“WQS”), identifying bodies of water that do not meet -3- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 4 of 39 PageID #: 6938 those standards, and developing TMDLs to bring those bodies of water into compliance. See § 1313; 40 C.F.R. § 130.7. EPA then reviews state action to ensure compliance with the CWA. Id. Before a state can begin development of TMDLs, the CWA requires each state to develop a WQS consistent with the CWA’s requirements. §§ 1313(a)(3)(A), (b), (c); see also 40 C.F.R. §§ 130.2(d), 131.4(a). A WQS identifies the “designated uses” for a particular waterbody (e.g., public water supply, support of aquatic life, or recreational uses) and a “water quality criteria” expressed as a numeric limit or narrative condition that must be met for the waterbody to support the identified uses (e.g., iron concentrations necessary for aquatic life). § 1313(c)(2); 40 C.F.R. § 131.3(i). When existing pollution controls in a waterbody are not stringent enough to meet applicable water quality standards and the waterbody therefore cannot support its designated uses, that waterbody must be classified by the state as “impaired.” § 1313(d)(1); 40 C.F.R. § 130.7. States must place all impaired waters on a list commonly referred to as a “303(d) List” for review and approval by EPA. Id. States, through 303(d) Lists, also “establish a priority ranking of impaired waterbodies, taking into account the severity of the pollution and the uses to be made of such waters.” § 1313(d)(1)(A); see also 40 C.F.R. § 130.7(b)(4). States then submit 303(d) Lists to EPA, and EPA must either approve or disapprove of the list within thirty days. § 1313(d)(2); 40 C.F.R. § 130.7(d)(2). The CWA does not require EPA to approve a state’s priority ranking. See § 1313. If EPA disapproves a state's 303(d) List, EPA must establish a list of waterbodies that should have been included in the state’s 303(d) List within thirty days of the disapproval. Id. Bodies of water listed on a state’s 303(d) List are, as demonstrated by their presence on the list, not meeting the applicable WQS. § 1313(d)(1)(A). Where a body of water is not meeting its WQS, the CWA requires states to develop a TMDL for that body of water in accordance with the -4- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 5 of 39 PageID #: 6939 waterbody’s priority ranking on the state’s 303(d) List. § 1313(d)(1)(C); 40 C.F.R. § 130.7(c)(1). “A TMDL establishes the maximum daily discharge of pollutants into a waterway” from all sources. Hayes v. Whitman, 264 F.3d 1017, 1021 (10th Cir. 2001) (citing Scott v. City of Hammond, 741 F.2d 992, 996 (7th Cir. 1984)). As a state develops a TMDL for a particular body of water in accord with that waterbody’s priority ranking, EPA regulations permit a state to produce a TMDL for each individual pollutant affecting a particular body of water or to produce a TMDL via a “biomonitoring” approach. 40 C.F.R. § 130.7(c)(1)(i). As with 303(d) Lists, states must submit their TMDLs to EPA, and EPA must approve or disapprove of the TMDLs within thirty days. § 1313(d)(2). If EPA disapproves of a TMDL, EPA must develop, submit for public comment, and finalize a TMDL within thirty days. Id. B. West Virginia’s statutory structure and history of biologic impairment TMDLs West Virginia’s water quality standards include two narrative water quality criteria, which are designed to protect uses of West Virginia’s streams related to aquatic life. Those criteria provide: 3.2. No sewage, industrial wastes or other wastes present in any of the waters of the state shall cause therein or materially contribute to any of the following conditions thereof: ... 3.2.e. Materials in concentrations which are harmful, hazardous or toxic to man, animal or aquatic life; ... 3.2.i. Any other condition, including radiological exposure, which adversely alters the integrity of the waters of the State including wetlands; no significant adverse impact to the chemical, physical, hydrologic, or biological components of aquatic ecosystems shall be allowed. W. VA. CODE R. §§ 47-2-3.2.e–3.2.i. From 2002 through 2010, West Virginia used the West Virginia Stream Condition Index (“WVSCI”) as its methodology for assessing compliance with the narrative criteria that protect aquatic life. EPA Enclosure 1 Review of W. Va.’s 2012 -5- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 6 of 39 PageID #: 6940 Section 303(d), J.A. 2597. Beginning in 1998 West Virginia used the health of the macroinvertebrate community to assess compliance with narrative water quality standards. Id. By 2002 West Virginia, in conjunction with Tetra Tech, Inc. and EPA, developed the WVSCI. Id.; The West Virginia GLIMPSS: Genus Level Index of Most Probable Stream Status, J.A. 2621 [hereinafter GLIMPSS]. “The WVSCI summarizes family [taxonomic] level identifications of benthic macroinvertebrate assemblages to bioassess the condition of wadeable streams.” GLIMPSS, J.A. 2621. WVSCI was developed using data collected by EPA and WVDEP from riffle habitats in wadeable streams in West Virginia. EPA Enclosure 1 Review of W. Va.’s 2012 Section 303(d), J.A. 2597. WVSCI is scaled from 0 (worst) to 100 (best). Id. The score is calculated through a combined multimetric index that consists of six benthic community metrics. Id. A threshold score of 68.0 was considered by WVDEP to represent a body of water fully supportive of narrative water quality standards. Id. WVDEP standardized a score of 68.0 to represent the fifth percentile of reference sites, meaning ninety-five percent of reference streams 1 had a higher score. Id. In WVDEP’s 2002 303(d) List through its 2010 list WVDEP incorporated a “gray zone” of 60.6 to 68.0. Id.; W. Va. 2002 Section 303(d) List, J.A. 1242. Bodies of water that scored in this range were not considered impaired. Id. WVDEP claimed that a score in the gray zone did not statistically support impairment; it believed that a score below 60.6 was a sign of statistically significant impairment. Id. EPA did not agree, and in comments to WVDEP on its 2008 and 2010 303(d) Lists, EPA explained that WVDEP’s gray zone was not statistically supportable. EPA Enclosure 1 Review of W. Va.’s 2012 Section 303(d), J.A. 2597–98 n. 3. Nonetheless, EPA approved 303(d) 1 “Reference conditions represent the characteristics of stream reaches that are least disturbed by human activities and are used to define benchmarks for chemical, biological, and habitat conditions for a region.” GLIMPSS, J.A. 2631. -6- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 7 of 39 PageID #: 6941 Lists that employed the grey zone. Id. In general, WVSCI was an approved and accepted methodology to determine biological impairment. See Id. at. 2597–98 (EPA used WVSCI without the gray zone to list streams on WVDEP’s 2012 303(d) List). Methodologies like WVSCI detect impairment, but “they do not identify the cause or causes of the impairment.” TMDL Guidance, Suppl. J.A. 4592. Accordingly, in order to identify causes of impairment—a necessary preliminary step for developing a TMDL—EPA has developed a “stressor identification process.” Id. at 4593. In the development of TMDLs, West Virginia routinely performs the stressor identification process. See, e.g., Dunkard Creek Watershed TMDL, J.A. 756. West Virginia considers ionic toxicity one stressor among many that cause biological impairment, and in numerous waterbodies the State has found ionic toxicity a significant stressor, giving rise to biological impairment. See, e.g., id. at 757–59. Since 2006, West Virginia has determined that ionic toxicity is the stressor causing biological impairment in at least 179 streams. W. Va. 2008 303(d) List, J.A. 1864, 1886, 1890, & 1897 (identifying ionic toxicity as the cause of biological impairment for four streams in the Upper Kanawha Watershed, seven streams in the Coal River Watershed, and six streams in the Gauley River Watershed); Upper Ohio South TMDL, J.A. 1097 (identifying ionic toxicity as the cause of biological impairment for nine streams in the Upper Ohio South Watershed); Dunkard Creek TMDL, J.A. 759 (identifying ionic toxicity as the cause of biological impairment for four streams in the Dunkard Creek Watershed); Lower Kanawha TMDL, J.A. 559. Plaintiffs also identify 396 other streams on West Virginia’s 2012 303(d) List which were listed as biologically impaired but for which a stressor identification has not been performed. Compl. ¶ 65, ECF No. 78. -7- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 8 of 39 PageID #: 6942 Although WVDEP has identified ionic toxicity as a cause of biological impairment in many streams throughout the state since at least 2006, WVDEP has not issued any TMDLs for ionic toxicity. See 2008 W. Va. Integrated Water Quality Monitoring and Assessment Report, J.A. 1864 (explaining that Stillhouse Branch was determined to be biologically impaired due to ionic stress as early as 2003 and placed on the 303(d) List in 2006 but TMDL development was deferred). WVDEP has issued TMDLs for bodies of water deemed to be biologically impaired from causes other than ionic toxicity, see, e.g., Upper Ohio South TMDL, J.A. 1097, but stopped issuing TMDLs addressing biologic impairment regardless of the cause in 2012—the decision that sparked this litigation, Letter from Randy C. Huffman, Cabinet Sec’y, WVDEP, to Jon M. Capacasa, Dir., Water Prot. Div., EPA Region III (Apr. 6, 2012), J.A. 3298. In the years leading up to that decision, it was WVDEP’s position that it had insufficient information “regarding the causative pollutants and their associated impairment thresholds for biological TMDL development for ionic toxicity stressed streams.” Upper Ohio South Watershed: TMDL Report Draft, Mar. 2009, J.A. 1024. Quizzically, it was also WVDEP’s position that “[a] strong presence of sulfates and other dissolved solids exists . . . in all streams where ionic toxicity has been determined to be a significant biological stressor. Id.; Dunkard Creek Watershed: TMDL Final Report, Sept. 11, 2009, J.A. 759. As early as 2008 WVDEP designated “mining” as the source of biological impairment for streams that had undergone a stressor identification process that identified ionic toxicity as a significant stressor. 2008 W. Va. Integrated Water Quality Monitoring and Assessment Report, J.A. 1864. In the same set of findings WVDEP explained “water quality data indicates elevated conductivity and sulfates contributed by mining discharges,” further supporting the correlation between sulfates and ionic toxicity. Id. TMDL completion dates for waters determined to be impaired by “mining” (ionic toxicity) -8- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 9 of 39 PageID #: 6943 ranged broadly from 2006 in WVDEP’s 2006 303(d) List, to 2016 in its 2008 List, to 2013 in its 2010 List, to “TBD” in its 2012 List for the same stream. See 2008 W. Va. Integrated Water Quality Monitoring and Assessment Report, J.A. 1864 (determining in 2003 to 2004 that ionic toxicity is a stressor in Stillhouse Branch and first listing it on a 303(d) List in 2006 but delaying TMDL development for ionic toxicity due to a lack of data on the causative pollutants of ionic toxicity); see, e.g., W. Va. 2006 Section 303(d) List, J.A. 1676 (Stillhouse Branch. Listed as “CNABiological”); W. Va. 2008 Section 303(d) List, J.A. 1897; W. Va. 2010 303(d) List, J.A. 2135; W. Va. 2012 Section 303(d) List, J.A. 2378.2 All bodies of water biologically impaired by ionic toxicity were, and still are, retained on WVDEP’s 303(d) Lists. Upper Ohio South Watershed: TMDL Report Draft, Mar. 2009, J.A. 1024; Dunkard Creek Watershed: TMDL Final Report, Sept. 11, 2009, J.A. 759. In the fall of 2010, EPA and WVDEP began a project to develop a pilot TMDL for ionic toxicity in four streams in the Upper Kanawha Watershed. W. Va. Ionic Stress Background Info., Suppl. J.A. 5230–31. The project called for completing pilot ionic toxicity TMDLs for those streams by August 2012. Id. In 2011, in the midst of the pilot project, EPA published “A FieldBased Aquatic Life Benchmark of Conductivity in Central Appalachian Streams,” (“Benchmark”) a peer-reviewed and highly respected study of the causes and effects of ionic toxicity in Appalachian streams. See Ohio Valley Envtl. Coal., Inc. v. Fola Coal Co., LLC, 82 F. Supp. 3d 673, 679–80 (S.D. W.Va. 2015), aff’d, ___ F.3d ___, No. 16-1024, 2017 WL 35726 (4th Cir. Jan. 2 Although Stillhouse Branch is listed as impaired by “CNA-Biological” in the earlier 303(d) Lists, WVDEP’s statements in its 2008 Integrated Report indicate that ionic toxicity was identified in the stream as early as 2003. See W. Va. 2006 Section 303(d) List, J.A. 1676; 2008 W. Va. Integrated Water Quality Monitoring and Assessment Report, J.A. 1864. It is not clear why WVDEP designated Stillhouse Branch using the broader designation for biological impairment, “CNA-Biological,” without also identifying the underlying impairment as ionic toxicity. -9- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 10 of 39 PageID #: 6944 4, 2017) [hereinafter OVEC]; A Field-Based Aquatic Life Benchmark for Conductivity in Central Appalachian Streams (Final Report), EPA/600/R-10/023F (2011), J.A. 3301 [hereinafter The Benchmark]. The Benchmark, among other things, identified the constituent salts (Ca, Mg, SO4, HCO3) that contribute ions resulting in ionic toxicity as well as their source—surface mining, valley fills, slurry impoundments, coal refuse fills, and deep mines. OVEC, F. Supp. 3d at 687. The Benchmark also found that the constituent salts and their sources were nearly uniform across “the examined region,” which includes Ecoregions 69 and 70 (most of West Virginia, excluding the Eastern Panhandle). Id.; WVDEP Aquatic Life Use Assessment Methodology for Wadeable Streams and Rivers, May 2016, J.A. 3971 (depicting a map designating West Virginia’s Level III Ecoregions). “It is precisely because water in the examined regions is so consistently and uniformly dominated by a distinct mixture of ionic pollutants that setting a benchmark for the Appalachian Region is possible.” Id. (citing The Benchmark, 4, J.A. 3320). The Benchmark ultimately concluded that when conductivity in a stream reaches 300 micro Siemens per centimeter (μS/cm), a measure of conductivity that demonstrates a presence of dissolved ions, the stream is biologically impaired. Id. (finding WVSCI score of 64 corresponds to 300 μS/cm. A score of 68 is the threshold below which a body of water is considered impaired.). In February 2012, after EPA published the Benchmark, WVDEP and EPA considered a TMDL endpoint for conductivity of 720 μS/cm, well above the threshold in the Benchmark. W. Va. Ionic Stress TMDL Dev., Suppl. J.A. 5224. Although WVDEP did not see the pilot project through to the end, Tetra Tech, the same company that helped develop WVSCI, prepared a draft report synthesizing what data the pilot project had collected up to that point. See WV Ion TMDL Endpoint Analysis, Draft Technical Document, Dec. 2012, J.A. 3079. The report concluded, as did -10- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 11 of 39 PageID #: 6945 the Benchmark, that total dissolved solids and certain mixes and concentrations of particular ions (Ca, Mg, SO4, HCO3) directly correlate with WVSCI scores—the higher the concentration, the lower the WVSCI score. Id. at J.A. 3099. The authors of the report also explained that due to the geologic and macroinvertebrate community similarities of ecoregions 69 and 70, the data collected from each was used to produce general recommendations to reduce ionic toxicity in all streams in those regions. Id., J.A. 3083–87. In April 2012, WVDEP ended its participation in the pilot program before a pilot TMDL could be fully developed. On April 6, 2012, WVDEP Secretary Randy Huffman sent a letter to EPA Region III terminating West Virginia’s participation in the pilot project, claiming passage of Senate Bill 562 (“SB 562”) precluded WVDEP from continuing to use WVSCI to determine biologic impairment and therefore WVDEP could not continue to develop a TMDL using WVSCI. Letter from Randy C. Huffman, Cabinet Sec’y, WVDEP, to Jon M. Capacasa, Dir., Water Prot. Div., EPA Region III, J.A. 3298–99. SB 562 requires WVDEP to develop a new assessment methodology for measuring compliance with West Virginia’s narrative water quality standards but the new methodology must be as protective or more protective than WVSCI. SB 562 states: (f) The secretary shall propose rules measuring compliance with the biologic component of West Virginia's narrative water quality standard [which] requires evaluation of the holistic health of the aquatic ecosystem and a determination that the stream: (i) Supports a balanced aquatic community that is diverse in species composition; (ii) contains appropriate trophic levels of fish, in streams that have flows sufficient to support fish populations; and (iii) the aquatic community is composed of benthic invertebrate assemblages sufficient to perform the biological functions necessary to support fish communities within the assessed reach, or, if the assessed reach has insufficient flows to support a fish community, in those downstream reaches where fish are present. The secretary shall propose rules for legislative approval in accordance with the provisions of article three, chapter twenty-nine-a of this code that -11- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 12 of 39 PageID #: 6946 implement the provisions of this subsection. Rules promulgated pursuant to this subsection may not establish measurements for biologic components of West Virginia's narrative water quality standards that would establish standards less protective than requirements that exist at the time of enactment of the amendments to this subsection by the Legislature during the 2012 regular session. 2012 W. Va. Acts 562 (codified at W. VA. CODE § 22-11-7b). WVDEP interprets SB 562 to preclude the use of the WVSCI methodology for finding biological impairment from the date SB 562 was passed. W. Va. Draft 2012 Integrated Water Quality Monitoring and Assessment Report, J.A. 2332. As a result, it is also WVDEP’s position that “the effect of [SB 562] necessarily requires WVDEP to postpone the TMDL development process” for all biologically impaired streams until a new methodology is developed and approved by the Legislature. Letter from Randy C. Huffman, Cabinet Sec’y, WVDEP, to Jon M. Capacasa, Dir., Water Prot. Div., EPA Region III, J.A. 3298–99. WVDEP applied its interpretation to both the 179 waterbodies determined to be biologically impaired from ionic toxicity and the 396 waterbodies biologically impaired but for which no stressor identification has been conducted. In response, EPA requested that WVDEP clarify its interpretation of SB 562 and its effect on West Virginia’s WQS and other CWA mandated programs. Letter From Shawn Garvin, Regional Adm’r, EPA Region III, to Randy C. Huffman, Cabinet Sec’y, WVDEP, (Nov. 6, 2012), J.A. 2740–41. The record before the Court does not contain a formal response. EPA, conversely, does not interpret SB 562 to preclude WVDEP from developing TMDLs for biologically impaired streams. Draft TMDL for Selected Streams in the Monongahela River Watershed, W. Va. EPA Comments—Oct. 24, 2013, J.A. 188. (“SB 562 does not appear to expressly preclude TMDL development.”). It was also EPA’s stated position as late as 2014 that -12- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 13 of 39 PageID #: 6947 the Genus Level Index of Most Probable Stream Status (“GLIMPSS”)3 is a valid methodology that could replace WVSCI and comply with SB 562. EPA’s Comments on W. Va’s 2014 Draft Section 303(d) List, Jul. 11, 2014, J.A. 2752. EPA has urged WVDEP to adopt GLIMPSS since 2010.4 Id. at 2751. To date, WVDEP has not adopted a new methodology to replace WVSCI. Due to WVDEP’s interpretation of SB 562, WVDEP stopped adding waters newly determined to be biologically impaired to its 303(d) List because it believed that it could no longer use WVSCI to determine biological impairment. EPA Enclosure 1, Review of W. Va. 2012 Section 303(d), J.A. 2603. WVDEP also did not set dates for completing TMDLs for any body of water already found to be biologically impaired. Rather, DEP noted that in light of SB 562 those TMDLs would be completed “TBD – To be determined. TMDLs will be developed as soon as practicable after the effective date of rules enacted pursuant to [SB 562].” W. Va. 2012 Draft 303(d) List, J.A. 2364, 2368. WVDEP submitted its 2012 303(d) List to EPA without adding new biologically impaired waterbodies. EPA Enclosure 2, EPA’s List Dev. Process, J.A. 2603. EPA partially disapproved that list because the State failed to “evaluate all existing and readily available water quality-related data and information, specifically, information related to whether certain waters are achieving 3 WVDEP and EPA developed GLIMPSS in a joint effort as a next generation index “designed to provide higher resolution than . . . family-level WVSCI.” EPA’s Comments on W. Va’s 2014 Draft Section 303(d) List, Jul. 11, 2014, J.A. 2751. EPA described GLIMPSS as using “41 different biological metrics . . . tested across seasonal and geographic strata, primarily to refine expectation criteria for aquatic life use attainment in WV.” Id. GLIMPSS provides better diagnostic capabilities than WVSCI and a more accurate index “that more directly measures specific aquatic life attainment in West Virginia streams as it accounts for natural variability driven by geographic location, seasonality, and waterbody size.” Id. GLIMPSS was developed with nearly 400 reference sites, whereas WVSCI was developed with only 107. Id. Finally, EPA believes GLIMPSS “is also better suited than WVSCI to detect biological changes due to climate change.” Id. 4 Kentucky, Ohio, Pennsylvania, Maryland, and Tennessee use a form of genus-level assessment to determine biologic impairment. Id. -13- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 14 of 39 PageID #: 6948 West Virginia’s narrative water quality criteria as applied to the aquatic life uses.” Id. In its letter disapproving the 2012 303(d) List, EPA stated, “[r]ecognizing WVDEP’s position that it is unable to carry out the requirement set forth in 40 CFR 130.7(b)(5), EPA has an obligation to take action to ensure that the federal requirement is satisfied.” Letter from Shawn M. Garvin, Regional Adm’r, EPA Region III, to Randy C. Huffman, Cabinet Sec’y, WVDEP (Mar. 25, 2013), J.A. 2584. EPA fulfilled its obligations by developing a 303(d) List of waters in West Virginia that are not achieving the narrative standards that protect the aquatic life use, i.e., by adding biologically impaired waterbodies to West Virginia’s 2012 303(d) List using WVSCI as the methodology to determine biologic impairment. Id. In response, WVDEP’s director of the Division of Water and Waste Management, Scott Mandirola, sent a letter to EPA’s Bill Richardson in Region III’s Water Protection Division. Mandirola explained: DEP is neither unable nor unwilling to carry out its responsibilities under the CWA, but it is both unwilling and unable to violate its obligations under [SB 562] or the West Virginia Administrative Procedures Act. [T]he West Virginia Legislature has made the policy decision that the biologic health of a stream . . . must be measured using more factors than simply a [WVSCI] score taken at one point in a watercourse. Letter from Scott G. Mandirola, Dir, Div. of Water and Waste Mgmt., WVDEP, to Bill Richardson, Water Prot. Div. EPA Region III (May 8, 2013), J.A. 2707. Despite WVDEP’s interpretation of SB 562, in its draft 2014 303(d) List WVDEP stated: “The DEP is proposing to retain most of the biological impairments identified in the Final West Virginia 2012 Section 303(d) List and to add new listings using the WVSCI.” W. Va. Draft 2014 Section 303(d) List, Suppl. J.A. 4955 (emphasis added). Even though WVDEP resumed use of WVSCI for identifying biologically impaired streams, WVDEP’s 2014 Draft 303(d) List again set -14- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 15 of 39 PageID #: 6949 no dates for completing TMDLs for biologically impaired streams, only noting that the date for TMDL development was “TBD.” Id. at Suppl. J.A. 4958, 4975. WVDEP also explained that it would not be developing biologic impairment TMDLs because state law precluded it from using WVSCI. Id. at 4958. Both EPA and Plaintiffs submitted comments to WVDEP recommending that WVDEP establish a schedule for developing TMDLs for biologically impaired streams. EPA’s Comments on W. Va. 2014 Draft Section 303(d) List, July 11, 2014, J.A. 2753, 2785. In its 2014 303(d) List submitted to EPA for final approval on April 13, 2015, WVDEP deferred completion dates for TMDLs for waterbodies that had previously been listed as “TBD” to dates ranging from 2020 to 2025. Letter from Scott Mandirola, Dir. Div. of Water and Waste Mgmt., WVDEP to Evelyn MacKnight, Assoc. Dir., Office of Standards, Assessments and TMDLs, EPA Region III, (Apr. 13, 2015), J.A. 3046; see also, e.g., W. Va. 2014 Section 303(d) List, J.A. 2805–06. Although WVDEP set dates for biological impairment TMDLs, EPA did not take any additional action on WVDEP’s schedule. From 2012 to the present the Court is unaware of any completed TMDLs for biologically impaired waterbodies even though EPA has approved numerous TMDLs for other pollutants with numeric limits, such as metals, pH, and chlorides. See, e.g., Letter from Jon M. Capacasa, Dir. Water Prot. Div., EPA Region III, to Scott Mandirola, Dir. Div. of Water and Waste Management, WVDEP, J.A. 38–39. WVDEP has also been unable to promulgate a new methodology to determine biologic impairment. In the initial aftermath of SB 562, WVDEP expected to develop a new methodology in a year’s time. In the WVDEP Director’s April 6, 2012 letter to the EPA Region III Administrator, the Director expected that a new methodology would be submitted to the Legislature for approval within a year.5 Letter from Randy C. Huffman, Cabinet Sec’y, WVDEP, 5 By way of background, in order for WVDEP’s new methodology to become a final -15- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 16 of 39 PageID #: 6950 to Jon M. Capacasa, Dir., Water Prot. Div., EPA Region III, J.A. 3298–99. WVDEP was not able to meet this goal. In 2014, WVDEP predicted that it would be able to submit its new methodology to the 2015 legislature, but then later that year determined that it would submit the methodology to the 2016 legislature. W. Va. Draft 2014 Section 303(d) List, Suppl. J.A. 4954; 2014 W. Va. Integrated Water Quality Monitoring and Assessment Report, J.A. 2767. To date, WVDEP has not finalized a new methodology and now predicts that “it will still be some time before [WVDEP has] sufficient data to compliment [sic] any benthic macroinvertebrate index.”6 EPA’s Surreply 4 n. 2, ECF No. 75 (quoting Letter from Scott Mandirola, Dir., Div. of Water and Waste Mgmt., WVDEP, to Bill Richardson, Water Prot. Div, EPA Region III (Jul. 5, 2016), https://www.epa.gov/ sites/production/files/2016-07/documents/wvdep_comments_re_epa_overlist_july_5_2016.pdf). Pursuant to West Virginia law, a new methodology, even if it were now finalized by WVDEP, would not be considered until the 2018 legislative session. See W. VA. CODE §§ 29A-311, 29A-3-12. At oral argument counsel for the Environmental Groups requested the Court take judicial notice of the list of regulations submitted to the West Virginia Legislature to be considered binding regulation, it must first submit the proposed regulation to the Legislative Rule-Making Review Committee. W. VA. CODE § 29A-3-11. The Committee may then make a recommendation to the West Virginia Legislature to approve the regulation with or without changes or amendments, or the Committee may simply reject the rule. Id. If the Committee submits the regulation to the Legislature, the Legislature may approve, disapprove, or amend the regulation. W. VA. CODE § 29A-3-12. 6 The letter from which the quote is taken is not part of the administrative record submitted to the Court. In its Surreply, EPA requests that the Court take judicial notice of the letter, which is publically available on EPA’s website. ECF No. 75. “A court may take judicial notice of information publically announced on a party’s web site, so long as the web site’s authenticity is not in dispute and it is capable of accurate and ready determination.” Jeandron v. Bd. of Regents of Univ. Sys. of Md., 510 Fed.Appx. 223, 227 (4th Cir. 2013) (citing Fed. R. Evid. 201(b); O’Toole v. Northrop Grumman Corp., 499 F.3d 1218, 1225 (10th Cir. 2007)). Plaintiffs do not dispute the authenticity of the of the website from which the letter was taken, and as a government record, it is capable of accurate and ready determination. The Court therefore takes judicial notice of the letter. -16- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 17 of 39 PageID #: 6951 in the 2017 legislative session. The list is maintained on the Secretary of State’s website.7 Upon inspection of the list, WVDEP has not submitted a proposed methodology to be considered by the legislature for the 2017 session. Pl’s. Ex. 1, ECF No. 84-1, http://www.sos.wv.gov/administrativelaw/modified/Pages/default.aspx. In order for the Legislative Rule-Making Review Committee and ultimately the West Virginia Legislature to consider a regulation in the 2017 legislative session, the regulation must have been submitted to the Secretary of State and the Committee by July 27, 2016. Summary of Regular Rule Making Steps, WEST VIRGINIA SECRETARY OF STATE, http://www.sos.wv.gov/administrative-law/rulemaking/Pages/stepsummary.aspx (last visited Nov. 16, 2016). The Court is unaware of any schedule, self-imposed or otherwise, for the completion of WVDEP’s new methodology. Plaintiffs filed suit pursuant to the citizen suit provision of the CWA, § 1365(a)(2), asserting that EPA had a nondiscretionary duty to promulgate TMDLs for biologically impaired bodies of water when WVDEP stated that it would not issue those TMDLs. Plaintiffs also brought suit pursuant to 5 U.S.C. § 706(2) of the Administrative Procedure Act, claiming that EPA’s approval of TMDLs for six watersheds was arbitrary and capricious, an abuse of discretion, or contrary to law because the TMDLs for each watershed did not include TMDLs for streams impaired by ionic toxicity. III. Legal Standard To obtain summary judgment, the moving party must show that there is no genuine issue as to any material fact and that the moving party is entitled to judgment as a matter of law. Fed. R. Civ. P. 56(a). In considering a motion for summary judgment, the Court will not “weigh the 7 In accord with the standard explained in footnote six, the Court takes judicial notice of the list of proposed regulations maintained on the West Virginia Secretary of State’s website. See id. -17- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 18 of 39 PageID #: 6952 evidence and determine the truth of the matter.” Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 249 (1986). Here the parties agreed that the facts consist solely of the administrative record and the parties do not dispute any relevant facts in that record. Summary judgment based on the facts presented in the administrative record is therefore appropriate. See Occidental Eng’g Co. v. INS, 753 F.2d 766, 769 (9th Cir. 1985) (“[T]he function of the district court is to determine whether . . . as a matter of law the evidence in the administrative record permitted the agency to make the decision it did.”). IV. Discussion A. The CWA claims and background of the Constructive Submission Doctrine When Congress passed the CWA, it set a strict deadline for development of 303(d) Lists and the attendant TMDLs. Under the previous water pollution control regime, states declined to follow federal directives to develop and submit water quality standards. Kingman Park Civic Assoc v. EPA, 84 F.Supp.2d 1, 7 (D.D.C. 1999). A central motivation for Congress to enact the CWA was to address this problem. Id. Congress lamented in 1971 that ‘[m]ore than 4 years after the deadline for submission of standards, only a little more than half of the states have fully approved standards.’” Id. (quoting S. REP. NO. 92-414, at 4 (1971), as reprinted in 1972 U.S.C.C.A.N. 3668, 3671. Consequently, the CWA commanded states to submit all TMDLs not more than one hundred eighty days after EPA published a list of pollutants to be addressed by TMDLs and then “from time to time” thereafter. 8 33 U.S.C. § 1313(d)(2). Congress gave EPA one year from 8 Worth mentioning is that before the first round of TMDLs could be implemented states would also have to develop WQS and secure approval from EPA for them. Had Congress’ original schedule been followed, both WQS and TMDLs would have been in place eighteen months after Congress passed the CWA. § 1313(a). -18- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 19 of 39 PageID #: 6953 October 18, 1972 to publish this list, making states’ 303(d) Lists and TMDLs due on April 16, 1974. See id. Congress, aware of the herculean task of collecting data on all pollutants affecting all waterbodies on a state’s 303(d) List in such a short time, decided in favor of speed of implementation rather than certainty, declaring that TMDLs should use “a margin of safety which takes into account any lack of knowledge concerning the relationship between effluent limitations and water quality.” § 1313 (d)(1)(c). Not to be rushed, however, EPA did not identify pollutants for TMDLs until December 28, 1978—more than six years after the CWA deadline. 43 Fed. Reg. 60,662 (Dec. 28, 1978). States were therefore not required to submit their TMDLs until June 26, 1979 and thereafter “from time to time.” See § 1313(d)(2); NRDC v. Fox, 909 F.Supp. 153, 157 (S.D.N.Y. 1995) (citing Scott, 741 F.2d at 996 n. 10). States and EPA largely ignored this deadline until the 1990’s and early 2000’s. See, e.g., San Francisco BayKeeper, 297 F.3d at 883 (California did not submit any TMDLs until 1991); Friends of the Wild Swan, Inc. v. EPA, 130 F. Supp. 2d 1184, 1189 (D. Mont. 1999) (from 1979 to 1992 Montana did not submit any 303(d) Lists or TMDLs); Kingman Park Civic Assoc., 84 F. Supp. 2d at 2 (The District of Columbia did not submit TMDLs for eighteen years); Am. Canoe Assoc., Inc. v. EPA, 30 F. Supp. 2d 908, 919 (E.D. Va. 1998) (Virginia did not submit any TMDLs for two decades); Sierra Club v. Hankinson, 939 F. Supp. 865, 868 (N.D. Ga. 1996) (Georgia did not submit a 303(d) List until 1992 and submitted its first TMDL in 1994); Idaho Sportsmen’s Coal. V. Browner, 951 F. Supp. 962, 964 (D. Idaho 1996) (Idaho did not submit a 303(d) List until 1989). Alaska Ctr. for the Env’t v. Reilly, 762 F. Supp. 1422, 1425 (W.D. Wash. 1991) (Alaska had not submitted any TMDLs by 1991). As an example of the scope of the noncompliance at the time the Alaska Center Court found “[i]n 1989 EPA Region IV approved 163 TMDLs, Region V -19- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 20 of 39 PageID #: 6954 approved 74, Region I approved 50, Region VIII approved 16, Region X approved 11. Regions II, III, and VII however, approved no TMDLs.” Alaska Ctr., 762 F. Supp. at 1425 (Alaska, Washington, and Idaho had not submitted any TMDLs by 1991).9 Congress apparently did not anticipate a total failure of compliance by EPA and the states, and accordingly made no explicit provision in the CWA to address when states fail to submit a TMDL. Congress, perhaps naively, expected the states to comply with the CWA and submit TMDLs to EPA for review. As such, it was not clear from the language of the CWA what recourse EPA had if states declined to submit TMDLs. Read literally, EPA’s duty to act was only triggered upon a submission from a state. See § 1313 (“Each State shall submit to the Administrator . . . , for his approval the waters identified and the loads established . . . .”). The majority of courts faced with this problem have adopted the “constructive submission” doctrine. See San Francisco BayKeeper, 297 F.3d at 882; Hayes, 264 F.3d at 1023; Kingman Park Civic Assoc, 84 F. Supp. 2d at 5. The doctrine, established by Scott v. City of Hammond, concludes that EPA’s duty to approve or disapprove a TMDL is triggered “if a state fails over a long period of time to submit proposed TMDLs.” 741 F.2d at 996 (finding the doctrine applicable after three years of inaction). The failure of a state to submit TMDLs to EPA is a “constructive submission” to EPA of no TMDLs. Id. The Seventh Circuit reasoned that although the CWA does not explicitly compel EPA to act when a state fails to act, it did not “believe that Congress intended that the 9 To put the number of TMDLs approved by the various regions in 1989 in perspective, Alaska alone has 3,000,000 lakes and thousands of rivers and streams that needed to be tested and incorporated into the state’s water quality management program. Alaska Ctr. for Env’t v. Browner, 20 F.3d 981, 985 (9th Cir. 1994). Out of those streams and lakes, even Alaska’s threadbare water quality program identified several hundred impaired bodies of water in need of TMDLs. Alaska Ctr., 762 F. Supp. at 1426 (emphasis added). No EPA region approved anywhere near several hundred TMDLs at the time. Id. Or, in Oklahoma’s case, the state had 1,500 outstanding TMDLs and planned to produce about 100 per year—as many or more TMDLs than the numbers approved by entire EPA Regions in 1989. See Hayes, 264 F.3d at 1024. -20- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 21 of 39 PageID #: 6955 states by inaction could prevent implementation of TMDLs.” Id. at 997. The Seventh Circuit noted, however, that the appeal came to it from a motion to dismiss and, with only the complaint before it, it could not determine whether the states (Illinois and Indiana) “are, or will soon be, in the process of submitting TMDL proposals.” Id. at 997 n. 11. If the states were in the process of submitting the missing TMDLs, then the constructive submission doctrine might not apply. Id. The Seventh Circuit’s condition has served as a guidepost for courts in their decisions addressing constructive submissions. Where plaintiffs have brought claims alleging that a state has failed to produce any TMDLs for any waters on its 303(d) List—a challenge to a statewide TMDL program—Courts have declined to find a constructive submission if the state has produced at least some TMDLs and has a credible plan to fulfill its CWA obligations. See Hayes, 264 F.3d at 1024. “If a state has submitted or soon plans to submit TMDLs for its impaired waterbodies, the constructive-submission analysis would be factually inapplicable.” Id. (finding a programmatic challenge to Oklahoma’s TMDL process did not rise to a constructive submission because Oklahoma had submitted “a number” of TMDLs and was making progress toward completing 1,500 in twelve years); see also San Francisco BayKeeper, 297 F.3d at 883 (finding a programmatic challenge to California’s TMDL program did not prove a constructive submission where California had submitted eighteen TMDLs and had a schedule to complete the rest). On the other hand, where plaintiffs brought a programmatic challenge and a state has not produced any TMDLs and is not in the process of producing them, courts have found a constructive submission. See Alaska Ctr., 762 F. Supp. at 1429 (holding Alaska had not submitted any TMDLs for any body of water in the state and had no plan to comply with the CWA, and therefore a constructive submission had occurred). -21- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 22 of 39 PageID #: 6956 The doctrine is equally applicable to a single missing TMDL as it is to a programmatic failure. “Naturally a state that has publically indicated . . . that it will not produce a specific TMDL has violated its statutory obligations with regard to that TMDL, no matter how robust its program otherwise is.” Sierra Club v. McLerran, No. 11-cv-1759, 2015 WL 1188522, at *7 (W.D. Wash Mar. 16, 2015) see also Scott, 741 F.2d at 996–97 (finding a constructive submission doctrine applicable where states did not submit TMDLs for a single body of water—Lake Michigan). Thus, a constructive submission occurs where a “state’s actions clearly and unambiguously express a decision to submit no TMDL for a particular impaired waterbody.” Hayes, 264 F.3d at 1024 (emphasis added). B. Application of the Constructive Submission Doctrine to this case By its own admission, WVDEP has chosen to pursue a new methodology to calculate biological impairment and has stopped submitting TMDLs for biologic impairment, claiming a state law does not permit it to comply with federal directives. It expressly informed EPA in 2012 that it would not be issuing any TMDLs for biologic impairment until it developed a new methodology to determine biologic impairment and secured legislative approval of that methodology. At the same time WVDEP withdrew from the very pilot program that, seen to completion, would have resulted in a TMDL for ionic toxicity—an impairment that has avoided any regulation thus far. For over four years, WVDEP has made successive promises to deliver the new methodology to the West Virginia Legislature and for four years has failed to even develop a methodology, to say nothing of actually submitting it to the Legislature. WVDEP has not produced a schedule for when the methodology might be finished. In the meantime, it has stopped complying with the CWA and denied it has the data to fulfill its duties, all the while possessing a significant amount of information and two accepted and -22- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 23 of 39 PageID #: 6957 valid methodologies (WVSCI and GLIMPSS) to determine biologic impairment. Moreover, in an apparent attempt to avoid EPA rejection of its 303(d) List, WVDEP used WVSCI to determine biologic impairment of new streams while at the same time arbitrarily denying that it has the authority to use WVSCI to determine biologic impairment for the development of TMDLs. All of WVDEP’s plodding and EPA’s appeasement have resulted in an abjuration of WVDEP’s and EPA’s duties committed to each by the CWA. WVDEP has publically stated that it will not develop TMDLs for biologic impairment and has continued to move the goalposts for when it will begin developing them once again. Consequently, WVDEP has constructively submitted no TMDLs for biologic impairment to EPA, triggering EPA’s duty to approve or disapprove of the submission. i. WVDEP’s biological impairment TMDL schedule EPA attempts to disguise WVDEP’s decision by arguing that WVDEP understands it has a duty to submit TMDLs for all waters listed on its 303(d) List, but that here WVDEP has only postponed or reprioritized TMDLs for biologic impairment—not refused to act. Thus, EPA contends, WVDEP will soon submit TMDLs and the constructive submission analysis is factually inapplicable. As evidence, EPA points to the terms “pause,” “defer,” and “suspend” used in WVDEP statements about its TMDL program. EPA also relies on WVDEP’s 2014 303(d) List where, after EPA and plaintiffs prodded, WVDEP added dates by which TMDLs would be issued for biologically impaired streams. The dates range from 2020 to 2025 with no explanation for how WVDEP could achieve them. WVDEP explained that after considering resources and available information the dates were “the next practical opportunity” based on WVDEP’s watershed-bywatershed approach to TMDL development. 2014 W. Va. Integrated Water Quality Monitoring and Assessment Report, JA 2785. There is ample evidence to doubt the credibility of these dates. -23- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 24 of 39 PageID #: 6958 West Virginia has a five-year schedule for developing TMDLs on a watershed-bywatershed approach. Draft 2012 Integrated Water Quality Monitoring and Assessment Report, J.A. 2349; see also EPA Memo. Supp. Cross-Mot. Summ. J. 9, ECF No. 73. “WVDEP’s TMDL development program has historically attempted to comprehensively address all streams and all impairments in a particular watershed simultaneously.” EPA Memo. Supp. Cross-Mot. Summ. J., at 9 (quoting Dunkard Creek Watershed: TMDL Report, Draft, J.A. 821). In other words, it takes WVDEP five years to develop TMDLs for any given watershed. Draft 2012 Integrated Water Quality Monitoring and Assessment Report, J.A. 2349. If West Virginia’s 2020 completion date for biologically impaired bodies of water is to be believed, West Virginia, by its own stated process, would have had to start developing biologic impairment TMDLs by 2015. Yet, currently, WVDEP believes it does not have an approved methodology to use to determine biologic impairment and it will be sometime before the legislature approves one. Even if WVDEP finalized a methodology today, the earliest that such a methodology could be used is 2018. WVDEP missed the deadline to submit regulations to the 2017 legislature and now must wait to submit any methodology to the 2018 legislature. Even 2018 assumes that the West Virginia Legislature will approve whatever final methodology WVDEP creates in the 2018 legislative session. There is obviously no guarantee that it will. Moreover, although belied by the scientific data in the record,10 WVDEP claims that it lacks necessary information to develop a TMDL for ionic toxicity (one of the impairments that 10 In section II.B of this Opinion, the Court recounted WVDEP’s estimation in 2006 that dissolved solids and sulfates are highly correlated to ionic toxicity. The section also summarizes EPA’s peer reviewed Benchmark for ionic toxicity in streams in the Appalachian region and Tetra Tech’s data analysis of data collected during WVDEP and EPA’s pilot ionic toxicity TMDL project. Although the data on ionic toxicity appears to be extensive, including determinations of levels of ionic toxicity and the causative ions which result in impairment, EPA credits WVDEP’s claims that it still lacks certain data necessary to create an ionic toxicity TMDL. -24- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 25 of 39 PageID #: 6959 leads to biologic impairment). Assuming the Legislature approves whatever methodology WVDEP ultimately develops, WVDEP will have to then spend time conducting research to address ionic toxicity. And, if EPA and WVDEP are to be believed, creating a TMDL for ionic toxicity is a complex task that cannot be rushed: very likely adding a year or more to the ultimate development of a TMDL for ionic toxicity. Given the statement by WVDEP in its most recent letter, however, that it will still be some time before a methodology is completed, that timeline may be overly optimistic, casting serious doubt on WVDEP’s ability to develop TMDLs by 2025—much less 2020. Naturally, the Court does not believe that WVDEP is obliged to produce a TMDL immediately upon the addition of a body of water to a 303(d) List, but it must at least have a plan to complete it. See Hayes, 264 F.3d at 1024. The distinction that has become apparent in TMDL jurisprudence is between states that have neither produced the delinquent TMDLs nor have any plan to do so, and those that at the very least have a concrete plan to comply with their duties. Compare, e.g., Sierra Club v. Browner, 843 F. Supp. 1304, 1313 (D. Minn. 1993) (finding a constructive submission of all TMDLs in the state did not occur because Minnesota had completed 18 TMDLs) with Alaska Ctr., 762 F. Supp. at 1429 (finding Alaska had not produced any TMDLs and had no plan to comply with its CWA duties). Put another way, courts that have found a constructive submission inapplicable have gone to great lengths to emphasize that at least something related to the development of the challenged TMDLs was being done. See Fox, 93 F. Supp. 2d at 540 (finding constructive submission doctrine inapplicable because New York submitted some of the missing TMDLs and formulated a plan to for finishing the rest of the missing TMDLs); Idaho Sportsman Coal., 951 F. Supp. at 967–68 (finding constructive submission inapplicable because Idaho had submitted three of the missing -25- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 26 of 39 PageID #: 6960 TMDLs and proposed a schedule for completing the rest); Hankinson, 939 F. Supp. at 871 n. 6 (finding although plaintiffs challenged Georgia’s statewide TMDL program, “[t]he Court does not find the constructive submission analysis to be appropriate for this case because Georgia has made some TMDL submission, albeit totally inadequate.” (emphasis in original)). “Even in Scott . . . the Seventh Circuit remanded the case to the district court for a determination whether the state was in the process of submitting any TMDLs . . . .” San Francisco BayKeeper, 297 F.3d at 883. The Ninth Circuit found this distinction dispositive. In San Francisco BayKeeper, plaintiffs brought a challenge claiming that California had not submitted any TMDLs statewide and therefore had made a constructive submission of no TMDLs in the state, requiring EPA to act. San Francisco BayKeeper, 297 F.3d at 883. The district court found, and the Ninth Circuit agreed, that no constructive submission had occurred because California had actually produced eighteen TMDLs and had established a “concrete” schedule for completing all TMDLs for the entire state. See, San Francisco BayKeeper, Inc. v. Browner, 147 F. Supp. 2d 991, 1000 (N.D. Cal 2001), aff’d sub nom. San Francisco BayKeeper v. Whitman, 297 F.3d at 883. In fact, California was in the process of conducting 200 TMDL studies. Id. Accordingly, the Ninth Circuit affirmed the district court’s holding that California had not made a statewide constructive submission of no TMDLs. San Francisco BayKeeper, 297 F.3d at 883; see also Hayes, 264 F.3d at 1023 (“If a state has submitted or soon plans to submit TMDLs for its impaired waterbodies, the constructive submission analysis would be factually inapplicable.”). Conversely, where Alaska failed to develop TMDLs and had no credible plan to complete them, the same scenario the court is faced with here concerning a subset of TMDLs, the district court found that a constructive submission had occurred. See Alaska Ctr., 762 F.Supp. at 1429.12 12 The Court is aware that, unlike this case, Hayes, San Francisco BayKeeper, and Alaska -26- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 27 of 39 PageID #: 6961 The facts presented to the Court in this case bare a much closer resemblance to those addressed by the Alaska Center Court than those presented to the San Francisco BayKeeper Court. WVDEP has not produced biological impairment TMDLs since 2012, stated that it cannot produce them, and has not proposed a credible schedule for producing them. In light of WVDEP’s statements that it will not be producing biological impairment TMDLs and its failure to produce a credible plan to comply with its CWA duties, it has made a constructive submission of no biological impairment TMDLs that triggered EPA’s duty under the CWA to approve or disapprove of the submission. ii. WVDEP’s reasons for not developing biological impairment TMDLs and EPA’s responses Further eroding WVDEP’s credibility, and calling into question EPA’s acquiescence, is WVDEP’s use of WVSCI in its 2014 303(d) List to determine biologic impairment. Since the passage of SB 562, it has been WVDEP’s stated position that the legislation barred WVDEP from using WVSCI, and because it could not use WVSCI to determine biologic impairment, it could not develop TMDLs for biologic impairment until a new methodology was developed. Then, without explanation, WVDEP stated in its 2014 303(d) List that it would use WVSCI to add new biologically impaired streams to that list. In that same list, however, WVDEP concluded that SB 562 precluded it from using WVSCI and therefore it could not develop TMDLs for biologic Center were all challenges alleging a complete statewide failure to produce any TMDLs for any body of water. 264 F.3d at 1023; 297 F.3d at 883; 762 F. Supp. at 1429. Nevertheless, the interpretation and application of Scott by these cases to require some amount of effort to complete missing TMDLs in order to avoid a constructive submission is not confined to a wholesale statewide failure. It is equally applicable to particular groups of TMDLs or even single missing TMDLs. See Scott, 741 F.2d at 992. The Court will take up this issue in more detail in Section IV.B.iii below. -27- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 28 of 39 PageID #: 6962 impairment. WVDEP’s position on SB 562 defies explanation and critically undermines WVDEP’s excuse for not developing the TMDLs at issue and accordingly EPA’s inaction. Moreover, SB 562, a state law, cannot be the justification for WVDEP’s failure to comply with the CWA. It is a bedrock principle of American federalism enshrined in Article VI of the Constitution that a state law in itself cannot limit the scope of an otherwise constitutional federal law. “[a]bsent statutory authority in the CWA . . . it cannot possibly be urged that . . . state law in itself can contradict or limit the scope of the CWA, for that would run squarely afoul of our Constitution’s Supremacy Clause. N. Plains Res. Council v. Fidelity Expl. & Dev. Co., 325 F.3d 1155, 1165 (9th Cir. 2003) (citing U.S. Const. art. VI, cl. 2; Nat’l Audubon Soc’y, Inc. v. Davis, 307 F.3d 835, 851 (9th Cir. 2002)). Yet, in effect, that is exactly what EPA and WVDEP are claiming when EPA accepts WVDEP’s position that state law precludes it from issuing TMDLs required by the CWA. Whatever the effect of SB 562, it cannot “contradict or limit the scope of the CWA.” Id. When it became clear to EPA that WVDEP would rely on state law to stop complying with federal law, EPA was obliged to act to ensure the CWA is enforced expeditiously. See § 1313. At first, EPA reminded WVDEP of the supremacy of federal law when it disapproved WVDEP’s 2012 303(d) List and overlisted a number of impaired bodies of water. “Even assuming that SB 562 . . . precludes WVDEP from assessing state waters against . . . narrative water quality criteria . . . , SB 562 is a state law that does not override federal requirements.” EPA Enclosure 1: Review of W. Va.’s 2012 Section 303(d), J.A. 2599. Yet, when considering TMDLs, EPA accepts WVDEP’s claim that state law precludes it from issuing TMDLs, choosing to construe it as a “pause” in TMDL development. -28- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 29 of 39 PageID #: 6963 For support EPA relies on language in the CWA that permits states to submit TMDLs “from time to time” to argue that the CWA imposes no specific schedule on the creation of TMDLs and therefore WVDEP may “pause” its program. Rather, EPA’s argument goes, states may produce TMDLs at a pace consistent with priority ranking and available resources, and here the state mandated search for a new methodology has consumed WVDEP’s resources. Although the Court seriously doubts that a state could avoid the mandatory duty to develop TMDLs by claiming that the CWA imposes no set schedule to develop them,13 EPA’s argument is inapplicable here. See Scott, 741 F.2d at 998 (“We cannot allow the states’ refusal to act to defeat the intent of Congress that TMDL’s be established promptly.” (emphasis added)). 13 The context of the statutory language cited by EPA runs counter to EPA’s intended meaning. The “from time to time” language is found in the same subsection where Congress required states to submit all TMDLs six months after EPA published its list of pollutants to be addressed by TMDLs. § 1313(d)(2). Rather that providing cover for states to plod along, the section indicates that Congress conceived the TMDL program as continuous and iterative. Once all TMDLs were promulgated and incorporated into a state’s water management plan where they would be used to set NPDES permit limits, states were expected to continuously monitor ambient water quality. § 1313(d)(1)(A); 40 C.F.R. § 130.7(d). Where changes in water quality were detected, old TMDLs would need to be revised or retired and new TMDLs might have to be issued if new pollutants started to appear. Id. It is quite obvious that Congress could not predict when and if circumstances for a particular impaired body of water would change and therefore could not command states to produce replacement or new TMDLs on any set schedule. What is clear from the passage of the CWA is that the impetus for amending the Federal Pollution Water Control Act was state intransigence. Kingman Park Civic Ass’n, 84 F.Supp.2d at 7. “Although 1965 legislation required the states to develop water-quality standards for interstate waters within their boundaries, Congress lamented in 1971 that ‘[m]ore than 4 years after the deadline for submission of standards, only a little more than half of the states have fully approved standards.’” Id. (quoting S. REP. NO. 92-414, at 4 (1971), as reprinted in 1972 U.S.C.C.A.N. 3668, 3671.) And lest EPA forget, the stated policy enacted by the CWA was to eliminate the discharge of pollutants by 1985. § 1295(a). The Court seriously doubts that Congress would have created a mandatory regulatory structure that it believed would cure the nation’s water quality ills in a short time only to permit states to develop TMDLs at whatever pace the state deemed fit. “[F]rom time to time” read in context, embodies Congress’ intent that once all TMDLs were issued, states would continue to monitor their waters and revise TMDLs as it became necessary—not that states can sit on their hands until they are ready to act. -29- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 30 of 39 PageID #: 6964 What EPA ignores about WVDEP’s actions since SB 562 is that WVDEP is not developing TMDLs for biologic impairment. It is developing a methodology to determine biologic impairment—the two are not equivalent. The CWA does not permit a state to discontinue developing required TMDLs to work on some other aspect of its water regulation scheme. It is beside the point that EPA believes that the CWA does not impose a set schedule for development of TMDLs. What is undisputed, and determinative, is that states must develop TMDLs. WVDEP has stated unequivocally that it is not developing TMDLs for biologically impaired bodies of water. If WVDEP had continued to work on these TMDLs while it developed a new methodology, or produced a credible schedule for developing TMDLs, EPA’s argument might be compelling. See San Francisco BayKeeper, 297 F.3d at 883; Hayes, 264 F.3d at 1023; Browner, 843 F. Supp. at 1313 (finding that a statewide challenge to Minnesota’s TMDL program failed because Minnesota had concrete plans approved by EPA to produce the missing TMDLs and had produced 43 already). The fact is WVDEP has done neither, and accordingly EPA’s argument is inapplicable.14 iii. EPA’s arguments concerning TMDL prioritization EPA counters that the Court cannot apply the constructive submission doctrine to a subset of TMDLs within a state without invading the discretion of the state committed to it by the CWA to prioritize the development of TMDLs. This same argument was rejected in McLerran, 2015 WL 1188522, at *7, and is now rejected here.15 There is no support in TMDL jurisprudence for EPA’s 14 Also troubling to the Court, but not addressed by either party, is that by refusing to use a methodology to determine biologic impairment, WVDEP has consequently stopped enforcing its narrative water quality standards codified at §§ 47-2-3.2.e–3.2.i of the West Virginia Code. Yet, SB 562 did not change the narrative water quality standards, and importantly, it requires that any enforcement of the narrative water quality standards through a new methodology is at least as protective as the use of WVSCI. W. VA. CODE § 22-11-7b. 15 The McLerran Court addressed many of the same arguments raised by EPA in this case -30- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 31 of 39 PageID #: 6965 contention. In fact, the case that established the constructive submission doctrine, Scott, 741 F.2d at 992, was a challenge to missing TMDLs for a single body of water, Lake Michigan. Presumably both Illinois and Indiana are required to develop TMDLs for more bodies of water than just Lake Michigan. The Scott Court had no trouble finding the doctrine applicable to the plaintiff’s challenge to a specific subset of TMDLs. See id. Moreover, later cases such as San Francisco BayKeeper and Hayes, on which EPA relies, addressed programmatic challenges to a statewide TMDL program. See 297 F.3d at 882–83; 264 F.3d at 1024. Thus, as already noted earlier, where plaintiffs had challenged the entire program, a and therefore provides persuasive authority for addressing EPA’s arguments here. See 2015 WL 1188522, at *5–*7. Although EPA has decided to argue this case in a similar manner, raising many of the same arguments, the outcome of McLerran rests on facts materially distinct from this case. The McLerran Court held that a constructive submission had not occurred where the state of Washington had not finalized a draft TMDL for PCBs in the Spokane River. Id. at *8–*9. Rather, Washington created a “task force” that would address PCBs in the Spokane. Id. at *5. After Washington formed the task force in lieu of the TMDL, the plaintiffs requested EPA to make a determination whether Washington had made a constructive submission. Id. EPA found that Washington’s delay of the TMDL was within its discretion if a TMDL was even needed because of gaps in data and the delays associated with preparing a TMDL. Id. The plaintiffs brought both CWA and APA claims and the court found that a constructive submission had not occurred because Washington did not know the source of 57% of the PCBs. Id. at *8. Thus, compliance with the TMDL would unfairly fall on identified sources. Id. The court further held that the plaintiffs’ assertion that Washington had all but completed the TMDL and then stepped away was incorrect. Id. at *9. The court credited Washington’s interpretation of a draft of the TMDL released as incomplete and preliminary and determined that Washington still needed to conduct additional analyses to complete the TMDL. Id. “Therefore, [Washington’s] failure to submit the PCB TMDL did not clearly and unambiguously indicate its intent to abandon the PCB TMDL.” Id. The court, however, found that EPA’s determination that the task force was a suitable alternative to a TMDL was contrary to law under the APA chiefly because the task force had no goal, no timeline, and did not explain whether it would ever submit a TMDL. Id. at *10–*11. In this case, EPA has made no formal determination of WVDEP’s decision to stop issuing biological impairment TMDLs, although it has voiced concern. More importantly, and as will be discussed in greater detail, gaps in data have not precluded the development of the missing TMDLs—a state law has. Even still, it appears to the Court that there has been a significant accumulation of high quality data identifying not only the sources but also the constituent ions and ion mixes that produce ionic toxicity. Finally, WVDEP has stopped issuing TMDLs for all biological impairments, not just ionic toxicity, but admittedly has sufficient information to issue those TMDLs as evidenced by WVDEP’s production of those TMDLs before SB 562 was passed. -31- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 32 of 39 PageID #: 6966 showing that the state was producing some TMDLs made a constructive submission inapplicable. See id. The question raised in those cases was whether California and Oklahoma had abandoned their TMDL programs. Id. Evidence that proved that both states had developed some TMDLs demonstrated that both states had functioning TMDL programs, however meager. Id. Importantly, those cases and others like them did not address the applicability of the doctrine to challenges aimed at some but not all TMDLs in a state. Thus, they do not support EPA’s contention. See id.; Fox, 93 F.Supp.2d at 542. But, as already noted, Scott was such a challenge, and the doctrine was found applicable. 741 F.2d at 992. More recently, the McLerran Court addressed the issue in more detail and wisely found that a state’s duty to develop a TMDL is not diluted by its ability to choose in which order it will develop required TMDLs. 2015 WL 1188522, at *7. Otherwise, the court reasoned, a state could continually use prioritization to avoid developing particular TMDLs—an absurd result in light of the text of the CWA. Id. It is undisputed that the CWA permits states to prioritize waters on its 303(d) List and commands states to develop TMDLs in accord with that prioritization. § 1313(d). Where a state explicitly announces that it is not developing TMDLs, the state has not reprioritized the TMDLs— it has instead determined that no TMDL will be adopted for an extended and indefinite period. Here, WVDEP has declared that it will not develop TMDLs for biologic impairment and has spun its wheels on something else entirely under the guise of a reprioritization. Reprioritization, however, implies that WVDEP is working on the missing TMDLs, just in a different order. But in reality, WVDEP is not working on the missing TMDLs at all. It is working on an assessment tool with no end in sight and no plan to develop the missing TMDLs anytime soon. “[I]t would be absurd for the Court to hold that a state could perpetually avoid this requirement under the guise -32- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 33 of 39 PageID #: 6967 of prioritization; such an administrative purgatory clearly contravenes the goal and purpose of the CWA.” Id. In a related argument, EPA maintains that only a programmatic failure of a state’s TMDLs can result in a constructive submission. EPA is flat wrong. To return to the Scott case, the Seventh Circuit dealt with a challenge to some, but not all, TMDLs in Illinois and Indiana. 741 F.2d at 992. The plaintiff did not argue that the two states had failed to produce any TMDLs, and the Scott Court did not believe it necessary for the plaintiff to prove a complete programmatic failure in order for the court to find that a constructive submission was applicable to the case. Id. at 998; see also McLerran, 2015 WL 1188522, at *7 (“[A] state’s discretion to prioritize TMDLs over other TMDLs does not remove its ultimate obligation to produce a TMDL for each water pollutant of concern in every 303(d) water segment.”). The Scott Court’s holding is consistent with the text of the CWA. Indeed, the CWA imposes a duty on states to produce TMDLs. See § 1313(d)(1)(C). It does not impose a duty on states to simply have a TMDL program. See id. Although the former may necessitate the latter, the converse is not true. The presence of a robust TMDL program does not discharge the specific duty imposed by the CWA to produce a TMDL for each body of water on a 303(d) List. McLerran, 2015 WL 1188522, at *7. The Court thus rejects the proposition that a TMDL challenge must challenge a statewide failure. iv. EPA’s argument concerning gaps in WVDEP’s data EPA’s final argument addresses TMDLs for biologic impairment caused by ionic toxicity and insists that although WVDEP’s stated policy is that SB 562 precludes the development of the missing TMDLs, WVDEP also lacks crucial information about the constituent pollutants that cause ionic toxicity. WVDEP, EPA argues, should be permitted to postpone TMDL development where -33- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 34 of 39 PageID #: 6968 it lacks necessary information without triggering EPA’s duty to develop TMDLs in their place. There are at least two problems with EPA’s argument, in addition to the previously discussed presence of high quality data about ionic toxicity in the record. See supra Section II.B. First, WVDEP created its own information gap by withdrawing from the pilot TMDL program with EPA. Had WVDEP continued its work with EPA, EPA might have a credible argument that ionic toxicity is a complex problem that requires years to develop a TMDL. As it stands, however, WVDEP’s claimed ignorance about ionic toxicity, and the best way to create a TMDL to address it, is self-imposed. States, and in this case EPA, cannot be permitted to hide behind “technical difficulties” of their own making to avoid federal statutory duties. EPA’s claim that it was justified in not assuming its duties because WVDEP should be permitted to conduct the research needed to develop an ionic toxicity TMDL is a diversion from the relevant issue in this case—WVDEP’s decision to stop issuing TMDLs for biologic impairment because of SB 562. The Court has no doubt that in 2010 WVDEP lacked necessary information to develop an ionic toxicity TMDL. A constructive submission suit at that time may have been premature. Nevertheless, in the last four years WVDEP has not completed the first step to fill in the blanks, claiming quite publically that state law barred it from continuing its work on any biologic impairment TMDL, including ionic toxicity TMDLs. It is this policy, and not technical issues, that has stopped WVDEP from developing required TMDLs. Accordingly, EPA’s arguments about missing data or the complexity of the problem miss the point that WVDEP stopped producing ionic toxicity TMDLs because of SB 562. Second, the excuse given for not developing any biologic impairment TMDLs (including ionic toxicity TMDLs) is SB 562, not gaps in crucial information. Gaps in information arguably exist, although created by WVDEP, but WVDEP has stopped developing all TMDLs that address -34- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 35 of 39 PageID #: 6969 biologic impairment. EPA and WVDEP only claim to lack information about ionic toxicity, not the other causes of biologic impairment. In fact, while WVDEP has never issued a TMDL for ionic toxicity, before SB 562, WVDEP issued numerous TMDLs to address other causes of biologic impairment. See, e.g., Decision Rationale Total Maximum Daily Loads for Selected Streams in the Upper Ohio South River Watershed W. Va., Sept. 2009, J.A. 1035–37. Surely informational deficiencies are not keeping WVDEP from developing TMDLs for other causes of biologic impairment. Thus, EPA’s argument again misses the mark. It was not gaps in data, or any other technical problem, that forced WVDEP to abandon biological impairment TIMDLs, it was SB 562. Consequently, the Court feels obliged to note that the text of SB 562 does not change the narrative WQS and requires that any new methodology “may not establish measurements . . . that would [be] less protective than [WVSCI].” This requirement sets a floor for whatever methodology is eventually developed, meaning that any new methodology should not generally result in delisting streams on West Virginia’s 303(d) List already found to be impaired as determined by WVSCI. Bodies of water that already have been determined to be biologically impaired as a result of their WVSCI score, the bodies of water at issue here, will therefore still be considered impaired when using a new methodology. Were WVDEP to develop TMDLs for these bodies of water using WVSCI, they would quite obviously meet West Virginia’s WQS and fulfill its CWA duties. Moreover, had WVDEP adopted GLIMPSS as a way to comply with SB 562, as EPA urged in 2014, and produced TMDLs based on GLIMPSS, these would also quite obviously comply with West Virginia’s WQS and the CWA. This is all to say that WVDEP could develop TMDLs with either of the two valid and accepted methodologies and still develop a new methodology with little consequence for TMDLs produced with the established methodologies. -35- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 36 of 39 PageID #: 6970 That approach, in practice, may mean that if a TMDL is produced using WVSCI, WVDEP might need to revisit that TMDL in the future, as it will have to do for all the biological impairment TMDLs issued before 2012 because they were developed using WVSCI. Of course, the use of GLIMPSS might prospectively solve that issue. Still, a review should not be a major impediment since it is WVDEP’s policy to conduct a TMDL review process in every watershed every few years as part of its TMDL program. WVDEP’s refusal to produce TMDLs, using either WVSCI or GLIMPSS, makes little sense when streams that already have been determined to be impaired will not be delisted with a new methodology and will still need a TMDL. Lastly and most importantly, the fundamental federal structure of our Constitution demands that West Virginia’s federal duties must prevail over SB 562. Notwithstanding, with GLIMPSS as a viable alternative that could comply with SB 562, it is not entirely clear what reason WVDEP has for not producing biological impairment TMDLs. What is clear, however, is that EPA must act when WVDEP’s actions either clearly run afoul of federal supremacy, if SB 562 is truly the reason for WVDEP’s inaction, or have no basis in fact, as when WVDEP laments that it has no methodology that it can use. Neither reason is a proper basis for EPA to accept inaction. C. Conclusion WVDEP has explicitly stated first that it cannot develop biologic impairment TMDLs due to state law and then proposed dates to develop TMDLs, which the Court finds are unsubstantiated and unrealistic. This is especially true for the subset TMDLs that address ionic toxicity. WVDEP has not even been able to produce a schedule for when it will finalize its new assessment methodology. It is clear and unambiguous that WVDEP has decided not to submit TMDLs for biologically impaired bodies of water. WVDEP’s complete abdication of its CWA duties and utter lack of a plan to comply with those duties is a constructive submission of no TMDLs for all -36- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 37 of 39 PageID #: 6971 biologically impaired bodies of water, triggering EPA’s nondiscretionary duty to approve or disapprove of the submission. D. The APA claims The Environmental Groups also move for summary judgment on their APA claims. EPA brought a cross motion for summary judgment directed at the Environmental Groups’ APA claims. The Environmental Groups’ APA claims challenge EPA’s approval of TMDLs for six watersheds that did not include TMDLs for ionic toxicity as arbitrary, capricious, or otherwise contrary to law. Although the APA claims are directed solely at TMDLs for ionic toxicity, the CWA claims address all biologically impaired bodies of water in West Virginia, of which all waterbodies impaired by ionic toxicity are a part. The APA claims are subsumed by the CWA claims and are duplicative. “Congress did not intend the general grant of review in the APA to duplicate existing procedures for review of agency action.” Bowen v. Massachusetts, 487 U.S. 879, 903 (1988). Indeed, the APA limits review of agency action to agency actions “for which there is no other adequate remedy in a court.” 5 U.S.C. § 704. “Because review of [the Environmental Groups’] claim is available under the Clean Water Act, it is not subject to review under the APA.” Hayes, 264 F.3d, 1025 (emphasis in original); see also Allegheny Cty. Sanitary Auth. v. EPA, 732 F.2d 1167, 1177 (3d Cir. 1984) (finding plaintiff could not sustain an APA claim when it had a valid claim pursuant to the CWA); Am. Canoe Ass’n, Inc. v. EPA, 30 F.Supp.2d 908, 915 n. 6 (E.D. Va. 1998) (same). Accordingly, the Court grants summary judgment in favor of EPA on the Environmental Groups’ APA claims. E. Remedy The Environmental Groups’ CWA claim was brought pursuant to 33 U.S.C. § 1365 which authorizes a citizen to bring suit against the administrator of the EPA “where there is alleged a failure of the Administrator to perform any act or duty under this chapter which is not -37- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 38 of 39 PageID #: 6972 discretionary.” EPA has a nondiscretionary duty to approve or disapprove within thirty days a TMDL submission by West Virginia. WVDEP’s decision not to submit biological impairment TMDLs to EPA for approval triggered EPA’s duty to either approve or disapprove of the submission that no TMDLs are necessary for biological impairment within thirty days. Accordingly, EPA must, within thirty days of this order, approve or disapprove of WVDEP’s submission of no TMDLs for all bodies of water identified as biologically impaired but lack a TMDL to address that impairment. The Environmental Groups urge the Court to require EPA to produce the missing TMDLs within thirty days as if EPA had disapproved WVDEP’s constructive submission because they believe that EPA cannot approve a submission of no TMDLs for bodies of water listed on a 303(d) List. There is a certain simplicity and harmony to the Environmental Group’s argument. For every body of water listed on West Virginia’s 303(d) List it must also produce a TMDL. It stands to reason then that EPA could never approve a submission that no TMDL is needed for a body of water on a 303(d) List. Nevertheless, the Court does not believe that it can order EPA to produce the TMDLs without EPA first disapproving West Virginia’s constructive submission. Section 1365(a)(2) permits suit to compel EPA to perform a nondiscretionary duty. EPA inherits a nondiscretionary duty to act (approve or disapprove) once a submission, or constructive submission, has been made. Scott, 741 F.2d at 998; Fox, 909 F. Supp. at 158. The CWA commits to EPA’s discretion its choice between approval and disapproval. § 1313(d)(2); see Hayes, 264 F.3d at 1023; Scott 741 F.2d at 995. Where it disapproves a submission, it then inherits a nondiscretionary duty to produce the TMDLs that were disapproved. § 1313(d)(2). If EPA were to approve a constructive submission, essentially deciding in this case that no biological impairment TMDLs are necessary, that decision -38- Case 3:15-cv-00271 Document 87 Filed 02/14/17 Page 39 of 39 PageID #: 6973 would then likely be subject to judicial review under the APA. At present, however, the Court can only order the EPA to perform the nondiscretionary duty which it has failed to assume—approve or disapprove WVDEP’s constructive submission. V. Conclusion For the reasons stated in this Memorandum Opinion and Order, the Court GRANTS in part and DENIES in part the Environmental Groups’ Motion for Summary Judgment, ECF No. 30, and GRANTS in part and DENIES in part EPA’s Cross Motion for Summary Judgment. ECF No. 38. EPA shall approve or disapprove WVDEP’s constructive submission of no TMDLs for all biologically impaired bodies of water for which no TMDL has been developed to address that impairment within thirty days of this order. EPA shall inform the Court upon reaching its decision. The Court will retain jurisdiction of the case until EPA submits its decision. The Court DIRECTS the Clerk to send a copy of this Order to counsel of record and any unrepresented parties. ENTER: -39- February 14, 2017