United States
Environmental Protection
Agency
Office of Water
Washington DC 20460
DRAFT
Technical
Support Document
for Water Quality-based
Toxics Control

-------
United States
Environmental Protection
Agency
Office of Water
Washington, DC
DRAFT
REVISED
TECHNICAL SUPPORT DOCUMENT FOR
WATER QUALITY-BASED TOXICS CONTROL
April 1990
Office of Water Enforcement and Permits
Office of Water Regulations and Standards
U.S. Environmental Protection Agency
Washington, D.C. 20460

-------
Foreword
The U.S. Environmental Protection Agency (EPA) and the State pollution
control agencies have been charged with enforcing the laws regarding pollution of
the natural environment. Environmental pollution is an urgent and continuing
problem and, consequently, the laws grant considerable discretion to the control
authorities to define environmental goals and develop the means to attain them.
Establishing environmentally protective levels and incorporating them in a
decision making process entails a considerable amount of scientific knowledge
and judgment. One area where scientific knowledge is rapidly changing concerns
the discharge of toxic pollutants to the Nation's surface waters.
This document provides technical guidance for assessing and regulating the
discharge of toxic substances to the waters of the United States. It was issued in
support of EPA regulations policy initiatives involving the application of
biological and chemical assessment techniques to control toxic pollution. The
recommendations contained in this document are not mandatory and are intended
to be suggestions for approaching problems which tend to be complex and site-
specific.
This document is expected to be revised periodically to reflect advances in
this rapidly evolving area. Comments from users will be welcomed.
iii

-------
Table of Contents
Foreword 			iii
Table of Contents	iv
Executive Summary		ix
List of Abbreviations		xii
Glossary	xvi
Introduction	 xxiv
CHAPTER 1. APPROACHES TO WATER QUALITY-BASED TOXICS
CONTROL		1
1.1	Introduction		1
1.2	Chemical-Specific Approach For Aquatic Life Protection		2
1.2.1	Correlation of Chemical-specific Measurements to
Actual Receiving Water Impacts 		3
1.2.2	Chemical-Specific Analytical Method Precision		3
1.3	Whole-Effluent Approach For Aquatic Life Protection		9
1.3.1	Correlation of Whole Effluent Toxicity Measurements
to Actual Receiving Water Impact 		12
1.3.2	Toxicity Test Method Precision 		18
1.3.3	Considerations Involved in the Implementation of the
Whole Effluent Toxicity Approach		23
1.4	Biocriteria/Bioassessment and Biosurvey Approach for
Aquatic Life Protection		30
1.4.1	Use of Biosurveys and Bioassessments in Water
Quality-based Toxics Control		31
1.4.2	Conducting Biosurveys 		32
1.5	Integration of the Whole-EfTluent, Chemical-Specific and
Bioassessment Approaches 		33
1.5.1	Advantages and Disadvantages of the Chemical-
Specific Approach		34
1.5.2	Advantages and Disadvantages of the Whole Effluent
Approach		34
1.5.3	Advantages and Disadvantages of the Bioassessment
Approach		36
1.5.4	Other Factors Influencing Water Quality-based Toxics
Control 		37
1.6	Human Health Protection		42
1.6.1 Types of Health Effects		42
iv

-------
CHAPTER 2. WATER QUALITY CRITERIA AND STANDARDS		50
2.1	Introduction		50
2.1.1	Overview of Regulatory Considerations 		50
2.1.2	Water Quality Standards and State Toxics Control
Programs 		52
2.2	General Considerations 		53
2.2.1	Magnitude, Duration, and Frequency		53
2.2.2	Mixing Zones 		55
23 Water Quality for Aquatic Life Protection 		56
2.3.1	Development Process for Criteria 		56
2.3.2	Magnitude for Single Chemicals 		56
2.3.3	Magnitude for Whole-effluent Toxicity 		57
2.3.4	Duration for Single Chemicals and Whole-effluent
Toxicity 		59
2.3.5	Frequency for Single Chemicals and Whole-effluent
Toxicity 		60
2.4 Water Quality Criteria for Human Health Protection		62
2.4.1	Overview 		 .	62
2.4.2	Magnitude and Duration 		62
2.4.3	Human Exposure Considerations		63
2.4.4	Fish Consumption Values 			63
2.4.5	Bioaccumulation Consideration		64
2.4.6	Updating Human Health Criteria and Generating
AACs Using IRIS		64
2.4.7	Calculating AACs for Non-Carcinogens 		67
2.4.8	Calculating AACs for Carcinogens		68
2.4.9	Deriving Dose Factors in the Absence of IRIS Values	70
2.4.10	Deriving Acceptable Tissue Concentrations for Monitoring
Fish Tissue		70
2.5. Biocriteria/Biological Criteria		71
2.5.1 Development of Biocriteria		71
2.5.1 Regulatory Bases for Biocriteria 		72
2.6 Sediment Criteria		73
2.6.1	Current Developments in Sediment Criteria		73
2.6.2	Approach to Sediment Criteria Development		73
2.6.3	Application of Sediment Criteria		74
2.6.5 Potential Uses 		74
2.6.5 Sediment Criteria Status 		75
CHAPTER 3. EFFLUENT CHARACTERIZATION		77
3.1	Introduction		77
3.1.1 Background for Toxic Effects Assessments on Aquatic
Life and Human Health 		77
3.2	Assessment of Aquatic Life Effects		79
3.2.1 General Considerations in Effluent Characterization ...	79
v

-------
3.2.2	Determining the Need for a Limitation Without
Effluent Testing Data		80
3.2.3	Determining the Need for a Limitation With Effluent
Data 		85
3.2.4	Recommendations for Whole Effluent Toxicity
Data Generation		89
3.2.5	Effluent Characterization for Specific Chemicals		99
3.3 Assessment of Human Health Effects 		^
CHAPTER 4. EXPOSURE AND WASTELOAD ALLOCATION- .	121
4.1	Introduction		121
4.2	Total Maximum Daily Load (TMDL) and Wasteload
Allocations (WLA) 		123
4.2.1	TMDLs 		123
4.2.2	Wasteload Allocation Schemes 		126
4.3	Incompletely Mixed, Discharge-Receiving Water Situations . . .	126
4.3.1	Determination of Boundaries		129
4.3.2	Prevention of Lethal Conditions for Aquatic Life	131
4.3.3	Prevention of Bioaccumulation Problems for Human
Health		133
4.4	Mixing Zone Analyses		133
4.4.1	General Recommendations for Outfall Design 		134
4.4.2	Critical Design Periods for Waterbodies		136
4.4.3	General Recommendations for Tracer Studies 		138
4.4.4	Discharge-Induced Mixing		142
4.4.5	Ambient-Induced Mixing		145
4.5	Completely Mixed Discharge-Receiving Water Situations	148
4.5.1	Wasteload Modeling Techniques 	148
4.5.2	Calculating the Allowable Effluent Concentration
Distribution and the Return Period 	158
4.5.3	General Recommendations for Model Selection		159
4.5.4	Specific Model Recommendations	160
4.5.5	Effluent Toxicity Modeling 		164
4.6	Human Health	169
4.6.1	Human Health Considerations	169
4.6.2	Determining the Wasteload Allocation for Human
Health Toxicants		169
CHAPTER 5. PERMIT REQUIREMENTS		182
5.1	Introduction	182
5.1.1 Regulatory Requirements 		182
5.2	Basic Principles of Effluent Variability 	183
5.2.1	Variations in Effluent Quality	183
5.2.2	Statistical Parameters and Relationship to Permit
Limits 		184
vi

-------
5.3	Ensuring Consistency with the Wasteload Allocation	188
5.3.1	Statistical Considerations of WLAs 		188
5.3.2	Types of Water Quality Models and Model Outputs . . .	189
5.4	Permit Limit Derivation 	192
5.4.1	Permit Limit Derivation from Single Steady State
Model Output		192
5.4.2	Permit Limit Derivation from Two Value Steady State
Outputs for Acute and Chronic Protection	193
5.4.3	Permit Limit Derivation from Dynamic Model
Outputs 		197
5.4.4	Special Permitting Applications	200
5.4.5	Other Approaches	202
5.5	Special Considerations In Use of Statistical Permit Limit
Derivation Techniques	204
5.5.1	Effect of Changes on Statistical Parameters on Permit
Limits 	204
5.5.2	Coefficient of Variation	209
5.5.3	Number of Samples 	209
5.5.4	Probability of Exceedence	210
5.6	Permit Documentation 	211
5.7	Expressing Limitations and Developing Monitoring
Requirements	211
5.7.1	Pollution Prevention/Energy Conservation
Considerations 	212
5.7.2	Considerations in the Use of Chemical Specific Limits .	213
5.7.3	Special Considerations in the use of Whole Effluent
Toxicity Limits 	213
5.7.4	Selection of Monitoring Frequencies 	215
5.8	Toxicity Reduction Evaluations 	216
5.8.1	TRE Guidance Documents 	217
5.8.2	Recommended Approach for Conducting TREs 	217
5.8.3	Circumstances Warranting a TRE 	224
5.8.4	Mechanisms for Requiring TREs	226
CHAPTER 6. COMPLIANCE MONITORING AND ENFORCEMENT ...	231
6.1	Overview 	231
6.2	Permit Requirements 	231
6.2.1 Permit Considerations	232
6.3	Compliance Monitoring	233
6.3.1	Self-monitoring Reports 	233
6.3.2	Discharge Monitoring Report/Quality Assurance 	234
6.3.3	Inspections	235
6.4	Violation Review 	235
6.5	Enforcement	237
6.5.1 Enforcement Precedence 	238
6.6	Reporting of Violations 	239
vii

-------
INDEX
241
APPENDICES	245
Appendix A-l
Appendix A-2
Appendix A-3
Appendix B-l
Appendix B-2
Appendix B-3
Appendix B-4
Appendix B-5
Appendix B-6
Appendix C
Appendix D
Appendix E
Appendix F
Appendix G
Appendix H
Toxicity Test Precision Data
Effluent Variability Data
Acute to Chronic Ratio Data
Summary of Clean Water Act Provisions
Policies for Toxics Control
New Regulations for Toxics Control
Whole Effluent Toxicity Permitting Principles and
Enforcement Strategy
Quality Control Fact Sheets
Case Decisions on Whole Effluent
Toxicity
Ambient Toxicity Testing and Data Analysis
Duration and Frequency
Lognormal Distribution and Permit Limit Derivations
Sampling
Biological Indicator Approach to Human Health Toxics
Control
Reference Dose: Description and Use
viii

-------
Executive Summary
The revised Technical Support Document (TSD) for Water Quality-based
Toxics Control provides States and Regions with an encyclopedia of procedures
and guidance to be used in the water quality-based control of toxic pollutants. It
presents recommendations to regulatory authorities faced with the task of
controlling the discharge of toxic pollutants to the nation's waters. The document
provides guidance for each step in the water quality-based toxics control process
from effluent characterization to compliance monitoring. Human health issues
are no longer addressed in a separate chapter, but are discussed throughout the
document. The overall approach taken in this revised document is to provide
additional explanations and rationales based on accumulated experience and data
for the various recommendations which were made in the original TSD. The
following is a brief synopsis of the guidance provided in the TSD.
Approaches to Water Quality-based Toxics Control
EPA's surface water toxics control regulation (54 FR 23868 (June 2, 1989)
established specific requirements that the "integrated"approach be used in water
quality-based toxics control. The "integrated"approach consist of both whole-
effluent and chemical specific approaches as means of protecting aquatic life and
human health. As techniques are made available for the implementation of
biocriteria, they too should be integrated into the water quality-based toxics
control, thus creating a triad of approaches: whole-effluent, chemical specific and
biological assessments. Each approach has its limitations and thus, exclusive use
of either approach cannot ensure maximum protection of aquatic life and human
health. The advantages/disadvantages of each approach and how the integrated
approach creates an effective toxics control program is discussed in the text.
The whole effluent approach to toxics control involves the use of toxicity
tests and a criterion for the parameter "toxicity" to assess and control the
aggregate toxicity of effluents. New references and information in support of the
whole effluent toxicity assessment and control approach have been included in
Chapter 1 and associated appendices (e.g., precision data, justifications for acute-
chronic ratio (ACR) recommendations, information on analytical variability in
toxicity testing). The chemical-specific approach to aquatic life toxics control
relies on numeric water quality criteria in State standards to assess and control
specific toxicants individually.
Water Quality Standards and Criteria
Where specific numerical criteria for a chemical or biological parameter
(such as toxicity) are absent, compliance with water quality standards must be
based on the general narrative criteria and on protection of the designated uses.
For many pollutants, EPA's recommended criteria may be used, or criteria may
ix

-------
be developed using data from IRIS, or data on the toxicological effects of the
pollutant found either in the literature or required of a discharger.
Aquatic impacts occur not only from the magnitude of a pollutant, but also
from the duration and frequency with which criteria are exceeded. EPA's
recommended aquatic life criteria for both individual toxicants and whole-effluent
toxicity are specified as two numbers: the criteria continuous concentration
(CCC), applied as a four-day average concentration; and the criteria maximum
concentration (CMC) applied as a one-hour average concentration. The
frequency with which criteria are allowed to be exceeded depends on site-specific
factors as explained in the text.
EPA's recommended criteria for whole-effluent toxicity are as follows: to
protect aquatic life against chronic effects, the ambient toxicity should not exceed
1.0 chronic toxic unit (TUc) to the most sensitive of at least three different test
species. For protection against acute effects, the ambient toxicity should not
exceed 0.3 acute toxic units (TUa) to the most sensitive of at least three different
test species.
EPA has a total of 100 recommended criteria (acceptable ambient
concentrations (AACs)) which are designed to protect human health. In the
absence of EPA's recommended criteria, States may calculate AACs based on the
equations in the text. In addition, the need for sediment and biological criteria in
State water quality standards is discussed.
Effluent Characterization
This chapter contains completely revised effluent characterization
discussions and recommendations. It includes streamlined procedures (as
compared to the original TSD) for predicting the likely impacts of toxic effluents
on aquatic life and human health. Recommendations are provided for
determining, either with or without actual effluent data, whether a discharge is
causing or has the "reasonable potential" of causing an excursion above a State
water quality standard. These effluent characterization procedures can be
performed in one step and do not include initial screening followed by definitive
data generation.
The revised effluent characterization procedures for assessing potential
human health impacts now focus on control of bioconcentratable chemicals and
reference a set of procedures which are described in more detail in EPA's [draft]
guidance entitled, "Guidance on Assessment, Criteria Development, and Control
of Bioconcentratable Contaminants in Surface Waters".
Exposure and Wasteload Allocation
A goal of permit writers is to determine what effluent composition will
protect aquatic organisms and human health. Exposure assessment includes an
x

-------
analysis of how much of the water body is subject to the criteria being exceeded,
for how long, and how frequently. The first step is to evaluate the effluent plume
dispersion. If mixing is not rapid and complete, the wasteload allocation must be
based on a mixing zone analysis (where State standards allow a mixing zone).
Means to assess dilution at the edge of a mixing zone are described in Chapter 5.
As with the original TSD, ambient criteria to control acute toxicity to aquatic life
may be met within a short distance of the outfall. However, this provision is no
longer restricted to outfalls which have a high rate diffuser.
If mixing is rapid and complete, there are several models that can be used
to assess exposure. Steady state models work on the assumption that the effluent
concentration is steady and that the duration and frequency with which criteria
are exceeded can be reflected entirely by selecting a critical flow condition in the
receiving water of appropriate duration and frequency.
Another means of modeling exposure is to use computer models that
incorporate that variability of the individual inputs (such as effluent flow and
concentration, receiving water flow, temperature, background concentration, etc.)
These models are termed dynamic models and are more accurate than steady
state models in reflecting or predicting exposure provided adequate data exist.
The acceptable effluent condition derived using these models is expressed as the
effluent long term average (LTA) and variance, which greatly simplifies
derivation of permit limits. Three dynamic modeling approaches are described
along with instructions for their use.
Permit Requirements
The requirements of a wasteload allocation (WLA) must be translated into
permit limitations in the wastewater discharge permit. In many cases permit
limits will be different than the WLA to reflect different assumptions and means
of expressing effluent quality. Three types of WLAs are identified, and
recommendations are provided for deriving permit limits to properly enforce each
type of WLA. Other permit-related issues such as permit documentation and
how to express limitations are discussed. In addition, approaches for requiring
and conducting toxicity reduction evaluations (TREs) are presented.
Compliance Monitoring
The compliance monitoring and enforcement process for water quality-
based permits summarized in Chapter 6 is based on existing regulation and
guidance. As with technology-based permits, any failure to meet a limit is a
violation, and every violation must be reviewed to determine the appropriate
response. Whole-effluent toxicity monitoring and enforcement concepts embodied
in the "Compliance Monitoring and Enforcement Strategy for Toxics Control
(January 19, 1989)" have been added to this revision.
xi

-------

List of Abbreviations
AA
atomic absorption
AAC
acceptable ambient concentration
ACR
acute-chronic ratio
ADI
acceptable daily intake
ATC
acceptable tissue concentrations
ATE
acute toxicity endpoints
BAF
bioaccumulation factor
BAT
best available technology
BCF
bioconcentration factor
BCT
best conventional technology
BMP
best management practices
BOD
biochemical oxygen demand
BPJ
best professional judgment
BPT
best practicable technology
CCC
criterion continuous concentration
CFR
Code Federal Regulations
CMC
criterion maximum concentration
CTE
chronic toxicity endpoint
CV
coefficient of variation
CWA
Clean Water Act
EC
effective concentration
DF
dilution factor
DMR
discharge monitoring report
DO
dissolved oxygen
EMS
Enforcement Management System
EP
equilibrium partitioning approach
ERL
Environmental Research Laboratory (EPA)
FAV
Final Acute Value
xii

-------
FGETS
Food and Gill Exchange of Toxic Substances
GC/MS
gas chromatograph/mass spectrometer
HPLC
high-pressure liquid chromatography
IC
inhibition concentration
IRIS
Integrated Risk Information System (EPA)
LA
load allocation
LC
lethal concentration (percentage of organisms killed)
LOAEL
lowest observed adverse effect level
LOEC
lowest observed effect concentration
LTA
long term average
NOAEL
no observed adverse effect level
NOEC
no observed effect concentration
NPDES
National Pollutant Discharge Elimination System
NTIS
National Technical Information Service
PCS
permit compliance system
POTW
publicly owned treatment works
ql*
cancer potency factor
QA/QC
quality assurance/quality control
QNCR
quarterly noncompliance report
QSAR
structure-activity relationships
Rfd
reference dose
RWC
receiving water concentration
SLSA
Simplified Lake/Stream Analysis
STORET
storage and retrieval of water quality information
TIE
toxicity identification evaluation
TMDL
total maximum daily load
TRE
toxicity reduction evaluation
TU
toxic unit
TUa
acute toxic unit
TUc
chronic toxic unit
WET
whole effluent toxicity
xiii

-------
WLA	wasteload allocation
Modeling Abbreviations
ARM	agricultural runoff model
CHNTRN Channel Transport Model
CETIS Complex Effluent Toxicity Information System
CIS	Chemical Information System
CORMIX 1 Cornell Mixing Zone Expert System
CTAP Chemical Transport and Analysis Program
DYNHYD4 hydrodynamic model
DYNTOX dynamic toxics model
EXAMS-II Exposure Analysis Modeling System
FCM2	WASP Food Chain Model
FETRA Finite Element Transport Model
FLOSTAT USGS computer program that estimates the arithmetic mean flow
and 7Q10 of rivers and streams
HHDFLOW historic daily flow program
HSPF Hydrologic Simulation Program - FORTRAN
MEXAMS Metals Exposure Analysis Modeling System
MINTEQA2 Equilibrium Metals Specification Model
MICH Michigan River Model
PSY	steady-state two-dimensional plume model
NPS	Non-point Source Model for Urban and Rural Areas
SARAH2 surface water assessment model for back calculating reductions in
biotic hazardous wastes
SERATRA Sediment Contaminant Transport Model
TODAM Transport One-Dimensional Degradation and Migration Model
TOXIWASP Chemical Transport and Fate Model
TOXI4 a subset of WASP4
TOXIC Toxic Organic Transport and Bioaccumulation Model
UDKHDEN three dimensional model used for single or multiple port diffusers
ULINE uniform linear density flume model
xiv

-------
UMERGE two dimensional model used to analyze positively buoyant discharge
UOUTPLM cooling tower plume model adapted for marine discharges
UPLUME numerical model that produces flux-average dilutions
WASP4 water quality analysis program
WASTOX Estuary and Stream Quality Model
WQAB FLOW water quality analysis system flow data subroutine
xv

-------
Glossary
absolute toxicity the toxicity of the effluent, without considering dilution.
acceptable ambient concentrations (AAC) the concentration of a chemical in
water which will not cause adverse impacts to human health; AAC is
expressed in units of mg/1.
acceptable tissue concentrations (ATC) the concentration of a chemical in edible
fish or shellfish tissue which will not cause adverse impacts to human
health when ingested. ATC is expressed in units of mg/kg/day.
acute involving a stimulus severe enough to rapidly induce a response; in aquatic
toxicity tests, a response measuring lethality observed in 96 hours or less is
typically considered acute. When referring to human health, an acute
effect is not always measured in terms of lethality.
acute-chronic ratio (ACR) the ratio of the acute toxicity of an effluent or a
toxicant to its chronic toxicity. It is used as a factor for estimating chronic
toxicity on the basis of acute toxicity data, or for estimating acute toxicity
on the basis of chronic toxicity data.
acutely toxic conditions conditions lethal to aquatic organisms that pass through
them.
acute toxicity endpoints (ATEs) described in terms of effluent concentrations.
Lethal concentration (LC) is the effluent concentration at which a certain
percentage of the test organisms are killed.
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.
ambient toxicity toxicity manifested by a sample collected from a waterbody.
antagonism the characteristic property of a mixture of toxicants that exhibits a
less-than-additive cumulative toxic effect.
antidegradation part of the water quality standards requirement# a designated
use is currently being attained, the waterbody may not be classified for a
less stringent use except as provided in 40 CFR Part 131.
aquatic community a biological association composed of all the interacting
populations of aquatic organisms that co-occur in a given waterbody or
habitat.
xvi

-------
bioaccumulation the process by which a compound is taken up by an aquatic
organism, both from water and through food.
bioaccumulation factor (BAF) measure of a chemical's tendency to
bioaccumulate.
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 frequently used in the
pharmaceutical industry to evaluate the potency of vitamins and drugs.
bioavailability a measure of the physicochemical access that a toxicant has to the
biological processes of an organism. The less the bioavailability of a
toxicant the less its toxic effect on an organism.
bioconcentration the process by which a compound is absorbed from water
through gills or epithelial tissues and is concentrated in the body.
bioconcentration factor (BCF) measure of a chemical's tendency to concentrate
in tissues of aquatic organisms
biocriteria numerical measures or narrative descriptions of the biological
integrity of unimpaired natural systems.
biological assessments evaluation of the biological condition of a waterbody
using biological surveys, chemical-specific analyses of pollutants known to
impact aquatic life, and/or toxicity tests.
biological criteria or biocriteria numerical measures or narrative expressions that
describe the biological integrity of aquatic communities inhabiting waters
of a given designated aquatic life use.
biological integrity the condition of the aquatic community inhabiting unimpaired
waterbodies of a specified habitat as measured by community structure and
function.
biological monitoring use of a biological entity as a detector and its response as
a measure to determine environmental conditions. Toxicity tests and
biosurveys are common biomonitoring methods.
biological survey or biosurvey collecting, processing, and analyzing a
representative portion of the resident aquatic community in order to
determine its structural and/or functional characteristics.
biomagniflcation the process by which the concentration of a compound
increases in different organisms, occupying successive trophic levels.
xvii

-------
cancer potency slope factor (ql*) an indication of a chemical's human cancer-
causing potential derived using animal studies or epidemiological data on
human exposure; based on extrapolation of high dose levels over short
periods of time to low dose levels and a lifetime exposure period through
the use of a linear model.
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
considered a relative term depending on the life span of an organism. The
measurement of a chronic effect can be reduced growth, reduced
reproduction, etc., in addition to lethality.
chronic toxicity endpoints (CTEs) include NOEC, LOEC, EC, and IC; based on
observations of reduced reproduction, growth, etc., in addition to lethality.
coefficient of variation (CV) standard statistical measure of the relative variation
of a distribution or set of data, defined as the ratio of the standard
deviation to the mean.
community component a general term that may pertain to the biotic guild (fish,
invertebrates, algae), the taxonomic category (order, family, genus,
species), the feeding strategy (herbivore, omnivore, predator), or the
organizational level (individual, population, assemblage) of a biological
entity within the aquatic community.
completely mixed condition no measurable difference in the concentration of a
pollutants that exists across a transect of the waterbody (e.g., does not vary
by 5%).
conservative pollutant a pollutant that is persistent and not subject to decay or
transformation.
continuous stimulation model a fate and transport model that uses time series
input data to predict receiving water quality concentrations in the same
chronological order as that of the input variables.
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 indefinitely 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.
xviii

-------
critical life stage the period of time in an organism's lifespan in which it is the
most susceptible to adverse effects caused by exposure to toxicants, usually
during early development (egg, embryo, larvae). Chronic toxicity tests are
often run on critical life stages to replace long duration, life cycle tests
since the most toxic effect usually occurs during the critical life stage.
design flow the critical flow used for steady state wasteload allocation modeling.
discharge length scale the square root of the cross-sectional area of any
discharge outlet.
diversity the number and abundance of species in a specified location.
duration the period of time over which the instream concentration is averaged
for comparison with criteria concentrations. This specification limits the
duration of concentrations above the criteria.
effective concentration (EC) a point estimate of the toxicant concentration that
would cause an observable adverse effect (such as death, immobilization,
or serious incapacitation) in a given percentage of the test organisms.
Calculated from curve fitting.
effluent biomonitoring the measurement of the biological effects of effluents
(such as toxicity, biostimulation, and bioaccumulation).
Equilibrium Partitioning Approach (EP) method for generating sediment criteria
that focuses on the chemical interaction between sediments and
contaminants.
Final Acute Value (FAV) an estimate of the concentration of the toxicant
corresponding to a cumulative probability of 0.05 in the acute toxicity
values for the genera for which acceptable acute tests have been conducted
on the toxicant.
frequency how often criteria can be exceeded without unacceptably affecting the
community.
harmonic mean flow elapsed time divided by the sum of the reciprocals of the
arithmetic mean flow.
Inhibition Concentration (IC) a point estimate of the toxicant concentration that
would cause a given percent reduction (e.g., IC25) in a non-lethal
biological measurement of the test organisms, such as reproduction or
growth. Determined using curve fitting with an assumption of a
continuous dose-response relationship.
xix

-------
LC the4oxicant concentration killing a certain percentage (e.g., LC50) of
exposed organisms at a specific time of observation.
lipophilic high affinity for lipids (fats).
load allocations (LA) the pattern of the loading capacity of a receiving water,
attributed either to one of its existing or future nonpoint sources of
pollution or to natural background sources.
lognormal probabilistic dilution model a model that calculates the probability
distribution of receiving water quality concentrations from the lognormal
probability distributions of the input variables.
lowest observed adverse effect level (LOAEL) the lowest concentration of an
effluent or toxicant that results in statistically significant adverse health
effects as observed in chronic or subchronic human epidemiology studies
or animal exposure.
lowest observed effect concentration (LOEC) the lowest concentration of an
effluent or toxicant that results in observable adverse effects in the aquatic
test population.
magnitude how much of a pollutant (or pollutant parameter such as toxicity),
expressed as a concentration or toxic unit is allowable.
mixing zone an area where an effluent discharge undergoes initial dilution and is
extended to cover the secondary mixing in the ambient waterbody. An
allocated impact zone is an area where chronic water quality criteria can
be exceeded as long as acutely toxic conditions are prevented.
Monte Carlo simulation a stochastic modeling technique that involves the
random selection of sets of input data for use in repetitive model runs in
order to predict the probability distributions of receiving water quality
concentrations.
n-octanol/water partition coefficient the ratio, in a two-phase system of n-
octanol and water at equilibrium, of the concentration of a chemical in the
n-octanol phase to that in the water phase.
No Observed Adverse Effect Level (NOAEL) a threshold dose of an effluent or a
toxicant below which no adverse biological effects are observed, as
identified from chronic or subchronic human epidemiology studies or
animal exposure studies.
No Observed Effect Concentration (NOEC) the highest concentration of an
effluent or a toxicant at which no adverse effects are observed on the
aquatic test organisms. Determined using hypothesis testing with the
xx

-------
assumption of a noncontinuous threshold model of the dose-response
relationship.
nonthreshold effects carcinogenicity.
permit averaging period the duration of time over which a permit limit is
calculated - day(s), week, or month.
persistence that property of a toxicant or an effluent that is a measurement of its
resistance to chemical and/or biological degradation. A persistent toxicant
or toxicity maintains effect after mixing, degrading slowly. A non-
persistent toxicant or toxicity may have a quickly reduced effect after
mixing as degradation processes such as volatilization, photolysis, etc.,
transform the chemical.
probability a number expressing the likelihood of occurrence of a specific event,
such as the ratio of the number of outcomes that will produce a given
event to the total number of possible outcomes.
probability distribution a mathematical representation of the probabilities that a
given variable will have various values.
reasonable potential an analysis of an effluent's capability to cause an excursion
above an applicable numeric or narrative water quality standard.
Receiving Water Concentration (RWC) the exposure concentration of a toxicant
or the parameter toxicity in the receiving water after mixing. (Formerly
termed "instream waste concentration," IWC.)
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.
reference dose (RfD) an estimate of the daily exposure to human population that
is likely to be without appreciable risk of deleterious effect during a
lifetime; derived from NOAEL or LOAEL.
relative toxicity the toxicity of the effluent when it is mixed with the receiving
water, or a dilution water of similar composition for toxicity testing.
slug flow sampling a monitoring procedure that follows the same slug of
wastewater throughout its transport in the receiving water. Water quality
samples are collected at receiving water stations, tributary inflows, and
point source discharges only when a dye slug or tracer passes that point.
xxi

-------
steady state model a fate and transport model that uses constant values of input
variables to predict constant values of receiving water quality
concentrations.
STORET EPA's computerized water quality data base that includes physical,
chemical, and biological data measured in waterbodies throughout the
United States.
sublethal involving stimulus below the level that causes death.
synergism the characteristic property of a mixture of toxicants that exhibits a
greater-than-additive cumulative toxic effect.
threshold effects acute, subacute, or chronic human health effects.
total maximum daily load (TMDL) the sum of the individual waste load
allocations for point sources and load allocations for nonpoint sources and
natural background.
toxicity identification evaluation (TIE) set of procedures to identify the specific
chemical(s) responsible for effluent toxicity.
toxicity reduction evaluation (TRE) a site specific study conducted in a step-
wise process designed to identify the causative agents of effluent toxicity,
isolate the sources of toxicity, evaluate the effectiveness of toxicity control
options, and then confirm the reduction in effluent toxicity.
toxicity test a procedure to determine the toxicity of a chemical or an effluent
using living organisms. A toxicity test measures the degree of response of
exposed test organisms to a specific chemical or effluent.
Toxic Units (TUs) 100 divided by the toxicity endpoint measured; expressed as
acute toxicity units (TUa) or chronic toxicity units (TUc).
toxic unit acute (TUa) the reciprocal of the effluent concentration that causes an
acute effect by the end of the acute exposure period (e.g., 100/LC50).
toxic unit chronic (TUc) the reciprocal of the effluent concentration that causes
no observable effect on the test organisms by the end of the chronic
exposure period (e.g., 100/NOEC).
wasteload allocation (WLA) the portion of a receiving water's total loading
capacity that is allocated to one of its existing or future point sources of
pollution.
water quality criteria scientifically derived ambient concentrations developed by
EPA or States for various pollutants of concern. Criteria developed by
xxii

-------
EPA are recommended levels which should not be exceeded in a body of
water in order to protect aquatic life and human health; all states have
both chemical-specific numeric criteria for individual toxic pollutants and
narrative "free from toxics in toxic amounts" criteria; criteria are subjected
to the regulatory process of regional, state, and public comment.
water quality limited any segment in which it is known that water does not meet
applicable water quality standards, and/or is not expected to meet
applicable water quality standards even after application of technology-
based effluent limitations.
water quality standard a law or regulation that consists of the beneficial
designated use or uses of a waterbody, the water quality criteria that are
necessary to protect the use or uses of that particular waterbody, and an
antidegradation statement.
whole effluent toxicity the aggregate toxic effect of an effluent measured directly
with a toxicity test.
xxiii

-------
Introduction
Purpose
The purpose of this revised Technical Support Document (TSD) for Water
Quality-Based Toxics Control is to provide the most current procedural
recommendations and guidance for identifying, analyzing and controlling adverse
water quality impacts caused by toxic discharges to the waters of the United
States. The original TSD was published in September 1985. Since then much
has changed. Tlie Clean Water Act was amended in 1987 with an emphasis on
controlling toxic pollutants, new policies and regulations have been promulgated
and a vast amount of knowledge and experience has been gained in controlling
toxic pollutants. Because of these changes, EPA revised and updated the TSD.
This guidance document is intended to support the implementation of the
EPA water quality-based approach to toxics control. As such, the
recommendations and guidance found in this document are not binding and
should be used by regulatory authorities with discretion. The guidance in this
document has been developed as the most current representation of knowledge in
the field of assessment and control of toxic discharges. Some of the guidance in
this document is based on ongoing research and development (bioconcentration
methods, Chapter 3) and should only be used with this understanding.
Background
The EPA surface water toxics control program, represented
diagrammatically in Figure 1, relies on portions of the national pretreatment
program, the effluent limitations guidelines program, the sludge program, the
combined sewer overflow program, the stormwater management program, the
304(1) program, the water quality standards program and the national pollutant
discharge elimination system (NPDES) program. States are authorized by EPA
to implement certain portions of the national toxics control program, such as the
NPDES program. Scientific and technical guidance is developed and published
by EPA to assist the States. EPA is required by the Clean Water Act (CWA)
and federal regulations to play an oversight role to assure that States authorized
to implement various program requirements, do so in a nationally consistent and
equitable fashion for the protection and improvement of the nation's surface
waters.
Throughout the evolution of the toxics control program, EPA has provided
guidance concerning new program initiatives, statutory developments, and
regulatory requirements. In 1980, EPA emphasized in its preamble to NPDES
regulations (45 FR 33520) that NPDES permit limitations must reflect the most
stringent of technology-based or water quality-based controls for toxic pollutants.
EPA reiterated the significance of surface water toxics control in 1984 through
xxiv

-------
the publication of its national "Policy for the Development of Water Quality-
Based Permit Limitations for Toxic Pollutants" (49 FR 9016, March 9, 1984).
EPA recommended the use of "biological techniques as a complement to
chemical-specific analyses to assess effluent discharges and express permit
limitations" (49 FR 9017). The preamble to additional regulations promulgated
in 1984 (49 FR 37998) stressed the importance of establishing effluent limitations
in NPDES permits to control toxic pollutants. Regulatory provisions promulgated
on June 2, 1989 clarify EPA's surface water toxics control program and the use of
whole effluent toxicity, and implement CWA { 304(1) concerning the identification
of impaired waters and the development of individual control strategies.
The control of toxic discharges to the Nation's waters is an important
objective of the NPDES program. To effectively accomplish this objective, EPA
recommends the use of an integrated water quality-based approach for controlling
toxic discharges. EPA's integrated "standards to permits" approach, illustrated in
Figure 1, starts with water quality criteria, objectives and standards and results in
NPDES permit limitations to control toxic pollutants through the use of both
chemical-specific and whole effluent toxicity limitations. Limitations are essential
for controlling the discharge of toxic pollutants to the Nation's water. Once
NPDES permit limitations are set, compliance is essential. Compliance can be
ascertained by continual routine monitoring of effluent quality. Water quality-
based effluent limitations, when developed in accordance with the procedures in
this document, will protect water quality and prevent the violation of State water
quality standards.
xxv

-------
Define water qoaBty
otoiectivas. criteria, and standards
EstabSah priority
wstar bodiM
Chapter 2
Chemical-specific Effluent
Characterization
I
p Evaluate for excursions above standards
I
	Determine 'reasonable potential"
I
Generate effluent data
Set permit limits directly
Whole Effluent Toxicity Effluent
Characterization
Evaluate for excursions above standards -

Determine 'reasonable potential'
Generate data
	1
Evaluate axpoaura
(critical flow, fata modeling.'
and mbang) and calculate
waststoed a«oc«t»on
Define raqurad dtodwga
charactariatica by the
wastetosd allocation
Set permit limits directly—J	
I —
f
I
Chapter 3
Chapter 4
Derive permit
raquiramenta
Evaiuata toxicity reduction
~
~
final permit writfi
monrtonnq raquaanierna
Compliance
Chapter 5
]
Chapter 6
Figure i: Overview of the water quality-based 'Standards to permits"
process for toxics control.
xxvi

-------
1. APPROACHES TO WATER QUALITY-BASED TOXICS CONTROL
1.1 INTRODUCTION
In this chapter, basic principles are presented which cover the protection
of aquatic life and the protection of human health from impacts caused by the
release of toxic pollutants to the nation's surface waters. TTie control of the
discharge of toxic pollutants is a paramount objective of the NPDES program.
The Clean Water Act and EPA regulations (described in Appendices B-l and B-
4, respectively) authorize and require the use of the "integrated strategy" to
achieve and maintain water quality standards. In addition, EPA policy and
guidance have long advocated this approach. For the protection of aquatic life,
the integrated strategy involves the use of three control approaches: the chemical
specific control approach, the whole effluent toxicity control approach and the
biocriteria/bioassessment and biosurvey approach. However, for the protection of
human health, technical constraints do not yet allow forfull reliance on an
integrated strategy, and thus primarily chemical-specific assessment and control
techniques should be employed.
The integrated approach to water quality-based toxics control, including
the use of toxicity testing and whole effluent toxicity limitations, chemical-specific
testing and limitations, and bioassessments/biosurveys, relies on the water quality
standards that each state has adopted. Virtually all states have water quality
standards consisting of both chemical-specific numeric criteria for individual toxic
pollutants, and narrative "free from toxics in toxic amounts" criteria. Currently,
very few States have developed biocriteria as a part of their water quality
standards.
The narrative water quality criterion in all states generally requires that
the state's waters be free from oil, scum, floating debris, materials that will cause
odors, materials that are unsightly or deleterious, materials that will cause a
nuisance, or substances in concentrations that are toxic to aquatic life, wildlife or
human health (emphasis added). The use of toxicity testing and whole effluent
toxicity limitations is based upon a state's narrative water quality criterion and in
some cases, a state numeric criterion for toxicity.
Chemical-specific limitations are based upon the numeric criteria adopted
by each state or upon EPA Water Quality Criteria [54, 55]. In many cases, states
have adopted EPA Water Quality Criteria as a part of their water quality
standards. (See Chapter 2 "Water Quality Criteria and Standards" for further
information).
The integrated approach must include the control of toxic pollutants
through implementation of the narrative standard (preferably through a numeric
1

-------
criteria for the parameter toxicity), as well as the control of individual pollutants
for which specific chemical water quality criteria exist in a state's standards.
Reliance solely on either chemical-specific numeric criteria or on the narrative
criterion would result in only a partially effective state toxics control program.
Each control approach is described in greater detail below.
1.2 CHEMICAL-SPECIFIC APPROACH FOR AQUATIC LIFE
PROTECTION
The chemical-specific approach to toxics control for the protection of
aquatic life, uses specific chemical effluent limitations in NPDES permits to
control the discharge of toxic pollutants. These limitations are generally
developed from laboratory-derived, biologically-based numeric water quality
criteria adopted within a state's water quality standards. Water quality criteria
are adopted by a state for the protection of the designated uses of the receiving
water. Chemical-specific water quality-based limitations in NPDES permits
involve a site-specific evaluation of the discharge and its effect upon the receiving
water. This may include collection of effluent and receiving water data,
development of a wasteload allocation (WLA) and a total maximum daily load
(TMDL) through modeling, a mixing zone analysis, and the calculation of permit
limitations. Once a numeric water quality criterion is adopted, chemical-specific
limitations can be developed to ensure that a permittee's discharge does not
violate acute and chronic water quality criteria for the pollutant in a receiving
water. These steps are discussed in Chapters 3, 4, and 5.
EPA water quality criteria for the protection of aquatic life are developed
under the requirements of Section 304(a)(1) of the Clean Water Act and are
published by EPA in separate criteria documents and summarized in the Quality
Criteria for Water [54]. Water quality criteria are scientifically derived and
attempt to consider a wide range of toxic endpoints including acute and chronic
impacts and bioaccumulation. Each criteria consists of two values, an acute and a
chronic value. Criteria are developed using the latest scientific knowledge on the
kind and extent of identifiable effects on organisms such as; plankton, fish,
shellfish, wildlife and plant life, which may be expected from the presence of
pollutants in any body of water. Water quality criteria also reflect the
concentration and dispersal of pollutants, or their by-products, through biological,
physical, and chemical processes, and the effects of pollutants on biological
community diversity, productivity, and stability of the receiving water [54]. A
more detailed discussion of the derivation of numeric criteria is provided in
Chapter 2. Recommendations for using chemical-specific data to determine
which individual toxicants need to be controlled are found in Chapter 3.
2

-------
1.2.1	Correlation of Chemical-specific Measurements to Actual Receiving
Water Impacts
EPA has conducted a series of studies to determine whether EPA Ambient
Water Quality Criteria are protective of receiving water systems. The first such
study was conducted at Shayler Run, Ohio, to evaluate the applicability of
laboratory generated toxicity data of copper to a natural stream [14]. The results
of the study indicated that the laboratory predictions of the toxicity of copper
underestimated the instream effects by approximately a factor of two. This
difference was caused by avoidance of organisms in the stream and was not
anticipated by the laboratory tests conducted.
Additional studies performed by EPA's Monticello Ecological Research
Station (MERS) indicate good agreement between the level of protection
expected from the application of EPA water quality criteria and the results in
experimental streams. If the criteria are met, then the receiving water system will
be protected from unacceptable impacts caused by the chemical of concern. The
studies conducted by MERS are described in Box 1-1.
1.2.2	Chemical-Specific Analytical Method Precision
Tables 1-3 through 1-5 illustrate the types of precision commonly seen in
inorganic, organic and non-metal inorganic chemical analyses which are routinely
used for determining concentrations of specific toxic pollutants in effluents.
These tables show the observed variability. The variability of chemical
measurements increases as one approaches the limit of detectability for a
chemical. Table 1-3 shows the precision, of ten metals, at the lower end of the
measurement detection range. The coefficient of variation, or CV (defined as the
standard deviation divided by the mean) for these analyses ranges from 18 to
129% [45]. Table 1-4 shows the precision associated with organic chemical
analyses. The coefficients of variation range from 12 - 91%. Table 1-5
demonstrates the precision of non-metal inorganic analyses at the lower end of
the measurement range. The CVs for this type of analyses ranges from 4.6-61%
[46]. (For comparison with toxicity test method precision, see Box 1-3).
3

-------
Box 1-1
Correlation of Chemical Specific
Criteria to Instream Impacts
In studying the field applicability of EPA's water quality criteria in freshwater systems,
MERS (Monticello Ecological Research Station) conducted studies in experimental streams [15-25]
to determine the level of protection provided by the individual chemical criteria. Each of the
streams was one-quarter mile long with alternating mud-bottomed pools and rocky riffles. Fish were
stocked into the streams to a known population density while other plants and animals were the
result of natural colonization.
The chemicals studied were ammonia, chlorine, chlorine combined with ammonia, selenium,
and pentachlorophenol. Some studies were conducted during a summer (pentachlorophenol) while
others continued for over two years (selenium IV). Sample data on ammonia and ammonia
combined with chlorine are shown in Tables 1-1 and 1-2. In all experiments, the streams were
continuously dosed with the chemical(s) being studied and the biological effects were determined
statistically by a comparison to the control streams. The concentration at which biological effects
occurred were then compared to the criteria continuous concentration (CCC) for that compound.
With the exception of chlorine in the presence of ammonia, the data from the other
experiments indicate that slight or no effects were found in the streams at the CCC. In the case of
chlorine combined with ammonia, a substantial impact was found, but only on one species, the
channel catfish. Because the CCC is designed to protect most, but not all of the species all of the
time (See discussion in Chapter 2 on EPA Ambient Water Quality Criteria), slight impacts are
expected.
4

-------
Table 1-1
Effects in Streams Exposed to Ammonia [19-24]



Effects
Indicator
Criteria?
3Xb
9Xf
Fish
fathead minnow
0*
0
0
bluegill
0
0
++
channelcatfish
+
~ +
+++
white sucker
0
0
0
rainbow trout
0
0
++
walleye
0
0
++
Benthic Invertebrates
0
+

Zooplankton
0
~
~
Notes
a	Criteria = 0.05 mg/l unionized ammonia at average stream pH and temperature; 1.0 mg/1 total ammonia was
added to reach this concentration; concentrations of unionized ammonia varied daily and seasonally due to
natural pH and temperature fluctuations.
b	3X = 3 times criteria concentration based on input of 3 mg/1 total ammonia.
c	9X = 9 times criteria concentration based on input of 9 mg/1 total ammonia.
d	0= no difference from controls;+'s represent gradation of differences from controls ranging from slight (+) to
dramatic (++++)•
5

-------
Table 1-2
Effects in Streams Exposed to Ammonia and Chlorine [19-24]
Indicator	Effects
4 ue/la	35 ug/1	122 ug/1
Fish
Channel catfish
++ b
++
+++
Bluegill
0
0
0
Benthic Invertebrates
0
+
++
Zoopiankton
0
0
0
Bacteria
+
++
+++
Periphyton
0
0
0
Primary production
D
0
0
Litter decomposition
+
+
++
Aquatic Plants
0
0
0
Notes
a	Average concentrations of TRC in presence of 2-3 mg/1 total ammonia; National criteria for chlorine = 11 ug/1.
b	0= no difference from controls;+'s represent gradation of differences from controls ranging from slight (+) to
dramatic (++++).
6

-------
Table 1-3
Precision of Inorganic Analysis
at the Low End of the Measurement Detection Range [45]
Analvte	No. of Labs	CV (%)
Aluminum
37
43
Cadmium
63
66
Chromium
72
40
Copper
86
36
Iron
78
38
Lead
64
46
Manganese
55
129
Mercury
76
79
Silver
50
18
Zinc
62
118
7

-------
Table 1-4
Precision Ranges for Organic Chemical Analysis
Chemical	No.
Labs
Benzene
4 Chlorobenzenes	20
Ethyl benzene
Toluene
23	Halocarbons	20
4 Halocarbons	20
11 Phenols	20
Benzidine	17
3,3-Dichlorozidine
6 Pthalate esthers	16
3 Nitrosamines	XI
24	Organochlorine	22
Pesticides and PCBs
16 PNAs	?
CV	% Data	EPA Document
(%)	Discarded*	Referenced
31-64	10	600/S4-84-064
16-29	?	600/S4-84-064
40-50	?
20-45	20	600/S4-84-044
38-64	?
38-69	?	600/S4-84-062
?	22	600/S4-84-056
?	19	600/S4-84-051
>12-45	?	600/S4-84-061
16-91	?	600/S4-84-063
Notes
* Discarded as outfiers
It is important to note that in many chemicalanalysesa decision may be made that certain anomalous data points, or
outliers, are unusable and are not reported as valid data points. This type of data evaluation is made because in chemical
analysesit is routine to repeat the analysis with the same sample and referencestandard until an acceptable result is
obtained.
8

-------
Table 1-5
Precision of Non-metal Inorganic Analyses
Over the Measurement Range [46]
Lab
Parameter
CV (%) Ranee
17
Alkalinity
4.9-14
>20
Residual chlorine
13-25
16
Ammonia nitrogen
15-58
6
Kjeldahl nitrogen, total
38-41
15
NO3 nitrogen
17-61
6
Total P
25-40
58
BOD
15-33
58
COD
6.9-34
21
TOC
4.6-70
1.3 WHOLE-EFFLUENT APPROACH FOR AQUATIC LIFE
PROTECTION
The whole effluent approach to toxics control for the protection of aquatic
life involves the use of acute and chronic toxicity tests to measure the toxicity of
wastewaters. Whole effluent toxicity is a useful parameter for assessing and
protecting against impacts upon water quality and designated uses caused by the
aggregate toxic effect of the discharge of toxic pollutants. Whole effluent toxicity
tests employ the use of standardized, surrogate freshwater or marine (depending
upon the receiving water) plants, invertebrates and vertebrates. EPA has
published extensive written protocols listing numerous marine and freshwater
species for toxicity testing [42, 43, 44].
Acute toxicity is defined as a test of usually less than 96-hours in duration
in which lethality is the measured endpoint. A chronic toxicity test is defined as a
long term test in which sublethal effects, such as, fertilization, growth and
reproduction are usually measured, although in highly toxic effluents lethality may
also result. Traditionally, chronic tests are full life cycle tests or at least 30 days
tests. However, the duration of most of the EPA chronic toxicity tests have been
shortened to seven days by focusing on the most sensitive life cycle stages. For
this reason the EPA chronic tests are called short-term chronic tests.
In an laboratory acute toxicity test, an effluent sample is collected, diluted
and placed in test chambers with the chosen test species. After 24, 48 and 96
9

-------
hours, the number of live organisms remaining in each test concentration and in a
control is recorded. In a laboratory chronic toxicity test, an effluent sample is
collected, diluted and placed in test chambers. Usually, the dilutions used in a
chronic test are 100%, 30%, 10%, 3%, 1%, and a control. Often the receiving
water is used to dilute the effluent because it more closely simulates
effluent/receiving water interactions. However, a standard control water an a
reference toxicant test is also recommended to ensure quality assurance in
chronic testing. Test organisms are placed in these test chambers for specified
periods of time. At various times during the exposure period, the organisms in
each chamber are observed and the lowest effluent concentration that causes a
significant adverse impact on the most sensitive endpoint for that test is
calculated (this endpoint can be mortality, reduced fertilization, lower fecundity,
reduced growth, etc.).
In chronic toxicity tests, the exposure duration is almost always assumed to
be the seven day short-term period unless otherwise noted. It is useful to note
that LCs, ECs and ICs are point estimates derived from a mathematical model
that assumes a continuous dose-response relationship. NOECs and LOECs, on
the other hand are not point estimates, and are determined using hypothesis
testing [43].
The toxicity test endpoint concentration becomes a quantified measure of
the concentration that would cause receiving water impact if exceeded for a
specific period of time. Acute toxicity endpoints (ATEs) commonly include LCs
and are described in terms of effluent concentrations. The lethal concentration
(LC), is the concentration of toxicant at which a certain percentage of the test
organisms die, e. g. the LC10 or LC50. An exposure duration is also included in
the endpoint such as 24, 48 or 96 hours (e.g. 96-hour LC50).
Commonly used chronic toxicity endpoints (CTEs) include the NOEC, the
LOEC, the EC and the IC. The no observed effect concentration (NOEC) is the
concentration of toxicant, in terms of percent effluent, to which the test organisms
are exposed that causes no observable adverse effect. The effects measured are
usually non-lethal such as decreases in reproduction and growth, however,
lethality is also measured if the effluent is highly toxic. The LOEC is the lowest
observed effect concentration, which is the lowest concentration of toxicant to
which the test organisms are exposed which causes an observed effect. Again, the
same effects are usually observed. The effective concentration (EC) is the
toxicant concentration that would cause a non-lethal adverse effect upon a certain
percentage of the test organisms, e.g. EC10 or EC50. The IC, or inhibition
concentration, is an estimate of the toxicant concentration that would cause a
given percent reduction in a biological measurement of the test organisms,
including reproduction, growth, fertilization or mortality. For example, an IC25
for reproduction would represent the effluent concentration at which a 25%
reduction in reproduction occurred.
10

-------
The new endpoint terminology (CTE and ATE) will facilitate the use of
more descriptive chronic toxicity endpoints, such as ICs. An IG endpoint is a
point estimate interpolated from the actual effluent concentrations at which the
measured effects occurred during the test. In addition, since an IC is a point
estimate, a coefficient of variation can be calculated from a given set of data. On
the other hand, an NOEC is determined using hypothesis testing and is
contingent upon the effluent concentrations chosen for the test. For example,
many chronic tests are conducted using 100, 30, 10, 3 and 1% effluent
concentrations. If the lowest observed effect concentration exhibited by the data
is at 10% effluent, the NOEC is estimated to be the next lowest dilution, or 3%.
The NOEC value is therefore dependent upon the test dilutions initially selected.
Yet the true NOEC value may lie somewhere between 10 and 3% effluent. In
addition, since NOECs are not point estimates, coefficients of variation cannot be
calculated for estimates of test precision. Furthermore, most of the chronic
toxicity test protocols allow up to a 20% effect level in the controls for a valid
test. Comparisons of both types of data indicate that an IC25 is approximately
the analogue of an NOEC. Therefore, test results previously expressed as
NOECs can be expressed as IC25s.
The measurement of whole-effluent aquatic toxicity can be used to limit
the discharge of toxicants in an effluent. Thus, toxicity itself is used as the
effluent parameter, and the toxicants creating the toxicity need not be specifically
identified or controlled unless a toxicity reduction evaluation is required to ensure
compliance with a toxicity limit in the permit. Permit limits can be expressed in
terms of a specific toxicity endpoint, or combination of endpoints, such as an
LC10, an LC50, an NOECs or an IC25. The derivation of toxicity-based permit
limits is described in Chapter 5.
Toxic Units
Since toxicity involves an inverse relationship (the lower the effect
concentration, the higher the toxicity of the effluent), it is more understandable to
translate concentration-based toxicity measurements into toxic units (TUs). In
this way, the potential confusion involving the inverse relationship is overcome
and the permit limit derivation process is better served. The number of toxic
units in an effluent is simply 100 divided by the toxicity endpoint measured:
TUa = 100/LC50
TUC = 100/NOEC (or 100/IC25)
(where LC50 or NOEC is expressed as percent effluent in dilution water and
IC25 can be substituted for NOEC). For example, an effluent with an acute
toxicity of an LC50 in 5% effluent is an effluent containing 20 TUaS.
A very important aspect of toxic units is that two different types are used
depending on whether acute or chronic aquatic toxicity is measured. The proper
expressions for toxic units are TUa, and TUC. TUa is the measurement of acute
11

-------
toxicity units, and TUC is a measurement of chronic toxicity units (see the glossary
for a definition of these terms). They are not the same measurement and should
not be used interchangeably. Acute and chronic TUs make it easy to quantify the
toxicity of an effluent and to specify water quality criteria based upon toxicity.
For example, an effluent sample that contains 20 TUcS is twice as toxic as an
effluent that contains 10 TUcS.
1.3.1 Correlation of Whole Effluent Toxicity Measurements to Actual Receiving
Water Impact
EPA conducted the Complex Effluent Toxicity Testing Program (CETTP)
which examined sites in both freshwater and saltwater systems to investigate
whether or not an evaluation of effluent toxicity, when adequately related to
receiving water conditions, can give a valid assessment of receiving system
impacts [1-10, 12, 13]. Summaries of these site studies are provided in Boxes 1-2
(freshwater) and 1-3 (saltwater). In addition, two other studies, presented in Box
1-2, were conducted to address this issue; a study conducted by the North
Carolina Division of Environmental Management [11], and a comparative
investigation, conducted by the University of Kentucky [53].
It is important to note that in these studies, different objectives were
addressed. The CETTP freshwater studies attempted to correlate ambient
receiving water toxicity to instream impacts (Figure 1-1). The CETTP saltwater
studies compared effluent toxicity to ambient receiving water toxicity utilizing dye
studies to measure receiving water concentrations of effluent (Figure 1-3). The
North Carolina study compared effluent toxicity to receiving water impact using
Ceriodaphnia chronic toxicity tests and receiving stream benthic
macroinvertebrates (shown in Figure 1-2). The Kentucky study examined the
relationship between effluent toxicity tests and instream ecological parameters.
Together, these studies comprise the largest available data base specifically
collected to determine the validity of toxicity tests to predict receiving water
community impact. It is essential to note that there is currently no direct method
by which to compare toxicity test data with receiving water population or
community measurements. Therefore, in order to address the correlation of
effluent and ambient toxicity tests to receiving water impacts, EPA evaluated the
results of the studies discussed above. The results, when linked together, clearly
show that if toxicity after considering dilution is present, impact will also be
present.
It is important to recognize that toxicity caused by contaminants in the
effluent, as measured by the whole-effluent toxicity tests, is only one of many
influences that determine the health of a biological community. Impact would
12

-------
Box 1-2
Correlation of Ambient Toxicity Measurements to
Receiving Water Impact (Freshwater)
EPA conducted eight freshwater site studies in which ambient toxicity was compared to the
receiving water biological impact. These site studies were a part of the Complex Effluent Toxicity
Testing Program (CETTP). Testing was done on-site concurrent with the field surveys. Sites in
Oklahoma, Alabama, Maryland, West Virginia, Ohio and Connecticut were included. Organisms
were exposed to samples of water from various stations and tested for toxicity. Biological surveys
(quantitative field sampling of fish, invertebrate, zooplankton, and periphyton communities in the
receiving water areas upstream at the discharge points and downstream) were made at these stations
at the same time the toxicity was tested to see how well the measured toxicity correlated to the
health of the community. These studies have been reviewed and published in the EPA publication
series [2-10].
Figure 1-1 illustrates the data from the CETTP studies. In essence, toxicity and impact were
normalized to a percent value and combined across 83 stream stations excluding the data from the
Baltimore Harbor site study because this was an estuarine site. For the ambient test data, statistical
analyses are not feasible because no dose response curve is obtained and a relevant control cannot
be obtained. A 20% toxic effect or a 20% reduction in taxa was used to define a positive value.
This value was chosen because it represents the allowable mortality of the controls for the tests used.
False positives (impact predicted but none found) occurred at only 3.6% of the 83 stations.
As discussed above, this is the only definitive error that can be identified in such comparisons. The
correct or non-contradictory findings were 96.4% of the stations. Because the 20% mortality value is
subjective and it is possible to use other values, larger percent values were evaluated. As the percent
mortality increases, the number of false positives remain nearly constant but the percent of cases
where impact is predicted and found, decreases markedly and the percent where impact occurs but is
not predicted, increases proportionately.
These analyses clearly show that even the definition of only 20% toxicity as a positive
toxicity finding substantially underpredicts toxicity. Even more clearly, the data show that the false
positives are essentially a non-occurrence. Therefore, a discharger's chance of being charged with a
toxic impact is remote if and only if dilution in the receiving water is considered.
Another study conducted by the North Carolina Department of Environmental Management
indicated the high accuracy of predicting receiving water impacts from whole effluent toxicity tests.
Forty-three comparisons were made between freshwater flowing streams using the Ceriodaphnia
dubia chronic test and a qualitative macroinvertebrate sampling. Overall there was 88% accuracy of
prediction (Figure 1-2) [11].
In addition, another comparative study [53] was conducted in the Kentucky River Basin.
This study consisted of a comparative ecological and toxicological investigation of a secondary
wastewater treatment plant and measured instream effects at 10 stations including reference sites.
The principal objective of the study was to assess downstream persistence of aquatic contaminants, to
quantify their effects on structure and function of aquatic communities and to evaluate the fathead
minnow embryo-larval test for measuring instream toxicity and estimating chronic effects on aquatic
biota. The results of the study indicate a good predictive correlation between embryo-larval survival
and independent ecological parameters, especially species richness of macroinvertebrates. The
correlation coefficients for species richness and embryo-larval survival was 0.96, and for embryo-
larval survival and diversity was 0.93. The estimated toxicity (LCI) correlated closely with the actual
percent instream effluent dilution observed at the first downstream station at which no ecological
impact was discernable.
13

-------
only be suspected where receiving water concentrations are at or above the
toxicity effect concentrations. Where effluent toxicity effects are not dominating,
influences from substrate differences, physical conditions such as dissolved oxygen,
temperature, channelization, flooding and weather cycles are also likely to affect
the biological community. Since these "non-toxicant" related effects may be
important, the value of the toxicity test is even greater because of its ability to
assess the impact of discharged toxicants. This allows regulatory authorities to
specifically control the impact caused by the discharge.
Box 1-3
Correlation of Effluent Toxicity Measurements to
Receiving Water Toxicity (Saltwater)
In saltwater systems, as in freshwater systems, receiving water impact should only be
suspected where receiving water waste concentrations are at or above the effect concentrations.
Dilutions in saltwaters may be greater than those seen in freshwater, and as a result there is a less
likely chance for receiving water impacts to be observed as predicted by toxicity tests.
Figure 1-3 illustrates the comparison between predictions of saltwater receiving water
toxicity and whole effluent toxicity. A total of 79 comparisons, or 89% of the receiving water sample
predictions were accurate. In only 6% of the cases did effluent toxicity tests predict receiving water
toxicity that was not present (false positive). In 5% of the cases was there a false negative
prediction, or the toxicity tests predicted no toxicity when the receiving water was toxic [12]. As
discussed in Box 1-2, if toxicity is only one possible adverse influence, then if one measures only
toxicity, a very high correlation should not necessarily be expected. All receiving water toxicity to
effluent toxicity correlations are based upon dye studies.
Most sites were selected to isolate potential impacts from a single source, unaffected by
other sources. Two studies with dye studies were conducted which provided enough data to derive
correlations. Both studies showed near-field effects, generally within the mixing zone.
One study conducted at Fernandina Beach, Florida [13], showed impacts outside the
proposed mixing zone. Results of another study (East Greenwich) showed the existence of poor
water quality well beyond the influence of the East Greenwich Sewage Treatment Plant and suggest
that other sources (point or non-point) may contribute significantly. This condition may be typical in
some of the more stressed estuaries.
14

-------
COMPARISON OF AMBIENT TOXICITY AND IMPACT
AT 83 STATIONS ON 8 STREAMS IN THE U.S.
IMPACT PREDICTED
MPACT PRESENT
ID
IMPACT PREDICTED
NO IMPACT PRESENT
IMPACT PREDICTED
NO IMPACT PRESENT
3.6%
4.8%
NO IMPACT PREDICTED
IMPACT PRESENT
22.9%
Figure 1-1: Results of the comparison of ambient toxicity to receiving water impact on
eight freshwater streams in"the U.S. (Complex Effluent Toxicity Testing Program, (1-10)).

-------
NORTH CAROLINA STUDY
43 Point Source Discharge Sites
No InBtream To* Jetty Predict
Impact Noted 50/
In9tream Toxicity Predicted
Impact Noted _ _ _
65%
«sD
f-H
S
No Inatream Toxicity Predicted
No Impact Noted 23^
Inatream Toxicity Predicted
No Impact Noted
Figure 1-2:
Comparison of effluent toxicity to receiving water impact using Ceriodaphnia
chronic toxicity tests and freshwater receiving stream benthic invertebrates
at 43 point source discharging sites in North Carolina. (11).

-------
SALT WATER STUDY
79 Ambient Stations and 4 Dischargers
PREDICTED AMBIENT TOXICITY
6% NO TOXICITY OBSERVED
NO AMBIENT TOXICITY PREDICTED
TOXICITY OBSERVED 55/
PREDICTED AMBIENT TOXICITY
TOXICITY OBSERVED
14%
NO AMBIENT TOXICITY PREDICTED
NO TOXICITY OBSERVED 75^
Figure 1-3: Comparison of predictions of receiving water toxicity based on effluent toxicity
and ambient receiving water testing, in saltwater environments. (Complex Effluent
Toxicity Testing Program, (12-13)).

-------
1.3.2 Toxicity Test Method Precision
Like all measurements, toxicity tests exhibit variability. Toxicity test
variability can be described in terms of two types of precision; "within"or intra-
laboratory precision, and round robin or inter-laboratory precision. Intra-
laboratory precision, simply defined, is the ability of trained laboratory personnel
to repeatedly obtain consistent results when performing the same test on the
same species using the same toxicant. Inter-laboratory precision (or round-robin)
tests are a measure of how reproducible a method is when conducted by a large
number of laboratories using the same method, species, and toxicant or effluent.
Generally, intralaboratory results are less variable than interlaboratory results.
Research into the precision of whole effluent toxicity methods by various
groups (including EPA) has shown that toxicity test procedures exhibit a
quantifiable range of variability [35-44]. In toxicity tests, variability is measured
close to the limit of detection because the endpoint of the test is already at the
lower end of the biological method detection range (i.e., an NOEC, IC25, or an
LCI). Coefficients of variation (CV) can not be calculated for NOEC endpoints
because a NOEC is not a statistical point estimate. However, CVs can be
calculated for IC, EC and LC endpoints because they are statistical point
estimates. A more detailed discussion of precision can be found in Box 1-4.
Table 1-6 summarizes the intra-laboratory precision for all 10 EPA short-term
chronic whole effluent toxicity tests and some acute toxicity tests. In addition,
Table 1-7 summarizes the inter-laboratory precision for three chronic test species
and two acute test species using a variety of different compounds.
In summary, whole effluent toxicity testing methods represent practical
tests that estimate potential receiving water impacts, with a quantifiable variation.
Limitations that are correctly developed from whole effluent toxicity tests should
protect aquatic biota if the discharged effluent meets the limitations.
18

-------
Table 1-6
Intra-laboratory Precision of Chronic and Acute Whole Effluent Toxicity Test Methods	
Test	NOEC	Mean	Mean	Water
Method	Ranee	IC25 CV{%) IC50 CV(%) Compound	Used
Cvnrinodon variegates
Survival & Growth
50+; 0 ug/l 4
70
41.8
130
40.8
Copper
SDS
AS
0.5 - 1.0 mg/I„
1.5
31.4
1.9
31.8
AS
31- 125 ug/l
300.4
33.0
396.9
19.2
Copper
NS
1.3 - 2.5 mg/1
2.2
27.6
Z6
35.3
SDS
NS
Embryo larval survival & teratogenicity
t EC10	EC50
200 - 240 ug/l" 202 2.8 233.5 2.5 Copper	AS
2.0 - 4.0 ug/1	1.9 35	11.7 2.9 SDS	AS
Menidia bervtlina - Survival & Growth
31 - 125 ug/1	209.9 43.7 340.8 50.7 Copper	NS
1.3 * 0 ug/1	1.3 43.2 1.9 9.4 SDS	NS
Mvsidopsis bahia —Survival, Growth & Fecundity
<0.3 - 5.0 ug/J 5.7 35.0 6.9 47.8 SDS	NS
63 - 125 ug/1 138.3 18.0 185.8 5.8 Copper	NS
Arbacia punctulata —Fertilization
5.3 - 12.5 ug/l*
23.5
54.6
45.7
47.9
Copper
1.2 - 3.3 mg/1 „
1.7
29.7
2.4
23.3
SDS
<6.1 - 24.4 ug/l
22.9
41.9
33.3
46.4
Copper
0.9 - 1.8 mg/1
2.54
30.3
3.2
33.1
SDS
Champia parvula - Reproduction
0.5 - 1.0 ug/1 „	0.93
0.09 - 0.48 mg/1	230
Pimephales promelas - Survival & Growth
128 - 256 ug/l * ^ -
0.011 - 0.013 mg/1 —
Embryo larval survival & teratogenicity
0.011-0.013 mg/I —
0.011 -0.013 mg/1 —
Ceriodaphnia dubia — Reproduction
0.10-0.30 mg/1*	0.22
Selenastnim capricomuturo- 96 hour Survival
0.49-10.0ug/l"*	0.23
63	1.4 38.6 Copper
60.9 0.3 24.1 SDS
NAPCP2
Cadmium
LCI
0.0068 62
1.51 41.3
41.13 0.3
27.9
Cadium
Diquat
NAPCP
102.6 0.38 75.6 Cadmium
chloride
AS/NS
AS/NS
FW
FW
FW
FW
FW
FW
Notes
* Difference of 1 test concentration AS—artificial seawater
** Differenceof 2 test concentrations NS-naturalseawater
!** Differenceof 4 test concentrations FW-freshwater
J Sodium dodecylsulfate
Sodium pentachlorophenol
— Data not available
19

-------
Box 1-4
Toxicity Test Method Precision
Precision can be described by the mean and relative standard deviation (percent coefficient
of variation, or CV = standard deviation/mean X 100) of the calculated end points from the
replicated toxicity tests. Several factors can affect the precision of the test, including test organism
age, condition, sensitivity, temperature control, salinity, pH control, handling and feeding of the test
organisms, and the training of laboratory personnel. For these reasons, it is recommended that
trained laboratory personnel carefully conduct the tests in strict accordance with the test manuals for
acute and chronic toxicity testing. In addition, acute and chronic toxicity testing quality assurance
practices should be fully performed. Simple quality assurance procedures, which are described at the
beginning of each manual, include:
o single laboratory precision determinations, using reference toxicants, on each of the
tests procedures to determine the ability of the laboratory personnel to obtain
consistent, precise results. These determinations should be made before attempting
to measure effluent toxicity, and routinely confirmed as long as routine whole
effluent toxicity tests are being conducted;
o the use of reference toxicants to routinely evaluate the quality and sensitivity of the
test organisms to be used in each test;
o the development of "Control Charts" which should be prepared for each reference
toxicant/organism/protocol combination to determine if the results are within
prescribed limits. The "Control Chart" consists of successive data added with each
reference toxicant test;
o and the minimum criteria of test acceptability specific for each protocol.
Guidelines for recommended quality assurance practices are found in each manual [42-44].
Within-laboratory precision data are routinely calculated on a minimum of two reference
toxicants as part of the EPA methods development process. These data have been established for
each of the four EPA freshwater chronic methods and each of the six marine/estuarine chronic
methods. Within-laboratory precision is discussed in detail at the end of each of the methods
sections in the Methods Manuals (above) and is summarized in Appendix A (Tables 1 through 17 for
the marine/estuarine methods and Tables 18 through 28 for the freshwater methods). Intra-
laboratory precision data is also presented for acute toxicity tests and is summarized in Table 1-6.
Each laboratory should be establishing a reference toxicant "record",including a control chart.
EPA's reference toxicant numbers are only meant to show precision of the methods within EPA
laboratories and to serve as guidance for other laboratories. Each laboratory's reference toxicant
data will reflect conditions unique to that facility, including dilution water, culturing, etc. However,
each laboratory's reference toxicant CVs should reflect good repeatability.
The CVs may be calculated for acute LC50 and chronic EC50, IC25 and IC50 data. A
mean and range is given for the chronic (NOEC) precision data because an NOEC is not a point
estimate and is dependent on the tightness of the concentration interval employed in the reference
toxicant tests (i.e., the closer the NOEC concentration range the more precise the test is for the
reference toxicant). The closer the CV is to 0, the better. However, CVs should only be compared
with the same test protocol/species tested against the same reference toxicant Estimates of
variability (CVs) should only be applied for specific protocols against a specific chemical using the
same concentration intervals.
Reference toxicant data should be required for each of the methods stipulated by the permit
authority as part of routine QA/QC for checking the reliability of the tests conducted by the
permittees. In addition, Criteria of Acceptability for each of the 10 chronic methods are listed in the
methods manuals, and should be used as a check for whether the compliance data submitted is
minimally acceptable [43] (See Table 1 of each of the 4 freshwater methods.) and (See Table 2 of
20

-------
each of the 10 marine/estuarine methods entitled "Summary of Recommended Effluent Toxicity Test
Conditions")[44],
To date, round robin tests have been completed for the 7-day Fathead Minnow Larval
Survival and Growth Test, the Cladoceran, Ceriodaphnia Survival and Reproduction Test, and the
Sheepshead Minnow Larval Survival and Growth Test. The results of these round-robin studies
show good reproducibility for these 3 methods. In addition inter-laboratory data is presented from
several acute toxicity tests [39]. The data from these round-robin tests can be found in Appendix A
(Tables 5, 21, 22, 23, 24, 25, 26 and 27) and is summarized below in Table 1-7.
21

-------
Table 1 -7
Summary of Inter-fab Variability Data for Whole Effluent Toxicity Test Methods [39. 42-44]
Chronic:
1.
Acute:
Test Method
Cyprinodon varieoatus
7 day growth & survival
Rmephales prometas
7 day growth & survival
Ceriodaphnia dubia
7 day reproduction
Ceriodaphnia dubia
7 day reproduction
Ceriodaphnia dubia
7 day reproduction
Cyprinodon varieoatus
96 hour static
96 hour flow-through
96 hour static
96 hour flow-through
Mvsidopsisbahia
96 hour static
96 hour flow-through
96 hour static
96 hour flow-through
NOEC Range
1 - 3.2% effluent
<3.0 - 6.0 mg/1
potassium chromate
0.25 - 0.30 tng/1
NAPCP(Z}
6 -12% effluent*
<0.25 - 1.0 g/1
NaCl
Toxicant
endosulfan
endosulfan
silver nitrate
silver nitrate
endosulfan
endosulfan
silver nitrate
silver nitrate
IC25 CV(1)
44.2%
31.0
41.1
29.0
IC50CV
56.9%
29.0
27.9
40.0
LC50CV
37.7%
46.2
34.6
50.1
59.5%
51.9
26.6
22.3
Notes
jCV-coefficientof variation
\ NAPCP-Sodium pentachlorophenol
This representsa difference of one exposure concentration
— Data unavailable
22

-------
1.3.3 Considerations Involved in the Implementation of the Whole Effluent
Toxicity Approach
An understanding of some basic considerations and toxicological principles
is important in order to routinely apply the whole effluent approach to the
assessment and control of municipal and industrial effluents. The following seven
sections provide a more in-depth discussion of each of these factors and
principles. (Specific details for characterizing an effluent and deriving permit
limits are discussed in Chapters 3 and 5).
On-site vs. Off-site Toxicity Testing
Comparisons of toxicity data between tests conducted on-site and tests
conducted off-site on samples shipped to Environmental Research Laboratory
(ERL)-Duluth and (ERL)-Narragansett via airfreight have, with a few exceptions,
shown little variation. For many effluents, on-site or off-site test data do not
appear to be significantly different. The major consideration is cost. Cost should
also be weighed against data needs to make the on-site/off-site determination.
For example, if the presence in the effluent of non-persistent compounds
(i.e., chlorine or other volatiles) is suspected or known, then the regulatory
authority may want to conduct on-site testing. If it is not considered important to
the analysis of toxic impact, off-site testing (which is cheaper and can result in the
generation of more data) is as acceptable as on-site testing. In general, offsite
testing would be acceptable for most effluents except those with volatiles. When
conducting flow-through toxicity tests which require a continuously pumped
sample, on-site testing is strongly recommended. Regardless, cost considerations
should not override the need to adequately characterize a given effluent and the
factors unique to the discharge situation.
Flow-through vs. Static and Renewal Toxicity Testing
Several factors must be considered in making the choice of toxicity test
system. These include: the type of toxicity being measured (i.e., is the effluent
highly variable or not; continuous or intermittent discharge?), the amount of data
which needs to be obtained (variable effluents may require more data), and
expense.
Two basic types of testing systems are available to measure effluent
toxicity; flow-through systems and static systems. A flow-through toxicity test is
conducted using a diluter system and a continuous feed of effluent and dilution
water. A static toxicity test is conducted in test chambers (without a serial diluter
delivery system) into which effluent and diluent are added manually. Usually,
only one effluent sample is collected and used at the beginning of a static test. A
variation of the static procedure is the renewal toxicity test which uses the same
delivery system as that of a static test but the test solutions are changed, or
23

-------
renewed, on a pre-determined schedule (i.e., every 24 hours). Fresh effluent
samples are collected to renew the test solutions.
On-line continuous flow-through testing can sample and measure "peaks"
of toxicity should they occur during the testing period. In variable effluents,
however, the test organisms would only be exposed to peak toxicity for periods
proportional to the flow-through rate, the duration of the peak in toxicity and
length of the test. Static and static renewal tests can also measure peaks in
effluent toxicity depending on the type of sampling used, and if the sampling
occurs at the time of the toxicity peak.
If the effluent is highly variable and continuously discharged, either a flow-
through or renewal test would be appropriate. However, the effluent sample for
the renewal should be a 24-hour composite. If the effluent is highly variable with
an intermittent discharge, a flow-through or a renewal test would also be
appropriate, however the effluent sample collected for the renewal test should be
a composite collected over the period of the discharge. If the effluent is not
considered variable, then a static or renewal test using a grab or 24-hour
composite sample would be an appropriate test system.
Cost is also a factor. Flow-through tests are more resource-intensive and
require complex delivery systems. Consequently, less data can be generated than
with static or renewal testing. Where more data at less cost are desirable, static
or renewal testing is probably more appropriate. Therefore, more samples using
renewal is preferable to fewer samples using flow-through for the same total cost
since this would allow better characterization effluent variability.
Grab sampling vs. composite sampling
The use of a grab sample or a composite sample is based upon the
objectives of the test and an understanding of the long-term operations and
schedules of the discharger. If the toxicity of the effluent is variable, grab
samples collected during the peaks of effluent toxicity provide a measure of
maximum toxic effect. Although a grab sample has the potential of revealing the
peak of toxicity in an effluent, the sample has to be collected at the time of the
toxicity spike. Therefore, in a variable effluent, the grab sample has a high
probability of missing the toxicity peak. On the other hand, a 24-hour composite
sample may more readily catch the toxicity peak(s), but the compositing process
may tend to dilute the toxicity resulting in a misleading measure of the maximum
toxicity of the effluent. Composited samples are, therefore, more appropriate for
chronic tests where peak toxicity of short duration is of lesser concern. More
detailed discussions of the type of toxicity tests and the best sampling methods
are provided in the manuals for the acute and chronic, freshwater and marine
toxicity testing procedures [42-44] and in Chapter 3.
24

-------
Variability
There are three important sources of differences in a water quality impact
analysis:
o Effluent variability caused by changes in the composition of the effluent.
Virtually all effluents vary in composition over time.
o Exposure variability caused by changes in flow rates of both effluent and
receiving water. There are also variable receiving water parameters that
are independent of flow, such as background toxicant levels, pH, salinity,
tides, suspended solids, hardness, dissolved oxygen, and temperature, that
can be important in assessing impact.
o Species sensitivity differences caused by the differences in response to
toxicants between species.
Each type of variability is discussed below.
1) Effluent Variability
Effluent variability is an important component in overall variability of
water quality impact analyses and should be adequately addressed in permitting
(See Chapter 5, Permit Requirements). Effluent variability can be addressed by
designing proper sampling and testing procedures. Sampling measurements
should be tailored to the toxic effect of concern (i.e., acute or chronic) and the
need to design testing that accounts for effluent variability. Chapter 3, Effluent
Characterization, describes recommendations for testing frequency designed to
assess variable effluents. Appendix F details suggested sampling procedures.
Appendix A-2 demonstrates the types of variability which may be seen in
POTW effluents as measured through toxicity testing of the effluents (Appendix
A-2, Tables 29 through 37). The CVs for POTW effluents are based on acute
data which range from 19.6-42% effluent, and for chronic data which range from
0-88%. Also in Appendix A-2, Tables 38 through 40 show acute and short-term
chronic effluent variability data from oil refineries on 3 species, fathead minnows,
Ceriodaphnia. and mysids. The CVs associated with this data range from 18.7 -
54% for the acute data, and from 29.8 - 59.6% for the chronic data. Data on
effluent variability in various types of manufacturing facilities is in Tables 41
through 46 of Appendix A-2. Acute toxicity test results show CVs ranging from
20.3 - >100.0%.
Tables 34 through 37 of Appendix A-2 demonstrate the effluent variability
of a POTW effluent over the course of a year in which gradual upgrading to full
secondary treatment was occurring. Four saltwater short-term chronic toxicity
tests were conducted on the POTW's effluent using the sea urchin fertilization
test (Arbacia punctulatal the red macroalga fertilization test (Champia parvulaV
25

-------
the mysid 7-day growth, fecundity and survival test fMvsidopsis bahja), and the
inland silverside 7-day larval growth and survival test (Menidia bervllinaV The
sea urchin and red macroalga tests were conducted daily during each of the four
7-day studies, and provide good examples of the daily variability of the effluent.
These results show that the effluents vary in toxicity and that any one
effluent can exhibit significantly varying toxicity to different test species over time.
The data also indicate that the effluents were rarely toxic below 10% effect
concentration and were not toxic below 0.1% effect concentration. This
information is discussed in the Chapter 3 recommendations for testing the toxicity
of effluents.
2)	Exposure Variability
Exposure variability is a complex factor that can be addressed in two ways.
First, the simplest, easiest applied approach is to assume a steady state exposure
condition (usually an estimate of presumed "worst case" exposure) using a critical
receiving water flow or condition and a typical effluent flow.
A second method is to attempt to estimate or actually measure the
variable exposure situation at the discharge site. This requires statistical analysis
and some form of dynamic modeling. Chapter 4, Exposure and Wasteload
Allocation, describes appropriate exposure assessment procedures for freshwater
and saltwater systems.
3)	Species Sensitivity Differences
One of the primary considerations in establishing a toxicity testing
requirement for a discharger is requiring a suitable test species. Different species
exhibit different sensitivities to toxicants. Often, differences of several orders of
magnitude exist for a given individual toxicant between the least sensitive and the
most sensitive species. This range varies greatly and can be narrow or wide
depending on the individual toxicant involved.
Since the measured toxicity of an effluent will be caused by unknown toxic
constituents, the relative sensitivities of various test species will also be unknown.
Therefore, proper effluent toxicity analysis requires an assessment of a range of
sensitivities of different test species to that effluent. A knowledge of the range is
necessary so that the regulatory authority can identify a test species among the
different species tested that will be most protective of the receiving water. The
only way to assess the range of sensitivities is to test a number of different
species from different taxonomic groups, as in the development of the National
Water Quality Criteria. To determine how many species to test, cost must be
balanced against reducing scientific uncertainty.
To best balance cost with toxicological certainty, EPA recommends an
optimum number of three species, representing three different phyla (i.e., a fish,
26

-------
an invertebrate, and an algae) be used to test an effluent for toxicity. However,
in some cases, the optimum number of species may be fewer or more depending
upon such factors as; how thoroughly the effluent has been characterized, the
available receiving water dilution,-the use classification and existing uses of the
receiving water, as well as other special considerations. For example, if an
effluent has been characterized as highly consistent, with little chance of variation
due to batch processes, changes in raw materials or changes in treatment
efficiency, then the use of the two most sensitive species, or even the one most
sensitive species, may be appropriate as determined on a case by case basis.
Since whole effluents are complex mixtures of toxicants, generalizations
about sensitive and non-sensitive species are difficult to make. For example, a
universally accepted generalization is that trout are considered sensitive organisms
requiring high quality water. However, this generalization may not apply in all
cases, as trout are very sensitive to oxygen depletion, but may be relatively
insensitive to certain toxicants. Another species, Daphnia magna, is very sensitive
when exposed to many toxicants, but relatively insensitive to exposure to the
pesticide endrin. Bluegills are very resistant to metals, particularly copper.
Conversely, bluegills are a sensitive test species for organophosphate pesticides.
Figures 1-4 through 1-6 show the differences in species sensitivities to
hexavalent chromium, dieldrin, and an effluent from a POTW, respectively [47].
The wide range between sensitivities for the different test species is shown.
Comparing the figures shows that the fish, invertebrates and algae shift relative
sensitivities to the effluents/toxicants. The fish are less sensitive to chromium but
more sensitive to dieldrin. For the cladocerans, the reverse is true. The results
of whole-effluent tests using 5 marine/estuarine short-term chronic test methods
also indicate that no species or test method is always the most sensitive. In a
total of 13 effluents tested on-site, Champia parvula was the most sensitive in
15%, Arbacia punctulata in 54%, mysids in 31% and fish in 15% of the cases
[12].
Analysis of species sensitivity ranges found in the National Water Quality
Criteria [54-55] indicates that if tests are conducted on three particular species
(Daphnia magna. Pimephales promelas. and Lepomis macrochirus). the most
sensitive of the three will have an LC50 within one order of magnitude of the
most sensitive of all species tested [48]. This was found to be true for 71 of the
73 priority pollutants tested with four or more species.
Sometimes, regulatory agencies require testing on representative resident
species under the assumption that such tests are needed to assess impact to local
biota. EPA considers it unnecessary to test resident species since standard test
species have been shown to represent the sensitive range of all ecosystems
analyzed [48]. Resident species toxicity testing is strongly discouraged unless it is
required by State statute or some other legally binding factor, or it has been
determined that a unique resident species would be far more protective of the
receiving water than the EPA surrogate species. Such testing is more costly,
27

-------
more difficult, and potentially subject to more variability (disease, age, etc.) than
standardized testing. In any case, organisms taken directly from the receiving
water itself should never be used. The use of other representative species should
be subjected to strict quality assurance and quality control procedures and should
follow rigorous test methodologies which are at least equivalent to EPA methods.
Quality assurance procedures should account for the use of the same species, the
same life stage and age of individuals, acclimation periods to avoid mortality due
to collection, seasonal variations in populations, habitat requirements, health of
the species cultured, as well as the use of reference toxicant tests and other
standard procedures.
Acute-Chronic Ratio
The Acute-Chronic Ratio (ACR) expresses the relationship between the
concentration of whole-effluent toxicity or a toxicant causing acute toxicity to a
species (expressed as an acute toxicity endpoint such as an LC50) and the
concentration of whole-effluent toxicity or a toxicant causing chronic toxicity to
the same species (expressed as a chronic toxicity endpoint such as an NOEC or
its equivalent, i.e., ACR = ATE/CTE or LC50/NOEC). An ACR is commonly
used to extrapolate to a "chronic toxicity" concentration using exposure
considerations and available acute toxicity data when chronic toxicity data for the
species, chemical or effluent of concern is unavailable. The ACR should be
greater than one, since the ratio compares an acute effect concentration with a
chronic effect concentration.
This parameter can be a source of uncertainty in predicting water quality
impact because the ACR varies both between species for a given chemical and,
for any one species, between different toxicants. The latter is a reason why the
ACR for a complex effluent may not be a constant. Regardless of this variability,
when faced with a limited amount of chronic toxicity data, the regulatory
authority must apply some ACR to an effluent or chemical (or decide to collect
more data) when converting wasteload allocations to common terms in the permit
limit derivation process described in Chapter 5.
The ACR may also be used in developing chronic toxicity limitations
where chronic toxicity is not directly measured, in order to minimize testing costs.
Likewise, if the toxicity is for the most part manifested in reproduction, growth
etc. (i.e., non-lethal) endpoints, an acute test may not be appropriate for
compliance monitoring. Of course, where acute and chronic toxicity data are
available, the ACR should be directly calculated for that specific effluent. In
addition to LC50/NOEC, the ACR can be based upon other measures of toxicity
such as chronic IC values; e.g. LC50/IC25.
Data on acute and chronic toxicity for complex effluents from different
categories of dischargers (i.e., POTWs, oil refineries, and chemical manufacturers)
28

-------
Figures 1-4 through 1-6. Differences in species sensitivity
to different toxicants (47).
¦S 3L
O J
£ i
0	2 L
5 i
1	1 L
z
Pimepha/es promeias
Ceriodaphma reticulata	/
• • • •<
J	L
-2.0
-1.0
2.0
0.0	1.0
Log LC50 in mg/1
Figure 1-4. Log of LC50s of freshwater species exposed to hexavalent
chromium.
i	i	'	'	i	i	i	
•3.0	-2.0	-1 0	0.0
Log LC50 in mg/l
Figure 1-5. Log of LC50s of freshwater species exposed to dieldrin.
20	40	GO	80
LC50 as Percent Effluent
Figure 1-6. LC50s of freshwater and saltwater species exposed to
a POTW effluent.
29

-------
show that ACRs for whole effluents range from < 1.0 - > 50.0, with the majority
of ACRs falling below 20 (see Appendix A-3). Acute to chronic ratios for oil
refinery data from 1 plant, based on three species ranged from 1.49 - > 10.0.
Acute to chronic ratios for a variety of chemical manufacturers, based on data
from two species ranged from < 1.0 - > 50.0. Acute to chronic ratios for
POTWs based on two species ranged from 1.4 - 16.1 (these data can be found in
Appendix A-3). Interestingly, this range of ACRs is virtually identical to ACRs
generated on a number of wastewater dischargers in the State of Sao Paulo,
Brazil. Although the acute and chronic toxicities measured in Brazil were
proportionally higher (more toxic) than those measured in the U.S., the ACRs
were quite similar (Appendix A-3, Table 47-49).
Based on experiences over the past five years, EPA recommends that in
the absence of a measured ACR for an effluent, regulatory authorities use an
ACR of 10. Given the protective margin of safety inherent with the use of a
critical flow for the calculation of a chronic receiving water waste concentration,
and given the fact that an NOEC is the next lowest test dilution concentration
below which an observed effect actually occurs, an ACR of 10 should provide
ample protection against chronic instream impacts.
1.4 BIOCRITERIA/BIOASSESSMENT AND BIOSURVEY APPROACH FOR
AQUATIC LIFE PROTECTION
Biological integrity is probably the single best indicator of overall
ecological integrity of aquatic environments because it can provide both a
meaningful goal and a useful measure of environmental status that relates directly
to the overall health of the Nation's waters. To better protect the biological
integrity of aquatic communities, EPA recommends that States develop and
implement biological criteria (biocriteria) in their water quality standards.
Biocriteria are numerical measures or narrative descriptions of biological
integrity. When formally adopted into State standards, biocriteria and aquatic life
use designations serve as direct, legal endpoints for determining aquatic life use
attainment/nonattainment. Per section 131.11(b)(2) of the Water Quality
Standards Regulation (40 CFR Part 131), biocriteria can supplement existing
chemical-specific criteria and provide an alternative to chemical-specific criteria
where such criteria cannot be established. Biocriteria can be quantitatively
developed by identifying unimpaired or least-impacted reference waters that
operationally represent best attainable conditions. Once candidate references are
identified, integrated biological assessments (bioassessments) and biological
surveys (biosurveys) are conducted to substantiate the unimpaired nature of the
reference and characterize the resident community. Because of the complexity of
fully characterizing the biological integrity of an entire aquatic community, State
standards should contain biocriteria that consider various components (measures
of structure and/or function) of the larger aquatic community.
30

-------
1.4.1 Use of Biosurveys and Bioassessments in Water Quality-based Toxics
Control
Bioassessments are the evaluation of the biological condition of a
waterbody using biological surveys, chemical-specific analyses of pollutants known
to impact aquatic life, and/or toxicity tests. Biosurveys are the collection,
processing and analysis of a representative portion of the resident aquatic
community in order to determine its structural and/or functional characteristics.
Biosurveys and bioassessments can be used to directly evaluate the overall
"health" (structure and/or functional characteristics) of an aquatic community.
Deviations from, and threats to, the biological integrity of an aquatic community
can be measured directly using bioassessments and biosurveys.
Biosurveys, however, measure the aggregate effect of impacts upon an
aquatic community where discharges are multiple, complex and variable and
where point, nonpoint and stormwater discharges are all affecting the biological
health of the receiving water. The resident community integrates the effects of
multiple stresses and sources on numerous interactive biological components over
a relatively long period of time. Because of this, biosurveys can not necessarily
measure the impacts of one particular effluent that is being discharged to the
receiving water. However, biosurveys provide a useful monitor of both aggregate
ecological impact and overall temporal trends in the condition of an aquatic
ecosystem. Biosurveys can detect aquatic life impacts that other available
assessment methods may miss, such as impacts caused by pollutants that are
difficult to identify chemically or characterize toxicologically, and impacts from
complex or unanticipated exposures. Perhaps most importantly, biosurveys can
detect impacts caused by habitat degradation such as channelization,
sedimentation and historical contamination which disrupt the interactive balance
among community components.
Biosurvey data should be applied towards:
o Refining use classifications among different types of aquatic systems
and within a given type of use category.
o Defining and protecting existing aquatic life uses and classifying
Outstanding National Resource Water under State antidegradation
policies as required by the Water Quality Standards Regulation.
o Identifying where site-specific criteria modifications may be needed
to effectively protect a waterbody.
o Improving use-attainability studies.
o Assessing impacts of certain nonpoint sources, and together with the
chemical-specific and whole effluent toxicity approaches, controlling
them.
31

-------
o Monitoring the ecological effects of regulatory action taken under
sections 401, 402 and 301(h) of the Clean Water Act.
o Evaluating the effectiveness and documenting the receiving water
biological benefits of pollution controls.
1.4.2 Conducting Biosurveys
As is the case with all types of water quality monitoring programs,
biosurveys should have clear data quality objectives, utilize consistent laboratory
and field methods, and include quality assurance and quality control. Biosurveys
should be tailored to the particular type of waterbody being assessed (e.g.,
wetland, lake, stream, river, estuary, etc.) and should focus on aquatic community
components that are representative of the larger ecosystem and which are
practical to measure. Biosurveys should be routinely coupled with basic chemical
and physical measurements and an objective evaluation of habitat quality.
Over the past several years, EPA's Office of Water has been developing
rapid bioassessment techniques to support biosurveys. Rapid bioassessments are
only one of several techniques available [27 through 32, 34]. The techniques are
an excellent supplementary tool to whole-effluent toxicity testing and chemical-
specific techniques. However, it is important that biosurveys include sampling at
all trophic levels with as many species as possible to accurately reveal receiving
water community impacts.
It is EPA's policy that biosurveys should be fully integrated with toxicity
testing and chemical-specific assessment methods in State water quality
programs. Biological surveys should be used together with toxicity testing and
chemical-specific analyses to assess attainment/nonattainment of designated
aquatic life uses in State water quality standards. Chemical-specific analyses,
toxicity testing and biosurveys can each provide a valid and independent
assessment of receiving water impairment. Note that each assessment also
provides different information. Biosurvey data, for example, measures impact
form all sources of perturbation such as habitat destruction and conventional
pollutants, not just impact from toxicants.
Whenever any one of the three types of assessments demonstrates that the
standard is not attained, appropriate action should be taken by the regulatory
authority. However, since each method has unique as well as overlapping
attributes, sensitivities and program applications, no single approach for
detecting impact should be considered uniformly superior to any other approach.
The inability to detect receiving water impacts using a biosurvey alone is
insufficient evidence to waive or relax a permit limit established using either of
the other methods. The most protective results from each assessment conducted
should be used in the effluent characterization process (Chapter 3). The results
of one assessment technique should not be used to contradict or overrule the
results of the other(s).
32

-------
Excellent examples of biosurvey/bioassessment data collected and used in
concert with toxicity test data are the site studies described in Boxes 1-3 and 1-4.
The toxicity test results and the ambient biosurvey data were based on the
recommended minimum of three 'trophic levels (a fish, invertebrate, and a plant)
to give a good overall picture of what was happening in the receiving water.
Recommended methodologies for conducting biosurveys are included in
references numbered 27 through 32 and 34.
1.5 INTEGRATION OF THE WHOLE-EFFLUENT, CHEMICAL-SPECIFIC
AND BIOASSESSMENT APPROACHES
To more fully protect aquatic habitats and provide more comprehensive
assessments of aquatic life use attainment/nonattainment, EPA recommends that
States fully integrate chemical-specific controls, whole effluent controls, and
bioassessment controls into their water quality-based toxics control programs.
The ultimate objective of the Clean Water Act in section 101(a) states: "The
objective of this Act is to restore and maintain the chemical, physical and
biological integrity of the Nation's waters". Taken together, chemical, physical
and biological integrity define the overall ecological integrity of an aquatic
ecosystem. Regulatory agencies should strive to fully integrate all three
approaches since each has its respective advantages and disadvantages. The
information summarized in Box 1-5, and discussed in detail below, explains how
each approach complements the other and why neither approach should be used
alone.
Box 1-5
Components of an Integrated Approach to Water Quality-based Toxics Control
Control Approach: What it Provides:
What it Doesn't Provide:
Chemical-Specific: -Human health protection.
-Complete toxicology.
-Straightforward treatability.
-Familiarity with control.
-Persistency coverage.
-All toxics present.
-Bioavailability.
-Interactions of mixtures (e.g., additivity)
-Poor trend analysis.
-Accurate toxicology (false assumptions)
Whole effluent toxicity: -Aggregate toxicity.
-All toxics present.
-Bioavailability.
-Accurate toxicology.
-Good trend analysis.
-Human health protection.
-Complete toxicology.
(few species may be tested)
-Simple treatability.
-Persistency coverage.
Bioassessments:
-Actual receiving water effects.
-Excellent trend analysis.
-Severity of impact.
-Total effect of all sources.
-Critical flow effects.
-Straightforward interprtation of results.
-Cause of impact.
-Differentiation of sources.
33

-------
A more detailed discussion of the advantages and disadvantages of the
three approaches is provided below.
1.5.1	Advantages and Disadvantages of the Chemical-Specific Approach
The principal advantages of the chemical-specific approach are:
o Treatment systems are more easily designed to meet chemical
requirements because; 1) more treatability data are available and,
2) treatment engineers and permit writers are more familiar with
the procedures.
o The fate of a pollutant can be more accurately predicted through
modeling.
o Chemical analyses are sometimes less expensive than toxicity testing
(i.e. ammonia and chlorine).
The principal disadvantages of the chemical-specific approach:
o All toxicants in complex wastewaters are not known and, therefore,
control requirements for all toxicants cannot be set. Toxicological
information on these unknown pollutants is unavailable.
o It is not always clear which compound(s) is causing toxicity in the
mixture.
o Measurement of individual toxicants, particularly where many are
present in the mixture, can be expensive. Organic chemicals, in
particular, can be costly to measure.
o The bioavailability of the toxicants at the discharge site are not
assessed, and the interactions between toxicants (e.g., additivity,
antagonism) are not measured or accounted for.
1.5.2	Advantages and Disadvantages of the Whole Effluent Approach
The principal advantages of whole-effluent techniques are:
o The aggregate toxicity of all constituents in a complex effluent is
measured, and toxic effect can be limited by limiting one
parameter- whole effluent toxicity.
34

-------
o Control of the toxicant(s) is not dependent upon established
toxicological information which may not yet be available for some
pollutants.
o Toxicity caused by compounds not commonly analyzed for in
chemical tests is detected.
o The bioavailability of the toxic constituents is assessed, and the
effects of interactions of constituents are measured.
The principal disadvantages of whole-effluent techniques are:
o There is less familiarity with the use of biological techniques for
analyzing effluents, and treatment engineers are less familiar with
designing or manipulating treatment systems to treat the parameter
toxicity.
o Where there are chemical/physical conditions present (pH changes,
salinity changes, photolysis, etc.) that act on toxicants in such a way
as to "release"toxicity away from the discharge point, such toxicity
may not be measured in the effluent. The opposite of this is also
possible; toxicity may degrade rapidly so there is no trace of it away
from the point of discharge.
o The whole effluent toxicity test directly measures only the
immediate bioavailability of a toxicant; it can not measure the
persistence and long-term cumulative toxicity of a compound. Thus,
bioaccumulative chemicals are not necessarily assessed or limited.
o Dischargers of low (non-lethal) concentrations of persistent
toxicants (i.e., metals) may be inadequately detected (e.g., due to
the time length of the toxicity tests) or limited even though the
toxicants can accumulate in sediment to toxic concentrations over a
period of time.
o EPA's water quality criteria are based on a minimum of eight
different species for the acute criteria and three different species for
the chronic criteria. Effluent aquatic toxicity is commonly measured
with only one, two, or three species. For some toxicants a wider
sensitivity range (more species) must be tested; particularly where
the mode of toxicity action is specific (such as diazinon or some
other pesticides).
o In multiple discharge situations, it is possible for the combined
effects of several initially non-toxic effluents to contribute to a
combined toxicity away from the outfall based on additivity of
common toxicants discharged. Limits based only on whole effluent
35

-------
toxicity might not prevent this situation since the effluents
themselves would not be individually toxic but once combined, in
the receiving water, would exceed the criterion for the specific
chemical involved.
o In some cases, TREs can be complicated and expensive.
1.5.3 Advantages and Disadvantages of the Bioassessment Approach
The principal advantages of the bioassessment approach are:
o Biological communities reflect overall ecological integrity.
Biosurvey results directly assess the status of a waterbody.
o Biological communities integrate the effects of different pollutant
stressors and thus provide a holistic measure of their aggregate
impact. Biological assessments also measure stresses over time and
can provide a measure of trends and fluctuating environmental
conditions.
o The status of a waterbody's biological health is of direct interest
and more meaning to the public as a measure of a pollution free
environment.
o Biosurveys can identify previously unknown sources of impairment
and may identify where site-specific chemical criteria are needed.
o Bioassessments can characterize the ecological value to high quality
waters under antidegradation.
The principal disadvantages of the bioassessment approach are:
o Biosurveys measure impact from any source and over long periods
of time and as such, the data cannot be related to a specific
discharge. Regulatory control action based upon such results is
difficult.
o Biosurveys conducted at the wrong time of the year and at wrong
flow conditions may only measure seasonal influences on aquatic
populations rather than impacts of pollutants.
o Biosurvey data are more subject to interpretive differences since the
biosurvey measures degree of impact rather than absolute endpoints
or concentrations.
36

-------
o Biosurvey data that are not comprehensive (i.e. assess as many
aquatic organism groups and their interactions as possible) may not
be fully valuative of impacts.
o Sensitivity of biosurvey data is low, especially in degraded
waterbodies, and decreases as water body size increase.
By using all three approaches, a State will more thoroughly protect aquatic
life. The chemical-specific approach allows for a high accuracy of analysis of the
individual chemical constituents, is highly acceptable among regulatory agencies
and the regulated community, and is generally lowest in cost because of market
availability. However, the protective coverage of the chemical-specific approach
can be low, especially if toxicants are present in an effluent for which no
chemical-specific criteria exists. In addition, some States have adopted very few
criteria as a part of their water quality standards. On the other hand, whole
effluent toxicity provides a high level of coverage by measuring the aggregate
effect of all toxicants, it provides accurate toxicology but can be higher in cost
and has experienced a somewhat lower level of acceptability by permittees and
regulatory authorities. Bioassessments also provide a high level of coverage and
allow for accurate trend analyses, but cost more and data interpretation can be
difficult. The results from bioassessments cannot be targeted at a specific
discharger and used as specific regulatory requirements. Therefore, the
integrated approach to water quality-based toxics control is essential for a strong
toxics control program.
1.5.4 Other Factors Influencing Water Quality-based Toxics Control
Behavior of Toxicants and Toxicity after Discharge
An understanding of the fate and behavior of both single toxicants and
whole-effluent toxicity after discharge can be important in the application of
water quality-based toxics controls. Evaluating the combined effects of interacting
toxic discharges may also be important in multiple discharge situations. When
evaluating the receiving water behavior of toxicants and toxicity, factors such as
toxicity degradation or persistence, and toxicant additivity, antagonism, and
synergism are important. Ambient toxicity tests can give some indication of the
importance of each of these factors:
o Toxicity Persistence - How long and to what extent (in terms of area),
does effluent toxicity or the toxicity of a single toxicant persist after
discharge? It is not reasonable to assume that in all cases the persistence
of both individual toxic chemicals and effluent toxicity is conservative. For
two effluents of equal initial toxicity the aquatic effects of an effluent
whose toxicity degrades rapidly will be different from an effluent whose
toxicity persists.
37

-------
o Additivity, antagonism, and synergism - When toxicants or effluents with
toxic properties mix in the receiving water, what is their combined fate and
toxic effects?
Each of these factors is discussed below.
1) Persistence
As soon as an effluent mixes with receiving water its properties begin to
change. The rate of change of toxicity in that effluent is a measure of its toxicity
persistence or degradation. After mixing, the level of toxicity in the receiving
water may either remain relatively constant (until further diluted), increase in
toxicity due to transformation, or degrade due to fate processes
(photodecomposition, microbial degradation) or compartmentalization processes
(particulate adsorption and sediment deposition, volatilization).
One disadvantage of the chemical-specific approach is that the
bioavailability of the toxicant after discharge is not measured. On-site toxicity
testing has indicated that the individual toxicants causing toxicity measured at
discharge sites tend to be relatively persistent near the point of discharge.
However, persistence of individual chemicals can be modeled and the persistence
of specific toxicants can also be accounted for in making impact predictions and
setting controls. A procedure to determine whether or not an effluent's toxicity is
persistent has been developed by EPA [49]. The procedure describes the steps
required to conduct a laboratory evaluation of the degradation of toxicity in
complex effluents that are released to receiving streams by simplistically
simulating a water body and discharge. EPA recommends this procedure be
conducted where the interaction of sources of toxicants is critical to establishing
controls.
This simple procedure is performed in a refrigerator-sized environmental
chamber in the laboratory using commonly available glassware and shipped
effluent samples. Toxicity is measured using conventional acute or short-term
chronic toxicity tests. The results are used to generate a toxicity degradation rate
for the effluent under representative environmental conditions. The procedure
has several applications, including: 1) measuring the decay of effluent toxicity in
a stream or lake, and 2) identifying the most important fate processes responsible
for toxicity decay (which may also be useful in treatability or toxicity identification
studies).
Mixing zones designated by State water quality standards, or developed on
a case-by-case basis, are usually restrictive enough to render toxicity assessments
as the limiting factor in near field situations. There may be little near field
degradation of some measured effluent toxicity. Degradation of toxicity away
from the outfall may not be an important consideration in these permitting
situations, and toxicity can be considered conservative (non-degrading) in most of
these near field cases.
38

-------
However, effluent toxicity can exhibit far field degradation. Typical
patterns of progressively decreasing downstream toxicity (similar to BOD decay)
have been observed in a number of freshwater situations [2 through 10]. If there
is reason to suspect that an effluent's toxicity is not persistent, several
techniques can be employed to measure changes of toxicity after discharge:
A.	Testing should be performed during various seasons of the year
corresponding to various receiving water flow regimes. The toxicity test
itself, when performed with dilution water immediately upstream or from
an uncontaminated area nearby, is an analogue of the mixing and fate
processes taking place in the receiving water. The types of rapid chemical
reactions found in the mixing zone can also be expected to take place to a
large extent when effluents and receiving waters are mixed for toxicity
tests. The effects on toxicity persistence of varying physical/chemical
conditions in the receiving water or in the effluent cannot, however, be
accurately predicted from these results.
B.	Ambient toxicity testing, as detailed in Appendix C, measures the ambient
interactions of effluent and receiving water and can be used to assess
toxicity persistence.
Toxicity persistence may present a more serious problem in estuarine or
lake receiving waters where the toxicity is not flushed away rapidly. In one study,
on a POTW effluent being discharged into a small cove off of Narragansett Bay,
the decay rate of the effluent was temperature-dependent and was reduced
markedly during the winter. However, persistence of the effluent in the receiving
water cove in the winter did present a problem because tidal flushing did not
remove the toxicity [50].
For coastal discharges, certain toxic compounds [58] are more often found
to cause impacts in estuaries and the marine environment. Due to the physical
and chemical processes that tend to trap pollutants in estuaries (sedimentation,
salinity flux, etc.), the discharge of these compounds, at very low concentrations,
over a long period of time, may allow them to accumulate to toxic concentrations.
For many of these compounds, applicable permit limits may need to be very
stringent to avoid chronic toxicity problems due to the persistence of these
compounds.
2) Additivity
Where multiple toxic effluents are discharged to a receiving water, the
resultant ambient toxicity may equal the sum of the toxicity of each effluent
(assuming no contribution from nonpoint sources and elsewhere). Since each
effluent is composed of individual toxic substances, a mixture of the effluents in a
receiving water produces a mixture of these individual pollutants (assuming
conservative behavior). Thus, additivity of toxic effluents is similar to the
additivity of mixtures of individual chemicals.
39

-------
An extensive review of the additivity phenomenon found:
"Examination of available data using this [additivity] model shows that for
mixtures of toxicants found in sewage and industrial effluents, the joint
acutely lethal toxicity to fish and other aquatic organisms is close to that
predicted assuming simple addition of the proportional contribution from
each toxicant. The observed median value for the joint effect of these
toxicants on fish is 0.95 of that predicted; the corresponding collective
value for sewage effluents, river waters and a few industrial wastes, based
on toxicity of their constituents, is 0.85, while that for pesticides is 1.3.
The less than additive effect of commonly occurring toxicants in some
mixtures may be partly attributable to small fractions of their respective
LC50s having little or no additional effects." (Figure 1-7 illustrates the
data summary). [51]
In relation to chronic toxicity, for the growth of fish, Alabaster and Lloyd
[51] conclude:
"...in the few studies on the growth of fish, the joint effect of toxicants has
been consistently less than additive which suggests that as concentrations
of toxicants are reduced towards the levels of no effect, their potential for
addition is also reduced. There appear to be no marked and consistent
differences between the response of species to mixtures of toxicants."
In summary, the available information tends to indicate that the combined
effects of individual acutely toxic agents are somewhat less than strictly additive.
In addition, chronic toxicities approaching the NOEC do not appear additive.
This suggests that additivity will produce an upper bound at the actual chronic
toxicity concentration.
3)	Antagonism
Cases in which one effluent or pollutant parameter (such as TSS)
ameliorated the toxicity of another effluent pollutant (antagonism) have been
observed. Testing procedures can be designed to measure such interactions. A
description of such a procedure is found in "Recommended Multiple-Source
Toxicity Test Procedures", Box 3-3.
4)	Synergism
Theoretically, under certain conditions, synergism, a greater than additive
increase in toxicity upon mixing, can occur. However, field studies of effluent
toxicity and laboratory experiments with specific chemicals imply that synergism
would be an extremely rare phenomenon. It has not been observed during
on-site effluent toxicity studies, and is not considered an important factor in the
toxicological assessment of effluents.
40

-------
Note
a) represents effluent having a high content of industrial
wastes including pesticides
legend-
- Constituents of sewage and mdustnal wastes
• River waters
® Sewage effluents
a Gas liquor
& Drilling fluid
— — Pesticides and other substances
Times as Toxic as Predicted
(from ITU I
Figure 1-7. Data summarizing the property of additivity (51).
41

-------
1.6 HUMAN HEALTH PROTECTION
Impacts on human health due to exposure to waterborne toxicants can
occur through three primary exposure routes: contact recreation, drinking water,
and the ingestion of contaminated fish and shellfish tissues. Contact recreation
may pose potential risks due to dermal absorption and incidental ingestion.
Exposure through drinking water is a significant concern but can be mitigated by
the application of drinking water criteria. The third exposure route, human
consumption of contaminated aquatic life, is of primary concern in this document
due to two factors: 1) the potentially high concentrations achieved in fish and
shellfish tissues due to bioconcentration, and 2) no permitting controls exist
between tissue contamination and human exposure. For these reasons, this
document focuses on prevention of bioconcentration as the principal way to
control human exposure to waterborne toxicants [52].
Currently, the regulation of human health impacts can be based only upon
the control of individual chemicals. EPA human health water quality criteria
protect against the consumption of contaminated water and aquatic life. There is
no mechanism like the aquatic toxicity test to determine the effect of a chemical
mixture like an effluent on human health. EPA has developed, however, a
preliminary approach to analyzing effluents for bioaccumulation potential through
the use of a whole effluent bioconcentration analysis followed by identification of
individual bioconcentratable pollutants. This procedure is described in Chapter 3.
1.6.1 Types of Health Effects
Health effects from toxic pollutants are divided into two categories:
nonthreshold effects ~ such as carcinogenicity, and threshold effects ~ such as
acute, subacute or chronic toxicity. Both terms are defined below.
EPA's approach to assessing the risks associated with nonthreshold human
carcinogens is different from the approach for threshold toxicants due to the
different mechanisms of action thought to be involved. In the case of
carcinogens, the Agency assumes that a small number of molecular events can
evoke changes in a single cell that can lead to uncontrolled cellular proliferation.
This mechanism for carcinogenesis is referred to as "nonthreshold",since there is
essentially no level of exposure for such a chemical that does not pose a small,
but finite, probability of generating a carcinogenic response. Genotoxic pollutants
are presumed to have no threshold level, but incremental risk levels can be
determined based on the carcinogenic potency of the chemicals.
Threshold toxicants, on the other hand, are generally treated as if there is
an identifiable exposure threshold (both for individuals and populations) below
which effects are not observable. Threshold toxicants are chemicals that give rise
to toxic endpoints other than cancer because of their effects on the function of
various organ systems. Such chemicals are presumed to have safe exposure
levels. This characteristic distinguishes threshold end-points from nonthreshold
42

-------
end points. However, it should be noted that chemicals which cause cancer and
mutations also commonly evoke other toxic effects (systemic toxicity). In the case
of systemic toxicity, compensating and adaptive "defense"mechanisms exist that
must be overcome before the toxic end point is manifested. For example, there
could be a large number of cells performing the same or similar function whose
population must be significantly altered before the effect is seen. The individual
threshold hypothesis holds that a range of exposures from zero to some finite
value can be tolerated by the organisms with essentially no chance of expression
of the toxic effect.
This threshold concept is important in the regulatory context. Further, it is
often prudent to focus on the most sensitive members of the population;
therefore, regulatory efforts are generally made to keep exposures below the
population threshold, which is defined as the lowest of the thresholds of the
individuals within a population.
Currently, the control of toxicants which bioconcentrate in edible tissues is
achieved in the NPDES program by limiting such pollutants individually. There
are whole effluent tests which can measure a wastewater's potential to cause
carcinogenicity or mutagenicity (e.g., Ames test). However, the application of
such data is experimental because of the difficulty in establishing cause/effect
relationships between exposure to wastewaters and human health problems.
Therefore, at this time EPA recommends regulatory authorities focus on controls
for bioconcentratable toxicants on a chemical-by-chemical control basis.
The remaining information regarding regulation of human health impacts
is contained in the following chapters: Chapter 2, Water Quality Standards,
discusses the development and updating of human health water quality criteria.
Chapter 3, Effluent Characterization, discusses the evaluation of effluents for
potential human health impacts. Chapter 4, Exposure and Wasteload Allocation,
contains information on design conditions and averaging periods. Finally, Chapter
5, Permit Requirements, discusses the derivation of permit limits protective
against human health impacts.
43

-------
REFERENCES
1.	Bergman, H., R. Kimerle, and A. W. Maki, (editors). 1985.
Environmental Hazard Assessment of Effluents. Pergamon Press, Inc.,
Elmsford, NY.
2.	Mount, D., N. Thomas, M. Barbour, T. Norberg, T. Roush, and R.
Brandes. 1984. Effluent and Ambient Toxicity Testing and Instream
Community Response on the Ottawa River, Lima, Ohio. Permits Division,
Washington, D.C., Office of Research and Development, Duluth, MN,
EPA-600/3-84-080, August, 1984.
3.	Mount, D. I., and T. J. Norberg-King, (editors). 1985. Validity of Effluent
and Ambient Toxicity Tests for Predicting Biological Impact, Scippo Creek,
Circleville, Ohio. U.S. Environmental Protection Agency,
EPA/600/3-85/044, June, 1985.
4.	Mount, D. I., et al., (editors). 1985. Validity of Effluent and Ambient
Toxicity Tests for Predicting Biological Impact, Five Mile Creek,
Birmingham, Alabama. U.S. Environmental Protection Agency, EPA
600/8-85/015.
5.	Mount, D. I., A.E. Steen, and T. Norberg-King, (editors). 1986. Validity
of Effluent and Ambient Toxicity Tests for Predicting Biological Impact,
Back River, Baltimore Harbor, Maryland. U.S. Environmental Protection
Agency, EPA 600/8-86/001, July 1986.
6.	Mount, D. I., T. Norberg-King, and A.E. Steen, (editors). 1986. Validity
of Effluent and Ambient Toxicity Tests for Predicting Biological Impact,
Naugatuck River, Waterbury, Connecticut. U.S. Environmental Protection
Agency, EPA 600/8-86/005, May 1986.
7.	Norberg-King, T.J., D.I. Mount (editors). 1986. Validity of Effluent and
Ambient Toxicity Tests for Predicting Biological Impact, Skeleton Creek,
Enid, Oklahoma. U.S. Environmental Protection Agency, EPA 600/8-
86/002, March 1986.
8.	Mount, D. I., A.E. Steen, T. Norberg-King, (editors). 1986. Validity of
Effluent and Ambient Toxicity Tests for Predicting Biological Impact, Ohio
River, Wheeling, West Virginia. U.S. Environmental Protection Agency,
EPA 600/3-85/071, March 1986.
44

-------
9.	Mount, D. I. and T. Norberg-King, (editors). 1986. Validity of Effluent
and Ambient Toxicity Tests for Predicting Biological Impact, Kanawha
River, Charleston, West Virginia. U.S. Environmental Protection Agency,
EPA 600/3-86/006, July 1986.
10.	U.S. Environmental Protection Agency. 1987. Biomonitoring to Achieve
Control of Toxic Effluents. EPA 625/8-87/013.
11.	Eagleson, K.W., D.L. Lenat, L. Ausley, and F. Winborne. 1988.
Comparison of measured in-stream biological responses with responses
predicted by Ceriodaphnia chronic toxicity tests. North Carolina Division
of Environmental Management, Raleigh, North Carolina.
12.	Schimmel, S.C., G.E. Morrison and M.A. Heber. 1989. Marine Complex
Effluent Toxicity Program: Test Sensitivity, Repeatability, and Relevance
to Receiving Water Impact. Env. Toxicol, and Chem., 5: 739-746.
13.	Schimmel, S.C., G.B. Thursby, M.A. Heber, and M.J. Chammas. 1989.
Case Study of a Marine Discharge: Comparison^ Effluent and Receiving
Water Toxicity. In: Aquatic Toxicology and Environmental Fate:
Eleventh Volume, ASTM STP 1007, G.W. Suter II and M.A. Lewis, Eds.,
American Society for Testing and Materials, Philadelphia, 1989, pp. 159-
173.
14.	Geckler, J.R., W.B. Horning, T.N. Neiheisel, Q.H. Pickering, E.L.
Robinson, C.E. Stephan. 1976. Validity of Laboratory Tests for Predicting
Copper Toxicity in Streams. EPA 600/3-76-116.
15.	Hedke, S., C. West, K.N. Allen, T. Norberg-King, D. Mount. 1986. The
toxicity of PCP to aquatic organisms under naturally varying and controlled
environmental conditions. Env. Toxicol. Chem. 5:531-542.
16.	Hedke, S.F. and J.W. Arthur. 1985. Evaluation of a site-specific water
quality criterion for pentachlorophenol using outdoor experimental
streams. Aquatic Toxiciolgy and Hazard Assessment: Seventh Symposium.
ASTM STP 854, R.D. Cardwell, T. Purdy, and R.C. Bahner, EDS.
American Society for Testing and Materials, Philadelphia, Pa. pp. 551-564.
17.	Yount, J.D. and J.E. Richter. 1986. Effects of pentachlorophenol on
periphyton communities in outdoor experimental streams. Arch. Environ.
Contam. Toxicol. 15; 51-60.
18.	Zischke. 1985. Effects of pentachlorophenol on invertebrates and fish in
outdoor experimental channels. Aquatic Toxicology 7:37-58.
45

-------
19.
20.
21.
22,
23.
24.
25.
26.
27.
28.
29,
30,
Arthur, J.W., C.W. West, K.N. Allen and S.F. Hedtke. 1987. Seasonal
toxicity of ammonia to five fish and nine invertebrate species. Bull.
Environ. Cont. Tox. 38(2): 324-331.
Hermanutz, R.O., S. F. Hedtke, J.W. Arthur, R.W. Andrew, and J.C.
Helgen. 1987. Ammonia effects on microinvertebrates and fish in outdoor
experimental streams. Environ. Pollut. 47: 249-283.
Zischke, J.W. and J.W. Arthur. 1987. Effects of elevated ammonia levels
on fingernail clams (Musculium transversum"! in outdoor experimental
streams. Arch. Environ. Cont. Tox. 16: 225-231.
Hermanutz, R.O., K.N. Allen, and S.F. Hedtke. (In Press) Toxicity and
fate of total residual chlorine in outdoor experimental streams.
Proceedings of Eigth Conference on Water Chlorination.
Monticello Ecological Research Station. 1988. The impact of
Chlorine/ammonia on ecosystem structure and function in experimental
streams. Report to the Office of Water, U.S. EPA. March, 1988. 44p.
Newman, R.M., J.A. Perry. E. Tam, and R.L. Crawford. 1987. Effects of
chronic chlorine exposure on litter processing in outdoor experimental
streams. Freshwater Biology 18: 415-428.
Perry, J.A., N.H. Troelstrup Jr., M. Newson and B. Shelley. Generality.
Wat. Sci. Tech. 19(11): 55-71.
U.S. Environmental Protection Agency. 1984. Technical Support
Document for Conducting Use Attainability Analysis. Office of Water
Regulations and Standards, Washington, D.C.
Plafkin, J.L. et. al. 1989. Rapid Bioassessment Protocols for Use in
Streams and Rivers. Office of Water Regulations and Standards, EPA
444/4-89-001.
Karr, J.R., et. al. 1986. Assessing Biological Integrity in Running Waters:
A Method and its Rationale, 111. Nat. Hist. Survey Special Publ. 5.
Ohio EPA. 1987. Biological Criteria for the Protection of Aquatic Life:
Vol I, II, and III. Division of Water Quality Monitoring and Assessment.
Columbus, Ohio.
Lenat, D.R. 1988. Water Quality Assessments of Streams Using a
Qualitative Collection Method for Benthic Macroinvertebrates, J.N. Am.
Benthol. Soc. 7:222.
46

-------
31.	Schackleford, B. 1988. Rapid Bioassessments of Macroinvertebrate
Communities: Biocriteria Development. Arkansas Dept. Poll. Contr. and
Ecol., Little Rock, AR.
32.	Maine DEP. 1987. Methods for Biological sampling and Analysis of
Maine's Waters Maine Bureau of Water Quality Control.
33.	Proceedings of the First National Workshop on Biological Criteria. 1988.
EPA 905/9-89/003. U.S. Environmental Protection Agency, Region 5:
Chicago, Illinois.
34.	Weber, C.I. 1973. Biological Field and Laboratory Methods for
Measuring the Quality of Surface Waters and Effluents. EMSL -
Cincinnati. EPA 670/4-73/001.
35.	Nebeker, A. 1982. Evaluation of a Daphnia magna. Renewal Life-cycle
Test Method with Silver and Endosulfan. Water Research. Vol. 16, pp.
739-744.
36.	Grothe, D., and R. Kimerle. 1985. Inter- and Intra- laboratory Variability
in Daphnia magna. Effluent Toxicity Test Results. Env. Tox. and Chem.
Vol. 4(2), pp. 189-192.
37.	Qureshi, A. D., K. W. Flood, S. R. Thompson, S. M., Junhurst, C. S. lnniss,
and D. A. Rokosh. 1982. Comparison of a Luminescent Bacterial Test
with Other Bioassays for Determining Toxicity of Pure Compounds and
Complex Effluents, pp. 179-195. In J. G. Pearson et al. (eds.), Aquatic
Toxicology Hazard Assessment: Fifth Conference, ASTM STP 766.
American Society for Testing and Materials, Philadelphia, PA.
38.	Strosher, M. T. 1984. A Comparison of Biological Testing Methods in
Association with Chemical Analyses to Evaluate Toxicity of Waste Drilling
Fluids in Alberta. Volume I. Canadian Petroleum Association, Calgary,
Alberta.
39.	Schimmel, S. C. 1981. Results: Interlaboratory Comparison of Acute
Toxicity Tests Using Estuarine Animals. EPA-600/4-81-003.
40.	U.S. Environmental Protection Agency. 1982. Pesticide Assessment
Guidelines. Office of Pesticide Programs, Washington, D.C.,
EPA/9-82-018 through 028.
41.	U.S. Environmental Protection Agency. 1982. Toxic Substances Test
Guidelines. Office of Toxic Substances, Washington, D.C.,
EPA/16-82-001 through 003.
47

-------
42.	Peltier, W. and C.I. Weber. 1985. Methods for Measuring the Acute
Toxicity of Effluents to Aquatic Organisms, Third Edition. Office of
Research and Development, Cincinnati, OH. EPA-600/4-85-013.
43.	Weber, C.I. et. al. (ed.). 1989. Short-Term Methods for Estimating the
Chronic Toxicity of Effluents and Receiving Waters to Freshwater
Organisms, Second Edition. Office of Research and Development,
Cincinnati, OH. EPA-600/4-89/001.
44.	Weber, C.I. et. al. (ed.). 1988. Short-Term Methods for Estimating the
Chronic Toxicity of Effluents and Receiving Waters to Marine and
Estuarine Organisms. Office of Research and Development, Cincinnati,
OH. EPA-600/4-87/028.
45.	U.S. Environmental Protection Agency. 1979. Handbook for analytical
quality control in water and wastewater laboratories. Environmental
Monitoring and Support Laboratory, Cinn., Ohio. EPA 600/4-79-019.
46.	U.S. Environmental Protection Agency. 1979. Methods for chemical
analysis of water and wastes. Environmental Monitoring and Support
Laboratory, Cinn., Ohio. EPA 600/4-79-020.
47.	LeBlanc, G. A. 1984. Interspecies Relationships in Acute Toxicity of
Chemicals to Aquatic Organisms. Env. Tox. and Chem. 3(1), pp.47-60.
48.	Kimerle, R. A., A. F. Werner, and W. J. Adams. 1984. Aquatic Hazard
Evaluation Principles Applied to the Development of Water Quality
Criteria. In Aquatic Toxicology and Hazard Assessment; Seventh
Symposium, ASTM STP 854. R. D. Cardwell, R. Purdy, and R. C. Bahner
(eds.) American Society for Testing and Materials, Philadelphia.
49.	U.S. Environmental Protection Agency. 1989. Method for Conducting
Laboratory Toxicity Degradation Evaluations with Complex Effluents.
Battelle Report, March 1989.
50.	Dettmen, E.H., J.F. Paul, J.S. Rosen, C.J. Strobel. 1989. Transport, Fate,
and Toxic Effects of a Sewage Treatment Plant Effluent in a Rhode Island
Estuary. U.S. EPA/ORD ERL-Narragansett, Contribution # 1003.
51.	Alabaster, J. 5., and R. Lloyd (editors). 1982. Water Quality Criteria for
Fish. 2nd edition. Butterworths, London, 361 pp.
52.	U.S. Environmental Protection Agency. 1990. Document in draft.
Guidance on Assessment, Criteria Development, and Control of
Bioconcentratable Contaminants in Surface Waters. Office of Water
Enforcement and Permits, Washington, D.C.
48

-------
53.	Birge, W.J., J.A. Black, T. M. Short and A.G. Westerman. 1989. A
Comparative Ecological and Toxicological Investigation of a Secondary
Wastewater Treatment Plant Effluent and its Receiving Stream.
Environmental Toxicology and Chemistry 8:437-450.
54.	U.S. Environmental Protection Agency. 1987. Quality Criteria for Water.
Office of Water Regulations and Standards, Washington, D.C. EPA
440/5-86-001.
55.	U.S. Environmental Protection Agency. 1983. Water Quality Standards
Handbook. Office of Water Regulations and Standards, Washington, D.C.
56.	U.S. Environmental Protection Agency. September, 1989. Biomonitoring
for the Control of Toxicity in Effluent Discharges to the Marine
Environment. EPA 625/8-89/015.
57.	Morrison, G., E. Torello, R. Comeleo, R. Walsh, A. Kuhn, R. Burgess, M.
Tagliabue, and W. Greene. 1989. Intralaboratory Precision of Saltwater
Short-term Chronic Toxicity Tests. Res. J. W.P.GF., 61 (11/12): 1707-
1710.
58.	U.S. Environmental Protection Agency. 1989. Pollutants of Concern.
Office of Marine and Estaurine Protection, Wash., D.C.
49

-------
2. WATER QUALITY CRITERIA AND STANDARDS
2.1 INTRODUCTION
The starting point for a water quality-based toxics control program is the
determination of the level of water quality adopted by a State for the waterbody.
The following discussion describes regulatory and technical considerations for
water quality criteria and standards designed to protect aquatic life and human
health.
2.1.1 Overview of Regulatory Considerations
Water quality criteria refer to scientifically derived ambient concentrations
which are developed by EPA or States for various pollutants of concern. Criteria
developed by EPA are recommended levels which should not be exceeded in a
body of water in order to protect aquatic life and human health. A criterion, in
some cases, may be a narrative statement instead of a concentration. EPA has
published a total of 35 recommended criteria for aquatic life and 100 for human
health. The method by which these numbers are derived are significantly
different and will be discussed respectively in sections 2.3 and 2.4.
A water quality standard defines the water quality goals of a water body,
by designating the use or uses to be made of the water, by setting criteria
necessary to fully protect the uses, and by preventing degradation of water quality
through antidegradation provisions. States adopt water quality standards to
protect public health or welfare, enhance the water quality, and serve the
purposes of the Clean Water Act (CWA). "Serve the purposes of Act" (as
defined in Section 101(a), 101(a)(2), and 303(c) of the CWA) means that water
quality standards should: 1) include provisions for restoring and maintaining
chemical, physical and biological integrity of State waters, 2) provide, wherever
attainable, water quality for the protection and propagation of fish, shellfish, and
wildlife and recreation in and on the water ("fishable/swimmable"), and 3)
consider the use and value of State waters for public water supplies, propagation
of fish and wildlife, recreation, agriculture and industrial purposes, and
navigation.
Use classifications describe the uses for which each state intends its waters
to be suitable (e.g., public water supplies). The CWA requires each state to
classify all of the waters within its boundaries according to intended use. Water
quality standards, including use classifications, are to be reviewed by the States
and, where appropriate, modified at least every three years.
State water quality standards require that if a designated use is currently
being attained, the water body may not be classified for a less stringent use.
50

-------
Procedures for reclassification are established under state laws and may require
hearings, state environmental agency review and sometimes legislative action.
All state reclassification of water quality uses are subject to EPA review and
approval in accordance with the provisions of the CWA.
A numerical water quality criterion within a State standard may use an
EPA water quality criterion as a basis for regulation or enforcement. However,
the State standards may differ from an EPA water quality criterion because of
prevailing local natural conditions. Some criteria apply State-wide, whereas
others are specific to particular designated uses or water bodies.
All States have both numeric and narrative toxics criteria (e.g., "free from
toxic pollutants in toxic amounts") in their standards. A State toxics control
program can use either type of criterion as the basis for assessment and control
of toxic pollutants and toxicity. Where States use narrative criteria, they must
identify a method by which the State intends to regulate point source discharges
of toxic pollutants on water quality-limited segments (40 CFR 131.11). EPA also
recommends that States develop a criterion for the parameter " toxicity" (See
section 2.3.3 for EPA's recommended criterion.). In the absence of any state
numerical standard for toxics, EPA criteria may be used to define expected levels
of toxicity and help interpret a narrative standard. Also, EPA's proposed revision
to the water quality standards regulation stipulates that States may adopt
biological criteria which include sufficient detail to describe attainable
characteristics of designated aquatic life uses based on appropriate reference
conditions and measure compliance with designated and/or existing aquatic life
uses.
In addition, the water quality standards regulation (40 CFR 131.11(b)(1)
allows states to develop site-specific criteria to control toxicity. Site-specific
criteria development are most appropriate on water quality limited water bodies
where: 1) Background water quality parameters, such as pH, hardness,
temperature, suspended solids, etc., appear to differ significantly from the
laboratory water used in developing the Section 304(a) criteria; or 2) The types of
local aquatic organisms in the region differ significantly from those actually tested
in developing the Section 304(a) criteria. In either case, such criteria must be
subjected to public comment and EPA approval prior to their use in setting
permit requirements (Section 303(c) of the CWA). The State should briefly
describe the process for development and adoption of site-specific criteria.
Numeric criteria can be used to limit specific chemicals where the cause of
toxicity is known or for protection against potential human health impacts. The
narrative standard can be the basis for limiting toxicity where a specific toxic
pollutant can be identified as causing the toxicity, but there is no numeric
criterion in State standards. The narrative standards can also be used to limit
whole effluent toxicity where it is not known which chemical or chemicals are
causing the toxicity.
51

-------
Section 308(d) of the Water Quality Act of 1987 (Pub. L. 100-4)
established a new Section 303(c)(2)(B) in the CWA (33 U.S.C. 1313(c)(2)(B) that
requires State adoption of numeric criteria for the toxic pollutants listed pursuant
to Section 307(a)(1) of CWA for which EPA numeric criteria are available. For
those pollutants for which Section 304(a) criteria have been published by EPA,
States must establish numeric criteria at levels necessary to support the
designated uses, where the discharge or presence of toxic pollutants could
reasonably be expected to interfere with designated uses. If Section 304(a)
criteria recommendations are not available for a specific toxic pollutant, the
amendment provides that States shall adopt criteria generated through the
application of biological testing procedures.
On December 2, 1988, EPA sent "Guidance for State Implementation of
Water Quality Standards for CWA Section 303(c)(2)(B)" to each of its Regions
and to each State water pollution control agency. The guidance contained three
options for implementing the new numeric criteria requirements of the Act: (1)
adopt Statewide numeric criteria in standards for all those priority pollutants for
which EPA has published national criteria; (2) adopt numeric criteria for only
those priority pollutants and those stream segments whete the discharge or
presence of the pollutant could reasonably be expected to interfere with
designated uses; or (3) adopt a specific procedure in the standards to "translate"
the State's narrative "free from toxics" standard to a derived numeric criteria.
2.1.2 Water Quality Standards and State Toxics Control Programs
Existing regulations at Section 131.11, as well as recent amendments to
EPA's regulations covering the national surface water toxics control program
(Section 122.44), include specific requirements for controlling toxic pollutants
where a state numeric criterion is available, as well as where no state numeric
criterion has yet been promulgated.
40 CFR 122.44(d)(l)(v) and (d)(l)(vi) provide the legal basis for assessing
problems and developing controls in the absence of a promulgated numeric
standard and when there is an excursion above a State narrative water quality
standard. There is flexibility in a State's interpretation of its narrative water
quality criterion provided under Section 122.44(d)(l)(v) and (vi). Section
122.44(d)(l)(v) requires the permitting authority to use effluent limitations on
whole effluent toxicity. If, however, chemical-specific effluent limitations are
demonstrated to be adequate to achieve all applicable water quality standards,
then Section 122.44(d)(l)(v) allows the permitting authority to forego a
limitation on whole effluent toxicity. Section 122.44(d)(l)(vi) requires that where
an individual chemical of concern has been identified, for which no national
numeric criterion exists, the permitting authority may use a proposed State
numeric criterion or an explicit State policy or a proposed or final State
regulation interpreting its narrative criteria, provided the interpretation will fully
protect the designated use.
52

-------
In addition, State policies and procedures on antidegradation and
antibacksliding are important to the States' toxics control programs. States should
have clearly documented policies and procedures for implementing
antidegradation and procedures for implementing antidegradation and
antibacksliding that conform to the requirements of the amendments to the CWA
in Sections 303(d) and 402.
The antidegradation policy reflects the objective of the Act to "... restore
and maintain the chemical, physical, and biological integrity of the Nation's
waters." EPA's regulations require each State to adopt, as part of their water
quality standards, an antidegradation policy consistent with 40 CFR 131.12. The
regulation also requires States to have implementation methods for their
antidegradation policies, i.e., decision criteria for assessing activities that may
impact the integrity of a water body. Activities covered by the antidegradation
policy and implementation methods include both point and nonpoint sources of
pollution.
Backsliding from water quality-based effluent limits is governed by section
402(o) of the Clean Water Act (CWA). This section establishes a conditional
prohibition against backsliding, unless the permittee is able to meet one of the
exceptions set forth in either section 303(d)(4), or section 402(o)(2). In no event,
however, may an effluent limitation be revised to be less stringent than necessary
to meet applicable water quality standards or effluent guidelines.
Water quality standards, WLAs and permit limitations must conform to
existing requirements governing both antidegradation and antibacksliding. States
must implement their antidegradation policy as an integral part of their toxics
control programs. It is important, at a minimum, for States to have a good
understanding of federal requirements regarding antidegradation and
antibacksliding since EPA may have to veto State permits that do not conform to
the relevant rules and procedures.
The remainder of this chapter discusses and explains in more detail the
development process of water quality criteria and standards. In particular, this
chapter addresses: 2.2 General Considerations; 2.3 Water Quality for Aquatic Life
Protection; 2.4 Water Quality for Human Health Protection; 2.5 Biocriteria; and
2.6 Sediment Criteria.
2.2 GENERAL CONSIDERATIONS
2.2.1 Magnitude, Duration, and Frequency
As stated earlier, criteria are specifications of water quality designed to
ensure protection of the designated use. EPA criteria are developed as national
recommendations to assist States in developing their standards and to assist in
interpreting narrative standards. EPA criteria consist of three components:
53

-------
o Magnitude - How much of a pollutant (or pollutant parameter such as
toxicity), expressed as a concentration, is allowable.
o Duration - The period of time (averaging period) over which the instream
concentration is averaged for comparison with criteria concentrations.
This specification limits the duration of concentrations above the criteria.
o Frequency - How often criteria can be exceeded without unacceptably
affecting the community.
In the case of criteria for protection of aquatic life, specifications for these
components are developed for two levels of aquatic life effects. These levels are
acute effects and chronic effects (the criterion for dissolved oxygen is further
subdivided). Thus, a criterion will consist of at least six specifications of
recommended magnitude, duration, and frequency values for both acute and
chronic levels of protection (See Appendix D for a detailed discussion on
Duration and Frequency).
A typical aquatic life water quality criteria statement contains a
concentration limit, averaging period, and return frequency and is stated in the
following format:
"The procedures described in the 'Guidelines for Deriving National
Water Quality Criteria for the Protection of Aquatic Organisms and Their
Uses'" indicate that, except possibly where a locally important species is
very sensitive, (1^ aquatic organisms and their uses should not be affected
unacceptably if the four-day average concentration of (2) does not exceed
ug/L more than once every three years on the average and if the one-
hour average concentration does not exceed (4^> ug/L more than once
every three years on the average."
In this example generic statement, the following terms are inserted:
at (1) - either "freshwater" or "saltwater",
at (2) - the name of the pollutant,
at (3) - the lower of the chronic-effect or residue-based concentrations as
the Criterion Continuous Concentration, and
at (4) - the acute effect-based Criterion Maximum Concentration.
Defining water quality criteria with an appropriate duration and frequency
of exceedence helps to ensure that criteria are appropriately considered in
developing WLAs which are then translated into permit requirements. Duration
and frequency may be defined in the design flow appropriate to the criterion.
However, in these cases, the State should provide an evaluation that the selected
design flow approximates the recommended duration and frequency.
54

-------
2.2.2 Mixing Zones
Many State water quality standards allow a regulatory zone of mixing in
which less stringent criteria apply than apply to the rest of the waterbody. The
rationale is that a small area of degradation can exist without causing adverse
effects on the overall waterbody. The criteria that apply to mixing zones vary
from State to State. In some States there are explicit requirements for water
quality within mixing zones (such as no acute toxicity, floating materials, or
deposit-forming solids). In other States there are no requirements or the
requirements are ambiguous. EPA's recommends that the States have a
definitive statement in their standards on whether or not mixing zones are
allowed. Where mixing zone provisions are part of the State standard, the State
should describe the methodologies and procedures used for developing mixing
zones.
Different considerations apply for aquatic life and human health protection
(see chapter 4 for a detailed discussion). For aquatic life protection a mixing
zone may be permitted as long as lethal conditions are prevented within the
mixing zone. Chronic aquatic life criteria are met at the edge of the mixing zone.
EPA has not established a national policy specifying the point of application in
the receiving water that should be used with human health criteria. However,
EPA's goal is the eventual elimination of the discharge of persistent
bioaccumulative pollutants. Determinations of the appropriateness of mixing
zones for bioaccumulative pollutants are at the discretion of the regulatory
authority. However, where unsafe fish tissue levels or other evidence indicates
that a bioaccumulative pollutant is consumed and incorporated into the aquatic
organisms, special care should be taken in calculating discharge limits. In
particular, relaxed discharge limits due to the provision of a mixing zone may not
be appropriate in these situations.
EPA's basis for not prohibiting mixing zones for bioaccumulative pollutants
is as follows. Although EPA relies on the best available scientific techniques, the
derivation and application of criteria do not precisely consider all factors relating
to a pollutant's environmental fate. Thus water quality modeling, including the
use of mixing zones, is an approximate, and subject to potential error when
depicting the real world. Given such circumstances, EPA believes that a blanket
prohibition on mixing zones for bioaccumulative pollutants is overly conservative.
Thus, regulatory authorities should carefully consider site specific factors, such as
the hydrology of the site, and ambient analytical data, in water or biota, on the
bioaccumulative pollutant. EPA recognizes that this may lead to the possibility of
State by State inconsistency, but EPA believes that this is one example of the
appropriateness of flexibility for the State to determine how environmentally
conservative they want to be, within the bounds allowed by the CWA. EPA's
policy is described in more detail in the Handbook [6] and technical
considerations in establishment of mixing zones are discussed in detail in
Chapter 4.
55

-------
2.3 WATER QUALITY FOR AQUATIC LIFE PROTECTION
2.3.1	Development Process for Criteria
The development of national numerical water quality criteria for the
protection of aquatic organisms is a complex process that uses information from
many areas of aquatic toxicology. (See the "Guidelines for Deriving Numerical
National Water Quality Criteria for the Protection of Aquatic Organisms and
Their Uses" [9] for a detailed discussion of this process.) After a decision is made
that a national criterion is needed for a particular material, all available
information concerning toxicity to, and bioaccumulation by, aquatic organisms is
collected and reviewed for acceptability. If enough acceptable data on acute
toxicity to aquatic animals are available, they are used to estimate the highest
one-hour average concentration that should not result in unacceptable effects on
aquatic organisms and their uses. If justified, this concentration is made a
function of a water quality characteristic such as pH, salinity, or hardness.
Similarly, data on the chronic toxicity of the material to aquatic animals are used
to estimate the highest four-day average concentration that should not cause
unacceptable toxicity during a long-term exposure. If appropriate, this
concentration is also related to a water quality characteristic.
Once a thorough review of all pertinent information (e.g., data on
acute/chronic toxicity to aquatic animal, data on toxicity to aquatic plants and
data on bioaccumulation by aquatic organisms) indicates that enough acceptable
data are available, numerical national water quality criteria are derived for fresh
water or salt water or both to protect aquatic organisms and their uses from
unacceptable effects due to exposure to high concentration for short periods of
time, lower concentration for longer periods of time, and combinations of the
two.
As discussed earlier, the water quality standards regulations allow States to
develop numerical criteria or modify EPA's recommended criteria to account for
site-specific factors. In cases where additional toxicological data are needed to
modify or develop criteria, the discharger may be required to generate the data.
Guidance on modifying national criteria is found in the Handbook [6]. When a
criterion must be developed for a chemical for which a national criterion has not
been established, the regulatory authority should refer to the Guidelines for
Deriving Criteria for Aquatic Life and Human Health (see 45 FR 79341:
November 28, 1980 and 50 FR 30784, July 29, 1985).
2.3.2	Magnitude for Single Chemicals
Water quality criteria for aquatic life contain two expressions of allowable
magnitude: a criterion maximum concentration (CMC) to protect against acute
(shorf term) effects; and a criterion continuous concentration (CCC) to protect
56

-------
against chronic (long term) effects. The two concentrations for the priority
pollutants are presented in EPA's criteria documents.
Most State standards only include numerical criteria for a limited number
of individual toxic chemicals. Therefore, evaluation and control of toxic
pollutants is based on maintenance of the designated use and often relies heavily
on the narrative criterion prohibiting toxic substances in toxic amounts. The
adverse effects of concern will depend on the designated use and the chemical.
Human health hazards, bioaccumulation of chemicals in aquatic organisms,
toxicity to these organisms, aesthetic factors, the potential for additivity,
antagonism, synergism and persistence of the chemicals may be important.
Available information on the toxic effects of the chemical is used when standards
do not include specific numerical criteria. Such information can include EPA
criteria documents, published literature reports, or studies conducted by the
discharger.
As discussed in Section 2.1.2, water quality-based controls may be based
directly on the State's technical determination of what concentration of a specific
pollutant meets the State's narrative "free from" toxics criterion. Although EPA
water quality standards regulation require that the State's process for
implementing its narrative criterion be described in the State standards, there is
no requirement that this concentration be adopted as a numerical criterion in
State water quality standards prior to use in developing water quality-based
controls and therefore a case-by-case interpretation of the narrative criterion may
be necessary.
2.3.3 Magnitude for Whole-effluent Toxicity
Criteria for toxicity in current State standards range from the narrative
prohibition (e.g., no discharge of toxic chemicals in toxic amounts) to detailed
requirements that specify the test species and the allowable toxicity level. Acute
and chronic toxicity units (TUs) are a mechanism for quantifying instream toxicity
using the whole effluent approach. The procedure to implement the narrative
criteria using a whole effluent approach should specify the testing procedure, the
duration of the tests (acute v. chronic), the test organism(s), and the frequency of
testing required.
EPA's recommended magnitudes for whole effluent toxicity are as follows
(again, two expressions of allowable magnitude are used): a CMC to protect
against acute (short term) effects; and a CCC to protect against chronic (long
term) effects). For acute protection, the CMC should not exceed 0.3 acute toxic
unit (TUa) to the most sensitive of at least 3 test species. Where 03 TUa is
applied for end-of-pipe compliance, there should be no significant statistical
difference between the control and acute toxicity. As explained in the previous
section, the selection of test species for testing the effluent is not critical provided
species from ecologically diverse taxa are used (e.g., a fish, an invertebrate, and a
plant). The factor of 0.3 is used to adjust the typical LC50 endpoint of an acute
57

-------
Figure

-------
toxicity test (50% mortality) to an LCI value (virtually no mortality). Specifically,
a factor of 0.3 was found to include 91% of observed LCI to LC50 ratios in 496
effluent toxicity tests as illustrated in Figure 2-1. This figure presents effluent
toxicity data from many years of toxicity testing of both industrial and municipal
effluents by the Environmental Services Division, U.S. EPA Region IV, Athens,
Georgia.
For chronic protection, the CCC should not exceed 1.0 chronic toxic unit
(TUc) to the most sensitive of at least three test species. The selection of test
organisms is as described above. A 1.0 TUC is applied at the edge of the mixing
zone to protect the receiving water from any chronic toxicity that may occur
outside the mixing zone.
2.3.4 Duration for Single Chemicals and Whole-effluent Toxicity
The quality of an ambient water typically varies in response to variations
of effluent quality, stream flow, and other factors. Organisms in the receiving
water are not experiencing constant, steady exposure but rather are experiencing
fluctuating short periods of exposure to high concentrations of toxicants which can
have adverse effects. Thus, EPA's criteria restrict the maximum time period over
which exposure is to be averaged, thereby limiting the magnitude and duration of
exposure (See Appendix D for more detail).
For the CMC, the maximum period in which to average the exposure is
one hour. Thus, the one-hour average exposure must be less than the CMC to
protect against acute effect. In practice, one-day periods are the shortest periods
for which wasteload allocation (WLA) modelers and enforcement personnel have
adequate data. Compliance with the duration criterion can be ensured by paying
particular attention to short term effluent variability and requiring measures to
control variability (e.g., installation of equalization basins) when needed.
For the CCC, the maximum period in which to average the exposure is
four days. Some pollutants may have magnitude criteria that correspond to
different periods. The toxicity criterion for chronic effects is specified as a
four-day average. Thus, the four-day average exposure must be equal to or less
than the CCC.
These specifications for duration apply to both chemical-specific and
whole-effluent toxicity criteria. A common error is to assume that, because some
chronic toxicity tests are 28 or 30 days in length, the CCC was meant to be used
as a 30-day average. However, the duration of a toxicity test has nothing to do
with the critical period of exposure to concentrations greater than the criteria.
Some chronic toxicity tests are of a one-year or longer duration, yet this does not
lead to the establishment of an averaging period of one year's duration.
Obviously, if a one-year averaging period were used, the CCC could theoretically
be exceeded by a factor of two for six months, a duration more than long enough
to cause an unacceptable chronic effect in a waterbody.
59

-------
For the CCC, the maximum period in which to average the exposure is
four days. Some pollutants may have magnitude criteria that correspond to
different periods. The toxicity criterion for chronic effects is specified as a
four-day average. Thus, the four-day average exposure must be equal to or less
than the CCC.
These specifications for duration apply to both chemical-specific and
whole-effluent toxicity criteria. A common error is to assume that, because some
chronic toxicity tests are 28 or 30 days in length, the CCC was meant to be used
as a 30-day average. However, the duration of a toxicity test has nothing to do
with the critical period of exposure to concentrations greater than the criteria.
Some chronic toxicity tests are of a one-year or longer duration, yet this does not
lead to the establishment of an averaging period of one year's duration.
Obviously, if a one-year averaging period were used, the CCC could theoretically
be exceeded by a factor of two for six months, a duration more than long enough
to cause an unacceptable chronic effect in a waterbody.
The toxicity tests used to establish the national criteria are conducted using
steady state exposure techniques. They do not account lor fluctuating exposures
such as those experienced by the instream biota due to effluent variability. As
the period of averaging increases, so too does the period of time the exposure
concentrations can be above the criterion concentration without exceeding the
average. The significant consideration involved in setting duration criteria is how
long the exposure concentration can be above the criterion concentration without
unacceptably affecting the endpoint of the test (e.g., survival, growth, or
reproduction). The period of time above the criteria concentration must be
limited. The four-day duration recommended by EPA specifies the period of
time over which to average exposure. This duration time was derived to limit the
duration above the CCC.
2.3.5 Frequency for Single Chemicals and Whole-effluent Toxicity
The application of criteria involves predictive modeling of ambient
conditions based on the assimilative capacity of the receiving water and sources of
pollution. An essential element in this modeling is an assessment of the
projected frequency with which criteria will be exceeded. There are two reasons
why this frequency is a factor in water quality criteria. First, it is statistically
impossible to project that criteria will never be exceeded (unless pollution sources
are removed entirely). Second, ecological communities are able to recover from
stresses. Since WLA modeling requires a specification for frequency, it is
important that ecologically sound recommendations be provided.
The frequency with which criteria are allowed to be exceeded depends on
site-specific factors. To implement the criteria, site specific factors described in
Appendix D must be taken into account. EPA's recommendations for frequency
are once in three years for both the CMC and the CCC. These recommendations
apply to both chemical-specific and whole-effluent approaches.
60

-------
As explained in Appendix D, field studies indicate that many discharge
situations are affected both by predictable and measurable discharges of toxicants
and by unpredictable spills of toxic substances. In most cases, the dischargers
were unaware that spills were occurring. These spills are a second source of
stress for the community and decrease recovery potential. An aggressive program
to minimize, contain, and treat spills should be in place at any plant where the
potential for spills exists.
The concentration, duration and frequency provisions of the criteria are
implemented through the development of wasteload allocations and water quality
based effluent limitations. As discussed in Chapter 4, the duration and frequency
recommendations are implemented directly if a dynamic modeling approach is
used to develop wasteload allocations and permit limitations. However, if a
steady state approach is used, a design flow is needed for the calculations.
For the protection of aquatic life, the duration and frequency
recommendations provided above have been used to develop recommended
design flows for steady-state modeling. These recommended design flows are
discussed in Chapter 4. For the protection of human health, the recommendation
for design flows is also discussed in Chapter 4.
It should be noted that EPA's water quality criteria for aquatic life
protection are applicable at all flow conditions - low as well as high. These
criteria and their specified duration and frequency may be used as the basis for
TMDL's after considering seasonal flow and loading scenarios. The
concentration, duration and frequency provisions of EPA's water quality criteria
can be modified to account for site-specific conditions.
Traditionally, most of water quality-based permits for point source
discharges had been tied to the biologically-based flow or the 7-day 10-year low
flow conditions. The reason for this that critical conditions for perennial point
source discharges occur, in general, during the low flow period. Currently, State
laws and regulations generally state that water quality standards are applicable to
the 7-day 10-year low flow, or higher flow conditions.
As States have started using the new two-number water quality criterion
for perennial as well as intermittent discharges such as combined sewer overflows,
urban runoff, etc., their proper use in the context of the total maximum daily
load/wasteload allocation/load allocation (TMDL/WLA/LA) process needs to be
emphasized. Permits for point sources based on low flow condition may not meet
water quality standards during higher flow conditions where loadings increase of
these higher flows.
61

-------
2.4 WATER QUALITY CRITERIA FOR HUMAN HEALTH PROTECTION
2.4.1	Overview
There are several key elements of State water quality standards relevant to
human health protection. States must first determine ambient standards for the
two primary human exposure routes ~ fish consumption and drinking water.
Next, the level of exposure (dose) upon which the acceptable ambient receiving
water concentrations will be based, and the appropriate risk level for carcinogens
must be determined. Finally, the State must establish whether a mixing zone will
apply, and, if so, determine the critical design conditions.
States, standards or their implementation procedures often specify the risk
level for carcinogens; methods for identifying compliance thresholds in permits
where calculated limits are below detection; and methods for selecting
appropriate hardness, pH, and temperature variables for criteria. However, if
State standards do not specify these items, then the permitting authority must
develop water quality-based effluent limitations based upon either an
interpretation of the State's water quality standards or EPA's procedures.
The purpose of the following section is to provide a review of EPA's
procedures used to develop assessments of human health effects in developing
water quality criteria. A complete human health effects discussion is included in
the (draft) "Guidelines and Methodology Used in the Preparation of Health
Effects Assessment Chapters of the Consent Decree Water Documents." by EPA's
Environmental Criteria and Assessment Office (ECAO). The procedures
contained in the ECAO document are used in the development and updating of
EPA water quality criteria and may be used in developing acceptable ambient
concentrations (AACs) for those pollutants lacking EPA human health criteria.
Although the same procedures are used to develop criteria and AACs, only those
values which are subjected to the regulatory process of regional, state, and public
comment can be considered "criteria." AACs may be applied as site-specific
numeric standards and as a basis for permit limitations.
Procedures are also provided in this chapter to develop values called
acceptable tissue concentrations (ATCs) which can be used in monitoring fish
tissues where the AAC is below the analytical detection limit.
2.4.2	Magnitude and Duration
Magnitude of human health criteria is the primary consideration.
Currently, national policy and prevailing opinion in the expert community dictate
that human health criteria for carcinogens are derived assuming lifetime
exposure: a 70 year time period. The duration of exposure assumed in deriving
criteria for non-carcinogens is more complicated due to a wide variety of
endpoints; some developmental (and thus age-specific and perhaps sex-specific),
some lifetime as with carcinogens, and some, such as organoleptic effects, not
62

-------
age-related at all. Thus, durations depend on the individual non-carcinogenic
pollutants being considered.
2.4.3	Human Exposure Considerations
A complete human exposure evaluation for toxic pollutants of
bioconcentration concern would not only encompass estimates of exposures due
to fish consumption, but also exposure due to background concentrations and
other exposure routes, including recreational contact, occupational, dietary intake
from other than fish, inhalation of air, etc. As stated in Chapter I's Overview,
Section 1.1, the focus of this document is on ingestion of contaminated fish, a
direct human exposure route of serious potential risk. Other lesser exposure
route information should be considered to the extent it is available (Sections 3.6
and 3.7).
Levels of actual human exposures from consuming contaminated fish vary
depending upon a number of case-specific consumption factors. These factors
include type of fish species consumed, type of fish tissue consumed, tissue lipid
content, consumption rate and pattern, and food preparation practices. In
addition, depending on the spatial variability in the fishery area, the behavior of
the fish species, and the point of application of the AAC or criterion, the average
exposure of fish may be only a small fraction of the expected exposure at the
point of application of the criterion. If an effluent attracts fish, it might be
greater than the expected exposure.
With shellfish, such as oysters, snails, and mussels, whole body tissue
consumption commonly occurs, whereas with fish, muscle tissue and roe are most
commonly eaten. This difference in the types of tissues consumed has
implications for the amount of available bioconcentratable contaminants likely to
be ingested. Whole body shellfish consumption presumably means ingestion of
the entire burden of bioconcentratable contaminants. However, with most fish,
selective cleaning and removal of internal organs, and sometimes body fat as well,
from edible tissues may result in removal of the lipid rich tissues that contain the
majority of bioconcentratable contaminants.
2.4.4	Fish Consumption Values
EPA's 1980 criteria assumed a human body weight of 70 kg and the
consumption of 0.0065 kg of fish and shellfish per day. The national fish
consumption value for freshwater and estuarine fish calculated in 1980 was 6.5
g/day with the 95th percentile for all seafood at 42 g/day (SRI Document). The
mean lipid content of fish tissue found to be consumed in the 1980 fish
consumption study was 3.0% (Stephens memo).
EPA recommends that the consumption values used in deriving AACs
from the formulas in this chapter reflect the most current relevant and/or site-
specific information available. For example, some States have developed their
63

-------
own fish consumption estimates, ranging from 20 g/day in Wisconsin, Louisiana,
Illinois and Arizona to 37 g/day in Delaware for salt water (Ref).
Currently, a range of four levels of fish consumption are provided in the
EPA guidance manual "Assessing Human Health Risk from Chemically
Contaminated Fish and Shellfish." These are:
o 6.5 g/day to represent an estimate of average consumption of fish and
shellfish from estuarine and fresh waters by the U.S. population (U.S.
EPA 1980b)
o 20 g/day to represent an estimate of the average consumption of fish and
shellfish from marine, estuarine, and fresh waters by the U.S. population
(USDA 1984)
o 165 g/day to represent average consumption of fish and shellfish from
marine, estuarine, and fresh waters by the 99.9th percentile of the U.S.
population (Finch 1973)
o 180 g/day to represent a "reasonable worst case" based on the assumption
that some individuals would consume fish at a rate equal to the combined
consumption of red meat, poultry, fish, and shellfish in the U.S. (EPA Risk
Assessment Council assumption based on data from the USDA Nationwide
Food Consumption Survey of 1977-1978; see Appendix F).
EPA is currently updating the national estuarine and freshwater fish and
shellfish consumption default values and will provide a range of recommended
national consumption values. This range will include: 1) mean values
appropriate to the population at large, and 2) values appropriate for those
fishermen who consume a relatively large proportion of fish in their diets
(maximally exposed individuals).
2.4.5	Bioaccumulation Consideration
Chemical residues in fish may increase above expected concentrations
based upon the bioconcentration factor (BCF) as the result of consuming residue-
containing food. The trophic level and age of the fish being consumed both
affect the accumulation of contaminants, especially with higher trophic level
predator fish. In addition, growth rate can also have an effect on contaminant
levels in organisms. For chemicals with log P (partition coefficient between
octanol and water) between 4.5 and 7.0, bioaccumulation above bioconcentration
increases from 3 times to 100 times for higher trophic level fishes.
2.4.6	Updating Human Health Criteria and Generating AACs Using IRIS
EPA recommends that the process of updating criteria and generating
AACs use the most current risk information. EPA's Integrated Risk Information
64

-------
System (IRIS) is a computer-accessed, electronically-communicated catalogue of
the latest EPA risk assessment and risk management information for chemical
substances. The risk assessment information contained in IRIS, except as
specifically noted, has been reviewed by intra-Agency review groups representing
all of EPA's program offices and represents an Agency-wide consensus. Risk
values contained in IRIS are approved for Agency-wide use [22].
IRIS is designed especially for Federal, State, and local environmental
agencies as a source of the latest information about Agency health assessments
and regulatory decisions for specific chemicals. The system is available to EPA
and some State users through the EPA electronic Mail system and to the public
through on-line networks such as DIALCOM, Inc., Public Health Network (PHN)
and TOXNET. The document entitle "Volume I: Supportive Documentation"
(users guide), March 1987 (EPA/600/6-86/032a) provides information on the use
of this system (Contact IRIS Usfer Support: 202/ITS 684-7254).
IRIS contains two types of dose response factors: the human reference
dose (RfD [formerly known as the Acceptable daily Intake, ADI]) and the
carcinogenic potency slope factor (ql*). The RfD applies to non-carcinogenic
(also known as target organ, or systemic) effects. The ql*, which indicates the
cancer-causing potential of a substance is used in calculating the AAC for
carcinogens. Appendix C contains the supporting information for derivation of
RfDs and ql*s.
EPA periodically updates risk assessment information including RfDs,
ql*s, and related information on contaminant effects, and reports the current
information in IRIS. A list of the IRIS RfD and ql* values current at time of
publication is included in Appendix B. Although Appendix B was current at the
time of publication, its inclusion in this document is intended for reference use
in initial screening only.
Since IRIS values may be updated at any time, current IRIS dose values
supersede those used in developing 1980 human health criteria. This means that
the 1980 human health criteria should be updated with the latest IRIS values.
The procedure for deriving an updated human health water quality criterion is to
insert the new IRIS data into the appropriate equation in Section 3.6 or 3.7.
Also, in the absence of a promulgated State numeric criterion, an AAC may be
calculated using these same formulas. States may also elect to use these
equations to update existing promulgated State numeric criteria. Appendix B
values should not be used to revise any criteria or derive AACs without checking
IRIS for the most current values.
Figure 2.2 shows the procedure for determining an updated criterion or
AAC using IRIS data. If a chemical has both carcinogenic and non-carcinogenic
effects; i.e., both a ql* and RfD, the carcinogen AAC formula in Section 2.7
should be used as it will result in the more stringent AAC of the two.
65

-------
Figure 2.2
Procedure for Updating an EPA Human Health Criterion
or Developing an Acceptable Ambient Concentration
A)
TO UPDATE AN EPA HUMAN
HEALTH CRITERION OR
TO DERIVE AN AAC
EPA'S WATER
QUALITY
CRITERION
AVAILABLE
B)
Al)
UPDATED
IRIS DATA
EXISTS
CALCULATE
USING IRIS
DATA
A2)
IRIS DATA
NOT UPDATED
USE
CURRENT
CRITERION
EPA'S WATER
QUALITY
CRITERION
UNAVAILABLE
Bl)
DATA IN
IRIS
CALCULATE
AACs
USING IRIS
DATA
B2)

NO DATA
IN IRIS


DEVELOP
A DOSE
FACTOR
OR AN
ATC
USING
AVAILABLE
DATA
(I.E.,
MCL,FISH
ADVISORY
FDA
ACTION
LEVEL,
ETC. )
66

-------
Figure 2.2 should be interpreted in the following manner:
A)	If an EPA human health criterion exists:
Al) But the dose response factor data on IRIS has been updated,
then a new criterion should be calculated using IRIS data
and Section 3.6 or 3.7.
A2) But there is no change to IRIS data, use the current
criterion;
B)	If no EPA human health criterion exists:
Bl) But there is dose response factor data on IRIS, use IRIS
data to derive an AAC (Section 3.6 or 3.7).
B2) But no data on IRIS exist, a dose factor or an acceptable
tissue concentration (ATC) may be developed by
extrapolation from available human, mammalian or other
dose data, including, for example, drinking water maximum
contaminant levels (MCLs), fish advisories or FDA Action
Levels (Section 2.4.9 or 2.4.10).
2.4.7 Calculating AACs for Non-Carcinogens
The RfD is an estimate (with uncertainty spanning perhaps an order of
magnitude) of the daily exposure to the human population (including sensitive
subgroups) that is likely to be without appreciable risk of deleterious effect during
a lifetime. The RfD is expressed in units of mg toxicant/kg human
body weight/day.
RfDs are derived from the "no observed adverse effect level"
(NOAEL) or the "lowest observed adverse effect level" (LOAEL) identified from
chronic or subchronic human epidemiology studies or animal exposure (mammal
LD50) studies. [Note: LOAEL and NOAEL refer to animal and human
toxicology and are therefore distinct from the aquatic toxicity terms "no observed
effect concentration (NOEC) and the "lowest observed effect concentration"
(LOEC)]. When no NOAEL is available, then the LOAEL is used with
additional uncertainty factors. Factors are applied to accommodate uncertainty
associated with variability among individuals, extrapolation from nonhuman test
species to humans, and data on other than long-term chronic exposures and
endpoints [8].
The RfD is a threshold below which effects are unlikely to occur. While
exposures above the RfD increase the probability of adverse effects, they do not
produce a certainty of adverse effects. Similarly, while exposure at or below the
RfD reduces the probability, it does not guarantee the absence of effects in all
67

-------
persons. Hie RfDs contained in IRIS are EPA's consensus best scientific
judgements.
For non-carcinogenic effects, an updated criterion or an AAC can be
derived using the following equation:
Equation 2.1: C or AAC = (RFD x WD - (DT + IN)
WI+ [FC x P x (FM x BCF)]
where: C = updated water quality criterion (mg/1)
AAC = acceptable ambient concentration (mg/1)
RfD = reference dose (mg toxicant/kg human body weight/day)
WT = weight of an average human adult [70 kg]
DT = dietary exposure [other than fish] (mg toxicant/kg human body
weight/day)
IN = inhalation exposure (mg toxicant/kg human body weight/day)
WI = average human adult water intake [2 liters/day]
FC = daily fish consumption (kg fish/day)
P = ratio of lipid fraction of fish tissue consumed to 3%
FM = food chain multiplier
BCF = bioconcentration factor (mg toxicant/kg fish divided by mg
toxicant/L water) for fish with 3% lipid
If the receiving water body is not used as a drinking water source, the
factor WI can be deleted. Where dietary and/or inhalation are unavailable, these
factors may be deleted from the above calculation. For identified non-
carcinogenic chemicals without known RfDs, extrapolation procedures can be
used to estimate the RfD (see Appendix H).
2.4.8 Calculating AACs for Carcinogens
Maximum protection of human health from the potential effects of
exposure to carcinogens via contaminated fish would require an ATC of zero.
The zero level is based upon the assumption of non-threshold effects (i.e, no safe
level exists) for carcinogens. However, because the zero level might never be
attainable, a cancer risk concentration is chosen instead. Cancer risk is calculated
by multiplying the experimentally derived potency, ql*, by the lifetime average
dose rate for a human consuming contaminated fish.
The ql*, called the cancer potency slope factor, is derived using animal
studies or epidemiological data on human exposure. High dose exposures are
extrapolated to low dose concentrations and a lifetime exposure period through
the use of a linear model. The model calculates the upper 95 percent confidence
limit of the slope of a straight line that can be extended to depict the dose-
response relationship at low doses [8]. In IRIS, the ql*'s are called Oral Slope
Factors.
68

-------
It is important to note that cancer potency slope factors may overestimate
actual risk. Such potency estimates are subject to uncertainty due to a number of
primary factors: 1) adequacy of the cancer data base and 2) limited information
regarding the mechanism of cancer causation, and 3) extrapolation from animals
to humans and high does to low doses. The actual risk may be much lower,
perhaps as low as zero, particularly for those chemicals for which human
carcinogenicity information is lacking. EPA cancer potency slope factors are
approved by an Agency workgroup made up of senior Agency scientists from all
the Program Offices and the Office of Research and Development. They are
used by Agency Offices whenever they develop regulations or guidance with a
cancer-based risk assessment component. EPA believes that it's procedures
produce a plausible upper limit for risk and that the procedures are appropriately
conservative given the uncertainty involved. EPA procedures also consider
concurrent human exposure assumptions (i.e., fish consumption levels, water
intake, etc.) that may not be representative of local conditions, resulting in AACs
that can overestimate or underestimate actual risk. EPA recommends that,
whenever possible, exposure values reflect the most current relevant and/or site
specific information available.
Risk factors are the proportion (e.g., one in 1,000,000, or 10"6, or 0.000001)
of humans exposed to a chemical who are expected to develop cancer as a result
of exposure. EPA's ambient water quality criteria risk levels range from 10'5 to
10"7. Risk levels of 10"5, 10"6, and 10"7 are often used by States in interpreting
their numeric standards. EPA considers risk to be additive, i.e., the risk from
individual chemicals is not necessarily the overall risk from exposure to water.
For example, an individual risk level of 10'* may yield an overall risk level of 10"
4 if 100 chemicals are of concern. Where there are no EPA - approved State
standards to guide the Agency's decisions, EPA will employ a general human
health risk range of 10~4 to 10'6, using 10'6 as a point of departure while allowing
case-specific factors to enter into the evaluation.
For carcinogenic effects, the AAC can be determined by using the
following equation [11]:
Equation 2.2: C or AAC = (RL x WD - DT	
ql* [WI + FC x P(FM x BCF)]
where: C = updated water quality criterion (mg/1)
AAC = acceptable ambient concentration (mg/1)
RL = risk level (10**)
WT = weight of an average human adult [70 kg]
DT = dietary exposure [other than fish] (mg toxicant/kg human body
weight/day)
ql* = carcinogenic potency factor (mg/kg/day)
WI = average human adult water intake [2 liters/day]
FC = daily fish consumption (kg fish/day)
P = ratio of lipid fraction of fish tissue consumed to 3%
FM = food chain multiplier [from Table 2.2]
BCF = bioconcentration factor (mg toxicant/kg fish divided by mg
toxicant/L water) for fish with 3% lipid
69

-------
If the receiving water body is not used as a drinking water source, the
factor WI can be deleted. Where dietary and/or inhalation are unavailable, these
factors may be deleted from the above calculation. For identified carcinogenic
chemicals without known ql* values, extrapolation procedures can be used to
estimate the ql* [see Appendix C].
2.4.9	Deriving Dose Factors in the Absence of IRIS Values
The RfDs or ql*s comprise the existing dose factors for developing AACs.
However, effluents may contain identified chemicals above 100 ng/L (the process
by which this number was derived is discussed in Section 3.3.1) for which no Rfd
or ql* is available. When IRIS data are unavailable, dose information may be
developed according to a State's own procedures. Some States have established
their own procedures whereby dose factors can be developed based upon
extrapolation of acute and/or chronic animal data to concentrations of exposure
protective of fish consumption by humans. Where no procedure exists, factors
may be based upon extrapolation from mammalian or other data using the IRIS
documentation in Appendix C. Finally, where no other relevant information or
procedure exists, a dose factor could be based upon such information as drinking
water maximum contaminant levels (MCLs) or FDA Action Levels.
Michigan's procedures for deriving a dose factor depend upon the type and
quality of the toxicity data base used. Wisconsin uses feeding and drinking rates
for mammalian and avian species to estimate ATCs for fish consumption by
wildlife. Where species specific data are lacking, Wisconsin applies an
uncertainty factor of 0.1 to account for differences in species sensitivity [19].
Readers are referred to the following references for additional information:
1)	Michigan Department of Natural Resources, "Guidelines for Rule
57(a) (5 and 6), Michigan Department of Natural Resources,
Environmental Protection Bureau, 1987.
2)	Wisconsin Department of Natural Resources, "Proposed Chapter
NR10S.08", Wisconsin Department of Natural Resources, Bureau of
Water Resources Management, 1987.
2.4.10	Deriving Acceptable Tissue Concentrations for Monitoring Fish Tissue
Where fish tissue monitoring is required because an AAC is below
analytical detection limits, the following formulas may be used to calculate an
acceptable tissue concentration (ATC). Readers should also consult EPA's
"Assessing Human Health Risks from Chemically Contaminated Fish and
Shellfish" [3].
^ The basic equations for deriving acceptable tissue concentrations use the
same variables as in Equations 2.1 and 2.2, where BCF is normalized at 3.0%
lipid:
70

-------
For non-carcinogens:
Equation 2.3: ATC (mg/kg) = RFD x WT - (DT + IN)
WI/BCF + FC
For carcinogens:
Equation 2.4: ATC (mg/kg) = RL x WT
ql* (WI/BCF) + FC
The above equations should be corrected for site specific lipid content and
bioaccumulation factors where data are available.
Again, some States have established their own procedures whereby ATCs
can be developed based upon extrapolation of acute and/or chronic animal data
to safe concentrations protective of fish consumption by humans. Where no other
relevant information or procedure exists, an ATC could be based upon such
information a fish advisory, as drinking water maximum contaminant levels
(MCLs), or FDA Action Levels.
2.5. BIOCRITERIA/BIOLOGICAL CRITERIA
As discussed in Chapter 1, to fully protect aquatic habitats and provide
more comprehensive assessments of aquatic life use attainment/nonattainment,
EPA encourages States to fully integrate chemical-specific techniques, toxicity
testing, biological surveys and biocriteria into their water quality programs. In
particular, the Agency's policy is that States should develop and implement
biological criteria in their water quality standards.
2.5.1 Development of Biocriteria
Biocriteria are numerical measures or narrative descriptions of the
biological integrity of unimpaired natural systems. The biological communities in
these waters become a reference and represent the best attainable conditions.
The reference site then becomes the basis for developing biocriteria for major
surface water types (streams, rivers, lakes, estuaries, wetlands, costal or marine
water).
The assessment of the biological integrity should include measures of the
structure and function of a community of species within a specified habitat.
Expert knowledge of the system is required for the selection of appropriate
biological components and measurement indices. The specific definition of the
biological integrity selected by a State will form the basis for comparing impacted
sites to an established reference condition. Therefore, it is essential that the
State develop a rigorous statement of biological integrity for each designated use
of surface waters within the State.
71

-------
The following four steps are required to develop criteria:
1)	Identify and locate the unimpaired reference condition for specific
ecoregions and habitat types.
2)	Characterize the reference condition using biological measures.
3)	Select appropriate spatial and temporal scales to assess system changes.
4)	Design standardized sampling and statistical protocols to detect real
differences between the reference condition and comparable impaired
waters.
2.5.1 Regulatory Bases for Biocriteria
The primary statutory basis for EPA's policy that States should develop
biocriteria is found in Section 303(c)(2)(B) of the Water Quality Act of 1987.
Section 303 requires EPA to develop criteria based on biological assessment
methods when numerical criteria are not established. In addition to Section 303,
Section 131.11 of the Water Quality Standards Regulation and Section 304 of the
Act provides the legal basis for biocriteria.
Once biocriteria are formally adopted into State standards, biocriteria and
aquatic life use designations serve as direct, legal endpoints for determining a
quality life use attainment/nonattainment. As stated in Section 131.11(b)(2) of
the Water Quality Standards Regulation (40 CFR Part 131), biocriteria should be
used as a supplement to existing chemical-specific criteria and provide an
alternative to chemical-specific criteria where such criteria cannot be established.
States are encouraged to implement and integrate all three approaches
(biosurvey, chemical-specific and toxicity testing methods) into their water quality
programs, applying them in combination or independently (providing the most
protective of the three methods is used) as site-specific conditions and assessment
objectives dictate.
Section 304(a) requires EPA to review and publish criteria to maintain the
biological integrity of the nation's waters, to publish the effects of pollutants on
biological diversity and productivity and other scientific information regarding a
number of water quality related matters, including:
o Effects of pollutants on aquatic community components ("Plankton, fish,
shellfish, wildlife, plant life...") and community attributes ("diversity,
productivity, and stability...");
o Factors necessary "to restore and maintain the chemical, physical,
biological integrity of all navigable waters...", and "for protections and
72

-------
propagation of shellfish, fish, and wildlife for classes and categories of
receiving waters...";
o Appropriate "methods for establishing and measuring water quality criteria
for toxic pollutants on other bases than pollutant-by-pollutant criteria,
including biological monitoring and assessment methods."
Section 304(a) of CWA has been historically cited as the basis for
publishing nation guidance on chemical-specific criteria for aquatic life, but is
equally applicable to the development and use of biological monitoring and
assessment methods and biocriteria.
2.6 SEDIMENT CRITERIA
2.6.1 Current Developments in Sediment Criteria
While ambient water quality criteria are playing an important role in
assuring a healthy aquatic environment, they alone have not been sufficient to
ensure appropriate levels of environmental protection. Sediment contamination,
which can involve deposition of toxicants over long periods of time is responsible
for water quality impacts in some areas.
EPA has authority to pursue the development of sediment criteria in
streams, lakes and other waters of the United States under section 104 and
304(a)(1) and (2) of the CWA as follows:
o Section 104(n)(l) authorizes the Administrator to establish national
programs that study the effects of pollution, including
sedimentation, in estuaries on aquatic life.
o Section 304(a)(1) directs the Administrator to develop and publish
criteria for water quality, including information on the factors
affecting rates of organic and inorganic sedimentation for varying
types of receiving waters.
o Section 304(a)(2) directs the Administrator to develop and publish
information on, among other things, "the factors necessary for the
protection and propagation of shellfish, fish, and wildlife for classes
and categories of receiving waters..."
To the extent that sediment criteria could be developed which addressed the
concerns of the section 404(b)(1) Guidelines (for discharges of dredged or fill
material under the CWA and the Marine Protection Research, and Sanctuaries
Act), they may also be incorporated into those regulations.
73

-------
2.6.2	Approach to Sediment Criteria Development
Over the past several years, sediment criteria development activities have
centered on evaluating and developing the Equilibrium Partitioning Approach for
generating sediment criteria. The Equilibrium Partitioning Approach focuses on
predicting the chemical interaction between sediments and contaminants.
Developing an understanding of the principal factors that influence the
sediment/contaminant interactions will allow for predictions to be made as to
what concentration of a contaminant benthic and other organisms may be
exposed to. Chronic water quality criteria, or possibly other toxicological
endpoints can then be used to predict potential biological effects. In addition to
the development of sediment criteria, EPA is also working to develop a
standardized sediment toxicity test that could be used with or independently of
sediment criteria and could be osed to assess chronic effects in fresh and marine
waters.
2.6.3	Application of Sediment Criteria
Any methods capable of generating sediment criteria that are reasonably
accurate in their ability to predict biological impacts or the lack thereof are likely
to be useful in a variety of activities currently being pursued by EPA and States.
If the Equilibrium Partitioning Approach is used, it is expected that generated
criteria are likely to play a significant role in the identification, monitoring and
cleanup of contaminated sediment sites on a national basis and in ensuring that
those sites that are uncontaminated will remain so. In some cases, sediment
criteria alone would be sufficient to identify and to establish clean up levels for
contaminated sediments. In other cases, the sediment criteria would be
supplemented with biological or other types of analysis before decisions can be
made.
Regulatory frameworks for the application of these criteria are currently
being considered by EPA. Public input is expected prior to the adoption of any
formal regulatory framework for deriving and implementing sediment criteria.
2.6.5 Potential Uses
The EP method is likely to be used in many of the activities being pursued
by EPA. EP based sediment quality criteria could play a significant role in the
identification, monitoring and clean up of contaminated sediment sites on a
national basis and in ensuring that those sites that are uncontaminated will
remain so. In some cases sediment criteria alone would be sufficient to identify
and to establish clean up levels for contaminated sediments. In other cases the
sediment criteria would be supplemented with biological sampling and testing or
other types of analysis before a decision could be made. Sediment criteria can
provide a basis for determining whether contaminates are accumulating in
sediments to the extent that an unacceptable contaminant level is being
74

-------
approached or has been exceeded. By monitoring contaminants in the vicinity of
a discharge, contaminant levels can be compared to sediment criteria to assess
the likelihood of impact.
EP based sediment criteria will be particularly valuable in site monitoring
applications where sediment contaminant concentrations are gradually
approaching a criteria over time. Comparison of field measurement to sediment
criteria will be a reliable method for providing early warning of a potential
problem. Such an early warning would provided an opportunity to take corrective
action before adverse impacts occur.
In many ways sediment criteria developed using EP are similar to existing
water quality criteria. However, in their application it is likely that they may vary
significantly. Contaminants at levels of concern in the water column in most
cases need only be controlled at* the source to eliminate unacceptable adverse
impacts. Contaminated sediments often have been in place for quite some time
and controlling the source of that pollution (if the source still exists) will not be
sufficient to alleviate the problem. The problem is compounded due to the fact
that the safe removal and treatment or disposal of contaminated sediments can
be difficult and expensive. For this reason it is not anticipated that sediment
criteria will be used as mandatory clean up levels, but as a means for predicting
or identifying the degree and spatial extent of contaminated areas such that
regulatory decisions can be made.
2.6.5 Sediment Criteria Status
Interim sediment criteria have been developed for 12 non-ionic
contaminants (acenaphthene, aniline, chlorpyrifos, DDT, dieldrin, endrin, ethyl
parathion, heptachlor, lindane, phenanthrene and PCB (1254)). The Science
Advisory Board is currently reviewing the methodology used to develop those 12
criteria. It is expected that the Science Advisory Board will review the methods
for generating sediment criteria for metal contaminants and procedures for
establishing standardized bioassays in 1991.
75

-------
REFERENCES
1.	Michigan Department of Natural Resources, "Guidelines for Rule 57(a) (5
and 6), Michigan Depart, of Natural Resources, Environmental Protection
Bureau, 1987.
2.	Wisconsin Dept. of Natural Resources, "Proposed Chapter NR10S.08",
Wisconsin Dept. of Natural Res., Bureau of Water Resources
Management, 1987.
3.	U.S. EPA. August 1988. "Guidance Manual for Assessing Human Health
Issues from Chemically Contaminated Fish and Shellfish." by Robert A.
Pastorok, PTI Environmental Services, Bellevue, WA 98006; for Batelle
New England Marine Research Laboratory, Duxbury, MA. Final Report,
PTI Contract No. C737-01.
4.	U.S. EPA. March 1987a. "Integrated Risk Information System. Volume I.
Supportive Documentation." Office of Health and Environmental
Assessment. (EPA/600/8-86/032a).
5.	U.S. EPA. March 1987b. "Integrated Risk Information System. Volume
II. Chemical Files." Office of Health and Environmental Assessment.
(EPA/600/8-86/032b).
6.	U.S. Environmental Protection Agency. 1984. Water Quality Standards
Handbook. Office of Water Regulations and Standards (WH-585),
Washington, D.C.
7.	U.S. Environmental Protection Agency. 1984. Technical Support
Document for Conducting Use Attainability Studies. Office of Water
Regulations and Standards (WH-585), Washington, D.C.
8.	U.S. Environmental Protection Agency. 1973. "Water Quality Criteria
1972". (EPA-R3-73-033 or NTIS-PB236199)
9.	U.S. Environmental Protections Agency. 1985. "Guidelines for Deriving
Numerical National Water Quality Criteria for the Protection of Aquatic
Organisms and Their Uses" (NTIS-PB85-227049)
76

-------
3. EFFLUENT CHARACTERIZATION
3.1 INTRODUCTION
Once a level of water quality adopted for the water body has been
determined, as was described in the previous chapter, the next step is to
characterize the effluent. The purpose of effluent characterization is to
determine whether or not the discharge of toxic materials causes, or has the
reasonable potential to cause adverse impact to the biota of the receiving water
system or to human health. The effluent characterization procedures described in
the following sections apply only to the water quality-based approach, not to end-
of-the pipe technology-based controls.
This chapter is divided into two parts: 3.2 Assessment of Toxic Effects on
Aquatic Life, including whole effluent toxicity characterization and chemical
specific characterization; based on the cumulative experience gained by EPA,
States, POTWs and industry when implementing the water quality-based
approach; and 3.3 Assessment of Human Health Effects, concentrating on whole
effluent bioconcentration characterization. The bioconcentration analysis
procedures are based on procedures that are currently advisory and subject to
validation. A whole effluent toxicity screening approach for human health
impacts (analogous to whole effluent toxicity testing) is in the development stage
and is not yet fully reviewed and verified. As such, these experimental techniques
are discussed in Appendix G. Each section contains a discussion of specific
measurement techniques and assessment procedures recommended by EPA.
3.1.1 Background for Toxic Effects Assessments on Aquatic Life and Human
Health
Two assessment techniques can be employed in effluent characterization
for aquatic life effects; whole effluent toxicity assessments and chemical specific
assessments. The type of effluent assessment procedure selected for water
quality-based toxics control should be tailored to each discharge situation. In
many cases, it would be beneficial for permittees to perform both types of
assessments. For example, where it is suspected or faiown that toxicity will
continue to be present after imposition of limits on specific chemicals, a whole-
effluent toxicity limit should be set to control the toxicity of the entire effluent.
An example of such a situation is where a known toxicant is discharged at
relatively high concentrations in a complex effluent known to cause a toxic effect;
for example, a publicly-owned treatment works (POTW) discharging ammonia at
toxic concentrations. A limit for ammonia would need to be set to reduce the
concentration of ammonia, along with a whole effluent toxicity limit to control
toxicity caused by any other toxicants present.
77

-------
Aquatic toxicity effects can be characterized by conducting a general
assessment of the effluent, and/or by measuring effluent toxicity or concentrations
of individual chemicals and comparing these measurements to the expected
exposure concentrations in the receiving water. The "receiving water
concentration" (RWC) is the exposure concentration of a toxicant or the
parameter toxicity (when dealing with the whole effluent toxicity) in the receiving
water after mixing.
As with aquatic life protection, there are two possible approaches to
characterizing effluents for human health effects: chemical-by-chemical and
whole effluent. However, only one approach is currently practical for assessing
and controlling human health impacts: the chemical-by-chemical approach.
Chapter 2 provides detailed information on how Acceptable Ambient
Concentrations (AACs) incorporate adequate levels of human health protection.
Appendix G discusses developing procedures on the assessment of human health
impacts from whole effluents.
The following important relationships apply to both aquatic life and human
health protection. With individual toxicants (or the parameter toxicity), the
potential for aquatic impact in the receiving water is minimized where the RWC
is less than the CCC or the CMC, and becomes maximized where the RWC
exceeds the CCC or CMC. Therefore, to prevent ambient aquatic life impacts,
the RWC of the parameter effluent toxicity or an individual toxicant (based on
allowable dilution for the criterion) must be less than the most limiting of the
applicable criterion. Protection of human health will be achieved where the
RWC is less than the EPA criterion or the AAC (AAC as used throughout this
Chapter incorporates EPA human health criteria, as well).
RWC < CCC (chronic aquatic life)
RWC < CMC (acute aquatic life)
RWC < AAC (human health)
It is important to recognize that the water quality analyst will utilize the
same basic components in the above-described relationship (i.e., critical receiving
water flows, ambient criteria values, measures of effluent quality) for both
effluent characterization and wasteload allocation development; albeit from
different perspectives. In the case of effluent characterization, the objective is to
project receiving water concentrations based upon existing effluent quality to
determine whether or not an excursion above ambient criteria is occurring, or has
the reasonable potential to occur. In developing wasteload allocations on the
other hand, the objective is to fix the RWC at the desired criteria level and
determine an allowable effluent loading which will not cause excursions above the
criteria.
78

-------
Recommendations for projecting the RWC are described below. Chapter
4,	Exposure Assessment and Wasteload Allocation, provides recommendations for
determining allowable effluent loadings to achieve established ambient criteria
and for calculating wasteload allocations for establishing permit limits. Chapter
5,	Permit Requirements, describes the actual calculation of permit limits after
effluent characterization and loadings and wasteload allocations are complete.
Box 3-1 summarizes the circumstances for which the whole effluent toxicity
and chemical specific aquatic life assessment approaches are best used.
Box 3-1
When to Use Aquatic Life Analysis and Chemical-specificAnalysis in an
Integrated Approach
Whole effluent analysis is most appropriate in the following discharge situations:
o Where an effluent's constituents are not completely known or where a complex mixture of
potentially additive, synergistic or antagonistic toxic pollutants are discharged.
o Where more than one discharger is located in a specific area and the potential exists for
effluent mixing and additive toxic effect.
o Where chemical-specific evaluation is impractical due to a lack of information about the
toxic effects of a chemical or a lack of the capability to model the chemical(s) present.
Chemical-specific analysis is most appropriate in the following discharge situations:
o Where an effluent contains only one or several well quantified pollutants for which
toxicological data are available or can be generated, and whole effluent toxicity data indicate
no other toxicity present, or where one or more known pollutants comprise a large
percentage of the waste flow.
o Where bioaccumulative chemicals with available criteria values are suspected in the effluent
and specific limits for each of these chemicals may be required.
o Where potentially toxic constituents of an effluent may be masked initially, but cause
delayed toxicity when acted on by downstream physical, chemical or biological processes.
3.2 ASSESSMENT OF AQUATIC LIFE EFFECTS
3.2.1 General Considerations in Effluent Characterization
There are two possible ways to proceed when characterizing an effluent for
the need for effluent limits for the protection of aquatic life. First, an assessment
79

-------
may be made without generating effluent data; or second, an assessment may be
conducted after effluent data have been generated. Regardless, it is now explicit
EPA regulation that water quality-based limits for specific toxicants and whole
effluent toxicity are required where necessary to achieve state water quality
standards (40 CFR 122.44(d)). Regulatory authorities must use an appropriate
procedure for determining whether a discharge is causing, or has the "reasonable
potential" to cause to an excursion above an applicable water quality standard
(either narrative or numeric). Reasonable potential is an analysis of an effluent's
capability to cause such excursions. Where the regulatory authority determines
that an excursion above a water quality criterion is either occurring or is
projected, 40 CFR 122.44(d) requires the development of a permit limit for the
pollutant or pollutant parameter of concern.
In making a determination of the need for a permit limitation for whole
effluent toxicity or an individual toxicant, the regulatory authority is required to
consider, at a minimum, existing controls on point and nonpoint sources of
pollution, the variability of the pollutant or pollutant parameter in the effluent,
the sensitivity of the species to toxicity testing (for whole effluent), and where
appropriate, the dilution of the effluent in the receiving water (40 CFR
122.44(d)(ii)). (Where applicable, existing water quality criteria and technology- .
based effluent limitations guidelines must also be considered).
In addition, the regulatory authority should consider all other available
data and information pertaining to the discharger to assist in making an informed
judgement. Where both effluent testing data and important other factors exist,
the regulatory authority will need to exercise discretion in the determination of
the need for a limitation. The authority should employ the principle of "joint
applicability" of the data and information which characterizes the effluent. In
other words, effluent data alone, showing absence of toxicity, may not be
adequate to overrule the need for a toxicity limit if other factors are prominent.
The regulatory authority will need to prioritize the importance of all data and
information used in making a determination on a case-by-case basis. To assist in
case-by-case determinations, recommended guidelines for characterizing an
effluent for the need for a permit limitation for whole effluent toxicity or
individual toxicants are discussed below and summarized in Boxes 3-2 and 3-4.
3.22 Determining the Need for a Limitation Without Effluent Testing Data
If the regulatory authority so chooses, or if the circumstances dictate, the
authority may decide to develop and impose a permit limitation for whole
effluent toxicity or individual toxicants, without effluent data, or prior to the
generation of effluent data. Water quality-based permit limits can be set for a
single toxicant or whole effluent toxicity (based on the available dilution and the
water quality criterion or State standard) in the absence of effluent data.
However, effluent testing is still used to ensure compliance with the permit limit.
80

-------
When conducting a non-database determination of whether or not an
effluent causes, or has the reasonable potential to cause, an excursion of a
numeric or narrative water quality criterion for individual toxicants or for toxicity,
the regulatory authority can use a variety of factors and information when
conducting an effluent safety assessment. (As discussed above, these factors
should also be considered in conjunction with available effluent data- See section
3.2.4). These factors should include but not be limited to the following.
o Dilution - Toxic impact is directly related to available dilution for the
effluent. Dilution is related to receiving stream flow and size of the
discharge. The lower the available dilution, the higher the potential for
toxic effect. If an effluent's concentration in a receiving water is expected
to reach 1% or higher during critical or worst case design periods, then
such an effluent may have a high priority for a toxicity limitation.
Assessment of the amount of stream dilution available should be made at
the annual low flow and at 7Q10. Figure 3-1 shows that a majority of
NPDES permittees nationwide discharge to areas during annual mean flow
ranging in dilution from 100-1000, however, during low flow conditions, the
majority of dischargers fall into the 1-10 dilution range.
o Type of industry - Although dischargers should be individually assessed
because toxicity problems are site-specific, the primary industrial categories
should be of principal toxicity concern. In general, secondary industrial
categories may have less potential for toxicity than primary industries.
However, based on experience, it is virtually impossible to generalize with
absolute certainty about the toxicity of effluents. If two plants produce the
same type of product, one effluent may be toxic while the other may not
be. The toxicity may simply be a function of the type and efficiency of the
treatment applied, general housekeeping practices, and the functional
target of the compound(s) being produced.
o Type of POTW - POTWs with large loadings frorti indirect dischargers
(particularly primary industries) may be high priority candidates for toxicity
limitations. However, absence of industrial input does not guarantee an
absence of POTW discharge toxicity problems. The types of industrial
users, their product lines, their raw materials, their potential and actual
discharges and their control equipment should be evaluated. POTWs
should also be assessed for the possibility of chlorination and ammonia
problems.
o Existing data on toxic pollutants - Discharge monitoring reports (DMRs)
and data from NPDES permit application forms 2C and 2A may provide
some indication of the presence of toxicants. The presence or absence of
the 126 "priority pollutants" may or may not be an indication of the
presence or absence of toxicity. There are thousands of "non-priority"
toxicants which may cause effluent toxicity. Also, combinations of several
81

-------
8,000-
j" 6,000-
"5 4.000
l 2.000
Low Row (7Q10)
15,863 <3,160 Majors)
Major*
j | Minors

2.467
2,771
2.084

¦
1,628
81
MM)



MM


1,241
i_L
470
1-10 10-20 20-60 50-100 100- 1,000- >10,000
1,000 10,000
Oilution <«treafn/affluent)
8,000-
6,000-
¦S 4.000-
* 2,000-
5,006
Annual Maan How
7,908
3,<
1,771
2,477
fl
3.848
2,066
M0 10-20 20-60 50-100 100- 1,000- >10,000
1.000 10.000
~Button (atraam/affluent)
Figure 3-1. National distribution of NPDES dilution conditions
at 7Q10 and at annual mean flow.
82

-------
toxicants can produce ambient toxicity where the individual toxicants would
not.
o History of compliance problems and toxic impact - Regulatory authorities
may consider particular dischargers with compliance problems or a history
of known toxicity impacts as probable priority candidates for effluent
toxicity limitations.
o Type of Receiving Water and Designated Use - Regulatory authorities
should consider water bodies which fall into the following categories (see
50 FR 23898, July 2, 1989):
1)	Waters where fishing or shellfish bans and/or advisories are currently in effect or
are anticipated;
2)	Waters where there have been repeated fishkills or where abnormalities (cancers,
lesions, tumors, etc.) have been observed in fish or other aquatic life during the last
ten years;
3)	Waters where there are restrictions on water sports or recreational contact;
4)	Waters identified by the state in its most recent state section 305(b) report as either
"partially achieving" or "not achieving" designated uses;
5)	Waters identified by the states under section 303(d) of the CWA as waters needing
water quality-based controls;
6)	Waters identified by the state as priority water bodies;
7)	Waters where ambient data indicate potential or actual exceedances of water quality
criteria due to toxic pollutants from an industry classified as a primary industry in
Appendix A of 40 CFR Part 122;
8)	Waters for which effluent toxicity test results indicate possible or actual exceedances
of state water quality standards, including narrative "free from" water quality criteria
or EPA water quality criteria where state criteria are not available;
9)	Waters with primary industrial major dischargers where dilution analyses indicate
exceedances of state narrative or numeric water quality criteria (or EPA water
quality criteria where state standards are not available) for toxic pollutants,
ammonia, or chlorine. These dilution analyses must be based on estimates of
discharge levels derived from effluent guidelines development documents, NPDES
permits or permit application data, or other available information;
10)	Waters with POTW dischargers requiring local pretreatment programs where
dilution analyses indicate exceedances of state water quality criteria (or EPA water
quality criteria where state water quality criteria are not available) for toxic
pollutants, ammonia, or chlorine. These dilution analyses must be based upon data
from NPDES permits or permit applications, Discharge Monitoring Reports, or
other available information;
11)	Waters with facilities not included in the previous two categories such as major
POTWs, and industrial minor dischargers where dilution analyses indicate
exceedances of numeric narrative state water quality criteria (EPA water quality
83

-------
criteria where state water quality criteria are not available) for toxic pollutants,
ammonia, or chlorine;
12)	Water classified for uses that will not support the "fishable/swimmable" goals of the
Clean Water Act;
13)	Waters where ambient toxicity or adverse water quality conditions have been
reported by local, state, EPA or other Federal Agencies, the private sector, public
interest groups, or universities;
14)	Waters identified by the state as impaired in its most recent Clean Lake
Assessments conducted under 314 of the Clean Water Act;
15)	Waters identified as impaired by nonpoint sources in the America's Clean Water;
The States' Nonpoint Source Assessments 1985 or waters identified as impaired or
threatened in a nonpoint source assessment submitted by the state to EPA under
section 319 of the Clean Water Act;
16)	Surface waters impaired by pollutants from hazardous waste sites on the National
Priority List prepared under section 105(8)(A) of CERCLA.
17)	Waterbody judged to be impaired as a result of a bioassessment/biosurvey.
Of course, the presence of a combination of a number of these factors,
such as low available dilution, high quality receiving water, poor compliance
record, and clustered industrial and municipal discharges, would constitute a high
priority for effluent limitations.
Regardless, the regulatory authority, if it chooses to impose an effluent
limitation after conducting an effluent assessment without data, will need to
provide adequate justification for the limitation in its permit development
rationale or in its permit fact sheet. A clear and logical rationale for the need
for the limitation will be necessary to defend the limitation should it be
challenged. In justification of a limitation, EPA recommends that the more
information the authority can acquire to support the limitation, the better a
position the authority will be in if a defense of the limitation becomes necessary.
In such a case, the regulatory authority may well benefit from the collection of
effluent data prior to establishing the limitation.
If the regulatory authority, after evaluating all available information on the
effluent, in the absence of effluent monitoring data, is not able to make a
decision on whether or not the discharge causes, or has the reasonable potential
to cause, an excursion of a numeric or narrative criterion for toxicity or for
individual toxicants, the authority should require whole effluent toxicity or
chemical-specific testing to gather further evidence. In such a case, the regulatory
authority can take the option of requiring the monitoring ahead of permit
issuance, if time exists, or it may require the testing as a condition of the issued
(reissued) permit.
84

-------
Box 3-2
Determining "Reasonable Potential'Tor Excursions Above Ambient Criteria Based on Factors
Other than Effluent Data
When determining the "reasonable potentiai"of a discharge to cause an excursion above a State
water quality standard, the regulatory authority should consider the guidelines described below on
a case-by-case basis.
Without Effluent Data: (If effluent data does not exist then evaluate the following factors*)
1.	Dilution.
2.	Receiving water type and designated/existing uses.
3.	Adopted State water quality criteria, or EPA criteria.
4.	Industry type- Primary, secondary, raw materials used, products produced,
BMPs, control equipment, treatment efficiency, etc.
5.	POTW type- Pretreatment, industrial loadings, no. of taps, unit processes,
treatment efficiencies, chlorination/ammonia problems, etc.
6.	Compliance history.
7.	Existing chemical data from DMRs and applications.
8.	Any other applicable case-specific factors.
*If effluent data also exist, see Box 3-4 and jointly evaluate the available effluent data with the above
factors.
Under these circumstances, the regulatory authority may find it to be protective
of water quality to include a "trigger" or a permit reopener for the imposition of a
effluent limit should the effluent testing establish an excursion of water quality
criteria. A discussion of these options is provided later in this Chapter.
3.2.3 Determining the Need for a Limitation With Effluent Data
If data already exist when characterizing an effluent for the need for a
whole effluent toxicity, or an individual toxicant limitation, the regulatory
authority has additional information upon which to base a decision. The
regulatory authority may already have effluent toxicity data available from
previpus monitoring, or it may decide to require the permittee to generate
effluent data prior to permit issuance, or as a condition of the issued permit.
85

-------
In the instance where the permittee is required to generate data in
advance, data collection should begin 12 - 18 months in advance of permit
development to allow adequate time for toxicity tests and chemical analyses to be
conducted. EPA recommends data be generated on effluent toxicity prior to
permit limit development for the following reasons: 1) the presence or absence of
effluent toxicity can be more clearly established or refuted; 2) where toxicity is
shown, effluent variability can be more clearly defined; and 3) the most sensitive
species for compliance monitoring and TREs can be identified. Several basic
factors which should be considered in generating effluent data are discussed
below.
Addressing Uncertainty in Effluent Characterization by Generating Effluent Data
Although quantitative uncertainty factors need not be used, some basic
principles still apply. All toxic effects testing and exposure assessment
parameters, for both effluent toxicity and individual chemicals, have some degree
of uncertainty associated with them. The more limited the amount of test data
available, the larger the uncertainty. The least amount of uncertainty of an
effluent's impact on the receiving water exists where: a) a complete data base is
available on acute and chronic toxicity on many indigenous species, b) there is a
clear understanding of ecosystem species composition and functional processes,
and c) actual measured exposure concentrations are available for all chemicals
during seasonal changes and dilution situations. The uncertainty associated with
such an ideal situation would be minimal. However, generation of these data
would be very resource intensive.
Uncertainty over receiving water effect concentrations is minimized when
chemical-specific limits are developed based on EPA water quality criteria,
because a large amount of data is collected for each chemical criterion. The
current method for calculating EPA's National Water Quality Criteria for a
specific toxicant utilizes laboratory acute and chronic testing of a large number
and variety of species to arrive at the criterion maximum concentration (CMC)
and the criterion continuous concentration (CCC). These concentrations, if not
exceeded, are expected to protect aquatic life (see 50 FR 30784, July 29, 1985)
and their designated uses.
An example of uncertainty associated with the amount of data collected
could be if a regulatory authority has only one piece of effluent data (i.e., an
LC50 of 50% for a chemical manufacturer). The effluent variability for this,
given the range of effluent toxicity variability seen in other effluents, may range
between 20% - 100% (Appendix A). It is impossible to determine from one
piece of data where in this range the effluent variability really falls. More data
would need to be generated to determine the actual variability of this effluent
and reduce this source of uncertainty. A detailed discussion of the principles
used in the generation of more effluent data for a more accurate effluent
characterization is described in section 3.2.4 below.
86

-------
Use of Toxicity Testing in Multiple-source Discharge Situations
Where more than one discharge is contributing to toxic impact, additional
testing may be needed to provide the permit authority with the information
necessary to separate the relative impact of each source. For purposes of this
discussion, a multiple-source discharge situation is defined as a site where impact
zones overlap. In multiple-source discharge situations, additivity, antagonism, and
persistence of toxicity often are of concern.
Assuming that screening has been conducted, two options for assessment
exist. The first option is for the permit authority to regulate each source
separately using the procedures for individual point sources. A second option is
to treat each discharge as an interactive component of a whole system. If this
second option is selected, the permit authority should require the testing
procedures presented
in Box 3-3.
Ambient Toxicity Testing
Another procedure, ambient toxicity testing, is useful in screening receiving
water bodies for existing toxic conditions. The procedure, described in Appendix
C, uses short term chronic toxicity tests (freshwater, marine, and estuarine) to
measure the toxicity of samples of receiving water taken above, at, and below (or
around) outfalls. The procedure must be conducted during an appropriate
low-flow or worst case design period.
The utility of the ambient toxicity screening approach is that actual
receiving water toxicity is measured, indicating the probability of toxic impact.
No extrapolation from exposure or acute-to-chronic ratio is needed. Further,
impact from multiple-source discharge situations, which may not be apparent
from individual discharger data, is identified. Finally, the technique can provide
an assessment of the persistence of effluent toxicity.
Special Considerations for Discharges to Marine and Estuarine Environments
Special problems are encountered when assessing and controlling impacts
of toxic pollutants discharging to marine and estuarine waterbodies. These
special problems include:
o physical characteristics of estuaries and the complex
mixing and effluent dilution situations for which
receiving water concentrations of effluents are difficult
to determine.
o generating toxicity data on non-saline effluents which
discharge to brackish or saline waters and establishing
cause-effect relationships on that basis.
87

-------
o assessing exposure and controlling impacts from
persistent toxicants accumulating in fish and shellfish
tissues and in sediments. These factors are
particularly important in estuaries and near coastal
waters due to high use of estuaries as breeding and
fishing areas for important commercial seafood
supplies and recreational fishing.
A more complete discussion of discharges to marine and estuarine
waterbodies can be found in the Permit Writers Guide to Water Oualitv-Based
Toxics Control for Marine and Estuarine Discharges [26].
Box 3-3
RECOMMENDED MULTIPLE-SOURCE TOXICITY TESTING PROCEDURES
o Tests
-	For effluent-dominated (where the effluent(s) make up 1% or greater of the flow)
receiving waters (assuming total flow input from all sources including other effluent flows is
measured), conduct chronic toxicity tests following the testing procedures described in
Recommendations for Whole Effluent Toxicity Data Generation.
-	For receiving water-dominated (where the effluent(s) make up less than 1% of the flow)
receiving waters (assuming total flow input from all sources including other effluent flows is
measured), conduct acute toxicity tests following the testing procedures described in the
Recommendations for Whole Effluent Toxicity Data Generation (Figure 3-2) determine if
any of the effluents are exhibiting toxicity.
-An additional data requirement in multiple-source discharge situations is the assessment of
relative and absolute toxicity of each source so that appropriate permit conditions can be set
for individual dischargers. The following procedure is suggested.
1.	Conduct one set of toxicity tests on the effluents using a control of reconstituted
or uncontaminated dilution water. The set of tests will give an absolute toxicity
measurement of the effluent.
2.	Run a parallel set of toxicity tests on the effluent using dilution water taken
directly upstream from the point of discharge or for estuarine waters, from an area
outside of the immediate discharge impact zone (this will have to be determined by
a dye study). This dilution water may be contaminated with upstream effluents or
other toxicant sources. The purpose of this test is to project toxic impact of the
effluent after it is mixed at its point of discharge. This is a relative effluent toxicity
measurement. The relative testing procedure could result in a change in the
standard concentration-effect curve generated by the testing.
The dilution water for the relative toxicity test may cause significant mortality,
growth or reproductive effects at the lower effluent concentrations (including the
100% diluent
88

-------
control concentration) if the diluent from the receiving water is toxic (from an
upstream discharge). Such mortality does not invalidate the test. Instead, analysis
of toxicity trends resulting from the relative toxicity tests can be used to assess the
effluent's toxicity in relation to other sources and ambient receiving water
conditions. However, a control dilution water with no toxicity must be used for
quality assurance and determination of absolute toxicity of the effluent.
3. Conduct ambient toxicity tests to: a) determine whether or not the effluent has a
measurable toxicity after mixing, b) measure persistence of toxicity from all sources
contributing to receiving water toxicity, and c) determine combined toxicity resulting
from the mixing of multiple, point and non-point sources of toxicity. See Appendix
C for a discussion of ambient toxicity testing procedures.
The ambient testing can be required of each discharger, and conducted during low flow or
worst case design periods.
o Frequency for Ambient Testing^
-	Conduct tests at appropriate frequencies as described in the previous section. All testing
should be conducted simultaneously by each discharger, if possible. At a minimum, the tests
should be conducted concurrently starting within a short time period (one to two days).
Repeated ambient toxicity analyses will be desirable when variable effluents are involved.
Effluent toxicity data showing variability can be used to assess what frequency will be most
applicable. The level of repetition for variability analysis should be similar to that used in
effluent variability analyses.
o Other Considerations
-	Dye studies of effluent dispersion for rivers, lakes, reservoirs, and estuaries are strongly
recommended. This allows analysis of effluent concentration at the selected sampling
stations above and below the discharge points.
-The procedures suggested in this multiple-source section are based on actual
multiple-source site investigations conducted under the Complex Effluent Toxicity Testing
Program. Site reports from that study can be used to obtain further description of the
toxicity testing procedures used to analyze multiple-source toxic impact [5,6].
3.2.4 Recommendations for Whole Effluent Toxicity Data Generation
Once an effluent has been selected for whole effluent toxicity
characterization, after consideration of the factors discussed above, the regulatory
authority should require toxicity testing in accordance with appropriate site-
specific considerations and the recommendations discussed below. In the past
five years, significant additional experience has been gained in generating effluent
toxicity data upon which to make decisions as to whether or not an effluent will
cause toxic effects in the receiving water in both fresh water and marine
environments.
89

-------
General Considerations and Assumptions
Three specific and important observations can be made which have led
EPA to modify its effluent toxicity data generation recommendations:
1.	Only extremely rarely have effluents discharged by NPDES permittees
been observed to have LC50's less than 1.0% or NOEC's less than 0.1%.
However, there is always a chance that an effluent could be toxic at such
low effluent concentrations.
2.	With the exception of a small number of "outliers"for which
confirmation is not possible, acute-to-chronic ratios for effluents
discharged by NPDES permittees have not been observed by EPA
above 20. However, higher ACRs may be observed for selected
facilities. (The majority bf observed ACRs are very seldom above
10.)
3.	The use of the three commonly used freshwater species,
Ceriodaphnia. fathead minnow, and Selanastrum: and three of the
five commonly used marine organisms, inland silverside, sheephead
minnow, mysid shrimp, Champia. and sea urchin; have generally
been sufficient to measure any effluent's toxicity for the purposes of
projecting effluent toxicity impact and making regulatory decisions.
(Other species may be used, provided they include a number of
species which represent a range of taxonomic phyla.) Data from
EPA's Complex Effluent Toxicity Testing Program as well as similar
data and results from the North Carolina DNR and other studies
(described in Chapter 1) all show that the three species testing
regimen typically employed (or fewer species where effluent
variability is adequately characterized) will be sufficient in a large
majority of cases to account for differences in species sensitivity.
Figure 3-2 is a flow chart of EPA's recommendations for data generation
for three different dilution scenarios. It is divided into three basic steps: initial
dilution determination, toxicity testing procedures, and triggers for permit limit
development. There are certain basic assumptions built into these flow charts.
First, the basic principle used in making decisions is to compare available
dilution to known or projected toxic effect concentrations to place an effluent into
one of three categories:
1. The effluent is causing an excursion of a water quality standard and
the permit requires a limit on toxicity.
90

-------
Figure 3-2
EFFLUENT CHARACTERIZATION FOR WHOLE EFFLUENT TOXICITY
STEP 1:
Dilution Determination 1
STEP 2:
STEP 3:
VP
Conduct Toxicity Testing
Based on Dilution Determination
(3 species at a min. of quarterly for 1 yr.)
acuta toxicity data or estimate based on ACR
I
Determine if CMC is Exceeded
i
yes
Develop
Permit
Limits
yes
no
Determine if
Reasonable
Potential 3
Exists
no
Require Monitoring
at Permit Reissuance
chronic toxicity data or estimate based on ACR
~	
Determine if CCC is Exceeded
yes



no
Develop
Permit
Limits
yes
Determine if
Reasonable
Potential 3
Exists
no
Require Monitoring
at Permit Reissuance
NOTES: 1 - Dilution Determinations: Should be performed for critical flows and any applicable mixing zones.
2	- Toxicity Testing Recommendations
a)	Dilution > 1000.1 : Acute Testing, check CMC only.
b)	100:1 < Dilution < 1000:1 . Acute or Chronic Testing, check CMC and CCC with data or ACR.
c)	Dilution < 100:1 : Conduct Chronic Testing, check CCC with data and CMC .using acute data or ACR.
3	- Reasonable Potential. Use Procedures In Boxes 3-2 and 3-4.

-------
2.	The effluent has a reasonable potential of causing an excursion of a water
quality standard and a limitation is required.
3.	The effluent has no, or only a very low, probability of causing an
excursion of a water quality standard and no limit is required.
This categorization is accomplished by using dilution estimates in the first
step and the results of the toxicity tests in the next steps. In addition, all these
impact estimates assume discharge at critical conditions and imposition of any
applicable mixing zone requirements. Therefore, a conservative assumption is
used to determine whether or not an impact is projected to occur. Estimates of
possible toxic impact are made assuming that the effluent is most toxic to the
most sensitive species or lifestage at the time of lowest available dilution.
The changes to the Agency's data generation recommendations are based
on the belief that it is no longer necessary to rely on the application of multiple
sets of safety margins. Rather, the application of general observations on effluent
toxicity described above now allow regulatory authorities to tighten the bounds of
the initial dilution categorization, eliminate the species sensitivity uncertainty
factor and target LC50s of 1.0% and NOECs of 0.1% as the most extreme
toxicity measurements which can normally be expected for the vast majority of
effluents discharged by NPDES permittees for acute and chronic toxicity,
respectively. For this data generation process, the use of single concentration
toxicity tests is strongly discouraged. Single concentration tests simply do no
provide enough information to make the decisions described below.
Since the new data generation requirements are much less expensive than
the previous requirements, tiered testing (less expensive single concentration
initial screening followed by increasingly expensive definitive data generation
using multi-concentration tests as described in the September 1985 version of the
TSD) is unnecessary. However, elimination of the requirement to conduct
toxicity testing on the basis of projections using dilution alone is not
recommended. The Clean Water Act requires protection of all State numeric
and narrative standards, including the "no toxics in toxic amounts" narrative
standard. This requirement as it relates to aquatic toxicity, coupled with the fact
that acute toxicity near the point of discharge is almost always of concern, suggest
that toxicity testing data are desirable to fully characterize discharges of concern.
Note that these effluent characterization recommendations rely on the
CMC and CCC as indicators of acute or chronic impacts respectively, and as
triggers for subsequent permit actions. As discussed in Chapter 2, the CMC and
CCC represent numeric expressions of the narrative water quality standard of "no
toxics discharged in toxic amounts." Therefore, the effluent characterization
procedures described below use the CMC and CCC as ambient numeric criteria
which, if shown to be exceeded (or where reasonable potential for exceedance
exists), trigger further regulatory action.
92

-------
Steps in Effluent Characterization Process
The following is a detailed description of the major steps presented in
Figure 3-2 and the logic behind each.
Step 1 - Dilution Determination
Three initial dilution categories are recommended for establishing data
generation requirements. These categories are use to determine which type of
toxicity tests are recommended to characterize an effluent:
Category A: Where effluent dilution at the edge of the mixing zone
is greater than 1000:1 (the RWC1 is less than 0.1%).
Categoiy B: Where efflnent dilution at the edge of the mixing zone is
greater than 100:1 but less than 1000:1 (RWC is between 1.0% and 0.1%).
Category C: Where effluent dilution at the edge of the mixing zone is less
than 100:1 (RWC is 1.0% or higher).
For each of these effluent dilution situations, toxicity testing
recommendations are specified. These data can then be used in making a
regulatory decision as to whether or not toxic impact will occur and toxicity
should be limited in the NPDES permit. The rationale for each of these
situations follows the same logic: where the difference between the projected
effluent concentration in the receiving water and effect concentration (as
indicated by the CCC and CMC for the parameter "toxicity") is large, no impact is
predicted. Where that difference is smaller, impact becomes more possible,
particularly after accounting for the "reasonable potential" to cause excursions.
Each of these dilution situations should take any applicable State mixing
zone requirements into consideration. Figure 3-3 shows a schematic
representation of typical mixing zone requirements. Development of mixing
zones for site-specific situations can be expensive. Simplified conservative mixing
zones may also be constructed at a reduced cost. However, for complex
situations, such as marine and estuarine waters or lakes, dye studies (or other
techniques used to assess mixing zones) may still be required.
1 The RWC represents the concentration of the effluent at the edge of the mixing zone at critical
flow conditions. The effluent's design flow or another higher effluent flow (such as the maximum
effluent flow expected during the life of the permit) should be used in this calculation, not the
average annual flow.
93

-------
Figure 3-3. Schematic representations of typical mixing zone
requirements.
94

-------
It is important to note that some State water quality standards do not
allow the use of mixing zones to control acute toxicity. For these States, the
following data generation recommendations must be modified. In these
situations, acute toxicity is often limited at the end of the pipe. In such cases, it
should be noted that chronic toxicity could still occur even where a discharger is
meeting acute toxicity limits and thus chronically toxic impacts must be assessed
as well in these situations. Note that permit limitations derived to enforce such
requirements would still be considered "water quality-based" since they would be
based upon an ambient criterion (as opposed to an arbitrary test endpoint). See
Chapters 4 and 5 and for additional discussions of this issue.
Step 2 - Toxicity Testing Procedures
In any case where toxicity tests are required in order to make decisions
regarding appropriate next steps in a screening protocol, EPA recommends that
three species (a vertebrate, an invertebrate, and a plant) be tested quarterly for a
minimum of one year. As discussed in Chapter 1, the use of three species is
strongly recommended since experience indicates that algal tests can be the most
sensitive test species in some situations. For both freshwater and marine
situations, the use of three species is more protective than two species since a
wider range of species sensitivity can be measured.
EPA recognizes the costs associated with extensive toxicity testing and
recommends that where test results clearly indicate one species to be most
sensitive to the effluent of concern and where effluent variability has been
adequately assessed, the above recommendations may be adjusted to require
testing of only the most sensitive of the species tested.
Conducting toxicity tests using three species quarterly for one year is
recommended to adequately assess the variability of toxicity observed in effluents.
Below this minimum, the chances of missing toxic events becomes increasingly
possible. Note that three toxicity tests using multiple species and the same
effluent sample should be considered a single test for purposes of applying the
"reasonable potential" procedure described in Box 3-4. The toxicity test result to
the most sensitive of the species tested for a particular effluent sample would be
recorded.
The data generation recommendations in Figure 3-2 represent minimum
testing requirements. Since uncertainty regarding whether or not an effluent
causes toxic impact is reduced with more data, EPA recommends that this test
frequency be increased where necessary to adequately assess effluent variability.
Note, however, that for variable effluents, if a choice must be made based upon
cost, it is preferable to use three species tested less frequently, than to test the
effluejit more frequently with only a single species, whose sensitivity to the
effluent is not well characterized.
95

-------
EPA recommends that the following types of toxicity tests be performed:
Category A - Dilution is > 1000:1
A discharger should be required to conduct acute toxicity testing if the
dilution of the effluent is greater than 1000:1 at the edge of the mixing zone.
A discharger should be considered a low priority for chronic toxicity testing
if the effluent dilution at the edge of the mixing zone is greater than 1000:1. The
rationale for this is that the effluent concentration would be below 0.1% at the
edge of the mixing zone and thus incapable of causing an excursion above the
CCC. (A worst case NOEC of 0.1% translates into 1000 TUc, which would result
in a concentration of less than 1.0 TUc at the edge of the mixing zone for this
dilution category.)
In these circumstances, EPA recommends that acute toxicity data be
generated. Comparison of test results with the CMC after consideration of any
allowable initial mixing will lead to one of the three outcomes discussed below
under "triggers for permit limit development."
Category B - Dilution is between 100:1 and 1000:1
A discharger should be required to conduct acute or chronic toxicity
testing if the dilution of the effluent falls between 100:1 and 1000:1 at the edge of
the mixing zone. (Where other factors are equal, EPA recommends the choice of
chronic testing since a chronic test gives data on acute toxicity as well. This acute
endpoint data can then be used to compare directly to the CMC without the need
for an ACR). Effluents have been shown to be both acutely and chronically toxic
within this range of receiving water dilution. (Under worst case scenarios, LC50s
of 1.0% and ACRs of 10 will result in excursions above both the CCC and CMC
at the edge of the regulatory mixing zone.)
EPA recommends that the most appropriate type of toxicity testing be
conducted in consideration of the dilution scenario for a particular discharger. At
the higher end of this dilution range (1000:1 or 0.1%), acute impact could occur
for reasons discussed under Category A above. At the lower end of this dilution
range (100:1 or 1.0%) the margin of safety shrinks and acute and chronic impacts
could occur during a period of lesser effluent toxicity (as compared to the higher
dilution scenario) or higher ACR.
Whichever type of toxicity test is specified should be used to determine
whether an excursion above the criterion associated with that type of test is
projected. Where excursions above either the CMC or CCC are projected based
on data collected, the regulatory authority should proceed directly to permit limit
development. Where excursions above either criterion are not projected based
upon the test data, the analyst should determine -- using the ACR or additional
data generated - whether an excursion above the other criterion is projected and
96

-------
thus, whether permit limits need to be developed. This may be done by first
performing toxicity tests on three species, or by applying the acute to chronic
ratio to the test results.
Note that these recommendations should not suggest that decisions on
permit actions be delayed as much as two years (e.g., a battery of acute tests
followed by chronic tests). Rather, the type of toxicity testing should be specified
by the regulatory authority at the outset and all requisite toxicity testing data
should be collected as a part of the permit application so that appropriate permit
limitations and conditions may be included at permit issuance or reissuance.
Category C - Dilution is < 100:1
A discharger should be required to conduct chronic toxicity testing if the
dilution of the effluent falls belew 100:1 at the edge of the mixing zone.
The rationale for this recommendation is that chronic toxicity is often
observed in effluents down to the 1.0% effect concentration. Therefore, chronic
toxicity tests, although somewhat more expensive to conduct, should be used
directly in order to make decisions about toxic impact. There are dilution
scenarios within this category where acute toxicity is of concern, even though the
CCC is shown not to be exceeded. These scenarios are more possible as the
100:1 dilution level is approached.
Thus, the recommended screening protocol shown in Figure 3-2 includes a
determination of whether or not excursions above the CMC are projected. This
analysis may be performed by assuming an ACR, applying this value to the
chronic toxicity testing data, and allowing for any allowable initial mixing.
Alternatively, the regulatory authority may require the generation of acute toxicity
test data and require the use of this information to perform the comparison.
However, as noted above, this is not to suggest that decisions on permit actions
be unnecessarily delayed.
If an excursion above the CCC is projected based upon the chronic toxicity
tests, the regulatory authority should proceed directly to permit limit
development. If an excursion above the CCC is not projected, the regulatory
authority should use the chronic toxicity testing data, together with the ACR or,
in the alternative, generate acute data to see if an excursion above the CMC is
projected.
Step 3 ¦ Triggers for Permit Limit Development
Once the toxicity data have been generated for a discharger, the regulatory
authority must decide whether or not the results show that the permittee causes,
has the reasonable potential to cause, or contributes to a toxic effect in the
receiving water and therefore needs to limit effluent toxicity based upon the
effluent data characterization. To do this, these data should be used to project
97

-------
receiving water concentrations and these concentrations compared to the CCC
and CMC One of four outcomes will be reached when following the screening
protocol shown in Figure 3-2:
1.	Excursion above CMC or CCC: Any one data point shows an
excursion above the State's numeric or narrative standard for
toxicity. Where the State's standard for the parameter toxicity is
shown to be exceeded, EPA regulations require a permit limit be
set for whole effluent toxicity (40 CFR 122.44(d)(l)(iv or v), see
Appendix B-3) unless limitations on a specific chemical will allow
the narrative water quality standard to be attained or maintained.
In the absence of a State numeric criterion for the parameter
toxicity, EPA recommends that 1.0 TUc and 0.3 TUa be used as the
CCC and CMC respectively. It is important to recognize that any
decision to develop permit limits based upon an excursion above
either the CMC or CCC should lead to development of
comprehensive permit limitations which provide for both acute and
chronic protection. (Note, however, that such permit limitations
need not require both acute and chronic toxicity testing for
compliance monitoring purposes; see discussion under Section
5.7.3.).
2.	Potential for Excursion above CMC or CCC: EPA believes, based upon
effluent data alone, "reasonable potential" is proved where an effluent is
projected to cause an excursion above the CCC or CMC based upon a
statistical analysis of available data which accounts for limited sample
size and effluent variability. EPA's detailed recommendations for making
a statistical determination based upon effluent data alone are shown in
Box 3-4. Where a regulatory authority finds that test results alone indicate
a "reasonable potential" to cause an excursion above a State water quality
standard in accordance with 40 CFR 122.44(d)(1)(H), a permit limitation
must be developed.
(In some cases the statistical analysis (Box 3-4) of the effluent data may
not actually show a reasonable potential for excursions above the CMC or
CCC but may be close. Under such conditions, reasonable potential
determinations will include an element of judgment on the part of the
regulatory authority, since a number of factors, in addition to effluent data,
will need to be considered and given appropriate weight in the decision-
making process, including: value of water body (e.g., high use fishery),
relative proximity to the CCC or CMC, existing controls on point and
nonpoint sources, information on effluent variability, compliance history of
the facility, and type of treatment facility, (these factors are summarized in
Box 3-2 and are discussed in detail in section 3.2.2). EPA recommends
regulatory authorities establish a written policy and procedure for making
determinations of "reasonable potential" under these circumstances.
98

-------
3.	No Excursions Above CMC or CCC (and no reasonable potential for
excursions above the CMC or CCC based on all available information): In
these situations, EPA recommends that the toxicity tests recommended
above be repeated at a frequency of at least once every five years as a part
of the permit application.
4.	Where a regulatory authority has inadequate information upon which to
make a decision of reasonable potential for an excursion of a numeric or
narrative water quality standard and there may still be a basis for concern
on the part of the authority: The permit should contain whole effluent
toxicity testing requirements and a reopener clause. This clause would
require reopening of the permit and establishment of a limit based upon
any test results, or other new factors, which substantiate that the effluent is
causing or has a reasonable potential of causing an excursion above the
CCC or CMC
Other Approaches
Some States use the RWC itself as the permit limit for toxicity and require
compliance monitoring for that limit using a toxicity testing method which
employs only a control and a single exposure at the RWC. Although single
dilution screening tests can be less expensive than full dilution series testing, Such
tests provide no knowledge as to the extent of toxicity present during the test and
therefore no guidance concerning the seriousness of the impact or the amount of
toxicity reduction necessary. The death of a single test animal can occur at any
concentration level beyond the lethality threshold for the test organism and such
a test is therefore much less powerful from a statistical standpoint. In addition, it
is not possible to determine dose-response relationships for the test organisms
without using multiple effluent concentrations. Dose-response curves are useful
in determining quality assurance of the tests, and in defining threshold dosages
for regulatory purposes. Regulatory authorities which use this approach should
require through permit triggers multi-concentration tests when the RWC is
determined to be exceeded by the single screening test. Single exposure testing
should not be used for effluent characterization because of the limited
information it provides.
3.2.5 Effluent Characterization for Specific Chemicals
The previous section discussed effluent characterization for whole effluent
toxicity. This section will describe EPA's recommendations for data generation to
be used to determine whether or not permit limits are needed to control specific
chemical pollutants in effluents. While many of the same principles apply when
developing chemical-specific limits, there are some differences based upon
regulatory and analytical considerations.
Characterization of impacts due to specific chemicals do not require a
determination of the type of testing based on available dilution as is required for
99

-------
whole effluent toxicity. This is because there is generally only one type of test for
specific chemicals which may be used to determine whether an excursion above a
criterion or the reasonable potential for an excursion exists. However, there are
some antecedent steps that are unique to effluent characterization for specific
chemicals: determination of the chemicals of concern and determination of
acceptable ambient levels (AAC, CMC, or CCC) for these pollutants.
Recommendations for Chemical-Specific Data Generation
Figure 3-5 illustrates EPA's recommendations for determining whether or
not permit limits need to be developed based upon an evaluation of a limited
data set. The following discussion corresponds to the various activities shown in
Figure 3-5. (Refer to the human health discussion in Section 3.3.1 for additional
details on procedures to characterize the bioconcentration potential of effluents).
Step 1 - Identify the Pollutants of Concern
This process should begin with an examination of existing data to
determine the presence of specific toxicants for which criteria, standards, or other
toxicity data are available. Sources of data include:
o Permit application forms, DMRs, PCS, and permit files,
o Pretreatment industrial surveys.
Purpose: This is an explicit procedure with which a "reasonable potential" to exceed an ambient
criterion can be determined from effluent data, based on limited sampling information and
probability models. (The results of this procedure should be assessed in combination with all other
reasonable potential factors discussed in section 3.2.2 before arriving at a final determination).
Box 3-4
Determination of "ReasonablePotential"for Excursions
Above Ambient Criteria Based Only on Effluent Data
Procedure:
Step 1:
¦ Determine the number of total observations for a particular set of effluent data
(concentrations or TUs) and the highest value from such a data set.
Step 2:
¦ Determine the coefficient of variation for the data set (actual or estimated).
Step 3:
¦ Select the probability basis (95% or 99%).
Step 4:
¦ Select using table 3-1 or 3-2 the multiplier of highest data point.
Step 5:
¦ Multiply highest value from data set and use with allowable dilution to project an
RWC which is then compared to the CMC, CCC, or AAC. Where such a effluent
100

-------
concentration is projected to cause an excursion above an ambient criterion,
"reasonable potential" is shown and a permit limitation should be developed.
A table of "reasonable potential" multiplication factors for 99% confidence and a 99%
probability basis is provided in Table 3-1 as a function of number of samples and assumed coefficient
of variation. Also, multiplication factors for 99% confidence and a 95% probability basis are
provided in Table 3-2.
Example
Consider the following results of toxicity measurements of an effluent which is being
characterized: 5 TUc, 2 TUc, 9 TUc, and 6 TUc. Assume that the discharge flow is 20 gpm and the
receiving stream critical flow is 900 gpm. The maximum of the 4 sample results is 9 TUc and the
multiplying factor is 4.7 based on an assumed discharge CV of 0.6 for chronic toxicity and a 99%
probability basis. The calculation determination whether or not reasonable potential exists is as
follows:
0.92 TUc = [9 TUc x 4.7] x [20 gpm/(20gpm + 900 gpm)]
The resultant calculation of 0.92 TUc is less that the ambient concentration of 1.0 TUc. Therefore,
for this example, there is not "reasonable potential" for the toxicity in the discharge to cause and
excursion above the CCC.
Rationale
The procedure is derived from the theories of probability applied to the lognormal
distribution which is used in describing effluent behavior. For a specified CV and assuming effluent
behavior can be described using a lognormal distribution an entire distribution of values can be
projected from limited data (See Figure 3-4). One can examine a particular point on the curve
(e.g., the 99 th percentile, or the point below which 99% of the values fall), and can compare this
point to other points in the distribution of values and make assumptions about that relationship with
some degree of confidence.
In the example above, once can state that the maximum of 4 samples is greater than the
31.6th percentile of all potential samples from the same population with 99% confidence. So, with
99% confidence it can be asserted that:
31.6th percentile < maximum of 4 samples
With an assumed coefficient of variation (CV), say 0.6, the lognormal distribution theory provides a
means of calculation of the ratio of any two percentiles of the distribution. In particular, the ratio of
the 99th percentile to the 31.6th percentile is 4.7 assuming the CV is 0.6. Therefore, we can state
with 99% confidence that
99th percentile value =
4.7 x (31.6th percentile value) < 4.7 x (maximum of 4 samples)
If the product on the right will not cause a concentration above the ambient criterion after
accounting for available dilution, then a pollutant concentration at the 99th percentile of the
pollutant or pollutant parameter distribution will not exceed the standard at critical flow condition
with 99% confidence.
101

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
_Q
3
5
0
1
6
5
2
8
0
6
5
6
9
2
7
2
8
4
i
8
Coefficient of Variation
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
1.8
2.5
6.0
13.2
26.5
48.3
81.4
128.0
190.3
269
2.0
4.0
7.4
12.7
20.2
30.3
43.0
58.4
76
1.9
3.3
5.6
8.9
13.4
19.0
25.7
33.5
42
1.7
2.9
4.7
7.2
10.3
14.2
18.6
23.6
29
1.7
2.7
4.2
6'. 2
8.6
11.5
14.8.
18.4
22
1.6
2.5
3.8
5.5
7.5
9.8
12.4
15.3
18
1.6
2.4
3.6
5.0
6.7
8.7
10.8
13.1
15
1.5
2.3
3.3 •
4.6
6.1
7.8
9.6
11.6
13
1.5
2.2
3.2
4.3
5.7
7.1
8.7
10.4
12
1.5
2.2
3.0
4.1
5.3
6.6
8.0
9.5
11
1.5
2.1'
2.9
3.9
5.0
6.2
7.4
8.8
10
1.4
2.0
2.8
3.7
4.7
5.8
7.0
8.1
9
1.4
2.0
2.7
3.6
4.5
5.5
6.5
7.6
8
1.4
2.0
2.6
3.4
4.3
5.2
6.2
7.2
8
1.4
1.9
2.6
3.3
4.1
5.0
5.9
6.8
7
1.4
1.9
2.5
3.2
4.0
4.8
5.6
6.5
7
1.4
1.9
2.5
3.1
3.8
4.6
5.4
6.2
7
1.4
1.8
2.4
3.0
3.7
4.4
5.2
5.9
6
1.4
1.8
2.4
3.0
3.6
4.3
5.0
5.7
6
1.3
1.8
2.3
2.9
3.5
4.2
4.8
5.5
6
Table 3-1. Reasonable Potential Multiplying Factors:
99% Confidence and 99% Probability Basis
9
6
3
1
4
3
6
6
2
0
1
4
7
2
7
3
0
7
4
1
102

-------
Number	Coefficient of Variation
of
SampLes
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
1
2.2
4.6
9.1
16.3
27.3
42.6
62. 7
87.9
118.5
154.4
2
1.8
3.1
5.1
7.8
11.4
15.8
21.1
27.0
33.6
40. 9
3
1.6
2.6
3.8
5.5
7.6
9.9
12.6
15.5
18.6
21.8
4
1.5
2.3
3.2
4.4
5.8
7.4
9.1
10.9
12.8
14. 7
5
1.5
2.1
2.9
3.8
4.9
6.0
7.2
8.5
9.8
11.2
6
1.4
2.0
2.6
3.4
4.2
5.1
6.1
7.1
8.0
9.0
7
1.4
1.9
2.4
3.1
3.8
4.5
5.3
6.1
6.8
7.6
8
1.3
1.8
2.3
2.9
3,5
4.1
4.7
5.4
6.0
6.6
9
1.3
1.7
2.2
2.7
3.2
3.7
4.3
4.8
5.3
5.9
10
1.3
1.7
2.1
2.5
3.0
3.5
3..9
4.4
4.8
5.3
11
1.3
1.6
2.0
2.4
2.8
3.2
3.6
4.0
4.4
4.8
12
1.3
1.6
1.9
2.3
2.7
3.0
3.4
3.8
4.1
4.4
13
1.2
1.5
1.9
2.2
2.5
2.9
3.2
3.5
3.8
4.1
14
1.2
1.5
1.8
2.1
2.4
2.7
3.0
3.3
3.6
3.9
15
1.2
1.5
1.8
2.0
2.3
2.6
2.9
3.1
3.4
3.6
16
1.2
1.5
1.7
2.0
2.2
2.5
2.8
3.0
3.2
3.4
17
1.2
1.4
1.7
1.9
2.2
2.4
2.6
2.9
3.1
3.3
18
1.2
1.4
1.6
1.9
2.1
¦2.3
2.5
2.7
2.9
3 1
19
1.2
1.4
1.6
1.8
2.0
2.3
2.4
2.6
2.8
3.0
20
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.5
2.7
2.9
Table 3-2. Reasonable Potential Multiplying Factors:
99% Confidence and 95% Probability Basis
103

-------
S 2
3
1
0
2
Long term average
95th percentile
99th percentile
0	1- . 2	3	4	5
Valu*
Figure 3-4a. Frequency Distribution of Values for a Lognormal
Distribution with a Mean of 1.0 and a Coefficient of Variation of 0.6
>
3
i 1



Long term average
p
/

\ CVm0.2


CV]°-9J\


\ cy-o.4

CV-0.6
0	1	2	3	4	5
Value
Figure 3-4b . Comparison of Relative Frequencies of Lognormal
Distributions with a Mean of 1.0 for Different Coefficients of
Variation
104

-------
Figure 3-5
EFFLUENT CHARACTERIZATION FOR SPECIFIC CHEMICALS
NOTES: 1 - AAC and/or CMC/CCC: Use State numeric criterion or InterpretState narrative criterion using 1 of 3 options
specified under 40 CFR 122.44(d).
2	- Dilution Determination: Should be performed for critical flow and any applicable mixing zones for aquatic life
and human health protection respectively.
3	- Reasonable Potential: Use Procedures In Boxes 3-2 and 3-4.

-------
o STORET for U.S. Geological Survey (USGS) flow data and ambient
monitoring data,
o Industrial effluent guidelines development documents,
o The Treatability Manual [13].
o Toxicity reduction evaluation results for selected industries.
Data on specific chemicals which are typically submitted with NPDES
application forms will consist of a limited number of analytical test results for
many of the reported parameters. Where the regulatory authority has reason to
believe that additional data for key parameters of concern are needed in order to
adequately characterize the effluent, this information should be requested as a
part of the application or, in some cases, through the use of Section 308 letters.
It is recommended that 8 to 12 samples be analyzed for key parameters of
concern. In some cases, special analytical protocols will need to be specified in
order to gather all appropriate information. For instance, the protocol described
in Section 3.3.1 is a whole effluent screen followed by analysis for specific
bioconcentratable chemicals present in the effluent.
Step 2 - Determine the Basis for Establishing AACs, CMCs, and CCCs for
the Pollutants of Concern
If a state does not have a numeric water quality criterion for the pollutant
of concern, then one of three options for using the narrative criterion may be
used (40 CFR 122.44(d)(l)(vi)) to determine whether a discharge is causing or
has the reasonable potential to cause an excursion above a narrative standard due
to an individual pollutant. Note that the provisions of 40 CFR 122.44(d)(l)(vi)
are presented in the regulation in the context of permit limit development.
However, these same considerations should be applied in characterizing effluents
in order to determine whether limits are necessary.
Option A allows the regulatory authority to establish limitations
using a "calculated numeric water quality criterion" which the
regulatory authority demonstrates will attain and maintain
applicable water quality standards. This option allows the
regulatory authority to use any criterion that protects aquatic life
and human health. This option would also allow the use of site
specific factors, including local human consumption rates of aquatic
foods, the state's determination of an appropriate risk level, and any
current data that may be available.
Option B allows the regulatory authority to establish effluent limits
using EPA's Water Quality Criteria guidance documents, if EPA has
published a criteria document for the pollutant (supplemented,
where necessary, by other relevant information). As discussed
earlier, EPA criteria documents provide a comprehensive summary
of available data on the effects of a pollutant.
106

-------
Option C may be used to develop limits based on an indicator
parameter under limited circumstances. (An example of an
indicator parameter is total toxic organics (TTO); effluent limits on
TTO are useful where an effluent contains organic compounds).
However, use of this option must be justified to show that controls
on one pollutant control one or more other pollutants. Where such
data are available, this option may be used provided several
conditions are met. (See 40 CFR 122.44(d)(l)(vi)(C)). It should be
noted that using this option in screening presents complications.
When trying to determine whether or not a pollutant of concern has
a reasonable potential to cause an excursion above the narrative
standard, development of specific information on the pollutant of
concern and comparison with an AAC (using option A or option B)
will normally be the preferred approach. Note that the
determination of the ambient criterion may be an AAC for human
health protection or a CCC and CMC established for aquatic life
protection — or both. The protocol described in Section 3.3.1 is
designed to identify specific bioconcentratable pollutants of concern
for which AACs can be developed using the equations found in
Chapter 2.
Step 3 - Dilution Determination
Once a basis for comparison has been established, the margin between
projected effluent chemical concentration and the AAC, CMC, or CCC after any
allowable mixing may be compared. As discussed in Chapter 2, the water quality
analyst should be careful to apply applicable mixing considerations for aquatic life
and human health, respectively. These comparisons will lead to one of the three
outcomes discussed below.
There are two levels of analysis for chemical data, both based on the
margin between measured chemical concentration and the AAC, CMC, or CCC
(e.g., State water quality standards or use of one of the above options). The first
level is to use simple fate models based on a dilution analysis and comparison
with the AAC, CMC, or CCC. The second level of analysis is to use more
complex fate models including dynamic models to estimate persistence and may
be applied to lakes, rivers, estuaries, and coastal systems using a desktop
calculator or microcomputer. EPA has supported development of a second level
of analysis that estimates point source wasteload allocations and nonpoint source
allocations and predicts the resulting pollutant concentrations in receiving waters
[12].
Step 4 - Triggers for Permit Limit Development
t After this dilution analysis has been performed, the projected RWC is
compared to the AAC, CMC, or CCC (either the state numeric criteria or an
interpretation of the narrative criteria as described earlier). Note that
107

-------
determinations of excursions above aquatic life effects should include evaluations
with respect to both the CCC and the CMC. Analysis of human health impacts
will only involve comparisons with the AAC. The three possible outcomes
discussed above under section 3.2.4 are:
1.	Excursion above the AAC, CMC or CCC.
2.	Potential for an excursion above the AAC, CMC or CCC.
3.	No excursion above the AAC, CMC, CCC.
If these evaluations do not project excursions above the AAC, CMC, or CCC, an
analysis, based upon a limited data set, of whether or not the effluent has a
"reasonable potential" to cause an excursion should be performed using the
procedures described in Box 3-4. If the effluent is shown to cause or has the
reasonable potential to cause an excursion above an ambient criterion, then a
permit limit is required (40 CFR 122.44(d)(l)(iii)). (See Chapter 4 for Waste
Load Allocation and Chapter 5 for limit derivations.) Effluents that are shown
not to cause or have a reasonable potential to cause, an excursion above an AAC,
CMC or CCC should be re-evaluated at permit reissuance.
Where chemical specific test results do not show a "reasonable potential"
but indicate a basis for concern in consideration of the various narrative factors
discussed under "reasonable potential", the permit should contain chemical testing
requirements and a reopener clause. This clause would require reopening of the
permit and establishment of a limit based upon any test results which show
effluent toxicity at levels which are causing or have a reasonable potential of
causing an excursion above the AAC, CCC, or CMC.
Analytical Considerations
Analysis of discharges for toxic substances requires special quality control
procedures beyond those necessary for conventional parameters. Toxicants can
occur in trace concentrations and are frequently volatile or otherwise unstable.
An EPA publication entitled, "Test Methods - Technical Additions to methods for
Chemical Analysis of Water and Wastes" (December 1982) [17] contains sampling
and handling procedures recommended by EPA for a number of toxic and
conventional parameters. Methods for analyses for all toxicants are described in
Standard Methods of Water and Wastewater Analyses (ASTM, 17th edition, 1989,
or most recent edition) and 40 CFR Part 136. For a more detailed discussion of
detection limits and sampling requirements, see
Chapter 4.
108

-------
3.3 ASSESSMENT OF HUMAN HEALTH EFFECTS
The previous section discussed assessing aquatic life effects and the use of
data and other information to determine whether or not the discharge is causing
or has the potential to cause excursions above an ambient criterion. This section
will describe the additional data needs and considerations used to assess human
health effects.
There are three potential exposure routes of human health concern: 1)
drinking water, 2) recreational contact, and 3) consumption of fish or shellfish
contaminated by toxicants. The first exposure route is addressed via public water
supply treatment regulations (maximum contaminant levels, MCLs). The second
exposure route is considered of little concern due to the small likelihood of
exposure to sufficiently hazardous concentrations.
The major focus of water quality-based toxics control efforts will be, in this
document, on the third exposure route: consumption of fish and shellfish
contaminated by toxicants that bioaccumulate. (Bioaccumulation is the process
by which a compound is taken up by an aquatic organism, both from water and
through food. A related phenomenon, bioconcentration, is the process by which a
compound is absorbed from water through gills or epithelial tissues and is
concentrated in the body. Biomagniflcation denotes the process by which the
concentration of a compound increases in different organisms, occupying
successive trophic levels.) Actual human exposure from fish and shellfish varies
depending upon species, tissue type, lipid content, consumption rate and pattern,
and food preparation practices.
This document focuses on control of the fish ingestion human exposure
route for several reasons. First, unlike public water supply treatment, no
intervening treatment system exists to eliminate toxic chemical residues in fish
and shellfish prior to human consumption. Second, even extremely low ambient
concentrations of some pollutants (sublethal to aquatic life) can result in chemical
concentrations in fish or shellfish tissue which can pose human health risk. Third,
preventing the formation of such harmful chemical residues in fish and shellfish
tissue is an important component of EPA's water quality-based toxics control
approach to protecting human health.
As with aquatic life, there are two approaches to characterizing effluents
for human health effects: chemical-specific and whole effluent. At present, only
the chemical-specific effluent characterization approach is widely used for the
following reasons:
o Measurements of individual chemicals can be readily made,
o Human health toxicological data are available for many chemicals.
109

-------
o Human exposure can be calculated using fish or shellfish consumption
estimates.
o Human health impacts of mixtures of individual chemicals are unknown
due to interactions such as additivity, antagonism, etc. that are not well
understood.
The primary disadvantages of the chemical-specific approach to assessing
potential human health impacts are like those of single chemical aquatic life
impact assessment: not all effluent chemicals of potential human health concern
are identified, nor are interactions among effluent components assessed. To help
address this gap, EPA recommends identifying pollutants of concern using the
protocol described in Section 3.3.1 and calculating "acceptable ambient
concentrations" (AACs) by applying the criteria amendment formulas discussed
in Chapter 2 to those organic pollutants lacking EPA human health criteria.
Whole effluent human health hazard assessment techniques are briefly
discussed in Appendix G. Due to the difficulty in establishing cause/effect
relationships between whole effluent human health surrogate measurements and
actual impacts on exposed individuals, EPA is not yet recommending the
application of these procedures to set permit limits. Some regulatory authorities
currently utilize such data in screening or impact assessment procedures to trigger
further analysis of toxic constituents of effluents.
3.3.1 Effluent Bioconcentration Evaluation
The following recommended evaluation procedures are based on a draft
EPA guidance document entitled, "Assessment, Criteria Development and Control
of Bioconcentratable Contaminants in Surface Waters". This document contains
information for regulatory authorities on effluent bioconcentration evaluation
principles, how to select dischargers needing to perform effluent bioconcentration
analysis, actual laboratory procedures, and how to analyze the resulting data and
develop acceptable ambient concentrations and permit limitations. The
procedures described in this document are currently being field validated. In
simple terms, this evaluation method is a whole effluent approach which produces
chemical-specific information. The method determines the bioconcentration
potential of a whole effluent (or any other aqueous sample) and culminates in
identification of the individual bioconcentratable chemicals in the mixture. This
information can then be used to develop acceptable ambient criteria and permit
limitations.
Principles of Bioconcentration Control
An effluent bioconcentration evaluation relies on several key principles
based on the nature of bioconcentratable pollutants. A bioconcentration factor
(BCF) is a measure of a chemical's tendency to concentrate in tissues of aquatic
110

-------
organisms. There are two primary ways to derive BCFs: calculation from
experimental measures or prediction using structure-activity relationships.
BCFs can be calculated from experimental measures via two methods.
The most common method is to calculate the BCF by dividing the measured
concentration of the chemical in the exposed tissue by the measured
concentration of the chemical in the exposure water, after a steady-state condition
is reached [28]. In equation form:
BCF = Concentration in Tissue
Concentration in Water
Bioconcentration factors can also be calculated experimentally by
measuring a chemical's kinetic uptake (kl) and elimination (k2) rates constants
and subsequently, dividing the uptake rate by the elimination rate [29]. In
equation form:
BCF = kl/k2
Alternatively, BCFs can be estimated (predicted) using structure-activity
relationships based upon a well-documented relationship between the BCF and
the n-octanol/water partition coefficient for organic chemicals (P) [28,30,3132].
EPA believes that BCF values for chemicals with high log octanol-water partition
coefficients (log P > 3.5) estimated in this way provide good, consistent BCF
values. The Quantitative Structure-Activity Relationships database (QSAR)
contains information useful in estimating BCF values. QSAR-estimated BCF
values form the basis for EPA's recommended bioconcentration analysis
procedure.
Some scientists prefer the use of experimentally measured BCF values.
The use of experimentally measured BCF values instead of the QSAR-estimated
values is not generally recommended by EPA since great variation in measured
BCF values exist for a given chemical. This variability arises from the existence
of either inappropriate experimental conditions during the BCF determination
and/or from metabolism of the chemical during the BCF determination. EPA
believes that reported BCF values which were determined under either or both of
the above conditions are incorrect, and recommends QSAR estimates to avoid
this problem.
It is important to recognize that the effluent bioconcentration analysis
procedure is subject to a number of basic principles and assumptions. These
principles and assumptions, described below, provide a number of constraints on
the application of the analytical procedure.
o The effluent bioconcentration analysis procedure does not detect and
quantify all bioconcentratable compounds.
Ill

-------
The analytical method contained within EPA's guidance document has
been specifically designed, constructed, and optimized for detecting highly
bioconcentratable organic chemicals. The procedure will not detect those
chemicals which are unstable under acidic conditions or those which have log P
values below the bioconcentration threshold of 3.5. (For further information on
this bioconcentration threshold, see EPA's guidance, "Assessment, Criteria
Development and Control of Bioconcentratable Contaminants in Surface
Waters.") Finally, this procedure will not detect those bioconcentratable
chemicals present at concentrations below the method detection limit.
o The effluent bioconcentration analysis procedure generally cannot be
applied successfully to hydrocarbons.
The procedure is optimized for detecting nonpolar organic chemicals with
high log P values. Hydrocarbons found in petroleum products such as oils,
paraffin waxes, lubricants, and coal are also nonpolar organics chemicals with
high log P values. Effluents from refineries and other industries which process or
use large quantities of hydrocarbon materials often contain relatively large
amounts of hydrocarbons in their effluent samples. When analyzed using the
procedure, such hydrocarbon-filled effluents often provide unusable data due to
the large hydrocarbon background observed in the gas chromatograph/mass
spectrometer (GC/MS) analysis.
o The effluent bioconcentration analysis procedure is applicable to nonpolar
organic chemicals that are stable under acidic conditions.
Bioconcentratable chemicals which are unstable in acidic conditions will
not be detected with the procedure. This is because an acid clean-up of the
effluent sample is included in the procedure to remove biologically-derived
organic chemicals commonly found in effluents. If not removed, these materials
(i.e., fatty acids, fatty acid esters, sterols, phthalates, and phenolic plant materials)
cause serious interferences in common chemical residue analysis procedures.
Fortunately, these materials which must be removed have low potential to
bioconcentrate since, in general, they are easily metabolized and/or have log P
values below the 3.5 threshold.
. o The effluent bioconcentration analysis procedure eliminates chemicals with
low potential to bioconcentrate (i.e., chemicals with log P values below the
3.5 bioconcentration threshold) from analysis.
The effluent analysis procedure uses a well-documented relationship
between log P and the retention time of chemicals on a reverse phase high
pressure liquid chromatography (HPLC) column to fractionate the effluent extract
[33,34]. Fractionation of the effluent extract allows chemicals with log P values
less than the bioconcentration threshold to be removed and allows the highly
bioconcentratable portion of the effluent extract to be subdivided.
112

-------
o The. effluent bioconcentration analysis procedure bases its initial
quantification measurements upon model compounds (surrogates) added to
the effluent sample in known amounts prior to extraction.
The quantifications on the GC/MS are performed by using the responses
of the surrogates to quantify all components in their respective fractions. This
method of quantification assumes that the analytical recoveries and mass spectral
responses are the same for all components within each fraction. These
assumptions can be expected to cause error in the quantifications. However, it is
important to remember that the effluent bioconcentration analysis procedure is a
screening method only. After initial compound identification and risk analysis,
selected dischargers should be required to perform confirmation and
quantification analyses using standards made from the pure compound (EPA
Method 680) [35].
o The effluent bioconcentration analysis procedure has detection limits
approaching 10 ng/L.
In the procedure, the surrogates are added to the effluent prior to
extraction at a concentration of 100 ng/L. Detection of the surrogates during the
GC/MS analysis insures that a minimum detection limit of 100 ng/L occurs in
each analysis. The 100 ng/L concentration threshold was selected after thorough
consideration of current residue chemistry techniques and represents a minimum
level of detection which can be easily obtained on a routine basis with contract
laboratories. However, performance of the effluent bioconcentration analysis
procedure with highly bioconcentratable compounds at 1, 10, 25, 50, and 100
ng/L concentrations demonstrates that the detection limits approaching 10 ng/L
are obtainable.
o The effluent bioconcentration analysis procedure assigns tentative
identifications to chemical components using mass spectral library
searching.
In the procedure, chemical components detected by the GC/MS are
searched against the EPA/NIH/NBS mass spectral library. Components with
library searching fits of 70% and above are considered tentatively identified. For
each component with tentative identifications, the ten best tentative compound
identifications with fits of 70% and above should be reported by the lab to the
permit authority. The recommended 70% library searching fit criterion was
chosen based upon best scientific judgement.
An overall summary of the evaluation method is shown in Figure 3-6.
113

-------
Figure 3-6.
PROCEDURES FOR OENIWCATlON AKO CONTROL OF BOCONCfNTR ATABLE
POLLUTANTS IN COM>l_£X EFFLUENTS
EFFLUENT
SAKPLE
SOLVENT EXTRACTION—
OF EFFLUENT COMPOUNOS
FRACTIONATION OF
EXTRACTED SAfcPLE via HpLC
r
FRACTION WTTH
loflP <3 5:COMPOlfCS
NOT 90CONCENTRATA&£
FRACTIONS WITH
ogPj 3 5
*
I
NO FURT>« ANALYSB
RECOMwCNOED
ANALYZE COMPOUNDS
VIA OC/M3
res
f
SEARCH TARGET LOT MASS spectral lbrary
conTBDFrr .ro*?
CO&POUM5
TOiTATNBLY-
L
NOFimcn
ANALYSS
rn
8 CONCENTRATION« 100
NO |
1
SEARCH »A/KH/I«
MASS SPECTRAL LflRARY:
OBtTOBflT >70%
C0*«0UM> UMOOMN
octbmmbcf
BCf VALUE KNOWK
FR0MQSAR
NOFumcn
ANALYSS
1
BCflNOOWK
EST14ATE FROM
AVBWOE FOR FRACTION
OCTtPfcM OR ESTMATI AAC
AtCftWCBftffC »AAC?
a cOfcPOUM)
TB4TATTVB.Y OBflVB?
NO
2^ t
LOW
COtfflM OOffTTY V* COMP AMBON
WnM KNOMH STAfCAMO CO*#OUM) ON QCA«
ANALYZE FURTVO
TO O04TPY COfcPOlM)
*1
"9 ^ nomttxr
ANALYSi
ORWC »AAC
OEVfLORWLA
OCVB.OP OR WOOTY PBMT LMT
MOMTOR R*C • n WC» AAC N R0MT LMT?—1
J
COfcWJANCE MOMTORMQ
114

-------
Selection of Dischargers for Analysis
The selection of NPDES permittees for bioconcentration analysis is the
responsibility of the NPDES permitting authority. Factors to consider in selecting
dischargers should include:
1.	Dischargers to receiving waters designated for:
a.	Actual or potential commercial or recreational fisheries.
b.	Exceptional resource value waters (e.g., park, wetland, etc.)
2.	Facilities known or suspected to discharge bioconcentratable chemicals,
e.g., organic chemical facilities producing chemicals with high BCFs, etc.
The NPDES permit application form (for direct dischargers), or the list of
industrial users (for POTWs), should be checked for indications of the
presence of known or suspected bioconcentratable pollutants. In addition,
some States may have industrial facility materials inventories, SARA Title
III data, previous effluent monitoring data (e.g., GC/MS scans), etc.,
should also be reviewed.
3.	Ambient monitoring data in conjunction with other effluent
characterization information. Information from sources such as the 304(1)
lists of impaired waters, fish flesh analyses, and sediment studies may
indicate likely dischargers of bioconcentratable pollutants.
EPA recommends that an effluent bioconcentration evaluation be
conducted by the selected permittees every three months for a period of one year.
If no quantifiable bioconcentratable pollutants are identified (i.e., log P > 3.5), no
further bioconcentration analysis is recommended unless a change in process or
discharge occurs.
Results of Effluent Bioconcentration Evaluation
These procedures yield four sets of data on chromatographic components
which should be supplied to the regulatory authority. The lab's report should
contain the following:
1)	Components which were tentatively identified using the CHC mass
spectral library.
2)	Components above 100 ng/L which were tentatively identified using
the EPA/NIH/NBS mass spectral library. (See "Assessment,
Criteria Development and Control of Bioconcentratable
Contaminants in Surface Waters" for more information.)
115

-------
3)	Components that are unidentified with effluent concentrations
greater than 100 ng/L, and with fits less than 70% in each library.
4)	The concentration of each identified and unidentified effluent
component along with the fraction number in which the component
was found. For each tentatively identified and unidentified
component, the concentration in whole effluent is calculated by
comparison to the surrogates added to the mixture initially.
It is important to note that most unidentified chemicals below a
concentration of 100 ng/L are considered by EPA to have relatively low priority
for further identification and may be eliminated from further analysis to conserve
permit authority resources. Alternatively, the regulatory authority may wish to
require or conduct an investigation of a direct discharge permittee or of the
industrial users at a publicly-owned treatment works (POTW) permittee to further
identify these chemicals.
The permit writer should determine a BCF for each tentatively
identified compound using the following procedures:
1)	Use the estimated BCFs from the Quantitative Structure Activity
Relationship (QSAR) data base [36].
2)	For compounds with laboratory-derived BCFs reported in IRIS
(proposed), use the IRIS values.
3)	For compounds with laboratory-derived (measured) and/or
estimated (predicted) log P values (laboratory-derived is preferred),
but unknown BCFs, use the following relationship to estimate the
BCF [37,38]:
log BCF = 0.79 log P - 0.40 - log (7.6/3.0)
4) For compounds with unknown BCF or log P values, use the average
BCF for the fraction in which the compound was found:
Average
Fraction	BCF
1	230
2	1,700
3	49,000
The BCFs derived above are used to calculate AACs and WLAs as
appropriate, following the procedures described in Chapters 2 and 4 respectively.
116

-------
REFERENCES
1. Bergman, H., R. Kimerle, and A. W. Maki (editors). 1985. Environmental
Hazard Assessment of Effluents. Pergamon Press, Inc., Elmsford, New
York.
2a. Kimerle, R., W. Adams, and D. Grothe. 1985. Tiered Assessment of
Effluents. In Environmental Hazard Assessment of Effluents. H. Bergman,
R. Kimerle, and A. Maki (eds.).
2b. Kimerle, R., A. Werner, and W. Adams. 1983. Aquatic Hazard
Evaluation, Principles Applied to the Development of Water Quality
Criteria. In Aquatic Toxicology and Hazard Assessment (7th Symposium).
R. Cardwell and R. Purdy, (eds.)
3.	Macek, K. 1985. Perspectives on the Application of the Hazard
Evaluation Process. In Environmental Hazard Assessment of Effluents.
H. Bergman, R. Kimerle, and A. Maki (eds.).
4.	DiToro, D. 1985. Exposure Assessment for Complex Effluents: Principles
and Possibilities. In Environmental Hazard Assessment of Effluents. H.
Bergman, R. Kimerle, and A. Maki (eds.).
5.	Mount, D., N. Thomas, M. Barbour, T. Norberg., T. Roush, and W.
Brandes. 1984. Effluent and Ambient Toxicity Testing and Instream
Community Response on the Ottawa River, Lima, Ohio. Permits Division,
Washington, D. G, Office of Research and Development, Duluth, MN,
EPA-600/2-84-080, August, 1984.
6.	Mount, D. I., and T. J. Norberg-King (editors). 1985. Validity of Effluent
and Ambient Toxicity Tests for Predicting Biological Impact, Scippo Creek,
Circleville, Ohio. U.S. Environmental Protection Agency,
EPA/600/3-85/044, June, 1985.
7.	Mount, D. et al. (editors). Validity of Effluent and Ambient Toxicity
Tests for predicting Biological Impact, Five Mile Creek, Birmingham,
Alabama. U.S. Environmental Protection Agency, EPA 600/8-85/015,
1985.
8.	Mount, D. I., et al. (editors). Validity of Effluent and Ambient Toxicity
Tests for Predicting Biological Impact, Back River, Baltimore Harbor,
Maryland. U.S. Environmental Protection Agency. EPA 600/8-86/001,
July, 1986.
117

-------
9.	Mount. D. I., et al. (editors). Validity of Effluent and Ambient Toxicity
Tests for Predicting Biological Impact, Naugatuck River, Waterbuiy,
Connecticut. U.S. Environmental Protection Agency. EPA 600/8-86/005,
May, 1986.
10.	Mount, D. I., et al. (editors). Validity of Effluent and Ambient Toxicity
Tests for Predicting Biological Impact, Skeleton Creek. Enid Oklahoma.
U.S. Environmental Protection Agency. EPA 600/8-86/002, March, 1986.
11a. Crane, J.L., A. Pilli:, and R. C. Russa 1984. CETIS: Complex Effluents
Toxicity Information System. Data Encoding Guidelines and Procedures.
Office of Research and Development, Duluth, MN, EPA-60018-84-029.
November, 1984.
lib. Crane, J. L, A. Pilli, and'R. C. Russa. 1984. CETIS: Complex Effluent
Toxicity Information System. CETIS Retrieval System User's Manual.
Office of Research and Development, Duluth, MN, EPA-60018-84-030,
November, 1984.
12.	Mills, W., et al. 1982. Water Quality Assessment: A Screening Procedure
Toxic and Conventional Pollutants. Parts 1 and 2. Office of Research and
Development, Athens, GA EPA 600/6-82-004 A,6 September, 1982.
13.	U.S. Environmental Protection Agency. 1983. The Treatability Manual.
Vol. IV. U.S. EPA Office of Research and Development. EPA
600/2-82-001 (revised January 24, 1983). GPO Stock No. 055-000-00237-1.
14.	U.S. Environmental Protection Agency. 1983. Technical Guidance
Manual for Performing Wasteload Allocations, Book II Streams and
Rivers. U.S. EPA, Office of Water Regulations and Standards,
Washington, D.C.
15.	U.S. Environmental Protection Agency. 1984. Technical Guidance
Manual for Performing Wasteload Allocations, Book III Estuaries. U.S.
EPA, Office of Water Regulations and Standards, Washington, D.C.
16.	U.S. Environmental Protection Agency. 1984. Technical Guidance
Manual for Performing Wasteload Allocations, Book IV. Lakes,
Reservoirs, and Impoundments. U.S. EPA Office of Water Regulations
and Standards, Washington, D.C.
17.	U.S. Environmental Protection Agency. 1982. Test Methods - Technical
Additions to Methods for Chemical Analysis of Water and Wastes. Office
of Research and Development, Cincinnati, OH EPA 600/4-82-055,
December, 1982.
118

-------
18.	U.S.-Environmental Protection Agency. 1989. Generalized Method for
Conducting Industrial Toxicity Reduction Evaluations (TREs). Office of
Research and Development, Cincinnati, OH. EPA 600/2-88/070. March
1989.
19.	U.S. Environmental Protection Agency. 1989. Toxicity Reduction
Evaluation Protocol for Municipal Waste Water Treatment Plants. Office
of Research and Development, Cincinnati, OH. EPA 600/2-88/062. April
1989.
20.	U.S. Environmental Protection Agency. 1989. Methods for Aquatic
Toxicity Identification Evaluations: Phase I Toxicity Characterization
Procedures. National Effluent Toxicity Assessment Center, ERL- Duluth,
MN. EPA 600/3-88/034. September 1988.
21.	U.S. Environmental Protection Agency. 1989. Methods for Aquatic
Toxicity Identification Evaluations: Phase II Toxicity Characterization
Procedures. National Effluent Toxicity Assessment Center, ERL-Duluth,
MN. EPA 600/3-88/035. February 1988.
22.	U.S. Environmental Protection Agency. 1989. Methods for Aquatic
Toxicity Identification Evaluations: Phase III Toxicity Characterization
Procedures. National Effluent Toxicity Assessment Center, ERL- Duluth,
MN. EPA 600/3-88/036. February 1988.
23.	Peltier, W. and C.I. Weber, 1985. Methods for Measuring the Acute
Toxicity of Effluents to Aquatic Organisms, Third Edition. Office of
Research and Development, Cincinnati, OH. EPA-600/4-85-013. (NEW
MANUAL DUE OUT IN NOVEMBER (1989))
24.	Weber, C.I. et. al. (ed.). 1989. Short-Term Methods for Estimating the
Chronic Toxicity of Effluents and Receiving Waters to Freshwater
Organisms, Second Edition. Office of Research and Development,
Cincinnati, OH. EPA-600/4-89/001.
25.	Weber, C.I. et. al. (ed.). 1988. Short-Term Methods for Estimating the
Chronic Toxicity of Effluents and Receiving Waters to Marine and
Estuarine Organisms. Office of Research and Development, Cincinnati,
OH. EPA-600/4-87/02
26.	U.S. Environmental Protection Agency. 1990. Permit Writer's Guide for
Marine and Estuarine Discharges. Office of Water Enforcement and
Permits, Wash., D:C. (IN DRAFT).
119

-------
27.	U.S. Environmental Protection Agency. 1990. Document in draft. Guidance or
Assessment, Criteria Development, and Control of Bioconcentratable
Contaminants in Surface Waters. Office of Water Enforcement and Permits,
Washington, D.C.
28.	Veith, G. D., D. L. DeFoe and B. V. Bergstedt. "Measuring and Estimating the
Bioconcentration Factor of Chemicals in Fish," J, Fish. Res. Board Can., 36:1040-
1048 (1979).
29.	Spacie, A. and J. L. Hamelink. "Alternative Models for Describing the
Bioconcentration of Organics in Fish," Environ. Toxicol, and Chem., 1:309-320
(1982).
30.	Chiou, C. T. "Partition Coefficients of Organic Compounds in Lipid-Water
Systems and Correlations with Fish Bioconcentration Factors," Environ. Sci. and
Technol., 19:57-62 (1985).
31.	Davies, R. P. and A. J. Dobbs. "The Prediction of Bioconcentration in Fish," J. of
Water Res., 18(10):1253-1262 (1984).
32.	Neely, W. B., D. R. Branson and G. E. Blau. "Partition Coefficient to Measure
Bioconcentration Potential of Organic Chemicals in Fish," Environ. Sci. and
Technol., 6(13):1113-1115 (1974).
33.	Burkhard, L. P., D. W. Kuehl and G. D. Veith. "Evaluation of Reverse Phase
Liquid Chromotography/Mass Spectrometry for Estimation of N-Octanol/Water
Partition Coefficients for Organic Chemicals," Chemosphere, 14(10):1551-1560
(1985).
34.	Veith, G. D., N. M. Austin and R. T. Morris. "A Rapid Method for Estimating
Log P for Organic Chemicals," Water Res. J., 13:43-47 (1978).
35.	Alford-Stevens, T. A. Bellar, J. W. Eichelberger, and W. L. Budde. Method 680.
Determination of Pesticides and PCBs in Water and Soil/Sediment by Gas
Chromatography/Mass Spectrometry. Physical and Chemical Methods Branch,
U.S. EPA, Cincinnati, OH. 1985.
36.	User Manual for QSAR System, Institute for Biological and Chemical Process
Analysis, Montana State University, October 1986.
37.	Hansch, C. and A. J. Leo. "Substituent Constants for Correlation Analysis in
Chemistry and Biology." John Wiley and Sons: New York, 1979.
38.	Veith, G. D., K. J. Macek, S. R. Petrocelli, and J. Carroll. "An Evaluation of
Using Partition Coefficients and Water Solubility to Estimate Bioconcentration
Factors for Organic Chemicals in Fish." In Aquatic Toxicology, J. G. Eaton, P. R.
Parish, and A. C. Hendricks, eds. ASTM STP 707, 116-129 (1980).
120

-------
4. EXPOSURE AND WASTELOAD ALLOCATION
4.1 INTRODUCTION
At this point in the toxics control process, a water quality problem has
been identified. Screening analyses may have been done to assess the extent of
toxicity, or the decision may have been made to proceed directly to development
of a wasteload allocation (WLA) based on an existing total maximum daily load
(TMDL). A TMDL is the sum of the individual wasteload allocations (WLAs)
for point sources and load allocations (LAs) for nonpoint sources of pollution and
natural background sources, tributaries, or adjacent segments. WLAs represent
that portion of a TMDL that is established to limit the amount of pollutants from
existing and future point sources so that surface water quality is protected at all
flow conditions.
The TMDL process is one that uses water quality analyses to predict water
quality conditions and pollutant concentrations. Limits on wastewater pollutant
loads are set and controls on nonpoint source loadings are established so that
predicted receiving water concentrations do not exceed water quality criteria.
TMDLs and WLAs/LAs should be established at levels necessary to attain and
maintain the applicable narrative and numerical water quality standard (WQS)
with seasonal variations and a margin of safety which take into account any lack
of knowledge concerning the relationship between point and nonpoint source
loadings and water quality. Determination of WLAs/LAs and TMDLs should
take into account critical conditions for stream flow, loading, and water quality
parameters. Criteria that will protect the receiving water have been determined
from the following sources (Chapter 2.0 and Appendix D):
o For specific toxicants, approved or proposed State standards
(numerical or narrative) have been used. Site-specific criteria may
have been derived with an EPA-approved methodology (USEPA
1983b).
o For effluent toxicity, approved or proposed State standards have
been used. Various species may have been tested for sensitivity to
toxicants, and criteria have been derived using an acute-chronic
ratio (ACR) and species sensitivity factors as appropriate.
o For both specific toxicants and effluent toxicity, appropriate
durations and frequencies for criteria compliance have been
selected.
121

-------
This chapter is divided into sections that explain the steps that precede
establishment of a WLA and then the methods and tools (models) that can be
used to determine the WLA.
Section 4.2 briefly discusses TMDLs and how they relate to waters
identified as requiring a water quality-based approach for toxics controls. The
section also discusses different WLA schemes. Sections 4.3 and 4.4 discuss mixing
zones, areas described as allocated impact zones where chronic water quality
criteria may be exceeded. Section 4.3 provides background information on mixing
zones and discusses EPA's mixing zone policy and how this policy affects the
allowable toxic load that can be discharged from a point source. State mixing
zone dimensions and the determination of mixing zone boundaries are also
discussed.
Section 4.4 discusses mixing zone analyses for situations in which the
discharge does not mix completely with the receiving water within a short
distance. Included in Section 4.4 are discussions of outfall designs that maximize
initial dilution in the mixing zone, critical design periods for mixing zone analyses,
and methods to analyze and model nearfield and farfield mixing.
Section 4.5 discusses the calculations of the WLA and LA process and the
types of EPA-recommended mathematical models that are available to determine
WLAs in completely mixed situations for both aquatic life and human health
(Section 4.6). The WLA models listed in Section 4.5 can be used to predict
exposure and to calculate the effluent quality required to meet the criteria and
protect designated and existing uses of the receiving water. The data
requirements of each of these models are also described so that the effluent
characterization procedures described in Chapter 3 can be designed to support
the specific types of WLA modeling selected by the regulator.
EPA is currently working on methods to develop sediment criteria. In the
future, point source discharges could be further limited to prevent accumulation
of pollutants in the bed sediment resulting in impairment of beneficial uses.
Although the criteria are not yet available for this document, they will be
addressed in future documents. In the meantime, some of the models discussed
in Section 4.5 are capable of simulating interactions between the water column
and sediment as well as toxic transport and transformation in the sediment, and
EPA is encouraging the States to consider the role of sediments in wasteload
allocation modeling.
122

-------
42 TOTAL MAXIMUM DAILY LOAD (TMDL) AND WASTELOAD
ALLOCATIONS (WLA)
4.2.1 TMDLs
The Federal Clean Water Act, under section 303(d), requires the
establishment of total maximum daily loads (TMDLs) for "water quality limited"
stream segments. In such segments, "water quality does not meet applicable
water quality standards, and/or is not expected to meet applicable water quality
standards, even after the application of the technology-based effluent limitations"
(40 CFR Parts 35 and 130, January 11, 1985). A TMDL is the amount of a
pollutant or property of a pollutant from point and nonpoint sources, including a
margin of safety, that may be discharged to a water quality-limited waterbody.
Any loading above this amount risks violating water quality standards. TMDLs
can be expressed in terms of chemical mass per time, toxicity, or other
appropriate measures. Permits should be issued based on TMDLs where
available.
The establishment of a TMDL for a particular waterbody will be
dependent on the location of sources, available dilution, water quality standards,
nonpoint source contributions, background conditions, and in-stream pollutant
reactions and effluent toxicity. All of these factors can affect the allowable mass
of the pollutant in the waterbody. Thus, two issues must be determined in
conjunction with the establishment of the TMDL. These are (1) the definition of
upstream and downstream boundaries of the waterbody for which the TMDL is
being determined and (2) the definition of critical conditions. For the following
discussion, the waterbody boundaries are delineated as the portion of the
waterbody between the most upstream pollutant source (whether point source or
nonpoint source) and the downstream point at which water quality has recovered
to the background quality found above the most upstream pollutant source. The
delineation of critical conditions is specific to the type of waterbody and is
discussed in Section 4.4.
TMDLs are established based on water quality criteria pertinent to the
designated and existing uses for the waterbody in question. TMDLs are
traditionally calculated using State water quality standards as applied to a specific
waterbody. Such a fitting of the TMDL to desired water quality criteria requires
information concerning the distribution of loadings within the waterbody, namely,
the locations and relative contributions of pollutant-specific loadings from point,
nonpoint, and background sources during all flow conditions. Low flow TMDS,
by themselves, will not be adequate in situations where nonpoint source loadings
(LAs) during high or intermediate flow conditions violate water quality standards.
Without this information, the TMDL would be determined based on the best
estimation of the relative proportions of these loading sources.
123

-------
TMDLs have been determined in many ways, but the most common is to
find the pollutant mass' that complies with applicable water quality criteria. For
example, in the Tualatin River Basin in Oregon, TMDLs were determined by
multiplying streamflow in critical flow periods by the pollutant water quality
standard (Oregon Environmental Quality Commission, 1988). Another method of
determining a TMDL is by quantifying instream toxicity. This method was used
in developing a TMDL for the Amelia River in Florida (USEPA, 1987a).
The allowable TMDL is defined as the sum of the individual wasteload
allocations (WLAs) and load allocations (LAs); a margin of safety is included
with the two types of allocations so that any additional loading, regardless of
source, would not produce a violation of water quality standards. The WLAs are
those portions of the TMDL assigned to point sources. The LAs are those
portions of the TMDL assignedto the sum of all nonpoint sources and
background sources. The background sources represent loadings to the specified
waterbody or stream segment that come from sources outside the defined
segment. For example, loadings from regions upstream of the segment and
estimated atmospheric deposition of the pollutant would constitute background
sources. Sediments that are highly contaminated from upstream discharges or
historical discharges may also act as a source of toxicants and contribute to the
background levels.
The TMDL represents a mass loading that may occur over a given time
period so long as receiving stream concentration limits are not exceeded. As a
result, the design flows under which the TMDL is determined can significantly
alter its value. This phenomenon results in a somewhat unusual dichotomy. The
design flows for aquatic life protection most applicable to point source loadings
(wasteload allocations) usually involve low flow events (e.g., 7Q10) because the
volumes associated with the point sources do not decrease, generally, with
decreased streamflow. As a result, the highest concentrations associated with
specific point source loads would be expected under low flow conditions.
Conversely, elevated nonpoint source pollutant loadings (i.e., urban, agricultural)
generally correspond to storm events. In fact, agricultural and urban runoff are
often minimal or nonexistent in the absence of precipitation (i.e., nonexistent
under low flow drought conditions).
Because the TMDL is a composite of the allowable loads associated with
point sources and nonpoint sources within the defined boundaries of the
waterbody segment, and the background loadings to that segment from upstream
and from inplace sediments, the TMDL should be evaluated under conditions
that reflect worst case (critical) conditions for both point source and nonpoint
source loadings (i.e., low flow drought and high flow conditions). Determination
of the TMDL under these two scenarios would result in the estimation of the
lower of the two assimilative capacities of the waterbody in question. This lower
assimilative capacity would represent the TMDL most protective of the waterbody
in question.
124

-------
In the case of design flows for human health protection, the harmonic
mean flow is being recommended as the basis for TMDLs for carcinogens.
Design flows for human health protection should consider worst conditions for
both point and nonpoint source loadings (See Section 4.6).
In many cases, load allocations (LA) for nonpoint sources are difficult to
assess because the information does not exist to describe the runoff associated
with the high flow storm events due to the high variability of the events. Because
of the importance of estimating the nonpoint contributions to the TMDL,
site-specific models may be required to estimate nonpoint source loadings. Even
then, detailed models are difficult to calibrate with accuracy without intensive
monitoring studies, and simplistic correlations between loadings and rainfall can
be, by their statistical nature, unreliable for estimating low frequency events (e.g.,
worst 10-year storm). The uncertainties associated with nonpoint source loadings
and background sources require* that the TMDL be determined with a sufficient
margin of safety to allow for significant variability in nonpoint source loadings.
Section 303(d) of the Clean Water Act and EPA regulations (40 CFR,
Parts 35 and 130, January 11, 1985) require that TMDLs contain a margin of
safety "which takes into account any lack of knowledge concerning the
relationship between effluent limitations and water quality." The margin of safety
is to take into account any uncertainties related to development of the water
quality-based control, including any uncertainties in the water quality criteria,
measurements of pollutant loadings, monitoring of ambient conditions, and water
quality analysis. The size of the required margin of safety can, of course, be
reduced by collecting additional information and thus reducing the amount of
uncertainty. The margin of safety can be provided for in the TMDL process by
one of the following:
1.	Allocating a portion of the TMDL to a separate margin of safety.
2.	Including a margin of safety within the individual WLAs for point
sources and the LAs for nonpoint sources and natural background.
Most TMDLs are developed using the second approach, most often through the
use of conservative design conditions.
In addition, all WLAs, LAs, and TMDLs must meet the State
antidegradation provisions developed pursuant to the Water Quality Standards
Regulation (Section 131.12 of 40 CFR, Part 131, November 8, 1983). This
regulation establishes explicit procedures that must be followed prior to a
decision to lower existing water quality which exceeds that necessary to support
the Section 101(a)(2) "fishable/swimmable" goal of the Act. WLAs, LAs, and
TMDLs cannot be established that allow such a decline in water quality unless
the applicable public participation and intergovernmental review requirements of
the antidegradation and antibacksliding provisions have been met and all existing
uses are fully maintained and protected.
125

-------
42.2 Wasteload Allocation Schemes
WLAs for water quality-based toxics permits must be set in accordance
with EPA regulations (USEPA 1983a; 1985) and should be developed using
EPA's WLA guidance documents. The USEPA Office of Water Regulations and
Standards Assessment and Watershed Protection Division maintains the latest
listing of all WLA guidance documents. Toxic WLA guidance documents are
currently available for rivers and streams (USEPA 1984), lakes and reservoirs
(USEPA 1986a), and estuaries (USEPA 1989). Guidance for the determination
of critical design conditions for steady-state modeling of rivers and streams is also
available (USEPA 1986b).
Table 4-1 lists 19 allocation schemes that may be used by the States.
States are not limited to these approaches and may use any reasonable allocation
scheme that meets the antidegradation provisions and other requirements of the
water quality standards (USEPA 1983a) and water quality management
regulations (USEPA 1985).
The most commonly used allocation methods have been (1) equal percent
removal, (2) equal effluent concentrations, and (3) a hybrid method. The equal
percent removal approach can be characterized in two ways: the overall removal
efficiencies of each pollutant source must be equal or the incremental removal
efficiencies must be equal. The equal effluent concentration approach can also
be applied in two acceptable ways -- equal final concentrations or equal
incremental concentration reductions. This method is similar
to the equal percent removal method if influent concentrations at all sources are
approximately the same. However, if one point source has substantially higher
influent levels, requiring equal effluent concentrations will result in higher overall
treatment levels than the equal percent removal approach.
The final commonly used method of allocating waste loads is a hybrid
method in which the criteria for waste reduction may not be the same for each
point source. One facility may be allowed to operate unchanged, whereas
another may be required to provide the entire load reduction. More often, a
proportionality rule can be assigned that requires the percent removal to be
proportional to the input loading. In these cases, larger sources would be
required to achieve higher overall removals.
4.3 INCOMPLETELY MIXED, DISCHARGE-RECEIVING WATER
SITUATIONS
Mixing zones are areas where an effluent discharge undergoes initial
dilution and are extended to cover the secondary mixing in the ambient
waterbody. The Water Quality Standards Handbook (USEPA 1983b) describes a
mixing zone as an allocated impact zone where chronic water quality criteria can
126

-------
Table 4-1
Wasteload Allocation Methods
1.	Equal percent removal (equal percent treatment)
2.	Equal effluent concentrations
3.	Equal total mass discharge per day
4.	Equal mass discharge per capita per day
5.	Equal reduction of raw load (pounds per day)
6.	Equal ambient mean annual quality (mg/1)
7.	Equal cost per pound of pollutant removed
8.	Equal treatment cost per unit of production
9.	Equal mass discharged per unit of raw material used
10.	Equal mass discharged per unit of production
11a.	Percent removal proportional to raw load per day
lib.	Larger facilities to achieve higher removal rates
12.	Percent removal proportional to community effective income
13a.	Effluent charges (dollars per pound, etc.)
13b.	Effluent charge above some load limit
14.	Seasonal limits based on cost-effectiveness analysis
15.	Minimum total treatment cost
16.	BAT (industry) plus some level for municipal inputs
17.	Assimilative capacity divided to require an "equal effort among all dischargers"
18a.	Municipal: treatment level proportional to plant size
18b.	Industrial: equal percent between BPT and BAT, i.e.,
Allowable = BPT - _x_ (BPT - BAT)
100
19. Industrial discharges given different treatment levels for different stream flows and seasons.
For example, a plant might not be allowed to discharge when stream flow is below a certain
value, but below another value, the plant would be required to use a higher level of
treatment than BPT. Finally, when stream flow is above an upper value, the plant would be
required to treat to a level comparable to BPT.
Source: Chadderton, et al. (1981).
127

-------
be exceeded as long as acutely toxic conditions are prevented and the water
remains free of the following:
o Materials in concentrations that settle to form objectionable
deposits;
o Floating debris, oil, scum, and other matter in concentrations that
form nuisances;
o Substances in concentrations that produce objectionable color, odor,
taste or turbidity; and
o Substances in concentrations that produce undesirable aquatic life
or result in a dominance of nuisance species.
Acutely toxic conditions are defined as those lethal to aquatic organisms
that may pass through the mixing zone. As discussed in Chapter 2, the underlying
assumption for allowing a mixing zone is that a small area of concentrations in
excess of chronic criteria but below acutely toxic releases can exist without
causing adverse effects to the overall waterbody. The State regulatory agency can
decide to allow or deny a mixing zone on a site-specific basis. In order for a
mixing zone to be permitted, the discharger should prove to the State regulatory
agency that all State requirements for a mixing zone are met.
When wastewater is discharged into a waterbody, its transport may be
divided into two stages with distinctive mixing characteristics. Mixing and dilution
in the first stage are determined by the initial momentum and buoyancy of the
discharge. This initial contact with the receiving water is where the concentration
of the effluent will be its greatest in the water column. The design of the
discharge outfall should provide ample momentum to dilute the concentrations in
the immediate contact area as quickly as possible.
The second stage of mixing covers a more extensive area in which the
effect of initial momentum and buoyancy is diminished and the waste is mixed
primarily by ambient turbulence. In large rivers or estuaries, this second stage
mixing area may extend for miles before uniformly mixed conditions are attained.
In some instances, such as larger lakes or coastal bays, completely mixed
conditions are never reached in the waterbody. The general definition for a
completely mixed condition is when no measurable difference in the
concentration of the pollutant (e.g., does not vary by more than five percent)
exists across a transect of the waterbody.
This section provides background information on the policy of mixing
zones and the means to characterize them for use in WLAs (Section 4.5). The
first subsection discusses the concerns that must be addressed when the
boundaries and restrictions of a mixing zone are determined. The second
subsection discusses the guidelines to prevent lethal conditions in the mixing
zone.
128

-------
4.3.1 Determination of Boundaries
The Water Quality Criteria - 1972 (NAS 1972) recommend that mixing
zone characteristics be defined on a case-by-case basis after it has been
determined that the assimilative capacity of the receiving system can safely
accommodate the discharge. This assessment should take into consideration the
physical, chemical, and biological characteristics of the discharge and the
receiving system, the life history and behavior of organisms in the receiving
system, and the desired uses of the waters. Nearly all States require such an
analysis before they allow a mixing zone (USEPA 1988a). Further, mixing zones
should not be permitted where they may endanger critical areas (e.g., drinking
water supplies, recreational areas, breeding grounds, areas with sensitive biota).
The EPA has developed'a holistic approach to determine whether a
mixing zone is tolerable (Brungs 1986). The method considers all the impacts to
the waterbody and all the impacts that the drop in water quality will have on the
surrounding ecosystem and waterbody uses. It is a multistep, data collection and
analysis procedure that is particularly sensitive to overlapping mixing zones. It
includes the identification of all surrounding waterbodies andthe ecological and
cultural data pertaining to them; collection of data on all present and future
discharges to the waterbody; the assessment of relative environmental value and
level of protection needed for the waterbody; and finally, the allocation of
environmental impact for a discharge applicant. Because of the difficulty in
collecting the data necessary for this procedure, and the general lack of
agreement concerning relative values, this method will be difficult to implement
in full. However, the method does serve as a guide on how to proceed in
allocating a mixing zone.
Most States allow mixing zones as a policy issue, but provide spatial
dimensions to limit the areal extent of the mixing zones. The mixing zones are
then allowed (or not) after case-by-case determinations. State regulations dealing
with streams and rivers generally limit mixing zone widths, cross-sectional areas,
and flow volumes and allow lengths to be determined on a case-by-case basis.
For lakes, estuaries, and coastal waters, dimensions are usually specified by
surface area, width, cross-sectional area, and volume.
If a mixing zone is allowed at a specific site, State standards generally
require that chronic criteria be met at the edge of that regulatory mixing zone
during chronic criteria design flow conditions (1) to provide a continuous zone of
passage that meets water quality criteria for free-swimming and drifting organisms
and (2) to prevent impairment of critical resource areas. Individual State mixing
zone dimensions are designed to limit the impact of a mixing zone on the
waterbody and ensure that a zone of passage for free-swimming organisms is
maintained. Furthermore, EPA's review of State WLAs should evaluate whether
assumptions of complete or incomplete mixing are appropriate based on State-
supplied data.
129

-------
In river systems, reservoirs, lakes, estuaries, and coastal waters, zones of
passage are defined as continuous water routes of such volume, area, and quality
as to allow passage of free-swimming and drifting organisms so that no significant
effects are produced on their populations. Transport of a variety of organisms in
river water and by tidal movements in estuaries is biologically important in a
number of ways: food is carried to the sessile filter feeders and other nonmobile
organisms; spatial distribution of organisms and reinforcement of weakened
populations is enhanced; and embryos and larvae of some fish species develop
while drifting (NAS 1972). Anadromous and catadromous species must be able
to reach suitable spawning areas. Their young (and in some cases the adults)
must be assured a return route to their growing and living areas. Many species
make migrations for spawning and other purposes. Barriers or blocks that
prevent or interfere with these types of essential transport and movement can be
created by water with inadequate chemical or physical quality.
As explained above, a State regulatory agency can always decide to deny a
mixing zone in a site-specific case. For example, denial should be considered
when bioaccumulative pollutants are in the discharge. Bioaccumulation in living
organisms is measured by (1) the bioconcentration factor (BCF), which is
chemical-specific and describes the degree to which an organism or tissue can
acquire a higher contaminant concentration than its environment (e.g., surface
water), (2) the duration of exposure, and (3) the concentration of the chemical of
interest. While any BCF value greater than 1 indicates bioaccumulation potential
exists, bioaccumulation potential is generally not considered to be significant
unless than BCF exceeds 100 or more. Thus, a chemical that is discharged to a
receiving stream resulting in low concentrations and having a low BCF value, will
not result in a bioaccumulation hazard through the food chain. Conversely, a
chemical that is discharged to a receiving stream resulting in a low concentration
but having a high BCF value may result in a bioaccumulation hazard. Also, some
chemicals of relatively low toxicity, such as zinc, will bioaccumulate in fish
without harmful effects resulting from human consumption.
Another example of when a regulator should also consider prohibiting a
mixing zone is in situations where an effluent is known to attract biota. In such
cases, provision of a continuous zone of passage around the mixing area will not
serve the purpose of protecting aquatic life. A review of the technical literature
on avoidance/attraction behavior revealed that the majority of toxicants elicited
an avoidance or neutral response at low concentrations (Versar 1984). However,
some chemicals did elicit an attractive response, but the data were not sufficient
to support any predictive methods. Temperature can be an attractive force and
may counter an avoidance response to a pollutant resulting in attraction to the
toxicant discharge. Innate behavior such as migration may also supersede an
avoidance response and cause a fish to incur a significant exposure.
130

-------
4.3.2 Prevention of Lethal Conditions for Aquatic Life
The Water Quality Standards Handbook (EPA 1983) indicates that
whether to establish a mixing zone policy is a matter of State discretion, but that
the State's policy must be consistent with the Clean Water Act and is subject to
approval of the Regional Administrator. The Handbook (EPA 1983) provides
the earliest available basis for and scientific desirability of the proposed mixing
zone policy.
Lethality is a function of the magnitude of pollutant concentrations and
the duration an organism is exposed to those concentrations. Requirements for
wastewater plumes that tend to attract aquatic life should incorporate measures
to reduce the toxicity (e.g., via pre-treatment, dilution, etc.) in order to minimize
lethality or any irreversible toxic effects on aquatic life.
EPA's water quality criteria provide guidance on the magnitude and
duration of pollutant concentrations causing lethality. The criterion maximum
concentration (CMC) is used as a means to prevent lethality or other acute
effects. As explained in Appendix D, the CMC is a toxicity level and should not
be confused with an LC50 level. The CMC is defined as one-half of the final
acute value (FAV) for specific toxicants and 0.3 acute toxicity units (TUa) for
effluent toxicity (see Chapter 2). The CMC describes the condition under which
lethality will not occur if the duration of the exposure to the CMC level is less
than one hour. The CMC for whole-effluent toxicity is intended to prevent
lethality or acute effects in the aquatic biota. The CMC for individual toxicants
prevents acute effects in all but a small percentage of the tested species. Thus,
the areal extent and concentration isopleths of the mixing zone must be such that
the duration of exposure to concentrations in excess of the CMC is kept to less
than one hour. The organism must be able to pass through quickly or flee the
high-concentration area. The objective of mixing zone water quality
recommendations is to provide time-exposure histories which produce negligible
or no measurable effects on populations of critical species in the receiving system.
The Water Quality Criteria - 1972 (NAS 1972) outlines a method to
determine if a mixing zone is tolerable for a free-swimming organism. The
method is a five-step procedure that incorporates mortality rates (based on
toxicity studies for the pollutant of concern and a representative organism) along
with the concentration isopleths of the mixing zone and the length of time the
organism may spend in each isopleth. The final calculation is as follows (NAS
1972):
2 T(n) 	
ET(X) at C(n)
whererT(n) is the time an organism is in isopleth n, and ET(X) is the effective
time of exposure to the concentration C in isopleth n which produces X percent
131

-------
response in a sample of organisms. ET(X) is determined experimentally, and the
measured response is usually mortality. If the sum is less than 1, then the X
percent response will not be realized.
Lethality can be prevented in the mixing zone in one of two ways. The
first method is to prohibit lethal concentrations (values in excess of the CMC) in
the pipe itself, as measured directly at the end of the pipe. As an example, the
CMC should be required in the pipe whenever a continuous discharge is made to
an intermittent stream. In addition, the CMC should be derived using the
assumption of additivity where there is more than one toxicant in the effluent.
The second approach is to require that the CMC be met within a very short
distance from the outfall during chronic design flow conditions for receiving
waters (See Section 4.4.2).
If the second alternative is selected, hydraulic investigations and
calculations indicate that the use of a high velocity discharge with an initial
velocity of three meters per second, or more, together with a mixing zone spatial
limitation of 50 times the discharge length scale, should ensure that the CMC is
met rapidly under practically all conditions. The discharge length scale is
defined as the square-root of the cross-sectional area of any discharge outlet. If
this high velocity discharge is nsi used, the discharger should provide data to the
State regulatory agency. Examples of such data include monitoring studies,
except for those situations where collecting chemical samples to develop
monitoring data would be impractical, such as at deep outfalls in oceans, lakes, or
embayments. Other types of data could include field tracer studies using dye,
current meters, other tracer materials, or detailed analytical calculations, such as
modeling estimations to verify that the dispersion occurs rapidly and close to the
outfall structure. These various types of data will provide information that
demonstrates the most restrictive of the following conditions are met for each
outfall:
o The CMC should be met within 10 percent of the distance from the
edge of the outfall structure to the edge of the regulatory mixing
zone in any spatial direction.
o The CMC should be met within a distance of 50 times the discharge
length scale in any spatial direction. In the case of a multiport
diffuser, this requirement must be met for each port using the
appropriate discharge length scale of that port. This restriction will
ensure a dilution factor of at least 10 within this distance under all
possible circumstances, including situations of severe bottom
interaction, surface interaction, or lateral merging.
o The CMC should be met within a distance of 5 times the local
water depth in any horizontal direction from any discharge outlet.
The local water depth is defined as the natural water depth
(existing prior to the installation of the discharge outlet) prevailing
132

-------
under mixing-zone design conditions (e.g., low flow for rivers). This
restriction will prevent locating the discharge in very shallow
environments or very close to shore, which would result in
significant surface and bottom concentrations.
4.3.3 Prevention of Bioaccumulation Problems for Human Health
States are not required to allow mixing zones. Where unsafe fish tissue
levels or other evidence indicates a lack of assimilative capacity in a particular
water body for a bioaccumulative pollutant, care should be taken in calculating
discharge limits for this pollutant or the additivity of multiple pollutants. In
particular, relaxed discharge limits due to the provision of a mixing zone may not
be appropriate in this situation.
4.4 MIXING ZONE ANALYSES
Proper design of a mixing zone study for a particular waterbody requires
estimation of the distance from the outfall to the point where the effluent mixes
completely with the receiving water. The boundary is usually defined as the
location where the concentrations across a transect of the waterbody differ by less
than five percent. The boundary can be determined based on the results of a
tracer study or the use of mixing zone models. Both procedures, along with
simple order-of-magnitude dilution calculations, are discussed in the following
subsections.
If the distance to complete mixing is insignificant, then mixing zone
modeling is not necessary and the fate and transport models in Section 4.5 can be
used to perform the WLA. It is important to remember that the assumption of
complete mixing is not a conservative assumption for toxic discharges; an
assumption of minimal mixing is the conservative approach. If completely mixed
conditions do not occur within a short distance of the outfall, the WLA study
should rely on mixing-zone monitoring and modeling. Just as in the case of
completely mixed models, mixing-zone analysis can be performed using both
steady-state and dynamic techniques. State requirements regarding the mixing
zone will determine how water quality criteria are used in the TMDL.
This section is divided into five subsections. The first discusses
recommendations for outfall designs and means to maximize initial dilution. The
second provides a brief description of the four major waterbody types and the
critical design period when mixing zone analysis should be performed for each.
The third provides a brief description of tracer studies and how they may be used
to define a mixing zone. The fourth and fifth subsections discuss simplified
methods and sophisticated models to predict the two stages of mixing (i.e.,
discharge-induced and ambient-induced mixing). For a detailed explanation of
the mechanisms involved in estimating both stages of mixing, two references are
recommended, Holley and Jirka (1986) and Fischer et al (1979). Although the
133

-------
models presented in subsections 4.4.4 and 4.4.5 simplify the mixing process, the
assessor should have an understanding of the basic physical concepts governing
mixing in order to use them appropriately. [The USEPA Center for Exposure
Assessment Modeling (CEAM) in Athens, GA provides an overview course that
teaches the basics of mixing and how they should be used for water quality
management.]
It is important to note that the mixing-zone models presented here attempt
to predict the dispersion and dilution of the effluent plume. They do not attempt
to predict any removal or transformation of the pollutants. In the near field,
dispersion and dilution caused by discharge-induced mixing and then
ambient-induced mixing will be the major cause of toxicity reduction. If
incomplete mixing persists downstream (such as in the case of shore-hugging
plumes), then some far-field processes will become important. Some of the
models described in Section 4.5 -which have sophisticated hydrodynamic
simulation routines coupled with fate simulation routines may be used for these
far-field, incomplete mixing analyses.
4.4.1 General Recommendations for Outfall Design
An important factor in maximizing the initial dilution of an effluent is the
design of the effluent outfall. There are three major types of outfall designs:
surface discharge from free flows in a pipe or canal, single port submerged
discharge, and multiport submerged discharge. The last type is often referred to
as multiport diffusers. Of the three, the surface discharge type is the least
favorable for toxic discharges since it offers the least initial mixing. In particular,
surface discharges at the shoreline of a waterbody usually have an impact along
the shoreline when there is significant crossflow, and thus yield high surface
concentrations.
Submerged discharges offer more flexibility in meeting the design goals for
toxic discharges. They may be in the form of a single pipe outlet or of multiport
discharges (diffusers) giving rise to one or several submerged discharge jets. A
typical diffuser section is illustrated in Figure 4-1. Submerged discharges allow
the effluent to be directed at different angles to the ambient flow to maximize
the initial dilution. Diffusers are particularly effective in counteracting the
buoyancy of the effluent. However, submerged multiport discharges are only
feasible in waterbodies that are of sufficient depth and are not subjected to
periodic dredging or considerable scour or deposition.
Many of the complexities of submerged diffusers have been summarized by
Jirka (1982), Holley and Jirka (1986), and Roberts et al. (1989a; 1989b; 1989c).
Submerged discharges should be designed to avoid direct surface impingement
and bottom attachment of the submerged jet or jets. Surface and bottom impacts
should be evaluated at critical design conditions (low flow or high stratification)
and at off-design conditions (higher flow or lower stratification) in order to
ensure the best placement and design of the diffuser. Diffusers provide more
134

-------
Figure 4-1: A typical diffuser section.
(Source: Mstcalf & Eddy, Inc. 1979)
135

-------
dilution than single outlets, but the alignment of the diffuser with the receiving
water flow direction influences how much dilution will be provided. If the outlet
structure is directed parallel to the direction of flow, dilution under high ambient
velocities (off-design conditions) may be lower than under low velocities (critical
design conditions).
In rivers, the preferred arrangement for a submerged discharge is to direct
the outlet into the current flow direction or vertically upward. In order to deal
with the reversing currents of estuaries and coastal bays, the preferred
arrangements for off shore discharges are parallel diffuser alignment (tee
diffuser) or perpendicular diffuser alignment (staged diffuser)(Jirka 1982). In
lakes and reservoirs, the preferred arrangement for a negatively buoyant
discharge is to direct the diffuser vertically upward. A positively buoyant,
vertically directed jet could penetrate stratification, so the preference for this type
of discharge is to orient the diffuser at a slight angle above the horizontal. For
ocean outfalls, initial dilution is improved by longer (perpendicular to the
shoreline) and deeper diffusers. Further, the ports of the diffuser should be
sufficiently separated to minimize merging of the separate plumes (USEPA 1982).
4.4.2 Critical Design Periods for Waterbodies
This section provides a brief description of the four major waterbody types
and defines the critical design periods that should be used when performing
mixing zone analyses in each of these waterbody types. Appendix D provides a
further discussion on the appropriate selection of design periods.
1) Rivers and Run-of-River Reservoirs
Rivers and run-of-river reservoirs are waterbodies that have a persistent
through-flow in the downstream direction and do not exhibit significant natural
density stratification. Recommendations for hydrologically-based and
biologically-based design flows for mixing zone and completely mixed, steady state
modeling of rivers are described in Appendix D of this document. The
biologically-based design flows are determined using the averaging periods and
frequencies specified in water quality criteria (USEPA 1986b). Also, the
hydrologically-based flows 7Q10 and 1Q10 for the CMC and CCC, respectively
have been used traditionally in the past and may continue to be used for steady-
state modeling. Run-of-river reservoirs with residence times less than 20 days at
critical conditions should also be analyzed using biologically- or
hydrologically-based design flows (see below). Regulated rivers may have a
minimum flow in excess of these toxicological flows. In such cases, the minimum
flow should be used in TMDL modeling.
136

-------
2)	Lakes and Reservoirs
This receiving water category encompasses lakes and reservoirs with
residence times in excess of 20 days at critical conditions (Ford and Johnson
1986). Seasonal variations in the water level, wind speed and direction, and
seasonal solar radiation should be determined to define the critical period (Ford
and Johnson 1986). In the case of long and narrow reservoirs, areas above the
plunge point (i.e., the area where no stream-like flow is present and waters are
mixed or stratified by density) can be analyzed as rivers. The areas below can be
analyzed as reservoirs. Since effluent density relative to the ambient water can
vary over seasons, no one season or stratification condition can be selected as the
most critical dilution situation for all cases. In general, all four seasons should be
analyzed to determine the most critical periods for mixing zone analyses. All
seasonal analyses should assume an ambient velocity of zero unless persistent
currents have been documented. Special attention should be given to rising water
level periods since pollutants can move back into coves and accumulate under
these conditions. Location of discharges in coves and dead-end embayments
should be prevented whenever possible.
3)	Estuary and Coastal Bays
This receiving water category encompasses estuaries, which are defined as
having a main channel reversing flow, and coastal bays, which are defined as
having significant two-dimensional flow in the horizontal directions. For both
waterbodies, the critical design conditions recommended here are based on
astronomical, not meteorological, tides.
Determining the nature and extent of the discharge plume is complicated
in marine systems by such conditions as differences in tides, riverine input, wind
intensity and direction, and thermal and saline stratification. Because of the tidal
nature of the estuaries and coastal systems and their complex circulation patterns,
dilution of discharges cannot be determined simply by calculating the discharge
rate and the rate of receiving water flow (e.g., the design flow). For example,
tidal frequency and amplitude vary significantly in different coastal regions of the
U.S. Furthermore, tidal influences at any specific location have daily and
monthly cycles. These and additional factors require that direct, empirical steps
be taken to ensure that basic dilution characteristics of a discharge to saltwater
be determined.
In estuaries without stratification, the critical dilution condition includes a
combination of low-water slack at spring tide for the estuary and design low flow
for riverine inflow. In estuaries with stratification, a site-specific analysis of a
period of minimum stratification and a period of maximum stratification, both at
low-water slack, should be made to evaluate which one results in the lowest
dilution. In general, minimum stratification is associated with low river inflows
137

-------
and large tidal ranges (spring tide), whereas maximum stratification is associated
with high river inflows and low tidal ranges (neap tide).
After either stratified or unstratified estuaries are evaluated at critical
design conditions, an off-design condition should be checked. The off-design
condition (e.g., higher flow or lower stratification) recommended for both cases is
the period of maximum velocity during a tidal cycle. This off-design condition
results in greater dilution than the design condition, but it causes the maximal
extension of the plume. Extension of the plume into critical resource areas may
cause more water quality problems than the high concentration, low dilution
situation.
Recommendations of a critical design for coastal bays are the same as for
stratified estuaries. The period of maximum stratification must be compared with
the period of minimum stratification in order to select the worst case. The off-
design condition of maximum tidal velocity should also be evaluated to predict
the worst-case extent of the plume.
4) Oceans
Critical design periods for ocean analyses are described in two separate
documents, the 301(h) Technical Support Document and the 301(h) document,
Initial Mixing Characteristics of Municipal Ocean Discharges (USEPA 1982;
Mullenhoff et al. 1985). The following subsection contains a summary from these
documents. Like estuaries, discharges to ocean waters will be subject to
two-dimensional horizontal flows. Oceanic critical design periods must include
periods with maximum thermal stratification, or density stratification. These
periods shorten the distance of vertical diffusion that occurs in the zone of initial
dilution. Thus, during these periods it is difficult to achieve the recommended
100 to 1 dilution that is to occur before the plume begins a predominantly
horizontal flow as compared to vertical flow. Periods when discharge
characteristics, oceanographic conditions (spring and neap tides currents), wet and
dry weather periods, biological conditions, or water quality conditions that
indicate that water quality standards are likely to be exceeded should also be
noted. The 10th percentile value from the cumulative frequency of each
parameter should be used to define the period of minimal dilution.
4.4.3 General Recommendations for Tracer Studies
A tracer or dye study can be used to determine the areal extent of mixing
in a waterbody, the boundary where the effluent has completely mixed with the
ambient water, and the dilution that results from the mixing. Analysis of the
mixing zone with a dye study that is supplemented with modeling should be
performed at flow conditions that approach critical flow. Some of those design
conditions are summarized above in the subsections dealing with specific
waterbodies. Once the critical design condition has been selected for a
waterbody, dye studies can be performed to provide data on the dimensions and
138

-------
dilution of the wastewater plume during this critical period. Tracer studies other
than dye studies (e.g., chloride, lithium) can be performed forcases in which the
receiving water is amenable to such a test.
For WLA studies in which a discharge is already in operation, tracer
studies can be used to determine specific concentration isopleths in the mixing
zone which reflect both discharge-induced and ambient-induced mixing. The
isopleth concentrations in conjunction with effluent toxic concentrations should be
superimposed over a map of the various resource zones of the waterbody. The
map will illustrate whether the State's mixing-zone dimensions are exceeded, if
the required zone of passage is provided, and whether the plume avoids critical
resource areas. The WLA can then be calculated to provide the appropriate
zone of passage and to prevent detrimental impacts on spawning grounds,
nurseries, water supply intakes, bathing areas, and other important resource areas.
Obviously, if the outfall is not yet in operation, it is impossible to
determine discharge-induced mixing by tracer studies. Tracer studies can be used
in these situations to determine characteristics of the ambient mixing. For
ambient mixing studies, the tracer release can be either- instantaneous or
continuous. Instantaneous releases are used frequently to measure longitudinal
dispersion, but can also be used to determine lateral mixing in rivers (Holley and
Jirka 1986) and lateral and vertical mixing in estuaries, bays, reservoirs, and lakes.
For waterbodies with significant flow velocities, continuous releases of tracer are
normally used to determine lateral and vertical mixing coefficients. Continuous
releases can also be used to determine three-dimensional concentration isopleths
for steady-state conditions. The tracer study must be made at critical design
conditions in order to use the results directly for WLAs. If a tracer study for
ambient mixing is conducted at near-to-design conditions, the observed data can
be used to determine dimensionless mixing coefficients. These coefficients can
then be extrapolated to critical conditions using hydraulic parameters (Holley and
Jirka 1986). A tracer study at near-to-critical conditions can also be used to
determine the computer model required to predict critical-condition mixing and
provide the coefficients needed for that TMDL model.
A number of references provide information concerning the design,
conduct, and analysis of tracer studies for mixing analyses. The USGS
Techniques for Water-Resources Investigations provide the best overview of how
to conduct tracer studies (USGS 1986; USGS 1985; USGS 1982). The
fluorescent dyes (usually Rhodamine WT), measuring equipment, fluorometers,
field and laboratory procedures, and calculation methods are all discussed. The
procedures essentially consist of adding dye to the waterbody and recording
concentrations of the dye at various stations at specific time intervals. Examples
of tracer studies for river systems are presented in Fischer (1968); Kisel (1964);
Holley and Jirka (1986); and Yotsukura et al. (1970). Examples of tracer studies
in ti
-------
coastal bay in Marin County, California; and Crocker et al. (1989), a study of
Corpus Christi Bay, Texas. Methods to perform a tracer study in a reservoir are
provided in Johnson (1984).
The dye study recommended for obtaining a quick marine/estuarine
dilution assessment is one in which Rhodamine WT dye is administered to a
discharge and monitored in the receiving waters for not less than 24 hours. The
basic goal of this study is to determine the near-field nature of the effluent
dilution, and not steady-state or far-field dilution. The environmental and
discharge conditions selected for the study should be those that would elicit
"worst case" conditions (i.e., highest ambient concentrations in the receiving
water). These include low wind, neap tide (tide of minimum range occurring
during the 1st and 3rd quarters of the moon), plume trapping by density
stratification, low rainfall and low riverine input, and, if possible, high effluent
discharge.
The dye should be administered to the effluent before discharge to the
receiving water in proportion to effluent flow rate. Dye should be maintained at
a concentration in the effluent sufficient to permit detection of the dilution ratio
of interest when amount and variability of background fluorescence in the
receiving water are taken into account. Measurements of dye concentration are
made using a fluorometer and should be corrected for water temperature.
A survey of background fluorescence and its variability in the anticipated
dilution zone must be conducted just prior to the beginning of the study in order
to permit correction of fluorescence data and to determine the dye concentration
required in the effluent. Since Rhodamine WT dye is bleached by free chlorine,
a preliminary study of the degree of dye bleaching by the effluent should precede
the study for chlorinated discharges to avoid underestimation of the extent of the
mixing zone. Dye concentrations should be surveyed for two successive,slack
tides, and for any other conditions that could lead to concentration maxima.
Surveys should extend from the point of discharge to a distance at which the
effluent dilution ratio of interest is attained. The dye fluorescence at this point
should be at least twice the variability in background fluorescence.
EPA has recently completed two TMDL studies to test the procedures
outlined in the previous version of this document. Both studies used dye to
determine the mixing zone and the dilution within it. The first study (USEPA
1987a) was performed on the Amelia River, an estuarine system in Florida, and
the second was on the Greenwich Cove, an embayment of Narragansett Bay in
Rhode Island (USEPA 1988b). In both studies, Rhodamine WT dye was
introduced continuously into the effluent and numerous stations were set up to
measure the spatial and temporal distribution of the dye. Both studies are good
examples of how to perform a dye study in complex tidal systems.
140

-------
Figure 4-2: Example of discharge-induced mixing
(Source: USEPA 1989 Draft)
ncn-vorttcol discharge
d > Shallow vvatsr. low buoyancy
non-vertical discharge
141

-------
4.4.4 Discharge-Induced Mixing
The first stage of mixing is controlled by discharge jet momentum and
buoyancy of the effluent (see Figure 4-2). This stage generally covers most of the
regulatory or near-field mixing zone. It is particularly important in lakes and
reservoirs and slow moving rivers since ambient mixing in those waterbodies is
minimal.
In shallow environments, it is important to determine whether nearfield
instabilities occur. These instabilities, associated with surface and bottom
interaction and localized recirculation cells extending over the entire water depth,
can cause build-up of effluent concentrations by obstructing the effluent jet flow.
There are no simple means to estimate dilution in these cases. Criteria for these
instabilities and specialized predictive models have been developed to address
these problems (Holley and Jirka 1986).
In the absence of nearfield instabilities, horizontal or nearly horizontal
discharges will create a clearly defined jet in the water column that will initially
occupy only a small fraction of the available water depth. The following
equations and models are designed to describe mixing under stable nearfield
conditions.
1) Use of a Simplistic Screening Equation
A minimum estimate of the initial dilution available in the vicinity of a
discharge can be made using the following equation derived from information in
Holley and Jirka (1986):
S= 0.3 jl
d
where
S = flux-averaged dilution;
x = distance from outlet;
d = diameter of outlet.
Coefficient 0.3 represents the average of two values derived from the literature,
0.28 (Fischer et al. 1979) and 0.32 (Albertson et al. 1950).
The equation provides a minimum estimate of mixing because it is based
on the assumptions that outlet velocity is zero and the discharge is neutrally
buoyant. The equation can be used in inverse form to solve for the discharge x
at which a desired solution ~ for example, that corresponding to the CMC - has
been achieved. The equation is valid only close to the discharge, up to a distance
142

-------
corresponding to several (2 to 3) water depths. At longer distances other factors
are of increasing importance in jet mixing and must be included.
Mixing graphs that include the effects of discharge buoyancy, ambient
velocity, and stratification can be found in Holley and Jirka (1986), Fischer et al.
(1979), and Wright (1977). They are useful to account for these other initial
dilution factors and can aid in determining whether criteria will be met at the
edge of the regulatory mixing zone.
2) Use of Detailed Computer Models
More detailed design data for the mixing zone can be obtained from the
use of computer models based on integral jet techniques. It is important to note
that most models represent an idealization of actual field conditions and must be
used with caution to ensure that"the underlying model assumptions hold for the
site-specific situation being modeled. In general these buoyant jet models require
the following input data: discharge depth, effluent flow rates, density of effluent,
density gradients in 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 dilution at each point.
Described below are six mixing zone models thaf are available through the
U.S. EPA. All of the models require a user who is well versed with mixing
concepts and the data necessary to run the models. The first model, CORMIX1 \
(Doneker and Jirka 1990, Akar and Jirka 1990) may be the most useful to
regulators since it is an expert system that guides the user in selecting an
appropriate modeling strategy for rivers or estuaries. It is available from NTIS,
and user support will be provided from the U.S. EPA Center for Exposure
Assessment Modeling in Athens, Georgia, beginning in summer 1990. The other^
models were developed and designed for ocean discharges. All but one can be
used on rivers, lakes, and estuaries with appropriate input modifications;
UPLUME is restricted to stagnant water environments where the ambient water
current velocity is zero (e.g., lakes and reservoirs).
These five models were designed for submerged discharges in oceans.^
They all report dilution and all terminate execution when the vertical ascent of
the plume is zero (e.g., when the plume reaches the surface, or when plume
density is equal ambient density in some stratified systems). With the exception
of CORMIX1, they all assume that there is a "deep" receiving stream (i.e., no
bottom interference). They too are available from NTIS and user support is
provided by the U.S. EPA Hatfield Marine Science Center in Newport, Oregon
(Muellenhoff et al. 1986). These five models have been modified such that the
user inputs the data into a universal data format that allows the user to apply anj
of thrive models with only minor input changes.
143

-------
CORMIX is a series of software elements for the analysis and
design of a submerged buoyant or nonbuoyant discharge containing
conventional or toxic pollutants and entering into stratified or
unstratified watercourses, with emphasis on the geometry and
dilution characteristics of the initial mixing zone. Subsystem
CORMIX1 deals with single port discharges, and subsystem
CORMIX2 addresses multiport diffusers. The system is
implemented on microcomputers with the MS-DOS operating
system. CORMIX1 can summarize dilution characteristics of the
proposed design, flag undesirable designs, give dilution
characteristics at specified boundaries (i.e., legal and toxic mixing
zones) and recommend design alterations to improve dilution
characteristics. The CORMIX 1 program guides the user, based on
the user's input, to appropriate analyses of design conditions and
mixing zone dimensions.
UPLUME is an initial dilution model that can be used for stagnant
waterbodies such as lakes and reservoirs where the ambient currents
can be assumed to be zero. The model simulates a submerged
single port discharge. The bouyancy between the effluent and
ambient water can be accounted for and the discharge can be given
a vertical angle. UPLUME calculates flux-averaged dilutions and,
for one output option, a centerline dilution.
UOUTPLM can be used in flowing and stagnant waterbodies. The
user specifies the current speed of the ambient water and this speed
is assumed to be constant with depth. The model simulates a single
submerged port discharge. Buoyancy between the effluent and
ambient water can be modeled as well as the discharge vertical
angle. The ambient current is assumed to be perpendicular to the
diffuser.
UMERGE is a model that can also be used for both flowing and
stagnant waters. It has capabilities that UOUTPLM does not have:
it considers multiple submerged ports, and the user can specify
arbitrary amibienf currenf speed^variations with depth. The ports
are assumed to be equally spaced. The model accounts for adjacent
plume interferences over the course of the plume trajectory and in
the subsequent dilution calculation. Positive buoyancy is accounted
for, and the discharge vertical angle can be modified. The ambient
current is assumed to be perpendicular to the diffuser.
UDKHDEN is a three-dimensional model that can be used for
flowing and stagnant waterbodies. It has all the capabilities of
UMERGE plus the ability to simulate instances where the ambient
current flow is not perpendicular to the diffuser.
144

-------
o ULINE models a vertical slot jet discharge into a flowing
waterbody. The discharge angle is assumed to be. perpendicular to
ambient current. The ambient current may vary with depth and the
axis of the diffuser may range rfrom parallel to perpendicular to the
ambient current. The buoyancy of the effluent can also be
modeled.
An evaluation and comparison of all these models can be found as part of the
Technical Guidance Manual on Waste Load Allocations in Estuaries (USEPA
1989, Draft).
4.4.5 Ambient-Induced Mixing
The equations for discharge-induced mixing can be used to predict
concentrations in the regulatory-mixing zone where strong jet mixing
predominates over ambient mixing. Beyond this point, the mixing is controlled by
ambient turbulence. Thus, ambient mixing models must be used to predict the
pollutant concentration distributions up to the stage of complete lateral mixing to
provide boundary conditions for the completely mixed fate and transport models
described in Section 4.5. This information may also be needed to estimate
concentrations encountered at important resource areas or at subsequent
downstream dischargers.
If there is no discharge-induced vertical mixing associated with the jet
action of the discharge, then mixing over the depth of the waterbody must be
accomplished by ambient mixing. For a neutrally buoyant, soluble effluent
discharged with low velocity at the surface or at the bed of a stream, the flow
distance required to achieve complete vertical mixing is on the order of 50 to 100
times the depth of water in that portion of the channel where the effluent is
discharged (Yotsukura and Sayre 1976). For a discharge that is either lighter
(positively buoyant) or heavier (negatively buoyant) than the ambient water but
still has no excess momentum, the flow distance for mixing over the depth will be
greater. In the normal case with a high-velocity jet designed to prevent lethality
in the mixing zone, mixing over the depth will be accomplished primarily by jet
action, and the distance required for this vertical mixing will be much shorter.
In general, ambient mixing must also accomplish mixing over the width of
a waterbody to bring the effluent to the completely mixed condition. For
situations where the width of the zone that is mixed by the discharge-induced
mixing is much smaller than the width of the river, the flow distance (X„,)
required to achieve the completely mixed condition may be estimated from an
equation of the form (Fischer et al. 1979):
Xn, = mW\i
Dy
145

-------
where
m
u
W
Dy
width of the river,
flow velocity for the critical design flow,
lateral dispersion coefficient as discussed below, and
a parameter whose value depends on the degree of
uniformity used to define "complete mixing" and on the
transverse location of the outfall in the stream.
If completely mixed conditions are defined as a 5 percent variation in
concentration across the stream width, the value of m would be 0.3 to 0.4 for a
discharge near the center of river flow (not the center of river width) and 0.4 to
0.5 for a discharge near the edge of the river. If, because of other uncertainties,
a 25 percent variation across the width is accepted as being completely mixed,
then the corresponding ranges of values for m for the two locations would be 0.1
to 0.2 and 0.2 to 0.3. For a very small stream, Xm may be only a few hundred
feet; for medium and large streams, Xn, is normally several miles to several tens
of miles.
The lateral dispersion coefficient (Dy) for most rivers can be calculated
with the following equation (Fischer et al. 1979):
where d = water depth at design flow and u* = shear velocity. The coefficient
(0.6) can vary from 0.3 to above 1.0 depending on the type and degree of
irregularity of the channel cross- sections. The more straight and uniform the
flow, the lower the value; the more irregular the flow (resulting from curves,
sidewall interference, etc.), the higher the value. Values approaching and
exceeding 1.0 are normally associated with significant channel meandering
(Yotsukura and Sayre 1976). The following equation for shear velocity should be
used (Fischer et al. 1979):
where
g = acceleration due to gravity,
s - slope of the channel, and
d = water depth.
For diffusers that initially spread the discharge across a significant part of
the river width or for cases where the discharge-induced mixing causes mixing
across a significant part of the river width, the values of m and X„, can be smaller
than the ones indicated here. For distances greater than X„» the models for
completely mixed effluents discussed in Section 4.5 can be used to calculate
concentrations at these distances. For shorter distances, maximum concentrations
Dy = 0.6 du* ± 50%
u* = (gds)1/2
146

-------
can be much greater than those predicted by "completely mixed" models and
should be estimated using the following equation:
Cx
CelkW	,
Qs(tt Dy X/u)1/2
where
Cx
Ce
Qe
Qs
Dy
X
w
u
maximum pollutant concentration distance x from the outlet,
effluent concentration,
effluent flow,
design stream flow,
lateral dispersion coefficient,
distance from the outlet,
stream width, and
flow velocity for the design flow.
It should be noted that this estimate of C*, is a worst-case prediction since the
equations assume no significant discharge-induced mixing and a neutrally buoyant
effluent. A more accurate way to predict concentrations within this second stage
of mixing is to use the methods of Yotsukura and Sayre (1976). In order to use
this approach, however, the value of Dy and pollutant concentrations after
discharge-induced mixing must be known from tracer studies and/or from the use
of one of the discharged-induced models.
The PSY model can be used to predict ambient mixing in shallow,
freshwater streams where water depth is small in proportion to the width. PSY is
a steady-state two-dimensional plume model that predicts dilution of a surface
discharge into a shallow receiving water where the plume attaches to both bottom
and nearshore (Pailey and Sayre 1978). Uniform vertical mixing is assumed to
occur at the point of discharge.
Ambient mixing is minor for lakes and reservoirs because flow velocity is
assumed to be minimal and mixing is accomplished from the discharge
momentum and buoyancy. For estuaries that are completely mixed with regard to
salinity, the equations presented above can be used to estimate concentrations
between the outlet and the point of complete mixing with a slight modification of
shear velocity. The above equations will be applicable to only unstratified
estuaries, since the time required to mix across the estuary must be significantly
less than: (1) the time required for the effluent to pass out of the unstratified
part of the estuary, (2) the time required for the effluent to pass into a segment
of greatly changed cross-section, or (3) the time required for the substance to
decay. When the above equations for estuaries are used, the velocity of the
design flow should include the velocity associated with the inflow of freshwater as
well as tidal velocity, thus ut, which is based on an average total velocity is
substituted for u in the equations and shear velocity becomes:
147

-------
u* = 0.10 ut.
The CORMIX expert system model can also be used to obtain predictions
or the ambient-induced mixing. In addition to the routines for discharge-induced
mixing, this model also includes predictive elements that apply to ambient mixing
in riverine, lake, or coastal situations.
4.5 COMPLETELY MIXED DISCHARGE-RECEIVING WATER
SITUATIONS
At the present time, most States and EPA Regions use steady-state models
that assume the wastewater is completely mixed with the receiving waters in order
to calculate WLAs for contaminants. This approach is appropriate for
conventional contaminants where critical environmental effects are expected to
occur far downstream from the source. WLAs for toxic chemicals require a
different approach, however, because critical environmental conditions occur near
the discharge before complete mixing with the receiving water occurs.
Consequently, mixing analyses should be performed because many of these
toxicants can exert maximal toxicity in a variety of regions spanning from the
discharge point to significant distances downstream.
If complete mixing occurs near the discharge point, such as in effluent-
dominated receiving streams, then steady-state models may be used to calculate
TMDLs. Recent EPA developments in the identification of critical design flows
based on toxicological concerns provide for better use of steady state models in
toxic WLAs. However, if complete mixing does not occur near the discharge
point and the effluent plume is discernable downriver, then modeling techniques
that can simulate and predict mixing conditions are more appropriate. The
mixing zone models presented in the previous section may be used to define the
mixing zone. However, they only determine the dispersion and dilution of the
effluent and do not account for chemical or biological processes in the mixing
zone. TMDL models are available that can simulate mixing processes and
predict areas of maximal concentrations in the receiving stream based on
chemical, biological, and physical processes.
4.5.1 Wasteload Modeling Techniques
1) Steady-State Modeling Techniques
A steady-state model requires single, constant inputs for effluent flow,
effluent concentration, background receiving water concentration, receiving water
flow, and meteorological conditions (e.g., temperature). The frequency and
duration of ambient concentrations predicted with a steady-state model must be
assumed to equal the frequency and duration of the critical receiving water
conditions used in the model. The variability in effluent flows and concentrations
148

-------
also affects receiving water concentrations, but these effects cannot be predicted
with constant inputs. Steady-state models can be improved for toxic WLAs by:
o using design flows that will ensure criteria compliance
at the appropriate duration and frequency, and
o calculating both acute and chronic WLAs.
EPA is encouraging the States to adopt two-number water quality criteria
and is using them in WLA studies. Ambient water quality criteria have been
established for numerous toxic pollutants. These criteria specify an acute
concentration (CMC) and a chronic concentration (CCC) for each toxicant, as
well as maximum tolerable durations and frequencies of exposure for the two
concentration levels. The design flows used in steady-state modeling should be
reflective of the CCC and CMC duration and frequencies. The duration of the
design flow is based on the maximum exposure time that will prevent acute and
chronic effects. The duration of flow is assumed to apply to the duration of the
allowable effluent concentration or load. For example, if the flow used is a 7-day
average value, the allowable load is considered to be a 7-day average. The return
period or frequency of the flow is based on the number of years required for
biological population recovery after criteria have been exceeded.
Appendix D describes the toxicological basis for selecting receiving stream
design flows for steady-state modeling and recommends specific design flows for
CCC and CMC calculation of TMDLs for rivers and streams. In summary, there
are two types of design flows, hydrologically based and biologically based. The
hydrologically based design flows are those traditionally used by the States where
the 7Q10 flow is used as the CCC design flow and the 1Q10 is used as the CMC
design flow. The biologically based method uses the 1-day 3-year
duration-frequency for determining the CMC design flow and the 4-day 3-year
duration-frequency for determining the CCC design flow. Consequently, the
biologically based design flows are based on specific toxicological effects of a
pollutant and biological recovery times from localized stresses (USEPA 1986b).
The advantages of both types, as well as how they may be calculated, are also
described in Appendix D.
At the present time, there are no recommended toxicological flows for
steady-state modeling of lakes, reservoirs, or estuaries. The critical conditions
recommended for these waterbodies in Section 4.4.2 are based on hydrological
and meteorological conditions rather than on toxicological duration and frequency
data. These conditions should be used until further guidance is provided.
Another improvement in steady-state toxics modeling can be realized by
performing two separate WLAs, one for the CMC and one for the CCC. Steady-
state^ WLA models should be used to calculate the allowable effluent load that
will meet the CMC at the acute design flow and the allowable load that will meet
the CCC at the chronic design flow. Calculation of these values will enable the
149

-------
permit writer to calculate the more limiting long-term average for the treatment
system and develop permit limits protective of both WLAs (see Chapter 5).
In addition to stream design flow, steady state models require design
temperature, pH, alkalinity, and hardness, depending on the pollutants modeled
at site-specific conditions. To determine stream design temperature, pH,
alkalinity, and hardness, a program called DESCON was developed. DESCON is
a computer program that estimates design conditions for WLA modeling. These
conditions are based on maintaining a desired limit on the frequency of water
quality excursions in a receiving water. DESCON considers the effect that daily
fluctuations in streamflow and water quality conditions, such as temperature and
pH, have on the variability of the capability of a receiving water to accept
pollutant loadings. It specifically accounts for the within-year correlations
observed between such variables as streamflow, temperature, pH, alkalinity,
hardness, and dissolved oxygen. DESCON determines design conditions using a
four step process (see Figure 4-3):
1)	A long-term record of observed stream flows and
pertinent water quality data are assembled or
synthesized.
2)	The maximum allowable pollutant load that the
receiving water can accept without causing a water
quality excursion is computed for each day of this
record.
3)	This synthesized record of allowable loads is searched
for the critical load, i.e., the load whose frequency of
not being exceeded matches the desired water quality
excursion frequency.
4)	Design conditions are then derived from receiving
water conditions realized during the period of record
when the computed allowable load was closest to the
critical load.
DESCON provides the same advantages as continuous simulation by considering
the joint occurrences of streamflow and other water quality parameters as
observed in the historical record. In addition, it is more computationally
efficient; it contains a facility for extracting and analyzing flow and water quality
data from STORET; it can analyze both the extreme value and the biologically-
based definitions of water quality excursions; and it is specifically designed to
handle such pollutants as ammonia, heavy metals, pentachlorophenol, and
biochemical oxygen demand whose water quality criteria are functions of such
design condition variables as temperature, pH, alkalinity, hardness, and dissolved
oxygen. The main limitations of DESCON are that it requires at least ten years
of historical daily flow data and it can only analyze a single discharger,
150

-------
©
©
SuumFloM
Ttmpatiturt
DAYS
Allowable Sltaam Uadiq
T^__
Critical
Load
DAYS
S tit am flow
Onitn Flaw
Anoxribl* Suiam j
Loading	j
©
| ©
Critical Load
Dititi Twnpuitm
Trial Cfilicrf LomI
Final
Critical
Load
I
| Allowed
I Mtmibar
NO. OF EXCURSIONS
DAYS
Critical Emm
Figure 4-3: Computational scheme for deriving design conditions.

-------
edge-of-mixing zone situations (or a simplified Streeter-Phelps dissolved oxygen
response for BOD).
2) Dynamic Modeling Techniques
Steady-state modeling considers only a single condition; effluent flow and
loading are assumed to be constant. The impact of receiving water flow
variability on the duration for which and frequency with which criteria are
exceeded is implicitly included in the design conditions if these conditions reflect
the desired toxicological effects regime. Dynamic modeling techniques explicitly
predict the effects of receiving water and effluent flow and concentration
variability. The three dynamic modeling techniques recommended by EPA for
wasteload allocations are continuous simulation, Monte Carlo simulation, and
lognormal probability modeling. These methods calculate a probability
distribution for receiving water concentrations rather than a single, worst-case
concentration based on critical conditions. Prediction of complete probability
distributions allows the risk inherent in alternative treatment strategies to be
directly quantified.
The use of probability distributions in place of worst-case conditions has
been accepted practice for years in water resource engineering, where it was
found to produce more cost-effective design of bridge openings, channel
capacities, floodplain zoning, and water supply systems. The same cost-
effectiveness can be realized for pollution controls if probability analyses are
used.
The dynamic modeling techniques have an additional advantage over
steady-state modeling in that they determine the entire effluent concentration
frequency distribution required to produce the desired frequency of criteria
compliance. Maximum and monthly average permit limits can be obtained
directly from this distribution. Generally, steady-state modeling has been used to
calculate only a chronic WLA. This generates a single allowable effluent value
and no information about effluent variability. If the steady-state model is used to
calculate both acute and chronic wasteloads, limited information will be provided,
and the entire effluent distribution will not be predicted. Steady-state WLA
values can be more difficult to use in permits and enforcement because of the
variable nature of the receiving waterbody and the effluent. The outcome of
probabilistic modeling can be used to ensure that permit limits are determined
based on best probability estimates of receiving water concentrations rather than
a single, worst-case condition. As a result, maximum and monthly average permit
limits, based on compliance with water quality criteria over a 3-year period, can
be obtained directly from the probability distribution.
Continuous Simulation Models. As shown in Figure 4-4, a continuous
simulation model uses daily effluent flows (Qe) and concentration data (Ce) in
conjunction with daily receiving water flow (Qs) and background concentration
data (Cs) to calculate downstream receiving water concentrations (Donigan and
152

-------
Waggy 1974). The model predicts these concentrations in chronological order
with the same time sequence as the input variables (Cb vs. time). The daily
receiving water concentrations can then be ranked from the lowest to the highest
without regard to time sequence. A probability plot can be constructed from
these ranked values, and the occurrence frequency of any 1-day concentration of
interest can be determined (Cb vs. frequency). Running average concentrations
for 4 days (i.e., the chronic design flow), or for any other averaging period, can
also be computed from the daily concentrations (Figure 4-5).
The probability plot generated by the continuous simulation model using
existing effluent data will indicate whether criteria are predicted to be exceeded
more frequently than desired. Appendix D discusses how to select the
appropriate occurrence frequency (or return period) based on the biological
recovery period required for a specific waterbody. If recurrence intervals of 10 or
20 years are desired, at least 30*years of flow data should be available to provide
a sufficient record to estimate the probability of such rare events. Of the 30
years of required flow data, at least 20 to 25 years should be continuous daily
data, with the remaining years represented with only intermittent data. The data
should be examined to verify that the receiving stream has not undergone
significant hydrological modification. The same data requirements are also true
for the lognormal probabilistic and Monte Carlo methods. However, except for
the continuous simulation models, other non-steady state models in this section
cannot be used to account for the duration and frequency provision of the two-
number WQC. Users are cautioned about the specific limitations of some of the
dynamic models included here.
Continuous simulation models have the following advantages compared to
steady-state formulations:
o the frequency and duration of toxicant concentrations in a receiving
water can be predicted;
o the cross-correlation and interaction of time-varying pH, flow,
temperature, pollutant discharges, and other parameters are
incorporated;
o the effect that the serial correlation of daily flows and other
parameters has on the persistence of criteria violations is
incorporated;
o long-term stream flow records for ungauged rivers using
precipitation and evapotranspiration data can be synthesized; and
o long simulation times can prevent the initial conditions used in the
model from affecting the calibration of fate and transport processes.
153

-------
Paream of Tim® Coneentneen it Lmb Than or Eou*
OS 1 2 91042
	1	I 1	U4I
R«eur«nea toawt (y«ar«l	20
Figure 4-4 • Continuous simulation modeling schematic.
154

-------
Piguri 4-5. Conctntration Fr«qu«ncy Cutvm.
"•LNaJuM-
«• L


®
^ U* ~Vk* "
1
r	Si
155

-------
Unlike steady-state models, continuous simulation models require significantly
more data to apply, calibrate, and/or verify a specific problem and require that
input information for the application of the model be time-series data.
Monte Carlo Simulation Models. Monte Carlo simulation combines
probabilistic and deterministic analyses since it uses a fate and transport
mathematical model with statistically described inputs. Monte Carlo simulations
have been the most frequently used approach in stochastic water quality studies
(Tiwari and Hobbie 1976, Malone et al. 1979, Hornberger and Spear 1980, Ford
et al. 1981, Scavia et al. 1981, Rose et al. 1989, Versar 1989). The probability
distributions of effluent flow, effluent concentration, and other model input must
be defined using the appropriate duration for comparison to the CMC and CCC.
If 1-day average receiving water concentrations must be predicted for CMC
comparisons, probability distributions of daily model input data are needed for
Monte Carlo simulation. If 4-d'ay average concentrations must be predicted for
CCC comparisons, the probability distributions of 4-day average input data are
required. The computer selects input values from these distributions using a
random generating function. The fate and transport model is repetitively run for
a large number of randomly selected input data sets. The result is a simulated
sequence of receiving water concentrations. These concentrations do not follow
the temporal sequence that is calculated with the continuous simulation model,
but they can be ranked in order of magnitude and used to form a frequency
distribution. Monte Carlo analyses can be used with steady-state or continuous
simulation models (Thornton et al. 1982).
The approach for calculating the allowable effluent concentration
distribution using Monte Carlo simulation is the same as that described for the
continuous simulation model. The advantages of Monte Carlo simulation are the
following:
o it can predict the frequency and duration of toxicant concentrations
in a receiving water;
o it can be used with steady-state or continuous simulation models
that include fate processes for specific pollutants;
o it can be used with steady-state or continuous simulation models
that include transport processes for rivers, lakes, and estuaries;
o it can be used with steady-state or continuous simulation models
that are designed for single or multiple pollutant source analyses;
o it does not require time series data;
o it does not require model input data to follow a specific statistical
distribution or function; and
156

-------
o it can incorporate the cross-correlation and interaction of
time-varying pH, flow, temperature, pollutant discharges, and other
parameters if the analysis is developed separately for each season
and the results are combined.
The primary disadvantage of Monte Carlo simulation is that it requires more
input, calibration, and verification data than do steady-state models.
Lognormal Probabilistic Dilution Model. Without resorting to the
continuous simulation method of computing receiving water concentrations in
temporal sequence, this probabilistic method uses the lognormal probability
distributions of the input variables to calculate probability distributions of output
variables (DiToro 1984). As a result, the method requires only the relevant
statistical parameters of the input variables (medians and coefficients of variation)
rather than the actual time series data needed for continuous simulation. If 1-day
average receiving water concentrations must be predicted for comparisons with
the CMC, lognormal probability distributions of daily input data are needed. If
4-day average concentrations must be predicted, the lognormal probability
distributions of 4-day average input data are required. Because this probabilistic
model cannot, as yet, incorporate fate and transport processes, it can be used to
predict the concentration of a substance only after complete mixing and before
degradation or transformation significantly alters the concentration.
The lognormal probabilistic dilution model has the following advantages:
o it can predict the frequency and duration of toxicant concentrations
in riverine environments;
o it does not require time series data; and
o it can incorporate the cross-correlation and interaction of
time-varying pH, flow, temperature, pollutant discharges, and other
parameters if the analysis is developed separately for each season
and the results combined.
The lognormal probability dilution model has the following disadvantages:
o it requires more input than a steady-state model;
o it does not include instream fate processes;
o it applies only to rivers and streams;
o it analyzes multiple pollutant sources inaccurately; and
o it requires model input data to be lognormally distributed.
157

-------
4.5.2 Calculating the Allowable Effluent Concentration Distribution and the
Return Period
Information concerning effluent concentration means and variabilities can
be obtained from databases on existing treatment plants and from Development
Documents for specific industrial point source categories. This information is
available from the Industrial Technology Division of the Office of Water
Regulations and Standards. These effluent data can be used with dynamic
models to determine what the effluent concentration distribution must be to meet
water quality standards. Two possible approaches can be taken to determine this
distribution regardless of the type of dynamic modeling technique (i.e.,
continuous, Monte Carlo, or lognormal probabilistic). One approach is based on
the simplifying assumption that treatment will change only the magnitude of
effluent concentrations. No changes are assumed to occur in effluent flows or in
the relative variability of effluent concentrations. With these assumptions, no
additional model runs are needed to determine the allowable distribution for
effluent concentrations. The other approach assumes that the required effluent
concentration distribution is the same as the existing distribution except that it is
reduced in magnitude by whichever is greater ~ the percentage necessary for the
1-day average concentrations to meet the CMC or the 4-day average
concentrations to meet the CCC at the desired recurrence interval. Chapter 5
includes details on how permit limits are derived from the mean and coefficient
of variation of effluent concentrations determined from this analysis.
The second approach for determining the allowable effluent concentration
distribution is based on the assumption that effluent concentrations after
treatment will not have the same coefficient of variation as concentrations before
treatment. Studies have documented that advanced secondary treatment
increases the coefficient of variation of BOD and TSS concentrations compared
to secondary treatment. Where feasible, investigations should be conducted to
evaluate how treatment processes for heavy metals, organic chemicals, and
effluent toxicity will change the variability of these constituents. The
Development Documents mentioned above also provide some variability data for
treatment processes. To account for a change in variability, an alternative
approach should be used to determine the allowable effluent distribution.
Iterative model runs can be performed using different concentration means with
the effluent "future treatment" variance until a mean is found that meets the
criteria at the desired recurrence intervals. These iterative model runs require
stochastic generation of effluent input data since daily effluent concentrations will
not be available for the hypothetical treatment schemes. The required "future
treatment" mean and coefficient of variation of effluent concentration can then be
used to set permit limits (see Chapter 5).
EPA's Office of Water Regulations and Standards developed an interactive
preprocessor for DYNTOX that (1) automatically creates input for continuous
simulation models, (2) randomly selects the sets of input data required for Monte
Carlo simulations, and (3) performs the numerical integration calculation for the
158

-------
lognormal probabilistic model. DYNTOX is available from the EPA Center for
Exposure Assessment Modeling at the Athens, Georgia, Environmental Research
Laboratory (ERL) (LimnoTech 1985). If the observed database is fairly complete
but missing a few points, a linear interpolation scheme is used to fill in the
missing data. If data are scarce, a lag-one Markov method is used to generate
daily data stochastically. The lag-one Markov method uses the mean, standard
deviation, and daily correlation coefficient of the observed data to create random
sequences of data having the same statistical properties. The interactive program
is written in FORTRAN and is available for use on mainframe or IBM
PC-compatible computers.
Two common methods exist to calculate the return period for a given
concentration from probabilistic modeling: the percentile method and the
extrema method. The percentile method used by DYNTOX ranks a listing of all
individual daily concentrations* The return period for a concentration is then
calculated based on the percentile occurrence. In the extrema method, only
annual extrema values are used in the ranking. The return periods calculated
from these two methods are equally valid statistical representations. When using
the percentile method, results express an average return period and multiple
occurrences within any year. The extrema method describes the return period for
an annual extreme and includes only the extreme of multiple occurrences within a
year.
4.5.3 General Recommendations for Model Selection
The reliability of the predictions from any of the modeling techniques
depends on the accuracy of the data used in the analysis. The minimum data
required for model input includes receiving water flow/volume, effluent flow,
effluent concentrations, and background concentrations. In many locations,
streamflow data should be sufficient for both steady-state and dynamic models.
At least 30 years of flow data should be available if violations of the CMC and
CCC must be evaluated at rare recurrence intervals of once in 10 or 20 years.
Measurements of effluent toxicity or individual toxicity can be much more limited.
If only a few toxicant or effluent toxicity measurements are available,
steady-state assessments should be used. Modeling should also be limited to
steady-state procedures if a daily receiving water flow record is not available;
however, in effluent-dominated situations, critical flow may be used to
characterize the receiving stream. Appendix D describes how to select
appropriate design flows if State regulations do not require a specific design flow
for river WLAs. Fate and transport models or dilution calculations can be used
for individual toxicants. At the present time, only dilution calculations or first-
order decay equations are recommended for effluent toxicity analyses. Chapter 1
discusses the conservative/additive assumption for toxicity.
If adequate receiving water flow and effluent concentration data are
available to estimate frequency distributions, one of the dynamic modeling
159

-------
techniques should be used to develop more cost-effective treatment requirements.
If the effluent data exhibit significant seasonal differences or batch process trends,
the continuous simulation approach may be the easiest dynamic modeling method
to use. The best results will, of course, be obtained if daily effluent flows and
concentrations are available for model input for an entire year. The lag-one
Markov technique can be used to generate daily effluent data for the entire
simulation as long as adequate measurements for the site-specific facility (or a
similar one) are available to estimate a day-to-day correlation coefficient and to
determine when seasonal or batch process changes in effluent quality occur.
If adequate receiving water flow and effluent concentration data are
available and if effluent data exhibit no seasonal or batch process trends,
lognormal and Monte Carlo methods may be easier and require less computer
time than the continuous simulation approach.
4.5.4 Specific Model Recommendations
The following section recommends models for toxicity and individual
toxicants for each type of receiving water -- rivers, lakes, and estuaries. Detailed
guidelines on the use of fate and transport models of individual toxicants are
included in the toxic TMDL guidance available from the Monitoring Branch of
EPA's Office of Water Regulations and Standards (USEPA 1984, USEPA 1986a,
USEPA 1989) and from the Office of Research and Development (USEPA
1987b). These manuals describe in detail the transport and transformation
processes involved in water quality modeling. Transport processes include the
dispersion and advection of a contaminant once it enters the receiving stream; its
volatilization from the water; and its sorption to suspended sediment, eventual
settling, and possible resuspension and diffusion from the sediment.
Transformation processes include the oxidation, hydrolysis, photolysis,
biodegradation, and bioaccumulation of the chemical.
Most water quality models were developed with an emphasis on the
dynamics in the water column and the eventual water column concentrations.
Several models, including some of those listed below (EXAMSII, WASP4) are
now capable of simulating water column-sediment interactions (resuspension,
settling, and diffusion). With the advent of sediment criteria in the next few
years, it will be necessary to use models that predict concentrations in both
receiving water and bed sediment. This will be of particular importance in areas
where the sediments are contaminated to the point at which they act as the
source of a pollutant to the water column. Table 4-2 lists and summarizes
models that may be used for predicting the fate and transport of toxicants and
that are supported by the EPA Center for Exposure Assessment Modeling
(CEAM) in Athens, Georgia (Ambrose and Barnwell 1989). All the models, plus
two bioaccumulation models, are briefly described below.
o DYNTOX (Limno Tech 1985) is a wasteload allocation model that
uses a probablistic dilution technique to estimate receiving water
160

-------
chemical concentrations or whole effluent toxicity fractions. The
model considers dilution and net first-order loss?but not sorption
and benthic exchange. The net loss rate must be determined
empirically on a case-by-case basis and cannot be extrapolated to
different conditions of flow, temperature, solids, pH, or light.
EXAMS-II (Burns and Cline 1985) is a compartment model that
can be used as either a steady-state or a quasi-dynamic model
designed for evaluation of the behavior of synthetic organic
chemicals in aquatic ecosystems. It simulates a toxic chemical and
its transformation products using second-order kinetics for all
significant organic chemical reactions. EXAMS-II does not simulate
the solids with which the chemical interacts. The concentration of
solids must be user-specified for each compartment. The model
accounts for sorbed chemical transport based on solids
concentrations and specified transport fields. Sediment exchanges
with the water column include pore-water advection, pore-water
diffusion, and solids mixing. The latter describes a net steady-state
exchange associated with solids that is proportional to pore-water
diffusion.
WASP4 (Ambrose et al. 1988) is a generalized modeling framework
for contaminant fate in surface waters. Based on the flexible
compartment modeling approach, WASP4 can be applied in one,
two, or three dimensions, given the transport of fluxes between
segments. WASP4 can read output files from the link-node
hydrodynamic model DYNHYD4, which predicts unsteady flow
rates in unstratified rivers and estuaries given variable tides, wind,
and inflow. TOXI4, a subset of WASP4, simulates up to three
interacting toxic chemicals and up to three sediment size fractions
in the bed and overlying waters. First- or second-order kinetics can
be used for all significant organic chemical reactions. Sediment
exchanges include pore-water advection, pore-water diffusion, and
deposition/scour. Net sedimentation and burial rates can be
specified or calculated. The output can be used with the two
bioaccumulation models FGETS and FCM2, which are described
below.
HSPF (Johansen et al. 1984) simulates watershed hydrology and
water quality for both conventional and toxic organic pollutants.
HSPF incorporates the watershed-scale ARM and NPS models into
a basin-scale analysis framework that includes transport and
transformation in one-dimensional stream channels. The simulation
provides a time history of the runoff flow rate, sediment load, and
nutrient and pesticide concentrations, along with a time history of
water quantity and quality at any point in a watershed. HSPF
simulates three sediment types (sand, silt, and clay) in addition to
161

-------
specific organic chemicals and transformation products of those
chemicals. The reaction and transfer processes included are
hydrolysis, oxidation, photolysis, biodegradation, volatilization, and
sorption. Sorption is modeled as a first-order kinetic process in
which a desorption rate and an equilibrium partition coefficient for
each of the three solid types must be specified. Resuspension and
settling of silts and clays (cohesive solids) are defined in terms of
shear stress at the sediment-water interface. For sands, the system's
capacity to transport sand at a particular flow is calculated and
resuspension or settling is defined by the difference between the
sand in suspension and the calculated capacity. Sediment exchanges
are modeled as sorption/desorption and deposition/scour with
surficial benthic sediments. Underlying sediment and pore water
are not modeled.
o SARAH2 (Vandergift and Ambrose 1988) is a steady-state,
near-field model for calculating acceptable concentrations of
hazardous organic chemicals discharged to land disposal or
wastewater treatment facilities. Acceptable leachate or treated
industrial waste discharge constituent concentrations are estimated
by a "back calculation" procedure starting from chemical safety
criteria in surface water, drinking water, or fish. For steady or
Table 4-2


Toxicant Fate and Transport Models

Model
Environment
Time Domain
Spatial Domain
Chemical
DYNTOX
river
dynamic
farfield;
1 dimensional
organic; metal
EXAMS-II
lake; river;
estuary
steady-state
quasi-dynamic
farfield;
3-dimensionaI
organic
WASP4
lake; river;
estuary
steady-state,
dynamic
farfield;
3-dimensional
organic; metal
HSPF
river
dynamic
farfield
1-dimensional
organic; metal
SARAH2
river
steady-state
treatment plant;
nearfield;
2-dimensional
organic
MINTEQA2
lake; river;
estuary
steady-state

metal
162

-------
batch waste streams, SARAH2 considers the following concentration
reductions: dilution and loss during treatment, initial Gaussian mixing at
the edge of a stream, lateral and longitudinal diffusion in the mixing zone,
sorption, volatilization, hydrolysis, and bioaccumulation in fish. The user
must specify appropriate instream criteria for protection of the aquatic
community and of humans exposed through consumption of fish and water.
The benthic community is not presently considered. Treatment loss is
handled empirically. SARAH2 contains data sets for three
disposal-watershed scenarios that can be easily modified and employed.
The model is designed for screening analysis and contains numerous
assumptions that should be verified before the model is used in actual
cases.
o MINTEQA2 is an equilibrium metals speciation model for dilute aqueous
systems (Brown and Allison 1987). It does not have any transport and
transformation processes and must be run in conjunction with one of the
above models. It can be used to calculate the mass distribution at
equilibrium among dissolved, absorbed, and solid phases and the species
distribution within each phase. MINTEQA2 contains a chemical
component data set for major ions commonly found in aqueous systems
(e.g., Ca, Fe, and S), trace metals/metalloids of pollution interest (e.g., Cd,
Cr, Ni, Pb, and Zn), and organic ligands of significant affinity for metal
complexation. The model can be used to calculate the concentrations of
adsorbed metals via any of seven different adsorption algorithms.
o FGETS (Food and Gill Exchange of Toxic Substances) is a toxicokinetic
model that simulates the bioaccumulation of nonpolar organic chemicals
by fish from both water and tainted food (Barber et al. 1988). Both of
these routes of exchange are modeled as diffusion processes that depend
upon physicochemical properties of the pollutant and
morphological/physiological characteristics of the fish. FGETS contains a
moderately sized database of allometric relationships for gill morphology
with which it can simulate the direct gill/water exchange of organic
chemicals for essentially any fish species, assuming certain default values.
FGETS also contains a limited database of physiological/- morphological
relationships that are used to parameterize food exchange. In addition to
simulating bioaccumulation of organic toxicants, FGETS can calculate time
to death from chemicals whose mode of action is narcosis. This
calculation is based on the existence of a single, lethal, internal chemical
activity for such chemicals. The concentrations of toxic chemical to which
the food chain is exposed may be specified by the user or may be taken
directly from the values calculated by the exposure concentration model
WASP4. Thus FGETS may be executed as a separate model or as a
post-processor to WASP4.
163

-------
o FCM2 (WASP Food Chain Model) is a generalized model of the uptake
and elimination of toxic chemicals by aquatic organisms (Connolly and
Thomann 1985). It generates a mass balance calculation in which the
rates of uptake and elimination are related to the bioenergetic parameters
of the species. A linear food chain or a food web may be specified. Fish
tissue concentrations are calculated as a function of time and age for each
species included. Exposure to the toxic chemical in food is based on a
consumption rate and predator-prey relationships that are specified as a
function of age. Exposure to the toxic chemical in water is functionally
related to the respiration rate. Steady-state concentrations also may be
calculated. The concentrations of the toxic chemical to which the food
chain is exposed may be specified by the user or may be taken directly
from the values calculated by the exposure concentration model WASP4.
Thus FCM2 may be executed as a separate model or as a post-processor
to WASP4. Migratory species, as well as nonmigratory species, may be
considered. Separate nonmigratory food chains may be specified, and the
migratory species is. exposed sequentially to each food chain based on its
seasonal movements.
4.5.5 Effluent Toxicity Modeling
To apply the steady-state, continuous simulation, or probabilistic methods
to effluent toxicity modeling, the percent effluent measurements should be
converted to toxic units (TUs). Chapter 1 discusses in detail TUs and their
derivation from LCstf and NOELs. Chapter 2 discusses the use of TUs in
establishing a criterion for whole effluent toxicity, and Chapter 3 discusses their
use in characterizing effluent toxicity. In essence, it is necessary to convert
toxicity to units that can be directly related to mass. When comparing toxicity
among chemicals, the relationship between toxicity and concentration is inverse;
chemicals that have toxic effects at low concentrations have a greater "toxicity"
than chemicals that have toxic effects at higher concentrations. The modeling of
toxic effluents is based on mass balance principles; therefore, toxicity needs to be
in units that increase when the percent of the effluent of the receiving stream
increases. Thus, a TU is the reciprocal of the dilution that produces the test
endpoint (e.g., NOEL). An acute toxic unit (TUa) is the reciprocal of the effluent
concentration that causes 50 percent of the organisms to die by the end of the
acute exposure period. A chronic toxic unit (TUC) is the reciprocal of the
concentration that causes no unacceptable effect on the organisms by the end of
the chronic exposure period. The WLA must ensure that the CMC and the CCC
are met in the receiving water at the desired duration and frequency. The CMC
for toxicity is recommended as 0.3 TUa- This is a value that should prevent
lethality unless the duration of exposure exceeds one hour.
164

-------
The CCC for toxicity measured with chronic tests is recommended as the
following:
CCC= 1.0 TUC.
The first step in the TMDL process is to calculate the allowable acute
effluent toxicity that meets the CMC in the receiving water at the duration and
frequency discussed in Appendix D.
The next step in the TMDL process is to calculate the allowable chronic
effluent toxicity that meets the CCC in the receiving water at the duration and
frequency discussed in Appendix D. To compare the allowable acute toxicity
value to the allowable chronic toxicity value, the numbers must be converted to
the same units as follows:
TUa = (ACR)(TUc)
where ACR is the acute-chronic ratio determined from tests on the effluent. It is
important that the ACR used for TMDL purposes be based on actual data and
not be assumed to be 10 or 20 as in the screening procedure (Chapter 3). The
value of this ratio will influence whether the acute or chronic TMDL is more
stringent and is used to calculate the permit limit using the methods described in
Chapter 5.
At the present time, the fate of effluent toxicity in a receiving water is not
fully understood. Even if a decay rate for toxicity can be measured on a given
day in a site-specific situation, there is no way as yet to know how this rate is
affected by temperature, pH, or other environmental conditions. There is also no
way to know how this rate may change when new treatment is installed. Instream
measurements of toxicity should be made at least once per season to identify any
time-varying trends in site-specific fate processes. These monitored decay rates
can then be used in steady-state or continuous simulation fate and transport
models to predict receiving water toxicity, assuming that the rates will not change
with future treatment.
Without specific information concerning the persistence of toxicity, it is
recommended that effluent toxicity be limited to dilution estimates and that
toxicity be assumed to be additive and conservative. Toxicity is expected to be
additive even when the toxicity of one effluent affects selected biota while the
toxicity of a downstream discharge affects different biota. For rivers and
run-of-river reservoirs with a detention time of less than 20 days, the following
dilution equation should be used, assuming completely mixed conditions:
CSQS + CeQe
C= 	
OeQs
165

-------
where:
C = downstream concentration (TUC or TUa),
Q = upstream concentration (TUC or TUa),
Qs = upstream flow (cfs),
Ce = effluent concentration (TUC or TUa), and
Qe = effluent flow (cfs).
For multiple dischargers, this equation must be applied sequentially to find
the concentration as a function of distance downstream. The equation can be
used for a steady-state analysis if Qs is set equal to the design flow, Qe is set
equal to the historical plant flow, and Ce is calculated to meet the CMC and
CCC. This equation can also be used with the continuous simulation, lognormal
probabilistic, or Monte Carlo methods. For these dynamic analyses, a series of
Ce QB Cs, and Qs values would be used.
If instream toxicity measurements are available and a first-order decay rate
for toxicity can be estimated, the following equation should be used:
C _ c
where:
C = downstream concentration (TUC or TUa),
C0 = concentration after the point source discharge has mixed completely
with the river (TUC or TUa),
x = distance downstream of complete mix point,
u = velocity of river, and
K = measured decay rate.
Additional statistical approaches are available that might provide better statistical
fits to the available data. However, these models are somewhat more limited
than the example provided above.
The same equations used for toxicity analyses in rivers can also be used in
steady-state, continuous simulation, or probabilistic analysis of long, narrow,
shallow impoundments with high inflow velocities. Wider, deeper lakes require
more complicated analyses since prolonged detention times (>20 days) and
stratification exert a significant impact on water quality. The prolonged detention
times make it essential that receiving water measurements of toxicity be available
to estimate decay factors. These measurements should be made at least once per
166

-------
season to identify any time-varying trends in toxicity fate processes. Steady-state
or continuous simulation fate and transport models for lakes can then be run with
monitored decay rates for toxicity. A simple steady-state analysis can be
performed using the following equations (Mills et al. 1982):
T„= V/Q
c = Cjr/(i+T jq
where:
Th =	mean hydraulic residence time,
V =	lake volume at critical conditions,
Q = mean total inflow rate at critical conditions,
C =	steady-state lake concentration (TUC or TUa),
Cirr	steady-state inflow concentration (TUC or TUa), and
K =	first-order decay rate.
If effluent is discharged into a stratified lake and mixes only with the
hypolimnion or epilimnion, the volume of the layer should be used only to
calculate mean hydraulic residence time (Tw). TTie mean total inflow rate (Q)
and the inflow concentration (C1r0 should be calculated as the sum of all sources
to the lake, including point source, nonpoint source, and tributary inputs.
Dilution calculations for effluent toxicity discharges to an estuary are
complicated by the oscillatory motion of the tides and possible stratification of
the estuary. The prolonged detention times make it essential that field
measurements of toxicity be available to estimate decay factors. These
measurements should be made at least once per season to identify any
time-varying trends in toxicity rate processes. Steady-state or continuous
simulation fate and transport models for estuaries can then be run with
monitored decay rates for toxicity. A simple steady-state analysis can be
performed using the following equations for each nonconservative pollutant
entering from the river at the head of an estuary (Mills et al. 1982):
C,= C,-l(f,)B,
(M
where:
Bi = r,
l-(l-ri) e
-kt
167

-------
n = exchange ratio for segment i as defined by modified tidal prism
method,
t = flushing time,
fi = fraction of freshwater in segment i,
Q = nonconservative pollutant concentration in segment i (TUa or TUC),
and,
k = decay rate of pollutant.
The following equations should be used for each nonconservative pollutant
entering along the side of an estuary:
For segments downstream of outfall:
Ci = C0 fj f =l,...,n	r^
f.	l-(l-r,)e •"
For segments upstream of outfall:
Cj Co Si j ~~ X,**7Taa«,n
So
l-(l-r,)e
-let
where:
Ci = nonconservative pollutant mean concentration in segment i (TUC or
TUa),
C0 = nonconservative pollutant mean concentration in segment of
discharge,
168

-------
r,- = exchange ratio for segment i as defined by the modified tidal prism
method,
n = number of segment away from outfall,
f< = fraction of freshwater in segment i,
fo = fraction of freshwater in segment with discharge,
Si = salinity in segment i,
S0 = salinity in segment of discharge,
k = decay rate, and
t = flushing time.
The details of how to calculate exchange ratios and flushing times for estuaries
are included in Part 2 of EPA's Water Quality Assessment Manual (Mills et al.
1982). This manual also describes how to perform these calculations for stratified
estuaries using a two-dimensional box model analysis.
4.6 HUMAN HEALTH
4.6.1	Human Health Considerations
Human exposure to pollutants should be evaluated as completely as
available information will allow. Exposure information is used in calculating the
human health acceptable ambient concentration (AAC) from the formulas in
Section 2, Water Quality Standards. This information should be used to estimate
exposures due to fish consumption and drinking water ingestion, background
concentrations, and other exposure routes, such as recreational, occupational,
drinking water, dietary (other than fish), and inhalation. Factors in the formulas
for which information is not available can be omitted from the calculation. If
states choose, bioaccumulation factors (BAF) can also be modified.
4.6.2	Determining the Wasteload Allocation for Human Health Toxicants
Wasteload allocations (WLAs) are necessary only where mixing is allowed.
Mixing zones are used at the discretion of the States. If a State does not allow a
mixing zone or the assumption of complete mixing, then the AAC is applied at
the end-of-pipe and no TMDL or WLA determination is necessary.
With persistent or bioconcentratable pollutants, special mixing zone
considerations apply. Bioconcentratable pollutant criteria exceedances within the
169

-------
mixing zone can potentially result in tissue contamination of organisms directly or
indirectly through contamination of bed sediments with subsequent incorporation
in the food chain. For discharge situations with incomplete mixing (e.g., large
rivers, lakes, estuaries, oceans), States need to carefully consider whether mixing
zones for persistent or bioconcentratable pollutants are appropriate. Where a
mixing zone is allowed, one WLA should be calculated to achieve the AAC or
criterion selected above (USEPA 1987). Because most human health criteria are
chronic only, an acute WLA will usually not be needed, although EPA's Office of
Drinking Water does have acute criteria for some pollutants.
For the purpose of the following discussion, use of simple, steady-state
dilution models is assumed. However, these models may be inappropriate for
certain situations where sediments serve as a sink for bioconcentratable pollutants
and where additional factors need to be considered. Dynamic models, where
available, are useful tools for accounting for an array of variables that may have
an impact on the fate of bioconcentratable pollutants in the food chain.
In simple situations, the TMDL is determined from the ACC and the
design flow of the receiving water. In more complicated situations, e.g., where
mixing is not rapid or where lakes or estuaries are involved, a spatial averaging
scale must be chosen. Selection of the spatial scale must be consistent with
reasonable assumptions about the behavior of aquatic organisms and the target
human population.
In some cases, it may be necessary to apply the chronic human health
criterion within this mixing zone, if it is reasonable to assume that (1) the
bioconcentrating aquatic organisms have little mobility, thus spending most of
their time within the mixing zone, and (2) the target human population
consistently consumes fish from the mixing zone (over a 70-year lifetime, for
carcinogenic risks).
The procedure for developing TMDLs/WLAs generally requires
determining values for the following parameters, based upon water quality
considerations: (1) the duration of the averaging period applicable to the WLA,
(2) design considerations, e.g., flow, (3) the discharge (WLA) concentration that
will result in meeting the ambient water quality criterion during the design
condition, and (4) the allowable probability (or frequency) of the discharge's
exceeding the WLA, averaged over the appropriate duration. The technical basis
for setting these values is discussed in the following sections.
1) Averaging Periods
The duration of the averaging period for the WLA should be selected to
be consistent with the assumptions used to derive the water quality criteria. Two
categories of pollutants should be recognized: carcinogens and noncarcinogens.
The human health criteria for carcinogens are derived assuming lifetime
exposure. The upper-bound risk is directly proportional to the lifetime arithmetic
170

-------
mean dose. The criteria thus applies,to the ambient water concentrations
averaged over a 70-year period.
The duration of exposure assumed in deriving criteria for noncarcinogens
may be ambiguous, particularly where the criterion is derived from animal
studies. Furthermore, the duration may be highly variable, ranging as high as 20
to 30 years for cadmium.
2) Dilution Design Conditions
a^ Carcinogens: Rivers and Stream Discharge Situations
In well-mixed situations, the receiving water concentration, C, is
determined by the pollutant load, W (mass/time), and the combined receiving
water plus effluent flow, Q, such that:
C = W/Q, (Qhn).
The long-term harmonic mean flow is recommended as the design flow for
carcinogens. The harmonic mean, or;
Qhm = n/E i=ni=i (1/Q)
(where n = the number of recorded flows) is always less than the arithmetic
mean. The arithmetic mean flow is not appropriate as the design flow, since it
overstates the dilution available. Extreme value statistics (such as 7Q10 or 30Q5)
are also not appropriate, since they have no consistent relationship with the long-
term mean dilution.
The harmonic mean flow, may be estimated by any of several methods
described in U.S. EPA, 1989 assuming that flows are approximately lognormally
distributed:
Qhm = Qgm
Qam
where is the geometric mean flow, and Qam is the arithmetic mean flow. For
USGS flow records, summaries of the statistical parameters needed to estimate
the harmonic mean can be quickly obtained from STORET, through a user-
friendly procedure for permit writers, as described in Appendix D.
Two software packages are available for computation of harmonic mean
flow: WQAB FLOW (described in Appendix D) and HHDFLOW (discussed
below). The HHDFLOW program should be used in conjunction with data that
are not lognormally distributed.
171

-------
In order to develop some quantitative sense of how a long-term harmonic
mean flow of any stream compares with its 7-day 10-year low flow, the
Assessment and Watershed Protection Division and the Risk Reduction
Engineering Laboratory at Cincinnati, Ohio analyzed flow records of 60 streams.
These are the same stream flow records that had been analyzed for stream design
flow condition for aquatic life protection as listed in Book VI, Design Conditions:
Chapter 1, Stream Design Flow for Steady-State Modeling, (Office of Water
Regulations and Standards, USEPA, August 1986). Based on the long-term
harmonic flow and 7-day 10-year low flow estimates for these 60 streams, the
long-term harmonic mean flows of all 60 streams were equal to or greater than 2
times the 7Q10 low flow. Fifty-four of the streams' harmonic mean flows were
equal to or greater than 2.5 times their 7Q10 low flows. Finally, 40 of the 60
streams' harmonic mean flows were equal to or greater than 3.5 times the 7Q10.
Based on the above observations, permit authorities may choose a
multiplication factor of 3 x 7Q10 to estimate stream design flow for human health
protection for carcinogenic pollutants. However, it is recommended that the
harmonic mean flow be calculated directly from the historical daily flow record, if
possible. A modified version of the DFLOW computer program (named
HHDFLOW) is available to make the calculation. Alternatively, the following
equation might be used to estimate harmonic mean flow [Rossman, 1989]:
Qhm = [1.194 - (QX'7) * [(7Q10)°-55^, r2 = .99.
In that equation, Qan, and 7Q10 are estimated using the USGS computer program,
FLOSTAT.
b1 Non-carcinogens: River and Stream Discharge Situations
The choice of return interval represents a level-of-protection consideration
inherent in the risk management decision to be made by the permitting agency.
If a short term duration of exposure is chosen (i.e., 90 days or less), design flows
may be appropriately based on extreme value statistics. Because the effects from
noncarcinogens are more often associated with shortened exposures, EPA
suggests the use of 30Q5. However, in the comparisons of flows for smaller rivers
(i.e., low flow 50 cfs), the 30Q5 flow was only 1.1 times that of the 7Q10. For
larger rivers (i.e., low flow of 600 cfs), the factor was 1.4 times.
172

-------
3) Point of Application of the Criteria
The point at which the chronic criteria are to be met in the receiving
water may be fixed by existing State standards or may be determined by
considerations for managing individual and aggregate risks. The several
possibilities include the following:
1.	Where State standards allow no mixing zone and no spatial averaging, the
criterion would be met at the end of the pipe.
2.	Where State standards specify that the criterion must be met at the end of
the mixing zone, the criterion would be applied at that point.
3.	Where State standards allow consideration of spatial averaging, the
criterion may be met as an average within a specified area, as appropriate
for the individual and aggregate risk scenarios underlying the application.
173

-------
REFERENCES
1.	Albcrtson ML, Dai YB, Jensen RA, and Rouse H. 1950. Diffusion
of Submerged Jets. Trans. Am. Soc. Civ. Eng. Vol.115. 639-664.
2.	Ambrose RB, and Barnwell TO. 1989. Environmental Software at
the USEPA Center for Exposure Assessment Modeling. U.S.
Environmental Protection Agency. Environmental Research
Laboratory. Athens, GA.
3.	Ambrose RB, Wool TA, Connolly JP, and Shantz RW. 1988.
WASP4, A Hydrodynamic and Water Quality Model-Model
Theory, User's Manual, and Programmer's Guide. U.S.
Environmental Protection Agency. Center for Exposure Assessment
Modeling. Athens, GA. EPA/600/3-87/039.
4.	Baily TE. Fluorescent-Tracer Studies of an Estuary. 1966. Journal
of the Water Pollution Control Federation. Vol. 38. No. 12.
p. 1988.
5.	Barber MC, Suarez LA, and Lassiter RR. 1988. FGETS (Food
and Gill Exchange for Toxic Substances): A Simulation Model for
Predicting Bioaccumulation of Nonpolar Organic Pollutants by Fish.
U.S. Environmental Protection Agency. Center for Exposure
Assessment Modeling. Athens, GA. EPA/600/3-87/038.
6.	Brown DS, and Allison JD. 1987. MINTEQA1 Equilibrium Metal
Speciation Model: A Users Manual. U.S. Environmental
Protection Agency. Environmental Research Laboratory. Athens,
GA. EPA/600/3-87/012.
7.	Brungs WA. 1986. Allocated Impact Zones for Areas of
Non-Compliance. U.S. Environmental Protection Agency. Region 1.
Water Management Division. Boston, MA.
8.	Burns LA, Cline DM. 1985. Exposure Analysis Modeling System
Reference Manual for EXAMS II. U.S. Environmental Protection Agency.
Environmental Research Laboratory. Athens, GA EPA 600/3-85-038.
9.	Callahan M, Slimak M, Gabel N, May I, Fowler C, Freed J, Jennings P,
Durfee R, Whitmore F, Maestri B, Mabey W, Holt B, and Gould C. 1979.
Water-related Fate of 129 Priority Pollutants. Volumes I and II. U.S.
Environmental Protection Agency. Office of Water Regulations and
Standards. Washington, DC EPA 440/4-79-029.
174

-------
10.	Chadderton RA, Miller AC, and McDonnell AJ. 1981. Analyses of Waste
Load Allocation Procedures. Water Resources Bulletin. Vol. 17. No. 5.
pp. 760-766.
11.	Connolly JP, and Thomann RV. 1985. WASTOX, A Framework for
Modeling the Fate of Toxic Chemicals in Aquatic Environments. Part 2:
Food Chain. U.S. Environmental Protection Agency. Environmental
Research Laboratory. Gulf Breeze, FL. EPA/600/4-85/040.
12.	Crocker PA, Morrison G, Rives SR, Shipley FS, Palachek R, Neleigh D,
and Schimmel S. 1989. Case Study to Evaluate Effluent and Receiving
Water Toxicity for an Estuarine Discharger: U.S. Naval Air Station,
Corpus Christi, Texas. Technical Section, Water Quality Management
Branch, U.S. Environmental Protection Agency, Region 6.
13.	DiToro D. 1984. Probability Model of Stream Quality Due to Runoff.
Journal of Environmental Engineering, Proc. ASCE. Vol. 110 (3).
14.	Donekar RL, and Jirka GH. 1988. CORMIX 1. An Expert System for
Mixing Zone Analysis of Toxic and Conventional Single Port Aquatic
Discharges. U.S. Environmental Protection Agency, Environmental
Research Laboratory. Athens, GA. PB 88-220 504/AS.
15.	Fischer HB. 1968. 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. U.S. Geological Survey. U.S. Department of the Interior. Reston,
VA.
16.	Fischer HB. 1972. A Langrangian Method for Predicting Pollutant
Dispersion in Bolinas Lagoon, Marin County, California. Dispersion in
Surface Water. U.S. Geological Survey Professional Paper 582-B. U.S.
Geological Survey. U.S. Department of the Interior. Reston, VA.
17.	Fischer HB, List EJ, Koh CY, Imberger J, and Brooks NH. 1979. Mixing
in Inland and Coastal Waters. Academic Press. New York.
18.	Ford DE, and Johnson LS. 1986. An Assessment of Reservoir Mixing
Processes. Technical Report 86-7 prepared by Ford, Thornton, Norton
and Associates, Ltd., Little Rock, AR, for the U.S. Army Engineer
Waterways Experiment Station, Vicksburg, MS.
19.	Ford DE, Thornton KW, Lessem AS, and Norton JL. 1981. A Water
Quality Management Model for Reservoirs. ASCE Proc. of Symposium on
Surface Water Impoundments. June 2-5, 1980. Minneapolis, MN. Stefan,
HG.
175

-------
20. Gobas, et al. nd. In Environmental Toxicology and Chemistry 8, p. 231.
21. Hetling LJ, and O'Connell RL. 1966. A Study of Tidal Dispersion in the
Potomac River. Water Resources Research. Vol. 2. No. 4. p. 825.
22o Holley ER, and Jirka GH. 1986. Mixing in Rivers. Technical Report
E-86-11. U.S. Army Engineer Waterways Experiment Station, Vicksburg,
MS.
23.	Jirka GH. 1982. Multiport Diffusers for Heat Disposal: A Summary.
Journal of the Hydraulics Division. Proc. ASCE. Vol. 108. No. HY12.
pp. 1425-1468.
24.	Johansen RC, Imhoff JC, Kittle JL, and Donigian AS. 1984. Hydrologic
Simulation Program—FORTRAN (HSPF): Users Manual for Release 8.0.
U.S. Environmental Protection Agency. Environmental Research
Laboratory. Athens, GA. EPA/600/3-84-066.
25.	Johnson MC. 1984. Fluorometric Techniques for Tracing Reservoir
Inflows. Instruction Report E-84-1. U.S. Army Engineer Waterways
Experiment Station. Vicksburg, MS.
26.	Kisiel CC. 1964. Dye Studies in the Ohio River. Proceedings of
Symposium on Diffusion in Oceans and Fresh Waters. Columbia
University.
27.	Limno Tech, Inc. 1985. Dynamic Toxics Wasteload Allocation Model
(DYNTOX): User's Manual. U.S. Environmental Protection Agency.
Office of Water Regulations and Standards. Washington, DC.
28.	Lyman W, Reehl WF, and Rosenblatt DH. 1982. Handbook of Chemical
Property Estimation Methods. McGraw-Hill Book Co. New York.
29.	Mabey W, Smith J, Podoll R, Johnson H, Mill T, Chou T, Gates J,
Partridge I, and Vandenberg D. 1981. Aquatic Fate Process Data for
Organic Priority Pollutants, U.S. Environmental Protection Agency. Office
of Water Regulations and Standards. Washington, DC.
EPA 440/4-81-014.
30.	Malone RF, Bowles DS", Grenney WJ, and Windham MP. 1979.
Stochastic Analysis of Water Quality. Utah State University.
UWRL/Q-79/01,
176

-------
31.	Mills WB, Dean JD, Porcella DB, Gherini SA, Hudson RJM, Frick WE,
Rupp GL, and Bowie GL 1982. Water Quality Assessment: A Screening
Procedure for Toxic and Conventional Pollutants. U.S. Environmental
Protection Agency. Environmental Research Lab. Athens, GA.
EPA 600/6-82-004 a and b.
32.	Mullenhoff WP, Soldate AM, Baumgartner DJ, Schuldt MD, Davis LR,
and Frick WE. 1986. Initial Mixing Characteristics of Municipal Ocean
Discharges: Volume I. Procedures and Applications. U.S. Environmental
Protection Agency. Environmental Research Laboratory. Narragansett,
RI. PB 86-137478.
33.	NAS. 1972. Water Quality Criteria, 1972. National Academy of Sciences.
U.S. Environmental Protection Agency. Washington, DC. EPA-R3-73-033.
34.	Natrella MG. 1963. Experimental Statistics. National Bureau of
Standards Handbook 91. U.S. Department of Commerce.
35.	Oregon Environmental Quality Commission. 1988. Memorandum from
Fred Hansen, Director, to the Commission. Dated July 8, 1988. Subject:
Proposed Adoption of Rules Defining Loading Capacity (LC), Waste Load
Allocation (WLA), Load Allocation (LA), and Total Maximum Daily
Loads (TMDL) (OAR 340-41-006) and Proposed Adoption of Rules
Establishing Total Maximum Daily Loads, Load Allocations, and Waste
Load Allocations for Total Phosphorus and Ammonia in the Tualatin
River Basin (340-41-470). Portland, OR.
36.	Pailey PP, and Sayre WW. 1978. Model for Shore-Attached Thermal
Plumes in Rivers. Journal of the Hydraulics Division. Proc. ASCE.
Vol. 104.
37.	Roberts PJW, Synder WH, and Baumgartmer DJ. 1989a. Ocean outfalls
I: Submerged Wastefield Formation. ASCE. J. Hydraulic Engineering.
Vol. 115. No. 1. pp. 1-25.
38.	Roberts PJW, Synder WH, and Baumgartmer DJ. 1989b. Ocean Outfalls
II: Spatial Evolution of Submerged Wastefield. ASCE. J. Hydraulic
Engineering. Vol. 115. No. 1. pp. 26-48.
39.	Roberts PJW, Synder WH, and Baumgartmer DJ. 1989c. Ocean Outfalls
III: Effect of Diffuser Design on Submerged Wastefield. ASCE. J.
Hydraulic Engineering. Vol. 115. No. 1. pp. 49-70.
40.	Rose KA, Summers JK, and McLean RI. 1989. A Monte Carlo
Bioaccumulation Model of Silver Uptake by the Oyster, Crassostrea
Virginia. Ecol. Mod. Vol. 42. pp. 10-21.
177

-------
41.	Rossman, Lewis A. Design Stream Flows Based on Harmonic Means.
(Submitted for publication).
42.	Scavia D, Powers WF, Canale RP, and Moody JL. 1981. Comparison of
First-order Error Analysis and Monte Carlo Simulations in
Time-dependent Lake Eutrophication Models. Water Resources Research.
Vol. 17. No. 4. pp. 1051-1059.
43.	Thornton KW, Ford DE, and Norton JL. 1982. Use of Monte Carlo
Water Quality Simulations to Evaluate Reservoir Operational and
Management Alternatives. Presented at Third International Conference
on State-of-the-Art Ecological Modeling, Colorado State University.
44.	Tiwari JL, and Hobbie JE. 1976. Random Differential Equations as
Models of Ecosystems: Monte Carlo Simulation Approach. Math
Bioscience. Vol. 28.
45.	USEPA. 1982. Revised Section 301(h) Technical Support Document.
U.S. Environmental Protection Agency. Office of Water Program
Operations. Washington, DC. EPA 430/9-82-011.
46.	USEPA. 1983a. 40 CFR Part 131, Revised Water Quality Standards
Regulation. Federal Register, Vol. 48, Nov. 8, 1983.
47.	USEPA. 1983b. Water Quality Standards Handbook. U.S.
Environmental Protection Agency. Office of Water Regulations and
Standards. Washington, DC.
48.	USEPA. 1984. Technical Guidance Manual for Performing Wasteload
Allocations. Book II Streams and Rivers. Chapter 3: Toxic Substances.
U.S. Environmental Protection Agency. Office of Water Regulations and
Standards. Washington, DC. EPA 440/4-84-022.
49.	USEPA. 1985. 40 CFR Parts 35 and 130, Final Rule. Water Quality
Planning and Management. Federal Register, Vol. 50, Jan. 11, 1985.
50.	USEPA. 1986a. Technical Guidance Manual for Performing Wasteload
Allocations. Book IV Lakes, Reservoirs, and Impoundments. Chapter 3
Toxic Substances. U.S. Environmental Protection Agency. Office of
Water Regulations and Standards. Washington, DC. EPA 440/4-87-002.
51.	USEPA. 1986b. Technical Guidance Manual for Performing Wasteload
Allocations. Book VI Design Conditions. Chapter 1 Stream Design Flow
for Steady-State Modeling. U.S. Environmental Protection Agency. Office
of Water Regulations and Standards. Washington, DC.
178

-------
52.	USEPA. 1987a. Reichenbach NG, Wickramanayake CB, Gavaskar AR,
Walsh JJ, Hayward RS, and Vaya RM. Amelia River Pilot Study for
Implementing EPA's Toxic Control Policy. U.S. Environmental Protection
Agency. Office of Water Regulations and Standards. EPA Contract
No. 68-01-6986. Work Assignment Nos. 1-72.
53.	USEPA. 1987b. McKeon TJ, and Segna JJ. Selection Criteria for
Mathematical Models Used in Exposure Assessments Surface Water
Models. U.S. Environmental Protection Agency. Office of Health and
Environmental Assessment. Washington, DC. EPA/600/8-87/042.
54.	USEPA. 1987c. Schoor JL, Chikashi S, McKechnie D, Schoo D.
Processes, Coefficients, and Models for Simulating Toxic Organics and
Heavy Metals in Surface Waters. U.S. Environmental Protection Agency.
Environmental Research Laboratory. Athens, GA. EPA 600/3-87-015.
55.	USEPA. 1987d. Permit Writers Guide to Water Quality-based Permitting
for Toxic Pollutants. U.S. Environmental Protection Agency. Office of
Water Enforcment and Permits, Washington, DC. EPA/440/4-87/005.
56.	USEPA. 1988a. Mixing Zones Water Quality Standards Criteria
Summaries: A Compilation of State/Federal Criteria. U.S. Environmental
Protection Agency. Office of Water Regulations and Standards.
Washington, DC. EPA 400/5-88/015.
57.	USEPA. 1988b. Dettmann EH, Paul JF, Rosen JS, and Strobel CJ.
Transport, Fate, and Toxic Effects of a Sewage Treatment Plant Effluent
in a Rhode Island Estuary. U.S. Environmental Protection Agency.
Environmental Research Laboratory. Narragansett, RI.
EPA/600/X-87-366.
58.	USEPA. 1988c. Technical Guidance Manual for Performing Wasteload
Allocations Book VII Design Conditions. Chapter 2 Technical Guidance
on Supplementary Stream Design Conditions for Steady-state Modeling.
U.S. Environmental Protection Agency. Office of Water Regulations and
Standards. Washington, DC.
59.	USEPA. 1989. Draft Technical Guidance Manual for Performing
Wasteload Allocations. Book III Estuaries. U.S. Environmental
Protection Agency. Office of Water Regulations and Standards.
Washington, D.C.
60.	USGS. 1982. Hubbard EF, Kiipatrick FA, Martens LA, and Wilson JF.
Measurement of Time of Travel and Dispersion in Streams by Dye
Tracing. Chapter A9. Techniques of Water-Resources Investigations of
the USGS. Book 3 Application of Hydraulics. U.S. Department of the
Interior. Reston, VA.
179

-------
61.	USGS. 1985. Kilpatrick FA, and Cobb ED. Measurement of Discharge
Using Tracers. Chapter A16. Techniques of Water-Resources
Investigations of the USGS. Book 3 Application of Hydraulics. U.S.
Department of the Interior. Reston, VA.
62.	USGS. 1986. Wilson JF, Cobb ED, and Kilpatrick FA. Fluorometric
Procedures for Dye Tracing. Chapter A12 Techniques of Water-Resources
Investigations of the USGS. Book 3 Application of Hydraulics. U.S.
Department of the Interior. Reston, VA.
63.	Vandergift SB, and Ambrose RB. 1988. SARAH2 A Near Field Exposure
Assessment Model for Surface Waters. U.S. Environmental Protection
Agency. Environmental Research Laboratory. Athens, GA.
EPA/600/3-88/020.
64.	Veith, GD, and Kosian P. 1982. Estimating Bioconcentration Potential
from Octanol/Water Partition Coefficients." In Physical Behavior of PCBs
in the Great Lakes. MacKay, Paterson, Eisenrich, and Simmons, eds. Ann
Arbor Science. Ann Arbor, MI. Chapter 15.
65.	Versar. 1984. Draft Assessment of International Mixing Zone Policies.
Avoidance/Attraction Characteristics, and Available Prediction
Techniques. Prepared for U.S. Environmental Protection Agency, Office
of Water Regulations and Standards, EPA Office of Pesticides and Toxic
Substances. Washington, DC. EPA Contract No. 68-01-6271. Assignment
No. 52.
66.	Versar. 1989. Prediction of Uncertainties Associated with Water Quality
Models. Prepared for U.S. Environmental Protection Agency, Office of
Research and Development and Office of Toxic Substances. Washington,
DC. EPA Contract No. 68-02-4254. Assignment No. 125.
67.	Wilson JF, Cobb ED, and Yotsukura N. 1969. Movement of a Solute in
the Potomac Estuary at Washington, D.C. at Low Flow Conditions. U.S.
Geological Survey Circular 529B. U.S. Geological Survey. U.S.
Department of the Interior. Reston, VA.
68.	Wright SJ. 1977. Mean Behavior of Buoyant Jets in a Crossflow. Journal
of the Hydraulics Division. Proc. ASCE. Vol. 103.
69.	Yotsukura N, and Cobb ED. 1972. Transverse Diffusion of Solutes in
Natural Streams. Dispersion in Surface Water. Geological Survey
Professional Paper 582-C. U.S. Geological Survey. U.S. Department of
the Interior. Reston, VA.
180

-------
70.	Yotsukura N, Fisher HB, and Sayre WW. 1970. Measurement of Mixing
Characteristics of the Missouri River between Sioux City, Iowa, and
Plattsmouth, Nebraska. Geological Survey Water-Supply Paper 1899-G.
U.S. Geological Survey. U.S. Department of the Interior. Reston, VA.
71.	Yotsukura N, and Sayre WW. 1976. Transverse Mixing in Natural
Channels. Water Resources Research. Vol. 12. No. 4. pp. 695-704.
181

-------
5. PERMIT REQUIREMENTS
5.1 INTRODUCTION
As the final step in the "standards-to-permits"process, development of
permit requirements is often the culmination of the activities discussed in the
preceding chapters. This chapter describes the basic principles of effluent
variability and permit limit derivation and provides recommendations for deriving
limits from various types of wasteload allocation outputs. It also addresses
important considerations in the expression of limits and other types of permit
requirements, including toxicity reduction evaluations.
5.1.1 Regulatory Requirements
A hallmark of the national surface water toxics control program is the use
of an integrated approach, consisting of both chemical and biological methods for
the assessment and control of toxic pollutants and toxicity. The integrated
approach to water quality-based permitting involves the development of permit
limits for specific chemicals, whole effluent toxicity, or both, depending upon the
requirements of specific situations.
It is important to recognize that there are both mandatory and
discretionary elements associated with the development of water quality-based
permit limits to control toxic pollutants and toxicity. The mandatory elements are
described in 54 EE 23868, June 2, 1989, the revisions to the NPDES Surface
Water Toxics Control Program regulations. The regulations at 40 CFR Part
122.44(d)(1) require that regulatory authorities first determine whether a
discharge is causing or has the potential to cause an excursion above a water
quality standard (narrative or numeric). In making these determinations, they
must use a procedure which accounts for effluent variability, existing controls on
point and nonpoint sources, available dilution, and (when using toxicity testing)
species sensitivity.
There is a degree of flexibility in the specific procedures a regulatory
authority uses in determining whether an excursion is occurring or is projected
and in the weight given to the various factors in conducting their evaluation of a
specific discharger. EPA's guidance for making these determinations is contained
in the data generation recommendations in Chapter 3.
Where the regulatory authority determines, using the procedures discussed
above, that an excursion above a numeric water quality standard (for an
individual chemical or the parameter "toxicity")is either occurring or is projected,
section 122.44(d) requires the development of a permit limitation for the
pollutant or pollutant parameter of concern. Where an excursion above a
narrative standard attributable to an individual chemical is occurring or is
182

-------
projected, and where a State does not have a numeric criterion for the pollutant
of concern, permit limits must be developed using one of three options (using
case-specific water quality criteria, EPA criteria, or an indicator parameter). In
addition, the regulation requires limits for whole effluent toxicity where a
narrative water quality standard is being exceeded except in the following
circumstance: limits on individual chemicals may be substituted where a State
does not have a specific numeric criterion for the parameter "toxicity" and where
limits on individual chemicals are shown to attain and maintain all applicable
water quality standards.
There are also several EPA policies that reflect these regulatory
requirements including the "National Policy for the Development of Water
Quality-based Limits for Toxic Pollutants" (Appendix B-l.l) and EPA's "Whole
Effluent Toxicity Permitting Principles and Enforcement Strategy," (Appendix B-
2) which requires that "all major permits and minors of concern must be
evaluated for potential or known toxicity (chronic or acute if more limiting)." In
addition, the policy states that "Final whole effluent toxicity limits must be
included in permits where necessary to ensure that State Water Quality Standards
are met. These limits must properly account for effluent variability, available
dilution, and species sensitivity."
There is an element of judgment inherent in the specific permit limit
derivation procedures used for an individual discharger once a decision has been
made to develop a specific type of limitation. Case-specific considerations will
usually dictate the most appropriate approach to be taken in individual situations.
In general, the various assumptions used in the permit limit development process
should be consistent with the assumptions and principles inherent in the effluent
characterization and exposure assessment steps preceding permit limit
development.
52 BASIC PRINCIPLES OF EFFLUENT VARIABILITY
An understanding of the basic principles of effluent variability is central to
water quality-based permitting. Many of the concepts are the same as those
considered in the development of technology-based limitations. However, the
process for applying the principles is substantially different, as explained below.
5.2.1 Variations in Eflluent Quality
Effluent quality varies over time in terms of volumes discharged and
constituent concentrations. Variations occur due to a number of factors,
including: changes in human activity over a 24 hour period (particularly for
POTWs), changes in production cycles for industries, variation in responses of
wastewater treatment systems to influent changes, variation in treatment system
performance, and climatological changes. Very few effluents remain constant
over long periods of time. Even in industries that operate continuous processes,
183

-------
variations in the quality of raw materials and activities such a "back-washing of
filters cause peaks in effluent constituent concentrations and volumes.
If effluent data for a particular pollutant or pollutant parameter for a
typical publicly owned treatment works (POTWs) are plotted against time, the
daily concentration variations can be seen (see Figure 5-1, left-hand graphs).
This behavior can be described by constructing frequency-concentration plots of
the same data (see Figure 5-1, right-hand graphs).
522 Statistical Parameters and Relationship to Permit Limits
Based upon the shape of the curve of a frequency-concentration plot, the
data can be described in terms of a particular type of statistical distribution. The
choices for statistical distributions include normal (bell shaped), lognormal
(positively skewed), or other variations on the lognormal distribution. From the
vast amount of data that EPA has examined in this manner, it is reasonable to
assume (unless specific data exist to show otherwise) that treated effluent data
follow a log-normal distribution. This is because effluent values are non-negative
and treatment efficiency at the low end of the concentration scale is limited,
while effluent concentrations may vary widely at the high end of the scale,
reflecting various degrees of treatment system performance and loadings. These
factors combine to produce the characteristically positively skewed appearance of
the log-normal curve when data are plotted in a frequency histogram. Appendix
E provides a discussion of the basis for concluding that effluent data are typically
lognormally distributed, as well as recommendations for handling data sets from
treatment plants that follow some other type of distribution.
Effluent data from any treatment system may be described using standard
descriptive statistics such as the mean concentration of the pollutant or pollutant
parameter (i.e., the long term average (LTA)) and the coefficient of variation
(CV)). The CV is a standard statistical measure of the relative variations of a
distribution or set of data, defined as the ratio of the standard deviation to the
mean. Using a statistical model such as the lognormal, an entire distribution of
values can be projected from limited data, and limits can be set at a specified
probability of occurrence. Figure 5-1 shows the frequency-concentration curve
and the relative positions of the concentrations corresponding to the LTA for the
data on the left hand side of the figure.
All permit limits, whether technology-based or water quality-based, are set
at the upper bounds of acceptable performance. Requirements are usually
expressed using two types of permit limits. The maximum daily permit limit is
the maximum allowable value for any single observation. The monthly average
permit limit is the maximum allowable value for the average of all observations
obtained during one month (limits for weekly periods are also specified in the
NPDES regulations for publicly owned treatment works (POTWs)).
184

-------
3
30
Carlodaphnla sp.
CV s 1.06
§
a
cr
•
>
5
c
20
T Carlodaphnla sp.
j\
CV s 1.06
I \
Long tatm average
'I
\
0.25 0.50 0.75 1.00
Chronic toxic units
1.25
Daphnla sp.
CV = 0.70
Daphnla sp.
CV = 0.70
— Long term average
2 4 6
Acuts toxic Units
Zinc
CV a 0.59
100
c
a 4
?
A 3
I 2
s
nr 1
Zinc
/ \
CV 3 0.59
/
—- Long term average
12 3 4
Concentration
Figure 5-1. Data and Relative Frequency Distributions for Cerlodaphnia Toxicity,
Daphnla Toxicity, and Zinc Concentrations for Three Different Effluents.
185

-------
The purpose of a permit limit is to specify an upper bound of acceptable
effluent quality. For technology-based requirements, the limits are based on
proper operation of a treatment system. For water quality-based requirements,
the limits are based on maintaining the effluent quality at a level that will comply
with water quality standards in the receiving water. These requirements are
determined by the wasteload allocation (WLA). The WLA dictates the required
treatment plant performance.
Setting permit limits to enforce this performance may be difficult. Typical
compliance monitoring requirements are relatively infrequent due to cost
considerations. Also, since effluent limits are typically a multiple of the long
term average, it may be fairly difficult to distinguish a wastewater treatment
facility that is not complying with expected performance. If permit limits are set
too high relative to the LTA, a discharger not complying with the desired level of
performance will not exceed the limits. If permit limits are set too low, a
discharger that is complying with the desired level of performance may frequently
still exceed the permit limits.
In the development of technology-based effluent limitations guidelines, the
operating records of various wastewater treatment facilities for a particular
category of discharger are examined. Based on the effluent data for the
treatment facilities, a composite long term average value for the parameter is
determined. This LTA, in conjunction with relevant estimates of variability, is
then used to derive effluent limitation guidelines which lead directly to permit
limits. In the case of water quality-based permit limits, the process essentially
operates in reverse: The WLA determines the appropriate discharge level, which
in turn determines the requisite LTA for the treatment facility in order to meet
that WLA. Permit limits may then be derived from this LTA and CV.
The WLA is only one factor in determining the LTA; the CV is also
important. The relationship among the various statistical parameters is illustrated
in Figure 5-2. As these figures show, highly variable effluents require a much
lower LTA in order to meet the WLA and account for the variability above the
LTA.
It is extremely important to recognize that the various statistical principles
and relationships discussed above operate in any discharge situation -- whether or
not they are specifically recognized or accounted for. Where a permit limit
derivation procedure does not Specifically address these principles, the permit
writer will be making an implicit assumption that there are enough conservative
assumptions built into other steps in the process (water quality models, "buffer"
between permit limitations and actual operating conditions, etc.) to ensure that
there will not be too many excursions above water quality standards.
186

-------
140
120
100
80 "
SO
40
20
Wasteload allocation
CV - 0.53
Long term average
10
20
30
40
Oaya
50
Days
Figure 5-2a Relationship between a Single Wasteload Allocation
and Two Long Term Averages for Different Coefficients of Variation
!i
fi
1.0
0.S
?i
is o.6
0.4
0.2
ao
Th# gr«aior coiffoant of
variation, ttw lower rh« long
tarmcvgraQa
.j-

' 1 « 1 1 1 ¦ '
-u
1 1
0.0 0.2 0.4 0.6 0.9 1.0 1.2 1.4 1.6 1.8 2.0
Coefficient of Variation
Figure 5-2b. Long Term Average per Unit Wasteload
Allocation as a Function of the Coefficient of Variation
187

-------
5.3 ENSURING CONSISTENCY WITH THE WASTELOAD ALLOCATION
The WLA provides a measure of effluent quality that is necessary to
protect the water quality of the receiving water. The WLA should be based on
ambient criteria and the exposure of the resident aquatic community or humans
to toxic conditions. Once a WLA has been developed which accounts for all
appropriate considerations, a water quality-based permit limit may be derived to
enforce the WLA.
It is important to consider how the WLA addresses variability in effluent
quality. For example, a WLA for bioconcentratable pollutants is typically
expressed as a single level of effluent quality necessary to provide protection
against long term or chronic effects. On the other hand, a WLA for toxic
pollutants affecting aquatic life (with corresponding duration and frequency
requirements) should describe levels necessary to provide protection against both
short term and long term effects.
5.3.1 Statistical Considerations of WLAs
Direct use of a WLA as a permit limit creates a significant risk that the
WLA will be incorrectly enforced, since effluent variability and the probability
basis for the limit are not specifically considered. For example, a steady state
WLA typically specifies an effluent value with the assumption that it is a value
never to be exceeded. The same value used directly as a permit limit could allow
the WLA value to be exceeded without observing permit violations if compliance
monitoring was infrequent. Confusion can also result in translating a longer
duration WLA requirement (e.g., for chronic protection) into daily maximum and
monthly average permit limits. It is therefore extremely important to ensure that
permit limits are derived to implement a WLA requirement correctly. Potential
problem areas are as follows:
o A WLA value must be enforced in a regulatory context by translation into
daily maximum and monthly average limits; compliance monitoring
associated with permit limitations allows the regulatory authority to
determine whether or not such permit limitations are violated.
o A WLA that assumes the discharge is steady state (i.e., not changing over
time) requires an assumption regarding how the effluent may vary.
o Daily maximum and monthly average limits must be developed so that
they are consistent with each other and reflect the required level of
wastewater treatment facility performance.
The objective is to establish permit limits that lead to the effluent meeting
the WLA under normal operating conditions. It is not possible to guarantee,
188

-------
through permit limits, that a WLA will never be exceeded. However, it is
possible, using the methodology described here to account for extreme values and
establish low probabilities of exceedence of permit limits and the WLA, This is
not to suggest that permit writers should assume a probability of exceedence of
the WLA, but rather, that they should develop limits which will make this a very
small likelihood.
Since effluents are variable, and permit limits are developed based on a
low probability of exceedence, the permit limits should take effluent variability
into consideration and ensure that the requisite loading from the model is not
exceeded under normal conditions. In effect then, the limits must "force"
treatment plant performance which, after taking into consideration acceptable
effluent variability, will only have a low statistical probability of exceeding the
WLA and will achieve the desired loadings.
Figure 5-3 shows a number of important aspects of the relationships
among the various statistical parameters: In this illustration, the most limiting
LTA (after comparing the LTAs derived from both acute and chronic WLAs) has
been chosen. The more restrictive LTA will automatically meet both WLA
requirements. If the effluent "fingerprint" for this LTA (and associated CV) is
projected, it can be seen that the distribution of daily effluent values will not
exceed the acute or chronic wasteload allocations for unacceptable periods of
time. In addition, the duration and frequency requirements of the acute and
chronic criterion for the pollutant or pollutant parameter will not be exceeded.
Finally, this figure also shows permit limits which have been derived from the
more limiting LTA. (Note that for the scenario depicted in Figure 5-3, the daily
maximum limit is lower than the acute WLA and the monthly average limit is
lower than the chronic WLA; This scenario will occur when a 99% probability
basis is used to calculate the LTA and a 95% probability basis used to calculate
the permit limits from a limiting LTA, based on acute protection. However, for
other probability assumptions, these relationships will differ.)
5 .32 Types of Water Quality Models and Model Outputs
Each of the major three types of water quality models and WLA outputs
has specific implications for the subsequent permit limit development process.
1) Single Value from a Steady State Analysis
In general, steady state analyses tend to be more conservative than
dynamic models, since they rely on "worst case" assumptions. Thus, permit
limitations derived from these outputs will generally be more stringent than
limitations derived from dynamic models. Generally, steady state modeling has
been used to calculate only a chronic WLA. Such models produce a single
effluent loading value and no information about effluent variability.
189

-------
Days
Figure 5-3. Relationship between Daily Concentrations, Long
Term Average, Wasteload Allocations, and Permit Limits
190

-------
Such WLAs are typically based upon older State water quality standards
which do not specify levels for both acute and chronic protection, but only include
one level of protection. Such outputs would also be found where a model is
based upon protection from bioconcentration, since only a single long term
ambient value is of concern.
2)	Two Values from Steady State Analysis
Steady state modeling can be improved significantly by performing two sets
of calculations: one for protection against acute effects and one for protection
against chronic effects. Such models must use a water quality criterion which
specifies two levels of protection. These models also include considerations of
mixing zones when developing WLAs to afford two levels of protection. Like the
single value steady state models, these models do not produce any information
about acceptable effluent variability and may require additional calculations to be
translated into permit limits. These models may also be augmented through the
use of biologically-based design flows. These design flows allow consistency
between definitions of acute and chronic criteria exposure durations and
hydrologic averaging periods.
For complex discharge situations (multiple dischargers, complex
environmental factors needing consideration), water quality models and associated
WLAs are typically developed by specialized water quality analysts in the
regulatory authority. However, the permit writer is often required to develop a
water quality model and wasteload allocation prior to permit limit derivation. In
the latter situation, water quality modeling usually consists of simple steady state
dilution models using worst case assumptions.
3)	Dynamic Models
Dynamic models use estimates of effluent variability and the variability of
receiving water assimilation factors to develop effluent requirements in terms of
concentration and variability. The outputs from dynamic models can be used to
base permit limits on probability estimates of receiving water concentrations
rather than worst case conditions. The advantages and disadvantages of various
types of dynamic models are provided in Chapter 4.
It is important to recognize that a long term average and coefficient of
variation can be derived from each type of dynamic model output, but some of
the outputs require some additional manipulation of the data to develop the LTA
and the CV. These parameters are also the starting point for the statistical
permit limit derivation procedures discussed in the next section. Continuous
Simulation models provide an array of effluent data which require further
manipulation to develop an LTA and a CV. Both Monte Carlo and Lognormal
Probabilistic models produce an LTA and CV which can be used directly in
developing permit limitations.
191

-------
As discussed in Chapter 4, steady state assessments should generally be
used where few or no whole effluent toxicity or specific chemical measurements
are available. Modeling should also generally be limited to steady state
procedures where daily receiving water flow records are not available. Two value
steady state models can provide toxicologically protective results and are
relatively simple to use. If adequate receiving water flow and effluent
concentration data are available to estimate their frequency distributions, one of
the dynamic modeling techniques should be used.
5.4 PERMIT LIMIT DERIVATION
There are a number of different approaches currently being used by
permitting authorities to develop water quality-based limitations for toxic
pollutants and toxicity. Differences in approaches are often attributable to the
need for consistency between permit limit derivation procedures and the
assumptions inherent in various types of water quality models and wasteload
allocation outputs. In addition, permitting authorities are also constrained by
legal requirements and policy decisions which may apply to a given permitting
situation.
The purpose of the following discussion is to clearly indicate the
advantages and disadvantages of various approaches. Permitting authorities
should choose procedures which are most appropriate for a particular application
and available information.
5.4.1 Permit Limit Derivation from Single Steady State Model Output
Many WLIas are reported as a single value for effluent quality. An
example of such a requirement is "copper concentration must not exceed 0.75
milligrams per liter (mg/1)." Steady state analyses assume that the effluent is
constant and, therefore, the WLA value will never be exceeded. This presents a
problem in deriving permit limits because permit limits must reflect variability.
The proper enforcement of this type of WLA depends on the parameter
limited. For nutrients and BOD, the WLA value has generally been used as the
average daily permit limit. However, the impact associated with toxic pollutants is
much more time dependent as reflected in the four-day average duration for the
CCC (see Chapter 2). Two options are possible:
Option 1
o Consider the single WLA to be the chronic WLA and derive an LTA for
this WLA using the procedures in Box 5-1 (steps 1 and 2).
192

-------
o Derive Daily Maximum and monthly average permit limitations using the
procedures in Box 5-1 (step 4)
The principal advantages and disadvantages of this procedure are similar to those
for the second permit limit derivation method discussed below, except that it does
not examine two WLAs.
Option 2
o The WLA value for toxic pollutants should be used as the daily maximum
permit limit.
o In the absence of other information, permit writers typically divide the
daily maximum limit by 1.5 or 2.0 to derive a monthly average limit
(depending on the expected range of variability).
The principal advantage of this 2nd option is that this procedure is very straight-
forward in terms of implementation and requires minimal resources. The
disadvantage of this option is that the monthly average limits must be derived
without any information about the variability of the effluent parameter and the
permit writer cannot be sure that these procedures are toxicologically protective.
On the other hand, Option 2 (or a variation of Option 1) is recommended
for addressing situations in which a single criterion is applied at the end of the
pipe and a single monthly sample is contemplated for compliance monitoring
purposes. Use of Option 1 in this case would result in both the monthly average
and the daily maximum limit being in excess of the criterion. (For example, for a
CCC of 1.0 TUc applied as a WLA at the end of the pipe, both the daily
maximum and monthly average permit limit would be 1.6 TUc; assuming CV =
0.6, n = 1, and 99% probability basis.) A discharger could thus comply with the
permit limitation and routinely exceed the criterion. In the alternative, Option 1
could be employed with an assumed number of samples for the monthly average
permit limit derivation.
5.4.2 Permit Limit Derivation from Two Value Steady State Outputs for Acute
and Chronic Protection
A number of WLAs are now being developed with two required results:
acute and chronic requirements. These types of allocations will be developed
more often as States begin to adopt both acute and chronic water quality
standards. These WLA outputs need to be translated into daily maximum and
monthly average permit limits. The following methodology is designed to derive
permit limits to enforce these WLAs.
o An effluent performance level (LTA and CV) that will meet the WLA
requirement is back-calculated. Where two requirements are specified
193

-------
BOX 5-1
Calculating Permit Limits Based on Two-value Wasteload Allocation
To set maximum daily and average
monthly permit limits based on
acute and chronic wasteload
allocations, use the following four
steps:
Convert the acute wasteload
1	allocation to chronic toxic
units.
Calculate the long term
average wasteload that will
2	satisfy the acute and chronic
wasteload allocations.
Determine the lower (more
3	limiting) of the two long term
averages.
Calculate the maximum daily
and average monthly permit
4	limits using the lower (more
limiting) long term average.
Term
Meaning
CV
Coefficient of vanation
a
Standard deviation
WLAa,c
Acute wasteload allocation
in chronic toxic units
WLAa
Acute wasteload allocation

in acute toxic units
WLAc
Chronic wasteload
allocation in chronic toxic
units
TU,
Acute toxic units
TUC
Chronic toxic units
ACR
Acute-chronic ratio
MDL
Maximum daily limit
AML
Average monthly limit
z
z statistic
Step 1
WLAa c (in TUC) - WLAa (in TUa) • ACR
Step 2
LTA
' WLAa c * e
10.5 o2 - z o]
a,c'
where o2 = /n ( CV2 + 1),
z « 1.645 for 95th percentile occurrence probability, and
z = 2.326 for 99th percentile occurrence probabilty
LTAr - WLAr • e
[ 0 5 c42 - z o4 ]
C "-"C
where O42 = In ( CV2 / 4 + 1),
z » 1.645 for 95th percentile occurrence probability, and
z » 2.326 for 99th percentile occurrence probabilty
Step 3
LTA (in TUc ) * min ( LTA,.. LTAa c )
Step 4
| z 0 - 0.5 o2 ]
MDL - LTA • e
where o2 « In ( CV2 + 1)
z - 1.645 for 95th percentile excoedence probability, and
z * 2.326 for 99th percentile excoedence probability
[ z on - 0 5 on2 ]
AML ¦ LTA • e
where on2 ¦ In ( CV2 / n ~ 1)
z • 1.645 for 95th percentile excoedence probability,
z • 2.326 for 99th percentile excoedence probability, and
n * number of samples per month
194

-------
based on different duration periods, two performance levels are back-calculated
(Steps 1 and 2; Box 5-1).
o Permit limits are then derived directly from whichever performance level
is more restrictive (Steps 3 and 4; Box 5-1).
Figure 5-4 presents a flow chart summarizing the various steps in this
procedure. In addition, the equations used in Box 5-1 are based on the
lognormal distribution which is explained in more detail in Appendix E. The
principal advantages of this procedure are described below.
o It provides a mechanism for setting permit limits which will be
toxicologically protective. A steady state WLA uses a single value to reflect the
effluent loading and thus is an inherent assumption that the actual effluent will
not exceed the calculated loading value. If the WLA is simply adopted as the
permit limit, the possibility exists for WLA impacts due to effluent variability.
Clearly, however, effluents are variable. In recognition of this fact, permit limits
are established using a value corresponding to a percentile of the required
probability distribution of the effluent (e.g., 95th or 99th percentile).
o It allows comparison of two independent WLAs to determine which is
more limiting for a discharge: The WLA output provides 2 numbers for
protection against two types of toxic effects; each based upon different
mixing conditions for different durations. Calculation of acute effects are
based upon one hour exposures at critical flow conditions, close to the
point of discharge, or where necessary, at the end of the pipe. Chronic
effects are limited based on four day exposures after mixing at critical flow
conditions. These requirements yield different effluent treatment
requirements that cannot be compared to each other without calculating
the long term average the plant would need to maintain in order to meet
each requirement. Without this comparison (or in the absence of
procedures which address this comparison), the WLA which represents the
more critical condition cannot be determined. A treatment system will
only need to be designed to meet one level of treatment for effluent
toxicity: treatment needed to control the most limiting toxic effect.
o The actual number of monthly samples are factored into permit limit
derivation procedures: The procedure provides the means to accurately
determine the average monthly permit limit based on the number of
observations that will be taken.
Some permit writers have indicated that additional mathematical
calculations associated with these procedures increase the burden for the permit
writer and add what is perceived to be an unnecessary step. However, as
discussed under advantages, this procedure provides the most toxicologically
sound approach. To help address the resource burden problem, EPA has
developed tables (see Table 5-1 and 5-2) to be used to quickly arrive at the
195

-------
f
Two-value Steady
State WLA
Figure 5-4. Flowchart for Calculating Permit Limits from
Two-value Steady State Wasteload Allocation
196

-------
necessary values. In addition, some permit authorities have developed programs
available on floppy disks which can be used with a personal computer to readily
compute the necessary information from the appropriate inputs.
An alternative permitting procedure which has been employed by some
permit writers for this type of output is direct application of WLAs as permit
limits: the WLA developed for protection against chronic effects becomes the
monthly average and the acute WLA becomes the daily maximum limit. There
are a number of inherent assumptions in such an approach and which need to be
recognized. These assumptions can prove to be fundamental weaknesses if not
properly accounted for.
Since effluent variability has not been specifically addressed with this
approach, a violation of either limit would entail automatically exceeding a WLA
(Whether actual in-stream impacts were caused under such conditions would
depend upon whether the conditions represented by the worst case input variables
to the model were also occurring at the same time.) By contrast, violations of
limits which were developed using statistical procedures do not automatically lead
to WLA violations since effluent variability is accounted for in deriving LTAs
associated with particular CVs (see Figure 5-3).
In addition, maintaining treatment plant performance at a level sufficient
to achieve one of the limits would not necessarily allow the discharger to meet
the other limit. The Two WLAs are based upon different effect levels and
different duration and frequency assumptions. Using the WLA for acute
protection as the daily maximum permit limit means that there could be
violations of the chronic WLA which would not be seen with monitoring in
connection with the acute WLA. Where the statistical relationship of the
monitoring frequencies to the limits has not been specifically addressed, it may be
much more difficult to distinguish a complying facility from a non-complying
facility.
5.4.3 Permit Limit Derivation from Dynamic Model Outputs
The least ambiguous way that a WLA can be specified is as the required
effluent performance in terms of the LTA and CV of the daily values. When a
. WLA is expressed as such, there is no confusion about assumptions used and the
translation to permit limits. A permit writer can readily design permit limitations
to achieve the WLA objectives. The types of exposure analyses that yield a WLA
in terms of required performance are the continuous simulation, Monte Carlo,
and lognormal probabilistic analyses. The permit limit derivation procedure is as
follows:
197

-------
Table 5-1
Back Calculation of Long Term Average Wasteload
! CV
WLA multipliers
[ 0.5 O*
e
2 -zo]
95th
percentile
99th
percentile
0.1
0.853
0.797
0.2
0.736
0.643
0.3
0.644
0.527
0.4
0.571
0.440
I 0.5
0.514
0.373
Q.6
0.468
0.321
0.7
0.432
0.281
o.a
0.403
0.249
0.9
0.379
0.224
i.°
0.360
0.204
n
0.344
0.187
I 1.2
0.330
0.174
"1.3
0.319
0.162
1.4
0.310
0.153
1.5
0.302
0.144
i 1.6
0.296
0.137
1.7
0.290
0.131
1.8
0.285
0.126
1.9
0.281
0.121
t 2-0
0.277
0.117
acute
I TA	IAII A	[ 0.5O2 - ZO]
LTAa = WLAa*e
where o2 = In [ CV2 + 1 ],
z = 1.645 for 95th percentile occurrence probability, and
2 = 2.326 for 99th percentile occurence probability


WLA multipliers

CV
[ 0.5 0
e
2-zo41


95th
99th
chronic

percentile
percentile



(4-day average)
0.1
0.2
0.922
0.853
0.891
0.797

0.3
0.791
0.715
[ 0.5 - 2 a a 1
LTAC = WLAC • e 4 4J
0.4
0.736
0.643
0.5
0.6
0.687
0.644
0.581
0.527

0.7
0.606
0.481
where o42 - In [ CV2 / 4 + 1 ],
0.8
0.571
0.440
0.9
0.541
0.404
z - 1.645 for 95th percentile occurrence probability, and
1.0
0.514
0.373
z - 2.326 for 99th percentile occurence probability
1.1
0.490
0.345

1.2
0.468
0.321

1.3
0.449
0.300

1.4
0.432
0.281

1.5
0.417
0.264

1.6
0.403
0.249

1.7
0.390
0.236

1.8
0.379
0.224

1.9
0.369
0.214

2.0
0.360
0.204
198

-------
Table 5-2
Calculation of Permit Limits
CV
LTA multipliers
[zo-
e
0.5 o2 ]
95th
percentile
99th
percentile
0.1
1.17
1.25
0.2
1.36
1.55
0.3
1.55
1.90
0.4
1.75
2.27
0.5
1.95
2.68
0.6
2.13
3.11
0.7
2.31
3.56
0.8
2.48
4.01
0.9
2.64
4.46
1.0
2.78
4.90
1.1
2.91
5.34
1.2
3.03
5.76
1.3
3.13
6.17
1.4
3.23
6.56
1.5
3.31
6.93
1.6
3.38
7.29
1.7
3.45
7.63
1.8
3.51
7.95
1.9
3.56
8.26
2.0
3.60
8.55
Maximum Daily Limit
MDL = LTA ~ e
[Zff-O.5 02]
where o2 » In [ CV2 + 1 ],
z = 1.645 for 95th percentile occurrence probability, and
z « 2.326 for 99th percentile occurence probability
Average Monthly Limit
AML = LTA•e
[zon-0.5on2]
where cn2 - In [ CV2 / n +1 ],
z - 1.645 for 95th percentile,
z a 2.326 for 99th percentile, and
n = number of samples/month

LTA multipliers

fl["n
e
0.5 on2 ]
CV
95th
99th

percentile
percentile

n=1 n=»2 n=4 n-10 n»30
n-1 n=2 n=4 n=10 n»30
0.1
1.17 1.12 1.08 1.06 1.03
1.25 1.*8 1.12 1.08 1.04
0.2
1.36 1.25 1.17 1.12 1.06
1.55 1.37 1.25 1.16 1.09
0.3
1.55 1.38 1.26 1.18 1.09
1.90 1.59 1.40 1.24 1.13
0.4
1.75 1.52 1.36 1.25 1.12
2.27 1.83 1.55 1.33 1.18
0.5
1.95 1.66 1.45 1.31 1.16
2.68 2.09 1.72 1.42 1.23
0.6
2.13 1.80 1.55 1.38 1.19
3.11 2.37 1.90 1.52 1.28
0.7
2.31 1 94 1.65 1.45 1.22
3.56 2.66 2.08 1.62 1.33
0.8
2.48 2.07 1.75 1.52 1.26
4.01 2.96 2.27 1.73 1.39
0.9
2.64 2.20 1.85 1.59 1.29
4.46 3.28 2.48 1.84 1.44
1.0
2.78 2.33 1.95 1.66 1.33
4.90 3.59 2.68 1.96 1.50
1.1
2.91 2.45 2.04 1.73 1.36
5.34 3.91 2.90 2.07 1.56
1.2
3.03 2.56 2.13 1.80 1.39
5.76 4.23 3.11 2.19 1.62
1.3
3.13 2.67 2.23 1.87 1.43
6.17 4.55 3.34 2.32 1.68
1.4
3.23 2.77 2.31 1.94 1.47
6.56 4.86 3.56 2.45 1.74
1.5
3.31 2.86 2.40 2.00 1.50
6.93 5.17 3.78 2.58 1.80
1.6
3.38 2.95 2.48 2.07 1.54
7.29 5.47 4.01 2.71 1.87
1.7
3.45 3.03 2.56 2.14 1.57
7.63 5.77 4.23 2.84 1.93
1.8
3.51 3.10 2.64 2.20 1.61
7.95 6.06 4.46 2.98 2.00
1.9
3.56 3.17 2.71 2.27 1.64
8.26 6.34 4.68 3.12 2.07
2.0
3.60 3.23 2.78 2.33 1.68
8.55 6.61 4.90 3.26 2.14
199

-------
o The permit limit derivation procedures described in Box 5-1, Step 4 are
used to derive daily maximum and monthly average limits from the
required effluent LTA and CV. Unlike these procedures, however, there
is only a single LTA which affords both acute and chronic protection and
therefore the comparison step indicated in Figure 5-4 and Box 5-1 is
unnecessary.
The principal advantages of this procedure are:
o Provides a mechanism for computing permit limits which are
toxicologically protective: As with the procedure summarized above for
two value steady state WLA outputs, the permit limit derivation
procedures which are used with this type of output take effluent variability
into consideration and derive permit limits from a single limiting LTA and
CV.
o Actual number of samples are factored into permit limit derivation
procedures: As discussed above, this procedure has the same elements as
discussed for the statistical procedures in section 5.4.2.
Concerns with the above procedures are generally the same
as those mentioned above for output type 2. Note, also that the permit
documentation (i.e., fact sheet) will need to clearly explain the basis for the LTA
and CV. In addition, as discussed previously, there are generally greater data
demands associated with dynamic models.
Example permit limit calculations are shown in Box 5-2 for each of the
principal types of permit limit derivation approaches discussed above under
Sections 5.4.1, 5.4.2, and 5.4.3.
5.4.4 Special Permitting Applications
There are special considerations associated with permit development for
certain types of receiving waters, for protection against particular routes of
exposure, and for certain types of discharges. These special situations are
discussed below.
Marine and Estuarine Permitting
Water quality-based permit development for discharges to marine and
estuarine waters follows the same basic steps as the water quality-based approach
for freshwater discharges. There are some differences, however, in the water
quality criteria used as the basis for protection, the designation of mixing zones,
and the water quality models used to develop wasteload allocations. (See
discussions of these elements in previous chapters.) In addition, there are some
special regulatory considerations associated with these types of dischargers,
including reviews of permits in conjunction with the Coastal Zone Management
200

-------
BOX 5-2
Sample Calculations of Permit Limits from Different Wasteload Allocation Data
Available data

Two-value wasteload
Dynamic model
Single wasteload

allocation
output
allocation
Wasteload Allocation (WLA)
—
—
14.3
Acute Wasteload Allocation (WLAa)
2.60
—
—
Chronic Wasteload Allocation (WLAC)
14.3
—
—
Acute-Chronic Ratio
4.62
—
—
Coefficient of Variation (CV)
0.8
0.8
0.8
Number of Samples per Month (n)
4
4
4
Long Term Average (LTA)
—
9.44
—
From two-value ateady state wasteload allocation
WLAac
. WLA, • ACR - 2.60 • 4.62
- 12.0
LTA,.
-WLAg-e'0 5 02 "232601 - 14.3 • 0.440 (from Table 5-1)
-6.29
LTAac
MDL
- WLA^c«e'0 5 042 2-328 °4' - 12.0 • 0.249 (horn Table 5-1)
12.320 0-0.5 o2]
- LTA^C • e - 2.99 • 4.01 (from Table 5-2)
.2.99
- 12.0
AML
12.326 o« • 0.5 o_21
-LTA-.-e - 2.99 • 2.27 (from Table 5-2)
-6.79
From dynamic model output
MDL
1T. 12.326 0-0.5 
-------
Program (CZMP). Some discharges also require an Ocean Discharge Criteria
Evaluation under section 403(c) of the Clean Water Act.
Permitting for Human Health Protection
Permit development to protect against certain routes of exposure is
another key consideration. Ingestion of contaminated fish and shellfish is a toxic
chemical exposure route of serious potential human health concern for which
there is no intervening treatment process, unlike the drinking water route of
exposure. Effluent limits designed to meet aquatic life criteria for individual
toxicants and whole effluent toxicity are not necessarily protective of toxic
pollutant residue formation in fish or shellfish tissue.
Developing permit limitations for bioconcentratable pollutants is somewhat
different from setting limitations for other pollutants because the averaging
period is generally longer than one month, and can be up to 70 years. Since
compliance with permit limitations is normally determined on a daily or monthly
basis, it is necessary to set permit limitations that meet a given WLA for every
month. If the procedures described above for aquatic life protection were used
for developing permit limitations on bioconcentratable pollutants, both daily
maximum and monthly average permit limits would exceed the WLA necessary to
meet instream criteria. Thus, even if a facility was discharging in compliance with
permit limits calculated using these procedures, it would be possible to always
exceed the WLA. This approach is clearly unacceptable.
The recommended approach for setting water quality-based limitations for
human health protection with statistical procedures is as follows:
o Set the monthly average limit equal to the WLA.
o Calculate the daily maximum limit based on effluent variability and the
number of samples per month using the multipliers provided in Table 5-3.
This approach ensures that the instream criteria will be met over the long term
and provides a defensible method for calculating a maximum daily permit limit
5.4.5 Other Approaches
There are other valid approaches for translating WLA outputs into permit
limitations. These methods typically combine appropriate elements of the
statistical procedures discussed above with specific technical and policy
requirements of the permitting authority to derive limitations which are protective
of water quality and consistent with the requirements of the WLA Such
approaches utilize simplified statistical procedures.
202

-------
Table 5-3 Multipliers for Calculating Maximum Daily Permit Limits from
Average Monthly Permit Limits
To obtain the maximum daily permit limit for a bioconcentratable pollutant,
multiply the average monthly permit limit (the wasteload allocation) by the
appropriate value in the following table.
Each value in the table is the ratio of the maximum daily permit limit, MDL, to
the average monthly permit limit, AML, as calculated by the following
relationship derived from step 4 of the statistically-based permit limit
calculation procedure (see Box 5-1).
MDL ^[10-0.50=1
AML I 05 On*]
e
where orn2 ¦ In (CV2 / n +1)
¦> /n( CV2 + 1)
CV « the coefficient of variation of the effluent concentration
n - the number of samples per month
z - 1.645 for 95th percentile exceedence probability, and
z • 2.326 for 99th percentile exceedence probability

Ratio between average monthly and maxirrum daily permit limits
CV
95th
99th

percentile
percentile

n»l rt»2 n-4 n-8 n-30
r>-1 r>—2 n«4 n-10 n-30
0.1
1.00 1.05 1.08 1.11 1.14
1.00 1.07 1.12 1.16 1.20
0.2
1.00 1.09 1.16 1.21 1.28
1.00 1.13 1.24 1.32 1.43
0.3
1.00 1.12 1.23 1.31 1.42
1.00 1.19 1.36 1.49 1.67
0.4
1.00 1.15 1-29 1.40 1.56
1.00 1.24 1.46 1.66 1.92
0.5
1.00 1.17 1.34 1.48 1.68
1.00 1.28 1.56 1.81 2.18
0.6
1.00 1.19 1.38 1.55 1.79
1.00 1.31 1.64 1.95 2.43
0.7
1.00 1.20 1.40 1.60 1.89
1.00 1.34 1.71 2.08 2.67
0.8
1.00 1.20 1.42 1.64 1.98
1.00 1.35 1.76 2.19 2.89
0.9
1.00 1.20 1.43 1.68 2.04
1.00 1.36 1.80 2.27 3.09
1.0
1.00 1.20 1.43 1.68 2.10
1.00 1.37 1.83 £34 3.27
1.1
1.00 1.19 1.43 1.68 2.14
1.00 1.37 1.84 2.39 3.43
1.2
1.00 1.18 1.42 1.68 2.17
1.00 1.36 1.85 2.43 3.56
1.3
1.00 1.17 1.41 1.68 2.19
1.00 1.36 1.85 2.45 3.68
1.4
1.00 1.17 "1.39 1.67 2.20
1.00 1.35 1.84 2.46 3.77
1.5
1.00 1.16 1.38 1.65 2.20
1.00 1.34 1.83 Z46 3.84
1.8
1.00 1.15 1.36 1.63 2.20
1.00 1.33 1.82 £46 3.90
1.7
1.00 1.14 1.35 1.61 2.19
1.00 1.32 1.80 £45 3.94
1.8
1.00 1.13 1.33 1.59 2.18
1.00 1.31 1.78 £43 3.97
1.9
1.00 1.12 1.31 1.57 2.16
1.00 1.30 1.76 2.41 3.99
2.0
1.00 1.11 1.30 1.55 2.14
1.00*. 1.29 1.74 2.38 4.00
203

-------
For example, for an assumed value for the CV, there is a corresponding
acute to chronic ratio, above which, the chronic WLA will always be more
limiting. Where such procedures are used, the need to compare LTAs derived
from acute and chronic steady state models would be avoided. Similarly, for
assumed values for n, CV, and exceedence probability, the various equations
shown in Box 5-1 can be further simplified, such that the monthly average limit
will always be a constant fraction of the daily maximum limit.
Such approaches allow the permit writer to rapidly and easily translate the
results of WLAs into permit limits. However, the permit writer should clearly
understand the underlying procedures and will need to carefully explain the basis
for the permit limit derivation process in the permit documentation. Appropriate
State or Regional guidance documents should also be referenced.
Recommendations
For the majority of permitting applications, EPA recommends that the
statistical permit limit derivation procedures discussed in section 5.4.2 and
section 5.4.3 (or appropriate variations of these methods as described above) be
used. Although there are advantages and disadvantages associated with each of
the procedures, EPA feels that the recommended procedures will result in the
most defensible and protective permit limits.
5J SPECIAL CONSIDERATIONS IN USE OF STATISTICAL PERMIT
LIMIT DERIVATION TECHNIQUES
The following is a summary of the effect of changes in the various
statistical parameters on the permit limits which are derived. An understanding
of these relationships is important for the permit writer. Additional
considerations of each of these parameters with respect to the statistical methods
for permit limit derivation are also discussed below.
5.5.1 Effect of Changes on Statistical Parameters on Permit Limits
o Effect of Changes in CV on derivation of LTA from WLA: As the CV
increases, the LTA decreases; and conversely, as the CV decreases, the
LTA increases. (See Figure 5-5.)
Reason: The ITA must be lower relative to the WLA to account for the
extreme values observed with high CVs. LTAs for data sets with a
relatively small amount of variability will be much closer to the WLA
o Effect of Changes in CV on Derivation of Permit Limits for a Fixed
Probability Basis: As the CV increases, the permit limits increase
204

-------
(become less stringent); and conversely, as the CV decreases, the permit
limits decrease (become more stringent). (See Figure 5-6.)
Reason: A higher value for the permit limit is produced
for the same LTAs as the CV increases in order to allow for fluctuations
about the mean. (Note, however, that the LTA will already have been
reduced to account for such fluctuations when the steps in Box 5-1 are
followed.)
o Effect of Changes in Number of Monthly Samples on Permit Limits: As
the value for "n"increases in the monthly average permit limit derivation
equations, the monthly average permit limit decreases to a certain point.
The effect on the monthly average limit is minimal for values of n greater
than approximately 10. Conversely, as the value for "n"decreases, the
monthly average permit limit increases until n = 1, at which point the
monthly average permit limit equals the daily maximum limit. (See Figure
5-7.)
Reason: As the number of observations (n) increases, the probability
distribution of the n-day average values becomes less variable (narrower)
around the LTA. Therefore, the 95th or 99th percentile value for an n-
day average deceases in absolute value as n increases. (See additional
discussion below in Section 5.53.)
o Effect of Changes in Probability Basis for Permit Limits: As the
probability basis for the permit limits expressed in percentiles (e.g., 95%,
99%, etc.) increases, the value for the permit limits increases (e.g.,
becomes less stringent). The converse is true as the probability basis
decreases. (See Figure 5-6.)
Reason: There is a higher probability that any randomly chosen effluent
sample will be in compliance with its permit limitations, if those limits are
statistically designed to be greater than a high percentage (e.g., 99%) of all
possible values for a given LTA and CV.
205

-------
.0	0.5	1.0	1.5	2.0
Coefficient of Variation
Figure 5-5. Long Term Average As a Function
of The Coefficient of Variation

-------
Coefficient of Variation
Figure 5-6. Maximum Daily Permit Limit as a Function of the
Coefficient of Variation
207

-------
i.2 r
1.0
0.8
0.6
0.4
0.2
0.0
0.0
The greater the number of
samples per month, the lower
the average monthly permit limit
95th percentile -
99th percentile
10	20
Number of Samples per Month
30
Figure 5-7. Relationship between Average Monthly Permit Limits and
Number of Samples per Month
208

-------
5.5.2 Coefficient of Variation
Use of the statistical method of permit limit derivation requires an
estimate of the coefficient of variation (CV) of the distribution of the daily
measurements of the parameter after the plant complies with the requirements.
If variability is mostly related to production, current data may be used to estimate
the CV. If future variability is expected to be substantially different, the CV must
be estimated. Discharges of toxic pollutants are generally more variable than
discharges of conventional pollutants. It is important to use an estimate of the
CV that is as low as can be reasonably achieved.
One concern with respect to using an appropriate CV in the statistical
limit derivation procedures is that CVs of regulated systems may be quite
different from nonregulated system. In other words, after permit limitations are
in place and the permittee is forced to operate in such a way as to achieve the
requisite limits, the variability associated with the parameter of concern may
change considerably. Where the permit writer has reason to believe that the CV
of the regulated system may behave differently from the non-regulated system
(e.g., where changes in the treatment facility are planned), information concerning
effluent concentration means and variability can be obtained from Effluent
Guideline documents for individual chemical parameters.
Variability associated with effluent levels of both individual chemicals and
whole effluent toxicity are difficult to predict for any individual situation.
However, it is important to recognize that failure to assign any CV to an
individual toxicant or the parameter toxicity involves an implicit assumption that
there is no effluent variability present. Based upon analyses of a wide variety of
data from various types of plants, EPA recommends a value of 0.6 as a default
CV, if the regulatory authority does not have any more accurate information on
the CV for the pollutant or pollutant parameter. Permit limits are usually not
extremely sensitive to small changes in the CV. The value of 0.6 is typical and
represents a reasonable degree of relative variability. However, wherever
possible it is recommended that data on effluent variability for the pollutant of
concern be collected to define a CV rather than selecting a default value.
5.53 Number of Samples
The statistically-based method for permit limit derivation results in a
maximum daily limit which does not depend on monitoring frequency. However,
the monthly average limit decreases as the monitoring frequency increases and a
greater number for "n" is inserted in the relevant equations. Some permit writers
have expressed concern in connection with this outcome that facilities which are
required to sample more frequently appear to be penalized, while permittees with
less frequent monthly sampling requirements appear to get a "break."
209

-------
The monthly average limit decreases as the number of monthly samples
increases because an average of 10 samples, for example, is closer to the long
term average than an average based on 4 samples. This phenomenon makes
monthly average limits based on 10 samples appear to be more stringent than the
monthly limit based on 4 samples. The stringency of these procedures, is
constant across monitoring frequencies because the probability basis is the same
regardless of the number of samples taken. Thus, a permittee performing
according to the long term average and variability associated with the wasteload
allocation will, in fact, meet either of these monthly average limits when taking
the corresponding number of monthly samples.
For water quality-based permitting, effluent quality is determined by the
underlying distribution of daily values which is determined by the long term
average associated with a particular WLA and by the CV of the effluent
concentrations. Increasing or decreasing monitoring frequency does not affect
this underlying distribution or treatment performance which should, at a
minimum, be targeted to comply with the values dictated by WLA. It is therefore
recommended that the actual planned frequency of monitoring normally be used
to determine the value of n for calculating the monthly average limit.
5.5.4 Probability of Exceedence
Selection of the probability of exceedence for use in the equations in Box
5-1 is primarily a policy decision. Where a permitting authority does not have a
specific recommendation for an exceedence probability, EPA recommends the
following:
o Daily Maximum: .01 exceedence probability (99th percentile confidence
level)
o Monthly Average: .05 exceedence probability (95th percentile confidence
level)
These levels have been used historically in connection with development of the
effluent limitations guidelines and have been well accepted. It is important to
note that these levels are statistical probabilities used as the basis for developing
limitations. The limitations determined in this manner represent values that a
well operated plant should be capable of achieving at all times. In practice, a
well operated treatment facility will take action to avoid or mitigate extreme high
values. In any event, enforcement discretion is exercised by regulatory authorities
to distinguish minor infrequent violations of permit limits from more serious
violations.
210

-------
5.6 PERMIT DOCUMENTATION
It is extremely important that the fact sheet and supporting documentation
accompanying the permit clearly explain the basis and the rationale for the
permit limitations. When the permit is in the draft stage, the supporting
documentation will serve to explain the rationale and assumptions used in the
deriving the limits to the permittee and the general public. When the permit is
issued, the administrative record for the facility (particularly the fact sheet) will
promote understanding of the permit requirements and facilitate subsequent
compliance with limits. This information will also serve to alert
compliance/enforcement personnel to any special considerations which were
addressed at the time of permit issuance. In addition, the accompanying
documentation will be extremely important during permit reissuance and will
assist the permit engineer in developing a revised permit.
Water quality-based treatment facilities should be designed to achieve a
performance level consistent with the LTA and associated CV. Sections 124.8
and 124.56 of 40 CFR Part 124 require a fact sheet containing "any calculations
or other necessary explanation of the derivation of specific effluent limitations ..."
for every draft permit. Accordingly, the WLA along with the required LTA and
CV, where available, should be included in the fact sheet. Where a permit limit
derivation method was used which does not explicitly account for these statistical
factors, any underlying assumptions which implicitly account for these factors
should be explained in the permit documentation. (Where a permitting authority
develops a standardized and simplified method for permit limit development as
discussed in Section 5.4.4, the permitting authority may not need to document all
of the underlying assumptions in the fact sheet -- provided that the fact sheet
references a written permit limit development protocol.)
5.7 EXPRESSING LIMITATIONS AND DEVELOPING MONITORING
REQUIREMENTS
The NPDES regulations (40 CFR 122.45(d)) require that all permit
limitations be expressed as both monthly average and daily maximum values.
Obviously, where only monthly sampling is required, the monthly average limit
becomes the limiting factor. It is extremely important that limitations be clearly
expressed in the NPDES permit so that they are clearly enforceable and
unambiguous. Section 6.2.1 discusses compliance monitoring and enforcement
problems that can result from improperly expressed limitations. All limitations,
both chemical specific and whole effluent, should appear in Part I of the permit.
Special considerations in the use of both chemical specific and whole effluent
toxicity limitations are discussed below.
211

-------
5.7.1 Pollution Prevention/Energy Conservation Considerations
The results of water quality-based permit limit calculations are
concentration-based, end-of-the-pipe limits. Such concentration-based limits do
not provide any incentive to dischargers to reduce wastewater flows. The reverse
is true; a more dilute effluent means water quality-based limits can be achieved
without reducing the mass loading of the pollutant involved.
Increased flow translates into increased power consumption for treatment
facilities. Significant power usage stems from pumping and mixing of volumes of
wastewater in treatment systems. If the volume of wastewater can be reduced,
power consumption can be reduced and less fossil fuel burned. Such reductions
can be expected to result in concomitant decreases in air pollution.
At a minimum therefore, EPA recommends that permit limits on both
mass and concentration be specified where appropriate to discourage permittees
from meeting concentration-based limits through effluent dilution. For example,
a permit limit of 10 mg/1 of cadmium discharged at an average rate 1 mgd would
also contain a limit of 38 kilograms/day of cadmium.
In addition to discouraging meeting concentration-based limits through
dilution, EPA recommends that flow reductions and energy savings be specifically
encouraged, where appropriate, by allowing water quality-based permit limits to
be mass-based and by allowing concentration-based limits to vary in accordance
with flow reduction requirements. The permit could also include an energy
savings analysis subject to approval by the permitting authority.
For example, a permitting authority could allow an effluent to be twice as
concentrated so long as effluent flow is halved. State water quality criteria would
still be maintained provided mixing is allowed and the State allows a small area
around the discharge point to be exposed to a more toxic wastewater prior to
dilution with the remainder of the receiving stream. Obviously, such permitting
strategies may be inappropriate in many water poor ares. In any case, resulting
downstream concentrations must still meet the CCC and CMC.
Specific examples of flow reduction provisions within the context of permit
limits are as follows. If a daily maximum limit for cadmium was 10 mg/1 and the
associated flow was 1 mgd, energy conservation could be encouraged by allowing
a limit of 20 mg/1, provided that the effluent flow was 0.5 mgd. Whole effluent
toxicity limits and associated requirements may also be written in such a way as
to encourage waste minimization and energy conservation. In these situations,
whole effluent toxicity limits expressed in toxic units should be specified in
conjunction with effluent flow requirements. For example, if 10 TUc was a daily
maximum limit associated with an effluent flow of 3 mgd, 20 TUc would be an
allowable daily maximum limit if effluent flows were reduced to 1.5 mgd.
212

-------
5.7.2	Considerations in the Use of Chemical Specific Limits
Detection Limits
A commonly encountered problem is the expression of calculated limits for
specific chemicals where the concentration of the limit is below the analytical
detection level for the pollutant of concern. This is particularly true for
pollutants which are toxic in extremely low concentrations (e.g., dioxin). The
recommended approach for these situations is to include, in Part I of the permit,
the appropriate permit limitation derived from the water quality model and the
wasteload allocation for the parameter of concern, regardless of the proximity of
the limit to the analytical detection level.
However, the limit should also contain an accompanying notation
indicating the specific analytical method which should be used for purposes of
compliance monitoring. The note should indicate that any sample which is
analyzed in accordance with the specified method and found to be below the
detection level will be deemed to be in compliance with the permit limit, unless
other monitoring information (as discussed below) indicates a problem.
The detection level for the analytical method cited in the permit should
clearly defined and quantified. For most NPDES permitting situations, EPA
recommends that the detection level be defined in the permit as the "minimum
level" (i.e., the level at which the entire analytical system gives recognizable mass
spectra and acceptable calibration points). The minimum level is developed
based on interlaboratory analyses of the analyte in the matrix of concern (i.e.,
wastewater effluents). The minimum level should not be confused with the
"method detection level", which is based on single laboratory analysis of the
analyte in distilled water.
Where water quality-based limitations below analytical detection levels are
placed in permits, it is recommended that permit special conditions also be
included in the permit to help ensure that the limitations are being met and that
excursion above water quality standards are not occurring. Examples of such
special conditions include: fish tissue collection and analysis; limitations and/or
monitoring requirements on internal waste streams; and limitations and/or
monitoring for surrogate parameters.
5.7.3	Special Considerations in the use of Whole Effluent Toxicity Limits
Units of Expression and Detection Levels
It is important to understand that it is the permit limit for toxicity itself
and the detection levels associated with the various types of toxicity tests which
determine the type of monitoring requirement which should be specified with the
213

-------
limitation. It is a misconception to think, for example, that only-acute toxicity
tests should be used where the WLA for acute protection is the more limiting
WLA or should always be used 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 is limiting. The limits are derived from the more limiting LTA.
Therefore, either acute or chronic tests might apply to a given situation.
For example, a limit of 5 TUc (NOEC of 20% or greater) would require
chronic toxicity testing where the ACR is 20 for that effluent. An acute test
would not be sensitive enough to measure effluent toxicity in this instance, since 5
TUc would be equivalent to 0.25 TUa. In this situation, an acute test used for
compliance monitoring purposes would need to be specified in terms of "no
statistical difference between 100% effluent sample and control." However, such
tests would not accurately quantify any levels of chronic toxicity present. There is
generally no reason to mix two types of monitoring requirements for the same
limitation. Doing so will confuse the results and complicate assessments of
monthly average limits where sampling frequency is greater than once per month.
Description of Limits
When toxicity limits are used, additional description of the limit is
required. The limit should be stated in Part I as "effluent toxicity" in the
parameter column with "maximum TUs," "minimum LC50," or "minimum NOEC"
in parenthesis underneath. The numerical values should be placed in the
appropriate concentration column followed by TU or a percent sign. A footnote
should direct the reader to Part III, "Additional Conditions." If the monitoring
requirements cannot be fitted into the columns in Part I, a reference such as "see
Part III" should be used.
The further description in Part III should accomplish the following:
o Explain how the limit is expressed (e.g., the limitation is the minimum
LC50 expressed as percent effluent, or the limitation is the maximum
acute toxic units [TUa])
o Specify the test organism and the test methods for compliance monitoring
purposes.
o Describe any special reporting or follow-up requirements (e.g.,
requirements to conduct a toxicity reduction evaluation).
The language in Part III should be modified as needed to suit the
situation. The following language is provided only as an example:
o The effluent toxicity limitation contained in Part I is the allowable chronic
toxicity to the most sensitive of three test species. It is expressed as the
allowable NOEL in percent effluent. The required test species and the
214

-------
procedures to follow are described in Short Term Methods for Estimating
the Chronic Toxicity of Effluents and Receiving Waters to Freshwater
Organisms, EPA/600/4-89/001, March, 1989.
o The permittee shall conduct monitoring of effluent toxicity once per
month. One 24-hour composite sample shall be collected and tested within
24 hours of collection. Results shall be reported as the NOEC. Any test
that does not meet quality control requirements as described in the above
referenced methods shall be repeated using a freshly collected sample as
soon as practicable
Ensuring consistency between whole effluent toxicity limitations and required
tests
The permit writer needs to ensure that the type of monitoring which is
specified is consistent with the assumptions inherent in the permit limitations.
Questions which arise include the following:
o Are the sample duration and frequency requirements associated with the
monitoring requirements consistent with the way the effluent limitation is
expressed?
o Can individual measurements be averaged to track conformance with the
monthly average limitation?
An important consideration in the use of whole effluent toxicity limits is
the relationship of the permit limitations, "daily maximum" and "monthly average"
to the specific testing protocols which will be required as a result of the
monitoring requirement. For the parameter toxicity, the time period implied by
the limitations may not precisely correspond to the requirements of the test
protocols. For example, a chronic toxicity test which is used to determine
compliance with a daily maximum limit would actually consist of samples from at
least three different days over a 7 day period. Acute tests, which can be
conducted in shorter time periods would more closely resemble the literal
meaning of daily maximum.
In any event, to address this situation, it is recommended that the permit
contain a notation indicating that, for the purposes of determining compliance
with the parameter toxicity, "daily maximum" shall be interpreted as signifying a
"maximum test result" for the month (whichever type of test is specified).
With respect to monthly average limits, toxicity test results (in toxic units) should
be averaged and reported in a manner similar to other parameters.
5.7.4 Selection of Monitoring Frequencies
There are a host of factors which need to be considered when establishing
monitoring requirements associated with permit limits. These factors include:
215

-------
o Type of treatment facility
o Environmental significance of the pollutant or pollutant parameter
o Cost of monitoring relative to the discharger's capabilities and benefit
obtained
o Compliance History
o Whether the number of monthly samples was used in developing the
permit limit (as discussed in Section 5.4.2)
Based upon an array of data analyzed for both individual chemicals and
whole effluent toxicity and, independently of cost, EPA has observed that 10 or
more samples per month provides the greatest likelihood that average of the
various monthly values will approach the monthly long term average value.
Ultimately, however, selection of monitoring frequencies will need to consider the
above factors and arrive at a reasonable compromise of the appropriate
considerations.
5.8 TOXICITY REDUCTION EVALUATIONS
Where monitoring indicates unacceptable effluent toxicity, the principal
mechanism for bringing a discharger into compliance with a water quality-based
whole effluent toxicity requirement is a toxicity reduction evaluation (TELE). The
purpose of a TRE is to provide the discharger with the opportunity to investigate
the causes and identify corrective actions for difficult effluent toxicity problems.
The permitting authority will require, either as a permit condition or through a
separate notice to the permittee, that the permittee conduct a TRE in those cases
where the discharger is unable to adequately explain and immediately correct any
exceedance of a whole effluent toxicity permit limit or requirement.
A TRE is a site specific study conducted in a step-wise process to narrow
the search for effective control measures for effluent toxicity. TREs are designed
to identify the causative agents of effluent toxicity, isolate the sources of the
toxicity, evaluate the effectiveness of toxicity control options, and then confirm
the reduction in effluent toxicity. The ultimate objective of a TRE is for the
discharger to achieve the limitations or permit requirements for effluent toxicity
contained in the permit and thereby attain the water quality standards for
receiving waters.
The requirement for a permittee to conduct a TRE may be written into
the Special Conditions section of a permit which contains whole effluent toxicity
limits. In some cases, the permit issuing authority may also utilize other legally
216

-------
binding mechanisms including 308 letters, Administrative orders, or Consent
Decrees to require a TRE.
5.8.1 TRE Guidance Documents
To assist permittees in conducting TREs and achieving compliance with
whole effluent toxicity limits, EPA has developed a series of three guidance
documents:
1)	Generalized Methodology for Conducting Industrial Toxicity Reduction
Evaluations (EPA/600/2-88/070)
2)	Toxicity Reduction Evaluation Protocol for Municipal Wastewater
Treatment Plants (EPA/600/2-88/062)
3)	Methods for Aquatic Toxicity Identification Evaluations
Phase I Toxicity Characterization Procedures (EPA/600/3-88/034)
Phase II Toxicity Identification Procedures (EPA/600/3-88/035)
Phase III Toxicity Confirmation Procedures (EPA/600/3-88/036)
These guidance documents describe the methods and procedures for
conducting TREs and Toxicity Identification Evaluations (TIEs). They are based
on the results of EPA's continuing efforts in TRE methods research and case
study applications. Separate TRE guidance has been developed for industrial
dischargers and municipal wastewater treatment plants to better address the
circumstances of each type of facility. Procedures for the characterization,
identification, and confirmation of the causative agents of effluent toxicity have
been developed and are described in a three-phased TIE methods manual. These
TIE methods are applicable to both industrial and municipal effluents and are an
integral part of the protocols for TREs described in the industrial and municipal
TRE guidance documents.
5.82 Recommended Approach for Conducting TREs
To ensure the successful completion of a TRE, the guidance documents
recommend a systematic, step-wise approach which narrows down the possible
causes or sources of toxicity until a solution or control method is determined.
The guidance documents discourage "playing hunches" or implementing extensive
control measures solely on the basis of unsubstantiated conclusions (e.g. selecting
and implementing a treatment plant upgrade without adequate information).
Experience shows that unnecessary delays and expenditures in achieving the
objective of the evaluation are avoided by building a sound scientific and
engineering basis for selection of a control method. This can best be done by the
logical interpretation of the information and data collected in a systematic
approach to a TRE. The solutions or control methods identified should then go
through a confirmation stage. This is especially important in cases where the
control method selected requires the construction of additional treatment. A
217

-------
flowchart, generalized from the guidance documents, for this approach to TREs is
presented in Figure 5-9. The steps in this flowchart can be summarized as
follows:
o Determination of TRE objectives and development of TRE plan
Obviously the success of any study is dependent on a clear understanding
of what is to be achieved and how these objectives are to be demonstrated and
measured. Typically, TRE objectives are set by the regulatory authority in terms
of a toxicity test endpoint (LC50 or CTE) in order to meet a limit or permit
condition. TRE plans should be submitted by the discharger 30-60 days following
notification that a TRE is required. These plans are important for ensuring that
the TRE objectives are well understood and that the TRE to be conducted is
thorough and represents a reasonable effort to achieve the required reduction in
effluent toxicity. An implementation schedule should also be developed which
describes the time frame for completion of the specific components of the TRE
plan by the required TRE completion date. This schedule should be submitted
for review in conjunction with the TRE plan. EPA recommends that the TRE
schedule should be set or approved by the regulatory agency. This approval of
the schedule and the completion date should not imply approval of the TRE plan
itself or the procedures and methods outlined in the plan. Instead the TRE plan
should only be reviewed and any comments provided to the permittee as needed.
218

-------
Figure 5-9 - Generalized TRE Flowchart
219

-------
To assist in this review, evaluation criteria for TRE plans are provided in
Box 5-2. The permitting authority should review the TRE plan and inform the
discharger of any apparent shortcomings or potential problems. The TRE should
not be delayed pending completion of the review of the plan. The specified
completion date for the TRE must still be met and the permittee should be
expected to begin steps to investigate and alleviate the effluent toxicity evaluation
as soon as possible following notification that a TRE is required. During the
course of the TRE, the regulatory agency should provide oversight, as time
permits, to make the TRE as effective as possible.
o Evaluation of existing site specific information
The next step involves the collection of any information and analytical data
relevant to the effluent toxicity. The permittee should begin collection and
evaluation of this information as soon as possible following notification that a
TRE is required. In some cases, this step may be conducted concurrently with
accelerated toxicity testing as part of the development of a TRE plan. For an
industrial discharger, this part of the evaluation would include information such
as plant and process information, influent and effluent physical and chemical
monitoring data, effluent toxicity data, and material use. For a POTW, additional
information such as industrial waste survey applications, local limits compliance
reports, and monitoring data should be collected. This information is used to
supplement the data generated in the later steps of the TRE and may be useful
at that stage to point to potential sources or treatment options.
o Evaluation of facility operations and maintenance practices
This part of the evaluation is performed in order to ascertain whether the
facility is consistently well operated and whether the effluent toxicity is the result
of periodic treatment plant upsets, bypass or some other operational deficiency
that may be causing or contributing to the effluent toxicity. This part of the TRE
should be initiated immediately after notification that a TOE is required.
Alternatively, the
permittee may begin to conduct this step at the same time that any accelerated
toxicity testing is required. At both municipal and industrial facilities, this step
would involve the evaluation of "housekeeping",treatment system operation, and
chemical use. In some cases, Best Management Practices (BMPs) may be
identified which would improve operations and effluent quality. However, the
effectiveness of BMPs in reducing effluent toxicity should be carefully confirmed
and it will usually be necessary to test a number of samples and perhaps to
conduct Phase I of the TIE to develop this level of certainty. The results of this
evaluation may lead to preliminary strategies for source reduction and pollution
prevention. TTiese might include spill or leak prevention, improvements in
material handling and disposal practices, or substitution or re-use of a compound
known to be highly toxic.
220

-------
Box 5-2
Evaluation Criteria for TRE Plans
o Are the objectives or targets of the TRE clearly and accurately stated?
o Are the schedule and milestones for accomplishing the tasks described in the study
plan?
o Are the final TRE report, progress reports, and meetings with the regulatory
authority included as part of the schedule?
o Are the approaches or methods to be utilized described to the extent that is
possible prior to beginning the TRE?
o Has available EPA guidance been utilized in the design of the TRE and the
development of the TRE plan (or if other methods are proposed, are these
sufficiently documented)?
o Does the TRE plan specify what results and data are to be included in the interim
and final reports?
o Does the TRE plan provide for arrangements for any inspections or visits to the
facility or laboratory which are determined by the regulatory authority to be
needed?
o Are the toxicity test methods and endpoints which will be used described or
referenced?
o Does the approach described build on the previous results and proceed by
narrowing down the possibilities in a logical progression?
o Does the plan provide for all test results to be analyzed and used to focus on the
most effective approach for any subsequent source investigations, treatability studies,
and control method evaluations?
o Are optimization of existing plant/treatment operations and spill control programs
part of the initial steps of the TRE?
o Does the TRE plan allow a sufficient amount of time and appropriate level of effort
for each of the components of the study plan?
o Does the TIE use broad characterization steps and consider effluent variability?
o Is toxicity tracked with aquatic organism toxicity tests throughout the analyses?
o Is the choice of toxicity ter :s for the TRE logical and will correlations be conducted
if the species used are different from those used for routine biomonitoring?
o Is the laboratory analytical capability and the expertise of the investigator broad
enough to conduct the various components of the evaluation?
221

-------
o Toxicity Identification Evaluation (TIE)
TIE procedures are performed in three phases: characterization,
identification, and confirmation. In each phase, aquatic organism toxicity tests
are used to track toxicity at each step of the procedure. In the toxicity
characterization phase, the general nature of the causative agents of effluent
toxicity or toxicants is determined. This is done by conducting a battery of tests
to characterize the physical/chemical characteristics of the toxicity: solubility,
volatility, decomposability, complexibility, filterability, and sorbability. This
information can then be used for deciding what chemical analytical methods will
be used in Phase II or it can be used for designing treatability studies.
The results of Phase I may also be used to provide additional confirmation
of the effectiveness of any BMP which was implemented in the previous step of
the TRE to reduce the effluent toxicity. This would require conducting at least
one Phase I analysis prior to implementation of the BMP (i.e. any source control
method implemented as a result of the evaluation of facility operation and
maintenance). The results of this analysis would then be compared with Phase I
results from samples taken after BMP implementation.
In Phase II of the TIE, the results of Phase I are built upon and the TIE
proceeds to chemical analyses designed to identify the specific chemicals which
are causing effluent toxicity. In Phase III, the toxicants which have been
identified are confirmed by utilizing a number of procedures including correlation
of toxicity with chemical concentration, spiking experiments, toxicity mass balance
and the use of additional test species and their symptoms.
The current version of the TIE methods utilizes acute toxicity tests to
characterize and identify the toxicants. In some cases, these methods may also be
used for TREs where the objective is to reduce chronic toxicity. However, in
order for these methods to be applicable, there must be some measurable acute
toxicity in the effluent samples which are to be characterized in Phase I and
analyzed in Phase II. If this approach is utilized, the appropriate chronic toxicity
test, as specified in the TRE objectives and permit requirements, should then be
used in the Phase III confirmation procedures. In this way a weight of evidence
can be developed which confirms that the toxicant(s) identified using acute tests
in Phases I and II, are indeed causing the whole effluent chronic toxicity which
must be reduced. Also, it is important to note that it is possible to use the
methods and procedures described in the other components of the overall TRE
with either acute or chronic toxicity tests. The fact that the current version of the
EPA TIE methods utilize acute toxicity tests should not be construed to mean
that TREs cannot be required or conducted for the reduction of chronic toxicity.
TIE methods which are designed specifically to utilize chronic toxicity tests for
the characterization and identification of effluent toxicants are being developed
by EPA. These methods will provide additional tools to assist permittees in the
reduction of whole effluent chronic toxicity.
222

-------
o Source investigation
Based on the results of the TIE, a decision is made on whether to conduct
treatability studies on the final effluent and/or conduct a source investigation. A
source investigation is most readily performed when the specific toxicants have
been identified and influent samples can be analyzed for the presence of these
compounds, or potential source streams can be selected for chemical analysis
(based on the results of the initial data acquisition step). However, in some cases
where the specific causative agents of effluent toxicity have not been identified in
the TIE, it may be possible to conduct a source investigation by "treating" influent
samples in bench scale models of the facility treatment plant, measuring the
toxicity of the treated sample, and then proceeding to track this toxicity to its
source.
Source investigations will lead to control methods such as chemical
substitution, process modification, treatment of process or influent streams
(pretreatment) and possible elimination of the process. For POTWs, source
investigations may lead to the development of local limits or to the requirement
that an indirect discharger evaluate and control their effluent so as to reduce its
toxicity and prevent pass-through at the POTW. The implementation of source
control methods can effectively reduce effluent toxicity and also avoid any cross-
media transfer of pollutants to air or sludge which may occur as a result of end of
pipe treatment. T^pes of source control methods that have proven to be effective
in reducing effluent toxicity are improvements in facility housekeeping, chemical
substitution, process optimization, reclamation/re-use, and pretreatment.
o Toxicity treatability evaluation
Toxicity treatability evaluations are conducted to identify possible
treatment methods that can effectively reduce effluent toxicity. These may
involve modifications or additions to the existing system. Treatability studies
generally utilize the same type of information on the nature of the chemicals to
be removed as is generated by Phase I of the TIE. These treatability tests should
be conducted on a bench scale initially and then a pilot scale prior to
construction of additional treatment or substantial modification of the existing
plant. The use of these bench and pilot scale tests, coupled with aquatic
organism toxicity tests, should be used to confirm the effectiveness of the
treatment option. Confirmation of the results of treatability studies is equally
important as it is for the TIE. Skipping this confirmation step is an invitation for
unwarranted expense.
o Toxicity control method selection and implementation
After the investigative steps of the TRE are completed, it is not unusual
for a number of possible control options to have been identified. At this point, a
223

-------
site specific selection must be made by the discharger based on the technical and
economic feasibility of the various alternatives. Following this selection, the
toxicity control method is implemented, or a compliance plan is submitted if
construction of additional treatment requires a substantial amount of time.
o Follow-up and confirmation
After the control method is implemented and the final TRE report is
submitted, the permitting agency should direct the permittee to conduct follow-
up monitoring to confirm that the reduction in effluent toxicity is attained and
maintained. Normally, this monitoring should follow an accelerated schedule,
weekly or biweekly toxicity tests, for a period of two to three months in order to
confirm the effectiveness of the controls implemented and the continued
attainment of the TRE objective. This follow-up monitoring should utilize the
same species as were specified for routine toxicity testing in the permit. The test
endpoints of these toxicity tests should be the same as those which were
calculated by the water quality-based permit limit derivation procedure used when
the permit was issued. Once the discharger has demonstrated the successful
completion of the TRE, the permitting agency should direct the discharger to
return to the routine permit monitoring schedule.
5.8J Circumstances Warranting a TRE
It is the responsibility of the permitting authority to determine if the
permit limits and/or the State water quality standards have been violated and to
notify the permittee if a TRE is to be required. In some cases, it may be
appropriate for the permitting authority to require additional, accelerated toxicity
testing following the initial exceedance or violation. This accelerated toxicity
testing may precede notification that a TRE will be required or it may be
considered as the initial part of the TRE and be conducted simultaneously with
TRE plan development and the evaluation of other existing site specific
information.
It is important to recognize that the purpose of this additional accelerated
toxicity testing is to determine the continued presence or absence of effluent
toxicity and the magnitude of that toxicity. This information can then be used to
determine the continued compliance or non-compliance with the limit or permit
conditions for effluent toxicity. These tests cannot serve to verify or confirm the
previous test results from an earlier sample. Instead, the results of these tests
should be used by the permit authority to determine if a TRE or some other
action is necessary.
If the permit has a limit for whole effluent toxicity, then generally, the
permit should not include specific conditions for accelerated toxicity test results
which would be used to trigger a TRE or some other action (e.g., exceedances in
2 consecutive tests or exceedances in any 3 out of 5 tests). Section 309 of the
Clean Water Act requires that any single violation of a permit limit may be
22^

-------
subject to enforcement. The EPA "Compliance Monitoring and Enforcement
Strategy for Toxics Control" (January 19, 1989, Appendix B) states that, "Each
exceedance of a directly enforceable whole effluent toxicity limit is of concern to
the regulatory agency and therefore qualifies as meeting the violation review
action criterion requiring professional review." Accelerated monitoring should
only be utilized to assist in this professional review. Permit language should state
that the permitting authority will determine what action will be required to return
the permittee to compliance with the limit or toxicity testing permit condition. It
will be necessary for the Region or State permitting authority to make this
determination on a case by case basis. This must be done in a manner consistent
with the priorities established in their respective toxics control strategies and
permitting procedures.
In situations where it is determined that accelerated testing is appropriate,
weekly tests for a minimum period of two months are recommended. This would
result in eight tests, plus the routine monitoring toxicity test which initially
indicated the exceedence or violation, for a total of nine tests in the series. As a
practical approach for determining if a TRE is an appropriate response, EPA
recommends if toxicity is repeatedly or periodically present at unacceptable levels
(i.e. greater than 20% of the time, or two out of the nine tests in the series) a
TRE should be required
In most cases, any one additional exceedance (beyond the initial routine
monitoring toxicity test result) in the accelerated toxicity tests should result in
notification of the permittee that a TRE is required. Deceptions to this guideline
might include cases where the permittee is able to adequately demonstrate that
the cause of the exceedances is known and corrective actions have been
implemented immediately, or cases where additional test quality assurance/quality
control (QA/QC)is necessary or desirable. The submittal of quality control fact
sheets for self-biomonitoring (e.g. Appendix B2) should always be recommended
to avoid QA/QC problems.
If the test results indicate that toxicity is not consistently or repeatedly
present in the test series, previous discharge monitoring reports (DMRs) should
be examined to ascertain if a recurrent problem exists. If the problem is
recurrent, a TRE should be required and the TRE plan should explain how the
design of the evaluation will address this periodic or recurrent effluent toxicity
problem. In these cases, more elaborate sampling design, and influent or process
stream monitoring mav be needed. It should be expected that TREs conducted
under these circumstances will probably require a more flexible schedule and
perhaps additional time before the required completion date.
If the accelerated testing and previous DMRs show the continued absence
of effluent toxicity, then the initial exceedance would be considered an episodic
event and a TRE should not be required. A TRE is not an appropriate response
to a single, episodic effluent toxicity event (e.g. a spill or a plant upset). By
conducting accelerated testing following a violation or exceedance of a permit
225

-------
condition the requirement of unnecessary TREs can be avoided. Similarly,
conducting accelerated testing as part of the initial steps of a TRE will allow for
the TRE to be ended in its very early stages should the toxicity be determined to
be episodic or nonrecurrent. By following the TRE guidance and incorporating
accelerated testing into the TRE, unnecessary analyses and expense can be
avoided.
It is also important to note that for the practical purposes of conducting a
TRE (as opposed to the purpose of determining if a TR£ should be required or
not), the magnitude of the effluent toxicity needed to conduct a TRE may be less
than the magnitude or level set as the permit limit or permit monitoring
condition. This is because if the limit or monitoring condition is water quality-
based then some amount of dilution will usually be incorporated in determining
the unacceptable level of effluent toxicity. In some cases it may be possible for
the TRE procedures to be carried out even if the toxicity does not actually
exceed this permitted level. This will be so as long as the effluent toxicity is
periodically or consistently present in measurable amounts in samples of 100%
effluent.
5.8.4	Mechanisms for Requiring TREs
There are a number of mechanisms that can be used to require a TRE. In
most cases, the TRE should be required through the permit. They can also be
required by a Section 308 letter, or by an enforcement action such as a 309
Administrative Order or a Consent Decree. Permit language for TREs can be
placed in the Special Conditions Section and include a compliance schedule or
TRE completion dates. When this mechanism is used, it is important that the
TRE not be an automatic requirement. The permittee should receive notification
of what response is required from the permit authority. This enables the permit
authority to assess if a TRE is the appropriate action to pursue and avoids
situations where a permittee automatically conducts more than one TRE during
the permit period. If effluent toxicity should reappear following the successful
completion of a TRE, then the permit authority should be able to review this
type of situation in order to determine if an additional TRE is appropriate or if
some other action is required. The permit should be issued with immediately
effective whole effluent toxicity limits in part I of the permit and TRE
requirements should be used where necessary to bring the permittee into
compliance with those limits. Box 5-3 provides model permit language developed
as part of the Whole Effluent Toxicity Basic Permitting Principles and
Enforcement Strategy (Appendix B2).
Permit language for requiring a TRE should be placed in each permit with
a whole effluent toxicity limit. This language should state that a TRE will be
conducted when the permit issuing authority's professional review determines that
a TRE is the appropriate regulatory action for addressing an exceedance of a
toxicity limit. This permit language should not stipulate that a TRE will be
automatic or that any subsequent exceedances will necessarily result in another
226

-------
TRE. Discretion should be maintained for the permit issuing authority to pursue
other enforcement actions if deemed appropriate. For this reason, it is
recommended that no numeric pattern of toxicity should be included in the
permit as an automatic trigger for a TRE requirement. Box 5-4 presents sample
Special Conditions permit language for requiring TREs.
This sample permit language should be tailored to fit the specific
permittee. The completion date should be specified on a case by case basis.
Factors to consider in setting this completion date include the type of facility, the
variability of the effluent and the previous compliance history. In general,
experience to date has indicated that reasonable timeframes for conducting a
TRE are 6-18 months for an industrial discharger and 12-24 months for a
municipal wastewater treatment plant. It should be recognized that extensions to
these initial timeframes may be granted if the progress reports demonstrate that
this is warranted. In situations where reductions in chemical concentrations to
meet chemical specific limits are needed as well as reductions in effluent toxicity,
the timeframes may be adjusted to enable those efforts to move forward
simultaneously.
227

-------
Box 5-3
MODEL PERMIT LANGUAGE FOR EFFLUENT TOXICITY LIMITS
Part IA. Final Effluent Limitations and Monitoring Requirements
During the period beginning on the effective date of this permit and lasting until the
expiration date, the permittee is authorized to discharge in accordance with the following limitations
and monitoring requirements from the following outfall(s): 001.
Effluent Characteristic
Reporting
code/units
Parameter
Discharge Limitation
Concentration
Daily MaxTMonthty
	Average
Monitoring Requirement
Measurement
Daily Monthly Frequency
Average
Sample
Type
	/TUc
Toxicity
1.2
.59
x/month
composite
Part I.B. Reporting Requirements
1. Toxicity Limitations
Where any one monitoring event shows a violation of thelimits in Part I.A of this permit, the
permittee shall be considered in violation of this permit and shall increase the frequency of toxicity
testing to once per week and submit the datawithin x days to the permitting authority. The
permitting authority will determine what action will be required to return the permittee to
compliance; whether the permittee must implement the requirements of Part ILIA of this permit
(TRE), or if the permittee has returned to compliance and may return to the monitoring
requirement in Part LA. The permittee shall use the testing and data assessment procedures
described in Part III.B of this permit.
228

-------
Box 5-4
Part m Special Conditions: Toxicity Reduction Evaluation
The Discharger shall demonstrate that effluent toxicity-based permit limitations described in Part
IA. of this permit are being attained and maintained through the application of all reasonable
treatment and/or source control measures. Upon identifying noncompliance with those limits
following the conditions of Part I.C1., the Discharger shall initiate corrective actions according to
the following schedule:
Task
4.
5.
6.
Take all reasonable measures
necessary to immediately reduce
toxicity
Where source of toxicity is known,
submit a plan and schedule to attain
continued compliance with the
effluent toxicity-based permit
limitations in Part LA., if immediate
compliance is not attained.
Where source is unknown and toxicity
cannot be immediately controlled
through operational changes, submit
a TRE study plan detailing the
toxicity reduction procedures to be
employed. EPA's Toxicity Reduction
Procedures, Phases 1,2, and 3 (EPA-
600/3-88/034,035, and 036) and TRE
protocol for POTWs (EPA- 600/2- 1
88/062) shall be the basis for this
plan and schedule.
Initiate TRE plan.
Comply with approved TRE schedule.
Submit results of the TRE; include
summary of findings, corrective
actions required, and data generated.
Implement TRE controls as
described in the final report.
Complete TRE implementation to
meet permit limits and conditions.
Deadline
Within 24 hours
Within 30 days
Within 45 days
Within 45 days
Immediately upon approval
Per approved schedule
On due date of final report per approved
schedule
Per approved schedule, but in no case later
than	months from initial noncompliance
229

-------
REFERENCES
1.	Aitchinson, J. and Brown, J.A.C. (1963), The Lognormal Distributions.
Cambridge University Press, London.
2.	Gilliom, R.H. and Helsel, D.R. (1986), "Estimation of Distributional
Parameters for Censored Trace Level Water Quality Data 1 and 2", Water
Resources Research 22, 135-155.
3.	Kahn, Henry D. and Rubin, Marvin B. (1989), "use of Statistical Methods
in Industrial Water Pollution Control Regulations in the United States",
Environmental Monitoring and Assessment 12, 129-148.
4.	Shumway, R.H., Azari, A.S., and Johnson, P. (1989), "Estimating Mean
Concentrations Under Transformation for Environmental Data with
Detection Limits". Technometrics 31(3), 347-356.
5.	U.S. Environmental Protection Agency (1987), Development Document for
Effluent Guidelines and Standards for the Organic Chemicals, Plastics and
Synthetic Fibers Point Source Category, Volumes I and II, EPA 440/1-
87/009.
6.	U.S. Environmental Protection Agency (1988), Generalized Methodology
for Conducting Industrial Toxicity Reduction Evaluations (EPA/600/2-
88/070).
7.	U.S. Environmental Protection Agency (1988), Toxicity Reductions
Evaluation Protocol for Municipal Wastewater Treatment Plants
(EPA/600/2-88/062).
8.	U.S. Environmental Protection Agency (1988), Methods for Aquatic
Toxicity Identification Evaluations:
Phase I Toxicity Characterization Procedures
(EPA/600/3-88/034)
Phase II Toxicity Identification Procedures
(EPA/600/3-88/035)
Phase III Toxicity Confirmation Procedures
(EPA/600/3-88/0360
230

-------
6. COMPLIANCE MONITORING AND ENFORCEMENT
6.1 OVERVIEW
Once a water quality-based permit containing limitations and conditions to
control effluent quality is issued, it is the responsibility of the permittee to attain,
monitor, and maintain compliance with the requirements of that NPDES permit
Failure to comply with any requirements stated in the permit is a violation of the
Clean Water Act.
It is the responsibility of the regulatory agency to track compliance with
and enforce NPDES permit requirements in its enforcement of the Clean Water
Act. Section 308 of the Clean Water Act and equivalent State statutes enable the
regulatory agency to verify compliance with permit conditions (including water
quality-based toxics limitations and compliance schedules) by authorizing the
imposition of requirements for sampling and analysis, record-keeping, and
reporting. Section 308 also authorizes access by EPA or State agencies to
facilities and records for the purpose of inspection to verify compliance with
permit conditions. All records associated with monitoring must be maintained by
the facility and available for inspection for a period of three years in conformance
with 40 CFR 122.41.
The authority to enforce water quality-based permit conditions is
established in the Clean Water Act. The ability to enforce water quality-based
permit conditions, however, relies on well-written, clearly stated permits. The
enforcement official must be familiar with the process by which permit
requirements were derived; this includes the procedures used to determine the
waste load allocation based on applicable water quality standards and the
procedures used to derive limitations from the waste load allocation.
6.2 PERMIT REQUIREMENTS
40 CFR Part 122 Subpart C sets forth the conditions that are to be
included in NPDES permits. In general, permits include effluent limitations,
schedules of compliance, and accompanying reporting requirements. Permits
should prescribe the self-monitoring procedures, a frequency of analysis no less
than monthly, sampling location and procedures, acceptable or required analytical
techniques, and frequency of reporting. Permits often require that analytical
methods referenced in 40 CFR 136 be used for analysis, but may specify
methodology not included in Part 136 for pollutants with no approved methods or
where the approved method is inappropriate for a particular permit limitation.
Permits should define any effluent limitations and explain any specific procedures
for calculating averages of data. The permit should identify what information
must be retained by the permittee, and what data must be submitted to EPA or
the State. Self-monitoring results are reported on Discharge Monitoring Reports
231

-------
(DMRs) which are completed monthly and generally required by the permit to be
submitted monthly or quarterly. Sampling and analysis that is done more
frequently than required by the permit must be included in the DMR. Permit
limitations must be expressed in terms of monthly average and either daily
maximum or, in the case of Publicly Owned Treatment Works, weekly average.
6.2.1 Permit Considerations
The procedures used to develop water quality-based toxic limitations for
permits as discussed in Chapter 5 provide for the development of both a monthly
average limitation and a daily maximum or weekly average limitation effective on
a specified date. Limitations that are established by other methods may present
unique problems in the areas of compliance monitoring and enforcement:
o Limits not expressed in terms of monthly average and daily maximum or
weekly average may not clearly specify the number of days represented by
the limit. Thus, the corresponding number of days of violation that can be
alleged in judicial actions is unclear (see Section 5.73).
o Tracking compliance in the Permit Compliance System (PCS) requires
limits expressed in terms of specific values (such as average, maximum,
and minimum) representing discrete time periods (such as monthly,
weekly, and daily).
o Limits must be effective as of a specific date (not triggered based on an
event) to allow proper tracking in PCS. Establishing "trigger levels" (i.e., a
level of toxicity above which a TRE is required) is typically not equivalent
to setting a limitation unless any effluent discharged with a toxicity above
that level is expressly prohibited in the permit; enforcement of such
"triggers" is complicated and may require enforcement of the narrative
standard.
Misunderstandings between the permit writer, enforcement authority, and
permittee can be minimized by well-written permits. For example, if a limit is
framed in terms of a monthly average, it is advisable that the permit state the
method by which numerous results be averaged to calculate the monthly average
value if different from arithmetic averaging. The permittee must be made aware
that, as with all required monitoring, additional samples and analyses may be
conducted to reduce the statistical difference between the averaged test data and
the average effluent quality.
Permit limitations that are not developed or expressed in terms of monthly
average and either weekly average or daily maximum are not in conformance
with 40 CFR 122.45 and should be returned by compliance personnel to the
permit writer for modification. Violations which occur prior to permit
modification must be alleged to represent days of violation comparable to the
frequency of monitoring. For example, monthly monitoring would equate to
232

-------
approximately 30 days of violation. This allegation is based on permit language
requiring that samples be representative of effluent quality. Even if the permit
does not explicitly state this, permittees have the right to conduct more
monitoring than required by their permit if they feel that particular results are
not representative of their effluent quality, or to prove that the violation did not
persist through-out the time-period represented by the required sample.
Permittees should be made aware of the penalty ramifications of relying on
infrequent monitoring that shows noncompliance. Permittees may incorrectly
assume that one noncompliant test result will equate with a statutory maximum
penalty of only $25,000. Permittee response to a violation may be expedited if
the appropriate statutory maximum is known (i.e., $750,000 for each violation of a
monthly average limitation, which is generally equated to 30 days).
Any permit conditions that are crafted as to be virtually unenforceable
should be brought to the attention of the permit writer for prompt modification.
6.3 COMPLIANCE MONITORING
Since most of the routine information gathered in compliance monitoring
results from permittee self-monitoring, quality assurance (QA) is as important as
compliance with limits. It is essential that permittees develop and adhere to a
quality assurance plan consistent with the required monitoring and analyses. If
the permit requires a testing methodology which includes QA procedures, the
permit should require compliance with those procedures. If the methodology
contains inadequate QA procedures, the permit should reference or establish
other QA procedures. The permittee is responsible for maintaining data to
demonstrate compliance with such procedures.
There are three ways of determining compliance with an NPDES permit
and assuring adequate QA: self-monitoring reports, Discharge Monitoring
Report/Quality Assurance results, and inspections. Each of these methods will
be discussed individually. .
6.3.1 Self-monitoring Reports
Self-monitoring reports provide much of the compliance data used by the
regulatory authority in the review of permittee compliance. These reports include
Discharge Monitoring Reports (DMRs) and Reports of Progress on Compliance
Schedules. DMRs contain information on the sampling method, frequency and
location, and analytical results of permittee self-monitoring. These data and data
from progress reports on major schedule milestones must be entered into the
Permit Compliance System (PCS), a computerized data base, by the State or EPA
[1]. When the required data are entered into the system, PCS is capable of
automatically flagging violations of permit limitations, compliance schedules, and
reporting requirements.
233

-------
In order to detect any problems with QA, it is often desirable to obtain
QA information with the self-monitoring data. For this reason, several States and
Regions have developed additional QA forms to accompany permittee self-
monitoring reports. This additional information may be required through the
permit or through a Section 308 order. The QA data are compared to a
reference QA data sheet which can be completed by the regulatory authority to
indicate acceptable ranges of values for the required protocol. An example of a
QA data sheet for a whole-effluent toxicity test is shown in Appendix B-2. Once
completed, this QA data sheet can be included in the compliance file for quick
reference by compliance personnel.
It is important to note that poor QA is a violation if the permit explicitly
specifies adequate QA or references an acceptable protocol with corresponding
QA procedures. It is also important to note that the signatory's certification of
effluent data certifies compliance with the specified protocols. A claim by the
permittee that the self-monitoring data are not complete or accurate due to
incorrect procedures or improper QA is an admission of false certification and
may be reason to contemplate criminal proceedings.
632 Discharge Monitoring Report/Quality Assurance
The Discharge Monitoring Report/Quality Assurance (DMR/QA)
program evaluates a permittee's ability to analyze and report accurate data. This
program is intended to improve overall laboratory analytical performance for self-
monitoring data. Authority for requiring participation is granted in Section 308
of the Clean Water Act. In the DMR/QA program, permittees are required to
analyze "blind" samples with constituents and concentrations that can be found in
their industrial or municipal wastewaters. The permittees' results are compared
to the known content of the sample, and an evaluation of the reported data is
sent to the permittees. Permittees are expected to use the same personnel and
methods employed for reporting NPDES data to analyze the samples; permittees
are also required to follow the instructions for reporting results and include a
signed certification statement in accordance with 40 CFR 122.22.
Regulatory agencies conduct follow up investigations to address poor or
incomplete DMR/QA results, failure to participate, or late submittal of
DMR/QA results. DMR/QA performance results are compiled annually.
In the past, only chemical-specific analyses were testjd in the DMR/QA
program. The Environmental Monitoring and Support Laboratory (EMSL) in
Cincinnati is developing a reference toxicant DMR/QA sample for permittees
with whole-effluent toxicity monitoring requirements. National implementation is
scheduled for 1991.
234

-------
6.3.3 Inspections
Inspections are conducted by the regulatory authority or its contractors to
address specific violations or problems, to prepare for permit reissuance, or to
verify permittee compliance with permit conditions and QA procedures.
Inspections may include reviewing records, inspecting treatment facilities,
assessing progress with compliance schedules, evaluating laboratory facilities and
performance, and collecting samples for analysis or "splitting"samples taken by
the permittee for concurrent analyses. EPA has defined several types of
inspections based on the tasks that are included [2]. Because regulatory
authorities are expected to inspect all major permittees annually, non-sampling
inspections (which are generally less resource-intensive) are encouraged for
routine evaluation of permittee performance. However, sampling inspections are
still encouraged to address permitting and enforcement priorities. For that
reason, it is important that the regulatory agency have the full capability to assess
effluent compliance through in-house resources or contract support.
Inspections that focus on toxics control can provide useful information for
water quality assessment and permit reissuance in addition to compliance data.
Procedures for inspecting facilities with toxicity testing requirements and
measuring effluent toxicity are detailed in Chapter Seven of the NPDES
Compliance Inspection Manual [2].
6.4 VIOLATION REVIEW
Review of permittee self-monitoring data to determine appropriate
enforcement response generally involves a two-tiered review. Tlie first tier is a
preliminary review for timely, complete data that indicates compliance with
permit requirements. Minor violations of requirements are often handled through
informal phone calls or warning letters that do not require extensive review or
oversight. As violations increase in magnitude, duration, or frequency, they are
generally escalated to personnel who are responsible for enforcing the permit
requirements. The guidelines for this escalation are presented in the
Enforcement Management System (EMS) [3], but the basic concepts of
responsible compliance tracking of water quality-based requirements are discussed
below.
When the initial review of effluent monitoring data indicates that
unacceptable analytical methods were used by a permittee or its contract
laboratory, the results should be escalated for review by personnel qualified to
determine the significance of the results. If the monitoring is insufficient to
determine compliance with effluent limitations, a warning letter or Section 308
letter requiring that the tests be repeated using acceptable procedures would be
an appropriate response.
235

-------
Tracking a p
requirements requi
acceptable quality 1
determine if the pt
This second review
qualified to make t
Likewise, m
effects that are eitt
bioconcentrate sho
enforcement perso
up monitoring whi<
sampling of aquati<
require this follow-
permit writer consi
protection is being
Testing and Mater:
with Fishes and Sa
analysis of the exp<
samples to an extn
organisms may be
of their tissue com
outfall. This apprc
effects of a single r
EPA has als
which effluent viol;
responsible for enf
Action Criteria anc
In the case •
no delineation betv
these Violation Re
average limitation
enforcement respoi
period. Seven day
reviewed if a minir
month.
In the case <
reviewed by a qual
States may wish to
chemical-specific li.
of chemicals with p
rmit or Section 308 letter that contains "monitor only"
s both a compliance review (e.g., to determine if results of
;re submitted on time), and an action review (e.g., to
nit should be modified or re-issued to include a limitation).
)ften requires escalation of the results to personnel who are
is regulatoiy decision.
litoring results for chemicals with potential human health
r below the current level of detection or which
d be reviewed by qualified personnel; in this instance,
iel should determine if the permit requires periodic follow-
includes more advanced analytical techniques and tissue
organisms exposed to the effluent. If the permit does not
o monitoring, enforcement personnel should request the
it requiring additional studies to ensure that adequate
rovided. One possible tool is the American Society of
!s Standard Practice for Conducting Bioconcentration Tests
vater Bivalve Mollusks [4]. This document details the
ed organisms for specific chemicals through subjecting tissue
tion procedure and GC/MS or AA analysis. Exposed, sessile
)llected down-stream of the outfall and the chemical analysis
ired to that of sessile aquatic biota collected up-stream of the
ch has the draw-backs of cost and difficulty in isolating the
urce in multiple-source discharge situations.
recommended minimum acceptable criteria for determining
ions must be escalated for review by a professional
cement. These criteria are known as the Violation Review
ire listed in the EMS.
a violation of a chemical-specific permit limitation, there is
en technology-based versus water quality-based limitations in
;w Action Criteria. EPA has recommended that monthly
Nations be reviewed by a professional for potential
2 whenever two or more violations occur in a six month
/erage and daily maximum violations should likewise be
m of two or four respectively occur during the course of one
a whole-effluent toxicity limitation, any violation must be
ed professional responsible for enforcement. Regions and
lopt the "any violation" criteria for water quality-based
tations as well. This may be especially advisable in the case
ential human health effects.
236

-------
6.5 ENFORCEMENT
Effective enforcement of toxic controls depends upon clearly expressed
requirements in NPDES permits. These controls are generally in the form of
numeric limits on specific toxic chemicals or whole-effluent toxicity and schedules
to initiate corrective actions if effluent violations exist.
Regardless of the basis used to develop a permit limitation, exceeding the
limitation is a violation. If any violation occurs, it is the responsibility of the
permittee to inform the regulatory agency. If the violation may endanger health
or the environment, the permittee must submit a noncompliance report within
five days of violation detection. If there is no danger to health or the
environment, the report must be submitted at the time monitoring reports are
submitted. These reports must include a description of the violation, its cause,
the period of noncompliance, and if the noncompliance has not been corrected,
the anticipated time when compliance will be achieved.
Available enforcement mechanisms include Section 308 orders, Section 309
Administrative Orders, Administrative Penalty Orders in conjunction with
Administrative Orders, or judicial action. Enforcement action must be tailored to
the specific violation and type of remedial action required. Enforcement actions
must be carefully worded so that they are clearly understood, easily tracked, and
expeditiously enforced.
As with other NPDES permit limitation violations, violation of a water
quality-based toxics limit should prompt immediate action on the part of the
permittee. Permittee response should include evaluation of the cause of the
violation, correction of operational deficiencies or operational changes, and any
other initial steps necessary to resolve the violation and mitigate the
environmental effects. These immediate investigatory and corrective steps should
also provide information that may be used in developing the compliance schedule
contained in an enforcement document.
When a water quality-based toxicity limit is violated, the regulatory agency
may require additional monitoring to determine the frequency and duration of
the violation. If the permit limit is not immediately met through improved
housekeeping, operation, or raw waste control (e.g., enforcement of pretreatment
requirements by POTWs or chemical substitution by industries), requiring a
Toxicity Reduction Evaluation (TRE) as discussed in Chapter 5 may be
appropriate. Where directly enforceable toxicity-based limitations are in effect,
the TRE alone is not sufficient enforcement response unless it includes corrective
action and requires expeditious compliance with the limit.
Violating limitations of chemicals with potential human health effects
should receive immediate enforcement attention to prompt rapid resolution of the
noncompliance. Immediate injunctive relief (such as a temporary restraining
237

-------
order or preliminary injunction) should be sought to protect public water supplies
and fish and shellfish areas.
In evaluating appropriate response to violations, EPA's Enforcement
Response Guide portion of the EMS should be used for guidelines on the
minimum acceptable response [3]. Further guidance on addressing violations of
whole-effluent toxicity limitations in particular is presented in the Compliance
Monitoring and Enforcement Strategy for Toxics Control [5] (Appendix B).
EPA has also developed guidance on the assessment of appropriate civil
penalties in both administrative and civil judicial actions. This guidance bases the
penalty amount on both the gravity of the violation and the economic benefit
enjoyed as a result of delayed compliance. This guidance is available in the
Clean Water Act Penalty Policy [6].
6.5.1 Enforcement Precedence
Many of the legal concepts involved with the enforcement of water quality-
based toxics requirements are the same concepts that have been tried and proven
for the NPDES program as a whole. Since the 1984 policy on the use of both
chemical and biological methods to protect water quality, additional court cases
have specifically addressed EPA's authority to establish limits on whole-effluent
toxicity in NPDES permits. A brief summary of some of the major legal bases
for chemical-specific toxics and whole-effluent toxicity follow:
NRDC v. EPA. 859 F.2d 156 (D.C. Cir. 1988)
The court upheld EPA regulations which authorize the use of effluent
limits framed in terms of toxicity.
Weyerhaeuser Co. v. Costle. 590 F.2d 1011 (D.C. Cir. 1978)
The court upheld other water characteristics (BOD, TTO, TSS, COD) as
permit limits.
Champion International Corp. v. EPA. 648 F. Supp. 1398 (W.D.N.C. 1987)
The court upheld EPAs authority to object to permits which do not
contain conditions adequate to achieve approved state water quality standards.
Trustees for Alaska v. EPA. 749 F.2d 549, 557 (4th Cir. 1984)
The court found that EPA as permit writer is required to establish
whatever permit limits are necessary to achieve water quality standards.
API v. EM, 787 F.2d 978 (5th Cir. 1986)
The court upheld EPAs use of a 96-hour LC50 as the most widely accepted
benchmark for toxicity evaluations by EPA and sustained EPAs choice of the test
for limiting effluent "mud" (drilling fluid) toxicity.
238

-------
Citizens to Preserve Overton Park v. Volpe. 401 U.S. 402, 416,
91 S.Ct. 814, 824, 28 L.Ed.2d 136 (1971)
The court deferred to the Agency's judgement in the settlement of
technical issues.
Given the on-going nature of enforcement of water quality-based permit
requirements, it is recommended that updated information be researched by the
reader through available means such as the use of LEXIS.
6.6 REPORTING OF VIOLATIONS
Part of the responsibility involved with the compliance monitoring and
enforcement of NPDES permits, is the requirement of reporting to the public on
permittees in violation. 40 CFR 123.45 establishes reporting requirements for the
Quarterly Noncompliance Report (QNCR) of major permittees in violation of
their NPDES permits. Reporting of violations of water quality-based monitoring,
limitations, schedules, and reporting requirements by major facilities must be
consistent with 40 CFR Part 123.45. Violations of permit or enforcement order
conditions by major permittees must be reported as follows [7]:
Effluent violations (chemical-specific and whole-effluent toxicity) must be
reported on the QNCR if the violation is determined through professional review
to have the potential to have caused a water quality impact.
Chemical-specific toxic permit limit violations must be reported on the
QNCR if 2 or more monthly average measurements in a 6 month period exceed
the limit by a factor of at least 1.2, or if 4 or more monthly average
measurements in a 6 month period exceed the limit by any amount. Any
violation during the quarter of an interim monthly average chemical-specific toxic
limit established in an administrative order or court order/consent decree must
be reported on the QNCR. (Note: whole-effluent toxicity is not characterized as
a Group I or Group II parameter, and as such, must be evaluated on a
professional judgement basis under the previous paragraph.)
Compliance schedule milestones that are not met within 90 days of the
scheduled date must be reported on the QNCR.
Failure to submit a report within 30 days of the due date must be reported
on the QNCR.
239

-------
REFERENCES
1.	Jensen, Lawrence J.,"Permit Compliance System (PCS) Policy Statement".
2.	NPDES Compliance Inspection Manual, May 1988.
3.	Enforcement Management System for the National Pollutant Discharge
Elimination System, September 1986.
4.	American Society of Testing and Materials Standard Practice for
Conducting Bioconcentration Tests with Fishes and Saltwater Bivalve
Mollusks, Designation E1022-84, 1986 Annual Book of ASTM Standards,
Volume 11.04 (PCN:01-110485-48), April 1985, pages 702-724.
5.	Rebecca W. Hanmer, "Whole Effluent Toxicity Basic Permitting Principles
and Enforcement Strategy", January 25, 1989.
6.	Lawrence J. Jensen, "Clean Water Act Penalty Policy for Civil Settlement
Negotiations", February 11, 1986.
7.	Rebecca W. Hanmer, "Guidance for Preparation of Quarterly and Semi-
Annual Noncompliance Reports", March 13, 1986.
240

-------
Index
AACs for Non-Carcinogens 67
Acceptable ambient concentrations 110
Acceptable tissue concentrations (ATCs) 62
Acute toxicity endpoints 10
Acute toxicity testing 96
Acute-Chronic Ratio 28
Additivity 39
Allowable Effluent Concentration Distribution 158
Ambient concentrations (AACs) 62
Ambient Toxicity Testing 87
Ambient-Induced Mixing 145
Ames test 43
Antagonism 40
Approach for Conducting TREs 217
Aquatic community 30
Aquatic Life Protection 56
Assessment of Human Health Effects 109
BCF 111
Bioaccumulation 64, 109
Bioaccumulation Consideration 64
Bioassessments 31
Bioconcentration 109
Bioconcentration factor (BCF) 110
Biocriteria 30, 71
Biomagnification 109
Biosurveys 31
Biosurveys/Bioassessments 31
Calculating AACs for Carcinogens 68
Carcinogenicity 42
CCC 57
Chemical-Specific Approach 2
Chemical-specific testing 1
Chronic toxicity endpoints 10
Chronic toxicity testing 97
Circumstances Warranting a TRE 224
Clean Water Act 1
CMC 57
Coefficient of variation 184, 209
Completley Mixed Discharge-Receiving Water Situations 148
Complex Effluent Toxicity Testing Program 12, 13
Compliance monitoring 233
Compliance problems 83
Consistency 215
Criterion continuous concentration 86
241

-------
Critical Design Periods for Waterbodies 136
Description of Limits 214
Designated Use 83
Detection Limits 213
Determining the Wasteload Allocation for Human Health Toxicant 169
Dilution 81
Dilution Determination 93
Discharge Monitoring Report/Quality Assurance (DMR/QA) 234
Discharge monitoring reports 81
Discharge-Induced Mixing 142
Discharges to Marine and Estuarine Environments 87
Duration 54
Duration for Single Chemicals and WET	59
Dye study 88
Effect of Changes on Statistical Parameters on Permit Limits 204
Effective concentration 10
Effluent Bioconcentration Evaluation 110
Effluent Caracterization 77
Effluent Characterization for Specific Chemicals 99
Effluent Characterization Process 93
Effluent variability 183, 189
Enforcement 237
Evaluation Criteria for TRE Plans	221
Excursion above CMC or CCC 98
Excursions Above Ambient Criteria 100
Fish Consumption Values 63
Flow-through 23
40 CFR 122.44(d)(l)(v) and (d)(l)(vi) 52
Frequency 54
Frequency for Single Chemicals and WET	60
GC/MS 113
General Recommendations for Outfall Design 134
General Recommendations for Tracer Studies 138
Generating Effluent Data 86
Genotoxic pollutants 42
Health effects 42
Human health 236, 237
Human Health Protection 42
Human health protection (WQC) 61
Incompletely Mixed, Discharge-Receiving Water Situations 126
Inhibition concentration 10
Inspections 235
Integrated Risk Information System (IRIS) 64
Integrated strategy 1
Intra-laboratory Precision 19
IRIS 116
LC10 10
242

-------
LC50 10
Lethal concentration 10
Log BCF 116
log-normal distribution 184
Long term average 184
Lowest observed effect concentration 10
Magnitude 54
Magnitude for Single Chemicals 56
Magnitude for Whole-effluent Toxicity 57
Magnitude, Duration, and Frequency (Criteria) 53
Mechanisms for Requiring TREs 226
Mixing Zone Analyses 133
Mixing Zones 55
Monticello Ecological Research Station 3
Multiple-source Discharge 87
Multiple-Source Toxicity Testing 88
N-octanol/partition coefficient (log P) 111
Narrative water quality criterion 1
No Excursions Above CMC or CCC 99
No observed effect concentration 10
No reasonable potential 99
NPDES program 1
Number of Samples 209
Numerical water quality criterion 51
Penalties 238
Permit Documentation 211
Permit Limit Derivation from Dynamic Model Outputs 197
Permit Limit Derivation from Single Steady State Model Output 192
Permit Limit Derivation from Two Value Steady State Outputs 193
Permit Limitations (WQS)
Permit Limitations 52
Pollution Prevention/Energy Conservation Considerations 212
Potential for Excursion above CMC or CCC 98
POTW 81
Precision 3, 20
Prevention of Bioaccumulation Problems for Human Health 133
Prevention of Lethal Conditions for Aquatic Life 131
Principles of Bioconcentration Control 110
Probability of Exceedence 210
Procedure for Updating an EPA Human Health Criterion 66
Ql* 65
Quality Assurance 234
Quantitative Structure-Activity Relationships database (QSAR) 111
Quarterly Noncompliance Report (QNCR) 239
Reasonable Potential 85
Reasonable potential 80, 98, 100
Receiving Water Concentration 78
243

-------
Regulatory Considerations (Criteria & Standards)
Rfd 50, 65
Renewal 23
Screening protocol 97
Sediment Criteria 73
Shayler Run, Ohio 3
Simplified statistical procedures. 202
Single exposure testing 99
Special Permitting Applications 200
Species Sensitivity 26
Static 23
Synergism 40
Total Maximum Daily Load (TMDL) 123
Toxic units 11
Toxicant Fate and Transport Models 162
Toxicity Identification Evaluation (TIE) 222
Toxicity Persistence 37
Toxicity Reduction Evaluation (TRE):(Enforcement)
Toxicity Reduction Evaluations 216
Toxicity Test Method Precision 18
Toxicity Testing Procedures 95
TRE Guidance Documents 217
Trigger 85
Triggers for Permit Limit Development 97
Variability 18, 25
Violation Review Action Criteria 236
Wasteload allocation 2
Wasteload Allocation Methods 127
Wasteload Allocations (WLA) 123
Wasteload Modeling Techniques 148
Water Quality Criteria 3, 50
Water Quality Standards 52
Whole effluent approach 9
Whole Effluent Toxicity Data Generation 89
WLA outputs 189
244

-------
APPENDIX A
Data for Chapter 1
Contents
Appendix A-l:
Appendix A-2:
Appendix A-3:
Toxicity Test Precision Data
Effluent Variability Data
Acute to Chronic Ratio Data

-------
Appendix A-l: Toxicity Test Precision Data
(Marine/ Estuarine and Freshwater Chronic Tests)

-------
Marine/ Estuarine Short-term Chronic Toxicity Tests

-------
SHEEPSHEAD MINNOW (CYPRONODON VARIEGATUS\
SEVEN-DAY LARVAL SURVIVAL AND GROWTH TEST
SINGLE LABORATORY PRECISION DATA
Table 1. Single laboratory precision of the sheepshead minnow
(Cvprinodon variegatus) larval survival and growth test
performed in forty fathoms artificial seawater, using
larvae from fish maintained and spawned in forty
fathoms artificial seawater, using copper as the
reference toxicant [3],
TEST	NOEC	IC25	IC50	MOST SENSITIVE
NUMBER (ug/1) (mg/1) (mg/1) END POINT
1	50	0.1133	0.1523	S
2	<50*	0.0543	0.0975	G
3	<50*	0.0418	0.0714	G
4	50	0.0632	0.0908	S
5	<50*	0.0577	0.0998	S
6	50	0.0483	0.1325	G
7	50	0.0796	0.1597	G
8	50	0.1235	0.2364	G
n:	5	8	8
Mean:	50	0.0727 0.1300
CV(%)	NA	41.82	40.77
NOEC Range: >50* - 50 ug/1 (* 50 ug/1 was the lowest
concentration tested)
Copper concentrations in Tests 1-6 were: 0.050, 0.10, 0.20, 0.40,
and 0.80 mg/1 and Tests 7-8: 0.025, 0.050, 0.10, 0.20, and 0.40
mg/1.
Prepared by Florence Kessler, TAI, Cincinnati, OH 1/11/90 (ICp
Program - version 1.1b)

-------
Table 2. Single laboratory precision of the sheepshead minnow
fCyprinodon variegatus) larval survival and growth
test performed in forty fathoms artificial seawater,
using larvae from fish maintained and spawned in forty
fathoms artificial seawater, using sodium dodecyl
sulfate (SDS) as the reference toxicant [3].
TEST
NOEC
IC25
IC50
MOST SENSITIVE
NUMBER
(ug/1)
(mg/1)
(mg/1)
END POINT
1
1.0
1.2799
1.5598
S
2
1.0
1.4087
1.8835
S
3
1.0
2.3051
2.8367
s
4
0.5
1.9855
2.6237
G
5
1.0
1.1901
1.4267
s
6
0.5
1.1041
1.4264
G
n:
6
6
6

Mean:
0.8
1.5456
1.9595

CV(%)
NA
31.44
31.82

NOEC Range: 0.5 - 1.0 ug/1 (this represents a difference of one
exposure concentration)
SDS concentrations in Tests 1-2 were: 1.0, 1.9, 3.9, 7.7, and
15.5 mg/1 and in Tests 3-6 were: 0.20, 0.50, 1.0, 1.9, and 3.9
mg/1.
Prepared by Florence Kessler, TAI, Cincinnati, OH 1/11/90 (ICp
Program - version 1.1b)

-------
Table 3. Single laboratory precision of the sheepshead minnow
(Cvprinodon varieqatus^ larval survival and growth
test performed in natural seawater, using larvae from
fish maintained and spawned in natural seawater, using
copper as the reference toxicant. [3]
TEST
NOEC
IC25
IC50
MOST SENSITIVE
NUMBER
(ug/1)
(ug/1)
(ug/l)
END POINT
1
125
320.3
437.5
S
2
31
182.3
323 .0
G
3
125
333.4
483.4
S
4
125
228.4
343.8
S
5
125
437.5
NC*
S
n:
5
5
4

Mean:
106.2
300.4
396.9

CV(%):
NA
33.0
19.2

NOEC Range: 31 - 125 ug/1 (this represents a difference of 2
exposure concentrations)
* A 50% Inhibition was not calculable for this test
Prepared by Elise Torello (SAIC) and Margarete Heber (EPA),
Narragansett, RI and Washington DC 2/90 (ICp Program version
1.1b)

-------
Table 4. Single laboratory precision of the sheepshead minnow
(Cvprinodon varieqatus) larval survival and growth
test performed in natural seawater, using larvae from
fish maintained and spawned in natural seawater, using
sodium dodecyl sulfate (SDS) as the reference toxicant.
[3]
TEST
NOEC
IC25
IC50
MOST SENSITIVE
NUMBER
(ug/1)
(ug/1)
(ug/l)
END POINT
1
2.5
2.9
3.6
S
2
1.3
NC*
NC*
G
'3
1.3
1.9
2.4
S
4
1.3
2.4
NC*
G
5
1.3
1.5
1.8
S
n:
5
4
3

Mean:
1.5
2.2
2.6

CV(%):
NA
27.6
35.3

NOEC Range: 1.3 - 2.5 ug/1 (this represents a difference of 1
exposure concentration)
* This value was not calculable for this test.
Prepared by Elise Torello (SAIC) and Margarete Heber (EPA),
Narragansett, RI and Washington DC 2/90 (ICp Program - version
l. lb)

-------
SHEEPSHEAD MINNOW ICYPRONODON VARIEGATUS^
SEVEN-DAY LARVAL SURVIVAL AND GROWTH TEST
INTER-LABORATORY PRECISION DATA
Table 5. Inter-laboratory precision of the sheepshead minnow
fCvprinodon varieoatus^ larval survival and growth
test, using an industrial effluent as the reference
toxicant. [3]
TEST	(MOST SENSITIVE ENDPOINT)
NUMBER NOEC	IC25	IC50
(%) (%) (%)
Laboratory
A




1
3 . 2(S, G)
7 . 4(S)
7.4(G)

2
3.2(S,G)
7 . 6(S)
14.3(G)
Laboratory
B




1
3 . 2(S, G)
5.7(G)
9.7(G)

2
3 . 2(S , G)
5.7(G)
8.8(G)
Laboratory
C




1
1.0(S)
4 • 7(S)
7.2(S)
Laboratory
D




1
3.2(S,G)
7.4(G)
24.7(G)

2
1.0(G)
5 . 2(S)
7•2(S)
n:

7
7
7
Mean:

2.6
5.5
11.3
CV(%):

NA
44.2
56.9
NOEC Range: 1.0 - 3.2 % (this represents a difference of one
exposure concentration)
Prepared by Elise Torello (SAIC) and Margarete Heber (EPA),
Narragansett, Rl and Washington DC 2/90 (ICp Program - version
1.1b)

-------
SHEEPSHEAD MINNOW (CYPRINODON VARIEGATUS ^
EMBR Y 0- LARVAL SURVIVAL AND TERATOGENICITY TEST
SINGLE LABORATORY PRECISION DATA
Table 6. Single laboratory precision of the sheepshead minnow
(Cvprinodon varieqatus) embryo-larval survival and
teratogenicity test performed in HW Marinemix
artificial seawater, using embryos from fish
maintained and spawned in HW Marinemix artificial
seawater, using copper as the reference toxicant.
[3]
TEST
EC1
EC5
EC10
EC50
NOEC
NUMBER
(ug/1)
(ug/1)
(ug/1)
(ug/1)
(ug/1)
1
173
189
198
234
240
2
*
.*
*
*
240
3
*
*
*
*
240
4
182
197
206
240
240
5
171
187
197
234
240
6
*
*
*
*
<200
7
*
*
*
*
220
8
195
203
208
226
220
n:
4
4
4
4
7
Mean:
180
194
202
233
234
CV(%):
6.1
3.8
2.8
2.5
NA
NOEC Range: 200 - 240 (this represents a difference of 2
exposure concentrations)
* Data do not fit the Probit model

-------
Table 7. Single laboratory precision of the sheepshead minnow
(Cvprinodon varieqatus) embryo-larval survival and
teratogenicity test performed in HW Marinemix
artificial seawater, using embryos from fish
maintained and spawned in HW Marinemix artificial
seawater, using sodium dodecyl sulfate (SDS) as the
reference toxicant. [3]
TEST	EC1	ECS	EC10	EC50	NOEC
NUMBER (mg/1) (mg/1) (mg/1) (mg/1) (mg/1)
1	1.7	2.0
2	*	*
3	0.4	0.7
4	1.9	2.2
5	1.3	1.7
2.2	3.1	2.0
*	*	4.0
0.9	2.5	2.0
2.4	3.3	2.0
1.9	3.0	2.0
n:	4	4	4	4	5
Mean:	1.3	1.6	1.9	2.9	2.4
CV(%): 51.2	41.6	35.0	11.7	NA
NOEC Range: 2.0 - 4.0 ug/1 (this represents a difference of one
exposure concentration)
* Data do not fit the Probit model

-------
INLAND SILVERSIDE (MI21 BERYLLINA)
SEVEN-DAY LARVAL SURVIVAL AND GROWTH TEST
SINGLE LABORATORY PRECISION DATA
Table 8« Single laboratory precision of the inland silverside
(Menidia bervllinal larval survival and growth test
performed in natural seawater, using larvae from fish
maintained and spawned in natural seawater, using
copper as the reference toxicant. [3]
TEST
NOEC
IC25
IC50
MOST SENSITIVE
NUMBER
(ug/1)
(ug/1)
(ug/1)
END POINT
1
63
96.2
148.6
S
2
125
207 o 2
NC*
S
3
63
218.9
493 .4
G
4
125
177.5
241.4
S
5
31
350.1
479.8
G
n:
5
5
4

Mean:
81.4
209.9
340.8

CV(%):
NA
43.7
50.7

NOEC Range: 31 - 125 ug/1 (this represents a difference of 2
exposure concentrations)
* A 50% Inhibition was not calculable for this test
Prepared by Elise Torello (SAIC) and1 Margarete Heber (EPA) ,
Narragansett, RI and Washington DC 2/90 (ICp Program - version
1.1b)

-------
Table 9. Single laboratory precision of the inland silverside
fMenidia bervllina) larval survival and growth test
performed in natural seawater, using larvae from fish
maintained and spawned in natural seawater, using
sodium dodecyl sulfate (SDS) as the reference toxicant.
[3]
TEST	NOEC	IC25	IC50	MOST SENSITIVE
NUMBER (ug/1) (ug/1) (ug/1) END POINT
1	1.3	0.3	1.7	S
2	1.3	1.6	1.9	S
3	1.3	1.5	1.9	S
4	1.3	1.5	1.9	S
5	1.3	1.6	2.2	S
n:	5	5	5
Mean:	1.3	1.3	1.9
CV(%): NA	43.2	9.4
NOEC Range: 1.3 ug/1
Prepared by Elise Torello (SAIC) and Margarete Heber (EPA),
Narragansett, RI and Washington DC 2/90 (ICp Program - version
1.1b)

-------
MYSID (MYSIDOPSIS BAHIA)
SEVEN-DAY SURVIVAL, GROWTH AND FEDUNDITY TEST
SINGLE LABORATORY PRECISION DATA
Table 10. Single laboratory precision of the mysid (Mvsidopsis
bahia) survival, growth and fecundity test performed in
natural seawater, using juveniles from mysids cultured
and maintained in natural seawater, using copper as the
reference toxicant. [3]
TEST
NOEC
IC25
IC50
MOST SENSITIVE
NUMBER
(ug/1)
(ug/1)
(ug/1)
END POINT
1
63
96.1
NC*
S
2
125
138.3
175.5
S
3
125
156.3
187.5
S
4
125
143.0
179.9
S
5
125
157.7
200. 3
S
n:
5
5
4

Mean:
112.6
138.3
185.8

CV(%):
NA
18.0
5.8

NOEC Range: 63 - 125 ug/1 (this represents a difference of 2
exposure concentrations)
* A 50% Inhibition was not calculable for this test
Prepared by Elise Torello (SAIC) and Margarete Heber (EPA),
Narragansett, RI and Washington DC 2/90 (ICp Program - version
1. lb)

-------
Table 11. Single laboratory precision of the mysid (Mvsidopsis
bahia) survival, growth and fecundity test performed in
natural seawater, using juveniles from mysids cultured
and maintained in natural seawater, using sodium
dodecyl sulfate (SDS) as the reference toxicant. [3]
TEST
NOEC
IC25
IC50
MOST SENSITIVE
NUMBER
(ug/1)
(ug/l)
(ug/1)
END POINT
1
2.5
4.5
NC**
S
2
<0.3
NC*
NC**
S
3
<0.6
NC*
NC**
S
4
5.0
7.8
NC**
S
5
2.5
3.6
4.6
S
6
5.0
7.0
9.3
S
n:
4
4
2

Mean:
3.8
5.7
6.9

CV(%):
NA
35.0
47.8

NOEC Range: <0.3 - 5.0 ug/1 (this represents a difference of 4
exposure concentrations)
* A 25% Inhibition was not calculable for this test
** A 50% Inhibition was not calculable for this test
Prepared by Elise Torello (SAIC) and Margarete Heber (EPA),
Narragansett, RI and Washington DC 2/90 (ICp Program - version
1.1b)

-------
SEA URCHIN (ARBACIA PUNCTULATA^
FERTILIZATION TEST
SINGLE LABORATORY PRECISION DATA
Table 12. Single laboratory precision of the sea urchin (Arbacia
punctulata) fertilization test performed in natural
seawater, using gametes from sea urchins maintained and
spawned in artificial seawater (Forty Fathoms), using
copper as the reference toxicant. [3]
TEST
NOEC
IC25
IC50
NUMBER
(ug/1)
(ug/l)
(ug/1)
1
5.0
8.92
29.07
2
12.5
26.35
38.96
3
<6.2
11.30
23.93
4
6.2
34 .28
61.75
5
12.5
36. 67
75.14
n:
4
5
5
Mean:
9.0
23.51
45.77
CV(%):
NA
54.60
47.87
NOEC Range: <6.2 - 12.5 ug/1 (this represents a difference of 1
exposure concentration)
Prepared by Florence Kessler, TAI, Cincinnati, OH 1/11/90 (ICp
Program - version 1.1b)
Copper concentrations in Test 1 were 2.5, 5.0, 10.0, 20.0 and
40.0 ug/1 and in Tests 2-5 were: 6.25, 12.5, 25.0, 50.0, and
100.0 ug/1.

-------
Table 13. Single laboratory precision of the sea urchin fArbacia
punctulata) fertilization test performed in natural
seawater, using gametes from sea urchins maintained and
spawned in artificial seawater (Forty Fathoms), using
sodium dodecyl sulfate (SDS) as the reference toxicant.
[3]
TEST
NOEC
IC25
IC50
NUMBER
(mg/1)
(mg/1)
(mg/1)
1
<0.9
1.11
1.76
2
0.9
1.27
1.79
3
1.8
2.26
2.87
4
0.9
1.90
2.69
5
1.8
2.11
2.78
n:
4
5
5
Mean:
1.4
1.73
2.38
CV(%):
NA
29.7
23 . 3
NOEC Range: 1.2 - 3.3 mg/1 (this represents a difference of 1
exposure concentration)
Prepared by Florence Kessler, TAI, Cincinnati, OH 1/11/90 (ICp
Program - version 1.1b)
SDS concentrations for all tests were 0.9, 1.8, 3.6, 7.2, and
14.4 mg/1.

-------
Table 14. Single laboratory precision of the sea urchin (Arbacia
punctulata) fertilization test performed in natural
seawater, using gametes from sea urchins maintained and
spawned in natural seawater, using copper as the
reference toxicant. [3]
TEST
NOEC
IC25
IC50
NUMBER
(ug/1)
(ug/1)
(ug/1)
1
12.2
14.2
18.4
2
12.2
32.4
50.8
3
24.4
30.3
46.3
4
<6.1
26.2
34.1
5
6.1
11.2
17.2
n:
4
5
5
Mean:
13 .7
22 .8
29.9
CV(%):
NA
41.9
48.2
NOEC Range: <6.1 - 24.4 ug/1 (this represents a difference of 2
exposure concentrations)
Prepared by Elise Torello (SAIC) and Margarete Heber (EPA),
Narragansett, RI and Washington DC 2/90 (ICp Program - version
1.1b)

-------
Table 15. Single laboratory precision of the sea urchin (Arbacia
punctulata) fertilization test performed in natural
seawater, using gametes from sea urchins maintained and
spawned in natural seawater, using sodium dodecyl
sulfate (SDS) as the reference toxicant. [3]
TEST
NOEC
IC25
IC50
NUMBER
(ug/1)
(ug/l)
(ug/1)
1
1.8
2.3
2.7
2
1.8
3.9
5.1
3
1.8
2 . 3
2.9
4
0.9
2 .1
2.6
5
1.8
2 . 3
2.7
n:
5
5
5
Mean:
1.6
2.58
3.2
CV(%):
NA
28.7
33 . 3
NOEC Range: 0.9 - 1.8 mg/1 (this represents a difference of 1
exposure concentration)
Prepared by Elise Torello (SAIC) and Margarete Heber (EPA),
Narragansett, RI and Washington DC 2/90 (ICp Program - version
1. lb)

-------
RED MACROALGAE (CHAMPIA PARVULA)
REPRODUCTION TEST
SINGLE LABORATORY PRECISION DATA
Table 16. Single laboratory precision of the red macroalga
(Champia parvula) reproduction test performed in 50/50
natural seawater and GP-2 artificial seawater. Copper
is the reference toxicant. [3]
TEST
NOEC
IC25
IC50
NUMBER
(ug/1)
(ug/i)
(ug/l)
1
1.0
1.67
2.35
2
1.0
1. 50
1.99
3
1.0
0.69
1.53
4
1.0
0.98
1.78
5
0.5
0.38
0.76
6
0.5
0.38
0.75
"n:	6	6	6
Mean:	0.83	0.93	1.5
CV(%): NA	59.6	43.7
NOEC Range: 0.5 - 1.0 ug/1 (this represents a difference of l
exposure concentration)
Prepared by Elise Torello (SAIC) and Margarete Heber (EPA),
Narragansett, RI and Washington DC 2/90 (ICp Program - version
1. lb)

-------
Table 17. Single laboratory precision of the red macroalga
(Champia parvula) reproduction test performed in 50/50
natural seawater and GP-2 artificial seawater. Sodium
dodecyl sulfate (SDS) is the reference toxicant. [3,
and personal communication with G. Thursby, SAIC,
Narragansett, RI]
TEST
NOEC
IC25
IC50
NUMBER
(ug/1)
(ug/1)
(ug/l)
1
<0.80
0.6
0.3
2
0.48
0.7
0.6
3
<0.48
0.4
0.2
4
<0.48
0.2
0.4
5
0.26
0.2
0.5
6
0. 09
0.1
0.3
7
0.16
0.2
0.3
8
0.09
0.1
0.2
9
<0.29
0.3
0.4
n:
5
9
9
Mean:
0.22
0.31
0.36
CV(%):
NA
69.0
37.0
NOEC Range: 0.09 - 0.48 mg/1 (this represents a difference of 2
exposure concentrations)
Prepared by Elise Torello (SAIC) and Margarete Heber (EPA),
Narragansett, RI and Washington DC 2/90 (ICp Program - version
1.1b)

-------
Freshwater Short-term Chronic Toxicity Tests

-------
FATHEAD MINNOW (PIMEPHALES PROMMjES)
SEVEN-DAY LARVAL SURVIVAL AND GROWTH TEST AND
EMBRYO-LARVAL SURVIVAL AND TERATOGENICITY TEST
SINGLE LABORATORY PRECISION DATA
Table 18. Single laboratory precision of the fathead minnow
(Pimephales promales) embryo-larval survival and
teratogenicity test performed in using Diquat as the
reference toxicant [2].
TEST	LCI
NUMBER	(mg/1)
1	0.58
2	2.31
3	1.50
4	1.71
5	1.43
n:
Mean:
CV(%):
5
1.51
41.3

-------
Table 19. Single laboratory precision of the fathead minnow
(Pimephales joromales) embryo-larval survival and
teratogenicity test performed in using cadmium chloride
as the reference toxicant [2].
TEST
LCI
NOEC
NUMBER
(mg/1)
(mg/1)
1
0.014
0.012
2
0. 006
0.012
3
0.005
0.013
4
0.003
0.011
5
0.006
0. 012
n:
5
5
Mean:
0.0068
0.012
CV(%):
62
NA
NOEC Range: 0.011 - 0.013
mg/1 (this represents a difference of 1
exposure
concentration)


-------
FATHEAD MINNOW (PIMEPHALES PROMALES)
SEVEN-DAY LARVAL SURVIVAL AND GROWTH TEST
SINGLE LABORATORY PRECISION DATA
Table 20.
Single laboratory precision of the fathead minnow

(Pimechales promales) larval survival and qrowth test

performed in using NAPCP as the reference toxicant [2].
TEST
NOEC*
NUMBER .
(mg/1)
1
256
2
128
3
256
4
128
5
128
n:
5
Mean:
179.2
CV(%):
NA
NOEC Range: 128 - 256 ug/1 (this represents a difference of 1
exposure concentration)
*Raw data unavailable, IC25 and IC50 values could not be
calculated

-------
.e 21. Results of the performance evalua n for contract laboratories conducted fc
the California Regional Water Quality Control Board. All tests were conducted
using potassium chromate (expressed as Cr+6) and testing the fathead minnow
(Pimephales promelas) in the seven day subchronic tests. [4]
Lab
Water
Food
Age
X Control
Weight
Ctrl,
n
IC25 (CI)
(mg/1 as Cr+6)
NOEC
Endpt.
IC50 (CI)
(mg/1 as Cr+6)
1
Tap8
2X
<24
0.590
3
3.7
(2.3-4.7)
3 G
5.4
(4.5-8.3)
2
MHRW
2X
<24
0. 623
3b
1.6f
(1.4-2.0)
<3 G
3.3
(2.8-4.0)
3
MHRW
3X
<24
0.274c
4
2. 2f
(1.7-3.1)
<3 G
4.7
(3.9-5.6)
4
Tapd
3X
<24
0.670
2
4.1
(2.3-5.0)
6 G
6.6
(5.0-8.4)
5
MHRW
__e
<24
0.773
4
1. 3f
(1.2-1.5)
<3 G
2.6f
(2.5-3.3)
6
MHRW
3X
y<24
0.635
2
7.1
(2.0-8.2)
6 G
9.9
(8.5-11)
7
MHRW
3X
<24
0. 390
3
4 . 5
(3.5-5.4)
3 G
7 . 4
(6.6-8.1)
8
Well9
2X
<24
0. 346
5
2 . 5f
(1.9-3.3)
<3 G
8.1
(6.4-15)
9
MHRW
3X
<24
0.415
4
6.6
(5.3-7.6)
6 G
9.2
(8.4-10)
10
MHRW
3X
<24
0.255
2
4.6
(4.1-5.9)
3 G
7. 8h
(5.2-12)
Mean	5.1	6.9
CV (%)	27	31
A Moderately hard tap water.
b Control with three replicates and all concentrations with two replicates.
c Weight measurements made with questionable techniques.
d Dechlorinated Lake Ontario tap water.
e Not reported.
f Value is extrapolated and is not included in coefficient of variation
calculation.
9 Well water mixed with spring water, moderately hard.
h Value may be skewed as middle concentration had 45% survival but no
weights reported.

-------
Table 22. Inter-laboratory precision data of the fathead minnow (Pimephales promales)
seven-day larval survival and growth test. Combined frequency distribution for
weight and survival NOECs for all participating laboratories. [2]
Sample
SURVIVAL	WEIGHT
NOEC Frequency (%) Distribution	NOEC Frequency (%) Distribution
Tests with 2 Reps Tests with 4 Reps	Tests with 2 Reps Tests with 4	Reps
Median ±l(a) >2(b) Median + 1(a) >2(b)	Median ±l(a) >2(b) Median +(a) >2 (b
Sodium
Pentachlorophenate (A)
35
53
12
57 29
14
59
41
57 43
0
Sodium
Pentachlorophenate (B)
42
42
16
56 44
37
63
22 45
33
Potassium
Dichromate (A)
47
47
75 25
35
47
18
88
0
12
Potassium
Dichromate (B)
41
41
18
50 50
12
47
41
63 25 12
Refinery Effluent 301
26
68
78 22
35
53
12
75 25
0
Refinery Effluent 401
37
53
10
56 44
37
47
16
33 56 11
Utility Waste 501
56
33
11
56 33 11
11
61
28
33 56 11
(a)	Percent of values with one concentration intervals of the median.
(b)	Percent of values within two or more concentrations intervals of the median,

-------
Table 23. Inter-laboratory precision data of the fathead
minnow (pimeohales promales) seven-day larval
survival and growth test. CV's for both IC50s
and IC25s have been calculated from the data
presented in Table 22 [4].	
Combined Tests CV (%'s)
Test	Battelle This Report
Material	IC50	IC25
101
PCP
Test 1
Test 2
32.4
55.2
201
KgCr^
Test 1
Test 2
27.5
27.5
301
Refinery Effueunt
Test 1
Test 2.
40.0
33.3
401
Refinery Effluent
Test 1
Test 2
22.4
27.6
501
Utility Effluent
Test 1
Test 2
61.9
82.2
601
KgCr^
Test 1
Test 2
88.0
70.1
701
PCP
Test 1
Test 2
37.6
67.5
Overall Mean
CV's (%)
44.6	51.9

-------
CERIODAPHNIA (CERIODAPHNIA DUBIA)
SEVEN-DAY REPRODUCTION TEST
SINGLE LABORATORY PRECISION DATA
Table 24. Single laboratory precision of (Ceriodaphnia dubia)
reproduction test performed in using NAPCP as the
reference toxicant [2].
TEST
NOEC
IC25
IC50
NUMBER
(mg/1)
(mg/1)
(mg/1)
19
0.30
0.3754
0.4508
4 6A
0.20
0.0938
0.2608
46B
0.20
0.2213
0.2897
49
0.20
0.2303
0.2912
55
0.20
0.2306
0.3177
56
0.10
0.1345
0.1744
57
0.20
0.2241
0.2827
n:
7
7
7
Mean:
0.20
0.2157
0.2953
CV(%):
NA
41.1
27.9
NOEC Range: 0.25 - 0.30 mg/1 (these values all fell within the
same concentration range)
Prepared by Florence Kessler, TAI, Cincinnati, OH 1/11/90 (ICp
Program -.version 1.1b)

-------
CERIODAPHNIA (CERIODAPHNIA DUBIA)
SEVEN-DAY LARVAL REPRODUCTION TEST
INTER-LABORATORY PRECISION DATA
Table 25.	Inter-laboratory precision of (Ceriodaphnia dubia)
reproduction test, using an industrial effluent as the
reference toxicant and a reference toxicant (NaCl).
Tests were conducted in May 1987. [Personal
Communication T.J. Norberg-King, U.S. EPA ERL-Duluth]

EFFLUENT
REFERENCE
TOXICANT
LAB
IC50 (%)
IC25 (%)
IC50 (%)
IC25 (%)
A
6.20
4.9
33 . 0
21.8
B
8.40
6.2
38.8
30.8
D
7.69
5.8
36.3
29.4
E
6.34
5.0
36.6
28.0
F
4.00
1.2
8.11
1.211
J
2.84
1.9
35.1
25.2
K
6.89
5.3
18.4
13.2
M
5.70
1.9
38.1
31.0
N
7.43
5.9
27.8
10.4
0
0.041
0.021
35.1
27.3
n:
Mean:
CV(%):
9
6.17
29
9
3.4
67
9
32.8
21
9
24.1
31
1 Outlier values are excluded from mean.

-------
Table 26. Results of the performance evaluation for contract laboratories conducted for the
California Regional Water Quality Control Board. All tests were conducted using
sodium chloride and testing Ceriodaphnia dubia in the seven day chronic tests.
X Young/ IC2 5 (Cl)	NOEC
Lab	Water	Food	Age Control (g/1 NaCl) Endpt.
1
Tap®
YCT/Algae
0-4;<24
17. 8
0. 20f
(0.14-0.35)
<0.25 R
2
Hard WWb
TF/Algae
0-4
26.5
1.3
(0.78-1.7)
1.0 R
3
DMW*
YCT/Algae
0-6
24.9
0. 21f
(0.17-0.54)
<0.25 R
4
Tapd
YCT
0-4
17. 2
0.49
(0.35-1.0)
0.5 R
5
HRW
YCT
0-4;<24
19.8
0.42
(0.20-1.1)
0.5 R
6
Surface®
YCT/Algae
0-6
14.8
0.90
(0.66-1.1)
0.25 R
7
MHRW
YCT/Algae
4-8
17.2
0.56
(0.24-0.64)
0.25 R
8
MHRW
YCT
<24
16. 8
0. 21f
(0.11-0.32)
0.25 R
9
MHRW
YCT
0-4
12.8
0.71
(0.56-0.81)
0.50 R
10
DMWC
YAT/Algae
0-4
31.5
0.91
(0.45-1.1)
1.0 R
Mean:	0.76
CV(%):	40
a Moderately hard tap water.
b Hard well water.
c 10% Diluted Mineral Water (DMW).
d Dechlorinated Lake Ontario tap
water.
e Briones reservoir water.
f Dose response curve limited.
R Reproductive Endpoint
MHRW = Moderately Hard Reconstituted
Water
HRW = Hard Reconstituted Water.
WW = Well Water
YCT = Yeast-Cerophyl-Trout chow
YAT = Yeast-Alfalfa-Trout chow
TF = Trout food suspension
Algae = Selenastrum capricornutum

-------
Table 27. Inter-laboratory precision data for Ceriodaphnia dubia
summarized for eight materials, including reference
toxicants and effluents. [5]

Test
Material
Mean
IC50
CV %
Mean
IC25
CV %
1
Sodium chloride
1.34
29.9
1. 00
34 . 3
2
Industrial
3.6
83.3
3.2
78. 1
3
Sodium chloride
0.96
57.4
0.90
44.4
4
Pulp & Paper
60.0
28.3
47. 3
27.0
5
Potassium dichromate
35.8
30.8
23.4
32.7
6
Pulp & Paper
70.2
7.5
55.7
12.2
7
Potassium dichromate
53 .2
25.9
29.3
46.8
8
Industrial
69.8
37.0
67.3
36.7
n:
Mean:
Standard Deviation:

8
37.5
23.0

8
39. 0
19.1

-------
SELENASTRUM CAPRICORNUTUM
GROWTH TEST
SINGLE LABORATORY PRECISION DATA
Table 28. Single laboratory precision of (Selenastrum
capricornutum) growth test performed in using cadmium as
the reference toxicant [2].
TEST	IC25	IC50
NUMBER (ug/1) (ug/1)
I	0.1929	0.3858
4	0.1723	0.3446
7	0.1747	0.3494
9	0.7137	0.9003
II	0.0429	0.0857
13	0.1121	0.2243
n:	6	6
Mean:	0.2348 0.3817
CV(%) :	102.6	72.57
Cadmium concentrations in Tests 1,4,7,9 were: 0.625, 1.25, 2.5,
5.0, and 10.0 ug/1; and in Test 11 were: 0.156, 0.312, 0.625, 1.25,
2.5, 5.0, and 10.0 ug/1 and in Test 13 were 0.312, 0.625, 1.25,
2.5, 5.0 ug/1.
Prepared by Florence Kessler, TAI, Cincinnati, OH 1/11/90 (ICp
Program - version 1.1b)

-------
Appendix A-2: Effluent Variability Data

-------
Table 29. Percent mortality in 100 % effluent to a POTW effluent
discharged via a common outfall, collected 1989 by grab
method [Personal communication W. Peltier, EPA-Athens,
GA]. Results indicate variability over 24 hours and
differences in species sensitivity over time. Tests
were conducted according to methods described in [1].
DATE
TIME
% MORTALITY
IN 100% EFFLUENT
P. promelas
D. culex
C. dubia
3/07/89
1230
0
15
100
3/07/89
1830
0
85
100
3/08/89
0030
0
65
100
3/08/89
0630
0
30
80
3/20/89
1230
0
0
100
3/20/89
1830
0
100
100
3/21/89
0030
0
95
100
3/21/89
0630
0
70
100
6/19/89
1230
0
5
100
6/19/89
1830
0
40
100
6/20/89
0030
0
100
100
6/20/89
0630
0
100
100
7/25/89
1230
0
0
100
7/25/89
1830
0
100
100
7/26/89
0030
0
100
100
7/26/89
0630
0
55
100

-------
Table 30. LC50s for a POTW effluent over seventeen months. All
tests were conducted using Ceriodaphnia dubia and tests
were run for 48 h. All tests were conducted according
to the methods described in [1]. Dates with roman
numeral notation mean that more than one sample was
collected at different times over a short interval (1-
2 days) .
Sample Date	48 h LC50 (%)
08/23/86-1
71
03/09/87-1
71
05/02/87-1
35
05/03/87-1
65
05/04/87-1
71
05/23/87-1
71
05/23/87-11
71
05/23/87-111
61
06/27/87-1
36
06/27/87-11
41
06/27/87-111
18
09/22/87-1
71
12/18/87-1
87
01/05/88-1
68
01/05/88-11
63
Mean LC50
CV (%):
n:
60.0
31.1
15

-------
Table 31. LC50s for a POTW effluent over seven months. All
tests were conducted using Ceriodaphnia dubia and tests
were run for 4 8 h. All tests were conducted according
to the methods described in [1], Dates with roman
numeral notation mean that more than one sample was
collected at different times over a short interval (1-
2 days).
Sample Date	48 h LC50 (%)
10/06/87-1
71
10/06/87-11
71
10/06/87-111
71
10/30/87-1
87
12/03/87-1
61
12/03/87-11
35
01/12/88-1
61
01/13/88-1
58
02/03/88-IX
50
02/03/88-X
50
03/03/88-111
87
03/03/88-IV
81
03/23/88-1
25
03/23/88-11
35
04/28/88-1
50
04/28/88-11
55
05/17/88-1
61
05/17/88-11
35
Mean LC50
CV (%):
n:
58.0
31.4
18

-------
Table 32. LC50s for a POTW effluent over twelve months. Tests
were conducted using either Ceriodaphnia dubia or
fathead minnows or both. Ceriodaphnia tests were
conducted for 48 h while fathead minnow tests were 96
h. Both the 48 h and 96 h fathead minnow results are
shown in order to evaluate how the LC50s for the two
species compare. All tests were conducted according to
the methods described in [1]. Dates with roman numeral
notation mean that more than one sample was collected
at different times over a short interval (1-2 days).
LC50 (%)
Sample Date	C.
C. dubia
48 h
P. promelas
48 h	96 h
03/16/88-1
06/09/88-1
09/08/88-1
10/04/88-1
10/04/88-11
12/14/88-1
12/14/88-11
02/17/89-1
02/17/89-11
03/22/89-1
03/22/89-11
62
18
68
61
63
70
17
35
35
35
47
35
1
25
>100
58
60
61
61
81
61
>100
34
41
39
37
64
40
Mean LC50: 46 59.6 40.0
CV (%): 42 22.4 29.7
n	;	11	7	7	
1 Data not available.
Note: Greater than values were excluded from the mean
LC50 calculation.

-------
Table 33. LC50s for a POTW effluent over four months. Tests were
conducted using either Ceriodaohnia dubia or fathead
minnows or both. Ceriodaphnia tests were conducted for
48 h while fathead minnow tests were 96 h. Both the 48
h and 96 h fathead minnow results are shown in order to
evaluate how the LC50s for the two species compare.
All tests were conducted according to the methods
described in [1]. Dates with roman numeral notation
mean that more than one sample was collected at
different times over a short interval (1-2 days).
	LC50 m	
Sample Date C. dubia P. promelas
	48 h	48 h	96 h
09/01/88-1
2.1
>100
77
11/15/88-1
92
67
37
11/16/88-1
61
>100
100
11/30/88-11
>100
>100
33
ll/30/88-III
95
>100
>100
12/08/88-1
100
87
54
12/08/88-11
>100
87
53
12/13/88-1
90
>100
77
12/13/88-11
87
85
51
•5
01/10/89-1
75
58
C
0l/lO/89—II
61
41

01/19/89-1
100
88
68
01/19/89-11
87
84
69
01/25/89-1
>100
87
64
01/25/89-11
95
85
56
01/31/89-1
90
70
60
01/31/89-11
63
70
60
Mean:
78.4
75.8
61.3
CV (%) :
33.1
19.6
27.7
n:
14
12
14
Not calculable
2 Not obtained
Note: Greater than values were not included in the
mean LC50 calculation.

-------
Table 34. NOECs for a POTW effluent conducted 20 times over one
year. All tests were conducted using Champia parvula
according to methods described in [3]. All effluent
samples were 24 hour composites collected post-
chlorination. (Data Source: Personnal Communication -
Glen Thursby, SAIC, Narragansett, RI.)
% EFFLUENT
Test Date
IC25
IC50
NOEC
12/09/85
0.65
1.23
1.25
12/10/85
0.38
0.76
1.25
12/11/85
0.69
1.50
2.50
12/12/85
0.41
0.82
1.25
12/13/85
3 . 09
4.12
5.00
12/15/85
2.16
4.09
5. 00
07/16/86
2.99
4.33
5.00
07/17/86
3.59
4.68
5.00
07/18/86
3.44
4.76
5.00
07/19/86
2.47
3.41
5.00
07/20/86
3.24
3.98
7.50
07/21/86
2.11
3.20
5.00
07/22/86
3.84
5.19
5.00
09/09/86
2.07
3.02
2.50
09/10/86
3. 17
4.13
7.50
09/11/86
2.73
3.62
7.50
09/12/86
1.57
1.89
1.25
09/14/86
1.25
1.76
2 . 50
n:
18
18
18
Mean:
2.2
3.1
4.2
CV (%):
52.8
46.8
NA

-------
Table 35. NOECs for a POTW effluent conducted over one year. All
tests were conducted using Arbacia punctulata according
to methods described in [3]. All effluent samples were
24 hour composites collected post-chlorination. (Data
source, ERL-Narragansett)
% EFFLUENT
Test Date	IC25	IC50	NOEC
12/09/85
1.09
1.71
0.65
12/10/85
1.41
2.84
0. 65
12/11/85
0.75
1.09
0.65
12/12/85
3.28
4.06
1.30
12/13/85
2 . 65
5.32
2.50
12/14/85
1.11
1.60
0. 65
12/15/85
1.29
1.84
0. 65
07/16/86
0.17
0.35
<0.30
07/17/86
0.21
0.46
<0.30
07/18/86
0.63
0.86
<0.30
07/19/86
1.09
1.68
<0.30
07/20/86
0.54
1.13
<0.30
07/21/86
0.40
0.58
<0.30
07/22/86
0.40
0.56
<0.30
09/09/86
0.31
0.41
<0.30
09/11/86
0.47
0.79
<0.60
09/12/86
0.21
0.48
<0.20
09/13/86
3 . 30
5.42
1.30
09/14/86
0.23
0.35
<0.20
09/15/86
0.10
0.15
<0.20
11/11/86
0.27
0.54
1.30
11/13/86
0.88
1.48
0.30
11/14/86
0.82
1.61
0.60
11/15/86
0.34
0.56
<0.30
Mean:
0.91
1.49
0.95
CV (%) :
101.3
96.9
NA
n:
24
24
11
Note: Less than (<) values were excluded from CV and
mean NOEC calculations.

-------
Table 36. NOECs for a POTW effluent conducted over one year. All
tests were conducted using Mvsidopsis bahia according
to methods described in [3], All effluent samples were
24 hour composites collected post-chlorination. (Data
source [ERL-Narragansett].)

%
EFFLUENT

Test Date
. IC25
IC50
NOEC
12/09 - 12/16/85
07/16 - 07/23/86
09/09 - 09/16/86
11/11 - 11/18/86
1.78(G)
2.75(R)
0.69(R)
0. 66(R)
2.93(G)
6.3(S)
20.1(S)
0.99(R)
1.0
3.2
10.0
3.2
Mean:
CV (%) :
n:
1.47
68.0
4
7.58
113 .8
4
4.4
NA
4
R-reproductive endpoint
S-survival endpoint
G-growth endpoint

-------
Table 37. NOECs for a POTW effluent conducted over one year. All
tests were conducted using Menidia bervllina according
to methods described in [3]. All effluent samples were
24 hour composites collected post-chlorination. (Data
source [ERL-Narragansett].)
% EFFLUENT
Test Date
IC25
IC50
NOEC
12/09 -
12/16/85
15.4
21.3
10. 0
07/16 -
07/23/86
15.2
21.0
10. 0
09/09 -
09/16/86
14.2
20.1
10. 0
11/11 -
11/18/86
NC
NC
10. 0
Mean:
CV (%):
n:

14 .9
4.3
3
20.8
3.0
3
10.0
NA
4
NC - value is not calculable

-------
Table 38. LC50s for a refinery effluent over fourteen months.
Tests were conducted using either Ceriodaphnia dubia or
fathead minnows Pimephales promeles or both.
Ceriodaphnia tests were conducted for 48 h while
fathead minnow tests were 96 h. Both the 48 h and 96 h
fathead minnow results are shown in order to evaluate
how the LC50s for the two species compare. All tests
were conducted according to the methods described in
[1]. Dates with roman numeral notation mean that more
than one sample was collected at different times over a
short interval (1-2 days).
	LC50 (%)	
Sample Date	C. dubia	P. promelas

48 h
48 h
96 h
12/01/87
15
35
16
01/05/88
35
36
19
02/09/88-1
35
35
<12
02/09/88-1
35
35
	1
<12
	1
03/02/88-1
17
03/02/88-11
<12
38
I
03/24/88-1
35
35
	1
1
4
05/06/88-1
35
I
07/14/88-1
55
61
25
07/28/88-1
37
35
22
07/28/88-11
28
31
<25
09/29/88-1
41
39
25
12/01/88-1
75
56
34
12/07/88-1
18
67
13
01/27/89-1
100
61
37
01/27/89-11
71
60
25
03/23/89-1
58
54
20
Mean LC50:
43
45
24
CV (%) :
54
28
32
n:
16
15
10
1 Data not available.
Note: Less than values excluded from mean LC50
calculations.

-------
Table 39. LC50s for a refinery effluent conducted over six months
using fathead minnows (Pimephales proroelas),
Ceriodaphnia dubia. and mysids (Mvsidopsis bahia).
(Data source [Dorn 1989].) according to methods
described in [1].
LC50
(% Effluent)
Test Date	C^_ dubia	P. promelas M. bahia
1/24/86
—
26.6
—
2/26/86
65.0
24.5
—
3/05/86
50.9
—
—
3/12/86
39.3
36.6
—
3/19/86
66.5
40.5
—
4/02/86
65.4
32.8
—
4/09/86
69.8
34.2
—
4/17/86
71.2
37.2
—
4/23/86
71.8
35.9
38.0
5/14/86
82 . 0
38.7
35.8
5/28/86
65.4
22.0
—
6/11/86
82.0
— —
24 .7
Mean NOEC:
66.3
32.9
32.8
CV (%) :
18.7
19.5
21.6
n:
11
10
3

-------
Table 40. NOECs for a refinery effluent conducted over six months
using fathead minnows (Pimephales promelasl,
Ceriodaphnia dubia. and mysids (Mvsidopsis bahia).
(Data source [Dorn 1989].) according to methods
described in [2,3].
NOEC
(% Effluent)
Test Date	Cj_ dubia	P. promelas M. bahia
1/24/86
—
14.1
—
2/26/86
10.1
7.1
—
3/05/86
5.6
—
—
3/12/86
10.1
14.1
—
3/19/86
10.1
14.1
—
4/02/86
18.0
14.1
—
4/09/86
10.1
14.1
—
4/17/86
10.1
7.1
—
4/23/86
10.1
7.1
24.0
5/14/86
31.7
14.1
24.0
5/28/86
18.0
7.1
—
6/11/86
31.7

13.4
Mean NOEC:
15.1
11.3
20.5
CV (%) :
59.6
31.9
29.8
n:
11
10
3

-------
Table 41. LC50s for a manufacturing effluent conducted over two
years. All tests were conducted using Daphnia magna
according to methods described in [1].
Test
Date

LC50
(% Effluent)
1982
(1st
quarter)
56
1982
(4 th
quarter)
90
1982
(4th
quarter)
70
1983
(3rd
quarter)
69
1983
(3rd
quarter)
36
1983
(3rd
quarter)
36
1983
(3rd
quarter)
32
1983
(3rd
quarter)
< 18
1983
(3rd
quarter)
28
1983
(3rd
quarter)
67
1983
(3rd
quarter)
< 10
1983
(4 th
quarter)
46
1983
(4th
quarter)
75
1983
(4 th
quarter)
78
1983
(4 th
quarter)
24
1983
(4 th
quarter)
26
1983
(4th
quarter)
32
1983
(4 th
quarter)
19
Mean
LC50
•
•
45.1 ± 24.3
CV (%) :

53.9
n:


18
Note: Less than (<) values were excluded from the mean
LC50 calculations.

-------
Table 42. LC50s for a manufacturing effluent conducted over eight
years. All tests were conducted using Pimephales
promales according to methods described in [1].
Test
Date

LC50
(% Effluent)
1979
(1st
quarter)
72.0
1979
(1st
quarter)
62 . 0
1979
(1st
quarter)
52.0
1979
(3rd
quarter)
39. 0
1981
(2nd
quarter)
64.0
1981
(4th
quarter)
70.0
1982
(2nd
quarter)
44 . 0
1982
(2nd
quarter)
66.0
1985
(1st
quarter)
59. 6
1985
(4 th
quarter)
>100.0
1986
(2nd
quarter)
49.2
1986
(2nd
quarter)
63.8
1986
(2nd
quarter)
50.0
1986
(3rd
quarter)
75.7
1986
(3rd
quarter)
80.0
1986
(3rd
quarter)
79.0
1986
(4 th
quarter)
71.0
Mean
LC50
»
•
64.5 ± 15.1
cv (%):

23.5
n:


17
Note: Greater than (>) values were excluded from the
mean LC50 calculations.

-------
Table 43. LC50s for a manufacturing effluent conducted over five
years. All tests were conducted using Pimephales
promales according to methods described in [1].
Test Date	LC50
"(% Effluent)
1980
(1st
quarter)
18.0
1980
(2nd
quarter)
11.0
1980
(3rd
quarter)
32.0
1980
(4th
quarter)
16.0
1981
(1st
quarter)
32.0
1981
(2nd
quarter)
23.0
1981
(3rd
quarter)
17.0
1981
(4 th
quarter)
46.0
1982
(1st
quarter)
9.0
1982
(2nd
quarter)
32.0
1982
(3rd
quarter)
28.0
1982
(4th
quarter)
52.0
1983
(1st
quarter)
34.0
1983
(2nd
quarter)
33 . 0
1983
(3rd
quarter)
20.0
1983
(4th
quarter)
43.0
1984
(1st
quarter)
45.0
1984
(2nd
quarter)
19.0
1984
(3rd
quarter)
61.0
1984
(4th
quarter)
20.0
Mean
LC50
•
•
29.6 + 14.2
CV (%):

47.9
n:	20

-------
Table 44. LC50s for a manufacturing effluent conducted over five
years. All tests were conducted using Daphnia magna
according to methods described in [1],
Test Date	LC50
(% Effluent)
1981
(2nd
quarter)
100.0
1981
(3rd
quarter)
>100.0
1982
(3rd
quarter)
>100.0
1984
(4th
quarter)
80.0
1985
(1st
quarter)
75.0
1986
(1st
quarter)
25.0
1986
(2nd
quarter)
82 .0
1987
(1st
quarter)
75.0
1987
(1st
quarter)
24.0
1987
(1st
quarter)
>100.0
1987
(1st
quarter)
>100.0
Mean LC50:	65.9 + 29.5
CV (%):	44.8
n:	11
Note: Greater than (>) values were excluded from the
mean LC50 calculations.

-------
Table 45. LC50s for a manufacturing effluent conducted over seven
years. All tests were conducted using Daphnia pulex
according to methods described in [1].
Test Date	LC50
(% Effluent)
1980
(1st
quarter)
55.0
1980
(4 th
quarter)
33.0
1981
(1st
quarter)
60.0
1981
(1st
quarter)
24.0
1981
(1st
quarter)
>100.0
1981
(2nd
quarter)
>100.0
1981
(3rd
quarter)
>100.0
1982
(3rd
quarter)
>100.0
1986
(3rd
quarter)
>100.0
1986
(3rd
quarter)
>100.0
Mean
LC50:

43.0
CV (%) :

40.2
n:	10
Note: Greater than (>) values were excluded from the
mean LC50 calculations.

-------
Table 46. LC50s for a manufacturing effluent conducted over three
months. All tests were conducted using Daphnia magna
according to methods described in [1],
Test Date	LC50
(% Effluent)
1982
(4 th
quarter)
>100.0
1982
(4 th
quarter)
81.0
1982
(4 th
quarter)
57. 0
1982
(4 th
quarter)
61.0
1982
(4th
quarter)
87.0
1982
(4 th
quarter)
90.0
1982
(4 th
quarter)
90. 0
1982
(4 th
quarter)
>100.0
1982
(4 th
quarter)
>100.0
1982
(4 th
quarter)
54.0
1982
(4 th
quarter)
74.0
1982
(4 th
quarter)
>100.0
Mean LC50:	74.3 + 15.1
CV (%):	20.3
n:	12
Note: Greater than (>)	values were excluded from the
mean LC50 calculations.

-------
Appendix A-3: Acute to Chronic Ratio Data

-------
Table 47.
EXAMPLES OF
ACUTE TO CHRONIC
RATIOS


OIL REFINERY*


FATHEAD MINNOW
CERIODAPHNIA
MYSIDS

1.89
9.09
1.58

3.47
3.89
1.49

2 . 60
6.58
1.84

2.87
3.63


2.33
6.91


2.43
7.05


5.26
7.11


5.08
3.63


2.74
2.59


3.11
5.5


5.1
4.4



>10.0



> 7.1



> 3.3



> 2.0



> 3.0



2.8



5.4**

Mean ACR:
3.3
5.3
1. 64
n:
11
13
3
Range:
1.89-5.26
2.59->10
.0 1.49-1.84
* Data Source: Personal communication P. Dorn
** Data Source: Personal communication M.L.C. Ramos and E.
Bertoletti
***Greater than(>) values were excluded from mean calculations

-------
Table 48.
EXAMPLES OF ACUTE TO CHRONIC RATIOS
CHEMICAL MANUFACTURER
FATHEAD MINNOW	CERIODAPHNIA
0.17	>1.0
0.07	>1.0
8.4	>10.0
7.6	>50.0
>3.0	>2.9
3.9	>1.4
>3.0	1.4
1.8	1.4
3.9
2.8
>	2.0
>	4.0
4 . 0
1.4
5.5
1.8
>	3.3
>	3.3
>	3.3
1.4
>	2.0
5.5
1.5
1.4
5.0
>10. 0
>	2.0
>	3.3
3.1*
14.0*
4.3*
2.5*
1.8*
5.5*
5.4*
Mean ACR:	3.7	3.7	**
n:	6	20
Range:	0.07-8.4	1.4- >50
* Data Source: Personal communication M.L. Ramos and E.
Bertoletti
** Greater than (>) values were excluded from the mean
calculation

-------
Table 49.
EXAMPLES OF ACUTE TO CHRONIC RATIOS
POTWS
FATHEAD MINNOW	CERIODAPHNIA
2.9
1.4
6.1
5.5
1.5
> 1.0
13.0
> 1.0
1.8
> 1.0
2.6
1.8
9.3
1.4
1.0
2.0
3.0
2.4
5.3
3.0
3.3
3.0
5.4
5.5
3.0
4.9
3.0
> 2.0

> 8.0

> 2.0

> 1.0

> 3.3

> 2.0

4.4

16.1

> 4.0

> 3.3

>10.0

2.6

5.7

2.8

>10.0

> 2.0

1.4

2.6

> 3.3

1.8

5.5

1.5

> 3.3

5.5
Mean ACR:
4.9
3.8*
n:
11
21
Range:
1.5 - 9.3
1.4 - 16.1
* Greater than
(>) values were
excluded from mean calculations

-------
REFERENCES
1.	Peltier, W. and C.I. Weber, 1985. Methods for Measuring the
Acute Toxicity of Effluents to Aquatic Organisms, Third
Edition. Office of Research and Development, Cincinnati, OH.
EPA 600/4-85/013.
2.	Weber, C.I. et. al. (ed.), 1989. Short-Term Methods for
Estimating the Chronic Toxicity of Effluents and Receiving
Waters to Freshwater Organisms, Second Edition. Office of
Research and Development, Cincinnati, OH. EPA 600/4-89/001.
3.	Weber, C.I. et. al. (ed.), 1988. Short-Term Methods for
Estimating the Chronic Toxicity of Effluents and Receiving
Waters to Marine and Estuarine Organisms. Office of Research
and Development, Cincinnati, OH. EPA 600/4-87/028.
4.	Personal communication with T. Norberg-King, EPA-ERL-Duluth.
5.	DeGraeve, G. M., J. D. Cooney, B. H. Marsh, T. L. Pollock,
and N. G. Reichenbach. 1989. Intra- and Interlaboratory
study to determine the reproducibility of the seven-day
Ceriodaphnia dubia survival and reproduction tests.
Battelle, Columbus Division, Columbus, Ohio (In preparation).

-------
APPENDIX B
Supporting Statute, Policies, Strategies and Regulations for
Water Quality-Based Toxics Control
Contents:
Appendix B-l:
Appendix B-2:
Appendix B-3:
Appendix B-4:
Appendix B-5:
Appendix B-6:
Statutory Provisions of the Clean Water Act Authorizing Whole
Effluent Toxicity Control
Policy for the Development of Water Quality-Based Limitations for
Toxic Pollutants
New Regulations Governing Water Quality-Based Permitting
Whole Effluent Toxicity Permitting Principles and Enforcement
Strategy
Quality Control Fact Sheets for Self-Biomonitoring Acute and
Chronic Toxicity Test Data
Important Case Decisions Regarding Whole Effluent Toxicity

-------
Appendix B-l: Statutory Provisions of the Clean Water Act Authorizing Whole
Effluent Toxicity Control

-------
Clean Water Act (33 U.S.C. 1251 et. seq.)
Statutory Authority for the Use of Toxicity Testing and Whole Effluent
Toxicity Limitations in NPDES Permits:
Over the years, a developmental process has occurred regarding the use of
biological techniques to assess effluent discharges and set permit limits. The
acquisition of data and the development of new techniques has contributed to the
refinement of toxicity testing methods, thus enabling EPA to more fully act in
accordance with its mandates to implement statutory requirements relating to the
attainment and maintenance of water quality.
Toxicity testing of whole effluents and whole effluent toxicity limitations in
National Pollutant Discharge Elimination System (NPDES) permits are essential
components in the control of the discharge of toxic pollutants to the nation's
waters. The use of toxicity testing and whole effluent toxicity limitations in the
NPDES program is clearly authorized by the Clean Water Act (CWA).
Relevant provisions that provide the statutory authority for using toxicity
testing and whole effluent toxicity limitations include the following:
CWA{ 101(a) sets forth not only the goal of restoring and
maintaining the "chemical, physical, and biological integrity of the
Nation's waters" (emphasis added), but also the national policy of
prohibiting the "discharge of toxic pollutants in toxic amounts"
(emphasis added).
As defined at CWA{ 502(15), biological monitoring means the
"determination of the effects on aquatic life, including accumulation
of pollutants in tissue, in receiving waters due to the discharge of
pollutants (A) by techniques and procedures, including sampling of
organisms representative of appropriate levels of the food chain
appropriate to the volume and the physical, chemical, and biological
characteristics of the effluent, and (B) at appropriate frequencies
and locations."
CWA {304(a)(8) requires EPA to develop information on methods,
including biological monitoring and assessment methods, to establish
and measure water quality criteria for toxic pollutants on bases
other than pollutant by pollutant criteria.
Biological assessment or monitoring methods may also be used
under { 303(c)(2)(B) by States to revise or set new water quality
standards for toxic pollutants listed under { 307(a) if the State
reasonably expects interference with designated uses, and EPA has
not published a water quality criteria document, nor are numerical
criteria available.

-------
Furthermore, CWA { 303(c)(2)(B) states, "Nothing in this section
shall he construed to limit or delay the use of effluent limitations or
other permit conditions based on or involving biological monitoring
or assessment methods . . (emphasis added).
Section 302(a) provides the authority to establish water quality-
based effluent limitations on discharges that interfere with the
attainment or maintenance of that water quality which shall assure
protection of public health, public water supplies, and the protection
and propagation of a balanced population of shellfish, fish and
wildlife.
Under CWA{ 301(b)(1)(C) and{ 402, all NPDES permits must
comply with any more stringent limitations necessary to meet
applicable water quality standards, whether numeric or narrative.
CWA { 308(a) and { 402 provide authority to EPA or the State to
require that NPDES permittees/applicants use biological monitoring
methods and provide chemical, toxicity, and instream biological data
when necessary for the establishment of effluent limits, the
detection of violations, or the assurance of compliance with water
quality standards.
Section 510 provides the authority for states to adopt or enforce any
standards or effluent limitations for the discharge of pollutants only
on the condition that such limitations or standards are no less
stringent than those in effect under the CWA.

-------
Appendix B-2: Policy for the Development of Water Quality-Based Limitations for
Toxic Pollutants

-------
9016
Federal Register / Vol. 49, No. 48 / Friday, March 9, 1984 / Notices
IW-fRL-2533-1]
vevelopment of Water Ouallty-Based
Permit Limitations tor Toxic Pollutants:
National Policy
AGENCY: Environmental Protection
Agency (EPA)
ACTION: Notice.
summary: EPA has issued a national
policy statement entitled "Policy for the
Development of Water Quality-Based
Permit Limitations for Toxic Pollutants."
This policy addresses the technical
approach for assessing and controlling
the discharge of toxic substances to the
Nation's waters through the National
Pollutant Discharge Elimination System
(NPDES) permit program.
FOR FURTHER INFORMATION CONTACT
Bruce Newton or Rick Brandes. Permits
Division (EN-336). Office of Water
Enforcement and Permits. U.S.
Environmental Protection Agency.
Washington. D C. 20460. 426-7010.

-------
Federal Register / Vol. 49. No. 48 / Friday, March 9, 1984 / Notices
9017
SUPPLKMCWTARY INFORMATION: As the
water pollution control effort in the
United States progresses and the
"traditional" pollutants (oxygen
demanding and eutrophying materials)
become sufficiently treated to protect
water quality, attention is shifting
towards pollutants that impact water
quality through toxic effects. Compared
with the traditional pollutants,
regulation of toxic pollutants is
considerably more difficult. The
difficulties include (1) the great number
of toxic chemicals that may potentially
be discharged to receiving waters and
the difficulties in their analysis; (2) the
changes in the toxic effects of a
chemical resulting from reactions with
the matrix of constituents in which it
exists; and (3) the inability to predict the
effects of exposure to combinations of
chemicals.
To overcome some of these problems,
EPA and the States have begun to use
aquatic toxicity tests and various human
health assessment techniques to
complement chemical analyses of
effluents and receiving water samples.
Because these techniques or their
application to effluent testing are new,
EPA and the States have been cautious
in their use. Based on EPA's evaluation
of these techniques and the experiences
of several States. EPA is now
recommeding the use of biological
techniques as a complement to
chemical-specific analyses to assess
effluent discharges and express permit
limitations. EPA has issued these
recommendations through a statement
of policy and is developing a technical
guidance document to help implement
the policy.
The complete test of the national
policy statement follows:
Policy for the Development of Water
Quality-Based Permit limitations for
Toxic Pollutants
Statement of policy
To control pollutants beyond Best
Available Technology Economically
Achievable (BAT), secondary treatment,
and other Clean Water Act technology-
based requirements in order to meet
water quality standards, the
Environmental Protection Agency (EPA)
will use an integrated strategy
consisting of both biological and
chemical methods to address toxic and
nonconventional pollutants from
industrial and municipal sources. Where
State standards contain numerical
criteria for toxic pollutants, National
Pollutant Discharge Elimination System
(NPDES) permits will contain limits as
necessary to assure compliance with
these standards. In addition to enforcing
specific numerical criteria, EPA and the
States will use biological techniques and
available data on chemical effects to
assess toxicity impacts and human
health hazards based on the general
standard of "no toxic materials in toxic
amounts."
EPA, in its oversight role, will work
with States to ensure that these
techniques are used wherever
appropriate. Under section 308 and
section 402 of the Clean Water Act (the
Act), EPA or the State may require
NPDES permit applicants to provide
chemical, toxicity, and instream
biological data necessary to assure
compliance with standards. Data
requirements may be determined on a
casa-;?y-case basis in consultation with -
the State and the discharger.
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
toxic 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
toxicity testing after installation of
treatment and may reopen the permit to
incorporate additional limitations if
needed to meet water quality standards.
(Toxicity data, which are considered
"new information" in accordance with
40 CFR 122.62(a)(2), could constitute
cause for permit modification where
necessary.)
To carry out this policy, EPA Regional
Administrators mil assure that each
Region has the capability to conduct
water quality assessments using both
biological and chemical methods and
provide technical assistance to the
States.
Background
The Clean Water Act establishes two
principal bases for effluent limitations.
First, existing dischargers an required
to meet technology-based effluent
limitations that reflect the best controls
available considering economic impacts.
New source dischargers must meet the
best demonstrated technology-based
controls. Second, where necessary,
additional requirements are imposed to
assure attainment and maintenance of
water quality standards established by
the States and approved by EPA. In
establishing or reviewing NPDES permit
limits, EPA must ensure that the limits
will result in the attainment of water
quality standards and protect
designated water uses, including an
adequate margin of safety.
For toxic and nonconventional
pollutants it may be difficult in some
situations to determine attainment or
nonattainment of water quality
standards and set appropriate limits
because of complex chemical
interactions which affect the fate and
ultimate impact of toxic substances in
the receiving water. In many cases, all
potentially toxic pollutants cannot be
identified by chemical methods. In such
situations, it is more feasible to examine
the whole effluent toxicity and instream
impacts using biological methods rather
than attempt to identify all toxic
pollutants, determine the effects of each
pollutant individually, and then attempt
to assess their collective effect.
The scientific basis for using
biological techniques has advanced
significantly in recent years. There is
now a general consensus that an
evaluation of effluent toxicity, when
adequately related to instream
conditions, can provide a valid
indication of receiving system impacts.
This information can be useful in
developing regulatory requirements to
protect aquatic life, especially when
data from toxicity testing are analyzed
in conjunction with chemical and
ecological data. Generic human health
effects methods, such as the Ames
mutagenicity test, and structure-activity
relationship techniques are showing
promise and should be used to identify
potential hazards. However, pollutant-
specific techniques are the best way to
evaluate and control human health
hazards at this time.
Biological testing of effluents is an
important aspect of the water quality-
based approach for controlling toxic
pollutants. Effluent toxicity data in
conjunction with other data can be used
to establish control priorities, assess
compliance with State water quality
standards, end set permit limitations to
achieve those standards. All States have
water quality standards which include
nanative statements prohibiting the
discharge of toxic materials in toxic
amounts. A few State standards have
criteria more specific than narrative
criteria (for example, numerical criteria
for specific toxic pollutants or a toxicity
criterion to achieve designated uses). In
States where numerical criteria are not
specified, a judgment by the regulatory
authority is required to set quantitative
water quality-based limits on chemicals
and effluent toxicity to assure

-------
9018
Federal Register / Vol. 49. No. 48 / Friday. March 9. 1904 / Notices
lpliance with water quality
jndards.
Note.—Section 306 of the Act and
corresponding State statutes authorize WA
and the States to require of the owner/
operator any information reasonably required
to determine permit limits and to determinp
compliance with standards or permit limitH
Biological methods are specifically
mentioned. Toxicity permit limits are
authorized under Section 301 and 402 unci
supported by Section 101.
Application
This policy applies to EPA and the
States. The policy addresses the use of
chemical and biological methods for
assuring that effluent discharges are
regulated in accordance with Federal
and State requirements. This policy was
prepared, in part, in response to
concerns raised by litigants to the
Consolidated Permit Regulations (see KK
52079, November 16.1982). Use of these
methods for developing water quality
standards and trend monitoring are
discussed elsewhere (see 48 FR 51400.
November 8,1983 and Basic Water
Monitoring Program EPA-440/9-7B-025J.
This policy is part of EPA's water
quality-based control program and does
not supersede other regulations, policy,
d guidance regarding use

-------
Federal Register / Vot. 49. No. 4* / Friday. March 9, 1964 / Notices
9019
quality standards violation, these
individual pollutants should be limited.
If a toxicity reduction evaluation
demonstrates that limiting an indicator
parameter will ensure attainment of the
water quality-based effluent toxicity
requirement, limits on the indicator
parameter should be considered in lieu
of limits on effluent toxicity. Such
indicator limits are not limits on
causative pollutants but limits
demonstrated to result in a specific
toxicity reduction.
Monitoring
Where pollution control requirements
are expressed in terms of a chemical or
toxicological parameter, compliance
monitoring must include monitoring for
that parameter. If an indicator
parameter is used based on the results
of a toxicity reduction evaluation,
periodic toxicity testing may be required
to confirm the adequacy of the indicator.
Where biological data were used to
develop a water quality assessment or
where the potential for water quality
standards violations exist, biological
monitoring (including instream
monitoring) may be required to ensure
continuing compliance with water
quality standards.
EPA believes that the intelligent
application of an integrated strategy
using both biological and chemical
techniques for water quality assessment
will facilitate the development of
appropriate controls and the attainment
of water quality standards. EPA looks
forward to working with the States in a
spirit of cooperation to further refine
these techniques.
Policy signed February 3.1984 by )ack E.
Ravan, Assistant administrator for Water.
Dated. February la. 19H4.
lack E. Ravan.
Assistant Administrator for Water.
[FH Doc. M 8443 Filled V-n-W «4i ,m|
ttLUM COOt ICtO 10 M

-------
ATTACHMENT 2
Background to the Compliance Monitoring and Enforcement
Strategy for Toxica Control
The Compliance Monitoring and Enforcement Strategy for
Toxica Control sets forth the Agency's strategy for tracking
compliance with and enforcing whole-effluent toxicity monitoring
requirements, limitations, schedules and reporting requirements.
The Strategy delineates the respective responsibilities of
permittees and permitting authorities to protect water quality
through the control of whole-effluent toxicity. It establishes
criteria for the review of compliance data and the quarterly
reporting of violations to Headquarters and the public. The
Strategy discusses the integration of whole-effluent toxicity
control into our existing inspection and quality assurance
efforts. It provides guidelines on the enforcement of whole-
effluent toxicity requirements.
The Strategy also addresses the concern many permittees
share as they face the prospect of new requirements in their
permit - the fear of indiscriminate penalty assessment for
violations that they are unable to control. The Strategy
recognizes enforcement discretion as a means of dealing fairly
with permittees that are doing everything feasible to protect
water quality. As indicated in the Strategy, this discretion
deals solely, with the assessment of civil penalties, however, and
is not an alternative to existing procedures for establishing
relief from State Water Quality Standards. The Strategy focuses
on the responsibility of the Agency and authorized States to
require compliance with Water Quality Standards and thereby
ensure protection of existing water resources.

-------
- 2 -
The second principle also deals with the specification of
test species and protocol. Clearly setting out the requirements
for toxicity testing and analysis is best done by accurately
referencing EPA's most recent test methods and approved
equivalent State methods. In this way, requirements which have
been published can be required in full, and further advances in
technology and science may be incorporated without lengthy permit
revisions.
The third and final permitting principle reinforces the
responsibility of the permittee to seek timely compliance with
the requirements of its NPDES permit. Once corrective actions
have been identified in a TRE, permittees cannot be allowed to
delay corrective actions necessary to comply with water quality-
based whole effluent toxicity limitations pending Agency review
and approval of voluminous reports or plans. Any delay on the
part of the permittee or its contractors/agents is the
responsibility of the permittee.
The final principle was written in recognition of the fact
that a full-blown TRE may not be necessary to return a permittee
to compliance in all cases, particularly subsequent to an initial
TRE. As a permittee gains experience and knowledge of the
operational influences on toxicity, TREs will become less
important in the day to day control of toxicity and will only be
required when necessary on a case-specific basis.

-------
ATTACHMENT 1
Explanation of the Basic Permitting Principles
The Basic Permitting Principles present the minimum
acceptable requirements for whole-effluent toxicity permitting.
They begin with a statement of the goal of whole-effluent
toxicity limitations and requirements: the protection of water
quality as established through State numeric and narrative Water
Quality Standards. The first principle builds on the Technical
Support Document procedures and the draft Section 304(1) rule
requirements for determining potential to violate Water Quality
Standards. It requires the same factors be considered in setting
whole-effluent toxicity based permits limits as are used to
determine potential Water Quality Standards violations. It
defines the universe of permittees that should be evaluated for
potential violation of Water Quality Standards, and therefore
possible whole-effluent limits, as all majors and minors of
concern.
The second permitting principle provides basic guidelines
for avoiding ambiguities that may surface in permits. Whole-
effluent toxicity limits should be listed in Part I of the permit
and should be derived and expressed in the same manner as any
other water quality-based limitations (i.e., Maximum Daily and
Average Monthly limits as required by Section 122.45(d)).
In addition, special re-opener clauses are generally not
necessary, and may mistakenly imply that permits may be re-opened
to revise whole-effluent limits that are violated. This is not
to imply that special re-opener clauses are never appropriate.
They may be appropriate in permits issued to facilities that
currently have no known potential to violate a Water Quality
Standard; in these cases, the permitting authority may wish to
stress its authority to re-open the permit to add a whole-
effluent limit in the event monitoring detects toxicity.
Several permittees have mistakenly proposed to conduct
additional nonitoring subsequent to a violation to "verify" their
results. It is not possible, to verify results with a subsequent
test whether a new sample or a split-sample which has been stored
(and therefore contains fewer volatiles) is used. For this
reason, any additional monitoring required in response to a
violation must be clearly identified as establishing continuing
compliance status, not verification of the original violation.

-------
BASIC PERMITTING PRINCIPLES FOR WHOLE EFFLUENT TOXICITY
1.	Permits must be protective of water quality.
a.	At a minimum, all major permits and minors of
concern must be evaluated for potential or known
toxicity (chronic or acute if more limiting).
b.	Final whole effluent toxicity limits must be
included in permits where necessary to ensure
that State Water Quality Standards are met.
These limits must properly account for effluent
variability, available dilution, and species
sensitivity.
2.	Permits must be written to avoid ambiguity and ensure
enforceability.
a.	Whole effluent toxicity limits must appear in part I
of the permit with other effluent limitations.
b.	Permits contain generic re-opener clauses which
are sufficient to provide permitting authorities
the means to re-open, modify, or reissue the
permit where necessary. Re-opener clauses covering
effluent toxicity will not be included in the
Special Conditions section of the permit where
they imply that limit revision will occur based
on permittee inability to meet the limit. Only
sch'edules or other special requirements will be
added to the permit.
c.	If the permit includes provisions to increase
monitoring frequency subsequent to a violation, it
must be clear that the additional tests only deter-
mine the continued compliance status with the limit;
they are not to verify the original test results.
d.	Toxicity testing species and protocols will be
accurately referenced/cited in the permit.
3.	where not in compliance with a whole effluent toxicity
limit, permittees must be compelled to come into compliance
with the limit as soon as possible.
a.	Compliance dates must be specified.
b.	Permits can contain requirements for corrective
actions, such as Toxicity Reduction Evaluations
(TREs), but corrective actions cannot be delayed
pending EPA/State approval of a plan for the
corrective actions, unless State regulations
require prior approval. Automatic corrective
actions subsequent to the effective date of a final
whole-effluent toxicity limit will not be included
in the permit.

-------
- 2 -
Water Enforcement and Permits. This expanded guidance will
include sample permit language and permitting/enforcement
scenarios.
Concurrent with this issuance of the Basic Permitting
Principles, I am issuing the Compliance Monitoring and
Enforcement Strategy for Toxics Control (Attachment 2). This
Strategy was developed by a workgroup of Regional and State
enforcement representatives and has undergone an extensive
comment period. The Strategy presents the Agency's position on
the integration of toxicity control into the existing National
Pollutant Discharge Elimination System (NPDES) compliance and
enforcement program. It delineates the responsibilities of the
permitted community and the regulatory authority. The Strategy
describes our current efforts in compliance tracking and quality
assurance of self-monitoring data from the permittees. It
defines criteria for review and reporting of toxicity violations
and describes the types of enforcement options available for the
resolution of permit violations.
In order to assist you in the management of whole effluent
toxicity permitting, the items discussed above will join the 1984
Policy as Appendices to the revised Technical Support Document
for Water Quality-based Toxics Control. To summarize, these
materials are the Basic Permitting Principles, sample permit
language, the concepts illustrated through the permitting and
enforcement scenarios, and the Enforcement Strategy. I hope
these additions will provide the needed framework to integrate
the control of'toxicity into the overall NPDES permitting
program.
I encourage you and your staff to discuss these documents
and the 1984 Policy with your States to further their efforts in
the implementation of EPA's toxics control initiative.
If you have any questions on the attached materials, please
contact James Elder, Director of the Office of Water Enforcement
and Permits, at (FTS/202) 475-8488.
Attachments
ccx ASWIPCA
Water Management Division Directors

-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
I 5SBJ	WASHINGTON, D.C. 20460
J?	January 25, 1 989

OFFICE OF
WATER
MEMORANDUM
SUBJECT:^_Whole Effluent Toxicity Basic Permitting Principles and
( enforcement Strategy
— MTWi Kwfi-
FROM: Rebecca W. Hanmer, Acting Assistant Administrator
Office of Water
TO:	Regional Administrators
Since the issuance of the "Policy for the Development of
Water Quality-based Permit Limitations for Toxic Pollutants" in
March of 1984, the Agency has been moving forward to provide
technical documentation to support the integrated approach of
using both chemical and biological methods to ensure the
protection of water quality. The Technical Support Document for
Water Quality-based Toxics Control (September, 1985) and the
Permit Writer's Guide to Water Quality-based Permitting for Toxic
Pollutants (July, 1987) have been instrumental in the Initial
implementation of the Policy. The Policy and supporting
documents, however, did not result in consistent approaches to
permitting and enforcement of toxicity controls nationally. When
the 1984 Policy was issued, the Agency did not have a great deal
of experience in the use of whole effluent toxicity limitations
and testing to ensure protection of water quality. We now have
more than four years of experience and are ready to effectively
use this experience in order to improve national consistency in
permitting and enforcement.
In order to increase consistency in water quality-based
toxicity permitting, I am issuing the attached Basic Permitting
Principles for Whole Effluent Toxicity (Attachment 1) as a
standard with which water quality-based permits should conform.
A workgroup of Regional and State permitting, enforcement, and
legal representatives developed these minimum acceptable
requirements for toxicity permitting based upon national
experience. These principles are consistent with the toxics
control approach addressed in the proposed Section 304(1)
regulation. Regions should use these principles when reviewing
draft State permits. If the final Section 304(1) regulations
include changes in this area, we will update these principles as
necessary. Expanded guidance on the use of these principles will
be sent out shortly by James Elder, Director of the Office of

-------
Appendix B-4: Whole Effluent Toxicity Permitting Principles and Enforcement
Strategy

-------
Federal Regiatar / VoL 54. No. 108 / Friday, Juae 2. 1989 f Rulea and Regulation*
238%
Hie notice of approval and disapproval
•hail	the following:
(1) The name and address of the EPA
office that reviews the state's
submittals.
(ii)	A brief description of the section
304(1) process.
(iii)	A list of waters, point sources and
pollutants disapproved under this
paragraph.
(iv)	If the Regional Administrator
determines that a state did not provide
adequate public notice and an
opportunity to comment on the lists
prepared under this section, or if the
Regional Administrator chooses to
exercise his or her discretion, a list of
waters, point sources, or pollutants
approved under this paragraph.
(v)	The name, address, and telephone
number of the person at the Regional
Office from whom interested persons
may obtain more information.
(vi)	Notice that written petitions or
comments are due within 120 days.
(11) As soon as practicable, but not
later than June 4.199a the Regional
Office shall issue a response to petitions
or comments received under paragraph
(d)(10) of this section. Notice shall be
given in the same manner as notice
described in paragraph (d)(10) of this
section, except for the following changes
to the notice of approvals and
disapprovals;
(i)	The lists of waters, point souk
and pollutants must reflect any changes
made pursuant to comments or petitions
received.
(ii)	A brief description of the
subsequent steps in the section 304(1)
process shall be included.
(FR Doc. 80-13100 Filed 8-1-89: 8:45 am|

-------
2
Ft 2 Begtofr / VoL 54> No. IPS / Friday. Jmtm 2. T9SB / Rales tad Regulations
(i)	Water* whan fishing or shellfish
bans and/at advisorita an currently in
effect or are antfdpated.
(ii)	Waters where than have been
tested Hahuiu or whan
/normalities (cancer*, lesion*, tumors,
etc) have bees observed is fiah or other
aquatic life during the last ten years.
(iii)	Watan when there are
restzidfoBs an water sport* or
noeatloiial caatacL
(Iv) Waters identified by the state in
its most recent state section 305(b)
report a* either "partially achieving" or
"not achieving" designated uses.
(v) Waters identified by the state*
under section 303(d) of the CWA as
waters needing wstBrquBty-bsaed
controls.
Waters identified by the state as
priority weitubudluai (State Water
QtaBfy Management plans often include
priority wataroody Hate which am those
waters that moot need water poOotion
control derisions to adders water
quality standards or goals.)
(vii) Walesa when anbient data
indicate potential «r actaal exceed awes
afwatsrqaalHy aitaria dae to toxic
potiataDtsfrooanuidastxydaaBifisdas
apei—yindastry tnAppendh AaSm
CF& tat IS.
(vi±i) Wanes far which afflaeiB
toxicity Met rasaba iarifcrata peaaihla or
actual exceedaaosa of state water
ality standards* iadading narrative
ai from" water quality catena or £PA
water quality aitaria when state
criteria an not available.
(ix) Watan with primary industrial
major dischargers whan dilution
analyses indicate exceedancea of stata
narrative or numaric water quality
criteria (or &A water qualify criteria
when state standard* an not available]
for toxic pollutants, ammonia, or
chlorine. Thesa dilution analyses must
be based an estimates of discharge
levels derived from affluent gnideiines
development documents, NFDES
permits or permit application data (eg.
Form 2Q. Discharge Muni taring Reports
(DMRs), or otbsr evaflaMe information.
M Waters with POTW dbdurgen
requiring local pntreetaeat programs
when dilution analyses indicate
exceedancaa of atata watar quality
criteria (or EPA water qaattty criteria
wh— state watar qoahty criteria eee
not e leilaWa) fas fade pottgtants.
aoaonia. or eUflite. Ihaaa dtefloa
analysee most be baaed span data from
NPDES permit* or permit applications
(e.g, Farm 2C). nin.liaigu Mentoring
Reports (DMRs); or other available
xi) Watara with farifttiea not
eluded ia thepnvioaa two categories
such as major POTWa, and industrial
minor dischaigm when dilution
analyses indicate exceedancaa of
numeric or narrative stata water quality
criteria, (or EPA water quality criteria
whan state water quality criteria an
not available) for toxic pollutants,
ammonia, or chlorine. These dilution
analyses must be baaed upon estimates
of discharge levels derived from effluent
guideline development documents.
Nnjts permits or permit application
data. Discharge Monitoring Reports
(OMRs], or other available information.
(xii) Waters classified for uses that
will not sapport the "flshaUe/
swimmabie" goals of the Clean Water
Act
(xiii} Waters when ambient toxicity
or advose water quality condtttons
have been reported by locaL state. EPA
or other Federal Agendas, the private
sector, public interest groups, or
universities. These organizations and
groups should be actively •solicited for
research they may be conducting or
reporting. For example, university
reseadura, the United State*
Depailmmlof Agricnlture; the National
Oceanic and Atnoepherfc
Adamfetratioa the United States
Geological Sway, and the United
States Fish and WQiflife Service ere
good soarees of field date sad research.
(xiy) Waters identified by the state as
impaired in its most recent Clean Lake
Assessments conducted under section
314 of the Qean Water Act
(xv)	Waters identified as impaired by
nonpoint suintes in the America's Clean
Waterr 77m States' Noapaiat Scarce
Assessments 1985 (Association of State
and Interstate Watar Pollution Control
Adxninistretaea (ASTWPCAJ) or waters
{AmmtMmA mm iwrn^miwrnA rm	ft
nonpoint totem aaeesnnwit submitted
by lbs state to EPA mdsr section 31S of
the dean Water Act.
(xvi)	Sadase watan impaired by
polhaants front haraidoaa waste sites
on the National Pdeitty list prepared
undar section lOSfBK A> of CEBCLA.
jyj Each stata shall provide
documentation to tha Regional
Administrator to sapport tha state's
detenainatibn to list or not to hat waters
as requited by paragraphs (d)(1). (d)(2)
and (d)(3) ef this section. Ibis
documentation akalL be submitted to the
Regional AdminiaCmtar together with
the lists required by pangnpha (d)(1).
(d)(2). and (d)(3) of this section and shall
include as a mjnimuin;
(i) A description of the methodology
used to develop each list
Iii} A description of the data and
information nd to identify wnfem
sources indoding a description of the '
data and information uasd by the stata
as required by paragraph (d](8) of this
section;
(iii) A rationale for any decision not to
use any one of the categories of existing
and readily available data required by
paragraph (d)(0) of this section: and
flvj Any other information requested
by the Regional Administrator that :s
reasonable or necessary to determine
the adequacy of a state's lists. Upon
request by the Regional Administrator
each state must demonstrate good cause
for not including a water or waters on
one or more Bate. Good cause includes,
bet is not limited to. more recent or
accurate data; more accurate water
quality modeling; flaws in the original
analysis that led to the water being
identified in a category in S 130.10(d)(6);
or changes in conditions, e.g^ new
control equipment or elimination of
discharges.
(8)	Tha Regional Administrator shall
approve or disapprove each list required
by paragraphs (d)(1). (d)(2). and (d)(3) of
this section no later than June 4.1389.
Ilia Regional Administrator shall
approve each list required under
paragraphs (d)(1). (d)(2). and (d)(3) of
this section only if it maeta the
repiiatcry requirements for listing under
paragraphs (dXl). (d)U). nd (d)(3) of
this section and if the state has met all
the requirements of paragraphs (d)(6)
and (d)(7) of this section.
(9)	If a state fails to submit lists m
accordance with paragraph (d) of this
section or the Regional Administrator
does not approve the lists submitted by
such state in accordance with this
parapaph. then not later than June 4.
1900. the Regional Administrator. :n
coopamlun with soch state, shall
iapfcnaat the requirements of CWA
section 3040) (1) sod (2) in such state.
(10)	If the Regional Administrator
dlsapptusas a state's decision with
respect to one or more of the waters
raqirirad under parapaph (d) (1). (2). or
(3) of this section, or one or mon of the
individual control strategies required
pursoant to section 304(1)(1)(D), then not
later than June 4,1988. the Regional
Administrator shall distribute the notice
of approval or disapproval given under
this paragraph to the appropriate state
Director. Tha Regional Administrator
shall also publish a notice of
availabfltty. ia a daily or weakly
newspaper with state-wide circulation
or in the Federal Register, for the notice
of approval or disapproval The
Regional Administrator shall also
provide written notice to each
discharger identified under section
304(1)(1)(Q. that EPA has listed the
discharger undar section 304(1)(1)(C).

-------
Federal ia&Mtm / Vol H. No. 106 / Friday, [una a. 1900 / Rules and Regulation*
23&
issue a final permit on or before
February 4,1988, an individual control
strategy may be a draft permit with an
attached achedule (provided the State
meets the schedule for laming the final
permit) <»wtir»Mwg that the permit will be
issued on or before February 4,1980. If a
point source is subject to section
304(I)(1](C1 of the CWA and is also
subject to an on-site response action
under sections KM or 106 of the
Cumprahenatvo Environmental
Response, Compensation, and Liability
Act Of 1980 (CEROA), (42 UAC 9001 «t
saq.), tn iadhridual control strategy may
be the decision document (which
incorporates the applicable or relevant
and appropriate requirements under the
CWA) prepared under sections 104 or
108 of CERCLA to address the release or
threatened release of hazardous
substances to the environment
(d)	A petition submitted pursuant to
section 304(1X3) of the CWA must be
submitted to the appropriate Regional
Administrator. Petitions mast identify a
watarbody in sufficient detail so that
EPAissble to determine the location
and boundaries of the watarbody. The
Ktion most also identify the list or
i for which the watarbody qualifies,
and the petition must erpiatn why the
watarbody satisfies the criteria for
listing under CWA section 304(1) and 40
CFR 130.10(d)(8).
(e)	If the Regional Administrator
disapproves one or more individual
control strategies, or if a State fails to
provide adequate public notice and an
opportunity to comment on the ICS*,
thin, not later than June 4, I960, the
Regional Administrator shall give a
notice of approval or disapproval of the
individual control strategies submitted
by each State pursuant to this section as
follows:
(I)	The notice of approval or
disapproval given under this paragraph
ihilliadiidi thi foUowiu*
(i) The name and address of the EPA
office that lertews the State's
submittals,
(II)	A brief deecription of the section
304(1) process.
(ill) A list of ICSs disapproved under
this section and a finding that the ICSe
will not meet all applicable review
criteria under this section and section
304(1) of the CWA.
(iv) If the Regional Administrator
determines that a State did not provide
adequate public notice and an
opportunity to comment on the waters,
point sources, or ICSs prepared pursuant
to section 304(1). or if the Regional
Administrator chooses to exercise his or
her disaation, a list of the ICSs
approved under this section, and a
finding that the ICSs satisfy all
applicable review criteria.
(v)	The location where interested
penons may examine EPA's records of
approval and disapproval.
(vi)	The name, address, and telephone
number of the person at the Regional
Office from whom interested persons
may obtain more information.
(vii)	Notice that written petitions or
comments are due within 120 days.
(2)	The Regional Administrator shall
provide the notice of approval or
disapproval given under this paragraph
to the appropriate State Director. The
Regional Administrator shall publish a
notice of availability, in a daily or
weekly newspaper with State-wide
circulation or in the Federal Register, for
the notice of approval or disapproval.
The Regional Administrator snail also
provide written notice to each
discharger identified under section
304(1)(1)(C). that EPA has listed the
discharger under section 304(1X1)(C).
(3)	As soon aa practicable bat not
later than June 4.1900. the Regional
-Offices shall iseae a response to
petitions or comments received under
section 304(1). The response-to
comments shall be given in the same
manner as the notice desaibed in
paragraph (e) of this section except for
the following changes:
(1) The lists of ICSs reflecting any
changes medo pursuant to comments or
petitions received.
(ii) A brief deecription of the
subsequent steps tn the section 304(1)
prectss*
(!) EPA shall review, end approve or
disapprove, the individual control
streteges prepared under section 301(1)
of the CWA. using the applicable
criteria set forth in section 304(1) of the
CWA. end in 40 CFR Part 122, tnriuding
112144(d). At any time after the
Regions! Administrator disapprovee an
ICS (or conditionally aprosee a draft
penit ee en ICS), the Regional Office
may submit a written notification to the
State that the Regional Office intends to
issue the ICS Upon mailing the
notification, and notwithstanding any
other regulation, exclusive authority to
issue the permit paasss to EPA.
4. Section 123,83 is emended by
adding paragraph (a)(8) to reed aa
follows:
J12X83 Criteria tor wOMrewel of ewe
(•)***
(S) Where the State Sails to develop en
adequate regulatory program for
developing water quality^ased effluent
limits is NPDES permits.
PART 130—WATEB QUALITY
PLANNING AND MANAGEMENT
1.	The authority citation for Par
continues to read as follows:
Authority; 33 U.S.G 1231 at seq.
• • • • •
2.	Section 130.10 is amended by
adding paragraphs (d)(4), (d)(5), (d)(8).
(d)(7), (d)(8), (d)(9), (d)(l0). and (d)(ll) tc
read as follows:
113110 State sufanfttto to EPA.
Cd) • * '
(4)	For the purposes of listing waters
under S 130.10(d)(2), "applicable
standard" means a numeric criterion for
a priority pollutant promulgated as pan
of a state water quality standard. Where
a state numeric criterion for a priority
pollutant is not promulgated as part of a
state water quality standard, for the
purposes of listing waters "applicable
standard" means the state narrative
water quality aitenon to control a
priority pollutant (e.g* no toxics in toxic
amounts) interpreted on a chemical-by-
chemkal basis by applying a proposed
state drterion. aa explicit state policy or
regulation, or an EPA national water
quiriity criterion, supplemented with
other relevant information.
(5)	If a water meets either of the
conditions listed below the water u
be listed under S 130.10(d)(2) on the
grounds that the applicable standard is
not achieved or expected to be achieved
due entirely or substantially :o
discharges from point sources.
(i) Existing or additional water
quality-based limits on one or more
point souxcee would result in the
achievement of aa applicable water
quality standard for a toxic pollutant; or
(U) The discharge of a toxic pollutant
from one or more point sources,
regardless of any nonpoint source
oontribetion of the seme pollutant is
sufficient to cause or is expected to
cause an excursion above the applicable
water quality standard for the toxic
pollutant
(8) Each state shall assemble and
. evaluate all existing and readily
available water quality-related data and
information and eech state shall develop
the lists required by paragraphs (d)(1).
(2). and (3) of this section based upon
this data and information. At a
minimnm. all existing and readily
available water quality-related data and
information includes, but is not liar
to, all of the existing end readily
available data about the fallowing
categories of wsters in the state:

-------
23836
iOQATAi	/ V Oi«	iNO* 1UO / fflQtiyi	m i«?Q& / Ad^td hiiu Acgu*o»*ui-j
reasonable potential to cause, or
contribute to an excunion above any
State water quality standard, including
State narrative criteria for water quality.
(iij When determining whether a
Kicharge causes, has the reasonable
potential to cause, or contributes to an
in-stream excursion above a narrative or
numeric criteria within a State water
quality standard, the permitting
authority shall use procedures which
account for existing controls on point
and nonpoint sources of pollution, the
variability of the pollutant or pollutant
parameter in the effluent the sensitivity
of the species to toxicity testing (when
evaluating whole effluent toxicity), and
where appropriate, the dilution of the
effluent in the receiving water.
(iii)	When the permitting authority
determines, using the procedures in
paragraph (d)(l)(ii) of this section, that a
discharge causes, has the reasonable
potential to cause, or contributes to an
in-stream excunion above the allowable
ambient concentration of a State
numeric criteria within a State water
quality standard for an individual
pollutant the permit must contain
effluent limits for that pollutant
(iv)	When the permitting authority
dfitaninMi asina ths proctdurw in
paragraph (d)(l)(ii) of this section, that a
discharge cause*, has the reasonable
potential to cause, or contributes to an
'a-stream excunion above the numeric
riterion for whole effluent toxicity, the
permit must contain effluent limits for
whole effluent toxicity.
(v)	Except as provided in this
subparagraph, when the permitting
authority determines, using the
procedures in paragraph (d)(l)(ii) of this
section, toxicity testing data, or other
information, that a discharge causes, has
the reasonable potential to cause, tfr
contributes to an in-stream excunion
above a narrative ait*rioa within an
applicable State water quality standard,
the permit must	effluent limits
for whole effluent toxicity. Limits on
whole toxicity are not necessary
where the permitting authority
demonstrates in the fact sheet or
statement of basis of the NFDES permit
the procedures in paragraph
(d)(l)(ii) of this section, that chemical-
specific limits for the effluent an
sufficient to attain and
applicable numeric and narrative State
water quality standards.
(vi)	When a State has not established
a water quality criterion for a specific
chemical pollutant that is present in an
effluent at a concentration that causes,
has the reasonable potential to cause, or
tributes to en excursion above a
sanative criterion within an applicable
itata water quality standard, the
permitting authority must establish
effluent limits uaing one or more of the
following options:
(A)	Establish effluent limits using a
calculated numeric water quality
criterion for the pollutant which the
permitting authority demonstrates will
attain and maintain applicable narrative
water quality criteria and will fully
protect the designated use. Such a
criterion may be derived using a
proposed State criterion, or an explicit
State policy or regulation interpreting its
narrative water quality criterion,
supplemented with other relevant
information which may include: EPA's
Water Quality Standards Handbook.
October 1963. risk assessment data,
exposure data, information about the
pollutant from- the Food and Drug
Administration, and current EPA criteria
documents; or
(B)	Establish effluent limits on a case-
by-case basis, using EPA's water quality
criteria, published under section 307(a)
of the CWA. supplemented where
necessary by other relevant information:
or
(C)	Establish effluent limitations on an
indicator parameter for the,pollutant of
concern, provided:
(1)	The permit identifies which
pollutants are intended to be controlled
dv tht qm of thi flffluint limitation:
(2)	Hie fact sheet required by 112158
sets forth the basis for the limit
inputting a	that compliance with
the effluent limit on the indicator
parameter will result in controls on the
pollutant of concern which an sufficient
to attain and maintain applicable water
quality standards: •
(J) The permit requires all effluent and
ambient monitoring necessary to show
that during the term of the permit the
limit an the indicator parameter
continue* to attain and maintain
applicable water quality standards: and
(V) Th* permit	a reopener
clause allowing the permitting authority
to modify or revoke and reissue the
permit if th* limits on th* indicator
parameter no longer attain and maintain
applicable water quality standards.
(vii) When developing water quality-
based effluent limits under this
paragraph the permitting authority shall
ensun that
(A)	The level of water quality to be
achieved by limits on point sources
established under this paragraph is
derived from, and complies with all
applicable water quality standards; and
(B)	Effluent limits developed to
protect a narrative water quality
criterion, a numeric water quality
criterion, or both, are consistent with the
assumptions and requirements of any
available wasteload allocation for the
discharge prepared by the State and
approved by EPA pursuant to 40 CFR
130.7.
• • • • •
4. The title of paragraph (e) of § 122.44
is revised to read as follows:
• • • • •
(e) Technology-based controls for
toxic pollutants. * * *
PART 123—STATE PROGRAM
REQUIREMENTS
1.	The authority citation for Part 123
continues to read as follows:
Authority: Clean Water Act 33 U.S.C 1251
etseq.
2.	Section 123.44 is amended by
adding paragraph (c)(8) to read as
follows:
f 12X44 EPA review of and objectlona to
Stat* permits
(c) * * •
(8) The effluent limits of a permit fail
to satisfy the requirements of 40 CFR
122.44(d).
•	•	•	t	t
3.	In { 123.46 paragraph (a) is revised
and paragraphs (c). (d), (e) and (fj are
added, as follows:
1123.4* Individual controi strategies.
(a) Not later than February 4.1989.
each State shall submit to the Regional
Administrator for review, approval, and
implementation an individual control
strategy for each point source identified
by the State punuant to section
304(1)(1)(C) of the Act which will
produce a reduction in the discharge of
toxic pollutants from the point sources
identified under section 304(1)(1)(C)
through the establishment of effluent
limitations under section 402 of the
CWA and water quality standards
under section 303(c)(2)(B) of the CWA.
which reduction is sufficient in
combination with existing controls on
point and nonpoint sources of pollution,
to achieve the applicable water quality
standard as soon as possible, but not
later than three yean after the date of
the establishment of such strategy.
•	• • • •
(c) For the purposes of this section the
term individual control strategy, as set
forth in section 304(1) of the CWA.
means a final NPDES permit with
supporting documentation showing that
effluent limits are consistent with an
approved wasteload allocation, or other
documentation which shows that
applicable water quality standards will
be met not later than three years after
the individual control strategy is
established Where a State is unable to

-------
turn* degmm / Vol Sfc No. IPS / Friday, June 2. nag / Rni and Ragglattops
2SCS5
b. Judicial Rsview of Dedsioa Under
section 304(1). As EPA stated in (ha
proposal. judicial review of a
disapproved ICS under section 509(b) of
the CWA is not available until EPA
makes a final decision with respect to
the ICS, Le* a final decision on the EPA-
issued NPDES permit under Part 124 of
EPA's regulations. One comnenter
argued that EPA's position oa this issue
farther	iti dflficitloxi of fln
ICS by	909(b)(1)(G) redundant.
Within the limits of the Act the Agency
has tha discretion to define ICSs as
discussed above in Section Gl.Tho
Agency continues to belleva that its
is the best reading of the
statute, and Congress gava EPA
discretion in determining what exact
definition to use. Congress' addition a!
section 309(b)(1)(C) to the Act shows
Congress1 intent that EPA's
promulgation of an ICS. however
defined, be reviewed in the courts of
appeals more than any intent to
preclude or restrict EPA's interpretation
that an NPDES permit be an ICS.
Therefore, EPA believee that the posits
that EPA issues as ICSs are reviewable
in the court of appeals. Review of any
other actions by EPA under taction
304(1) mast be obtained in a district
court
IV.	EBbUivo Date
This rule bees me affective on May 28,
1989. Title S U3.C 553(d) provides that
regulations should take effect 30 days
after their publication in the Federal
Register, unlets EPA finds and publishes
good causa for a shorter time. In
determining that good cause exists in
this casa. EPA weighed the necessity for
an immediate effective date against
problems it would causa for those
subject to the rules. The need far an
effective date ansae from the
statute's dsariline for EPA decisions on
state submissions, ft is aiticsl that
today's rale be effective when EPA's
Ragfanal Offices make their decisions
under section 3040). Today's regulations
are important to ensure consistency end
certamty in regional decisions.
V.	Refulstocy Analyse
A. Executive Order 12201
Under section 3(b) of Executive Order
12291 the agency must judge whether a
regulation is major sad thus subject to
the requirements of a Regulatory Impact
Analysis. The regulation published
today is not msjor because the rale will
not result in an effect on the economy of
$100 million or more, will sot result in
increased costs or prices, will not have
significant adverse effects on
competition, employment Investment
productivity, and inxumttion. and will
not significantly disrupt domestic or
export markets.
EPA received comments arguing that
the regulations were in fact msjor. EPA
disagrees. The regulations specify what
factors states must use to determine
whether permits will achieve water
o.uality standards, while the
determination of what the standards are
and what permit limits are necessary to
comply with the standards remains
principally with the states. Compliance
with the water quality standards has
been required by the Clean Water Act
sine* July 1.1977. Many of the limits that
are imposed as a result of the
procedures in today's rale are to
implement standards that were in place
long ago. The reporting requirements
discussed in todiay's rule reqmre no
additional monitoring, and preparing the
reports will not result ia an effect on the
economy of S100 million or more.
Therefore, the Agency has not
prepared a Regulatory Impact Analysis
under the Executive Order. EPA
submitted this regulation to the Office of
Msnsgemaiil sad Badget (Oiffi) for
review as required by Executive Order
12291.
ILPapermui ReducticmAct
There is ¦> information. coUeetkni.
requirement after the effective date af-
this rulemaking, and. therefore, no
lafmlmtUo	rS^MSt
clearance are needed. An information
collection request for the proposed
rulemaking, submitted by EPA to the
Office of Mansgement and Budget*
(OMB) was disapproved by OMB
because the information had slready
been submitted to EPA pursuant to the
statutory deadline of February 41989,
and EPA did not formally submit the
ICR in a timely manner after the
piupossd rate was published (see PRA
tiooaS CFR1330113 (b) end (d)).
comments from OMB rsgsnlbm
ths ICR Cor the prapoeed rub-are
available hum tha Chief. information
Policy Branch. HC ITT U.S. EPA, 401M
Street SW„ Washington. DC 20460; and
tha Office of Information and Regulatory
Affairs. Office of Management and
Budget. Washington, DC 20303.
C Regulatory Flexibility Act
Under the Regulatory Flexibility Act
of 1980 (5 UAC 801 et seq.). Federal
agencies mast when developing
regulations, analyze their impact on
small entities (small businesses, small
government jurisdictions, snd small
organizations). This analysis ia
unnecessary, however, where the
agency's administrator certifies that tha
rule will not have a significant economic
effect on a substantial number of small
entities. The agency has concluded that
this rule will not have a significant
economic effect on a substantial number
of small entities because today's
rulemaking imposes no new
requirements for the regulated
community. Today's regulations merely
establish the procedures for
implementing section 30*(l) cf the
CWA and clarify certain elements of
EPA's surface water toxica control
program.
List of Subjects
40 CFR Part 122
EPA Administered Permit Programs:
The National Pollutant Discharge
Elimination System.
40 CFR Part 123
State program requirements.
40 CFR Part 130
Water quality planning and
Date May mass.
F. IIbbh Hishkht O,
Action Administrator.
FART 123—CPA ADMINISTERED
PERMTT PROGRAMS: THE NATIONAL
POLLUTANT DISCHARGE
SUMMATION SYSTEM
1.	The authority citation for Part::
continues to read as follows:
Authority: Ths Gean Water Act 3: U.S.C.
1231 at t*q.
2.	Section 1222 is amended by adding
in alphabetical order a new definition as
follows:
I t&j Oeflnttena.
Whole effluent toxicity means the
aggregate toxic effect of an effluent
measured directly by a toxicity test
3. Paiegraph (d)(1) of 1122.44 is
revised to end as follows:
>12144 IsttftWamnglBHtaeowa.
Ml o0Mf psnnft conditions
(appAeafels to Sttto NPOCS programs.
112X25V
• • • • •
(d) * • •
(1) Achieve water quality standards
established under section 303 of the
CWA including State narrative critera
for water quality..
(i) Limitations must control all
pollutants or pollutsnt parameters
(either conventional, nonconventional.
or toxic pollutants) which the Dirasj
determines are or may be discharger
a level which will causa, have the

-------
5
reissue the permits with water quality-based effluent limitations
which achieve water quality standards. The specific requirements
in §122.44(d) are structured in a way that implements EPA's
Policy for the Development of Water Quality-Based Permit
Limitations for Toxic Pollutants (49 Fed. Reg. 9016 March 9,
1984). Second, Regions will need to look closely at each state's
surface water toxics control program to ensure that the state's
regulations, policies and technical guidance result in the
consistent and comprehensive development of NPDES permits which
achieve the state's water quality standards. Where this does not
occur, each Region should work with the state to rectify the
problem and, after these negotiations and where necessary,
investigate the possibility of withdrawing the NPDES program.
I hope these regulations will assist you in developing water
quality-based effluent limits and will support your efforts to
implement surface water toxics control programs. If you have
questions or need more information about these requireraents/
please contact Cynthia Dougherty at FTS 475-9545 or have your
staff contact Rick Brandes at FTS 475-9537.
cc: Permits Branch Chiefs, Regions I - X
Martha Prothro, OWRS

-------
ST,,
	 sr
} A \	UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
'	WASHINGTON. D.C. 20460

AUG 2 I ;C53	o^ejo.
MEMORANDUM
SUBJECT: New Regulations Governing Water Quality-Based
Permitting in the NPpES Permitting Program
FROM:	Jame*^1T7'*in.d^'r-r "Director
Office of Water Enforcement
and Permits
TO:
Water Management Division Directors
Regions I - X
On May 26# 1989 the Deputy Administrator signed regulations
that implement section 304(1) of the CWA. The regulations became
effective upon his signature and were published in the Federal
Register on June 2, 1989 (54 Fed. Reg. 23868). This rulemaking
also clarified and reinforced EPA's existing regulations
governing water quality-based permitting. The purpose of this
memorandum is to describe the significance of these
clarifications to EPA's baseline water quality-based permitting
regulations.
CHANGES TO 40 C.F.R. PART 122
Section 122.44 covers the establishment of limitations,
standards, and other permit conditions in NPDES permits.
Subsection (d) covers water quality standards and state
requirements. Prior to the promulgation of these new regulations
the Subsection was non-specific, requiring only that NPDES
permits be issued with requirements more stringent than
promulgated effluent guidelines as necessary to achieve water
quality standards. We have strengthened considerably the
requirements of §122.44(d). The new language is very specific
and requires water quality-based permit limits for specific
toxicants and whole effluent toxicity where necessary to achieve
state water quality standards. The following is a section-by-
section description of the new requirements.

-------
2
1.	§122.44(d)(1)(i)
This new paragraph provides that all pollutants that cause,
have the reasonable potential to cause, or contribute to an
excursion above a water quality standard must be controlled to
achieve all applicable water quality standards, including
narrative water quality criteria. We added this paragraph so
that our regulations would reflect EPA's approach to water
quality-based permitting.
2.	§122.44(d)(1)(ii)
Subparagraph (ii) of the new regulations requires the states
to use valid procedures to determine whether a discharge causes,
has the reasonable potential to cause, or contributes to an
excursion above a water quality standard. These procedures must
account for existing controls on point and nonpoint sources of
pollution, the variability of the pollutant in the effluent, the
sensitivity of the test species (when evaluating whole effluent
toxicity), and where allowed by state water quality standards,
the dilution of the effluent in the receiving water. The purpose
of this new regulation is to require the states to use
technically valid procedures when determining whether a discharge
is exceeding a numeric or narrative water quality criterion.
When the permitting authority determines, using these procedures,
that a discharge causes, has the reasonable potential to cause,
or contributes to an excursion above a water quality criterion,
that permit must include one or more water quality-based effluent
limits established under subparagraphs (iii) - (vi).
Subparagraphs (iii) and (iv) deal with water quality-based
limitations where the state has numeric water quality criteria;
subparagraphs (v) and (vi) deal with a state's narrative water
quaity criteria.
3.	§122.44(d)(1)(iii)
This paragraph requires NPDES permits to include effluent
limitations for every individual pollutant that causes, has the
reasonable potential to cause, or contributes to an excursion
above a numeric water quality criterion. Thus, when a state has
adopted a water quality criterion for an individual pollutant and
the state determines under subparagraph (ii) that an effluent
limit is necessary, subparagraph (iii) requires an effluent limit
for that individual pollutant.
4.	§122.44(d)(1)(iv)
Subparagraph (iv) requires effluent limitations on whole
effluent toxicity when a discharge is exceeding a state numeric
criteria for whole effluent toxicity. This paragraph is applied

-------
3
where a state has adopted a numeric criterion for whole effluent
toxicity (e.g. a discharge must achieve an LC50 of 50% or
higher).
5.	§122 .44(d)(1)(v)
When the state determines that a discharge exceeds a
narrative water quality criterion, subparagraph (v) requires
effluent limitations on Whole effluent toxicity. If, however,
chemical-specific effluent limitations are demonstrated to be
sufficient to achieve all applicable water quality standards,
then subparagraph (v) allows the permitting authority to forego a
limitation on whole effluent toxicity. It may be necessary for
you to work with an individual state to ensure that they have the
necessary protocols to support whole effluent toxicity limits.
6.	§122.44(d)(l)(vi)
Where an actual or projected excursion above a narrative
water quality criterion is attributable to a particular
pollutant, but the state has not adopted a water quality
criterion for the pollutant of concern, this new regulation
requires water quality-based effluent limitations which will
control the pollutant of concern. Subparagraph (vi) establishes
three options for developing such limitations. Under these
options, a state may: 1) calculate a numeric criterion for the
pollutant; 2) use EPA's water quality criterion for the pollutant
of concern; or 3) establish effluent limits on an indicator
parameter.
By an indicator parameter we mean a pollutant or pollutant
parameter for which control of this indicator will result in
control of the pollutant of concern. For example, it may be
shown that a more stringent control on total suspended solids
will reduce discharge of a metal to a level which achieves the
water quality standard. Subparagraph (vi) also sets out four
provisions which must be met to allow the use of an indicator:
1)	The permit must identify which pollutants are intended
to be controlled by a limit on the indicator parameter.
2)	The fact sheet must set forth the basis for the limit,
including a finding that compliance with the limit will
result in controls on the pollutant of concern that are
sufficient to achieve the water quality standard.
3)	The permit must require all monitoring necessary to
show continued compliance with water quality standards.

-------
4
4) The permit must contain a reopener clause allowing for
changes in the permit as needed to achieve water
quality standards.
A state's narrative water quality criterion serves as the legal
basis for establishing such effluent limits.
7. §122.44(d)(l)(vii)
Subparagraph (vii) requires that all water quality-based
effluent limitations adhere to two fundamental principles: 1) the
effluent limitations must be derived from and comply with all
applicable water quality standards; and 2) the effluent
limitations are consistent with the assumptions and requirements
of an applicable wasteload allocation (WLA) if a WLA is available
for the pollutant.
CHANGES TO 40 C.F.R. PART 123
We amended the permit objection regulations at 40 C.F.R. .
§123.44 to reflect the amendments to §122.44(d)(1). Under
§123.44(c)(8) EPA may now object to a state-issued permit if the
permit does not meet the requirements of §122.44(d)(1). Thus, if
a state does not use technically sound procedures for evaluating
the need for water quality-based effluent limitations then EPA
may object to the permit. Also, if a state fails to include
chemical-specific or whole effluent toxicity limitations in a
permit as required by paragraphs (iii) - (vi), then EPA may
object to the permit. Finally, if a water quality based effluent
limitation is not derived according to the principles in
subparagraph (vii) then EPA may object to the permit.
If a state's surface water toxics control program is not
adequate to implement these requirements, the new regulations at
40 C.F.R. §123.63 expand EPA's criteria for withdrawing a state's
NPDES program. Under the new regulations (§123.63(a)(5)), EPA
may withdraw a state's NPDES program if the state fails to
develop an adequate regulatory program for developing water
quality-based effluent limitations. In November 1987,
Headquarters provided procedural and technical guidance to the
Regions on conducting state toxics control program reviews to
assess the adequacy of state water quality-based control
programs. This guidance sets guidelines for assessing whether or
not a state's regulations, policies, and technical guidance
constitute an adequate program.
The significance of these additions to Part 123 is twofold.
First, the Regions must issue permits which comply with these
requirements and must work with the NPDES states to insure they
also issue permits which comply with these regulations. If the
states do not issue permits consistent with Part 123, the Region
must veto insufficient permits and work with the states to

-------
Appendix B-3: New Regulations Governing Water Quality-Based Permitting

-------
01/19/89
COMPLIANCE MONITORING AND ENFORCEMENT STRATEGY
FOR TOXICS CONTROL
I. Background
Issuance of NPDES permits now emphasises the control of toxic
pollutants, by integrating technology and water quality-based
permit limitations, best management practices for toxic discharges,
sludge requirements, and revisions to the pretreatment implementa-
tion requirements. These requirements affect all major permittees
and those minor permittees whose discharges may contribute to
impairment of the designated use for the receiving stream. The
goal of permitting is to eliminate toxicity in receiving waters
that results from industrial and municipal discharges.
Major industrial and municipal permits will routinely contain
water quality-based limits for toxic pollutants and in many cases
whole effluent toxicity derived from numerical and narrative
water quality standards. The quality standards to establish NPDES
permit limits are discussed in the "Policy for the Development of
Water Quality-based Permit Limits for Toxic Pollutants," 49FR 9016,
March 9, 1984. The Technical Support Document for Water Quality-
based Toxics Control, EPA >440/44-85032, September, 1985 and the
Permit Writer's Guide to Water Quality-based Permitting for Toxic
Pollutants, Office of Water, May, 1987, provide guidance for inter-
preting numerical and narrative standards and developing permit
limits.
The Water Quality Act (WQA) of 1987 (PL 100-4, February 4,
1987) further directs EPA and the States to identify waters that
require controls for toxic pollutants and develop individual
control strategies including permit limits to achieve control of
toxics. The WQA established deadlines, for individual control
strategies (February 4, 1989) and for compliance with the toxic
control permit requirements (February 4, 1992). This Strategy
will support the additional compliance monitoring, tracking, evalu-
ation, and enforcement of the whole effluent toxicity controls
that will be needed to meet the requirements of the WQA and EPA's
policy for water quality-based permitting.
It is the goal of the Strategy to assure compliance with
permit toxicity limits and conditions through compliance inspec-
tions, compliance reviews, and enforcement. Water quality-based
limits may include both chemical specific and whole effluent toxi-
city limits. Previous enforcement guidance (e.g., Enforcement
Management System for the National Pollutant Discharge Elimination
System, September, 1986; National Guidance for Oversight of NPDES
Programs, May, 1987; Guidance for Preparation of Quarterly and
Semi-Annual Noncompliance Reports, March, 1986) has dealt with

-------
- 2 -
chemical-specific water quality-based limits. This Strategy will
focus on whole effluent toxicity limits. Such toxicity limits may
appear in permits, administrative orders, or judicial orders.
II. Strategy Principles
This strategy is based on four principles!
1)	Permittees are responsible for attaining, monitoring,
and maintaining permit compliance and for the quality
of their data.
2)	Regulators will evaluate self-monitoring data quality
to ensure program integrity.
3)	Regulators will assess compliance through inspections,
audits, discharger data reviews, and other independent
monitoring or review activities.
4)	Regulators will enforce effluent limits and compliance
schedules to eliminate toxicity.
Ill. Primary Implementation Activities
In order to implement this Strategy fully, the following
activities are being initiated:
A. Immediate development
1.	The NPDES Compliance Inspection Manual was
revised in May 1988 to include procedures for
performing chronic toxicity tests and evaluating
toxicity reduction evaluations. An inspector
training module^waa also developed in August
1988 to support inspections for whol# effluent
toxicity.
2.	The Permit Compliance System (the national NPDES
data base) was modified to allow inclusion
of toxicity limitations and compliance schedules
associated with toxicity reduction evaluations.
The PCS Steering Committee will review standard
data elements and determine if further modifi-
cations are necessary.
3.	Compliance review factors (e.g., Technical
Review Criteria and significant noncompliance
definitions) are being proposed to evaluate
violations and appropriate response.
4.	A Quality Assurance Fact Sheet has been developed
(Attached) to review the quality of toxicity test
results submitted by permittees.

-------
- 3 -
5. The Enforcement Response Guide in the Enforcement
Management System will be revised to cover the use
of administrative penalties and other responses to
violations of toxicity controls in permits. At
least four types of permit conditions are being
examined: (1) whole-effluent toxicity monitoring
(sampling and analysis), (2) whole effluent
toxicity-based permit limits, (3) schedules to
conduct a TRE and achieve compliance with water
quality-based limits, and (4) reporting requirments.
B. Begin development in Spring 1989
With the assistance of the Office of Enforcement and
Compliance Monitoring (OECM), special remedies and model forms
will be developed to address violations of toxicity permit
limits (i.e., model consent decrees, model complaints, revised
penalty policy, model litigation reports, etc.)
IV. Scope and Implementation of Strategy
A. Compliance Tracking and Review
1. Compliance Tracking
The Permits Compliance System (PCS) will be
used as the primary system for tracking limits and
monitoring compliance with the conditions in NPDES
permits. Many new codes for toxicity testing have
already been entered into PCS. During FY 89, head-
quarters will prbvide additional guidance to Regions
and States on PCS coding to update existing documenta-
tion. The Water Enforcement Data Base (WENDB)
requirements as described in the PCS Policy Statement
already require States and Regions to begin
incorporating toxicity limits and monitoring information
into PCS.
In addition to guidance on the use of PCS,
Headquarters has prepared guidance in the form
of Basic Permitting Principles for Regions and
States that will provide greater uniformity
nationally on approaches to toxicity permitting.
One of the major problems in the tracking and
enforcement of toxicity limits is that they differ
greatly from State-to-State and Region-to-Region.
The Permits Division and Enforcement Division in
cooperation with the PCS Steering Committee will
establish standard codes for permit limits and
procedures for reporting toxicity results based on
this guidance.

-------
- 4 -
Whole effluent toxicity self-monitoring data
should undergo an appropriate quality review. (See
attached checklist for suggested toxicity review
factors.) All violations of permit limits for
toxics control should be reviewed by a professional
qualified to assess the noncompliance. Regions and
States should designate appropriate staff.
2. Compliance Review
Any violation of a whole effluent toxicity
limit is of concern to the regulatory agency and
should receive an immediate professional review.
In terms of the Enforcement Management System (EMS),
Any whole effluent violation will have a violation
review action criterion (VRAC) of 1.0. However, the
appropriate initial enforcement response may be to
require additional monitoring and then rapidly
escalate the response to formal enforcement if the
noncompliance persists. Where whole effluent
toxicity is based on a pass-fail permit limitation,
any failure should be immediately targeted for
compliance inspection. In some instances, assessment
of the compliance status will be required through
issuance of Section 308 letters and 309(a) orders to
require further toxicity testing.
Monitoring data which is submitted to fulfill
a toxicity monitoring requirement in permits that do
not contain an independently enforceable whole-effluent
toxicity limitation should also receive immediate
professional review.
The burden for testing and biomonitoring is on
the permittee; however, in some instances. Regions and
States may choose to respond to violations through
sampling or performance audit inspections. When an
inspection conducted in response to a violation identi-
fies noncompliance, the Region or State should
initiate a formal enforcement action with a compliance
schedule, unless remedial action is already required
in the permit.
B. Inspections
EPA/State compliance inspections of all major permittees
on an annual basis will be maintained. For all facilities
with water quality-based toxic limits, such inspections should
include an appropriate toxic component (numerical and/or
whole effluent review). Overall the NPDES inspection and
data quality activities for toxics control should receive
greater emphasis than in the present inspection strategy.

-------
- 5 -
1.	Regional/State Capability
The EPA's "Policy for the Development of Water
Quality-based Permit Limits for Toxic Pollutants"
(March 9, 1984 Federal Register) states that EPA
Regional Administrators will assure that each
Region has the full capability to conduct water
quality assessments using both biological and chemi-
cal methods and provide technical assistance to the
States. Such capability should also be maintained
for compliance biomonitoring inspections and toxics
sampling inspections. This capability should include
both inspection and laboratory capability.
2.	Use of Nonsampling Inspections
Nonsampling inspections as either compliance
evaluations (CEIs) or performance audits (PAIs) can
be used to assess permittee self-monitoring data
involving whole effluent toxicity limits, TREs, and
for prioritization of sampling inspections.* As
resources permit, PAIs should be used to verify
biomonitoring capabilities of permittees and
contractors that provide toxicity testing self-
monitoring data.
3.	Quality Assurance
All States are encouraged to develop the
capability for acute and chronic toxicity tests
with at least one fish and one invertebrate species
for freshwater and saltwater if appropriate. NPDES
States should develop the full capability to assess
compliance with the permit conditions they establish.
EPA and NPDES States will assess permittee
data quality and require that permittees develop
quality assurance plans. Quality assurance plans
must be available for examination. The plan should
include methods and procedures for toxicity testing
and chemical analysis; collection, culture, mainte-
nance, and disease control procedures for test
organisms; and quality assurance practices. The
* Due to resource considerations, it is expected that sampling
inspections will be limited to Regional/State priorities in
enforcement and permitting. Routine use of CEIs and PAIs should
provide the required coverage.

-------
- 6 -
permittee should also have available quality control
charts, calibration records, raw test data, and
culture records.
In conjunction with the QA plans, EPA will
evaluate pernittee laboratory performance on EPA
and/or State approved aethods. This evaluation is
an essential part of the laboratory audit process.
EPA will rely on inspections and other quality
assurance measures to maintain data quality. However,
States may prefer to implement a laboratory certifi-
cation program consistent with their regulatory
authorities. Predetermined limits of data accepta-
bility will need to be established for each test
condition (acute/chronic), species-by-species.
C. Toxicity Reduction Evaluations (TREs)
TREs are systematic investigations required of permittees
which combine whole effluent and/or chemical specific testing
for toxicity identification and characterization in a planned
sequence to expeditiously locate the source(s) of toxicity and
evaluate the effectiveness of pollution control actions and/or
inplant modifications toward attaining compliance with a permit
limit. The requirement for a TRE is usually based on a
finding of whole effluent toxicity as defined in the permit.
A plan with an implementation schedule is then developed to
achieve compliance. Investigative approaches include
causative agent identification and toxicity treatability.
1.	Requiring TRE Plans
TRE's can be triggered: 1) whenever there is a
violation of a toxicity limit that prompts enforcement
action or 2) from a permit condition that calls for a
toxicity elimination plan within a specified time
whenever toxicity is found. The enforcement action
such as a 309(a) administrative order or State
equivalent, or judicial action then directs the
permittee to take prescribed steps according to a
compliance schedule to eliminate the toxicity. This
schedule should be incorporated into the permit, an
administrative order, or judicial order and compliance
with the schedule should be tracked through PCS.
2.	Compliance Determination Followup
Compliance status must be assessed following the
accomplishment of a TRE plan using the most effi-
cient and effective methods available. These methods
include site visits, self-monitoring, and inspections.

-------
- 7 -
Careful attention to quality assurance will assist in
minimizing the regulatory burden. The method of
compliance assessment should be determined on a
case-by-case basis.
0. Enforcing Toxic Control Permit Conditions
Enforcement of toxic controls in permits depends upon a
clear requirement and the process to resolve the noncompli-
ance. In addition to directly enforceable whole effluent
limits (acute and chronic, including absolute pass-fail
limits), permits have contained several other types of
toxic control conditions! 1) "free from" provisions,
2) schedules to initiate corrective actions (such as TREs)
when toxicity is present, and/or 3) schedules to achieve
compliance where a limit is not currently attained.
Additional requirements or schedules may be developed
through 308 letters, but the specific milestones should be
incorporated into the permit, administrative order or
State equivalent mechanism, or judicial order to ensure
they are enforceable.
1. The Quarterly Noncompliance Report (QNCR)
Violations of permit conditions are tracked and
reported as follows:
a.	Effluent Violations
Each exceedance of a directly enforceable whole
effluent toxicity limit is of concern to the
regulatory agency and, therefore, qualifies
as meeting the VRAC requiring professional
review (see section IV.A.2.).
These violations must be reported on the QNCR
if the violation is determined through profes-
sional review to have the potential to have
caused a water quality impact.
All QNCR-reportable permit effluent violations
are considered significant noncompliance (SNC).
b.	Schedule Violations
Compliance schedules to meet new toxic controls
should be expeditious. Milestones should be
established to evaluate progress routinely and
minimize delays. These milestones should be
tracked and any slippage of 90 days or more
must be reported on the QNCR.

-------
- 8 -
The following milestones are considered SNC when
90 days or more overdue» subnit plan/schedule
to conduct TRE, initiate TRE, submit test results,
submit implementation plan/schedule (if appro-
priate} , start construction, end construction,
and attain compliance with permit.
c. Reporting/Other Violations
Violation of other toxic control requirements
(including reports) will be reported using
criteria that are applied to comparable NPDES
permit conditions. For example, failure to
submit a report within 30 days after the due
date or submittal of an inaccurate or inadequate
report will be reportable noncompliance (on
the QNCR).
Only failure to submit toxicity limit self-
monitoring reports or final TRE progress reports
indicating compliance will be SNC when 30 days
or more overdue.
Resolution (bringing into compliance) of all three
types of permit violations (effluent, schedule,
and reporting/other) will be through timely and
appropriate enforcement that is consistent with
EPA Oversight Guidance. Administering agencies
are expected to bring violators back into compliance
or take formal enforcement action against facilities
that appear on the QNCR and are in SNC; otherwise,
after two or more quarters the facility must be
listed on the Exceptions List.
2. Approaches to Enforcement of Effluent Limitations
In the case of noncompliance with whole effluent
toxicity limitations, any formal enforcement action
will be tailored to the specific violation and remedial
actions required. In some instances, a Toxicity
Reduction Evaluation (TRE) may be appropriate. However,
where directly enforceable toxicity-based limits are
used, the TRE is not an acceptable enforcement response
to toxicity noncompliance if it requires only additional
¦onitoring without a requirement to determine appropriate
remedial actions and ultimately compliance with the
limit.
If the Regions or States use administrative
enforcement for violations of toxic requirements,
such actions should require compliance by a date
certain, according to a set schedule, and an

-------
- 9 -
administrative penalty should be considered.!
Failure to comply with an Administrative Order
schedule within 90 days indicates a schedule delay
that may affect the final compliance date and a
judicial referral is the normal response. In instances
where toxicity has been measured in areas with potiential
impacts on human health (e.g., public water supplies,
fish/shellfish areas, etc.), regions and states
should presume in favor of judicial action and seek
immediate injunctive relief (such as temporary
restraining order or preliminary injunction).
In a few highly unusual cases where the permit-
tee has implemented an exhaustive TRE plan^, applied
appropriate influent and effluent controls^, maintained
continued compliance with all other effluent limits,
compliance schedules, monitoring, and other permit
requirements, but is still unable to attain or maintain
compliance with the toxicity-based limits, special
technical evaluation may be warranted and civil penalty
relief granted. Solutions in these cases could be
pursued jointly with expertise from EPA and/or the
States as well as the permittee.
Some permittees may be required to perform a
second TRE subsequent to implementation of remedial
action. An example of the appropriate use of a
subsequent TRE is for the correction of new violations
of whole effluent limitations following a period of
^Federal Administrative penalty orders must be linked to violations
of underlying permit requirements and schedules.
^See Methods for Aquatic Toxicity Identification Evaluations,
Phase I, Toxicity Characterization Procedures, EPA-600/3-88/035,
Table 1. An exhaustive TRE plan covers three areas: causative
agent identification/toxicity treatability; influent/effluent
control; and attainment of continued compliance. A listing of
EPA protocols for TREs can be found in Section V (pages 11 and
12).
^For industrial permittees, the facility must be well-operated
to achieve all water quality-based, chemical specific, or BAT
limits, exhibit proper 0 & M and effective BMPs, and control
toxics through appropriate chemical substitution and treatment.
For POTW permittees, the facility must be well-operated to
achieve all water quality-based, chemical specific, or secondary
limits as appropriate, adequately implement its approved pretreat-
ment program, develop local limits to control toxicity, and
implement additional treatment.

-------
10 -
sustained compliance (6 months or greater in duration)
indicating a different problem from that addressed
in the initial TRE.
3* Enforcement of Compliance Schedule and Reporting
Requirements
In a number of instances, the primary
requirements in the permits to address toxicity
will be schedules for adoption and implementation
of biomonitoring plans, or submission of reports
verifying TREs or other similar reporting require-
ments. Regions and States should consider any
failure (1) to conduct self-monitoring according
to EPA and State requirements, (2) to meet TRE
schedules within 90 days, or (3) to submit reports
within 30 days of the specified deadline as SNC.
Such violations should receive equivalent enforce-
ment follow-up as outlined above.
4.	Use of Administrative Orders With Penalties
In addition to the formal enforcement actions
to require remedial actions. Regions and States
should presume that penalty AO's or State equiva-
lents can be issued for underlying permit violations
in which a formal enforcement action is appropriate.
Headquarters will also provide Regions and States
with guidance and examples as to how the current
CWA penalty policy can be adjusted.
5.	Enforcement Models and Special Remedies
OWEP and OECM will develop standard pleadings
and language for remedial activities and compliance
milestones to assist Regions and States in addres-
sing violations of toxicity or water quality-based
permit limits. Products will include model litiga-
tion reports, model complaints and consent decrees,
and revised penalty policy or penalty algorithm
and should be completed in early FY 1989.

-------
- 11 -
V. Summary of Principal Activities and Products
A.	Compliance Tracking and Review guidance
1.	PCS Coding Guidance - May, 1987; revision
2nd Quarter 1989
2.	Review Criteria for Self-monitoring Data (draft
attached)
B.	Inspections and Quality Assurance
1. Revised NPDES Compliance Inspection Manual -
May 1988.
2* Quality Assurance Guidance - 3rd Quarter FY 1989.
3.	Biomonitoring Inspection Training Module -
August 1988.
4.	Additions of a reference toxicant to DMRQA program -
(to be determined)
C.	Toxics Enforcement
1. Administrative and Civil Penalty Guidance - 4th
Quarter FY 1989
2., Model Pleadings and Complaints - 2nd Quarter 1989
3. EMS Revision - 2nd Quarter FY 1989
D.	Permitting Consistency^
1. Basic Permitting Principles - 2nd Quarter FY 1989
E.	Toxicity Reduction Evaluations
1.	Generalized Methology for Conducting Industrial
Toxicity Reduction Evaluations - 2nd Quarter
FY 1989
2.	Toxicity Reduction Evaluation Protocol for
Municipal Wastewater Treatment Plants - 2nd Quarter

-------
12 -
3. Methods for Aquatic Toxicity Indentlfication
Evaluations
Phase I.
b. Phase 11.
Toxicity Characterization
Procedures, EPA-600/3-88/034-
September 1988
Toxicity Identification
Procedures» EPA-600/3-88/035-
2nd Quarter 1989
Phase III. Toxicity Confirmation Procedures*
EPA-600/3-88/036- 2nd Quarter
FY 1989

-------
Attachment
QUALITY CONTROL FACT SHEET FOR SELF-BIOMONITORING
ACUTE/CHRONIC TOXICITY TEST DATA
Permit No. 	
Facility Name		
Facility Location 	
Laboratory/Investigator _____	
Permit Requirements:
Sampling Location 	 Type of Sample	
Limit 		Te8t Duration 	
Type of Test 		Test Organism Age 	
Test Results:
LC50/EC50/NOEL 	 95% Confidence Interval 	
Quality Control Summaryi
Date of Sample:	Dates of Test: 			
Control Mortality:	%	Control Mean Dry Weight 	
Temperature maintained within +2»C of test temperature? Yes No
Dissolved oxygen levels always greater than 40% saturation?
Yes	 No	
Loading factor for all exposure chambers less than or equal to
maximum allowed for the test type and temperature? Yes	 No
Do the test results indicate a direct relationship between effluent
concentration and response of the test organism (i.e.# more deaths
occur at the highest effluent concentrations)? Yes No

-------
Appendix B-5: Quality Control Fact Sheets for Self-Biomonitoring Acute and
Chronic Toxicity Test Data

-------
Attachment
QUALITY CONTROL FACT SHEET FOR SELF-BIOMONITORING
ACUTE/CHRONIC TOXICITY TEST DATA
Permit No. 	
Facility Name 	
Facility Location 	
Laboratory/Investigator 	
Permit Requirements:
Sampling Location 		Type of Sample	
Limit 		Test Duration 	
Type of Test 		Test Organism Age
Test Results?
LC50/EC50/NOEL 		95% Confidence Interval 	
Quality Control Sbmmary:
Date of Sample: 		Dates of Test: 	
Control Mortality:	%	Control Mean Dry Weight 	
t
Temperature maintained within +2-C of test temperature? Yes	 No
Dissolved oxygen levels always greater than 40% saturation?
Yes	 No	
Loading factor for all exposure chambers less than or equal to
maximum allowed for the test, type and temperature? Yes	 No	
Do the test results indicate a direct relationship between effluent
concentration and response of the test organism (i.e., more deaths
occur at the highest effluent concentrations)? Yes No	

-------
Appendix B-6: Important Case Decisions Regarding Whole Effluent Toxicity

-------
Case Summary:
Natural Resources Defense Council v. EPA. 859 F.2d 156 (D.C. Cir. 1988).
The consolidated cases arose from EPA's promulgation of NPDES
regulations. In discussing state program requirements, the Court in NRDC
analyzed the legislative history of the CWA. Particularly in relation to
effluent limitations, the Court found that:
Uniformity is indeed a recurrent theme in the Act, a direct
manifestation of concern that the permit program be standardized
to avoid the "industrial equivalent of forum shopping" and the
creation of "pollution havens" by migration of dischargers to areas
having lower pollution standards.
NRDC 859 F.2d at 174, and accompanying footnotes numbered 17-20
referencing, in part, 1972 Legislative History at 356, 263, 452-53, 473, 577,
1472, 172, 309, 378. The Court's discussion of EPA's refusal to provide an
affirmative upset defense to noncompliance with water quality-based limits
is significant since much emphasis is placed on Permittee's need to satisfy
"any more stringent limitation ... necessary to meet water quality
standards" under CWA §301(b)(l)(C). Congress* "firm conviction of need
for technology-forcing measures" to upgrade waters was underscored
(NRDC. 895 F.2d at 208), although in this instance, the EPA was found to
have acted arbitrarily in summarily dismissing the defense as impracticable.
NRDC 895 F.2d at 210.
More importantly, the Court of Appeals held that EPA has the authority
to frame effluent limits in terms of toxicity. The fact that toxicity was an
attribute of pollutants rather than a pollutant itself was found not to
preclude the Agency's use of toxicity as a measure to regulate effluents
that are pollutants. The Court disagreed with industry's assertions that
• EPA's 1984 policy statement ("Development of Water Quality-Based
Permit Limitations for Toxic Pollutants: National Policy," 49 Fed. Reg.
9016 (March 9, 1984)) and draft technical support document (draft "TSD")
were "rules" requiring notice and comment. TTie Court noted that informal
rulemaking regarding 40 CFR 1253(c)(4), which was pending between
1980 and 1984, did not limit agency information gathering to the issuance
of new or revised notices of proposed rulemaking. Unlike the VHS model
at issue in McLouth Steel Products Corp. v. Thomas. 838 F.2d 1317, 1324-
25 (D.C Cir. 1988), which was found to constitute a legislative rule
promulgated without adherence to §553 of the Administrative Procedure
Act ("APA"),5 U.S.C. §553 (1982), the EPA national policy and TSD were
not binding norms but rather general statements of policy leaving the
agency discretionary power in individual cases, open to challenges
regarding various applications and uses. Biomonitoring was to be used
where "appropriate"on a case by case basis.

-------
APPENDIX C.
Ambient Toxicity Testing and Data Analysis

-------
Ambient Toxicity Testing and Data Analysis
Ambient Toxicity Analysis
Ambient toxicity testing procedures are useful where measurement of toxicity levels after
discharge is important in the assessment of toxic effluent impact. This is particularly
true where impact is caused by the presence of multiple point sources. The purpose of
this testing is to provide an analysis of toxicity levels instream from whatever sources of
toxicity are affecting the receiving water.
Procedures
The basic ambient toxicity testing procedure is to expose test organisms to receiving
water samples taken from selected sampling stations above, at, and below the discharge
point(s). Since effluent concentrations after discharge are often relatively low, chronic
toxicity tests should be conducted so that the tests are sensitive enough for the purpose.
The methods available for chronic testing of sufficiently short duration are limited. Two
organisms for which short term chronic toxicity tests are available are Pimephales
promelas and Ceriodaphnia sp.
The following Procedures are used:
o Select instream sampling stations based on the mixing characteristics involved in
the specific discharge situation,
o Collect a daily grab sample or a daily composite sample of receiving water from
each station.
o Use a renewal testing method to expose test organisms to the daily samples
collected at each station. Use an appropriate number of replicates (10 for
Ceriodaphnia^ for each sampling station. No dilution series is required where
screening is the primary goal.
o Testing must be conducted at a low flow period. However, it is not necessary to
conduct the tests at the critical low flow period. Testing is best when relatively
stable flow occurs during the test period.
o Record the results of the testing in the format shown in Table C-l. The survival
of the test organisms and the effect on their growth or reproduction are used as
endpoints. Figure C-l plots the results in graphic form so that the pattern of
ambient toxicity can be observed.

-------
Table C-1. Young Production and Percent Survival of CaHodaphnia In Ambiant Toxicity Taata at Ottawa River, Lima, Ohio


River
Mffe
Young
Female

Final
Survival


Daily Survival


Station
Station Description
SD
1
2
3
4
5
6
7
1
Above Lime
46.0
15.5
8.0
90
100
100
100
90
90
90
90
2
Above STP
37.7
14.1
2.1
0
100
100
100
100
90
10
0
3
Below STP
37.4
0
-.
0
100
100
10
0
0
0
0
3A
Midway between STP and refinery
37.3
0
-
0
100
100
10
0
0
0
0
3B
Above refinery
37.1
0.4
-
0
90
90
40
0
0
0
0
4
Above chemical plant
36.9
7.5
3.6
10
100
100
100
100
100
50
10
4A
Below chemical plant
36.3
11.1
4.6
30
100
100
100
100
100
40
30
5
Shawnee Bridge
36.4
5.7
4.0
0
90
90
90
90
90
60
0
6
Route 117
32.6
12.6
3.8
10
100
100
100
100
100
100
10
7
AHentown
28.8
16.8
6.1
100
100
100
100
100
100
100
100
8
Rimer
16.0
17.4
9.5
80
100
90
90
90
90
80
80
8A
"Boonie" Station
8.0
25.0
3.3
100
100
100
100
100
100
100,
100
9
Kabda
1.0
25.6
5.5
100
100
100
100
100
100
100
100
Figure C-1. Cariodaphnia young production In water from various
atream itatione on the Ottawa River. Uma. Ohio.

-------
Selecting Sampling Stations
The selection of sampling stations is determined by the characteristics of the site. When
determining stations, the following factors should be considered:
o Mixing and flow - The mixing characteristics of the discharge site are useful to
determine the placement of sampling stations. Knowledge of concentration
isopleths allows the regulatory authority to place stations at locations instream
that correspond to concentrations measured in the dilution series in the effluent
tests. For example, where effluent testing shows the effluent NOEC is 10%, an
instream station should be placed where dilution is estimated to create a 10%
instream waste concentration. In this way, the size of a toxic plume can be
measured. Sampling stations should be placed where the effluents exist at
relatively constant and relatively specific concentrations. Test at specific low flow
conditions, if possible. Presence of tributaries or other sources of dilution will
influence positions and numbers of stations. Where smaller tributaries have
several point sources on them, treat the tributary as a point source. Obvious
non-point source areas should also be used to set stations.
o Existing biological data - Where biosurvey data are available, sampling station
location should be influenced by the more obvious trends in impact. In
particular, control stations and recovery stations can be determined by biosurvey
data.
o Single point sources - Single point source situations should be bracketed with an
above station, an immediate mixing station, several intermediate stations
corresponding to different instream concentrations, and a recovery station. Of
course, a control station should be established.
o Presence of other point sources - Multiple point source situations require the
placement of more stations between discharge points. Each source should be
„ bracketed by sampling stations.
There are four areas or zones that can be recognized when establishing the sampling
stations for ambient toxicity testing:
o Zone 1 - An upstream zone before the effluent enters.
o Zone 2 - A zone of mixing.
o Zone 3 - A zone after mixing and before additional dilution water enters.
o Zone 4 - A zone where additional dilution occurs either from effluents or
tributaries.
All possible combinations of occurrences are not practical to discuss but must be sorted
out for each site. Some generalizations are important to mention:
o Any upstream sources of contaminants, such as other discharges, will confound
the individual effects of a downstream discharge. For example, Zone 3 of the

-------
downstream discharge may occur in Zone 4 of an upstream discharge. This does
not invalidate the measurement of ambient toxicity. It only makes it difficult to
attribute amounts of response to each individual discharge. Response to the
instream mixture is what is measured.
o Careful location of sampling stations in Zone 3 is critical. Zone 3 is the only
place where toxicity decay rates of any one discharger can be measured and then
only if there are no upstream discharges, or if there are, only if that upstream
effluent is stable in that reach.
o In Zone 4, not only is degradation of the effluent toxicity occurring, but there is
dilution of it by other effluents and tributaries. Depending on the site
circumstances, one may not be able to learn anything about the ambient toxicity
characteristics of the effluent of concern in this zone.
o To emphasize, what can be measured in each zone depends on the above
considerations. In the more complex situation, only an estimate of ambient
toxicity at each station can be obtained. No information about one effluent's
toxicity decay rate will be available where several toxic effluents mix. In the
most simple situation of one discharge and no dilution downstream for a long
distance, Zone 3 will be large enough to get a good measure of toxicity decay.
Analysis of Ambient Toxicity Measurement
o When used in screening, the ambient toxicity data can identify areas in receiving
waters where ambient toxicity exists instream. Attributing such impact to specific
point sources (particularly where several sources discharge) may require effluent
toxicity testing.
o Except when used for screening purposes, ambient toxicity measurements must be
interpreted in conjunction with effluent toxicity test data if conclusions are to be
drawn concerning changes in toxic effect after discharge. The same species must
be used in both the ambient and the effluent toxicity tests.
o When analyzing the data, the performance of the animals at each station
downstream is compared to that of the animals exposed to receiving water
without the effluent of concern in it but containing all other upstream additions.
The result is an integration of effects from all contaminants and components
and represents not only the toxicity of the effluent of concern but also the
interactions of it with other effluents.
o Where the downstream stations show toxic effect at the concentrations measured
as toxic in the effluent toxicity tests, effluent toxicity can be considered to be
occurring instream, after discharge.
o Where the toxic effect decreases from station to station downstream in the
absence of further dilution, the effluent toxicity is degrading. If the decay rate is
rapid (eg., no toxicity at the closest instream station to the discharge point), the
effluent has a non-persistent toxicity. Where the decay rate is more gradual,

-------
toxicity is more persistent. The rate of decay of toxicity together with mixing data
allows the regulatory authority to approximate a receiving water toxicity impact
area. That impact area can then be compared to the appropriate state water
quality standards when establishing control requirements.
o In some cases, ambient toxicity may increase in relation to effluent toxicity
measurements. Either upstream sources of toxicity exist or some factor in the
receiving water is reacting with the effluent to increase its toxicity. Again, the
pattern and magnitude of change in toxicity should be analyzed. Differences in
toxicity levels between stations will reveal what is happening to the effluent as it
is mixed instream and interacts with the constituents of the receiving water.
o Trend analysis in the raw test data is of importance in the interpretation of
ambient toxicity data. As used in this context, trend analysis means observing
toxic effect as it occurs in the test itself and relating it to what is occurring
instream (plug flow, intermittent discharge, peak toxicity of effluents). Using
time-of-travel data or receiving water flow rates and patterns, observe effects on
the test organisms from day to day. There may be a pattern of mortality which
can be linked to discharge events. For example, in C-l the data indicate late
mortality at downstream stations on days 6 and 7. Flow rates for the river in this
example correlated this mortality to the downstream movement of a toxic slug
illegally discharged upstream.

-------
APPENDIX D
Duration and Frequency

-------
Duration and Frequency
As discussed on pages 7 through 13 of the "Guidelines for Deriving Numerical
National Water Quality Criteria for the Protection of Aquatic Organisms and Their
Uses" (1), the format used to express water quality criteria for aquatic life should take
into account toxicological and practical realities. Because of variation in the flows of the
effluent and the upstream receiving water as well as variation in the concentrations of
pollutants in the effluent and in the upstream receiving water, a simple format, such as
specifying a concentration that must not be exceeded at any time or place, is not
realistic. Furthermore, such a simple format does not take into account the fact that
aquatic organisms can tolerate higher concentrations of pollutants for short periods of
time than they can tolerate throughout a complete life cycle. The format that was
selected for expressing water quality criteria for aquatic life consists of recommendations
concerning concentrations, durations of averaging periods, and average frequencies of
allowed excursions. Use of this concentration-duration-frequency format allows water
quality criteria for aquatic life to be adequately protective without being as
overprotective as would be necessary if criteria were expressed using a simpler format.
In addition, this format can be applied directly to hydrological data and to the flow of,
and concentrations of pollutants in, effluents using both dynamic and steady-state
modeling (2,3).
In aquatic life criteria for both individual chemicals and whole effluents, the
recommended concentrations are the CMC and CCC. For individual chemicals the
CMC and CCC are derived using the procedures described by Stephan et al. (1). As
described in chapter 3 of this document, the CMC and CCC for whole effluents
can be specified generically in terms of toxic units. Alternatively, for a particular
effluent the CMC can be specified in terms of either a LC50 or an EC50, and the CCC
can be specified in terms of either a NOEC or an ICxx, if the LC50, EC50, NOEC, and
ICxx were obtained from appropriate toxicity tests conducted on the effluent with
sensitive species.
The CCC is intended to be the highest concentration that could be maintained
indefinitely in a receiving water without causing an unacceptable effect on the aquatic
community or its uses. Any concentration above the CCC, if maintained indefinitely, is
expected to cause an unacceptable effect. Due to the four sources of variation
mentioned above, the concentration in the receiving water will not be constant. Because
organisms can tolerate higher concentrations for short periods of time, it is expected that
the concentration of a pollutant in a body of water can exceed the CCC without causing
an unacceptable effect if (a) the magnitudes and the durations of exceedences are
appropriately limited and (b) there are compensating periods of time during which the
concentration is below the CCC. These goals are accomplished by specifying a duration
of an averaging period over which the average concentration should not exceed the
CCC. For example, if the concentration is twice the CCC for one-half the specified
averaging period, it must be zero for the rest of the averaging period if the average
concentration is not to exceed the CCC. Thus both the magnitude and duration of an
exceedence .are limited and there must be a compensating period of time during the
averaging period when the concentration is below the CCC. Because exceedences are
defined to be due to usual variation, most exceedences will be small, with larger
exceedences becoming increasingly rare (1,2).

-------
Although an exceedence is defined to occur whenever the instantaneous
concentration is above the CCC, an excursion is defined to occur only when the average
concentration over the duration of the averaging period is above the CCC. It is
expected that excursions can occur without causing unacceptable effects if (a) the
frequency of such excursions is appropriately limited and (b) all other average
concentrations are below the CCC. The recommended average frequency of allowed
excursions is intended to appropriately limit the frequency of excursions. Because
excursions are the highest average concentrations that occurred due to usual variation,
all other average concentrations will be less than the CCC. As above for exceedences,
because excursions are defined to be due to usual variation, most excursions will be
small, with larger excursions becoming increasingly rare. The duration of the averaging
period is intended to limit the impact of exceedences, whereas the average frequency of
allowed excursions is intended to limit the impact of excursions. (Note: The words
"exceedence"and "excursion"are used slightly differently here than in references 1 and
2.)
Although spills can impact aquatic communities, they are not considered
exceedences or excursions because they are not part of the usual variation in the
concentrations of pollutants in effluents. In the Complex Effluent Toxicity Testing
Program, eight field studies were conducted to evaluate the use of toxicity tests to
diagnose the cause of biological impact, and ambient toxicity measurements were taken
over a seven-day period. During two of these studies (4,5) spills of pollutants resulted in
acute toxicity. This suggests that the impacts caused by spills might be as important as
impacts caused by variation in the compositions and flows of the effluent and the
receiving water.
The primary purpose of this appendix is to present the rationale for the
recommendations of the U.S. EPA concerning duration and frequency in national water
quality criteria for aquatic life. The recommended duration is based on data from
laboratory toxicity tests, whereas the recommended frequency is based on field data.
With the concurrence of the U.S. EPA, States may adopt site-specific criteria, rather
than national criteria, in their State standards. Such site-specific criteria may include
not only site-specific concentrations, but also site-specific, and possibly pollutant-specific,
durations of averaging periods and average frequencies of allowed excursions. If
adequate justification is provided, site-specific and/or pollutant-specific concentrations,
durations, and frequencies may be higher or lower than those given in national water
quality criteria for aquatic life. A secondary purpose of this appendix is to discuss
rationale that might be used as a basis for electing alternative durations of averaging
periods and average frequencies of allowed excursions.
Duration
In order for this concentration-duration-frequency format to allow water quality
criteria for aquatic life to be adequately protective without being unnecessarily
overprotective, the duration of the averaging period must allow some exceedences above
the CCC without allowing unacceptable effects. Thus the averaging period must

-------
appropriately limit the magnitude and duration of exceedences and provide
compensating periods of time during which the concentration is below the CCC.
Even though only a few tests have compared the effects of a constant
concentration with the effects of the same average concentration resulting from a
fluctuating concentration, nearly all the available comparisons have shown that
substantial fluctuations result in increased adverse effects (6-16). Thus the duration of
the averaging period must be shorter than the duration of the chronic tests on which the
CCC is based so that the averaging period does not allow substantially more adverse
effect than would have been caused by a continuous exposure to the same average
concentration. Life-cycle tests with species such as mysids and daphnids and early life-
stage tests with warmwater fishes usually last for 20 to 30 days, whereas life-cycle tests
with ceriodaphnids usually last for seven days. If the duration of the averaging period is
too short, however, it will not allow any meaningful exceedences and will, in effect,
defeat the purpose of the concept of the averaging period. For example, because few
effluents are monitored more often than once a day, an averaging period of 24 hours
would effectively mean that for most effluents each individual sample that was above the
CCC would be considered an excursion.
For the following reasons, a 4-day averaging period is recommended for
application of the CCC in aquatic life criteria for both individual pollutants and whole
effluents:
1.	It is substantially shorter than the 20 to 30-day duration of most chronic
tests and is somewhat shorter than the 7-day duration of the ceriodaphnia
life-cycle test.
2.	The results of some chronic tests are apparently due to an acute effect on
a sensitive life stage that occurs at some time during the test, rather than
being caused by either long-term stress or long-term accumulation of the
test material in the organisms. Horning and Neiheisel (17) documented
one such situation, and others ate probably the cause of,at least some of
the acute-chronic ratios that are not much greater than unity.
3.	For both endrin and fenvalerate, Jarvinen et al. (18) found that a 72-hr.
exposure caused about the same amount of effect on the growth of fathead
minnows in early life-stage tests as did a 30-day exposure to the same
concentration.
4.	In some life-cycle tests on effluents with ceriodaphnids, concentrations of
effluents that were a factor of 1.8 greater than the CCC caused
unacceptable effects in four or five days (5,19,20).
5.	It is not so short as to effectively defeat the purpose of the concept of the
averaging period.
As discussed below, other averaging periods might be acceptable on a site-specific or
pollutant specific basis.

-------
Just as the concept of exceedences can be applied to the CCC, it can also be
applied to the CMC. As with the CCC, the CMC averaging period should be
substantially less than the lengths of the tests on which the CMC is based, i.e.,
substantially less than 48 to 96 hours. Because 4 to 8-hr LC50s are about the same as
the 96-hr LC50 for some materials (21-27), the duration of the averaging period for the
CMC must be less than 4 hours. One hour is probably an appropriate duration of the
averaging period for the CMC because concentrations of some materials that are only a
factor of two higher than the 96-hr LC50 cause death in one to three hours (25). Even
when organisms do not die within the first hour or so, it is not known how many
organisms might have died due to the delayed effects of the short exposure (28-31). If
the one-hour average does not exceed the CMC, it is unlikely that the concentration of
the pollutant in the receiving water can fluctuate rapidly enough during the hour to
cause additional adverse effect. Thus, it seems inappropriate to apply the CMC to
instantaneous concentrations.
With adequate justification, the CMC and/or CCC averaging periods may be
increased or decreased on a site-specific or pollutant-specific basis. A possible site-
specific justification for increasing the duration of the CCC averaging period would be
that the variation in the concentration of the pollutant in the receiving water is low.
Where variation is demonstrated to be
consistently low, a longer averaging might be acceptable because the magnitudes and
durations of exceedences above the CCC would be limited. A possible pollutant-specific
justification for a longer averaging period would be demonstration that the LC50
decreases substantially as the length of the exposure increases. For example, an 8-hr
averaging period might be justified for the CMC if it were shown that 24-hr exposures of
a variety of sensitive species resulted in 96-hr LC50s that were substantially above the
96-hr LC50s obtained from continuous exposure to
a constant concentration for 96 hours.
Regardless of what averaging periods are used, exact calculation of the number of
excursions would require continuous monitoring of the concentration in the receiving
water, which is not feasible in most cases. A valid alternative would be to use a
statistically designed monitoring program and a statistical interpretation of the measured
concentrations. The one-hour averaging period for the CMC would imply that the
samples analyzed should be one-hour composites; the four-day averaging period would
imply that concentrations in all samples obtained within any four-day period should be
averaged, preferably using a time-weighted average. If information is available
concerning the discharge pattern of a particular effluent, it might be possible to design a
monitoring program that is specifically appropriate for that effluent.
Unless critical species are especially sensitive to particular toxicants, most
excursions of criteria should have minor impacts on aquatic communities. However,
whereas excursions above the CCC will probably reduce growth and reproduction,
excursions above the CMC will probably cause death and other severe acute effects. In
addition, special care should be exercised when many outfalls exist in a small reach of a
receiving water, because if low flow causes an excursion for one discharge, that same low
flow will probably also cause excursions for other discharges at the same time. Several
"minor" excursions might thus add up to a "major"one.

-------
Frequency
The purpose of the average frequency of allowed excursions is to provide an
appropriate average period of time during which the aquatic community can (a) recover
from the effect of an excursion and then (b) function normally for a period of time
before the next excursion. The average frequency is intended to ensure that the
community is not constantly recovering from effects caused by excursions of aquatic life
criteria. Because most regulated discharges are to flowing water (lotic) systems, this
discussion will emphasize discharges to rivers and streams rather than to lakes, ponds,
and reservoirs.
General considerations for setting frequency with which criteria mav be exceeded:
Not long ago ecological communities were thought to be largely in equilibrium
and their structure and function determined primarily by internal interactions between
species, such as competition and predation. Communities were considered to be
analogous to "super-organisms",with close parallels to organisms in their response to
stress and in "health". Current understanding is that external factors, including
disturbances, often play a major role in the structure of communities (32,33). The
frequency of disturbance can affect a community not only by decreasing the fitness of
component species, but also, by causing a natural selection of species or phenotypes
having characteristics that allow them to tolerate or even thrive under the disturbance
regime. Natural disturbances such as floods and droughts are common in lotic systems
(32) and vary in intensity not only between headwater streams and large rivers, but also
between similar sized lotic communities in different climatic regions. Rather than
requiring more time to recover from the effects of additional anthropogenic
disturbances, lotic communities with high natural background disturbance frequencies
are actually predisposed to recover more rapidly because only species that are able to
recolonize and reproduce quickly, or perhaps to avoid disturbances, can persist there
(34-37). This does not imply that they are also more resistant to novel anthropogenic
disturbances with which they have had no previous evolutionary experience; it only
implies that they are predisposed to recover quickly once the disturbance is gone. The
question then is how frequently can aquatic communities experience these additional
disturbances (excursions of criteria) without being unacceptably affected.
In an extensive review of the published literature Niemi et al. (38) identified
more than 150 case studies of freshwater systems in which some aspect of recovery from
the impact of a disturbance was reported. A case study was used only if the disturbance
caused a death or displacement of organisms. This restriction was necessary because it
was rarely possible to determine if an event was outside the normal intensity range (a
common alternate definition of disturbance), mainly because it is usually difficult to
define the normal intensity range. It also permitted the inclusion of natural as well as
anthropogenic events. Approximately 80% of these systems were lotic, and the
remainder were lentic (lakes and ponds). The impacts were due to such disturbances as
persistent and non-persistent chemicals, logging, flooding, channelization, dredging, and
drought. Reported endpoints for recovery were sparse for phytoplankton, periphyton,
and macrophytes, but were numerous for macroinvertebrates and fishes. Because more
than one recovery endpoint was reported for most studies, the number of endpoints

-------
greatly exceeded the number of case studies. For short-term (non-persistent)
disturbances, approximately 85% of all macroinvertebrate endpoints indicated recovery
in less than two years. Macroinvertebrate biomass, density, and taxonomic richness
recovered in less than one year for approximately 95% of reported endpoints.
Dipterans (flies, mosquitos, midges, etc.), which generally have short generation times 01'-
high dispersal ability, recovered most rapidly, whereas stoneflies and caddisflies
recovered least rapidly. Fishes recovered in two years or less for over 85% of reported
endpoints; however there were significant exceptions, as discussed below.
Most excursions of criteria will be minor and their impacts will therefore be
difficult to detect. Although most disturbances in the above case studies caused more
severe impacts than most criteria excursions are expected to cause, CMC excursions will
result in death of some organisms. However, these data indicate that as a general rule,
the purpose of the average frequency of allowed excursions will be achieved if the
frequency is set at once eveiy three years on the average. Excursions of the CCC are
more difficult to evaluate because non-lethal excursions could not be evaluated from the
data used by Niemi et al. (38). However, it is reasonable to expect that cumulative
effects from too frequent excursion of the CCC will also result in unacceptable
degradation of lotic communities.
Considerations for proposing site-specific increases or decreases in the average frequency
of allowed excursions:
Although an average frequency of one criterion excursion every three years
should usually be protective of lotic communities, more frequent excursions might be
acceptable in certain situations. Sedell et al. (39) have shown that lotic systems with
refugia (areas of refuge) such as well-developed riparian zones, connected flood plains
and meanders, snags, etc., recover more rapidly from disturbances than segments without
such refugia, because organisms are better able to avoid disturbances and return or
repopulate. However, many of these refugia are likely to be most restricted and
vulnerable during the low-flow periods wh^n criteria excursions are also most likely to
occur. Evidence of action to preserve refugia, particularly during low-flow periods, or to
create or restore them, might be grounds for demonstrating that an excursion frequency
of more than once every three years on the average is acceptable.
Schlosser (36) found that lower-order (i.e., headwater) streams, because of their natural
high variability, contain communities consisting of species that have short life cycles
and/or high dispersal ability and can recover from major disturbances in a year or even
less. Thus, many lower-order streams, particularly those for which refugia are available,
should be able to tolerate a minor criteria excursion every two years or even every year,
unless other considerations are important. For example, discharges to lower-order
streams sometimes constitute a large fraction of the stream flow for most of the year.
Although lower-order streams are naturally highly variable and can therefore
tolerate higher disturbance frequencies, the converse is true for higher-order lotic
systems for at least two partially related reasons: (1) reaches with tributaries draining a
large watershed will be buffered from the effects of localized droughts in a portion of
the watershed, and will therefore experience a less severe natural disturbance regime,
and (2) organisms inhabiting these reaches will therefore not be adapted to disturbances

-------
that are as frequent or severe as those in lower-order reaches. Fish in particular will be
larger and have longer generation times in larger streams and rivers; consequently it will
take longer for these populations to reproduce and regain predisturbance densities and
size class distributions. Schlosser (36) suggests that, based on such life-history
characteristics, fish communities in larger rivers might take 20 to 25 years to re-establish
the predisturbance age and size structure of their component populations after a severe
disturbance such as a drought or major spill.
Extreme cases in which recovery has taken much longer than three years usually
involve spills of persistent chemicals or severe habitat modification, such as stream
channelization or clear-cutting of a watershed (38). If the chemical contaminant is not
widespread, recovery is limited primarily by the rate of disappearance of the chemical
rather than by strictly ecological processes. Widespread contamination can affect
recovery by increasing the distance over which recolonizers must travel. Watershed
clear-cutting reduces the input of organic matter that provides the food base of streams
in forested watersheds and also provides woody debris and snags that serve as refugia.
Channelization and dredging reduce the in-stream habitat diversity and thereby decrease
refugia. In addition to these anthropogenic disturbances, multiple excursions during a
drought, due to low flow conditions, can result in a severe cumulative impact on
sensitive species even if the individual excursions are small. Special measures, such as
plant shutdowns, might be required in extreme cases. Finally, severe chemical spills,
which cannot be regulated but which will occur in any highly industrialized river reach,
will affect aquatic life over a large area. If maintenance of long-lived fish species in
these reaches is desired, recovery times up to 25 years might be necessary.
Based on the above considerations, recovery periods of much greater than three
years might be necessary after multiple minor excursions or after a single major
excursion or spill during a low-flow period in medium to large river reaches, and up to
25 years where long-lived fish species are to be protected. Even longer times might be
necessary as the size of the affected area or the persistence of the pollutant increases.
Calculation of Design Conditions
" The use of aquatic life criteria for developing water quality-based permit limits
and for designing waste treatment facilities requires the selection of an appropriate
wasteload allocation model. Dynamic models are preferred for the application of
aquatic life criteria in order to make best use of the specified concentrations, durations,
and frequencies. If dynamic models cannot be used, then an alternative is steady-state
modeling. Because steady-state modeling is based on various simplifying assumptions, it
is less complex, and might be less realistic, than dynamic modeling.
An important step in the application of steady-state modeling to streams is the
calculation of the design flow. The procedures outlined in the EPA document Technical
Guidance Manual for Performing Waste Load Allocation. Book VI. Design Conditions:
Chapter 1 - Steam Design Flow for Steadv-State Modeling (USEPA 1986) are
recommended for calculating design flows for rivers and streams. States may use other
methods so long as the methods are technically defensible. The document discusses and
recommends two methods for determining design flows, the hydrologically-based method

-------
and the biologically-based method, and the flows that should be used for the CCC and
CMC for both methods.
The hydrologically-based design flow method is the one presently used by many
States. It is based on the selection and identification of an extreme value, e.g., the 7Q10
flow. The underlying assumption of this method is that the design flow will occur X
number of times in Y years. Thus, this method limits the number of years in which one
or more excursions below the design flow can occur. The method has several
advantages: (1) the log-Pearson Type III flow estimating technique or other extreme
value analytical techniques that are used to calculate flow statistics from daily flow data
are consistent with past engineering and statistical practice, (2) many people that have
used it do not understand it, States currently use this method, and (3) the USGS
provides technical support for this method. The disadvantage of this method is that it is
essentially independent of biological considerations. Design flows calculated using this
method might allow more or fewer excursions than once every three years on the
average. In addition, it is difficult to use site-specific durations and frequencies with this
method. For toxic WLA studies in which the hydrologically-based method is used, EPA
recommends the use of the 1Q10 flow as the design flow for the CMC and the 7Q10 as
the design flow for the CCC.
The biologically-based design flow method was developed by the USEPA Office
of Research and Development and directly uses the averaging periods and frequencies
specified in the water quality criteria for individual pollutants and whole effluents for
determining design flows. The method is an empirical iterative convergence procedure
that includes the calculation of harmonic means of the flow to determine the total
number of excursions. The method makes exact use of whatever duration and frequency
are specified for the CMC and CCC. These might be one-day and 3-years for the CMC
and four-day and 3-years for the CCC or site-specific durations and frequencies.
A comparison of the two methods reveal them to be similar with neither method
being more protective than the other on the average. The methods were used on
approximately 60 different rivers and the hydrologically-based 1Q10 and 7Q10 design
flows were found to be similar on the average to the biologically-based l-day/3-year and
4-day/3-year design flows. Thus the hydrologically-based design flows will, on the
average, result in the allowed number of excursions if site-specific durations and
frequencies are not used. For some of the 60 rivers, however, the 1Q10 and 7Q10
allowed substantially more or fewer excursions than the intended number of excursions.
Because the biologically-based method calculates the design flow directly from the
national or site-specific duration and frequency, it always provides the maximum allowed
number of excursions and never provides more excursions than allowed.
EPA provides software tools to calculate both types of design flows via the
STORET environment on its NCC-IBM mainframe. Biologically-based design flows can
be calculated using the program DFLOW. The hydrologically-based design flows can be
calculated using FLOSTAT or DFLOW; the latter uses a simplified version of the
log-Pearson T^pe III method. Both programs access the STORET Row file that
contains daily flow records for USGS gaging stations. They are easy to use and the user
simply needs to know the identification number of the gaging station. To obtain further
information on the STORET environment and the programs, contact:

-------
Mr. Philip Taylor
USEPA Office of Water Regulations and Standards
401 M Street, S.W.
Washington, D.C. 20460
(202) 382-7046
The methods described above use daily flow data to determine design flow, but
they do not consider any other physical or chemical condition that might affect toxicity.
EPA has prepared a supplementary method and a software tool that incorporate such
supplemental water quality parameters as temperature, pH, alkalinity, hardness, and
dissolved oxygen to determine design conditions. The method and software are
described in two documents available from the Assessment and Watershed Protection
Division of the Office of Water Regulations and Standards:
o Technical Guidance on Supplementary Stream Design Conditions for
Steady State Modeling [3].
o DESCON Users Manual
The supplementary method is consistent with the hydrologically-and
biologically-based methods described above; it simply extends them to include other
conditions besides streamflow. The advantage of considering multiple conditions is that
the critical conditions necessary to protect water quality criteria might not occur when
the streamflow is critical; e.g., low DO or high temperatures might occur at times other
than when the flow is low.
This supplementary method can be used for five pollutant categories with the
physical-chemical parameters described above. The pollutant categories are: general
toxicant; ammonia; heavy metals (Cd, Cr , Cu, Pb, Ni, Zn); pentachlprophenol; and
ultimate oxygen demand.
The software tool to facilitate this method is called DESCON. It is on EPA's
IBM mainframe and is available through the STORET environment. DESCON accesses
the STORET Flow file for the daily flow record and the water quality file for data on
the physical-chemical parameters. Options are available to the user if the area of
concern has no flow record or if no water quality data are available.

-------
REFERENCES
1.	Stephan, C.E., D.I. Mount, D.J. Hansen, J.H. Gentile, G.A. Chapman, and W.A.
Brungs. 1985. Guidelines for Deriving Numerical National Water Quality
Criteria for the Protection of Aquatic Organisms and Their Uses. PB85-227049.
National Technical Information Service, Springfield, VA.
2.	U.S. EPA. 1986. Technical guidance manual for performing waste load
allocation. Book VI. Chapter 1. Stream design flow for steady-state modeling.
Monitoring and Data Support Division, Office of Water Regulations and
Standards, Office of Water, Washington, DC.
3.	U.S. EPA. 1988. Technical guidance on supplementary stream design conditions
for steady state modeling. Available from: Assessment and Watershed
Protection, Office of Water Regulations and Standards, Office of Water,
Washington, DC.
4.	Mount, D.I., N.A. Thomas, T.J. Norberg, M.T. Barbour, T.H. Roush, and W.F.
Brandes. 1984. Effluent and ambient toxicity testing and instream community
response on the Ottawa River, Lima, Ohio. EPA/600/3-84/080. National
Technical Information Service, Springfield, VA.
5.	Norberg-King, T.J. and D.I. Mount (Eds.). 1986. Validity of effluent and
ambient toxicity tests for predicting biological impact, Skeleton Creek, Enid,
Oklahoma. EPA/600/8-86/002. National Technical Information Service,
Springfield, VA.
6.	Brown, V.M., D.H.M. Jordan, and B.A. Tiller. 1969. The acute toxicity to
rainbow trout of fluctuating concentrations and mixtures of ammonia, phenol, and
zinc. J. Fish. Biol. 1:1-9.
7.	Coffin, D.L., D.E. Gardner, G.I. Sidorenko, and M.A. Pinigin. 1977. Role of
time as a factor in the toxicity of chemical compounds in intermittent exposures.
* J. Toxicol. Environ. Health 3:821-828.
8.	Smith, L.L., Jr., SJ. Broderius, D.M. Oseid, G.L. Kimball, W.M. Koenst, and D.T.
Lind. 1979. Acute and chronic toxicity of HCN to fish and invertebrates. EPA-
600/3-79-009. National Technical Information Service, Springfield, VA.
9.	Thurston, R.V., C. Chakoumakos, and R.C. Russo. 1981. Effect of fluctuating
exposures on the acute toxicity of ammonia to rainbow trout (Salmo gairdneri)
and cutthroat trout (S* clarki). Wat. Res. 15:911-917.
10.	Buckley, J.T., M. Roch, J .A. McCarter, C.A. Rendell, and A.T. Matheson. 1982.
Chronic exposure of coho salmon to sublethal concentrations of copper -1.
Effect on growth, on accumulation and distribution of copper, and on copper
tolerance. Comp. Biochem. Physiol. 72C: 15-19.
11. Ingersoll, C.G. and R.W. Winner. 1982. Effect on Daphnia pulex (De Geer) of

-------
daily pulse exposures to copper or cadmium. Environ. Toxicol. Chem. 1:321-327.
12.	Hodson, P.V., B.R. Blunt, U. Borgmann, C.K. Minns, and S. McGaw. 1983.
Effect of fluctuating lead exposures on lead accumulation by rainbow trout
(Salmo gairdneriV Environ. Toxicol. Chem. 2:225-238.
13.	Seim, W.K., L.R. Curtis, S.W. Glenn, and G.A. Chapman. 1984. Growth and
survival of developing steelhead trout (Salmo gairdnerfl continuously or
intermittently exposed to copper. Can. J. Fish. Aquat. Sci. 41:433-438.
1^ Curtis, L.R., W.K. Seim, and G.A. Chapman. 1985. Toxicity of fenvalerate to
developing steelhead trout following continuous or intermittent exposure. J.
Toxicol. Environ. Health 15:445-457.
15.	Siddens, L.K., W.K. Seim, L.R. Curtis, and G.A. Chapman. 1986. Comparison of
continuous and episodic exposure to acidic, aluminum-contaminated waters of
brook trout (Salvelinus fontinalisY Can. J. Fish. Aquat. Sci. 43:2036-2040.
16.	Brooks, A.S., D.C. Szmania, and M.S. Goodrich. 1989. A
comparison of continuous and intermittent exposures of
four species of aquatic organisms to chlorine. Center
for Great Lake studies, University of Wisconsin-Milwaukee,
WI.
17.	Horning, W.B. and T.W. Neiheisel. 1979. Chronic effect of copper on the
bluntnose minnow, Pimephales notatus (Rafinesque). Arch. Environ. Contam.
Toxicol. 8:545-552.
18.	Jarvinen, A.W., D.K. Tanner, and E.R. Kline. 1988. Toxicity of chlorpyrifos,
endrin, or fenvalerate to fathead minnows following episodic or continuous
exposure. Ecotoxicol. Environ. Safety 15:78-95.
19.	Mount, D.I., A.E. Steen, and TJ. Norberg-King (Eds.). 1985. Validity of effluent
and ambient toxicity testing for predicting biological impact on Five Mile Creek,
Birmingham, Alabama. EPA/600/8-85/015. National Technical Information
Service, Springfield, VA.
20.	Mount, D.I., T.J. Norberg-King, and A.E. Steen (Eds.). 1986. Validity of effluent
and ambient toxicity tests for predicting biological impact, Naugatuck River,
Waterbury, Connecticut. EPA/600/8-86/005. National Technical Information
Service, Springfield, VA.
21.	Cardwell, R.D., D.G. Foreman, T.R. Payne, and DJ. Wilbur. 1976. Acute
toxicity of selected toxicants to six species of fish. EPA-600/3-76-008. National
Technical Information Service, Springfield, VA.
22.	Thurston, R.V., G.R. Phillips, R.C. Russo, and S.M. Hinkins. 1981. Increased
toxicity of ammonia to rainbow trout (Salmo gairdnerO resulting from reduced
concentrations of dissolved oxygen. Can. J. Fish. Aquat. Sci. 38:983-988.

-------
23.	Brooke, L.T., D.J. Call, D.L. Geiger, and C. E. Northcott (Eds.). 1984. Acute
toxicities of organic chemical to fathead minnows (Pimephales promelas).
Volume I. Center for Lake Superior Environmental Studies, Univeristy of
Wisconsin-Superior, Superior, WI.
24.	Geiger, D.L., C.E. Northcott, D.J. Call, and L.T. Brooke (Eds.). 1985. Acute
toxicities of organic chemicals to fathead minnows (Pimephales promelas).
Volume II. Center for Lake Superior Environmental Studies, University of
Wisconsin-Superior, Superior, WI.
25.	Bailey, H.C., D.H.W. Liu, and H.A. Javitz. 1985. TimeAoxicity relationships in
short-term static, dynamic, and plug-flow bioassays. In: Aquatic Toxicology and
Hazard Assessment: Eighth Symposium. ASTM STP 891. R.C. Bahner and D.J.
Hansen (Eds.). American Society for Testing and Materials, Philadelphia, PA.
pp.193-212.
26.	Geiger, D.L., S.H. Poirier, L.T. Brooke, and D.J. Call (Eds.). 1986. Acute
toxicities of organic chemicals to fathead minnows (Pimephales promelasl
Volume III. Center for Lake Superior Environmental Studies, University of
Wisconsin-Superior, Superior, WI.
27.	Geiger, D.L., D.J. Call, and L.T. Brooke. 1988. Acute toxicities of organic
chemicals to fathead minnows fPimephales promelas). Volume IV. Center for
Lake Superior Environmental Studies, Univerisity of Wisconsin-Superior,
Superior, WI.
28.	Abel, P.D. 1980. Toxicity of hexachlorocyclohexane (lindane) to Gammarus
pulex: mortality in relation to concentration and duration of exposure.
Freshwater Biol. 10:251-259.
29.	Abel, P.D. 1980. A new method fbr assessing the lethal impact of short-term,
high-level discharges of pollutants on aquatic animals. Prog. Wat. Tech. 13:347-
352.
30.	Abel, P.D., and S.M. Garner. 1986. Comparisons of median survival times and
median lethal exposure times for Gammarus pulex exposed to cadmium,
permethrin and cyanide. Wat. Res. 20:579-582.
31.	Heming, T.A., A. Sharma, and Y. Kumar. 1989. Time-
toxicity relationships in fish exposed to the organochlroine
pesticide methoxychlor. Environ. Toxicol. Chem. 8:923-932.
32.	Resh, V.H., A.V. Brown, A.P. Covich, M.E. Gurtz, H.W. Li, G.W. Minshall, S.R.
Reice, A.L. Sheldon, J.B. Wallace, and R.C. Wissmar. 1988. The role of
disturbance in stream ecology. J. North Am. Benthol. Soc. 7:433-455.

-------
33.	Reice, S.R., R.C. Wissmar, and R.J. Naiman. 1989. The influence of spatial and
temporal heterogeneity and disturbance regime on the recovery of animal
communities in lotic ecosystems. IN: Yount, J.D., and GJ. Niemi (Eds.).
Recovery of lotic communities and ecosystems from disturbance: Theory and
applications. Environ. Management (submitted).
34.	Steinman, A.D. 1989. Recovery of lotic periphyton communities after disturbance.
IN: Yount, J.D., and G.J. Niemi (Eds.). Recovery of lotic communities and
ecosystems from disturbance: Theory and applications. Environ. Management
(submitted).
35.	Wallace, J.B. 1989. Recovery of lotic invertebrate communities from disturbance.
IN: Yount, J.D., and GJ. Niemi (Eds.). Recovery of lotic communities and
ecosystems from disturbance: Theory and applications. Environ. Management
(submitted).
36.	Schlosser, I.J. 1989. Environmental variation, life history attributes, and
community structure in stream fishes: Implications for environmental
management and assessment. IN: Yount, J.D., and G.J. Niemi (Eds.). Recovery of
lotic communities and ecosystems from disturbance: Theory and applications.
Environ. Management (submitted).
37.	Poff, N.L. and J.V. Ward. 1989. The physical habitat template of lotic systems:
Recovery in the context of historical pattern of spatio-temporal heterogeneity.
IN: Yount, J.D., and G.J. Niemi (Eds.). Recovery of lotic communities and
ecosystems from disturbance: Theory and applications. Environ. Management
(submitted).
38.	Niemi, G.J., P. DeVore, N. Detenbeck, D. Taylor, J.D. Yount, A. Lima, J. Pastor,
and R.J. Naiman. 1989. An overview of case studies on recovery of aquatic
systems from disturbance. IN: Yount J.D., and G.J. Niemi (Eds.). Recovery of
lotic communities and ecosystems from disturbance: Theory, and applications.
Environ. Management (submitted).
39.	Sedell, J.R., F.R. Hauer, C.P. Hawkins, and J.A. Stanford. 1989. The role of
refugia in recovery from disturbance: Modem fragmented and disconnected
river systems. IN: Yount, J.D., and GJ. Niemi (Eds.). Recovery of lotic
communties and ecosystems from disturbance: Theory and applications. Environ.
Management (submitted).

-------
APPENDIX E
Lognormal Distribution and Permit Limit Derivations

-------
TSD APPENDIX E
Introduction
This appendix provides supporting information for the statistical methodology used in
permit limit calculations. The methodology described in this appendix applies to many
types of data including data that are used to develop both techiiology-based and water
quality-based permit limits. The appendix is divided into two sections. The first section
gives an overview of permit limits: the derivation of water quality-based limits, and the
consistency among different permit limits. The second section describes the statistical
methodology for: the normal distribution, the lognormal distribution, the delta-lognormal
distribution, methods of checking distributional assumptions, and correlation. This
section also provides guidance on the application of each distribution to permit limits.
Tables 2, 3 and 4 at the end of the appendix summarize the permit limit calculations.
SECTION 1: OVERVIEW OF PERMIT LIMITS
Two types of permit limits are contained in the effluent guidelines regulations: daily
maximum limits, and maximum monthly average limits. The daily maximum permit limit
is the maximum allowable value for any daily sample. The daily maximum limits are
usually based on the 99th percentile of the distribution of daily measurements. The
maximum monthly average permit limit is the maximum allowable value for the average
of all daily samples obtained during one month. Maximum monthly average limits are
in most cases based on the 95th percentile of the distribution of averages of daily values.
The following two sub-sections discuss the derivation of water quality-based limits and
the consistency among different permit limits.
Derivation of water qualitv-based limits
Water quality-based limits are derived from the required treatment system performance
necessary to comply with the wasteload allocation (WLA). Technology-based effluent
limits are derived from treatment system performance. The mathematical expressions
for water quality-based limits are the same as those for technology-based effluent limits;
the major difference is that the means and standard deviations in those expressions are
derived from the WLA This topic is discussed in detail in Chapter 5.
Consistency among different permit limits
The current TSD procedures provide consistency among different permit limits. The
stringency of permit limits is independent of monitoring frequency and is determined
entirely by the Wasteload Allocation (WLA) and permit limit derivation procedures.
The daily maximum limit is constant regardless of monitoring frequency. The numerical
value of the monthly average limit decreases as monitoring frequency increases only
because averages become less variable as the number of values included in the average
increases. For example, an average of 10 samples is less variable than an average based
on 4 samples. This phenomenon makes monthly average permit limits based on 10
samples appear to be more stringent than the monthly limit based on 4 samples. A

-------
TSD APPENDIX E
permittee performing according to the WLA specifications will in fact be equally capable
of meeting either of these monthly average limits when talcing the corresponding number
of samples. The stringency of the TSD procedures, accordingly, is constant across
monitoring frequencies.
SECTION 2: STATISTICAL METHODOLOGY
The statistical procedures which are used in permit limit development involve fitting
distributions to effluent data. The estimated upper percentiles of the distributions form
the basis of the limits. This section describes the statistical methodology applied to
permit limits in the following sub-sections: the normal distribution, the lognormal
distribution, the delta-lognormal distribution, methods of checking distributional
assumptions, and correlation. Before discussing these topics several definitions are made
for: notation, assumptions, coefficients of variation, and variability factors. Table 1
below summarizes the differences between the distributions and when to use each in
permit limits.
TABLE 1: CHARACTERISTICS OF THE DISTRIBUTIONS
Distribution
Shape
Use for Adjusts for
Range of Sample non-detect
Values	Size	values
Normal
Lognormal
Bell
skewed
(-00,00)	>io
(9,00)	detection
limit
Yes

-------
TSD APPENDIX E
Notation
In the calculations in this appendix, natural logarithms (i.e., logarithms to the base e),
denoted by ln(x), are used. The calculations can be modified to use logarithms to the
base 10 by replacing log10(x) for ln(x) in the formulas.
Assumptions
The distribution fitting methods assume that the daily measurements are independent,
uncorrected observations.
The fundamental assumptions underlying the limits are:
-	daily pollutant measurements are approximately lognormally distributed for values
above the detection limit
-	maximum n-day monthly averages for n<10 are approximately lognormally
distributed above the detection limit
-	maximum n-day monthly averages for n>10 are normally distributed.
Coefficients of variation
The coefficient of variation (denoted by cv) is the ratio of the standard deviation to the
mean. Thus, the cv is a dimensionless measure of the relative variability of a
distribution. Estimates of the cv can be used when the actual cv cannot be calculated or
if the available datasets for calculating the cv are small. In such cases, different values
for the cv should be used in the permit calculations to assess the effect of the cv on the
final permit limit. Typical values of the cv for effluent data usually range from 0.2 to
1.2. In many cases, changes in the cv will have little impact on the final permit limit. In
assessing the sensitivity of the permit limit to the cv, the calculations may include cv=0.6
as a. conservative estimate (assumes relatively high variability). If the final permit
values vary greatly with different cv values either of two approaches may be used. The
first approach is to use a conservative estimate of the cv which assumes relatively high
variability (e.g., cv=0.6) in the final permit limit. The second approach is to collect
additional data to obtain a more definitive value for the cv.
Variability Factors
An important component of the process used by EPA for developing technology-based
limits are variability factors. The variability factor is the ratio of a large concentration
level of a pollutant to the average level determined from that particular plant. The
ratio expresses the relationship between the average treatment performance level and
large values that would be expected to occur only on rare occasion in a well-designed
and operated treatment system. Such factors are useful in situations where little data
are available to characterize the long-term performance of a plant.

-------
In cases where only a small number of observations are available from a plant, EPA has
been reluctant to estimate a variability factor. In the Organic Chemicals, Plastics and
Synthetic Fibers (OCPSF) rulemaking [1], a minimum of 7 daily observations from a
plant, with at least 3 of the 7 above the detection limit, was established for calculation
of a plant level priority pollutant variability factor. However, EPA has not established a
minimum number of observations required for calculating variability factors for all
pollutants in all industries.
The calculations for variability factors for the daily maximum and the maximum monthly
average are included in the discussion of the different distributions below.
Normal Distribution
The normal distribution plays a central role in the methods described in this appendix.
The normal distribution is related to the lognormal distribution which is used to
establish many permit limits. In most cases, the simple logarithmic transformation of
effluent and water quality-based data results in data distributions that are normally
distributed. This type of data is referred to as being lognormally distributed. Since the
normal and lognormal distributions are related in a straight-forward manner, the
methods of analysis for normal and lognormal data are also easily related.
The normal probability distribution is encountered in a number of applications. The
bell-shaped curve of the normal distribution is shown below in Figure 1. Excellent
introductions and reviews of the normal distribution are found in numerous statistical,
engineering, and scientific texts, as for example in Reference [2]. Only a brief review is
given here.
FIGURE 1
A sample of independent observations, denoted by x,, x^...^ from a normally
distributed population can be used to estimate the mean, ji, and variance, a2, according
the formulas below:
H = estimated mean
= Z[x,]/k, l
-------
a2 =	estimated variance
=	s[(x,-M)2]/(k-l), l
-------
Maximum Monthly Average permit limits based on the normal distribution
The normal distribution can be used to model the averages of the individual
measurements for a wide range of circumstances. Although the normal distribution is
usually not an appropriate model for individual pollutant measurements, the averages of
those individual measurements can often be modeled by the normal distribution. This
sub-section explains the theory behind using the normal distribution for averages and
provides the general formulas.
A powerful statistical result, called the Central Limit Theorem, provides theoretical
support for determining limits based on averages of individual measurements.
According to the Central Limit Theorem, when the sample size n is large enough, the
average of the n sample values will be approximately normally distributed regardless of
the distribution of the individual measurements. In determining permit limits, the
calculations incorporate the number of samples that will be required for monitoring
purposes during the specified time period (usually a month). For the purposes of permit
writing, monitoring sample sizes greater than 10 are considered to be "large enough" to
assume the sample average is approximately normally distributed. The above formulas
can be modified for finding the estimated mean and variance for the average from a
sample of size n (e.g. for 14-day monthly average, n=14 samples during the month for
monitoring purposes). The parameters nn and cr „ denote the mean and variance,
respectively, of the distribution of the average of n values. The estimates of the n-day
average and the variance of the n-day average are denoted by nn and a2n, respectively.
H = estimated mean of distribution of X
a2 = estimated variance of distribution of X
Hn = mean of distribution of the n-day monthly average
= M
a£n = variance of distribution of the n-day monthly average
= a2/n
an = standard deviation
= (^n)1/2
cvn = coefficient of variation
=
The upper percentile limits are:
X p = pth percentile maximum n-day monthly average limit
= un + Z-On
'n p n

-------
E(X) =	daily average
=	exp(/x+cr 2/2)
V(X) =	variance
=	exp(2^+a 2)[exp(a2)-l]
cv(X) =	coefficient of variation
= [exp(a2)-l](1/2)
Daily Maximum permit limits based on the lognormal distribution
The upper percentile limits for the random variable X (which is lognormally distributed)
are:
X p = pth percentile daily maximum limit = exp[ju + z_a]
where zp is the pth percentage point of the standard normal distribution.
For example:
X95 = 95th percentile daily maximum limit = exp[/x + (1.645)a]
X 99 = 99th percentile daily maximum limit = exp[/x + (2.326)ff]
Note:
Zg5= 1.645
z99=2-326
The daily variability factors (denoted by VF.,) are estimated by:
daily maximum 95th percentile VF., = X 95/E (X)
daily maximum 99th percentile VI^ = X gg/E (X).
Maximum Monthly Average permit limits based on the lognormal distribution
This subsection contains the formulas required to approximate the distribution of the
average of a small number of lognormally distributed values with another lognormal
distribution. Although, the Central Limit Theorem holds that the average of a sample
of independent measurements is normally distributed provided that the number of
measurements, n, is sufficiently large, the minimum value for n required in specific cases
may vary considerably. In cases where the individual values are lognormally distributed,
the minimum required for the average to be normally distributed may be quite large.
As a consequence, the distribution of the average of a small number of lognormally
distributed values may be better approximated by another, related lognormal distribution

-------
(see Barakat [3]). For sample sizes larger than ten when the data are lognormally
distributed, the calculations given in Table 4 should be used. For the purposes of
permit writing, monitoring sample sizes of 10 or less are considered to be "small enough"
to assume the sample average is approximately lognormally distributed. The mean,
variance^and coefficient of variation of the distribution of the average of n daily values
are un, a2n, and cv, estimated by:
o n= variance
=	ln{V (X)/[n[E (X)]2] + 1}
Hn =	n-day monthly average
=	ln(E(X)) - 0.5?„
an =	standard deviation
_ -
cvn= coefficient of variation
= [exp(a2n)-l](1/2)
where E(X) = exp(n+a 2fl)
V(X) = exp(2/i+a 2)[exp(a2)-l]
The upper percentile limits of the maximum n-day monthly average are:
X p = pth percentile maximum n-day monthly average limit
= exp[Mn+ zpCTn]
where Zp is the pth percentage point of the standard normal distribution.
For example:
X 95 = 95th percentile maximum n-day monthly average limit
= exp[Mn+ (1.645)an]
Xgg = 99th percentile maximum n-day monthly average limit
= exp[Mn+ (2.326)ffn]
Note:
z^l.645
Zgg=2.326
The variability factors are:
monthly maximum 95th percentile VFn =
monthly maximum 99th percentile VFn =
^.95^ n
X.99/^ n

-------
where Zp is the pth percentage point of the standard normal distribution.
For example:
X 95 = 95th percentile maximum n-day monthly average limit
The above discussion of the normal distribution can be modified for data from the
lognormal distribution. The next sub-section explains the modifications.
Lognormal Distribution
Experience has shown that daily pollutant discharges are generally lognormally
distributed. The distributional fit of the data varies somewhat from application to
application, but not enough to alter the conclusion that effluent pollutant discharges are
generally lognormally distributed. Ambient water quality data are also often lognormally
distributed. The following figure (Figure 2) displays the positively skewed shape of the
lognormal distribution.
FIGURE 2
Note:
Z95=1-645
2^=2.326
The monthly maximum variability factors (denoted by VFn) are estimated by:
monthly maximum 95th percentile VFn = X95//x
monthly maximum 99th percentile VFn = X 99/m
1
6
a


-------
The distribution fitting methods assume that the daily measurements are independent,
uncorrelated observations. Although, in general, this assumption is not satisfied exactly,
the lognormal distribution has been used in the effluent guidelines program primarily
because it consistently provides a reasonably good fit to observed effluent data
distributions. Figure 3 shows the lognormal distribution applied to data used in the
development of the Organic Chemicals, Plastics and Synthetic Fibers (OCPSF) effluent
guidelines regulation. [1]
FIGURE 3	B0D5 freQWENCY DISTRIBUTION
PLANT C
>.
o
2
3
O
iu
ec
u.
180
100
so
¦ • 1

1 1—
( i i r —r i" T- r
-


-





-




-







,
	l_.
	1 .
,
.1
i 1 i i i	1— i —i—
1« 11 27 99 99 4« 81 C7 M
CONCENTRATION IN mg/l
The logarithmic transformation of the random variable X, Y=ln(X) results in a random
variable Y that is normally distributed. Therefore, the analysis procedures for analyzing
lognormal data are similar to those for the normal distribution. The mean and variance
from the normal distribution of the random variable Y are n and a2 respectively. These
parameters can be estimated by:
M = Z(y,)/k
and a2 = Z[(y,-^)2]/(k-l), respectively
where yj = ln(Xj) for i=l,2,...k.
These values from the normal distribution can then be used to calculate the mean,
variance, and coefficient of variation for the random variable X which is lognormally
distributed. The mean, variance, and coefficient of variation of the random variable X
may be estimated by E(X), V(X), and cv(X) respectively.

-------
Delta-Lognormal Distribution
The delta-lognormal distribution is a generalization of the lognormal distribution. The,
delta-lognormal distribution may be used when the data contain a mixture of non-detect
values and values above the detection limit. The delta-lognormal distribution can be
used to model non-detects in water quality-based limits. The values above the detection
limit are assumed to be lognormally distributed values. The delta-lognormal distribution
can be used in setting daily maximum limits and for setting limits on maximum monthly
averages with the required number of monitoring samples being ten or less.
The delta-lognormal distribution models data as the combination of two distributions:
the lognormal distribution, and a distribution with discrete probability of obtaining
observations at or below the detection limit. The lognormal distribution models the
observations above the detection limit. The non-detect values are modeled by the
distribution with discrete probability of obtaining observations at or below the detection
limit. The organic priority pollutant data set shown in Figure 4 contains a number of
observations that were reported as "non-detect". These detection limit measurements
are observations that are censored at the detection limit and are represented by the left-
most bar in the histogram. Data sets of this form are fairly typical of organic chemicals
in wastewater. The delta-lognormal distribution provides a convenient model for dealing
with these data sets.
FIGURE 4
160
ORGANIC PRIORITY POLLUTANT
FREQUENCY DISTRIBUTION
PLANT B
>
o
z
ui
3
O
UJ
K
120
80
40 -
-10 90 70 110 180 190
i»(f S70 >10 880
CONCENTRATION IN ug/l

-------
The estimation procedure for the delta-Iognormal distribution assumes that a certain
proportion, 5, of values are at the detection limit, which is denoted by D. (The
estimation procedure when D=0 is covered in further detail by Aitchison and Brown
[4].) These values set to D are observations that can only by quantified as non-detect
(ND) at some minimum level. This minimum level is the detection limit as established
by the laboratory performing the chemical analysis.
Let xv x2,...pcrpcr+1,...pck denote a random sample of size k, with r observations recorded
as non-detects, and k-r observations greater than the detection limit. The k-r positive
observations are assumed to follow a lognormal distribution. The entire data set is
assumed to follow the delta-Iognormal distribution with censoring point equal to the
detection limit D. Let v and o2 be the sample mean and variance of the distribution of
the logarithmic transformation Y=ln(X). Let 6 be the sample proportion of non-
detects. Then the estimates of the mean and variance of the delta-Iognormal
distribution are estimated by:
E(X*) = daily average
= 6D + (l-f)exp(M+0-5a 2)
V(X*) = variance
= (l-£)exp(2/i+a 2) [exp(cx2) - (1-*)] +
6(1-5)D[D-2exp(/i+0.5a 2)]
cv(X*) = coefficient of variation
= [V(X*)](1/2) / E(X*)
where
k = number of samples
D = detection limit
. r = number of non-detect values in sample
k-r = number of values greater than the detection limit
y, = ln(Xj)
r+l
-------
Daily Maximum permit limits based on the delta-lognormal distribution
The 95th and 99th upper percentile limits for the random variable X (which is delta-
lognormally distributed) are given by the following formulas:
The estimated 95th percentile daily maximum limit is:
r D	6>0.95
X .95 = ^
L max [D, exp(/x + z*a)]	&<0.95
where
z = *-1[(0.95-5)/(l-fi )].
The estimated 99th percentile daily maximum limit is:
r D	*>0.99
* 99- = H
L max [D, exp(M + z* a)]	6<0.99
where
z = *-1[(0.99-£)/(l-5)].
$'1[ • ] is the mathematical notation for Z-scores. For example, when 6=0, then the
corresponding value is *"1[.99]=z99=2.326. Values of 4"1[-] are available from tables of
the normal distribution (available in most statistical textbooks and references).
The variability factors (denoted by VF) are estimated by:
daily maximum 95th percentile VF = X95/E (X)
daily maximum 99th percentile VF = Xgg/E (X).
Delta-Lopnormal Distribution of Averages
The derivation of the formulas for the averages is computationally difficult and beyond
the scope of this appendix. However, the formulas for n-day averages are included in
Table 3. The derivation of 4-day monthly averages using the delta-lognormal
distribution is available in Appendix VII-F of the Development Document for the
OCPSF regulation [1].

-------
Checking Distributional Assumptions
Two methods of checking distributional assumptions are: goodness-of-fit, and probability,
plots. When checking distributional assumptions, the sample size must be large enough.
Small sample sizes may lead to erroneous conclusions.
Goodness-of-fit tests
In some cases, statistical goodness-of-fit tests may indicate that a particular distribution
provides a reasonable fit to a dataset of pollutant measurements. Such cases should be
evaluated carefully to verify that the frequency curve for the data also show the shape
characteristic of the distribution.
Probability plots
Use of probability plots is one method of determining whether a normal distribution is
appropriate for modeling a population using only a limited set of measurements. The
set of measurements should have at least 20 observations (Johnson & Wichern [5]).
Consider an independent sample of size k, labeled	Let u1,u2,...,uk be the
ordered sample of x-values in ascending order in which u^i^...^. Now for each uj(
find z, from the normal table (in any statistical reference or textbook) such that
P[Z
-------
Other distributions
If the probability plots or the log-probability plots show serious deviation from straight
lines, other distributions should be considered. Non-parametric methods, which do not
require the assumption that the data follow a particular distributional form, are often
useful for this type of data. Further details are available in many statistical references
(e.g. reference [6]).
Correlation
Up to this point in the appendix, we have assumed that all the observed pollutant levels
are independent, i.e. uncorrelated. In the case of the maximum monthly average limit
derivation, this assumption can be quite critical. If the effluent levels are correlated, the
actual monthly average limit can be substantially higher than that derived from the
analysis based on the independence assumption. However, correlation has essentially no
effect on the calculated daily permit limits. This sub-section provides guidance on
determining when levels may be correlated, and adjusting the sample size.
A major factor that determines whether effluent levels are highly correlated is the
retention time of the wastewater treatment system. If the retention time is large relative
to the time between effluent samples, then those samples will be correlated in most
cases. In municipal systems, for example, the retention time is frequently a matter of
days, and sampling is often conducted on a daily basis. The effluent levels,
consequently, may be substantially correlated. However, in many industrial systems, for
instance a physical/chemical treatment system for electroplating wastewaters, the
treatment system retention time is relatively short: four to eight hours. Daily effluent
levels from these kinds of systems are generally uncorrelated; i.e., statistically
independent. These general patterns are the same irrespective of the kind of pollutant
in question. Correlation, when present, should be factored into the limit.
Several different methods can be used to adjust the sample size for correlation. One
method is statistical time series analysis. Another method would be to use a covariance
matrix type of approach. Help in adjusting the sample size for correlation is available
from the OWRS Statistics Section (phone number: 202-382-5397).

-------
TABLE 2: DAILY MAXIMUM PERMIT LIMIT CALCULATIONS
The daily maximum permit limit is usually the 99th upper percentilevalue of the pollutant distribution. In certain cases the 95th percentile value
may be allowable. TTie following gives the formulas:
\Vi 1H ALL MEASUREMENTS > Db'l fcCl ION LIMIT fbased on loenormal distribution1):
X„ = 95th percentile daily maximum limit
= exp[£ + (1.645)&]
X„ = 99th percentile daily maximum limit
» exp[/i + (2326)tf
where:
x, « daily pollutant measurement i
y, » ln(x.)
k « sample size of data set
H = 2(y,)/k	l«i exp(M+ff 72)
V(X) - exp(2u*a *)[exp(ff,)-l]
cv(X) = [exp^n]""'
WITH SOME MEASUREMENTS < Db l fcCl lON LIMIT fbased on delta-loenormal distribution-):
X„ = 95th percentile daily maximum limit
r D	ii0.95
1
L max [D, exp(£ + za)}	£<0.95
with z'« •'1[(0.95-S)/(l-£)].
X „ = 99th percentile daily maximum limit
r D	S> 0.99
X.„ H ,
1 max [D, exp(M + z'o)]	<5 <0.99
with z* = »"lK0.99-5);(W)]
where:
x, « daily pollutant measurement!
k « sample size of data set
D • detection limit (as established by the laboratory)
r » number of non-detects (x,^,,...^ are D)
y, = ln(x,) for r+lsi
-------
TABLE 3: MAXIMUM MONTHLY AVERAGE PERMIT LIMIT CALCULATIONS
FOR TEN SAMPLES OR LESS
The maximum monthly average permit limit is usually based on the estimates of the 95th percentile of the distribution of the average of the daOy
effluent values. For sample sizes less than or equal to 10, the data are assumed to be lognormally distributed (or delta-lognormally distributed if
the data includesnon-detects).
AM MEASUREMENTS > DETECTION LIMIT (based on lognormal distribution):
X „ = 95th percentile maximum n-day monthly average limit
¦ explM, + (1.645&]
X.„ = 99th percentile maximum n-day monthly average limit
= exp[i, ~ (2.326fo]
where:
x, = daily pollutant measurement
y, = ln(x,)
k = sample size of data set
M = Z(y,)/k.	l(2i+<7"1)[exp(i7i)-l]
«*.- lnfV(X)/(n[E(X)]>l}
Mn - ln(E(X)) - 0.5yn
cv„- [exp{dJ„)-l](,/11
SOME MEASUREMENTS < Dfa l hL'l lON LIMIT (based on delta-lognormal distribution):
X „ = 95th percentile maximum n-day monthly average limit
r D	£>0.95
X.„ = -I
L max [D, exp(k + z'd„)]	i<0.95
with z = #-'[(0.95-£)/(l-£)].
X„ = 99th percentile maximum n-day monthly average limit
r D	£>0.99
X „ = -I
L max [D, exp(i, + z'an)]	£<0.99
with z' = •"'[(0.99-J)/(l-£)]
where:
x, = daily pollutant measurement i
k = sample size of data set
D ¦ detection limit (as established by the laboratory)
r ¦= number of non-detects (x„x,,...,x, are D)
y, a ln(x,) for r+lsiik
£ = r/k
M= Z(y,)/(k-r) r+lii £"D)J
C= (2TDy(E(X*)-rD)
k = ln[(E(X-) - £"D)/(1 - £")] - 0.5cr\

-------
TABLE 4: MAXIMUM MONTHLY AVERAGE PERMIT LIMIT CALCULATIONS FOR MORE
THAN TEN SAMPLES
The maximum monthly average permit limit is usually based on the estimates of the 95th percentileof the distribution of the average of the daily
effluent values. These daily values are assumed to be lognormalty distributed. For sample sizes larger than 10, the averages (represented by the
random variable X„) are assumed to be normally distributed.
X „ » 95th percentile maximum n-day monthly average limit
- E(X.) + 1.645[V(X„)]'"
X „ - 99th percentile maximum n-day monthly average limit
= E(X.) ~ 2.326[V(X.)]"'
where:
x, ¦ daily pollutant measurement i
y, ¦ ln(x,)
k = sample size of data set
M= L(y,)/k,	liiik
if - X[Cyt-M)J]/(k-l) liiik
E(X) » exp(M+c V2)
V(X)- oip(2**0[exp<*»)-l]
E(X,) - E(X)
V(X„) = V(X)/n
cv(X,) . [V(X„)"«]/ E(X.)

-------
Cited References
1.	USEPA: October 1987a, Development Document for Effluent Limitations
Guidelines and Standards for the Organic Chemicals, Plastics and Synthetic
Fibers, EPA 440/1-87/009.
2.	Mendenhall, W., Scheaffer, R., and Wackerly, D.: 1981, Mathematical Statistics
with Applications. Second Edition, Wadswoith, Massachusetts.
3.	Barakat, R.: 1976, 'Sums of Independent Lognormally Distributed Random
Variables', J. Optical Soc. Am. 66, 211-216.
4.	Aitchison, J. and Brown, J.: 1963, The Lognormal Distribution. Cambridge
University Press.
5.	Johnson, R. and Wichern, D.:1982, Applied Multivariate Statistical Analysis.
Prentice-Hall, New Jersey.
6.	Hollander, M., and Wolfe, D.: 1973, Nonparametric Statistical Methods. Wiley,
New York.
Other References
1.	Kahn, H. and Rubin, M.:1989, 'Use of Statistical Methods in Industrial Water
Pollution Control Regulations in the United States', Environmental Monitoring
and Assessment 12: 129-148.
2.	Kahn, H.: 1989, Memorandum: 'Response to Memorandum from Dr. Don Mount
of December 22, 1988', U.S. EPA, Washington, D.C., to J. Taft, U.S. EPA,
Permits Division, Washington, D.C., August 30, 1989.

-------
APPENDIX F.
Sampling

-------
Sampling
The objective of an effluent or instream sampling program is to obtain a sample (or
samples) from which a representative measure of the parameter of interest can be
obtained. Unfortunately, many of the industrial and municipal NPDES sampling
protocols presently in use are carryovers from schemes used for calculating loadings of
nutrients and oxygen-demanding substances, or were developed to evaluate treatment
plant operational efficiency. Sampling for individual toxicants and particularly for
effluent toxicity can require more specific land thus different) sampling procedures.
Wastewater variability is an important consideration in selecting the method and
frequency of sampling for both chemical analysis and toxicity testing. Industrial waste
characteristics have been shown to vary in frequency, intensity, and duration (1). As
noted by Bender (2), the sources of effluent variability include both random and
systematic components that influence both daily and annual characteristics of waste
discharges. Although toxic pollutant loading may be of primary concern in assessing
human health impact or bioaccumulation, loading may be of lesser importance in toxicity
assessment than frequency, intensity, and duration of peak toxic discharge. Sampling
must be tailored to measure the type of toxicity of importance for that discharge: either
long term (chronic) impact, which is a more constant effect, or short term (acute)
impact, which is more variable and subject to peaks of intensity.
There are several chemical parameters for which continuous analysis is possible These
include pH, temperature, dissolved oxygen, and other parameters involving instantaneous
measurement. All other types of measurement involve some time period over which the
analysis is conducted. Toxicity tests require an exposure period. Chemical tests require
sample preparation and analysis. There is no continuous analysis method for toxicity.
It should be noted that although it is difficult to design a representative sampling
program for toxicity analysis, the problems are of no greater magnitude than similar
problems associated with obtaining a representative sample for conventional pollutants.
Sampling Methods
Continuous Flow Samples
For toxicity testing, the test organisms may be exposed to serial dilutions of a sample
continuously pumped from the effluent pipe or ditch. In the case of effluents, if optimum
accuracy is desired, then the ratio of effluent flow/test chamber volume can be scaled to
simulate the time-varying concentration at the mixing zone boundary.
Although flow-through methods can provide a realistic simulation of time-varying
exposure, they are relatively expensive and are usually conducted on site. Therefore,
flow-through methods, may only be practical where the goals of the analysis of impact
require this type of testing or where treatment costs are sufficiently high that this type of
analysis can be required. A flow-through exposure method is not a continuous analysis
because only one result or data point is obtained at the end of the test. However, the
continuous exposure does provide some measure of time-varying exposure effects.

-------
Discrete Samples
Grab or flow composited sampling provide a discrete sample for chemical analysis or
toxicity testing. Static or renewal toxicity tests using discrete samples result in exposure
of test organisms to a constant effluent composition over the period of the tests, or for
the period between renewals.
If discrete samples are collected during peaks of effluent toxicity then constant
concentration exposure static tests provide a measure of maximum effect.
Depending on the duration of a peak and the compositing period, composited samples
may not be useful for examining toxicity peaks because the compositing process tends to
dilute the peaks. Composited samples are usually appropriate for chronic tests where
peak toxicity of short duration is of less concern. The averaging effect of compositing
may be misleading when testing for acute toxicity.
Grab samples must be collected at sufficiently frequent intervals to provide a high
probability of sampling daily peaks. Fortunately static toxicity tests are relatively
inexpensive and can be done on shipped samples; thus, it may be cost effective to
conduct individual tests on a series of grab samples collected over a 24-hour period.
Sampling Frequency
Non-random effluent variability, resulting from batch processing, variable loadings, etc.,
is often known or can be determined. Therefore, the first step in designing a sampling
program for chemical analysis or toxicity testing is to select the annual sampling
frequency based on available site-specific operational information. This is important in
selecting sampling periods for both continuous flow and discrete sampling methods.
If discrete sampling methods (grabs or composites) are used, then random variations
between and within days for each sampling period must be considered. It is important
to recognize the tradeoff between the long term (between days) frequency and short
term '(within days) frequency of sample collection and analysis for toxics. At present, the
permit requirements for sampling and analyzing chemical parameters are site specific
and generally involve a single grab or 24-hour composite sample collected at daily,
weekly, or monthly intervals. Unfortunately, a sampling scheme involving a single daily
grab or a 24-hour composite sample can conceal the presence of those daily extreme
values that may be of importance To optimize sampling cost and effectiveness, it may be
desirable to reduce long term frequency so that daily frequency can be increased.
For example, a weekly grab or composite involves 52 analyses per year. It may be more
efficient to reduce the annual frequency to monthly or bimonthly, but collect and
analyze four or eight grabs daily. Either scheme (12 x 4 or 6 x 8) would involve 48
analyses per year vs. 52 for the weekly single sample approach. Assuming that daily toxic
events of environmentally significant intensity and duration would not be masked by
short
The initial sampling design step should involve stratification of sampling periods to

-------
account for non- random sources of variation (e.g., batch processing). The second step
includes selection of the frequency and the method of sampling to be conducted within
each sampling period. Depending on site-specific considerations, several options are
available term composites, it might be more efficient to collect eight samples each
composited over a three-hour interval.
If costs or other constraints prohibit satisfactory daily and annual replication of
sampling, then a level of uncertainty must be introduced into the calculations used to
evaluate waste toxicity Box F-l presents EPA's recommendations on sampling methods.
Box F-l
RECOMMENDATIONS
o Flow-through Methods - Ideally, for both acute and chronic effluent toxicity tests, the exposure of
biota should simulate the time-varying concentration at a predetermined point in the receiving water.
For regulatory purposes, the critical point is often the edge of the mixing zone where the waste
should exhibit neither acute nor chronic toxicity. Therefore, if warranted by site-specific factors, it is
recommended that test biota be exposed to a continuously collected flow-through sample of serially
diluted effluent. If no systematic annual variations (e.g., batch processing) are known or suspected,
flow-through testing can be conducted at a minimum of quarterly intervals for at least one year.
o Grab Sample Methods - Grab samples are recommended for chemical analyses and for acute and
chronic toxicity tests where site conditions (such as wastewaters that are known to have relatively
constant composition) do not require use of continuous flow methods. Grab samples of effluent or
receiving water may be used for static or renewal acute toxicity tests, which may be conducted on
site or at a remote lab. The design of a toxics grab-sampling program must take into account the
trade-off between long term and short term sampling intensity. Where there is no ponding of wastes
or retention time is insufficient for thorough mixing, it is important to collect or analyze a sufficient
number of samples to provide a measure of daily spikes. Therefore, to minimize analytical costs
where daily fluctuations are known or suspected, the annual sampling frequency should be reduced
in favor of more intensive daily sampling. It is recommended that on an annual cycle, grab sampling
and analysis include a minimum of four to six daily grabs collected monthly. An option could include
the use of short term (four-hour) composites rather than grabs. If site-specific data are available to
indicate that treatment system retention time is adequate to minimize daily variations, then the daily
replicates may be omitted in favor of more frequent annual sampling (e.g., weekly or semimonthly
rather than monthly). If, to minimize costs in screening tests, only single samples are collected at
infrequent intervals (eg, quarterly) an uncertainty factor for variability should be used in the toxicity
evaluation (see Section 3).
o Composite Sample Methods - If static or renewal methods are used for evaluation of toxicity, it is
recommended that 24-hour, continuous-flow composite samples be collected. Considerations of
annual frequency are the same as those for grab samples.

-------
REFERENCES
1.	Nemetz, P.N., and H.D. Dreschler. 1978. The role of effluent monitoring in
environmental control. J. Water, Air, and Pollution. 10:477-97.
2.	Bender, E.S. 1984. Sources of variations in effluent toxicity tests. In:
Environmental Hazard Assessment of Effluents. (H.Bergman, R. Kimerle, and
A.w. Maki, eds.) Proceedings of the Fifth Pellston Environmental Workshop.
Cody, WY, August, 1982.

-------
APPENDIX G.
The Development of a Biological Indicator Approach to Water
Quality-Based Human Health Toxics Control

-------
The Development of a Biological Indicator Approach to Water
Quality-Based Human Health Toxics Control
Current Approach
With one exception (New Jersey), the chemical-specific approach to protecting
human health is currently the only method used to regulate human health toxicants in
effluents. The chemical-specific approach identifies the individual chemicals in an
effluent and regulates them based upon health risk assessment information for each
individual chemical. Where data are available for such human health toxicants, the
chemical-specific approach can be used to develop permit limits.
However, the complex characteristics of effluent mixtures limit the effectiveness
of the single-chemical approach. When used as the sole basis for identifying effluents of
human health concern, the chemical-specific approach can overlook wastewaters
potentially toxic to humans for the following reasons:
1.	Analytical methods may not be sensitive enough to detect extremely small
quantities of chemicals which may exert their effects on human health after a
long latency period.
2.	Human health data are limited or lacking for many of the §307(a) "priority"
pollutants. Moreover, the number of human health toxicants discharged far
exceeds the "priority" pollutants list.
3.	The various chemical constituents of an effluent may resulting in synergistic,
additive or antagonistic chemical effects.
As a result of these limitations, biological indicator tests have been developed for
human health impact effluent screening, including both in vitro and in vivo tests.
Though not yet widely implemented, biological indicator test results can be important
supplements to a chemical-by-chemical effluent characterization.
Short-term biological indicator tests for human health impact screening are based
on cellular-level responses, indicating whether the substances being tested are
biologically active, and providing some measure of that activity. While these tests do
not quantify the degree of toxicity to humans, they can be used to identify effluents with
potential human health impacts, and regulatory priority-setting and targeting of
dischargers for futher chemical-specific analyses. Research is currently underway within
EPA and in the private sector to evaluate various biological indicator test batteries for
whole effluent analysis.
Biological Indicator Tests
Biological indicator tests include in vitro (test tube) and in vivo (whole animal)
tests which can help form the first tiers of a single chemical evaluation process. A
battery of simple biological tests can be used to test for the major types of effects which
are underlying causes of potential health impact, since each biological test measures a

-------
different type of response. The results of these tests can be used to decide whether
more definitive (and more resource-intensive) testing is needed to identify actual
problem pollutants.
Test results can serve as. triggers to additional chemical-specific analysis or more
sophisticated definitive biological tests. Where results of these screening tools indicate
potential health hazards, further characterization of the effluents, and regulation based
upon toxicological data and/or chemical structure-activity relationships can proceed. If
an effluent is extremely variable in other parameters, screening assays should be
repeated periodically to ensure that potentially hazardous discharges are detected. Two
types of biological indicator tests are discussed below: tests for non-threshold (no safe
level exists) chemicals and tests for threshold (a safe level is presumed to exist)
chemicals.
Genotoxicitv Tests for Non-Threshold Chemicals
Genotoxicity is the ability of a substance to damage an organism's genetic
material (its DNA). Certain positively-charged compounds tend to bind to DNA and
may lead to permanent changes in the genetic information. Such damage to the DNA
of reproductive (germ) cells can impair reproductive ability or can produce a change in
the DNA structure that could be passed on to offspring as a heritable mutation.
Alterations in the DNA of somatic cells can result in cancer or other diseases.
Interpretation of genotoxicity test results assumes that DNA damage in
nonhuman cells may be predictive of latent diseases in humans such as genetic disorders,
birth defects, and cancer. EPA believes that genotoxicity tests for point mutations,
numerical and structural chromosome aberrations, DNA damage/repair and in vitro
transformation provide supportive evidence of carcinogenicity [U.S. EPA, 1979 and
1987c]. In addition, wastewater mutagenicity tests could be used to detect genotoxic
activity which can adversely affect aquatic biota [Black, et al., 1980]. Several short-term
assays have been developed which can assess genotoxic effects (discussed below).
.For example, a correlation has been established between animal carcinogens and
positive mutagenic responses in the Ames Test The Ames test is often used to assess
point mutation effects. The original correlation study revealed that 90% of tested
carcinogens were detected as mutagens, while 87% of noncarcinogens were identified as
nonmutagens. Other studies have determined that between 77% and 91% of tested
carcinogens produce positive responses in the Ames test. The Ames Test has been used
in over 2,000 laboratories worldwide for drug and food additive screening, product
development, and environmental testing [New Jersey DEP, 1983].
To assess clastogenic effects (chromosomal breakage)
either the mammalian sister chromatid exchange test or a mammalian cell chromosomal
aberrations test can be conducted. Both of these tests typically use Chinese hamster
ovary cell cultures and involve cytologic examination after exposure to determine if
chromosomal effects are evident. The Organization of Economic Cooperation
Development (OECD) test methodology is recommended [OECD]. EPA's Office of
Toxic Substances and Office of Pesticides Programs also have published test methods
[U.S. EPA 1982a and 1982b] that are consistent with the OECD tests.

-------
Most effluent samples need special preparation (for example, concentration) to
produce a measurable biological indicator test response for human health effects. When
samples are concentrated, the response is calculated in terms of the pre-concentration
sample. In addition, for genotoxicity tests, because many chemicals are not actively
mutagenic in humans until they enter the body and are metabolized, many in vitro tests
are supplemented with extracts from mammalian livers which act as a source of
enzymes. The extract enzymes act to mimic metabolic activation of procarcinogens and
promutagens in humans, providing a more realistic picture of potential effects [U.S.
EPA, 1979].
A number of genetic toxicity assay batteries have been suggested in order to
address the many potential effects produced by nonthreshold chemicals (for which no
safe level exists) [U.S. EPA, 1979; Lave and Omenn; Environment Canada]. In addition
to providing assays that detect different endpoints, a battery of tests can also be
structured to minimize effort at the screening level while supplying more definitive data
for samples failing the initial tier of testing. Positive results can lead to further effluent
characterization, including priority and other pollutant chemical analyses, or
mutagenicity testing of specific processes or effluent fractions. Another approach would
be to evaluate the effects of various treatment or waste segregation techniques on
mutagenicity [McGeorge, et. al., 1985].
Many of the proposed test batteries utilize the Ames Assay as a screening level
test because of its relatively high degree of sensitivity (i.e. a high percentage of
carcinogens are Ames positive) and specificity (i.e. a high percentage of noncarcinogens
are Ames negative) [Tennant, et. al., 1987]. One study of 28 selected industrial
discharges revealed that 11 of the 28, or 39%, produced positive results using the Ames
Test (described below). Other test endpoints frequently covered in the initial tier of
testing include mammalian cell chromosomal effects, mammalian gene mutation and
microbial and mammalian cell DNA damage.
Results of a recent National Toxicology Program project suggest that
combinations of four of the most commonly used short-term tests covering these
endpoints did not show significant differences in individual concordance with rodent
carcinogenicity results for pure chemicals [Tennant, et. al., 1987]. This suggests that if a
sample causes only one type of endpoint as measured by several screening level tests, its
potential to cause human health effects should not be disregarded.
To assess the potential carcinogen hazard, subsequent tests focusing on effluent-
induced malignant changes in mammalian cells in vitro can be conducted. Higher levels
of testing may include in vivo rodent testing or the Medaka (fish) tumor assay, for
example. It should be noted that under existing guidelines, in vivo mammalian tumor
assays are necessary to establish a material as a possible human carcinogen. Results
from short term tests alone are considered as inadequate to establish human
carcinogenicity [U.S. EPA, 1986c]. Guidelines for risk assessment of individual
compounds are covered in U.S. EPA, 1986b and 1987c.
In vivo tests on complex mixtures are extremely complicated and expensive given
the variability intrinsic to effluents. As a result, it is recommended that after each tier

-------
of biological indicator testing, the cost of further refining the weight of evidence for
carcinogenesis or mutagenesis be balanced against the cost of conducting a causative
agent identification evaluation. Given the identity of the substance leading to positive
results in short term in vitro tests, it should not be necessary to generate in vivo dose-
response data for risk characterization if these data are already available in the
literature for the specific chemical.
In addition, causative agent identification studies may be unnecessary if
information on the physical and/or chemical characteristics of the toxicant is obtained.
Such information may provide clues to appropriate effluent treatment technologies
needed to reduce effluent mutagenicity.
In weighing the need for more definitive biological assays against causative agent
evaluation, the frequency (i.e., how often the effluent tests positive) and intensity (e.g.,
revertants/iiter) of the effluent's mutagenicity must be considered. As a default
assumption, a high dose of a carcinogen received over a short period of time is
equivalent to a low dose spread over a life-time [U.S. EPA, 1986c]. While effluents
which are highly variable in their mutagenicity are of concern, they will be more difficult
and costly to deal with in subsequent phases of study.
Accordingly, the initial tier of qualitative tests for human health effects
assessment can be relatively inexpensive, rapid, and have a low rate of false negative
results. Subsequent tests can be designed to increase confidence in the predictive nature
of the results. Additional levels of testing may also provide diagnostic information on
the characteristics of the causative agent(s) in the effluent.
Subsequent tiers of testing should focus on a more concise assessment of risk.
Such an assessment can be used to delineate hazard type; in effect, to separate germ cell
mutations (heritable genetic risk) from carcinogen risk. Thus, to assess heritable
mutation, subsequent testing should focus on mammalian germ cells, ultimately tested in
vivo [U.S. EPA, 1986b]. To assess potential carcinogen hazard, subsequent tests
focusing on effluent-induced malignant chaffges in in vitro mammalian systems should be
conducted. Ultimately, testing must result in a dose-response assessment to be used
with an exposure assessment in characterizing risk [U.S. EPA, 1987a].
EPA's Region V (Chicago), New Jersey, and Environment Canada have been
conducting mutagenicity testing at selected facilities. In Region V Ames test results are
used to suggest the need for more intensive chemical-specific analyses of the effluent.
New Jersey has incorporated a prohibition against discharging mutagenic compounds in
amounts that are mutagenic into its "New Jersey Administrative Code" [N.J.S.A. Section
7:9-4.5 (a)4, May 1985].
For both types of endpoints (genotoxicity and carcinogenesis), hazard
identification should be followed by quantitative risk assessment which includes
assessment of doseresponse (requiring in vivo data) and human exposure. Human
exposure assessment typically considers the composition and size of the population
exposed and the types, magnitude, frequency and duration of exposures [U.S. EPA,
1986d].

-------
Evaluation of Effluent Genotoxicitv Screening Results
Control of human health hazards depends upon assessment of both the
toxicological properties of the pollutants and the level of exposure. The permit
authority should review the results of a human health toxicant effluent screening
program and establish the actions triggered by each level of potential risk indicated! For
example, a discharger with either a high exposure risk or a high effects risk might
automatically be required to conduct a detailed assessment or institute controls. A
medium risk in both exposure and effects might require further review of the data and a
case-specific decision about whether to require additional assessment. A medium effects
risk and a low exposure risk might indicate the need for limited testing to ensure that
the low is really indicative of the risk. Low risk in both exposure and effects should
receive low priority for further assessment. The bioconcentration evaluation procedures
can be used to aid in defining exposure risk, as well as determining receiving water
concentration.
One possible tool for evaluating results of biological indicator effluent screening
is the "relative potency approach," a concept used rather widely in radiation biology and
chemical pharmacology. The relative potency of an effluent is the dose of a reference
agent needed to produce an effect of a given magnitude in a particular bioassay, divided
by the dose of the effluent needed to produce the same magnitude of the same effect in
the same bioassay. A predictive battery of several short-term biological tests, when
standardized to a reference agent, could provide a rank or comparative estimate of the
hazard posed by an effluent in the context of measures of other known hazards [Glass,
1988]. It should be recognized that this approach does not consider exposure through
bioaccumulation.
When screening has indicated a high potential for health hazard, further
assessment should be required. A chemical-specific approach is recommended to
evaluate and regulate the discharge constituents. The first half of this process involves
characterizing the composition of the effluent. Typically, only a small fraction of the
total organic carbon (TOC) can be accounted for as individual chemicals. Therefore,
effort should be placed on identifying constituents through means other than chemical
analysis, such as through a detailed process evaluation and/or toxicant characterization
evaluation.
A process evaluation is a study in which components in the wastewater are
determined from an analysis of feedstocks, manufacturing processes, products, by1
products, and pollution control in place. The result is a list of compounds or classes of
compounds with a high probability of t>eing present in the wastewater. Chemical
analysis can also be conducted for not only the priority pollutants but also nonpriority
pollutant peaks and bioconcentratable chemicals [EPA/600/xx-xx]. IRIS and SAR can
be used to determine the likelihood that a given compound is causing positive results in
the bioassay. The toxicant characterization evaluation can provide information on the
physical/chemical nature of the chemical producing positive bioassay results.

-------
Summary of Current Biological Indicator Tests for Non-Threshold Human Health
Toxicants
The following tests are currently in use or under development for assessing
carcinogenicity or mutagenicity:
Salmonella tvphimurium Assay (Ames Test) [U.S. EPA, 1985 and 1983]
Background: Strains of Salmonella requiring the amino acid, histidine, are
exposed to a solvent extract of the effluent. Tests are performed with and
without added rat liver enzyme for activation of indirect mutagens. The bacteria
are grown on histidine-free medium; colony formation indicates the effluent
contains mutagenic compounds capable of genetically altering the bacteria.
Endpoint: Gene mutation; response measured in revertant colonies/L effluent.
Advantages: Test is rapid, relatively inexpensive. The Ames Test has been shown
to have broad application for the assessment of the mutagenic activity of a
diversity of industrial effluent types [McGeorge, et. al., 1985]. Test sensitivity and
specificity are documented [Ashby and Tennant, 1988].
Disadvantages: Requires metabolic activation and several different strains of
Salmonella to detect a broad range of compounds, requires extrapolation from
prokaryot, use of effluent extract may exclude certain types of compounds,
epigenetic carcinogens not detected.
Cost: Approximately $1200 [Lave and Omenn, 1986]
Escherichia coli SOS Assay (SOS Chromotest) [Quillardet, et.al., 1985].
Background: All cells contain an "SOS"enzymatic system for detecting and
correcting errors in their genetic material. A strain of E. coli has been
genetically engineered so that DNA damage ultimately results in production of an
enzyme which reacts with test reagents to form a blue color. Bacteria are
exposed to effluent or an extract of the effluent, with or without added rat liver
enzyme for activation indirect mutagens. The intensity of color produced
indicates the extent to which the effluent contains mutagenic compounds capable
of damaging bacterial DNA.
Endpoint: DNA damage; response measured as the change in optical density.
Advantages: Simple kit commercially available, test requires <8 hrs to perform,
relatively inexpensive. Test sensitivity, specificity docuented [Quillardet, et.al.,
1985].
Disadvantages: Requires metabolic activation, extrapolation from prokaryot, use
of effluent extract may exclude certain types of compounds, epigenetic
carcinogens not detected, measurement of effect must be referenced to known

-------
genotoxic compound.
£q§I: ??
Sister-Chromatid Exchange Assay (SCE) [Eckl, et. al., 1987]
Background: Sister chromatid exchange occurs when damaged DNA is replicated
during cell division. Recent advances allow the use of cultured rat hepatocytes in
detecting SCE formation, thus precluding the need to add rat liver enzyme for
metabolic activation. Hepatocyte exposure to the sample is effected by using
filter sterilized effluent in preparing the cell culture medium. Exposed cells are
lysed and genetic material fixed in order to count SCEs.
Endpoint: DNA damage; response measured in SCE per chromosome/L
effluent.
Advantages: Test is rapid, relatively inexpensive, does not require metabolic
activation (therefore more realistic). Uses mammalian cells, therefore results
more readily applicable to humans.
Disadvantages: Sensitivity, specificity not well documented, test more complex
relative to prokaryotic systems, filter sterilization may remove some genotoxic
compounds from the sample, epigenetic carcinogens not detected.
Cost: $5000 [Jirtle, 1989]
HGPRT Assay with Chinese Hamster Ovary Cells (HGPRT/CHO) [Hsie, et. al., 1981]
Background: Strains of Chinese Hamster Ovary cells in culture are exposed to
the effluent or an extract of the effluent, with or without added rat liver enzyme.
Mutagen interactions with certain sections of the DNA make the cell resistant to
toxicants like 6-thioguanine. Cell survival is used to indicate both cytotoxicity
(cell death) and genetic mutations resulting from effluent components.
Endpoint: Gene mutation; response measured in % survival/L.
Advantages: Test is rapid and uses a mammalian system.
Disadvantages: Sensitivity, specificity not well documented, use of effluent extract
may exclude certain types of compounds, epigenetic carcinogens not detected,
requires metabolic activation.
Cost: $6500
Medaka Tumor Assay. [U.S. EPA, 1988; U.S. EPA, 1989b.]
Background: Larval fish are exposed to nonlethal concentrations of effluent for
one month, this period is followed by a 5-month grow out period in clean water.
At six months, fish are sacrificed and submitted for histopathological studies.

-------
Endpoint: Tumor formation, response measured in frequency of tumors at a
given site/effluent concentration.
Advantages: Use of whole effluent, whole organism, oncogenic endpoint
Disadvantages: Carcinogen levels in unconcentrated effluent may not be high
enough to produce tumors at a detectable frequency in exposed populations,
effluent must not be toxic to Medaka, requires extrapolation from non-
mammalian system, relatively expensive, length of test, endpoint requires
pathologist experienced in fish cancers, method still in developmental stages.
Cost: $20,000 [Johnson, 1989].
Other Human Health Effects
Toxicants present in effluents may produce a variety of effects in humans besides
genotoxicity or carcinogenicity via exposure through ingestion of water and/or
contaminated fish and shellfish. Potential health effects could include suppression of the
immune system, neurotoxicity, specific organ toxicity, or developmental toxicity. These
effects occur after exposure above a presumed safe (threshold) level and are referred to
as "systemic."
Formerly, the only means to assess systemic effects was by using subchronic
toxicity procedures designed to determine the effects that may occur with repeated
exposure over a part of the average life span of an experimental animal. However, such
studies are expensive ($100,000 and over) and beyond the cost constraints for most
effluent analyses. As an alternative, a number of short-term in vitro tests utilizing
mammalian cells have been developed [U.S. EPA, 1978; Wilson, 1978; Kimmel, et. al.,
1982; Brown and Fabro, 1982; Borenfreund and Puerner, 1985]. Test endpoints
include cytotoxicity, effects on cell growth, division, structure, metabolism and function,
alterations in enzyme activities, and metabolite formation.
As with the nonthreshhold assays previously discussed, these in vitro assays only
serve to qualify potential human health hazards. In the case of positive in vitro results,
tests on intact mammals can be pursued in order to confirm screening test findings and
establish a dose-response relationship. Alternatively, causative agent evaluations '
resulting in either the identity of the toxicant or toxicity treatability data may be
pursued.
Current Limitations of the Biological Regulatory Approach
At present, the use of biological indicator tests as a regulatory tool is limited for
a number of reasons. First, biological indicator information must be linked to human
exposure to wastewater components. To date, no definitive mechanism exists for
interpreting the human health hazard implications of the biological test results. While
many in vitro (i.e. test tube) human health assays provide data about cellular changes
relative to the dose delivered to the target tissue, they do not provide the information

-------
necessary to correlate environmental exposure to target tissue dose or cellular change to
ultimate human health effects (e.g., cancer). The higher animal testing necessary to
quantify the dose-response relationship (or "potency"of the effluent) would be extremely
costly.
Second, as with aquatic organism toxicity tests, a human health hazard test must
be capable of dealing with intra- and interspecies sensitivity variability. This concern is
particularly relevant for those effluents containing chemicals which only become
carcinogenic upon metabolism by mammalian systems (i.e. procarcinogens). The use of
cultured human liver cells (hepatocytes), currently being tested, would eliminate the
need for interspecies extrapolation.
Finally, whole effluent testing to assess potential human health impacts presents
several unique practical problems such as the continual change in composition typical for
most effluents, the need to concentrate samples to obtain a dose-response curve, and the
need to compensate for or eliminate interferences from cytotoxic (toxic to cells)
components of the effluent. Only those components which occur in the relatively
nonvolatile, nonpolar organic fraction of the effluent sample are conventionally
measured. [Anderson-Carnahan, article in preparation].
Until additional research resolves these difficulties, biological indicator tests will
be most useful as screening tools, with actual regulation of effluents posing potential
health hazards likely to remain on a chemical-by-chemical basis.

-------
REFERENCES
1.	Anderson-Carnahan, L., P. M. Eckl, and R. L. Jirtle. Sister Chromatid Exchange
in Screening Wastewaters. In preparation.
2.	Ashby, J. and R.W. Tennant, 1988. Chemical Structure, Salmonella Mutagenicity
and Extent of Carcinogenicity as Indicators of Genotoxic Carcinogens Among 222
Chemicals Tested in Rodents by the U.S. NCI/NTP. Mutation Research 204:17-
115.
3.	Black, J., P. Dumerski, and Zapisek, W. 1980. "Fish Tumor Pathology and
Aromatic Hydrocarbon Pollution in a Great Lakes Estuary." In B. Afghan and D.
MacKay (eds.), Hydrocarbons and Halogenated Hydrocarbons in the Aquatic
Environment. Plenum Press, New York, 1980.
4.	Borenfreund, E. and J. Puerner. 1985. Toxicity Determined m vitro by
Morphological Alterations and Neutral Red Absorption. Toxicology Letters 24.
pp. 119-124.
5.	Brown, N.A. and S.E. Fabro. 1982. The in vitro Approach to Teratogenicity
Testing. In: K. Snell. ed. Developmental Toxicology. London, England: Croom-
Helm, p.p. 31057.
6.	Eckl, P.M., S.C. Strom, G. Michalopoulos, and R.L. Jirtle. 1987. Induction of
Sister Chromatid Exchanges in Cultured Hepatocytes by Directly and Indirectly
Acting Mutagens/Carcinogens. Carcinogenesis 8:1077-1083.
7.	Environment Canada. 1986. Guidelines on the Use of Mutagenicity Tests in the
Toxicological Evaluation of Chemicals. Health and Welfare Canada. Ottawa,
Canada
8.	Glass, L.R. "Background and Rationale for Relative Potency Framework for
Evaluating Hazards Associated with Waste Water Samples." Appendix B in
"Health Hazard Evaluation of Waste Water Using Bioassays: Preliminary
Concepts". C. E. Easterly, et. al. Oak Ridge National Laboratory, Oak Ridge,
TN 37831-6101 and U.S. EPA Office of Research and Development, Health
Effects Research Laboratory, Cincinnati, Ohio 45268. July 1988.
9.	Hsie, A.W., D.A. Casciano, D.B. Coach, D.F. Krahn, J.P. O'Neill and B.L.
Whitfield. 1981. The Use of Chinese Hamster Ovary Cells to Quantify Specific
Locus Mutation and to Determine Mutagenicity of Chemicals. Mutation
Research 86:193-214.
10.	Jirtle, R., Duke University Medical School, Durham, NC. Personal
communication, April 24, 1989.

-------
11.	Johnson, Rodney. Office of Research and Development, Environmental
Research Laboratory, Duluth, MN. Personal communication, June 1, 1989.
12.	Kimmel, G.L., K. Smith, D.M. Kochhar, and R.M. Pratt. 1982. Overview of m
vitro Teratogenicity Testing: Aspects of Validation and Application to Screening.
Teratogenesis. Carcinog. Mutagen. 2. pp. 221-229.
13.	Lave, L. B. and G. S. Omenn. 1986. Cost-Effectiveness of Short-Term Tests for
Carcinogenicity. Nature 324, 6:29-34. Note: Costs based on 1981 figures for
pure chemicals.
14.	Marx, J.L. 1989. Detecting Mutations in Human Genes. Science 243. pp. 737-
738.
15.	McGeorge, Leslie J., Judith B. Louis, Thomas B. Atherholt, and Gerard J.
McGarrity. 1985. "Mutagenicity Analyses of Industrial Effluents: Results and
Considerations for Integration into Water Pollution Control Programs," in Short-
Term Bioassays in the Analysis of Complex Environmental Mixtures, IV. Edited
by Michael D. Waters, Shahbeg S. Sandhu, Joellen Lewtas, Larry Claxton, Gary
Strauss and Stephen Nesnow. (Hearst Publishing Corp., 1985).
16.	Miller, J.A., and E.C. Miller. 1977. Ultimate Chemical Carcinogens as Reactive
Mutagenic Electrophiles, in H.H. Hiatt, J.D. Watson and J.A. Winston (Eds.),
Origins of Human Cancer, Cold Springs Harbor Laboratory, pp. 605-628.
17.	New Jersey Department of Environmental Protection. Mutagenicity Analyses of
Industrial Effluents: Background and Results to Date. Office of Science and
Research. August 1983.
18.	Quillardet, P., C. de Bellecombe, and M. Hofnung. 1985. "The SOS Chromotest,
a Colorimetric Bacterial Assay for Genotoxins: Validation Study with 83
Compounds". Mutation Research, 147. pgs. 79-95.
19.	Organization of Economic Cooperation and Development (OECD). 1984.
Guidelines for Testing Chemicals. Section 4 -- Health Effects. Director of
Information, OECD, 2, rue Andre-Pascal 75775 Paris CEDEX 16, France.
20.	Tennant, R.W., B.H. Margolin, M.D. Shelby, E. Zeiger, J.K. Haseman, J.
Spalding, W. Caspary, M. Resnick, S. Stasiewicz, B. Anderson, and R. Minor.
1987. Prediction of Chemical Carcinogenicity in Rodents From in vitro Genetic
Toxicity Assays. Science 236:943-941.
21.	U.S. Environmental Protection Agency. 1978. Directory of Short Term Tests for
Health and Ecological Effects. Health Effects Research Laboratory. EPA 600/1-
78-052.
22.	U.S. Environmental Protection Agency. 1979a. Environmental Assessment:
Short-term Tests for Carcinogens, Mutagens and other Genotoxic Agents. Health
Effects Research Laboratory. Research Triangle Park, N.C.

-------
23.	U.S. Environmental Protection Agency. 1979b. Short Term Tests for
Carcinogens, Mutagens, and other Genotoxic Agents. Health Effects Research
Laboratory, Research Triangle Part. EPA 625/9-79-003.
24.	U.S. Environmental Protection Agency. 1982a. Pesticide Assessment Guidelines,
Office of Pesticide Programs. EPA/9-82-018 through 028.
25.	U.S. Environmental Protection Agency. 1982b. Toxic Substances Test Guidelines,
Office of Toxic Substances. EPA/6-82-001 through 003.
26.	U.S. Environmental Protection Agency. 1983. Interim Procedures for Conducting
the Salmonella/Microsomal Mutagenicity Assay - Ames Test. EPA 600/4-82-068.
27.	U.S. Environmental Protection Agency. 1985. Guidelines for Preparing
Environmental and Waste Samples for Mutagenicity (Ames) Testing: Interim
Procedures and Panel Meeting Proceedings. Office of Research and
Development. EPA 600/4/85-058.
28.	U.S. Environmental Protection Agency. 1986a. Guidelines for the Health Risk
Assessment of Chemical Mixtures. Federal Register 51 (185). pp. 34014-34025.
29.	U.S. Environmental Protection Agency. 1986b. Guidelines for Mutagenicity Risk
Assessment. Federal Register 51 (185). pp. 34006-34012.
30.	U.S. Environmental Protection Agency. 1986c. Guidelines for Carcinogen Risk
Assessment. Federal Register 51(185). pp. 33932-34003.
31.	U.S. Environmental Protection Agency. 1986d. Guidelines for Exposure
Assessment. Federal Register 51(185). pp. 34042-34054.
32.	U.S. Environmental Protection Agency. 1988. Validation of the Medaka Assay
for Chemical Carcinogens: A Progress Report (Deliverable # 8095A). Office of
Research and Development, Environmental Research Laboratory, Duluth, MN.
August 1988.
33.	U.S. Environmental Protection Agency. 1989a. Draft Guidance on Assessment,
Criteria Development, and Control of Bioconcen-tratable Contaminants in
Surface Waters.
34.	U.S. Environmental Protection Agency. 1989b. The Medaka Carcinogenesis
Model: A Progress Report (Deliverable # 8094A). Office of Research and
Development, Environmental Research Laboratory, Duluth, MN. February 1989.
35.	Wilson, J. G. 1978. Survey of in vitro Systems: Their Potential Use in
Teratogenicity Screening. In J.G. Wilson and F.C Fiaser, eds. Handbook of
Teratology. Vol. 4. New York, NY. Plenum Press, pp. 135-153.

-------
APPENDIX H.
Reference Dose(RfD):
Description and Use in Health Risk Assessments

-------
I. INTRODUCTION
This concept paper describes the U S. Environmental Protection Agency's principal approach to and
rationale for assessing risks for health effects other than cancer and gene mutations from chronic
chemical exposure By outlining principles and concepts that guide EPA risk assessment for such
systemic* effects, the report complements the new risk assessment guidelines, which describe the
Agency's approach to risk assessment in other areas (carcinogenicity, mutagenicity, developmental
toxicity, exposure, and chemical mixtures.) See the IRIS glossary for a description and citation of each
guideline.
A.Background
Chemicals that give rise to toxic end points other than cancer and gene mutations are often referred
to as "systemic toxicants" because of their effects on the function of various organ systems. It should
be noted, however, that chemicals which cause cancer and gene mutations also commonly evoke
other toxic effects (systemic toxicity). Generally, based on our understanding of homeostatic and
adaptive mechanisms, systemic toxicity is treated as if there is an identifiable exposure threshold
(both for the individual and for the population) below which effects are not observable. This
characteristic distinguishes systemic end points from carcinogenic and mutagenic end points, which
are often treated as nonthreshold processes.
Systemic effects have traditionally been evaluated in terms of concepts such as "acceptable daily
intake" and "margin of safety." The scientific community has identified certain limits on some of
these approaches, and these limits have been borne out in EPA's experience. Nonetheless, EPA is
called upon to apply these concepts in making and explaining decisions about the significance for
human health of certain chemicals in the environment.
To meet these needs, the RfD Work Group has drawn on traditional concepts, as well as on
recommendations in the 1983 National Academy of Sciences (NAS) report on risk assessment, to
more fully articulate the use of noncancer. nonmutagenic experimental data in reaching decisions
on the significance of exposures to chemicals. In the process, the Agency has coined new terminology
to clarify and distinguish between aspects of risk assessment and risk management. EPA has tested
and implemented these innovations in developing consistent information for several recent
regulatory needs, for instance under RCRA.
B.	Overview
This Appendix consists of four parts in addition to this introduction. In Section II, much of the
traditional information on assessing risks of systemic toxicity is presented, with the focus on the
concepts of "acceptable daily intake (ADD" and "safety factor (SF)." Issues associated with these
approaches are identified and discussed.
In Section III, the Agency's approach to assessing the risks of systemic toxicity is presented in the
context of the NAS scheme of risk assessment and risk management in regulatory decision-making.
This approach includes recasting earlier AOI and SF concepts into the less value-laden terms
"reference dose (RfD)" and "uncertainty factor (UF)." A new term, "margin of exposure,"** as
utilized in the EPA regulatory context, is introduced to avoid some of the issues associated with-the
traditional approach.
Section IV examines how these new concepts can be applied in reaching risk management decisions,
while Section V briefly mentions some of the additional approaches the Agency is using and
exploring to address this issue. Section VI provides a sample RfD calculation.
'in this document the term "systemic" refers to an effect other than carcinogenicity or mutagenicity induced by a tome
chemical
••in this Appendix, the ratio of the NOAEL to the estimated exposure (often referred to as "margin of safety*) is referred to
as the "margin of exposure (MOE)" in order to avoid confusion with the original use of the term "margin of safety" m
pharmacology (i e . the ratio of the toxic dose to the theraputic dose) and to avoid the use of the value-laden term "safety *

-------
II. TRADITIONAL APPROACH TO ASSESSING SYSTEMIC (NONCARC1NOGENIC) TOXICITY
The Agency's approach to assessing the risks associated with systemic toxicity is different from that
for the risks associated with carcinogenicity. This is because different mechanisms of action are
thought to be involved in the two cases. In the case of carcinogens, the Agency assumes that a small-
number of molecular events can evoke changes in a single cell that can lead to uncontrolled cellular'
proliferation. This mechanism for carcinogenesis is referred to as "nonthreshold," since there is
essentially no level of exposure for such a chemical that does not pose a small, but finite, probability
of generating a carcinogenic response. In the case of systemic toxicity, organic homeostatic,
compensating, and adaptive mechanisms exist that must be overcome before the toxic end point is
manifested. For example, there could be a large number of cells performing the same or similar
function whose population must be significantly depleted before the effect is seen.
The threshold concept is important in the regulatory context. The individual threshold hypothesis
holds that a range of exposures from zero to some finite value can be tolerated by the organism with
essentially no chance of expression of the toxic effect. Further, it is often prudent to focus on the
most sensitive members of the population; therefore, regulatory efforts are generally made to keep
exposures below the population threshold, which is defined as the lowest of the thresholds of the
individuals within a population
A. The Traditional Approach
In many cases, risk decisions on systemic toxicity have been made by the Agency using the concept of
the "acceptable daily intake (ADl)." This quantity is derived by dividing the appropriate "no-
observed-adverse-effect level (NOAEL)" by a "safety factor (SF)" as follows:*
ADl (human dose) = NOAEL (experimental dose)/SF	(1)
The ADl is often viewed as the amount of a chemical to which one can be exposed on a daily basis
over an extended period of time (usually a lifetime) without suffering a deleterious effect. Often,
the ADl has been used as a tool in reaching risk management decisions; e.g., establishing allowable
jevels of contaminants in foodstuffs and water.
Once the critical study demonstrating the toxic effect of concern has been identified, the selection of
the NOAEL derives from an essentially objective, scientific examination of the data available on the
chemical in question.
Generally, the SF consists of multiples of 10, each factor representing a specific area of uncertainty
inherent in the available data. For example, an SF may be developed by taking into account the
expected differences in responsiveness betweerfhurrians and animals in prolonged exposure studies;
i.e., a 10- fold factor. In addition, a second factor of 10 may be introduced to account for variability
among individuals within the human population. For many chemicals, the resultant SF of 100 has
been judged to be appropriate. For other chemicals, with a less complete data base (e.g., those for
which only the results of subchronic studies are available), an additional factor of 10 (leading to an
SF of 1,000) might be judged to be more appropriate. On the other hand, for some chemicals, based
on well-characterized responses in sensitive humans (e.g.. effect of fluoride on human teeth), an SF
as small as 1 might be selected.
• A NOAEL is an experimentally determined dose at which there was no statistically or biologically significant indication of
the toxic effect of concern in an experiment with several NOAEU. the regulatory focus is normally on the highest one.
leading to the common usage of the term NOAEL as the highest experimentally determined dose without statistical or
adverse biological effect In some treatments, the NOAEL for the critical toxic effect is simply referred to as the NOEL This
latter term, however, invites Ambiguity in that there may be observable effects which are not of toxicologic significance,
i e . they are not 'advene " in order to be explicit, this Appendix uses the term NOAEL and it refers to the highest NOAEL m
an experiment Further, m cases m which a NOAEL has not been demonstrated experimentally, the formulation calls for use
of the "lowest-observed-adverse-effect level (LOAEL)" in order to focus on the maior concepts, however, we will use
NOAEL as a general exampte

-------
While the original selection of SFs appears to have been rather arbitrary (Lehman and Fitzhugh,
1954)*, subsequent analysis of data as reviewed by Dourson and Stara (1983) lends theoretical (and
in some instances experimental) support for their selection. Further, some scientists, but not all,
within the EPA interpret the absence of widespread effects in the exposed human populations as
evidence of the adequacy of the SFs traditionally employed.
B. Some Difficulties in Utilizing the Traditional Approach
1.	Scientific Issues
While the traditional approach has performed well over the years and the Agency has sought to be
consistent in its application, observers have identified scientific shortcomings of the approach.
Examples include the following:
•	By focusing on the NOAEL, information on the shape of the dose-response curve is ignored. Such
data could be important in estimating levels of concern for public safety.
•	As scientific knowledge is increased and the correlation of precursor effects (e.g., enzyme
induction) with frank toxicity becomes known, questions about the selection of the appropriate
"adverse effect" arise.
•	Guidelines have not been developed to take into account the fact that some studies have used
larger numbers of animals and, hence, are generally more reliable than other studies.
These and other "generic issues" are not susceptible to immediate resolution, because the data base
needed is not yet sufficiently developed or analyzed. Therefore, these issues are beyond the scope of
this Appendix. However, the Agency has established a work group to consider them.
2.	Management-related Issues
a. The use of the term "safety /actor"
The term "safety factor" suggests, perhaps inadvertently, the notion of absolute safety, i.e., absence
of risk. While there is a conceptual basis for believing in the existence of a threshold and "absolute
safety" associated with certain chemicals, in the majority of cases a firm experimental basis for this
notion does not exist.
b The implication that any exposure in excess of the ADI is "unacceptable" and that any exposure
less than the ADI is "acceptable" or "safe"
In practice, the ADI is viewed by many as an "acceptable" level of exposure, and, by inference, any
exposure greater than the ADI is seen as "unacceptable." This strict demarcation between what is
"acceptable" and what is "unacceptable" is contrary to the views of most toxicologists, who typically
interpret the ADI as a relatively crude estimate of a level of chronic exposure not likely to result in
adverse effects to humans. The ADI is generally viewed as a "soft" estimate, whose bounds of
uncertainty can span an order of magnitude. That is, within reasonable limits, while exposures
somewhat higher than the ADI are associated with increased probability of adverse effects, that
probability is not a certainty. Similarly, while the ADI is seen as a level at which the probability of
adverse effects is low, the absence of risk to all people cannot be assured at this level.
c. Possible limitations imposed on risk management decisions
Awareness of the "softness" of the ADI estimate (see b. above) argues for careful case-by-case
consideration of the implications of the toxicological analysis as it applies to any particular situation.
To the degree that ADls generated by the traditional approach are the determining factors in risk
'Lehman, AJ and Fitzhugh, 0 G (1954) Association of Food Drug Officials USQ Bulletin 18 33-35

-------
management decisions, they can take on a significance beyond that intended by the toxicoiogist or
merited by the underlying scientific support.
Further, in administering risk/benefit or cost/benefit statutes, the risk manager is required to
consider factors other than risk (e g , estimated exposures compared to the ADl) in reaching a
decision The ADl is only one factor in a management decision and should not prevent the risk
manager from weighing the full range of factors.
d Development of different ADIs by different programs
In addition to occasionally selecting different critical toxic effects. Agency scientists have reflected
their best scientific judgments in the final ADl by adopting factors different from the standard
factors listed in Table A-1. For example, if the toxic end point for a chemical in experimental animals
is the same as that which has been established for a related chemical in humans at similar doses, one
could argue for an SF of less than the traditional 100. On the other hand, if the total toxicologic data
base is incomplete, one could argue that an additional SF should be included, both as a matter of
prudent public policy and as an incentive to others to generate the appropriate data.
Such -jractices, as employed by a number of scientists in different programs, exercising their best
scientific judgment, have in many cases resulted in different ADIs for the same chemical. The fact
that different ADIs were generated (e.g., by adopting different SFs) can be a source of considerable
confusion when the ADIs are applied'in risk management decisionmaking (see c. above). For
example, although they generally agree on the experimental data base for 2,3,7,8-TCDD, regulatory
agencies within the United States and around the world have generated different ADIs by selecting
different "safety factors"; specifically, 1000, 500, 250, and 100. These different ADIs have been used
to justify different regulatory decisions. The existence of different ADIs need not imply that any of
them is more "wrongM--or Mright"--than the rest. It is more nearly a reflection of the honest
difference m scientific judgment.
These differences, which may reflect differences in the interpretation of the scientific data, can also
be characterized as differences in the management of the risk. As a result, scientists may be
inappropriately impugned,and/or perfectly just'fiable risk management decisions may be tainted by
charges of "tampering with the science." This unfortunate state of affairs arises, at least in part,
from treating the ADl as an absolute measure of safety.
III. EPA ASSESSMENT OF RISKS ASSOCIATED WITH SYSTEMIC TOXICITY
In 1983, the National Academy of Sciences published a report* which discusses the conceptual
framework within which regulatory decisions on toxic chemicals are made; see Figure A-1. The
determination of the presence of risk and its potential magnitude is made during the risk assessment
process, which consists of hazard identification, dose-response assessment, exposure assessment, and
risk characterization. Having been apprised by the risk assessor that a potential risk exists, the risk
manager answers the question: "What, if anything, are we going to do about it?"
A. Hazard Identification
1. Evidence
a. Type of effect
Exposure to a given chemical, depending on the dose employed, may result in a variety of toxic
effects. These may range from gross effects, such as death, to more subtle biochemical, physiologic,
or pathologic changes. The risk assessor considers each of the toxic end points from all studies
evaluated in assessing the risk posed by a chemical, although primary attention usually is given to
the effect exhibiting the lowest NOAEL, often referred to as the critical effect. For chemicals with a
limited data base, there may be a need for more toxicity testing.
*NAS Risk Aueument in the Federal Government Mtntging the Proceu (NAS Press, 1983)

-------
FIGUREH -1

Dose-response
Assessment (e.g. RfD)
^ \


Hazard
Idendification
Exposure
Assessment
V
Risk
Characterization
(e.g. criterion)
Regulatory
* Decision
(e.g. RgD, Standard)
//



Control /
Options /
Non-risk /
Analyses
b.	Principal studies
Principal studies are those that contribute most significantly to the qualitative assessment of
whether or not a particular chemical is potentially a systemic toxicant in humans. In addition, they
may be used in the quantitative dose-response assessment phase of the risk assessment. These
studies are of two types:
(1) Human studies
Human data are often useful in qualitatively establishing the presence of an adverse effect in
exposed human populations. Further, when there is information on the exposure level associated
with an appropriate end point, epidemiologic'studies can also provide the basis for a quantitative
dose-response assessment. Use of these latter data avoids the necessity of extrapolating from
animals to humans, and therefore, human studies, when available, are given first priority, with
animal toxicity studies serving to complement them.
In epidemiologic studies, confounding factors that are recognized can be controlled and measured,
within limits. Case reports and acute exposures resulting in severe effects provide support for the
choice of critical toxic effect, but they are often of limited utility in establishing a quantitative
relationship between environmental exposures and anticipated effects. Available human studies on
ingestion are usually of this nature. Cohort studies and clinical studies may contain exposure-
response information that can be used in estimating effect levels, but the method of establishing
exposure must be evaluated for validity and applicability.
(2> Animal studies
Usually, the data base on a given chemical lacks appropriate information on effects in humans. In
such cases, the principal studies are drawn from experiments conducted on non-human mammals,
most often the rat, mouse, rabbit, guinea pig, hamster, dog, or monkey.
c.	Supporting studies
Supporting studies include information from a wide variety of sources. For example, metabolic and
other pharmacokinetic studies can provide insights into the mechanism of action of a particular
compound. By comparing the metabolism of the compound exhibiting the toxic effect in the animal

-------
with the metabolism found in humans, some light may be cast on the potential for the toxic
manifestation in humans or for estimating the equitoxic dose in humans.
Similarly, m vitro studies can provide insights into the compound's potential for biological activity,
although a definite connection to the human experience cannot be drawn. Under certain
circumstances, consideration of structure-activity relationships between the chemical under test and
the effects of structurally related agents can provide a clue to the biological activity of the former.
At the present time, these data are supportive, not definitive, in assessing risk. However, there is
focused activity aimed at developing more reliable in vitro tests to minimize the need for live-animal
testing. Similarly, there is increased emphasis on generating mechanism-of-action and
pharmacokinetic information as a means of increasing the fundamental understanding of toxic
processes in humans and nonhumans. it is expected that in the future these considerations will play a
larger role in our determination of toxicity of chemicals.
d.	Route of exposure
The Agency often approaches the investigation of a chemical with a particular route of exposure in
mind; e.g., an oral exposure for a drinking water contaminant or a residue in food. Although the
route of exposure is oral in both cases, specific considerations may differ For example, the
bioavailability of the chemical administered in food may differ from that when administered in
water or inhaled. Usually, the toxicologic data base on the compound does not include detailed
testing on all possible routes of administration.
In general, it is the Agency's view that the potential for toxicity manifested by one route of exposure
is relevant to any other route of exposure, unless convincing evidence exists to the contrary.
Consideration is always given to potential differences in absorption or metabolism resulting from
different routes of exposure, and whenever appropriate data (e.g., comparative metabolism studies)
are available, the quantitative impacts of these differences on the risk assessment are fully
delineated.
e.	Length of exposure
The Agency is concerned about the potential toxic effects in humans associated with all possible
exposures to chemicals. The magnitude, frequency, and duration of exposure may vary considerably
in different situations. Animal studies are conducted using a variety of exposure durations (e.g.,
acute, subchronic, and chronic) and schedules (e.g., single, intermittent, or continuous dosing)
Information from all of these studies is useful in the hazard identification phase of risk assessment.
For example, overt neurological problems identified in high-dose acute studies tend to reinforce the
observation of subtle neurological changes seen in a low-dose chronic study. Special concern exists
for* low-dose, chronic exposures, however, since such exposures can elicit effects absent in higher-
dose, shorter exposures, through mechanisms such as accumulation of toxicants in the organisms.
f.	Quality of the study
Evaluation of individual studies in humans and animals requires the consideration of several factors
associated with a study's hypothesis, design, execution, and interpretation. An ideal study addresses
a clearly delineated hypothesis, follows a carefully prescribed protocol, and includes sufficient
subsequent analysis to support its conclusions convincingly.
In evaluating the results from such studies, consideration is given to many other factors, including
chemical characterization of the compound(s) under study, the type of test species, similarities and
differences between the test species and humans (e.g., chemical absorption and metabolism), the
number of individuals in the study groups, the number of study groups, the spacing and choice of
dose levels tested, the types of observations and methods of analysis, the nature of pathologic
changes, the alteration in metabolic responses, the sex and age of test animals, and the route and
duration of exposure.
2. Welght-of-Evidence Determination
As the culmination of the hazard identification step, a discussion of the weight-of-evidence
summarizes the highlights of the information gleaned from the entire range of principal and

-------
supporting studies. Emphasis in the analysis is given to examining the results from different studies
to determine the extent to which a consistent, plausible picture of toxicity emerges. For example, the
following factors add to the weight of the evidence that the chemical poses a hazard to humans:
similar results in replicated animal studies by different investigators; similar effects across sex, strain,
species, and route of exposure; clear evidence of a dose-response relationship; a plausible relation
between data on metabolism, postulated mechanism-of-action, and the effect of concern; similar
toxicity exhibited by structurally related compounds; and some link between the chemical and
evidence of the effect of concern in humans. The greater the weight-of-evidence, the greater one's
confidence in the conclusions drawn.
B. Dose-Response Assessment
1.	Concepts and Problems
Empirical observation generally reveals that as the dosage of a toxicant is increased, the toxic
response (in terms of severity and/or incidence of effect) also increases. This dose-response
relationship is well-founded in the theory and practice of toxicology and pharmacology. Such
behavior is observed in the following instances: in quantal responses, in which the proportion of
responding individuals in a population increases with dose; in graded responses, in which the
severity of the toxic response within an individual increases with dose; and in continuous responses,
in which changes in a biological parameter (e.g., body or organ weight) vary with dose.
However, in evaluating a dose-response relationship, certain difficulties arise. For example, one must
decide on the critical end point to measure as the "response.'' One must also decide on the correct
measure of "dose." In addition to the interspecies extrapolation aspects of the question of the
appropriate units for dose, the more fundamental question of administered dose versus absorbed
dose versus target organ dose should be considered. These questions are the subject of much current
research.
2.	Selection of the Critical Data
a. Critical study
Often animal data are selected as the governing information for quantitative risk assessments, since
available human data are generally insufficient for this purpose. These animal studies typically
reflect situations in which exposure to the toxicant has been carefully controlled and the problems of
heterogeneity of the exposed population and concurrent exposures to other toxicants have been
minimized, in evaluating animal data, a series of professional judgments are made that involve,
among others, consideration of the scientific quality of the studies. Presented with data from several
animal studies, the risk assessor first seeks to identify the animal model that is most relevant to
humans, based on the most defensible biological rationale, for instance using comparative
pharmacokinetic data, in the absence of a clearly most relevant species, however, the most sensitive
species (i.e., the species showing a toxic effect at the lowest admininistered dose) is adopted as a
matter of scientific policy at EPA, since no assurance exists that humans are not innately more
sensitive than any species tested. This selection process is made more difficult if animal tests have
been conducted using different routes of exposure, particularly if the routes are different from those
involved in the human situation under investigation.
In any event, the use of data from carefully controlled studies of genetically homogeneous animals
inescapably confronts the risk assessor with the problems of extrapolating between species and the
need to account for human heterogeneity and concurrent human exposures to other chemicals,
which may modify the human risk.
While there is usually a lack of well-controlled cohort studies that investigate non-cancer end points
and human exposure to chemicals of interest, in some cases human data may be selected as the
critical data (e.g., in cases of cholinesterase inhibition). Risk assessments based on human data have
the advantage of avoiding the problems inherent in interspecies extrapolation. In many instances,
use of such studies, as is the case with the animal investigations, involves extrapolation from
relatively high doses (such as those found in occupational settings) to the low doses found in the
environmental situations to which the general population is more likely to be exposed. In some

-------
cases, a well-designed and well-conducted epidemiologic study that shows no association between
known exposures and toxicity can be used to directly project an RfD (as has been done in the case of
fluoride)
b Critical data
in the simplest terms, an experimental exposure level is selected from the critical study that
represents the highest level tested in which "no adverse effect" was demonstrated. This "no-
observed-adverse-effect level" (NOAEL) is the key datum gleaned from the study of the dose-
response relationship and, traditionally, is the primary basis for the scientific evaluation of the risk
posed to humans by systemic toxicants. This approach is based on the assumption that if the critical
toxic effect is prevented, then all toxic effects are prevented.
More formally, the NOAEL is defined in this discussion as the highest experimental dose of a chemical
at which there is no statistically or biologically significant increase in frequency or seventy of an
adverse effect between individuals in an exposed group and those in its appropriate control. (See
also discussion in the footnote on page A-4). As noted above, there may be sound professional
differences of opinion in judging whether or not a particular response is adverse. In addition, the
NOAEL is a function of the size of the population under study. Studies with a small number of
subjects are less likely to detect low-dose effects than studies using larger numbers of subjects. Also,
if the interval between doses in an experiment is large, it is possible that the experimentally
determined NOAEL is lower than that which would be observed in a study using intervening doses
c. Critical end point
A chemical may elicit more than one toxic effect (end point), even in one test animal, or in tests of
the same or different duration (acute, subchronic, and chronic exposure studies). In general, NOAELs
for these effects will differ. The critical end point used in the dose-response assessment is the one at
the lowest NOAEL.
3. Reference Dose (RfD)
In response to many of the problems associated with ADIs and SFs, which were outlined in Section II,
the concept of the "reference dose (RfD)" and "uncertainty factor (UF)" is recommended. The RfD is
a benchmark dose operationally derived from the NOAEL by consistent application of generally
order of magnitude uncertainty factors (UFs) that reflect various types of data used to estimate RfDs
(for example, a valid chronic human NOAEL normally is divided by an UF of 10) and an additional
modifying factor (MF), which is based on a professional judgment of the entire data base of the
chemical.* See Table A-1.	**
The RfO is determined by use of the following equation:
RfD = NOAEL/(UF x MF)	(2)
which is the functional equivalent of Eq. (1). In general, the RfD is an estimate (with uncertainty
spanning perhaps an order of magnitude ) of a daily exposure to the human population (including
sensitive subgroups) that is likely to be without an appreciable risk of deleterious effects during a
lifetime. The RfD is appropriately expressed in units of mg/kg-bw/day.
The RfD is useful as a reference point for gauging the potential effects of other doses. Usually, doses
that are less than the RfD are not likely to be associated with any health risks, and are therefore less
likely to be of regulatory concern. However, as the frequency of exposures exceeding the RfD
increases, and as the size of the excess increases, the probability increases that adverse effects may be
observed in a human population. Nonetheless, a clear conclusion cannot be categorically drawn that
all doses below the RfD are "acceptable" and that all doses in excess of the RfD are "unacceptable."
"uncertainty factor" is the new description applied to the term 'safety factor" (see Page -4) This new name is more
descriptive in that these factors represent scientific uncertainties, and avoids the risk management connotation of 'safety "
The "modifying factor" can range from greater than zero to 10. and reflects qualitative professional judgements regarding
scientific uncertainties not covered under the standard UF. such as the completeness of the overall data base and the
number of animals in the study

-------
TABLE H-1.
GUIDELINES FOR THE USE OF UNCERTAINTY FACTORS IN DERIVING REFERENCE DOSE (RfD)
Standard Uncertainty Factors (UFs)
Use a 10-fold factor when extrapolating from valid experimental results from studies, using
prolonged exposure to average healthy humans. This factor is intended to account for the
variation in sensitivity among the members of the human population. [10H]
Use an additional 10-fold factor when extrapolating from valid results of long-term studies on
experimental animals when results of studies of human exposure are not available or are
inadequate. This factor is intended to account for the uncertainty in extrapolating animal data to
the case of humans. [10A]
Use an additional 10-fold factor when extrapolating from less than chronic results on experimental
animals when there are no useful long-term human data. This factor is intended to account for the
uncertainty in extrapolating from less than chronic NOAELsto chronic NOAELs. 110S]
Use an additional 10-fold factor when deriving a RfD from a LOAEL. instead of a NOAEL. This factor
is intended to account for the uncertainty in extrapolating from LOAELs to NOAELs. [10L]
Modifying Factor (MF)
Use professional judgment to determine another uncertainty factor (MF) which is greater than
zero and less than or equal to 10. The magnitude of the MF depends upon the professional
assessment of scientific uncertainties of the study and database not explicitly treated above; e.g.,
the completeness of the overall data base and the number of species tested. The default value for
the MF is 1.
SOURCE: Adapted from Dourson, M L.; and Stara, J.F. (1983) Regulatory Toxicology and
Pharmacology 3:224-238.
(This is a consequence of the inability of either the traditional or the RfD approach to completely
address the question of dose-response extrapolation.)
The Agency is attempting to standardize its approach to determining RfDs. The RfD Work Group has
developed a systematic approach to summarizing its evaluations, conclusions, and reservations
regarding RfDs in a "cover sheet" of a few pages in length. The cover sheet includes a statement on
the confidence the evaluators have in the stability of the RfD: high, medium, or low. High
confidence indicates that the RfD is unlikely to change in the future because there is consistency
am&ng the toxic responses observed in different sexes, species, study designs, or in dose-response
relationships, or the reasons for differences, if any, are well understood. Often, high confidence is
given to RfDs that are based on human data for the exposure route of concern, because in such cases
the problems of interspecies extrapolation are avoided. Low confidence indicates that the RfD may
be especially vulnerable to change if additional chronic toxicity data are published on the chemical,
because the data supporting the estimation of the RfD are of limited quality and/or quantity.
C. Exposure Assessment
The third step in the risk assessment process focuses on exposure issues. For a full discussion of
exposure assessment, the reader is referred to EPA's recently published guidelines on the subject (SI
Federal Register 34042-340S4, Sept. 24, 1986). There is no substantive difference in the conceptual
approach to exposure assessment in the case of systemic toxicants and of carcinogens.
In brief, the exposure assessment includes consideration of the populations exposed and the
magnitude, frequency, duration and routes of exposure, as well as evaluation of the nature of the
exposed populations.

-------
D. Risk Characterization
Risk characterization is the final step in the risk assessment process and the first step in the risk
management process. Its purpose is to present to the risk manager a synopsis and synthesis of all the
data that contribute to a conclusion on the risk, including:
•	The qualitative ("weight-of-evidence") conclusions about the likelihood that the chemical may
pose a hazard to human health.
•	A discussion of dose-response and how this information, through the use of particular uncertainty
and modifying factors, was used to determine the RfD.
•	Data such as the shapes and slopes of the dose-response curves for the various toxic end points,
toxicodynamics (absorption and metabolism), structure-activity correlations, and the nature and
severity of the observed effect. These data should be clearly discussed by the risk assessor, since
they may influence the final decision of the risk manager (see below).
•	The estimates of exposure, the nature of the exposure, and the number and types of people
exposed, together with a discussion of the uncertainties involved.
•	A discussion of the sources of uncertainty, major assumptions, areas of scientific judgment, and,
to the extent possible, estimates of the uncertainties embodied in the assessment.
In the risk characterization process, comparison is made between the RfD and the estimated
(calculated or measured) exposure dose (EED), which should consider exposure by all sources and
routes of exposure. The risk assessment should contain a discussion of the assumptions underlying
the estimation of the RfD (nature of the critical end point, nature of other toxic end points, degree
of confidence in the data base, etc.), and the degree of conservatism in its derivation. The
assumptions used to derive the EED should also be discussed. If the EED is less than the RfD, the need
for regulatory concern is likely to be small.
An alternative measure that may be useful to some risk managers is the "margin of exposure (MOE)"
(see footnote on p. A-3), which is the magnitude by which the NOAEL of the critical toxic effect
exceeds the estimated exposure dose (EED), where both are expressed in the.same units:
MOE = NOAEL (experimental dose)/EED (human dose)	(3)
In parallel to the statements above on EED and RfD, the risk assessment should contain a discussion
of the assumptions underlying the estimates of the RfD and the degree of possible conservatism of
the UF and MF. It can be noted that when the MOE is equal to or greater than UF x MF, the need for
regulatory concern is likely to be small.
Section VI contains an example of the use of the concepts of NOAEL, UF, MF, RfD, and MOE.
IV. APPLICATION IN RISK MANAGEMENT
Once the risk characterization is completed, the focus turns to risk management. In reaching
decisions, the risk manager must consider a number of risk factors, nonrisk factors, and regulatory
options that influence the final judgment. It is generally useful to the risk manager to have
information regarding the contribution to the RfD from various environmental media. Such
information can provide insights that are helpful in choosing among available control options.
However, in cases in which site-specific criteria are being considered, local exposures through various
media can often be determined more accurately than exposure estimates based upon generic
approaches. In such cases, the exposure assessor's role is particularly important. For instance, at a
given site, consumption of fish may clearly dominate the local exposure routes, while, on a national
basis, fish consumption may play a minor role compared to ingestion of treated crops.
RfDs should be apportioned by route of exposure. Where specific exposure analysis can be made,
such apportionment is readily performed. If exposure information is not available, assumptions must
be made concerning the relative contributions from different routes of exposure. At present,
different EPA offices use assumptions that differ to some degree. These assumptions are being
reviewed by an Agency risk assessment group.

-------
As illusirated in Figure A-1, the risk manager utilizes the results of risk characterization, other
technological factors, and nontechnical social and economic considerations in reaching a regulatory
decision. Some of these factors include efficiency,-timeliness, equity, administrative simplicity,
consistency, public acceptability, technological feasibility, and legislative mandate.
Because of the way these risk management factors may impact different cases, consistent-but >^ot
necessarily identical-risk management decisions must be made on a case-by-case basis. For example,
the Clean Water Act calls for decisions with "an ample margin of safety"; the Federal Insecticide,
Fungicide and Rodenticide Act (FIFRA) calls for "an ample margin of safety," taking benefits into
account; and the Safe Drinking Water Act (SOWA) calls for standards that protect the public "to the
extent feasible." Consequently, it is entirely possible and appropriate that a chemical with a specific
RfD may be regulated under different statutes and situations through the use of different
"regulatory doses (RgDs)".
Expressed in general terms, after carefully considering the various risk and nonrisk factors,
regulatory options, and statutory mandates in a given case (i), the risk manager decides upon the
appropriate statutory alternatives to arrive at an "ample" or "adequate" margin of exposure
[MOE(i)], thereby establishing the regulatory dose, RgD(i) (e.g., a tolerance under FIFRA or a
maximum contaminant level under SDWA), applicable to that case:
RgD(i) = NOAEl/MOE(i)	(4)
Note that, for the same chemical (with a single RfD), the risk manager(s) can develop different
regulatory doses for different situations that may involve different exposures, available control
options, alternative chemicals, benefits, and statutory mandates. Also note that comparing the RfD
to a particular RgD(i) is equivalent to comparing the MOE(i) with the UF x MF:
RfD/RgD(i) = MOE(i)/UFxMF	(5)
In assessing the significance of a case in which the RgD is greater (or less) than the RfD, the risk
manager should carefully consider the case-specific data laid out by the risk assessors, as discussed in
in Section III. D. 4. In some cases this may require additional explanation and insight from the risk
assessor, in any event, the risk manager has the responsibility to clearly articulate the reasoning
leading to the final RgD decision.
V. OTHER DIRECTIONS
While the Agency is in the process of systematizing the approach outlined in this Appendix, risk
assessment research for systemic toxicity is also being conducted along entirely separate lines. For
example, the Office of Air Quality Ptanning'and Standards is using probabilistic risk assessment
procedures for criteria pollutants. This procedure characterizes the population at risk, and the
likelihood of various effects occurring, through the use of available scientific literature and
elicitation of expert judgment concerning dose-response relationships. The dose-response
information is combined with exposure analysis modeling to generate population risk estimates for
alternative standards. These procedures present the decisionmaker with ranges of risk estimates, and
explicitly consider the uncertainties associated with both the toxicity and exposure information. The
Office of Policy, Planning, and Evaluation is investigating similar procedures in order to balance
health risk and cost. In addition, scientists in the Office of Research and Development have initiated
a series of studies that should lead to future improvements in risk estimation. First, they are
investigating the use of extrapolation models as well as the statistical variability of the NOAEL and
underlying UFs as means of estimating RfDs. Second, they are exploring procedures for less-than-
lifetime health risk assessment. Finally, they are working on ranking the severity of toxic effects as a
way to further refine EPA's health risk assessments. While these procedures are promising, they
cannot be expected at this time to serve as a foundation of a generalized health risk assessment for
systemic toxicity in the Agency.

-------
VI. HYPOTHETICAL, SIMPLIFIED EXAMPLE OF DETERMINING AND USING RfD
Suppose the Agency had a sound 90-day subchronic gavage study in rats with the following data:
A. Experimental Results
Dose
(mg/kg-day)
0
1
25
Observation	Effect Level
Control - no adverse effects observed
No statistical or biological significant	NOEL
differences between treated and control animals
2% decrease* in body weight gain (not	NOAEL
considered to be of biological significance)
increased ratio of liver weight to body weight
Histopathology indistinguishable from controls
Elevated liver enzyme levels
20% decrease* in body weight gain	LOAEL
Increased* ratio of liver weight to body weight
Enlarged, fatty liver with vacuole formation
Increased* liver enzyme levels
* = Statistically significant compared to controls.
B. Analysis
7. Determination of the Reference Dose (RfD)
a.	From the NOAEL
UF = 10H x 10A x 10S = 1000
MF = 0.8, a subjective adjustment based on the fact that the experiment involved an
astonishing 250 animals per dos« group.
Therefore UF x MF = 800, so that
RfD = NOAEU(UFxMF) = 5 mg/kg-day / 800 = 0.006 mg/kg-day
b.	From the LOAEL (i.e., if a NOAEL is not available)
if 25 mg/kg-day had been the lowest dose tested,
UF = 10Hx 10Ax lOSx 10L = 10,000
MF = 0.8
Therefore UF x MF = 8,000, so that
RfD = LOAEU(UFxMF) = 25 (mg/kg-day) / 8000 = .003 mg/kg-day)
2. Risk Characterization Considerations
Suppose the estimated exposure dose (EED) for humans exposed to the chemical under the
proposed use pattern were .01 mg/kg-day; i e ,
EED > RfD

-------
Viewed alternatively, the MOE is:
MOE = NOAEL/EED = 5 mg/kg-day / 0.01 mg/kg-day = 500
Because the EEO exceeds the RfD (and the MOE is less than the UF x MF), the risk manager will
need to look carefully at the data set, the assumptions for both the RfD and the exposure
estimates, and the comments of the risk assessors, in addition, the risk manager will need to weigh
the benefits associated with the case, and other nonrisk factors, in reaching a decision on the
regulatory dose (RgD)

-------