United States Court of Appeals
                        For the First Circuit


No. 11-1474
         UPPER BLACKSTONE WATER POLLUTION ABATEMENT DISTRICT,
                             Petitioner,
                                  v.
            UNITED STATES ENVIRONMENTAL PROTECTION AGENCY,
                             Respondent.

No. 11-1610
                  CONSERVATION LAW FOUNDATION, INC.
                             Petitioner,
                                  v.
         UPPER BLACKSTONE WATER POLLUTION ABATEMENT DISTRICT,
                             Intervenor,
            UNITED STATES ENVIRONMENTAL PROTECTION AGENCY,
                             Respondent.


     PETITIONS FOR REVIEW OF A FINAL PERMIT DECISION BY THE
          UNITED STATES ENVIRONMENTAL PROTECTION AGENCY



                                Before
                         Lynch, Chief Judge,
                     Souter, Associate Justice,*
                      and Stahl, Circuit Judge.




     *
          The Hon. David H. Souter, Associate Justice (Ret.) of the
Supreme Court of the United States, sitting by designation.
          Robert D. Cox, Jr., with whom Douglas T. Radigan,
Bowditch & Dewey, LLP, Fredric P. Andes, and Barnes & Thornburg
LLP, were on brief, for petitioner Upper Blackstone Water Pollution
Abatement District.
          Christopher M. Kilian, with whom Anthony N.L. Iarrapino
was on brief, for petitioner Conservation Law Foundation.
          Madeline Fleisher, with whom Ignacia S. Moreno, Assistant
Attorney General, U.S. Department of Justice, Environment and
Natural Resources Division, and Samir Bukhari, Ira W. Leighton,
Karen A. McGuire, U.S. Environmental Protection Agency, were on
brief, for respondent.
          Donald L. Anglehart on brief for City of Marlborough,
amicus curiae.
          David M. Moore, City Solicitor, and Jennifer H. Beaton,
Assistant City Solicitor, on brief for City of Worcester, amicus
curiae.
          Karma B. Brown, Brooks M. Smith, Hunton & Williams LLP,
and Nathan Gardner-Andrews on brief for National Association of
Clean Water Agencies, amicus curiae.




                          August 3, 2012




                               -2-
           LYNCH, Chief Judge.                 These petitions seek review of

certain    effluent       limitations          imposed     by    the     Environmental

Protection      Agency    (EPA)      in    a    National        Pollutant      Discharge

Elimination System (NPDES) permit on the discharges of Upper

Blackstone Water Pollution Abatement District, a sewage treatment

plant located in central Massachusetts.

           The District's discharges are into the headwaters of a

polluted river which, in due course, flows into other rivers, and

ultimately      empties      into    Narragansett         Bay.         The   states   of

Massachusetts and Rhode Island each have strong interests in the

health of these waters and generally have supported the EPA's

decisions during the permitting process.                  The District, supported

by its member towns, has an interest in avoiding compliance costs

associated      with   the    permit      and    has     challenged      the    effluent

limitations as premature and unsupported by the scientific record.

           We have stayed enforcement of the permit during this

appeal    and    while     the      parties      were     engaged      in    settlement

negotiations in a court-sponsored settlement program.                        We now lift

the stay, deny the petitions, and find no error in the EPA's final

permit decision.

                                          I.

           The Blackstone River is a major, interstate freshwater

river which runs south from Worcester, Massachusetts, crosses the

border into Rhode Island, and continues on to Pawtucket Falls.


                                          -3-
There, it reaches sea level, becomes tidal, and changes its name to

the Seekonk River, which, in turn, flows into the Providence

River,1    and   ultimately   empties    into   Narragansett   Bay.   The

Blackstone River provides a significant source of freshwater to the

Bay.

            At the peak of the industrial revolution, water-powered

textile mills lined the Blackstone River; dams, millponds, and

canals altered its natural course and halted its flow at points.

Toxic sediments of heavy metals and other industrial waste products

released into the River accumulated behind its many impoundments

and damaged its ecology.      Today, industry has moved on; its legacy

remains in leftover dams and the toxic sediments held in place

behind them.

            With the discontinuation of industrial river dumping, the

River's health has dramatically improved.         Massachusetts and Rhode

Island now seek to put the River to new economic and recreational

uses including tourism, recreation, and commercial fishing.           The

new limiting factor, and the subject of dispute in this case, is

not the River's industrial legacy, but sewage treatment.               As

population has increased along the River, sewage processing has not

kept apace.      An influx of nitrogen and phosphorus from sewage




       1
          We will refer to the Blackstone River as "the River" on
occasion; the Seekonk and Providence Rivers, by their full names
only; and all three together as "the three rivers."

                                   -4-
treatment plants is causing serious problems for the River's waters

and those downstream.

           The    Blackstone,     Seekonk,     and    Providence   Rivers,    and

Narragansett Bay, all suffer from severe cultural eutrophication,

a process fueled by unnaturally high concentrations of nitrogen and

phosphorus.   When excessive levels of these chemical nutrients are

introduced into a water system, algae populations rapidly multiply

to nuisance levels.         As populations "bloom" and die-off in quick

succession,      dead     algae   accumulate       and     decompose   --   their

nutrient-laden remains further enriching the immediate environment,

thereby perpetuating the eutrophication cycle.               Increased rates of

respiration   and       decomposition    deplete     the   available   dissolved

oxygen in the water, threatening other plant and animal life in the

system.    When oxygen saturation levels drop below what is needed

by fish and invertebrates to breathe, the waters become host to

fish kills, red tides, and shellfish poisonings, events which can

pose threats to human health as well.

           Phosphorus drives cultural eutrophication in freshwater

systems and nitrogen drives the same process in marine waters. The

Blackstone River currently suffers from severe phosphorus-driven

cultural eutrophication.          Algae blooms, thick, cloudy waters,

putrid smells, and sudden fish kills periodically contaminate its

waters.   The numerous dams and impoundments along the River create

areas of stagnant water where nutrients collect and cultural


                                        -5-
eutrophication flourishes.     The toxicity build-up behind the dams

and   serious   concerns   about   resuspension   of   the   contaminated

sediments rule out an easy solution to this problem.

           Narragansett Bay and the Seekonk and Providence Rivers,

in turn, are each affected by the Blackstone's degraded waters.

Narragansett Bay, the ultimate depository for all the nutrients

carried by the Blackstone, suffers from severe nitrogen-driven

cultural eutrophication.       The Seekonk River, which forms the

uppermost part of the Bay, is the most seriously impaired by the

Blackstone's nitrogen loadings.

           Conditions in the three rivers and the Bay have been

deteriorating for many years. Increased domestic waste inputs into

the rivers are worsening their nutrient-related problems.              Among

the numerous events documented in the record, severely hypoxic

(waters characterized by levels of dissolved oxygen below what is

needed by aquatic organisms to breathe) to nearly anoxic (waters

completely depleted of dissolved oxygen) conditions, along with

associated fish kills, were observed in upper Narragansett Bay,

including the Providence River, in the summers of 2001 and 2002.

August 2003 witnessed one of the Bay's largest fish kills in

history, when more than one million fish died in anoxic water

conditions near East Greenwich, Rhode Island.

           The Rhode Island Department of Environmental Management

(RIDEM)   has   set   up   response   teams   which    monitor   the    Bay


                                   -6-
continuously and publish public notices when bacterial or pollution

conditions pose a threat to public health and commercial fishing.2

In recent years, the state has been forced to close down some of

the Bay's beaches and commercial fishing grounds entirely, measures

which damage state tourism and recreation businesses, and which

place the state's commercial fishing and shellfishing industries in

jeopardy.

            Recognizing     the    watershed's      growing       problems,   and

motivated    by    the    desire    to    improve       its   resource    value,

Massachusetts      and    Rhode     Island       have     begun     implementing

comprehensive plans to rehabilitate the three rivers and the Bay.

These efforts build on decades of work by both government actors

and private groups to study and address nutrient-related problems

in the watershed.

            Congress designated the Blackstone River Valley as a

National Heritage Corridor in 1986 for the purpose of recognizing

the   historical    significance     of    the    River    and    restoring   its

watershed.    The EPA formed the Narragansett Bay Project in the

1980s and the Blackstone River Initiative in the 1990s, to study,



      2
          See   Bay   Assessment    &  Response   Team,    RIDEM,
http://www.dem.ri.gov/bart/index.htm   (information     regarding
monitoring and closure of Narragansett Bay beaches and fisheries
when bacterial or pollution levels threaten public health) (last
visited   Aug.   2,  2012);   Office  of   Water  Res.,    RIDEM,
http://www.dem.ri.gov/programs/benviron/water/shellfsh/index.htm
(information regarding monitoring and closure of shellfishing
grounds) (last visited Aug. 2, 2012).

                                     -7-
among other issues, the impacts of cultural eutrophication on the

water systems.       The Governors of Massachusetts and Rhode Island

first signed a Memorandum of Understanding in 1992 to underscore

the    two   states'   commitments    to   studying       and   restoring   the

watershed.     In 1998, President Clinton designated the Blackstone

River an American Heritage River.          Bills "[t]o establish the John

H. Chafee Blackstone River Valley National Historic Park" are

currently before both houses of Congress.             S. 1708, 112th Cong.

(2011); H.R. 3191, 112th Cong. (2011).

             Federal, state, and local governments, businesses, and an

array of outside groups and coalitions have funded and conducted

numerous scientific studies on nutrient-related problems in the

three rivers and the Bay.     The EPA considered many of these studies

in setting the 2008 permit limits; just a few of those included in

full   in    the   administrative    record   in   this    case   are   studies




                                     -8-
conducted by Massachusetts3 and Rhode Island,4 as well as the U.S.

Army Corps of Engineers5 and the EPA.6

          Although nitrogen and phosphorus end up in the rivers and

the Bay from diverse sources, including storm run-off, agricultural


     3
          See Fiorentino, Div. of Watershed Mgmt., Mass. Dep't of
Envtl. Prot., Blackstone River Watershed 2003 Biological Assessment
(2006) (comprehensive report on Blackstone River water quality
incorporating historical perspectives and previous studies,
including 2003 biomonitoring survey); Tamul, Div. of Watershed
Mgmt., Mass. Dep't of Envtl. Prot., Blackstone River Watershed 2003
DWM Water Quality Monitoring Data (2005) (biomonitoring survey of
water quality including nitrogen and phosphorus inputs and related
effects); Weinstein et al., Div. of Watershed Mgmt., Mass. Dep't of
Envtl. Prot., Blackstone River Basin 1998 Water Quality Assessment
(2001) (comprehensive evaluation of water quality in Blackstone
River and related tributaries, and specific recommendations for
managing nitrogen- and phosphorus-related water quality problems).
     4
          See Nixon et al., Anthropogenic Nutrient Inputs in
Narragansett Bay, A Twenty-five Year Perspective: A Report to the
Narragansett Bay Commission and Rhode Island Sea Grant (2005)
(study of nitrogen and phosphorus sewage inputs into Narragansett
Bay over a twenty-year period, with measurements taken in 1975,
1976, 1983, 1991, 1992, 2003, and 2004); Governor's Narragansett
Bay Watershed Planing Comm'n, Nutrient and Bacteria Pollution Panel
Initial Report (2004) (study and management plan for addressing the
problems with cultural eutrophication in the Bay); RIDEM,
Evaluation of Nitrogen Targets and WWTF Load Reductions for the
Providence and Seekonk Rivers (2004) (reporting results of Rhode
Island's TMDL efforts and a management plan for addressing cultural
eutrophication in the Bay).
     5
          See Wright et al., Dry Weather Water Quality Sampling and
Modeling, Blackstone River Feasibility Study (2004) (for U.S. Army
Corps of Engineers) (study of water quality conditions in
Massachusetts segment of the Blackstone River for future use in
developing a TMDL).
     6
          See Wright et al., Blackstone River Initiative, Water
Quality Analysis of the Blackstone River Under Wet and Dry weather
Conditions (2001) (for EPA New England) (integrated water quality
study and report on both Massachusetts and Rhode Island segments of
the River and Narragansett Bay).

                               -9-
fields, and construction sites, sewage treatment facilities are the

primary source of anthropogenic nutrient inputs into the Seekonk

and Providence Rivers and the Bay.        Thus, a critical component of

both states' rehabilitation plans has been to impose tighter

limits, under the Clean Water Act (CWA or the "Act"), on the

amounts of nitrogen and phosphorus that sewage treatment facilities

may discharge into the rivers and the Bay.

           The CWA was enacted by Congress to address the serious

threats water pollution poses to public health, economic activity,

and the long-term viability of the Nation's water resources.             33

U.S.C. §§ 1252(a), 1313(c)(2)(A).         The Act's primary goal is "to

restore   and   maintain   the   chemical,   physical,     and   biological

integrity of the Nation's waters."        

and the

federal government share responsibility for achieving this goal.

v. Oklahoma, 

, 101 (1992).

           States have primary responsibility for designating the

ambient water quality of the waters within their territory.              33

U.S.C. § 1313(c)(1), (2)(A).      These "water quality standards" are

expressed as "designated uses" of water bodies (such as propagation

of aquatic life, recreation, aesthetics, and use as public water

supply), and as numeric or narrative "criteria," which specify the

amounts of pollutants that may be present in these water bodies

without impairing their designated uses.       

addition to incorporating state water quality standards, the Act


                                   -10-
also employs federal, technology-based effluent limitations on

individual discharges of pollution into navigable waters.                          

       State    water       quality     standards      generally

supplement these effluent limitations, so that where one or more

point   source    dischargers,         otherwise       compliant      with     federal

conditions, are nonetheless causing a violation of state water

quality standards, they may be further regulated to alleviate the

water quality violation.        

shall be

achieved . . . any more stringent limitation, including those

necessary    to meet    water    quality          standards   .   .   . established

pursuant to any State law or regulations . . . ."); see also 

1313(d)(1)(A), (d)(2), (e)(3)(A).

             "[A]ny person" who wishes to discharge "any pollutant"

from a "point source" into the navigable waters must obtain an

NPDES permit.     

              NPDES permits bring both

state ambient water quality standards and technology-based effluent

limitations to bear on individual discharges of pollution, 

and tailor these to the discharger through

procedures laid out in the Act and in EPA regulations, 

permits may be administered by the EPA or by an authorized

state   or   Indian   tribe.         

   40    C.F.R.

§ 123.31.     To date, the EPA has authorized forty-six states to

administer    their    own    NPDES    permit       programs,     including      Rhode




                                           -11-
Island.7   Massachusetts has not received authorization, and so the

EPA administers NPDES permits in that state.

           The CWA also requires states to identify the waters

within their boundaries that fail to meet their designated water

quality standards and rank these in order of priority, taking into

account "the severity of the pollution and the uses to be made of

such waters."      33 U.S.C. § 1313(d)(1)(A).    States must then begin

the planning process for bringing these waters into compliance with

water    quality    standards.8   

  40   C.F.R.

§ 122.44(d)(1).




     7
          See   EPA,   NPDES  Specific   State   Program   Status,
http://cfpub.epa.gov/npdes/statestats.cfm?view=specific      (last
visited Aug. 2, 2012); see also 49 Fed. Reg. 39,063 (Oct. 3, 1984)
(approving Rhode Island's NPDES program).
     8
          Part of this process requires the development of Total
Maximum Daily Loads (TMDLs) for each pollutant that is responsible
for a violation of water quality standards.             33 U.S.C.
§ 1313(d)(1)(C). A TMDL is a calculation of the maximum quantity
of a pollutant that may be added to a water body from all sources
without exceeding applicable water quality standards including "a
margin of safety which takes into account any lack of knowledge
concerning the relationship between effluent limitations and water
quality." 

and resources to develop and have
proven to be difficult to get just right; thus, under EPA
regulations, permitting authorities must adopt interim measures to
bring water bodies into compliance with water quality standards.

C.F.R. § 122.44(d); see also, e.g., 43 Fed.
Reg. 60,662, 60,665 (Dec. 28, 1978) ("EPA recognizes that State
development of TMDL's and wasteload allocations for all water
quality limited segments will be a lengthy process. Water quality
standards will continue to be enforced during this process.
Development of TMDL's . . . is not a necessary prerequisite to
adoption or enforcement of water quality standards . . . .").

                                  -12-
           In some circumstances, discharge into the waters of one

state may cause a violation of water quality standards in a

downstream state.      The CWA anticipates conflicts over pollution

discharges between upstream and downstream states.           See Milwaukee

v. Illinois, 

, 325-26 (1981).          When an application is

made for a discharge which may affect the water quality of a

downstream state, the EPA is required to notify both the origin

state and the downstream state.          33 U.S.C. § 1341(a)(2).     If the

downstream state then determines that the discharge will violate

its water quality standards, it may submit its objections and

request a public hearing.       

has held that the CWA grants the EPA

authority to require in NPDES permits conditions which ensure

compliance    with   the   water   quality   requirements    of   downstream

states.   

; see 33 U.S.C. § 1341(a)(2)

("[The permitting agency] shall condition such license or permit in

such   manner   as   may   be   necessary    to   insure   compliance   with

applicable water quality requirements.             If the imposition of

conditions cannot insure such compliance such agency shall not

issue such license or permit.").        EPA regulations have so required

since 1973.     See 40 C.F.R. § 122.4(d) ("No permit may be issued .

. . [w]hen the imposition of conditions cannot ensure compliance

with the applicable water quality requirements of all affected

States . . . .").


                                      -13-
                In this case, both Massachusetts and Rhode Island have

listed the Blackstone River as "impaired" under Section 1313(d) of

the CWA; Rhode Island has also listed the Seekonk and Providence

Rivers and Narragansett Bay as impaired.

               Massachusetts has designated the Blackstone River for

primary and secondary contact uses, including swimming, fishing,

and boating, and as habitat for fish and other wildlife. 314 Mass.

Code Regs. 4.05(3)(b).9            Under Massachusetts' narrative water

quality       standards,   the    Blackstone   River   must   be   "free   from

pollutants in concentrations or combinations that settle to form

objectionable deposits; float as debris, scum or other matter to

form       nuisances;   produce   objectionable   odor,   color,     taste   or

turbidity; or produce undesirable or nuisance species of aquatic

life;" "free from pollutants in concentrations or combinations or

from alterations that adversely affect the physical or chemical

nature of the [river's] bottom, interfere with the propagation of

fish or shellfish, or adversely affect populations of non-mobile or

sessile benthic organisms;" "free from pollutants in concentrations

that are toxic to humans, aquatic life or wildlife," and "free from

nutrients in concentrations that would cause or contribute to

impairment of existing or designated uses," 

all times, even under low flow conditions, 

           See Massachusetts Water Quality Designations, available
at   http://www.mass.gov/dep/water/laws/regulati.htm#wqual   (last
visited Aug. 2, 2012).

                                      -14-
          Massachusetts has determined that the Blackstone River

fails to meet state water quality standards.        In its testing and

analysis of the contaminants in the River, the Massachusetts

Department of Environmental Protection (MassDEP) has documented

multiple impairments including unknown toxicity, priority organics,

metals, ammonia,    chlorine,    nutrients,   organic   enrichment,   low

dissolved oxygen, flow and other habitat alterations, pathogens,

suspended solids, turbidity, objectionable deposits, and taste,

odor, and color objections.      Watershed Planning Program, Office of

Watershed Mgmt., Massachusetts Year 2006 Integrated List of Waters

81-82 (2006).

          Rhode Island has designated the Blackstone, Seekonk, and

Providence Rivers and Narragansett Bay for primary and secondary

contact recreational uses and as habitat for wildlife.10            Rhode

Island's narrative water quality criteria require that all three

rivers and the Bay be free of pollutants in concentrations that

adversely affect the composition of fish and wildlife; adversely

affect the physical, chemical, or biological integrity of the

habitat; interfere with the propagation of fish and wildlife; or

adversely alter the life cycle functions, uses, processes, and

activities   of   fish   and   wildlife.   With   respect   to   nutrient

pollution, Rhode Island requires that the three rivers and the Bay


     10
          See Rhode Island Water Quality Regulations, available at
http://www.dem.ri.gov/pubs/regs/regs/water/h2oq10.pdf (last visited
Aug. 2, 2012).

                                   -15-
be free of nutrients "in such concentration that would impair any

[designated uses] . . . or cause undesirable or nuisance aquatic

species associated with cultural eutrophication."

              Rhode Island has determined that all three rivers and the

Bay fail to meet its water quality standards. RIDEM has documented

numerous impairments in the Rhode Island segment of the Blackstone

River including ammonia, copper, lead, pathogens, nutrients, low

dissolved oxygen, and biodiversity impacts.                Rhode Island monitors

Narragansett Bay and its extensions particularly closely due to

their importance to state industries.                 It has documented numerous

impairments     to    the    Seekonk     and    Providence     Rivers,      including

nutrient pollution, low dissolved oxygen, and excessive algae

growth, and similar impairments to the Bay, including nutrients,

low dissolved oxygen, and pathogens.

              In order to address these impairments, Rhode Island has

issued several Rhode Island Pollutant Discharge Elimination System

permits (RIPDES permits) to the major sewage treatment facilities

along   the     rivers      and   Bay,   which       tighten   nitrogen      effluent

limitations.        The two largest treatment facilities, both on the

Providence River, Narragansett Bay Commission Fields Point and

Narragansett        Bay   Commission     Bucklin      Point,    are    designed    to

discharge      65    million      gallons      per   day   (mgd),     and   31    mgd,

respectively.        As part of its major nitrogen removal initiative,

RIDEM has issued both facilities nitrogen effluent limitations of


                                         -16-
5.0 mg/L.    RIDEM has also set a 5.0 mg/L nitrogen limit for East

Greenwich, a much smaller facility, with an average daily flow of

approximately 1.7 mgd, but which is located on a particularly

impaired portion of Narragansett Bay.           The Woonsocket facility,

which, behind the petitioner District in this case, is the second-

largest sewage treatment plant discharging into the Blackstone

River, has been given a 3.0 mg/L nitrogen limit as part of a

consent agreement.     Five other much smaller facilities have been

given nitrogen limits of 8.0 mg/L.

            As part of this process, in an effort to reduce the

incoming nitrogen into the Bay, Rhode Island also requested and

recommended to the EPA that the nitrogen limits on Massachusetts

dischargers into the Blackstone River be tightened as well.           While

nitrogen discharge does not cause cultural eutrophication in the

Blackstone River's fresh-waters, the discharge is swiftly carried

downstream to Rhode Island's saltwater rivers and the Bay, where it

produces severe cultural eutrophication and resulting violations of

Rhode Island's water quality standards.

                                   II.

            Against   this   complex   backdrop,   the   present    dispute

arises.     The petitioner in this case, Upper Blackstone Water

Pollution Abatement District (the "District"), is the largest

sewage treatment plant along the Blackstone River.          It is located

in   Millbury,   Massachusetts,   very   near    the   Blackstone   River's


                                  -17-
headwaters.   It discharges approximately 34 to 43 mgd of treated

domestic and industrial sewage11 into the River.12

          The District's discharge represents approximately seventy

percent of the total municipal wastewater flow into the Blackstone

River, making it the dominant discharger of both nitrogen and

phosphorus into the River's waters.

          The District's plant came online in 1976, and only

recently went through its first major upgrade.   This comprehensive

upgrade was completed pursuant to an administrative consent order

issued to the District by the EPA after the District had violated

its September 30, 1999, NPDES permit, as modified by an August 3,

2001, Settlement Agreement (the "2001 permit").        The upgrade

involved extensive plant renovations implemented over an eight-year

period, through which the District adapted its facilities to comply

with the 2001 permit's 0.75 mg/L limit on phosphorus, and, although

the permit did not limit nitrogen discharge, a 8.0 - 10.0 mg/L

limit on nitrogen in anticipation of future nitrogen controls.13


     11
          More than 200 industrial users contribute wastewater to
the District's facilities, thirty-three of which currently qualify
for the "pre-treatment" program under 40 C.F.R. § 403.3(v).
     12
          This is somewhat below its design flow capacity of 56
mgd. The next largest sewage treatment plant on the Blackstone
River is located in Woonsocket, Rhode Island, and has a design flow
of 16 mgd, and an actual average discharge of 7 mgd.
     13
          These needed upgrades to the District's aging facility
cost $180 million and resulted in rate increases for the District's
customers. However, even with these upgrades, and as was noted in
the administrative record, relative to other Massachusetts

                               -18-
            On November 8, 2005, while the upgrade was still ongoing,

the District submitted a timely application to the EPA for renewal

of the 2001 permit.14         As part of the permit reissuance process,

the EPA evaluated a variety of factors, including the District's

expected future discharge -- accounting for the upgrade -- and the

state of the receiving waters.          The EPA found that all three rivers

and   the   Bay   exhibited    severe    nitrogen-   and   phosphorus-driven

cultural eutrophication, and that the District's discharge was the

predominant point source of both phosphorus and nitrogen in the

Blackstone River.

            Applying Massachusetts and Rhode Island water quality

requirements, the EPA determined that the District's nitrogen and

phosphorus discharges "will cause, have the reasonable potential to

cause, or contribute to an excursion above" applicable state water

quality standards.      40 C.F.R. § 122.44(d)(1)(i).           Based on its

comprehensive analysis of these and the other required factors, the

EPA concluded that lower limits on the District's nitrogen and

phosphorus discharge were necessary to achieve compliance with



residents, the District's ratepayers pay significantly less than
the average sewage rate. For 2011 figures, see Mass. Water Res.
Auth. Advisory Bd., Annual Water and Sewer Retail Rate Survey
(2011), available at http://mwraadvisoryboard.com/wp-content/uploa
ds/2012/01/0-COMBINED-MASTER2.pdf (last visited Aug. 2, 2012).
      14
          Although it expired in 2006, during the permit reissuance
process that followed, the 2001 permit was administratively
continued, and remained in effect after this court granted the
District's motion to stay the 2008 permit on April 29, 2011.

                                     -19-
state   water   quality   standards.     See   

both Massachusetts and Rhode Island employ narrative water

quality criteria for the relevant pollutants, the EPA translated

these into numeric limits under its procedures set out in 40 C.F.R.

§ 122.44(d)(1)(vi).

           On March 23, 2007, the EPA published a draft permit that

limited total phosphorus discharge to 0.1 mg/L from April 1 through

October 31, and 1.0 mg/L from November through March, and limited

total nitrogen to 5.0 mg/L from May 1 through October 31, and

imposed a narrative criteria for nitrogen during the remaining

months.

           As part of the lengthy, public permitting process that

followed, the EPA published the draft permit and its accompanying

rationale in full, accepted public comments -- extending the time

for these from thirty to sixty-four days -- and held a public

hearing on the permit.     See 40 C.F.R. § 124.10.        The EPA received

and considered thirty-four sets of written comments from a variety

of stakeholders, interested parties, individuals, and researchers,

including the District, the states of Massachusetts and Rhode

Island, several municipalities, and numerous other organizations.

The EPA responded to each set of comments at length.

           On August 22, 2008, the EPA issued the final permit,

which contained the same limits on phosphorus and nitrogen proposed

in the draft permit.       In addition, on April 15, 2009, the EPA


                                  -20-
issued a draft permit modification proposing an effluent limitation

for aluminum discharge in order to comply with Massachusetts'

aluminum criterion.     After public comment, the EPA issued a final

permit modification adopting this limitation into the final permit.

           On September 15, 2008, the District filed a petition for

review of the permit with the EPA's highest adjudicative body, the

Environmental Appeals Board (the "EAB" or "Board"), see 40 C.F.R.

§   1.25(e),   appealing,   among    other    provisions,     the   permit's

phosphorus, nitrogen, and aluminum15 discharge limits.          Seven other

parties also filed petitions for review with the Board: MassDEP,

the Massachusetts towns of Holden, Millbury, and Worcester, the

Conservation    Law   Foundation    (CLF),   the   Northern   Rhode   Island

Chapter of Trout Unlimited, and Cherry Valley Sewer District.

MassDEP raised several objections to the methodology employed by

the EPA in setting the nitrogen limit.             RIDEM filed an amicus

curiae brief in support of the permit's nitrogen limit, citing the

comparable nitrogen limits it had imposed on similarly situated

Rhode Island facilities.     CLF contended that both the nitrogen and

phosphorus limits were too high.

           The District challenged multiple aspects of the permit,

re-raising many of the points it had made in its comments on the

draft permit.     In particular, the District challenged: (1) the


     15
          The District filed a separate petition for review of the
aluminum limit, which the Board consolidated with the District's
original petition.

                                    -21-
EPA's decision to tighten the nitrogen and phosphorus limits before

the District had fully completed its facility upgrades and without

more data on nutrient impairment in the Bay; (2) the EPA's refusal

to delay issuance of the permit until a new computer model of the

Blackstone River, then under development by the District, could be

completed,    and   (3)   the    EPA's    conclusion   that     the    District's

aluminum    discharge     had    a    reasonable    potential     to    cause    or

contribute to a violation of Massachusetts' water quality criterion

for aluminum.

            On May 28, 2010, the Board issued a 106-page decision

upholding the permit, with the exception of a provision that made

several other municipal entities "co-permittees," which the Board

remanded to the EPA for further action.             In re: Upper Blackstone

Water Pollution, Abatement Dist., Nos. 08-11 et al., 2010 EPA App.

LEXIS 17 (EAB May 28, 2010).           After thorough review of the record

materials, the Board considered and addressed each of the parties'

various    objections     to    the   permit's   nitrogen,    phosphorus,       and

aluminum limits.        

found the that the available

science and data concerning both the District's discharge as well

as the quality of the affected waters supported the EPA's judgment

to impose the tighter permit limits on the three chemical elements.

argument that the EPA should have delayed the

permit    until   the   District's       computer   model   was   complete      and

declined to consider some very preliminary outputs from that model


                                       -22-
because the "development and testing of the model ha[d] not been

completed" and were "not utilized in setting [the limits] for this

Permit."16    

the whole, it found the EPA's actions

reasonable and supported by the record.           

the

Board denied further review, the EPA provided the District with

notice of its final permit decision on April 6, 2011.

             On April 29, 2011, the District filed a petition for

review with this court along with an emergency motion for a stay of

the new permit during the pendency of the appeal.                This court

granted the stay that same day as to each permit condition cited by

the District in its motion.        On May 27, 2011, CLF filed a petition

for review of the new permit, and on June 22, 2011, this court

consolidated    CLF's   petition    with    the   District's   petition   for

purposes of briefing and oral argument.

             The court received extensive briefing from the District,

the EPA, CLF, and amici curiae17 on the issues in this case.              The

District and amicus curiae, the City of Worcester, filed additional

briefing on the HSPF water quality model the District was still in

the process of completing.      The District had raised the issue of


     16
          In 2004, the District began the lengthy process of
developing a Hydrological Simulation Program--Fortran (HSPF)
computer model of the Blackstone River watershed.       The model
remained incomplete through the permitting and EAB review process.
     17
          The court acknowledges the assistance provided by the
amici curiae in this case: City of Marlborough, City of Worcester,
and National Association of Clean Water Agencies.


                                     -23-
the then-unfinished model multiple times during the permitting

process and on review before the EAB, arguing that, once completed,

the model's results might justify a material change in the permit's

conditions.       See 

the District could not

estimate when the model would be finished, and the EPA declined to

delay the permit for an indefinite period until the District could

complete its model.      Instead, it instructed the District to file a

permit modification request when the model was complete, and

therein submit the model's results for consideration.                    See 40

C.F.R. §§ 122.62(a)(2), 124.5.            The District represented that it

would file such a request.18

                                      III.

           In     its   petition,   the    District   challenges    the    2008

permit's   effluent      limitations      for   nitrogen,   phosphorus,     and

aluminum. It argues that key parts of the scientific record before

the EPA    were    inadequate   and    unreliable, and      that   the   agency

irrationally based the permit's limitations on this flawed record.



     18
          By the time of briefing in this case, the preliminary
results of the District's model were available. After reviewing
the parties' arguments as well as the permit modification
mechanism, this court issued an order on January 24, 2012,
directing the parties to participate in a court-sponsored Civil
Appeals Management Program (CAMP). CAMP provided the parties with
an opportunity to resolve the issues in this case more quickly and
easily than proceeding with the appeal. Despite their good-faith
efforts to do so, the parties were unable to resolve their
differences and so informed the court on June 12, 2012. The court
received additional filings from the parties on June 20 and 25,
2012.

                                      -24-
It also argues that the EPA acted irrationally in refusing to delay

the permit until the District could complete both its facility

upgrade, then ongoing, and a new water quality model. CLF supports

the   science   in   the   record,   but    takes   issue   with   the   EPA's

interpretation of one report, arguing that a proper analysis of the

report requires a more stringent nitrogen limitation.

           The formulation of the 2008 permit's effluent limitations

for the three chemical elements at issue required substantial

scientific and technical expertise.            Our review of the EPA's

decision is deferential.       See 33 U.S.C. § 1369(b)(1)(F); City of

Pittsfield, Mass. v. EPA, 

, 10 (1st Cir. 2010).                Under

the Administrative Procedure Act, we ask whether the EPA's actions

were "arbitrary, capricious, an abuse of discretion, or otherwise

not in accordance with law."         5 U.S.C. § 706(2)(A).

           We will not set aside those actions unless the agency

"has relied on factors which Congress has not intended it to

consider, entirely failed to consider an important aspect of the

problem, offered an explanation for its decision that runs counter

to the evidence before the agency, or is so implausible that it

could not be ascribed to a difference in view or the product of

agency expertise."     Motor Vehicle Mfrs. Ass'n v. State Farm Mut.

Auto. Ins. Co., 

, 43 (1983).        We will "uphold a decision

of less than ideal clarity" where it finds support in the record

and has a rational basis.       FCC v. Fox Television Stations, Inc.,


                                     -25-


, 513-14 (2009) (quoting Bowman Transp., Inc. v.

Ark.–Best Freight Sys., Inc., 

, 286 (1974)) (internal

quotation mark omitted); Adams v. EPA, 

, 49 (1st Cir.

1994).

          This deference goes to the entire agency action, which

here includes both the EPA's permitting decision and the EAB's

review   and    affirmance    of   that        decision.     See   33    U.S.C.

§ 1369(b)(1); see also 5 U.S.C. §§ 551(13), 704.

          Our    scope   of   review      is    further    modulated    by   the

scientific and technical nature of the EPA's decisionmaking here.

; P.R. Aqueduct & Sewer Auth. v. EPA, 

, 604 (1st Cir. 1994); see also Balt. Gas & Elec. Co. v. Natural

Res. Def. Council, Inc., 

, 103 (1983) ("[A] reviewing

court must remember that [where the agency] is making predictions,

within its area of special expertise, at the frontiers of science

. . . . as opposed to simple findings of fact, a reviewing court

must generally be at its most deferential."); Coal. for Responsible

Regulation, Inc. v. EPA, Nos. 09-1322 et al., 

, at

*7 (D.C. Cir. June 26, 2012) (to be published in F.3d) ("[W]e give

an extreme degree of deference to the agency when it is evaluating

scientific data within its technical expertise." (quoting Am. Farm

Bureau Fed'n v. EPA, 

, 519 (D.C. Cir. 2009)) (internal

quotation marks omitted)).




                                   -26-
            We also defer to the EPA's reasonable interpretation of

the CWA.     Fed. Express Corp. v. Holowecki, 

, 397

(2008).    This deference increases where the EPA interprets its own

regulations, 

; generally speaking, the agency's

interpretation will be "controlling unless 'plainly erroneous or

inconsistent with the regulation,'" Auer v. Robbins, 

,

461 (1997) (quoting Robertson v. Methow Valley Citizens Council,

, 359 (1989)).

            Below we consider the petitioners' respective arguments,

in turn, and affirm the EPA's decision as to each.

A.          The EPA Did Not Commit Error by Issuing the 2008 Permit
            Without Waiting for Additional Information

            The District argues that the EPA should have waited to

issue the 2008 permit until after the District could complete both

its facility upgrades and its new computer model of the Blackstone

River.     Instead of waiting for the "latest and best data," the

District argues, the EPA "rush[ed] to issue the permit" in a

"mechanical desire to reach a rapid conclusion without regard to

whether the result is sound."    Br. of Pet'r Upper Blackstone Water

Pollution Abatement District, at 22 (quoting P.R. Sun Oil Co. v.

EPA, 

, 79 (1st Cir. 1993)) (internal quotation marks

omitted).    The District's arguments here fail.

            The District first argues that the EPA should have waited

to reissue the new permit until after the District had fully

implemented its facility upgrades to comply with the 2001 permit

                                 -27-
and 2002 administrative consent order.       These upgrades reduced the

District's nitrogen and phosphorus discharge down from prior levels

to 8.0 - 10.0 mg/L and 0.75 mg/L, respectively, and the District

argues that the EPA should have first assessed any water quality

gains from these reductions before tightening the limits further

still.

            The 2002 consent order was issued under 33 U.S.C. § 1319

for the District's violation of the 2001 permit and established a

schedule for construction of new facilities designed to achieve

compliance with the 2001 permit.          As sometimes happens when a

permit requires new construction, the compliance schedule extended

beyond the five-year expiration date of the actual permit.           In its

own words, the order provided "a schedule for compliance that the

Director . . . has determined to be reasonable."

            The order did not purport to alter the EPA's duties under

the CWA to review and reissue permits every five years.             See 33

U.S.C. § 1342(a)(3), (b)(1)(B) (permits "are for fixed terms not

exceeding five years").      EPA regulations provide that no permit's

term may be extended beyond this five-year statutory deadline

except where administratively continued by the EPA during the

permit reissuance process.      See 40 C.F.R. § 122.46(a)-(b) ("NPDES

permits shall be effective for a fixed term not to exceed 5 years

. . . . Except as provided in § 122.6, the term of a permit shall

not   be   extended   by   modification   beyond   the   maximum   duration


                                   -28-
specified in this section."); 

conditions of

an expired permit continue in force under 5 U.S.C. § 558(c) until

the effective date of a new permit . . . ."); see also Natural Res.

Def. Council, Inc. v. EPA, 

, 212-14 (D.C. Cir. 1988)

(per curiam) (upholding continuation provision). Here, neither the

CWA nor EPA regulations allow the District's requested delay.

            In addition, the record reflects that in formulating the

2008    permit    limits,    the    EPA   found   that   even    with    the    fully

completed facility upgrades, the District's discharge would still

"cause, have the reasonable potential to cause, or contribute to"

a     violation     of      water    quality      standards.19          40     C.F.R.

§ 122.44(d)(1)(i).          As to nitrogen, the EPA determined that "a

seasonal reduction to no more than 5.0 mg/l is required . . . to

achieve water quality standards," and that "[t]here is no realistic

likelihood . . . that water quality standards could be met with a

less stringent nitrogen limit."              As to phosphorus, the EPA found

that the 2001 permit's 0.75 mg/L limit is "inadequate for ensuring

the    water      quality     standards      related     to     the   control      of

eutrophication."


       19
           The argument has been raised in the briefing that water
quality in the Blackstone River would be improved by modifying or
removing the many industrial-era dams to allow the River's flow to
move faster. No evidence has been presented to the EPA as to this
proposal, so we may not address it.      Moreover, it may be that
since the build-up of toxic sediments behind the dams poses a
serious toxic risk to water quality, this proposal may exacerbate
the River's water quality problems and so should be studied
carefully.

                                          -29-
             As to the District's computer model, neither the CWA nor

EPA regulations permit the EPA to delay issuance of a new permit

indefinitely until better science can be developed, even where

there is some uncertainty in the existing data. The five-year term

limit requires the EPA or state permitting authority to re-ensure

compliance with the Act whenever a permit expires and is renewed.

33 U.S.C. § 1342(a)(3), (b)(1)(B); 40 C.F.R. § 122.46(a), (b).

Thus, in regular intervals, the Act requires reevaluation of the

relevant factors, and allows for the tightening of discharge

conditions.     The Act's goal of "eliminat[ing]" the discharge of

pollutants by 1985 underscores the importance of making progress on

the available data.     33 U.S.C. § 1251(a)(1).

             In this case, the District overstates the availability of

its data during the 2008 permit process.          Although it was working

on a computer model during the permitting process, the District did

not present any data from the unfinished model during the sixty-

four-day public comment period, and could not provide an estimated

date for the model's completion.       Indeed, it was uncertain during

permitting whether the District would be able to successfully

complete the model at all. Multiple previous attempts by state and

federal actors to develop similar computer models had failed,

leading RIDEM's experts to conclude that the Blackstone watershed

"was   too   complicated   to   simulate   with   available   mathematical

models."     The EPA took into account this prior experience as well


                                   -30-
as the uncertainty surrounding the District's efforts to develop

the model when it declined to delay issuance of the permit until

some indefinite point in the future.

            The EPA also concluded based on the extensive scientific

record before it that even with the District's completed computer

model, there was "no reasonable likelihood that a less stringent

limit   will    meet       [state   water     quality]     standards."     This

determination is entitled to deference.

            The EPA's decision entailed not only an evaluation of the

sufficiency of the available scientific record, but also a risk

analysis of the consequences of waiting.             Nitrogen-based cultural

eutrophication becomes more difficult to address the longer it is

left unchecked.      Nitrogen loadings accumulate and persist in water

systems in a way that can exacerbate future water quality problems.

The EPA found that both the severity of the existing water quality

problems,    and     the   potential   for     aggravated    future   problems,

"counsel[ed] in favor of imposing a nitrogen limit . . . based on

information currently available."             This type of risk assessment is

within the EPA's policymaking discretion, and its judgment here is

entitled to respect.        See Ethyl Corp. v. EPA, 

, 28 (D.C.

Cir. 1976) (en banc).

            The District argues that delay was especially warranted

here because the existing science was old and unreliable, and the

District's     new    model   could    offer    superior    information.    We


                                       -31-
addressed   and rejected   a   similar   challenge in   Sur    Contra   La

Contaminación v. EPA, 

(1st Cir. 2000).           A community

organization, SURCCo, had challenged the EPA's issuance of a

Prevention of Significant Deterioration permit under the Clean Air

Act as arbitrary and capricious, on the grounds that the analysis

relied upon by the EPA was faulty and the EPA should have required

an alternative analysis to be conducted.         We rejected SURCCo's

argument, finding it was rational for the EPA "to prefer its own

model, [and] to reject SURCCo's proposed alternative modeling."

also rejected SURCCo's argument that the EPA had

"relied on outdated -- and perhaps incorrect" data, and that the

agency instead should have "relied on more recent data," available

either from the state or from an analysis yet to be conducted when

the permit was issued.   

credited the EPA's response

that it had "no reason to question the continuing validity" of the

data on which it relied.   

omitted).20




     20
          The District also relies on this court's decision in
Puerto Rico Sun Oil Co. v. EPA, 

(1st Cir. 1993), for this
point. However, at issue in that case was the EPA's rebuff of a
state environmental quality board's request to delay the issuance
of an NPDES permit until it could assess a possible mistake it had
made in certifying the permit under 33 U.S.C. § 1341. However, the
EPA ignored the state board's request and issued a final permit
"with no explanation for its refusal to 

. This
court held that the EPA's failure to provide any justification for
its actions was arbitrary and capricious.     

this
instance, the EPA articulated its reasons for proceeding with
issuance of the permit.

                                 -32-
           In almost every case, more data can be collected, models

further calibrated to match real world conditions; the hope or

anticipation   that    better   science   will    materialize   is   always

present, to some degree, in the context of science-based agency

decisionmaking. Congress was aware of this when it nonetheless set

a firm deadline for issuing new permits.

           As in many science-based policymaking contexts, under the

CWA the EPA is required to exercise its judgment even in the face

of some scientific uncertainty.      The Supreme Court has recognized

this dimension of EPA decisionmaking in the context of the Clean

Air Act.   In Massachusetts v. EPA, 

(2007), the Court

held that the EPA cannot "avoid its statutory obligation by noting

the [presence of] uncertainty."           

"scientific

uncertainty is so profound that it precludes EPA from making a

reasoned judgment . . . EPA must say so.         That EPA would prefer not

to regulate greenhouse gases because of some residual uncertainty

. . . is irrelevant.    The statutory question is whether sufficient

information exists to make an endangerment finding." Id.; see also

Miami-Dade County v. EPA, 

, 1065 (11th Cir. 2008)

(holding that the "EPA is compelled to exercise its judgment in the

face of scientific uncertainty unless that uncertainty is so

profound that it precludes any reasoned judgment"); Ethyl 

("[R]ecognizing . . . the developing nature of [the

field] . . . . [t]he [EPA] Administrator may apply his expertise to


                                  -33-
draw conclusions from suspected, but not completely substantiated,

relationships         between      facts,    from       trends     among       facts,    from

theoretical      projections         from    imperfect          data,    from     probative

preliminary data not yet certifiable as 'fact,' and the like.").

The EPA did not act irrationally here by issuing the permit in the

face of some scientific uncertainty.

           Both       the    CWA    and     EPA    regulations          provide    for    the

incorporation of new information into a permit once it has issued.

The   District    will       have   multiple        opportunities         to    submit   new

information      to    the   EPA    during        the    lengthy    permit       compliance

process, which entails a period of close collaboration between the

state, permittee, and EPA.                The EPA noted during the permitting

process that it intended "to establish a reasonable schedule for

[the District] to come into compliance with the new nutrient

limits."   It has now drafted a compliance schedule, which provides

the District with more than five years to implement the upgrades.

A full twenty-one months are allocated to the District on the front

end of this schedule for conducting testing and investigation into

what measures should be implemented to comply with the permit's

conditions.       The District is required to submit its plans for

implementing the new measures to the EPA one year later, and

commence   construction         five      months        after   that.       The    EPA    has

explained that "it may be appropriate to allow some period of time

to operate the new plant before making a final decision on all


                                            -34-
aspects of additional treatment facilities to enable [the District]

and   its    consultants   to   determine   the   most   cost-effective

technologies for achieving the new limits."

            In addition to this schedule for compliance, the CWA and

EPA regulations provide procedures for the modification of issued

permits where, in the EPA's view, change is warranted.       33 U.S.C.

§ 1342(a)(4), (b)(1)(C); 40 C.F.R. §§ 122.62(a)(2), 124.5. The EPA

has stated that "if the model being developed for [the District],

together with any other relevant evidence, makes it clear that

alternative limits will result in attainment of water quality

standards, EPA will modify the permit accordingly."       The District

has already submitted a permit modification request to the EPA

based on its computer model and additional measurements it has

conducted, and the EPA may consider that request in the normal

course.     The modification request is not before us, but both that

procedure and the compliance process are relevant to our evaluation

of the District's more extreme claims that it is being harmed by

the EPA's decision not to delay the 2008 permit.

            The EPA did not act arbitrarily here in deciding to issue

the permit when it did.

B.          The EPA Did Not Act Arbitrarily in Setting the 2008
            Permit Limits

            1.      Nitrogen Limit

            Both the District and CLF challenge the 5.0 mg/L seasonal

nitrogen limit included in the 2008 permit.        The District argues

                                  -35-
that the EPA arbitrarily selected this limit based on an unreliable

scientific model and without making the touchstone finding, under

33 U.S.C. § 1311(b)(1)(C), that the limit is "necessary to meet

water     quality   standards."        CLF    argues    that    the   EPA   drew    an

unreasonable inference from the scientific record and, as a result,

set a limit that is too lenient.             We affirm the EPA's decision.

             a.      The District's Challenges

             The District first attacks a scientific model the EPA

incorporated into its analysis of the nitrogen-fueled cultural

eutrophication in Narragansett Bay.            The District argues that this

model, which was created by the University of Rhode Island's Marine

Ecosystems Research Laboratory (MERL) in the 1980s to simulate

water     quality   conditions    in    the    Bay,21   is     so   unreliable     and

unrepresentative of actual Bay conditions as to entirely undermine

the EPA's nitrogen analysis.

             The MERL model was peer-reviewed and published in a

scientific journal. Oviatt et al., Patterns of Productivity During



     21
          MERL developed a physical, enrichment gradient model of
the Bay through a series of tank experiments, which were designed
to simulate the impact of nutrient loadings in lower Narragansett
Bay and measure certain effects from these loadings, including
dissolved oxygen impairments and chlorophyll a production.      The
experiment used multiple large tanks, each designed to model the
temperature, mixing, turnover, and light conditions of Narragansett
Bay; several tanks were used as controls, with conditions "similar
to a relatively clean Northeast estuary with no major sewage
inputs." Nutrients were added to the remaining tanks at varying
levels designed to reflect a range of sewage-discharge scenarios in
the Bay.

                                       -36-
Eutrophication: A Mesocosm Experiment, 28 Marine Ecol. Progress

Series 69 (1986).          It has been used by the EPA, RIDEM, and other

groups     to    better    understand    the    causal    relationship    between

nitrogen loadings and cultural eutrophication in Narragansett Bay.

The EPA recently used the model in developing national guidance for

nutrient reduction in water systems.22

                RIDEM used the model to set nitrogen effluent limitations

for   Rhode      Island    sewage   treatment    plants    situated   along   the

Blackstone River and Narragansett Bay.              As part of this process,

RIDEM first conducted an extensive water quality study of the

Blackstone, Seekonk, and Providence Rivers and the Bay, sampling

and testing waters at various sites over a two-year period.                    It

then compared the results of this study with the outputs of the

MERL model, and concluded that both the study's results and the

model showed that higher nitrogen levels led to increased cultural

eutrophication       and    a   less   stable   system    in   general.     RIDEM

published these results, as well as its rationale for imposing new

nitrogen limits on Rhode Island sewage treatment facilities, in a

2004 Report.        RIDEM, Evaluation of Nitrogen Targets and WWTF Load

Reductions for the Providence and Seekonk Rivers (2004) ("RIDEM

Report").



      22
          See EPA, Nutrient Criteria, Technical Guidance Manual:
Estuarine and Coastal Waters, at 2-11, 2-16 (2001), available at
http://water.epa.gov/scitech/swguidance/standards/criteria/nutrie
nts/marine/index.cfm (last visited Aug. 2, 2012).

                                        -37-
           The EPA consulted both the RIDEM Report and the Nutrient

Criteria Technical Guidance Manual in setting the nitrogen limit

for the District's 2008 permit.         The record demonstrates that the

EPA carefully analyzed the MERL model during the permitting process

and   compared   its   results   with    the   outputs   of   water   quality

measurements taken from the three rivers and the Bay.                 The EPA

found that "[b]oth the MERL tank experiments and the data from the

Providence/Seekonk River system confirm a clear correlation between

nitrogen loadings, dissolved oxygen impairment, and chlorophyll a

levels" in those water bodies.          Both the MERL model and the field

measurements     demonstrated    that    as nitrogen     loadings increase,

dissolved oxygen decreases and chlorophyll a increases, with both

becoming less stable and subject to greater swings at higher levels

of nitrogen.     The EPA concluded that the basic causal relationship

demonstrated in the MERL experiments "corresponds to what is

actually occurring in the Providence/Seekonk River system."

           All of the parties agree that MERL's physical model did

not perfectly capture Bay conditions.          The EPA recognized that the

model's flushing rate was lower than the Bay's natural flushing

rate, which may have caused the model to overestimate the impacts

of nitrogen loadings, while the model's mixing rate was higher than

that in the Bay, which may, on the other hand, have caused the

model to underestimate the impacts of nitrogen loadings. The model

did not generate the precise maximum level of nitrogen loading at


                                    -38-
which Rhode Island's water quality standards would be maintained.

It did, however, generate a range of nitrogen loading scenarios

which the EPA used in calculating the numeric limit in the permit.

            The District argues that the discrepancies between the

Bay's   actual   conditions   and   the    conditions   under   which   the

experiments were conducted are so great as to render the model

wholly unreliable.      It argues that based on these "material[]"

differences, and based on the fact that the model cannot predict

the level of nitrogen control needed to meet state standards, it

was irrational for the EPA to consider the model in setting the

2008 permit's nitrogen limit.

           The EPA responds that the model provided one source of

"useful information" in a multi-factored analysis and that to the

extent it did rely on the model, it fully accounted for the model's

shortcomings, and ultimately selected a nitrogen limit based on a

less stringent nitrogen loading scenario than the model, considered

alone, would warrant.

           Our task is not to engage in a "de novo" evaluation of

what scientific evidence was before the EPA, but to look instead

for whether the EPA engaged in the proper decisionmaking process,

and whether its decision finds support in the record.               Motor

Vehicle Mfrs. 

; Kennecott v. EPA, 

, 449 (4th Cir. 1985) (Wilkinson, J.) ("The court best acts as

a check on agency decisionmaking by scrutinizing process and by


                                    -39-
determining whether 'the decision was based on a consideration of

the relevant factors and whether there has been a clear error of

judgment.'" (quoting Citizens to Preserve Overton Park, Inc. v.

Volpe, 

, 416 (1971))).

          Where the agency follows the proper procedures and acts

with a reasonable basis, both its choice of scientific data and

interpretation   and   application   of   that   data   to   real   world

conditions are entitled to deference. Sur Contra La 

; P.R. Aqueduct & Sewer 

;

see also Coal. for Responsible Regulation, 

, at *7.

          Although the District singles out the EPA's reliance on

the MERL model, the EPA used many sources of information in

formulating the nitrogen limits, including both Massachusetts and

Rhode Island reports on nitrogen loadings in the Bay, water quality

studies evaluating nitrogen levels and response variables in the

Bay, and national nitrogen guidance.23     One significant source of



     23
          The EPA consulted multiple sources, including: Deacutis
et al., Hypoxia in the Upper Half of Narragansett Bay, RI, During
August 2001 and 2002, 13 Ne. Naturalist (Special Issue 4) 173
(2006); RIDEM, Plan for Managing Nutrient Loadings to Rhode Island
Waters (2005); Governor's Narragansett Bay Watershed Planing
Comm'n, Nutrient and Bacteria Pollution Panel Initial Report
(2004); RIDEM, Evaluation of Nitrogen Targets and WWTF Load
Reductions for the Providence and Seekonk Rivers (2004); Howes et
al., Massachusetts Estuaries Project: Site-Specific Nitrogen
Thresholds for Southeastern Massachusetts Embayments: Critical
Indicators, Interim Report (2003) (submitted to MassDEP). The EPA
argues, and the record reflects, that the MERL model provided one
source of "useful information" in an analysis that incorporated a
wide variety of information.

                                -40-
information the EPA examined was Rhode Island's own in-state limits

for nitrogen discharge into the relevant waters.         As noted above,

Rhode Island has imposed nitrogen limits equivalent to or stricter

than the District's 5.0 mg/L limit on similarly situated sewage

treatment facilities discharging into the three rivers and the Bay.

           Where the EPA did rely on the MERL model, the record

reflects that it fully accounted for the model's shortcomings. The

EPA delved into "the assumptions and methodology used in preparing

the model," Sierra Club v. Costle, 

, 333 (D.C. Cir.

1981), and, far from ignoring the differences between the model and

real world conditions, it highlighted and responded to these during

the permitting process.       Specifically, the EPA found that the

difference in flushing rates tended to overestimate the effects of

nutrient loadings, while the different stratification levels tended

to underestimate those effects.      It did not blindly follow RIDEM's

conclusion, based on the model, that a 3.0 mg/L nitrogen limit

might be needed, but, "conscious[] of the limits of its model," 

EPA "chos[e] a nitrogen limit based on a less stringent

loading scenario."    See Am. Coke & Coal Chems. Inst. v. EPA, 

, 943 (D.C. Cir. 2006) (holding that EPA's use of the

challenged model was not arbitrary or capricious in part based on

EPA's   extensive   efforts   to   compare   model   assumptions   against

real-world data).




                                   -41-
            The District's argument that the MERL model should have

been excluded from consideration entirely is without merit.              The

EPA is not limited to models which perfectly replicate real world

conditions.   A model does not have to precisely predict the actual

or an average future to increase understanding of a particular

process or the role that different elements play in that process.

The District's objection that the MERL model does not predict the

level of nitrogen control needed misstates and misunderstands the

different roles that scientific models may play in informing

science-based decisions.        Here, the EPA states, and the record

reflects, that the MERL model demonstrated the relationship between

nitrogen loading, dissolved oxygen, and chlorophyll a production

for a range of loading scenarios in a water environment similar to

the Bay's.

            The EPA also followed the proper procedures for ensuring

that the model received scrutiny not only from the permittee, but

from the scientific community and the public.          The EPA highlighted

the model's potential shortcomings in the draft permit documents it

published     for     public    comment.         Numerous   stakeholders,

organizations, and individuals submitted support for and criticism

of the model.       In its detailed and extensive responses to these

comments,    the    EPA   carefully   reviewed   and   responded   to   each

criticism raised.         The EAB further reviewed the EPA Region's

analysis of the model, and found no reason to fault that analysis.


                                      -42-
"[A]dmission of uncertainties where they exist," "public exposure

of the assumptions and data incorporated into the analysis," "the

acceptance and consideration of public comment," and, ultimately,

a decision that reflects the rule of reason, are the structural

features of reasoned, publicly accountable science-based agency

decisionmaking.       Sierra 

& n.130; see also

Nat'l Mar. Safety Ass'n v. Occupational Safety & Health Admin., 

, 752 (D.C. Cir. 2011), cert. denied, 

(2012).    The EPA incorporated these structural safeguards into its

decisionmaking process.

               The EPA's determination, based on its analysis of the

evidence before it as a whole, that a nitrogen limit of 5.0 mg/L

was necessary to achieve Rhode Island's water quality standards was

not   a   "hunch[]   or    wild   guess[]"   but   a   rational    exercise   of

judgment.      Ethyl 

.

               The District's second challenge to the 2008 permit's

nitrogen limit is that the EPA failed to prove that the limit is

either "necessary" or "sufficient" to attain Rhode Island water

quality standards.         The CWA requires the EPA to impose certain

types     of   discharge   limitations   on    point    source    dischargers,

including publicly owned sewage treatment facilities, such as the

District's,      "including   those   necessary    to   meet     water   quality

standards."      33 U.S.C. § 1311(b)(1)(C).        The District argues that

the EPA failed to make specific findings either that the nitrogen


                                      -43-
limit   is   "necessary"   to    achieve   Rhode    Island    water    quality

standards or that it will "in fact" do so.

             We reject the first claim, since the EPA expressly found

that the 5.0 mg/L limit was necessary to meet state standards, and

that a higher limit would not achieve those standards.              In the Fact

Sheet that accompanied the original draft permit, the EPA found

"[b]ased on the available evidence, including nitrogen loadings

from [the District] and the discharge of the Blackstone River to

the Seekonk River, where the greatest impacts have been measured,

. . . [the] seasonal reduction of nitrogen to no more than 5.0 mg/l

is required at [the District's facility] in order to achieve water

quality standards."        The   EPA   reiterated   that     this    limit    was

"necessary" to attain water quality standards at multiple other

points in the draft permit and during the permitting process.

             As to the second objection, the District argues that the

EPA never found the nitrogen limit "sufficient" to attain water

quality standards and that a still lower effluent limit may be

needed.      The EPA noted in the draft permit's Fact Sheet the

possibility     that   further   monitoring    "will    demonstrate          that

additional pollutant reductions are ultimately needed to meet water

quality standards."     This review and potential tightening of the

conditions in NPDES permits is a basic feature of the CWA that the




                                   -44-
District does not dispute.24               See 33 U.S.C. §§ 1251(a), 1313,

1342(b).

               The District's argument seems to go to the precision of

the permit's nitrogen limit.              But where a complex administrative

statute, like those the EPA is charged with administering, requires

an agency to set a numerical standard, courts will not overturn the

agency's choice of a precise figure where it falls within a "zone

of reasonableness."         See, e.g., Nat'l Mar. Safety 

; Solite Corp. v. EPA, 

, 488 (D.C. Cir. 1991)

(per curiam) (judicial deference is warranted where EPA chooses "a

numerical      standard .     .    .   within    a   'zone   of   reasonableness'"

(omission in original) (quoting Small Refiner Lead Phase-Down Task

Force     v.   EPA,   

,    525   (D.C.   Cir.     1983)   (internal

quotation marks omitted))); 

(EPA's

"conclusions with respect to data and analysis need only fall

within a 'zone of reasonableness'" (quoting Reynolds Metals Co. v.

EPA, 

, 559 (4th Cir. 1985)) (internal quotation mark

omitted)); Hercules, Inc. v. EPA, 

, 117 (D.C. Cir. 1978)

(holding that within the zone of reasonableness, "the choice of a


     24
          The District's "sufficiency" argument is actually
directed toward a very different point. The District argues that
in addition to determining whether specific effluent limits will
attain state water quality standards, the EPA must make a companion
determination that the state standards are in fact "attainable,"
and if not, the EPA should work with the state to revise the
standard "to reflect what can be attained." However, the District
does not argue that Rhode Island's water quality standards here are
unattainable, so we reject this argument.

                                          -45-
precise figure is left to EPA").       The nitrogen limit the EPA chose

here    is   justified   by   the   record    and   within   the   zone   of

reasonableness.25    The District's challenges to the limit fail.

             b.     CLF's Challenges

             CLF does not challenge the EPA's reliance on either the

RIDEM Report or the MERL model, but attacks the inferences the

agency drew from these sources.            CLF argues that because RIDEM


       25
          The District makes a cursory argument that the EAB
applied an excessively deferential standard in its review of the
2008 permit's nitrogen and phosphorus limits. Since 1980, EPA's
regulations have provided that the EAB's review of NPDES permitting
decisions at the regional level shall be deferential to the EPA
Regions' determinations, see 40 C.F.R. § 124.19(a); see also 45
Fed. Reg. 33,290, 33,412 (May 19, 1980), particularly where these
involve science-based and technical judgments, see In re: NE Hub
Partners, L.P., 7 E.A.D. 561, 567-68 (EAB 1998), review denied sub
nom. Penn Fuel Gas, Inc. v. EPA, 

(3d Cir. 1999). This
deference is not unbounded, however; in its review of petitions,
the Board carefully examines the permit decisionmaking process and
the full record.    See, e.g., In re: City of Marlborough, Mass.
Easterly Wastewater Treatment Plant, 12 E.A.D. 235, 248-52 (EAB
2005) (remanding permit because "the Region [had] not sufficiently
explained where or how [the compliance finding] is reflected in the
record").
          In this case, the EAB exhaustively reviewed the EPA
Region's permitting decision in a thorough and exacting 106-page
opinion. The EAB carefully addressed each of the arguments of the
parties to this appeal, as well as those of seven other entities,
including the states of Massachusetts and Rhode Island.         The
Board's opinion, upholding the permit in part, and remanding to the
EPA for further proceedings in part, reviewed the analysis and
methodology employed by the EPA Region in full. The Board's review
of the permit decision on this record was reasonable.        To the
extent that amicus curiae, City of Marlborough, makes additional
broader arguments about the EAB's standard of review in this case,
amicus is not a party and we do not engage those arguments.
Downing/Salt Pond Partners, L.P. v. Rhode Island, 

, 28
(1st Cir. 2011), cert. denied 

(2011).



                                    -46-
determined in its Report that a 5.0 mg/L nitrogen limit "would not

be acceptable as [a] water quality goal[] for the area," and that

instead, a limit of at least 3.0 mg/L was necessary to ensure

compliance with Rhode Island water quality standards, the EPA acted

irrationally choosing a 5.0 mg/L limit over a more stringent limit.

            The EPA responds, and the record reflects, that RIDEM

noted in its Report that "some uncertainty remains regarding

predicted    water     quality   improvements    and    loading    reductions

necessary to meet water quality standards. . . . For these reasons,

evaluation    of     phased   implementation    is   indicated."      RIDEM,

Evaluation of Nitrogen Targets and WWTF Load Reductions for the

Providence and Seekonk Rivers at 27 (2004).            RIDEM, in fact, set a

limit of 5.0 mg/L for two sewage treatment facilities located along

the Providence River, both of which are comparable in size to the

District.    The EPA took these factors, and many others, including

the additional studies and data we have referenced above, into

account in setting the nitrogen limit.

            CLF's argument that the EPA should have interpreted

RIDEM's Report to require a 3.0 mg/L limit amounts to an attack on

the EPA's interpretation and application of the scientific data

before it to real world conditions.        We give the EPA substantial

deference in this area. See Coal. for Responsible Regulation, 

, at *7; 

; P.R. Aqueduct & Sewer

; 

.            Here, the EPA


                                    -47-
independently analyzed the model and data utilized in RIDEM's

Report and reasonably concluded that certain aspects of each

warranted a slightly higher limit than the Report recommended.

This decision is entitled to deference.

          CLF makes an additional argument that the EPA made the

impermissible assumption in setting the nitrogen limit that the

District's discharge will remain below its design flow of 56 mgd.

See 40 C.F.R. § 122.45(b)(1) ("POTW[] [publicly owned treatment

work] effluent limitations . . . shall be calculated based on

design flow.").   However, CLF has waived this argument by failing

to present it either to the EPA Region during the permitting

process or during the initial round of briefing before the EAB.   40

C.F.R. § 124.13 ("All persons . . . who believe any condition of a

draft permit is inappropriate . . . must raise all reasonably

ascertainable issues and submit all reasonably available arguments

supporting their position by the close of the public comment period

. . . .").   As a result, the EPA did not have the opportunity to

assess or respond to CLF's objection on the record.

          "Simple fairness to those who are engaged in the tasks of

administration, and to litigants, requires as a general rule that

courts should not topple over administrative decisions unless the

administrative body not only has erred but has erred against

objection made at the time appropriate under its practice." United

States v. L. A. Tucker Truck Lines, Inc., 

, 37 (1952);


                               -48-
see also Pepperell Assocs. v. EPA, 

, 27 (1st Cir. 2001).

The waiver rule serves particularly important purposes in the

administrative review context, both in that it accords respect to

the agency decisionmaking process by providing the agency with the

"opportunity to address a party's objections, . . . apply its

expertise, exercise its informed discretion, and create a more

finely tuned record for judicial review," and in doing so, guards

against a system in which regulated parties "simply turn to the

courts as a tribunal of first resort."        Mass. Dep't of Pub. Welfare

v. Sec'y of Agric., 

, 523-24 (1st Cir. 1993). Because

CLF failed to abide by this rule, its argument is waived.

             2.     Phosphorus Limit

             The District objects to the 2008 permit's imposition of

a 0.1 - 1.0 mg/L seasonal limit on its phosphorus discharge.                In

the way of brief background to this challenge, the District's 2001

permit limited phosphorus discharge to 0.75 mg/L in order to

address      low   dissolved   oxygen       levels,    but     not    cultural

eutrophication, in the Blackstone River.         Around the time the 2001

permit issued, the EPA was in the process of studying nutrient-

related issues more closely in water systems across the country.

In   2001,   the   EPA   published   a   national     action   plan   for   the

development and establishment of numeric nutrient criteria as well




                                     -49-
as recommended numeric criteria for most water systems.26                   In

conjunction with these ongoing efforts, the EPA specifically noted

during the 2001 permitting process that more stringent phosphorus

limits might be necessary in future permits to address cultural

eutrophication impacts in the Blackstone River.

             Subsequently, in order to address the severe and ongoing

phosphorus-driven cultural eutrophication in the Blackstone River,

the EPA incorporated a more stringent phosphorus limit into the

2008 permit.       In formulating this limit, the EPA considered the

national and regional guidance criteria and recommended values it

had recently published.        See, e.g., Buck et al., Office of Water,

EPA,    Nutrient   Criteria    Technical     Guidance   Manual:   Rivers   and

Streams     (2000);   Office   of   Water,   EPA,   Ambient   Water   Quality

Criteria Recommendations: Information Supporting the Development of

State and Tribal Nutrient Criteria: Rivers and Streams in Ecoregion

XIV (2000) (guidance document on river watersheds in eastern

coastal states, including Massachusetts); Barbour et al., Office of

Water, EPA, Rapid Bioassessment Protocols For Use in Streams and

Wadeable Rivers: Periphyton, Benthic Macroinvertebrates, and Fish

(2d ed. 1999).        It also looked at older studies, see Office of

Water Regulations & Standards, Quality Criteria for Water (1986)


       26
          See Grubbs, Office of Sci. & Tech., Development and
Adoption of Nutrient Criteria into Water Quality Standards (2001),
available at, http://water.epa.gov/scitech/swguidance/standards/
criteria/nutrients/upload/2009_01_21_criteria_nutrient_nutrientsw
qsmemo.pdf (last visited Aug. 2, 2012).

                                     -50-
(the "Goldbook"), and considered site-specific data from studies

conducted after the 2001 permit had issued.

             The District argues that the 2008 permit's phosphorus

limit is arbitrary because the EPA considered national guidance on

phosphorus reduction, and other regional and area studies which,

the District argues, have no proven connection to the Blackstone

River.

             The EPA did not act irrationally by considering its

national and regional phosphorus guidance criteria in addition to

site-specific data. The guidance documents helped inform the EPA's

background understanding of phosphorus-driven eutrophication and

recommended "a range of ambient phosphorus concentrations that

[would be] sufficiently low to prevent cultural eutrophication" in

river    systems    similar   to    the   Blackstone.        See   40    C.F.R.

§ 122.44(d)(1)(vi).

             Of the documents the EPA considered which recommended

specific numeric phosphorus limits, the Nutrient Criteria Technical

Guidance Manual recommends an in-stream phosphorus concentration of

0.01     -   0.09   mg/L,     the   Ambient     Water   Quality         Criteria

Recommendations      for    Ecoregion     XIV   recommends    an    in-stream

concentration of 0.024 mg/L, and the Goldbook recommends an in-

stream concentration of 0.05 mg/L for any stream entering a lake or

reservoir, and 0.1 mg/L for any stream not discharging into an

impounded waterbody (the EPA noted that the Blackstone River is


                                     -51-
characterized by multiple impoundments). The EPA did not blindly

follow      any    of    these   recommended        limits,   but    after   examining

additional        site-specific        data,    including     local    water      quality

studies, selected a phosphorus limit designed to ensure an in-

stream concentration of 0.1 mg/L.27

              The EPA also analyzed various site-specific phosphorus

load data produced after 2001, including studies conducted by

MassDEP, EPA New England, and the U.S. Army Corps of Engineers.28

To    account      for    the    District's      phosphorus      treatment     upgrade,

implemented as part of the 2001 permit and 2002 consent order, the

EPA also examined data collected when the District's phosphorus

discharge was comparable to what it would be with the upgrade.                       The

EPA examined data collected by MassDEP under low flow conditions in

August of 2003, when the District's average monthly discharge was

0.8   mg/L,       very   close    to   the     2001   permit's      0.75   mg/L   limit.

See Fiorentino, Div. of Watershed Mgmt., Mass. Dep't Envtl. Prot.,


       27
          In this instance, the EPA determined that a monthly
average total phosphorus limit of 0.1 mg/L was necessary from April
1 through October 31 in order to ensure an in-stream concentration
of not more than 0.1 mg/L due to the lack of any significant
dilution in the River's waters downstream during these months.
       28
          See Fiorentino, Div. of Watershed Mgmt., Mass. Dep't of
Envtl. Prot., Blackstone River Watershed 2003 Biological Assessment
(2006); Tamul, Div. of Watershed Mgmt., Mass. Dep't of Envtl.
Prot., Blackstone River Watershed 2003 DWM Water Quality Monitoring
Data (2005); Wright et al., Dry Weather Water Quality Sampling and
Modeling, Blackstone River Feasibility Study (2004) (for U.S. Army
Corps of Engineers); Wright et al., Blackstone River Initiative,
Water Quality Analysis of the Blackstone River Under Wet and Dry
weather Conditions (2001) (for EPA New England).

                                             -52-
Blackstone     River    Watershed     2003    Biological     Assessment (2006).

MassDEP nonetheless observed a "luxuriant algal community" and

measured levels of phytoplankton which were "extremely abundant,

covering virtually the entire river bottom."                 The EPA reasonably

determined, and the record reflects, that the 2001 permit's 0.75

mg/L phosphorus limit would thus be insufficient to reduce cultural

eutrophication and bring the River into compliance with state water

quality standards.

             To the extent the District challenges the precision of

the 2008 permit's numeric limit, we have already recognized that

the EPA's choice of a precise numeric value will be affirmed where

it is within the zone of reasonableness.               See Nat'l Mar. Safety

.        The permit's phosphorus limit is within

this zone of reasonableness.

             The District also alleges that the EPA was required to

demonstrate     both    that   the     phosphorus      limit    "would    have    a

substantial impact on the cultural eutrophication of the Blackstone

River"   and     that     it   will     alleviate      not     merely     cultural

eutrophication but "a specific impairment in designated uses."                   In

other words, the District argues that any effluent limitation

imposed upon it must cure (or nearly so) the water quality problem.

             The CWA quickly disposes of these arguments.                The Act's

TMDL   and   interim    planning     process    both   contemplate       pollution

control where multiple point sources cause or contribute to water


                                       -53-
quality standard violations.        33 U.S.C. § 1313(d), (e).            Under

earlier legislation, including the 1965 Federal Water Pollution

Control Act, when a water body failed to meet its state-designated

water quality standards, pollution limits could not be strengthened

against   any   one   polluter   unless    it   could   be   shown   that   the

polluter's discharge had caused the violation of quality standards.

See EPA v. California ex rel. State Water Res. Control Bd., 

, 202-03 (1976).         This standard was ill-suited to the

multifarious nature of modern water pollution and prevented the

imposition of effective controls.         

declared

that the system was "inadequate in every vital aspect," and had

left the country's waterways "severely polluted" and "unfit for

most purposes."       S. Rep. No. 92-414, at 3674 (1971).             The CWA

rejected the earlier approach and, among other things, introduced

individual pollution discharge limits for all point sources.                 33

U.S.C. 1311(b). To maintain state water quality standards, the Act

establishes the TMDL and continuing planning processes, which

target pollution from multiple sources.          

           EPA regulations require permitting authorities to include

in NPDES permits conditions which "control all pollutants or

pollutant parameters . . . [that] are or may be discharged at a

level which will cause, have the reasonable potential to cause, or

contribute to an excursion above any State water quality standard,

including State narrative criteria for water quality."               40 C.F.R.


                                   -54-
§ 122.44(d)(1)(i); see also 54 Fed. Reg. 23,868, 23,873 (June 2,

1989).   We thus reject the notion that in order to strengthen the

District's discharge limits, the EPA must show that the new limits,

in and of themselves, will cure any water quality problems.

           3.      Aluminum Limit

           Finally, the District challenges the limit placed on

aluminum discharge, arguing that the EPA assembled and then relied

upon an erroneous data set in deriving the limit.            Specifically,

the District argues that the EPA should have excluded an "outlier

data point of 344 µg/L" in calculating the District's average daily

aluminum discharge.     The EPA responds that the atypical data point

was properly included in its analysis of in-stream concentrations

of   aluminum   since   the   District   did   not   demonstrate   that the

conditions which led to the high discharge will not reoccur, and

that, in any case, the District has waived its argument.

           We find that the District has waived the argument by

failing to raise it during the public comment period of the

permitting process.      See 40 C.F.R. § 124.13; L. A. Tucker Truck

; Pepperell 

; Mass.

Dep't of Pub. 

.            In its February 27,

2009, comments on the draft modification for aluminum discharge,

the District objected that the EPA "used and relied upon incomplete

and incorrect data and as a result reached incorrect conclusions."

However, this comment was directed at another argument, that the


                                   -55-
EPA had failed to include data from certain years, and did not

state or imply that the EPA should have excluded the 344 µg/L data

point.     By failing to give the EPA an opportunity to address the

argument during the permitting process, the District has waived its

claim.

                                   IV.

            The   District's   responsibility     for    serious   pollution

problems in the important waterways of two states is clear, and its

challenge to the 2008 permit has no merit.              As the District has

recognized, cost considerations may not be considered by the EPA in

the setting of permit limits to assure compliance with state water

quality    standards.     33   U.S.C.    §§   1311(b)(1)(C),    1342(a)(2);

Defenders of Wildlife v. Browner, 

, 1163 (9th Cir.

1999); U.S. Steel Corp. v. Train, 

, 838 (7th Cir.

1977).     We trust that the District, as well as the EPA, will now

act with expedition to address these problems.

            The District and CLF's petitions are denied.           The stay

granted by this court on April 29, 2011, is lifted.            No costs are

awarded.

            So ordered.




                                   -56-