-------

-------
                              TABLE OF CONTENTS
EXECUTIVE SUMMARY	       i
CHAPTER ONE:  INTRODUCTION	;	       1
    Legislative History	        2
    General Approach and Assumptions in the Mater Quality
      Improvement Study	        5
    Report Organization	       12
CHAPTER TWO:  METHODOLOGY	      13
    Overview	      13
    Information Sources	      16
    Description of Methods to Assess Improvements	      23


CHAPTER THREE:  RESULTS	      29
    Improvements in Water Quality -  Water Quality  Model	      29
    Improvements in Water Quality -  Ambient  Monitoring  Data	      35
    Improvements in Water Quality -  Case Studies	      39
CHAPTER FOUR:  CONCLUSIONS	     49


CHAPTER FIVE:  REFERENCES	     54

-------
                               LIST  OF FIGURES
NUMBER                                 TITLE                              PAGE

 2-1   Summary of Facilities Screening  Process  	      20
 3-1   Summary of Water Quality Modeling:   Overall  Results	      34
 3-2   Summary of Ambient Monitoring  Data Analysis:  Overall Trends
         for BAT Reaches 	      38
 3-3   Trends in Metal  Concentrations in Oysters  -  Long  Island Sound.      41
 3-4   Trends in Metal  Concentrations for the South Fork Coeur d'Alene
         River	      43

-------
                                LIST  OF TABLES
NUMBER                                 TITLE                               PAGE


  1    Major Assumptions and Limitations	       vi

 1-1   Promulgation History of BAT Regulations	        4

 1-2   Comparison of RAW and BAT Pollutant Loadings for BAT
         Industries	        7

 2-1   Frequency of Occurrence of Selected Priority Pollutants in
         BAT Industrial Category Wastewater Discharges	       15

 2-2   Summary of Screening Process Used to Select BAT Facilities
         and Reaches for Inclusion in the Water Quality Improvement
         Study	       19

 2-3   EPA Ambient Water Quality Criteria Used in the Water Quality
         Improvement Study	       22

 3-1   Summary of Water Quality Modeling Results:  Compliance with
         Criteria (At Low Stream Flow)	       31

 3-2   Summary of Water Quality Modeling Results:  Compliance with
         Criteri a (At Average Stream Fl ow)	       32

 3-3   Summary of Ambient Monitoring Data Analysis:   Pollutant
          Trends	       37

 3-4   Water Quality Model  Results for the Delaware Estuary	       46

 3-5   Ambient Water Quality Monitoring  Data Summary for Delaware
         Estuary	       48

-------
                                  APPENDICES
Volume II contains the technical appendices,  providing the backup material and
outputs from the water quality model  and ambient data evaluation, and is
available under a separate cover.
    Appendix  A - Descriptions of EPA Data Bases and Files Used in the Water
                  Quality Improvement Study

    Appendix  B - RAW and BAT Pollutant Concentrations by Industrial  Category

    Appendix  C
    Appendix  D


    Appendix  E


    Appendix  F


    Appendix  G
EPA Ambient Water Quality Criteria for States
(For Heavy Metals with Hardness-Based Criteria)

Reach-by-Reach Listing of BAT Industrial  Facilities
by State

Results of Water Quality Modeling of BAT  Industrial
Dischargers by State (At Low Stream Flow)

Results of Water Quality Modeling of BAT  Industrial
Dischargers by State (At Average Stream Flow)

Results of Ambient Water Quality Monitoring Data Analysis for
BAT Reaches by State

-------
                                                              Executive Summary
            Report to Congress on the Water Quality Improvement Study

                              EXECUTIVE SUMMARY

    The  objective of this report is to identify improvements in water quality
 attributable to the application of best available technology economically
 achievable  (BAT).  This study is required by Section 308(g) of the Water
 Quality  Act of 1987.  Section 308(g) also requires an evaluation  of the water
 quality  program (including Section 302(a) site-specific water quality
 determinations) and recommendations of methods for improving such program.
 This study does not include an evaluation of non-BAT aspects of the water
 quality  program since this aspect is now the subject of a major concurrent
 effort mandated under Section 308(a) of the 1987 Water Quality Act [Section
 304(1) of the Clean Water Act].

    This Report to Congress presents the results and general conclusions of
 this study on the effectiveness of BAT effluent limitations for controlling
 pollutant discharges from industrial sources and the resulting improvements  in
 chemical and biological quality of streams and rivers receiving discharges
 from these industrial sources.  This study does not address the benefits of
 controls other than the national categorical effluent guidelines,  such  as
 "best professional judgment" or water quality-based controls implemented at
 the individual  permit level, pretreatment requirements,  nonpoint  source
 controls, and treatment requirements for publicly owned treatment  works.  Nor
 does this study address the non-water quality impacts that have resulted from
 improved wastewater treatment (e.g., generation of solid wastes or increased
 air pollutant emissions due to volatization or stripping).

                    LEGISLATIVE HISTORY  AND BACKGROUND

    The Clean Water Act establishes as a national  goal  the restoration  and
maintenance of "the chemical,  physical,  and biological  integrity of the

-------
                                                              Executive Summary
Nation's waters."  As a primary means of achieving this goal, Congress
prohibited "the discharge of any pollutant by any person" into the waters of
the United States unless that discharge complies with the specific
requirements of the Act Section 301 (a), 33 U.S.C. 1311(a).  Compliance may be
achieved by obtaining a permit issued pursuant to Section 402 of the Act.
Permits must incorporate applicable technology-based effluent limitations
guidelines promulgated by EPA on a nationwide, industry-by-industry basis
under Sections 301(b) and 304, and if EPA has not promulgated guidelines,
limitations are applied in individual permits on a case by case basis (Section
302).  Effluent limitations are implemented in two stages.  By July 1, 1977,
dischargers to surface waters were required to achieve effluent limitations
applicable to best practicable control technology currently available (BPT).
For toxic pollutants and nonconventional pollutants, they are required, by
1984, to achieve effluent limitations applicable to best available technology
economically achievable (BAT).  Separate requirements were established for
industrial facilities that discharged wastewaters to publicly owned treatment
works (POTWs) and for new sources.

    The development of categorical standards, including BPT and BAT
regulations, has spanned a period of 16 years.  Roughly half of the original
BPT regulations promulgated in the mid-1970s have remained unchanged, despite
additional regulatory activity for industrial categories.  For some of these
categories, the BPT effluent limitations have equaled BAT.  Therefore, the
technology basis for limitations for some industrial categories has been
unchanged since the mid-1970s; for others,  however, the statutory deadlines
for compliance with BAT limitations have not yet passed.

    "Technology-based" limitations are used by permitting authorities to set
minimum pollution control  requirements for all dischargers other than publicly
owned treatment works.  In some instances,  however, additional  controls may be
necessary to meet water quality standards.   Section 304(1) of the Clean Water
                                       11

-------
                                                              Executive Summary


Act  (added by the 1987 Amendments) requires States to identify waterbodies
that are impaired by toxic discharges from point sources and develop
strategies to control these discharges.  States were to identify waterbodies
and develop individual control strategies (ICSs) by February 4, 1989 and must
achieve applicable water quality standards on or before June 4, 1992.  EPA
approved or disapproved the ICSs on June 4, 1989, and where EPA disapproved an
ICS, the revised ICS must achieve applicable water quality standards on or
before June 4, 1993.

                                 METHODOLOGY

    The limitations required under EPA's effluent guideline regulations have
resulted in a significant reduction in the amounts of toxic, conventional, and
nonconventional pollutants that are discharged to the nation's waters.  While
these reductions in themselves have improved water quality, the extent of such
improvements has not been examined closely.  This study attempts to determine
the extent of these improvements, on a nationwide scale, through the
examination and use of currently existing EPA data sources.  However, this
study does not attempt to quantify benefits which result from improved water
quality.  These include increased sport and commercial  fishing, and improved
recreational opportunities.

    Three methods, or components, were used to identify and/or project the
extent of improvements in water quality that can be attributed to the
implementation of BAT effluent regulations.  These are:   (1) water quality
modeling of pollutant discharges by industries regulated by the BAT
limitations; (2) analysis of ambient water quality monitoring data for stream
segments receiving discharges from BAT facilities; and  (3)  summarization of
water quality improvement case studies relevant to the  BAT regulations.  For
the purpose of this study, a water quality improvement,  as it pertains to the
first two components, was defined by the Agency as:

    •  Complying with EPA national  ambient water quality criteria after the
       implementation of BAT, having exceeded (not complied with) water
       quality criteria prior to BAT.
                                       111

-------
                                                              Executive Summary
    •  A decrease  in  in-stream pollutant concentration after implementation
       of BAT.
    For the first  component of the study, modeling was performed to project
 in-stream pollutant concentrations of ten selected pollutants (six heavy
 metals, cyanide, and  three organic chemicals) for comparison with EPA ambient
 water quality criteria.  These pollutants were selected because they are the
 most common toxic  chemicals present in the wastewater discharges of BAT
 facilities.  Although national water quality criteria are guidance and may be
 more or less stringent than applicable State water quality standards, national
 water quality criteria were selected as the basis for comparison for two
 reasons.  First, national water quality criteria provide a nationally
 consistent measure for comparison and second, they account for the fact that
 many States have not yet adopted all necessary numerical water quality
 standards for toxic pollutants.

    Only direct discharging industrial facilities in those categories having
 BAT regulations that  specifically regulate toxic pollutants (i.e., those
 toxic, or priority, pollutants identified in the Clean Water Act of 1977) were
 included.  Indirect dischargers (i.e., those that discharge to POTWs) were not
 evaluated in the model.  The model focused on determining the impacts of these
 discharges on the unique stream segments (or reaches) that receive the
wastewater discharges.  No effort was made to determine downstream or
 cumulative effects.  The model also evaluated only streams and rivers; lakes,
 estuaries, and shorelines were not included because of the difficulty involved
 in determining the dilution characteristics.

    Two treatment levels were used to project the benefits of the categorical
 limitations:  (1) raw, or untreated, discharge levels; and (2)  levels
 representing compliance with BAT limitations.  Industry-wide pollutant
concentrations representing long-term average discharge levels were used.

    In the second component of the study, ambient water quality monitoring data
were analyzed for those reaches receiving discharges from the modeled BAT
facilities.   Since the methods available for analyzing these monitoring data

                                       iv

-------
                                                              Executive Summary
are best suited for determining relative differences in concentrations over
specific time frames, this evaluation focused on showing trends in the
in-stream levels for the ten selected pollutants between two time periods
selected to represent pre- and post-BAT conditions.  The major assumptions and
limitations that pertain to the first two components of this study are
summarized in Table 1.

    The last component of this study involved the presentation of actual case
studies showing that the implementation of BAT resulted in improvements in the
chemical and biological quality of the receiving waters.  Most water quality
improvement studies have focused on conventional pollutants discharged from
municipal facilities, so few studies that target toxic pollutants from
industrial discharges exist.  Several studies, however, are pertinent to the
BAT regulations and resultant improvements in water quality, and these are
summarized below.

    A special case study was developed by the Agency to evaluate the
improvements resulting from BAT on a typical estuary.  This study used a water
quality model to project theoretical improvements, which were verified through
the review of ambient monitoring data.

                            RESULTS  OF THE STUDY

    The results of the three methods indicate that water quality improvements
have occurred as a result of the application of BAT.  The model  evaluated
2,490 individual BAT facilities, discharging to 1,546 unique reaches (totaling
24,289 river miles).  On a pollutant-by-pollutant basis, pre-BAT compliance
with water quality criteria ranged from 43 percent (copper) to 97 percent
(toluene) in the river miles assessed under low flow conditions.   Compliance
with criteria after BAT increased by an average of 20 percent for each
pollutant.  Improvements under average flow conditions are less  pronounced;
compliance rates ranged from 75 (cyanide) to 100 (phenol,  toluene,  and
benzene) percent prior to BAT and from 90 (mercury) to 100 (nickel,  phenol,
toluene, and benzene) percent after BAT.

-------
                                                       Executive Summary
                             TABLE 1
              MAJOR ASSUMPTIONS  AND LIMITATIONS

Study is limited to only the reaches where BAT industries are located
because of the uncertainty in modeling the fate and transport of
pollutants without extensive site-specific information.  Consequently,
some reaches may have been excluded that benefited from BAT.

Analysis focused only on direct dischargers.  Indirect dischargers
were not included.  By not including indirect dischargers, BAT industry
plant population was reduced and potential water quality improvements
underestimated.

Projection of BAT benefits is limited to the ten most frequently
regulated and monitored pollutants instead of all priority pollutants
to facilitate data collection and analysis.  This approach was
considered adequate to determine trends in water quality improvements
while maintaining a manageable data base.   However, it also excludes
potential BAT benefits for overlooked pollutants.

Water quality model used untreated treatment levels to represent
industry's pre-BAT discharge levels due to the lack of representative
data on treatment in place prior to the BAT.  This approach appears
reasonable because only limited toxics controls existed in 1970s.   But
some BAT benefits may have been overestimated by not considering
existing treatment in-place.  The BAT treatment levels were used to
represent post-BAT conditions.

Water quality model used industry-wide average pollutant
concentrations for BAT industrial categories to represent individual
facility discharge levels.  This assumes every facility in a particular
industry discharges the same pollutants at the same concentrations.

The time period from 1970 to 1980 was selected to represent pre-BAT
conditions; from 1985 to 1988 to represent post-BAT conditions
(intervening years assumed as transition)  based on the BAT promulgation
dates (with an assumed 2-year implementation period) in our ambient
water quality analysis.  Consequently,  the STORET monitoring data
available for analysis were considerably reduced in scope.   Also,  full
benefit of BAT promulgated in the mid-1980s is not yet fully reflected
in monitoring data.

Ambient monitoring data was collected for  only those reaches
receiving discharges from BAT facilities.   The location of the
individual  monitoring stations in relation to the discharge(s) was not
readily available.

National water quality criteria were used  rather than individual
State standards because many States do not yet have numerical  standards
for many toxic pollutants, and EPA criteria provide a consistent basis
for nationwide in-stream pollutant concentration comparison.  Existing
State criteria may be more or less stringent than EPA criteria.

                                vi

-------
                                                              Executive Summary
    Prior to the implementation of BAT, only 29 percent of the assessed river
miles were projected to comply with all the criteria for the ten selected
pollutants under low stream flow conditions (58 percent of the river miles
were projected to comply under average flow conditions).  After industries
meet the discharge requirements of BAT, an additional 29 percent of the
assessed river miles are projected to comply with criteria (at both average
and low flow conditions).  Individually, the pollutants most associated with
noncompliance are mercury, copper, and lead.  After the implementation of BAT,
this study projects that 42 percent of the river miles assessed may not comply
with criteria for one or more of the modeled pollutants under low flow (12
percent at average receiving stream flow).  This projection,  however, is not a
precise measure of BAT derived improvements, since it is dependent on the
number of pollutants evaluated.  These river miles not complying with criteria
may require further water quality-based toxic controls depending on State
water quality requirements, including site-specific water quality criteria.

    Ambient monitoring data were available for both time periods evaluated for
at least one of the selected inorganic pollutants for 429 (8,434 river miles)
of the 1,546 reaches evaluated by the model.  Limited ambient data were
available for the three selected organic pollutants (toluene, benzene, and
phenol) but unfortunately, none of the reaches had monitoring data for both
time periods for any of the organics.  As a consequence, this monitoring data
analysis was reduced in scope, as compared to the discharge model.

    On an individual pollutant basis, the ambient monitoring  data analysis
found reductions in in-stream concentrations for each of the  seven monitored
pollutants.  These reductions ranged in scope from 69 percent of the monitored
river miles for zinc to 87 percent of the monitored river miles for mercury.
In terms of overall  pollutant reduction trends (considering all  monitored
pollutants),  improvements were shown for 76 percent (6,397 miles)  of the 8,434
river miles assessed.

-------
                                                              Executive Summary


    However, ambient monitoring data also indicated that for 14 percent (1,147
miles) of the assessed river miles, overall pollutant concentrations increased
between the two time periods.  Eleven percent of the river miles (890) were
found not to have a significant difference between the pre- and post-BAT
average concentrations.  Possible explanations for this deterioration or lack
of progress include:  variability in stream conditions; discharges from new
sources; changes in upstream pollutant inputs; pre-BAT treatment may have been
equivalent to BAT (thus not changing the overall performance); or sources that
either met required limitation prior to ambient sampling or have not yet
attained these limitations.

    The six case studies included in this study present various types of water
quality improvements that can be attributed to BAT.  Three show reduced heavy
metals concentrations, resulting from BAT-type limitations, in the receiving
stream and the return of less pollutant-tolerant aquatic biota.  Another case
study presents data on oysters which shows a decrease in metal concentrations
between samples collected in the 1970s and those collected in the 1980s (BAT
controls on metal industries were implemented on upstream sources in the
mid-1970s).  Two studies show that reduced pollutant discharges from pulp and
paper mills improved oxygen conditions, causing a return of indigenous fish
species.

    In addition, the special case study verified that toxic pollutant levels
in the Delaware estuary have been reduced, at least partially as a result of
BAT.

                         CONCLUSIONS OF  THE STUDY
    The three components of this study (e.g.,  water quality modeling,
monitoring data evaluation and case study reviews) indicate that water quality
has improved significantly as a result of the  implementation of the BAT
effluent limitation regulations.
                                      viii

-------
                                                              Executive Summary


    The water quality model projects that during average stream flow
conditions, compliance with WQC for each of the evaluated pollutants under BAT
is 90 percent or higher.  For four of the evaluated pollutants, compliance is
100 percent.  In addition, 88 percent of all stream miles are projected to
meet WQC for all 10 evaluated pollutants, compared to 59 percent prior to BAT.

    During low stream flow conditions, compliance with individual pollutants
ranges from 73 to 100 percent.  The percentage of waters meeting WQC for all
10 pollutants is projected to be 58 percent with BAT, compared to 29 percent
prior to BAT.

    Under BAT, 12 percent of all river miles are projected to exceed WQC for
one or more pollutants during normal stream flow; 42 percent of all river
miles may exceed one or more WQC during low stream flow.  Depending on
individual States WQC, additional water quality controls for some pollutants
may be necessary to bring those reaches completely into compliance.

    Ambient monitoring data analysis also shows significant improvements, in
terms of pollutant reductions trends.  Seventy-six percent of river miles in
the data base showed an overall decrease (improvement) in pollutant
concentrations after BAT.  At the same time, 14 percent of monitored river
miles were shown to increase in concentrations from pre-BAT to post-BAT
period.  The overall reductions in pollutant levels, however, reflect not only
the benefits of BAT controls but also the benefits from water quality-based
limitations, municipal treatment improvements, controls on non-BAT industries
and nonpoint sources, and, to an unknown extent, lower detection limits
between the two time periods.

    The full benefit of the implementation of the BAT regulations is not yet
reflected in the ambient monitoring analysis.  National categorical standards
have recently been promulgated for one major industry, organic chemicals
manufacturing, which will reduce the industry's toxic pollutant direct
                                       IX

-------
                                                              Executive Summary
discharge loadings by an estimated 1.1 million pounds per year.  Standards
have yet to be promulgated for another industrial category  (pesticides
manufacturing).  While the ambient data do reflect some of  the benefits
attributable to BAT, the full effect will not be evident until the early 1990s,

    Actual cases where the implementation of BAT has resulted in water quality
improvements present the best illustration of the effectiveness of the
categorical standards; however, few such cases are available.  Those studies
that do exist have shown that the BAT regulations have had  a positive impact
on water quality of the receiving stream, both in terms of  lower chemical
concentrations and increased biological activity.  In most  instances, the
streams/rivers assessed in these case studies were highly polluted prior to
BAT, and even after the discharge levels had been reduced to BAT levels,
additional water quality-based controls were and will continue to be
necessary.  In all cases, the implementation of BAT has resulted in
considerable improvements in the chemical and/or biological quality (as
measured by an increase in less pollution-tolerant aquatic  life) of the
receiving waters.

    The site-specific modeling of the Delaware River estuary predicts BAT will
achieve compliance with water quality criteria for the 10 modeled pollutants.
Ambient monitoring data also show a decreasing pollutant concentrations
trend.  However, full benefits of BAT are not yet reflected in monitoring data
because of the number of organic chemical facilities on this estuary.  Organic
chemical BAT guidelines were recently (11/87) promulgated,  consequently many
of these facilities may not yet be operating at BAT levels.

    Considering the results of each component of this analysis, together with
their respective assumptions and limitations, the BAT regulations have been an
effective step toward improving the quality of our nation's surface waters.
The extent of their effectiveness is difficult to assess, however,  given the
limits of existing EPA data sources and the fact that BAT controls have yet to

-------
                                                              Executive Summary
be universally applied.  Also, by not considering improvements resulting from
the categorical pretreatment requirements, the extent of the improvements
resulting from the overall national water quality program are underestimated.
There may be a need for additional water quality-based controls beyond BAT in
some cases to meet State water quality standards.  In addition, as required by
Section 304(m) of the Water Quality Act of 1987, the Agency will establish a
schedule for:  (1) the annual review and revision of promulgated effluent
guidelines, and (2) the promulgation of regulations for industrial categories
identified as sources of toxic and nonconventional pollutants for which
guidelines have not previously been established.
                                       XI

-------
                                                                   Introduction
                                   Chapter  One

                                 INTRODUCTION

    This report presents the findings of the Water Quality Improvement Study,
conducted by the U.S. Environmental Protection Agency (EPA) as required by the
amendments to the Federal Water Pollution Control Act (Clean Water Act).
These amendments, known as the Water Quality Act of 1987 (PL 100-4), included
several requirements for studies of the Nation's water quality and the
regulations controlling pollutant discharges.  Congressional intent was to
determine the state of the Nation's water quality and to review the
effectiveness of regulatory programs protecting or improving that water
quality.  Section 308 of the Water Quality Act requires the Administrator of
the EPA to report to Congress on the effectiveness of the water quality
improvement program, especially the application of best available technology
(BAT) effluent limitations to control pollutant discharges from industrial
sources.  This report addresses the requirement for a Water Quality
Improvement Study.

    The specific areas to be addressed by the Water Quality Improvement Study,
as specified in Section 308(g), are:

    "The Administrator shall study the water quality improvements which have
    been achieved by the application of best available technology economically
    achievable pursuant to Section 301(b)(2) of the Federal Water Pollution
    Control  Act.  Such study shall include, but not be limited to,  an analysis
    of the effectiveness of the application of best available technology
    economically achievable (BAT) pursuant  to such section in attaining
    applicable water quality standards (including the standard specified in
    Section 302(a) of such Act) and an analysis of the effectiveness of the
    water quality program under such Act and methods of improving such
    program,  including site specific levels of treatment which will  achieve
    the water quality goals of such Act.
                                        -1-

-------
                                                                   Introduction
    Not later than 2 years after the date of the enactment of this Act, the
    Administrator shall submit a report on the results of the study ...
    together with recommendations for improving the water quality program and
    its effectiveness to the Committee on Public Works and Transportation of
    the House of Representatives and the Committee on Environment and Public
    Works of the Senate."
    EPA's Office of Water was directed by the Administrator to prepare this
Report to Congress.
                             LEGISLATIVE HISTORY

    The Clean Water Act established a comprehensive program to "restore and
maintain the chemical, physical, and biological integrity of the Nation's
water."   Under this program, industrial dischargers were required to achieve
"effluent limitations requiring the application of the best practicable
control technology currently available" (BPT) by July 1, 1977, and "effluent
limitations requiring the application of the best available technology
economically achievable (BAT) ... which will result in reasonable further
progress toward the national goal of eliminating the discharge of all
pollutants," by July 1, 1984 (under Sections 301(b)(l)(A) and 301(b)(2)(A),
respectively).  New industrial  direct dischargers were required to comply with
Section 306 new source performance standards (NSPS) based on best available
demonstrated technology; existing and new dischargers to publicly owned
treatment works (POTW) were subject to pretreatment standards under
Sections 307(b) and (c), respectively, of the Act.

    The requirements for direct dischargers are incorporated into the National
Pollutant Discharge Elimination System (NPDES) permits issued under
Section 402 of the Act.  Permits must incorporate applicable technology-based
effluent limitations guidelines promulgated by EPA on a nationwide,
industry-by-industry basis under Sections 301(b) and 304, and if EPA has not
promulgated guidelines, limitations are applied in individual permits on a
                                        -2-

-------
                                                                   Introduction
case by case basis  (Section 302).  Effluent limitations are implemented in two
stages.  By July 1, 1977, dischargers to surface waters were required to
achieve effluent limitations applicable to best practicable control technology
currently available (BPT).  For toxic pollutants and nonconventional
pollutants, they were required, by 1984, to achieve effluent limitations
applicable to best available technology economically achievable (BAT).
Separate requirements were established for industrial facilities that
discharged wastewaters to publicly owned treatment works (POTWs) and for new
sources.  Pretreatment standards were made enforceable directly against
dischargers to a POTW (indirect dischargers).

    As a result of the Clean Water Act, BAT effluent limitation guidelines
have been promulgated for 24 industrial categories (for one category,
pesticides manufacturing, the BAT guidelines are currently being developed).
The promulgation of the effluent limitation guidelines for these industries
has occurred in two "phases."  The first "phase" was prior to the passage of
the Federal Water Pollution Control Act (PL 95-217), commonly referred to as
the Clean Water Act of 1977, with the limitations focusing mostly on
conventional pollutants.  Many of the BAT limitations promulgated at this time
were withdrawn or rescinded, while some of the BPT limitations remained in
effect.  The second "phase" regulations were promulgated after the 1977
amendments and reflect additional controls on toxic pollutants.  Most of the
second "phase" regulations were promulgated in the early 1980s (Table 1-1).
However, these regulations for one major industry (organic chemicals
manufacturing) were not promulgated until  1987, and one regulations for the
pesticide manufacturing category are presently being prepared.

    The effluent limitations specified under Section 301(b)(l)(A), (B) and (2)
of the Act are "technology-based" limitations.  These limitations are used by
permitting authorities (EPA or the States) to set minimum pollution control
requirements for all industrial dischargers without regard to the quality of
                                         -3-

-------
                                                                   Introduction
                                    Table 1-1
                     Promulgation History of BAT Regulations
           Industry Name
Promulgation Dates for BAT Regulations
     Phase I              Phase II
Aluminum Forming
Battery Manufacturing
Coal Mining
Coil Coating
Copper Forming
Electrical & Electronic Components
Foundries
Inorganic Chemicals
Iron & Steel
Leather Tanning
Metal Finishing
Nonferrous Metals
Nonferrous Metals Forming
Ore Mining
Organics/P&SF
Pesticides
Petroleum Refining
Pharmaceuticals
Plastics Molding & Forming
Porcelain Enameling
Pulp & Paper
Steam Electric
Textiles
Timber Products
_ *
_ *
04/77
. *
. *
_ *
_ *
03/74
06/74
04/74
03/74
04/74
-
11/75
04/74
11/76
05/74
11/76
-
_ *
05/74
10/74
07/74
04/74
09/83 0
02/84 0
11/82 0
11/82 & 11/83 0
08/83 0
03/83 & 11/83 0
10/85
06/82 & 07/84
05/82 0
11/82 0
07/83 0
02/84 & 08/85 0
07/85
11/82
11/87
WITHDRAWN 12/86
09/82 0
09/83
12/84
11/82 0
10/82 0
11/82 0
08/82
01/81 0
KEY

* These categories represented a portion of the Machinery and Mechanical
  Products category during Phase I (no regulations promulgated).

0 Minor changes (amendments) have been made to these industries after
  promulgation as a result of litigation.
                                         -4-

-------
                                                                   Introduction
receiving waters.  In some cases, however, technology-based standards  are  not
enough.  If water quality does not support the designated use even  after every
discharger meets his technology-based standards,  additional controls must  be
applied.  Section 301(b)(l)(C) of the Act requires "any more stringent
limitations," including those necessary to meet water quality standards,
whenever meeting the technology-based standards in Section 301  fail to attain
or maintain the water quality called for in the river or stream (water
quality-based limitations).  Section 302(a) also  authorizes the Administrator
to establish effluent limitations more stringent  than those necessary  to meet
quality standards in order to meet the water quality goals of the Act.

    In the Water Quality Act of 1987, Congress added new Section 304(1), which
requires States to develop lists of impaired waters,  lists of point sources
and amounts of pollutants causing toxic impacts and "individual  control
strategies" for such point sources.  These new requirements should  aid  in  the
identification of waters that will need water quality-based standards.  These
provisions direct immediate attention to establishing controls  where there are
known impacts due entirely or substantially to point sources of Section 307(a)
toxic pollutants.  The identification of waterbodies and point  sources  and the
development of control strategies is the subject  of a separate  Agency  effort.
The statutory deadline for identification of waterbodies and development of
individual  control  strategies (ICSs) was February 4,  1989,  and  must achieve
applicable water quality standards on or before June 4,  1992.   EPA  approved or
disapproved the ICSs on June 4,  1989, and where EPA disapproved  an  ICS, the
revised ICS must achieve applicable water quality standards on  or before June
4, 1993.

                     GENERAL APPROACH AND  ASSUMPTIONS
                 IN  THE  WATER  QUALITY IMPROVEMENT  STUDY

    The objective of this report is to identify improvements in  water quality
attributable to the application  of best available technology economically
                                        -5-

-------
                                                                   Introduction
achievable  (BAT).  This study is required by Section 308(g) of the Water
Quality Act of 1987.  Section 308(g) also requires an evaluation of the water
quality program  (including Section 302(a) site-specific water quality
determinations)  and recommendations of methods for improving such program.
This study does  not include an evaluation of the water quality program since
it is now the subject of a major concurrent effort mandated under
Section 308(a) of the 1987 Water Quality Act [Section 304(1) of the Clean
Water Act].

    Because of the implementation of BAT by industrial dischargers and the
subsequent reduction in pollutant loadings, it can be inferred that the water
quality in the nation has improved.  As illustrated in Table 1-2, the Agency
estimates that,  the result of treatment including BAT, organic priority
pollutant discharges have been reduced by 99 percent from untreated levels and
that inorganic priority pollutant discharges have been reduced by almost
98 percent.  These estimates of pollutant reductions do not account for
treatment in-place prior to the implementation of BAT, and therefore, may
overestimate BAT benefits.  However, the extent of these pollutant reductions
and their contribution to attaining the goals of the national  water pollution
control  program  have not been determined.  This study attempts to determine
the extent of these improvements on a national  scale, through  the examination
and use of currently existing data sources.  However, this study does not
attempt to quantify benefits which result from improved water  quality.   These
include increased sport and commercial  fishing, and improved recreational
opportunities.

    For purposes of this study,  "water quality improvement" was defined as one
of the following:

    •  Compliance with EPA national water quality criteria (WQC) after BAT,
       having not complied with WQC prior to BAT.
    •  A decrease in in-stream pollutant concentration after implementation
       of BAT.
                                         -6-

-------
                                                   Table 1-2  Comparison of RAU and BAT Pollutant Loadings for BAT Industries
INDUSTRIAL INFORMATION

PROCESS UASTEUATER
NUMBER FLOW (USD)
INDUSTRIAL CATEGORY OF PLANTS RAU BAT
ALUMINUM FORMING 42 10.6 7.3
BATTERY MANUFACTURING 15 1.2 0.1
COAL MINING 10,375 6,417.0 6,417.0
COIL COATING 32 3.3 1.1
COPPER FORMING 37 15.1 2.2
ELECTRICAL COMPONENTS 84 12.8 12.8
FOUNDRIES 301 177.3 7.0
INORGANIC CHEMICALS 149 197.8 166.1
IRON & STEEL 738 5,199.2 1,497.7
LEATHER TANNING 17 4.2 4.2
METAL FINISHING 2,800 365.9 365.9
NONFERROUS METALS 112 55.0 8.7
NONFERROUS METALS FORMING 51 2.1 0.3
ORE MINING 506 1,359.6 1,359.0
ORGANIC CHEMICALS * 301 394.3 394.3
PESTICIDES MANUFACTURING 42 4.6 4.6
PETROLEUM REFINING 164 312.0 312.0
PHARMACEUTICALS ** 52 24.9 24.9
PLASTICS MOLDING & FORMING 810 88.5 88.5
PORCELAIN ENAMELING 28 2.2 1.8
PULP & PAPER 355 3,746.4 3,746.4
TEXTILES 229 177.7 177.7
zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz
BAT INDUSTRIES TOTAL: 17,240 18,571.6 14,599.5
=======r==-===--========-=================r=rz::r=zzzzzzzzzzzzzzzz

TfiTAI SIKPFWHFn Cfll THQ
1 U IRL OUOrtnUCU 9UL1L/3
LOADINGS (Ib/day)

RAU BAT
4,059 727
7,544 11
224,811,488 1,672,004
2,978 113
24,108 223
4,292 1,282
1,151,113 197
7,255,677 60,883
5,141,618 135,470
87,277 1,944
2,059,357 54,348
62,457 199
793 18
178,579,200 95,845
398,349

239,804 67,957
125,693 13,603
63,775 3,593
76,629 179
10,847,704 889,338
171,521 72,842
IZZZZSZZZZZZZZZZZZZZZZZZZZZ
431,115,436 3,070,776
IZZZZ2ZZZZZZZZZZZZZZZZZZZZZ

RTfirUFMTPAl HYVCFN RFMAUn
DiULnuniUnu UA i vtn vtnnni/
LOADINGS (Ib/day)

RAU BAT
—
—
..
—
„
--



53,901 1,173


-- --

245,956

346,817 35,150
237,825 16,573
8,635 2,604
—
8,014,703 541,732
489,032 33,273
9,396,869 630,505
PRIORITY POLLUTANT

ORGANICS

RAU BAT
9

400 133
16
858 11
409 75
2,248 21

105,296 262
303 6
9,343 162
495 3
6
.. ..
483,546 1,097

17,119 103
6,425 232
141 124
„
32,794 3,355
3,656 691
663,064 6,275
LOADINGS (Ib/day)

INORGANICS

RAU BAT
2,943 63
1,109 31
134,373 7,401
968 7
41,813 34
250 152
51,307 64
166,723 12,616
917,027 2,551
3,595 67
240,178 6,555
3,033 132
2,525 2
2,265 2
83,304 794
„
4,077 796
229 81
135 95
--
8,501 8,166
2,358 1,438
1,666,713 41,047
SOURCE:  USEPA (1988).

-------
                                                                   Introduction
    To determine water quality improvements, the approach used had three major
components:   (1) projecting in-stream concentrations of selected pollutants by
modeling pre- and post-BAT effluent discharges of industries regulated by BAT
to determine  theoretical improvements based on compliance with water quality
criteria;  (2) reviewing available ambient water quality monitoring data to
determine  trends in chemical water quality; and (3) providing actual case
studies on the implementation of the national effluent regulations for
industrial dischargers and a discussion of any improvements in either the
chemical or biological quality of specific waters as a result of such
implementation.

    The modeling and ambient data review effort primarily focused on:
(1) facilities discharging directly to surface waters that are required, under
the BAT regulations, to control toxic pollutants; and (2) nationwide
improvements.  Indirect facilities (i.e., those that discharge their
wastewater to POTWs), which are controlled by pretreatment standards, were
excluded for  two reasons:  (1) there was insufficient information on indirect
facilities within the EPA national data bases, and (2) these facilities are
not controlled by State or EPA-issued permits (NPDES), but rather by local
limits developed by the POTW receiving the discharge.  This study also
excludes industries which are not regulated by the national categorical
(technology-based) effluent limitation guidelines, but rather by "best
professional judgment" permits.  By not considering indirect facilities and
improvements  in water quality resulting from pretreatment requirements or
controls on industries without additional categorical effluent limitation
regulations, this study underestimates the extent of the improvements
resulting from the national water quality program.  The effectiveness of the
BAT regulations in improving water quality can be addressed on a national
level  because site-specific and/or regional factors are not taken into account
in establishing these technology-based effluent limitations.  The
site-specific nature of water quality-based limits prevents a theoretical
estimate of their effectiveness.   Specific instances where water quality
improvements are attributable to indirect facilities will  be discussed, when
applicable, in the case studies.
                                         -8-

-------
                                                                   Introduction
    The first two components (or analyses) of this study made use of existing
EPA computerized information to expedite the nationwide study.  Ideally, such
information sources would have historical information available for each of
the industrial facilities affected by the BAT regulations.  Ideally, this
information would include pollutant loading information prior to, and after,
implementation of BAT treatment technologies; the date when BAT treatment was
implemented; and in-stream ambient monitoring data specific to each facility.
Unfortunately, existing data bases were developed primarily for the purpose of
and tracking ambient water quality for localized pollution control purposes
and not specifically for performing a nationwide water quality improvement
study.  Thus, the ideal data base described above is not available.

    Other factors also need to be considered, including the long time frame
over which BAT regulations have been implemented.  In addition, for a number
of years some States have been writing water quality-based permits for
industrial dischargers that are generally more stringent than the treatment
requirements of BAT.  Differentiating between the time periods that represent
pre- and post-BAT and which facilities have effluent limitations based on
technology or water quality is not feasible using these sources of data.

    Several assumptions were necessary in order to complete the first two
components of this study using the available EPA information:

    1.  All industrial facilities covered by the BAT regulations will be
        evaluated assuming that technology-based limitations are the basis for
        their respective discharge levels.
    2.  Pre-BAT effluent discharges will  be represented by untreated, or raw,
        levels as determined by EPA sampling programs.
    3.  The pre-BAT period for evaluating ambient monitoring data is from 1970
        to 1980.
    4.  The post-BAT period is from 1985  to the present.
                                         -9-

-------
                                                                   Introduction
    5.  Ten pollutants were selected to represent toxic pollutant discharges
        from BAT-regulated facilities.  These selected pollutants are the most
        frequently discharged and regulated priority pollutants.
    6.  Only stream segments receiving direct discharges for BAT facilities
        were evaluated.
    One limitation of this study is that it is not currently feasible to
determine which BAT industrial facilities actually have and meet permit
limitations based on the national effluent guideline (technology-based)
regulations.  However, by not considering more stringent controls imposed by
water quality-based permits and assuming that all BAT industries meet the
requirement of the technology-based limitations only, the overall need for the
more stringent controls can be more accurately assessed on a national level.
This is one of the requirements of the water quality improvement study.

    In evaluating the improvements resulting from the BAT regulations, the
ideal scenario would consider discharge levels prior to the implementation of
treatment requirements mandated by the Federal water quality program (i.e.,
pre-BAT), which began in 1972.  Unfortunately, no accurate method exists to
determine or estimate these discharge levels, especially in the area of toxic
pollutant discharges.  In some instances, facilities had treatment of their
wastewater discharges.  In others, untreated wastewaters were discharged
directly to receiving waters.  In addition, EPA's initial  sampling of
industrial facilities in the mid-1970s to define effluent characteristics
focused on conventional pollutants and a few toxic metals.  Only later, in
response to the NRDC Settlement Agreement and subsequent amendments to the
Clean Water Act, did the Agency focus its regulatory efforts on the toxic, or
priority, pollutants referenced in the Act.  For this reason, and because it
is Agency policy when determining benefits of BAT guidelines, untreated
wastewater was used as the basis for determining the improvements.  This
assumption will overestimate the actual benefits of the technology-based
program, but the extent of this overestimate is uncertain.
                                        -10-

-------
                                                                    Introduction
     It was difficult to select appropriate time periods for comparing pre- and
post-BAT ambient water quality monitoring data because BAT treatment levels
for  some industries (such as metal finishing/electroplating, pulp and paper,
and  textiles) have remained essentially the same since the mid-1970s, while
for  others, BAT limitations have yet to be fully implemented or promulgated
(organic chemicals and pesticide manufacturing, respectively).  Taking into
consideration, however, the fact that monitoring for toxic pollutants was not
as common prior to the mid-1970s as it is today and that detection limits for
individual chemical parameters have changed as technology has improved, the
years from 1970 to 1980 were selected to serve as the pre-BAT period.  During
this period some reductions in ambient pollutant concentrations would be
reflected in the monitoring data, but the overall average for that period
should tend to be representative of levels prior to the implementation of
technology-based regulations.  To reflect post-BAT levels, 1985 was selected
as the start date.  A majority of the "Phase II" BAT regulations for the
industrial categories were promulgated in the early 1980s (Table 1-1), so the
period from 1981 to 1984 was considered a transition period and was not
evaluated.

    By using only a select group of toxic pollutants, the overall  trends in
water quality can be identified without the need for voluminous amounts of
data.  However, by not considering improvements resulting from reductions of
other toxic pollutants, the full  benefits of BAT are underestimated.

    Only those stream segments receiving discharges from BAT-regulated
facilities were included in the first two components of this study.   Because
of difficulty in modeling the fate and transport of these pollutants, no
impacts to downstream stream segments were considered,  thus reducing the scope
of potential improvements.

    The first two components of this study project and  estimate improvements
in water quality resulting from the application of BAT.   While these analyses
                                        -11-

-------
                                                                   Introduction
provide a good picture of overall national trends, case studies best present
water quality improvements since such improvements are "real  life," not
projected or estimated.  Case studies reviewed for this study had to show a
direct benefit in water quality (either chemical or biological) as a result of
facilities implementing limitations set forth by BAT.  The focus of such
improvements or benefits was the reduction of toxic pollutants.  However,
other aspects of technology-based limitations, such as reductions in
conventional pollutant discharges for industrial sources were also considered,
as well as the beneficial results of the implementation of pretreatment
program for indirect dischargers to publicly-owned treatment  works.  However,
this study does not attempt to quantify benefits which result from improved
water quality.  These include increased sport and commercial  fishing,  and
improved recreational opportunities.

                             REPORT ORGANIZATION

    The remainder of this report consists of four chapters that address the
objective of Section 308(g) of the Water Quality Act of 1987  relevant  to the
BAT regulations.   Chapter 2 presents a general discussion of  the data  sources
and methodology used to identify water quality improvements.   Chapter  3
summarizes the results of the analyses.   Chapter 4 addresses  the study
findings and the evaluation of the effectiveness of the BAT regulations.
Chapter 5 lists the various references used in this study. Volume II  of this
report contains the technical appendices, providing the backup material  and
outputs from the water quality model and ambient data evaluation.   Volume II
is available under a separate cover.
                                        -12-

-------
                                                                    Methodology
                                   Chapter  Two

                                  METHODOLOGY

                                   OVERVIEW

    Three different components were used to identify water quality improvements
resulting from the application of best available technology economically
achievable (BAT):  (1) projection of theoretical improvements in water quality
for stream reaches with BAT industries (those industries covered by the BAT
regulations) using water quality modeling;  (2)  review of ambient water quality
data for pre- and post-BAT time periods; and (3) summary of specific instances
(case studies) where water quality has improved because of BAT.   Specifically,
this methodology included the following:

    Hater quality modeling of industrial facilities with BAT effluent
limitations to control toxics.  The model predicted in-stream concentrations
of ten selected pollutants at pre-BAT and post-BAT treatment levels at low and
average receiving stream flow conditions.  Concentrations were then compared
to EPA water quality criteria to determine  the  potential/theoretical  water
quality improvements.

    Analysis of STORET ambient water quality monitoring data,  for the same
ten selected pollutants, for stream segments receiving BAT industrial
discharges for time periods assumed to reflect  pre-BAT (1970 to  1980)  and
post-BAT (1985 to 1988) conditions.  The average values for these two time
periods were used to determine trends in ambient pollutant levels.

    Summary of case studies that associate  the  implementation of national
technology-based effluent limitations with  improvements in the chemical  and/or
                                        -13-

-------
                                                                    Hethodology


biological quality of the receiving waters.  All improvements that can be
attributable to BAT regulations are considered, primarily reductions in toxic
pollutant discharges but also including controls on conventional pollutants and
indirect dischargers.  A special case study was developed by the Agency to
evaluate the benefits of BAT regulations in estuaries.

    The ten pollutants selected to be evaluated were representative of toxic
pollutants regulated by BAT and could be considered "indicator" pollutants.
These pollutants are the most frequently discharged and regulated priority
pollutants found in the wastewater of the BAT facilities and include six heavy
metals (cadmium, mercury, copper, lead, nickel, and zinc), cyanide, and three
organic chemicals (phenol, toluene, and benzene).  Table 2-1 lists these
pollutants and their frequency of occurrence in the BAT industries effluent.
Also shown are the industries in which these pollutants are regulated.   In
addition, limiting the number of pollutants also reduced the modeling effort
to a manageable size.

    The modeling and ambient data analyses are limited to only those reaches
receiving BAT industrial discharges.  (Reaches are unique segments of streams
and rivers that have been delineated by EPA for the purpose of integrating
water quality and facility information.)  It was beyond the scope of this study
to address the effects of the discharges on reaches downstream of the
industries, even though the downstream reaches are likely to be directly
affected.

    The methodology used to identify water quality improvements is presented in
the following sections:  (1) Information Sources, which presents a brief
description of all the EPA information sources used for the first two
components of the study; and (2) Description of Methods to Assess Improvements,
which presents general discussions of the components of this study - the water
quality modeling analysis, the ambient water quality data evaluation,  and
site-specific case studies.   The latter section also provides,  for each
                                        -14-

-------
                                                                                                   Methodology
                                                       Table 2-1

                                Frequency of Occurrence of Selected Priority Pollutants
                                    in BAT  Industrial Category Wastewater Discharges
Occurrence of
Industrial Category
Aluminum Forming
Battery Manufacturing
Coil Coating
Copper Forming
Electrical & Electronic Components
Metal Molding & Casting (Foundries)
Inorganic Chemicals
Iron & Steel
Leather Tanning & Finishing
Metal Finishing
Nonferrous Metals
Nonferrous Metals Forming
Ore Mining & Dressing
Organic Chemicals. Plastics. &
Synthetic Fibers
Petroleum Refining
Pharmaceuticals Manufacturing
Porcelain Enameling
Pulp & Paper
Textiles
Total Occurrences
Total Regulated
Cadmium
1
R
1
1
1
1
R
1
1
R
1
R
R


1

1

1
16
5
Mercury

R
1

1
1
R
1
1

1

1


1
1

1
1
13
2
Copper
1
R
R
R
1
R
R
1
1
R
R
R
R

R
1
1
1
1
1
19
10
Lead
1
1
1
R
1
R
R
R
1
R
R
R
R

R
1
1
R
1
1
19
10
Pollutant in
Nickel
1

1
R
1
1
R
R
1
R
R
R
R


1
1
R
1
1
17
8
Zinc
R
R
R
R
1
R
R
R
1
R
R
R
R

R
1
1
R
R
1
19
14
Industrial Category
Cyanide
R
R
R
1
1

R
R
1
R
R
R
1

R
1
R

1
1
17
10
Phenol
1

1

(R)
1

1
1
(R)
l



R
1
1

1
1
13
1(2)
Toluene


(R)
1
(R)
1

1
1
(R)

1


R
1
1

1
1
13
1(3)
Benzene



1

1

R
1
(R)




R
1
1

1
1
10
2(1)
KEY

R - Regulated pollutant in  industrial category.
1 - Detected, but not regulated,  in Mastewater of industrial category.

Values in parentheses denote when pollutant is regulated as part of  "Total Toxic Organics.'

SOURCE:  U.S. EPA,  1986.
                                                    -15-

-------
                                                                    Methodology



component, a summary of the approach and purpose of the analysis, assumptions

used and limitations in performing the analysis, and the actual analysis

conducted.  Outputs from the water quality model and the ambient water quality

monitoring data analyses are included in the technical appendices (Volume II).


                             INFORMATION SOURCES


    To perform the first two components of this study, readily available and

accessible information contained in the following EPA data bases were used:

Permit Compliance System (PCS), Industrial Facilities Discharge (IFD) file,

GAGE file, REACH file, and STORE! Water Quality file.  A brief description of

each data base and how it was used is presented below:


    •  Permits Compliance System (PCS) - PCS is a computerized
       management information system for tracking permit status data for the
       National Pollutant Discharge Elimination System (NPDES).  This system
       was used to identify BAT facilities by SIC code and to determine their
       discharge status (active or inactive).

    •  Industrial Facilities Discharge (IFD) file - The IFD file
       provides a comprehensive data base of industrial and municipal point
       source dischargers,  including discharge flow and location information,
       standard industrial  classification (SIC) codes, and categorization of
       discharge types.  This file was used to locate the BAT facilities,
       identified by PCS, on specific reaches and to provide wastewater
       discharge flows for the water quality model.

    •  GAGE file - The GAGE file stores data from stream gaging
       stations, including:  station location, types and frequency of data
       collected, and stream flow data (mean, annual, and low flow).   The GAGE
       file was used to provide stream flows (both average and low)  for the
       water quality model.

    •  REACH file - The REACH file is a digital data base of streams,
       rivers,  reservoirs,  lakes,  and estuaries in the contiguous U.S.  divided
       into unique segments called "reaches."  The reaches allow EPA (and
       other system users)  to integrate data from different files and data
       bases by assigning unique reach numbers to individual  water body
       segments.  This file identified the type of reach (stream or  nonstream).
                                        -16-

-------
                                                                    Nethodology
    •  STORE! Mater Quality file - STORE! is a data base composed of
       several individual, but related, files, and includes data on:  stream
       flow; physical and chemical characteristics of streams, fish tissue,
       and sediment; municipal waste sources and disposal systems; and
       pollution-caused fish kills.  STORET was used to obtain ambient water
       quality information for the ten selected pollutants.


    All information contained in these data bases was retrieved during

December 1987, with the exception of the STORET data, which was retrieved in

July 1988.


Screening and Evaluation

    Data were retrieved from the EPA data bases and subsequently compiled and

analyzed using a computerized software system.  The files were then screened

to select only those facilities that would be covered by the BAT regulations

and could be evaluated by the water quality model.  All currently active

facilities were also assumed to be active prior to BAT implementation,

therefore neglecting the impacts on water quality for new sources and plant
closings.  The reaches identified as receiving BAT discharges in water quality

model  will serve as the basis of retrieving ambient monitoring data.  The
following is a brief description of the screening process:


    1.  The EPA data bases contain information on 120,992 industrial dischargers
       and roughly 68,000 reaches.  Of these dischargers, 46,467 are inactive
       (e.g., closed) or their discharge status is unknown, and 53,621 reaches
       do not have any assigned dischargers on them.   All inactive/unknown
       facilities and reaches without dischargers were excluded from this
       study.

    2.  Of the 74,525 active dischargers, 59,338 facilities are not covered by
       the BAT regulations and were not included in this study.  This also
       removed 8,556 reaches from consideration.

    3.  Five BAT industrial categories (coal  mining, steam electric, plastics
       molding and forming, timber products, and pesticide manufacturing) were
       not evaluated in this study because their effluent regulations either:
       (1) do not specify toxic pollutants;  (2) control only a small volume of
       the total  discharge; or (3) have been rescinded. Therefore, an
       additional 8,353 dischargers and 1,593 reaches were excluded from the
       model/ambient analyses.
                                        -17-

-------
                                                                     Hethodology
     4.  Because  it  is  necessary  to  correlate  dischargers with  stream  locations
        (i.e., reach numbers), those  facilities  in the  EPA data  bases without
        assigned  reach numbers were excluded,  resulting in the removal  of  1,837
        facilities  from the  study.  This  step also removed all facilities  in
        Hawaii, Alaska,  and  the  U.S.  Territories, since streams  in these areas
        are  not currently  hydrologically  linked.

     5.  The  water quality  model  can evaluate  only facilities located  on
        hydraulic transport  ("stream") type reaches; therefore,  all reaches
        that are  nonhydraulic or boundaries (e.g., coastlines, estuaries,
        lakes, and  shorelines) were excluded.  This screening  step eliminated
        683  facilities and 322 reaches from consideration in the study.

     6.  The  remainder  of the facilities and reaches were then  screened  to
        determine if flow  data were available.  One hundred and  forty-four
        reaches were excluded because they did not have both average  and low
        flows  (which also  eliminated  484  facilities) and 1,232 industrial
        facilities  did not have  process flows  (which removed 497 reaches from
        consideration).  The final  step eliminated facilities  that may  have
        erroneous process  flows  stored in the data base.  If the process flow
        exceeded  the 95th  percentile  flow for a particular industrial category,
        the  facility and its corresponding reach were excluded.  This step
        removed 108 facilities and  53 reaches because of possible erroneous
        process flows.

     This screening process, summarized in Table 2-2 and Figure  2-1,  identified
2,490 facilities and  1,546  reaches (representing 24,289 river miles) to be

evaluated in the water  quality  modeling component of the Water Quality

Improvement Study  (WQIS).   The  1,546 reaches were also used as the basis for
retrieving  ambient monitoring data for the second component of this  study.


Additional  Sources

     Additional EPA sources  of information (for the model  and  ambient analysis)

include (1) industry-wide effluent characteristics by  industrial category and
(2)  EPA ambient  water quality criteria.


     The industry-wide effluent  characteristics were used in the water quality

modeling effort  to determine the pollutant concentrations discharged by each

of the facilities evaluated.  The  source of these concentrations was the

Monitoring  and Data Support Division (MDSD) report,  Summary of Effluent
                                        -18-

-------
                                                                    Methodology
                                    Table 2-2

                          Summary of Screening Process
                    Used to Select  BAT  Facilities  and Reaches
              for Inclusion in the Water Quality Improvement Study
Description of Screening Step
Total facilities/reaches in EPA data bases
No. of reaches without assigned facilities
No. of inactive facilities excluded
No. of non-BAT facilities/reaches excluded
No. of BAT facilities/reaches not evaluated*
No. of BAT facilities not assigned reaches
No. of BAT facilities/reaches excluded
because of non-steam reach type**
No. of BAT facilities/reaches excluded
because of unavailable reach flow
No. of BAT facilities/ reaches excluded
because of unavailable process flow***
No. of BAT facilities/reaches excluded
because of "erroneous" process flow
No. of
Facilities
120,992
0
46,467
59,338
8,353
1,837
683
484
1,232
108
No. of
Reaches
-68,000
-53,621
1,668
8,556
1,593
0
322
144
497
53
No. of BAT facilities/reaches evaluated                 2,490        1,546


  *Coal/steam/plastics molding and forming/timber products/pesticides
   manufacturing categories.
 **Lakes/coast-shoreline/estuary/artificial/etc.
***Includes "zero discharge" facilities (those facilities that do not
   discharge process wastewater).
                                        -19-

-------
                      Total Facilities Screening

                              (120,992)
BAT Facilities Screening

        (15,157)
r\»
o
                             Inactive
                             46.467
                        N on-BAT
                         59,338
                                                  BAT Facilities
                                                     15,187
Not Evaluated

    8,353
                                                                                         BAT Facilities
                                                                                             2,490
                         Erroneous Flow
                              108
                         No Process Flow
                              1,232

                        No Stream Flow
                             484
                      Non-stream Reaches
                                                                                              683
                 No Reaches
                    1,837
            Contains the 4 subcategories not evaluated.
                               Figure  2-1.

                               Summary  of  Facilities Screening Process

-------
                                                                    Hethodology


Characteristics and Guidelines for Selected Industrial Point Source
Categories:   Industry Status Sheets (U.S. EPA, 1986).  For direct dischargers,
pollutant concentrations are provided for four treatment levels:  RAW
(untreated wastewater), CURRENT (the average level of discharge determined by
sampling during the second phase of the BAT rulemaking process), BPT, and BAT
(the expected BPT and BAT discharge levels based on the required treatment
technology specified in the effluent limitations).  These industry-wide
effluent concentrations, developed by EPA's Industrial Technology Division,
were used to provide a consistent source of data for the selected treatment
levels and represent long-term averages.  RAW and BAT levels were used in the
water quality model analysis.

    The calculated in-stream pollutant concentrations from the water quality
model analysis were compared to EPA ambient water quality criteria (either
chronic freshwater aquatic life or human health ingestion of organisms only
criteria) for the ten selected pollutants.  The more stringent human health
ingestion of water and organisms criteria were not used because exposures due
to drinking untreated surface waters (most surface water sources are treated)
are not as likely.

    The human health criterion for benzene (a human carcinogen) represents a
risk level of 10   (1 excess cancer death per 1,000,000 people exposed over
a 70 year period).  For each pollutant, the lower of the two criteria was
used.  Table 2-3 shows the criteria used in this study.  Four of the metals
have hardness-specific criteria.  For these metals, the median hardness value
(as determined through a separate STORET analysis) for each State was used to
calculate State-specific criteria.

    The study used EPA criteria rather than individual State water quality
standards because:  (1) most States do not yet have numerical standards for
toxic pollutants; and (2) EPA criteria provided a consistent basis for
comparison with in-stream pollutant concentrations, even though some existing
State standards may be more or less stringent than EPA criteria.
                                        -21-

-------
                                                                    Nethodology
                                    Table  2-3

                       EPA Ambient Water Quality Criteria
                   Used in  the  Water Quality Improvement  Study
Pollutant          Criteria (ug/1)                      Type


Cadmium          Hardness-specific*       Freshwater aquatic life-chronic
Mercury              0.012                Freshwater aquatic life-chronic
Copper           Hardness-specific*       Freshwater aquatic life-chronic
Lead             Hardness-specific*       Freshwater aquatic life-chronic
Nickel             100                    Human health-ingesting organisms only
Zinc             Hardness-specific*       Freshwater aquatic life-chronic
Cyanide              5.2                  Freshwater aquatic life-chronic
Phenol             750                    Freshwater aquatic life-chronic**
Toluene            650                    Freshwater aquatic life-chronic**
Benzene             40                    Human health-ingesting organisms only


 *  Hardness-specific criteria are calculated as follows:

       Cadmium  e(°-7852nn(hardness)]-3.49)

       Copper   e(°-8545nn(hardness)]-1.465)

       Lead     e(1.273[ln(hardness)]-4.705)

       Zinc     e(0.8473[ln(hardness)]+0.7614)

**  Lowest Reported Toxic Concentration:

SOURCE:  EPA Ambient Water Quality Criteria Documents (various  dates).
                                        -22-

-------
                                                                    Methodology
             DESCRIPTION  OF METHODS  TO ASSESS  IMPROVEMENTS

    The following section contains brief descriptions of the three methods used
to assess water quality improvements that have resulted from the implementation
of BAT treatment technologies.

Mater Quality Modeling Analysis
    Water quality modeling was performed to identify theoretical water quality
improvements by pollutant, as well as by reach.   The analysis consisted of
modeling industrial facilities with BAT effluent limitations that directly
control toxic (priority) pollutants.  The model  calculated theoretical
in-stream concentrations of the ten selected pollutants at pre-  and post-BAT
treatment levels under low and average receiving stream flow conditions.   The
in-stream concentrations were then compared to applicable water  quality
criteria to identify improvements in meeting these criteria.  Of the 6,834 BAT
facilities with toxic limitations (not including the 8,353 facilities in  the
coal mining, steam electric, timber products,  pesticides manufacturing,  and
plastics molding and forming categories)  identified in the EPA data bases,
2,490 were evaluated in the modeling procedure.

    Assumptions and Limitations
    In conducting the water quality modeling,  a  number of assumptions were
made.  These assumptions,  together with their limitations,  are as follows:

    1. Industry-wide pollutant concentrations  for the BAT industrial
       categories were used to represent  individual  facility discharge
       levels.   This approach assumes that every facility in a particular
       industrial category discharges the same pollutants at the same
       concentrations; the only difference between facilities in the same
       category is the volume of wastewater discharged.   This assumption  does
       not reflect "real life" conditions, since each facility is different
       and effluent levels vary across categories.   Since this is a national
       study,  however, the differences in levels (some higher, some lower)
       will  tend to cancel each other out.
                                        -23-

-------
                                                                    Methodology
     2. Two treatment  levels  from U.S.  EPA  (1986) were used  in the water
       quality model:   RAW and BAT.  The RAW treatment level was used to
       represent  a particular industry's discharge level prior to the
       promulgation of  BAT.  The use of RAW as the pre-BAT discharge level
       overestimates  the actual pre-BAT (1972) discharge levels because it
       does not credit  industry with treatment in place at  that time.  Since
       it is not  known  how much treatment was occurring prior to BAT, RAW
       levels were used.  The BAT treatment level was used  to represent the
       levels at  which  an industry should be discharging after the
       implementation of the BAT regulations (post-BAT).

     3. Process wastewater flows contained in the IFD data base were used to
       represent  pre-BAT (RAW) flows.  Where flow reduction was required by
       the BAT limitations,  this was taken into account by  reducing the
       pollutant  concentration by a proportional amount.  Flow reduction was
       required for the following industrial categories:  aluminum forming,
       battery manufacturing, coil coating, copper forming, foundries,
       inorganic  chemicals,  iron and steel, nonferrous metals, nonferrous
       metals forming,  and porcelain enameling.

     4. BAT treatment levels were assumed to be the only effluent limitation
       imposed on the industries.  Water quality-based limitations were not
       considered since the  site-specific nature of such limitations prevents
       a theoretical estimate of their effectiveness.

     5. The model  assumed the pollutant was completely mixed in the receiving
       stream and that no fate-related removal  (e.g., sedimentation,
       biodegradation, volatilization) occurred.  While fate-related removal
       could be significant  (especially for the organics), this assumption is
       partially  offset by not considering background concentrations.

     6. The model  assumed that all currently active BAT facilities (as
       designated in PCS) were also active prior to BAT (i.e., plant closings
       and new sources were not accounted for).


    Analysis

    The objective of the water quality modeling was to project pre- and

post-BAT in-stream pollutant concentrations for each reach in the contiguous

U.S. that received wastewater discharges from BAT industries.  There are

currently 24 major industrial categories for which BAT effluent limitation
guidelines have been promulgated (as presented in Table 1-1).  The effluent

discharge water quality modeling was performed for 19 of the 24 industries
using industry-wide concentrations for the ten selected pollutants.   Five of
                                        -24-

-------
                                                                    Methodology
the  24  industries were excluded because their effluent regulations either did
not  specifically regulate toxic pollutants  (coal mining, timber products, and
plastic molding and  forming); controlled toxics discharge only for a small
volume  of the total  discharge (steam electric); or have been rescinded
(pesticide manufacturing).

     For each of the  facilities identified as BAT facilities from the EPA data
bases,  individual plant loadings were calculated (the product of industry-wide
effluent concentrations and process wastewater flows) for each of the ten
pollutants.  To calculate in-stream concentrations, the individual plant
loadings were summed for each pollutant for all facilities on a particular
reach.  These total  loadings were then divided by the stream flow (either low
or average, depending on the analysis) and sum of the plant flows.  The
following equation illustrates this procedure:

                             z (Ce x Qe)

                             Qs + s (Qe)
                        where: C^ = In-stream pollutant concentration (ug/1)
                               Ce = Effluent pollutant concentration (ug/1)
                               Qe = Process wastewater flow (MGD)
                               Qs = Receiving stream flow (MGD).
    This procedure was followed for each pollutant on each reach for the pre-
(RAW) and post-BAT (BAT) effluent levels and summarized on a reach-by-reach
basis (included in Volume II).  The resulting in-stream pollutant
concentrations were also compared to water quality criteria (WQC) to determine
compliance.

Ambient Water Quality Monitoring Data Analysis
    An analysis of ambient water quality monitoring data for those reaches
identified as having BAT discharges was also conducted to determine water
quality improvements.  Ambient water quality monitoring data (from STORET) for
the ten selected pollutants for time periods reflecting pre-BAT (1970 to 1980)
                                        -25-

-------
                                                                    Hethodology



and post-BAT  (1985 to  1988) were used to determine trends in ambient pollutant

concentrations.  Corresponding monitoring data  (i.e., data from both time

periods) was  available for at least one pollutant for 429 of 1,546 reaches

receiving BAT discharges.


    Assumptions and Limitations

    The following is a brief discussion of the  specific assumptions, and their

limitations,  made in performing the ambient water quality monitoring analysis:


    1. The monitoring period from 1970 to 1980  represents pre-BAT conditions
       and the period from 1985 to 1988 represents post-BAT conditions.  The
       intervening years were considered a time of transition and were not
       addressed.  All monitoring data for the  individual time periods were
       averaged together.  Improvement trends within each time period are not
       considered as well as trends during the  transition period.

    2. Monitoring data reported below detectable levels were included only
       where  detected levels were also available.  In those instances, the
       monitored values were set equal to one-half the detection limit.  It
       can not be determined whether or not this practice overestimates or
       underestimates the actual pollutant concentration.

    3. The STORET water quality monitoring data should be used with some
       caution, in that:

       •  The origins of the pollutants monitored include all  upstream
          sources, including facilities and sources not evaluated in this
          study (e.g., POTWs, non-BAT industries, natural and nonpoint
          sources);

       •  Information on the location of the monitoring station(s) in
          relation to the BAT industrial facilities (i.e., upstream or
          downstream) was not readily available; and

       •  The flow conditions during sampling were unknown.


    Analysis

    To aid in the determination of the water quality improvements attributable
to the BAT guidelines, ambient in-stream water quality monitoring data for the

ten selected pollutants were analyzed for two monitoring periods - pre-BAT and
post-BAT.   The ambient pollutant data were obtained from EPA's STORET Water
                                        -26-

-------
                                                                    Methodology


Quality file.  Monitoring data that were "unremarked" (i.e., pollutant
concentration was quantifiable) and data "remarked"  (pollutant concentration
less than the detection limit) were used in this analysis.  The individual
monitoring values were retrieved from STORE! and aggregated on a reach basis.
The average concentration for a particular pollutant for each of the two time
periods was calculated by using both the unremarked and the remarked data
(which was set to one-half the detection limit).  This averaging procedure was
used only when there was at least one unremarked value.

    In order to compare the two time periods, it was necessary to exclude
average values where data were available for only one of the two time
periods.  Of the 1,546 BAT reaches, 429 had ambient water quality monitoring
data available for at least one pollutant for both time periods.  The average
concentration for the two time periods were used to determine trends in
chemical water quality.  Three different classifications were used to define
pollutant concentration trends:  improved,  deteriorated, and no change.  An
"improved" trend signifies that the pollutant concentration decreased by more
than 10 percent between the two time periods.  Likewise, a "deteriorated"
trend denotes an increase in the post-BAT concentration of more than 10
percent.  A "no change" designation signifies that the pollutant concentration
did not change by more than 10 percent between the two periods.

Water Quality Improvement Case Studies
    The third approach to determining ambient water quality improvements was
to identify actual  improvements (case studies) that can be attributed to the
implementation of BAT requirements.  The focus of this effort was different
from that used for the water quality and ambient monitoring data analyses
since all  types of improvements would be applicable, not just improvements
from direct dischargers of toxic pollutants.   Such improvements could include
reductions in conventional  (suspended solids  or biochemical oxygen demand) or
nonconventional (ammonia and chlorine) pollutants and reductions in toxic
pollutants discharged from POTWs as a result  of the implementation of
                                        -27-

-------
                                                                    Nethodology


pretreatment programs.   Improvements could be shown through reduced in-stream
pollutant concentrations, attainment of designated water use, or improved
biological integrity of  the receiving stream.

    In order to identify possible case studies, the 1988 State Water Quality
Assessment [305(b)] Reports were reviewed.  From these reports, potential case
studies were selected for further investigation.  Of the thirty-six 305(b)
reports reviewed, nine States were identified as having potential case
studies.  Based on contacts with these nine States, five case studies were
selected to illustrate water quality improvements attributed to the
application of BAT regulations.  An additional case study was provided by EPA
Region X.  The summary of these case studies are presented in Chapter 3.

    A special case study was developed by the Agency to represent improvements
resulting from BAT on a typical estuary.  This case study used a water quality
model  to project theoretical improvements (similar to the nationwide model)
and analyzed ambient monitoring data to verify these projected improvements.
This special  study is also discussed in Chapter 3.
                                        -28-

-------
                                                                        Results
                                  Chapter Three

                                    RESULTS

    The overall results of the three methods used to identify water quality
improvements that are attributable to BAT are presented in this  chapter.   For
the water quality model and ambient monitoring data, water quality
improvements are presented in terms of river miles that (1)  complied with
water quality criteria or (2) showed a decrease in ambient pollutant
concentrations, respectively.  Only nationwide summaries are presented  here,
along with summaries of specific cases studies in which the  implementation of
the BAT regulations have resulted in water quality improvements.  The results
of the model/ambient analyses, on a reach-by-reach/State basis,  are provided
in Volume II, Technical Appendices.

          IMPROVEMENTS IN  WATER QUALITY - WATER  QUALITY MODEL

    As defined in Chapter 1,  a water quality improvement can mean either
(1) compliance with water quality criteria after  implementation  of  BAT
treatment technology when criteria had been exceeded prior to BAT,  or (2)  a
reduction in in-stream pollutant concentrations after BAT.   The  water quality
model addresses improvements  only in water quality criteria  compliance, since
theoretically all reaches receiving discharges from BAT facilities  have
improved in terms of pollutant concentration reductions (e.g., all  industrial
categories evaluated in this  study reduced their  discharge levels from pre-BAT
to post-BAT).

    The model assessed 2,490  BAT facilities on 1,546 reaches (totaling 24,289
river miles).  The model calculated theoretical in-stream concentrations for
the ten selected pollutants using industry-wide effluent concentrations,
process wastewater discharge  flows, and low (7-Q-10) and average receiving
                                        -29-

-------
                                                                        Results
stream flows.  The results of the water quality model are presented on both an
individual pollutant basis and a reach basis (to determine if all modeled
pollutants comply with their respective criteria).

Pol 1utant-by-Pol1utant
    Nationwide summaries of the river miles complying (instream concentration
below criteria) and not complying (instream concentration at or above
criteria) with water quality criteria (WQC) for the individual pollutants,
based on pre- and post-BAT discharge levels, using low and average receiving
stream flow are shown in Tables 3-1 and 3-2, respectively.  Under low flow
conditions, pre-BAT discharges are projected to result in less than half of
the river miles complying with the WQC for copper (43 percent), lead (47
percent), and cyanide (47 percent).  No pollutant had 100 percent compliance
under low flow conditions prior to BAT.  After BAT,  compliance ranged from
73 (mercury) to 100 percent (phenol, toluene, benzene).   The average increase
in compliance was 20 percent.   On a pollutant-by-pollutant basis, the
additional percentage of river miles complying with WQC after BAT is shown
below:

                                     Additional Percentage
                                 of River Miles Complying with
              Pollutant             WQC After BAT (Low Flow)
Cadmium
Mercury
Copper
Lead
Nickel
Zinc
Cyanide
Phenol
Toluene
Benzene
25
7
33
28
22
30
34
8
3 *
9
              * 97 percent compliance prior to BAT.
                                        -30-

-------
                                                               Results
                 Table 3-1

 Suanary of Water Quality Modeling Results:
Compliance with Criteria (at Low Stream Flow)
Pre-BAT

Cadmium

Mercury

Copper

Lead

Nickel

Zinc

Cyanide

Phenol

Toluene

Benzene


River Miles
Percent
River Miles
Percent
River Miles
Percent
River Miles
Percent
River Miles
Percent
River Miles
Percent
River Miles
Percent
River Miles
Percent
River Miles
Percent
River Miles
Percent
Not
Complying
w/HQC
9,030.2
37X
8.176.4
34X
13.791.9
57X
12.864.3
53X
6,265.7
26X
9.748.1
40X
12,916.3
53X
1,836.1
8X
730.3
3X
2,283.8
9X
Complying
w/HQC
15.258.5
63X
16,112.3
66X
10,496.8
43X
11,424.4
47X
18.023.0
74X
14.540.6
60X
11.372.4
47X
22.452.6
92X
23.558.4
97X
22.004.9
91X
Post-BAT
Not
Complying
w/HQC
3.038.2
12X
6,523.8
27X
5.752.0
24X
6.174.2
25X
1.043.2
4X
2,469.0
10X
4.562.1
19X
0.0
OX
0.0
OX
37.5
OX
Total
Complying River Miles
w/HQC Assessed
21,250.5
88X
17.764.9
73X
18.536.7
76X
18,114.5
75X
23.245.5
96X
21.819.7
90X
19.726.6
81X
24.288.7
100X
24,288.7
100X
24.251.2
100X
24,288.7

24.288.7

24,288.7

24.288.7

24.288.7

24.288.7

24,288.7

24,288.7

24.288.7

24,288.7

                  -31-

-------
                                                                                         Results
                                         Table 3-2

                        Summary of Water Quality Modeling Results:
                     Compliance with Criteria  (at Average  Stream Flow)
                              Pre-BAT                  Post-BAT
                           Not                      Not                       Total
                         Complying    Coop lying   Conplying    Complying   River Miles
                           w/WQC        w/VQC       w/WQC        w/WQC       Assessed
Cadmium   River Miles      2.807.4      21.481.3      332.9      23.955.8       24,288.7
          Percent            12X           88X         IX           99X

Mercury   River Miles      4,704.3      19,584.4    2.488.8      21.799.9       24,288.7
          Percent            19X           SIX        10X           90X

Copper    River Miles      5,421.6      18.867.1      923.1      23.365.6       24,288.7
          Percent            22X           78X         4X           96X

Lead      River Miles      5,764.3      18.524.4    1.007.6      23.281.1       24,288.7
          Percent            24X           76X         4X           96X

Nickel    River Miles      1,328.7      22.960.0        2.6      24.286.1       24.288.7
          Percent             5X           95X         OX          100X

Zinc      River Miles      2,915.2      21.373.5      338.7      23.950.0       24,288.7
          Percent            12X           88X         IX           99X

Cyanide   River Miles      6.059.2      18.229.5      436.2      23.852.5       24.288.7
          Percent            25X           75X         2X           98X

Phenol    River Miles        115.0      24.173.7        0.0      24.288.7       24.288.7
          Percent             OX          100X         OX          100X

Toluene   River Miles         50.3      24.238.4        0.0      24.288.7       24.288.7
          Percent             OX          100X         OX          100X

Benzene   River Miles         44.5      24.244.2        0.0      24.288.7       24,288.7
          Percent             OX          100X         OX          100X
                                       -32-

-------
                                                                         Results
     Improvements  in meeting WQC  are  less pronounced under average receiving
 stream flow conditions.   Prior to BAT, between 75  (cyanide) and  100  (phenol,
 toluene,  and benzene) percent of the river miles assessed complied with WQC.
 After the implementation  of BAT, compliance ranged from 90  (mercury) to
 100  (nickel, phenol, toluene, and benzene) percent.  The average increase  in
 compliance as a result of the implementation of BAT, at average  flow, was  10
 percent.   Individually, the additional percentage of river miles complying
 with WQC  as a result of BAT for  each of the pollutants, is shown below:

                                     Additional Percentage
                                 of  River Miles Complying with
              Pollutant           WQC After BAT (Average Flow)
Cadmium
Mercury
Copper
Lead
Nickel
Zinc
Cyanide
Phenol
Toluene
Benzene
11
9
18
20
5
11
23
0 *
0 *
0 *
              * 100 percent compliance prior to BAT.

Overall Reach
    The second method of evaluating the water quality improvements attributable
to BAT, based on the water quality model, examines the reach as a whole.  If
the reach is to meet WQC, then all modeled pollutants on that reach must comply
with their respective criteria.  These "overall" reach evaluations, therefore,
assess the effects of BAT on individual reaches for all the selected
pollutants.  The results of assessing water quality improvements using this
methodology, under low and average flow conditions, are shown in Figure 3-1.
                                        -33-

-------
                          LOW  STREflM FLOW
                         RVERRGE  STREAM  FLOW
            7 , 13b
co
*»
                                              TOTflL RIVER MILES flSSESSEDs 24,289
                                                 EXCEEDING  ONE  OR MORE  WQC

                                                 PRIOR TO flND flFTER  BflT



                                                 flODITIONflL RIVER MILES MEETING

                                                 flLL  WQC DUE TO BflT
                                             |   |  MEETING flLL WQC PRIOR

                                                  TO flND flFTER BflT
                                                                                                                             73
                                                                                                                             m
                                                                                                                             
                                                                                                                             c
                       F i gure  3-1  .
Summary  of   Water  duality

Model ing:   OveralI   Results

-------
                                                                        Results
    Under low flow conditions, only 29 percent of the river miles were
projected to meet all WQC prior to the implementation of BAT controls.  After
attaining discharge levels required under the BAT regulations,  the model
predicts that 58 percent of the river miles would meet criteria,  that is,  an
additional 29 percent of the river miles would be in compliance.   Even after
BAT, 42 percent of the river miles assessed are projected to exceed criteria
for one or more of the ten pollutants.  Under this scenario, mercury and lead
are the main causes of noncompliance.

    At average stream flow, only 59 percent of the river miles  were projected
to comply with WQC for all the modeled pollutants prior to the  implementation
of BAT.  After the implementation of BAT, an additional  29 percent of river
miles would comply with WQC; 12 percent will still not comply with criteria.
The major cause of noncompliance after BAT is mercury.

       IMPROVEMENTS  IN WATER QUALITY - AMBIENT MONITORING DATA

    Improvements in water quality,  as determined through analysis of ambient
monitoring data, focus on trends in pollutant concentrations (both on an
individual and on a reach basis) as opposed to comparison with  water quality
criteria.  The primary reasons for this approach center on the  general  lack  of
monitoring data for reaches evaluated using the water quality model  and the
method of determining average pollutant concentrations for the  pre-  and
post-BAT time periods.

    Of the 1,546 reaches (24,289 river miles) assessed by the water quality
model, 429 (totaling 8,434 river miles) had monitoring data for at least one
of the selected pollutants for both time periods.   None of the  evaluated
reaches,  unfortunately, had monitoring data for both the pre-BAT  and post-BAT
time periods for the selected organic chemicals (phenol,  toluene,  and benzene).
                                        -35-

-------
                                                                        Results
Pol 1utant-by-Pol1utant
    Table 3-3 presents a summary of the in-stream concentration trends for each
of the seven monitored pollutants.  Between the pre-BAT (1970 to 1980) and
post-BAT (1985-1988) time periods, each of the pollutants showed a marked
decrease in in-stream concentration.  On the average, each of the pollutant
concentrations decreased in 78 percent of the monitored river miles.  Cadmium
and mercury showed the greatest concentration decreases (or improvement) in
monitored levels (84 and 87 percent of the river miles improved).  Zinc levels
reflected the least improvement (69 percent improvement).   The extent of no
significant change in monitored pollutant concentrations between the two time
periods ranged from 1 percent of the river miles (for mercury) to 11 percent
of river miles (for copper).  A deterioration, or increase in concentration,
occurred along some reaches for each pollutant.  The extent of deteriorations
ranged from 11 percent of the river miles (for cadmium) to 25 percent of river
miles (zinc).

Overall  Reach
    Using a method to show overall trends on a reach basis (similar to the
method used in presenting the results of the water quality model for overall
compliance with WQC), the ambient monitoring data were analyzed to determine
overall  trends in pollutant concentrations for the monitored pollutants.
Using this method,  about 76 percent of the river miles with monitoring data
available showed an overall improvement (or net decrease in pollutant
concentrations).  Roughly 14 percent of the river miles showed a net increase
(deterioration) in monitored concentrations and 11 percent showed no
significant change.  Figure 3-2 illustrates these overall  trends based on the
ambient  monitoring data analysis.
                                        -36-

-------
                                                             Results
                Table 3-3

Sunnary of Ambient  Monitoring Data Analysis:
             Pollutant Trends
Improved No Change
Cadmiun

Mercury

Copper

Lead

Nickel

Zinc


Cyanide

River Miles
Percent
River Miles
Percent
River Miles
Percent
River Miles
Percent
River Miles
Percent
River Miles
Percent

River Miles
Percent
3,822.6
84X
2,807.4
87X
4,349.9
70%
4,659.7
82X
3,590.4
72X
5,296.2
69X
0
1.228.2
SOX
193.0
4X
31.5
IX
667.3
11X
279.3
5X
229.8
5X
498.1
6X

110.7
7X
Deteriorated
519.5
11X
375.6
12X
1,238.2
20X
766.1
13X
1.172.9
23X
1,889.7
25X

205.3
13X
Total
Monitored
4,535.1

3,214.5

6,255.4

5,705.1

4,993.1

7,684.0


1.544.2

              -37-

-------
                                                          Results
       6.  397
       (76%)
                                                       890
                                                      (11%)

                                                     1, 147
                                                      (14%)
                  TOTAL RIVER MILES ASSESSED:  8434
                           NO SIGNIFICANT CHANGE


                      mm DETERIORATED
                      IUUUB

                      ||  IMPROVED
* River miles  for reaches with BAT  facilities and  monitoring data
  for one or more of the ten toxic  pollutants.
Note:  Percentages add up to  more than 100 percent  due to rounding.
 Figure  3-2.
Summary  of
Ana lys is:
Reaches.
Amb ient
Overa11
Monitoring  Data
Trends  for  BAT
                                -38-

-------
                                                                        Results
              IMPROVEMENTS IN WATER QUALITY  - CASE STUDIES

    Potential case studies screened for this study could represent any
improvement in water quality attributable to the implementation of BAT
regulations, including controls on conventional pollutants and regulations
applicable to indirect dischargers.  However, the focus was on toxic pollutant
controls for direct discharging facilities.  After a review of a number of
State 305(b) reports as well as conversations with State officials,  it appears
that the focus of case studies have been on municipal  discharges,  where
controls on oxygen-demanding pollutants and nutrients  have resulted  in
improved oxygen levels in streams.  The major concern  with toxic chemicals,  as
evidenced by the 305(b) reports, is sediment contamination, primarily because
of PCBs and pesticides.  Little information is available concerning
improvements that have resulted from toxic pollutant discharges,
especially instances involving technology-based (i.e., BAT) controls.
However, six studies have been reported that indicate  improvements in water
quality resulting from BAT or BAT-type controls.

    In addition to the above case studies, a detailed  water quality  model  and
ambient monitoring data analysis was performed by the  Agency on an estuary
(Delaware River Estuary) to highlight the effects of BAT regulations on these
types of waterbodies.

Long Island Sound - Connecticut
    In 1985, the Connecticut Department of Environmental  Protection  (CT DEP)
initiated a study to collect current fish tissue contaminant data  for
comparison with historical  data in order to show temporal  trends.  Of the  many
sites and organisms selected for testing, oyster data  obtained from  Bridgeport
Harbor and the Housatonic River areas indicated that these contaminated areas
had improved over the past decade.  Both Bridgeport Harbor and the Housatonic
River have a heavy concentration of metals-related industries (metal
                                        -39-

-------
                                                                        Results


finishing, copper forming, and foundries).  The comparison of 1972-1974 oyster
metal concentrations to the data collected in 1985-1986, shown in Figure 3-3,
indicates that the "levels of cadmium, chromium, copper, mercury and zinc were
currently lower than the lowest concentrations observed in an intensive study
from the early 1970s.  Although the 1985-1986 survey was not detailed enough
to permit rigorous statistical analysis, the disparity in metals levels
strongly suggests a reduction in metals contamination of oyster tissues in the
Bridgeport and Housatonic Rivers" (NY DEC, 1988).  Since Connecticut currently
does not write water quality-based permit limitations, but instead bases its
industrial discharge levels on BAT-type standards, the improvements in metals
concentrations can be at least partially attributed to reductions in discharge
levels from the metal industries located in these areas.

Naugatuck River - Connecticut
    Another example of water quality improvements in Connecticut is the
Naugatuck River.   According to CT DEP (1988), the Naugatuck was once
considered one of the most polluted rivers in the nation.   From 1973 to 1976,
CT DEP issued abatement orders to 77 industrial  dischargers along the river.
Metal finishers,  the most prevalent type of industry on the river,  were
required to neutralize acids, destroy cyanide water, and precipitate heavy
metals (BAT-type treatment technologies).   "Monitoring data has shown a marked
reduction in heavy metals, such as copper and zinc,  and improved pH levels.
While the river was virtually devoid of aquatic life in 1970,  water quality
has now improved to the point where the upper 22 miles ...  have been stocked
with trout on an experimental basis.  The river has been identified by CT DEP
Fisheries Unit as a potentially valuable resource for cold  water and anadromous
fisheries" (CT DEP,  1988).  Water quality problems still exist on the lower
portions of the river,  which will  require more stringent (e.g.,  water
quality-based) permits in order to achieve water quality goals.
                                        -40-

-------
                                                      Results
z  •
o +•
- 3
I-

-------
                                                                        Results
South  Fork Coeur d'Alene River  -  Idaho
    The  South  Fork Coeur d'Alene  River has had serious pollution problems for
many decades,  which were the result of ore mining and related activities.  At
the time of  the enactment of the  Federal Water Pollution Control Act of  1972,
heavy  metals concentrations in  the river reached 23,000 ug/1 for total zinc,
200 ug/1 for total cadmium, and more than 500 ug/1 for total lead during the
summer low flow periods.  These levels are roughly two orders of magnitude
higher than  EPA's acute (short-term) water quality criteria for protection of
aquatic  life.  As a result of the 1972 Act, effluent limits for industrial
dischargers were required.  EPA's Region X Office initiated a monitoring
program  in 1972 to document the improvements in the South Fork Coeur d'Alene
River  and identify any remaining  sources of heavy metals.  Figure 3-4 shows
the trends in  zinc, cadmium, and  lead at the mouth of the South Fork Coeur
d'Alene  River  from 1972 to 1986 (USEPA, 1987).  Each metal shows a decrease of
roughly  90 percent during this  period.  While metal concentrations still
exceed criteria levels, conditions of the South Fork are now suitable for many
of the less sensitive indigenous  species of aquatic biota, and conditions of
the mainstream are enabling game  fish to return downstream of the South Fork
confluence.  Data now indicate  that nonpoint sources are responsible for 50 to
90 percent of the metals.

Lower  Fox River - Wisconsin

    The  Fox River Valley is heavily industrialized, especially with paper
mills, and 50 percent of all point source discharges occur in the lower
portion  (from Depere to Green Bay).  High pollutant loadings from the paper
mills have contributed greatly to the historically low dissolved oxygen levels
in the Lower Fox River.  As a result of the Clean Water Act of 1972, improved
wastewater treatment systems,  which began operation in the 1970s,  have
resulted in the attainment of the 5 mg/1  dissolved oxygen standard for much of
the time period after the treatment systems became operational.   Permitted
effluent levels for the paper mills have been set at the limits  established by
                                        -42-

-------
                                                              Results
     zs, oo>o —|
     201, 0)0)01 —
     15, 0)0)0) _
      5, 0)010) _
                             ZINC
           1172
                                                         38b   1488

-------
                                                                        Results
the national categorical standards  (BPT/BAT), however, many of the mills are
discharging at lower levels.  The Wisconsin Department of Natural Resources
(WDNR)  is currently revising its effluent permits to reflect water
quality-based limits for toxic pollutants; nonetheless, the national effluent
standards have had a beneficial impact on the Lower Fox River.  According to
historical trends related to macroinvertebrates, there has been a recent
increase in pollution  intolerant aquatic life (WDNR, 1985).

Other Case Studies
    There are several  other case studies in which the water quality
improvements can be at least partially attributed to the implementation of
national categorical effluent standards.  The first involves the Scioto River
in Ohio.  Using a macroinvertebrate index as a measure of overall water
quality, the Ohio Environmental Protection Agency (OEPA) determined that prior
to 1976, the Scioto River had marginal attainment of a warm water habitat.  A
great change in the index was noted between 1977 and 1978, reflecting an index
indicative of exceptional macroinvertebrate fauna.  These exceptional
conditions continued to exist through 1985.  The changes in water quality "are
most attributable to improvements in wastewater treatment" at a major pulp and
paper facility on the  river (OEPA, 1988).

    Using the same macroinvertebrate index, OEPA has shown an improving trend
in biological  conditions on the Mohican River.  "These improvements may be
attributable to industrial  waste pretreatment (electroplaters) requirements in
the cities of Mansfield and Ashland, as well  as wastewater treatment
improvements by various industries and WWTPs" (OEPA, 1988).

Delaware River Estuary
    To determine the effects of the BAT regulations on estuaries a case study
of the Delaware Estuary was performed.  The general  methodology for this case
study follows that used in the nationwide analyses of BAT facilities and
reaches.  A modified water quality model, developed for EPA [from information
                                        -44-

-------
                                                                        Results
obtained from the National Oceanic and Atmospheric Administration (NOAA)], for
use in determining 304(1) lists of waters impaired by toxics, was used to
project in-stream pollutant concentrations from pre- and post-BAT effluent
discharges from BAT industries.  This model used the same industry-wide
concentrations of the ten selected pollutants as the nationwide water quality
model analysis.  Ambient water quality monitoring data were also retrieved
from EPA's STORET Water Quality File to show trends in selected pollutant
levels between the pre-BAT (1970-1980) and post-BAT (1985-1988) time periods
and to verify projected compliance with WQC (no State standards available).

    This analysis was based on estuary zones rather than reaches.  There are
three zones in any estuary: a freshwater tidal zone, a saltwater tidal zone,
and a mixing zone located between the two.  For this analysis, the in-stream
(or in-estuary) pollutant concentration was calculated using "pollutant
concentration potentials," a procedure developed by NOAA, that takes into
consideration both the flow available for dilution and the zone characteristics
(salinity).  The concentration potentials do not consider pollutant fate.  A
more detailed explanation of this approach can be found in U.S. EPA (1988).

    The Delaware estuary water quality model evaluated 64 BAT facilities
discharging at pre- and post-BAT levels.  Twenty-two discharged to the
freshwater zone (Zone 1), 39 discharged to the mixing zone (Zone 2), and
3 discharged into the saltwater zone (Zone 3).  The model evaluated each zone
independently and did not account for inputs from upstream zones or other
sources.  Table 3-4 presents the results of the water quality model  for the
Delaware estuary.  Prior to the implementation of BAT,  Zone 1 was projected to
not comply with freshwater criteria for mercury, copper,  lead, zinc, and
cyanide.  After the facilities met the discharge requirements of BAT,  all
pollutants were projected to comply with of WQC.  Likewise,  in Zone 2, four
pollutants (mercury,  copper,  lead, and cyanide) did not comply with WQC prior
to BAT and full compliance was projected after BAT.  The WQC used for this
zone was the more stringent of the two freshwater and saltwater criteria.  All
                                        -45-

-------
                                                                          Table 3-4

                                                    Water Quality Model Results for the Delaware Estuary
                                                                       Pollution Concentration  (uq/11
     Number of    Cadmium       Mercury        Copper         Lead       Hickel         Zinc         Cyanide         Phenol        Toluene        Benzene
Zone Facilities  RAW   BAT    RAW     BAT    RAW     BAT   RAW     BAT   RAW   BAT   RAW      BAT   RAW    BAT   RAW     BAT   RAW    BAT    RAW      BAT


 1       22      0.15  0.016  0.41  *  0.001  20.3 *  0.35  24.9 *  0.13  1.07  0.05  118.3 *  0.90  36.7  *  0.14  165.5   0.05  26.1   0.04   21.95    0.03

 2       39      0.31  0.037  0.90  *  0.008    6.5 *  0.18    7.1 *  0.18  2.22  0.13   30.4    1.02  11.0  *  0.32    43.9   0.09   8.0   0.01    5.6     0.02

 3        3      0.0   0.000  0.00     0.000    0.04   0.00    0.01   0.00  0.04  0.00    0.1    0.00    0.04  0.00     0.07   0.00   0.01  0.00    0.01    0.00
Fresh WQC
Salt WQC
1.1
9.3
0.012
0.025
11.4
2.9
3.0
5.6
100
8.3
102
86
5.2
1.0
750
290
650
3200
40
40
* Predicted in-stream concentration exceeds WQC.

NOTE:  Fresh WQC compared to in-stream concentrations  in  Zone 1.
       Salt WQC compared to in-stream concentrations  in Zone 3.
       The more stringent of the two WQC was compared  to  in-stream concentrations  in  Zone  2.

-------
                                                                        Results
pollutants in Zone 3 were projected to comply with WQC (saltwater) both before
and after the implementation of BAT.

    The ambient monitoring data analysis used information obtained from
monitoring stations designated as "estuarine," as opposed to ambient "stream"
stations used in the nationwide evaluation.  Values that were reported as
"less than detection limit" were handled in the same manner as the nationwide
evaluation.  Data from two stations were used to represent average pollutant
concentrations in Zone 1 and seven stations were averaged for Zone 2 (see
Table 3-5).  No ambient estuarine data were available for Zone 3 and limited
data (in terms of the number of pollutants) were available for Zones 1 and 2.
The results of the monitoring data analysis compare favorably to the water
quality model.  In Zone 1, the average pre-BAT concentrations for copper and
lead did not comply with WQC.  After BAT, only lead still did not comply with
criteria, but only by a slight margin.  All monitored levels decreased by at
least 83 percent.  In Zone 2, all pollutants showed a marked decrease in
average pollutant concentration.  Cadmium decreased by 91 percent, mercury by
55 percent, copper and lead by over 90 percent each, and zinc by 81 percent.
The pre-BAT average concentrations for cadmium, mercury, copper, and lead were
above WQC levels.  These pollutants were also above WQC after BAT, but by a
much smaller margin.  One possible reason to account for the non-compliance is
that 12 of the 39 BAT facilities were in the organic chemicals category and
may not have fully implemented the requirements of the BAT regulations (the
phase II regulations were promulgated in November 1987).
                                        -47-

-------
                                                                                                 Results
                                               Table 3-5

                   Ambient Water Quality Monitoring Data Sunraary for  Delaware Estuary

Number of
Zone Facilities
1 22
2 39
3 3
Fresh WQC
Salt WQC
Average Pollutant Concentration (ua/1)
Cadmium Mercury Cooper Lead
70-80 85- 70-80 85- 70-80 85- 70-80 85-
49.7 * 2.70 53.2 * 5.1 *
29.2 * 2.52 * 1.6 * 0.722 * 51.5 * 2.92 * 53.8 * 5.1 *

1.1 0.012 11.4 3.0
9.3 0.025 2.9 5.6

Zinc
70-80 85-
61.0 10.4
64.0 11.9

102
86
* Average concentration exceeds WQC.

NOTE:  Fresh WQC compared to in-stream concentrations in Zone 1.
       Salt WQC compared to in-strean concentrations in Zone 3.
       The more stringent of the two WQC Mas compared to in-stream concentrations in Zone  2.
       70-80 represents the pre-BAT time period (1970-1980).
       85-   represents the post-BAT time period (1985-present).
                                                 -48-

-------
                                                                    Conclusions
                                  Chapter Four

                                 CONCLUSIONS

    The three components of this study indicate that water quality has improved
as a result of the implementation of the BAT effluent limitation regulations:

Water Quality Model.

    The results of the water quality modeling effort, which evaluated 2,490
BAT facilities impacting 24,289 total river miles (1,546 unique reaches),  show
that under low stream flow conditions 14,169 river miles (58 percent) comply
with all the water quality criteria for the ten selected pollutants after  the
implementation of BAT (an additional 29 percent improvement over pre-BAT
conditions).

    Under average receiving stream flow conditions,  the model  predicts that
59 percent of the river miles modeled will  comply with all  criteria prior  to
the implementation of BAT.  After BAT, 88 percent of the river miles assessed
(an additional 29 percent) were projected to meet all ten criteria.  The major
causes of noncompliance with criteria are discharges of mercury, lead, and
copper.  All other pollutants comply with criteria in at least 81 percent  of
the assessed river miles at low flow (98 percent at  average flow).

    The use of the water quality model does have certain limitations.  The
model does not consider upstream sources of pollutants, nor does it consider
the discharge of other "nonselected" pollutants by the BAT facilities.  These
sources could impact the extent of compliance with criteria.  This  limitation
is offset to a certain extent by the fact that pollutant fate  is also not
considered.  The model also assumed all  facilities,  within a particular
                                        -49-

-------
                                                                    Conclusions
category, discharged the same pollutants at the same concentrations.  While
this is not a particularly valid assumption when dealing with an individual
facility, on a nationwide basis, the tendency to overrepresent or
underrepresent actual discharge levels is, at least partially, eliminated.
Finally, this modeling effort does not take into consideration the treatment
technologies in place prior to the Clean Water Act, and thus the actual
improvements should be somewhat less than projected.

    Based on the results of the water quality model, the national categorical
effluent standards program (BAT) has been found to be an effective tool for
improving water quality up to a point.  However, there may be a need for
additional water quality-based controls beyond BAT in some cases in order to
meet State water quality standards.  The model predicts that 42 percent of the
assessed river miles may exceed EPA national water quality criteria under low
stream flow conditions after BAT is in place (12 percent of the river miles do
not comply at average flow conditions).  This projection, however,  is not a
precise measure, since it is dependent on the number of pollutants evaluated.
In many cases, water quality-based NPDES permit limits may have been already
developed in order to meet locally applicable State water quality standards,
and in other cases new water quality-based permit limits may be needed.

Ambient Water Quality Monitoring Data

    Improvements in water quality, based on the ambient water quality
monitoring data analysis, are not as evident.   Appropriate ambient monitoring
data were available for only 35 percent of the river miles assessed by the
model,  and no comparable data (same reach for both the periods)  were available
for the organic chemicals selected.  This lack of ambient monitoring data is
expected,  especially for the organic chemicals,  since the Agency's major focus
during the early and mid-1970's was on conventional pollutants.   Only after
the Clean Water Act of 1977 were toxic pollutants emphasized.  It is also
important to note that ambient monitoring data is collected for many other
purposes that just to determine the effectiveness of controls placed on
                                        -50-

-------
                                                                    Conclusions


industrial discharges.  Nevertheless, it is apparent that a more goal-oriented
and focused effort needs to be made to truly evaluate controls on industrial
discharges.

    The monitoring data analysis, however, did show a great improvement, in
terms of pollutant reductions, between the pre- and post-BAT time periods.
About 76 percent of the river miles (assessed in this evaluation) indicated an
overall decreasing trend in in-stream concentrations of toxic pollutants.
However, about 14 percent showed an increase.  Concentrations of individual
pollutants (cadmium, mercury, lead, and cyanide) improved in 80 percent or
more of the assessed river miles, while zinc, nickel, and copper showed
increases (deterioration) in 20 to 25 percent of the assessed miles.

    The full benefit of the implementation of BAT is not reflected in the
ambient monitoring analysis.  National categorical  standards have recently
been promulgated for one major industry, organic chemicals manufacturing,
which will reduce the industry's toxic pollutant direct discharge loadings by
1.1 million pounds per year.  Standards for another category (pesticides
manufacturing) are currently being prepared.  While the ambient data do
reflect some of the benefits attributable to BAT, the full effect will  not be
evident until the early 1990s.  The second component of this study has  several
other limitations.  The sources of pollutants represented in the monitoring
data are not known, although these sources should include the BAT facilities
evaluated in the model.  Other sources could include upstream BAT facilities,
municipal facilities, hazardous waste sites, nonpoint,  and natural
(background) sources.  Reductions in the monitored  pollutants could possibly
be attributed to controls on these sources.  Another factor that could  account
for some of the decreasing pollutant concentration  trends is the increase in
accuracy of the analytical techniques used to determine the pollutant
concentrations.  In the past decade, increased sophistication of the
laboratory equipment has enabled the detection limits for all  pollutants to be
1owered.
                                        -51-

-------
                                                                    Conclusions
    The statistical significance of comparing monitoring data from time
periods of different spans was not evaluated.  Also, the naturally occurring
variability in monitored pollutant levels was not assessed; however, the
effect of such variability should be reduced by using average values over the
periods evaluated.

Case Studies

    Actual cases where the implementation of BAT has resulted in water quality
improvements present the best illustration of the effectiveness of the
categorical standards.  The few case studies that are available show that the
BAT regulations have had a positive impact on the receiving stream quality, in
terms of both chemical and biological improvements.  In most instances, the
streams/rivers assessed in these case studies were highly polluted prior to
BAT.  Even after the discharge levels were reduced to levels proscribed by
BAT, additional water quality-based controls may be needed in order to meet
State water quality standards.  In all cases, the implementation of BAT has
resulted in considerable improvements in the biological  quality (as measured
by an increase in less pollution-tolerant aquatic life)  of the receiving
waters.

    The special case study predicted and verified that improvements have
occurred in the chemical water quality of the Delaware estuary.   All
pollutants, as projected by the water quality model, that exceeded criteria
prior to BAT complied with criteria after the implementation of these
regulations.  The ambient monitoring data analysis verified these
improvements.   However, in Zone 2, the area most influenced by organic
chemical  manufacturing discharges, the predicted improvements are not yet
fully realized (e.g., the BAT regulations for the organic chemicals category
are not yet fully implemented by industry).
                                        -52-

-------
                                                                    Conclusions
Summary of Water Quality Improvements

    Considering the results of each component of this analysis, together with
their respective assumptions and limitations, the BAT regulations have been an
effective step toward improving the quality of our nation's waters.  The
extent of this effectiveness is difficult to assess using the existing EPA
data sources and considering the fact that the full benefit of BAT has not yet
been realized.  Also, by not considering improvements resulting from the
categorical pretreatment requirements, the extent of the improvements
resulting from the overall national water quality program are underestimated.
There may be a need for additional water quality-based controls beyond BAT in
some cases to meet State water quality standards.  In addition, as required by
Section 304(m) of the Water Quality Act of 1987, the Agency will establish a
schedule for:  (1) the annual review and revision of promulgated effluent
guidelines, and (2) the promulgation of regulations for industrial categories
identified as sources of toxic and nonconventional pollutants for which
guidelines have not previously been established.
                                        -53-

-------
                                                                     References
                                  Chapter Five
                                  REFERENCES
Connecticut Department of Environmental Protection, 1988.  State of
    Connecticut 1988 Water Quality Report to Congress.  Water Compliance Unit.

New York Department of Environmental Conservation, 1988.  New York State Water
    Quality 1988.  Division of Water, Bureau of Monitoring and Assessment.

Ohio Environmental Protection Agency, 1988.  Ohio's Water Quality Inventory -
    1988 305(b) Report, Volume I.  Division of Water Quality Monitoring and
    Assessment.  Columbus, Ohio.

U.S. Environmental Protection Agency, 1988.  Summary of Effluent
    Characteristics and Guidelines for Selected Industrial Point Source
    Categories:  Industry Status Sheets.  Interim Final Report.  Office of
    Water Regulations and Standards, Monitoring and Data Support Division.
    Washington, D.C.

U.S. Environmental Protection Agency, 1987.  Coeur d'Alene Basin - EPA Water
    Quality Monitoring (1972-1986), Draft Final.  EPA Region X.

U.S. Environmental Protection Agency, 1988.  Estuarine Dilution Analyses to
    Estimate Toxic Substance Impairment for 304(1) Identification, Draft
    Report.  Office of Marine and Estuarine Protection and Office of Water
    Regulations and Standards.  Washington, D.C.

U.S. Environmental Protection Agency (various dates).   Ambient Water Quality
    Criteria Documents.  Office of Water Regulations and Standards,  Criteria
    and Standards Division.  Washington, D.C.

U.S. Environmental Protection Agency.  National Primary Drinking Water
    Regulations.  40 CFR Parts 141 and 142.

Wisconsin Department of Natural Resources, 1985.  Lower Fox River -  Depere to
    Green Bay:  Water Quality Standards Review.  Bureaus of Water Resources
    Management and Fish Management.
                                        -54-

-------