United States
            Environmental Protection
            Off ice of
            Washington DC 20460
July 1987
Permit Writer's Guide
to Water Quality-Based
Permitting for
Toxic Pollutants


       Permit Writer's Guide
To Water Quality-Based Permitting
        For Toxic Pollutants
        U.S. Environmental Protection Agency
              Office of Water

                July 1987


The National Pollutant Discharge Elimination System (NPDES} permits program
has begun to focus increased regulatory attention on the control of toxic pollutants
to protect water quality. Recent passage of amendments to the Clean Water Act
and the acquisition of increasing amounts of data on the toxicity of effluents point
to the need for an increased effort to control the discharge of toxic  pollutants.
The Permit Writer's Guide to Water Quality-Based Permitting for Toxic Pollutants
provides procedural recommendations to State and Federal NPDES permit writers
on setting water quality-based permit limits for toxic pollutants. The recommen-
dations contained in this document are not mandatory and are intended to be sug-
gestions for setting such permit limits.
This document is expected to be revised periodically to reflect advances in this
area. Comments from users will be welcomed. They should be sent to Rick
Brandes,  U.S. EPA, Office of Water Enforcement and Permits (EN-336), 401 M
St. S.W., Washington, D.C. 20460.


                          Table of Contents

   Foreword	       iii
   Table of Contents	      v
   Acknowledgements	     vii
   Glossary	      ix
   Introduction	       xi
1.  Basis for Toxics  Control	      1
2.  Components of Water Quality-Based Toxics Control	      3
     EPA Water Quality Criteria for Aquatic Life	      3
     EPA Water Quality Criteria for Human Health	      3
     State Water Quality Standards	      4
     Mixing Zones	      4
     Design Flow of Receiving Waters	      5
     Wasteload Allocations	      6
     Lake, Estuarine, and Marine Discharges	      6
     Effluent Flow	      6
     Recommended  Toxicity Criteria	      6
3.  Permitting  Procedures	      9
   Subsection  3.1—Implementing  Narrative  Standards	      9
     A.  Approach 1 —Setting Limits	      9
     B.  Approach 2 —Setting Monitoring  Requirements	     14
   Subsection 3.2 —Implementing Numeric Criteria	     18
     A.  Limits from  Two Number  State Criteria	     18
     B.  Limits  from  Single  Number State  Criteria	     18
     C.  Limits Based on Unspecified Toxicity Value	     19
     D.  Limits Based on Whole Effluent Numeric Value	     19
   Subsection 3.3 —Recommended Permitting Procedures	     19
     A.  Issuance Involving Toxicity Limits	     19
     B.  Issuance Not Involving Specific Limits	     20
4.  Case Example	     21
5.  Toxicity  Reduction  Evaluations  (TRE)	     25
     Application of TREs and Example Language	     25
     Legal  Basis	     26
     Conducting  TREs	     26
   Appendix A —Example Special Conditions  Permit Language
   Appendix B —Example §308 Letter
   Appendix C—Overview of Selected Available Tools


                                      A cknowledgements
The principal authors of the Permit Writer's Guide to
Water Quality-Based Permitting for Toxic Pollutants
were Rick Brandes and Tom Wall of the Permits Divi-
sion and  Bruce Newton of the Monitoring and Data
Support Division. Hiranmay Biswas, Monitoring and
Data  Support Division,  contributed the multiplier
tables used in the permit limit derivation process.
Comments on the draft Guide were provided by EPA
Regions, the Association of State and Interstate Water
Pollution Control  Administrators  (ASIWPCA),  a
number of States,  and a number of individual EPA
water quality experts in the Regions, the Office of
Research and Development, and EPA Headquarters.
Their efforts are greatly appreciated.


acute  Involving a stimulus severe enough to rapidly
   induce a  response; in toxicity tests, a response
   observed  in 96 hours or less typically is considered
   acute. An acute effect is not always measured in
   terms of  lethality; it can measure a variety of ef-
   fects. Note that acute means short, not mortality.

acute-chronic ratio (ACR)  The ratio of the acute tox-
   icity (expressed as an LC50) of an effluent or a tox-
   icant to its chronic toxicity (expressed as an NOEL).
   It is used  as a factor for estimating chronic toxici-
   ty on the basis  of acute toxicity data.

additivity  The characteristic property of a mixture of
   toxicants that exhibits a cumulative toxic effect
   equal to the arithmetic sum of the  effects of the
   individual toxicants.

antagonism   The characteristic property of a mixture
   of  toxicants  that  exhibits  a less-than-additive
   cumulative toxic effect.

Average Monthly  The highest allowable average of
   "daily discharges" over a calendar month as de-
   fined in 40 CFR  § 122.2, calculated as the sum of
   all "daily  discharges" measured during a calendar
   month divided by the number of "daily discharges"
   measured during that month.

bioaccumulation   Uptake and retention of substances
   by an organism  from its surrounding medium and
   from food.

bioassay   A test used to evaluate the relative potency
   of a chemical by comparing its effect on a living
   organism  with the effect of a standard preparation
   on the same type of organism. Bioassays are fre-
   quently used  in the  pharmaceutical  industry to
   evaluate the potency of vitamins and drugs. "Bio-
   assay" and "toxicity test" are not synonymous.

chronic  Involving a stimulus that lingers or continues
   for a relatively long period of time, often one-tenth
   of the  life span  or more. Chronic should be con-
   sidered a  relative term depending on the life span
   of an organism.  A chronic effect can be lethality,
   growth, reduced reproduction, etc. Chronic means
Coefficient of Variation (CV)  The statistical measure-
   ment  of variability  calculated  as   CV =
   [V(X)/E(X)2]°-5 where V(X) equals the variance and
   E(X} equals the  long term average.
Criteria Continuous Concentration (CCC)  The EPA
  national water quality criteria recommendation for
  the highest instream concentration of a toxicant or
  an effluent to which organisms can be  exposed in-
  definitely without causing unacceptable effect.
Criteria  Maximum  Concentration (CMC)  The EPA
  national water quality criteria recommendation for
  the highest instream concentration of a toxicant or
  an effluent to which organisms can be exposed for
  a brief period of time without causing mortality.
design flow The critical flow used for steady state
  wasteload allocation modeling.
duration  The period of time over which the instream
  concentration is averaged for comparison  with
  criteria concentrations. This specification limits the
  duration of concentration above the criteria.
frequency  How  often  criteria  can  be exceeded
  without unacceptably affecting the designated use
  of the receiving water.
LC50  The  toxicant concentration  killing 50%  of
  exposed organisms at a specific time of obser-
magnitude   How much of  a pollutant (or pollutant
  parameter such  as toxicity), expressed as a con-
  centration or toxic unit is allowable.
Maximum  Daily   The  highest  allowable  "daily
  discharge"  permit  limit  as defined  in 40  CFR
  § 122.2 which should not be exceeded by the per-
No  Observed  Effect  Level  (NOEL)  The  highest
  measured continuous concentration of an effluent
  or a toxicant that causes no observed  effect on a
  test organism.
recurrence  interval  The  average number of years
  within which a variable will be less than or equal
  to a specified value. This term is synonymous with
  return period.
steady state model  A fate  and transport model that
  uses constant values of input variables to predict
  constant values of receiving water quality concen-
Total Maximum Daily Load (TMDL)  The sum of the
  individual wasteload allocations for point sources
  and load allocations for nonpoint and natural
  background  sources.

toxicity test  The means to determine the toxicity of
  a chemical or an effluent using living organisms. A
  toxicity test measures the degree of  response of
  an exposed test organism to a specific chemical or
Toxic Unit acute(TUa)  The reciprocal of the effluent
  dilution that causes the acute effect on the test
  organisms by the end of the acute exposure period.
Toxic Unit chronic(TUc)   The reciprocal of the ef-
  fluent dilution that causes no unacceptable effect
  on the test organisms by the end of the chronic
  exposure period.

wasteload allocation (WLA)  The portion of a receiv-
  ing water's total maximum daily pollutant load that
  is allocated to one of its existing or  future point
  sources  of pollution.

whole effluent toxicity  The aggregate toxic effect of
  an effluent measured directly with a toxicity test.

The Permit Writer's Guide to Water Quality-Based Per-
mitting For Toxic Pollutants provides State and Federal
NPDES permit writers and water quality management
staff a reference to water quality-based permit  is-
suance procedures. It supports the implementation of
existing State water quality standards and current per-
mitting techniques. As such, it presents fundamental
concepts and procedures in detail and simply refers
to those  more advanced toxics control procedures,
such as dynamic modeling of complex discharge situa-
tions, which may not yet be incorporated into many
State programs. The Guide is meant to explain aspects
of water quality-based toxics control in terms of what
a permit  writer currently needs to know to issue a
water quality-based toxics control NPDES permit. It
augments the Office of Water's Technical Support
Document for Water Quality-Based  Toxics  Control,
EPA 440/4-85-032, September, 1985, (the TSD)
which provides more detailed discussion of much of
the contents of this document. The user of this Guide
should have  read the TSD and  have a copy  of  it
available when implementing recommendations of this
The NPDES permits program is now focused on con-
trol of toxic pollutants. This document is directed at
supporting these toxics control efforts. Water quali-
ty problems related to conventional pollutants, such
as those associated with point source contributions
to oxygen depletion, are addressed in other guidance
In this guide, three types of toxic effects are ad-
dressed: toxic effects on aquatic life, toxic effects on
human health, and toxic effects due  to the bioac-
cumulation of specific chemicals.  Each of these
effects must be dealt with on an individual basis us-
ing available data and available tools. In addition, this
guide seeks to catalogue the principal procedures and
tools available and present them to the permit writer
in the context of their relationship to permit issuance.
An integrated toxics control strategy using both whole
effluent toxicity-based assessment procedures and
pollutant-specific assessment procedures is strongly
encouraged for most permitting situations. Both pro-
cedures are needed to enforce  State  water quality
standards. Both have benefits and disadvantages, and
so both are often needed in the toxics control process.
Chemical-specific controls will almost  always be
needed  to  meet State  standards for  individual tox-
icants and to assess an effluent for human health and
bioaccumulation  problems.  Toxicity testing  will
usually be necessary to assess overall toxicity since
NPDES  effluents  are  often  complex mixtures of


                              Section 1.  Basis for Toxics  Control
Permit writers have a clear, unambiguous basis for
setting  water quality-based permit limits for toxic
pollutants: existing State water quality  standards.
Both the Clean Water Act and promulgated Federal
regulations require that all NPDES permits include
limitations to achieve all applicable State water quality
standards. Further, NPDES permits must include limit-
ations which reflect any total maximum daily loads or
wasteload allocations set by EPA or States to achieve
applicable water quality standards.

Two forms of State water quality standards for tox-
icants can be  used to set  NPDES  permit  limits:
numerical  standards and   narrative  standards.
Numerical standards for some individual toxicants are
contained in virtually all State water quality standards.
They are usually expressed as an instream "not-to-
be-exceeded" concentration of a toxicant  (e.g. 0.019
mg/l for total residual chlorine).

All States also have narrative standards for pollutants.
The  most common narrative standards are the four
"free-froms" which normally require that a State's
waters  shall be free from:

  (a) substances that will cause  the formation of
      putrescent or otherwise objectionable bottom

  (b} oil, scum, and floating debris in amounts that
      are unsightly or deleterious;

  (c)  materials that cause odor, color or other condi-
      tions in such a degree to cause a  nuisance;

  (d) substances in concentrations or combinations
      toxic to humans, wildlife, or aquatic life.

The standard under (d) above pertains to toxic effects
and is an important element in any effective toxics
control  strategy.  The standard should be  used by
States  and  EPA Regions to  limit both individual
toxicants (where a toxic effect  can be traced  to a
specific chemical for which no standard or criteria ex-
ist) and  whole effluent toxicity (where it is not obvious
which chemicals are causing toxicity or  where the
limitation  of  generic effluent toxicity is more  ap-
propriate to that particular discharge situation).

EPA's policy and legal basis regarding the use of State
water quality standards to set NPDES permit limits on
toxicants is provided by the Office of Water's "Policy
for the Development of Water Quality-Based Permit
Limits for Toxic Pollutants," 49 FR 9016, March 9,
1984 {see Appendix A of the TSD). In part, that policy
states that:

  Where violations of water quality standards are
  identified or projected, the State will be expected
  to develop water quality-based effluent limits for
  inclusion in any issued permit. Where necessary,
  EPA will develop these limits in consultation with
  the State. Where there is a significant likelihood of
  toxic effects to biota in the  receiving water, EPA
  and the States may Impose permit limits on effluent
  toxicity and may require an NPDES permittee to
  conduct a toxicity reduction evaluation. Where tox-
  ic effects  are  present but there  is a  significant
  likelihood that compliance with technology-based
  requirements will sufficiently mitigate the effects,
  EPA and the States may require chemical and tox-
  icity testing after installation of treatment and may
  reopen the permit to incorporate additional limita-
  tions if needed to meet water quality  standards.
  [Toxicity data, which are  considered "new infor-
  mation" in accordance with 40 CFR 122.62(a)(2),
  could  constitute cause for  permit  modification
  where necessary]  (emphasis added).

NPDES  permits  must be developed and  issued in
accordance with current permit issuance policies, in-
cluding current Agency Operating Guidance, permit
issuance strategies, and State-specific agreements
and workplans. Applicable water quality standards
and site  specific water quality data as well as in-
dividual discharge data should be assessed during the
permit issuance process to determine whether water
quality-based permit  requirements  for toxics  are
necessary for a particular discharge.

To make this determination, the permit writer should
consider data from all available sources including:

   1)  Lists   of   waters needing  toxics control
      developed by States as  required by the 1 987
      Clean Water Act amendments.

   2) Water body-specific control strategies devel-
      oped by States as required by the 1987 Clean
      Water Act amendments.

   3) Form  2C of the permit application.

   4) State or Federal compliance inspections, per-
      formance audits, or stream use designation

   5) Discharge  Monitoring  Reports (DMR),  data
      generated in excursion reports, and any other
      chemical or biological data generated by the
   6) For  POTWs,  industrial  user  surveys  and
      chemical/ toxicological analyses of influent, ef-
      fluent, and sludge samples collected as part of
      an approved pretreatment program.
   7) Complex Effluent Toxicity Information System
      (CETIS) database.

   8) State 305(b) reports.
   9} State  401 (c)  certifications   (non-delegated
      States only).
  10) Any available stream  survey  data including
      tissue residue studies.
  11) Other additional information obtained from the
      permittee through use of a § 308 letter or other
  12) Other  additional  information  available  from
      water quality management specialists  in the
      Region or State.
The use of this data and the selection of candidates
for water quality-based toxics control is site specific.
It is  becoming increasingly clear that the presence
or absence   of  unacceptable effluent  toxicity is
sometimes highly variable. The toxicity of an effluent
(and  the subsequent need for toxics control) is depen-
dent on many factors including:
  •  manufacturing process employed
  •  toxicity of products, intermediates, and ancillary
     chemicals used
  •  treatability of chemicals  in the effluent
  •  soundness of best management practices
  * variability of effluent composition and concen-
  • contribution of indirect wastewaters (for POTWs)
  • capacity and real retention time of treatment

Experience of States, Regions, and EPA's Office of
Research and Development has shown that the use
of accepted generalizations about toxicants and tox-
icity of NPDES effluents can be inappropriate in tox-
ics control situations. For example, not all effluents
from organic chemical plants exhibit toxicity; only
some do. Not all facilities in primary industries are in
need of water quality-based toxics  control.  Not all
secondary industries should  be exempted from con-
sideration. Many POTWs (up to 40% in some areas)
have had measurably toxic effluents; even  some
POTWs without significant industrial contributions.

The important point in this context is that combina-
tions of factors must be considered. The question to
be answered is "is it reasonable  to assume that the
subject discharger might discharge a toxic effluent?"

One further note on effluent toxicity is important. The
list of 126 priority pollutants  is not an all-inclusive list
of toxic chemicals. This list was  based on a number
of factors such as national production level and poten-
tially high human health impacts. Some very impor-
tant factors  (including  the  treatability of  the
compounds by typical treatment systems) were not
considered in the development of the list. Therefore,
effluent toxicity estimates cannot be and should not
be  made  based  on  the analysis  of the priority
pollutants alone. Absence of any or all of the priority
pollutants does not mean that further investigation
into effluent  toxicity,  human  health problems, or
bioaccumulation potential can be terminated.

              Section 2.  Components of Water Quality-Based Toxics Control
Permit writers should become familiar with certain
aspects of the water quality management program
which may ultimately lead to requirements established
in NPDES permits. These components are EPA water
quality criteria for aquatic life, EPA water quality
criteria  for human  health.  State  water quality
standards, mixing  zones,  design flow of  receiving
waters, wasteload allocation/total maximum daily load
(WLA/TMDU,  exposure for  lake,  estuarine,  and
marine discharges, effluent flow, and recommended
criteria for whole effluent  toxicity.
EPA Water Quality Criteria for Aquatic Life
EPA's water quality criteria represent a large database
on  the effects  of  toxicants  and  other  pollution
parameters. They are issued in the form of individual
criteria documents  (e.g.. Ambient Water Quality
Criteria for Mercury, EPA 440/5-84-026, January
1985.) Each document catalogues available data on
the effects of each  chemical on aquatic life. Bioac-
cumulation is considered where FDA action levels or
wildlife feeding studies are available. Ambient concen-
trations are  then calculated which, if met instream,
will not adversely affect aquatic life.
EPA criteria for aquatic life are now expressed as two
numbers: a number  that will protect against chronic
effects (termed the  Criterion Continuous Concentra-
tion or CCC), and a  number that will protect against
acute effects (termed the Criterion Maximum Concen-
tration or CMC). Two concentrations need to be
specified because aquatic organisms can  tolerate
higher concentrations  for short periods of time than
they can tolerate for  long periods of time. Severe
adverse effects can quickly occur at instream concen-
trations that exceed  the CMC. A CMC is set as an up-
per concentration which, if not exceeded, will insure
that no unacceptable effects will occur due to a short
(i.e., one hour) exposure.
Adverse effects can also  occur at  instream concen-
trations between the CCC and the CMC if the biota
are exposed for a sufficiently long period of time. The
CCC  can be exceeded for a short period  of  time
without unacceptable harm to the ecosystem. Longer
exposure, however,  will cause unacceptable toxic ef-
fects  (such as impairment of survival, growth  or
reproduction). The  CCC is established to  protect
against effects caused by long-term exposure to con-
centrations below the CMC. Thus, two numbers are
needed to provide protection against toxic effects.
A criterion consists  of three parts. For example, the
criterion for chlorine is presented as a concentration
(11 /ig/L for the CCC, 19 /ig/L for the CMC), a dura-
tion of exposure (four-day average for the CCC, one-
hour average for the  CMC),  and a frequency  of
allowed excursion  (once every three years on the
average for both the CCC and the CMC). Criteria are
stated in terms of acceptable excursions because we
know there is variability in nature and random excur-
sions of criteria will occur. The criteria set conditions
which will be protective of the ecosystem even if ex-
cursion  of  the  criteria occur.  The  concept  of
establishing criteria in terms of allowable excursions
is discussed in much more detail in the TSD (see Sec-
tion 2 and Appendix D) and in Stream Design Flow
for Steady State Modeling (August, 1986), available
from EPA's Monitoring and Data Support Division.
Concentration, duration, and frequency are used in
steady state modeling  to calculate the allowable
ambient concentration of the toxicant (or toxicity) on
the basis of a  design flow of the receiving water (e.g.,
7Q10). The design flow should be established so that
it will properly account for the duration and frequency
aspects of the criteria. It specifies a  volume of flow
occurring at a critical period for the receiving water
and limits discharge based on this period. A permit
limit is back-calculated to determine the concentra-
tion of the toxicant or toxicity in  the effluent which
will not cause those ambient levels to be exceeded
more often than allowed.
The design flow can be calculated in several ways.
This is discussed below in the subsection on design

EPA Water  Quality Criteria for Human  Health

EPA has published human health  criteria on all but a
few of the 65 classes of priority pollutants. For poten-
tial carcinogens, the  individual criteria are  usually
presented as single  values,  often  correlated to

estimated cancer  risk levels of  10~5, 10~6,  and
10~7. The human  health criteria should be used as
the principal means of controlling toxicants of concern
to human health.  Permit writers should use these
criteria to set limits on individual toxicants using the
permit limit derivation procedures described in Sec-
tion 3 of this document.

The use of a "whole effluent" human health criterion,
similar to the effluent toxicity criteria described in this
section,  is an  attractive  option  when controlling
effluents for human health  impact. Such a criterion
would control the impact of all  hazardous materials
in a complex mixture like an effluent. However, unlike
the whole effluent toxicity criteria for aquatic life, no
whole  effluent criteria  for  human health can  be
calculated since the direct human health effects of
effluents cannot be tested. Effects of  complex mix-
tures (like effluents} on human health can be screened
using tests like the Ames test but limits based upon
Ames test results cannot be assured to be protective
of all human health concerns. Ames test data (or any
other human health test data) can lead to identifying
individual toxicants of  human  health  concern but
limiting the mutagenicity or carcinogenicity of a whole
effluent based on such limited data is not generally
recommended. Therefore, toxicants of human health
concern should generally be limited on a  chemical-
specific basis using laboratory-de rived effect concen-

Exposure to toxicants of human health  concern is dif-
ficult to project because human health effects have
long latency periods. Effects from exposure may not
show up for years. Modeling the actual routes of ex-
posure is difficult since toxicants often change form
and therefore bioavailability upon discharge. Further,
for drinking water exposure there is almost always a
treatment system between toxicants discharged to a
receiving water and human exposure.

Human health (and bioaccumulation) impact protec-
tion involves specific exposure considerations which
affect how WLAs  are subsequently calculated. For
example, cancer in humans probably results from ex-
posure to toxicants over a period of years,  not days.
Therefore,  design  flows used to protect against
human health effects  should reflect longer term ex-
posure conditions (e.g., 30Q5 as a design flow) than
the design flows used for aquatic life  protection.
State Water Quality Standards
State water quality standards represent conditions
which must be met instream to support the desired
water  quality.  The  standard  includes  both  a
designated use (e.g., protection of aquatic life) and
criteria (e.g., copper of 5.6/*g/l) that, if not exceeded,
will protect that use. Different criteria apply to dif-
ferent uses. A permit writer must know the designated
use prior to setting water quality-based permit limits.

Numeric criteria can be established for individual tox-
icants or for generic toxicity. To date, few States have
formally adopted the two number format for their tox-
icant criteria. Most State criteria are listed as not-to-
be-exceeded  values. This  presents a  problem for
water quality analysts and  permit writers following
EPA's two number criteria and duration/frequency for-
mat and will require some adjustment in procedure.
Section 3 of this Guide describes how a two number
permit limit can be derived from a single number State

Some States have numeric criteria for whole effluent
toxicity, often stated as end-of-the-pipe acute toxicity
limits. Section 3 also describes how these  numeric
criteria for toxicity can be used.
Mixing Zones
Many State standards also specify allowable mixing
zones. Since all water quality-based controls are by
definition meant to protect ambient conditions, effects
after discharge are the effects of concern. This means
that regulatory authorities should  consider effluent
mixing. Many State standards allow a zone of mixing
in which less stringent criteria apply. The rationale is
that a small area of degradation  can exist without
causing effects to the overall waterbody.

EPA's policy on mixing zones is described in the 1983
Water Quality Standards Handbook. This document
states that any mixing zone should be free  from
materials which cause acute toxicity to aquatic life.
Acute toxicity is of  particular  concern  because
organisms passing through the mixing zone can be ex-
posed to lethal concentrations of toxicants in a par-
tially diluted effluent.

The TSD gives three options for applying  criteria in
mixing zones. In complete mix conditions, the criteria
can be applied using the entire  streamflow. If  com-
pletely mixed conditions do not occur within a  short
distance of the outfall, the criteria should be applied
at specific  points in the mixing zone. The TSD gives
detailed guidance on these approaches (see page 32
of the TSD). Finally, a third option  is to disallow any
mixing and meet criteria at the end-of-the-pipe.

Design Flow of Receiving Waters
Design  flow is  a hydrological  condition  which
describes a low flow of a flowing waterbody. It is
calculated from the historical flow record. Since am-
bient toxic effect  is directly related to available dilu-
tion, an analysis of available dilution over a period of
time is essential. Only by knowing what dilution is ex-
pected and when it is expected can a dilution analysis
and corresponding permit condition for allowable tox-
ics discharge be set. The purpose of the design flow
is to  set a  low  flow condition against which an
allowable effluent discharge can be calculated.

The design flow for a receiving water is of significance
only in steady state wasteload allocation modeling
because effluent composition is assumed constant.
Worst case conditions can thus be calculated and pro-
tected against. In dynamic modeling, the actual dilu-
tion situation is modeled so that ambient conditions
can be predicted for all days. A single "worst case"
estimate of flow is not needed. In steady state model-
ing, the actual dilution  patterns  are not precisely
known. The level of effluent control must be set on
a "worst case" basis to insure that ambient criteria
are not exceeded during low flow conditions. Upon
determining  the extent and duration of dilution ex-
pected, the regulatory authority can perform a dilu-
tion analysis and  set the corresponding WLA.

The design flow is important to permit writers because
it represents the flow against which available dilution
is calculated for steady state WLAs and corresponding
water quality-based permit limits. It is in this calcula-
tion that the regulatory authority must assure that the
duration and frequency recommendations for acute
and chronic  protection are achieved. When specify-
ing and using a design flow, the regulatory authority
must  assure that  an effluent will not produce a con-
centration instream which will exceed the concentra-
tion, duration, and frequency specified by the criteria.
To do this, the regulatory authority must specify a
design flow (and then calculate permit limits) to pro-
tect, in a worst case flow situation, against a permit-
tee reaching an acute impact concentration instream
over a one-hour period more than once in three years
on the average and against a permittee reaching a
chronic impact concentration over a four-day period
more  than once every three years on the average.
Design flow of a receiving water can be expressed in
an xQy format. X refers to the averaging period (in
days) of the flow measurement and represents the
averaging period of concern. Daily flows over x days
are averaged to determine the  flow. Y refers to the
return period. This is the number of years over which
the flow is expected to occur at 1/y probability. The
most commonly used design flow, 7Q10, represents
the lowest flow averaged over any consecutive seven
day period that is expected to occur in a typical ten
year period. It is expressed in cubic feet per second
(cfs) or million gallons per day (MGD). One general
misconception about such flows is that the flow exists
only once in ten years. The 7Q10 flow may occur on
a number of days in a non-consecutive manner dur-
ing any one year. It just does not exist for seven full
days in a row. It can also occur over consecutive
periods of time within the ten year period since it is
based on probability alone.

The procedures in the recent EPA guidance document
Stream  Design  Flow  for Steady-state Modeling
(August, 1986) are recommended  for  calculating
design flow for flowing waters. Two  options are

   1} The hydrologically-based design flow method

     This  option recommends that the 1Q10 be
     used as the design flow for calculating dilution
     for the CMC and  that 7Q10  be used as the
     design flow for calculating dilution for the CCC.
     The distinction between stressed and unstress-
     ed systems described in the  TSD has been

   2) The biologically-based design flow method

     This option involves the use of  DFLOW, a com-
     puter program which employs an iterative con-
     vergence procedure for calculating the two
     design flows. DFLOW is installed on EPA's NCC-
     IBM  computer and  is run under the  TSO
     operating  environment.  The procedure  is
     described in the document. Copies of the FOR-
     TRAN source code can be obtained from Lewis
     Rossman, WERL, U.S.  EPA, 26 West St. Clair
     Street,  Cincinnati, Ohio 45268. Phone

This document supercedes the TSD's steady-state
design flow recommendations.
Wasteload allocation modeling for streams usually
uses flow  data obtained  from  the United States
Geological  Survey gauging stations. If sufficient flow
data are not available for a stream of interest, data
must be extrapolated from other streams having
hydrologic characteristics similar to those of the
stream of interest. Also, for intermittent streams or
streams  where  the  design  flows  equal zero, the
calculation of WLAs are site-specific, depending on
the State's designated use of the receiving water.
Where such a receiving  water has a designated use
which  must be protected, water quality-based limits

would be applied at the end-of-the-pipe. No dilution
would be allowed.

Once a regulatory authority specifies a design flow for
a steady state modeling case, the permit writer no
longer has  to  be  concerned  with duration  and
frequency. They are taken care of by specifying the
design flow. Dilution calculations can then proceed.

Wasteload Allocation/Total
Maximum Daily Load

A WLA is the site-specific effluent loading a permit
writer seeks to achieve when writing a permit to im-
plement  the applicable water  quality  standard(s).
There are two forms of WLAs: WLA based on steady
state modeling, and WLA based on dynamic model-
ing. Only steady state WLAs will be discussed in this

A steady state WLA is calculated (usually by State or
Federal water quality analysts working with permit
writers) by setting an appropriate design flow, deter-
mining allowable mixing characteristics for acute and
chronic protection, and calculating allowable loadings
based on the concentration and flow of the effluent.
In the case of multiple discharges, loadings must be
allocated to  each discharger depending on how the
State determines allotment.

The permit writer subsequently takes this target WLA
and begins the permitting procedure, designed to set
limits which will ensure compliance with the WLA and
thus  the standards.   This  permitting procedure is
described in detail in Section 3 of this document.

Exposure for Lake, Estuarine,
and Marine Discharges

Lake, estuary, and marine discharge situations require
specific procedures for establishing dilution on which
to base steady-state WLAs.  Biologically-based design
flows, such  as those  described  above,  are  not
currently available for such receiving waters. Only
estimates of allowable dilution can be made for lake,
estuary, and marine discharges. These are used to
calculate  concentrations of effluent at the edge of the
mixing zone. Again, the CCC is to be met at the edge
of the mixing zone. If near-field dilution is of concern,
dye studies can be required to measure  the concen-
tration of the effluent within the mixing zone and at
the edge of the mixing zone. The TSD provides some
guidance on how to analyze mixing for lakes and reser-
voirs (page 37), estuaries (page 38), and oceans (page
38).  The recommendations for marine discharges
comes from the 301 (h) publication "Initial  Mixing
Characteristics of Municipal Ocean Discharges" {see
Appendix C.VII.B for cite).

Effluent Flow
Since a critical flow period is of concern, the flow of
the effluent during this period should be used. If this
is not available, use the average annual  flow of the
treatment plant. For POTWs, the design flow should
be used if the plant flow is expected to  continue to
increase with time. If not, an annual average flow or
the permitted flow specified by the State would be
appropriate. Since effluents can be highly variable, the
effluent flow which could cause the greatest impact
should be used.

Recommended  Criteria for the Parameter
Effluent Toxicity
Just as a regulatory agency can set permit limits for
an element (like lead) or an effluent parameter {like
BOD), so too can  "whole effluent" toxicity be used
as an effluent parameter for which  controls are set
(40 FR 9016, March 9, 1984. See Appendix A of the
TSD). The individual causative agents of the effluent
toxicity need not  be identified in advance for this
approach to be effective.

For simplicity, toxicity measurements are converted
to Toxic Units. Toxic Units are calculated by dividing
100 by the LC50 or the NOEL. There are acute Toxic
Units (TUa)  and   chronic Toxic Units  (TUC).  An
effluent with an LC50 of 5% contains 100/5 or 20
  •  Acute Criterion
As described in detail in Section 2 of the TSD, the
recommended criterion for acute toxicity {the CMC}
is 0.3 TUa. The adjustment factor of one-third is used
to extrapolate the LC50 to an LC1 (concentration at
which 1 % of the  test organisms die) and thus limit
the lethality involved in setting criteria based on an
LC50 (where 50% die).
  •  Recommended Point of  Application for Acute
The  criterion  can  be  applied  at different  points
depending on the  dilution situation.  In a  situation in
which the effluent is not completely mixed (a mixing
zone is involved), effluent in the pipe itself must meet
the 0.3 TUa criterion  unless a high rate diffuser
disperses the effluent. If a high rate diffuser is used,
0.3 TUa must be met within a short distance of the
outfall as described in the TSD. In rapid, complete mix
discharge situations, the criterion must be met after
complete mix dilution is calculated at the design flow.

Here, dilution of the effluent at the acute design flow
must be at least threefold for cases where the LC50
of the effluent is greater than 100%. Only if there is
more dilution available can the allowable effluent
acute toxicity be greater.
  •  Chronic Criterion
For chronic toxrcity, the recommended criterion (the
CCC} is 1.0 TUC.
  •  Recommended Point of Application for Chronic
In incomplete mix situations, the criterion must be met
after dilution at the chronic design flow at the edge
of the mixing zone. This means that to meet this am-
bient criterion, the MATC or the NOEL of the effluent
must be no higher than the dilution achieved at the
edge of the mixing zone. If the dilution available at the
edge of the mixing zone  is  10:1, then the allowable
NOEL of the effluent is >  10%. No toxic effect to
aquatic life is expected  to occur at this concentration.
In this dilution situation, a NOEL of 2% would not be
acceptable because the concentration of the effluent
at the edge of the mixing zone  (10%) is five times
higher than the concentration at which chronic effect
to aquatic life occurs. In rapid, complete mix discharge
situations, the criterion must be met after complete
mix  dilution is calculated at the design flow.
In summary, the criteria for toxicity in complete mix
situations are 0.3 TUa and 1.0 TUC applied after com-
plete mixing. The criteria for toxicity in incomplete mix
situations can be diagrammed as follows:
 No high rate diffuser:
                              mixing zone boundary
  High rate diffuser (> 3m/sec):

                             mixing zone boundary
                 CMC = 0.3 TUe
                 CCC = 1.0 TIL
It should be noted that the recommendations for the
point of application for these whole effluent toxicity
criteria can be applied to individual toxicants as well.


                               Section 3. Permitting Procedures
This section describes permit limit derivation pro-
cedures. The procedures described here are the same
as those described in  Section 6 in the TSD but are
listed in a step-wise manner. They apply to permits
issued to both industrial and municipal dischargers.
Steady-state  dilution  WLA derivation  procedures,
usually conducted by water quality specialists rather
than permit writers, are also described in this section
to show the basis for permit limit derivation. However,
the process will  apply regardless of the  specific
organizational  or functional  arrangements of  the
regulatory authority.
Only steady-state dilution WLAs and corresponding
permit limit derivation procedures are shown. The
more advanced  WLAs, which  involve the fate of
pollutants and the use of dynamic models, are detailed
in other guidance documents (including the TSD). See
Appendix C for additional references.
The procedural recommendations which  cover  the
water quality-based permitting process are arranged
in three subsections as follows:
  Subsection  3.1—Implementing  Narrative Water
    Quality Standards
    A.  Approach  1—Set  permit limits using a
        statistical  limit derivation approach and
        steady-state WLAs
    B. Approach  2—Set toxicity  monitoring  re-
        quirements prior to setting permit limits
  Subsection  3.2—Implementing  Numeric  Water
    Quality Criteria
    A.  Deriving permit limits where a two-number
        water  quality  criteria exist for a toxicant
    B. Deriving permit limits where a single-number
        water  quality  criterion exists for a toxicant
    C.  Deriving permit limits based on an unspecified
        numeric toxicity value
    D.  Deriving  permit limits  based  on a whole
       effluent numeric toxicity value
Subsection 3.3 — Recommended  Water Quality-
    Based  Permit Issuance Procedures
    A. Issuance of permits involving specific toxicity-
       based  discharge limitations
    B. Issuance of permits  which do not involve
       specific toxicity-based discharge limitations
In a typical case, the water quality-based permitting
process follows a consistent pattern. First, since there
are two numbers for ambient water quality criteria (the
CMC and the CCC), two WLAs for acute and chronic
protection  are calculated  using  the State's water
quality criteria  (or EPA criteria) for the parameter, the
mixing zone, and the design flow specified. If more
than one discharger is present, these other potential
sources must also be considered. One of these two
WLAs becomes limiting and thus becomes the basis
for  control  requirements. Next, permit  limits are
calculated  which will implement the more limiting
WLA. The limits  are put into the permit and the per-
mittee is required to meet them. Compliance monitor-
ing requirements for the parameter limited are also put
into the permit.

EPA Regional  Offices and  State Agencies have
adopted a wide variety of organizational arrangements
for  development of standards, WLAs, and  permit
limits.  This document  is  not  intended to influence
established arrangements, but rather to encourage a
teamwork approach to toxics control. Experience has
shown that a variety of disciplines are necessary  to
assess and control toxicity effectively. Therefore, per-
mit writers should  routinely  consult water  quality
analysts during the development of water quality-
based  permit  limits, and  encourage  water  quality
analysts to reciprocate  when executing their duties.

3.1  Implementing Narrative Water Quality

Regulatory authorities have two approaches they can
use to implement narrative standards: A) set permit
limits using a statistical limit derivation approach and
steady-state WLAs or; B) set toxicity monitoring re-
quirements, through a §308 letter, or in the Special
Conditions section of the permit itself. This monitor-
ing information is then subsequently used to develop
WLAs  and  permit limits, if  necessary,  either for
effluent toxicity  or for individual toxicants found  to
cause the toxicity.  .

A. Approach  1 — Set Permit Limits Using a
  Statistical Limit Derivation Approach and
  Steady-State WLAs

The following  procedure is described  in terms  of
setting permit  limits for effluent toxicity.  Note that

any parameter for which  a criterion or standard is
available can be substituted for toxicity in the limit
derivation process.
Under this procedure, no initial toxicity testing or tox-
icity data are  necessary. Since acceptable effluent
toxicity is  a function of  dilution and  the  ambient
criteria, the available dilution will drive the WLA and
thus the permit limit. The key to this procedure is to
determine the dilution factor.

Steady-State WLA Derivation Procedure
The following equation can be used to derive the
steady-state WLAs  needed as the basis for setting
limits on toxicity:
  effluent toxicity WLA  < toxicity criterion X  a
                           dilution factor
             criterion = 0.3 TUafor acute toxicity
                        1.0 TUC for chronic toxicity
  dilution factor {d.f,} = the factor by which the ef-
    fluent is diluted in the receiving water. The d.f.
     = Qs /Qe where the plant's water source is the
     receiving water. The d.f.  = 06  + 05 / Qg where
     the plant's water source is not the receiving
                  QS = receiving water design flow
                  Qe = effluent design flow
  The two WLAs are calculated in the following way,
  depending on the dilution situation:
    •  where mixing is rapid and complete (involving
       the entire streamflow)
       —  set an acute toxicity WLA using 0.3 TUa
          at the acute design flow (e.g., 1Q10if the
          hydrologically-based design flow option is
       —  set a chronic toxicity WLA using 1.0 TUC
          at the chronic design flow {e.g., 7Q10 if
          the hydrologically-based design flow op-
          tion is used)
    •  where mixing is not rapid and complete and the
       entire streamflow should  not be provided  to
       the discharger
       —  set an acute toxicity WLA of 0.3 TUa at
          the end-of-the-pipe where no diffuser  is
          present. Where a diffuser is present, use
          0.3 TUa times the dilution after initial ef-
          fluent mixing.
       —  set a chronic toxicity  WLA  of  1.0 TUC
          times the dilution factor at the edge of the
          State designated  mixing zone  at the
          chronic design flow.
Users of this procedure should note that this process
assumes upstream pollutant/toxicity concentrations
are zero. When writing permit limits, background con-
centrations should be assessed to determine if a site-
specific adjustment of allowable loading should be
Multiple Sources
The above WLA derivation equation assumes there is
no existing ambient concentration of the parameter
being limited. In multiple source situations where
steady-state  WLAs are used, two additional con-
siderations are involved. First, additivity and conser-
vation of the toxicity from different sources should
be assumed.  Second, the flows should be summed
and allowable  loads  distributed to the individual
sources using some option for allocating loads. These
allocation options are discussed in detail in  the TSD
(page 47).
The equation for multiple sources which can be used
to derive the two steady-state WLAs needed for set-
ting limits on toxicity is expressed as:

             &- < criterion  (0.3 TUc and 1.0 TUc)
       Qe  = individual effluent flow (or upstream
       Te  = individual effluent toxicity (in TUas or
             TUcs).Te must be adjusted by a WLA
             factor (a factor used for load allocation
             among multiple sources).
       Qs  = acute or chronic design flow at the
             point of reference below the last source
An example of this calculation  is given in the TSD in
Section 8,  Case Example, on page 71.
Zero Flow Streams
Where intermittent or zero flow streams are present
(design flow equals zero) and the above procedure is
to be used, the effluent toxicity requirements would
be met at the end-of-the-pipe. Regulatory authorities
would have to make a site-specific determination of
the validity of this water quality requirement based
on the  designated use of such  waters. Protection of
such waters from toxic impact due to point source
discharge would typically be a decision made through
the  State  water  quality standards  development
process rather than the  NPDES  permit  issuance
Lake,  Estuary/ and Marine Discharges
As  noted  in  Section 2, lake,  estuary, and  marine
discharge situations require somewhat different ex-
posure assessments than flowing waters. In setting
WLAs  in these situations, the regulatory authority
establishes a  mixing zone and then requires  the
chronic criteria or standard to be met at the edge of
that mixing zone. If near-field control is considered im-
portant, the regulatory authority may need to use the
more complex modeling approach  discussed in the
TSD and other guidance documents. The permit limit
derivation  procedure,  however, is  the same for all

Permit Limit Derivation Procedures

Once the WLA targets are set, the permit limit deriva-
tion process begins. This calculation is performed in-
dependently of water  quality  considerations. The
remaining procedures  seek a  level  of treatment
(governed by permit limits) which is designed to pro-
tect against both acute and chronic instream effects.
Note that it is not correct to say that Maximum Daily'
limits protect against acute toxicity and  Average
Monthly1 limits protect against chronic toxicity. The
limits derivation process calculates both  permit limits
based  on the more stringent of the  two WLAs.
Remember, all remaining calculations are designed to
define  treatment performance  (i.e., effluent quality
targets), not ambient water quality targets.

The procedures below are taken directly from Section
6  of the TSD.  The procedure  itself,  despite  ap-
pearance, is simple. Permit writers should be familiar
with the statistical details of the procedure so that
questions on the process which might be raised dur-
ing permit reviews or hearings can be answered.

  Permit limit derivation has two basic elements:

    •  An effluent  performance level  (Long  Term
       Average [LTA]), that will meet each WLA re-
       quirement is back-calculated. Since two WLA
       requirements are specified based on different
       duration periods (acute  [CMC]  and chronic
       [CCC] protection), two performance levels are

    •  Permit limits  are then derived directly from
       whichever  performance  level  is  more

  step 1: Convert the acute WLA to chronic toxic
     units (TUC) using the acute-chronic ratio {ACR}
     for the effluent (use ACR of 10 where unknown).
     TUa's are multiplied by 10 to get a TUC value.
     It is necessary  to get both toxicity WLAs ex-
     pressed in the same units so that it is possible
     to determine which is limiting and thus which
     value will  drive treatment requirements. This
     step is only needed where toxicity is the control

  step 2: Back-calculate the long term average (LTA)
     that will meet the above WLAs. Since the two
     WLAs are based on different duration periods,
     two sets of equations are used.

    a)  The LTA  for the one-hour or  acute  WLA is
      calculated as follows:

                  LTA  = e'^ + -5°2}

                                                       /i =  In (acute WLA) - Za

                                                       Z =  the Z score for the probability basis for the
                                                         WLA value. Use the 99th percentile occur-
                                                         ance probability (0.01} i.e., Z = 2.326
                                                 NOTE: acute WLA is expressed in TUcs
                                                                 (CV2 + 1)
                                                   Coefficient of
                                                   Variation (CV) = 0.6 unless data are available to
                                                     calculate a CV (use the CV calculation procedure
                                                     on page  53 of the TSD)

                                                     NOTE: 0.6 is  recommended  based  on some
                                                       typical  CV's observed for effluent toxicity. A
                                                       permittee can be required to generate a CV for
                                                       effluent toxicity or a toxicant.

                                                     b) The  LTA for the four-day chronic WLA is
                                                       calculated as follows:
                                                                   LTA = ef^ + -5ff21

                                                       where: Z, CV, and a are as described above
                                                         and  where ji is calculated as follows:

                                                       First,  calculate  the  distribution of average
                                                       values that  will meet the chronic WLA:

                                                       /*4 =  In {chronic WLA)
                                                                      - Z  yin[1
                                                        Second, convert the distribution of average
                                                        values to a distribution of daily values:
                                                                        +  .5 in [1 +  {{eff2- 1)/4|]
1 See the Glossary for definition of these terms.
NOTE: Even if it is assumed  that the 7Q10  was
  acceptable for calculating the WLA, the WLA should
  be applied as a 4-day average for the effluent. This
  recommendation is made because effluent toxicity
  tends to be more variable than streamflow in the
  directions of concern; i.e., toward lower stream
  flows and toward higher effluent toxicity (see page
  54 of the TSD, WLA Output Type  3).

  Step 3: Select the lower LTA. Of  the two WLA
    specifications, the one with the lower LTA is more
    limiting for the plant in question. This toxicity
    value will drive the required treatment to meet the
    permit limits. Two permit limits will be derived in
    the next step from the more limiting of the acute
    or chronic LTA

  step 4: Using the equations below and the LTA and
    CV selected,  derive the  Maximum Daily and
    Average Monthly permit limits. The 95th percen-
    tile is generally used for both permit limits unless
    monitoring will be so frequent (e.g., daily values)
    as to provide sufficient confidence that the true
    performance will be known. In such cases the
    99th percentile may be used. Base the Average
    Monthly permit limit on the monitoring frequency
    {where n  =  required frequency in samples per
    month) described on page  52 of  the TSD.

a) calculate the Maximum Daily permit limit as:

            Maximum Daily  = ete + Zo>

      Z  =  1.645 for the 95th percentile
      Z  =  2.326 for the 99th percentile

      p.  =  In (LTA)  -  .5a2
     a2  =  In (CV2 +  1)

b) calculate the Average Monthly permit limit
  and Z described above) as:
Step 4. Calculate the Maximum Daily and the Average Monthly per-
mit limits for the 95th percentile by multiplying the LTA by the
appropriate multiplier (and CV) in the multiplier table below:
          Average Monthly  =
     Mn  = n +  (a*  -

      a* = In [1  +  ((e°
                                   + Zff
      n  = number of effluent sampling
           observations per month

Simplified Limit Derivation Procedure

To increase the ease of application of this permit limit
derivation procedure. Table 3-1,  was developed. It
provides sets of multipliers which accomplish the
same result  as  the  equations above: translating
steady-state  model outputs into LTAs and deriving
permit limits based on the limiting LTA. The procedure
in Table  3-1 can be  used for permit issuance  on a
routine basis.
Table 3.1. Permit limit derivation process for routine application
         of the procedure described in Subsection 3.1 .A.

Step 1. Convert the acute WLA from TUa to TUC by multiplying the
TUa by an appropriate  acute-chronic ratio  
discharge point or chronic toxicity effects over a larger
area of the receiving water), cannot be determined.

A treatment system will only need to be designed to
meet one level of treatment for effluent toxicity: the
treatment needed to control the most limiting toxic
effect. It is misleading and unnecessary to set acute
toxicity controls as Maximum Daily limits based on the
acute WLA and chronic toxicity controls as Average
Monthly  limits based on the chronic WLA. Such an
approach assumes that acute and chronic toxicity are
different parameters. They are basically  the same
parameter, just measured by toxicity tests of different
duration. The treatment system,  however, will  be
designed to control whichever of these is the more
limiting in terms of toxic effect.

Third, this procedure provides the means to accurately
determine the Average Monthly permit limit based on
the  number  of observations  that  will be taken.
Statistically, the proper monthly permit limit decreases
as the number of samples taken  increases. If the pro-
cedure recommended is not used, the alternative is
to set the Maximum Daily equal to the more stringent
WLA. If monitoring is more frequent than once per
month, the Average Monthly limit should be different
than the  Maximum Daily limit. However, there is no
way to derive the limit short of dividing by a profes-
sional judgment-derived value unless the recommend-
ed procedure is used.

This recommended permit  limit derivation procedure
is applicable to any effluent parameter limited on a
water quality basis, including any individual toxicant
for which standards are available.  Further, where
toxicity-based limits lead to the identification of a
causative agent of effluent toxicity, chemical-specific
limits on those toxicants should be calculated. How-
ever, due to  effluent variability,  the regulatory
authorities may choose to  retain toxicity limits even
where causative agents of effluent toxicity have been
identified and limited.

Other Permit Limit  Derivation Approaches

The recommended permit limit derivation procedure
is not the only procedure which can be used to set
permit limits based on steady  state WLAs.  Some
States and  Regions have successfully used other
approaches to establishing permit limits.  However,
permit writers should be aware of the strengths and
limitations of these different permit limit derivation
procedures. On balance,  the  statistical  approach
described above has sufficient strengths to make it
the recommended procedure.
Another  permit limit procedure is described in  the
TSD, Section 6, page 54. Under this approach, a WLA
(or the most stringent of two WLAs) is used directly
as a Maximum Daily permit  limit.  There are certain
limitations to this approach which are discussed in the
Monitoring Considerations
When setting monitoring requirements for any water
quality-based effluent parameter limited in the permit,
the permit writer  must specify the frequency  of
testing, type of sample, and type of test {acute or
chronic). If the parameter is effluent toxicity, the kind
and number of species to test should also be specified.
  • Frequency
Monitoring frequency is a compromise between need
and cost. Ideally, continuous monitoring would be re-
quired. Such data generation requirements, however,
are prohibitively expensive. A minimum recommend-
ed frequency of testing is once  per month. Higher
monthly monitoring frequency should conform to the
number of observations in) specified in the Permit limit
derivation process. Again, it is suggested that initial
monitoring frequencies be relatively high until contin-
uing compliance allows the reduction of monitoring
  • Sampling
In general, where toxicity limits  are in place, com-
posite samples should be used. Further discussion of
sampling is found in Appendix B in the TSD and in
EPA's standard test methods  (referenced in Appen-
dix C of this manual}.
  • Type of Test (acute or chronic)
It is a misconception to think that only acute toxicity
tests should be used where acute toxicity is limiting
or to monitor for the Maximum Daily permit limits. It
is a similar misconception to think that only chronic
tests should be used where  chronic toxicity were
limiting. The limits are derived  from the more limiting
LTA; so either acute or chronic tests can be used for
compliance  monitoring. It is the  toxicity limit itself
which  influences the choice of test.
For example, limits less than 10 TUC (10% NOEL or
greater] require chronic toxicity testing where the ACR
is 10 or greater for the effluent. An acute toxicity test
would  simply not be sensitive enough to  measure
effluent toxicity at this level (an  LC50 of 100%
effluent or 1 TUa  is the detection  level for acute
tests). Thus, for an effluent where the ACR is 10 and
the NOEL is 20%, an acute test would not measure
the toxicity. Acute toxicity here would be greater than
100%  and beyond the acute detection level. Limits
above this level may be monitored using acute tests.
There is generally  no reason  to  mix two  types  of
monitoring for the same outfall. Doing so will confuse
results  and  complicate assessments for  Average
Monthly limits (for > 1/mo. monitoring).
  • Number of Species to Test
It is essential to specify the test species to be used
in a toxicity  testing requirement.  The actual species
use.d is not as important as the number of species
used (although standardized test species are strongly
encouraged for several reasons: reduced test variabil-

ity; availability of test  organisms,  etc.}.  Different
species will exhibit different sensitivity to an NPDES
effluent. Which species  is the most sensitive to any
effluent will not be known {except  where previous
toxicity testing has consistently identified a sensitive
test species) and so a range of species sensitivities
is analyzed.

Test species sensitivity is important to a permit writer
because species sensitivity varies greatly and the use
of a non-sensitive test species on a particular effluent
may result  in false negative  results and a toxic
discharge  situation will go uncorrected.

To avoid this, it is recommended that a permit writer
require at least three test species initially. Use the test
species specified in EPA's standard methods reference
in Appendix C. It is the best balance between cost and
sufficient toxicity data. Once the sensitive species of
these three species is positively identified,  the other
species can be dropped from the monitoring program.
A permit writer should never allow the use of just one
test species unless the most sensitive test species is
found by testing the toxicity of the effluent.

Another important misconception is that the species
used in this toxicity monitoring program must be resi-
dent species in the receiving water. It is not necessary
to  use resident  species.  In  fact,  it  is strongly
discouraged because it is usually more expensive to
test nonstandardized species. Resident species test
may have  high variability associated with test results
due to a lack of standard methods. However, if the
State water quality standards require a sensitive resi-
dent species, use one resident species together with
two standard test species in the testing program.

The test method(s) for  effluent toxicity testing are
published  by EPA's Office of Research and Develop-
ment and are listed in Appendix C of this document.
The appropriate  toxicity test methods  must  be
specified by the permit writer in the permit to conform
with 40 CFR 122.41 (j). Note that it is not necessary,
nor has it ever been a legal requirement, to have a test
method listed at 40 CFR Part  136 (pursuant to
§ 304{h) of the Clean Water Act)  available to test an
effluent or limit the parameter.

B.  Approach  2 — Set  Monitoring  Re-
    quirements Prior to Setting Permit Limits
The second  approach to implementing the narrative
water  quality standard, involves the use of data
generating monitoring procedures as a first step to set-
ting water quality-based controls for toxics.  If the
regulatory authority believes that data are needed on
effluent  toxicity or  individual  toxicants  in  the
wastewater prior to setting control requirements, then
initial monitoring  must be required.

There are three principal reasons for generating data
prior to setting limits:
  1) to confirm that a permittee exceeds the narrative
     no toxics water quality standard and thus needs
     water quality-based permit limits for toxicants.
  2) to identify a sensitive test species for toxicity
     monitoring purposes.
  3) to generate data on the variability of effluent
It should be noted that it is not necessary to test an
effluent's toxicity in order to set permit limits.
This section is arranged in the following way. First,
two mechanisms for data generation are described,
either a § 308 letter or in the Special Conditions sec-
tion  of the NPDES permit itself. Next, the  data  re-
quirements  recommended for  both  of these
mechanisms  are  discussed.   Then,  the  options
available to the permit writer after data generation are
Under this approach to implementing the narrative
standard, all types of toxic   effects can  be  in-
vestigated. Further, the  permit  writer  could  set
monitoring requirements  for   bioaccumulation  or
human health monitoring and, at the same time, set
limits for aquatic toxicity in  the permit.

Data Generation Mechanisms

   1.  The § 308 letter
      Section 308 of the Clean Water Act authorizes
      EPA and the States to impose monitoring  re-
      quirements on any point  source discharge so
      long as the data generation conforms to the
      criteria of reasonableness. Biological monitor-
      ing is specifically listed in  § 308. The § 308 let-
      ter is perhaps the best means to get toxicity
      information on an effluent.
      An example §308 letter, taken from the TSD,
      is provided in Appendix B. Note that the exam-
      ple requires  both toxicity testing and whole
      effluent bioaccumulation  analyses (the proce-
      dures listed are not necessarily recommended,
      they are only examples).
   2.  NPDES Permit, Special Conditions
      Permits can be  and are routinely issued with
      data generation  requirements described in the
      Special Conditions section of the permit itself.
      They are written to augment the limits imposed
      on other parameters. These testing procedures
      require permittees  to  generate  data  on their
      effluents so that the permit writer  can deter-
      mine if additional permit limits or controls will
      be  necessary to meet  other  statutory  re-
      quirements, such as water quality standards.
      There is a  time  lag associated  with this
An example of a monitoring requirement that can be
placed in the Special Conditions  section of an NPDES
permit is provided in Appendix  A.

Either of these data generation mechanisms should
subsequently result in modification of the NPDES
permit involved if the data generated show unaccep-
table toxicity or other unacceptable toxic effect (see
§122.62(a)(2) of the  NPDES permit regulations).
However, as noted above, it is not necessary to have
toxicity data prior to  setting toxicity-based permit
Recommended Monitoring Procedures

The permit writer has different effluent testing pro-
cedures available for  use  in setting monitoring re-
quirements for permittees. They are presented below
according to the type of toxicity  being investigated.

  • Aquatic life toxicity
    The TSD describes a tiered toxicity testing pro-
    cedure which is designed to assess the poten-
    tial of  an effluent to cause a toxic impact and to
    generate sufficient  data  to set  limits,  if
    necessary.  It allows for decisions to  be made
    regarding toxic impact early in the testing pro-
    cess.  Effluents with low potential for instream
    toxicity are  eliminated early from consideration.

    Those effluents with a higher  potential for  in-
    stream toxicity are required to generate addi-
    tional data so that a decision can be made. These
    procedures, described in Figures  3-1  and 3-2
    can be adopted for use in § 308 letters or in the
    Special Conditions section of the permit. For fur-
    ther discussion  of these procedures, consult Sec-
    tion 3, page 13,  of the TSD.

  • Bioaccumulation
    Monitoring for bioaccumulative chemicals in ef-
    fluents can  take several forms. First, the permit
    writer  can  require ambient  analysis as de-
    scribed in the ASTM standard practice {see Ap-
    pendix C.VI.A). The organisms exposed to the
    effluent are then analyzed for specific chemicals
    by subjecting their tissues to GC/MS  analysis.

    Second, the permit writer can require the permit-
    tee to conduct a bioaccumulation analysis using
    a High Pressure Liquid Chromatograph (HPLC)
    procedure. This procedure will be available from
    EPA's laboratory in Duluth, Minnesota in the near
    future. Effluents with bioaccumulative chemicals
    are then required  to be analyzed for those
    pollutants identified in the HPLC  procedure.

  * Human Health  Analysis
    The TSD recommends three screening tests for
    human health hazards: the Ames test, the mam-
    malian sister chromatid test and the mammalian
    cell chromosomal aberration  test. Currently, one
    of the tests for human health effects is more
    commonly used to analyze  whole effluents  of
    unknown composition: the Ames test. This test
    measures the mutagenic potential of  effluents
    and can be  used as a monitoring and screening
    tool in an NPDES permit. However, interpreta-
    tion of Ames test data for effluents is difficult
    due to the current inability to identify causative
    agents of mutagenic response.

    The Ames test method is described in Appendix

Monitoring frequency for these last two toxic effects
is site-specific. Quarterly sampling and analysis is

Post-Monitoring Procedures

Once any or all of these monitoring requirements are
completed by the permittee, the permitting authority
analyzes the data and determines what follow-up ac-
tion  is  required. Where unacceptable toxicity  to
aquatic  life or human health is identified, the data
should be used to develop water quality-based limita-
tions  and modify the NPDES  permit to meet  the
applicable water quality standard.

  • Writing Permit Limits Based on Monitoring Data

    With the monitoring data generated, a  permit
    writer will  have  the following  information:
    knowledge that the effluent is toxic; identity of
    a sensitive  species; some  idea  of effluent
    variability; and data on the presence of bioac-
    cumulative  chemicals. The human   health
    analysis  recommended above is of  limited scale
    and should be considered a screening tool only.

     — permit limits for aquatic life protection

    The permit limit derivation procedure described
    in Sub-section 3.1 .A which uses a CV and a LTA
    should be used to generate toxicity-based per-
    mit limits.

     — permit limits for bioaccumulative chemicals

    The EPA Criteria  Documents set acceptable
    ambient concentrations  for  some   bioac-
    cumulative toxicants. For other bioaccumulative
    pollutants where no criteria are available but an
    analysis of bioaccumulation has been conducted,
    the identification of a bioaccumulation potential
    should lead to an investigation of what pollu-
    tant(s) in the effluent are  bioaccumulative and
    an analysis of  their bioconcentration  factor
     (BCF). With a BCF available for a pollutant, con-
    trols can be determined. Two approaches  are

    1) where the bioaccumulative pollutants have a
       maximum acceptable tissue  concentration
       available, such as an Acceptable Daily Intake,
       limit derivation is as follows:
       maximum acceptable
       tissue concentration

Rgure 3-1. Recommended procedures for toxicity screening (taken from the TSD).
Recommendations for Whole-Effluent Toxicity Screening
« Individual Dischargers—Compare receiving water flow rate
   (in terms of whatever water quality-based design low flow
   is specified by the State) to average effluent flow rate.
   —If dilution exceeds 10,000 to 1, and there is a reasonably
     rapid mix of the effluent outside of the rapid  initial dilu-
     tion area in the receiving water, then the effluent should
     be given a  low priority for any  further attention.
   —if dilution is less than 10,000 to 1, or mixing is not rapid
     and toxicity within  a plume is of  concern, then toxicity
     screening tests should be performed.
   -If dilution is between  1,000 to  1  and 10,000 to 1, or a
     poorly mixed effluent plume  in a large receiving water
     t>10,000 to 1 dilution) is of concern, conduct acute tox-
     icity screens  as  follows:
      1. Collect four to six effluent samples on one day (grab
         or short term composite), quarterly. Conduct screen-
         ing tests (24-hour) in 100% effluent, using a daphnid
         and a fish, on each sample.
      2. If 50% mortality  or greater is observed  in three
         samples, the potential for toxicity is assumed and fur-
         ther testing is required.

      3. If 50% mortality or greater is observed for two or
         fewer samples, the discharge should be given a low
         priority for further analysis.
   —If dilution is less than 1,000 to 1, conduct chronic toxi-
     city screens (short term chronic tests are recommended)
     as follows:
      1. Collect four to six effluent samples (24-hour com-
         posite) on four to six successive days. Conduct static
         screening tests (seven-day) in 100% effluent, using
         a cladoceran and a fish,  on each sample.
     2.  If a 50% of greater effect is observed between con-
        trols and test organisms, the potential for toxicity is
        assumed and further testing is required.
     3.  If less than 50% effect is observed, the discharge
        should be given a low priority for further analysis.
    Acute tests can be used in these dilution situations, but
    is should be noted that there will be cases where no acute
    toxicity is measured but the effluent is chronically toxic.
  —Where dilution is less than 100 to 1, the use of toxicity-
    testing-based screening procedure is not recommended.
    Screening has already been accomplished through dilution
    analysis.  Even in discharge situations where no toxicity
    is observed in screening tests, the narrow margin between
    effect concentration and available dilution suggests more
    complete effluent toxicity characterization is mandatory.
    If uncertainty factors are applied in a 100 to 1 discharge
    situation, dilution alone would mandate further testing.
    Where very limited dilution is available, it is recom-
    mended that toxicity-testing screening be skipped and the
    discharger  be  required  to  begin DEFINITIVE  DATA
    GENERATION procedures. An example of this situation is
    described in Section 8.
Ambient Toxicity Analysis—Use ambient toxicity analysis to
  identify areas of instream  toxicity associated with specific
  dischargers. This analysis may be most useful when con-
  ducted by the regulatory agency, but dischargers may be
  required to  conduct the tests in conjunction with effluent
  tests.  A systematic plan  for identifying  problem areas is
  recommended. This procedure is useful for multiple source
  discharge situations. The analysis should be conducted con-
  currently with discharge-specific screening and must be done
  at low flow  conditions. A procedure is described in Appen-
  dix C.

Figure 3-2. Recommended toxicity testing procedures (taken from the TSD).
Recommendations for Whole Effluent Definitive Data Generation
Requirements for toxicity analysis are site specific. Recommen-
dations for testing in this section will be presented in terms of
eliminating levels of uncertainty. Analytical procedures will be
presented together with those uncertainty factors which can
be eliminated if those procedures are used. Again, a simple rela-
tionship can be applied to determine whether to require more
data, whether to stop testing and begin the  process of
establishing permit conditions, or  whether to cease analysis
because of  a  wide  margin between toxicity and  the  IWC:
        LC50 or NOEL (in %)
> level of uncertainty
where the level of uncertainty equals the uncertainty factors
multiplied together. It should be stressed that this equation is
used to evaluate the need for additional analysis. The relation-
ship is not the basis for the development of permit limits or con-
ditions. These procedures are discussed in Section 6.
• Initial or "Baseline" Testing
   — Tests—Conduct acute toxicity tests on two species. A fish
     and an invertebrate are recommended.
   —Frequency—Conduct  monthly  on  grab  or  composite
     samples, whichever is appropriate. Where court orders, ad-
     ministrative orders, or legal deadlines require permits to be
     issued within a shorter timef rame, schedule these tests to
     conform with the deadline.
   — Uncertainty factors— 10X to 100X for effluent variability,
     10X for species  sensitivity  variability, and 10X for ACR
     where chronic toxicity is of concern, but no chronic data
     are available. Level of uncertainty  =  1,000 to  10,000.
• Eliminate Effluent Variability Factor
   — Tests—Conduct acute toxicity tests on two species. A fish
     and an invertebrate are recommended.
   —frequency—Examine manufacturing processes, treatment
     plant design and retention times (actual, not design), and
     variability of measured parameters to estimate the variability
     of toxicity. It may be impossible in practical sense to assess
     toxicity variability, but an estimate should be made initial-
     ly. This factor is perhaps the most important source of
     uncertainty in the toxicity assessment process.
     For effluents exhibiting short term (12 to 48 hours) variabili-
     ty, measure the  highest toxicity during a short period of
     duration (one discharge cycle or a 24-hour period). Four to
     six samples (grab or short term composite) taken during this
  period should be tested. Tests are conducted monthly for
  at least one year. Another option is to require a continuously
  pumped, flow-through acute test on each species. This will
  integrate the effects of a variable toxic concentration but
  will not provide a quantification of effluent variability (see
  Appendix B).
  For effluents exhibiting suspected  long term variability,
  schedule testing frequency to conform to expected changes
  (weekly, monthly, seasonal, process changes) in effluent
  composition, if known (see Appendix B). For long term
  variability, a year-long monitoring program may be required
  to determine long term variability.
— Uncertainty Factor— 10X for species sensitivity and 10X
  for ACR where chronic toxicity is of concern but no chronic
  data are available. Level of uncertainty  = 100,
Eliminate Species Sensitivity Factor
— Tests—Conduct acute toxicity tests on a total of three to
  five species. The species should be representative of several
  groups including fish, invertebrates, and plants.
—Frequency— Same as above.
— Uncertainty Factor— 10X where chronic toxicity is of con-
  cern but no chronic data are available.
  Level of uncertainty =  10.
Eliminate ACR Factor
— Tests—Conduct short term chronic toxicity tests on three
  species. Since the available test procedures short enough
  to be practical in effluent  characterization are limited, to
  eliminate the species sensitivity factor where chronic testing
  is mandatory,  the use of Ceriodaphnia, fathead minnow
  growth tests, or short other term chronic tests currently
  available is  recommend.
—Frequency—Same as above.
— Uncertainty Factor—Level of uncertainty equals one if these
  data are generated.
Other Considerations
— Use of Dye Studies—Dye studies are strongly recommended
  for effluent characterization. They are relatively inexpen-
  sive and provide data needed tojilan an assessment of an
  effluent's impact orTChe receiving~water. A dye study should
  be included in any Tier 2 analysis unless mixing is known
  to be rapid and complete. Procedures for conducting a dye
  study in an effluent characterization are described in the
  Lima, Ohio, report [5].

       NOTE: in the absence of a laboratory derived
       BCF, a BCF can be estimated by QSAR, the
       structure/activity relationship  computer
       system. Consult QSAR (described in Appen-
       dix C) for the BCFs of the toxicants identified
       as bioaccumulative.
        Acceptable Effluent
        Concentration (AEC)
  In stream
dilution factor
       where the dilution factor is based on complete
       dilution at a longer duration design flow such
       as 30Q5.

     Next, the AEC is used directly as the WLA for
     the facility. The WLA can then be used in the per-
     mit limit derivation procedure. Again, a CV must
     be estimated  {use 0.6  where  unknown) or
     calculated  from effluent data.

     2) for bioaccumulative chemicals with no  ac-
        ceptable tissue concentration available, a
        conservative control option can be exercised.
        The permitting authority will determine a BCF
        which is not to be exceeded. A BCF of 1000
        or greater might be used as an unacceptable
        degree of bioaccumulation. Any pollutant ex-
        hibiting a BCF of 1000 or greater would be
        given a zero discharge limit. This approach
        assumes that there is no safe ambient con-
        centration for such chemicals and that their
        discharge should be minimized to the fullest
        extent  possible. Application of this option
        could be controversial.

     —permit limits for human health protection
for implementing these numeric water quality stand-
ards in NPDES permits are described.

At present, most State numeric water quality criteria
are expressed as single values which are not to be ex-
ceeded instream. This creates a difficulty for permit
writers who are attempting to use EPA's current water
quality criteria guidance in which two number criteria
and the corresponding biologically-based duration and
frequency recommendations are specified.

There are four types of numeric State  water quality
criteria discussed in this  Sub-section:

  A) two value State water quality criteria
  B| single value State water quality criteria
  C) unspecified  numeric  standards  based  on a
     measurement of the toxicity of any pollutant
  D} numeric water quality values for toxicity.

A.  Deriving Permit Limits Where Two Value
   State  Water Quality  Criteria  Exist for a
                  Some States have begun to use the two number for-
                  mat for their State water quality  criteria. In those
                  States, the difficulties associated  with single value
                  criteria are obviously not  applicable. The  recom-
                  mended permit limit derivation procedure for using
                  these State criteria is the same as the process de-
                  scribed in Sub-section 3.1.A.

                  B. Deriving Permit  Limits  Where  a  Single
                    Value State Water Quality Criterion Exists
                    for a Toxicant
     Where an Ames test {or other acceptable human
     health effects test) shows a positive response
     (indicates mutagenic activity), the permitting
     authority may wish to investigate the specific
     toxicants which can be identified in the effluent.
     Again, since no whole  effluent human  health
     criterion can be established, limits on human
     health impact toxicants must be set on individual
3.2—Implementing Numeric Water Quality

Virtually all  State water quality standards  contain
numeric water quality criteria for one or more tox-
icants. Some State  standards also include numeric
values for toxicity (e.g., effluent shall not exhibit an
LC50 of less than 50%). In this section, procedures
                  Two options are presented for a permit writer con-
                  fronted with a  single, never-to-be-exceeded State

                    Option 1) use the single value State criterion in the
                      steady-state WLA derivation process described
                      in Sub-section 3.1 .A and derive two number per-
                      mit limits based on that single WLA.

                    Option 2) use the single value State criterion in the
                      steady-state WLA derivation procedure describ-
                      ed in Sub-section 3.1. A and place that value in
                      the permit as a Maximum Daily permit limit. The
                      WLA used as a limit in this way should never be
                      used as an Average Monthly permit limit. Doing
                      so would allow averaging of monitoring data and
                      could result in unacceptable excursions of the
                      criterion which would not be violations of the

Option 1 —Use the single State criterion to derive two
  number permit limits
  step 1.  Take the  single  value  criterion  and
    generate a single steady-state WLA using the
    WLA procedure described in Sub-section 3.1 .A.
    Use this WLA as a chronic WLA (use the chronic
    design flow and the chronic mixing zone in the
    WLA calculation).
  step 2.  With the WLA set, follow the permit limit
    derivation procedure described in  Sub-section
    3.1 .A.  The LTA from this WLA is the most limiting
    LTA. Thus, the Maximum Daily and the Average
    Monthly permit limits will be based on this value.

Option 2—Use the single criterion as the Maximum
  Daily permit limit
  step 1.  Establish a WLA for the toxicant using the
    WLA procedure described in Sub-section 3.1.A.
  step 2.  Use this number (it is most often a chronic
    protection value) directly as the Maximum Daily
    permit limit.

If this option is used there are at least two  reasons
for using the single number generated as a Maximum
Daily  only.
First,  steady state WLAs assume the effluent is con-
stant  when in fact  it is usually variable. Variability is
not taken  into account in  this simple  permit  limit
derivation option. Since there will be varying effluent
composition, an Average Monthly limit cannot be
allowed.  If it were, the permittee could  average
monitoring  values (using some effluent values below
the average  and  some  above), comply with the
Average Monthly permit limit, but exceed instream the
water quality criteria the limit is designed to  protect.
Second, monitoring frequency is so low in most cases
that violations of a time-dependent limit (above the
Average  Monthly for four days or longer  but below
the Maximum Daily) will not be detected (you would
need  continuous monitoring to detect  excursions
above the Average Monthly). Setting the limit as a
Maximum Daily limit insures that a violation of this
limit will  be detected as a violation.

C.  Deriving Permit  Limits  Based  on an
   Unspecified Numeric  Value for Effluent

One numeric value for toxics often is found in many
State  water quality standards. It is usually stated as:
   "Any pollutant or combinations of pollutants shall
   not exceed  one-tenth of  the  96-hour  median
   tolerance limit (TLm) or LC50 for any representative
   aquatic species"
This requirement was put into place early in the water
quality standards program development process by
many States. It is the same standard basically as the
whole effluent toxicity criteria described in Section 2
in that it specifies the toxicity to a test species of a
toxicant or combination of toxicants (like an effluent),
whatever that toxicity might be, as the value. It is
termed an "unspecified" value here because that tox-
icity is not known until the toxicity testing  is per-
However, there are several differences which need to
be pointed out. First, any water quality criteria are best
applied as ambient criteria. The above requirement
should follow the mixing zone recommendations in
Sub-section 3.1.A and be applied accordingly.
Second, testing of one representative test species is
insufficient. This  criterion should  be interpreted as
using or requiring the data from at least three species.
The most sensitive  species would be considered
This criterion could be used by the regulatory authority
by  defining  it as  a  toxicity-based  chronic numeric
standard for an effluent (or a toxicant in that effluent
known to cause the  effluent's toxicity).  Then, that
single value chronic numeric criterion is used in the
same procedure described in 3.2.B, where a single
number water quality criterion is used in the WLA and
permit limit development process.

D.  Deriving Permit Limits Based on a Whole
  Effluent Numeric Toxicity  Value

In a few States, a numeric  criterion for toxicity has
been established.  It is usually stated as a  not-to-be-
exceeded value:
  "The discharge shall not exhibit an LC50  of less
  than 50% at the end-of-the-pipe to a representative
  aquatic species"
This value should be used directly in a permit and does
not need a WLA development step since it is an end-
of-the-pipe limitation.  It can be a limit (Maximum Daily)
or a special condition. It may not,  however, provide
protection from chronic toxicity. Therefore, the per-
mitting authority  should consider  adding  a limit on
chronic toxicity in those discharge situations where
chronic toxicity may cause impact {limited dilution
situations). The recommended criterion of 1.0 TUC
can be used as the basis for the chronic toxicity limit.
The narrative toxic "free from" standard would be the
legal basis for this additional chronic limit.

3.3 Recommended Water Quality-Based
Permit Issuance Procedures
A.  Issuance of Permits Involving  Specific
   Toxicity-Based  Discharge Limitations

Once water quality-based permit limits for toxicity are
calculated,  permittees should be required to take all
necessary steps to comply with those limits. An en-

forceable mechanism should be used for achieving the
limitations  which could include  permit conditions.
Administrative Orders, or State authorities as  ap-
propriate. An approach involving the following pro-
cedure is suggested:
   1)  Issue the final permit containing water quality-
      based permit limits and include specific condi-
      tions for achieving compliance with the limits.
   2)  Include language in the Special Conditions sec-
      tion of the permit which sets initial monitoring
      requirements to measure whether or not  the
      limits are exceeded. Initial monitoring should be
      relatively  intensive. Where the limits are  ex-
      ceeded, a Toxicity Reduction Evaluation (TRE)
      is triggered.
   3)  The TRE, as specified in the Special Conditions
      section of the permit, includes  development of
      a study plan for determining how to achieve the
      permit limits and a schedule for implementing
      the TRE procedures. Within a period of time
      (e.g., within one year), attainment of the limits
      should be required. Example language is provid-
      ed in Section  5.
   4}  The  permit  limit derivation calculations,
      monitoring requirements, and  TRE considera-
      tions together with the rationales for each of
      these elements are written into a permit fact
      sheet and included in the issuance of the per-
      mit. It is important to document these factors
      in a fact sheet when issuing a water quality-
      based permit.
It should be noted that limits based on water quality
do not  supercede  technology-based limits if such
limits are more stringent.
Appendix A gives an example of monitoring language
which can be used in monitoring for compliance with
toxicity-based limits.

B. Issuance of Permits Which Do Not Involve
  Specific   Toxicity-Based  Discharge
Another approach to using monitoring data is to re-
quire a permittee to conduct a TRE in the absence of
corresponding toxicity-based permit limits. Under this
approach, a TRE is required through an Administrative
Order, permit condition, or other appropriate mech-
anism upon demonstration that the effluent is unac-
ceptably toxic. The  permittee is required only  to
develop and implement a TRE, not meet toxicity-based
The drawback of this approach is that the absence
of limits restricts the ability of the regulatory authori-
ty to enforce the effluent quality. The regulatory office
is put in the position of trying to determine if the ac-
tions taken by the permittee  are sufficient in the
absence of specific treatment  requirements.

                                   Section 4.  Case Example
Complex Chemicals, Inc. is a hypothetical specialty
chemicals producer in the Gum and Wood Chemicals
industrial  category. The company produces rosin-
based derivatives and other products from the waste
liquors of the kraft pulping process. Approximately 2.5
MGD of effluent (3.9 cfs) is discharged to the Black
River from an aerated lagoon treatment system. The
Black River is designated a "fishable, swimmable"
State water. Current State water quality standards for
toxic pollutants are  limited to  metals and  five
pesticides. No other dischargers are present.

The  Gum and Wood Chemicals  industrial category
(SIC 2861) was analyzed by EPA for possible develop-
ment of  BAT  guidelines.   However,  final BAT
guidelines were not promulgated. No national effluent
guidelines limitations beyond BPT exist.

The  current NPDES permit is up for renewal. Draft
technology-based BAT and BCT limits for the permit
have already been calculated using BPJ for COD, BOD,
TSS, and  Oil and Grease. These BPJ limits are iden-
tical to the BPT limits in  the previous permit and the
company  is currently meeting those BPT limits, the
Company submitted an  application for  renewal and
provided the chemical information required on Form
2C.  Aside from low levels of six metals, the only
priority pollutants detected in the wastestream were
total phenols (390 ppb), methylene chloride (40 ppb),
toluene (17.9 ppb), and di-N-octylphthalate (44 ppb).

The State permitting office decided that an investiga-
tion  into  the  water quality  impact  of Complex
Chemicals was warranted. There were three reasons:
   1) the  effluent is complex, containing numerous
     chemicals of unknown  but potentially high tox-
     icity. No water quality standards were available
     to indicate the toxicity of the pollutants known
     to be present.
   2) no further technology-based controls for toxic
     pollutants were available.
   3) instream concentrations of effluent were high
     enough to cause toxic impact if the effluent ex-
     hibited toxicity. The IWC is 12.5% (dilution fac-
     tor  = 8) at 7Q10.

Therefore, the State instituted the following  process
for investigating the potential water quality impact of
the company and insuring the narrative "no toxics in
toxic amounts" State water quality standard was
being met.
State Toxics Control Procedure
First, a three month effluent toxicity testing program
would be required through the State equivalent of a
§ 308 letter. This letter would require monthly chronic
toxicity testing using three standard test species.
Composite samples (24:hour) would be required from
the plant's normal operating period and Black River
water would be used as the dilution water. The total
of nine tests was estimated to cost $5,000.

Second, if the effluent is shown to  be toxic in this
monitoring program,  toxicity-based permit  limits
would be calculated using Approach 1 as described
in Section 3.1 .A of the Permit Writer's Guide to Water
Quality-Based Permitting for Toxicants. These limits
would be written into Part I of the permit.

Third, the Special Conditions section of the permit
would be modified to include an initial weekly toxic-
ity monitoring requirement, which would extend for
three months, and TRE requirements which would be
triggered upon demonstration of noncompliance with
the final permit limits. The TRE would include a re-
quirement to submit a TRE study plan within 45 days
of demonstrated noncompliance, implementation of
treatment remedies  identified  in  the study which
would be  needed to achieve compliance with final
limits, and a schedule for achieving the final toxicity

If the initial toxicity monitoring showed no effluent
toxicity, the toxicity monitoring requirement would be
reduced to a quarterly toxicity monitoring requirement.

The  permittee would be required to achieve com-
pliance with the final toxicity-based permit  limits
within 270 days of the determination of what controls
are needed to meet the final limits but not longer than
three years after permit issuance.

Permittee's Data Generation Results
Complex  Chemicals tested the  toxicity of their
effluent. As specified in the § 308 letter,  three test
species were used; a daphnid,  a fish, and an algae.
Short-term chronic tests were employed and com-
posite samples were taken. The effluent was found
to be acutely toxic to the daphnids (the NOEL was not
calculated because the test organisms all died after
24 hours of exposure in the lowest effluent concen-
tration tested). The fish and the algae were not sen-
sitive. Repeat testing over the three month period
indicated the toxicity was not a transient problem.

On the basis of these results, the State implemented
the second phase of the permitting process. Toxicity-
based permit limits were calculated using the recom-
mended EPA whole effluent criteria {0.3 TUa and 1.0
TUC)  and steady-state WLA derivation procedures.
The narrative "free from" toxics in toxic amounts was
used as the applicable State standard. Toxicity limits
were  considered necessary because they provided the
legal  basis for the narrative standard, they provided
a target  toxicity reduction, and they provided  the
framework for assessing continued compliance with
the toxicity reduction plan.

Calculation of the WLA  for Effluent Toxicity
The State wasteload  allocation staff was asked to
develop a WLA for effluent toxicity using the equation:

   wasteload allocation <  criterion x dilution factor

The conditions at the site were as follows:

   —no upstream sources present
   —two WLAs  calculated for  acute and chronic
    toxicity using 1Q10  and 7Q10 as  required in
    State standards
   —effluent mixing was complete and rapid

Using the WLA approach  presented in Sub-section
3.1.A of this Guide, two WLAs were calculated:
    WLAa  < 0.3 TUa  x

           < 0.3 TUa  x

           < 0.3 TUa  x

    WLAa  < 2.3 TUa

    WLAC  < 1.0 TUC  x

           < 1.0 TUC  x

           < 1.0 TUC  x
    WLAr  < 8.0 7U.
10.10 + effluent flow
    effluent flow
 26.1 cfs +  3.9 cfs
       3.9 cfs
7Q10 +  effluent flow
    effluent flow
 27.3 cfs +  3.9 cfs
       3.9 cfs
Calculation of Permit Limits
With the two WLAs  calculated,  the results were
used by the permitting authority to develop toxicity-
based permit limits. The procedure described in Sub-
section 3.1 .A was used to calculate the limits. First,
the permit writer changed the acute WLA to chronic
toxic units  so that both WLAs are expressed  in the
same units. Next, the two effluent performance levels
{long term  averages or LTAs) that will meet  these
WLAs were backcalculated. A CV is needed to per-
form the calculation. A CV of 0.6 was used because
it represents the average of the effluent toxicity CVs
available to the State  permit writers.
                            step 1: Convert acute WLA to TUcs

                              2.3 TUa  x ACR (acute/chronic ratio 10)  = 23 TUC

                            step 2: Back-calculate the long term average (LTA)
                             which will meet the above WLAs

                             one-day (acute) LTA was calculated as:

                                     LTA  =  e«* + -5<7''
                                       &  = In  (acute  WLA  in  TUC) — Z
Calculate the Maximum Daily as:

    Maximum Daily  = e1^ + Zo)
                 Z  =  1.645 for the 95th
                 p.  =  In (LTA)  -  .5<72
                    =  1.435 - 0.153
                    =  1.28
                    =  8.9 TUC
  Calculate the Average Monthly  as:

    Average Monthly = e'^ + Zffn>
                  n = number of effluent sampl-
                        ing  observations/month
                        (four is used for this ex-
                        ample  calculation)

                 oj  = In [1 + ((e°2- 1)/4)]
                     = 0.086
Average Monthly =
                     =  1.28 + 0.11
                     =  1.39
                            + Zo4i
                     —  e(1.39 - 0.47)
                     =  6.5  TUC

For illustrative purposes, the calculations above are
conducted using the multiplier tables provided in Table
3-1. If the permitting authority in this case wished
to use a quicker limit derivation process, the  use of
the multiplier tables would be as follows.

  step 1 : Convert the acute WLA from TUa to TUC
    by multiplying the TUa  by the acute-chronic ratio
    {here  10 is used because the  actual ACR  is

            2.3TUa  x  10 =  23.0 TUC

      note: the chronic  WLA is 8.0 TUC

  step  2: Convert the acute and chronic WLAs into
    LTA values by multiplying the WLAs by the
    multipliers in Table 3-1 corresponding to the CV
    of 0.6.

              23.0 TUC X 0.32 = 7.4 TUC
     and      8.0TUC X 0.53 = 4.2 TUC
     where the averaging period is four days (n = 4)

  step  3:  Select  the lower (controlling) LTA

       4.2 TUC is the lower (controlling) LTA
                for this effluent

  step  4:  Calculate the Maximum Daily and Average
    Monthly permit limits by multiplying the controll-
    ing LTA by the multiplier from Table 3-1  cor-
    responding to  a CV of  0.6.
      Maximum Daily permit limit:
              4.2 TUC x  2.13 =  8.9 TUC
      For the Average Monthly four samples  per
      month effluent monitoring is required.
      Average Daily permit limit:
            4.2 TUC x 1.55 =  5.5 TUC

Upon calculation of the permit limits, the permitting
authority issued the final permit with these toxicity-
based limits: Maximum Daily — 8.9 TUC (or NOEL >
11.2%  if  Toxic Units are  converted to percent ef-
fluent);  Average Monthly  =  6.5 TUC (or  NOEL >
1 5.4% if converted to percent effluent). By meeting
these permit limits, the  water quality requirements of
the WLAs would be achieved. Weekly chronic toxicity
testing was set as the compliance monitoring require-
ment using Ceriodaphnia as the test species.

Chronic toxicity tests were required as the toxicity
tests to use  in compliance monitoring in this case
because the permit limits were in the range where if
the TUcs were converted to TUas (using an ACR of
10; thus 8.9 TUC/10 = 0.89 TUa), the value (0.89
TUa)  would  be out of the detection range of  the
acute toxicity test (remember 1.0 TUa is the detec-
tion limit for acute toxicity tests since 1.0 TUa equals
100% effluent).

The toxicity monitoring data had shown the effluent
to be acutely toxic. No LC50 could be calculated since
all test organisms died in less than 24 hours. Chronic
toxicity testing would indicate very high toxicity,
significantly less  than  1% (greater than  100 TUC)
since chronic tests would be more  sensitive than
acute tests.

State's Toxicity Reduction Requirements

The State determined that noncompliance with  the
calculated permit limits  was proved and the company
was required to implement the third phase of toxics
control: submit a toxicity reduction evaluation study
plan within 45 days of issuance of the final permit.
The plan must show how the company was to comply
with the TRE's schedule for attainment of the limits.
The company submitted a  study plan describing  the
steps it would take to meet the toxicity limits. The plan
contained a description of  a fractionation procedure
the company would employ to try to identify causative
agents of  toxicity. If the specific toxicants causing
toxicity could not be positively identified, the company
would attempt to trace individual in-plant processes
to identify the most toxic  wastestreams by bench-
scale toxicity treatabilrty  studies.  Those internal
wastestreams shown to be causing effluent toxicity
would be given separate treatment prior to or in addi-
tion to the biological treatment. These procedures
would lead to compliance with the toxicity-based per-
mit limits  within 270 days of permit issuance.


                      Section 5.  Toxicity Reduction Evaluations  (TREs)
When  toxicity  testing shows  that  a permittee's
discharge contains toxicity at unacceptable levels, the
Regional Office or the State agency with responsibility
for that permit should require the permittee to reduce
toxicity so that no harmful effects occur instream.
This section describes the application of site investiga-
tions (TREs) which use toxicity  testing and physical
and chemical analysis  of an effluent  to zero in on
causative toxicants or on treatment methods that will
reduce effluent toxicity. It describes the legal basis
for requiring permittees to perform TREs and it pro-
vides a general description of how a permittee might
perform a TRE. Several examples of TREs performed
by Regional, State, and EPA Office of  Research and
Development personnel are also provided. The con-
tributions  by these  scientists  and regulators  are
gratefully acknowledged.
Application of TREs and Example Language

Permit writers can ensure that permittees with unac-
ceptably toxic discharges reduce effluent toxicity by
writing toxicity reduction requirements  into  the
Special Conditions section of every permit that in-
cludes toxicity limits,  or by utilizing other legally
binding mechanisms including Administrative Orders
or Consent Decrees.  A  TRE may be used as the
technical basis for a compliance schedule leading to
the attainment  of final limits by a certain date: the
effective date  of  the permit limit,  or to correct
instances of noncompliance after the limit is in effect.
An example of a toxicity reduction requirement that
could be written into the Special Conditions section
of an  NPDES permit follows.  To determine  non-
compliance with the Effluent Toxicity limitations, the
EPA Region or State should apply the principles of its
Enforcement Management System to evaluate the
violation or series of violations. They should consider
the compliance history  of the permittee  and other
aspects  of their  compliance to  determine  an
appropriate response to the violations. The permitting
agency may use additional monitoring inspections,
enforcement actions, compliance schedules or some
combination of these activities to attain compliance
and establish appropriate schedules:

  a.  If the  initial monitoring requirement  de-
     monstrates noncompliance with the  Effluent
     Toxicity limitations in Part I of this permit, the
     permittee shall conduct a Toxicity Reduction
     Evaluation (TRE) to determine how the permit-
     tee can achieve the Effluent Toxicity limitations.

  b.  Within X days  {e.g.,  45 days)  of  the test
     demonstrating noncompliance with the Effluent
     Toxicity limitations, the permittee shall submit
     a TRE study plan detailing what toxicity reduc-
     tion  procedures the  permittee  will employ.
     Within X days (e.g., 270 days) of submittal of
     the study plan, the permittee shall complete im-
     plementation of those measures identified in the
     study as necessary to attain compliance with
     the Effluent Toxicity limitations.

  c.  If, for any reason, the implemented measures do
     not result in compliance with the Effluent Tox-
     icity limitations, the permittee shall continue the
     TRE. The TRE shall not be complete until the per-
     mittee has attained compliance with the Effluent
     Toxicity  limitations  of this permit.

Reducing effluent toxicity to acceptable levels is solely
the permittee's responsibility. The permit writer's only
responsibility is to require permittees to perform TREs
if their effluents are unacceptably toxic. However, if
time permits, the permit writer or compliance section
personnel may be able to  provide the permittee with
assistance that will quicken the permittee's TRE and
make the TRE more efficient. For  example, permit
writers and compliance personnel can choose to pro-
vide  permittees  with  information  on   potential
approaches to TREs (such as this section of the Per-
mit Writer's Guide and the Toxicity Characterization
Procedures Manual that  will be available in  early
1987). They can choose  to identify successful TRE
case examples and suggest that the permittee con-
tact  persons who are familiar with these examples.

It may be  particularly useful  if the permit writer and
other regulatory agency staff review and comment on
the permittee's draft TRE workplan. However, if agen-
cy staff review a draft TRE workplan, they should
make clear to the permittee that their review does not
imply agency approval of the workplan and agency
responsibility for success of the TRE. If the TRE does
not result in reduction of  effluent toxicity to accep-
table levels, the permittee must continue with addi-
tional efforts to reduce toxicity until an acceptable
level of effluent toxicity is achieved.

Legal Basis for TREs

The Clean Water Act and the NPDES regulations pro-
vide a sound basis for requiring NPDES permit holders
to perform TREs. 40 CFR 122.44(d) requires that per-
mits must contain limits that will achieve water quality
standards. Permit limits for toxicity are authorized
under sections 301  and 402 of the Act.
U.S.  EPA has identified TREs as an effective means
to ensure that water quality standards are achieved
because  they focus attention on  toxicity reduction
procedures which will enable the permittee to come
into compliance with its permit limits. In the example
just provided, once the regulatory authority has deter-
mined that permit limits for toxicity have been violated
the permitee is required to submit a TRE plan and a
schedule for attaining compliance with  its permit
limits. Requiring  the submission of such reports and
information is authorized by Section 308 of the Clean
Water Act and analogous provisions under State law.
(Section  308 requires that the permittee provide any
information, make reports,  or maintain monitoring
equipment or methods (including biological monitor-
ing methods} that EPA may reasonably require to set
permit limits and determine compliance with permit
limits and standards. Section 402(b)(2)(A} requires
that approved NPDES States also have this authority.)

Conducting Toxicity Reduction Evaluations
While permittees can choose to use whatever means
they wish to  reduce effluent  toxicity,  U.S. EPA
believes  that it will  be very helpful to have proven,
cost-effective TRE methodologies available to assist
permittees in their evaluations. Thus, the Agency is
participating in a number of TREs and  is sponsoring
research  to  develop  some of the tools that will  be
useful. This section discusses two general approaches
to TREs that have proven successful in a  number of
instances.  Case  examples  are  also  provided for

A TRE is  a step-wise process which combines toxici-
ty testing and analysis of the physical and chemical
characteristics of causative toxicants to zero in on the
toxicants causing effluent toxicity and/or on treatment
methods  which will reduce the effluent toxicity.  In
most  cases, the  process  proceeds  from  simple
assessments that use the quickest, most inexpensive
methods  (e.g.,  pre-chlorination   effluent  toxicity
testing and post-chlorination toxicity testing} to more
complex  analyses (e.g.,  effluent fractionation and
subsequent toxicity testing/chemical identification of
Following  an  initial  characterization  of the
physical/chemical properties of the causative tox-
icants  in the  effluent and  an  assessment of the
variability in these properties, a TRE will utilize two
basic  approaches for identifying  and controlling
effluent toxicity.
In the causative agent approach, investigators iden-
tify the  toxicants  causing or  contributing most
significantly to effluent toxicity. Once likely causative
agents are identified, the  permittee can pursue a
number of options to reduce effluent toxicity. For
example,  the  permittee  can  modify  production
processes to keep the agents out of the wastestream
or  substitute  for  the  raw materials,  solvents,
intermediates and other chemicals that are suspected
of significantly contributing to  effluent toxicity. If
process modification or compound substitution is not
practical,  the  permittee might consider pretreating
wastestreams containing the toxicants. A municipal
permittee may be able to require its industrial users
to cease discharging the causative toxicants in unac-
ceptable quantities. The permittee may find that revis-
ing best  management practices (BMPs)  to  control
spills  and leaks from material storage, loading  and
unloading  areas,  in-plant transfer,  process  and
materials handling areas, sludge and hazardous waste
disposal areas, and plant site runoff is the most cost-
effective way  to keep the causative toxicants out of
the wastestream.

In the toxicity treatability approach investigators
determine the  physical and chemical characterisitics
of the toxicants causing  effluent toxicity and employ
various treatment options, typically in bench scale or
pilot plant studies, to determine which treatment prac-
tices will most cost-effectively reduce effluent tox-
icity.  As with studies of  how to reduce biological
oxygen demand, it is not necessary to know which
particular compounds are causing toxicity to identify
effective treatment options. However, the more com-
pletely the causative toxicants have been character-
ized, the greater the certainty that a cost-effective
treatment option  has not been overlooked.

Based on Regional, State, and permittee experiences
to date, the causative agent identification approach
is the preferred approach. It is more cost-effective to
keep a toxicant out of a wastestream than to treat
wastewater. Even if a set of causative toxicants or
an individual compound causing toxicity  cannot be
identified, the  process wastestream or industrial user
contributing the toxicant  may be identified. Treatment
can then be applied to that wastestream and not the
whole effluent. However,  if the causative toxicants
are highly variable, the toxicity treatability approach
may be the only option  available to the permittee.

Initial Analysis
As  a first step in the TRE, a relatively inexpensive
series of  tests should be performed to determine the
physical/chemical characteristics of the causative tox-
icants in the effluent. To assess the variability in the
magnitude of  effluent toxicity and in the properties
of the causative toxicants, these tests  should be
repeated on a  number of effluent samples. Compar-
ing shifts in the magnitude of effluent toxicity or in

the properties of the causative toxicants to operational
changes may lead to an understanding of what proc-
esses are contributing to effluent toxicity. Variability
assessment also  increases  the liklihood that  all
causative toxicants will be identified. If the causative
agents of toxicity are readily identifiable, these initial
tests may provide a strong indication of what they are.

In TREs conducted to date, investigators have com-
bined physical/chemical manipulation of  an effluent
with short duration toxicity tests {timed lethality tests,
48- or 96-hour static acute tests or short-term chronic
tests) to determine which  broad classes of com-
pounds are contributing most significantly to effluent
toxicity. As a first step, a  timed lethality test (toxic-
ity test of one to eight hours duration in which the end-
point is rapidly induced toxicity} is performed on the
whole effluent sample to  determine whether or not
the effluent is toxic. It may be necessary to perform
definitive acute or short-term chronic tests to deter-
mine toxicity.

If an effluent sample exhibits toxicity, the sample is
split into a number of small volumes. One volume is
used as a control and a separate treatment is applied
to the other volumes.  The treatments used are:
   • filtration;
   • air stripping, under acidic, basic and neutral con-
   • addition of a reducing agent to bind chlorine and
    similar electrophiles;
   • addition of a chelating agent to bind cations; and
   • passing the effluent through a column that ex-
    tracts organics, under acidic, basic and  neutral

The toxicity of the  effluent after  each  of  these
treatments can then be tested and compared to the
toxicity of the control sample. To conserve resources,
simple,  inexpensive procedures are  used wherever
possible. For example, timed lethality tests rather than
definitive toxicity tests should be used in the first tests
to determine which effluent fractions contribute to
effluent toxicity. If one or several fractions  exhibit
substantial toxicity, then definitive tests to confirm
toxicity would only be performed  on these toxic

The toxicity characterization procedures may provide
important information about the treatability  of  an
effluent's toxicity and may also indicate which broad
class(es) of compounds contribute most significantly
to effluent toxicity.  For example, if filtration removes
most of an effluent's toxicity, then the filterable solids
should be closely examined. If, instead, air stripping
significantly reduces the effluent's toxicity, then the
possibility that the sample contains significant levels
of volatile  organics,  ammonia,  hydrogen  sulfide,
hypochlorous acid or other volatile compounds should
be investigated. Toxicity removal by the electrophile
test might  indicate chlorine  toxicity; removal  of
toxicity by the chelating test might indicate that ca-
tionic toxicants such as metals might be responsible
for the effluent's toxicity. If the toxicity in a sample
that has been filtered and air stripped is removed when
the sample is passed through the column that extracts
organics, then soluble organics may be the cause of
the effluent's  toxicity.  If  the  toxicity  decreases
between the initial timed lethality test and the test on
the control sample, the toxicity may be degradable.
It is very important  that the toxicity characterization
tests  be performed  on enough samples to determine
whether or not the causative toxicants in the effluent
vary.  Otherwise, the permittee may pursue expensive
control  options that are  only partially effective in
reducing effluent toxicity.
Many industrial and municipal permittees have the
expertise to characterize the toxicity of their effluents
using in-house staff. To assist permittees who have
unacceptably toxic effluents  and are  therefore  re-
quired to perform  a TRE, EPA is now preparing a
guidance manual describing in detail  how these tests
might be performed.

In some instances, a permittee's staff or consultants
may be very certain that one or several compounds
are causing effluent toxicity, and the permittee may
choose to try a product substitution, a BMP change
or to  additionally treat its effluent without perform-
ing toxicity characterization tests first. In EPA's ex-
perience in resolving toxicity problems, however, the
Agency has found that an effluent's toxicity is often
not caused by  the  toxicants that best professional
judgement initially  identified as the causative tox-

Causative Agent Identification

Generally, the most cost effective approach to reduc-
ing effluent toxicity is to  identify the toxicants that
are causing toxicity  and remove them from the
wastestream. Building on the toxicity characterization
procedures just described, the investigator can employ
additional techniques to separate toxicants and use
short duration toxicity tests to identify  the causative
toxicants.   These  additional  analyses   are  often
necessary because  the toxicity characterization pro-
cedures do not definitively identify all of the causative
For example, if the characterization procedures show
that nonpolar organics consistently contribute most
of the toxicity in an effluent, the investigator could
separate the nonpolar organics  by  their relative
solubilities in a solvent and in  water. Before this
"effluent  fractionation" procedure, a  POTW or  in-
dustrial effluent  may contain hundreds of organic
compounds; determining which of these compounds
cause effluent  toxicity  may be   impossible.  By
separating the organic compounds into effluent frac-
tions  containing smaller,  analytically manageable

numbers of compounds, attention can be focused on
a subset of toxicants, some of which are causing
effluent toxicity.
The first step in the effluent fractionation procedure
is to pass the effluent through a column that extracts
organics. Next, the column is eluted  with solutions
containing  various  concentrations  of water and  a
relatively non-toxic solvent such as methanol. If the
concentration of methanol is kept low in the first elu-
tion volume and increased in each following elution,
more polar organics will elute from the column in the
first several effluent  fractions, while less  polar
organics will stay on the column until elution volumes
with higher methanol concentrations extract them into
solution. At this point, the investigator can perform
simple, modified toxicity tests on the effluent frac-
tions. If one or several of the sample fractions exhibit
considerable toxicity, chemical-specific analyses can
be used to identify  the toxicants in those fractions.
The investigator may need to use  additional tech-
niques to further separate the toxicants. For exam-
ple, if gas chromatography or high performance liquid
chromotography shows fifteen "peaks" in the most
toxic effluent fraction, heating and aerating the frac-
tion may remove the more volatile organics. After this
step, only four compounds may remain in the fraction.
To identify the  causative toxicants, the investigator
should compare the concentrations of the compounds
in the most toxic fractions of various samples of the
effluent to each sample's EC50 concentration. If the
effluent samples are ranked in order of highest EC50
to lowest, the  suspected toxicant's concentrations
should increase in a direct relationship, while the con-
centrations of other compounds will not correlate with
sample toxicity. Figure 6-1 provides an example of
this type of analysis.
                               Literature data on compound toxicity may also  be
                               helpful in identifying causative toxicants. For exam-
                               ple, if the pesticide dichlorvous is identified in toxic
                               samples of  the effluent at concentrations of about
                               60 /ig/liter and its toxicity to the test species is about
                               1  /xg/l, the compound is  probably contributing  to
                               effluent toxicity. In some cases, if literature data  on
                               the compounds suspected of causing toxicity are not
                               available, it may be useful to obtain pure samples of
                               these compounds and perform toxicity tests to deter-
                               mine the concentrations at which these compounds
                               cause effects in  test  organisms. However, the  in-
                               vestigator should be  extremely  cautious in hypo-
                               thesizing that a compound is contributing to toxicity
                               because it is present in high concentrations. Unless
                               a compound is biologically available it cannot cause
                               toxicity. For example, a particular metal may be pre-
                               sent in an effluent at very high concentrations but may
                               be  bound into a complex and contribute minimally to
                               the effluent's toxicity. Confirmatory data is needed
                               such as a statistical relationship between compound
                               concentration and  effluent toxicity  and, perhaps,
                               toxicity tests demonstrating that a test species known
                               to be more or less sensitive to the suspected toxicant
                               than the test species used in the analyses is also more
                               or  less sensitive to toxic  samples of the effluent.
                               Identifying toxicants is an attractive option. If the
                               investigator can identify one or more toxicants, an in-
                               dustrial facility may be  able to substitute a similar but
                               less toxic compound or modify process kinetics  so
                               that less toxicant enters the wastestream. A POTW
                               may be able to use data on the compounds that  in-
                               dustrial users discharge or chemically  analyze  its
                               influent at various points in the collection system to
                               locate one or several industrial users that are discharg-
                               ing the toxicant(s). Assuming  that a violation of the
Figure 6-1 Correlation of Effluent Toxicity with Compound Concentration (Example using organic toxicants detected by GC).

                                                Gas Chromatograph of Toxic Fraction
Sample Collection

Sample EC^

Ai j

POTWs permit provided the impetus for the TRE, the
POTW could then use its authority to prohibit toxicity
pass-through to require the IDs to reduce discharge
of the toxicants to acceptable levels. Identifying one
or several toxicants and eliminating their discharge by
process modifications, compound substitutions  or
altering best management practices may be more cost
effective than trying to remove the compounds from
the wastestream by effluent treatment.

Causative Agent Identification Case Studies
1. Small POTW
A good example of the usefulness of the causative
agent  identification  approach  involves  a  small
municipal  sewage  treatment plant in the State  of
Florida. Routine testing of this plant's effluent by the
Florida  Department  of Environmental  Regulation
showed that the effluent contained more toxicity than
its receiving waters could assimilate.
In an effort to determine the cause of the  plant's tox-
icity, EPA Region IV used a Section 308  letter to re-
quire  the  POTW to  perform  chemical-specific
analyses. For four months, effluent samples were col-
lected  and analyzed for the 126 priority pollutants.
Only ammonia, zinc, nickel and 2-chlorophenol were
present and these were in non-toxic concentrations.
The Regional office requested a TRE be conducted  by
EPA's  Duluth laboratory.  In  Duluth, a step-wise
process was employed to try to identify the causative
agents of  effluent toxicity.  Aeration did  not remove
effluent toxicity, suggesting that toxicity was not due
to volatile toxicants. Next, an organic solvent was
used to remove non-volatile organic compounds from
the effluent sample. After this step, the effluent was
no longer toxic, suggesting that a  non-volatile organic
compound might be causing the effluent's toxicity.
To identify the causative toxicants, the effluent was
analyzed for all organic compounds that could be iden-
tified  using GC/MS techniques.  The scan showed
more than 300 nonpolar organic compounds present.
Because it would be impossible  to identify each  of
these compounds  and determine which ones were
present in toxic concentrations, chemical separation
techniques were performed to split the effluent into
fractions.  Each fraction could then be tested for tox-
icity. After solid-phase extraction with a C-18 column
and elution of the column with varying concentrations
of methanol, the resulting effluent fractions were
tested for  toxicity. One fraction was consistently
toxic. A chemical-specific GC scan of this fraction
revealed 36 separate compounds. Of the five com-
pounds present  in highest  concentration, one (the
pesticide Diazinon), was present in concentrations
sufficient to cause toxicity.
In subsequent effluent samples, Diazinon, while not
the only source of toxicity  in the effluent, has con-
sistently been present.
2. Industrial Discharger
Another example of the utility of the causative agent
identification approach involves a textile manufacturer
in the Middle-Atlantic states which had a highly toxic
discharge. In this case, investigators first filtered an
aliquot of an effluent sample that was acutely toxic
to the test species, Ceriodaphnia dubia, to remove
suspended solids and compared its toxicity to an un-
filtered aliquot. No difference in toxicity was found.
Another  sample aliquot  was  filtered and passed
through  a solid phase extraction column and  the
sorbed chemicals were eluted using a number of con-
centrations of methonal in water.

Toxicity was not retained on the column but passed
through, suggesting the toxicants were water solu-
ble and  not associated with  the solids.  EDTA,  a
chelating agent, was then added to  an  aliquot of
100% effluent. After the addition of EDTA, no acute
toxicity was found in the sample, suggesting that
metals were the cause of toxicity. AA analyses of the
effluent sample revealed zinc concentrations in excess
of 1000 jug/liter; all other metals were below analytical
detection limits.

Next,  investigators  analyzed  additional  effluent
samples  collected  on different days.  This step was
taken because, even though zinc was present in  suf-
ficiently high concentrations to cause toxicity, the  zinc
might be bound  into  a  complex and biologically
unavailable. In these additional samples, zinc concen-
trations correlated with effluent toxicity. As further
confirmation,  toxicity tests  were performed using
fathead minnows,  a test species which is  markedly
less sensitive  to zinc than Ceriodaphnia dubia.  The
samples were  not toxic to fathead minnows, consist-
ent with the suspicion that zinc was the major toxi-
cant. As a final confirmation, zinc was added to  one
mildly toxic sample to raise its  zinc concentration to
the level present in a much more toxic sample.  The
toxicity of the two samples  were almost  identical.

The correlation of toxicity to zinc concentration, the
lesser sensitivity of fathead minnows to the effluent
and the  agreement between  spiked and  unspiked
samples  all confirm zinc as the toxicant. However, to
definitively  prove  that zinc  is  the toxicant in  the
discharger's effluent, a greater number of effluent
samples should be measured for both toxicity and  zinc
concentrations to establish a statistically significant
correlation between the zinc  and  toxicity meas-
urements. The State and the  discharger are  now
negotiating a testing plan.

3. Industrial Discharger
A third example demonstrates  how causative agent
identification can be followed by chemical substitu-
tions to reduce effluent toxicity. In North Carolina, the
State Division  of Environmental Manangement (DEM)
conducted  a toxicity screening test  using Daphnia

pulex as the test species on a hosiery manufacturer
and found toxicity was greater than the lowest test
concentration (LC50<5%). A second screening test
showed even greater toxicity (LC50<2.5%).

As a result of these initial tests, the DEM conducted
an onsite flow-though test and calculated a 96-hour
LC50  of 1 %  for fathead  minnows.  At the facility
design flow and the stream's 7Q10, the instream
waste concentration of the effluent was calculated
to be 22%,  indicating that the toxic effects on the
receiving stream were, in all liklihood, severe. Benthic
macro in vertebrate testing conducted at the site con-
firmed the toxic effects of the discharge: below the
point of discharge, there was a large decrease in taxa
richness and total  numbers of individuals.

A survey of the manufacturing process revealed the
use of a biocide for bacterial control with the active
ingredients  bis(tri-n-butyltin)oxide  (TBTO)  and
5-chloro-2-(2,4-dichlorophenoxy}phenol.  Subsequent
testing showed that the biocide  and chlorine con-
tributed most significantly to the  effluent toxicity:
     12 hour


Washwater (sock) with
Washwater without biocide
Washwater with biocide,
chlorine removed
Washwater without
biocide, chlorine removed
A 96-hour flow-through test using fathead minnows
was performed at another hosiery mill that also used
the biocide but did not chlorinate. At this plant, the
LC50 was 0.88%. Analysis of the wastewater for
TBTO  and 5-chloro-2-(2,4-dichlorophenoxy)phenol
showed  concentrations  of 260 /ig/liter and  4.3
/ig/liter, respectively. U.S. EPA has not  published
aquatic toxicity guidelines for these compounds, so
acute toxicity data provided by the manufacturers and
obtained from literature were reviewed. These data
indicated that TBTO was present at acutely toxic
levels;  5-chloro-2-(2,4-dichlorophenoxy)phenol was

As  a result of these studies, the State worked with
the manufacturers to identify less toxic subsitutes for
the biocide. After these substitutes were introduced,
toxicity was reduced. At the first facility, the average
LC5Q has risen from 11 % before the substitution to
28%. The facility is now discharging to a POTW. At
the second facility, the average LC5o has risen from
4% to  58%.

Toxicity Treatability
The toxicity treatability approach also builds on the
Phase I physical/chemical characterization of effluent
                                  toxicity. Typically, permittees will want to use bench-
                                  scale or pilot plant tests to determine what changes
                                  in effluent treatment may reduce toxicity to accept-
                                  able levels. For example, if the toxicity characteriza-
                                  tion procedure in which a reducing agent was added
                                  to effluent samples consistently reduced effluent tox-
                                  icity, the  permittee may want to consider adding a
                                  dechlorination unit to the effluent treatment train.
The effluent treatment option may be particularly at-
tractive to permittees with a small volume discharge
and permittees that can track toxicity back to one or
several  processes but  cannot identify a  specific
causitive toxicant. However, investigators should be
very cautious in efforts to track toxicity back to proc-
ess lines: the toxicity in process wastewaters that
demonstrate extreme toxicity prior to treatment may
be very treatable by the plant's treatment train, while
the "mildly" toxic effluent from another process may
resist treatment and  be  responsible for much of the
toxicity in the final effluent. If tracking toxicity back
to process wastestreams is attempted, treatment that
closely mimics the treatment provided by the plant's
treatment train must be applied to the wastestream
before toxicity testing.

While substantially altering effluent treatment may be
expensive for a POTW or other facility with a large
wastewater stream that is unacceptably toxic, in-
creasing effluent treatment may be the only practical
approach if the physical and chemical properties of
an effluent's toxicity are highly variable. Similar to
BOD reduction, treatment must be identified that will
consistently  reduce  toxicity, no  matter what  the
source of the toxicity.

If it can  be  determined that organic pollutants are
causing effluent toxicity, switching from batch to con-
tinuous discharge may be a cost effective treatment
option for indirect dischargers or direct dischargers
who utilize biological treatment. Adding a holding tank
and metering flow may  allow the biological popula-
tion to acclimate to the toxicants.
                                   EPA's Office of Research and Development is examin-
                                   ing the effectiveness of various treatment options in
                                   removing effluent toxicity and is evaluating tests that
                                   may help POTWs determine which industrial users are
                                   contributing to interference or toxicity pass through.
                                   In the meantime, many consultants and industrial and
                                   municipal plant employees can perform bench scale
                                   and pilot plant treatment tests and are learning (or
                                   already know) how to do the toxicity tests needed to
                                   evaluate the effectiveness of treatment options in
                                   removing toxicity.


A. Example Special Conditions Permit Language
  —  Permit Language for Toxicity-based Permit Limits Monitoring
  —  Permit Language for Data Generation Monitoring
B. Example §308 Letter
C. Overview of Selected  Available Tools


                        Appendix A
                Example Special Conditions
                     Permit Language
Permit Language for Toxicity-Based Permit Limits Monitoring
Used when the permit contains toxicity-based permit limits
Permit Language for Characterizing Effluent Toxicity
Used when the permit is employed  as the vehicle for requiring data

                                       Special Conditions
   Compliance Blomonitoring Requirements for Part I Toxicity-Based Permit Limitations
The permittee shall perform toxicity tests, as describ-
ed below, on the discharge from outfall(s)	.
1.  The permittee shall initiate the following series of
   tests within	days of the effective date of this
   Part to evaluate wastewater toxicity. Such testing
   will determine if an appropriately dilute effluent
   sample affects  the  survival,  reproduction,  or
   growth of the test organisms. All  tests will be
   conducted  on 24-hour  composite samples  of
   	% final effluent. A minimum of four replicates
   will be used in each  of the following tests. The
   Student's t  test shall  be used to determine
   whether differences in control  and  effluent  data
   are significant.
  a.  The  permittee shall conduct a  7-day  Cerio-
     daphnia survival  and  reproduction  test on
     samples of  	%  final  effluent (diluted by
     appropriate  control water).  Toxicity will be
     demonstrated if there is a statistically significant
     difference at the 95% confidence level  in sur-
     vival or growth between Ceriodaphnia exposed
     to an appropriate control water and	% final
     effluent. All test solutions shall be renewed
     daily. If, in any control, more than 20% of the
     test  organisms die, that test  (control  and
     effluent) shall be repeated.
  b.  The  permittee  shall conduct a 7-day fathead
     minnow larval survival  and growth test on
     samples of  	%  final  effluent (diluted by
     appropriate  control water).  Toxicity will be
     demonstrated if there is a statistically significant
     difference at the 95% confidence level in sur-
     vival or growth between Pimephales promelas
     exposed to an  appropriate  control water and
     	% final effluent. All test solutions shall be
     renewed daily. If, in any control, more than 20%
     of the test organism die, that test (control and
     effluent) shall be repeated.

2. The toxicity tests specified in Paragraph (1) above,
   shall be conducted once every month for a period
   of one year following initiation of the tests and
   once every 6 months thereafter for the duration
   of the permit. Results shall be reported according
   to  EPA/600/4-85/014,  Section   10,  Report
   Preparation, and shall be submitted to EPA with
   the monthly discharge monitoring report. If any
   one test indicates the presence of toxicity, another
   confirmatory chronic toxicity  test using the
   specified methodology and same test species shall
   be conducted within 2 weeks.

3. All test organisms, procedures  and  quality
   assurance criteria used shall be in accordance with
   Short-term Methods for Estimating the Chronic
   Toxicity  of  Effluents and Receiving Waters to
   Freshwater Organisms, Section 13; Ceriodaphnia
   Survival and Reproduction Test Method 1002.0,
   Section  11;  Fathead  Minnow  (Pimephales
   promelas) Larval Survival and Growth Test Method
   1000.0, EPA/600/4-85/014. The selection of an
   appropriate control water for the toxicity tests
   shall be submitted to EPA for review and approval
   prior to use.

                                       Special Conditions
         Chronic Toxicity Testing Requirement for Characterizing Effluent Toxicity
The  permittee  shall  perform  toxicity testing,  as
described below, on the discharge from outfall(s)	.
1.  The permittee shall initiate the following series of
   tests within 90 days of the effective date of this
   Part to evaluate wastewater toxicity. Such testing
   will determine if an appropriately dilute effluent
   sample affects the survival, growth, or reproduc-
   tion of the test species. All tests will be conducted
   on 24-hour composite samples. A minimun of four
   replicates will be used in each  of the following
   tests. The Student's t test shall be used to deter-
   mine whether differences in control and effluent
   data are signficant.
  a.  the permittee shall conduct a 7-day Ceriodaphnia
     survival and reproduction toxicity test on the
     final effluent  diluted  by  appropriate control
     water. Toxicity will be demonstrated if there is
     a statistically significant difference at the 95 per-
     cent confidence level in survival or reproduction
     between Ceriodaphnia exposed to an appro-
     priate control water and  the final effluent.  All
     test  solutions  shall  be  renewed  using  an
     approved renewal schedule. If, in any control,
     more than 20% of the test organisms  die, that
     test  shall be repeated.
  b.  the  permittee shall conduct a 7-day fathead
     minnow survival and growth toxicity test on the
     final effluent  diluted  by appropriate control
     water. Toxicity will be demonstrated if there is
     a statistically significant difference at the 95 per
  cent  confidence level  in survival  or  growth
  between  fathead  minnows exposed  to  an
  appropriate control water and the final effluent.
  All test solutions shall be renewed using an
  approved renewal schedule. If, in any control,
  more than 20% of the test organisms die, that
  test shall be repeated.

The toxicity tests specified in paragraph (1} above,
shall be conducted once per month for a period of
one year following initiation of the tests and once
every six months thereafter for the duration of the
permit. Results  shall  be reported according to
EPA/600/4-85/014, Section 10  Report Prepara-
tion, and shall be submitted to EPA with  the
monthly Discharge Monitoring Report. If any  one
test indicates  the effluent  is  toxic, another
confirmatory chronic toxicity test using the same
species and the same  methodology shall  be
conducted within one week.

All test species, procedures, and quality assurance
criteria used shall be in accordance  with Short
Term Methods for Estimating the Chronic Toxicity
of Effluents  and Receiving Waters to  Freshwater
Organisms, Section 13; Ceriodaphnia Survival  and
Reproduction Test Method 1002.0; Section  11;
Fathead  Minnow (Pimephales prome/as) Larval
Survival and Growth Test Method 1000.0, EPA
600/4-85/04.  The selection of an  appropriate
control water  for all toxicity tests shall  be sub-
mitted to EPA for review and approval prior to use.


    Appendix B

Example § 308 Letter


                        WASHINGTON. D.C. 20460

Ms. Ann Powell
Plant Manager
Chemico Corporation
Anytown, USA
Dear Ms. Powell:
                                           RE: NPDES No. XX000123
                                               Chemico Anytown
     The U. S. Environmental Protection Agency (EPA) is preparing
to reissue the NPDES permit for the Chemico Anytown facility.
Preliminary analyses conducted by EPA and water quality monitoring
conducted by the  [State Water Control Authority!  indicate that the
discharges from the Chemico Anytown facility may be violating
State water quality standards.  In order for EPA to fulfill its
responsibilities under the Clean Water Act, 33 u.S.C.  1251, et seq.,
additional information regarding the nature of the discharges from
the Chemico facility is required.

     Therefore, you are hereby required (1) to perform the sampling
and analysis programs described below, (2) to maintain for possible
use by EPA all records regarding plant operations during the sampling
wmcn couj.a resu.it in tfie judicial imposition of civil or crimir
penalties.  Please refer to the attachment to this letter for a
summary of our authority and your rights to confidential treatment
of certain information.
Effluent Toxicity Characterization Study -

minimum the following test organisms: Pimepttales prgrnglas  (fatnead
minnow) from one hour to 30 days old and paphhia" spp. (water flea)
from zero to 24 hours old.  Results shall be expressed as  the 48-hr
LC50 with the 95 percent confidence interval.  Dilution water used
in the analysis may be either receiving water collected immediately
upstream of the discharge or another source of clean water such as
the water used for culturing organisms.  However, the analysis of
at least one sample shall be replicated for each test organism
using the other source of dilution water.

     Toxicity analyses shall be conducted in accordance with
Methods for Measuring the Acute Toxicity of Effluents^to Aquatic
        s EPA-600/4-7B-012, Revised July 1978.Each~sample shall
        analyzed for the parameters currently limited in the

NPDES permit for outfall 001 and the following parameters;
chlorine, cyanide, total dissolved solids, chloride, and ammonia.
Associated production levels, ancillary activities, and the
performance of the treatment systems shall be recorded for each
wastewater sampling.

     The results of the toxicity analysis together with the
required chemical analyses and plant operations information shall
be provided for the first two samples within 45 days of initiation
of the sampling program and for subsequent samples every 30 days

Bioaccumulative Pollutant Characterization Study -

     In order to identify potentially bioaccumulative pollutants
discharged from Outfall 001, Cheraico shall analyze effluent
samples as described below.  Wastewater samples for analysis
shall be composited over 24 hours once per month for a period of
four months commencing within 30 days of receipt of this letter.

     High performance liquid chromatography techniques shall be
used to separate the sample constituents.  Methods shall conform
to EPA draft method CG-1410 Partition Coefficient (n^pctanol/
water) Estimation by Liquid Chromatography Test Guideline^
DSF.Pft^  office o£ Toxic Substances,  Washington DC [copy attached) .
All chemicals corresponding to log KQW greater than 4 shall be
collected and analyzed using gas chromatography/mass spectroscopy.
The identification and quantification of pollutants shall conform
to EPA proposed method 625 (44 PR 69540).  Alternate methods may
be substituted with prior approval by EPA.  In addition to
measuring the priority pollutants, a reasonable effort shall be
made to identify and quantify no fewer than the 20 largest non-
priority pollutant peaks on the total ion plot (reconstructed
chromatogram) except that peaks less than 10 times the peak to
peak background noise need not be identified.

      Identification of  the non-priority pollutant peaks  shall  be
attempted by  reference  to the  EPA/NIH  computerized library of
mass  spectra,  with visual confirmation  by an experienced analyst;
quantification may be an order of magnitude estimate based upon
the  response  of an internal standard.   If more  than 20 peaks
exceed 10 times the background noise,  those not identified shall
be reported by relative retention time  and magnitude based upon
the  response  of an internal standard.

      The results of the analysis together with  the total ion
plots and description of the analysis  shall be  provided  within
t5 days of each sample  collection-

      If you have any questions please contact Kermit Riter at
213-456-7890.   Thank you for your cooperation.
Sincerely yours.
Water Division

cc:  [state contacts]
 require the ounce or operator of any point source to  (I) establish and

 maintain such monitoring equipment and methods (including where appropriate,
 biological monitoring methods), (iv) sample such effluents...and (v) provide

 his authorized representative, upon presentation of his credentials, shall

 to and copy any records...and sample any effluents.—•

 federal prosecution undec IB U.S.c. 1001 and that this or any other failure to
 comply with the requirements of Section 308 as requested by U.S. EPA nay

 fiatf: Act, which provides for specified civil and/or  criminal penalties.


 tr.S.  EPA tabulations concerning confidentiality and treatnent of business
 infornation are contained in 40 CFR Part 2, Subpart a.  Information nay not be
 withheld fron the Administrator or his authorized representative because you

 accordingly, if disclosure would divulge methods or processes entitled to
 protection as trade secrets (33 O.S.C. 1318(b) and 18 D.S.C. 1905), eict.pt
 that  effluent data (as defined in 40 CFR 2.302(a)(2M may not be considered by
 O.S.  EPA as confidential.

 such  claing is specified In 40 CFR 2.203(b).  In the  event that a request is
 made  for release of infornation covered by your claim of confidentiality or

 information is entitled to confidential treatment, notice will be provided to

 confidentiality is made when information is furnished to U.S. EPA, any
 information submitted to the Agency nay be made available to the public
 without prior notice to you.

       the Paperwork Reduction ftct of 19BO, 44 O.s.C. chapter 35.

            Appendix C

Overview of Selected Available Tools

                                            Appendix C
                             Overview of Selected Available Tools
This appendix summarizes some of the important tools
that can be used to identify and control unacceptable
toxicity in effluents.
For whole effluent characterization, methods are de-
scribed which:
  • determine toxicity to aquatic organisms;
  • screen for bioconcentrating materials;
  • screen for mutagenicity;  and
  • determine the physical/chemical properties and
    variability of causative toxicants.
In a number of situations, the permit writer or the per-
mittee will want to assess the effects of individual
compounds in an effluent. This section describes pro-
cedures for:
  • reviewing existing toxicity data and screening
    compounds for which little toxicity information
    is available;
  " determining  bioconcentration  factors;  and
  • screening for mutagenicity.
In many cases, water quality  analysts will need to pro-
vide permit writers with effluent concentrations that
will ensure that acceptable instream concentrations
of pollutants or of toxicity are not exceeded. Methods
for analyzing mixing conditions and an alternative pro-
cedure for calculating steady-state modeling design
flows are described. References for a number of im-
portant waste load allocation manuals are provided.
Figure C-1 presents an overview of the tools dis-
cussed in this Appendix.
    Whole Effluent
Toxicity Testing   —Chronic Tests
               —Acute Tests
BJoaccumulation   —Whole effluent HPLC
Human Health    —Ames Test
               —Other Mutagenic
                 Screening Tests
Toxicity Reduction —Phase I Toxicity
Evaluations        Characterization
	           Procedures Manual
Toxicity Data


Human Health
-Log P
 Compound HPLC
-Ames Test
-Other Mutagenic
 Screening Tests
                                       Mixing Zone
                                    IV. Calculating
                                       Design Flows
                                     Dye Studies

                                     Steady-State Design Flow Manual
V. Other Waste Load Allocation Manuals
Figure C-1. Overview of Selected Available Tools

1. Whole Effluent Toxicity Testing Procedures
and Data Bases

A. Short Term Chronic Tests

   1. Freshwater Species
   Reference: Short-Term Methods for Estimating the
   Chronic Toxicity of Effluents and Receiving Waters
   to Freshwater Organisms, EPA/600/4-85/014,
   December 1985.

   To Obtain: National Technical Information Service,
   5285 Port  Royal Road, Springfield, VA 22161
   ({703}  487-4650).  NTIS  Document   No.:
   PB-86158474. Cost: $16.95.

   Use: Quick,  reliable  and  inexpensive tests for
   measuring the chronic toxicity of effluents to one
   fish, one  invertebrate and  one  algae species.
   The permit  writer should reference this manual in
   permits that include testing requirements and/or
   limits for chronic toxicity {e.g., "the permittee shall
   determine compliance with chronic toxicity  limits
   by using the fathead  minnow  growth test and
   the  Ceriodaphnia test described  in Short-Term
   Methods . . .").  If the  permit writer chooses not
   to reference the manual in the permit, the writer
   should, at least, informally recommend that the per-
   mittee  use  these methods. The manual contains
   QA/QC procedures.

   Cost: Each  test may  cost from  $300 to $2,200,
   in addition to sample collection costs, depending
   on the test  design  and the contractor's  price

   Cautions: Chronic tests may be more costly than
   acute tests.

  2. Marine/Estuarine Species
  Reference:  Aquatic  Toxicity  Testing  Seminar
  Manual,  Marine/Estuarine  Effluent Toxicity
  Workshop, Narragansett, Rhode Island, October
  16-17, 1985. Final Draft  Guidance Manual for
  Rapid  Chronic Toxicity Tests  on  Effluents and
  Receiving Waters with Larval In/and Silversides
  (Menid/a beryllina). ERL-Narragansett Contribution
  No. 792.
  To Obtain:  Stephen Bugbee,  Permits Division
  (EN-336), Office of Water Enforcement and Per-
  mits,  401 M St.  SW, Washington, D.C.  20460,
  FTS 475-9515,  (202) 475-9515
  Use:  Quick, reliable  and inexpensive tests  for
  measuring the chronic toxicity of effluents to two
  fish, two invertebrate and one macroalgae species.
  These tests are  applicable  to  marine discharge
  situations where the permitting authority wishes to
  approximate effluent/receiving water interactions
  and so the dilution water used in the test itself is
  saline. To protect aquatic life,  the permit writer
  should reference this manual in permits that include
  testing requirements and/or limits for chronic tox-
  icity (e.g., "the permittee shall determine com-
  pliance with chronic toxicity limits by using  the
  sheepshead minnow test,  the  Champia  parvula
  (seaweed) test and the Mysidopsis bah/a test
  described in Aquatic Toxicity Testing . . ."). If the
  permit writer chooses not to reference the manual
  in the permit, the writer should, at least, informal-
  ly recommend  that the  permittee  use  these
  methods.  The manual contains QA/QC procedures.
  Cost: Each test may cost from $300 to $2,200,
  depending on the test design and the contractor's
  price  schedule.
  Cautions: ERL—Narragansett has not validated
  these methods. Few contract labs have experience
  with these tests.

B. Acute Tests for Freshwater and Marine/Estuarine
  1. Organism Tests
  Reference: Methods for Measuring the Acute Tox-
  icity  of  Effluents to  Freshwater  and Marine
  Organisms,  Third  Edition   EPA/600/4-85/013,
  March 1985.
  To Obtain: ORD Publications Office, Center for En-
  vironmental Research Information, 26 St. Clair
  Street, Cincinnati, Ohio 45268,  (513) 569-7562,
  FTS 684-7562.
  Use:  Quick, reliable and inexpensive tests for
  measuring the acute toxicity  of  effluents. The
  guidance manual lists 23 recommended freshwater
  and 28 marine fish and invertebrate test species.
  Test species geographic distribution, life cycle, tax-
  onomy and  culture methods are provided for one
  freshwater invertebrate (daphnia), one freshwater
  fish (fathead minnow),  one marine invertebrate
  (mysid) and one marine fish (silversides). When pro-
  tection of aquatic life from acute toxicity is war-
  ranted, the  permit  writer should reference this
  manual in permits that include testing requirements
  and/or limits for acute toxicity (e.g., "the permit-
  tee shall determine compliance with acute toxicity
  limits using  the test  procedures  described  in
  Methods for Measuring and the following  test
  species . . ."}. If the permit writer chooses not to
  reference the manual in the permit,  the writer
  should, at least, informally recommend that the per-
  mittee use these methods. The  manual contains
  QA/QC procedures.
  Cost: Each test may cost from $150 to  $950,
  depending on the test design (for example, whether
  the contractor runs a static-renewal test in the lab
  or performs a flow-through test  on  site) and the
  contractor's price schedule.
  Cautions: While acute tests are generally cheaper
  and quicker than chronic tests, they cannot predict
  the long-term effects to  organisms of exposure to
  sub-acute concentrations of effluents. Acute tests
  should be used primarily to predict the effects of
  discharges to  relatively high dilution  capacity
  receiving waters and discharges which cause short-
  term  high concentration exposure of  effluent  to
  aquatic organisms.

  2. Microbial tests
  Reference: Contact manufacturers' representatives
  To Obtain:  Contact manufacturers'  represen-

  Use:  A number of  microbial toxicity tests  are
  available. For example,  Beckman Instruments of-
  fers Microtox, Polybac Corporation offers Polytox
  and Oregenics Limited offers the Toxi-Chromotest.
  Some regulatory authorities have used  microbial
  tests as an inexpensive tool to test for effluent tox-
  icity. In  some cities, microbial  tests  have been
  employed to assess influent toxicity to POTWs both
  to track  toxicity sources and  to  estimate  in-
  terference potential.
  Cost: Contact manufacturers representatives.

  Cautions: Like any test organisms, microbial strains
  have profiles of sensitivity to the toxicants in a
  complex effluent. Some regulatory authorities are
  unwilling to use microbial tests  as  the basis for
  writing permit limits on effluent toxicity, or for re-
  quiring installation of expensive  pollution control
  technology. With some tests, effluents must  be
  altered (salinated) prior  to testing.

C. Computerized Complex Effluent Toxicity Data Base
  Reference: CETIS: Complex Effluent Toxicity Infor-
  mation System Data Encoding, Entry and Retrieval

  To Obtain: Ann Pilli, CETIS Coordinator, EPA En-
  vironmental Research Laboratory, 6201 Congdon
  Blvd., Duluth MN 55804, (218) 720-5714, FTS

  Use: The CETIS Retrieval system allows permit
  writers  to review the results of toxicity tests on
  complex effluents nationwide. Among the standard
  reports  available, permit writers may request tox-
  icity testing data by industry, geographic area and
  by effluent treatment. The user can design and im-
  plement specific retrievals. Statistical Applications
  Software (SAS) format ouput is available if the per-
  mit writer wants to perform further analysis  using
  SAS. Because the CETIS program submits a batch
  request, it may take several hours or overnight to
  generate reports unless the user selects a high
  priority  job number.

  Cost: The user must have an account with the Na-
  tional Computer Center  (FTS 629-7862; (919)
  541-7862). If standard CETIS  reports and job
  priority  are requested, CETIS is very inexpensive.
  Printing and mailing costs ($2.00 per thousand
  lines) are usually the most costly element of the job.
  For example, a recent request for all toxicity infor-
  mation on facilities from every SIC code in the  State
  of Ohio cost $18.50. A request for all toxicity in-
  formation on steam electric generating facilites na-
  tionwide cost less than $4.00.

  Cautions: Reporters submit data to CETIS volun-
  tarily. Much more toxicity data is being generated
  nationwide than appears in the CETIS data base.
  Generalizations about types of dischargers which
  will exhibit toxicity should be made with care  since
  toxicity problems can be very site-specific.

D. Toxicity Reduction  Evaluation Methods
  Reference: Draft  Toxicity Reduction Evaluation
  Methods, Phase I: Characterization of Effluent Tox-
  icity, January 1987.

  To Obtain: Permits Division (EN-336), U.S.  EPA,
  401 M  Street, S.W.  20460.

  Use: Methods manual for a relatively inexpensive
  and quick battery of tests designed to provide in-
  formation on the physical/chemical nature of the
  causative toxicant(s) in an industrial or  municipal
  effluent  and the variability  associated with the
  causative toxicants. Each test in the battery neu-
  tralizes  or renders biologically unavailable a dif-
  ferent group of toxicants. Aquatic organisms are
  used as "detectors" of toxicity. By repeating the
  battery  of tests on a number of effluent samples
  collected over a period of time, information on the
  type of causative toxicants and the variability in the
  cause of effluent toxicity is  gained.

  Cost:  Based on very limited  data, it appears that
  each battery of tests will cost the discharger about
   $ 1,500 (i.e., about $1,500 per sample). Total cost
   will be a function of causative toxicant variability
   and effluent complexity. Complex effluents with
   highly variable causative toxicants may be costly
   to characterize.

   Cautions: While the discharger can be required to
   reduce  toxicity to acceptable levels, permitting
   authorities cannot specify  that  dischargers use
   these methods to evaluate  their effluents.
II. Whole Effluent Bioaccumulation Tests
  Techniques under development by Dr. Gil Veith,
III. Whole Effluent Mutagenicity Test

A. Ames Test
   Reference: Interim Procedures for Conducting the
   Salmonella/Microsomal Mutagenicity Assay (Ames
   Test),  EPA/600/4-82/068, March  1983.

   To Obtain: Dr. Lewellyn Williams, Environmental
   Monitoring Systems Laboratory, P.O. Box 1 5027,
   Las Vegas, NV 89114, (702) 798-21338,  FTS

   Use:  Screening  tool for effluent  point  muta-
   genicity. Where employed, the Ames test is typical-
   ly used as one of a number of screening tests for
   unacceptable  effluent effects.  Positive results
   should trigger a close, chemical-specific examina-
   tion of the effluent for mutagenic compounds.

   Cost: A single test  may cost from several hundred
   to several thousand dollars.

   Cautions:  The degree  of correlation  between
   positive Ames test results and mutagenicity in other
   species including humans  and aquatic organisms
   is unclear. For some classes of compounds the cor-
   relation between the Ames test and mammalian cell
   mutagenicity is quite good. For some other classes
   of compounds (for  example, the anilines), the cor-
   relation is poor. The group of compounds causing
   postive Ames test results  and the group of com-
   pounds considered to be potential carcinogenic by
   U.S. EPA are not mutually inclusive.

B. Other Whole Effluent Mutagenetic Screening Tests
   The Organization for Economic Cooperation and
   Development  has  prepared test  methodologies
   (mammalian sister chromatid exchange  test and
   mammalian cell chromosomal aberration test) for
   determining clastogenic activity. The  Office of
   Pesticide  Programs  and  the  Office  of Toxic
   Substances have prepared documents based on the
   OECD guidelines  (References 18 and 19 in Section
   1 of the TSD).

IV. Chemical-Specific Toxicity Data Bases

A. Aquatic Information Retrieval Toxicity Data Base

   Reference: AQUIRE: Aquatic Information Retrieval
   Toxicity Data Base Project Description, Guidelines
   and Procedures, EPA/600/8-84/021 June 1984.
   Chemical Information Systems, Inc.  (CIS) supplies
   a streamlined user's manual when the user signs
   an  agreement for access to the chemical infor-

   To Obtain: ORD Publications Office,  Center for En-
   vironmental Research Information, 26  St. Clair
   Street, Cincinnati, Ohio 45268, (513)  569-7562,
   FTS 684-7562. For the CIS Manual and user's con-
   tract, contact Ms. Laurie Donaldson, CIS Inc., 7215
   York  Road, Baltimore, MD 21212,  (SOO)-CIS-
   USER or (301)  247-8737.

   Use: The AQUIRE data base provides a  comprehen-
   sive compilation of over  68,000 aquatic toxicity
   tests on more than 4,179  individual chemicals.
   AQUIRE  staff  review   scientific  publications
   worldwide for test data. Acute, sublethal and bioac-
   cumulation effects data are reported. For  individual
   compounds known to be present in an effluent, per-
   mit writers can use AQUIRE to examine available
   toxicity data and determine if the compounds are
   present in concentrations which could potentially
   cause effuent toxicity.

   Cost: To access AQUIRE, the user must subscribe
   to the Chemical Information System ($300  per
   year). AQUIRE costs $55 per on-line hour. EPA per-
   mit writers may contact Mary  Patterson (FTS
   382-5929) of the Information Management and
   Services Division to set up an account. State agen-
   cy permit writers should communicate directly with

   Cautions: Investigators  must use caution in at-
   tempting to relate the concentration of a compound
   in an effluent to effluent toxicity. Compounds that
   are present in seemingly toxic concentrations may
   be biologically unavailable.

B.  Quantitative  Structural Activity Relationship
(QSAR) System

   Reference: User Manual for QSAR System, Insti-
   tute for Biological and Chemical Process  Analysis,
   Montana State  University, October, 1986.

   To  Obtain: Frank Culver, Institute  for Biological
   and Chemical Process Analysis, Montana State
   University, Bozeman,  Montana 59717,  (406)
   994-5387. User Manual  contain the forms needed
   to apply for a user's number.

   Use: Tens of thousands of chemicals  have not been
   studied for their environmental effects and fate.
   When the permit writer must decide whether to set
  limits for a compound for which little aquatic tox-
  icity or fate information is available or must decide
  whether to require further testing by a permittee,
  the writer can use QSAR as a screening tool to
  predict chemical properties including partition and
  persistence in the environment, bioaccumulation
  and toxicity to certain aquatic organisms. QSAR
  will also screen for mutagenic functional groups.
  Under current discount rates, OSAR also provides
  EPA and state environmental regulatory agencies
  with inexpensive access to AQUIRE, a comprehen-
  sive data base of aquatic toxicity tests on individual

  Cost: $ 100 per hour on-line. U.S. EPA users receive
  a 80% discount. Other federal  agency users and
  state agencies receive a 50% discount. Hard copies
  of QSAR data are expensive. As an alternative to
  requesting printouts from QSAR, the permit writer
  may print information off their terminal screen.

  Cautions: At this time, the QSAR predictive model
  should not be used to write permit limits without
  confirmatory  information.

V. Chemical-Specific Bioaccumulation Tests

A. American Society of Testing and Materials
  Reference: "Standard  Practice  for Conducting
  Bioconcentration Tests with Fishes and Saltwater
  Bivalve Molluscs," Designation E 1022-84, 1986
  Annual  Book of ASTM Standards,  vol. 11.04,
  Publication Code Number (PCN): 01-110485-48,
  April 1985.

  To Obtain: ASTM, Customer Service, 1916 Race
  Street, Philadelphia, PA 19103, (215) 299-5400,
  FTS 299-5400. It is not necessary to order the An-
  nual Book of Standards. A copy of Standard Prac-
  tice E 1022-84 costs $8.00.

  Use: Where protection against bioaccumulating
  compounds is warranted and individual, potential-
  ly bioaccumulative compounds in the effluent can
  be identified, the permit writer should consider re-
  quiring the permittee to perform bioconcentration
  tests on these compounds  using the methods
  described in Standard Practice  E  1022-4.

  Cost: The ASTM bioconcentration test is very time-
  consuming and costly.

  Cautions:  Some  techniques described in the
  method were developed for tests on nonionizable
  organic chemicals and may not apply to ionizable
  or inorganic  compounds.  The  bioaccumulation
  potential of many nonionizable organic compounds
  can  be  predicted much  more  cheaply  using
  physical-chemcial properties such as the octanol-
  water correlation coefficient ("log P").  For exam-
  ple, QSAR (See IV. B.} uses log P to  determine
  bioconcentration factors.

B. Log P (Octonol-Water Correlation Coefficient)
   Reference: See Item IV.B., User Manual for QSAR
   System.  QSAR uses the log P model  to predict
   bioconcentration factors and is  a  simple,  cost-
   effective tool for predicting the bioconcentration
   potential of individual compounds.

VI. Chemical Specific Mutagenicity Screening
A. Ames Test—See Item III.A., Interim Procedures for
   Conducting the Salmonella/Microsomal Mutagenic-
   ity Assay (Ames Test)
B. Mammalian Sister Chromatid Exchange Test and
   Mammalian Cell Chromosomal Aberration Test-
   See References 18 and 19 in Section 1 of the TSD.

   VII. Mixing  Zone Analyses
   A. Dye Studies
   1. Mixing in Rivers and Lakes
   References: Measurement of Mixing Charac-
   teristics of the Missouri River Between Sioux City,
   Iowa and Plattsmouth, Nebraska.  U.S. Geological
   Survey Water-Supply  Paper  1899-G, 1970.
   Methods for Predicting Dispersion Coefficients in
   Natural Streams,   with  Applications  to Lower
   Reaches  of the Green  and Duwamish  Rivers,
   Washington. U.S. Geological Survey Professional
   Paper 582-A, 1968.
   To Obtain: U.S. Geological Survey,  Books and Open
   File  Reports —Federal Center,  Building 41,  Box
   25425,   Denver,   Colorado  80225,   (303)
   236-7476, FTS 776-7476.
   Use: Describes how to conduct tracer  studies to
   determine concentration gradients within the mix-
   ing zone of a wastewater discharge to a freshwater
   Cost: Each publication costs $1.75. The cost of
   performing a dye study on a lake  or  river will de-
   pend on the complexity of the hydrodynamics in the
   site-specific situation.
   Cautions: Dye studies can only be performed on ex-
   isting dischargers. Mixing zone models must be us-
   ed to predict expected dilution contours within the
   mixing zone of a proposed discharger.
   2. Mixing in Estuaries
   References: "A Study of Tidal Dispersion in the
   Potomac River." Water Resources Research, Vol.
   2,  1966.  "Fluorescent-Tracer  Studies of  an
   Estuary." Journal of the Water Pollution Control
   Federation, Vol. 38, 1966.
   To Obtain: Copies of these  journals  are available
   in many libraries.
   Use: Describes how to conduct tracer  studies to
   determine concentration gradients within the mix-
   ing zone of a wastewater discharge to an estuary.
  Cost: Small fee for reproducing the journal articles.
  The cost of performing a tracer study on an estuary
  will  depend on  the  complexitiy of  the  hydro-
  dynamics  in the  site-specific situation.
  Cautions: Dye studies can only be performed on ex-
  isting dischargers.  Mixing zone  models must be
  used to predict expected dilution contours within
  the mixing zone  of a proposed discharger.

B. Computer Models for Freshwater, Marine and
Estaurine Discharges

  Reference:  Initial Mixing  Characteristics of
  Municipal  Ocean Discharges. Volumes  I  and II.
  EPA/600/3-85/073 a and b, November 1985.

  To Obtain: State and Regional EPA  Staff can ob-
  tain copies of the reference manual and most of
  EPA's mixing zone models (except PDS and  PDSM)
  from  Bryan  Coleman at EPA's  Marine Science
  Center in Newport,  Oregon (503) 867-4035). Re-
  quests for models  must be  accompanied by an
  IBM-PC compatible diskette which will be  return-
  ed with the  model  source codes copied onto the
  disc. State and Regional EPA staff who would like
  to obtain the PDS and PDSM models should send
  an IBM-PC  compatible diskette  to  David  Eng of
  EPA's Permits Division in Washington, DC (202)
  475-9522), FTS 475-9522.

  Representatives of non-governmental organizations
  seeking the  referenced manual and models must
  purchase them from the National Technical Infor-
  mation Service, 5285 Port Royal Road,  Springfield,
  VA 22161 ((703) 487-4650). The  NTIS Publica-
  tion Numbers for the manuals and  PC programs are:
  Volume I  (PB86-137478);  Volume  II (PB86-
  137460);  and  IBM-PC compatible  diskette
  Use: The referenced manual provides instructions
  on how to  use the mixing zone models available on
  the PC diskette.  All the  models  can be used for
  rivers, lakes and estauries as well as oceans. In
  general, these models require the following input
  data: discharge depth, effluent flow  rates, density
  of effluent, density gradients in the receiving water,
  ambient current  speed and direction,  and outfall
  characteristics (port size, spacing, and orientation).
  Model output includes the dimensions of the plume
  at each integration step, time of travel to points
  along the  plume  centerline, and the average dilu-
  tion at each point.

  Cost: The  manual and the models are provided free
  of charge  to all government users. NTIS charges
  non-government  users $16.95 for  the Volume  I
  manual, $  16.95 for the Volume II manual and $75
  for the IBM-PC compatible diskette containing the
  model source codes.

  Cautions:  Each mixing zone model represents an
  idealization of actual field conditions. Users should

  carefully consider whether the underlying model
  assumptions are valid for the situation being model-
  ed. Mixing zone models should be used to predict
  expected dilution contours during critical design
  flow conditions.

VIII. Calculating Steady-State Design Flows
  Reference:  Technical Guidance Manual for Perfor-
  ming Waste Load Allocations. Book VI: Design Con-
  ditions,  Chapter  /-Stream Design Flow for
  Steady-State Modeling, Office of Water Regula-
  tions and Standards, Washington, DC, and Office
  of  Research  and Development  Environmental
  Research Laboratory, Duluth, Minnesota, August
  To Obtain: Hiranmay Biswas, Monitoring  Data and
  Support Division (WH-553), U.S. EPA, 401 M St.,
  SW, Washington, DC, 20460, (202) 382-7012,
  FTS 382-8012.
  Use: Describes  and compares two methods that
  can be used to calculate stream design flows to be
  used in steady state modelling for any pollutant or
  effluent for which a two-number water quality
  criterion for  the protection  of  aquatic  life  is
  available. One method is the hydrologically-based
  method. The other method is a biologically-based
  method developed by the Office of Research and
  Cautions: State-specified design flows will preempt
  any design flow recommended  in this  guidance
  unless the state chooses to use either of these two

IX.  References for Waste Load  Allocation
A. Identifying Waters Needing Water Quality-Based
   Controls for Toxics (draft). Monitoring  and Data
   Support Division, Office of Water Regulations and
   Standards, U.S. EPA, December 1986.

B. Technical Guidance Manual for Performing Waste
   Load Allocations.  Book II: Streams and Rivers,
   Chapter 3—Toxic Substances, Monitoring and
   Data  Support Division, Office of Water Re-
   gulations   and   Standards,   U.S.    EPA,
   EPA/400/4-84/022, June 1984.

C. Technical Guidance  Manual  for Performing
   Wasteload Allocations. Book IV: Lakes, Reservoirs
   and Impoundments, Chapter 3— Toxic Substances
   Impact, Monitoring and Data Support Division, Of-
   fice of Water Regulations and  Standards, U.S.
   EPA, 440/4-87/002, December 1986.

D. Technical Guidance Manual for Performing Waste
   Load Allocations: Simplified Analytical Method for
   Determining  NPDES  Effluent  Limitations  for
   POTWs Discharging into  Low-Flow  Streams,
   Monitoring and Data Support Division, Office of
   Water  Regulations and Standards, U.S. EPA,
   September 1980.

E. Technical Guidance Manual for Performing Waste
   Load Allocations. Book VII: Permit Averaging
   Periods, Monitoring and Data Support Division, Of-
   fice of Water Regulations and  Standards, U.S.
   EPA, EPA/440/4-84/023, June 1984.

F. Technical Guidance Manual for Performing Waste
   Load Allocations. Book VIII. A Screening Pro-
   cedure for Toxic and Conventional Pollutants in
   Surface and  Ground Water—Parts I and II, En-
   vironmental  Research Laboratory, Athens, GA,
   EPA/600/6-85/002a/002b/002c, September