United States Office of Water
Environmental Protection Washington, DC
TECHNICAL SUPPORT DOCUMENT FOR
WATER QUALITY-BASED TOXBCS CONTROL
This copy represents the second printing of this document.
Changes made to this document reflect corrections of typographical errors and the following update
of the interim guidance on criteria for metals: The Agency has issued "Interim Guidance Interpretation
and Implementation Aquatic Life Criteria for Metals." The interim guidance supersedes criteria document
statements expressing criteria in terms of a acid soluble analytical method and also the metals
discussion of Section 5.7.3. The availability of this document appeared in the
June 5, 1992 Federal Register (Vol. 57, No. 7 09, pg. 24401).
March 1991
Office of Water Enforcement and Permits
Office of Water Regulations and Standards
U.S. Environmental Protection Agency
Washington, DC 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 decisionmaking 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 and policy initiatives involving the
application of biological and chemical assessment techniques to control toxic pollution to surface waters. This
document is agency guidance only. It does not establish or affect legal rights or obligations. It does not establish
a binding norm and is not finally determinative of the issues addressed. Agency decisions in any particular case
will be made applying the law and regulations on the basis of specific facts when permits are issued or regulations
promulgated.
This document is expected to be revised periodically to reflect advances in this rapidly evolving area. Comments
from users will be welcomed. Send comments to U.S. EPA, Office of Water Enforcement and Permits, 401 M
Street, SW, Mailcode EN366, Washington, DC 20460.
James R. Elder, Director Martha G. Prothro, Director
Office of Water Enforcement and Permits Office of Water Regulations and Standards
-------
-------
TABLE OF CONTENTS
Section Page
Foreword iii
Acknowledgment xiii
Executive Summary xiv
List of Abbreviations xvii
Glossary xix
Introduction xxiii
1. APPROACHES TO WATER QUALITY-BASED TOXICS CONTROL 1
1.1 INTRODUCTION 1
1.2 CHEMICAL-SPECIFIC APPROACH FOR AQUATIC
LIFE PROTECTION 1
1.2.1 Correlation of Chemical-specific Measurements to Actual
Receiving Water Impacts 2
1.2.2 Chemical-Specific Analytical Method Precision 2
1.3 WHOLE EFFLUENT APPROACH FOR AQUATIC
LIFE PROTECTION 4
1.3.1 Toxic Units 6
1.3.2 Correlation of Whole Effluent Toxicity Measurements to
Actual Receiving Water Impact 6
1.3.3 Toxicity Test Method Precision 11
1.3.4 Considerations Involved When Implementing the Whole
Effluent Toxicity Approach 11
1.4 BIOLOGICAL CR1TERIA/BIOASSESSMENT AND BIOSURVEY
APPROACH FOR AQUATIC LIFE PROTECTION 18
1.4.1 Use of Biosurveys and Bioassessments in Water Quality-
based Toxics Control , 18
1.4.2 Conducting Biosurveys 19
1.5 INTEGRATION OF THE WHOLE EFFLUENT, CHEMICAL-
SPECIFIC, AND BIOASSESSMENT APPROACHES 20
1.5.1 Capabilities and Limitations of the Chemical-Specific Approach 20
1.5.2 Capabilities and Limitations of the Whole Effluent Approach 21
1.5.3 Capabilities and Limitations of the Bioassessment Approach 22
-------
Table of Contents (Continued)
Section Page
1.6 OTHER FACTORS INFLUENCING WATER QUALITY-BASED
2.
1 .6.1 Persistence i
1 .6.2 Additivity, Antagonism, and Synergism
1 .6.3 Test Interferences
1 .7 HUMAN HEALTH PROTECTION
1.7.1 Types of Health Effects
REFERENCES
WATER QUALITY CRITERIA AND STANDARDS
2.1 INTRODUCTION :
2.1 .1 Overview of Water Quality Standards ;
2.1 .2 Water Quality Standards and State Toxics Control Programs
2.2 GENERAL CONSIDERATIONS ;
2.2.1 Magnitude, Duration, and Frequency
2.2.2 Mixing Zones 4.
2.3 WATER QUALITY CRITERIA FOR AQUATIC LIFE PROTECTION
2.3.1 Development Process for Criteria .!
2.3.2 Magnitude for Single Chemicals
233 Magnitude for Whole Effluent Toxicity
2.3.4 Duration for Single Chemicals and Whole Effluent Toxicity
2.3.5 Frequency for Single Chemicals and Whole Effluent Toxicity .....
2.4 WATER QUALITY CRITERIA FOR HUMAN HEALTH PROTECTION
.
• r . •
2 4.1 Overview
2.4.2 Magnitude and Duration
2.4.3 Human Exposure Considerations
2.4.5 Bioaccumulation Considerations for Reference Ambient
2.4.6 Updating Human Health Criteria and Generating RACs
Using IRIS
2.4.7 Calculating RACs for Non-carcinogens
2.4.8 Calculating RACs for Carcinogens
2.4.9 Deriving Quantitative Risk Assessments in the Absence
of IRIS Values
2.4.1 0 Deriving Reference Tissue Concentrations for Monitoring
Fish Tissue
. .'. ; . 23
24
24
24
.25
! ..26
29
: 29
. 29
30
.'. 31
. 31
33
. 34
34
34
. 35
35
. , 36
'.'.'.' i 36
36
. 37
. . . . ! 37
. 37
38
38
39
40
40
41
VI
-------
Table of Contents (Continued)
Section Page
2.5. BIOLOGICAL CRITERIA . . 41
2.5.1 Regulatory Bases for Biocriteria 41
2.5.2 Development and Implementation of Biocriteria .41
2.6. SEDIMENT CRITERIA 42
2.6.1 Current Developments in Sediment Criteria '......'• 42
2.6.2 Approach to Sediment Criteria Development 42
2.6.3 Application of Sediment Criteria ...;.., 43
2.6.4 Sediment Criteria Status 43
REFERENCES 45
3. EFFLUENT CHARACTERIZATION 47
3.1 INTRODUCTION 47
3.1.1 NPDES Regulation Requirements 47
3.1.2 Background for Toxic Effects Assessments on Aquatic Life
and Human Health 48
3.1.3 General Considerations in Effluent Characterization 49
3.2 DETERMINING THE NEED FOR PERMIT LIMITS WITHOUT
EFFLUENT MONITORING DATA FOR A SPECIFIC FACILITY 50
3.3 DETERMINING THE NEED FOR PERMIT LIMITS WITH
EFFLUENT MONITORING DATA 51
3.3.1 General Considerations .51
3.3.2 Addressing Uncertainty in Effluent Characterization by
Generating Effluent Monitoring Data ..:....... .". . 52
3.3.3 Effluent Characterization for Whole Effluent Toxicity ."53
3.3.4 Use of Toxicity Testing in Multiple-source Discharge
Situations - 59
3.3.5 Ambient Toxicity Testing 61
3.3.6 Special Considerations for Discharges to Marine and
Estuarine Environments .61
3.3.7 Using a Chemical-specific Limit to Control Toxicity 61
3.3:8 Effluent Characterization for Specific Chemicals 62
3.3.9 Effluent Characterization for Bioconcentratable Pollutants .. 64
3.3.10 Analytical Considerations for Chemical 65
REFERENCES 66
VII
-------
Table of Contents (Continued)
Section ; Page
4. EXPOSURE AND WASTELOAD ALLOCATION ........... J 67
I
4.1 INTRODUCTION 67
4.2 TOTAL MAXIMUM DAILY LOADS AND WASTELOAD
ALLOCATIONS J . ". •. , 67
4.2.1 Total Maximum Daily Loads ..,......,....,...,. 67
4.2.2 Wasteload Allocation Schemes f. 69
4.3 INCOMPLETELY MIXED, DISCHARGE RECEIVING WATER SITUATIONS .69
4.3.1 Determination of Mixing Zone Boundaries 70
4.3.2 Mimimizing the Size of Mixing Zones 71
4.3.3 Prevention of Lethality to Passing Organisms 71
4.3.4 Prevention of Bioaccumulation Problems for Human Health 72
4.4 MIXING ZONE ANALYSES ;..... : 72
4.4.1 General Recommendations for Outfall Design
4.4.2 Critical Design Periods for Waterbodies ....
4.4.3 General Recommendations for Tracer Studies
4.4.4 Discharge-induced Mixing
4.4.5 Ambient-induced Mixing
4.5 COMPLETELY MIXED DISCHARGE-RECEIVING WATER
SITUATIONS
4.5.1 Wasteload Modeling Techniques
4.5.2 Calculating the Allowable Effluent Concentration Distribution
and the Return Period
4.5.3 General Recommendations for Model Selection .. 83
83
85
4.5.4 Specific Model Recommendations
4.5.5 Effluent Toxicity Modeling
73
73
74
75
77
78
78
82
4.6 HUMAN HEALTH 87
4.6.1 Human Health Considerations ..;.;.. ;. 87
4.6.2 Determining the TMDL Based on Human Health Toxicants 87
REFERENCES .. 90
5. PERMIT REQUIREMENTS ..... ..93
5.1 INTRODUCTION ,: . . . r 93
5.1.1 Regulatory Requirements 93
VIII
-------
Table of Contents (Continued)
Section Page
5.2 BASIC PRINCIPLES OF EFFLUENT VARIABILITY . . 93
5.2.1 Variations in Effluent Quality 93
5.2.2 Statistical Parameters and Relationship to Permit Limits 95
5.2.3 Expression of Permit Limits 96
5.3 ENSURING CONSISTENCY WITH THE WASTELOAD
ALLOCATION 96
5.3.1 Statistical Considerations of WLAs ..'...- 96
5.3.2 Types of Water Quality Models and Model Outputs 96
5.4 PERMIT LIMIT DERIVATION 98
5.4.1 EPA Recommendations for Permitting for Aquatic Life
Protection 98
5.4.2 Other Approaches to Permitting for Aquatic Life ..103
5.4.3 Special Permitting Requirements 104
5.4.4 EPA Recommendations for Permitting for Human Health
Protection 104
5.5 SPECIAL CONSIDERATIONS IN USE OF STATISTICAL PERMIT
LIMIT DERIVATION TECHNIQUES 105
5.5.1 Effect of Changes of Statistical Parameters on Permit Limits 105
5.5.2 Coefficient of Variation 106
5.5.3 Number of Samples 107
5.5.4 Probability Basis 110
5.6 PERMIT DOCUMENTATION 110
5.7 EXPRESSING LIMITS AND DEVELOPING MONITORING
REQUIREMENTS 110
5.7.1 Mass-based Effluent Limits 110
5.7.2 Energy Conservation Ill
5.7.3 Considerations in the Use of Chemical specific Limits 111
5.7.4 Considerations in the Use of Whole Effluent Toxicity Limits 112
5.7.5 Selection of Monitoring Frequencies ..113
5.7.6 Analytical Variability 113
5.7.7 Antibacksliding ..113
5.8 TOXICITY REDUCTION EVALUATIONS 114
5.8.1 TRE Guidance Documents 114
5.8.2 Recommended Approach for Conducting TREs 114
5.8.3 Circumstances Warranting a TRE 117
5.8.4 Mechanisms for Receiving TREs 118
REFERENCES 121
IX
-------
Table of Contents (Continued)
Section , Page
6. COMPLIANCE MONITORING AND ENFORCEMENT 123
6.1 INTRODUCTION . . I'. 123
6.2 PERMIT REQUIREMENTS • 123
6.3 COMPLIANCE MONITORING . 123
6.3.1 Self-monitoring Reports 123
6.3.2 Discharge Monitoring Reports/Quality Assurance 124
6.3.3 Inspections \ 124
6.4 VIOLATION REVIEW i 124
6.5 ENFORCEMENT 125
6.6 REPORTING OFVIOLATIONS j 126
i
I
REFERENCES .127
7. CASE EXAMPLES .. .129
7.1 INTRODUCTION I 129
7.2 CASE 1: INDUSTRIAL DISCHARGE i . i'. '. 129
7.2.1 General Site Description and Information .... [ 129
7.2.2 Effluent Characterization for Specific Chemicals 129
7.2.3 Effluent Characterization for Whole Effluent Toxicity 131
7.2.4 Determine Wasteload Allocations 132
7.2.5 Develop Permit Limits 132
7.2.6 Determining and Expressing the Controlling Effluent Limit 133
7.2.7 Comparing Different Limit Development Methods '133
7.3 CASE 2: POTW DISCHARGE L 134
7.3.1 General Site Description and Information .... i 134
7.3.2 Effluent Characterization for Specific Chemicals! 134
7.3.3 Effluent Characterization for Whole Effluent Toxicity 136
7.3.4 Determine Wasteload Allocations i 136
7.3.5 Develop Permit Limits 136
7.3.6 Determining and Expressing the Controlling Effluent Limits 137
7.3.7 Comparing Different Limit Development Methods 137
7.4 CASE 3: MULTIPLE DISCHARGERS INTO THE SAME REACH 137
i
7.4.1 Effluent Characterization : 137
7.4.2 TMDLs and WLAs 138
7.3.3 Permit Limit Development 139
INDEX 140
-------
APPENDICES
Section
Page
Appendix A-l: Toxicity Test Precision Data .- A-1
Appendix A-2: Effluent Variability Data , , A-2-1
Appendix A-3: Acute to Chronic Ratio Data A-3-1
Appendix B-1: Summary of Clean Water Act Provisions B-1 -1
Appendix B-2: Policies for Toxics Control ; B-2-1
Appendix B-3: Regulations for Toxics Control B-3-1
Appendix B-4: Whole-Effluent Toxicity Permitting Principles and Enforcement
Strategy .,. B-4-1
Appendix B-5: Quality Control Fact Sheets B-5-1
Appendix B-6: Case Decisions on Whole-Effluent Toxicity B-6-1
Appendix C: Ambient Toxicity Testing and Data Analysis C-1
Appendix D: Duration and Frequency D-1
Appendix E: Lognormal Distribution and Permit Limit Derivations E-1
Appendix F: Sampling ;...... F-1
Appendix G: The Development of a Biological Indicator Approach to Water
Quality-based Human Health Toxics Control . G-l
Appendix H: Reference Dose (RfD): Description and Use in Health Risk
Assessments H-T
Appendix I: Chemicals Available in IRIS '. 1-1
XI
-------
-------
ACKNOWLEDGMENT
The preparation of the revised .Technical Support Document for Water
Quality-based Toxic Control began with a 3-day conference held in
Williamsburg, Virginia, in December 1988. Representatives of
EPA Headquarters and Regions, States, private industry, munici-
palities, academia, and various interest groups attended this
meeting and provided valuable input. The principal authors of
this document were Bill Swietlik, James Taft, Jacqueline Romney,
Kathryn Smith, John Cannell, Robert Wood, James Pendergast,
and Rick Brandes of the Permits Division; Sheila Frace and Margarete
Heber of the Enforcement Division, Elizabeth Southerland and
Richard Healy of the Assessment and Watershed Protection Divi-
sion; and Charles Delos, Warren Banks, and Robert April of the
Criteria and Standards Division. Listed below are the contributors
to specific chapters of this document.
Approaches to Water Quality-Based Toxics Control
Margaret Heber, U.S. EPA, Enforcement Division
Kathryn Smith, U.S. EPA, Permits Division
Water Quality Criteria and Standards
Charles Delos, U.S. EPA, Criteria and Standards Division
Warren Banks, U.S. EPA, Criteria and Standards Division
Kathy Barylski, U.S. EPA, Criteria and Standards Division
Robert April, U.S. EPA, Criteria and Standards Division
David Moon, U.S. EPA, Criteria and Standards Division
Jacqueline Romney, U.S. EPA, Permits Division
Effluent Characterization
Robert Wood, U.S. EPA, Permits Division
Bill Swietlik, U.S. EPA, Permits Division
James Pendergast, U.S. EPA, Permits Division
Bill Swietlik, U.S. EPA, Permits Division
James Pendergast, U.S. EPA, Permits Division
John Cannell, U.S. EPA, Permits Division
Compliance Monitoring and Enforcement
Sheila Frace, U.S. EPA, Enforcement Division
Theodore Coopwood, U.S. EPA, Enforcement Division
Human Health Component of AH Chapters
John Cannell, U.S. EPA, Permits Division
Katherine Dowell, U.S. EPA, Permits Division
William Morrow, U.S. EPA, Permits Division
Case Examples Workgroup
Bill Swietlik, U.S. EPA, Permits Division
James Pendergast, U.S. EPA, Permits Division
Charles Delos, U.S. EPA, Criteria and Standards Division
Jacqueline Romney, U.S. EPA, Permits Division
Appendices
Appendix A: Margaret Heber, U.S. EPA, Enforcement Division
Appendix B: U.S. EPA, Permits Division
Appendix C: U.S. EPA, Permits Division
Appendix D: Nelson Thomas, U.S. EPA, ERL/ORD, Duluth, MN
Appendix E: Henry Kahn and Maria Smith, U.S. EPA, Analysis
and Evaluation Division
Appendix F: U.S. EPA, Permits Division
Appendix G: U.S. EPA, Permits Division
Appendix H: U.S. EPA, RfD Workgroup
Exposure and Wasteload Allocation
Elizabeth Southerland, U.S. EPA,
Protection
Richard Healy, U.S. EPA, Assessment and Watershed Protection
Elizabeth Southerland, U.S. EPA, Assessment and Watershed
Protection
XIII
-------
EXECUTIVE SUMMARY
The revised Technical Support Document for Water Quality-based
Toxics Control (TSD) provides States and Regions with guidance
on procedures for use in the water quality-based control of toxic
pollutants. It presents recommendations to regulatory authorities
faced with the task of controlling the point source 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 standards development to compliance monitoring.
Both human health and aquatic toxicity issues are incorporated
fnto the discussions throughout the document. The overall ap-
proach in this revised document provides additional explanations
and rationales based on accumulated experience and data for the
various recommendations that 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
The Environmental Protection Agency's (EPA) surface toxics con-
trol 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 consists
of whole effluent and chemical-specific approaches as a means of
protecting aquatic life and human health. As techniques are
made available for implementing biocriteria, they too should be
integrated into the water quality-based toxics control, thus creat-
ing a triad of approaches: whole effluent, chemical-specific, and
biological assessments. Each approach has its limitations and
thus, exclusive use of one approach alone cannot ensure required
protection of aquatic life and human health. The advantages/
disadvantages of each approach and how the integrated ap-
proach creates an effective toxics control program are discussed
In the text.
The whole effluent approach to toxics control involves the use of
toxicity tests and water quality criteria for the parameter "toxic-
ity" to assess and control the aggregate toxicity of effluents. New
references and information in support of the whole effluent toxic-
ity assessment and control approach have been included in Chap-
ter 1 and associated appendices (e.g., precision data, justifications
for acute-to-chronic ratio recommendations, information on ana-
lytical variability in toxicity testing). The chemical-specific approach
to aquatic life toxics control relies on numeric water quality
criteria in State standards and interpretations of State narrative
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 be
developed using data from the Integrated Risk Information Sys-
tem, 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 pollut-
ant, 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 criterion continuous concentration is applied
as a 4-day average concentration; and the criterion maximum
concentration is applied as an 1 -hour average concentration. The
frequency with which criteria are allowed to be exceeded de-
pends on site-specific factors as explained in the text.
j - - . - .. , . ': :..•:
Strictly speaking the term "criteria" means EPA guidance formally
published under the authority of Section 304(a) of the Clean
Water Act. The toxicity level recommendations have not been so
published. However, they represent EPA's carefully developed
technical recommendation, and so are referred to in this docu-
ment in the same manner as other criteria.
!•
EPA's recommended criteria for whole effluent toxicity are as
follows: to protect aquatic life against chronic effects, the ambi-
ent itoXicity should not exceed 1.0 chronic toxic unit (TUJ to the
most sensitive of at least three different test species. For protec-
tion against acute effects, the ambient toxicity should not exceed
0.3 jacute toxic units (TUa) to the most sensitive of at least three
different test species. '
[
EPA has developed recommended human health criteria, which
are tailed reference ambient concentrations (RACs). In the ab-
sence of EPA's recommended criteria, States may calculate RACs
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 proce-
dures (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 causes, has the
reasonable potential to cause, or contributes to an excursion
above a State water quality standard. These effluent characteriza-
tion procedures can be performed in one step and do not include
initial screening followed by definitive data generation as was
recommended in the original TSD.
i.
The revised effluent characterization procedures for assessing po-
tential human health impacts now include control of
bioaccumulative chemicals.
i'
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 analysis of how much of the waterbody is
subject to the exceedance of criteria, for how long, and how
frequently. The first step is to evaluate the effluent plume disper-
sion. If mixing is not rapid and complete and if State standards
allow a mixing zone, the wasteload allocation also must be based
XIV
-------
on a mixing zone analysis. Chapter 5 describes the means to
assess dilution at the edge of a mixing zone. As with the original
TSD, ambient criteria to control acute toxicity to aquatic life may
be pnet within a short distance of the outfall. However, this
provision is no longer restricted to outfalls that 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 assume that the
effluent concentration is constant and that the duration and
frequency with which criteria are exceeded can be reflected en-
tirely by selecting a design flow in the receiving water of appropriate
averaging period and frequency.
Another means of modeling exposure is to use computer models
that incorporate variability of the individual inputs (such as efflu-
ent flow and concentration, receiving water flow, temperature,
background concentration, etc.). These models are termed dy-
namic 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 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 trans-
lated into a permit limit fn 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 documenta-
tion and how to express limitations are discussed. In addition,
guidance for requiring and conducting toxicity reduction evalua-
tions is presented.
Compliance Monitoring
The compliance monitoring and enforcement process for water
quality-based permits summarized in Chapter 6 is based on exist-
ing 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.
xv
-------
-------
LIST OF ABBREVIATIONS
AA atomic absorption
ACR acute-to-chronic ratio
ADI acceptable daily intake
AML average monthly limit
ATC acceptable tissue concentration
ATE acute toxicity endpoint
AVS acid volatile sulfides
BAF bioaccumulation factor
BAT best available technology
BCF bioconcentration factor
BCT best conventional technology
BMP best management practice
BOD biochemical oxygen demand
BPJ best professional judgment
BPT best practicable technology
CCC criteria continuous concentration
CEAM Center for Exposure Assessment Modeling (EPA)
CETTP Complex Effluent Toxicity Testing Program
CFR Code of Federal Regulations
CHC chemical of highest concern
CMC criteria maximum concentration
CTE chronic toxicity endpoint
CV coefficient of variation
CWA Clean Water Act
DF dilution factor
DMR discharge monitoring report
DO dissolved oxygen
EC effect concentration
ECAO Environmental Criteria and Assessment Office
EMS Enforcement Management System
EP equilibrium partitioning
EPA Environmental Protection Agency
ERL Environmental Research Laboratory (EPA)
FAV final acute value
FDA Food and Drug Administration
FM food chain multipliers
GC/MS gas chromatograph/mass spectrometer
HHC human health criteria
HPLC high-pressure liquid chrpmatography
1C inhibition concentration
IRIS Integrated Risk Information System (EPA)
LA load allocation
LC lethal concentration
LOAEL lowest observed adverse effect level
LOEC lowest observed effect concentration
LTA long-term average
M0- maximum contaminant levels
MDL maximum daily limit
MERS Monticello Ecological Research Station
ML minimum level
NOAEL no observed adverse effect level
NOEC no observed effect concentration
NPDES National Pollutant Discharge Elimination System
NTIS National Technical Information Service
ONRW outstanding national resource waters
PCS Permit Compliance System
POTW publicly owned treatment works
PQL practical quantitation limit
ql* cancer potency factor
QA/QC quality assurance/quality control
QNCR quarterly noncompliance report
QSAR quantitative structure-activity relationships
RAC reference ambient concentration
Rf D reference dose
RWC receiving water concentration
SQC sediment quality criteria
STORET storage and retrieval of water quality information
TIE toxicity identification evaluation
TMDL total maximum daily load
TRE toxicity reduction evaluation
TSD technical support document
TSS total suspended solids
TTO total toxic organics
TU toxic unit
TUa acute toxic unit
TUC chronic toxic unit
WQS water quality standard
WLA wasteload allocation
XVII
-------
MODELING ABBREVIATIONS
ARM agricultural runoff model
CHNTRN Channel Transport Model
CETIS Complex Effluent Toxicity Information System
CIS Chemical Information System
CORMIX1 Cornell Mixing Zone Expert System
CTAP Chemical Transport and Analysis Program
DESCON computer program that estimates design condi-
tions
DFLOW computer program that calculates biologically
based design flows
DYNHYD4 hydrodynamic model
DYNTOX dynamic toxics model
EXAMS-II Exposure Analysis Modeling System
FCM2 WASP Food Chain Model
FETRA Finite Element Transport Model
FGETS Food and Gill Exchange of Toxic Substances
FLOSTAT U.S. Geological Survey 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 Speciation Model
MICH Michigan River Model
NPS Nonpoint Source Model for Urban and Rural Ar-
eas
PSY steady-state, two-dimensional plume model
SARAH2 surface water assessment model for back calculat-
! ing reductions in biotic hazardous wastes
SERATRA Sediment Contaminant Transport Model
SLSA Simplified Lake/Stream Analysis
TODAM Transport One-Dimensional Degradation anh Mi-
gration Model
TOpdWASP Chemical Transport and Fate Model "
TOXI4 asubsetofWASP4
TOXIC Toxic Organic Transport and Bioaccumulation
.1 Model
UDKHDEN three-dimensional model used for single or mul-
: tiple port diffusers • . \ , ,
ULINE uniform linear density flume model
UMERGE 'two-dimensional model used to analyze positively
; buoyant discharge
UOUTPLM cooling tower plume model adapted for maririe
discharges
UPLUME , numerical model that produces flux-average dilu-
J- ' ' tions ' • ' ' ; • • '' • ' ' • • • •
! I * > : '. . - >»«".'
VV/VSP4 water quality analysis program
W^STOX . Estuary and Stream Quality Model
WQAB FLOW water quality analysis system flow data subroutine
xviii
-------
GLOSSARY
absolute toxicity is the toxicity of the effluent without considering
dilution.
acute means a stimulus severe enough to rapidly induce an effect;
in aquatic toxicity tests> an effect observed in 96 hours
or less typically is considered acute. When referring to
aquatic toxicology or human health, an acute affect is
not always measured in terms of lethality.
acute-to-chronic ratio (ACR) is the ratio of the acute toxicity of
an effluent or a toxicant to its chronjc 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 are those acutely toxic to aquatic
organisms following their short-term exposure within
an affected area.
acute toxicity endpoints (ATE) are toxicity test results, such as
an LCso (96 hours) and ECsq (48 hours), which describe
a stimulus severe enough to rapidly induce an effect on
aquatic organisms.
additivity is the characteristic property of a mixture of toxicants
that exhibits a total toxic effect equal to the arithmetic
sum of the effects of the individual toxicants.
ambient toxicity is measured by a toxicity test on a sample
collected from a waterbody.
antagonism is the characteristic property of a mixture of toxicants
that exhibits a less-than-additive total toxic effect.
antidegradation policies are part of each State's Water quality
standards. These policies are designed to protect water
quality and provide a method of assessing activities that
may impact the integrity of the waterbody.
aquatic community is an association of interacting populations
of aquatic organisms in a given waterbody or habitat.
averaging period is the period of time over which the receiving
water concentration is averaged for comparison with
criteria concentrations. This specification limits the
duration of concentrations above the criteria.
bioaccumulation is the process by which a compound is taken up
by an aquatic organism, both from water and through
food.
bioaccumulation factor (BAF) is the ratio of a substance's
concentration in tissue versus its concentration in ambient
water, in situations where the organism and the food
chain are exposed.
bioassay is a test used to evaluate the relative potency of a
chemical or a mixture of chemicals by comparing its
effect on a living organism with the effect of a standard
preparation on the same type of organism. Bioassays
frequently are used in the pharmaceutical industry to
evaluate the potency of vitamins and drugs.
bioavailability is 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 is the process by which a compound is absorbed
from water through gills, or epithelial tissues and is
concentrated in the body. . ;,
bioconcentration factor (BCF) is the ratio of a substance's
concentration in tissue versus its concentration in water,
in situations where the food chain is not exposed or
contaminated. - For nonmetabolized substances, it
represents equilibrium partitioning between water and
organisms.
biological assessment is an evaluation of the biological condition
of a waterbody using biological surveys and other direct
measurements of resident biota in surface waters.
biological criteria, also known as biocriteria, are narrative
expressions or numeric values of the biological
characteristics of aquatic communities based on
appropriate reference conditions. Biological criteria
serve as an index of aquatic community health.
biological integrity is the condition of the aquatic community
inhabiting unimpaired waterbodies of a specified habitat
as measured by community structure and function.
biological monitoring, also known as biomonitoring, describes
the living organisms in water quality surveillance used to
indicate compliance with water quality standards or
effluent limits and to document water quality trends.
Methods of biological monitoring may include, but are
not limited to, toxicity testing such as ambient toxicity
testing or whole effluent toxicity testing.
biological survey or biosurvey is the collecting, processing, and
analyzing of a representative portion of the resident
aquatic community to determine its structural and/or
functional characteristics.
biomagnif ication is the process by which the concentration of a
compound increases in species occupying successive
trophic levels.
cancer potency slope factor (ql *) is an indication of a chemical's
human cancer-causing potential derived using animal
studies or epidemiological data on human exposure. It
is based on extrapolating 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 means 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 (CTE) are results, such as a no
observed effect concentration, lowest observed effect
concentration, effect concentration, and inhibition
concentration based on observations of reduced
reproduction, growth, and/or survival from life cycle,
partial life cycle, and early life stage tests with aquatic
animal species.
XIX
-------
coefficient of variation (CV) is a standard statistical measure of
the relative variation of a distribution or set of data,
defined as the standard deviation divided by the mean.
community component is 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 means no measurable difference in
the concentration of a pollutant exists across a transect
of the waterbody (e.g., does not vary by 5 percent).
continuous simulation model is 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) is 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) is' 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 an acute effect.
critical life stage is 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 is the flow used for steady-state wasteload allocation
modeling.
designated uses are those usesspecified in waterquality standards
for each waterbody or segment whether or not they are
being attained.
discharge length scale is the square root of the cross-sectional
area of any discharge outlet.
diversity is the number and abundance of biological taxa in a
specified location.
effect concentration (EC) is 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.
equilibrium partitioning (EP) is a method for generating
sediment criteria that focuses on the chemical interaction
between sediments and contaminants.
final acute value (FAV) is an estimate of the concentration of the
toxicant corresponding to a cumulative probability of
0.05 in the acute toxicity values for all genera for which
acceptable acute tests have been conducted on the
toxicant.
frequency is how often criteria can be exceeded without
unacceptably affecting the community.
genotoxic is the ability of a substance to damage an organism's
i genetic material (DNA).
harmonic mean flow is the number of daily flow measurements
divided by the sum of the reciprocals of the flows. That
is, it is the reciprocal of the mean of reciprocals.
inhibition concentration (1C) is a point estimate of the toxicant
concentration that would cause a given percent reduction
(e.g., IC25) in a nonlethal biological measurement of the
test organisms, such as reproduction or growth.
lethal concentration is the point estimate of the toxicant
! concentration thatwould be lethal to a given percentage
of the test organisms during a specific period.
lipophilic is a high affinity for lipids (fats).
load allocations (LA) are the portion of a receiving water's total
maximum daily load that is attributed either to one of its
existing or future nonpoint sources of pollution or to
natural background sources.
lognormal probabilistic dilution model calculates the
probability distribution of receiving water quality
concentrations from the lognormal probability
; distributions of the input variables.
log P (also expressed as log ROW or as n-octanal/water
partition coefficient) is 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.
lowest observed adverse effect level (LOAEL) is 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.
magnitude is how much af a pollutant (or pollutant parameter
such as toxicity), expressed as a concentration or toxic
unit is allowable.
minimum level (ML) refers to the level at which the entire
i analytical system gives recognizable mass spectra and
acceptable calibration points when analyzing for
pollutants of concern. This level corresponds to the
lowest pointat which the calibration curve is determined.
mixing zone is an area where an effluent discharge undergoes
initial dilution and is extended to cover the secondary
mixing in the ambient waterbody. A mixing zone is an
• allocated impact zone where water quality criteria can
be exceeded as long as acutely toxic conditions are
prevented.
Monte Carlo simulation is 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.
xx
-------
no observed adverse effect level (NOAEL) is a tested 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) is the highest tested
concentration of an effluent or a toxicant at which no
adverse effects are observed on the aquatic test organisms
at a specific time of observation. Determined using
hypothesis testing.
nonthreshold effects are associated with exposure to chemicals
that have no safe exposure levels (i.e., cancer).
permit averaging period is the duration of time over which a
permit limit is calculated (days, weeks, or months).
persistent pollutant is not subject to decay, degradation,
transformation, volatilization, hydrolysis, or photolysis.
priority pollutants are those pollutants listed by the Administrator
under CWA Section 307(a).
probability is 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 is a mathematical representation of the
probabilities that a given variable will have various
values.
practical quantitation limit (PQL) is a correction factor,
sometimes arbitrarily defined, used to account for
uncertainty in measurement precision.
reasonable potential is where ah effluent is projected "or
calculated to cause an excursion above a water quality
standard based on a number of factors including, as a
minimum, the four factors listed in 40 CFR
receiving water concentration (RWC) is the concentration of a
toxicant or the parameter toxicity in the receiving water
after mixing (formerly termed "instream waste
concentration" [IWC]).
recurrence interval is the average number of years within that a
variable will be less than or equal to a specified value.
This term is synonymous with return period.
reference ambient concentration (RAC) is the concentration of
a chemical in water that will not cause adverse impacts
to human health. RAC is expressed in units of mg/l.
reference tissue concentration (RTC) is the concentration of a
chemical in edible fish or shellfish tissue that will not
cause adverse impacts to human health when ingested.
RTC is expressed in units of mg/kg.
reference dose (RfD) is an estimate of the daily exposure to
human population that is likely to be without an
appreciable risk of deleterious effect during a lifetime;
derived from nonobserved adverse effect level or lowest
observed adverse effect level.
relative toxicity is 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 is 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.
steady-state model is a fate and transport model that uses
constant values of input variables to predict constant
values of receiving Water quality concentrations.
STORET is EPA's computerized water quality data base that
includes physical, chemical, and biological data measured
in waterbodies throughout the United States.
subiethal means a stimulus below the level that causes death.
synergism is the characteristic property of a mixture of toxicants
that exhibits a greater-than-additive total toxic effect.
.threshold effects result from chemicals that have a safe level (i.e.,
acute, subacute, or chronic human health effects).
total maximum daily load (TMDL) is the sum of the individual
wastelpad allocations and load allocations. 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.
toxicity identification evaluation (TIE) is a set of procedures to
identify the specific chemicals responsible for effluent
toxicity.
toxicity reduction evaluation (TRE) is a site-specific study
conducted in a stepwise 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 is a procedure to determine the toxicity of a chemical
or an effluent using living organisms. A toxicity test
measures the degree of effect on exposed test organisms
of a specific chemical or effluent.
toxics are those pollutants that have a toxic effect on living
organisms. TheCWASection307(a) "priority" pollutants
are a subset of this group of pollutants.
toxic pollutants are those pollutants listed by the Administrator
under CWA Section 307(a).
toxic units (TUs) are a measure of toxicity in an effluent as
determined bytheacutetoxicityunitsorchronictoxicity
units measured. :
toxic unit acute (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 (i.e., 100
LC50).
toxic unit chronic (TUC) is the reciprocal of the effluent
concentration that causes no observable effect on the
test organisms by the end of the chronic exposure
period (i.e., 100/NOEC).
water quality assessment is an evaluation of the condition of a
waterbody using biological surveys, chemical-specific
analyses of pollutants in waterbodies, and toxicity tests.
-------
wasteload allocation (WLA) is the portion of a receiving water's
total maximum daily load that is allocated to one of its
existing or future point sources of pollution.
water quality criteria are comprised of numeric and narrative
criteria. Numeric criteria arescientifically derived ambient
concentrations developed by EPA or States for various
pollutants of concern to protect human health and
aquatic life. Narrative criteria are statements that describe
the desired water quality goal.
water quality limited characterizes a stream segment in which it
is known that water does not meet applicable water
quality standards, and/or is not expected to meet
applicable waterqualitystandardsevenafterapplication
of technology-based effluent limitations.
water quality standard is a law or regulation that consists of the
beneficial designated use or uses of a waterbody, the
numeric and narrative water quality criteria that are
• necessary to protect the use or uses of that particular
I waterbody, and an antidegradation statement.
whole effluent toxicity is the total toxic effect of an effluent
! measured directly with a toxicity test.
xxii
-------
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, ana-
lyzing, and controlling adverse water quality impacts caused by
tbxic discharges to the surface waters of the United States. The
original TSD was published in September 1985. Since then, the
Clean Water Act (CWA) 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
experienced has been gained in controlling toxic pollutants. Be-
cause of these changes, EPA revised and updated the TSD.
This guidance document is intended to support the implementa-
tion of the CWA water quality-based approach to toxics control.
As such, the recommendations and guidance found in this docu-
ment are not binding and should be used by regulatory authori-
ties 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 (bioaccumulation methods, Chapter 3) and should
not be used until the procedures are finalized.
Background
The EPA surface water toxics control program, represented dia-
grammatically in the figure, relies on portions of the national
pretreatment program, the effluent limitations guidelines pro-
gram, 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 Dis-
charge Elimination System (NPDES) program. States are authorized
by EPA to implement certain portions of the national toxics con-
trol program, such as the NPDES program. Scientific and techni-
cal guidance is developed and published by EPA to assist the
States. EPA is required by the CWA and federal regulations to play
an oversight role to ensure that States authorized to implement
various program requirements do so in accordance with federal
regulations.
States are given discretion in the CWA to establish and implement
water quality standards. As such, there may be differences in
toxics control programs between States. EPA's oversight role is to
ensure that each State's program is technically sound and that
each State fully implements its program.
Throughout the evolution of the toxics control program, EPA has
provided guidance concerning new program initiatives, statutory
developments, and regulatory requirements. In 1980, EPA em-
phasized in its preamble to NPDES regulations (45 FR 33520) that
NPDES permit limitations must reflect the most stringent of tech-
nology-based, water quality-based controls, or other standards
required by the CWA (e.g., ocean discharge requirements under
Section 403 and toxics standards or prohibition under Section
307[a]). EPA reiterated the significance of surface water toxics
control in 1984 through the publication of its national policy
statement entitled, "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 limita-
tions in NPDES permits to control toxic pollutants. Regulatory
provisions promulgated on June 2,1989 (54 FR 23868), clarify EPA's
surface water toxics control program and the use of whole effluent
toxicity, and implement CWA Section 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 CWA. To effectively accomplish this
objective, EPA recommends the use of an integrated water qual-
ity-based approach for controlling toxic discharges. EPA's inte-
grated "standards to permits" approach, illustrated in the figure,
starts with water quality criteria, objectives, and standards and
results in NPDES permit limits 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
limits are set, compliance is essential. Compliance can be ascer-
tained 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.
xxiii
-------
Define water quality
objectives, criteria, and standards
Establish priority
waterbodies
Chapter 2
Chemical-specific Effluent
Characterization
I
- Evaluate for excursions above standards
I
,_.___ Determine "reasonable potential"
I
Generate effluent data
Whole Effluent Toxicity Effluent
Characterization
Evaluate for excursions above standards -
' ' I
Determine "reasonable potential"
I
Generate data
Evaluate exposure
(critical flow, fate modeling,
and mixing) and calculate
wasteload allocation
Define required discharge
characteristics by the
wasteload allocation
Derive permit
requirements
Evaluate toxicity reduction
i
Investigate indicator
parameters
Final permit with
monitoring requirements
Compliance
i
Chapter 3
Chapter 5
Chapter 6
Overview of the Water Quality-based "Standards to Permits" Process for Toxics Control
XXIV
-------
1. APPROACHES TO WATER QUALITY-BASED TOXBCS
CONTROL
1,1 INTRODUCTION
In this chapter, basic principles are presented that cover the
protection of aquatic life and the protection of human health
from impacts caused by the release of toxics to the Nation's
surface waters. Protection against toxic releases is called for under
Section 101 (a)(3) of the Clean Water Act (CWA), which states that
"it is the national policy that the discharge of toxic pollutants in
toxic amounts be prohibited/' In addition, CWA ^Section 303(c)
requires States to develop water quality standards to protect the
public health or welfare, enhance the quality of water, and serve
the purposes of the CWA. The control of the discharge of toxics is
a paramount objective of the National Pollutant Discharge Elimi-
nation System (NPDES) and water quality standards programs.
The CWA and Environmental Protection Agency (EPA) regulations
(described in Appendices B-l and B-4, respectively) authorize and
require the use of the "integrated strategy" to achieve and main-
tain water quality standards. In addition, EPA policy and guidance
have long advocated this approach (see Appendices B-2 and B-3).
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
biological criteria/bioassessment and biosurvey approach. How-
ever, for the protection of human health, technical constraints do
not yet allow for full 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
limits, chemical-specific testing and limits, and biological criteria
using bioassessments/biosurveys, relies on the water quality stan-
dards that each State has adopted. All States have water quality
standards consisting of both chemical-specific numeric criteria for
individual pollutants, and narrative "free from toxics in toxic'
amounts" criteria. Currently, a few States have incorporated bio-
logical criteria into water quality standards.
The narrative water quality criteria in all States generally require
that the State 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. The use of toxicity testing and whole effluent toxicity
limits is based upon a State's narrative water quality criterion and/
or in some cases, a State numeric criterion for toxicity.
. - . ,-. - ,. ;,,>.. ,., ... ,•
Chemical-specific numeric criteria have been adopted by each
State. In many cases, States have adopted EPA-recommended
water quality criteria as a part of their water quality standards [1,
2]. (See Chapter 2, Water Quality Criteria and Standards, for
further information.) These State-adopted numeric chemical cri-
teria provide the basis upon which specific chemicals can be
limited in permits. Where States have not developed chemical-
specific numeric criteria, States may interpret their narrative stan-
dards for specific chemicals by using EPA criteria updated with
current quantitative risk values.
Biological criteria provide a direct measure of ambient aquatic life
and overall biological integrity in a waterbody. Biological criteria
constitute one basis for limits that will protect the biological
integrity of a surface water.
The integrated approach must include the control of toxics through
implementation of the narrative "no toxics" criterion and/or nu-
meric criteria for the parameter toxicity, the control of individual
pollutants for which specific chemical water quality criteria exist in
a State's standards, as well as use of biological criteria. Reliance
solely on the chemical-specific numeric criteria or the narrative
criterion or biological criteria would result in only a partially
effective State toxics control program. In the discussion that
follows, each control approach is described in greater detail as
well as how each of the approaches complement the other two
by providing additional information for the protection of water
quality.
1.2 CHEMICAL-SPECIFIC APPROACH FOR AQUATIC
LIFE PROTECTION
The chemical-specific approach to toxics control for the protec-
tion of aquatic life uses specific chemical effluent limits in NPDES
permits to control the discharge of toxics. These limits are
developed from laboratory-derived, biologically based numeric
water quality criteria adopted within a State's water quality stan-
dards. Water quality criteria are adopted by a State for the
protection of the designated uses of the receiving water. Chemi-
cal-specific water quality-based limits 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 and resultln the development of a wasteload
allocation (WLA) and a total maximum daily load (TMDL) through
modeling, a mixing zone analysis, and the calculation of permit
limits. Once a numeric water quality criterion is adopted, chemi-
cal-specific limits must be developed in NPDES permits to ensure
that a permittee's discharge does not exceed acute or chronic
water quality criteria for the pollutant in a receiving water if there
- is a reasonable potential for that discharge to cause or contribute
to excursions of the criterion. These steps are discussed in Chap-
ters 3,4, and 5.
EPA water quality criteria for the protection of aquatic life are
developed under the requirements of CWA Section 304(a)(1) and
are published by EPA in separate criteria documents and summa-
rized in the Quality Criteria for Water [1 ]. Water quality criteria
are derived scientifically 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 concen-
tration and dispersal of pollutants, or their byproducts, through
biological, physical, and chemical processes, and the effects of
pollutants on biological community diversity, productivity, and
stability of the receiving water [1 ]. They can be used to assess and
control a variety of water quality impacts. Chapter 2 provides a
more detailed discussion of the derivation of numeric criteria.
Recommendations for using chemical-specific data to determine
which individual toxicants need to be controlled are found in
Chapter 3. Legal requirements, including chemical-specific limits
in permits, are found in Chapter 5.
1,2.1 Correlation of Chemical-specific Measurements to Actual
Receiving Water Impacts
EPA has conducted a series of studies to determine whether its
water quality criteria concentrations are protective of aquatic life
in receiving water systems. The first study was conducted at
Shayler Run, Ohio, to evaluate the applicability of laboratory-
generated toxicity data to a natural stream artificially dosed with
copper to provide steady concentrations [3]. The results of the
study indicate that several characteristics of site-specific water
quality affect the toxicity of copper. The results also indicate that
avoidance of elevated concentration areas by instream organisms
can produce observable ecological changes at concentrations
below those found to be harmful in laboratory toxicity tests. No
instream effects were observed at continuous exposure concen-
trations near EPA's current chronic criterion, applied at the water
hardness of Shayler Run.
Studies performed on experimental streams at EPA's Monticello
Ecological Research Station (MERS) indicate good agreement be-
tween EPA's criteria concentrations and the instream concentra-
tions producing aquatic life effects under steady exposure condi-
tions [4-13]. EPA's water quality criteria are not threshold levels
abo^e which definite measurable instream effects are always ex-
pected. Rather, the criteria embody conservative Assumptions
such that small excursions above the criteria should,not result in
measurable environmental impacts upon the biota. The data
indicate that if the ambient water quality criteria are met, then the
biota in the receiving water system will be protected from unac-
ceptable impacts caused by the chemical of concern. The studies
conducted by MERS are described in greater detail in Box 1 -1 and
Tables 1-1 and 1-2.
1.2,2 Chemical-specific Analytical Method Precision
Tables 1-3 to 1-5 illustrate the types of precision commonly seen
in inorganic, organic, and nonmetal inorganic chemical analyses
that are routinely used for determining concentrations of specific
pollutants in effluents. These tabjes show the observed variability.
The variability of chemical measurements increases as one ap-
proaches the limit of detectability for a chemical. Table 1 -3 shows
the interlaboratory precision of 10 metals. The coefficient of
variation (CV), defined as the standard deviation divided by the
mean x 100, for these analyses ranges from 18 percent to
129 percent [15]. Table 1 -4 shows the interlaboratory precision
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 [4-14] 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 pentachloro-
phenol. Some studies were conducted during a summer (pentachlorophenol) while others continued for more
than 2 years (selenium IV). Tables 1-1 and 1-2 show sample data on ammonia and ammonia combined with
chlorine. In all experiments, the streams were dosed continuously 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 EPA 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. this indicates that the CCC is providing chronic
protection at the recommended concentration for that particular chemical. 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 may be expected under continuous exposure conditions.
-------
Table 1-1. Effects in Streams Exposed to Ammonia [8-13]
Table 1-3. Intel-laboratory Precision of Inorganic Analysis
at the Low End of the Measurement Detection Range [15]
Indicator
Fish
Fathead minnow
Bluegill
Channel catfish
White sucker
Rainbow trout
Walleye
Benthic Invertebrates
Zooplankton
Effects
Criteria3
Od
0
+
0
0
0
0
0
3Xb
0
0
++
0
0
0
'+
-
9XC
0
++
-H-+
0
++
++
-
+'
Notes
a Criteria = 0.05 mg/l unionized ammonia (NHs) at average stream pH and tem-
perature; 1.0 mg/i 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 = three times criteria concentration based on input of 3 mg/l total ammonia.
c 9X = Nine times criteria concentration based on input of 9 mg/l total ammonia.
d 0 s No difference from controls; +'s represent gradation of differences from
controls ranging from slight (+•) to dramatic (++++).
Analyte
Aluminum
Cadmium
Chromium
Copper
Iron
Lead
Manganese
Mercury
Silver
Zinc
No. of Labs
37
63
72
86
78
64
55
76
50
62
CV(%)
43
66
40
36
38
46
129
79
.18
118
Table 1-4. Interlaboratory Precision Ranges for Organic
Chemical Analysis
Table 1-2. Effects in Streams Exposed to Ammonia
and Chlorine [8-13]
Indicator
Fish
Channel catfish
Bluegill
Benthic invertebrates
Zooplankton
Bacteria
Periphyton
Primary production
Litter decomposition
Aquatic plants
Effects
4 ug/la
--H-b
0
0
0
+
0
0
+ -
0
35 ug/l
++
0
+
0
++
0
0
+
0
122 ug/l
+++
0
++
0
+++
0
0
++
0
Notes
a Average concentrations of TRC in presence of 2mg/l to 3mg/l total ammonia;
national criteria for chlorine = 11 ug/l.
** 0 = No difference from controls; +'s represent gradation of differences from
controls ranging from slight (+) to dramatic (++++).
Chemical
Benzene
4 Chlorobenzenes
Ethyl benzene
Toluene
23 Halocarbons
4 Halocarbons
11 Phenols
Benzidine
3,3-Dichlorozidine
6 Pthalate esthers
3 Nitrosarhines
24 Organochlorine
Pesticides and PCBs
16PNAs
No.
Labs
20
20
20
20
17
16
17
22
?
CV
31-64
16-29
40-50
20-45
38-64
38-69
?
7
>12-45
16-91
% Data
Discarded*
10
\
20
7
7
22
19
?
7
EPA
Document
Referenced
600/S4-84-064
600/S4-84-064
600/S4-84-044
600/S4-84-062
600/S4-84-056
600/S4-84-051
600/S4-84-061
600/S4-84-063
' Discarded as outliers.
It is important to note that in many chemical analyses a decision may be made
that certain anomalous data points, or outliers, are unusable and are not re-
ported as valid data points. This type of data evaluation is made because in
chemical analyses it is routine to repeat the analysis with the same sample and
reference standard until an acceptable result is obtained.
-------
Table 1-5. Intel-laboratory Precision of Nonmetal Inorganic
Analyses Over the Measurement Range [15]
No.
Lab
17
>20
16
6
15
6
58
58
21
Parameter
Alkalinity
Residual chlorine
Ammonia nitrogen
Kjeldahl nitrogen, total
N03 nitrogen
Total P
BOD
COD
TOC
CV(%) Range
4.9-14
13-25
15-58
38-41
17-61
25-40
15-33
6.9-34
4.6-70
associated with organic chemical analyses. The CVs range from
12 percent to 91 percent. Table 1-5 demonstrates the
interlaboratory precision of nonmetal inorganic analyses at the
lower end of the measurement range. The CVs for this type of
analyses range from 4.6 percent to 61 percent [15]. The data in
Tables 1-3 to 1-5 reflect testing in reagent grade water. Actual
CVs from testing effluents can be higher due to matrix effects.
However, in 40 CFR Part 136 analytical methods, matrix effects
are acknowledged.
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 pollutants [16]. Whole effluent
toxicity tests employ the use of standardized, surrogate freshwa-
ter or marine (depending upon the mixture of effluent and receiv-
ing water) plants, invertebrates, and vertebrates. EPA has published
extensive written protocols listing numerous marine and freshwa-
ter species for toxicity testing [17,18,19].
An acute toxicity test is defined as a test of 96-hours or less 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, in addition to lethality. Traditionally, chronic tests are
full life-cycle tests or a shortened test of about 30 days known as
an early life stage test. However, the duration of most of the EPA
chronic toxicity tests have been shortened to 7 days by focusing
on the most sensitive life-cycle stages. For this reason the EPA
chronic tests are called short-term chronic tests. Box 1 -2 summa-
rizes the short-term chronic tests currently recommended by EPA.
The acute and short-term chronic methods recommended by EPA
are presented in three methods manuals [17,18,19].
In a laboratory acute toxicity test, an effluent sample is collected,
diluted, and placed in test chambers with the chosen test species.
After 24, 48, 72, and 96 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 col-
lected, diluted, and placed in test chambers. An example of a di-
lution series used in chronic or acute tests is 1 00, 50, 25, 1 2.5,
and 6.25 percent, and a control. 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. In the short-term chronic tests, at test termination, 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 fecun-
dity, reduced growth, etc.). In the acute tests, at test termination,
the number of dead organisms are recorded and an LC$Q is cal-
culated.
Dilution water is an important part of toxicity testing. Dilution
water may either be standard laboratory water and/or the receiv-
ing water. Sometimes the receiving water is used to dilute the
effluent because it more closely simulates effluent/receiving water
interactions. This may be especially important in the case of saline
receiving waters. The salinity of the receiving water should be
matched as closely as possible to the salinity in the test chambers
(wifhin the salinity range constraints of a particular method) for
the purposes of conducting the tests.
Quality control and quality assurance are an integral part of whole
effluent toxicity testing. Use of a standard control water and a
reference toxicant test are both recommended to ensure quality
assurance in chronic testing. It is important to understand that
each of the chronic tests has minimum criteria of acceptability for
each endpoint that is measured in the controls (i.e., 80 percent
survival and minimum criteria for growth, reproduction, and
fertilization). The acute tests also have criteria of acceptability
measured in the controls.
Acute toxicity endpoints (ATEs) commonly include lethal concen-
trations (LCs) and are described in terms of effluent concentra-
tiops. The LC is the concentration of toxicant at which a certain
percentage of the test organisms die, e.g., the LC-jo or LC5Q. An
exposure duration also is included in the endpoint such as 24, 48,
72, or 96 hours (e.g., 96-hour
Commonly used chronic toxicity endpoints (CTEs) include the no
observed effect concentration (NOEC), the lowest observed effect
concentration (LOEC), and the effect concentration (EC). The
NOEC is the highest concentration of toxicant, in terms of per-
cent effluent, to which the test organisms are exposed that causes
no observable adverse effect. The effects measured may include
decreases in reproduction and growth, or lethality. The LOEC is
the lowest concentration of toxicant to which the test organisms
are exposed that causes an observed effect. Again, the same
effects are usually observed. The EC is the toxicant concentration
that would cause an adverse effect upon a certain percentage of
the test organisms, (e.g., ECfo
In chronic toxicity tests, the exposure duration in the EPA testing
protocols is almost always assumed to be the 7-day short-term
period unless otherwise specified in the protocol. For example,
the Ceriodaphnia test must be continued until at least 60 percent
-------
Box 1-2. Short-term Chronic Toxicity Methods
Species/Common Name
Test Duration
Freshwater Species
Ceriodaphnia dubia
Cladoceran
Pimephales promelas
Fathead minnow
Pimephales promelas
Fathead minnow
Selenastrum capricomutum
Freshwater algae
Marine/Estuarine Species
Arbada punctulata
Sea urchin
Champia parvula
Red macroalgae
Mysidopsis bahia
Mysid
Cyprinodon variegatus
Sheepshead minnow
Cyprinodon variegatus
Sheepshead minnow
Menidia beryllina
Inland silverside
Approximately 7 days
(until 60 percent of control
have 3 broods)
7 days
7-9 days
96 hours
1.5 hours
7-9 days
7 days
7 days
7-9 days
7 days
Test Endpoints
Survival, reproduction
Larval growth, survival
Embryo-larval survival,
percent hatch,
percent abnormality
Growth
Fertilization
Cystocarp production
(fertilization)
Growth, survival, fecundity
Larval growth, survival
Embryo-larval survival,
percent hatch,
percent abnormality
Larval growth, survival
of the females produce three broods. This may require more or
less than 7 days to occur.
It is useful to note that LCs and ECs are point estimates statistically
derived from a mathematical model, that assumes a continuous
dose-response relationship. NOECs and LOECs, statistically deter-
mined using hypothesis testing, are not point estimates [18]. In
order to overcome the difficulty in statistically deriving the NOEC
using hypothesis testing, a new statistical procedure has been
developed. This procedure, referred to as the inhibition concen-
tration (1C), is a point estimate interpolated from the actual
effluent concentrations at which measured effects occurred dur-
ing a chronic test. The 1C is an estimate of the toxicant concentra-
tion 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 re-
production would represent the effluent concentration at which a
25-percent reduction in reproduction occurred.
Since the 1C is a point estimate, a CV can be calculated. A CV
cannot be calculated if hypothesis testing is used because results
are only available for the effluent concentrations used. For this
-------
reason, estimates of test precision cannot be calculated for NOECs
derived by hypothesis testing.
The 1C also Is not dependent upon the selection of the effluent
concentrations. In contrast, NOECs calculated by hypothesis
testing are dependent upon the concentrations initially selected.
For example, if a chronic test is conducted using TOO, 50, 25,
12.5, and 6.25 percent effluent concentrations, and the LOEC
exhibited by the data is at 25 percent effluent, the NOEC calcu-
lated by hypothesis testing is estimated to be the next lowest
dilution, or 12.5 percent However, the true NOEC value may lie
somewhere between 25 percent and 12.5 percent effluent
Comparisons of both types of data indicate that an NOEC derived
using the IC2S is approximately the analogue of an NOEC derived
using hypothesis testing (see Figure 1-1). For the above reasons,
if possible, the 1C25 is the preferred statistical method for deter-
mining the NOEC.
Another important issue in conducting both acute and short-term
chronic toxicity tests is the dilution series. The EPA methods
manuals recommend six dilutions, including the control. The
only exception to this is a toxicity test conducted on ambient
receiving waters. Then, each ambient receiving water is com-
pared statistically to the control without dilutions. It is not
accurate to assume that two dilutions (the receiving water con-
centration [RWC] and control) are all that are ultimately necessary
for determining compliance with a toxicity limit. If the toxicity
tests are conducted with only the control and one effluent con-
centration (I.e., the RWQ, the error and variability associated with
this type of statistical analysis is large [20].
For the above reasons, EPA recommends the use of five effluent
concentrations and a control to determine the magnitude of
toxicity. When conducting compliance monitoring, an option is
to choose the five concentrations that bracket the RWC (two
concentrations above and two below). This would result in the
determination of compliance status as well as a statistically valid
estimation of the NOEC. The information provided from the full
dilution series would indicate how close the test endpoints are to
the permit limit and how close to violating the limit the discharger
is, and, if measured over time, the variability of the effluent.
1.3.1 Toxic Units
Since toxicity involves an inverse relationship to EC (the lower the
EC, the higher the toxicity of the effluent), it is more understand-
able to translate concentration-based toxicity measurements into
toxic units (IDs). In this way, the potential confusion involving
the inverse relationship is overcome and the permit limit deriva-
tion process is better served. The number of toxic units in ah
effluent is defined as 100 divided by the EC measured:
TUa = 100/LC50
TUC=TOO/NOEC.
For jCxample, an effluent with an acute toxicity of an LC$Q in
5 percent 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
100-,
90-
80-
1
! 70~
8 60-
50-
40-
30-
20-
10-
n-
4.3%
I n»1 I
26%
n=6
34.8%
n=8
17.4% i7.4o/0
n=4 n=4
n=0
IC10
IC2o
IC30
IC50
Figure 1-1. This figure represents the percentage of the time the mean NOEC was approximately equivalent to an lOjo, ICjs, I
IC25/ IC3o, and IC5Q for all 23 effluent and reference toxicant data sets analyzed. The data sets included short-term chronic
toxicity test for Ceriodaphnla dubla, Plmephales promelas (fathead minnows), Arbada punctulata (sea urchin), Cyprinodon varlegatus
(sheepshead minnows), and Champla parvula (red algae) [21].
-------
TUC. TUa is the measurement of acute toxicity units and TUC is a
measurement of chronic toxicity units. (See the glossary for a
definition pf 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 efflu-
ent sample that contains 20 TUcs is twice as toxic as an effluent
that contains 10 TUcs.
1.3.2 Correlation of Whole Effluent Toxicity Measurements to
Actual Receiving Water Impact
EPA conducted the Complex Effluent Toxicity Testing Program
(CETTP) that 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
(i,ew temperature, pH, salinity), can give a valid assessment of
receiving system impacts on waters that support aquatic biota
[22-25]. Summaries of these site studies are provided in Box 1-3
(freshwater) and .Box 1-4 (saltwater). In addition, three other
studies, presented in Box 1-3, were conducted to address this
issue: a comparative investigation conducted by the University of
Kentucky [26], a second study on the Trinity River in Texas
conducted by the University of North Texas [27], and a third
study conducted by the North Carolina Division of Environmental
Management [28]. , . ......
It is important to note that in these studies, different objectives
were-addressed.. The CETTP freshwater studies attempted to
correlate receiving water chronic tpxicity measured by EPA toxic-
ity tests to instream observed impacts (Figure 1 -2). The CETTP
saltwater studies compared effluent toxicity to ambient receiving
water toxicity using dye studies to measure receiving water con-
centrations of effluent. The North Carolina study compared
effluent toxicity to receiving water impact • using Ceriodaphnia
chronic toxicity tests and receiving stream benthic
macroinvertebrates (Figure 1 -3). The Kentucky study examined
the relationship between effluent toxicity tests and instream eco-
logical parameters. The Trinity River study attempted to spatially
compare the biological, physical, and chemical water quality and
sediment quality of Trinity River reaches above and below the
Dallas/Fort Worth area (Figure 1-4).
Together, these studies comprise a large data base specifically
collected to determine the validity of toxicity tests to predict
receiving water community impact. |n order to address the
correlation of effluent and ambient toxicity tests to receiving
water impacts, EPA evaluated the results of the studies discussed
above [29]. The results, when linked together, clearly show that if
toxicity is present after considering dilution, impact will also be
present.
Parkhurst et al., were requested by representatives of industrial
and municipal discharges to critique the CETTP studies [30]. One
major criticism was that the EPA study sites were not selected
randomly and .therefore the results,of the.studies cannot be
extended to all waters. EPA agrees that the CETTP sites were not
selected to represent a statistically valid sampling of all types of
waterbodies in the United States. A representative sampling of
receiving water would require assessment of more sites than EPA
could study in a comprehensive manner. Such a sampling was
beyond the capability of EPA's resources. However, the CETTP
and corresponding studies such as the Trinity River study [27] did
show unequivocally that a strong correlation exists between tox-
icity and a biological impact. :
EPA believes that it is reasonable to assume in the absence of data
showing otherwise that this relationship is basically independent
Box 1-3. Correlation of 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 onsite concurrent with the field surveys. Sites exhibiting biological impacts 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 and downstream
of the discharge points) 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 [23, 31-38].
Figure 1-2 illustrates the data from the CETTP studies. A robust canonical correlation analysis was performed to
determine whether or not statistically significant relationships existed between the ambient toxicity tests and
instream biological response variables and to identify which variables played an important ro|e in that relation-
ship [29]. Influential variables were then used, to classify stations as either impacted or not. Ceriodaphnia dubia
productivity,and/or Pimephales promelas weight were used as the basis for predicting impact. Fish richness was
used to classify streams as impact observed or impact not observed.
-------
Box 1-3. Correlation of Toxicity Measurements to Receiving
Water Impact (Freshwater) (continued)
Classification was based on the relative performance of the stations on each stream in the study. Percentiles of
the appropriate distribution (normal for toxicity variables, and Poisson for fish richness) were used to set cutoffs
for classification. Two-way contingency tables representing stations as impact predicted or not, and impact
observed or not were prepared from a variety of cutoffs (percentages). The exact test for independence was
performed on each contingency table.
If toxicity test results were used to classify sites as impacted or not (predicted classification) and if a strong
relationship does exist between ambient toxicity and biological response, then the classification of stations
according to biological response should closely match the predicted classification. Hence, the errors in
misclassification should be small.
Figure 1-2, developed using a 95 percent-95 percent cutoff, shows that false positives (impact predicted but
none found) occurred at 7.5 percent of the 80 stations. The probability of getting no more than 7.5 percent false
positives under the null hypothesis that there is no relationship between ambient toxicity and biological response
is less than p=0.001. As discussed above, this is the only definitive error that can be identified in such
comparisons. The correct or noncontradictory findings (no measured toxicity but observed impacts) were
92.5 percent of the stations. A variety of other cutoff criteria combinations were evaluated and the number of
false positives remained in the 7 percent to 8 percent range. Therefore, a discharger's chance of being charged
incorrectly with causing instream toxicity is low if and only if dilution in the receiving water is considered.
I
A comparative time series study conducted on the Trinity River in Texas that used the same classification method
as the CETTP studies also showed a strong relationship between ambient toxicity and instream biological
response (Figure 1-2). False positives (impact predicted but not observed) had a frequency of 8.3 percent.
Overall there was a 91.7-percent accuracy of prediction or noncontradictory findings [29], and the probability of
a false positive (impact predicted but not observed/impact predicted) ranged from 8 percent to 11 percent in
these studies.
Another study conducted by the North Carolina Division 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 Cerlodaphnia dubia chronic test and a qualitative
macroinvertebrate sampling. Overall there was 88 percent accuracy of prediction (Figure 1 -3) [28].
In addition, another comparative study was conducted in the Kentucky River Basin [26]. 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, it was 0.93. The estimated toxicity (LG|) correlated closely with the actual
percent instream effluent dilution observed at the first downstream station at which no ecological impact was
discernafale.
Using the statistical classification previously described in the CETTP and Trinity River studies, an analysis was
conducted on the combined data sets of the CETTP, Trinity River, and Kentucky River Basin data. Because the
North Carolina study was based on the Ceriodaphnia dubia chronic test and a qualitative macroinvertebrate sam-
pling, the data were not amenable to this type of statistical analysis. This combined analysis is illustrated in Figure
1 -5. The probability of getting no more than 9.4 percent false positives (impact predicted/impact not observed)
when the null hypothesis (no relationship between ambient toxicity and biological response) is less than
p=0.0028.
-------
Box 1-4. Correlation of Effluent Toxicity Measurements to
Receiving Water Toxicity (Saltwater)
In saltwater systems, as in freshwater systems, receiving water impact should only be seen where receiving water
waste concentrations are at or above the effect concentrations. Dilution in marine and estuarine systems may be
greater due to large and/or complex mixing than most freshwater systems. As a result, there is a less likely chance
for receiving water impacts to be observed in saltwater systems as predicted by toxicity tests.
Figure 1 -6 illustrates the comparison between predictions of saltwater receiving water toxicity and whole effluent
toxicity. Toxicity test data from 79 ambient stations (four study sites) were compared to effluent toxicity test
results from an isolated discharge at each site. All receiving water toxicity to effluent toxicity correlations are
based on dye studies conducted at each of the four sites to determine the actual dilution.
Most of the sites were selected because the discharge was isolated from other point sources and potential
impacts from other point sources was anticipated to be negligible. Two of these studies indicated near-field
effects, generally within the mixing zone. One study conducted at Fernandina Beach, Florida [25], showed
impacts outside the proposed mixing zone. Results of another study (East Greenwich) indicated the existence of
poor water quality well beyond the influence of the East Greenwich Sewage Treatment Plant and suggests that
other sources (point or nonpoint) may contribute significantly [25, 39, 40]. This condition may be typical in
some of the more stressed estuaries.
In a total of 79 comparisons, 11 out of 15 (73 percent) of the receiving water samples predicted to be toxic were
toxic. This constitutes 14 percent of the total comparisons. Toxicity was not predicted in the receiving water and
toxicity was not seen in the receiving water 59 out of 64 times (92 percent). This constitutes 75 percent of the
total comparisons.
In 5 percent of the total comparisons there was a false negative prediction, or the toxicity tests predicted no
toxicity when the receiving water was toxic [24]. As previously discussed, toxicity is only one possible adverse
influence. Since only toxicity is measured, a very high correlation should not be expected necessarily because
receiving water biological impacts may be attributed to other sources or factors.
The results of the studies at these four sites indicates a 94 percent accuracy when using the marine and estuarine
toxicity tests to predict receiving water impacts. In only 6 percent of the cases did effluent toxicity tests predict
receiving water toxicity that was not present (false positive).
of waterbody type. Also, this was not the objective of the CETTP
studies. The CETTP purpose was to determine if toxicity and
impacts to biological communities are found concurrently in
receiving waters. Therefore, EPA disagrees that this is a reason to
conclude that the CETTP studies failed to show the validity of
toxicity tests to predict water quality impact.
Another criticism was the studies did not investigate replication of
results over time. However, toxicity results cannot be expected to
be replicated over time in waters where river flow and other time-
variant factors change the degree of ambient toxicity. Indeed,
the Kanawa River and Five-Mile Creek data showed that ambient
toxicity did not occur at high river flows whereas it was found at
low flows; this was an expected result. The objective of the CETTP
studies was to see if impact was present when effluent toxicity
exceeds the available effluent dilution. This objective was achieved
by the studies.
Another major criticism was the correlation between toxicity tests
and biological impact relied extensively upon maximum impact
responses and that correlation was poor when data from high
flow events and lesser toxicity discharges (minimal impact re-
sponses) were added. EPA acknowledges that impact correlations
will be higher where higher toxic impact occurs and lower where
impacts are expected to be minimal. Such a response is expected
given the complexity of ecosystems and that biological communi-
ties and species have different sensitivities to toxicants and may
respond differently. Also, higher river dilution will reduce the
potential instream impact from effluent toxicity. However, this
observation does not disprove that the CETTP and other studies
showed a statistically sound relationship to correlate toxicity to
the existence of a biological ambient impact. Therefore, EPA still
concludes that control of toxicity is a valid approach for protect-
ing ambient water quality.
In addition, other studies confirm that effluent toxicity, when
adequately related to ambient conditions, can give a valid assess-
-------
3.8%
7.5%
2.5%
n
Impact not predicted/
Impact observed
Impact predicted/
Impact observed
Impact not predicted/
Impact not observed
Impact predicted/
Impact not observed
Figure 1-2. Comparison of Ambient Toxicity and Instream
Impact—EPA Study [23, 31-38]
5.6%
8.3%
11.1%
Impact not predicted/
Impact observed
Impact predicted/ * •
Impact observed
Impact riot predicted/
Impact not observed •
Impact predicted/
Impact no't observed
Figure 1-4. Comparison of Ambient Toxicity and Instream
Impact—Trinity River [29]
23.0%
No Instream toxiclty
predicted, Impact
noted
Instream toxiclty
predicted, Impact
noted
No Instream toxiclty
predicted, no Impact
noted
Instream toxiclty
predicted, no Impact
noted
65.0%
Figure 1-3. Comparison of Effluent Toxicity of Receiving
Water Impact Using Cerlodaphnla Chronic Toxicity Tests and
Freshwater Receiving Stream Benthic Invertebrates at 43
Point Source Discharging Sites In North Carolina [28]
4.4%
9.4%
6.3%
H Impact not predicted/
Impact observed
_- Impact predicted/
•• Impact observed
~ Impact not predicted/
M Impact not observed
_ Impact predicted/ ;
^ Impact not observed
Figure 1-5. Comparison of Ambient Toxicity and Instream
Impact—EPA Study, Trinity River Study, and
Kentucky Study [26]
6.0%
5.0%
75.0%
14.0%
No ambient toxlcily
predicted, toxicity
observed :
Predicted ambient
toxicity, toxiclty
observed
No ambient toxicity
predicted, no toxiclty
observed
Predicted ambient
toxicity, no toxiclty
observed
Figure 1-6. Comparison of Predictions of Receiving Water toxicity Based on Effluent Toxicity and Ambient
Receiving Water Testing in Saltwater Environments: 79 Ambient Stations
and 4 Dischargers [24, 25, 39, 40]
10
-------
ment of receiving water impact [3, 24, 26-29, 39, 41]. These
studies tested waters other than those studied under CETTP.
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 from toxics would only be sus-
pected where effluent concentrations after dilution are at or
above the toxicity effect concentrations. Influences from sub-
strate differences and physical conditions, such as dissolved oxy-
gen, temperature, channelization, flooding and weather cycles,
also can affect the biological community adversely. These other
types ,of influences may be better evaluated by using a
bioassessment approach. However, the existence of these other
factors concurrently with toxicity does not absolve a regulatory
authority from controlling the discharge of toxicity if the State has
established a designated use to protect aquatic biota.
The value of the toxicity test is its ability to assess the impact of
discharged toxicants independent of effects from other factors.
This allows regulatory authorities specifically to identify and con-
trol the portion of the impact caused by the discharge. Biological,
physical, and chemical factors of the community can influence
the actual effects that effluent toxicity may cause in the receiving
water, and further emphasize the need for a totally integrated
water quality-based approach.
7.3.3 Toxicity Test Method Precision
Like all measurements, toxicity tests exhibit variability. Toxicity
test variability can be described in terms of two types of preci-
sion^-"within" or intralaboratory precision, and round robin or
interlaboratory precision. Intralaboratory precision is the ability of
trained laboratory personnel to obtain consistent results repeat-
edly when performing the same test on the same species using
the same toxicant. Interlaboratory precision (or round robin tests)
is 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. ;
EPA believes that several toxicity test methods have a precision
profile that can be reasonable to evaluate compliance with NPDES
permits. The appropriateness of a given method can be deter-
mined in a permit proceeding or, in part, by rulemaking. EPA has
proposed a range of whole effluent toxicity test procedures in 40
CFR136 and may promulgate these methods soon. Current data,
however, show that the precision profiles of a number of whole
effluent toxicity tests is similar to already approved chemical-
specific methods.
Research into the precision of whole effluent toxicity methods by
various groups (including EPA) has shown that toxicity test proce-
dures exhibit variability [17-18, 19, 42-49]. In chronic 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). This is in
contrast to acute toxicity tests where the test endpoint is normally
calculated at midrange (i.e., LCso), DUt is sometimes calculated at
the lower end of the biological detection range (i.e., LGi). CVs
cannot be calculated for NOEC endpoints determined using an
analysis of variance (hypothesis testing) because this procedure
does not produce a statistical point estimate. However, CVs can
be calculated for NOECs if they are determined using the 1C
statistical procedure, and for EC and LC, endpoints because they
are all statistical point estimates. ;
To facilitate the comparability between'different NOEC calcula-
tions using the IC25 and the analysis of variance (hypothesis test-
ing), Appendices A-1 and A-2 list NOEC results in terms of both.
In some instances the IC2S could not be calculated based on sta-
tistical assumptions and available data. In addition, there are
some instances where an IC25 cannot be calculated because there
was no toxic effect. In these' cases, the CV for a method and
reference toxicant was calculated using only data where IC25S could
be calculated.
A more detailed discussion of precision can be found in Box 1-5.
Tables 1 -6 and 1 -7 summarize the intralaboratory precision for all
10 EPA short-term chronic whole effluent toxicity tests and some
acute toxicity tests. In addition, Table 1-8 summarizes the
interlaboratory precision for three chronic test species and two
acute test species using a variety of different compounds.
In summary, whole effluent toxicity testing methods can repre-
sent practical tests that estimate potential receiving water im-
pacts. ' Permit limits that are developed correctly from whole
effluent toxicity tests should protect aquatic biota if the discharged
effluent meets the limits. It is important not to confuse permit
limit variability with toxicity test variability. Chapter 5 discusses
permit limit variability. , :
1.3.4 Considerations Involved When Implementing the Whole
Effluent Toxicity Approach
An understanding of some basic considerations and toxicological
principles., is important in order to apply routinely the whole
effluent approach to the assessment and control of municipal and
industrial effluents. The following sections provide a more indepth
discussion of each of these factors and principles. (Chapters 3 and
5 discuss specific details for characterizing an effluent and deriv-
ing permit limits.) . • • , ,
~ ) -1 -, * •
Onsite versus Offsite Toxicity Testing
Comparisons of toxicity data between tests conducted onsite and
tests conducted offsite on samples shipped to Environmental
Research Laboratory (ERL)-Duluth and (ERL)-Narragahsett via air-
freight have, with a few exceptions, shown little variation. For
many effluents, onsite or offsite test data do not appear to be
significantly different. The major consideration is cost. Cost also
should be weighed against data needs to make the onsite/offsite
determination.
For example, if the presence in the effluent of nonpersistent
compounds (i.e., chlorine or other volatiles) is suspected or known,
then the regulatory authority may want to conduct onsite testing.
If it is not considered important to the analysis of toxic impact,
offsite testing is as acceptable as onsite 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, onsite testing is strongly
recommended. Regardless, cost considerations should not over-
11
-------
Box 1-5. 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 endpoints 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 accor-
dance 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:
• 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.
• Use of reference toxicants to routinely evaluate the quality and sensitivity of the test organisms to be used in
each test.
• Development of "control charts" should be prepared for each reference toxicant/organism/protocol combi-
nation to determine if the results are within prescribed limits. The control chart consists of successive data
added with each reference toxicant test, and is the basis for evaluating data once the control chart" is
established.
• The minimum criteria of test acceptability specific for each protocol.
Guidelines for recommended quality assurance practices are found in each manual [17,18,19].
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 detailed
at the end of each of the methods sections in the methods manuals [17,18,19] and is summarized in Appendix
A (Tables A-1-1 to A-1-18 for the marine/estuarine methods and Tables A-1-19 to A-1-31 for the freshwater
methods) and summarized in Tables 1-6 and 1-7. Intralaboratory precision data also are presented for acute
toxicity tests and are summarized in Table 1-8. 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 LC5Q and chronic ECsg, IC25, and \C$Q data. A mean and range is given for
the chronic no observed effect concentration (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 zero, 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 quality assurance/quality control (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 usec1 as a check for whether the compliance data submitted is minimally acceptable [18,
19]. (See Table 1 of each of the 4 freshwater methods and Table 2 of each of the 10 marine/estuarine methods
entitled, "Summary of Recommended Effluent Toxicity Test Conditions.")
To date, interlaboratory precision (round robin) tests have be?n completed for the 7-day Fathead Minnow Lar-
val 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 three methods. Results of the round robin testing will show greater variability (i.e., larger
CVs) due to a larger number of variables introduced by many round robin laboratories participating. Researchers
12
-------
Box 1-5. Toxicity Test Method Precision (continued)
have found that a two- to threefold increase in CV values is acceptable with biological testing [46, 50, 51 ].
Interlaboratory data also are presented from several acute toxicity tests [46]. The data from these round robin
tests can be found in Appendix A (Tables A-1 -5, A-1 -23, A-l -24, A-1 -27, A-1 -28, and A-1 -30) and are summarized
in Table 1-8.
Researchers agree that the precision of these tests is acceptable. Rue, Fava, and Grothe concluded that whole
effluent toxicity test methods "are comparable to accepted analytical methodologies" [50]. Another study by
Grothe, Kimerle, and Malloch also concluded that when comparing "...CVs for select effluent toxicity test
methods and commonly accepted analytical methods...the precision of both techniques is similar" [51 ]. This has
led the Agency to conclude "...that toxicity test methods, where properly followed, exhibit an acceptable range
of variability" (see the discussion of toxicity testing requirements for POTWs, 55 ffl 30082 at 30112, July 24,1990)
[52]. '
ride the need to characterize adequately a given effluent and the
factors unique to the discharge situation.
Flow-through versus Static anil Renewal Toxicity Testing
Several factors should 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; is the dis-
charge continuous or intermittent?); the amount of data needed
(variable effluents may require more data); and, as between differ-
ent systems that will provide adequate data, expense.
Two basic types of testing systems are available to measure efflu-
ent 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 otie effluent sample is collected and used at the
beginning of a static test. A variation of the static procedure is the
renewal toxicity test. This test uses the same delivery system as
that of a static test but the test solutions are changed, or renewed,
on a predetermined schedule (i.e., every 24 hours). Fresh effluent
samples generally'are collected to renew the test solutions.
Online 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 also can 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. If the
effluent is highly variable with an intermittent discharge, a flow-
through or a renewal test also would 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, such as a discharge from a 30-
day retention basin, then a static or renewal test using a grab or
24-hour composite sample would be an appropriate test system.
For a chronic toxicity test, a 24-hour composite effluent sample is
most appropriate. For an acute test, four grab samples taken 6
hours apart or four 6-hour composite samples are most appropri-
ate to measure the peaks of toxicity in an effluent.
Cost also is a factor. Flow-through tests are more resource
intensive and require complex delivery systems. Consequently,
less data can be generated per unit cost than with static or
renewal testing. Where more data at less cost are desirable, static
or renewal testing probably is more appropriate. Typically, more
samples using renewal is preferable to fewer samples using flow-
through for the same total cost since this would allow better
characterization of effluent variability.
Grab Sampling versus 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.
Collection of grab samples may be necessary if there is little
dispersion or mixing of the effluent in the receiving water. In
these instances the peaks could persist in the receiving water.
Although a grab sample has the potential of revealing the toxicity
peak 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 maxi-
mum 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 [17,18,19] and in Chapter 3.
Variability
There are three important sources of differences in a water quality
impact analysis:
13
-------
Table 1-6. Intralaboratory Precision of Chronic Whole Effluent Toxicity Test Methods
Test NOEC Mean
Method Range IC25
Cyprinodon variegatus— Survival and Growth
>0.05 - 0.05 mg/l 0.07
0.5 -1.0 mg/l1 1.5
31 -125 ug/l2 300.4
1.3 -2.5 mg/l1 2.2
Embryo larval survival and teratogenicity
EC10
200 - 240 ug/l2 202
2.0 -4.0 mg/l1 1.9
Menldia beiyllina — Survival and Growth
31 -125 ug/l2 209.9
1.3 + 0 mg/l 1.3
Mysldopsls bahla — Survival, Growth, and Fecundity
<0.3 - 5.0 mg/l4 5.7
63 -125 ug/l1 138.3
Arbacla punctulata — Fertilization
5.0 -12.5 ug/l1 23.5
1.2 -3.3 mg/l1 1.7
<6.1 - 24.4 ug/l2 22.9
0.9 -1.8 mg/l1 2.58
Champfa pamila— Reproduction
0.5 -1.0 ug/l1 1.79
0.5 -1.0 ug/l1 0.93
0.09 - 0.48 mg/l2 0.31
0.1 5 -0.60 mg/l2 0.46
Pimephales promelas — Survival & Growth
128 -256 ug/l1 — 5
0.011 - 0.01 3 mg/l1 — s
Embryo larval survival and teratogenicity
0.011 -0.01 3 mg/l —
0.011 -0.01 3 mg/l —
Ceriodaphnia dubia — Reproduction
0.1 0-0.30 mg/l1 0.22
0.25 -1.00 mg/l 0.91
Selenastnim capricomutum — 96-hour Survival
2.1-2.8g/l4 —
CV(%)
'
41.8 !
31.4
33.0 ;
27.6
1
2.8
35 ;
43.7
43.2
F
35.0
18.0 r
i
54.6
29.7 ;
41.9
28.7 ;
61.09 :
63
69.0
62.3
1
41.13
20.5
—
Mean
IC50
0.13
1.9
396.9
2.6
ECIso
233.5
11.7
340.8
1.9
6.9
185.8
45.7
2.4
29.9
3.2
3.35
1.4
0.36
0.75
5
5
LCi
0.0068
1.51
0.3
1.24
LC50
2.4
CV(%)
40.8
31.8
19.2
35.3
2.5
2.9
50.7
9.4
47.8
5.8
47.9
23.3
48.2
33.3
34.5
38.6
37.0
22.92
^_-r*
——
62
41.3
27.9
15.2
10.2
Compound
Copper
SDS3
Copper
SDS
Copper
SDS
Copper
SDS
SDS
Copper
- •
Copper
SDS
Copper
SDS
Copper
Copper
SDS
SDS
NAPCP6
Cadmium
Cadmium
Diquat
NAPCP
Sodium
Chloride
Sodium
chloride
Water
Used
AS
AS
NS
NS
AS
AS
NS
NS
NS
NS
AS
AS
NS
NS
NS
AS/NS
AS/NS
NS
FW
FW
FW
FW
FW
FW
1 Difference of one test concentration.
2Diffcrcncc of two test concentrations.
3Sod!um dodccyl sulfate.
^Difference of four test concentrations.
sR*w data were unavailable, so ICjs and ICjo could not be calculated.
^Sodium pcntachlorophenol.
AS-^-artificial seawater.,
NS^natural seawater.
FW—freshwater.
—: Data not available.
Note: Data used in this table are found in Appendix A-1.
14
-------
Table 1-7. Intralaboratory Precision of Acute Whole Effluent Toxicity Test Methods
N
Pimephales promelas*
(96-hour)
Daphnia pulex*
(48-hour)
Daphnia magna*
(48-hour)
(number of tests)
12
9
9
14
10
9
13
8
8
CV(%)
40
22
86
36
43
21
10
29
72
Compound
NAPCP
SDS
Cadmium
NAPCP
SDS
Cadmium
NAPCP
SDS
Cadmium
•Data taken from Draft 1990 Acute Manual.
Table 1-8. Summary of Interlaboratory Variability Data for Whole Effluent Toxicity Test Methods [17,18,19, 46]
Test Method
NOEC Range IC2sCV(%)1
Chronic
1.
2.
3.
4.
5.
6.
Cyprinodon variegatus
7-day growth and survival
Pimephales promelas
7-day growth and survival
Ceriodaphnia dubia
7-day reproduction
Ceriodaphnia dubia
7-day reproduction
Ceriodaphnia dubia
7-day reproduction
Ceriodaphnia dubia
7--day reproduction
Acute
7.
8.
Cyprinodon variegatus
96-hour static
96-hour flow-through
96-hour static
96-hour flow-through
Mysidopsis bahia
96-hour static
96-hour flow-through
96-hour static
96-hour flow-through
1 - 3.2% effluent2 44.2
<3.0 - 6.0 mg/l2 31 .0
potassium chromate
0.25 - 0.30 mg/l 41 .1
NAPCP3
6 -12% effluent2 —
<0.25 - 1 .0 mg/l 29.0
sodium chloride
0.25-1 .0 mg/l 20.5
sodium chloride
Toxicant
endosulfan
endosulfan
silver nitrate
silver nitrate
endosulfan
endosulfan
silver nitrate
silver nitrate
LC50 CV(%)
37.7
46.2
34.6
50.1
59.5
51.9
26.6
22.3
^ CV—coefficient of variation.
^This represents a difference of one exposure concentration.
3NAPCP—Sodium pentachlorophenol.
—. Data unavailable.
Note: Data summarized in this table were taken from Appendix A-1.
15
-------
• Effluent variability is caused by changes in the composi-
tion of the effluent. Virtually all effluents vary in composi-
tion over time.
• Exposure variability is caused by changes in flow rates of
both effluent and receiving water. There also are variable
receiving water parameters that may be independent of
flow, such as background toxicant levels, pH, salinity, tides,
suspended solids, hardness, dissolved oxygen, and tem-
perature, that can be important in assessing impact.
• Species sensitivity differences are caused by the differ-
ences in response to toxicants between species. -
Each type of variability is discussed below.
Effluent Variability
Effluent variability is an important component in overall variability
of water quality impact analyses and should be addressed ad-
equately in permitting (see Chapter 5, Permit Requirements).
Effluent variability can be addressed by designing proper sam-
pling 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 a testing frequency designed to assess variable effluents. Ap-
pendix F details suggested sampling procedures.
Appendix A-2 demonstrates the types of effluent variability that
may be seen in publicly owned treatment works (POTW) effluents
as measured through toxicity testing of the effluents (see Appen-
dix A-2, Tables A-2-1 to A-2-9). The CVs (effluent variability) for
POTW effluents are based on acute LC50 data that range from 19.6
percent to 42 percent effluent, and for IC25 chronic data that range
from 52.8 percent to 101.3 percent. Also in Appendix A-2, Tables
A-2-10 to A-2-12 show acute and short-term chronic effluent
variability data from oil refineries on three species, fathead min-
nows, Ceriodaphnia, and mysids. The CVs associated with this
effluent variability data range from 18.7 percent to 54 percent for
the acute LC$o data, and from 29.8 percent to 59.6 percent for
the chronic NOEC data. Data on effluent variability in various
types of manufacturing facilities are in Appendix A-2, Tables A-2-
13 to A-2-18. Acute toxicity test results show CVs for effluent
variability ranging from 20.3 percent to >53.9 percent.
Tables A-2-6 to A-2-9 in Appendix A-2 illustrate 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
(Arbada punctulatd), the red macroalga fertilization test (Champia
pamild), the mysid 7-day growth, fecundity and survival test
(Mysldopsis bahld), and the inland silverside 7-day larval growth
and survival test (Menidia beryilind). 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 percent effect concentration and were
not toxic below 0.1 percent effect concentration. This informa-
tion is discussed in Chapter 3, Recommendations for Testing the
Toxicity ,of Effluents section.
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 ex-
posure assessment procedures for freshwater and saltwater sys-
tems.
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 indi-
vidual 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 also will 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 protect aquatic
organisms. 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 ambient water quality
criteria. ,
To provide sufficient information for making permitting deci-
sions, EPA recommends a minimum number of three species,
representing three different phyla (e.g., a fish, an inverte-
brate, and a plant) 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
thejeffluent 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 mate-
rials 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, generali-
zations about sensitive and nonsensitive species are difficult to
make. For example, one generalization is that trout are consid-
ered sensitive organisms requiring high-quality water. However,
this; generalization may not apply in all cases; trout are very
sensitive to oxygen depletion but may be relatively insensitive to
16
-------
certain toxicants. Another species, Daphnia magna, is very sensi-
tive when exposed to many toxicants, but relatively insensitive
when exposured to the pesticide endrin. Bluegills are very resis-
tant to metals, particularly copper. Conversely, bluegills are a
sensitive test species for organophosphate pesticides.
Figures 1 -7 to 1 -9 show the differences in species sensitivities to
hexavalent chromium, dielorin, and an effluent from a POTW,
respectively [53]. 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 five 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 onsite, Champia parvula was the most sensitive in 15 per-
cent, Arbacia punctulata in 54 percent, mysids in 31 percent and
fish in 15 percent of the cases [24].
Analysis of species sensitivity ranges found in the national ambient
water quality criteria [1,2] indicates that if tests are conducted on
three particular species (Daphnia magna, Pimephalespromelas, and
Lepomis macrochirus), the most sensitive of the three will have an
LCjo within one order of magnitude of the most sensitive of all
species tested [54]. 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 [54].
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. The use of other representative species should be sub-
jected to strict quality assurance and quality control procedures
and should follow rigorous test methodologies that 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. To use a resident
organism, a facility would have to develop a protocol to culture
the organism and to assess intra- and interlaboratory variability.
Such testing is more costly, more difficult, and potentially subject
to more variability (disease, age, etc.) than standardized testing.
In any case, organisms collected directly from the receiving water
itself should never be used because existing impairment may
mask any toxicity.
Acute-to-Chronic Ratio
The acute-to-chronic ratio (ACR) expresses the relationship be-
tween the concentration of whole effluent toxicity or a toxicant
causing acute toxicity to a species (expressed as an acute toxicity
endpoint such as an LCso) and the concentration of whole efflu-
ent toxicity or a toxicant causing chronic toxicity to the same
species (expressed as a chronic toxicity endpoint such as an
Q.
CO
Pimephales promelas -
Ceriodaphnia reticulaia
-2.0 -1.0 0.0 1.0
Log LCso in mg/l
2.0
Figure 1-7. Log of LCsgs of Freshwater Species Exposed to
Hexavalent Chromium
8 3
8
* 2
"5
I'
Pimephales promelas
Daphnia pulex
-3.0 -2.0 -1.0 0.0
Log LCso in mg/1
Figure 1-8. Log of LCsos of Freshwater Species Exposed
to Dieldrin
CO
•5 3
Ceriodaphnia reticulata
• Arbacia punctulata
/Pimephales promelas
20 40
Log LCso in m
60 80
l as Percent Effluent
Figure 1-9. Log of LCsns of Freshwater and Saltwater
Species Exposed to a POTW Effluent
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 are 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 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
17
-------
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 also may be used in developing chronic toxicity limits
where chronic toxicity is not measured directly, in order to mini-
mize testing costs. Likewise, if the toxicity is for the most part
manifested in reproduction, growth, etc. (i.e., nonlethal) end-
points, an acute test may not be appropriate for compliance
monitoring. Where acute and chronic toxicity data are avail-
able, the ACR should be calculated directly for that specific
effluent.
Data on acute and chronic toxicity for complex effluents from
different categories of dischargers (i.e., POTWs, oil refineries, and
chemical manufacturers) show that ACRs for whole effluents range
from <1.0 to >50.0, with the majority of ACRs falling below 20
(see Appendix A-3). Acute to chronic ratios for oil refinery data
from one plant, based on three species ranged from 1.49 to
>10.0. Acute to chronic ratios for a variety of chemical manufac-
turers, based on data from two species ranged from <1.0 to
>SO.O. Acute to chronic ratios for POTWs based on two species
ranged from 1.4 to 16.1 (these data can be found in Appendix A-
3). Interestingly, this range of ACRs virtually is identical to ACRs
generated on a number of wastewater dischargers in the State of
Sao Paulo, Brazil (Appendix A-3, Tables A-3-1 and A-3-2). Al-
though the acute and chronic toxicities measured in Brazil were
proportionally higher (more toxic) than those measured in the
United States, the ACRs were quite similar (Appendix A-3, Tables
A-3-1 toA-3-3).
EPA recommends that regulatory authorities use a measured
ACR. In the absence of data to develop an ACR, EPA's data
suggests that an ACR of 10 could be used (see Appendix A-3).
This represents the upper 90th percentile of all the ACR data in
Appendix A-3. Given the protective margin of safety inherent
with the use of a critical flow for the calculation of a chronic
receiving water waste concentration, an ACR of 10 should provide
ample protection against chronic instream impacts.
1.4 BIOLOGICAL CRITERIA/BIOASSESSMENT AND
BIOSURVEY APPROACH FOR AQUATIC LIFE
PROTECTION
As Illustrated In Figure 1-10, ecological integrity is attainable
when chemical, physical, and biological integrity occur simul-
taneously [55]. Biological integrity is a good indicator of overall
ecological integrity of aquatic environments because it can pro-
vide both a meaningful goal and a useful measure of environmen-
tal status that relates directly to the overall integrity of the Nation's
waters. To better protect the biological integrity of aquatic
communities, EPA recommends that States begin to develop
and Implement biological criteria in their water quality stan-
dards. Biological criteria, or "biocriteria," are numerical values or
narrative statements that describe the reference biological integ-
rity of aquatic communities inhabiting waters of a given desig-
nated aquatic life use. When formally adopted into State stan-
dards, biological criteria and aquatic life use designations serve as
Figure 1-10. The Elements of Ecological Integrity
direct, legal endpoints for determining aquatic life use,
noriattainment. Per Section 131.11(b)(2) of the Water Quality"
Standards Regulation (40 CFR Part 131), biological criteria can
supplement existing chemical-specific criteria and provide an ak
ternative to chemical-specific criteria where such criteria cannot
be established. Biological criteria quantitatively are developed by
identifying unimpaired or least-impacted reference waters that
operationally represent best attainable conditions. Once candidate
references are identified, integrated biological surveys (biosurveys)
are used to characterize the resident community. Because of the
complexity of fully characterizing the biological integrity of an
entire aquatic community, State standards should contain bio-
logical criteria that consider various components (measures of
structure and/or function) of the larger aquatic community.
When biological criteria are incorporated into water quality pro-
grams, the biological integrity of surface waters' may be directly
evaluated and protected. Biological criteria also provide addi-
tional benefits by requiring an evaluation of physical integrity and
providing a monitoring tool to assess the effectiveness of current
chemically based criteria. Table 1-9 summarizes how biological
criteria directly and indirectly protect the elements of ecological,
integrity [55].
7.4.7 Use of Biosurveys and Bioassessments in Water Quality-
\ based Toxics Control
A biological assessment, or "bioassessment," is an evaluation of
the biological condition of a waterbody using biological surveys
and other direct measurements of resident biota in surface waters.
A biological survey, or "biosurvey," consists of collecting, process-
ing, and analyzing representative portions of a resident aquatic
community to determine the community structure and function.
Biosurveys and bioassessments can be used directly to evaluate
the overall biological integrity (structure and/or functional charac-
teristics) of an aquatic community. Deviations from the biological
integrity of an aquatic community can be measured directly using
bioassessments and biosurveys only when the impacted commu-
18
-------
Table 1-9. Water Quality Programs That Incorporate Biological Criteria to Protect
Elements of Ecological Integrity
Elements of
Ecological Integrity
Directly Protects
Indirectly Protects
Chemical Integrity
Physical Integrity
Biological Integrity
Chemical-specific criteria (toxics)
Whole effluent toxicity (toxics)
Criteria for conventionals
(pH, tempature, dissolved oxygen)
Biocriteria (biota response in
surface water)
Biocriteria
(identification of
impairment)
Biocriteria
(habitat evaluation)
Chemical/whole
effluent testing (biota
response in laboratory)
nity is compared against a predetermined reference condition.
Without proper quality controls (i.e., reference conditions),
biosurveys tend to underestimate impairment.
Biosurveys assess or detect the aggregate effect of impacts upon
an aquatic community where discharges are multiple, complex,
and variable and where point, nonpoint, and stormwater dis-
charges are all affecting the biological condition of the receiving
water. The resident community integrates the effects of multiple
stresses and sources on numerous interactive biological compo-
nents over time. Because of this, biosurveys necessarily cannot
measure the impacts of one particular effluent that is being
discharged to the receiving water. Chemical-specific analyses of
pollutants known to impact aquatic life and whole effluent toxic-
ity tests are predictive water quality assessment tools used to
evaluate biological integrity. At the present time, biological sur-
veys and biological assessments cannot be used as predictive
water quality assessment tools.
Biosurveys provide a useful monitor of both aggregate ecological
impact and historical trends in the condition of an aquatic ecosys-
tem. Biosurveys can detect aquatic life impacts that other avail-
able assessment methods may miss, such as impacts caused by
pollutants that are difficult to identify chemically or characterize
lexicologically, and impacts from complex or unanticipated ex-
posures. Perhaps most importantly, biosurveys can detect impacts
caused by habitat degradation such as channelization, sedimen-
tation, and historical contamination that disrupt the interactive
balance among community components.
Biosurvey data should be applied towards:
• Refining use classifications among different types of aquatic
systems and within a given type of use category.
• Defining and protecting existing aquatic life uses under
State antidegradation policies as required by the water
quality standards regulation.
• Classifying outstanding national resource waters.
• Identifying where site-specific criteria modifications may be
needed effectively to protect a waterbody.
• Improving use-attainability studies.
• Assessing impacts of certain nonpoint sources and, to-
gether with the chemical-specific and whole effluent toxic-
ity approaches, assist in controlling them.
• Monitoring the ecological effects of regulatory action taken
under CWA Sections 401, 402, and 301 (h).
• 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 consis-
tent laboratory and field methods, and include quality assurance
and quality control. Biosurveys should be tailored to the particu-
lar type of waterbody being assessed (e.g., wetland, lake, stream,
river, or estuary) and should focus on aquatic community compo-
nents that are representative of the larger ecosystem and that are
practical to measure. Biosurveys should be coupled routinely with
basic chemical and physical measurements and an objective
evaluation of habitat quality.
EPA's Office of Water and several State water quality programs
have developed techniques as guidance to support biosurveys
and bioassessments [56-62]. The techniques are an excellent
supplementary tool to whole effluent toxicity testing and chemi-
cal-specific techniques. However, it is important that biosurveys
include sampling of as many species at different trophic levels as
possible to reveal accurately receiving water community impacts.
Excellent examples of biosurvey/bioassessment data collected and
used in concert with ambient or effluent 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 recom-
mended minimum of three trophic levels (a fish, invertebrate, and
a plant) to give a good overall picture of what was happening in
19
-------
the receiving water. Recommended methodologies for conduct-
ing biosurveys are included in References 56 through 62.
Treatment systems are more easily designed to meet
chemical requirements because more treatability data are
available.
1.5 INTEGRATION OF THE WHOLE EFFLUENT,
CHEMICAL-SPECIFIC, AND BIOASSESSMENT
APPROACHES
Section 101 (a) of the CWA states: "The objective of this Act is to
restore and maintain the chemical, physical and biological integ-
rity 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 capa-
bilities and limitations. Table 1-10 shows EPA guidance, State
implementation, and State application of each approach [55].
The information summarized in Box 1-6, and discussed in detail
below, explains how each approach complements the other and
why no one of the approaches should be used alone.
A more detailed discussion of the capabilities and limitations of
the three approaches is provided below.
1.5.1 Capabilities and Limitations of the Chemical-specific
Approach
The principal capabilities of the chemical-specific approach are:
• At present, protection of human health only can be achieved
by control of specific chemicals.
• A more complete understanding is available on the toxicol-
ogy of specific chemicals. EPA acute ambient water quality
criteria are based on protecting up to a minimum of eight
different organisms including fish, invertebrates, and plants;
a minimum of three organisms are used to develop chronic
criteria. Considerable information is available in the scien-
tific literature on toxicity caused by specific chemicals.
• More information is available on the fate of a pollutant in
receiving waters so that the pollutant fate can be conve-
, niently predicted through modeling. Persistence and deg-
; radation can be factored into the evaluation.
• Chemical analyses are sometimes less expensive than toxic-
: ity testing and biological surveys, if there are only a few
toxicants present. This is more pertinent if only chlorine
and ammonia are present in an effluent or ambient water.
• This approach allows prediction of ecological impacts be-
. fore they occur. NPDES permit limits can therefore be
developed before an actual ecological impact occurs.
The; principal limitations of the chemical-specific approach are:
• All toxicants in complex wastewaters are not known and,
] therefore, control requirements for all toxicants cannot be
: set. Toxicological information on these unknown pollut-
ants is often unavailable.
• The bioavailability of the toxicants at the discharge site are
>'. • '<• typically not assessed, and the interactions between toxi-
cants (e.g., additivity, antagonism) are not measured or
: accounted for. As a result, the controls may be either under
protective or overly protective.
• Direct biological receiving water impact and impairment is
i not typically measured. There is no way to ascertain di-
rectly if the chemical controls adequately are protecting
aquatic life.
• Complete measurement of all individual toxicants, particu-
| larly where many are present in the mixture, can be expen-
sive. Organic chemicals, in particular, can be costly to
;. measure. ' . •. ..
Table 1-10. Process for Implementation of Water Quality Standards
Criteria
EPA Guidance
State Implementation
State Application
Chemical-Specific
Narrative "Free Frorns"
Biological
Pollutant-specific
numeric criteria
Whole effluent toxicity
guidance
Biosurvey minimum
requirement guidance
State Standards
-use designation
-numeric criteria
-antidegradation
Water Quality Narrative
-no toxic amounts translator
State Standards
-refined use
-narrative/numeric criteria
-antidegradation •••'
Permit limits monitoring
Best management practices
Wasteload allocations
Permit limits monitoring
Wasteload allocation
Best management practices
Permit conditions monitoring
Best management practices
Wasteload allocation
20
-------
Box 1-6. Components of an Integrated Approach to Water Quality-based Toxics Control
Control Approach
Capabilities
Limitations
Chemical-Specific
Whole effluent toxicity
Bioassessments
-Human health protection
-Complete toxicology
-Straightforward treatability
-Fate understood
-Less expensive testing if only
a few toxicants are present
-Prevents impacts
-Aggregate toxicity
-Unknown toxicants addressed
-Bioavailability measured
-Accurate toxicology
-Prevents impacts
-Measures actual receiving
water effects
-Historical trend analysis
-Assesses quality above standards
-Total effect of all sources,
including unknown sources
-Does not consider all toxics present
-Bioavailability not measured
-Interactions of mixtures (e.g., additivity)
unaccounted for
-Complete testing can be expensive
-Direct biological impairment not
measured
-No direct human health protection
-Incomplete toxicology
(few species may be tested)
-No direct treatment
-No persistency or sediment coverage
-Conditions in ambient may be different
-Incomplete knowledge of causative
toxicant
-Critical flow effects not always assessed
-Difficult to interpret impacts
-Cause of impact not identified
-No differentiation of sources
-Impact has already occurred
-No direct human health protection
1.S.2 Capabilities and Limitations of the Whole Effluent
Approach
The principal capabilities of whole effluent techniques are:
• 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.
• Toxicity caused by compounds commonly not analyzed for
in chemical tests is detected. Control of the toxicant(s) is
not dependent upon established toxicological information
that may not yet be available for some pollutants.
• The bioavailability of the toxic constituents is assessed, and
the effects of interactions of constituents are measured.
Additivity, synergism, and antagonism between compounds
in an effluent are addressed implicitly by whole effluent
toxicity. -.-.'.
• The toxicity of the effluent or ambient water is measured
directly for the species tested.
• This approach allows prediction of ecological impacts be-
fore they occur. NPDES permit limits can therefore be
developed before an actual ecological impact occurs.
The principal limitations of whole effluent techniques are:
• The approach only measures and controls toxicity to aquatic
organisms. It does not protect human health from expo-
sures through ingestion of fish. This is particularly impor-
tant for carcinogens.
• EPA's water quality criteria are based on a minimum of
eight different species for the acute criteria and three differ-
ent species for the chronic criteria. Effluent aquatic toxicity
commonly is measured with only one, two, or three spe-
cies. 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).
• There is less knowledge on designing or manipulating treat-
. ment systems to treat the parameter toxicity. Investigate
tools for identifying causative toxicants only .have been
recently developed and may not easily identify all causative
toxicants. As a result, identification and proper control may
be difficult and expensive.
• The whole effluent toxicity test directly measures only the
immediate bioavailability of a toxicant; it cannot measure
the persistence "downstream" and long-term cumulative
toxicity of a compound. Thus, bioaccumulative chemicals
necessarily are not assessed or limited. Toxicants can accu-
mulate in sediment to toxic concentrations over a period of
time.
• Where there are chemical/physical conditions present (pH
changes, hardness changes, solids changes, salinity changes,
photolysis, etc.) that act on toxicants in such a way as to
21
-------
"release" toxicity away from the discharge point, such tox-
idty may not be measured in the effluent. The opposite of
this also is possible; toxicity may degrade rapidly so there is
no trace of it away from the point of discharge. For
example, the actual pH and temperature in an ambient
water may be sufficiently low to preclude toxicity from
ammonia whereas the higher pH and temperature of the
toxicity test may induce toxicity from ammonia.
* It is not always clear which compound or mixture of com-
pounds is causing toxicity in the mixture. The causative
toxicant may be difficult to identify for control.
7.5.3 Capabilities and Limitations of the Bloassessment
Approach
The principal capabilities of the bioassessment approach are:
• Biological communities reflect overall ecological integrity.
Biosurvey results therefore directly assess the status of a
waterbody. The status of a waterbod/s biological health
may be of direct interest and more meaningful as a mea-
sure of a pollution-free environment.
• Biological communities integrate the effects of different
pollutant stressors and thus provide a holistic measure of
their aggregate impact. Biological assessments also mea-
sure stresses over long time periods and can measure his-
torical trends and fluctuating environmental conditions.
• Biosurveys can identify previously unknown sources of im-
pairment and may identify where site-specific chemical
criteria are needed. Bioassessments can be useful in charac-
terizing ecological impacts to a waterbody in multiple di.c-
charge situations.
• Bioassessments can characterize the ecological value of
ambient waters that are in attainment of the standards. As
such, bioassessments provide a means to determine com-
pliance with State antidegradation requirements .in stan-
dards.
The principal limitations of the bioassessment approach are:
• Bioassessments conducted at critical low flow conditions
may be difficult to accomplish.
• Biosurvey data cannot fully characterize impairment until
after suitable blocriteria are developed. Biosurvey data may
not be sufficient to detect impairments without appropriate
reference conditions.
• Bioassessments measure integrated impacts over long peri-
ods of time. Multiple factors can contribute to measured
impacts. However, bioassessments cannot isolate the caus-
ative factor leading to the impairment nor predict future
impairment.
• Bioassessments measure impact from any source and as
such, the data bracketing a discharge used to assess im-
pacts may be influenced by pollutant sources further up-
stream. Causes of biological impairment may not be as-
; signed readily to any one discharger.
» Bioassessments identify water quality problems after they
have occurred; they currently are not predictive of water
quality problems. By design, bioassessments are limited m
their ability to identify waters that are not impaired.
» The approach only measures biological impairments to
} aquatic organisms. It does not protect human health from
exposures through ingestion.of fish.
By using all three approaches, a State will more thoroughly pro-
tect aquatic life. The chemical-specific approach provides a high
accuracy of analysis of the individual chemical constituents, has
been used by regulatory agencies, and is generally lowest in cost
because of market availability. However, the level of protection of
the chemical-specific approach can be low 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 protection by measuring the
aggregate effect of all toxicants. It provides accurate toxicology,
but it can be higher in cost and has been historically less Widely
used by regulatory authorities. Bioassessments also provide a
coverage of many biological impacts and allow for accurate his-
torical trend analyses. However, bioassessments cost more and
data interpretation can be difficult. Therefore, the integrated
approach to water quality-based toxics control is essential for a
strong toxics control program.
To! more fully protect aquatic habitats and provide more compre-
hensive assessments of aquatic life use nonattainment, EPA rec-
ommends that States fully integrate chemical-specific, whole
effluent, and bioassessment approaches into their water qual-
ity-based toxics control programs. It is EPA's position that the
concept of "independent application" be applied to water
quality-based situations. Since each method has unique as
well as overlapping attributes, sensitivities, and program ap-
plications, no single approach for detecting impact should be
considered uniformly superior to any other approach. For
example, 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 (see
Chapter 3). The results of one assessment technique should
not be used to contradict or overrule the results of the other(s).
(For more information see Reference 55.)
i
Whenever there are discrepancies between the findings of the
approaches, regulatory agencies may'need to re-examine the
findings to determine if simplifications or assumptions may have
caused the difference. The State of Ohio found in 60 percent of
the sites where they collected bioassessment data, a biological
impact occurred when chemical-specific data predicted no im-
pact. The reverse also can occur—biosurveys may not show any
impact in a stream whereas effluent data modeled at low flow
project an exceedance of a chemical-specific criterion. In this
instance, the regulatory authority may need to consider a more
detailed monitoring and modeling of chemical fate and transport
22
-------
(which could include probabilistic modeling) to determine if sim-
plifications in dilution calculations projected higher concentra-
tions than would be expected using the detailed model. The
authority also would need to examine concurrently the sampling
approach and analysis of the biosurvey data to determine if it
appropriately characterized the water. If there was still a difference,
then the regulatory authority will need to use the more protective
approach as the basis to determine necessary regulatory controls.
1.6 OTHER FACTORS INFLUENCING WATER QUALITY-
BASED TOXICS CONTROL
An understanding of the fate and behavior of both single toxi-
cants and whole effluent toxicity after discharge can be important
in the application of water quality-based toxics controls. Evaluat-
ing the combined effects of interacting toxic discharges also may
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: •-••-,
-• 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.
• Additivity, Antagonism, and Synergism—When toxicants
•,.: or effluents with toxic properties mix in the receiving water,
what.is their combined fate and toxic effects?
1 ''.?"'-'=-.
.• Test .Interferences—This includes pH, temperature, salin-
ity, hardness, and metals.
Each of these.factors is discussed below.
1.6.1 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 (photode-
compqsition, microbial degradation) or compartmentalization
processes (particulate adsorption and sediment deposition, vola-
tilization).
Qne disadvantage of the chemical-specific approach is that the
bipavailability of the toxicant after discharge is not measured.
Onsite toxicity testing has indicated that the individual toxicants
causing toxicity measured at discharge sites tend relatively to be
persistent near the point of discharge [23, 31-38]. However,
persistence of individual chemicals can be modeled and the per-
sistence of specific toxicants also can be accounted for in making
impact predictions and setting controls. A procedure to deter-
mine whether or not an effluent's toxicity is persistent has been
developed by EPA [63]. The procedure describes the steps re-
quired to conduct a laboratory evaluation of the degradation of
toxicity in complex effluents that are released to receiving waters
by sirnplistically simulating a water body and discharge. EPA
recommends this procedure be conducted where the interac-
tion of sources of toxicants is critical to establishing controls.
This simple procedure is performed in a refrigerator-sized environ-
mental 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 pro-
cedure has several applications, including measuring the decay of
effluent toxicity in a stream or lake, and identifying the most
important fate processes responsible for toxicity decay (which
also may 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 typically small enough that
toxicity evaluations need only consider near field situations. Con-
tinuous discharges continually can introduce toxic pollutants into
a receiving water. Although these pollutants can decay over time,
this decay will occur downstream or away from the discharge.
The receiving water concentrations at the point of discharge
continually are being refreshed. In these instances, toxicity can be
considered conservative and persistent (nondecaying) in the near
field.
However, effluent toxicity can exhibit far field decay. Typical
patterns of progressively decreasing downstream toxicity (similar
to biochemical oxygen demand decay) have been observed in a
number of freshwater situations [23, 31-38]. This is of concern
when evaluating the combined toxicity of sources located far
apart. 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:
• 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 also can 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.
, • Ambient toxicity testing, as detailed in Appendix C, mea-
sures 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 estua-
rine 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
23
-------
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 [39].
For coastal discharges, certain toxic compounds are more often
found to cause impacts in marine and estuarine environments
[64]. 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.
1.6.2 Additivity, Antagonism, and Synergism
Where multiple toxic effluents are discharged to a receiving wa-
ter, the resultant ambient toxicity is of interest. 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). The over-
all ambient toxicity could be equal to the sum of each discharge's
toxicity (additivity), less than the sum (antagonism), or greater
than the sum (synergism).
Alabaster and Lloyd [65] observed from their data that the com-
bined acutely lethal toxicity to fish and other aquatic organisms is
approximately the simple addition of the proportional contribu-
tion from each toxicant The median value of the effect on fish is
0.95 of that predicted; the collective value for sewage effluents,
river waters and a few industrial wastes is 0.85. The range for
effluents, river wastes, and industrial wastes is 0.4 to 2.8. (Figure
1-11 illustrates the data summary.)
Pesticides and Other Substances
•Regression Line of Points
•Sewage anil Industrial Wastes
O Effluent With High Industrial
nd Pesticide Components
0.4 1 2 48
Times as Toxic as Predicted from Summed Toxic Units
Figure 1-11. Data Summary on Additivity [65]
In relation to chronic toxicity, for the growth offish, Alabaster and
Lloyd [65] 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.
Cases in which one effluent or pollutant parameter (such as total
suspended solids) 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, Chapter 3.
I
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 onsite efflu-
ent toxicity studies, and is not considered an important factor in
the toxicological assessment of effluents.
In summary, the available information indicates that the com-
bined effects of individual acutely toxic pollutants are from 0.4 to
2.8 times the effects predicted by adding the individual effects.
The median combined effect is approximately additive. For this
reason, EPA recommends in the absence of site-specific data
that regulatory authorities consider combined acute toxicity
to be additive. Since the data shows no such additivity for
chronic toxicity, EPA recommends that chronic toxicity not be
considered as additive.
1.15.3 Test Interferences
Environmental conditions such as pH, temperature, salinity, hard-
ness, and solids concentration can influence the toxicity test. For
example, higher ambient solids concentrations provide more sur-
faces for toxicants to be adsorbed and can tend to reduce toxicity.
In;addition, toxicity caused by ammonia is controlled by the
ambient pH and temperature. As a normal part of the whole
effluent toxicity testing procedure, it is very important to
replicate closely the "worst case" receiving water conditions
in the testing conditions.
There may be a few unusual situations where the pH, tempera-
ture, hardness, salinity, and solids requirements of the testing
procedures differ greatly from the worst environmental condi-
tions for these parameters. In these situations, the effluent toxic-
ity tests may either over or under predict the toxicity in the
ambient receiving water. An example of this is where ammonia is
present and the highest expected ambient water temperature is
20°C whereas the chronic toxicity test must be conducted at
25°C. Since a higher temperature causes more ammonia toxicity,
the temperature requirements of the test may induce toxicity not
found in the ambient water. In such an instance, the regulatory
authority must look carefully at the test protocols and all the data
collected to determine if the facility is actually contributing to
toxicity in the ambient water. A toxicity identification evaluation
24
-------
may be necessary to make this determination. If this analysis
shows a toxicity test result to be artificial due to environmental
parameters, then that test should be overridden by subsequent
valid toxicity tests conducted.
1.7 HUMAN HEALTH PROTECTION
Impacts on human health due to exposure to waterborne toxi-
cants 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 miti-
gated for specific chemicals by applying drinking water criteria.
The third exposure route, human consumption of contaminated
aquatic life, is of primary concern in this document due to the
potentially high concentrations achieved in fish and shellfish tis-
sues from bioconcentration, and because no NPDES permitting
controls exist between tissue contamination and human exposure.
For these reasons, this document focuses on prevention of con-
taminated aquatic life from bioconcentration as the principal way
to control human exposure to waterborne toxicants.
Currently, the regulation of human health impacts typically are
based only upon the control of individual chemicals. EPA human
health water qualify 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 is developing,
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 [66]. This procedure is described in
Chapter 3. Once this method is reviewed (both internally and
externally) and finalized, it will provide another way for regulatory
authorities to assess bioconcentratable pollutants.
1.7.1 Types of Health Effects
Health effects from toxics are divided into two categories:
nonthreshold effects, such as carcinogenicity, and threshold ef-
fects, 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 deter-
mined 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. Thresh-
old 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 expo-
sure levels. This characteristic distinguishes threshold endpoints
from nonthreshold endpoints. However, it should be noted that
chemicals that 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 endpoint is manifested.
For example, there could be a large number of cells performing
the same or similar function whose population must be signifi-
cantly 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.
Currently, the control of toxicants that bioconcentrate in edible
tissues is achieved in the NPDES program by limiting such pollut-
ants individually. There are whole effluent tests that can measure
a wastewater's potential to cause carcinogenicity or mutagenicity
(e.g., Ames test). However, the application of such data is experi-
mental because of the difficulty in establishing cause/effect rela-
tionships 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 Charac-
terization, discusses the evaluation of effluents for potential hu-
man health impacts. Chapter 4, Exposure and Wasteload Alloca-
tion, contains information on design conditions and averaging
periods. Finally, Chapter 5, Permit Requirements, discusses the
derivation of permit limits protective against human health im-
pacts.
25
-------
CHAPTER 1
REFERENCES
1. U.S. EPA. 1987. Quality Criteria for Water. Office of Wter
Regulations and Standards, Washington, DC. EPA 440/5-
86-001.
2. U.S.EPA. 1983. Water Quality Standards Handbook. Office of
Water Regulations and Standards, Washington, DC.
3. Geckler, J.R., W.B. Homing, T.N. Neiheisel, Q.H. Pickering,
E.L Robinson, and C.E. Stephan. 1976. Validity of Labora-
tory Tests for Predicting Copper Toxicity in Streams. EPA 600/
3-76-116.
4. Hedke, S., C. West, K.N. Allen, T. Norberg-King, and D.
Mount 1986. The Toxicity of PCP to Aquatic Organisms
Under Naturally Varying and Controlled Environmental
Conditions. Env. Toxicol. Chem. 5:531 -42.
5. Hedke, S.F., and J.W. Arthur. 1985. Evaluation of a Site-
Specific Water Quality Criterion for Pentachlorophenol Us-
ing Outdoor Experimental Streams. Aquatic Toxicology
and Hazard Assessment: Seventh Symposium. ASTM STP
854. Ed. R.D. Cardwell, T. Purdy, and R.C. Banner. Ameri-
can Society for Testing and Materials, Philadelphia, PA.
6. Yount, J.D., and J.E. Richter. 1986. Effects of Pentachlorophe-
nol on Periphyton Communities in Outdoor Experimental
Streams. Arch. Environ. Contam. Toxicol. 15:51 -60.
7. Zlschke. 1985. Effects of Pentachlorophenol on Invertebrates
and Fish in Outdoor Experimental Channels. Aquatic Toxi-
cology 7:37-58.
8. Arthur, J.W., C.W. West, K.N. Allen, and S.F. Hedtke. 1987.
Seasonal Toxicity of Ammonia to Five Fish and Nine Inverte-
brate Species. Bull Environ. Cont. Tox. 38(2):324-31.
9. 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.
47:249-83.
10-Zischke, J.W., and J.W. Arthur. 1987. Effects of Elevated
Ammonia Levels on Fingernail Clams (Musculium
transversurri) in Outdoor Experimental Streams. Arch. Environ.
Cont, Tox. 16:225-31.
11. 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.
12. 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.
13. 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-
28.
14. Perry, J.A., N.H. Troelstrup, Jr., M. Newson, and B., Shelley.
Generality. Wat. Sci. Tech. 19(11 ):55-71.
15. U.S. EPA. 1979. .Methods for Chemical Analysis of Water and
; Wastes. Environmental Monitoring and Support Labora-
| tory, Cincinnati, OH. EPA 600/4-79-020 (revised March
1983).
16. Wang, W. 1990. Toxicity Assessment of Pretreated Industrial
Wastewaters Using Higher Plants. Res. \. Water Poll. Control
Fed. 62:853-60.
17. U.S. EPA. 1991. Methods for Measuring the Acute Toxicity of
Effluents to Aquatic Organisms. 4th. ed. Office of Research
and Development, Cincinnati, OH. EPA-600/4-90-027.
~
18. U.S. EPA. 1991. Short-Term Methods for Estimating the Chronic
Toxicity of Effluents and Receiving Waters to Freshwater Or-
ganisms. 3d. ed. Office of Research and Development,
! Cincinnati, OH. EPA-600/4-91/002. , ..
19. U.S. EPA. 1991. Short-Term Methods for Estimating the Chronic
' Toxicity of Effluents and Receiving Waters to Marine and Estua-
! fine Organisms. 2d. ed. Office of Research and Develop-
\ ment, Cincinnati, OH. EPA-600/4-91/003.
• • '•...• r-,
20. Warren-Hicks, William J. 1990. Biological Test Methods-
Performance Considerations. Kilkelly Environ. Assoc. Re-
port no. 1342-144-01/01/29/90-02F.
21. Norberg-King, T.J. 1991. Calculations of ICp Values of 1C-] 5,
IC2Q/ IC25, ICso, and ICso for Appendix A of the Revised
Technical Support Document. Memorandum to M. Heber,
. [ u.s. EPA. ;,
- i
22. Bergman, H., R. Kimerle, and A.W. Maki, eds. 1985. Environ-
: mental Hazard Assessment of Effluents. Elmsford, NY:
Pergamon Press, Inc. : . ,
23. U.S. EPA. 1987. Biomonitoring to Achieve Control of Toxic
Effluents. EPA 625/8-87/013.
24. Schimmel, S.C., G.E. Morrison, and M.A. Heber; 1989. Ma-
rine Complex Effluent Toxicity Program: Test Sensitivity,
Repeatability, and Relevance to Receiving Water Impact.
Env. Toxicol. and Chem. 5:739-46.
26
-------
25. Schimmel, S.C., G.B. Thursby, M.A. Heber, and M.|. Chammas.
1989. Case Study of a Marine Discharge: Comparison of
Effluent and Receiving Water Toxicity. In Aquatic Toxicol-
ogy and Environmental Fate: Eleventh Volume. ASTM STP
1007. Ed. G.W. Suter II and M.A. Lewis. American Society
for Testing and Materials, Philadelphia, PA.
26. Birge, W.J., J.A. Black, T.M. Short, and A.G. Westerman. 1989.
A Comparative Ecological and lexicological Investigation
of a Secondary Wastewater Treatment Plant Effluent and Its
Receiving Stream. Envi. Toxicol. and Chem. 8:437-50.
27. Dickson, K.L., W.T. Waller, J.H. Kennedy, W.R. Arnold, W.P.
Desmond, S.D. Dyer, J.F. Hall, J.T. Knight, jr., D. Malas, M.L.
Martinez, S.L Matzner. 1989. A Water Quality and Ecologi-
cal Survey of the Trinity River, volume 1. Report conducted
by Institute of Applied Sciences, University of N. Texas and
Graduate Program in Environmental Sciences, University of
Texas at Dallas.
28. Eagleson, K.W., D.L. Lenat, L. Ausley, and F. Winborne. 1990.
Comparison of Measured Instream Biological Responses
with Responses Predicted byCeriodaphnia Chronic Toxicity
Tests. Env. Toxicol. and Chem. 9:1019-28.
29. Dickson, K.L., W.T. Waller, L.P. Ammann, and j.H. Kennedy.
1991. Examining the Relationship Between Ambient Toxicity
and Instream Impact. Submitted to: Environmental Toxi-
cology and Chemistry.
30. Parkhurst, B.R., M.D. Marcus, and L.E. Noel. 1990. Review of
the Results ofEPA's Complex Effluent Toxicity Testing Program.
Utility Water Act Group.
31. 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, DC, Office of Research
and Development, Duluth, MN. EPA-600/3-84-080, Au-
gust 1984.
32. Mount, D.I., and T.J. Norberg-King, eds. 1985. Validity of Ef-
fluent and Ambient Toxicity Tests for Predicting Biological Im-
pact, Scippo Creek, CirdevHIe, Ohio. U.S. EPA. EPA/600/3-85/
044, June 1985.
33. Mount, D.I., et al., eds. 1985. Validity of Effluent and Ambient
Toxicity Tests for Predicting Biological Impact, Five Mile Creek,
Birmingham, Alabama. U.S. EPA. EPA 600/8-85/015.
34. Mount, D.I., A.E. Steen, and T. Norberg-King, eds. 1986.
Validity of Effluent and Ambient Toxicity Tests for Predicting
Biological Impact, Back River, Baltimore Harbor, Maryland.
U.S. EPA. EPA 600/8-86/001, July 1986.
35. Mount, D.I., T. Norberg-King, and A.E. Steen, eds. 1986.
Validity of Effluent and Ambient Toxicity Tests for Predicting
Biological Impact, Naugatuck River, Waterbury, Connecticut.
U.S. EPA. EPA 600/8-86/005, May 1986.
36. Norberg-King, T.J., and D.I. Mount, eds. 1986. Validity of Ef-
fluent and Ambient Toxicity Tests for Predicting Biological Im-
pact, Skeleton Creek, Enid, Oklahoma. U.S. EPA. EPA 600/8-
86/002, March 1986.
37. Mount, D.I.; A.E. Steen, and T. Norberg-King, eds. 1986.
Validity of Effluent and Ambient Toxicity Tests for,Predicting
Biological Impact, Ohio River, Wheeling, West Virginia. U.S. EPA.
EPA 600/3-85/071, March 1986,
38. Mount, D.I., and T. Norberg-King, eds. 1986. Validity of Ef-
fluent and Ambient Toxicity Tests for Predicting Biological Im-
pact, Kdnawha River, Charleston, West Virginia. U.S. EPA. EPA
600/3-86/006, July 1986.
39. Dettmen, E.H., J.F. Paul, J.S. Rosen, and 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 No. 1003.
40. U.S. EPA. 1989. Biomonitoring for the Control of Toxicity in
Effluent Discharges to the Marine Environment. EPA 625/8-
89/015,. September 1989.
41. Nimmo, D.R., M. Dodson, P.H. Davies, J.C. Greene, and M.A.
Kerr. 1990. Three Studies Using Ceriodaphnia to Detect
Nonpoint Sources of Metals from Mine Drainage. Res. ].
Water Poll. Control Fed. 62:7-14.
42. Nebeker, A. 1982. Evaluation of a Daphnia magna, Renewal
Life-Cycle Test Method with Silver and Endosulfan. Water
Research 16:739-44.
43. Grothe, D., and R. Kimerle. 1985. Inter-and Intra-Laboratory
Variability in Daphnia magna. Effluent Toxicity Test Results.
Env. Tox. and Chem. 4(2)189-92.
44. Qureshi, A.D., K.W. Flood, S.R. Thompson, S.M. Junhurst, C.S.
Inniss, and D.A. Rokosh. 1982. Comparison of a Lumines-
cent Bacterial Test with Other Bioassays for Determining
Toxicity of Pure Compounds and Complex Effluents. Ed.
J.G. Pearson, et al. In Aquatic Toxicology Hazard Assessment:
Fifth Conference. ASTM STP 766. American Society for
Testing and Materials, Philadelphia, PA.
45. Strosher, M.T. 1984. A Comparison of Biological Testing Meth-
ods in Association with Chemical Analyses to Evaluate Toxicity
of Waste Drilling Fluids in Alberta, volume 1. Canadian Pe-
troleum Association, Calgary, Alberta.
46. Schimmel, S.C. 1981. Results: Interlaboratory Comparison of
: Acute Toxicity Tests Using Estuarine Animals. EPA-600/4-81 -
003.
47. U.S. EPA. 1982. Pesticide Assessment Guidelines. Office of
Pesticide Programs, Washington, DC. EPA/9-82-018-028.
48. U.S. EPA. 1982. Toxic Substances Test Guidelines. Office of Toxic
Substances, Washington, DC. EPA/16-82-001-003.
27
-------
49. 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
ToxicityTests. Res.J. W.P.CF. 61 (11/12):1707-10.
50, Rue, W.J., J.A. Fava, and D.R. Grothe. 1988. A Review of
Inter- and Intralaboratory Effluent Toxicity Test Method
Variability. Aquatic Toxicology and Hazard Assessment: 10th
Volume. ASTMSTP971.
51. Grothe, D.R., R.A. Kimerle, and CD. Malloch. 1990. A
Perspective on Biological Assessments. Water Environment
and Technology.
52. U.S. EPA. 1990. EPA Administered Permit Programs, the
National Potlutant Discharge Elimination System, General
Pretreatment Regulations for Existing and New Sources,
Regulations to Enhance Control of Toxic Pollutant, and
Hazardous Waste Discharges to Publicly Owned Treatment
Works. July 24,1990,55 FR 30082.
53. LeBlanc, G.A. 1984. Interspecies Relationships in Acute Tox-
fcity of Chemicals to Aquatic Organisms. Env. Tax. and Chem.
3(1):47-60.
54. 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. Ed. R.D.
Cardwell, R. Purdy, and R.C. Banner. American Society for
Testing and Materials, Philadelphia, PA.
55. U.S. EPA. Biological Criteria, National Program Guidance for
Surface Waters. EPA 440/5-90-004, April1990.
56. 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.
57. Karr, J.R., et. al. 1986. Assessing Biological Integrity in Running
''- Waters: A Method and Its Rationale. III. Nat. Hist. Survey
Special Publ. 5.
58. Ohio EPA. 1987. Biological Criteria for the Protection of Aquatic
: Life: Volumes 1, 2, and 3. Division of Water Quality Moni-
toring and Assessment. Columbus, OH.
59. Lenat, D.R. 1988. Water Quality Assessments of Streams
Using a Qualitative Collection Method for Benthic
Macroinvertebrates. J.N. Am. Benthol. Soc. 7:222.
I • . .
60. Schackleford, B. 1988. Rapid Bioassessments of
Macroinvertebrate Communities: Biocriteria Development.
Arkansas Dept. Poll. Contr. and Ecol., Little Rock, AR.
61. Maine DEP. 1987. Methods for Biological Sampling and Analy-
sis of Maine's Waters. Maine Bureau of Water Quality Con-
| trol.
62. 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.
63. U.S. EPA. 1989. Method for Conducting Laboratory Toxicity
'', Degradation Evaluations with Complex Effluents. Battelle Re-
; port. March 1989.
64. U.S. EPA. 1989. Pollutants of Concern. Office of Marine and
Estuarine Protection, Washington, DC.
65. Alabaster,)., and R. Lloyd, eds. 1982. Water Quality Criteria
for Fish. 2d ed. Butterworths, London.
66. U.S. EPA. Draft 1991. Guidance on Assessment and Control of
Bioconcentratable Contaminants in Surface Waters. Office of
Water Enforcement and Permits, Washington, DC.
ADDITIONAL REFERENCES
U.S. EPA. 1990. Intralaboratory Study of the Short-Term Chronic
Test with Mysidopsis Bahia. Office of Marine and Estuarine
Protection, Washington, DC.
U.S. EPA. 1988. Proceedings of the First National Workshop on
Biological Criteria. 1988. U.S. EPA, Region 5: Chicago,
Illinois. EPA 905/9-89/003.
U.S. EPA. 1984. Technical Support Document for Conduct-
ing Use Attainability Analysis. Office of Water Regulations
and Standards, Washington, DC.
28
-------
2. WATER QUALITY CRITERIA AND STANDARDS
2.1 INTRODUCTION
The foundation of a water quality-based toxics control program
consists of the State water quality standards applicable to the
waterbody. The following discussion describes the regulatory and
technical considerations for application of water quality stan-
dards.
2.1.1 Overview at Water Quality Standards
A water quality standard defines the water quality goals of a water
body, or portion thereof, by designating the use or uses to be
made of the water, by setting criteria necessary to protect the
uses, and by establishing antidegradation policies and implemen-
tation procedures that serve to maintain and protect water qual-
ity. States adopt water quality standards to protect public health
or welfare, enhance the quality of water, and serve the purposes
of the Clean Water Act (CWA). "Serve the purposes of the Act"
(as defined in Sections 101 (a), 101(a)(2), and 303(c) of the Act)
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.
The CWA describes various uses of waters that are considered
desirable and should be protected. These uses include public
water supply, recreation, and propagation of fish and wildlife. The
States are free to designate more specific uses (e.g., cold water
and warm water aquatic life), or to designate uses not mentioned
in the CWA, with the exception that waste transport and assimila-
tion is not an acceptable designated use (see 40 CFR 131.10(a)).
EPA's regulations emphasize the uses specified in CWA Section
101 (a)(2), but do not preclude other beneficial uses and subcat-
egories of uses as determined by the State.
When designating uses, States should give careful consideration
to whether uses that will support the "fishable and swimmable"
goal of Section 101 (a)(2) are attainable. If the State does not
designate uses in support of this goal, the State must perform a
use attainability analysis under Section 131.10(j) of the standards
regulation. States should designate uses for the waterbody that
the State determines can be attained in the future. "Attainable
uses" are those uses (based on the State's system of water use
classification) that can be achieved when effluent limits under
CWA Section 301(b)(1)(A) and (B) and Section 306 are imple-
mented for point source discharges and when cost-effective and
reasonable best management practices are implemented for
nonpoint sources. The Water Quality Standards regulation speci-
fies the conditions under which States may remove uses or estab-
lish subcategories of uses. Among these are that the State must
provide opportunity for public hearing. In addition, uses that
have been attained in the waterbody on or after November 28,
1975, whether or not they are included in the water quality
standards, may not be removed unless a use requiring more
stringent criteria is added. These uses are the "existing uses" as
defined in 40 CFR 131.3(e). Also, uses that are attainable, as
defined above, may not be removed. Removal of a "fishable/
swimmable" use, or adoption of a subcategory of a "fishable/
swimmable" use that requires less stringent criteria, requires the
State to conduct a use attainability analysis. Technical guidance
on conducting use attainability analyses is available from EPA
(e.g., Chapter 3 of the Water Quality Standards Handbook (1983)
[1 ], and Technical Support Manual: Waterbody Surveys and Assess-
ments for Conducting Use Attainability Analyses (1983) [2],
In the Water Quality Standards regulation, Section 131.11 en-
courages States to adopt both numeric and narrative criteria.
Aquatic life criteria should protect against both short-term (acute)
and long-term (chronic) effects. Numeric criteria particularly are
important where the cause of toxicity is known or for protection
against pollutants with potential human health impacts or
bioaccumulation potential. Numeric water quality criteria also
may be the best way to address nonpoint source pollution prob-
lems. Narrative criteria can be the basis for limiting toxicity in
waste discharges where a specific pollutant can be identified as
causing or contributing to the toxicity but there are no numeric
criteria in the State standards or where toxicity cannot be traced
to a particular pollutant. Section 131.11 (a)(2) requires States to
develop implementation procedures that explain how the State
will ensure that narrative toxics criteria are met.
EPA's water quality standards regulation requires each State to
adopt, as part of its water quality standards, an antidegradation
policy consistent with 40 CFR 131.12 and to identify the methods
it will use for implementing the policy. Activities covered by the
antidegradation policy and implementation methods include both
point and nonpoint sources of pollution. Section 131.12 effec-
tively sets out a three-tiered approach for the protection of water
quality.
"Tier I" (40 CFR 131.12(a)(1)) of antidegradation maintains and
protects existing uses and the water quality necessary to protect
these uses. An existing use can be established by demonstrating
that fishing, swimming, or other uses have actually occurred since
November 28,1975, or that the water quality is suitable to allow
such uses to occur, whether or not such uses are designated uses
for the waterbody in question. (Compare Sections 131.3(e) and
131.3(f) of the existing regulation.) For example, in an area
where shellfish are propagating and surviving in a biologically
suitable habitat, the shellfish use is existing, whether or not people
are harvesting the shellfish. The aquatic life protection use is a
broad category requiring further explanation, which may be found
in the Water Quality Standards Handbook.
-------
Tier II" (Section 131.12(a)(2)) protects the water quality in wa-
ters whose quality is better than that necessary to protect "fishable/
swimmable" uses of the waterbody. 40 Cffi 131.12(a)(2) requires
that certain procedures be followed and certain showings be
made before lowering water quality in high-quality waters. These
showings may be called an "antidegradation review." In no case
may water quality on a Tier II waterbody be lowered to the level at
which existing uses are impaired. The Tier II protection usually is
applied on a parameter-by-parameter basis (called the defini-
tional approach to Tier II). This approach is applied on a case-by-
case basis so that, if the level of any parameter is better than water
quality standards for that waterbody, then an antidegradation
review will be performed for any activity that could reduce the
level of that parameter.
Outstanding national resource waters (ONRWs) are provided the
highest level of protection under the antidegradation policy (Tier
III); no degradation is allowed. ONRWs include the highest-
quality waters of the United States. However, the ONRW
antidegradation classification also offers special protection for
waters of "exceptional ecological significance," i.e., those
waterbodies that are important, unique, or sensitive ecologically,
but whose water quality, as measured by the traditional param-
eters such as dissolved oxygen or pH, may not be particularly
high. Waters of exceptional ecological significance may also
Include waters whose characteristics cannot be described ad-
equately by traditional parameters (such as wetlands and estuaries).
States may, at their discretion, adopt certain policies in their
standards affecting the application and implementation of stan-
dards. For example, policies concerning mixing zones, variances,
low-flow exemptions, and schedules of compliance for water
quality-based permit limits may be adopted. Although these are
areas of State discretion, EPA retains authority to review and
approve or disapprove such policies (see 40 CFR 131.13). Guid-
ance on these subjects is available from EPA's Office of Water
Regulations and Standards, Criteria and Standards Division.
2,1.2 Water Quality Standards and State Toxics Control
Programs
Applicable requirements for State adoption of water quality crite-
ria for toxicants vary depending upon the toxicant. The reason
for this is that the 1983 water quality standards regulation and the
1987 amendments to the CWA (Pub. L. 100-4) include more
specific requirements for the particular toxicants listed in CWA
Section 307(a). For regulatory purposes, EPA has translated the
65 compounds and families of compounds listed in Section 307(a)
into 126 specific substances that EPA refers to as priority toxic
pollutants. The 126 priority toxic pollutants are listed in Appendix
A of 40 CFR Part 423. Because of the more specific requirements
for priority toxic pollutants, it is convenient to organize the re-
quirements applicable to State adoption of criteria for toxicants
Into three categories:
• Requirements applicable to priority toxic pollutants that
have been the subject of CWA Section 304(a)(1) criteria
guidance
• Requirements applicable to priority toxic pollutants that
have not been the subject of CWA Section 304(a)(1) criteria
guidance and
[• Requirements applicable to all other toxicants (i.e.,
nonpriority toxic pollutants).
The criteria requirements applicable to priority toxic pollutants
(i.e., the first two categories above), are specified in CWA Section
30?(c)(2)(B). On December 2, 1988, EPA sent "Guidance for
State Implementation of Water Quality Standards for CWA Sec-
tion 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
priprity toxic pollutants for which EPA has published national
criteria; (2) adopt numeric criteria, for only those priority toxic
pollutants and those stream segments where the discharge or
presence of the pollutant could reasonably be expected to inter-
fere with designated uses; or (3) adopt a specific procedure in the
standards to "translate" the State's narrative "free from toxics"
stahdard to derived numeric criteria. , ,
The transmittal memorandum for the Section 303(c)(2)(B) na-
tional guidance expresses the Office of Water position regarding
priority toxic pollutants that may "reasonably be expected" to
interfere with designated uses. That memorandum and guidance
established a rebuttable presumption that any information indi-
cating that such pollutants are discharged or present in surface
waters (now or in the future) is sufficient justification to require
adoption or derivation of numerical criteria. The goal is not just to
identify pollutants that are already impacting surface waters, but
rather to identify pollutants that may be impacting surface waters
noyv, or have the potential to do so in the future. Lack of detailed
or widespread monitoring data is not an acceptable basis to omit
numerical (or derived numerical) criteria from water quality stan-
dards under Options 2 and 3. Even a limited amount of monitor-
ing data indicating the discharge or presence of priority toxic
pollutants in surface waters is sufficient basis to conclude that
numerical (or derived numerical) criteria are necessary.
Wljere States select an Option 2 or 3 approach, States must
include, as part of the rationale supporting the adopted stan-
dards, the information used in determining which priority toxic
pollutants require criteria. Where there is uncertainty about the
need for criteria for specific priority toxic pollutants, the State
should adopt (or derive) criteria for such pollutants so as to err on
the side of environmental protection and pollution prevention.
This approach is appropriate given the general lack of monitoring
data for priority toxic pollutants; it will provide maximum protection
to the environment by anticipating, rather than reacting to, water
quality problems.
For priority toxic pollutants for which EPA has not issued Section
304(a)(1) criteria guidance, CWA Section 303(c)(2)(B) requires
States to adopt criteria based on biological monitoring or assess-
ment methods. The phrase "biological monitoring or assessment
methods" includes (1) whole effluent toxicity control methods,
(2) biological criteria methods, or (3) other methods based on
biological monitoring or assessment. The phrase "biological
monitoring or assessment methods" in its broadest sense also
includes criteria developed through translator procedures. This
broad interpretation of that phrase is consistent with EPA's policy
of applying chemical-specific, biological, and whole effluent tox-
icity methods independently in an integrated toxics control pro-
gram. It also is consistent with the intent of Congress to expand
30
-------
State standards programs beyond chemical-specific approaches.
Where EPA has not issued Section 304(a) criteria guidance, but
available laboratory toxicity (bioassay) data are sufficient to sup-
port derivation of chemical-specific criteria, States should consider
deriving and adopting numeric criteria for such priority toxic
pollutants. This is particularly important where other compo-
nents of a State's narrative criterion implementation procedure
(e.g., whole effluent toxicity controls or biological criteria) may
not ensure full protection of designated uses. For some pollutants,
a combination of chemical-specific and other approaches is nec-
essary (e.g., pollutants where bioaccumulation in fish tissue or
water consumption by humans is a primary concern).
":>''•' .
Criteria requirements applicable to toxicants that are not priority
toxic pollutants (i.e., the third category above), are specified in
the 1983 water quality standards regulation (see 40 CFR131.11).
Under these requirements, States must adopt criteria based on
sound scientific rationale that cover sufficient parameters to pro-
tect designated uses. Both numeric and narrative criteria are
'addressed by these requirements.
Numeric criteria are required where such criteria are necessary to
protect designated uses. Numeric criteria to protect aquatic life
should be developed to address both short-term (acute) and
long-term (chronic) effects. Saltwater species, as well as freshwa-
ter species, must adequately be protected. Adoption of numeric
criteria 'is particularly important for toxicants known to be impair-
ing surface waters and for toxicants with potential human health
impacts (e.g., those with high bioaccumulation potential). Hu-
man health should be protected from exposure resulting from
consumption of water and fish or other aquatic life (e.g., mussels,
crayfish). Numeric water quality criteria also are useful in address-
ing nonpoint source pollution problems.
In evaluating whether chemical-specific numeric criteria for toxi-
cants are required, States should consider whether other ap-
proaches (such as whole effluent toxicity criteria or biological
controls) will ensure full protection of designated uses. As men-
tioned above, a combination of independent approaches may be
required in some cases to support the designated uses and com-
ply with the requirements of the water quality standards regula-
tion (e.g., pollutants where bioaccumulation in fish tissue or water
consumption by humans is a primary concern).
To supplement numeric criteria for toxicants, all States also have
adopted narrative criteria for toxicants. Such narrative criteria are
statements that describe the desired water quality goal, such as
the following:
All State waters must, at all times and flows, be free from
• substances that are toxic to humans or aquatic life.
EPA considers that the narrative criteria apply to all designated
uses at all flows unless specified otherwise in a State's water
quality standards. EPA also believes that no acutely toxic condi-
tion may exist in any State waters regardless of designated use (54
FR 23875).
Narrative criteria can be the basis for establishing chemical-spe-
cific limits for waste discharges where a specific pollutant can be
^•identified as causing or contributing to the toxicity and the State
has not adopted chemical-specific numeric criteria. Narrative
criteria also can be the basis for establishing whole effluent toxic-
ity controls required by EPA regulations at 40 CFR 122.44(d)(1 )(v).
To ensure that narrative criteria for toxicants are attained, the
water quality standards regulation requires States to develop
implementation procedures (see 40 CFR 131.11(a)(2)). Such
implementation procedures (Box 2-1) should address all mecha-
nisms used by the State to ensure that narrative criteria are
attained. Because implementation of chemical-specific numeric
criteria is a key component of State toxics control programs,
narrative criteria implementation procedures must,describe or
reference the State's procedures to implement such chemical-
specific numeric criteria (e.g., procedures for establishing chemi-
cal-specific permits limits under the NPDES permitting program).
Implementation procedures also must address State programs to
control whole effluent toxicity and may address programs to
implement biological criteria, where such programs have been
developed by the State. Implementation procedures therefore
serve as umbrella documents that describe how the State's vari-
ous toxics control programs are integrated to ensure adequate
protection for aquatic life and human health and attainment of
the narrative toxics criterion. In essence, the procedure should
apply the "independent application" principle, which provides for
independent evaluations of attainment of a designated use based
on chemical-specific, whole effluent toxicity, and biological crite-
ria methods (see Chapter 1, Reference 56).
EPA encourages, and may ultimately require, State implementa-
tion procedures to provide for implementation,of biological crite-
ria. However, the regulatory basis for requiring whole effluent
toxicity controls is clear. EPA regulations at 40 CFR 122.44(d)(1 )(v)
require NPDES permits to contain whole effluent toxicity limits
where a permittee has been shown to cause, have the reasonable
potential to cause, or contribute to an in-stream excursion of a
narrative criterion. Implementation of chemical-specific controls
also is required by EPA regulations at 40 CFR 122.44(d)(1). State
implementation procedures should, at a minimum, specify or
reference methods to be used in implementing chemical-specific
and whole effluent toxicity-based controls, explain how these
methods are integrated, and specify needed application criteria.
In addition to EPA's regulation at 40 CFR Part 131, EPA has regu-
lations at 40 CFR 122.44 that cover the National Surface Water
Toxics Control Program. These regulations intrinsically,are linked
to the requirements to achieve water quality standards, and spe-
cifically address the control of pollutants both with and without
numeric criteria. For example, Section 122.44(d)(1 )(vi) provides
the permitting authority with several .options for establishing
effluent limits when a State does not have a chemical-specific
numeric criteria for a pollutant present in an effluent at a concen-
tration that causes or contributes to a violation of the State's
narrative criteria.
2.2 GENERAL CONSIDERATIONS
2.2.1 Magnitude, Duration, and Frequency
As stated earlier, criteria are specifications of water quality de-
signed to ensure protection of the designated use. EPA criteria are
31
-------
Box 2-1. Components of an Ideal State Implementation Procedure
• Specific, scientifically defensible methods by which the State will implement its narrative toxics standard for all
toxicants, including:
- Methods for chemical-specific criteria, including methods for applying chemical-specific criteria in per-
mits, developing or modifying chemical-specific criteria via a "translator procedure" (defined and
discussed below), and calculating site-specific criteria based on local water chemistry or biology
- Methods for developing and implementing whole effluent toxicity criteria and/or controls
- Methods for developing and implementing biological criteria.
• Integration of these methods in the State's toxics control program (i.e., how the State will proceed when the
specified methods produce conflicting or inconsistent results).
• Application criteria and information that are needed to app|y numerical criteria, for example:
- Methods the State will use to identify thosepollutants to be regulated in a specific discharge
- An incremental cancer risk level for carcinogens
- Methods for identifying compliance thresholds inpermits where calculated limits are below
detection
- Methods for selecting appropriate hardness, pH, and temperature variables for criteria
expressed as functions
- Methods or policies controlling the size and in-zone quality of mixing zones
- Design flows to be used in translating chemical-specific numeric criteria for aquatic life and human
health into permit limits [
- Other methods and information that will be needed to apply standards on a case-by-case basis.
developed as national recommendations to assist States in devel-
oping their standards and to assist in interpreting narrative stan-
dards. EPA criteria or guidance consist of three components:
• Magnitude—How much of a pollutant (or pollutant param-
eter such as toxicity), expressed as a concentration, is allow-
able.
• Duration—The period of time (averaging period) over which
the instream concentration is averaged for comparison with
criteria concentrations. This specification limits the dura-
tion of concentrations above the criteria.
• Frequency—How often criteria can be exceeded.
A typical aquatic life water quality criteria statement contains a
concentration, averaging period, and return frequency, 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 (3) jig/L more than once every
three years on the average and if the one-hour average
concentration does not exceed (4) |jg/L more than once
every three years on the average.
In this example generic statement, the following terms are in-
serted at:
I (1) — either "freshwater" or "saltwater"
, (2) — the name of the pollutant
' (3) — the lower of the chronic-effect or residue-based
concentrations as the criterion continuous con-
i centration (CCC)
(4) — the acute effect-based criterion maximum con-
centration (CMC).
Deffning water quality criteria with an appropriate duration and
frequency of excursions helps to ensure that criteria appropriately
are jconsidered in developing wasteload allocations (WLAs), which
are then translated into permit requirements. Duration and fre-
quency may be defined in the design stream flow appropriate to
the icriterion. However, in these cases, the State should provide
an evaluation that the selected design stream flow approximates
the recommended duration and frequency.
32
-------
2.2.2 Mixing Zones
It is not always necessary to meet all water quality criteria within
the discharge pipe to protect the integrity of the waterbody as a
whole. Sometimes it is appropriate to allow for ambient concen-
trations above the criteria in small areas near outfalls. These areas
are called mixing zones. Since these areas of impact, if dispropor-
tionately large, could potentially adversely impact the productiv-
ity of the waterbody, and have unanticipated ecological conse-
quences, they should be carefully evaluated and appropriately
limited in size. As our understanding of pollutant impacts on
ecological systems evolves, there may be cases identified where
no mixing zone is appropriate.
To ensure mixing zones do not impair the integrity of the
waterbody, it should be determined that the mixing zone will not
cause lethality to passing organisms and, considering likely path-
ways of exposure, that there are no significant human health risks.
One means to achieve these objectives is to limit the size of the
area affected by the mixing zones.
For application of two-number aquatic life criteria, there may be
up to two types of mixing zones (Figure 2-1). In the zone
immediately surrounding the outfall, neither the acute nor the
chronic criterion is met. The acute criterion is met at the edge of
this zone. In the next mixing zone, the acute, but not the
chronic, criterion is met. The chronic criterion is met at the edge
of the second mixing zone.
In the general case, where a State has both acute and chronic
aquatic life criteria, as well as human health criteria, indepen-
dently established mixing zone specifications may apply to each
of the three types of criteria. The acute mixing zone may be sized
to prevent lethality to passing organisms, the chronic mixing zone
Chronic criteria met
Figure 2-1. Diagram of the Two Parts of the Mixing Zone
sized to protect the ecology of the waterbody as a whole, and the
health criteria mixing zone sized to prevent significant human
risks. For any particular pollutant from any particular discharge,
the magnitude, duration, frequency, and mixing zone associated
with each of the three types of criteria will determine which one
most limits the allowable discharge.
Mixing zone allowances will increase the mass loadings of the
pollutant to the waterbody, and decrease treatment require-
ments. They adversely impact immobile species, such as benthic
communities, in the immediate vicinity of the outfall. Because of
these and other factors, mixing zones must be applied carefully,
so as not to impede progress toward the CWA goals of maintain-
ing and improving water quality. EPA recommendations for
allowances for mixing zones, and appropriate cautions about
their use, are contained in this section.
The CWA allows mixing zones at the discretion of the State [1 ].
EPA recommends that States have a definitive statement in
their standards on whether or not mixing zones are allowed.
Where mixing zones provisions are part of the State standards,
the State should describe the procedures for defining mixing
zones.
To determine that a mixing zone is sized appropriately for aquatic
life protection, water quality conditions within the mixing zone
may be compared to laboratory-measured or predicted toxicity
bench marks as follows:
It is not necessary to meet chronic criteria within the
mixing zone, only at the edge of the mixing zone.
Conditions within the mixing zone would thus not be
adequate to ensure survival, growth, and reproduction
of all organisms that might otherwise attempt to reside
continuously within the mixing zone.
If acute criteria (CMC derived from 48- to 96-hour expo-
sure tests) are met throughout the mixing zone, no
lethality should result from temporary passage through
the mixing zone. If acute criteria are exceeded no more
than a few minutes in a parcel of water leaving an outfall
(as assumed in deriving the Section 4.3.3 options for an
outfall velocity of 3 m/sec, and a size of 50 times the
discharge length scale), this likewise assures no lethality
to passing organisms.
If a full analysis of concentrations and hydraulic resi-
dence times within the mixing zone indicates that or-
ganisms drifting through the plume along the path of
maximum exposure would not be exposed to concen-
trations exceeding the acute criteria when averaged
over the 1 -hour (or appropriate site-specific) averaging
period for acute criteria, then lethality to swimming or
drifting organisms ordinarily should not be expected,
even for rather fast-acting toxicants. In many situations,
travel time through the acute mixing zone must be less
than roughly 15 minutes if a 1 -hour average exposure is
not to exceed the acute criterion.
Where mixing zone toxicity is evaluated using the probit
approach described in the water quality criteria
"Bluebook" [3], or using models of toxicant accumula-
33
-------
tion and action in organisms (described by Mancini [4]
or Erickson et al. [5]), the phenomenon of delayed mor-
tality should be taken into account before judging the
mixing zone concentrations to be safe.
The above recommendations assume that the effluent is repul-
sive, such that free-swimming organisms would avoid the mixing
zones. While most toxic effluents are repulsive, caution is neces-
sary in evaluating attractive mixing zones of known effluent toxic-
ity, and denial of such mixing zones may well be appropriate. It
also is important to ensure that concentration isopleths within any
plume will not extend to restrict passage of swimming organisms
Into tributary streams.
In all cases, the size of the mixing zone and the area within certain
concentration isopleths should be evaluated for their effect on the
overall biological integrity of the waterbody. If the total area
affected by elevated concentrations within all mixing zones com-
bined is small compared to the total area of a waterbody (such as
a river segment), then mixing zones are likely to have little effect
on the integrity of the waterbody as a whole, provided that they
do not impinge on unique or critical habitats. EPA has developed
a multistep procedure for evaluating the overall acceptability of
mixing zones [6],
For protection of human health, the presence of mixing zones
should not result in significant health risks, when evaluated using
reasonable assumptions about exposure pathways. Thus, where
drinking water contaminants are a concern, mixing zones should
not encroach on drinking water intakes. Where fish tissue resi-
dues are a concern (either because of measured or predicted
residues), mixing zones should not be projected to result in
significant health risks to average consumers offish and shellfish,
after considering exposure duration of the affected aquatic or-
ganisms in the mixing zone, and the patterns of fisheries use in
the area.
While fish tissue contamination tends to be a far-field problem
affecting entire waterbodies rather than a narrow-scale problem
confined to mixing zones, restricting or eliminating mixing zones
for bioaccumulative pollutants may be appropriate under condi-
tions such as the following:
• Mixing zones should be restricted such that they do not
encroach on areas often used for fish harvesting particularly
of stationary species such as shellfish.
• Mixing zones might be denied where such denial is used as
a device to compensate for uncertainties in the protective-
ness of the water quality criteria or uncertainties in the
assimilative capacity of the waterbody.
2.3 WATER QUALITY CRITERIA 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 Referi
ence 7 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 for 48- to 96-hour
toxicity tests on aquatic animals are available, they are used to
derive the acute criterion. If sufficient data on the ratio of acute to
chronic toxicity concentrations are available, they are used to
derive the chronic or long-term exposure criteria. If justified, one
or both of the criteria may be related to another water quality
characteristic, such as pH, temperature, or .hardness. Separate
criteria are developed for freshwaters and saltwaters.
,. . , . tb- .. • ^,,-'
The water quality standards regulation allows States to develop
numerical criteria or modify EPA's recommended criteria to, ac-
count for site-specific or other scientifically defensible 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 [1], When a criterion must be developed for a chemi-
cal for which a national criterion has not been established,, the
regulatory authority should refer to the Guidelines for Deriving Cri-
terib for Aquatic Life and Human Health (see 45 FR 79341, Novem-
ber 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 CMC to protect against acute, (short-
term) effects and a CCC to protect against chronic (long-term)
effects. EPA derives acute criteria from 48- to 96-hour tests of
lethality or immobilization. EPA derives chronic criteria from
longer-term (often greater than 28-day) tests that measure sur-
vival, growth, reproduction, or in some'cases, bioconcentration.
Most State standards include numerical criteria for a limited num-
ber of individual toxic chemicals. Therefore, evaluation and con-
trol of toxic pollutants is based on maintenance of the designated
use and often relies on the narrative criterion prohibiting toxic
substances in toxic amounts. The adverse effects of concern will
depend on the designated use and the chemical. Bioaccumulation
of chemicals in aquatic organisms, toxicity to these organisms,
the! potential for additivity, antagonism, synergism, and persis-
tence 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 mentioned 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 requires 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 develop-
ing water quality-based controls and therefore a case-by-case
interpretation of the narrative criterion may be necessary.
34
-------
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. At present, there are no
national criteria developed under CWA Section 304(a) for whole
effluent toxicity. Acute and chronic toxicity units (Tils) 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 or chronic), the test
species, 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 protec-
tion, the CMC should be set at 0.3 acute toxic unit (TUa) to the
most sensitive of at least three test species.
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 LCsg endpoint of an acute toxicity test (50
percent mortality) to an LCi value (virtually no mortality). Spe-
cifically, a factor of 0.3 was found to include 91 percent of
observed LC-| to LCso ratios in 496 effluent toxicity tests as illus-
trated in Figure 2-2. This figure presents effluent toxicity data
from many years of toxicity testing of both industrial and munici-
pal effluents by the Environmental Services Division, U.S. EPA
Region IV, Athens, Georgia.
130 -
120 -
110 -
100 -
90 -
a-80"
i 70 -
|f 60 -
so -
40 -
20 -
10 -
o-
4
o
^
§
12
o
CM
9
5
125 • ,
89
67
40
29
g
o
TT
o
in
o
CD
8
86
o
CO
42
2
f==j
O
O) O
2 22 22 2 2 £
ey 55 5 S ?^'5 5
G>oddc5ddp
LCi/LC5o Ratio
Figure 2-2. LCitoLCjo Ratios for Effluent Toxicity Tests
For chronic protection, the CCC should be set at 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 prevent any
chronic toxicity in the receiving water outside the mixing zone.
2.3.4 Duration far 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. Or-
ganisms in the receiving water are not experiencing constant,
steady exposure but rather are experiencing fluctuating exposures,
including periods of high concentrations, which may have adverse
effects. Thus, EPA's criteria indicate a time period over which
exposure is to be averaged, as well as a, maximum concentration,
thereby limiting the duration of exposure to elevated, concentra-
tions.
For acute criteria, EPA recommends an averaging period of 1
hour. That is, to protect against acute effects, the 1 -hour average
exposure should not exceed the CMC. The 1 -hour acute averag-
ing period was derived primarily from data on response time for
toxicity to ammonia, a fast-acting toxicant. The 1-hour averaging
period is expected to be fully protective for the fastest-acting
toxicants, and even more protective for slower-acting toxicants.
Scientifically justifiable alternative (site-specific) averaging periods
can be derived from (1) data relating toxic response to exposure
time, if coupled with considerations of delayed mortality (mortality
occurring .after exposure has ended), or (2) models of toxicant
uptake and action, such as presented by Erickson [5] and Mancini
etal. [4].
In practice, 1-day periods are the shortest periods for which WLA
modelers and enforcement personnel have adequate data. Attain-
ment of 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 chronic criteria, EPA recommends an averaging period of 4
days. That is, the 4-day average exposure should not exceed the
CCC. Different chronic averaging periods could be derived, de-
pending on the nature of the pollutant and the toxic endpoint of
concern (e.g., the rate of uptake and accumulation, and the mode
of action).
The toxicity tests used to establish the national criteria are con-
ducted using steady exposure to toxicants usually for at least 28
days. The test concentrations do not fluctuate as much as typically
occurs instream. 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 sig-
nificant consideration involved in setting duration criteria is how
long the exposure concentration can be above the criterion con-
centration without unacceptably affecting the endpoint of the test
(e.g., survival, growth, or reproduction). EPA selected the 4-day
averaging period based on the shortest duration in which chronic
effects are sometimes observed for certain species and toxicants,
and thus should be fully protective even for the fastest-acting
toxicants.
35
-------
2.3.5 Frequency for Single Chemicals and mole Effluent
Toxlcity
To predict or ascertain the attainment of criteria it is necessary to
specify the allowable frequency for exceeding the criteria. This is
because it is statistically impossible to project that criteria will
never be exceeded. As ecological communities are naturally
subjected to a series of stresses, the allowable frequency of pollut-
ant stress may be set at a value that does not significantly increase
the frequency or severity of all stresses combined.
EPA recommends a once in 3-year average frequency for
excursions of both acute and chronic criteria. These recom-
mendations apply to both chemical-specific and whole effluent
approaches. However, the allowable frequency depends on site-
specific factors. To implement alternative frequencies, site-spe-
cific factors (see Appendix D) or other data or analyses should be
taken into account. In all cases, the recommended frequency
applies to actual ambient concentrations, and excludes the influ-
ence of measurement imprecision.
EPA established its recommended frequency as part of its Guidelines
for Deriving Criteria, last issued in 1985 [8]. EPA selected the 3-
year return interval with the intent of providing a degree of
protection roughly equivalent to a 7Q10 design flow condition,
and with some consideration of rates of ecological recovery from
a variety of severe stresses. Because of the nature of the ecological
recovery studies available, the severity of criteria excursions could
not be related rigorously to the resulting ecological impacts.
Nevertheless, EPA derives its criteria intending that a single mar-
ginal criteria excursion (i.e., a slight excursion over a 1-hour
period for acute or over a 4-day period for chronic) would result in
little or no ecological effect and require little or no time for
recovery. If the frequency of marginal criteria excursions is not
high, it can be shown that the frequency of severe stresses,
requiring measurable recovery periods, would be extremely small.
EPA thus expects the 3-year return interval to provide a very high
degree of protection.
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 WLAs and
water quality-based effluent limits. As discussed in Chapter 4, the
duration and frequency recommendations are implemented di-
rectly if a dynamic modeling approach is used to develop WLAs
and permit limits. However, if a steady-state approach is used, a
design condition 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. Chapter 4
discusses these recommended design flows.
Traditionally, most water quality-based permits for point source
discharges had been tied to the 7-day, once in 10-year, low-flow
conditions. The reason for this is 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.
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, if
adopted into or used to interpret State water quality standards,
may be used as the basis for total maximum daily load (TMDL)
after considering seasonal flow and loading scenarios. The con-
centration, duration, and frequency provisions of EPA's water
quality criteria can be modified to account for site-specific condi-
tions. As States have started using the new two-number water
quality criteria for perennial as well as intermittent discharges
such as combined sewer overflows, urban runoff, etc., their proper
use in the context of the TMDL/WLA process needs to be empha-
sized.
2.4 WATER QUALITY CRITERIA FOR HUMAN HEALTH
PROTECTION
2.4.1 Overview
(
There are a number of key elements of State water quality stan-
dards and implementation procedures relevant to human health
protection. States must determine ambient standards for the two
primary human exposure routes, fish consumption and drinking
water. States must then establish whether mixing zones will
apply, and, if so, determine the design conditions.
State 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 tem-
perature variables for criteria. However, if State standards do not
specify these items, then the permitting authority must develop
water quality-based effluent limits based upon either an interpre-
tation of the State's water quality standards or EPA's criteria and
procedures.
i
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 and reference ambient
concentrations. 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 As-
sessment 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 reference
ambient concentrations (RACs) for those pollutants lacking EPA
human health criteria. Although the same procedures are used to
develop criteria and RACs, only those values that are subjected to
the'regulatory process of regional, State, and public comment
canlbe.considered "criteria." RACs may be applied as site-specific
interpretations of narrative standards and as a basis for permit
limits under 40 CFR 122.44 (d)(1)(vi).
36
-------
Procedures also are provided in this chapter to develop values
called reference tissue concentrations (RTCs) that can be used in
assessing or monitoring fish tissues for unacceptable residues.
2.4.2 Magnitude anil Duration
Water quality criteria for human health contain only a single
expression of allowable magnitude; a criterion concentration gen-
erally to protect against long-term (chronic) human health effects.
Currently, national policy and prevailing opinion in the expert
community dictate that the duration for human health criteria for
carcinogens be derived assuming lifetime exposure, taken to be a
70-year time period. The duration of exposure assumed in deriv-
ing criteria for noncarcinogens is more complicated due to a wide
variety of endpoints: some developmental (and thus age-specific
and perhaps sex-specific), some lifetime, and some, such as or-
ganoleptic effects, not duration-related at all. Thus, appropriate
durations depend on the individual noncarcinogenic pollutants
and the endpoints or adverse effects being considered.
2.4.3 Human Exposure Considerations
A complete human exposure evaluation for toxic pollutants of
concern for bioaccumulation would not only encompass esti-
mates of exposures due to fish consumption, but also exposure
due to background concentrations and other exposure routes,
including recreational and occupational contact, dietary intake
from other than fish, inhalation of air, and drinking water. How-
ever, the focus of this document is on ingestion of contaminated
fish tissue, a direct human exposure route of potentially significant
risk. (For the human health sections in this document the term
"fish" generally is used to mean both fish and shellfish.) The
consumption of contaminated fish tissue is of serious concern
since the presence of even extremely low ambient concentrations
of bioaccu-mulative pollutants (sublethal to aquatic life) in surface
waters, can result in residue concentrations in fish tissue that can
pose a human health risk. Other exposure route information
should be considered and incorporated in human exposure evalu-
ations to the extent it is available.
Levels of actual human exposures from consuming contaminated
fish vary depending upon a number of case-specific consumption
factors. These factors include type offish species consumed/type
of fish tissue consumed, tissue lipid content, consumption rate
and pattern, and food preparation practices. In addition,' de-
pending on the spatial variability in the fishery area, the behavior
of the fish species, and the point of application of the RAC 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, the average exposure 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 bioaccumulative contaminants likely to be ingested.
Whole body shellfish consumption presumably means ingestion
of the entire burden of bioaccumulative 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 much of the iipid material in which bioaccumulative
contaminants tend to concentrate.
2.4.4 Fish Consumption Values
EPA's human health criteria have assumed a human body weight
of 70 kg and the consumption of 0.0065 kg of fish and shellfish
per day. Based on data collected in 1973-1974, the national per
capita consumption of freshwater and estuarine fish was esti-
mated to average 6.5 g/day. Per capita consumption of all
seafood (including marine species) was estimated to average 14.3
g/day. The 95th percentile for consumption of all seafood by
individuals over a period of 1 month was estimated to be 42
g/day [9]. The mean lipid content of fish tissue consumed in this
study was estimated to be 3.0 percent [10].
Currently, four levels of fish consumption are provided in the EPA
guidance manual, Assessing Human Health Risk from Chemically
Contaminated Fish and Shellfish. These are:
• 6.5 g/day to represent an estimate of average consump-
tion offish and shellfish from estuarine and freshwaters
by the entire U.S. population [9]. This fish consumption
level is based on the average of both consumers and
nonconsumers of fish.
• 20 g/day to represent an estimate of the average con-
sumption of fish and shellfish from marine, estuarine,
: and freshwaters by the U.S. population [11]. This average
fish consumption level also includes both consumers and
nonconsumers of fish.
• 165 g/day to represent consumption of fish and shellfish
' from marine, estuarine, and freshwaters by the 99.9th
percentile of the U.S. population consuming the most fish
or seafood [12].
• 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 United States (EPA Risk
Assessment Council assumption based on data from the
U.S. Department of Agriculture Nationwide Food Con-
sumption Survey of 1977-1978).
EPA currently is 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 mean values appropriate to the population at large,
and values appropriate for those individuals who consume a
relatively large proportion offish in their diets (maximally exposed
individuals).
Many States use the EPA's 6.5 g/day consumption value. How-
ever, some States (e.g., Wisconsin, Louisiana, Illinois, and Arizona)
use the above mentioned 20 g/day value. For salt waters Delaware
uses another EPA value, 37 g/day [13]. In general, EPA recom-
mends that the consumption values used in deriving RACs from
the formulas in this chapter reflect the most current relevant and/
or site-specific information available.
37
-------
2.4.5 Bloaccumulatlon Considerations for Reference Ambient
Concentration Development
The ratio of the contaminant concentrations in fish tissue versus
water is termed either the bioconcentration factor (BCF) or the
bloaccumulation factor (BAF). Bioconcentration is defined as
involving contaminant uptake from water only (not from food).
Bioaccumulation is defined as involving contaminant uptake from
both water and food. Under laboratory conditions, measure-
ments of tissue/water partitioning generally are considered to
involve uptake from water only. On the other hand, both process
are likely to apply in the field since the entire food chain is
exposed.
Table 2-1 shows the ratio of the BAF to the BCF as a function of
the trophic level of the aquatic organism, and the log P (log
octanol-water partition coefficient) of the chemical [14]. .The
BAF/BCF ratio ranges from 1 to 100, with the highest ratios
applying to organisms in higher trophic levels, and to chemicals
with log P close to 6.5. For chemicals with log P values greater
than about 7, there is some uncertainty regarding the degree of
bioaccumulation, but generally, trophic level effects appear to
decrease due to slow transport kinetics of these chemicals in fish,
the growth rate of the fish, and the chemical's relatively low
bioavailability.
Care must be taken in assigning the trophic level since certain fish
species may inhabit one source area of contaminated food for
only a portion of their life. Under such conditions of migration,
fish would only receive a small portion of the chemical and never
come into equilibrium. In addition, trophic level for a given fish
species will vary with life stage and structure of the food chain.
In this document, bioaccumulation considerations are integrated
Into the RAC equations in Sections 2.4.7 and 2.4.8 by using food
chain multipliers (FMs) with the BCF. The bioaccumulation and
bioconcentration factors for a chemical are related as follows:
BAF = FM x BCF
By incorporating the FM and BCF terms into the RAC equations,
bioaccumulation is addressed.
In this process, bioaccumulation considerations are included by
incorporating the FM term with the BCF in calculating the RTCs
and RACs. In Table 2-1, FM values derived from the work of
Thomann [14,15] are listed according to log P value and trophic
level of the organism. Trophic level 4 organisms are typically
the most desirable species for sport fishing and therefore,
FMs for trophic level 4 generally should be used in the equa-
tions for calculating RTCs and RACs. In those very rare situations
where only lower trophic level organisms are found, e.g., possibly
oyster beds, an FM for a lower trophic level may be used in
calculating the RTCs and RACs.
Measured BAFs (especially for those chemicals with log P values
above 6.5) reported in the literature should be used when avail-
able. To use experimentally measured BAFs in calculating the
RAC or RTC, the (FM x BCF) term, is replaced by the BAF in the
equations in Sections 2.4.7 and 2.4.8. Relatively few BAFs have
been measured accurately and reported, and their application to
sites other than the specific ecosystem where they were devel-
Table 2-1. Estimated Food Chain Multipliers
- \
LogP
3.5
3.6
3.7
3.8
3.9
4.0
4.1
; 4.2
4.3
4.4
. 4.5
4.6
4.7
4.8
4.9
5.0
5.1
5.2
5.3 ,
5.4
5.5
5.6
5.7
5.8
5.9
6.0
6.1
6.2
6.3
6.4
6.5
j >6-5
Trophic
2
1.0
1.0
1.0
1.0
1.0
1.1
1.1 .
1 .1
1.1
1.2
1.2 '
1.2
1.3
1.4
1.5
1.6
1.7
1.9
2.2
2.4
2.8
3.3
3.9
4.6
5.6
6.8
8.2
10
13
15
19
19.2*
Levels
3
1.0
1.0
1.0
1.0
1.0
1.0
1.1
1.1
1.1
1.1
1.2
1.3
1.4
1.5
1.8
2.1
2.5
3.0
3.7
4.6
5.9
7.5
9.8
13
17
21
25
29
34
39
45
45*
4
1.0
1.0
1.0
1.0
1.0
1.0
1.1
1.1
1.1
1.1
1.2
1.3
1.4
1 .6
2.0
2.6
3.2
4.3
5.8
8.0
11
16
23
33
47
67
75
84
92
98
100
100*
* T|hese recommended FMs are conservative estimates; FMs for log P
values greater than 6.5 may range from the values given to as low as
0.1 for contaminants with very low bioavailability.
oped is problematic and subject to uncertainty. The option also is
available to develop BAFs experimentally, but this will be ex-
tremely resource intensive if done on a site-specific basis with all
this necessary experimental and quality controls.
2.4.6 Updating Human Health Criteria anil Generating RACs
Using IRIS
ERA recommends using the most current risk information
when updating criteria and generating RACs. The Integrated
Risk Information System (IRIS) is an electronic online data base of
the U.S. EPA that provides chemical-specific risk information on
trite relationship between .chemical exposure and estimated hu-
man health effects [16]. Risk assessment information contained in
the IRIS, except as specifically noted, has been reviewed and
agreed upon by an interdisciplinary group of scientists represent-
ing various program offices within the Agency and represent an
Agencywide consensus. Risk assessment information and values
are updated monthly and are approved for Agencywide use.
38
-------
The IRIS is .intended to make risk assessment information readily
available to those individuals who must perform risk assessments
and also to increase consistency among risk assessment/risk man-
.agement decisions. The IRIS is available to Federal and some State
and local environmental agencies through the EPA's electronic
MAIL system and also is available to the public through the Public
Health Network and TOXNET. Since IRIS is designed to be a
. publicly available data base, interested parties may submit studies
• or documents for consideration by the appropriate interdiscipli-
nary review group for chemicals currently on the IRIS or scheduled
for review. Information regarding the submission of studies of
chemicals may be obtained from the IRIS Information Submission
Desk. In addition to chemical-specific summaries of hazard and
: dose-response assessments, the IRIS contains a series of sections
identified by service codes that serve as a user's guide as well as
provide background documentation on methodology. Addi-
tional information is available from IRIS Users Support: 513/FTS
684-7254.
The IRIS contains two types of quantitative risks values: reference
dose (RfD) and the carcinogenic potency estimate or slope factor.
The RfD (formerly known as the acceptable daily intake or ADI) is
• the human health hazard assessment for noncarcinogenic (target
• organ) effects. The carcinogenic potency estimate (formerly known
•as ql*) represents the upper bound cancer causing potential
.. resulting from lifetime exposure to a substance. The RfD or the
oral carcinogenic potency estimate are used in the derivation of
an RAC. Appendix H contains the supporting information for
derivation of RfDs. ,
EPA periodically updates risk assessment information including
RfDs, cancer potency estimates, and related information on con-
taminant effects, and reports the current information on IRIS.
Since the IRIS contains the Agency's most recent quantitative risk
assessment values, current IRIS values should be used in develop-
ing new RACs. 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
would require inserting the current RfD or carcinogenic potency
estimate on the IRIS into the appropriate equation in Section
2.4.7 or 2.4.8.
Figure 2-3 shows the procedure for determining an updated
criterion or RAC using IRIS data. If a chemical has both carcino-
genic and noncarcinogenic effects, i.e., both a cancer potency
estimate and RfD, the carcinogen RAC formula in Section
2.4.8 should be used as it will result in the more stringent RAC
of the two.
2.4.7 Calculating RACs for Noncarcinogens
The RfD is an estimate of the daily exposure to the human
population that is likely to be without appreciable risk of causing
EPA's
water .quality \ NO
criterion ,
vailable?
Evaluate other
sources of data:
HEAST,
Risk Assistant,
drinking water
MCLs, fish
consumption
advisory levels,
FDA action levels,
etc. i
Figure 2-3. Procedure for Revising an EPA Human Health Criterion or Developing a Reference Ambient Concentration
39
-------
deleterious effects during a lifetime. The RfD is expressed in units
of mg toxicant per kg human body weight per 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 stud-
ies or animal exposure (mammal LDso) studies. [Note: LOAEL
and NOAEL refer to animal and human toxicology and are there
fore distinct from the aquatic toxicity terms "no observed effect
concentration" (NOEC) and the "lowest observed effect concen-
tration" (LOEC)]. Uncertainty factors are then applied to the
NOAEL or LOAEL to account for uncertainties in the data associ-
ated with variability among individuals, extrapolation from non-
human test species to humans, data on other than long-term
exposures, and the use of an LOAEL [17]. An additional uncertainty
may be applied to account for significant weakness or gaps in the
data base.
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. Simi-
larly, while exposure at or below the RfD reduces the probability,
it does not guarantee the absence of effects in all persons. The
RfDs contained in the IRIS are values that represent EPA's consen-
sus (and have uncertainty spanning perhaps an order of magni-
tude).
For noncarcinogenic effects, an updated criterion or an RAC can
be derived using the following equation:
CorRAC(mg/l) = (RfD x WT) - (DT + IN") x WT
Wl + [FC x L x FM x BCF]
where
C s updated water quality criterion (mg/l)
RAC = reference ambient concentration (mg/l)
RfD s 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 body human weight/day)
IN = inhalation exposure
(mg toxicant/kg body human weight/day)
Wl = average human adult water intake
(2 liters/day)
FC = daily fish consumption (kg fish/day)
L = ratio of lipid fraction of fish tissue consumed to
3 percent
FM = food chain multiplier (from Table 3-1)
BCF = bioconcentration factor (mg toxicant/kg fish divided
by mg toxicant/I water) for fish with 3 percent lipid.
If the receiving waterbody is not used as a drinking water source,
the factor Wl can be deleted. Where dietary and/or inhalation
exposure values are unknown, these factors may be deleted from
the above calculation. For identified noncarcinogenic chemicals
without known RfDs, extrapolation procedures can be used to
estimate the RfD (see Appendix H).
2.4.8 Calculating RACs for Carcinogens
Any human health criterion for a carcinogen is based on at least
three interrelated considerations: potency, exposure, and risk
characterization. States may make their own judgments on each
of these factors within reasonable scientific bounds, but docu-
mentation to support their judgments must be clear and in the
public record.
Maximum protection of human health from the potential effects
of Exposure to carcinogens via contaminated fish would require
an RAC of zero. The zero level is based upon the assumption of
nonthreshold effects (i.e., no safe level exists below which any
increase in exposure does not result in an increase in the risk of
cancer) for carcinogens. However, because safety does not re-
quire the absence of all risk, a numerical estimate of risk (in |ig/l)
that corresponds to a given level of risk for a population of a
specified size is selected instead. A cancer risk level is defined as
the- number of new cancers that may result in a population of
specified size due to an increase in exposure (e.g., 10"° risk level =
1 additional cancer in a population of 1,000,000). Cancer risk is
calculated by multiplying the experimentally derived cancer po-
tency estimate by the concentration of the chemical in the fish
arid the average daily human consumption of contaminated fish.
The risk for a specified population (e.g., 1,000,000 people or 10"6)
is then calculated by dividing the risk level by the specific cancer
risj<. EPA's ambient water quality criteria documents provide risk
levels ranging from 10"5 to 10"7 as examples.
len the cancer potency estimate, or slope factor (formerly
known as the q1*), is derived using animal studies, high-dose
exposures are extrapolated to low-dose concentrations and ad-
justed to a lifetime exposure period through the use of a linearized
multistage model. The model calculates the upper 95 percent
confidence limit of the slope of a straight line that the model
postulates to occur at low doses. When based on human (epide-
miological) data, the slope factor is based on the observed in-
crease in cancer risk, and is not extrapolated. For deriving RACs
fof carcinogens, the oral cancer potency estimates or slope factors
from the IRIS are used.
I
It Jis important to note that cancer potency factors may overesti-
mate actual risk. Such potency estimates are subject to great
uncertainty due to two primary factors: (1) adequacy of the
cancer data base (i.e., human versus animal data) and (2) limited
information regarding the mechanism of cancer causation. The
actual risk may be much lower, perhaps as low as zero, particu-
larly for those chemicals for which human carcinogenicity infor-
mation is lacking. Risk levels of 10-5,1Q"6, and 10'7 are often used
by States as minimal risk levels in interpreting their standards. EPA
considers risks to be additive, i.e., the risk from individual chemi-
cals is not necessarily the overall risk from exposure to water. For
example, an individual risk level of 10~6 may yield a higher overall
risk level if multiple carcinogenic chemicals are present.
For carcinogenic effects, the RAC can be determined by using the
following equation:
Cor RAC (mg/l) =
(RLxWD
ql* [Wl + FC x L x (FM x BCF)]
40
-------
where :
C = updated water quality criterion (mg/l)
RAC = reference ambient concentration (mg/l)
RL = risk level (10-x)
WT = weight of an average human adult (70 kg)
q1 * = carcinogenic potency factor (kg day/mg)
Wl = average human adult water intake (2 liters/day)
FC = daily fish consumption (kg fish/day)
L = ratio of lipid fraction of fish tissue consumed to
3 percent
FM = food chain multiplier (from Table 3-2)
BCF = bioconcentration factor (mg toxicant/kg fish divided
by mg toxicant/I water) for fish with 3 percent lipid.
If the receiving waterbody is not used as a drinking water source,
the factor Wl can be deleted. For identified carcinogenic chemi-
cals without known cancer potency estimate values, extrapolation
procedures can be used to estimate the cancer potency.
2.4.9 Deriving Quantitative Risk Assessments in the Absence
of IRIS Values
The RfDs or cancer potency estimates comprise the existing dose
factors for developing RACs. When IRIS data are unavailable,
quantitative risk level 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 con-
centrations of exposure protective of fish consumption by hu-
mans. -Where no procedure exists, factors may be based upon
extrapolation from mammalian or other data using IRIS docu-
mentation or information available from other EPA risk data bases.
Also, where no other information or procedure exists, drinking
water maximum contaminant levels (MCLs) or Food and Drug
Administration (FDA) action levels may be used as guidance in
developing numerical estimates.
2.4.10 Deriving Reference Tissue Concentrations for Monitoring
Fish Tissue
Where fish tissue evaluations have been used for assessing human
health risks, or, perhaps, used for additional routine monitoring
where a chemical is below analytical detection limits, the follow-
ing formulas may be used to calculate an RTC. Readers also
should consult EPA's Assessing Human Health Risks from Chemically
Contaminated Fish and Shellfish [17].
The basic equations for deriving RTC (in mg/kg) use the same
parameters as in equations 2.1 and 2.2, where BCF is normalized
at 3.0 percent lipid:
For noncarcinogens:
RTC (mg/kg) = (RFD x WD - (DT + IN) x WT
[WI/(BCF x FM x L)] + FC
For carcinogens:
RTC (mg/kg) = RLxWT
ql * [WI/(BCF x FM x L) + FC]
The above equations should be corrected for site-specific lipid
content and faioaccumulation factors where data are available.
Again, some States have established their own procedures whereby
RTCs can be developed based upon extrapolation of acute and/or
chronic animal data to safe concentrations protective of fish
consumption by humans. Where additional risk information is
needed, an RTC could be based upon other information such as
drinking water MCLs or FDA action levels.
2.5 BIOLOGICAL CRITERIA
As discussed in Chapter 1, to fully protect aquatic habitats and
provide more comprehensive assessments of aquatic life use at-
tainment/nonattainment, States are 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 (see Chapter 1,
Reference 55).
2.5.1 Regulatory Bases for Biocriteria
The primary statutory basis for EPA's policy that States should
develop biocriteria is found in Sections 101 (a) and 303(c)(2)(B) of
the Water Quality Act of 1987. Section 101 (a) of the CWA gives
the general authority for biological criteria. It establishes as the
objective of the Act the restoration and maintenance of the
chemical, physical, and biological integrity of the Nation's waters.
To meet this objective, water quality criteria should address bio-
logical integrity. Section 101 (a) includes the interim water quality
goal for the protection and propagation of fish, shellfish, and
wildlife.
Section 304 of the Act provides the legal basis for the develop-
ment of informational criteria, including biological criteria. Spe-
cific directives for the development of regulatory biocriteria can
be found in Section 303, which requires EPA to develop criteria
based on biological assessment methods when numerical criteria
are not established.
Once biocriteria formally are 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
as criteria where such chemical-specific criteria have not been,
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 pro-
tective of the three methods is used) as site-specific conditions
and assessment objectives dictate.
Section 304(a) directs EPA to develop and publish water quality
criteria and information on methods for measuring water quality
and establishing water quality criteria for toxic pollutants on bases
other than pollutant-by-pollutant, including biological monitor-
ing and assessment methods that assess:
• The effects of pollutants on aquatic community compo-
nents ("... plankton, fish, shellfish, wildlife, plant life ...")
41
-------
and community attributes (". . . biological community
diversity, productivity, and stability . . ."); in any body of
water.
Factors necessary "... to restore and maintain the chemi-
cal, physical, and biological integrity of all navigable waters
..." for"... the protection of shellfish, fish, and wildlife for
classes and categories of receiving waters "
These elements serve as an interactive network that is particularly
important during early development of .biological criteria where
rapid accumulation of information is effective for refining both
designated uses and developing biological criteria values.
2.6 SEDIMENT CRITERIA
2.5.2 Development and Implementation of Blocriteria
Biocriteria are numerical values or narrative expressions that de-
scribe the reference biological integrity of aquatic communities
inhabiting unimpaired waters of a designated aquatic life use.
The biological communities in these waters represent the best
attainable conditions. The reference site conditions then become
the basis for developing biocriteria for major surface water types
(streams, rivers, lakes, wetlands, estuaries, or marine waters).
Biological criteria support designated aquatic life use classifica-
tions for application in State standards. Each State develops its
own designated use classification system based on the generic
uses cited in the Act (e.g., protection and propagation of fish,
shellfish, and wildlife). Designated uses are intentionally general.
However, States may develop subcategories within use designa-
tions to refine and clarify the use class. Clarification of the use
class is particularly helpful when a variety of surface waters with
distinct characteristics fit within the same use class, or do not fit
wellinto any category.
For example, subcategories of aquatic life uses may be on the
basis of attainable habitat (e.g., cold versus warmwater communi-
ties dominates by bass versus catfish). Special uses also may be
designated to protect particularly unique, sensitive, or valuable
aquatic species, communities, or habitats.
Resident biota integrate multiple impacts over time and can
detect impairment from known and unknown causes. Biocriteria
can be used to verify improvement in water quality in response to
regulatory efforts and detect continuing degradation of waters.
They provide a framework for developing improved best manage-
ment practices for nonpoint source impacts. Numeric criteria can
provide effective monitoring criteria for inclusion in permits.
The assessment of the biological integrity should include mea-
sures of the structure and function of an aquatic 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 development and
Implementation of biological criteria requires:
• Selecting unimpaired (minimal impact) surface waters, to
use as the reference condition foi each designated use
• Measuring the structure and function of aquatic communi-
ties in reference surface waters to establish biological crite-
ria
• Establishing a protocol to compare the biological criteria to
biota in impacted waters to determine whether impairment
has occurred.
2.S.1 Current Developments In Sediment Criteria
While ambient water quality criteria are playing an important role
injassuring a healthy aquatic environment, they alone have not
been sufficient to ensure appropriate levels of environmental
protection. Sediment contamination, which can involve deposi-
tion 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
CWA Sections 104, and 304(a)(1) and (2) as follows:
Section 104(n)(1) authorizes the Administrator to establish
national programs that study the effects of pollution, in-
cluding sedimentation, in estuaries on aquatic life. ,-.,-.•
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 sedi-
mentation for varying types of receiving waters.
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 that
address the concerns of the Section 404(b)(1) guidelines for
discharges of dredged or fill material under the CWA or the
Marine Protection Research, and Sanctuaries Act, they also could
be incorporated into those regulations.
t •
2.S.2 Approach to Sediment Criteria Development
Over the past several years, sediment criteria development activi-
ties have centered on evaluating and developing the equilibrium
partitioning approach for generating sediment criteria. The equi-
librium partitioning approach focuses on predicting the chemical
interaction between sediments and contaminants. Developing
an understanding of the principal factors that influence the sedi-
ment/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 also is working to develop
a standardized sediment toxicity test that could be used with or
independently of sediment criteria and could be used to assess
chronic effects in freshwater and marine water.
42
-------
Equilibrium partitioning (EqP) sediment quality criteria (SQC)
are the EPA's best recommendation of the concentration of a
substance in sediment that will not unacceptably affect benthic
organisms or their uses.
Methodologies for deriving effects based SQC vary for different
classes of compounds. For non-ionic organic chemicals the meth-
odology requires normalization to organic carbon. A methodol-
ogy for deriving effects based sediment criteria for metal con-
taminants is under development and is expected to require nor-
malization to acid volatile sulfide. EqP SQC values can be derived
for varying degrees of uncertainty and levels of protection thus
permitting use for ecosystem protection and remedial programs.
2*6.3 Application of Sediment Criteria
SQC would provide a basis for making more informed decisions
on the environmental impacts of contaminated sediments. Exist-
ing sediment assessment methodologies are limited in their ability
to identify chemicals of concern, responsible parties, degree of
contamination, and zones of impact. EPA believes that a compre-
hensive approach using SQC and biological test methods is pre-
ferred in order to make the most informed decisions.
Sediment criteria will be particularly valuable in site monitoring
applications where sediment contaminant concentrations are
gradually approaching a criteria over time. Sediment criteria also
are valuable as a preventative tool to ensure that point and
nonpoint sources of contamination are controlled to ensure
uncontaminated sediments remain uncontaminated. Also, com-
parison of field measurements to sediment criteria will be a reli-
able method for providing early warning of a potential problem.
An early warning would provide an opportunity'to take corrective
action before adverse impacts occur. For the reasons mentioned
above it has been identified that SQC are essential to resolving key
contaminated sediment and source control issues in the Great
Lakes.
Specific Applications
Specific applications of sediment criteria are under development.
The primary use of EqP-based sediment criteria will be to assess
risks associated with contaminants in sediments. The various
offices and programs concerned with contaminated sediment
have different regulatory mandates and thus, have different needs
and areas for potential application of sediment criteria. Because
each regulatory need is different, EqP-based sediment quality
criteria designed specifically to meet the needs of one office or
program may have to be implemented in different ways to meet
•the needs of another office or program.
One mode of application of EqP-based numerical SQC would be
in a tiered approach. In such an application, when contaminants
in sediments exceed the SQC, the sediments Would be considered
as causing unacceptable impacts. Further testing may or may riot
be required depending on site-specific conditions and the degree
in. which a criteria has been violated. (No additional testing
would be required in locations where contamination significantly
exceeds a criterion. Where sediment contaminant levels are close
to a criteria, additional testing may be necessary.) Contaminants
in a sediment at concentrations less than the sediment criteria
would not be of concern. However, in some cases the sediment
could not be considered safe because they may contain other
contaminants above safe levels for which no sediment criteria
exist. In addition, the synergistic, antagonistic, or additive effects
of several contaminants in the sediments may be of concern.
Additional testing in other tiers of an evaluation approach, such as
bioassays, could be required to determine if the sediment is safe.
It is likely that such testing would incorporate site-specific consid-
erations. Examples of specific applications of sediment criteria
after they are developed are as follows:
• Establish permit limits to ensure that uncontaminated sedi-
ments remain uncontaminated or sediments already con-
taminated have an opportunity to cleanse themselves. This
would occur only after criteria and the means to tie point
sources to sediment deposition are developed.
• Establish target levels for nonpoint source causes of sedi-
ment contamination. •
• For remediation activities, SQC would be valuable in identi-
fying:
- Remediation need •'••••-
- Spatial extent of remediation area
- Benefits derived from remediation activities •
- Responsible parties . ...:•,
- Impacts of depositing contaminated sediments in
water environments •
- Success of remediation activities.
• In tiered testing sediment evaluation processes, sediment
criteria and biological testing procedures work very well
together. : - ' ' ' -
2.6.4 Sediment Criteria Status
Science Advisory Board Review
The Science Advisory Board has completed its review and issued a
favorable report on the EqP for assessing sediment quality. The
Subcommittee found the EqP "to have major strengths in its
foundation in chemical-theory, its ease of calculation, and its
ability to make use of existing data... The conceptual basis of the
approach is supported by the Subcommittee; however, its appli-
cation at this time is limited." ;
The Science Advisory Board also identified the need for "a better
understanding of the uncertainty around the assumptions inher-
ent in the approach, including assumptions Of equilibrium,
bioavailability, and kinetics, all critical to the application of the
EqP." An uncertainty analysis and a guidance document to assist
in the regulatory application of developed criteria are under de-
velopment and expected to be completed in 1991.
43
-------
Sediment Criteria Documents and Application Guidance
EPA efforts at producing sediment criteria documents are being
directed first toward phenanthrene, fluoranthene, DDT, dieldrin,
acenaphthene and endrin. Efforts also are being directed to
produce a guidance document, Application of Sediment Quality
Criteria for the Protection of Aquatic Life, scheduled for release in
1991.
Methodology for Developing Sediment Criteria for Metal
Contaminants
EPA is proceeding with a methodology for developing sediment
criteria for metal contaminants, with key work focused on identi-
fying and understanding the role of acid volatile sulfides (AVS) in
controlling the bioavailability of metal contaminants. A variety of
field and laboratory verification studies are underway to add
additional support to the methodology. Standard AVS sampling
and analytical procedures are under development [18]. Presenta-
tion of the metals methodology to the Science Advisory Board for
review is scheduled for 1991.
44
-------
CHAPTER 2
REFERENCES
1. U.S. EPA. 1984. Water Quality Standards Handbook. Office of Water
Regulations and Standards (WH-585), Washington, DC.
2. U.S. EPA. 1984. Technical Support Document for Conducting
Use Attainability Studies. Office of Water Regulations and
Standards (WH-585), Washington, DC.
3. National Academy of Science. 1973. Water Quality Criteria
1972. EPA-R3-73-033 or NTIS PB236199.
4. Mancini, J.L 1983. A Method for Calculating Effects on
Aquatic Organisms of Time-Varying Concentrations. Wa-
ter Res. 17:1355-61.
5 Erickson, R., C. Kleiner, J. Fiandt, and T. Highland. 1989.
Report on the Feasibility of Predicting the Effects of Fluctuating
Concentrations on Aquatic Organisms. U.S. EPA, ERL-Duluth.
6. Brungs, W.A. 1986. Allocated Impact Zones for Areas of Nan-
Compliance. U.S. EPA, Region 1. Water Management
Division, Boston, MA.
7. U.S. EPA. 1985. Guidelines for Deriving Numerical National
Water Quality Criteria for the Protection of Aquatic Organisms
and Their Uses. NTIS PB85-227049.
8. U.S. EPA. 1987. Integrated Risk Information System. Volume
2, Chemical Files. Office of Health and Environmental
Assessment. EPA/600/8-86/032b. March 1987b.
9. Javitz, H.S. 1980. Letter to H. Kahn (EPA). SRI International.
10. Stephan, C.E. 1980. Per Capita Consumption of Non-
Marine Fish and Shellfish. Memorandum to J. Stara. U.S.
EPA, ERL-Duluth.
11. U.S. Department of Agriculture. 1984. Agricultural Statistics.
U.S. DA, Washington, DC.
12. Finch, R. 1973. The MECCA Project: Effects of Regulatory
Guidelines on the Intake of Mercury from Fish. Fisheries
Bulletin 71:615-26.
13. U.S. EPA. 1989. Exposure Factors Handbook. OHEA, Wash-
ington, DC. EPA/600/8-89/043.
14. Thomann, R.V. 1989. Bioaccumulation Model of Organic
Chemical Distribution in Aquatic Food Chains. Environ. Sci.
Technol. 23:699-707.
15. Thomann, R.V. 1987. A Statistical Model of Environmental
Contaminants Using Variance Spectrum Analysis. Report to
National Science Foundation. NTIS PB88-235130/A09.
16. U.S. EPA. 1987. Integrated Risk Information System. Volume
1, Supportive Documentation. Office of Health and Envi-
ronmental Assessment. EPA/600/8-86/032a. March 1987a.
17. U.S. EPA. 1988. Guidance Manual for Assessing Human Health
Issues from Chemically Contaminated Fish and Shellfish.
Submitted by Robert A. Pastorok, PTI Environmental Ser-
vices, Bellevue, WA; for Battelle New England Marine Re-
search Laboratory, Duxbury, MA.
18. University of Delaware Department of Civil Engineering. 1990.
Development of an Analytical Method of the Determination of
Add Volatile Sulfide in Sediment. Submitted by Battelle to
U.S. EPA, Criteria and Standards Division, Washington, DC.
45
-------
-------
3. EFFLUENT CHARACTERIZATION
3.1 INTRODUCTION
Once the applicable designated uses and water quality criteria for
a waterbody are determined, the effluent must be characterized
and the permitting authority must determine the need for permit
limits to control the discharge. The purpose of effluent character-
ization is to determine whether the discharge causes, has the
reasonable potential to cause, or contributes to an excursion of
numeric or narrative water quality criteria. Once the permitting
authority determines that a discharge causes, has the reason-
able potential to cause, or contributes to the excursion of
water quality criteria, the permitting authority must develop
permit limits that will control the discharge. At a minimum, the
permitting authority must make this determination at each permit
reissuance. The effluent characterization procedures described in
the following sections apply only to the water quality-based ap-
proach, not to end-of-the-pipe technology-based controls.
Although many waterbodies receive discharges from only single
point sources, permitting authorities will also occasionally encoun-
ter receiving waters where several dischargers are in close proxim-
ity. In such situations, the permitting authority may find that each
discharger alone does not cause, have the reasonable potential to
cause, or contribute to an excursion above water quality criteria.
Yet, the dischargers may collectively cause, have the reasonable
potential to cause, or contribute to an excursion. Under these
circumstances, limits must be developed for each discharger
to protect against collective excursions of applicable water
quality standards consistent with the Environmental Protec-
tion Agency's (EPA) existing regulations in 40 CfR
122.44(d)(1)(ii) for controlling multiple discharges. The terms
"cause," "reasonable potential to cause," and "contribute to" are
the terms used in the National Pollutant Discharge Elimination
System (NPDES) regulations for conditions under which water
quality-based limits are required. Permitting authorities are re-
quired to consider each of these concepts when performing efflu-
ent characterizations.
This chapter is divided into two parts: Section 3.2, Determining
the Need for Permit Limits Without Effluent Data, and Section 3.3,
Determining the Need for Permit Limits With Effluent Data. Sec-
tion 3.3 includes effluent characterization for whole effluent toxic-
ity and for specific chemicals (including those for human health
protection) and is based on the cumulative experience gained by
EPA, States, publicly owned treatment works (POTWs), and indus-
try when implementing the water quality-based approach to toxics
control. The effluent bioconcentration evaluation procedures de-
scribed in the section on human health are currently draft and are
subject to further validation before being used. Until the proce-
dures are fully developed, reviewed, and finalized, permitting
authorities should not use them to characterize effluents.
3.1.1 NPDES Regulation Requirements
Effluent characterization is an essential step in determining the
need for an NPDES permit limit. NPDES regulations under 40
CFR 122.44(d)(1) specify the minimum requirements and gen-
eral types of analyses necessary for establishing permit limits.
Each of these regulations is described below.
40CFR122.44(d)(1)(ii)
When determining whether a discharge 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 exist-
ing controls on point and nonpoint sources of pollu-
tion, the variability of the pollutant or pollutant param-
eter 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.
This regulation requires at a minimum the consideration of each
of these elements in determining the need for a limit.
4O CFR 122.44(d) (1) (iii)
When the permitting authority determines, using the
procedures in paragraph (d)(1 )(ii) of this section, that a
discharge causes, has the reasonable potential to cause,
or contributes to an in-stream excursion 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.
Under this regulation, permitting authorities need to investigate
for the existence of pollutants in effluents if there is a numeric
water quality criterion for that pollutant and to implement limits
for those pollutants where necessary.
40CFR122.44(d)(l)(iv)
When the permitting authority determines, using the
procedures in paragraph (d)(1 )(ii) of this section, that a
discharge causes, has the reasonable potential to cause,
or contributes to an in-stream excursion above the
numeric criterion for whole effluent toxicity, the permit
must contain effluent limits for whole effluent toxicity.
Under this regulation, permitting authorities need to investigate
for the existence of whole effluent toxicity in effluents if there is a
numeric water quality criterion for that parameter and to imple-
ment whole effluent toxicity limits where necessary.
47
-------
40 CFfl122.44(d)(1)(v)
Except as provided in this subparagraph, when the
permitting authority determines, using the procedures
in paragraph (d)(1)(ii) of this section, toxicity testing
data, or other information, that a discharge causes, has
the reasonable potential to cause, or contributes to an
in-stream excursion above a narrative criterion within
an applicable State water quality'standard, the permit
must contain effluent limits for whole effluent toxicity.
Limits on whole effluent toxicity are not necessary where
the permitting authority demonstrates in the fact sheet
or statement of basis of the NPDES permit, using the
procedures in paragraph (d)(1)(ii) of this section, that
chemical-specific limits for the effluent are sufficient to
attain and maintain applicable numeric and narrative
State water quality standards.
Under this regulation, permitting authorities need to investigate
for the existence of whole effluent toxicity in effluents. If the
permitting authority can demonstrate that control of specific
chemicals is sufficient to control toxicity to the point of achieving
compliance with the water quality criteria, then chemical-specific
permit limits alone will be sufficient to comply with the regula-
tion.
40CFfi122.44(d)(l)(vi)
Where a State has not established a water quality crite-
rion for a specific chemical pollutant that is present in
an effluent at a concentration that causes, has the
reasonable potential to cause, or contributes to an ex-
cursion above a narrative criterion within an applicable
State water quality standard, the permitting authority
must establish effluent limits using one or more of the
following [three] options:....
Under this regulation, permitting authorities need to investigate
for the existence of specific chemicals in effluents for which trie
State has not adopted numeric criteria, but which may be con-
tributing to aquatic toxicity or impairment of human health.
Narrative criteria apply when numeric criteria do not protect all
the designated or existing uses. For example, the narrative
criteria need to be used to protect human health if a State has
only adopted a numeric criteria for protecting aquatic life. Con-
versely, the narrative criteria need to be used to protect aquatic
life if a State has only adopted a numeric criteria for protecting
human health. Once the permitting authority determines that
one or more specific chemicals in an effluent must be controlled,
the authorities can use EPA's national criteria, develop their own
criteria, or control the pollutant through use of an indicator
pollutant, as provided in subparagraph (d)(1)(vi). In any case,
the permitting authority will need to characterize the effluent in a
manner consistent with the selected approach for controlling the
pollutant. . .
3.1.2 Background for Toxic Effects Assessments on Aquatic
Life and Human Health
Aquatic toxicity effects can be characterized by conducting a
general assessment of the effluent, 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 measured or projected exposure concentration of a toxicant
or 1:he parameter toxicity (when dealing with the whole effluent
toxicity) in the receiving water after mixing. The RWC is calcu-
lated at the edge of a mixing zone if such a zone is allowed by a
State's water quality standards.
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 the chemical-by-
chemical approach currently is practical for assessing and control-
ling human health impacts. Appendix G discusses developing
procedures for assessing human health impacts from whole efflu-
, ents.
i •
A fundamental principle in the development of water quality-
based controls is that the RWC must be less than the criteria that
comprise or characterize the water quality standards. With indi-
vidual toxicants (or the parameter toxicity), the potential for
toxicity in the receiving water is minimized where the RWC is less
than the criterion continuous concentration (CCC), the criterion
maximum concentration (CMC), and the reference ambient con-
centration (RAC). Toxicity becomes maximized where the RWC
exceeds these criteria. Therefore, to prevent impacts to aquatic
life or human health, 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, as indicated below. (The RAC as used
throughout this chapter incorporates EPA human health criteria
and State standards as well.)
RWC < CCC (chronic aquatic life)
RWC < CMC (acute aquatic life)
RWC < RAC (human health)
The water quality analyst will use the same basic components in
the above-described relationship (i.e., critical receiving water flows,
ambient criteria values, measures of effluent quality) for both
effijjent characterization and wasteload allocation (WLA) develop-
ment, albeit from different perspectives. In the case of effluent
characterization, the objective is to project receiving water con-
centrations based upon existing effluent quality to determine
whether or not an excursion above ambient criteria occurs, or has
the reasonable potential to occur. In developing WLAs, on the
other hand, the objective is to fix the RWC at the desired criteria
level and determine an allowable effluent loading that will not
cause excursions above the criteria.
i
Recommendations for projecting the RWC are described within
this chapter. Chapter 4, Exposure Assessment and Wasteload
Allocation, provides recommendations for determining allowable
effluent loadings to achieve established ambient criteria and for
calculating WLAs for establishing permit limits. The procedures
described within Chapter 4 can also be used to calculate the
dilution for analyses within Chapter 3. Chapter 5, Permit Require-
mehts, describes the actual calculation of permit limits after efflu-
ent characterization and loadings, as well as WLAs, are complete.
48
-------
3. 1.3 General Considerations in Effluent Characterization
There are two possible ways to characterize an effluent to deter-
mine the need for effluent limits for the protection of aquatic life
and human health. First, an assessment may be made without
generating effluent data; second, an assessment may be con-
ducted after effluent data have been generated. Regulatory au-
thorities must determine whether a discharge causes, has the
"reasonable potential" to cause, or contributes to an excursion
above an applicable narrative or numeric water quality criterion.
An analysis of "reasonable potential" determines an effluent's
capability to cause such excursions.
In determining the need for a permit limit for whole effluent
toxicity or for 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 involved
species to toxicity testing (for whole effluent), and, where appro-
priate, the dilution of the effluent in the receiving water (40 CFR
The regulatory authority is also required by NPDES regulations to
consider whether technology-based limits are sufficient to main-
tain State water quality standards. There are two possibilities that
will need to be assessed. First, if the limits based on appropriate
treatment technology have already been specified in a previous
permit, and if the facility is operating at the required level, then
historical effluent and receiving water information can be used.
Second, if the facility has yet to achieve the required technology
performance (best available technology or best conventional tech-
nology), the regulatory authority will need to assess the technol-
ogy-based limit for reasonable potential for causing or contribut-
ing to an excursion above the water quality standard.
In addition, the regulatory authority should consider all other
available data and information pertaining to the discharger to
assist in making an informed judgment 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 limit. The authority should employ the prin-
ciple of "independent application" of the data and informa-
tion that characterizes the effluent. In other words, effluent
data alone, showing toxicity at the RWC, may be adequate to
demonstrate the need for a limit for toxicity or for individual
toxicants. Likewise, other factors may form an adequate basis for
determining that limits are necessary. For example, where avail-
able dilution is low and monitoring information shows that toxic
pollutants are frequently discharged at concentrations that have
caused toxicity when discharged from similar facilities, the per-
mitting authority may reason that a whole effluent toxicity limit is
necessary even without whole effluent toxicity data from the
specific facility. In all cases, the decision must be based upon
consideration of factors cited in 40 CFR 122.44(d)(1)(ii)- The
regulatory authority will need to prioritize, on a case-by-case
basis, the importance of all data and information used in making
a determination. To assist in case-by-case determinations, rec-
ommended guidelines for characterizing an effluent for the need
for a permit limit for whole effluent toxicity or individual toxi-
cants are discussed below and summarized in Boxes 3-1 through
3-3.
Box 3-1. Determining "Reasonable Potential" for Excursions Above Ambient Criteria Using
Factors Other than Facility-specific Effluent Monitoring Data
When determining the "reasonable potential" of a discharge to cause an excursion above a State water quality
standard, the regulatory authority must consider all the factors listed in 40 CFR 122.44(d)(1 )(ii). Examples of the
types of information relating to these factors are listed below.
Existing controls on point and nonpoint sources of pollution
• Industry type: Primary, secondary, raw materials used, products produced, best management practices,
control equipment, treatment efficiency, etc.
• Publicly owned treatment work type: Pretreatment, industrial loadings, number of taps, unit processes,
treatment efficiencies, chlorination/ammonia problems, etc.
Variability of the pollutant or pollutant parameter in the effluent
• Compliance history
• Existing chemical data from discharge monitoring reports and applications.
Sensitivity of the species to toxicity testing
• Adopted State water quality criteria, or EPA criteria
• Any available in-stream survey data applied under independent application of water quality standards
• Receiving water type and designated/existing uses
Dilution of the effluent in the receiving water
• Dilution calculations
49
-------
3.2 DETERMINING THE NEED FOR PERMIT LIMITS
WITHOUT EFFLUENT MONITORING DATA FOR A
SPECIFIC FACILITY
If the regulatory authority so chooses, or if the circumstances
dictate, the authority may decide to develop and impose a
permit limit for whole effluent toxicity or for individual toxicants
without facility-specific effluent monitoring data, or prior to the
generation of effluent data. Water quality-based permit limits
can be set for a single toxicant or for whole effluent toxicity based
on the available dilution and the water quality criterion or the
State standard in the absence of facility specific effluent monitor-
ing data. However, in doing so, the regulatory authority must
satisfy all the requirements of 40 CFR 122.44(d)(1)(ii).
When determining whether or not a discharge causes, has the
reasonable potential to cause, or contributes to an excursion of a
numeric or narrative water quality criterion for individual toxi-
cants or for toxicity, the regulatory authority can use a variety of
factors and information where facility-specific effluent monitor-
ing data are unavailable. These factors also should be considered
with available effluent monitoring data. Some of these factors are
the following:
• Dilution—Toxic impact is directly related to available dilu-
tion for the effluent. Dilution is related to the receiving
stream flow and the size of the discharge. The lower the
available dilution, the higher the potential for toxic effect.
If an effluent's concentration at the edge of a mixing zone
in a receiving water is expected to reach 1 percent or
higher during critical or worst-case design periods, then
such an effluent may require a toxicity limit (see discussion
in Section 3.3.3). Assessment of the amount of stream
dilution available should be made at the conditions re-
quired by the water quality standards or, if not specified in
the standards, at the harmonic mean flow and the 7Q10
flow. Figure 3-3 (Pg. 57) shows that, whereas a majority of
NPDES permittees nationwide discharge to areas during
annual mean flow ranging in dilution from 100 to 1,000,
the majority of dischargers fall into the 1 to 10 dilution
range during low-flow conditions.
• Type of industry—Although dischargers should be indi-
vidually characterized because toxicity problems are site-
specific, the primary industrial categories should be of
principal toxicity concern. EPA's treatment technology
data base generally suggests that secondary industrial cat-
egories may have less potential for toxicity than primary
industries. However, based on experience, it is virtually
impossible to generalize the toxicity of effluents with any
certainty. If two plants produce the same type of product,
one effluent may be toxic while the other may not be toxic
due to the type and efficiency of the treatment applied,
general materials handling practices, and the functional
target of the compound(s) being produced.
• Type of POTW—POTWs with loadings from indirect dis-
chargers (particularly primary industries) may be candi-
dates for toxicity limits. However, absence of industrial
input does not guarantee an absence of POTW discharge
toxicity problems. For example, commercial pesticide ap-
plicators often discharge to POTWs, resulting in pesticide
concentrations in the POTWs effluent. Household disposal
of pesticides, detergents, or other toxics may have a similar
effect. 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 characterized for the possibility of chlorine and
ammonia problems.
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 pres-
ence 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
"nonpriority" toxicants that may cause effluent toxicity.
Also, combinations of several toxicants can produce ambi-
ent toxicity where the individual toxicants would not. EPA
regulations at 40 CFR 122.21 (j) require POTWs with design
flows equal to or greater than 1 MGD and POTWs with
approved pretreatment programs, or POTWs required to
develop a pretreatment program, to submit the results of
whole effluent toxicity tests with their permit applications.
These regulations also provide discretion to the permitting
authority to request such data from other POTWs at the
time of permit application.
History of compliance problems and toxic impact—Regu-
latory authorities may consider particular dischargers that
have had difficulty complying with limits on toxicants or
that have a history of known toxicity impacts as probable
priority candidates for effluent toxicity limits.
Type of receiving water and designated use—Regulatory
authorities may compile data on water quality. Examples of
available data include fish advisories or bans, reports of fish
kills, State lists of priority waterbodies, and State lists of
waters that are not meeting water quality standards. Regu-
latory authorities should use this information as a means of
identifying point sources that discharge to impaired
waterbodies and that thus may be contributing to this
impairment. One source of this information is the lists of
waters generated by states to comply with Section 304(1)
regulations at 40 CFR 130.10(d)(6); 50 FR 23897-98, June 2,
1989:
1) Waters where fishing or shellfish bans and/or
advisories are currently in effect or are antici-
pated;
2) Waters where there have been repeated fish
kills 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 re-
cent state section 305(b) report as either "par-
tially achieving" or "not achieving" designated
uses;
50
-------
5) Waters identified by the states under section
303(d) of the Clean Water Act as waters need-
ing water quality-based controls;
6) Waters identified by the state as priority water
bodies;
7) Waters where ambient data indicate potential
or actual excur Jons of water quality criteria
due to toxic pollutants from an industry classi-
fied as a primary industry in Appendix A of 40
CFR Part 122;
8) Waters for which effluent toxicity test results
indicate possible or actual excursions of state
water quality standards, including narrative
"free from" water quality criteria or EPA water
quality criteria where state criteria are not avail-
able;
9) Waters with primary industrial major discharg-
ers where dilution analyses indicate
exceedances of state narrative or numeric wa-
ter quality criteria (or EPA water quality criteria
where state standards are not available) fortoxic
pollutants, ammonia, or chlorine;
10) Waters with POTW dischargers requiring local
pretreatment programs where dilution analy-
ses 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;
11) Waters with facilities not included in the previ-
ous two categories such as major POTWs, and
industrial minor dischargers where dilution
analyses indicate exceedances of numeric or
narrative state water quality criteria (or EPA
water quality 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 wa-
ter quality conditions have been reported by
local, state, EPA or other Federal Agencies, the
private sector, public interest groups, or uni-
versities;
14) Waters identified by the state as impaired in its
most recent Clean Lake Assessments conducted
under 314 of the Clean Water Act; and
15) Surface waters impaired by pollutants from
hazardous waste sites on the National Priority
List prepared under section 105(8)(A) of
CERCLA.
16) Waters judged to be impaired as a result of a
bioassessment/biosurvey.
The presence of a combination of these factors, such as low
available dilution, high-quality receiving water, poor compli-
ance record, and clustered industrial and municipal discharges,
could constitute a high priority for effluent limits.
Regardless, the regulatory authority, if it chooses to impose an
effluent limit after conducting an effluent assessment without
facility-specific monitoring data> will need to provide adequate
justification for the limit in its permit development rationale or
in its permit fact sheet. A clear and logical rationale for the need
for the limit covering all of the regulatory points will' be neces-
sary to defend the limit should it be challenged. In justification
of a limit, EPA recommends that the more information the
authority can acquire to support the limit, the better a
position the authority will be in to defend the limit if neces-
sary. In such a case, the regulatory authority may well benefit
from the collection of effluent monitoring data prior to estab-
lishing the limit. ..••...••.
If the regulatory authority, after evaluating all available informa-
tion on the effluent, in the absence of effluent monitoring data,
is not able to decide whether the discharge causes, has the
reasonable potential to cause, or contributes to, an excursion,
above a numeric or narrative criterion for whole effluent 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 require
the monitoring prior to permit issuance, if sufficient time exists,
or it may require .the testing as a condition of the issued/
reissued permit.
Under these circumstances, the regulatory authority may find it
protective of water quality to include a permit reopener for the
imposition of an effluent limit should the effluent testing estab-
lish that the discharge causes, has the reasonable potential to
cause, or contributes to excursion above a water quality criteria.
A discussion of these options is provided later in this chapter.
3.3 DETERMINING THE NEED FOR PERMIT
LIMITS WITH EFFLUENT MONITORING DATA
3.3.1 General Considerations
When characterizing an effluent for the need for a whole efflu-
ent toxicity limit, and/or an individual toxicant limit, the regula-
tory authority should use any available effluent monitoring
data, together with any information like that discussed under
Section 3.2 above, as the basis for a decision. The regulatory
authority may already have effluent toxicity data available from
previous monitoring, or it may decide to require the permittee
to generate effluent monitoring data prior to permit issuance or
as a condition of the issued permit. EPA regulations at 40 CFR
122.21 (j) require POTWs with design flows equal to or greater
than 1 MGD and POTWs with approved pretreatment pro-
grams, or POTWs required to develop a pretreatment program,
to submit the results of whole effluent toxicity tests with their
permit applications. These regulations also provide discretion
to the permitting authority to request such data from additional
POTWs at the time of permit application.
51
-------
In the instance where the permittee is required to generate data in
advance, data collection should begin 12 to 18 months in advance
of permit development to allow adequate time for conducting
toxicity tests and chemical analyses. The type of data, including
toxicity testing data, should be specified by the regulatory author-
ity at the outset so that decisions on permit actions will not be
delayed. EPA recommends monitoring 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 and (2)
where toxicity is shown, effluent variability can be more clearly
defined. Several basic factors that should be considered in gener-
ating effluent monitoring data are discussed below.
3.3.2 Addressing Uncertainty to Effluent Characterization
by Generating Effluent Monitoring Data
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 (1) a complete data base is available on the effects of acute
and chronic toxicity on many indigenous species, (2) there is a
clear understanding of ecosystem species composition and func-
tional processes, and (3) 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 can be very
resource intensive.
An example of uncertainty that results from limited monitoring
data is if a regulatory authority has only one piece of effluent data
(e.g., an LCso of 50 percent) for a facility. Effluent variability in
such a case, given the range of effluent toxicity variability seen in
other effluents, may range between 20 percent and 100 percent
(see Appendix A). It is impossible to determine from one piece of
monitoring data where in this range the effluent variability really
falls. More monitoring data would need to be generated to
determine the actual variability of this effluent and reduce this
source of uncertainty.
To better characterize the effects of effluent variability and reduce
uncertainty in the process of deciding whether to require an
effluent limit, EPA has developed the statistical approach described
below and in Box 3-2. This approach combines knowledge of
effluent variability as estimated by a coefficient of variation with
the uncertainty due to a limited number of data to project an
estimated maximum concentration for the effluent. The estimated
maximum concentration is calculated as the upper bound of the
expected lognormal distribution of effluent concentrations at a
high confidence level. The projected effluent concentration after
consideration of dilution can then be compared to an appropriate
water quality criterion to determine the potential for exceeding
that criterion and the need for an effluent limit.
The statistical approach has two parts. The first is a characteriza-
tion of the highest measured effluent concentration based on the
desired confidence level. The relationship that describes this is the
following:
pn = (1 - confidence level)1/"
whpre pn is the percentile represented by the highest con-
centration in the data and n is the number of samples. The
following are some examples of this relationship at a 99
perjcent confidence level:
• The largest value of 5 samples is greater than the 40
percentile
• The largest value of 10 samples is greater than the 63
percentile
• The largest value of 20 samples is greater than the 79
percentile
,• The largest value of 100 samples is greater than the 96
percentile.
The second part of the statistical approach is a relationship
between the percentile described above and the selected
upper bound of the lognormal effluent distribution. EPA's
effluent data base suggests that the lognormal distribution
well characterizes effluent concentrations (see Appendix E).
For example, if five samples were collected (which repre-
sents a 40th percentile), the coefficient of variation is 0.6,
and the desired upper bound of the effluent distribution is
the 99th percentile, then the two percentiles can be related
using the coefficient of variation (CV) as shown below:
C99 exp(2.326a - 0.502)
exp(-0.258
-------
Box 3-2. Determining "Reasonable Potential" for Excursions Above
Ambient Criteria Using Effluent Data Only
EPA recommends finding that a permittee has "reasonable potential" to exceed a receiving water quality
standard if it cannot be demonstrated with a high confidence level that the upper bound of the lognormal
distribution of effluent concentrations is below the receiving water criteria at specified low-flow conditions.
Step 1 Determine the number of total observations ("n") for a particular set of effluent data (concentrations or
toxic units [Tils]), and determine the highest value from that data set.
Step 2 Determine the coefficient of variation for the data set. For a data set where n<10, the coefficient of
variation (CV) is estimated to equal 0.6, or the CV is calculated from data obtained from a discharger.
For a data set where n>10, the CV is calculated as standard deviation/mean (see Figure 3-1). For less
than 10 items of data, the uncertainty in the CV is too large to calculate a standard deviation or mean
with sufficient confidence.
Step 3 Determine the appropriate ratio from Table 3-1 or 3-2.
Step 4 Multiply the highest value from a data set by the value from Table 3-1 or 3-2. Use this value with the
appropriate dilution to project a maximum receiving water concentration (RWC).
Step 5 Compare the projected maximum RWC to the applicable standard (criteria maximum concentration,
criteria continuous concentration [CCC], or reference ambient concentration). EPA recommends that
permitting authorities find reasonable potential when the projected RWC is greater than an ambient
criterion.
Example
Consider the following results of toxicity measurements of an effluent that is being characterized: 5 TUC, 2 TUC/ 9 TUC,
and 6 TUC. Assume that the effluent is diluted to 2 percent at the edge of the mixing zone. Further assume that the
CV is 0.6, the upper bound of the effluent distribution is the 99th percentile, and the confidence level is 99 percent.
Step 1 There are four samples, and the maximum value of the sample results is 9 TUC.
Step 2 The value of the CV is 0.6.
Step 3 The value of the ratio for four pieces of data and a CV of 0.6 is 4.7.
Step 4 The value that exceeds the 99th percentile of the distribution (ratio times xmax) after dilution is calcu-
lated as:
[9 TUC x 4.7 x 0.02] = 0.85 TUC.
Step 5 0.85 TUC is less than the ambient criteria concentration of 1.0 TUC. There is ho reasonable
potential for this effluent to cause an excursion above the CCC.
3.3.3 Effluent Characterization for Whole Effluent Toxicity
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 accor-
dance with appropriate site-specific considerations and the rec-
ommendations discussed below. In the past 5 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
freshwater and marine environments.
General Considerations and Assumptions
EPA has revised its initial effluent toxicity data generation recom-
mendations based on three observations made over the last 5
years:
1) Only rarely have effluents discharged by NPDES permittees
been observed to have LCsrjS less than 1.0 percent or no
observed effect concentrations (NOECs) less than 0.1 per-
cent. However, there is always a chance that an effluent
could be toxic at such low effluent concentrations.
53
-------
Table 3-1. Reasonable Potential Multiplying Factors: 99% Confidence Level and 99% Probability Basis
Number of
Samples
1
2
3
4
S
6
7
8
9
10
11
12
13
14
IS
16
17
18
19
20
Coefficient of Variation
0.1 0.2 0.3 0.4 O.S
1.6 2.5 3.9 6.0 9.0
1.4 2.0 2.9 4.0 5.5
1.4 1.9 2.5 3.3 4.4
1.3 1.7 2.3 2.9 3.8
1.3 1.7 2.1 2.7 3.4
1.3 1.6 2.0 2.5 3.1
1.3 1.6 2.0 2.4 2.9
1.2 1.5 1.9 2.3 2.8
1.2 1.5 1.8 2.2 2.7
1.2 1.5 1.8 2.2 2.6
1.2 1.5 1.8 2.1 2.5
1.2 1.4 1.7 2.0 2.4
1.2 1.4 1.7 2.0 2.3
1.2 1.4 1.7 2.0 2.3
1.2 1.4 1.6 1.9 2.2
1.2 1.4 1.6 1.9 2.2
1.2 1.4 1.6 1.9 2.1
1.2 1.4 1.6 1.8 2.1
1.2 1.4 1.6 1.8 2.1
1.2 1.3 1.6 1.8 2.0
0.6 0.7 0.8 0.9 1.0
13.2 18.9 26.5 36.2 48.3
7.4 9.8 12.7 16.1 20.2
5.6 7.2 8.9 11,0 13.4
4.7 5.9 7.2 8.7 10.3
4.2 5.1 6.2 7.3 8.6
3.8 4.6 5.5 6.4 7.5
3.6 4.2 5.0 5.8 6.7
3.3 3.9 4.6 5.3 6.1
3.2 3.7 4.3 5.0 5.7
3.0 3.5 4.1 4.7 5.3
2.9 3.4 3.9 4.4 5.0
2.8 3.2 3.7 4.2 4.7
2.7 3.1 3.6 4.0 4.5
2.6 3.0 3.4 3.9 4.3
2.6 2.9 3.3 3.7 4.1
2.5 2.9 3.2 3.6 4.0
2.5 2.8 3.1 3.5 3.8
2.4 2.7 3.0 3.4 3.7
2.4 2.7 3.0 3.3 3.6
2.3 2.6 2.9 3.2 3.5
1.1 1.2 1.3 1.4 1.5
63.3 81.4102.8128.0157.1
24.9 30.3 36.3 43.0 50.4
16.0 19.0 22.2 25.7 29.4
12.2 14.2 16.3 18.6 21.0
10.0 11.5 13.1 14.8 16.6
8.6 9.8 11.1 12.4 13.8
7.7 8.7 9.7 10.8 12.0
6.9 7.8 8.7 9.6 10.6
6.4 7.1 7.9 8.7 9.6
5.9 " 6.6 A3 8.0 8.8
5.6 6.2 6.8 7.4 8.1
5.2 5.8 6.4 7.0 7.5
5.0 5.5 6.0 6.5 7.1
^.8 5.2 5.7 6.2 . 6.7
4.6 5.0 5.4 5.9 6.4
k.4 4.8 5.2 5.6 6.1
4.2 4.6 5.0 5.4 5.8
jki 4.4 4.8 5.2 5.6
4.0 4.3 4.6 5.0 5.3
3.8 4.2 4.5 4.8. 5.2
1.6 1.7 1.8 1.9 2.0
90.3 227.8 269.9 316.7 368.3
58.4 67.2 76.6 86.7 97.5
33.5 37.7 42.3 47.0 52.0
23.6 26.3 29.1 32.1 35.1
18.4 20.4 22.4 24.5 26.6
15.3 16.8 18.3 19.9 21.5
13.1 14.4 15.6 16.9 18.2
11.6 12.6 13.6 14.7 15.8
10.4 11.3 12.2 13.1 14.0
9.5 10.3 11.0 11.8 12.6
8.8 9.4 10.1 10.8 11.5
8.1 8.8 9.4 10.0 10.6
7.6 8.2 8.7 9.3 9.9
7.2 7.7 8.2 8.7 9.2
6.8 7.3 7.7 8.2 8.7
6.5 6.9 7.3 7.8 8.2
6.2 6.6 7.0 7.4 7.8
5.9 6.3 6.7 7.0 7.4
5.7 6.0 6.4 6.7 7.1
5.5 5.8 6.1 6.5 6.8
Table 3-2. Reasonable Potential Multiplying Factors: 95% Confidence Level and 95% Probability Basis
Number of
SampScs
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Coefficient of Variation
0.1 0.2 0.3 0.4 0.5
1.4 1.9 2.6 3.6 4.7
1.3 1.6 2.0 2.5 3.1
1.2 1.5 1.8 2.1 2.5
1.2 1.4 1.7 1.9 2.2
1.2 1.4 1.6 1.8' 2.1
1.1 1.3 1.5 1.7 1.9
1.1 1.3 1.4 1.6 1.8
1.1 1.3 1.4 1.6 1.7
1.1 1.2 1.4 1.5 1.7
1.1 1.2 1.3 1.5 1.6
1.1 1.2 1.3 1.4 1.6
1.1 1.2 1.3 1.4 1.5
1.1 1.2 1.3 1.4 1.5
1.1 1.2 1.3 1.4 1.4
1.1 1.2 1.2 1.3 1.4
1.1 1.1 1.2 1.3 1.4
1.1 1.1 1.2 1.3 1.4
1.1 1.1 1.2 1.3 1.3
1.1 1.1 1.2 1.3 1.3
1.1 1.1 1.2 1.2 1.3
0.6 0.7 0.8 0.9 1.0
6.2 8.0 10.1 12.6 15.5
3.8 4.6 5.4 6.4 7.4
3.0 3.5 4.0 4.6 5.2
2.6 2.9 3.3 3.7 4.2
2.3 2.6 2.9 3.2 3.6
2.1 2.4 2.6 2.9 3.1
2.0 2.2 2.4 2.6 2.8
1.9 2.1 2.3 2.4 2.6
1.8 2.0 2.1 2.3 2.4
1.7 1.9 2.0 2.2 2.3
1.7 1.8 1.9 2.1 2.2
1.6 1.7 1.9 2.0 2.1
1.6 1.7 1.8 1.9 2.0
1.5 1.6 1.7 1.8 1.9
1.5 1.6 1.7 1.8 1.8
1.5 1.6 1.6 1.7 1.8
:1.4 1.5 1.6 1.7 1.7
1.4 1.5 1.6 1.6 1.7
1.4 1.5 1.5 1.6 1.6
1.4 1.4 1.5 1.5 1.6
H.1 1.2 1.3 1.4 1.5
1
1JS.7 22.3 26.4 30.8 35.6
8.5 9.7 10.9 12.2 13.6
5.8 6.5 7.2 7.9 8.6
4.6 5.0 5.5 6.0 6.4
b.9 4.2 4.5 4.9 5.2
3.4 3.7 3.9 4.2 4.5
3.1 3.3 3.5 3.7 3.9
2.8 3.0 3.2 3.3 3.5
2-6 2.8 2.9 3.1 3.2
2.4 2.6 2.7 2.8 3.0
2.3 2.4 2.5 2.7 2.8
2.2 2.3 2.4 2.5 2.6
2.1 2.2 2.3 2.4 2.5
2.0 2'.1 2.2 2.3 2.3
1.9 2.0 2.1 2.2 2.2
1.9 1.9 2.0 2.1 2.1
1.8 1.9 1.9 2.0 2.0
1.7 1.8 1.9 1.9 2.0
1.7 1.8 1.8 1.9 1.9
1.7 1.7 1.8 1.8 ' 1.8
1.6 1.7 1.8 1.9 2.0
40.7 46.2 52.1 58.4 64.9
15.0 16.4 17.9 19.5 21.1
9.3 10.0 10.8 11.5 12.3
6.9 7.4 7.8 8.3 8.8
5.6 5.9 6.2 6.6 6.9
4.7 5.0 5.2 5.5 5.7
4.1 4.3 4.5 4.7 4.9
3.7 3.9 4.0 4.2 4.3
3.4 3.5 3.6 3.8 3.9
3.1 3.2 3.3 3.4 3.6
2.9 3.0 3.1 3.2 3.3
2.7 2.8 2.9 3.0 3.0
2.5 2.6 2.7 2.8 2.9
2.4 2.5 2.6 2.6 2.7
2.3 2.4 2.4 2.5 2.5
2.2 2.3 2.3 2.4 2.4
2.1 2.2 2.2 2.3 2.3
2.0 2.1 2.1 2.2 2.2
2.0 2.0 2.0 2.1 2.1
1.9 1.9 2.0 2.0 2.0
54
-------
Long-term average
CD
3
,CT
.1
s
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
Value
Figure 3-1 a. Frequency Distribution of Values for a
Lognormal Distribution with a Mean of 1.0 and a
Coefficient of Variation of 0.6
100-,
80-
•5 60-
(D
40-I
20-
Percentile = (1 - 0.99)
1/n
(JO
Number of Samples
100
Figure 3-1 c. Relationship Between the Largest Value of n
Samples and the Percentile It Exceeds
with 99 Percent Confidence
(Long-term average
CV=0.2
CT
CD
UL
CD
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
CO
O 1.4,
CD
1.2-
^ 1_
1 0.8-
1
I °-
-------
STEP1
STEP 2
Dilution
determination1
Conduct toxicity testing2 based
on dilution determination (3 species
at a minimum of quarterly for 1 year)
Acute toxicity data or
estimate based on ACR
STEPS
Chronic toxicity data or
estimate based on ACR
Develop permit
limits
Has
CCC been
exceeded?
/ Do
reaso
poter
\exi
eshX YES
nable \ ^
itiaP /
°X>/
NO
Develo
lin
D permit
tits
Require
monitoring at
reissuance
YES / Does N
^ / reasonable
\potei
\exi
NO
itiaP
st?/
Notes:
1 Dilution determinations should be performed for critical flows and any applicable mixing zones.
^Toxicity testing recommendations
a. Dilution > 1000:1: acute testing, check CMC only.
b. 100:1 < Dilution < 1000:1: acute or chronic testing, check CMC arid CCC with data or ACR.
c. Dilution < 100:1: conduct chronic testing, check CCC with data and CMC using acute data or ACR.
^Reasonable potential: Use procedures in Box 3-3.
Figure 3-2. Effluent Characterization for Whole Effluent Toxicity
1) The effluent causes or contributes to an excursion of a
numeric or narrative water quality criterion and the permit
requires a limit on toxicity.
2) The effluent has a reasonable potential of causing or con-
tributing to an excursion of a numeric or narrative water
quality criterion and a limit is required.
3) The effluent has a very low probability of causing or con-
tributing to an excursion of a water quality standard and
no limit is required.
This categorization is accomplished by using dilution esti-
mates 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.
56
-------
15,863 (3,159 Majors)
6000 -
4000 -
2000 -
2,467 2'771
2,084 __ j^B
H n n n B -
Dilution (stream/effluent)
(a) AtLowFlow(7Q10)
7,908
Q 4000 -
o
2000 -
5,006
fl
I
3,450
1,771
3,849
'''0 %, ** N, '% X0 *
'%'
Dilution (stream/effluent)
(b) At Annual Mean Flow
Figure 3-3. National Distribution of NPDES Dilution
Conditions at 7Q10 and at Annual Mean Flow
The changes to the EPA's data generation recommendations
eliminate the application of multiple sets of safety margins
that was proposed in the 1985 version of this document.
Rather, 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 LCsns of 1 percent
and NOECs of 0.1 percent as the most extreme toxicity
measurements that can normally be expected for the vast
majority of effluents discharged by NPDES permittees for
acute and chronic toxicity, respectively. The observation of
toxicity was based on multiple dilution tests. The same
observation may not hold for toxicity measured with single
dilution tests (pass/fail). As reflected in Chapter 1, single
dilution toxicity tests are much more variable than multiple dilu-
tion tests. Therefore, the use of single concentration toxicity
tests is strongly discouraged for this data generation process.
Since the new data generation requirements are much less expen-
sive than the previous requirements, tiered testing (less expensive,
single-concentration, initial screening followed by increasingly
expensive definitive data generation, using multiconcentration
tests, as described in the September 1985 version of the technical
support document) is unnecessary. However, elimination of the
requirement to conduct toxicity testing on the basis of projec-
tions using dilution alone is not recommended. Although EPA's
data review suggests that an \-Cso °f 1 percent and an NOEC of
0.1 percent are the lower bounds on effluent toxicity, there may
be other effluents that are presently unmeasured that are more
toxic. Testing data are always desirable for fully characterizing
discharges of concern.
Steps in Whole Effluent Characterization Process
The following is a detailed description of the major steps pre-
sented in Figure 3-2 and the rationale behind each.
Step 1: Dilution Determination
The initial step is to determine the dilution of the effluent at the
edge of the mixing zone, assuming the State allows mixing zones.
Figure 3-4 shows a schematic representation of typical mixing
zone requirements for both acute and chronic toxicity. Calculat-
ing the dilution at the edges of mixing zones for site-specific
situations can be complicated. Modeling can be employed using
either steady-state or dynamic approaches to calculate the dilu-
tion (see Chapter 4). However, for complex situations, such as
marine and estuarine waters or lakes, dye studies (or other tech-
niques used to assess mixing zones) may still be required.
Some State water quality standards do not allow the use of
mixing in the control of acute toxicity. For these States, acute
toxicity is often limited at the end of the pipe. Permit limits
derived to enforce such requirements would be considered "wa-
ter quality-based" because they would be based upon an ambient
criterion (as opposed to an arbitrary test endpoint). Regardless,
both chronic and acute toxicity must be assessed in these situa-
tions.
Step 2: Toxicity Testing Procedures
Where toxicity tests are required in order to make decisions
regarding appropriate next steps in a screening protocol, EPA
recommends as a minimum that three species (for example, a
vertebrate, an invertebrate, and a plant) be tested quarterly
for a minimum of 1 year. As discussed in Chapter 1, the use of
three species is strongly recommended. Experience indicates that
marine algae can be a highly sensitive test species for some
effluents. Using a surrogate species of the plant kingdom adds
another trophic level to the testing regimen. 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 is continuing to develop toxicity test methods
using additional organisms including plants. In addition, EPA has
revised the test for Selenastnum, which has improved the test
precision.
57
-------
No initial mixing allowed:
river width
flow >•
Initial mixing allowed:
river width
flow
where
CMC = 0.3 TUa
CCC = 1.0TUC
Figure 3-4. Schematic Representation of Mixing Zone Areas
Where the CMC and CCC Apply
EPA recommends against selecting a "most sensitive" species
for toxldty testing. For one organism to consistently be the
most sensitive in a battery of toxicity tests, two conditions must
occur: (1) the toxicants causing toxicity must remain the same,
and (2) the ratios of the toxicants in the effluent (if more than
one) must remain the same. Based on EPA's experience at the
Duluth research laboratory, neither of these conditions is likely to
occur. For example, the causes of effluent toxicity in POTWs can
vary on a seasonal basis. Toxicity in the summer can be caused
by pesticides to which invertebrates are most sensitive. However,
the winter toxicity could be caused by ammonia to which fathead
minnows will respond most sensitively. The most sensitive spe-
cies for an effluent actually may not exist and at best is difficult to
Identify.
Conducting toxicity tests using three species quarterly for 1
year Is recommended to adequately assess the variability of
toxldty observed In effluents. Below this minimum, the chances
of missing toxic events increase. The toxicity test result for the
most sensitive of the tested species is considered to be the
measured toxicity for a particular effluent sample.
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 in-
creased where necessary to adequately assess effluent vari-
ability. If less frequent testing is required in the permit, it is
preferable to use three species tested less frequently than to test
the effluent more frequently with only a single species whose
sensitivity to the effluent is not well characterized.
EPA recommends that a discharger conduct acute toxicity
testing if the dilution of the effluent is greater than 1000:1 at
the edge of the mixing zone [3]. Such a discharger would be
considered a low priority for chronic toxicity testing. The rationale
for this is that the effluent concentration would be below 0.1
percent at the edge of the mixing zone and thus incapable of
causing an excursion above the CCC. A worst case NOEC of 0.1
percent translates into 1,000 TUC, which would result in a concen-
tration of less than 1.0 TUC at the edge of the mixing zone for this
dilution category. The test results would be compared to the CMC
after consideration of any allowable mixing.
EPA recommends that a discharger conduct either acute or
chronic toxicity testing if the dilution of the effluent falls
between 100:1 and 1,000:1 at the edge of the mixing zone.
Effluents have been shown to be both acutely and chronically toxic
within this range of receiving water dilution. Under worst-case
scenarios, LCsrjs of 1.0 percent and ACRs of 10 will result in
excursions above both the CCC and CMC at the edge of the
regulatory mixing zone.
• «
Although either acute or chronic testing can be required within
this dilution range, acute testing Would be more appropriate at the
higher end of this dilution range (1,000:1 or 0.1 percent). At the
lower end of this dilution range (100:1 or 1.0 percent), chronic
tests may be more appropriate. Where other factors are equal,
chronic testing may be preferable since the interim results in a
chronic test gives data on acute toxicity as well. The acute
endpoint data can then be used to compare directly to the CMC
w|thout the need for an ACR.
I .'''.'''''
Whichever type of toxicity test (either acute or chronic) is speci-
fied, the results from that test should be compared to the criterion
associated with that type of test. For example, a chronic test
would be compared to the CCC. Comparisons to the other criteria
ca|n be made by using the ACR or additional data generated to
convert a chronic test result to an acute endpoint and vice versa.
For example, a chronic NOEC of 5 percent effluent (or 20 TUJ
represents an acute LC^Q of 50 percent (or 2 TUa) at an ACR of 10.
EPA recommends that a discharger conduct chronic toxicity
testing if the dilution of the effluent falls below 100:1 at the
edge of the mixing zone. The rationale for this recommendation
is that chronic toxicity has been observed in some effluents down
to the 1.0 percent effect concentration. Therefore, chronic toxicity
tests, although somewhat more expensive to conduct, should be
used directly in order to make decisions about toxic impact.
i
There is a potential for acute toxicity within this dilution range,
although this is less likely as the 100:1 dilution level is approached.
Thus, the recommended screening protocol shown in Figure 3-2
includes a determination of whether excursions above the CMC
are projected [4]. 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 regula-
tory authority may use the interim results in the chronic test to
calculate the acute toxicity.
58
-------
Both the chronic and acute toxicity test data would be compared
to their respective criterion. The chronic test results would be
compared to the CCC, and the acute results, regardless of how
calculated, would be compared to the CMC.
Step 3: Decision Criteria 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 an excursion of an applicable numeric or narrative
water quality criterion and therefore needs to limit effluent toxic-
ity. To do this, these data should be used to project receiving
water concentrations, which are then 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 — Where any one data point
shows an excursion above the State's numeric or narrative
criterion for the parameter toxicity, EPA regulations require a
, permit limit be set for whole effluent toxicity (40 CFR
1 22.44(d)(1 )(iv or v)), unless limits on a specific chemical
will allow the narrative water quality criterion 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.
The decision to develop permit limits based upon an excur-
sion above either the CMC or CCC will lead to protection
against both acute and chronic toxicity if the permit deriva-
tion procedures in Chapter 5 are used to set effluent limits.
2) Reasonable Potential for Excursion Above CMC or CCC—
EPA believes that "reasonable potential" is shown where
an effluent is projected to cause an excursion above the
CCC or CMC. This projection is based upon a statistical
analysis of available data that accounts for limited sample
size and effluent variability. EPA's detailed recommenda-
tions for making a statistical determination based upon
effluent monitoring data alone are shown in Box 3-2. Where
a regulatory authority finds that test results alone indicate a
"reasonable potential" to cause an excursion above a State
water quality criterion in accordance with 40 CFR
1 22.44(d)(1 )(ii), a permit limit must be developed.
A regulatory authority may select an alternative approach
; for assessing reasonable potential. For example, an author-
ity may opt to use a stochastic dilution model that incorpo-
rates both ambient dilution and effluent variability for deter-
mining reasonable potential. Such an approach is analo-
gous to the statistical approach shown in Box 3-2. Whatever
approach selected by the authority, it must use all the
factors that account for all the factors listed in 40 CFR
In some cases the statistical analysis of the effluent data may
not actually project an excursion above the CMC or CCC
but may be close. Under such conditions, reasonable poten-
tial determinations will include .an element of judgment on
the part of the regulatory authority. Other factors will need
to be considered and given appropriate weight in the
decisionmaking process, including value of waterbody (e.g.,
high-use fishery), relative proximity to the CCC or CMC,
existing controls on point and nonpoint sources, informa-
tion on effluent variability, compliance history of the facil-
ity, and type of treatment facility. These factors are
summarized in Box 3-2 and are discussed in detail in
Section 3.1. EPA recommends regulatory authorities
establish a written policy and procedure for making
determinations of "reasonable potential" under these
circumstances.
3) No Reasonable Potential for Excursions Above CMC or
CCC—In these situations, EPA recommends that the
toxicity tests recommended above be repeated at a
frequency of at least once every 5 years as a part of
the permit application. Such testing is required for
certain POTWs under 40 CFR 122.21 (j).
4) Inadequate Information—Where a regulatory authority
has inadequate information to determine reasonable po-
tential for ah excursion of a numeric or narrative water
quality criterion, there may still be a basis for concern on
the part of the authority. The permit should contain
whole effluent toxicity monitoring 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 causes, has the reasonable potential of caus-
ing, or contributes to an excursion above the CCC or
CMC.
3.3.4 Use of Toxicity Testing in Multiple-source Discharge
Situations
Where more than one discharge to the same receiving waterbody
contributes, or has the reasonable potential to contribute to an
excursion of water quality standards,, permit limits must be
developed for each individual discharger on that waterbody.
For the regulatory authority to make this assessment/additional
testing may be needed to provide the authority with the infor-
mation necessary to assess the relative impact of each source.
For purposes of this discussion, a multiple-source discharge
situation is defined as a situation where impact zones overlap, dr
where ambient receiving water concentrations of a pollutant
are elevated due to upstream discharges. In multiple-source
discharge situations, additivity, antagonism, and persistence of
toxicity can be of concern. To collect additional data, the permit
authority should employ the toxicity testing procedures for
multiple dischargers described in Box 3-3. In addition, ambient
toxicity testing, as described below, could be used.
Assuming that screening has been conducted that reveals the
need for permit limits, two options for controlling the dis-
charges exist. The first option is for the permit authority to
regulate each source separately using the procedures for indi-
vidual point sources. In this option, the permitting authority
would require use of upstream ambient water as a diluent in the
toxicity test so as to be able to evaluate the contributions of
upstream sources of toxicity. A second option is to treat each
discharge as an interactive component of a whole system. In
this option, the permit writer would determine a total maxi-
mum daily load for the receiving waterbody and develop indi-
vidual wasteload allocations for each discharger using the pro-
cedures discussed in Chapter 4.
59
-------
Box 3-3. Recommend Multiple-source Toxicity Testing Procedures
Tests
Where the combined effluents make up 1 percent or greater of the receiving waters, conduct chronic toxicity
tests following the testing procedures described in Section 3.3.3.
•
Where the combined effluents make up less than 1 percent of the receiving waters, conduct acute toxicity tests
following the testing procedures described in Section 3.3.3 (see Figure 3-2) to determine if any of the effluents
are exhibiting toxicity. j
An additional data requirement 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 concentra-
tions (including the 100 percent diluent 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 nonpoint 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. i
Frequency for Ambient Testing
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 (1 to 2 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. I
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 [1,2].
60
-------
3.3.5 Ambient Toxieify Testing
Ambient toxicity testing also is useful in screening receiving
water bodies for existing toxic conditions. The procedure de-
scribed in Appendix C uses short-term chronic toxicity tests to
measure the toxicity of samples of receiving water taken above,
at, and below outfalls. It can be used in freshwater, marine, and
estuarine systems. 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 directly measured. No extrapo-
lation from exposure or ACR is needed. Further, impact from
multiple source discharge situations, which may not be apparent
from individual discharger data, is identified. Finally, the tech-
nique can provide an assessment of the persistence of effluent
toxicity.
3.3.6 Special Considerations for Discharges to Marine and
Estuarine Environments
Special problems are encountered when assessing and control-
ling impacts of toxic pollutants discharged to marine and estua-
rine waterbodies. These special problems include the following:
• Determining the physical characteristics of estuaries and
the complex mixing and effluent dilution situations for
RWCs of effluents.
• Generating toxicity data on nonsaline effluents that dis-
charge to brackish or saline waters and establishing cause-
effect relationships on that basis.
• Assessing exposure and controlling impacts from persis-
tent toxicants accumulating in fish and shellfish tissues
and in sediments. These factors are particularly important
in estuaries and near coastal waters because of high use of
estuaries as breeding and fishing areas for important com-
mercial seafood supplies and recreational fishing, and be-
cause many estuaries and near coastal waters act as sinks
for pollutants that accumulate in sediments.
Where these special problems are encountered, additional infor-
mation may need to be gathered to better quantify dilution, to
determine metals partitioning, and to identify potential interfer-
ences in whole effluent toxicity tests.
To characterize the type of whole effluent toxicity that is most
relevant for a particular discharge to marine and estuarine wa-
ters, the following questions should be considered [5]:
• What is the salinity of the receiving water, and is this
important in terms of the State standards?
• What is the appropriate test organism to require for toxic-
ity testing under differing salinity conditions?
The answers to these questions will enable the permitting au-
thority to determine what type of toxicity testing is most suitable
for effluent characterization and whole effluent toxicity control.
For most marine and estuarine discharges the choice of test
species and dilution water should be made based on the charac-
teristics of the receiving water at the critical conditions for flow,
mixing, and salinity. Foremost in this determination should be
the salinity of the receiving water and, to a lesser extent, the
salinity of the effluent itself.
The primary objective of whole effluent toxicity tests is to identify
sources of toxicity that can potentially cause an excursion of a
State's narrative or numeric water quality criteria. For this reason,
the toxicity tests should reflect the natural conditions of the
receiving water so to be able to measure any effluent characteris-
tic that could contribute to ambient toxicity. The marine toxicity
test methods identify 1,000 mg/l as the point at which salinity
begins to exert an effect on freshwater species. As a general
rule, EPA recommends that freshwater organisms be used
when the receiving water salinity is less than 1,000 mg/l, and
that marine organisms be used when the receiving water
salinity equals or exceeds 1,000 mg/l.
Saline Effluent Discharges to Saltwater
The dissolved salts in the effluent are pollutants. These salts may
or may not be the same as those present in the receiving water.
Also, the proportion of dissolved salts in the effluent may be
different from that of the salts in the receiving water. In this case,
the toxicity test needs to be able to determine if these salts
contribute to ambient toxicity. For this reason, marine organ-
isms are needed.
Saline Effluent Discharged to Freshwater
In this case, the dissolved salts in the effluent is a pollutant that
does not exist in the receiving water. The toxicity test needs to
determine whether the dissolved salts can be one of the toxicants
that contribute to ambient toxicity. For this reason, freshwater
organisms are needed.
Freshwater Effluent Discharged to Saltwater
In this instance, the lack of dissolved salts in the effluent can
cause an apparent toxic effect to the marine organisms in the
toxicity test. However, in contrast to the instances presented
above, the toxicity test does not need to be able to measure this
effect because a lack of salts is not a pollutant. The marine
toxicity test methods account for this by requiring that the
salinity of the effluent be adjusted to approximate the salinity of
the receiving water. As an alternative to using a marine organism,
a freshwater organism can be used if the test is being conducted
only on a 100-percent effluent sample and if State water quality
standards do not require that a marine organism be used.
3.3.7 Using a Chemical-specific Limit to Control Toxicity
EPA regulations at 40 CFR 122.44(d)(1 )(v) provide that limits on
whole effluent toxicity are not necessary where the permitting
authority demonstrates in the fact sheet or statement of basis of
the NPDES permit that chemical-specific limits for the effluent are
sufficient to attain and maintain applicable numeric and narrative
State water quality criteria. To make this demonstration that
chemical-specific limits are sufficient, additional effluent informa-
tion will be needed. EPA recommends that the discharger
conduct a toxicity identification evaluation to identify the
causative agent(s) in the effluent. Where the permitting au-
thority determines that the demonstration required by 40 CFR
122.44(d)(1 )(v) has been made, limits on whole effluent toxicity
61
-------
need not be imposed. Effluent limits on the controlling chemical
with concurrent whole effluent monitoring will be sufficient. Where
subsequent whole effluent toxicity testing reveals the presence of
toxicity In the effluent, the above process will need to be repeated,
or alternatively a whole effluent toxicity limit will be needed. If
continued toxicity testing shows that additional chemical-specific
effluent limits are insufficient to control whole effluent toxicity,
then toxicity limits may be the only practical way to control
toxicity.
3.3.8 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 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 re-
quire a determination of the type of testing as is required for whole
effluent toxicity because there is generally only one type of test for
specific chemicals. However, there are some antecedent steps that
are unique to effluent characterization for specific chemicals: de-
termination of the chemicals of concern and determination of
acceptable ambient levels (RAC, CMC, or CCC) for these pollut-
ants.
Steps for Chemical-specific Effluent Characterization Process
Figure 3-5 illustrates EPA's recommendations for determining
whether or not permit limits need to be developed according to
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 3.3.9 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 the following:
• Permit application forms, DMRs, permit compliance systems
(PCS), and permit files
• Pretreatment industrial surveys
• STORET for ambient monitoring data
• SARA Title III Toxic Chemical Release Inventory
• Industrial effluent guidelines development documents
• The Treatability Manual [6]
• Effluent bioconcentration assessment (see Section 3.3.9).
Data on specific chemicals that 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 regula-
tory authority has reason to believe that additional data for key
parameters of concern are needed in order to adequately charac-
terize 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 ana-
lyzed for key parameters of concern. In some cases, special
analytical protocols will need to be specified in order to gather all
appropriate information.
Step 2: Determine the Basis for Establishing RACs, CMCs, and
CCCs for the Pollutants of Concern
I
The second step is to identify the appropriate water quality stan-
dard, including designated or existing use, and criteria for use.
Ideally, the State water quality standards include aquatic life and
human health criteria 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 crite-
rioh may be used (40 Cffl 122.44(d)(1 )(vi)) to determine whether
a discharge causes, has the reasonable potential to cause, or
contributes to an excursion above a narrative criteria because of
an individual pollutant. Although the provisions of 40 CFR
122.44(d)(1)(vi) are presented in the regulation in the context of
permit limit development, these same considerations should be
applied in characterizing effluents in order to determine whether
limits are necessary. The options available are as follows:
» Option A allows the regulatory authority to establish limits
using a "calculated numeric water quality criterion" that
the regulatory authority demonstrates will attain and main-
tain applicable narrative water quality criteria and fully
protect the designated use. This option allows the regula-
tory authority to use any criterion that protects aquatic life
and human health. This option also allows the use of site-
specific factors, including local human consumption rates
of aquatic foods, the State's determination of an appropri-
ate risk level, and any other current data that may be
available.
» Option B allows the regulatory authority to establish efflu-
ent limits using EPA's Water Quality Criteria guidance docu-
i ments, if EPA has published a criteria document for the
pollutant supplemented where necessary by other relevant
I information. As discussed earlier, EPA criteria documents
provide a comprehensive summary of available data on the
effects of a pollutant.
«* Option C may be used to develop limits for a pollutant of
concern based on an indicator parameter under limited
circumstances. An example of an indicator parameter is
total toxic organics 0"TO); 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 to a
level that will attain and maintain applicable State narrative
water quality criteria and will protect aquatic life and hu-
man health (see 40 CFfl 122.44(d)(1 )(vi)(Q). Use of this
option is restricted by regulation to those instances where it
can be demonstrated that controls on indicator pollutants
serve to control the toxicant of concern. Using Option A or
Option B is a more direct and perhaps more defensible
approach.
62
-------
STEP1
STEP 2
Identify pollutants
of concern
RAG available
Determine RAC and/or
CMC/CCC for pollutants
of concern 1
CMC and/or
CCC available
STEP 3
Dilution determination
for human health
impacts2
Dilution determination
for aquatic life
impacts2
Does
reasonable
potential
exist?3
Select the most
restrictive
Develop permit
limits
Require
monitoring at
reissuance
YES /Mas CMC
or CCC been
exceeded?
Require
monitoring at
reissuance
Notes:
1 RAC and/or CMC/CCC: Use State numeric criterion or interpret State narrative criterion using one of three options specified under 40 CFR
122.44(d).
2 Dilution determination: Perform for critical flow and for any applicable mixing zones for aquatic life and human health protection procedures,
respectively.
3 Reasonable potential: Use procedures in Boxes 3-2 and 3-4.
Figure 3-5. Effluent Characterization for Specific Chemicals
Step 3: Dilution Determination
The third step is to calculate the effluent dilution at the edge of
the mixing zone. The pertinent factors for consideration here are
the same as were previously presented for whole effluent toxicity
with one difference: there are two levels of dilution analysis for
chemical data. The first level is to use simple fate models based
on a dilution analysis and comparison with the RAC, 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 sup-
ported development of a second level of analysis that estimates
point source wasteload allocations and nonpoint source alloca-
tions and predicts the resulting pollutant concentrations in re-
ceiving waters [7].
Step 4: Decision Criteria for Permit Limit Development
After this dilution analysis has been performed, the projected RWC
is compared to the RAC, CMC, or CCC (either the State numeric
criteria or an interpretation of the narrative criteria as described
earlier). Whereas analysis of aquatic impacts should include evalu-
ations with respect to both the CCC and the CMC, analysis of
human health impacts will only involve comparisons with the RAC.
The four possible outcomes discussed above in the triggers for
permit limit development discussion in Section 3.3.3 also apply
here:
• Excursion above the RAC, CMC, or CCC
• Reasonable potential for excursion above the RAC, CMC, or
CCC
63
-------
• No reasonable potential for excursion above the RAC, CMC,
CCC
• Inadequate information.
If these evaluations project excursions or the reasonable potential
to cause or contribute to an excursion above the RAC, CMC/or
CCC, then a permit limit is required (40 Cffi 122.44(d)(1 )(iii)).
The statistical approach shown in Box 3-2 or an analogous ap-
proach developed by a regulatory authority can be used to deter-
mine the reasonable potential. Effluents that are shown not to
cause or that have a reasonable potential to cause or contribute to
an excursion above an RAC, CMC, or CCC should be reevaluated
at permit reissuance.
Where chemical-specific test results do not show a reasonable
potential but indicate a basis for concern after consideration of the
other factors discussed in Section 3.2, or if there were inadequate
information to make a decision, the permit should contain chemi-
cal testing requirements and a reopener clause. This clause would
require reopening of the permit and establishment of a limit based
upon any test results that show effluent toxicity at levels that cause
or have a reasonable potential to cause or contribute to an excur-
sion above the RAC, CCC, or CMC.
3,3.9 Effluent Characterization for Bioconcentratable
Pollutants
The previous section discussed how to characterize effects of
specific chemicals, including those that may threaten human health,
to determine whether or not a discharge causes, has the reason-
able potential to cause, or contributes to excursions above an
water quality criterion. The primary disadvantage of this approach
is that it does not identify all effluent chemicals of potential con-
cern for human health. To help address this gap, EPA is develop-
ing a procedure for identifying pollutants with the propensity to
btoconcentrate in fish tissue. This procedure is presently in draft
form and should not be used for establishing NPDES permit limits
until EPA releases the final document on the procedure. This
section describes the outline of this procedure.
The overall approach illustrated in Figure 3-6 is a seven-step proce-
dure that starts with collecting samples and ends with developing
permit effluent limits. The effluent characterization step unique to
this approach lies in Step 3. There are two alternatives under this
step: fish tissue residue and effluent assessment. An analytical
chemistry laboratory with residue chemistry and gas chromato-
graph/mass spectometer (GC/MS) capability is needed to conduct
the analytical methods for both alternatives. A summary of the
alternatives follows:
• Tissue Residue Alternative: This alternative measures the con-
centrations of organic bioconcentratable chemicals in tissue
samples of indigenous organisms from the receiving water.
This analysis involves the collection of fish or shellfish samples,
the extraction of the organic chemicals from the tissue and
the analysis of these extracts with GC/MS to identify and
quantify the bioconcentratable contaminants. The procedure
provides recommendations to sort the results of this screening
analysis in order to determine which of the contaminants pose
a hazard and require regulatory action. The approach recom-
mends that the identity of those contaminants then be con-
firmed prior to taking subsequent action.
Select dischargers
and/or
receiving waters
Select assessment
alternative
Effluent
bioconcentration
alternative
Tissue
residue
alternative
Are bio-
concentratable
contaminants
present?
Does
reasonable
potential
exist?
No further
regulatory action
No further
regulatory action
Figure 3-6. Procedure for Assessment and Control of
Bioconcentratable Contaminants in Surface Waters
• Effluent Alternative: This alternative measures the concen-
trations of organic bioconcentratable chemicals in effluent
samples from point source dischargers. This analysis in-
volves the collection of effluent samples, the extraction of
the organic chemicals from the effluent sample, and the
separation of the chemicals that have characteristics known
to result in bioconcentration from the other chemical com-
ponents of the effluent sample. This separation is achieved
by way of an analytical chemistry methodology called hig'h-
64
-------
pressure liquid chromotography (HPLC). The HPLC
also separates (fractionates) an effluent sample into
three subsamples or "fractions." These three fractions
contain chemicals with increasing potential to
bioconcentrate, with the third fraction containing those
chemicals with the highest bioconcentration rates.
Following HPLC fractionation, each fraction is then
analyzed with GC/MS to identify and quantify the
bioconcentratabL- contaminants. The effluent proce-
dure also provides recommendations to sort the re-
sults of the initial screening analysis to determine
which of the contaminants pose a hazard and require
subsequent regulatory action. The approach then
recommends that the identity of those contaminants
then be confirmed prior to taking further regulatory
action.
While both of the assessment alternatives described above
may be used for a given discharger, generally one of these
alternatives may be preferred by the regulatory authority.
The regulatory authority would select the assessment ap-
proach based on the available site- and facility-specific infor-
mation and the objectives of the application.
Although the approach provides a means to identify chemicals
that can bioconcentrate, it does not identify all bioconcentratable
chemicals. Chemicals that bioconcentrate include many organic
compounds, and a small number of metals (e.g., mercury and
selenium) and organometals (e.g., tributyltin). The new approach
is limited to nonpolar organic chemicals that produce measurable
chemical residues in aquatic organisms or that have log octanol-
water partition coefficients greater than 3.5.
3.3.10 Analytical Considerations for Chemicals
Analysis of discharges for toxic substances requires special quality
control procedures beyond those necessary for conventional pa-
rameters. Toxicants can occur in trace concentrations and are
frequently volatile or otherwise unstable. An EPA publication en-
titled, Test Methods—Technical Additions to Methods for Chemical
Analysis of Water and Wastes [8], contains sampling and handling
procedures recommended by EPA for a number of toxic and
conventional parameters. Additional methods for analyses for
toxicants are described in Standard Methods of Water and Waste-
water Analyses (ASTM, 17th edition, 1989, or most recent edition)
and 40 CFR Part 136. Chapter 5 discusses detection limits and
sampling requirements.
65
-------
CHAPTERS
REFERENCES
1. Mount, D., N. Thomas, M. Barbour, T. Norberg, T. Roush, and
W. Brandes. 1984. Effluent and Ambient Toxlclty Testing and
Instream Community Response on the Ottawa River, Lima, Ohio.
Permits Division, Washington, DC, Office of Research and
Development, Duluth, MN. EPA-600/2-84-080, August 1984.
2. Mount, D.I., and T.J. Norberg-King, eds. 1985. Validity of
Effluent and Ambient Toxlclty Tests for Predicting Biological
Impact, Sclppo Creek, Circleville, Ohio. U.S. EPA. EPA/600/3-
85/044, June 1985.
3. Peltier, W., and C.I. Weber. 1985. Methods for Measuring the
Acute Toxkity of Effluents to Aquatic Organisms, 3d ed. Of-
fice of Research and Development, Cincinnati, OH. EPA-
600/4-85-013.
4. Weber, C.I., et al., eds. 1989. Short-Term Methods for Esti-
mating the Chronic Toxlclty of Effluents and Receiving Waters
to Freshwater Organisms, 2d ed. Office of Research and
Development, Cincinnati, OH. EPA-600/4-89/001.
5. Weber, C.I., etal., eds. 1988. Short-Term Methods for Estimat-
ing 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.
6. U.S. EPA. 1983. The Treatability Manual, Volume 4. U.S. EPA
Office of Research and Development. EPA 600/2-82-001
(revised January 24,1983).
7. 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, September 1982.
8. U.S. EPA. 1982. Test Methods - Technical Additions to Methods
for Chemical Analysis of Water and Wastes. Office of Re-
[ search and Development, Cincinnati, OH. EPA 600/4-82-
055, December 1982.
ADDITIONAL REFERENCES
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.
Crane, J.L., A. Pilli, and R.C. Russa. 1984. CETIS: Complex
Effluents Toxicity Information System. Data Encoding Guide-
tines and Procedures. Office of Research and Development,
Duluth, MN. EPA-60018-84-029. November 1984.
DiToro, D. 1985. Exposure Assessment for Complex
Effluents: Principles and Possibilities. In Environmental
Hazard Assessment of Effluents. Eds. H. Bergman, R.
Kimerle, and A. Maki.
'
Macek, K. 1985. Perspectives on the Application of the
Hazard Evaluation Process. In Environmental Hazard
Assessment of Effluents. Eds. H. Bergman, R. Kimerle,
and A. Maki.
66
-------
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 a wasteload allocation (WLA)
based on an existing total maximum daily load (TMDL) may
already have been established. A TMDL is the sum of the indi-
vidual 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 uses water quality analyses to predict water
quality conditions and pollutant concentrations. Limits on waste-
water pollutant loads are set and nonpoint source allocations 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 standards, with
seasonal variations and a margin of safety that takes 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 condi-
tions for stream flow, loading, and water quality parameters.
Conditions that will protect the receiving water have been deter-
mined from State numeric or narrative water quality criteria.
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 control.
The section also discusses different WLA schemes. Sections 4.3
and 4.4 discuss mixing zones, areas described as allocated impact
zones where acute and chronic water quality criteria may be
exceeded. Section 4.3 provides background information on mix-
ing 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 determina-
tion 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 near-field and far-field mixing.
Section 4.5 discusses the calculations of the WLA and LA and the
types of EPA-recommended mathematical models available to
determine WLAs in completely mixed situations for both aquatic
life and human health. The WLA models listed in Section 4.5 can
be used to predict ambient concentrations and to calculate the
effluent quality required to meet the criteria and protect desig-
nated and existing uses of the receiving water. The data require-
ments 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. Section 4.6 discusses human health
considerations and how to determine WLAs for human health
toxicants.
EPA is currently working on methods to develop sediment criteria.
Once developed, point source discharges could be further limited
to prevent accumulation of pollutants in the bed sediment; such
accumulation impairs 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 dis-
cussed in Section 4.5 are capable of simulating interactions between
the water column and sediment and between toxic transport and
transformation in the sediment. EPA is encouraging the States to
consider the role of sediments in WLA.
4.2 TOTAL MAXIMUM DAILY LOADS AND WASTELOAD
ALLOCATIONS
4.2.1 Total Maximum Dally Loads
The Federal Clean Water Act (CWA), under Section 303(d), re-
quires the establishment of 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. A TMDL includes a
determination of the amount of a pollutant, or property of a
pollutant, from point, nonpoint, and natural background sources,
including a margin of safety, that may be discharged to a water
quality-limited waterbody. Any loading above this loading capac-
ity risks violating water quality standards. TMDLs can be expressed
in terms of chemical mass per unit of time, by toxicity, or by other
appropriate measures. Permits should be issued based on TMDLs
where available.
The establishment of a TMDL for a particular waterbody is depen-
dent on the location of point sources, available dilution, water
quality standards, nonpoint source contributions, background
conditions, and instream 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: (1) the defini-
tion 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 be-
67
-------
Iween the pollutant source (whether point source or nonpoint
source) that is farthest upstream and the downstream point at
which water quality has recovered to the background quality
found above the pollutant source that is farthest upstream. The
delineation of critical conditions for stream flow, loading, and
water quality parameters may be 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 stan-
dards as applied to a specific waterbody. Such a fitting of the
TMDL to desired water quality criteria requires information con-
cerning the distribution of loadings within the waterbody, namely,
the locations and relative contributions of pollutant-specific load-
ings from point, nonpoint, and background sources during all
flow conditions (40 CFR 130.2(f)). Low-flow TMDLs, by them-
selves, will not be adequate in situations where nonpoint source
loadings (LAs) during high or intermediate flow conditions cause
excursions above water quality standards (40 CFR 130.2(f)).
The loading capacity of TMDLs have been determined in many
ways, but the most common method is to find the pollutant
loading that will attain and maintain applicable water quality
criteria. For example, in the Tualatin River Basin in Oregon,
loading capacity was determined by multiplying stream flow in
critical flow periods by the pollutant water quality standard [1].
Another method of determining a loading capacity is by quantify-
ing instream toxicity. This method was used in developing a
TMDL for the Amelia River in Florida [2].
The allowable TMDL is defined as the sum of the individual WLAs
and LAs; a margin of safety can be included with the two types of
allocations to ensure that allocated loads, regardless of source,
would not produce an excursion above water quality standards.
The WLAs are those portions of the TMDL assigned to point
sources; the LAs are those portions of the TMDL assigned to the
sum of all nonpoint sources and background sources (40 CFR
130.2(0). 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 histori-
cal discharges might also act as a source of toxicants and contribute
to the background levels; these sediments also may be part of the
nonpoint sources.
The TMDL represents a mass loading that may occur over a given
time period to attain and maintain water quality standards. As a
result, the design flows under which the TMDL is determined can
significantly alter its value. This phenomenon results in a some-
what unusual dichotomy. The design flows for aquatic life protec-
tion most applicable to point source loadings (WLAs) usually
involve low-flow events (e.g., 7Q10) because the volumes associ-
ated with the point sources generally do not decrease with de-
creased stream flow. 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., non-
existent under low-flow drought conditions).
The TMDL is a composite of the allowable loads associated with
point sources and nonpoint sources within the defined bound-
aries of the waterbody segment and the background loadings to
that segment from upstream and from in-place sediments.
Therefore, the TMDL should be evaluated under conditions that
reflect worst-case (critical) conditions for both point and nonpoint
source loadings (i.e., low-flow drought and high flow conditions).
Determination of the TMDL under these two scenarios would
identify the lower of the two loading capacities of the waterbody.
This lower capacity is necessary to protect the waterbody in
question.
'
In the case of design flows for human health protection, the
' harmonic mean flow is recommended as the basis for TMDLs for
carcinogens. Design flows for human health protection should
consider worst-case conditions for both point and nonpoint source
loadings under this flow condition (see Section 4.6).
In many cases, LAs for nonpoint sources are difficult to assess
because the information needed to describe the runoff associated
with the high-flow storm events does not exist. This lack of
information is due to the high variability of the events. Because of
the importance of estimating the nonpoint contributions to the
waterbody, 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.
CWA Section 303(d) and EPA regulations (40 CFR Parts 35 and 130,
January 11,1985) require that TMDLs contain a margin of safety
"wtych 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 pollutant loadings, ambient conditions, and the
model analysis. The size of the required margin of safety can, of
course, be reduced by collecting additional information, which
reduces the amount of uncertainty. The margin of safety can be
proyided for in the TMDL process by one of the following:
Reserving a portion of the loading capacity to a separate
margin of safety.
i
« Including a margin of safety within the individual WLAs for
point sources and within the LAs for nonpoint sources and
background sources.
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 prusuant to the Water
Quality Standards Regulation (Section 131.12 of 40 CFR Part 131,
68
-------
November 8, 1983). This regulation establishes explicit proce-
dures that must be followed prior to lowering existing water
quality to a level that still supports the Section 101 (a)(2) "fishable/
swimmable" goal of the Act. WLAs, LAs, and TMDLs that allow
such a decline in water quality cannot be established unless the
applicable public participation and intergovern-mental review
requirements of the antigradation provisions have been met and
all existing uses are fully maintained and protected.
4.2.2 Wasteload Allocation Schemes
WLAs for water quality-based toxics permits must be set in accor-
dance with EPA regulations [3, 4]. EPA has developed a number
of WLA guidance documents to assist regulatory authorities in
developing TMDLs and WLAs. The EPA Office of Water Regula-
tions and Standards, Assessment and Watershed Protection Divi-
sion, maintains the latest listing of all WLA guidance documents.
Toxic WLA guidance documents are currently available for rivers
and streams [5], lakes and reservoirs [6], and estuaries [7]. Guid-
ance for the determination of critical design conditions for steady-
state modeling of rivers and streams also is available [8].
Table 4-1 lists 19 allocation schemes that may be used by the
States to develop WLAs. This is not intended to be a complete list
of approaches; regulatory authorities may use any reasonable
allocation scheme that meets the antidegradation provisions and
other requirements of State water quality standards [3].
The most commonly used allocation methods have been equal
percent removal, equal effluent concentrations, and a hybrid
method. The equal percent removal approach can be applied 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 also
can be applied in two acceptable ways—equal final concentra-
tions or equal incremental concentration reductions. This method
is similar to the equal percent removal method if influent concen-
trations 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 treat-
ment levels for that source than the equal percent removal ap-
proach.
The final commonly used method of allocating wasteloads 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, while another may be required to provide
the entire load reduction. More often, a proportionality rule that
requires the percent removal to be proportional to the input
loading can be assigned. 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. A mixing zone is an allocated impact
zone where acute and chronic water quality criteria can be ex-
ceeded as; long as a number of protections are maintained, in-
cluding freedom from the following:
• Materials in concentrations that settle to form objection-
able deposits
• Floating debris, oil, scum, and other matter in concentra-
tions that form nuisances
Table 4-1. Wasteload Allocation Methods [9]
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/l)
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
11 a. 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. Best availability technology (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 best practicable tech-
nology (BPT) and BAT, i.e., Allowable wasteload alloca-
tion:
(WLA) = BPT- -^-.(BPT-BAT)
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 treat-
ment than BPT. Finally, when stream flow is above an
upper value, the plant would be required to treat to a
level comparable to BPT.
69
-------
• Substances in concentrations that produce objectionable
color, odor, taste, or turbidity
• 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 acute and
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. For a mixing zone to be permitted, the dis-
charger 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 character-
istics. 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 pos-
sible.
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 measur-
able difference in the concentration of the pollutant (e.g., does
not vary by more than 5 percent) exists across any 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 for preventing lethal conditions in the mixing zone.
4.3.1 Determination of Mixing Zone Boundaries
Allowable mixing zone characteristics should be established to
ensure the following: . .
• Mixing zones do not impair the integrity of the waterbody
as a whole.
• There is no lethality to organisms passing through the
mixing zone.
• There are no significant health risks, considering likely path-
ways of exposure (see Section 2.2.2).
The Water Quality Criteria — 1972 [10] recommends 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 bio-
logical 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 [11]. 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).
has developed a holistic approach to determine whether a
mixing zone is tolerable [12]. 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 upstream and down-
stream waterbodies and the ecological and cultural data pertain-
ing to them; the collection of data on all present and future
discharges to the waterbody; the assessment of relative environ-
mental value and level of protection needed for the waterbody;
and, finally, the allocation of environmental impact for a discharge
apJDlicant. 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.
Hojwever, 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 allowed) 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.
Where a mixing zone is allowed, water quality standards are met
at the edge of that regulatory mixing zone during design flow
conditions and generally, (1) provide a continuous zone of pas-
sage that meets water quality criteria for free-swimming and
drifting organisms and (2) prevent impairment of critical resource
areas. Individual State mixing zone dimensions are designed to
limit the impact of a mixing zone on the waterbody. Furthermore,
EPA's review of State WLAs should evaluate whether assumptions
of I complete or incomplete mixing are appropriate based on
available data.
•In Iriver 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 impor-
tarit in a number of ways: food is carried to the sessile filter
70
-------
feeders and other nonmobile organisms, spatial distribution of
organisms and reinforcement of weakened populations are en-
hanced, and embryos and larvae of some fish species develop
while drifting [11]. 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 may decide to
deny a mixing zone in a site-specific case. For example, denial
should be considered when bioaccumulative pollutants are in the
discharge. The potential for a pollutant to bioaccumulate 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 concentra-
tion 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 that bioaccumulation
potential exists, bioaccumulation potential is generally not con-
sidered to be significant unless the BCF exceeds 100 or more.
Thus, a chemical that is discharged to a receiving stream, result-
ing in low concentrations, and that has a low BCF value will not
create,a bioaccumulation hazard. Conversely, a chemical that is
discharged to a receiving stream, resulting in a low concentration
but having a high BCF value, may cause in a bioaccumulation
hazard. Also, some chemicals of relatively low, toxicity, such as
zinc, will bioconcentrate in fish without harmful effects resulting
from human consumption.
Another example of when a regulator should 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 [13].
However, some chemicals did elicit an attractive response, but the
data were npt sufficient to support any predictive methods. Tem-
perature 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 fish to incur a significant expo-
sure. ,, ,
4.3.2 Minimizing the Size of Mixing Zones
Concentrations above the chronic criteria are likely to prevent
sensitive taxa from taking up long-term residence in the mixing
zone. In this regard, benthic organisms and territorial organisms
are likely to be of greatest concern. The higher the concentra-
tions occurring within an isopleth, the more taxa are likely to be
excluded, thereby affecting the structure and function of the
ecological, community. It is thus important to minimize the
overall size of the mixing zone and the size of elevated concentra-
tion isopleths within the mixing zone.
4.3.3 Prevention of Lethality to Passing Organisms
The Water Quality Standards Handbook [14] indicates that whether
to establish a mixing zone policy is a matter of State discretion,
but that any State policy allowing for mixing zones must be
consistent with the CWA and is subject to approval of the Re-
gional Administrator. The handbook provides additional discus-
sion regarding the basis for a State mixing zone policy.
Lethality is a function of the magnitude of pollutant concentra-
tions and the duration an organism is exposed to those concen-
trations. Requirements for wastewater plumes that tend to attract
aquatic life should incorporate measures to reduce the toxicity
(e.g., via pretreatment, dilution) 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
LCso level. The CMC is defined as one-half of the final acute value
for specific toxicants and 0.3 acute toxic unit (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 1 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 1-hour average exposure of organisms passing
through the mixing zone is less than the CMC. The organism
must be able to pass through quickly'or flee the high-concentra-
tion area. The objective of developing water quality recommen-
dations for mixing zones is to provide time-exposure histories that
produce negligible or no measurable effects on populations of
critical species in the receiving system.
Lethality to passing organisms can be prevented in the mixing
zone in one of four ways. The first method is to prohibit concen-
trations 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 met in the pipe whenever a continuous discharge is made to
an intermittent stream. 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 dis-
charge with an initial velocity of 3 meters per second, or
more, together with a mixing zone spatial limitation of 50
times the discharge length scale in any direction, should
ensure that the CMC is met within a few minutes under
practically all conditions. The discharge length scale is defined
as the square root of the cross-sectional area of any discharge
pipe.
A third alternative (applicable to any waterbody) is not to use a
high-velocity discharge. Rather the discharger should provide
71
-------
data to the State regulatory agency showing that the most restric-
tive of the following conditions are met for each outfall:
• 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.
• 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 circum-
stances, including situations of severe bottom interaction,
surface interaction, or lateral merging.
• The CMC should be met within a distance of five 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 under mixing zone design con-
ditions (e.g., low flow for rivers). This restriction will pre-
vent locating the discharge in very shallow environments or
very close to shore, which would result in significant surface
and bottom concentrations.
A fourth alternative (applicable to any waterbody) is for the
discharger to provide data to the State regulatory agency show-
ing that a drifting organism would not be exposed to 1-hour
average concentrations exceeding the CMC, or would not receive
harmful exposure when evaluated by other valid toxicological
analysis, as discussed in Section 2.2.2. Such data should be
collected during environmental conditions that replicate critical
conditions.
For the third and fourth alternatives, 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 de-
tailed analytical calculations, such as modeling estimations of
concentration or dilution isopleths.
The Water Quality Criteria—1972 [11] outlines a method, appli-
cable to the fourth alternative, to determine whether a mixing
zone is tolerable for a free-swimming or drifting organism. The
method 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
intent of the method is to prevent the actual time of exposure
from exceeding the exposure time required to elicit an effect [10]:
ET(X)atC(n)
where T(n) is the exposure time an organism is in isopleth n, and
ET(X) is the "effect time." That is, ET(X) is the exposure time
required to produce an effect (including a delayed effect) in X
percent of organisms exposed to a concentration equal to C(n),
the concentration in isopleth n. ET(X) is experimentally deter-
mined; the effect is usually mortality. If the summation of ratios of
exposure time to effect time is less than 1, then the percent effect
will not occur.
4.3.4 Prevention of Bioaccumulatton 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 waterbody for a bioaccumulative pollut-
ant, care should be taken in calculating discharge limits for this
pollutant or the additivity of multiple pollutants. In particular,
relaxing discharge limits because of 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 concen-
trations across a transect of the waterbody differ by less than 5
percent. The boundary can be determined based on the results of
a tracer study or the use of mixing zone models. Both proce-
dures, along with simple order-of-magnitude dilution calcula-
tions, 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
described 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 ap-
proach. 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 regard-
ing 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 mix-
ing zone analysis should be performed for each. The third pro-
vides 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 [15] and Fischer et al. [16].
Although the models presented in Sections 4.4.4 and 4.4.5 sim-
plify the mixing process, the assessor should have an understand-
ing of the basic physical concepts governing mixing to use these
72
-------
models appropriately. (The U.S. EPA Center for Exposure Assess-
ment Modeling [CEAM] in Athens, Georgia, provides an overview
course that teaches the basics of mixing and how the basics
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 transfor-
mation 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 that have sophisti-
cated hydrodynamic simulation routines coupled with fate simu-
lation 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 efflu-
ent 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 sub-
merged 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 signifi-
cant cross-flow and thus yield high surface concentrations.
Submerged discharges offer more flexibility in meeting the design
goals for toxic discharges. Submerged discharges 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 to considerable scour or deposition.
0.20 n
90° re
0.15 m cast-iron pipe
: J/ ^~ BraCe
Jill S' i
vTl X
/ *\ \ f ° '•'•-•
\
nx 0.15m
ducer elbow Range
op of tremie V"\ Flange joint
ncasement U N /
_| t'
HI . HI
1 " * ^^
ti
/ ^ 0.45 m cast-iron blind /
/ flange with 0.1 5m cast- BMom of tremie ./
I iron trap at invert encasement
0.1 5m 90" elbow
Figure 4-1. A Typical Diffuser Section [17]
Many of the complexities of submerged diffusers have been
summarized by jirka [18], Holley and Jirka [15], and Roberts et al.
[19,20,21 ]. Submerged discharges should be designed to avoid
direct surface impingement and bottom attachment of the sub-
merged jet or jets. Surface and bottom impacts should be
evaluated at critical design conditions (low flow or high stratifica-
tion) and at off-design conditions (higher flow or lower stratifica-
tion) to ensure the best placement and design of the diffuser.
Diffusers provide more dilution than single outlets, but the align-
ment of the diffuser with the receiving water flow direction influ-
ences 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. To deal with the reversing currents of estuaries and
coastal bays, the preferred arrangements for offshore discharges
are parallel diffuser alignment (tee diffuser) and perpendicular
diffuser alignment (staged diffuser) [18]. 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 prefer-
ence 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 [22].
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 appropri-
ate selection of design periods.
1) Rivers and Run-of-River Reservoirs
Rivers and run-of-river reservoirs are waterbodies that have a
persistent throughflow in the downstream direction and do not
exhibit significant natural density stratification. Recommenda-
tions for hydrologically based and biologically based design flows
for completely mixed, steady-state modeling of rivers are de-
scribed in Appendix D of this document. The biologically based
design flows are determined using the averaging periods and
frequencies specified in water quality criteria [8]. Also, the hydro-
logically based flows 1Q10 and 7Q1 0 for the CMC and CCC,
respectively, have been used traditionally and may continue to be
used for steady-state modeling. Run-of-river reservoirs with resi-
dence times less than 20 days at critical conditions also should 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.
2) Lakes and Reservoirs
This receiving water category encompasses lakes and reservoirs
with residence times in excess of 20 days at critical conditions
73
-------
[23], Seasonal variations in the water level, wind speed and
direction, and seasonal solar radiation should be determined to
define the critical period [23]. In the case of long and narrow
reservoirs, areas above the plunge point (i.e., areas 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 ambi-
ent water can vary over seasons, no one season or stratification
condition can be selected as the most critical dilution situation for
alt 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 atten-
tion should be given to periods of rising water level since pollut-
ants 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) Estuaries 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 astro-
nomical, 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 estuar-
ies 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 (i.e., the
design flow). For example, tidal frequency and amplitude vary
significantly in different coastal regions of the United States.
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 salt water are 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 strati-
fication 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 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 maxi-
mum 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 for a critical design for coastal bays are the
same as for stratified estuaries. The period of maximum stratifica-
tion 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 Section 301 (h) Technical Support Docu-
ment [22] and the Section 301 (h) document, Initial Mixing Char-
acteristics of Municipal Ocean Discharges [24]. The following sub-
section contains a summary from these documents. Like dis-
charges to estuaries, discharges to ocean waters are 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 verti-
cal diffusion that occurs in the zone of initial dilution. Thus,
during these periods it is difficult to achieve the recommended
1pO-to-1 dilution that is to occur before the plume begins a
predominantly horizontal flow as compared to vertical flow. Peri-
ods when discharge characteristics, oceanographic conditions
(spring tide and neap tide currents), wet and dry weather periods,
bjological conditions, or water quality conditions that indicate
that water quality standards are likely to be exceeded should also
bb noted. The 10th percentile value from the cumulative fre-
quency of each parameter should be used to define the period of
rninimal dilution.
4.4.3 General Recommendations for Tracer Studies
i
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
ait flow conditions that approach critical flow. Some of those
design conditions are summarized above in the subsections deal-
ing 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 dilution of the wastewater
plume during this critical period. Tracer studies other than dye
studies (e.g., chloride, lithium) can be performed for cases in
Which the receiving water is amenable to such tests.
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 that reflect both discharge-induced
and ambient-induced mixing. The isopleth concentrations, 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, whether the required zone of passage is provided, and
vyhether the plume avoids critical resource areas. The WLA can
then be calculated to provide the appropriate zone of passage
abd to prevent detrimental impacts on spawning grounds, nurs-
eries, 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
74
-------
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 [15] and
lateral and vertical mixing in estuaries, bays, reservoirs, and lakes.
For waterbodies with significant flow velocities, continuous re-
leases 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 condi-
tions, the observed data can be used to determine dimensionless
mixing coefficients. These coefficients can then be extrapolated
to critical conditions using hydraulic parameters [15]. A tracer
study at near-to-critical conditions also can 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 de-
sign, conduct, and analysis of tracer studies for mixing analyses.
Techniques of Water-Resources Investigations of the USCS provides
the best overview of how to conduct tracer studies [25, 26, 27].
The fluorescent dyes (usually Rhodamine WT), measuring equip-
ment, fluorometers, field and laboratory procedures, and calcula-
tion 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 [28]; Kisiel
[29]; Holley and jirka [15]; and Yotsukura, Fisher, and Sayre [30].
Examples of tracer studies in tidal systems are presented in Wilson,
Cobb, and Yotsukura [31] and Hetling and O'Connell [32], both
of which are studies of the Potomac River estuary; Baily [33], a
study of Suisun Bay in California; Fischer [34], a study of Bolinas
Lagoon, a coastal bay in Marin'County, California; and Crocker et
al. [35], a study of Corpus Christi Bay, Texas. Methods to perform
a tracer study in a reservoir are provided in Johnson [36],
The dye study recommended for obtaining a quick saltwater
dilution assessment is one in which Rhodamine WT dye is admin-
istered 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, not the 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 re-
ceiving water). These include low wind, neap tide (tide of mini-
mum 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 the
amount and variability of background fluorescence in the receiv-
ing water are taken into account. Measurements of dye concen-
tration are made using a fluorometer and should be corrected for
water temperature.
A survey of background fluorescence and its variability in the
anticipated mixing zone must be conducted just prior to the
beginning of the study in order to permit correction of fluores-
cence data and to determine the dye concentration required in
the effluent. Since Rhodamine WT dye is bleached by free chlo-
rine, 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 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 was performed on the Amelia River,; an estuarine
system in Florida [2]; the second was performed on the Green-
wich Cove, an embayment of Narragansett Bay in Rhode Island
[37]. In both studies, Rhodamine WT dye was introduced con-
tinuously 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.
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 gener-
ally 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
near-field instabilities occur. These instabilities, associated with
surface and bottom interaction and localized recirculation cells
extending over the entire water depth, can cause buildup 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 [13].
In the absence of near-field instabilities, horizontal or nearly hori-
zontal discharges will create a clearly defined jet in the water
column that will initially occupy only a small fraction bf. the
available water depth. The following equations and models are
designed to describe mixing under stable near-field 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) [15]:
s-'0.3-3-
where
S = flux-averaged dilution
x = distance from outlet
d = diameter of outlet.
75
-------
receiving water surface
diffusor
receiving water bottom
c) Deep-water, high-buoyancy, nonvertical discharge
receiving water surface
dilfuser
X 1 ful1
J vertical
I mixing
receiving
water
bottom
d) Shallow-water, low-buoyancy, nonvertical discharge
Figure 4-2. Example of Discharge-Induced Mixing [7]
The coefficient 0.3 represents the average of two values derived
from the literature, 0.28 [16] and 0.32 [38].
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. Dilution may be underestimated
for partially full pipes because the equation assumes a fully flow-
ing pipe. 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 corresponding
to several (two to three) 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 [15], Fischer et al. [16], and Wright [39]. 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 tech-
niques. 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 buoy-
ant jet models require the following input data: discharge depth,
effluent flow rates, density of effluent, density gradients in receiv-
ing water, ambient current speed and direction, and outfall char-
acteristics (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 that are available
through EPA. All of the models require a user who is well versed in
mixing concepts and the data necessary to run the models. The
first model, CORMIX [40, 41], may be the most useful to regula-
tors since it is an expert system that guides the user in selecting an
appropriate modeling strategy for rivers or estuaries. It is available
from the National Technical Information Service (NTIS), and user
support is available from the U.S. EPA CEAM. 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 environ-
ments where the ambient water current velocity is zero (e.g.,
lakes, 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 to
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 [24]. These five
models have been modified such that the user inputs the data
into a universal data format that allows the user to apply any of
the five models with only minor input changes.
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 operates on microcomputers with the
MS-DOS operating system. CORMIX1 can summarize dilu-
tion 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 CORMIX1
program guides the user, based on the user's input, to
76
-------
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 submerged single-port discharge. Buoy-
ancy between the effluent and ambient water can be mod-
eled, 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 ambient current speed varia-
tions with depth. The ports are assumed to be equally
spaced. The model accounts for adjacent plume interfer-
ences over the course of the plume trajectory and in the
subsequent dilution calculation. Positive buoyancy is ac-
counted for, and the discharge vertical angle can be modi-
fied. 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 capa-
bilities of UMERGE plus the ability to simulate instances
where the ambient current flow is not perpendicular to the
diffuser.
• ULINE models a vertical slot jet discharge into a flowing
waterbody. The discharge angle is assumed to be perpen-
dicular to ambient current. The ambient current may vary
with depth, and the axis of the diffuser may range from
parallel to perpendicular to the ambient current. The buoy-
ancy of the effluent can also be modeled.
An evaluation and comparison of all these models can be found in
the Technical Guidance Manual for Performing Wasteload Alloca-
tions—Book 3, Estuaries [7].
4.4.5 Ambient-induced Mixing
The equations for discharge-induced mixing can be used to pre-
dict 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 concentra-
tion 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 also
may be needed to estimate concentrations encountered at impor-
tant 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 [42]. 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 prima-
rily 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 (Xm) required to achieve
the completely mixed condition may be estimated from an equa-
tion of the form [16]:
v mW2u
where
W = width of the river
u = flow velocity for the critical design flow
Dy = lateral dispersion coefficient as discussed below
m = 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 varia-
tion in concentration across the stream width, the value of m
would be approximately 0.1 for a discharge near the center of
river flow (not the center of river width) and approximately 0.4 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 values for m
would be approximately 0.06 for a discharge near the center of
river flow and approximately 0.24 for a discharge near the edge of
the river. For a very small stream, Xm may be only a few hundred
feet; for medium and large streams, Xm is normally several miles
to several tens of miles.
The lateral dispersion coefficient (Dy) for most rivers can be
calculated with the following equation [16]:
Dy =0.6du*±50%
where
d = water depth at design flow
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
77
-------
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 [42]. The following equation for shear velocity should
be used [16]:
u* = (gds)1/2
where
g a acceleration due to gravity
s = slope of the channel
d s 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 Xm can be smaller than the ones indicated
here. For distances greater than Xm, the models for completely
mixed effluents discussed in Section 4.5 can be used to calculate
concentrations at these distances. For shorter distances, maxi-
mum concentrations can be much greater than those predicted
by "completely mixed" models and should be estimated using
the following equation:
where
Cx =
Ce:
Qe =
Qs =
Dy =
X =
w =
U :
QsOrDyX/u)1/2
maximum pollutant concentration distance x from
the outlet
effluent concentration
design effluent flow
design stream flow
lateral dispersion coefficient
distance from the outlet
stream width
flow velocity for the design flow.
It should be noted that this estimate of Cx is a worst-case predic-
tion since the equation assumes 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 [42]. To use this approach,
however, the value of Dy and pollutant concentrations after dis-
charge-induced mixing must be known from tracer studies and/
or from the use of one of the discharge-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 receiv-
ing water where the plume attaches to both bottom and nearshore
[43]. 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 by
means of 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 modifi-
cation 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 the time required for
the effluent to pass out of the unstratified part of the estuary, the
time required for the effluent to pass into a segment of greatly
changed cross-section, or the time required for the substance to
depy. 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 the tidal velocity; thus Uj,
which is based on an average total velocity; is substituted for u in
the equations and shear velocity becomes
i1
\ u* = 0.10ut.
The CORMIX expert system model can also be used to obtain
predictions for 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 down-
stream from the source. WLAs for toxic chemicals require a
different approach, however, because critical environmental con-
ditions 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 calculat-
ing toxic WLAs. However, if complete mixing does not occur near
the discharge point and the effluent plume is discernible downriver,
then modeling techniques that can simulate and predict mixing
conditions are more appropriate. The mixing zone models pre-
sented 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, bio-
logical, and physical processes.
i •
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 concen-
tration (RWC), receiving water flow, and meteorological condi-
tions (e.g., temperature). The frequency and duration of ambient
concentrations predicted with a steady-state model must be as-
sumed to equal the frequency and duration of the critical receiv-
78
-------
ing water conditions used in the model. The variability in effluent
flows and concentrations also affects RWCs, but these effects
cannot be predicted with constant inputs. Steady-state models
can be improved for toxic WLAs by means of the following:
• Using design flows that will ensure criteria compliance at
the appropriate duration and frequency.
• Calculating both acute and chronic WLAs.
EPA is encouraging the States to adopt two-number aquatic life
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 (criteria continuous concentration,
or CCC) for each toxicant, as well as 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 durations 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 frequency 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, in which 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 [6]. The advan-
tages of both types, as well as how they may be calculated, also
are described in Appendix D.
A 4-day, 3-year biological design flow does not equate to a 4Q3
hydrological design flow. EPA has determined that a 4Q3 design
flow would result in an excessive number of water quality criteria
exceedances. As explained in Appendix D, a hydrologically based
7Q10 will, for most streams, be similar to a biologically based 4-
day, 3-year design flow.
At the present time, there are no recommended toxicological
flows for steady-state modeling of lakes, reservoirs, or estuaries.
The design conditions recommended for these waterbodies in
Section 4.4.2 are based on hydrological and meteorological con-
ditions rather than on toxicological duration and frequency data.
These conditions should be used until further guidance is pro-
vided.
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 permit writer to calculate the more limiting long-term average
(LTA) for the treatment system and develop permit limits protec-
tive 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 pro-
gram called DESCON was developed. (See Appendix D for
additional information.) 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 stream flow 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 stream flow, temperature, pH, alkalin-
ity, 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 fre-
quency.
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 stream flow 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 use both the extreme value and the biologically
based methods of calculating of water quality excursions; and it is
specifically designed to handle such pollutants as ammonia, heavy
metals, pentachlorophenol, and biochemical oxygen demand
(BOD) for which 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 10 years of historical daily flow data and it can
only analyze a single discharger, 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
79
-------
Stream Flow
Allowable Stream Loading
Temperature
Dilution model [and
water quality criteria
(WQC) equatibns
Days
Days
Stream Flow
Design Temperature
Critical event
Number of Excursions
Days
Figure 4-3. Computational Scheme for Deriving Design Conditions
In the design conditions if these conditions reflect the desired
toxfcological effects regime. Dynamic modeling techniques ex-
plicitly predict the effects of receiving water and effluent flow and
of concentration variability. The three dynamic modeling tech-
niques recommended by EPA for WLAs are continuous simulation,
Monte Carlo simulation, and lognormal probability modeling.
These methods calculate a probability distribution for RWCs rather
than a single, worst-case concentration based on critical condi-
tions. 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 condi-
tions 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 daily and
monthly average permit limits can be obtained directly from the
efficient LTA concentration and coefficient of variation (CV) that
characterize this distribution. Generally, steady-state modeling
has been used to calculate only a chronic WLA. Steady-state
modeling 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 RWCs
rather than a single, worst-case condition. As a result, maximum
daily and monthly average permit limits, based on compliance
80
-------
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) with daily receiving water flow (Qj) and
background concentration data (Cs) to calculate downstream
RWCs [44]. The model predicts these concentrations in chrono-
logical order with the same time sequence as the input variables
(Cb versus time). The daily RWCs can then be ranked from the
lowest to the highest without regard to time sequence. A prob-
ability plot can be constructed from these ranked values, and the
occurrence frequency of any 1 -day concentration of interest can
be determined (Cb versus frequency). Running average concen-
trations for 4 days (i.e., the chronic design flow), or for any other
averaging period, also can be computed from the daily concen-
trations (Figure 4-5).
The probability plot generated by the continuous simulation model
using existing effluent data will indicate whether criteria are pre-
dicted to be exceeded more frequently than desired. Appendix D
discusses how to select the appropriate allowed frequency of
excursions 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 data also should be
examined to determine if there were any long-term changes due
to technology-based treatment or periodic changes due to indus-
trial or municipal plant closings or expansions. The same data
requirements are also true for the lognormal probabilistic and
Monte Carlo methods. However, except for the continuous
simulation models, other nonsteady-state models in this section
O 20
99 99.5 100
Percent of Time Concentration Is Less Than or Equal
to Concentration Plotted
0.5 ,1 2 51042
i__ t : 1 1 UJi
f ; 20 '
Recurrence Interval (years)
Figure 4-4. Frequency of Occurrence of Concentrations in
Receiving Waters and Recurrence Intervals Generated
by a Continuous Simulation Model
cannot be used to account for the duration and frequency provi-
sion of the two-number water quality criteria. 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:
• The frequency and duration of toxicant concentrations in a
receiving water can be predicted.
• The cross-correlation and interaction of time-varying pH,
flow, temperature, pollutant discharges, and other param-
eters are incorporated.
• The effect that the serial correlation of daily flows and other
parameters has on the persistence of criteria excursions is
incorporated.
• Long-term stream flow records for ungauged rivers using
precipitation and evapotranspiration data can be synthe-
sized.
• Long simulation times can prevent the initial conditions
used in the model from affecting the calibration of fate and
transport processes.
Unlike steady-state models, continuous simulation models require
significantly more data to apply, to calibrate, and/or to verify a
specific problem and require that input information for the appli-
cation of the model be time-series data. Also, the model results
need manipulation to calculate the effluent LTA concentration
and CV for use in developing effluent limits.
Monte Carlo Simulation Models. Monte Carlo simulation com-
bines 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 [45-51 ]. 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
RWCs must be predicted for CMC comparisons, probability distri-
butions of daily model input data are needed for Monte Carlo
simulation. If 4-day 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 RWCs. These concentrations do not follow the
temporal sequence that is calculated with the continuous simula-
tion 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 [52].
The approach for calculating the allowable pollutant load distri-
bution using Monte Carlo simulation is the same as that described
for the continuous simulation model. The advantages of Monte
Carlo simulation are the following:
• It can predict the frequency and duration of toxicant con-
centrations in a receiving water.
81
-------
• It can be used with steady-state or continuous simulation
models that include fate processes for specific pollutants.
• It can be used with steady-state or continuous simulation
models that include transport processes for rivers, lakes,
and estuaries.
• It can be used with steady-state or continuous simulation
models that are designed for single or multiple pollutant
source analyses.
* It does not require time series data.
* It does not require model input data to follow a specific
statistical distribution or function.
• 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 disadvantages of Monte Carlo simulation are that it
requires more input, calibration, and verification data than do
steady-state models, and the model results need manipulation to
calculate the effluent LTA concentration and CV to develop efflu-
ent limits.
lognonmaLEfobabJUstic Dilution Model. Without resorting to
the continuous simulation method of computing RWCs in tempo-
ral sequence, this probabilistic method uses the lognormal prob-
ability distributions of the input variables to calculate probability
distributions of output variables [53]. As a result, the method
requires only the relevant statistical parameters of the input vari-
ables (medians and coefficients of variation) rather than the actual
time series data needed for continuous simulation. If 1-day
average RWCs 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 concentra-
tion of a substance only after complete mixing and before degra-
dation or transformation significantly alters the concentration.
The lognormal probabilistic dilution model has the following
advantages:
• It can predict the frequency and duration of toxicant con-
centrations in riverine environments.
• It does not require time series data.
• 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 lognormal probability dilution model has the following disad-
vantages:
• It requires more input than a steady-state model.
Qs
Cs
Qs
Frequency
; Figure 4-5. Concentration Frequency Curves
V It does not include instream fate processes.
1
« It applies only to rivers and streams.
'
« It analyzes multiple pollutant sources inaccurately.
« It requires model input data to be lognormally distributed.
4.5,2 Calculating the Allowable Effluent Concentration
Distribution ami the Return Period
Information concerning effluent concentration means and vari-
abilities can be obtained from data bases 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 ap-
proaches 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 mag-
82
-------
nitude 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 efflu-
ent concentrations. The other approach assumes that the re-
quired effluent concentration distribution is the same as the exist-
ing distribution except that it is reduced in magnitude by which-
ever is greater—the percentage necessary for the 1 -day average
concentrations to meet the CMC, or the 4-day average concen-
trations 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 con-
centration distribution is based on the assumption that effluent
concentrations after treatment will not have the same CV as
concentrations before treatment. Studies have documented that
advanced secondary treatment increases the CV of BOD and total
suspended solids concentrations compared to secondary treat-
ment. Where feasible, investigations should be conducted to
evaluate how treatment processes for heavy metals, organic chemi-
cals, 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 efflu-
ent input data since daily effluent concentrations will not be
available for the hypothetical treatment schemes. The required
"future treatment" mean and CV 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 automatically creates
input for continuous simulation models, randomly selects the sets
of input data required for Monte Carlo simulations, and performs
the numerical integration calculation for the lognormal probabi-
listic model. DYNTOX is available from the EPA CEAM, Environ-
mental Research Laboratory (ERL) [54]. If the observed data base
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 ex-
press 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 occur-
rences within a year.
4.5.3 General Recommendations lor Motel Selection
The reliability of the predictions from any of the modeling tech-
niques depends on the accuracy of the data used in the analysis.
The minimum data required for model input include receiving
water flow, effluent flow, effluent concentrations, and background
concentrations. In many locations, stream flow data should be
sufficient for both steady-state and dynamic models. At least 30
years of flow data should be available if excursions of the CMC
and CCC must be evaluated at rare frequency 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 avail-
able, steady-state assessments should be used. Modeling also
should 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 dy-
namic modeling techniques should be used to develop more
cost-effective treatment requirements. If the'effluent data exhibit
significant seasonal differences or batch process trends, the con-
tinuous simulation approach may be the easiest dynamic model-
ing 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 simula-
tion approach.
4.5.4 Specific Model Recommendations
The following section recommends models for toxicity and indi-
vidual 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 [5, 6, 7] and Office of Research
and Development [55]. These manuals describe in detail the
83
-------
transport and transformation processes involved in water quality
modeling. Transport processes include the dispersion and advec-
tion of a contaminant once it enters the receiving stream; its
volatilization from the water; and its sorption to suspended sedi-
ment, eventual settling, and possible resuspension and diffusion
from the sediment. Transformation processes include the oxida-
tion, 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 sorne of those listed
below (EXAMS-II, WASP4) are now capable of simulating water
column-sediment interactions (resuspension, settling, and diffu-
sion), however, additional work needs to be completed on the
mechanisms of sediment-water column exchange before the mod-
els can be validated for predictive applications involving sedi-
ments. 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 CEAM [56]. All the models, plus two
bioaccumulation models, briefly are described below.
• DYNTOX [54] is a WIA model that uses a probabilistic
dilution technique to estimate receiving water 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, tempera-
ture, solids, pH, or light.
• EXAMS-II [57] is a compartment model that can be used as
either a steady-state or 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 trans-
port fields. Sediment exchanges with the water column
include pore-water advection, pore-water diffusion, and
solids mixing. The last describes a net steady-state ex-
change associated with solids that is proportional to pore-
water diffusion.
WASP4 [58] is a generalized modeling framework for con-
taminant -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 estuar-
ies, 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. Sedi-
ment 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 [59] 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 trans-
port and transformation in one-dimensional stream chan-
nels. The simulation provides a time history of the runoff
flow rate, sediment load, and nutrient and pesticide con-
centrations, 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 specific
Table 4-2. Toxicant Fate and Transport Models
Model
DYNTOX
EXAMS-II
WASP4
HSPF
SARAH2
MINTEQA2
Environment
river
lake, river,
estuary
lake, river,
estuary
river
river
lake, river,
estuary
Time Domain
dynamic
steady-state, j
quasi-dynamic
steady-state,
dynamic
dynamic
steady-state
steady-state
84
Spatial Domain
far field,
1 -dimensional
far field,
3-dimensional
"far field,
3-dimensional
far field
1 -dimensional
treatment plant,
near field,
2-dimensional
~~
Chemical
organic,
metal
organic
organic,
metal
organic,
metal
organic
metal
-------
organic chemicals and transformation products of those
chemicals. The reaction and transfer processes included are
hydrolysis, oxidation, photolysis, biodegradation, volatiliza-
tion, and sorption. Sorption is modeled as a first-order
kinetic process in which a desorption rate and an equilib-
rium 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 ca-
pacity 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 capac-
ity. Sediment exchanges with surficial benthic sediments
are modeled as sorption/desorption and deposition/scour.
Underlying sediment and pore water are not modeled.
• SARAH2 [60] is a steady-state, near-field model for calculat-
ing acceptable concentrations of hazardous organic chemi-
cals 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 batch waste streams, SARAH2 considers the following
concentration reductions: dilution and loss during treat-
ment, initial Gaussian mixing at the edge of a stream,
lateral and longitudinal diffusion in the mixing zone, sorp-
tion, volatilization, hydrolysis, and bioaccumulation in fish.
The user must specify appropriate concentrations for pro-
tection of the aquatic community and of humans exposed
through consumption of fish and water. The benthic com-
munity is not presently considered. Treatment loss is handled
empirically. SARAH2 contains data sets for three disposal-
watershed scenarios that can be easily modified and em-
ployed. The model is designed for screening analysis and
contains numerous assumptions that should be verified
before the model is used in actual cases.
• M1NTEQA2 is an equilibrium metals speciation model for
dilute aqueous systems [61 ]. It does not have any transport
and transformation processes and must be run 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.
« FGETS is a toxicokinetic model that simulates the
bioaccumulation of nonpolar organic chemicals by fish from
both water and food [62]. Both of these routes of ex-
change are modeled as diffusion processes that depend
upon physicochemical properties of the pollutant and mor-
phological/physiological characteristics of the fish. FGETS
contains a moderately sized data base of allometric relation-
ships for gill morphology with which it can simulate the
direct gill/water exchange of organic chemicals for essen-
tially any fish species, assuming certain default values, FGETS
also contains a limited data base of physiological/morpho-
logical relationships that are used to set parameters for 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 calcula-
tion 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 postprocessor to WASP4.
• FCM2 is a generalized model of the uptake and elimination
of toxic chemicals by aquatic organisms [63]. 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. Expo-
sure to the toxic chemical in food is based on a consump-
tion 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 concen-
trations 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 concen-
tration model WASP4. Thus FCM2 may be executed as a
separate model or as a postprocessor to WASP4. Migratory
species, as well as nonmigratory species, may be consid-
ered. 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 mea-
surements should be converted to toxic units (TUs). As discussed
in Chapters 1, 2, and 3, 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 concen-
tration is inverse; chemicals that have toxic effects at low concen-
trations 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, i.e., acute toxicity end-
point (ATE) or chronic toxicity endpoint (CTE). An acute toxic
unit (TUa) is the reciprocal of an ATE. A chronic toxic unit (TUC) is
the reciprocal of a CTE. The TMDL 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 1 hour.
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)CTUc)
where the acute-to-chronic ratio (ACR) is 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 1 0 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 Chapters.
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 toxic-
ity, it is recommended that effluent toxicity be limited to dilution
estimates and that toxicity be assumed to be additive and conser-
vative. Toxicity is expected to be additive even when the toxicity
of one effluent affects selected biota while the toxicity of a down-
stream 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:
If jnstream toxicity measurements are available and a first-order
decay rate for toxicity can be estimated, the following equation
should be used:
where
C
Cs
Qs
Ce
Qe + Qs
downstream concentration (TUC or TUa )
upstream concentration CTUC or TUa )
upstream flow (cfs)
effluent concentration (TUC or TUa ) and
effluent flow (cfs).
For multiple dischargers, this equation must be applied sequen-
tially to find the concentration as a function of distance down-
stream. 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, log-
normal probabilistic, or Monte Carlo methods. For these dy-
namic analyses, a series of C& Qe, Cs, and 0$ values would be
used.
where
C = downstream concentration (TUC or TUa)
j Co = 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
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.
Trie same equations used for toxicity ahalyses in rivers can also be
used in steady-state, continuous simulation, or probabilistic analy-
sis of long, narrow, shallow impoundments with high inflow
velocities. Wider, deeper lakes require more complicated analy-
ses since prolonged detention times (>20 days) and stratification
exert a significant impact on water quality. The prolonged deten-
tion times make it essential that receiving water measurements of
toxicity be available to estimate decay factors. These measure-
ments should be made at least once per 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 equa-
tions [64]:
TW = V/Q
,. C =Cin/(1+TwK)
E • • '
where
Tw = mean hydraulic residence time
V = lake volume at design conditions
! Q ,= mean total inflow rate at design conditions
C = steady-state lake concentration CTUC or TUa)
' Cjn = steady-state inflow concentration (TUC or TUa)
; 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). The
miean total inflow rate (Q) and the inflow concentration (Cjn)
should be calculated as the sum of all sources to the lake, includ-
ing point source, nonpoint source, and tributary inputs.
t
Dilution calculations for effluent toxicity discharges to an estuary
are complicated by the oscillatory motion of the tides and pos-
sible 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 moni-
tored decay rates for toxicity. A simple steady-state analysis can
be performed using the following equations for each
nbnconservative pollutant entering from the river at the head of
ar) estuary [64]:
86
-------
(fi)
where
t
f I
Cj
1 1-(1-n)e-kt
exchange ratio for segment i as defined by modified
tidal prism method
flushing time
fraction of freshwater in segment i
nonconservative pollutant concentration in segment
i(TUaorTUc)
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:
For segments upstream of outfall:
where
Cj = nonconservative pollutant mean concentration in
segment i (TUC or TUg)
C0 = nonconservative pollutant mean concentration in
segment of discharge
TJ = exchange ratio for segment i as defined by the
modified tidal prism method
n = number of segment away from outfall
fj = fraction of freshwater in segment i
f0 = fraction of freshwater in segment with discharge
Sj = salinity in segment i
S0 = salinity in segment of discharge
k = decay rate
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 assess-
ment manual [64]. 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 reference ambient concentration
(RAC) from the formulas in Chapter 2, Water Quality Standards.
This information should be used to estimate exposures due to fish
consumption and drinking water ingestion, background concen-
trations, and other exposure routes, such as recreational, occupa-
tional, 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 also can be modified.
4.6.2 Determining the TMDL Based on Human Health Toxicants
TMDLs are typically 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 RAC is applied at the end of pipe and no TMDL
determination is typically necessary.
With persistent or bioconcentratable pollutants, special mixing
zone considerations apply. Bioconcentratable pollutant criteria
exceedances within the mixing zone can potentially result in
tissue contamination of organisms directly or indirectly through
contamination of bed sediments with subsequent incorporation
into 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 TMDL should be calculated to achieve the
RAC or criterion selected above [65]. Because most human health
criteria are chronic only, a TMDL to protect against acute effects
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. These models may be used by States for surface waters in
appropriate instances.
In simple situations, the TMDL is determined from the RAC 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 a mixing zone if it is reasonable to assume
that the bioconcentrating aquatic organisms have little mobility,
thus spending most of their time within the mixing zone; and 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
87
-------
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 mean dose. The criteria thus apply to
the ambient water concentrations averaged over a 70-year pe-
riod.
The duration of exposure assumed in deriving criteria for
noncarcinogens may be ambiguous, particularly where a 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: River and Stream Discharge Situations
In well-mixed situations, the RWC, C, is determined by the pollut-
ant load, W (mass/time), and the combined receiving water plus
effluent flow, Q, such that, C = W/Q.
The long-term harmonic mean flow is recommended as the
design flow for carcinogens. The recommendation of long-term
harmonic mean flow has been derived from the definition of the
human health criteria (HHC) for carcinogenic pollutants. The
adverse impact of carcinogenic pollutants is estimated in terms of
receptors (human) lifetime intakes. To be within the acceptable
level of life-time body-burden of any carcinogen, such intakes
should not exceed the HHC during the average life-time of the
receptor. A life-time for exposure to carcinogenic pollutants is
defined as 70 years, or approximately 365 (days/year) multiplied
by 70 years.
The HHC for carcinogenic pollutants can be numerically expressed
as:
HHC = C (design) = (q + C2 + C3 + — + Cn )/n
where
n = (365 days/year) x 70 years
C « concentrations
Based on an assumption of a constant daily load from a treatment
facility, the fully mixed instream concentration will go up or down
Inversely with the ups and downs of receiving water flows. There-
fore, instream concentration is a function of, and inversely pro-
portional to, the streamflow downstream of the discharge. Using
this concept, 1/Q can be substituted for C, as follows:
1 /Q (design) = (1 /Qt + 1 /Qa + 1 /Q3 + — + 1 /Qn)/n.
The stream design flow (Q design) can then be shown as follows:
Q (design) = n/(1/Ql+1/Q2 + 1/Q3 + — + 1/Qn)
The harmonic mean is expressed as follows:
I n
, Q (design) = n/I(1/Qj)
where
n = the number of recorded flows.
i
The harmonic mean is always less than the arithmetic mean. The
harmonic mean is the appropriate design flow for determining
long-term exposures using steady-state modeling of effluents.
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.
However, for situations involving seasonably variable effluent dis-
chaVge rates, hold-and-release treatment systems, and effluent-
dominated sites, the harmonic mean may not be appropriate. In
these cases, the effluent load and downstream flow are not inde-
pendent (i.e., they are correlated). Modeling techniques that can
calculate an average daily concentration over a long period of
time are more appropriate to determine the long-term exposure
in these cases.
The harmonic mean flow may be estimated by any of several
methods [8], assuming that flows are approximately lognormally
distributed: Q 2
Qhm="Q!r7
where
9gm is tr)e geometric mean flow
Qam is the arithmetic mean flow.
For U.S. Geological Survey flow records, summaries of the statisti-
cal 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.
WQAB DFLOW is a software package available for computation
of harmonic mean flow. The DFLOW program (as discussed
below and described in Appendix D) should be used with data
that are not lognormally distributed.
To develop some quantitative sense of how a long-term harmonic
mean flow of any stream compares with its 7Q10 flow, the
Assessment and Watershed Protection Division and the Risk Re-
duction Engineering Laboratory at Cincinnati, Ohio, analyzed
flow records of 60 streams selected at random throughout the
United States. These are the same stream flow records that had
been analyzed for stream design flow condition for aquatic life
protection as listed in EPA guidance [8]. 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 two 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. How-
88
-------
ever, it is recommended that the harmonic mean flow be calcu-
lated directly from the historical daily flow record, if possible.
Alternatively, the following equation might be used to estimate
harmonic mean flow [66]:
Qhm = [1 -194 * (Qam)0'473] * K?Q10)°-552], r2 = 0.99.
In this equation, Qam and 7Q10 are estimated using the U.S.
Geological Survey computer program, FLOSTAT.
b) Noncardnogens: River and Stream Discharge Situations
The choice of average period 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 short-
ened exposures, EPA suggests the use of 30Q5. However, in the
comparisons of flows for smaller rivers (i.e., low flow of 50 cfs), the
30Q5 flow was, on the average, only 1.1 times that of the 7Q10.
For larger rivers (i.e., low flow of 600 cfs), the factor was, on the
average, 1.4 times. If the effects from certain noncarcinogens
are manifested after a lifetime of exposure, then a harmonic
mean flow may be appropriate.
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:
• Where State standards allow no mixing zone and no spatial
averaging, the criterion would be met at the end of the
pipe.
• 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.
• Where State standards allow consideration of spatial aver-
aging, the criterion may be met as an average within a
specified area, as appropriate for the individual and aggre-
gate risk scenarios underlying the application.
89
-------
CHAPTER 4
REFERENCES
1. Oregon Environmental Quality Commission. 1988. Memo-
randum 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 Phospho-
rus and Ammonia in the Tualatin River Basin (340-41 -470).
Portland, OR.
2. U.S. EPA. 1987. Reichenbach N.G., C.B. Wickramanayake,
A.R. Gavaskar, J.J. Walsh, R.S. Hayward, and R.M. Vaya.
Amelia River Pilot Study for Implementing EPA's Toxic Control
Policy. U.S. EPA, Office of Water Regulations and Stan-
dards, Washington, DC.
3. U.S. EPA. 1983. 40 CFR Part 131, Revised Water Quality
Standards Regulation. Federal Register, volume 48, No-
vember 8,1983.
4. U.S. EPA. 1985. 40 CFR Parts 35 and 130, Final Rule. Water
Quality Planning and Management. Federal Register, vol-
ume 50, January 11,1985.
5. U.S. EPA. 1984. 'Toxic Substances." Technical Guidance
Manual for Performing Wasteload Allocations. Book 2, Streams
and Rivers. U.S. EPA, Office of Water Regulations and
Standards, Washington, DC. EPA 440/4-84-022.
6. U.S. EPA. 1986. 'Toxic Substances." Technical Guidance
Manual for Performing Wasteload Allocations. Book 4, Lakes,
Reservoirs, and Impoundments. U.S. EPA, Office of Water
Regulations and Standards, Washington, DC. EPA 440/4-
87-002.
7. U.S. EPA. 1989. Draft Technical Guidance Manual for Perform-
ing Wasteload Allocations. Book 3, Estuaries. U.S. EPA,
Office of Water Regulations and Standards, Washington,
DC.
8. U.S. EPA. 1986. "Stream Design Flow for Steady-State
Modeling." Technical Guidance Manual for Performing
Wastehad Allocations. Book 6, Design Conditions. U.S. EPA,
Office of Water Regulations and Standards, Washington,
DC.
9. Chadderton, R.A., A.C. Miller, and A.J. McDonnell. 1981.
Analyses of Waste Load Allocation Procedures. Water Re-
sources Bulletin 17(5):760-66.
10. NAS. 1972. Water Quality Criteria, 1972. National Academy
of Sciences. U.S. EPA, Washington, DC. EPA-R3-73-033.
11. U.S. EPA. 1988. Mixing Zones Water Quality Standards
Criteria Summaries: A Compilation of State/Federal Criteria.
U.S. EPA, Office of Water Regulations and Standards, Wash-
ington, DC. EPA 400/5-88/015.
12. Brungs, W.A. 1986. Allocated Impact Zones for Areas of Non-
Compliance. U.S. EPA, Region 1. Water Management
Division, Boston, MA.
13. Versar. 1984. .Draft Assessment of International Mixing Zone
Policies. Avoidance/Attraction Characteristics, and Available
' Prediction Techniques. U.S. EPA, Office of Water Regula-
tioqs and Standards and EPA Office of Pesticides and Toxic
] Substances, Washington, DC.
if U.S. EPA. 1983. Water Quality Standards Handbook. U.S.
| EPA, Office of Water Regulations and Standards, Washing-
i ton, DC.
15. Holley, E.R., and G.H. Jirka. 1986. Mixing in Rivers. Technical
Report E-86-11. U.S. Army Engineer Waterways Experi-
I ment Station, Vicksburg, MS.
16. Fischer, H.B., E.J. List, C.Y. Koh, J. Imberger, and N.H. Brooks.
1979. Mixing in Inland and Coastal Waters. New York:
Academic Press. ,
17. Metcalf & Eddy, Inc. 1979. Wastewater Engineering: Treat-
ment, D/sposo/,/tewse. New York: McGraw-Hill Book Com-
; pany.
ib. Jirka, G.H. 1982. Multiport Diffusers for Heat Disposal: A
Summary. /. of the Hydraulics Division. Proc. ASCE,
: 108(HY12): 1425-68.
19. Roberts, P.J.W., W.H. Synder, and D.J. Baumgartmer. 1989a.
| Ocean Outfalls 1: Submerged Wastefield Formation. ASCE.
| /. Hydraulic Engineering 15(1 ):1 -25.
2(3. Roberts, P.J.W., W.H. Synder, and D.). Baumgartmer. 1989b.
i Ocean Outfalls 2: Spatial Evolution of Submerged
: Wastefield. ASCE. /. Hydraulic Engineering, 115(1 ):26-48.
21. Roberts, P.J.W., W.H. Synder, and D.). Baumgartmer. 1989c.
i Ocean Outfalls 3: Effect of Diffuser Design on Submerged
•[ Wastefield. ASCE. /. Hydraulic Engineering 115(1 ):49-70.
I
22. U.S. EPA. 1982. Revised Section 301 (h) Technical Support
: Document. U.S. EPA, Office of Water Program Operations,
Washington, DC. EPA 430/9-82-011.
23. Ford, D.E., and LS. Johnson. 1986. An Assessment of Reser-
voir 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.
24., Muellenhoff, W.P., A.M. Soldate, D.J. Baumgartner, M.D.
Schuldt, L.R. Davis, and W.E. Frick. 1985. Initial Mixing
Characteristics of Municipal Ocean Discharges: Volume 1,
Procedures and Applications. U.S. EPA, Environmental Re-
search Laboratory, Narragansett, Rl. PB 86-137478.
25. Hubbard, E.F., F.A. Kilpatrick, LA. Martens, and J.F. Wilson.
i "Techniques of Water-Resources Investigations of the
USGS." Measurement of Time of Travel and Dispersion in
i Streams by Dye Tracing. Book 3, Application of Hydraulics.
j U.S. Department of the Interior, Reston, VA.
26. Kilpatrick, F.A., and E.D. Cobb. "Techniques of Water-Re-
; sources Investigations of the USGS." Measurement of Dis-
': charge Using Tracers. Book 3, Application of Hydraulics. U.S.
• Department of the Interior, Reston, VA.
90
-------
27. Wilson, J.F., E.D. Cobb, and F.A. Kilpatrick. "Techniques of
Water-Resources Investigations of the USGS." Fluorometric
Procedures for Dye Tracing. Book 3, Application of Hydraulics.
U.S. Department of the Interior, Reston, VA.
28. Fischer, H.B. 1968. Methods for Predicting Dispersion Coeffi-
cients in Natural Streams with Applications to Lower Reaches
of the Green and Duwamish Rivers, Washington. U.S. Geo-
logical Survey Professional Paper 582-A, U.S. Geological
Survey, U.S. Department of the Interior, Reston, VA.
29. Kisiel, C.C. 1964. Dye Studies in the Ohio River. Proceedings
of Symposium on Diffusion in Oceans and Fresh Waters.
Columbia University.
30. Yotsukura, N., H.B. Fisher, and W.W. Sayre. 1970. Measure-
ment of Mixing Characteristics of the Missouri River Between
Sioux City, Iowa, and Plattsmouth, Nebraska. Geological
Survey Water-Supply Paper 1899-G. U.S. Geological Sur-
vey, U.S. Department of the Interior, Reston, VA.
31. Wilson, J.F., E.D. Cobb, and N. Yotsukura. 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.
32. Hetling, L.J., and R.L O'Connell. 1966. A Study of Tidal
Dispersion in the Potomac River. Water Resources Research
2(4): 825.
33. Baily, T.E. 1966. Fluorescent-Tracer Studies of an Estuary. /.
of the Water Pollution Control Federation 38(12):1988.
34. Fischer, H.B. 1972. A Langrangian Method for Predicting
Pollutant Dispersion in Bolinas Lagoon, Mann County, Califor-
nia. Dispersion in Surface Water. U.S. Geological Survey
Professional Paper 582-B. U.S. Geological Survey, U.S.
Department of the Interior, Reston, VA.
35. Crocker, P.A., G. Morrison, S.R. Rives, F.S. Shipley, R. Palachek,
D. Neleigh, and S. Schimmel. 1989. Case Study to
Evaluate Effluent and Receiving Water Toxicity for an Estua-
rine Discharger: U.S. Naval Air Station, Corpus Christi, Texas.
Technical Section, Water Quality Management Branch,
U.S. EPA, Region 6.
36. Johnson, M.C. 1984. Fluorometric Techniques for Tracing
Reservoir Inflows. Instruction Report E-84-1. U.S. Army
Engineer Waterways Experiment Station, Vicksburg, MS.
37. Dettmann, E.H., J.F. Paul, J.S. Rosen, and C.J. Strobel. 1988.
Transport, Fate, and Toxic Effects of a Sewage treatment
Plant Effluent in a Rhode Island Estuary. U.S. EPA, Environ-
mental Research Laboratory, Narragansett, Rl. EPA/600/X-
87-366.
38. Albertson, M.L., Y.B. Dai, R.A. Jensen, and H. Rouse. 1950.
Diffusion of Submerged Jets. Trans. Am. Soc. Civ. Eng.
(15):639-64.
39. Wright, S.J. 1977. Mean Behavior of Buoyant Jets in a
Crossflow. y. of the Hydraulics Division. Proc. ASCE, vol-
ume 103.
40. Donekar, R.L., and G.H. Jirka. 1988. CORMIX1. An Expert
System for Mixing Zone Analysis of Toxic and Conventional
Single Port Aquatic Discharges. U.S. EPA, Environmental
Research Laboratory, Athens, GA. PB 88-220 504/AS.
41. Akar and Jirka. 1990. An Expert System for Mixing Zone
Analysis of Toxic and Conventional Single Port Aquatic Dis-
charges (CORMIX2). Report under peer review.
42. Yotsukura, N., and W.W. Sayre. 1976. Transverse Mixing in
Natural Channels. Water Resources Research 12(4): 695-
704.
43. Pailey, P.P., and W.W. Sayre. 1978. Model for Shore-
Attached Thermal Plumes in Rivers. ]. of the Hydraulics
Division. Proc. ASCE, volume 104.
44. Danigan, A., and W. Waggy. 1984. Simulation—A Tool for
Water Resource Management. Water Resources Bulletin
10(2).
45. Tiwari, J.L., and J.E. Hobble. 1976. Random Differential
Equations as Models of Ecosystems: Monte Carlo Simula-
tion Approach. Math Bioscience, volume 28.
46. Malone, R.F., D.S. Bowles, W.J. Grenney, and M.P. Windham.
1979. Stochastic Analysis of Water Quality. Utah State
University. UWRL/Q-79/01.
47. Hornberger, G.M., and R.C. Spear. 1980. Eutrophication in
Peel Inlet—I. The Problem—Defining Behavior and a
Mathmatical Model for the Phosphorous Scenario.' Water
Research 14:29-42.
48. Ford, D.E., K.W. Thornton, A.S. Lessem, and J.L. Norton.
1981. A Water Quality Management Model for Reservoirs.
ASCE Proc. of Symposium on Surface Water Impound-
ments, June 2-5,1980, at Minneapolis, MN.
49. Scavia, D., W.F. Powers, R.P. Canale, and J.L. Moody 1981.
Comparison of First-order Error Analysis and Monte Carlo
Simulations in Time-dependent Lake Eutrophication Mod-
els. Water Resources Research 17(4):1051 -59.
50. Rose, K.A., J.K. Summers, and R.I. McLean. 1989. A Monte
Carlo Bioaccumulation Model of Silver Uptake by the Oys-
ter, Crassostrea Virginia. Ecol. Mod. 42:10-21.
51. Versar. 1989. Prediction of Uncertainties Associated with
Water Quality Models. U.S. EPA, Office of Research and
Development and Office of Toxic Substances. Washing-
ton, DC.
52. Thornton, K.W., D.E. Ford, and J.L. Norton. 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 Ecologi-
cal Modeling, Colorado State University.
53. DiToro, D. 1984. Probability Model of Stream Quality Due
to Runoff. ]. of Environ. Engineering. Proc. ASCE, vol. 110,
no. 3.
54. Limno Tech, Inc. 1985. Dynamic Toxics Wasteload Allocation
Model (DYNTOX): User's Manual. U.S. EPA. Off ice of Water
Regulations and Standards, Washington, DC.
91
-------
55. McKeon, T.J., and J.J. Segna. 1987. Selection Criteria for
Mathematical Models Used in Exposure Assessments Surface
Water Models. U.S. EPA, Office of Health and Environmen-
tal Assessment, Washington, DC. EPA/600/8-87/042.
56. Ambrose, R.B., and T.O. Barnwell. 1989. Environmental
Software at the U.S. EPA Center for Exposure Assessment
Modeling. U.S. EPA, Environmental Research Laboratory,
Athens, GA.
57. Bums, LA, and D.M. Cline. 1985. Exposure Analysis Model-
ing System Reference Manual for EXAMS II. U.S. EPA, Envi-
ronmental Research Laboratory, Athens, GA. EPA 600/3-
85-038.
58. Ambrose, R.B., T.A. Wool, J.P. Connolly, and R.W. Shantz.
1988. WASP4, A Hydrodynamic and Water Quality Model-
Model Theory, User's Manual, and Programmer's Cuide. U.S.
EPA, Center for Exposure Assessment Modeling, Athens,
GA. EPA/600/3-87/039.
59. Johansen, R.C., J.C. Imhoff, J.L. Kittle, and A.S. Donigian.
1984. Hydrologic Simulation Program—FORTRAN (HSPF):
Users Manual for Release 8.0. U.S. EPA, Environmental
Research Laboratory, Athens, GA. EPA/600/3-84-066.
60. Vandergift, S.B., and R.B. Ambrose. 1988. SARAH2 A Near
Field Exposure Assessment Model for Surface Waters. U.S.
EPA, Environmental Research Laboratory, Athens, GA. EPA/
600/3-88/020.
61, Brown, D.S., and J.D. Allison. 1987. MINTEQA1 Equilibrium
Metal Spec/at/on Model: A Users Manual. U.S. EPA, Environ-
mental Research Laboratory, Athens, GA. EPA/600/
3-87/012. ,
62. Barber, M.C., L.A. Suarez, and R.R. Lassiter. 1988. FCETS
(Food and CHI Exchange for Toxic Substances): A Simulation
Model for Predicting Bioaccumulation of Nonpolar Organic
, Pollutants by Fish. U.S. EPA, Center for Exposure Assess-
'. ment Modeling, Athens, GA. EPA/600/3-87/038.
63. Connolly, J.P., and R.V. Thomann. 1985. WASTOX, A Frame-
• • work for Modeling the Fate of Toxic Chemicals in Aquatic
Environments. Part 2: Food Chain. U.S. EPA, Environmental
I Research Laboratory, Gulf Breeze, FL. EPA/600/4-85/040.
64. Mills, W.B., j.D. Dean, D.B. Porcella, S.A. Gherini, R.j.M.
! Hudson, W.E. Frick, G.L. Rupp, and G.L. Bowie. 1982.
\ Water Quality Assessment: A Screening Procedure for Toxic
I, and Conventional Pollutants. U.S. EPA, Environmental Re-
! search Laboratpry. Athens, GA. EPA 600/6-82-004
, a and b.
65. U.S. EPA. 1987. Permit Writers Cuide to Water Quality-based
Permitting for Toxic Pollutants. U.S. EPA, Office of Water
Enforcment and Permits, Washington, DC. EPA/440/4-87/
I 005.
66\ Rossman, L.A. Design Stream Flows Based on Harmonic
; Means. /. of Hydraulic Engineering 116(7):946-5Q.
92 ;
-------
5. PERMIT REQUIREMENTS
5.1 INTRODUCTION
As the final step in the "standards-to-permits" process, develop-
ment 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 such that
water quality standards are protected. It also addresses important
considerations in the expression of limits and other types of
permit requirements, including toxicity reduction evaluations.
The first portion of the chapter deals principally with aquatic life
protection. Permitting for protection of human health is found in
Section 5.4.4.
5.1.1 Regulatory Requirements
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 the revisions to the National Pollutant Discharge
Elimination System (NPDES) Surface Water Toxics Control Pro-
gram regulations (54 FR 23868, June 2,1989). The regulations at
40 CFR122.44(d)(1) require that regulatory authorities first deter-
mine whether a discharge causes, has the reasonable potential to
cause, or contributes to an excursion above water quality stan-
dards (narrative or numeric). In making these determinations,
regulatory authorities must use a procedure that accounts for
effluent variability, existing controls on point and nonpoint sources
of pollution, available dilution, and (when using toxicity testing)
species sensitivity. Each of these regulations were previously
discussed in Chapter 3.
There is a degree of flexibility in the specific procedures a regula-
tory authority uses in determining whether an excursion occurs or
is reasonably expected to occur and in the weight given to the
various factors in conducting the evaluation of a specific dis-
charger. The Environmental Protection Agency's (EPA) guidance
for making these determinations is contained in the recommen-
dations in Chapter 3.
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-
2) and EPA's "Whole Effluent Toxicity Permitting Principles and
Enforcement Strategy," (Appendix B-4). This strategy states 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 strategy states that "[f]inal whole effluent toxicity
limits must be included in permits where necessary to ensure that
State Water Quality Standards are met. These limits must prop-
erly 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 limit.
Case-specific considerations will usually dictate the most appro-
priate approach to be taken in individual situations. Nevertheless,
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. The permit limit
derivation procedure used by the permitting authority should
be fully enforceable and should adequately account for efflu-
ent variability, consider available receiving water dilution,
protect against acute and chronic impacts, account for com-
pliance monitoring sampling frequency, and protect the
wasteload allocation (WLA) and ultimately water quality stan-
dards. To accomplish these objectives, EPA recommends that
permitting authorities use the statistical permit limit deriva-
tion procedure discussed in Section 5.4 with the outputs from
either steady state or the dynamic wasteload allocation mod-
eling.
5.2 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 technol-
ogy-based limits. However, the process for applying the prin-
ciples is substantially different, as explained below.
5.2.1 Variations in Effluent Quality
Effluent quality and quantity vary 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 for publicly owned treatment works (POTWs),
changes in production cycles for industries, variation in responses
of wastewater treatment systems to influent changes, variation in
treatment system performance, and changes in climate. Very few
effluents remain constant over long periods of time. Even in
industries that operate continuous processes, variations in the
quality of raw materials and activities, such as 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 POTW 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-concentra-
tion plots of the same data (see Figure 5-1, right-hand graphs).
93
-------
I
! 10-
O
Ceriodaphnia sp.
CVs1.06
i
10
20
30
40
50 60
Days
Ceriodaphnia sp.
CV = 1.06
8-
CD
IT
2
0)
CC
5.0 7.5 10.0
Chronic Toxic Units
12.5
15.00
12-1
10-
8-
§
CT
I
Daphnia sp.
CV = 0.70
Long-Term Average
Days
Acute Toxic Units
10 -i
8-
6-
Ztnc
CV = 0.59
Zinc
CV = 0.59
Long-Term Average
100
Figure 5-1. Data Relative Frequency Distributions for Ceriodaphnia Toxicity, Daphnia Toxicity,
and Zinc Concentrations for Three Different Effluents
94
-------
5.2.2 Statistical Parameters ami 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, it is reasonable to assume
(unless specific data show otherwise) that treated effluent data
follow a lognormal 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 ap-
pearance of the lognormal curve when data are plotted in a
frequency histogram. Appendix E discusses the basis for conclud-
ing 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 posi-
tions of the concentrations corresponding to the mean for the
data.
All permit limits, whether technology-based or water quality-
based, are set at the upper bounds of acceptable performance.
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 systerr
For water quality-based requirements, the limits are based on
maintaining the effluent quality at a level that will comply with
water quality standards, even during critical conditions in the
receiving water. These requirements are determined by the WLA.
The WLA dictates the required effluent quality which defines the
desired level of treatment plant performance or target LTA.
In the development of technology-based effluent limits guide-
lines, the operating records of various wastewater treatment facili-
ties for a particular category of discharger are examined. Based
on the effluent data for the treatment facilities, a composite mean
or LTA value for the parameter is determined. This LTA value,
with relevant estimates of variability, is then used to derive efflu-
ent limit guidelines, which lead directly to permit limits.
In contrast, the process operates in reverse for water quality-based
permit limits. The WLA, determined from water quality stan-
dards, defines the appropriate discharge level, which in turn
determines the requisite target LTA for the treatment facility in
order to meet that WLA. Permit limits may then be derived from
this targeted LTA and CV. Figure 5-2 illustrates the relationship
among the various statistical parameters. As these figures show,
highly variable effluents require a much lower targeted LTA to
meet the WLA and account for the variability that occurs in
effluent concentration above the LTA.
It is extremely important to recognize that the various statistical
principles and relationships discussed above operate in any dis-
1.0-
16-r
14-
= 12-
•| 10-
1 8-
I 6-
8 4-
2-
0
Days
Wasleload Allocation
Long-Term Average
Days
O u
S1 —
5 <
ll
i s
1 13
-------
charge situation—whether or not they are specifically recognized
or accounted for. Where a permit limit derivation procedure does
not address these principles specifically, the permit writer will be
implicitly assuming that there are enough conservative assump-
tions built into other steps in the process (e.g., water quality
models, "buffer" between permit limits and actual operating
conditions) to ensure that there will be no reasonable potential for
excursions above water quality standards.
5.2.3 Expression of Permit Limits
The NPDES regulations at 40 CFR122.45(d) require that all permit
limits be expressed, unless impracticable, as both average monthly
and maximum daily values for all discharges other than POTWs
and as average weekly and average monthly limits for POTWs.
The maximum daily permit limit (MDL) is the highest allowable
discharge measured during a calendar day or 24-hour period
representing a calendar day. The average monthly permit limit
(AML) is the highest allowable value for the average of daily
discharges obtained over a calendar month. The average weekly
permit limit (AWL) is the highest allowable value for the average
of daily discharges obtained over a calendar week.
EPA believes that a maximum daily permit limit can be directly
used to express an effluent limit for all toxic pollutants or pollutant
parameters except chronic whole effluent toxicity. The typical
toxicity test used to measure chronic toxicity consists of samples
collected from at least 3 different days over a 7-day period.
Therefore, the test does not measure toxicity in any given 24-hour
period or calendar day, but rather measures toxicity over a 7-day
period. The toxicity could be caused by any one sample or a
combination of samples. To address this situation, EPA recom-
mends that the permit contain a notation indicating that
when chronic toxicity tests are required in a permit, the MDL
should be interpreted as signifying the maximum test result
for the month.
Additionally, in lieu of an AWL for POTWs, EPA recommends
establishing an MDL (or a maximum test result for chronic toxic-
ity) for toxic pollutants and pollutant parameters in water quality
permitting. This is appropriate for at least two reasons. First, the
basts for the 7-day average for POTWs derives from the secondary
treatment requirements. This basis is not related to the need for
assuring achievement of water quality standards. Second, a 7-day
average, which could comprise up to seven or more daily samples,
could average out peak toxic concentrations and therefore the
discharge's potential for causing acute toxic effects would be
missed. A MDL, which is measured by a grab sample, would be
toxicologically protective of potential acute toxicity impacts.
5.3 ENSURING CONSISTENCY WITH THE
WASTELOAD ALLOCATION
The WLA provides a definition of effluent quality that is necessary
to meet the water quality standards of the receiving water. The
WLA is based on ambient criteria and the exposure of the resident
aquatic community or humans to toxic conditions. Once a WLA
has been developed, accounting for all appropriate consider-
ations, a water quality-based permit limit may be derived to
enforce the WLA. The method used to derive the permit limits
must be consistent with the nature of the WLA.
The WLA addresses variability in effluent quality. For example, a
WLA for human health pollutants is typically expressed as a single
level of receiving water 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 nec-
essary 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 enforced incorrectly, since effluent variability and
the probability basis for the limit are not considered specifically.
For example, the use of a steady state WLA typically establishes a
level of effluent quality 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 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 maximum daily and average
monthly permit limits. The permit writer must ensure that permit
limits are derived to implement a WLA requirement correctly.
Potential problem areas are as follows:
I
[• The WLA must be enforced in a regulatory context by
' translating it into MDLs and AMLs; then and only then, will
', compliance monitoring associated with permit limits allow
' the regulatory authority to determine whether or not such
: permit limits are violated.
• The WLA that assumes that the discharge is steady state
i (i.e., not changing over time) requires a limit derivation
• ' assumption regarding how the effluent may vary.
'• MDLs and AMLs average monthly limits must be developed
so that they are consistent with each other and mandate
I the required level of wastewater treatment facility perfor-
' mance.
!• If the acute WLA is used alone directly as the MDL, the limit
( will not necessarily be protective against chronic effects. If
the acute WLA is used alone directly as the AML, the limit
can allow excursions above the WLA within each month.
[• If the chronic WLA is used alone as an MDL, the limit will be
protective against acute and chronic effects but at the
| expense of being overly stringent. If the chronic WLA is
i used alone as the AML, the limit may be protective against
| acute and chronic effects depending upon effluent variabil-
ity.
The objective is to establish permit limits that result in the effluent.
meeting the WLA under normal operating conditions virtually all
the time. It is not possible to guarantee, through permit limits,
that a WLA will never be exceeded. It is possible, however, using
the recommended permit limit derivation procedures, to account
for extreme values and to establish low probabilities of exceedence
96
-------
of the WLA in conformance with the duration and frequency
requirements of the water quality standards. This is not to sug-
gest that permit writers should assume a probability of exceedence
of the WLA, but rather, that they should develop limits that will
make a. i exceedance a very small likelihood.
Since effluents are variable and permit limits are developed based
on a low probability of exceedence, the permit limits should
consider effluent variability and ensure that the requisite loading
from the WLA is not exceeded under normal conditions. In effect
then, the limits must "force" treatment plant performance, which,
after considering 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 relation-
ships 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 for the chronic
limiting condition. 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.
The duration and frequency requirements of the acute and chronic
criteria for the pollutant or pollutant parameter will not be ex-
ceeded. This figure also illustrates permit limits derived from the
more limiting LTA. (Note that for the scenario depicted in Figure
5-3, the MDL is lower than the acute WLA and the average
monthly limit is lower than the chronic WLA. This scenario will
occur when a 99-percent probability basis is used to calculate the
LTA and a 95-percent probability basis is used to calculate the
permit limits from the lower of the acute and chronic LTA. For
other probability assumptions, these relationships will differ.)
5.3.2 Types of Water Quality Models and Model Outputs
Each of the two major types of water quality models, steady-state
and dynamic, and their WLA outputs have specific implications
20-1
Acute Wasteload Allocation
Maximum Daily Permit Limit
Chrome Wasteload Allocation
Long-Term Average
Days
Figure 5-3. Relationship Between Daily Concnetrations,
Long-Term Average, Wasteload Allocations,
and Permit Limits
for the subsequent permit limit development process. These
implications are discussed in detail below. EPA recommends
that steady-state WLA analyses generally be used by permit-
ting authorities in most cases and especially where few or no
whole effluent toxicity or specific chemical measurements are
available, or where daily receiving water flow records are not
available. Two-value, steady-state models, although potentially
more protective than necessary, can provide toxicologically pro-
tective results and are relatively simple to use. If adequate
receiving water flow and effluent concentration data are avail-
able to estimate frequency distributions, EPA recommends
that one of the dynamic WLA modeling techniques be used to
derive WLAs that will more exactly maintain water qualify
standards.
Steady-State Modeling
Traditional single-value or two-value steady-state WLA models
calculate WLAs at critical conditions, which are usually combina-
tions of worst-case assumptions of flow, effluent, and environ-
mental effects. For example, a steady-state model for ammonia
considers the maximum effluent discharge to occur on the day of
lowest river flow, highest upstream concentration, highest pH,
and highest temperature. Each condition by itself has a low
probability of occurrence; the combination of conditions may
rarely or never occur. Permit limits derived from a steady-state
WLA model will be protective of water quality standards at the
critical conditions and for all environmental conditions less than
critical. However, such permit limits may be more stringent than
necessary to meet the return frequency requirements of the water
quality criterion for the pollutant of concern.
On the other hand, a steady-state model approach may involve
simplifying assumptions for other factors, such as ambient back-
ground concentrations of a toxicant, multiple source discharges
of a toxicant, number of pollutants causing toxicity, incorrect
effluent variability assumptions, and infrequent compliance moni-
toring. The effect of these types of factors, especially if unaccounted
for in the WLA determination, can reduce the level of protective-
ness provided by the critical condition assumptions of the steady-
state model approach. Therefore, when using a steady-state WLA
model, the permitting authority should be aware of the different
assumptions and factors involved and should consider these as-
sumptions and factors adequately consideration when develop-
ing permit limits. : • •
In general, steady-state analyses tend to be more conservative
than dynamic models because they rely on worst case assump-
tions. Thus, permit limits derived from these outputs will gener-
ally be lower than limits derived from dynamic models.
a) Single Value From a Steady-State Analysis
Some single-value, steady-state modeling has been used to calcu-
late only chronic WLAs. These models produce a single effluent
loading value and no information about effluent variability. Single
value WLAs are typically based upon older State water quality
standards that do not specify levels for both acute and chronic
protection but only include one level of protection. Such outputs
also would be found where a model is based upon protection of
human health, since only a single long-term ambient value is of
concern.
97
-------
b) Two Values from Steady-State Analysis
Steady-state modeling for protection of aquatic life can specify
two sets of calculations—one for protection against acute effects
and one for protection against chronic effects. These models
must use water quality criteria specifying two levels of protection.
In addition, these models 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 pro-
duce any information about acceptable effluent variability and
may require additional calculations to be translated into permit
limits.
For complex discharge situations (i.e., multiple dischargers or
complex environmental factors needing consideration), water qual-
ity models and associated WLAs are typically developed by spe-
cialized water quality analysts in the regulatory authority. How-
ever, the permit writer is often required to develop a water quality
model and WLA prior to permit limit derivation. In the latter
situation, water quality modeling usually consists of simple steady-
state dilution models using worst-case assumptions.
Dynamic Modeling
Dynamic models use estimates of effluent variability and the
variability of receiving water assimilation factors to develop efflu-
ent 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.
In general, dynamic models account for the daily variations of and
relationships between flow, effluent, and environmental condi-
tions and therefore directly determine the actual probability that a
water quality standards exceedence will occur. Because of this,
dynamic models can be used to develop WLAs that maintain the
water quality standards exactly at the return frequency require-
ments of the standards. Since this return frequency is usually one
event in 3 years, WLAs developed by dynamic models are typically
higher than those developed by steady-state models.
A targeted long-term average performance level and coefficient of
variation can be derived from each type of dynamic model out-
put, but some of the outputs require some additional manipula-
tion 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 Simula-
tion models offer an array of effluent data that 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 limits. Chapter 4
details the different dynamic models. Specific instructions for the
use of dynamic models are available in the references listed at the
end of Chapter 4.
5.4 PERMIT UMIT DERIVATION
There are a number of different approaches currently being used
by permitting authorities to develop water quality-based limits 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 varir
typjas of water quality models and WLA outputs. In addition,
permitting authorities also are constrained by legal requirements
and policy decisions that may apply to a given permitting situa-
tion. In some instances, however, permitting procedures have
been adopted without careful consideration of the toxicological
principles involved or the advantages and disadvantages of the
procedure.
To avoid this problem, EPA recommends that the statistical
permit limit derivation procedure described in this chapter be
used for the derivation of both chemical-specific and whole
effluent toxicity limits for NPDES permits. The type of WLA
chosen from which to derive the limits is a matter of case-by-case
application, as determined by the permitting authority. Although
there are advantages and disadvantages associated with each of
the; procedures, EPA believes that the statistical derivation proce-
durjes will result in the most defensible and protective water
quality-based permit limits for both specific chemicals and whole
effluent toxicity.
I
The following section explains EPA's recommended permitting
procedures and highlights advantages and disadvantages of vari-
ous other approaches. With this information, permitting authori-
ties will be better informed when deciding on the most appropri-
ate permit limit derivation approach. For example, permitting
authorities may decide to derive water quality-based permit limits
for all dischargers using a steady-state WLA model as a baseline
limit determination. If time and resources are available or if the
discharger itself takes the initiative (after approval by the regula-
tory authority), dynamic modeling could be conducted to further
refine the WLA from which final permit limits would be derived.
Box 5-1 presents example permit limit calculations for each of the
principal types of WLA outputs discussed in Section 5.4.1. Permit
limits derived from dynamic modeling are usually higher than
those based upon steady-state modeling. The difference is re-
flected in Box 5-1 and has been observed in actual applications [1,
2, 3]. In addition, the case studies in Chapter 7 illustrate how
water quality-based permit limits are derived and compare the
results of limits derived from steady state and dynamic wasteload
allocations*.
5.4.1 EPA Recommendations for Permitting for Aquatic
Life Protection
Perjnit Limit Derivation from Two-Value, Steady-State Out-
puts for Acute and Chronic Protection
A number of WLAs have two results: acute and chronic require-
ments. These types of allocations will be developed more often as
States begin to adopt water quality standards that provide both
acuj:e and chronic protection for aquatic life. These WLA outputs
need to be translated into MDLs and AMLs. The following
methodology is designed to derive permit limits for specific chemi-
cals as well as whole effluent toxicity to achieve these WLAs.
« A treatment performance level (LTA and CV) that will allow
the effluent to meet the WLA requirement is calculated.
98
-------
Box 5-1. Sample Calculations of Permit Limits for Whole Effluent Toxictty
from Different Wasteload Allocation Data
Available Data
Two Value wasteload
allocation
Wasteload Allocation (WLA)
Acute Wasteload Allocation (WLAa) 2.60
Chronic Wasteload Allocation (WLAc) 14.3
Acute-Chronic Ratio 4.62
Coefficient of Variation (CV) 0.8
Number of Samples per Month (n) 4
Long Term Average (LTA)
Dynamic model
output
0.8
4
9.44
Single wasteload
allocation
14.3
0.8
4
From two-value steady state wasteload allocation
WLAa(C= WLAa*ACR =2.60*4.62 =12.0
LTAc'=WLAc«e[0.5a42-2.326o4] = 14.3-0.440 (from Table 5-1) =6.29
LTAajC = WLAa/c*e[0.5o2-2.326c]= 12.0*0.249
(from Table 5-1) =2.99'
MDL = LTAa/c*e [2.326a-0.5c2] = 2.99*4.01 (from Table 5-2) =12.0
AML = LTAa'c*e [2.326an-0.5an2]= 2.99*2.27 (from Table 5-2) =6.79
From dynamic model output
MDL = LTAc*e [2.326a-0.5o2]= 9.44*4.01 (from Table 5-2)= 37.9
AML = LTAc-e [2.326an-0.5an2]= 9.44*2.27 (from Table 5-2)= 21 .4
From single wasteload allocation
Option 1
LTA = WLA*e [0.5c2-2.326a] = 14.3*0.440 (from Table 5-1) =6.29
MDL = LTA*e [2.326a-0.5o2] = 6.29*4.01 (from Table 5-2) = 25.2
AML = LTA«e [2.326an-0.5an2] = 6.29*2.27 (from Table 5-2) =14.3
Option 2
MDL =WLA =14.3
AML =MDL/2 =7.15
Note: All calculations use the 99th
percentile z statistic for calculation
of long-term averages and permit
limits.
Where two requirements are specified based on different
duration periods, two performance levels are calculated
(Box 5-2, Step 2).
• For whole effluent toxicity only, the acute WLA is converted
into an equivalent chronic WLA by multiplying the acute
WLA by an acute-to-chronic ratio (ACR). This ratio should
optimally be based on effluent data, but also can be esti-
mated as 10, based on the information presented in Chap-
ter 1 and Appendix A.
• Permit limits are then derived directly from whichever per-
formance level is more protective (Box 5-2, Steps 3 and 4).
Figure 5-4 presents a flow chart summarizing the various steps in
this procedure. In addition, the equations used in Box 5-2 are
based on the lognormal distribution, which is explained in more
detail in Appendix E. The principal advantages of this procedure
are described below.
• This procedure provides a mechanism for setting permit
limits that will be lexicologically 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
exceedance of the WLA due to effluent variability. Clearly,
however, effluents are variable. Therefore, permit limits are
established using a value corresponding to a percentile of
the selected probability distribution of the effluent (e.g.,
95th or 99th percentile).
It allows comparison of two independent WLAs (acute and
chronic) to determine which is more limiting for a dis-
charge. The WLA output provides two numbers for protec-
tion against two types of toxic effects, each based upon
different mixing conditions for different durations. Acute
effects are limited based upon 1 -hour exposures at critical
conditions, close to the point of discharge, or where neces-
sary, at the end of the pipe. Chronic effects are limited
based on 4-day exposures after mixing at critical condi-
tions. These requirements yield different effluent treatment
requirements that cannot be compared to each other with-
out calculating the LTA performance level the plant would
need to maintain in order to meet each requirement. With-
out this comparison (or in the absence of procedures that
address this comparison), the WLA representing the more
critical condition cannot be determined. A treatment sys-
tem will only need to be designed to meet one level of
99
-------
Box 5-2. Calculating Permit Limits Based on Two-Value Wasteload Allocation ,
4
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. Skip to Step 2 for
chemical-specific limits.
Calculate the long-term
n average wasteload that will
^ satisfy the acute and chronic
wasteload allocations.
Determine the lower (more
3 limiting) of the two long-term
averages.
Calculate the maximum daily
M and average monthly permit
limits using the lower (more
limiting) long-term average.
Term Meaning
CV Coefficient of variation
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
LTAac Acute long-term average
wasteload in chronic units
LTA0 Chronic long-term average
wasteload
TUa Acute toxic units
TUC Chronic toxic units
ACR Acute-to-chronic ratio
MDL Maximum daily limit
AML Average monthly limit
z z statistic
Step 1 (for whole effluent toxicity only)
WLAac (in TUJ = WLAa (in TUa) • ACR
I . .
Step 2 (start here for chemical specific limits)
where b2 = ln(CV2+1) .
z = 1 .645 for 95th percentile probability basis, and
z = 2.326 for 99th percentile probability basis
I TA - Wl A • P t°-5CT4 - Z04l
L_ 1 ^^— ~~ • • "**»« ^ ' '
where o42 = ln(CV2/4 +1)
z = 1 .645 for 95th percentile probability basis, and
z = 2.326 for 99th percentile probability basis
Step3
LTA = mm (LTAC, LTAa c)
, , :'. '•;. .1 '•••-. '.. . . - ; •; ...
Step 4
where p^lnfCVVl)
z = 1 .645 for 95th percentile probability basis, and
z = 2.326 for 99th percentile probability basis
where ;rnz = ln(CV2/n +1)
z = 1 .645 for 95th percentile probability basis, and
z = 2.326 for 99th percentile probability basis
• ' ., • • • ••
Full details of this procedure are found in Appendix E. '
100
-------
Figure 5-4. Flowchart for Calculating Permit Limits From
Two-Value, Steady-State Wasteload Allocation
for Aquatic Life Protection
treatment for effluent, toxicity—treatment needed to control
the most limiting toxic effect.
• The actual number of samples can be factored into permit limit
derivation procedures.. The procedure provides the means to
accurately determine the.AML based on the number of obser-
vations that will be taken.
The principal disadvantages of this approach are:
• Some permit writers have indicated that additional math-
ematical calculations associated with these procedures increase
the burden for the permit writer and add what is perceived to
be an unnecessary step.
• The use of a steady-state WLA may result in permit limits that
are more conservative due to the assumption of critical condi-
tions. However, these limits are still protective of water quality
criteria. The level of conservatism may be necessary in those
instances where limited data prevent a more precise evaluation
of a WLA.
This procedure provides a toxicologically sound approach. To
help the permit writer, EPA has developed tables (see Tables 5-1
and 5-2) to be used to quickly determine the necessary values. In
addition, some permit authorities have developed their own com-
puter programs to readily compute the necessary information
from the appropriate inputs.
Permit Limit Derivation From Dynamic Model Outputs
The least ambiguous and most exact way that a WLA for specific
chemicals or for whole effluent toxicity can be specified by using
dynamic modeling from which the WLA is expressed as a 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 limits to achieve the WLA
objectives. The types of dynamic exposure analyses that yield a
WLA in terms of required performance are the continuous simula-
tion, Monte Carlo, and lognormal probabilities analyses. Chapter
4 provides a general discussion of these models. Guidance manu-
als for developing WLAs are listed in the references at the end of
Chapter 4. Once the WLA is determined, the permit limit deriva-
tion procedure which can be used for both whole effluent toxicity
and specific chemicals, is as follows:
• The WLA is first developed by iteratively running the dy-
namic model with successively lower LTAs until the model
shows compliance with the water quality standards.
• The effluent LTA and CV must then be calculated from the
model effluent inputs used to show compliance with the
water quality standards. This step is only necessary for the
Monte Carlo and continuous simulation methods.
• The permit limit derivation procedures described in Box 5-
2, Step 4 are used to derive MDLs and AMLs from the
required effluent LTA and CV. Unlike these procedures for
steady-state WLAs, there is only a single LTA that provides
both acute and chronic protection, and, therefore, the
comparison step indicated in Figure 5-4 and Box 5-2 is
unnecessary.
The principal advantages of this procedure are:
• It provides a mechanism for computing permit limits that
are toxicologically protective. As with the procedure sum-
marized below for two-value, steady-state WLA outputs,
the permit limit derivation procedures used with this type
of output consider effluent variability and derive permit
limits from a single limiting LTA and CV.
• Actual number of samples is factored into permit limit
derivation procedures. This procedure has the same ele-
ments as discussed for the statistical procedures in Option 2
below.
• Dynamic modeling determines an LTA that will be ad-
equately protective of the WLA, which relies on actual flow
data thereby reducing the need to rely on worst case critical
flow condition assumptions.
101
-------
Table 5-1. Back Calculations of Long-Term Average
cv
0.1
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2.0
WLA Multipliers
[0.5 02-2C]
e
95th
Percentile
0.853
0.736
0.644
0.571
0.514
0.468
0.432
0.403
0.379
0.360
0.344
0.330
0.319
0.310
0.302
0.296
0.290
0.285
0.281
0.277
99th
Percentile
0.797
0.643
0.527
0.440
0.373
0.321
0.281
0.249
0.224
0204
0.187
0.174
0.162
0.153
0.144
0.137
0.131
0.126
0.121
0.117
LTAa,c=Vl
Acute
[0.5 02-ZO]
LAa,c*e
where a2i/n[CV<: +1],
z = 1 .645 for 95th percentile occurrence probability, and
z = 2.326 for 99th percentile occurrence probability
I
i
.
Chronic
(4-day average)
;
1 TA \A/I A '°'5 °4 " Z °4 '
L 1 AC — vvLAc • 8 ,
where cr42=/n[CV2/ 4 + 1],
z a 1 .645 for 95th percentile occurrence probability, and
z = 2.326 for 99th percentile occurrence probability !
t
1 ' • '
[
•
f .
•
cv
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
'1.9
2.0
WLA Multipliers
e[0.5a42-Za4,
95th
Percentile
0.922
0.853
0.791
0.736
0.687
0.644
0.606
0.571
0.541
0.514
0.490
0.468
0.449
0.432
0.417
0.403
0.390
0.379
0.369
0.360
99th
Percentile
0.891
0.797
0.715
0.643
0.581
0.527
0.481
0.440
0.404
0.373
0.345
0.321
0.300
0.281
0.264
0.249
0.236
0.224
0.214
0.204
The principal disadvantages of this procedure are:
• Necessary data for effluent variability and receiving water
flows may be unavailable, which prevents the use of this
approach.
• The amount of staff resources needed to explain how the
limits were developed and to conduct the WLA also is a
concern. The permit documentation (i.e., fact sheet) will
need to clearly explain the basis for .the LTA and CV and this
can be resource intensive.
Permit Limit Derivation From Single, Steady-State Model
Output
Some State water quality criteria and the corresponding WLAs are
reported as a single value from which to define an acceptable
level of effluent quality. For example, "copper concentration
must not exceed 0.75 milligrams per liter (mg/l) instream." 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 need to consider
effluent variability.
102
-------
Table 5-2. Calculation of Permit Limits
CV
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2.0
LTA multipliers
[ZCT-0.502]
6
95th
Percentile
1.17
1.36
1.55
1.75
1.95
2.13
2.31
2.48
2.64
2.78
2.91
3.03
3.13
3.23
3.31
3.38
3.45
3.51
3.56
3.60
, 99th
Percentile
1.25
1.55
1.90
2.27
2.68
3.11
3.56
4.01
4.46
4.90
5.34
5.76
6.17
6.56
6.93
7.29
7.63
7.95
8.26
8.55
Maximum Daily Limit
, fzc-OSo2!
MDL = LTA • e
, where a2 = In [ CV2 + 1 ],
. z = 1 .645 for 95th percentile occurrence probability, and
z = 2.326 for 99th percentile occurrence probability
Average Monthly Limit
21
AML=,LTA»e °n" ' °n
where an2 = In [ CV2 / n + 1 ],
z = 1 .645 for 95th percentile,
z = 2.326 for 99th percentile, and
n = number of samples/month
CV
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
1.3
1.4
1.5
1,6
1.7
1.8
1.9
2.0
LTA Multipliers
. e[zon-0.5on2]
95th
Percentile
n=1 n=2 n=4 n=10 n=30
1.17 1.12 1.08 1.06 1.03
1.36 1.25 1.17 1.12 1.06
1.55 1.38 1.26 1.18 1.09
1.75 1.52 1.36 1.25 1.12
1.95 1.66 1.45 1.31 1.16
2.13 1.80 1.55 1.38 1.19
2.31 1.94 1.65 1.45 1.22
2.48 2.07- 1.75 1.52 1.26
2.64 2.20 1,85 1.59 1.29'
2.78 2.33 1.95 1.66 1.33
2.91 2.45 2.04 1.73 1.36
3.03 2.56 2.13 1.80 1.39
3.13 2.67 2.23 1.87 1.43
3.23 2.77 2.31 1.94 1.47
3.31 2.86 2.40 2.00 1.50
3.38 2.95 2.48 2.07 1.54
3.45 3.03 2.56 2.14 1.57
, 3.51 3.10 2.64 2.20 1.61
3.56 3.17 2.71 2.27 1.64
3.60 3.23 2.78 2.33 1.68
99th
Percentile
n=1: n=2 n=4 n=10 n=30
1.25 1.18 1.12 1.08 1.04
1.55 1.37 1.25 1.16 1.09
1.90 1.59 1.40. 1.24 1.13
2.27 1.83 1.55 1.33 1.18
2.68 2.09 1.72 1.42 1.23
3.11 2.37 1.90 1.52 1.28
3.56 2.66 2.08 1.62 1.33
4.01 2.96 2.27 1.73 1.39
4.46 3.28 2.48 1.84 1.44
4.90 3.59 2.68 1.96 1.50
5.34 3.91 2.90 2.07 1.56
5.76 4.23 3.11 2.19 1.62
6.17 4.55 3.34 2.32 1.68
6.56 4.86 3.56 2.45 1.74
6.93 5.17 3.78 2.58 1.80
7.29 5.47 4.01 2.71 1.87
7.63 5.77 4.23 2.84 1.93
7.95 6.06 4.46 2.98 2.00
8.26 6.34 4.68 3.12 2.07
8.55 6.61 4.90 3.26 2.14
The proper enforcement of this type of WLA depends on the
parameter limited. For nutrients and biochemical oxygen de-
mand (BOD), the WLA value generally has been used as the
average daily permit limit. However, the impact associated with
toxic pollutants is more time dependent, as reflected in the 4-day
average duration for the criteria continuous concentration (CCC)
(see Chapter 2). Where there is only one water quality criterion
and therefore only one WLA, permit limits can be developed
using the following procedure:,,. , . . ... .. -. , v . :
• Consider the single WLA to be the chronic WLA and derive
an chronic LTA for this WLA using the procedures in Box 5-
2 (Step 2, Part 2).
• Derive MDLs and AMLs using the procedures in Box 5-2
(Step 4).
The principal advantages and disadvantages of this procedure are
similar to those for the two-value permit limit derivation method
discussed previously except that it does not examine two WLAs.
5.4.2 Other Approaches to Permitting for Aquatic Life
Other approaches for translating WLA outputs into permit limits
have been used by some permitting authorities. These methods
may combine elements of the statistical procedures discussed
earlier with specific' technical and policy requirements of the
permitting authority to derive limits that may be protective of
water quality and consistent with the requirements of the WLA.
Such approaches may use simplified statistical procedures.
103
-------
For example, some permitting authorities assume a value for the
CV and an acute to chronic ratio above which the chronic WLA
will always be more limiting. Where such simplifying assumptions
are used, the need to compare LTAs derived from acute and
chronic steady-state models is unnecessary. Similarly, for as-
sumed values for n, CV, and exceedence probability, the various
equations shown in Box 5-2 can be simplified further, such that
the AMI will always be a constant fraction of the MDL.
These approaches allow the permit writer to rapidly and easily
translate the results of WLAs into permit limits. However, the
permit writer clearly should understand the underlying proce-
dures and carefully explain the basis for the chosen assumption.
Appropriate State or regional guidance documents also should be
referenced.
Another approach used by some permit authorities involves the
direct use of the WLA as a permit limit. This approach sometimes
Involves the following steps:
• The WLA value for toxic pollutants is used as the MDL
* In the absence of other information, permit writers typically
divide the MDL by 1.5 or 2.0 to derive an AML (depending
on the expected range of variability).
The principal advantage of this approach is that it is very straight-
forward to implement and requires minimal resources. The disad-
vantage of this option is that the average monthly limits must be
derived without any information about the variability of the efflu-
ent parameter; therefore, the permit writer cannot be sure that
these procedures are protective of water quality criteria. Con-
versely, limits derived from this approach may be overly stringent
and subject to challenge.
The direct application of both the acute and chronic WLAs as
permit limits is another approach that has been used. The WLA
developed for protection against chronic effects becomes the
average monthly limit and the acute WLA becomes the MDL.
EPA discourages the use of this approach. Since effluent vari-
ability has not been specifically addressed with this approach,
compliance with the monthly average (30-day) effluent limit
during critical conditions could exceed the chronic (4-day) WLA.
Whether standards are violated with excessive frequency under
such conditions would depend upon whether the conditions
represented by the worst-case assumptions of the model also
were occurring at the same time. By contrast, compliance with
limits that were developed using statistical procedures have a low
chance of leading to WLA excursions before effluent variability is
accounted for in deriving the limits (see Figure 5-3).
Another permitting approach is to use a narrative "no toxicity"
limit that is measured using a toxicity testing method that em-
ploys only a control and a single exposure at the receiving water
concentration (RWQ. This is sometimes referred to as a "pass/
fall" toxicity test. Although these tests can be less expensive than
full dilution series testing, they provide no knowledge as to the
extent of toxicity present during the test and therefore no data
concerning the seriousness of the impact or the amount of toxic-
ity reduction necessary. The death of a single test animal can
occur at any concentration level beyond the lethality threshold for
the test organism; therefore, such a test is much less powerful
from a statistical standpoint. In addition, it is not possible to
determine dose-response relationships for the test organisms with-
out using multiple effluent concentrations. Dose-response curves
are useful in determining quality assurance of the tests and in
defining threshold dosages for regulatory purposes. Because the
drawbacks of the approach generally outweigh the benefits, EPA
recommends that whole effluent toxicity limits be established
using a statistical derivation procedure that adequately ac-
counts for effluent variability and that monitoring for compli-
ance with whole effluent toxicity limits be conducted using a
full dilution series. '
i
When setting a whole effluent toxicity limit to protect against
acute effects, some permitting authorities use an end-of-pipe
approach. Typically, these limits are established as an LC5Q>100-
percent effluent at the end of the pipe. These limits are routinely
set without any consideration as to the fate of the effluent and the
concentrations of toxicant(s) after the discharge enters the receiv-
ing water. Limits derived in this way are not water quality-based
limits and suffer from significant deficiencies since the toxicity of a
pollutant depends mostly upon concentration, duration of expo-
sure, and repetitiveness of the exposure. This is especially true in
effluent dominated waters. For example, an effluent that has an
LC5o=100 percent contains enough toxicity to be lethal to up to
50 percent of the test organisms. If the effluent is discharged to a
low-flow receiving waterbody that provides no more than a three-
fold dilution at the critical flow, significant mortality can occur in
the receiving water. Furthermore, such a limit could not assure
protection against chronic effects in the receiving waterbody.
Chronic effects could occur if the dilution in the receiving water
multiplied by the acute to chronic ratio is greater than TOO
percent. Therefore, in effluent dominated situations, limits set
using this approach may be severely underprotective. In contrast,
whole effluent toxicity limits set using this approach in very high
receiving water flow conditions may be overly restrictive. Be-
cause of these problems, EPA recommends that all whole
effluent toxicity limits be set as water quality-based limits and
that to do so, the statistical permit limit derivation procedures
discussed in Section 5.4.1 be followed.
5A3 Special Permitting Requirements
Water quality-based permit limit development for discharges to
marine and estuarine waters follows the same basic steps as the
water quality-based spproach for freshwater discharges. There
are some differences in the water quality criteria used as the basis
for protection, the designation of mixing zones, and the water
quality models used to develop WLAs; however these differences
are addressed in the WLA. (See discussions of these elements in
previous chapters.) In addition, there are some special regulatory
considerations associated with these types of dischargers, includ-
ing special reviews of permits with such programs as the Coastal
Zone Management Program. Some discharges also require an
Ocean Discharge Criteria Evaluation under Section 403(c) of the
Clean Water Act (CWA).
i
i
5.4.4 EPA Recommendations for Permitting for Human
\ Health Protection
Permit development to protect against certain routes of exposure
is another key consideration. Ingesting contaminated fish and
shellfish is a toxic chemical exposure route of serious potential
104
-------
human health concern for which there is no intervening treat-
ment process, unlike the drinking water route of exposure. Efflu-
ent 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 limits for pollutants affecting human health is
somewhat different from setting limits for other pollutants be-
cause the exposure period is generally longer than 1 month, and
can be up to 70 years, and the average exposure rather than the
maximum exposure is usually of concern. Because compliance
with permit limits is normally determined on a daily or monthly
basis, it is necessary to set human health permit limits that meet a
given WLA for every month. If the procedures described previ-
ously for aquatic life protection were used for developing permit
limits for human health pollutants, both MDLs and AMLs would
exceed the WLA necessary to meet criteria concentrations. Thus,
even if a facility was discharging in compliance with permit limits
calculated using these procedures, it would be possible to con-
stantly exceed the WLA. This approach clearly is unacceptable. In
addition, the statistical derivation procedure is not applicable to
exposure periods more than 30 days. Therefore, the recom-
mended approach for setting water quality-based limits for hu-
man health protection with statistical procedures is as follows:
• Set the AMI equal to the WLA
• Calculate the MDL based on effluent variability and the
number of samples per month using the multipliers pro-
vided 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
MDL. Both an MDL (weekly average limit for POTWs) and a
monthly average limit are required by EPA regulations, unless
impracticable (40 CFR 122.45(d)) and are applicable for human
health protection. The MDL sets an upper bound on effluent
values used to determine the monthly average and provides a
measure of effluent compliance during operational periods be-
tween monthly sampling.
5.5 SPECIAL CONSIDERATIONS IN USE OF
STATISTICAL PERMIT LIMIT DERIVATION
TECHNIQUES
The following discussion summarizes the effect of changes in the
various statistical parameters on the permit limits that are derived.
An understanding of these relationships is important for the per-
mit writer. Additional considerations of each of these parameters
with respect to the statistical methods for permit limit derivation
also are discussed below.
5.5.1 Effect of Changes of Statistical Parameters on Permit
Limits
• 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 LTA must be lower relative to the WLA to
account for the extreme values observed with high CVs. An
LTA with a zero CV equals the WLA.
• Effect of changes in CV on derivation of permit limits for
a fixed probability basis: As the CV increases, the permit
limits increase (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. Following the steps in Box 5-
2 to derive the LTA will account for such fluctuations.
• Effect of changes in number of monthly samples on
permit limits: As the value for "n" (number of observa-
tions) increases in the average monthly permit limit deriva-
tion equations, the average monthly permit limit decreases
to a certain point. The effect on the average monthly limit
is minimal for values of n greater than approximately 10.
Conversely, as the value for "n" decreases, the AML in-
creases until n=1, at which point the AML equals the MDL
(see Figure 5-7).
Reason: As 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 decreases in absolute value as n
increases. (See additional discussion in Section 5.5.3.)
• Effect of changes in probability basis for permit limits:
As the probability basis for the permit limits expressed in
percentiles (e.g., 95 percent and 99 percent) increases, the
value for the permit limits increases (becomes less strin-
gent). The converse is true as the probability basis de-
creases (see Figure 5-6).
As the coefficient of variation
increases, the long-term
average decreases
1.0
Coefficient of Variation
Figure 5-5. Long-Term Average as a Function of the
Coefficient of Variation
105
-------
Reason: There is a higher probability that any randomly
chosen effluent sample will be in compliance with its permit
limits, if those limits are statistically designed to be greater
than a high percentage (e.g., 99 percent) of all possible
values for a given LTA and CV.
The overall combination of the coefficient of variation, number of
samples, and the assumed probability basis for calculating the LTA
from the WLA, and the most limiting LTA, has different effects on
the derived limits depending upon the selection made for each.
To help illustrate the combined effect of these factors, Figure 5-8
illustrates how the CV, number of samples and probability basis
affect the derivation of the AML. Figure 5-9 illustrates the com-
bined effect of the CV and the probability basis on the derivation
oftheMDL
S.5.2 Coefficient of Variation
Use of the statistical method of permit limit derivation requires an
estimate of the CV of the distribution of the daily measurements
of the parameter after the plant complies with the requirements.
Table 5-3. Multipliers for Calculating Maximum Daily Permit Limits From Average Monthly Permit Limits
- 0.5a2]
AML exp [ZjOn - 0.5on2j
To obtain the maximum daily permit limit (MDL) for a bioconcentratable pollutant, multiply the average monthly permit limit
(AML) (the wasteload allocation) by the appropriate value in the following table.
Each value in the table is the ratio of the MDL to the AML as calculated by the following relationship derived from Step 4 of the
statistically based permit limit calculation procedure.
MDL= exp
where
cn2=
a2
CV
n
zm
za
ln(CV2/n + 1)
In (CV2 +1)
the coefficient of variation of the effluent concentration
the number of samples per month '
the percentile exceedance probability for the MDL
the percentile exceedance probability for the AML.
CV
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2.0
Ratio
Between Maximum
Maximum = 99th percentile
Average = 95th percentile
n=1
1.07
1.14
1.22
1.30
1.38
1.46
1.54
1.61
1.69
1.76
1.83
1.90
1.97
2.03
2.09
2.15
2.21
2.27
2.32
2.37
n=2
1.13
1.25
1.37
1.50
1.622
1.73
1.84
1.94
2.03
2.11
2.18
2.25
2.31
2.37
2.42
2.42
2.52
2.56
2.60
2.64
n=4
1.16
1.33
1.50
1.67
1.84
2.01
2.16
2.29
2.41
2.52
2.62
2.70
2.77
2.83
2.89
2.89
2.98
3.01
3.05
3.07
n=8
1.18
1.39
1.60
1.82
2.04
2.25
2.45
2.64
2.81
2.96
3.09
3.20
3.30
3.39
3.46
3.46
3.57
3.61
3.65
3.67
I;
Daily and Average Monthly
r
E
n=30
1.22 1
1.46 [
1.74 f
2.02 }
2.32
2.62 !
2.91 '•
3.19 j
3.45 \
3.70
3.93
4.13
4.31
4.47 '
4.62
4.62
4.85 ;
4.94 ;
5.02
5.09 ;
I
n=1
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
Permit Limits
Maximum
Average =
n=2
1.07
1.13
1.19
1.24
1.28
1.31
1.34
1.35
1.36
1.37
1.37
1.36
1.36
1.35
1.34
1.33
1.32
1.31
1.30
1.29
= 99th percentile
99th percentile
n=4
1.12
1.24
1.36
1.46
1.56
1.64
1.71
1.76
1.80
1.83
1.84
1.85
1.85
1.84
1.83
1.82
1.80
1.78
1.76
1.74
n=8
1.16
1.32
1.49
1.66
1.81
1.95
2.08
2.19
2.27
2.34
2,39
2.43
2.45
2.46
2.46
2.46
2.45
2.43
2.41
2.38
n=30
1.20
1.43
1.67
1.92
2.18
2.43
2.67
2.89
3.09
3.27
3.43
3.56
3.68
3.77
3.84
3.90
3.94
3.97
3.99
4.00
106
-------
10-
*-> 03
1 S1
rT
I!
6--
4
2-
As the coefficient of variation
Increases, the maximum daily
permit limit per unit LTA Increases
0.0
0.5 1.0 1.5
Coefficient of Variation
2.0
Figure 5-6. Maximum Daily Permit Limit as a Function of the
Coefficient of Variation
0.2-
The greater the number of
samples per month, the lower
the average monthly permit limit
10 20
Number of Samples Per Month
30
Figure 5-7. Relationship Between Average Monthly Permit
Limits and Number of Samples Per Month
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 the best estimate of
the CV that can be reasonably achieved. As explained in Chapter
3, EPA's review of the uncertainly associated with effluent variabil-
ity suggests that a minimum of 10 samples is needed to reason-
ably quantify the CV.
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 systems. In
other words, after permit limits are in place and the permittee is
operating 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 regu-
lated system may behave differently from the nonregulated sys-
tem (e.g., where changes in the treatment facility are planned),
information concerning effluent concentration means and vari-
ability can be obtained from effluent guideline documents for
individual chemical parameters.
Variability associated with effluent levels of both individual chemi-
cals and whole effluent toxicity is 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 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 of the range of
variability of effluents measured by EPA (see Appendix A) and
represents a reasonable degree of relative variability. However,
wherever possible, it is recommended that data on effluent vari-
ability for the pollutant of concern be collected to define a CV
rather than selecting a default value.
5.5.3 Number of Samples
The statistically based method for permit limit derivation results in
an MDL that does not depend on monitoring frequency. How-
ever, the AMI decreases as the monitoring frequency increases,
and a greater number for "n" is inserted in the relevant equations.
Some permit writers are concerned with this outcome because
facilities with more frequent sampling requirements appear to
receive more stringent permit limits than those with less frequent
monthly sampling^requirements.
The AMI decreases as the number of monthly samples increases
because an average of 10 samples, for example, is closer to the
LTA than an average based on 4 samples. This phenomenon
makes AMLs based on 10 samples appear to be more stringent
than the monthly limit based on 4 samples. However, the strin-
gency of these procedures is constant across monitoring frequen-
cies because the probability basis and the targeted LTA perfor-
mance are the same regardless of the number of samples taken.
Thus, a permittee performing according to the LTA and variability
associated with the wasteload allocation will, in fact, meet either
of these AMLs 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 LTA 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 treat-
ment performance, which should, at a minimum, be targeted to
comply with the values dictated by the WLA. Therefore, it is
recommended that the actual planned frequency of monitor-
ing normally be used to determine the value of n for calculat-
ing the AML. However, in situations where monitoring fre-
quency is once per month or less, a higher value for n must be
assumed for AML derivation purposes. This is particularly
applicable for addressing situations such as where a single crite-
rion is applied at the end of the pipe and a single monthly sample
107
-------
3.5
3-
a. LTA at 95% Probability Basis
AML at 99% Probability Basis
i l i i
0.0 0.5 1.0 1.5
Coefficient of Variation
2.0
3.5-
3-
c. LTA at 95% Probability Basis
AML at 95% Probability Basis
2,5-
O
2-
0.5 1.0 1.5
Coefficient of Variation
2.0
b. LTA at 99% Probability Basis
AML at 95% Probability Basis
0.5 1.0 1.5
Coefficient of Variation
d. LTA at 99% Probability Basis
AML at 99% Probability Basis
0.5 1.0 1.5
Coefficient of Variation
Figure 5-8. Effect of Coefficient of Variation on Average Monthly Limits
108
-------
3.5-
3-
0.5
a. LTA at 95% Probability Basis
MDL at 99% Probability Basis
o.o
0.5 1.0 1.5
Coefficient of Variation
2.0
D
»
3.5
2.5-
1.5-1
b. LTA at 99% Probability Basis
MDL at 95% Probability Basis
0-1—L.
chronic
o.o 0.5 1.0 1.5
Coefficient of Variation
2.0
3.5-
2.5
3
(X
Q
s
1.5
0.5
0.0
c. LTA at 95% Probability Basis
MDL at 95% Probability Basis
chronic'
acute.
0.5 1.0 1.5
Coefficient of Variation
2.0
3.5
3 _
2.5
i "
°- 1.5-
0-5
d. LTA at 99% Probability Basis
MDL at 99% Probability Basis
o.o
chronic
acute
0.5 1.0 1.5
Coefficient of Variation
2.0
Figure 5-9. Effect of Coefficient of Variation on Maximum Daily Limits
109
-------
Is contemplated for compliance monitoring purposes, or where
monitoring frequency is only quarterly. In this case, both the
average monthly and the MDL would exceed the criterion. (For
example, for a CCC of 1.0 chronic toxic unit [TUJ applied as a
VVLA at the end of the pipe, both the MDL and AML would be 1.6
TUc; assuming CV=0.6, n=1, and a 99-percent probability basis.)
A discharger could thus comply with the permit limit but rou-
tinely exceed the criterion. Under these circumstances, the
statistical procedure should be employed using an assumed
number of samples of at least four for the AML derivation.
5,5.4 Probability Basis
Selection of the probability basis for use in the equations in Boxes
5-1 and 5-2 is a permitting authority decision necessary for estab-
lishing statistically derived permit limits. Where a permitting
authority does not have specific guidance for the probability
basts, EPA recommends the following:
For calculation of the LTAs from the WLAs (Box 5-2):
• Both acute and chronic WLA—.01 probability (99th per-
centile level).
For calculation of permit limits from the most limiting LTA (Box 5-
1):
• MDL—.01 probability basis (99th percentile level)
• AML—.05 probability basis (95th percentile level).
The probability levels for deriving permit limits have been used
historically in connection with development of the effluent limits
guidelines and have been upheld in legal challenges to the guide-
lines [4]. It is important to note that these levels are statistical
probabilities used as the basis for developing limits. The goal in
establishing these levels is to allow the regulatory agency to
distinguish between adequately operated wastewater treatment
plants with normal variability from poorly operated treatment
plants and to protect water quality criteria.
The level for the calculation of the LTA from the WLA is based
upon EPA's interpretation of the steady state model used to
develop the WLA. EPA considers the WLA to produce an effluent
condition that should never be exceeded whenever the critical
design conditions occur. To characterize this effluent condition,
EPA uses the 99th percentile concentration from the upper tail of
the effluent probabilistic distribution curve. The selection of this
value is one which can have a significant influence on the level of
conservatism In the permit limits. Permit authorities should con-
sider Figures 5-8 and 5-9 to understand the effect of this decision
along with other decisions on the AMLs and MDLs.
5.6 PERMIT DOCUMENTATION
The fact sheet and supporting documentation accompanying the
permit must clearly explain the basis and the rationale for the
permit limits. When the permit is in the draft stage, the support-
ing documentation will serve to explain the rationale and assump-
tions used in deriving the limits to the permittee and the general
public in order to allow public comment on the draft permit.
When the permit is issued, the administrative record for the
facility (particularly the fact sheet) will be the primary support for
defending the permit in administrative appeals including
evidentiary hearings. This information also will serve to alert
compliance/enforcement personnel to any special considerations
that 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 writer in devel-
oping a revised permit.
In 40 Cffl Part 124.56, a fact sheet containing "[a]ny calculations
or other necessary explanation of the derivation of specific efflu-'
ent limitations" for many draft permits is required. Accordingly,
the WLAs along with the required LTA and CV used and the
calculations deriving them must be included or referenced in the
fact sheet. The permit limit derivation method used must also be
explained in the permit documentation. Where a permitting
authority develops a standardized and simplified method for per-'
mit limit development as discussed in Section 5.4.2, the permit-
ting authority may not need to document all of the underlying
assumptions in the fact sheet, provided that the fact sheet refer-
ences a written permit limit development protocol. Any other
guidance used must also be cited.
5.7 EXPRESSING LIMITS AND DEVELOPING
> MONITORING REQUIREMENTS
I " .
Limits must be expressed clearly in the NPDES permit so that they
clearly are enforceable and unambiguous. Chapter 6 discusses
compliance monitoring and enforcement problems that can re-
sult from improperly expressed limits. All limits, both chemical-
specific and whole effluent, should appear in Part 1 of the permit.
Special considerations in the use of both chemical-specific arid
whole effluent toxicity limits are discussed below.
.
.
5.7.1 Mass-based Effluent Limits
Mass-based effluent limits are required by NPDES regulations at
40 \CFR 122.45(f). The regulation requires that all pollutants
limited in NPDES permits have limits, standards, or prohibitions
expressed in terms of mass with three exceptions, including one
for pollutants that cannot be expressed appropriately by mass.
Examples of such pollutants are pH, temperature, radiation, and
whole effluent toxicity. Mass limitations in terms of pounds per
day or kilograms per day can be calculated for all chemical-
specific toxics such as chlorine or chromium. Mass-based limits
should be calculated using concentration limits at critical flows.
For example, a permit limit of 10 mg/l of cadmium discharged at
an average rate of 1 million gallons per day also Would contain a
limit of 38 kilograms/day of cadmium.
j
Mass-based limits are particularly important for control of
bioconcentratable pollutants. Concentration-based limits will not
adequately control discharges of these pollutants if the effluent
concentrations are below detection levels. For these pollutants,
controlling mass loadings to the receiving water is critical for
preventing adverse environmental impacts.
However, mass-based effluent limits alone may not assure attain-
ment of water quality standards in waters with low dilution. In
110
-------
these waters, the quantity of effluent discharged has a strong effect
on the instream dilution and therefore upon the RWC. At the
extreme case of a stream that is 100 percent effluent, it is the effluent
concentration rather than the effluent mass discharge that dictates
the instream concentration. Therefore, EPA recommends that per-
mit limits on both mass and concentration be specified for
effluents discharging into waters with less than 100 fold dilution
to ensure attainment of water quality standards.
5.7.2 Energy Conservation
Water quality-based permit limits by themselves do not provide any
incentive to dischargers to reduce wastewater flows. The reverse is
true; a more dilute effluent means water quality-based limits are
more easily achieved. However, increased flow translates into in-
creased 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.
Therefore, EPA recommends that flow reductions and energy savings
be specifically encouraged where appropriate (usually in dilutions
greater than 100:1) 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 also could
include an energy savings analysis subject to approval by the permit-
ting authority.
5.7.3 Considerations In the Use of Chemlcal-speclfle Limits
Metals
Another common problem encountered in expressing permit limits
occurs for metals. Some water quality standards express numeric
criteria for metals in terms of the dissolved or acid soluble phase of
the metal. NPDES regulations at 40 CFR 122.45(c) require permit
limitations for metals to be expressed in terms of total recoverable
metal unless (1) an effluent guideline requires the use of another
form, (2) technology-based limits are established on a case-by-case
basis, or (3) the approved analytical method measures only the
dissolved form.
Where State water quality standards are expressed directly as total or
total recoverable metals, the permit limit can be established directly.
Where the water quality standards are expressed as dissolved or acid
soluble metal, the permit writer will need to reconcile the different
expressions of metals when establishing the permit limits. Some
State water quality standards implementation policies or procedures
provide the requirements for this conversion. In instances where a
State has no policy or procedure, the permit writer can take one of
four approaches. First, the permit writer could assume no difference
between the dissolved or acid soluble phases and the total recover-
able phase. This is the most stringent approach and would be most
appropriate in waters with low solids, where the discharged form of
the metal was mostly in the dissolved phase, or where data to use the
other options are unavailable. Second, the permit writer could
develop a site-specific relationship between the phases of metals by
developing a relationship through review of information on instream
metal concentrations. This approach requires concurrent sampling
of both metal phases during periods reflective of the environmental
conditions used to determine the WLA. Third, the permit writer
could use a relationship developed by EPA from national data;
this relationship is described in the national guidance for deter-
mining WLAs for toxic metals in rivers. This relationship re-
quires knowledge of instream concentrations of total suspended
solids at the environmental conditions used to determine the
WLA. Fourth, the permit writer could use a geochemical
model, such as the equilibrium metal speciation model
MINTEQA2 (see Chapter 4). However, the input data require-
ment of this model are equivalent to collecting site-specific
data under Option 2. These options will be expressed in more
detail in subsequent guidance issued by EPA.
Update: The Agency has issued "Interim Guidance on Interpreta-
tion and Implementation Aquatic Life Criteria for Metals'." See the
update notice in front of this document for availability.
Detection Level Limits
A commonly encountered problem is the expression of calcu-
lated 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 that are toxic
in extremely low concentrations or that bioaccumulate.
The recommended approach for these situations is to in-
clude in Part 1 of the permit the appropriate permit limit
derived from the water quality model and the WLA for the
parameter of concern, regardless of the proximity of the
limit to the analytical detection level. The limit also should
contain an accompanying requirement indicating the specific
analytical method that should be used for purposes of compli-
ance monitoring. The requirement should indicate that any
sample is analyzed in accordance with the specified method
and found to be below the compliance level will be deemed to
be in compliance with the permit limit unless other monitoring
information (as discussed below) indicates a violation. Sample
results reported at or above the compliance level should be
reported as observed whereas samples below the compliance
level should be reported as less than this level.
The level of compliance cited in the permit must be clearly
defined and quantified. For most NPDES permitting situa-
tions, EPA recommends that the compliance level be de-
fined in the permit as the minimum level (ML). The ML Is
the level at which the entire analytical system gives recog-
nizable mass spectra and acceptable calibration points.
This level corresponds to the lowest point at which the calibra-
tion curve is determined based on analyses for the pollutant of
concern in a reagent water. The ML has been applied in
determinations of pollutant measurements by gas chromatog-
raphy combined with mass spectrometry. The concept of a
minimum level recently was used in developing the Organic
Chemicals, Plastics, and Synthetic Fibers effluent guidelines
[5].
The minimum level is not equivalent to the method detection
level, which is defined in 40 CFR Part 136 Appendix B as the
minimum concentration of a substance that can be measured
and reported with 99-percent confidence that the analyte con-
centration is greater than zero and is determined from the analy-
sis of a sample in a given matrix containing the analyte. EPA is
not recommending use of the method detection level because
quantitation at the method detection level is not as precise as at
111
-------
the ML It is not similar to the practical quantitation limit (PQL),
which is typically set as a specific (and sometimes arbitrary)
multiple of the method detection level. Because the PQL has no
one definition, EPA is not recommending its use in NPDES permit-
ting. Nor is it similar to other terms such as the limit of detection,
limit of quantitation, estimated quantitation limit, or instrument
detection limit
The permitting authority may choose to specify another level at
which compliance determinations are made. Where the permit-
ting authority so chooses, the authority must be assured that the
level is quantifiable, defensible, and close as possible to the permit
level.
Where water quality-based limits below analytical detection
levels are placed in permits, EPA recommends that special
conditions also be included in the permit to help ensure that
the limits are being met and that excursions above water
quality standards are not occurring. Examples of such special
conditions include fish tissue collection and analyses, limits and/or
monitoring requirements on internal waste streams, and limits
and/or monitoring for surrogate parameters. This information
can be used to help support reopening the permit to establish
more stringent effluent limits if necessary.
5.7.4 Considerations In the Use of Whole Effluent Toxicify
Umfts
Test Methods
NPDES regulations at 40 CFR 122.44(i)(l)(iv) require that meth-
ods approved under 40 CFR Part 136 be used for compliance
monitoring, and in the absence of an approved methbd, the
permit must specify the method to be used. The permit should
also carefully consider any other case-specific aspects of the whole
effluent toxicity test method that should be designated in the
permit. Such aspects as the dilutions at which testing will be
conducted, the different species to be used, the specific end-
points, the statistical procedures for analyzing the data, quality
assurance, and other factors should be clearly stated as a permit
condition to assure that the whole effluent toxicity testing that is
performed to ascertain compliance with a limit or monitoring
requirement is the test procedure the regulatory authority desires.
In some instances, promulgated methodologies allow significant
flexibility and choice in how the method is actually conducted. A
simple reference to the methodology in the permit may not result
in the test being conducted as intended.
Units of Expression and Detection Levels
The permit limit for toxicity itself and the detection levels, or
sensitivity levels, associated with the various types of toxicity tests
determine the type of monitoring requirement, which should be
specified with the limit. It is a misconception to think, for ex-
ample, that only acute toxicity tests should be used where the
WLA for acute protection is used to derive the more limiting LTA
or should always be used to monitor for the MDL. It is a similar
misconception to think that only chronic tests should be used
where chronic LTA is limiting or should always be used to monitor
for the average monthly limit The MDLs and AMLs are derived
from the more limiting of the two LTAs. Therefore, either acute or
chronic tests might apply to a given situation depending upon
the test detection levels or test sensitivity.
For example, a limit of 5 TUC (no observed effect concentration
[NOEC] of 20 percent 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 in-
stance, since 5 TUC would be equivalent to 0.25 TUa. Conversely,
if the ACR was 2, then an acute test could be used because 5 TUC
would be equal to 2.5 TUa. Generally, there is no reason to mix
two types of monitoring requirements for the same limit when
limits are derived from the most limiting LTA. Doing so will
confuse the results and complicate assessments of average monthly
limits where sampling frequency is greater than once per month.
The acute toxicity test, when using an LC5Q as the test endpoint,
ha? an upper sensitivity level of 100-percent effluent, or 1.0 TUa.
If less than 50 percent pf the test organisms die at 100-percent
effluent an LC^Q cannot be determined from the test data, and
the true LC$Q value for the effluent cannot be measured. In this
situation, an acute test could still be used for compliance monitor-
ing purposes but the endpoint would need to be changed to a
greater level of sensitivity. The endpoint could be specified in
terms of "ho statistically significant difference in acute toxicity
between 100 percent effluent sample and the control." This is
the most sensitive application of an acute test and could be used
for monitoring compliance with a limit that, because of lack of
available dilution, applies the EPA recommended acute criterion
of 0.3 TUa at the end of the pipe.
However, these tests would not accurately quantify any level of
chronic toxicity present. For chronic testing, an effluent with an
NOEC of greater than 100 percent presents a similar test sensitiv-
ity problem. An effluent with an NOEC of greater than 100
percent contains less than 1.0 TUC and would meet the EPA
recommended chronic criterion for toxicity at the edge of the
mixing zone, if dilution were available, as well as at the end of the
pipe if no dilution were available.
Description of Limits
When toxicity limits are used, additional description of the limit is
required. The limit should be stated in Part 1 as "effluent toxicity"
in the parameter column with "maximum TUs," "minimum ATE
[acute toxicity endpoint]," or "minimum NOEC" in parentheses
underneath. The numerical values should be placed in the appro-
priate concentration column followed by TU or a percent sign. A
footnote should direct the reader to Part 3 for specific require-
ments on how to conduct the tests. The description in Part 3
should accomplish the following:
» Explain how the limit is expressed (e.g., the limit is, the
minimum ATE expressed as percent effluent or the limit is
the maximum TUa)
• Specify the test species and the test methods for compli-
ance monitoring purposes
Describe any special reporting or followup requirements
(e.g., requirements to conduct a toxicity reduction evalua-
tion).
112
-------
The language in Part 3 should be modified as needed to suit the
situation. The following example language is provided only for
purposes of illustration:
• 'The effluent toxicity limit contained in Part 1 is the allow-
able chronic toxicity to the most sensitive of three test
species. It is expressed as the allowable NOEC in percent
effluent. The required test species and the procedures to
follow are described in Short Term Methods for Estimating
the Chronic Toxicity of Effluents and Receiving Waters to Fresh-
water Organisms, EPA/600/4-89/001, March 1989."
• '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 col-
lected sample as soon as practicable."
5.7.5 Selection of Monitoring Frequencies
There is no fixed guidance on establishment of monitoring fre-
quencies. The decision on the monitoring frequency is case-
specific and needs to consider a number of factors, including
those listed below:
• Type of treatment process, including retention time
• Environmental significance and nature of the pollutant or
pollutant parameter
• Cost of monitoring relative to the discharger's capabilities
and benefit obtained
• Compliance history
• Number of monthly samples used in developing the permit
limit
• Effluent variability.
Based upon an array of data analyzed for both individual chemi-
cals and whole effluent toxicity, and independent of other consid-
erations, EPA has observed that ideally 10 or more samples per
month provides the greatest statistical likelihood that the average
of the various monthly values will approach the true monthly LTA
value. In practice, however, selection of monitoring frequencies
will need to consider the previously mentioned factors and arrive
at a reasonable compromise of the appropriate considerations.
5.7.6 Analytical Variability
Permits require monitoring to establish whether a facility is dis-
charging at a level that complies with the permit limits. All
monitoring includes analytical variability. The true concentration
in a sample can be higher or lower than the measured one due to
this variability; however, there is no way to predict which way it
will go.
Historically, EPA has not directly considered analytical variability
from monitoring methods when establishing permit limits. If the
upper bound of the analytical variability was added to the limit,
there would be a higher potential that the permit limit would fail
to protect the wasteload allocation. This would not be consistent
with 40 CFR122.44(d)(1). On the other hand, if the lower bound
of the analytical uncertainty was subtracted from the limit, there
would be better assurance that the limit achieved the WLA. This
approach could be overly conservative given the other factors
used to develop permit limits. EPA believes that its recommended
approach provides a balance between these two extremes.
5.7.7 Antibacksliding
CWA Section 402(o) establishes express statutory language pro-
hibiting the relaxation of permit limits based on water quality.
Under the statute, relaxation of water quality-based limits is per-
missible only if either the requirements of Sections 402(o)(2) or
303(d)(4) are met. These two provisions constitute independent
exceptions to the prohibition against relaxation of permit limits. If
either is met, relaxation is permissible.
Relaxation of Water Quality-based Limits Under
Section303(d)(4)
Section 402(o)(1) prohibits the establishment of less stringent
water quality-based effluent limitations "except in compliance
with Section 303(d)(4)." Section 303(d)(4) has two parts: Para-
graph (A), which applies to "nonattainment waters" and Para-
graph (B), which applies to "attainment waters."
• Nonattainment waters: Section 303(d)(4)(A) allows estab-
lishment of less stringent water quality-based effluent limi-
tations in a permit for discharge into a nonattainment
water only if (1) the existing permit limitation must have
been based on a total maximum daily load (TMDL) or other
WLA established under Section 303, and (2) attainment of
water quality standards must be assured.
• Attainment waters: Section 303(d)(4)(B) allows establish-
ment of less stringent water quality-based effluent limita-
tions in a permit for discharge into an attained water as
long as the revised permit limit is consistent with a State's
antidegradation policy. This is not restricted to limits based
on a TMDL or WLA.
Relaxation of Water Quality-based Limits Under
Section 402
Section 402(o)(2) also outlines exceptions to the general prohibi-
tion against establishment of less stringent water quality-based
permit limits in a permit. Under Section 402(o)(2), the establish-
ment of less stringent limits based on water quality may be
allowed where:
1) There have been material and substantial alterations or
additions to the permitted facility which justify this relax-
ation.
2) Good cause exists due to events beyond the permittee's
control (e.g., acts of God) and for which there is no reason-
ably available remedy.
3) The permittee has installed and properly operated and
maintained required treatment facilities but still has been
unable to meet the permit limitations (relaxation may only
be allowed to the treatment levels actually achieved).
113
-------
4) New Information (other than revised regulations, guidance,
or test methods) justifies relaxation of water quality-based
permit limitations.
This last exception applies to water quality-based permit
limitations only where the revised limitations result in a net
reduction in pollutant loadings and are not the result of
another discharger's elimination or substantial reduction of
its discharge for reasons unrelated to water quality (e.g.,
operation termination).
Although Paragraph 402(o)(2) lists two additional exceptions,
one for technical mistakes and mistakes of law and one for permit
modifications or variances, the statute provides that these excep-
tions do not apply to water quality-based effluent limitations. As a
result, these exceptions do not provide a basis for relaxing water
quality-based limitations.
Relaxation of Water Quality-Based Permit Conditions or Stan-
dards
The provisions in Section 402(o) discussed previously only ad-
dress the relaxation of effluent limits based on water quality. The
relaxation of other permit conditions or standards based on water
quality are governed by EPA's existing antibacksliding regulations
at 40 CfR 122.44(0(1 )• Under these regulations when a permit is
renewed or reissued, interim effluent limitations, standards, or
conditions must be at least as stringent as the final effluent
limitations, standards, or conditions in the previous permit "un-
less the circumstances on which the previous permit was based
have materially and substantially changed since the time the
permit was issued and would constitute cause for permit modifi-
cation...". In other words, unless cause for permit modification is
present, relaxed conditions or standards are not permissible. EPA
regulations setting forth cause for permit modification can be
found at 40 Cf/M 22.62.
Restrictions of Backsliding
Even if any of the backsliding exceptions outlined in the statute or
regulations are applicable and met, Section 402(o)(3) acts as a
floor and restricts the extent to which water quality-based permit
limitations may be relaxed. Paragraph (o)(3) prohibits the relax-
ation of water quality-based permit limitations in all cases if there
will be a violation of applicable effluent limitation guidelines or
water quality standards, including antidegradation requirements.
This requirement affirms existing provisions of the CWA that
require permit limits, standards, and conditions to ensure compli-
ance with applicable technology-based limits and water quality
standards.
5.8 TOXICITY REDUCTION EVALUATIONS
Where monitoring indicates unacceptable effluent toxicity, one
principal mechanism for bringing a discharger into compliance
with a water quality-based whole effluent toxicity requirement is a
toxicity reduction evaluation (TRE) [6], The purpose of a TRE is to
Investigate the causes and to identify corrective actions for diffi-
cult effluent toxicity problems. The permitting authority may
require that the permittee conduct a TRE in those cases where the
discharger is unable to explain adequately and immediately cor-
reci exceedances of a whole effluent toxicity permit limit or
requirement.
A TRE is a site-specific study conducted in a stepwise process to
narrow the search for effective control measures for effluent toxic-
ity. TREs are designed to identify the causative agents of effluent
toxicity, isolate the sources of the toxicity, evaluate the effective-
ness of toxicity control options, and then confirm the reduction in
effluent toxicity. The ultimate objective of a TRE is for the dis-
charger to achieve the limits 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 use other legally binding mechanisms, includ-
ing Section 308 letters, Administrative Orders, or Consent De-
crees, to require a TRE.
I
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 [6, 7, 8]:
i)
i
Generalized Methodology for Conducting Industrial Toxicity
Reduction Evaluations (EPA/600/2-88/070)
Toxicity Reduction Evaluation Protocol for Municipal Wastewa-
ter Treatment Plants (EPA/600/2-88/062)
3) Methods for Aquatic Toxicity Identification Evaluations:
Phase 1 Toxicity Characterization Procedures (EPA/600/3-
88/034)
Phase 2 Toxicity Identification Procedures (EPA/600/
3-88/035)
Phase 3 Toxicity Confirmation Procedures (EPA/600/
3-88/036).
These guidance documents describe the methods and proce-
dures 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 mu-
nicipal wastewater treatment plants to better address the circum-
stances of each type of facility. Procedures for the characteriza-
tion, identification, and confirmation of the causative agents of
effluent acute toxicity have been developed and are described in a
three-phased TIE methods manual. These TIE methods are appli-
cable to both industrial and municipal effluents and are an inte-
gral part of the protocols for TREs described in the industrial and
municipal TRE guidance documents. TIE methods using chronic
toxicity tests for identifying toxicants will soon be developed and
available in a draft guidance document.
5.8.2 Recommended Approach for Conducting TREs
To ensure the successful completion of a TRE, the guidance
documents recommend a systematic, stepwise approach that
114
-------
eliminates 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 causes 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 flow chart,
generalized from the guidance documents, for this approach to
TREs is presented in Figure 5-10., The steps in this flow chart are
summarized in the following discussion.
Determination of TRE Objectives and Development of the 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 (ATE or chronic toxicity endpoint [CTE]) in order to
TRE Regulatory
Requirements
Information and
Data Acquisition
Toxicity
Treatablllty
Evaluation
Facility Operation
and Maintenance
Evaluation
Toxicity
Identification
Evaluation
Control Method
Selection and
Implementation
1
Source
Investigation
1
Followup and
Confirmation
meet a limit or permit condition. TRE plans should be submitted
by the discharger as soon as possible. In some cases, this could be
30 to 60 days following notification that a TRE is required. In
other instances, this period could be longer. These plans are
important for ensuring that the TRE objectives are well under-
stood and that the TRE to be conducted is thorough and repre-
sents a reasonable effort to achieve the required reduction in
effluent toxicity. An implementation schedule should also be
developed describing the timeframe for completion of the specific
components of the TRE plan by the required TRE completion
date. This schedule should be submitted for review in conjunc-
tion with the TRE plan. EPA recommends that the TRE schedule
should be set or approved by the regulatory agency. Approval
of the schedule and the completion date should not imply ap-
proval 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.
To assist in this review, Box 5-3 provides evaluation criteria for TRE
plans. 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 as soon as
possible following notification that a TRE is required. During the
course of the TRE, the regulatory agency should provide over-
sight, as time permits, to make the TRE as effective as possible.
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 collecting and evaluating 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 informa-
tion 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.
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 TRE 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)
Figure 5-10. Generalized TRE Flow Chart
115
-------
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 1 of the TIE to develop this level of certainty. The results of
this evaluation may lead to preliminary strategies for source re-
duction and pollution prevention, including spill or leak preven-
tion, improvements in material handling and disposal practices,
or substitution or re-use of a compound known to be highly toxic.
Toxicity Identification Evaluation
TIE procedures are performed in three phases: characterization,
identification, and confirmation [7]. In each phase, aquatic or-
ganism toxicity tests are used to track toxicity at each step of the
procedure. In most cases, these are abbreviated or shortened
toxicity tests. 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 to decide
which chemical analytical methods will to use in Phase 2 or it can
be used to design treatability studies.
i
i
The results of Phase 1 also may be used to provide additional
confirmation of the effectiveness of any BMP that was imple-
mented in the previous step of the TRE to reduce the effluent
toxicity. This would require conducting at least one Phase 1
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 1 results from samples taken
after BMP implementation.
Box 5-3. Evaluation Criteria for TRE Plans
Are the objectives or targets of the TRE stated clearly and accurately?
Are the schedule and milestones for accomplishing the tasks described in the study plan?
Are the final TRE report, progress reports, and meetings with the regulatory authority included as part
of the schedule?
Are the approaches or methods to be used described to the extent possible prior to beginning the
TRE? !
'
Has available EPA guidance been used in designing the TRE and developing the TRE plan (or if other
methods are proposed, are these sufficiently documented)?
Does the TRE plan specify what results and data are to be included in the interim and final reports?
Does the TRE plan provide for arrangements for any inspections or visits to the facility or laboratory
that are determined to be necessary by the regulatory authority?
Are the toxicity test methods and endpoints to be used described or referenced?
Does the approach described build on previous results and proceed by narrowing down the possibili-
ties in a logical progression?
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 evalua-
tions?
Are optimization of existing plant/treatment operations and spill control programs part of the initial
steps of the TRE?
Does the TRE plan allow a sufficient amount of time and appropriate level of effort for each of the
components of the study plan?
Does the TIE use broad characterization steps and consider quantitative and qualitative effluent
variability? , , ,
Is toxicity tracked with aquatic organism toxicity tests throughout the analyses?
Is the choice of toxicity tests for the TRE logical and will correlations be conducted if the species used
are different from those used for routine biomonitoring? I
Is the laboratory analytical capability and the expertise of the investigator broad enough to conduct
the various components of the evaluation? i
116
-------
In Phase 2 of the TIE, the results of Phase 1 are built upon, and the
TIE proceeds to chemical analyses designed to identify the specific
chemicals causing effluent toxicity. In Phase 3, the identified
toxicants are confirmed using a number of procedures, including
correlation of toxicity with chemical concentration, spiking ex-
periments, toxicity mass balance, and additional test species and
their symptoms.
The current version of the TIE methods uses 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. In order for these methods to be appli-
cable, however, there must be some measurable acute toxicity in
the effluent samples that are to be characterized in Phase 1 and
analyzed in Phase 2. If this approach is used, the appropriate
.chronic toxicity test, as specified in the TRE objectives and permit
requirements, should then be used in the Phase 3 confirmation
procedures. This will confirm that the toxicant(s) identified using
acute'tests in Phases 1 and 2, are indeed causing the whole
effluent chronic toxicity, which must be reduced.
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 previous version of the EPA TIE
methods use acute toxicity tests should not be construed to mean
that TREs cannot be required or conducted for the reduction of
chronic toxicity. These methods provide additional tools to assist
permittees in the reduction of whole effluent chronic toxicity.
Phase 1 procedures that use chronic toxicity tests will soon be
available in draft EPA guidance. These TIE methods are applicable
to freshwater discharges to either saltwater or freshwater receiv-
ing waters. The use of these methods for saltwater receiving
waters may require their adaption for use with marine test species
or, preferably, an initial correlation of the recommended freshwa-
ter TIE test species to the marine species used for monitoring.
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 per-
formed when the specific toxicants have been identified and
influent samples can be analyzed for the presence of these com-
pounds or when potential source streams can be selected for
chemical analysis (based on the results of the initial data acquisi-
tion 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
tracking this toxicity to its source.
Source investigations will lead to control methods, such as chemi-
cal substitution, process modification, treatment of process or
influent streams (pretreatment), and possible elimination of the
process. For POTWs, source investigations may lead to the devel-
opment of local limits or to the requirement that an indirect
discharger evaluate and control their effluent so as to reduce its
toxicity and prevent passthrough at the POTW. The implementa-
tion of source control methods can effectively reduce effluent
toxicity and also can avoid any cross-media transfer of pollutants
to air or sludge, which may occur as a result of end of pipe
treatment. Types of source control methods that have proven to
be effective in reducing effluent toxicity are improvements in
facility housekeeping, chemical substitution, process optimiza-
tion, reclamation/re-use, and pretreatment.
Toxicity Treatability Evaluation
Toxicity treatability evaluations are conducted to identify possible
treatment methods that can effectively reduce effluent toxicity
and may involve modifications or additions to the existing system.
Treatability studies generally use the same type of information on
the nature of the chemicals to be removed as is generated by
Phase 1 of the TIE. These treatability tests should be conducted
on a bench-scale initially and then a pilot scale prior to construc-
tion 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.
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 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 submit-
ted if construction of additional treatment requires a substantial
amount of time.
Followup and Confirmation
After the control method is implemented and the final TRE report
is submitted, the permitting agency should direct the permittee
to conduct followup monitoring to confirm that the reduction in
effluent toxicity is attained and maintained. Normally, this moni-
toring should follow an accelerated schedule, weekly or biweekly
toxicity tests, for a period of 2 to 3 months to confirm the
effectiveness of the controls implemented and the continued
attainment of the TRE objective. This followup monitoring should
use 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.
5.8.3 Circumstances Warranting a TRE
It is the responsibility of the permitting authority to determine if
the permit limits and/or the State water quality criteria have been
threatened or violated and to notify the permittee if a TRE is
required. It is appropriate for the permitting authority to require
additional toxicity testing following the initial exceedance or vio-
lation. This additional testing may precede notification that a TRE
will be required or it may be considered as the initial part of the
117
-------
TRE and be conducted simultaneously with TRE plan develop-
ment and the evaluation of other existing site-specific informa-
tion.
It Is Important to recognize that the purpose of this additional
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 compli-
ance or noncompliance with the limit or permit conditions for
effluent toxicity. These tests do not serve to verify or confirm the
Initial test results from an earlier sample. Instead, the permit
authority shall use the results of these tests to determine if a TRE or
some other action is the appropriate response to the initial occur-
rence of toxicity.
If the permit has a limit for whole effluent toxicity, then generally,
the permit should not include any specific conditions for acceler-
ated toxicity testing or for triggering a TRE or some other action
(e.g., exceedances in two consecutive tests or exceedances in any
three out of five tests). CWA Section 309 requires that any single
violation of a permit limit may be subject to enforcement. The
EPA Compliance Monitoring and Enforcement Strategy for Toxics
Control (January 19, 1989, Appendix B-4) 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 VRAC [violation review action criterion] requiring
professional review." Accelerated monitoring should only be
used to assist In this professional review to determine what, if any,
enforcement response is necessary, including the need for the
permittee to conduct a TRE. It will be necessary for the Region or
State regulatory authority to determine this on a case-by-case
basis. This must be done in a manner consistent with the priori-
ties established in their respective toxics control strategies and
permitting procedures.
In situations where it is determined that accelerated testing is
appropriate, a maximum of weekly tests for a minimum period of
2 months is recommended. This would result in eight tests, plus
the routine monitoring toxicity test that 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 periodi-
cally present at levels above the effluent limits more than 20
percent of the time, a TRE should be required. With toxicity
present at this rate, the TRE protocols will be useful.
In most cases, any one additional exceedance (beyond the initial
routine monitoring toxicity test result) in the accelerated toxicity
tests could result in notification of the permittee that a TRE is
required. Exceptions 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 im-
mediately implemented or cases where additional test quality
assurance/quality control (QA/QQ is necessary or desirable. The
submittal of QC fact sheets for self-biomonitoring (e.g., Appendix
B-2) 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 monitor- *
Ing 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 wi|l address this periodic or recurrent effluent toxicity
problem. In these cases, more elaborate sampling design and
influent or process stream monitoring may 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 per-
mit condition, unnecessary TREs can be avoided. Similarly, con-
ducting accelerated testing as part of the initial steps of a TRE will
allow for the TRE to be ended in its very early stages if the toxicity
is immediately controlled or determined to be episodic or nonre-
current. By following the TRE guidance and incorporating accel-
erated testing into the TRE, unnecessary analyses and expense can
be avoided.
It also is important to note that for the practical purposes of
conducting a TRE (as opposed to the purpose of determining if a
TRE should be required or not), the magnitude of the effluent
toxicity needed to conduct a TRE may be less than the magnitude
or1 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 incor-
porated 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 the case as long as the effluent
toxicity is periodically or consistently present in measurable
amounts in samples of 100-percent effluent. •
[
It also is reasonable for a discharger to initiate a TRE prior to the
establishment of a permit limit for toxicity if unacceptable levels of
toxicity are found in the effluent through routine monitoring or
through inspection and compliance sampling by the regulatory
authority. Under these circumstances the regulatory authority
Will need to identify what constitutes unacceptable levels of toxic-
ity since this will not be defined by a permit limit (see Chapter 3
on determining the reasonable potentialfor excursions of water
quality standards). It also is not unreasonable for the discharger
to voluntarily initiate a TRE under these circumstances.
I
5,8.4 Mechanisms far Requiring TREs
There are a number of mechanisms that can be used to require a
TRE. In most cases, the TRE should be required by a Section 308
letter or by an enforcement action, such as a Section 309 Admin-
istrative Order or a Consent Decree. The permittee should receive
nbtification from the permit authority of what response is re-
qbired. This enables the permit authority to assess whether a TRE
is the appropriate action to pursue. If effluent toxicity reappears
following the: successful completion of a TRE, then the permit
authority should be able to review this type of situation to deter-
rrjiine if an additional TRE is appropriate or if some other action is
required. In general> when the permit is issued with whole
effluent toxicity limits in Part 1 of the permit, TRE requirements
should be used where necessary to bring the permittee into
compliance with those limits. Box 5-4 provides example lan-
118
-------
guage for effluent toxicity limits, developed as part of the Whole
Effluent Toxicity Basic Permitting Principles and Enforcement Strat-
egy (Appendix B-4).
Box 5-5 presents sample language for use in requiring TREs by a
Section 308 letter or a Section 309 Order. This sample language,
especially the reporting dates, 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 order to conduct a TRE, reason-
able timeframes are 6 to 18 months for an industrial discharger
and 12 to 24 months for a municipal wastewater treatment plant.
For POTWs, it may take longer to conduct a TRE due to lengthy
government contracting procedures, large sewer collection sys-
tems, and less influent constituent control. It should be recog-
nized that extensions to these initial timeframes may be granted if
the progress reports demonstrate that this is warranted. In situa-
tions where reductions in chemical concentrations to meet chemi-
cal-specific limits are needed as well as reductions in effluent
toxicity, the timeframes may be adjusted to enable those efforts
to proceed simultaneously.
Box 5-4. Model Permit Language for Effluent Toxicity Limits
Part 1 .A. Final Effluent Limits 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 limits and monitoring requirements
from the following outfall(s): 001.
Effluent Characteristic
Discharge Limit Concentration Monitoring Requirement
Reporting
Code/Units Parameter
Daily
Maximum
Monthly
Average
Measurement
Frequency
Sample
Type
—Til
Toxicity
10.0
5.0
x/month composite
The permittee shall use the toxicity testing and data assessment procedures described in Part 3.B of this
permit.
Box 5-5. Example Language for Requiring Toxicity Reduction Evaluations
The discharger shall demonstrate that effluent toxicity-based permit limits described in Part 1 .A. of the
permit are being attained and maintained through the application of all reasonable treatment and/or
source control measures. Upon identifying noncompliance with those limits the discharger shall initiate
corrective actions according to the following schedule:
Task Deadline
1. Take all reasonable measures necessary to
reduce toxicity immediately.
2. Submit a plan and schedule to attain continued
compliance with the effluent toxicity-based permit
limits in Part I.A.,where source of toxicity is known,
if immediate compliance is not attained.
3. Submit a TRE study plan detailing the toxicity
eduction procedures to be employed where source is
unknown and toxicity cannot be immediately controlled
through operational changes. EPA's Toxicity Reduction
Procedures, Phases 1, 2, and 3 (EPA-600/3-88/034, 035,
and 036) and TRE protocol for POTWs (EPA-600/2-88/062)
shall be the basis for this plan and schedule.
Within 24 hours
Within 30 days
Within 45 days
119
-------
Box 5-5. Example Language for Requiring Toxicity Reduction
Evaluations (continued)
4. Initiate TRE plan.
5. Comply with approved TRE schedule.
6. Submit results of the TRE, including summary of
findings, corrective actions required, and data generated.
7. Implement TRE controls as described in the final report.
8. Complete TRE implementation to meet permit limits
and conditions.
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 XX
months from initial noncom-
pliance.
120
-------
CHAPTERS
REFERENCES
1. Marr, J.K., and R.P. Canale. 1988. Load Allocation for Toxics 5.
Using Monte Carlo Techniques. Journal WPCF 60(5):659-
66.
2. Freedman, P.L., j.F. Pendergast, C. Wilber, and S.C. Chang.
1988. Seasonal Changes and Effluent Limits, journal WPCF 6.
60(3):317-23.
3. Parkerton, T.F., S.M. Stewart, K.L. Dickson, J.H. Rodgers, and
F.Y. Saleh. 1989. Derivation of Site-Specific Water Quality 7.
Criteria for Zinc: Implications for Wasteload Allocation.
Research ]oumal WPCF 61 (11,12):1636-44.
4. Shell Chemical Company NPDES Permit Nos. LA0005762,
LA0050962, TX00048663, Appeal No. 85-14, 85-15, 85-
16 before the Administrator, U.S. EPA. U.S. EPA Judicial 8.
Officer decision, October 20,1987.
U.S. EPA. 1987. Development Document for Effluent Guide-
lines and Standards for the Organic Chemicals, Plastics and
Synthetic Fibers Point Source Category, volumes 1 and 2; EPA
440/1-87/009.
U.S. EPA. 1988. Toxicity Reductions Evaluation Protocol for
Municipal Wastewater Treatment Plants. EPA/600/2-88/
062.
U.S. EPA. 1988. Methods for Aquatic Toxicity Identification
Evaluations: Phase 1, Toxicity Characterization Procedures
(EPA/600/3-88/034); Phase 2, Toxicity Identification Proce-
dures (EPA/600/3-88/035); and Phase 3, Toxicity Confirma-
tion Procedures (EPA/600/3-88/0360.
U.S. EPA. 1988. Generalized Methodology for Conducting In-
dustrial Toxicity Reduction Evaluations. EPA/600/2-88/070.
ADDITIONAL REFERENCES
Aitchinson, ]., and J.A.C Brown. 1963. The Lognormal Distri-
butions. London: Cambridge University Press.
Gilliom, R.H., and D.R. Helse. 1986. Estimation of Distribu-
tional Parameters for Censored Trace Level Water Quality
Data 1 and 2. Water Resources Research 22:135-55.
Kahn, H.D., and M.B. Rubin. 1989. Use of Statistical Meth-
ods in Industrial Water Pollution Control Regulations in the
United States. Environmental Monitoring and Assessment
12:129-48.
Shumway, R.H., A.S. Azari, and P. Johnson. 1989. Estimat-
ing Mean Concentrations Under Transformation for Envi-
ronmental Data with Detection Limits. Technometrics
31(3):347-56.
121
-------
-------
6. COMPLIANCE MONITORING AND ENFORCEMENT
6.1 INTRODUCTION
Once a water quality-based permit containing limitations and
conditions to control effluent quality is issued, the permittee is
responsible for attaining, monitoring, and maintaining compli-
ance with the requirements of that National Pollutant Discharge
Elimination System (NPDES) permit. Failure to comply with any
requirements stated in the permit is a violation of the Clean Water
Act (CWA).
The Environmental Protection Agency (EPA) and authorized State
agencies are responsible for tracking compliance with and enforc-
ing NPDES permit requirements in the enforcement of the CWA.
Section 308 of the CWA and equivalent State statutes enable the
regulatory agency to verify compliance with permit conditions
(including water quality-based toxics limitatipns and compliance
schedules) by authorizing the agency to impose on permittees
requirements for sampling and analysis, record-keeping, and re-
porting. Section 308 also authorizes access by EPA or State
agencies to facilities and records for verifying compliance with
permit conditions. All records associated with monitoring must
be maintained by the facility and available for a 3-year inspection
period in conformance with 40 CFR Part 122.41.
The CWA establishes the authority to enforce water quality-based
permit conditions. 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, including
the procedures used to determine the wasteload allocation based
on applicable water quality standards and the procedures used to
derive limitations from the wasteload allocation.
6.2 PERMIT REQUIREMENTS
The conditions that are to be included in NPDES permits are
described in 40 CFR Part 122 Subpart C. In general, permits
include effluent limitations, schedules of compliance, and accom-
panying reporting requirements. Permits should prescribe the
self-monitoring procedures, frequency of analysis, sampling loca-
tion 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 inappro-
priate for a particular permit limitation. Permits should define any
effluent limitations and explain specific procedures for calculating
averages of data if different from arithmetic averaging. Permits
should identify what information must be retained by the permit-
tee, and what data must be submitted to EPA or the State. Results
from self-monitoring required by the permit are reported on
discharge monitoring reports (DMRs) that generally are submit-
ted monthly. Sampling and analysis that is done more frequently
than required by the permit must be included in the DMR.
6.3 COMPLIANCE MONITORING
Since most of the routine information gathered in compliance
monitoring results from permittee self-monitoring, quality assur-
ance (QA) is as important as compliance with limits. It is essential
that permittees develop and adhere to a QA plan consistent with
the required monitoring and analyses. The permittee is responsible
for maintaining data to demonstrate compliance with QA proce-
dures established in the test methodology or as specified in the
permit.
The regulatory agency generally has three ways of determining
compliance with an NPDES permit and assuring adequate QA:
self-monitoring reports, DMR/QA results, and inspections. Each
of these methods is discussed below.
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 com-
pliance. These reports include DMRs and reports of progress on
compliance schedules. DMRs contain information on the sam-
pling 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 will automatically "flag" violations of permit limitations,
compliance schedules, and reporting requirements.
In order to detect any problems with the quality of the sample
analysis, 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 re-
quired through the permit or through a Section 308 order. The
QA data are compared to a reference QA data sheet that can be
completed by the regulatory authority to indicate acceptable
ranges of values for the required protocol. Appendix B-5 provides
an example of a reference QA data sheet for a whole effluent
toxicity test. 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 pro-
tocol with corresponding QA procedures. It also is important to
note that the signatory's certification of effluent data certifies
123
-------
compliance with the specified protocols. Any problems with QA
should be reported at the time of DMR submission and the testing
repeated.
6.3.2 Discharge Monitoring Report/Quality Assurance
(DMR/QA)
The 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 CWA Section
308. 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 permit-
tees' 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 required to follow the instructions for reporting
results and include a signed certification statement in accordance
with 40 CFR 122.22.
Regulatory agencies conduct followup 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 tested in the
DMR/QA program. The Environmental Monitoring and Support
Laboratory (EMSL) in Cincinnati has developed a reference toxicant
DMR/QA sample for permittees with whole effluent toxicity
monitoring requirements. National implementation is occurring
in 1991.
6.3.3 Inspections
Inspections are conducted by the regulatory authority or its con-
tractors to address specific violations or problems and 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 concur-
rent analyses. EPA has defined several types of inspections based
on the tasks that are included in the NPDES Compliance Inspection
Manual[2]. Because regulatory authorities are expected to inspect
all major permittees annually regardless of compliance status,
nonsampling inspections (which are generally less resource-in-
tensive) are encouraged for routine evaluation of permittee per-
formance. However, sampling inspections are still encouraged to
address permitting and enforcement priorities. For that reason,
the regulatory agency must have the full capability to assess
effluent compliance through inhouse resources or contract support.
Inspections that focus on toxics control can provide useful infor-
mation 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 the NPDES Compliance Inspection Manual, Chapter
7 [2].
6.4 VIOLATION REVIEW
i
Review of permittee self-monitoring data to determine appropri-
ate enforcement response generally involves a two-tiered review.
The 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 over-
sight. As violations increase in magnitude, duration, orfrequency,
they generally are assigned to personnel who are responsible for
the second-tier review (determining what enforcement action, if
any, is appropriate). The guidelines for this process 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. Section 6.5 discusses
the enforcement decision process.
•
V\fhen 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 assigned for review by
personnel qualified to determine the significance of the results. If
the monitoring is insufficient to determine compliance with efflu-
ent limitations, a warning letter or Section 308 letter requiring
that the tests be repeated using acceptable procedures would be
an appropriate response.
Tracking a permit or Section 308 letter that contains "monitor
only" requirements requires both a compliance review (e.g., to
determine if results of acceptable quality were submitted on
time), and an action review (e.g., to determine if the permit
should be modified or re-issued to include a limitation). This
second review should be assigned to personnel who are qualified
to make this regulatory decision.
In addition to the guidelines for reviewing monitoring data in the
absence of a specific effluent limitation, EPA also has recommended
a criterion for determining which effluent violations must be as-
signed for review by a professional who will determine if a formal
enforcement action is needed, or if a phone call, warning letter, or
Section 308 letter is more appropriate. These criteria are known
as the Violation Review Action Criteria and are listed in the EMS.
In the case of a whole effluent toxicity limitation, any violation
must be reviewed by a qualified professional responsible for the
enforcement decision. EPA makes this recommendation to ensure
that adequate attention is given to QA and to ensure that additional
testing is required if permitted testing frequency is less than once
p&r month.
i
In the case of a violation of a chemical-specific permit limitation,
EpA recommends that monthly average limitation violations be
reviewed by a professional for potential enforcement response
whenever two or more violations occur in a 6-month period.
Seven-day average and daily maximum violations should likewise
be reviewed if a minimum of two or four, respectively, occur
during the course of 1 month. Although there is no delineation
between technology-based versus water quality-based limitations
in these Violation Review Action Criteria, Regions and States may
wjsh to adopt a criteria of "any violation" for all water quality-
124
-------
based, chemical-specific limitations as these criteria are solely to
determine the level of review and do not prescribe enforcement
action.
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 construction or
other compliance measures.
Exceeding a permit limitation is a violation subject to enforce-
ment. Some members of the regulated community have expressed
concerns that single violations of stringent water quality-based
limitations will result in unreasonable enforcement actions. EPA's
guidance outlines a systematic review of all violations to determine
the appropriate level of response. This guidance generally suggests
an informal response for minor or infrequent violations, escalating
to formal enforcement and perhaps penalties for more frequent
and environmentally harmful violations.
In evaluating appropriate response to violations, EPA's "Enforce-
ment Response Guide" 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 [4] (see Appendix B-
4). This strategy expects that all available avenues to compliance
will be explored by the permittee, that the treatment facility is
designed, constructed, maintained, and operated to achieve all
water quality-based, chemical-specific or best available technology/
secondary treatment limitations, that chemical or process substi-
tutions have been attempted and pretreatment explored, and
that, in the case of publicly owned treatment works (POTWs),
pretreatment program requirements and local limits have been
established and enforced. The strategy further expects that the
permittee will pursue a Toxicity Reduction Evaluation (TRE) as
discussed in Chapter 5 in compliance with enforcement require-
ments or under its own initiative. If all of these expectations have
been met and the facility is unsuccessful in identifying the cause,
source, or treatability of toxicity despite making good-faith efforts
to do so, the strategy allows for relief from civil penalties. The
underlying responsibility to achieve compliance with the permit
limitation remains in effect.
Some members of the regulated community have requested EPA
and several State agencies to define more clearly enforcement
discretion with respect to violations of whole effluent toxicity
limitations. To define enforcement discretion would in effect
make it no longer discretionary. Furthermore, the purpose of
such guidance would be questionable as individual enforcement
responses by EPA and the States are open to review by the public
and the courts. In lieu of such additional guidance on enforce-
ment discretion, it is recommended that Regions and States
adhere to the principles presented in the EMS, the strategy, and in
this document.
EPA also has developed a policy [5] on the assessment of appro-
priate civil penalties in both administrative and civil judicial ac-
tions in response to any CWA violation. This policy bases the
penalty amount on the seriousness of the violation, the economic
benefit enjoyed as a result of delayed compliance, any history, of
such violations, any good-faith efforts to comply, and the violator's
ability to pay. In no instance can this calculated penalty exceed
the statutory maximum penalties defined in CWA Section 309.
If any violation occurs, the permittee has the responsibility of
informing the regulatory, agency. If the violation potentially
endangers health or the environment, the violation must be
verbally reported to the regulated agency within 24 hours and the
permittee must submit a noncompliance report within 5 days of
violation detection. If there is no danger to health or the envi-
ronment, the written 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.
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 improvement of treatment efficiency, and any
other initial steps necessary to resolve the violation and mitigate
the environmental effects. These immediate investigatory and
corrective steps also should provide information that may be used
in developing a compliance schedule if the violation is not resolved
quickly.
When a water quality-based toxicity limit is violated, the regulatory
agency may require additional monitoring to determine the fre-
quency and duration of the violation. If the permit limit is not met
quickly through improved housekeeping, operation, or raw waste
control (e.g., POTW enforcement of pretreatment requirements,
or chemical substitution by industries), requiring a TRE as discussed
in Chapter 5 may be appropriate. Where toxicity-based limitations
are in effect, the enforcement response must require expeditious
compliance with the limit.
Available enforcement mechanisms include Section 308 orders,
Section 309 Administrative Orders, Administrative Penalty Orders
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 worded carefully
so that they clearly are understood, easily tracked, and expeditiously
enforced.
Violating limitations of pollutants at concentrations that pose a
threat to human health should receive immediate enforcement
attention to prompt rapid resolution of the noncompliance. The
regulatory agency should consider the pollutant concentration,
exposure route, and whether or not the pollutant exhibits a
threshold response in determining if a schedule may be allowed.
Immediate injunctive relief (such as a temporary restraining order
or preliminary injunction) should be sought when necessary to
protect public water supplies and fish and shellfish areas from
imminent or substantial impairment.
125
-------
6.6 REPORTING OF VIOLATIONS
The regulatory authority is responsible for reporting to the public
on permittees in violation. Reporting requirements for the Quar-
terly Noncompliance Report (QNCR) of major permittees in vio-
lation of their NPDES permits are established in 40 CFR 1 23.45.
Reporting of violations of water quality-based monitoring, limita-
tions, schedules, and reporting requirements by major facilities
must be consistent with 40 CFR 123.45. Violations of permit or
enforcement order conditions by major permittees must be re-
ported as follows [6]:
• Effluent violations (chemical-specific and whole effluent tox-
icity) must be reported on the QNCR if the violation has the
potential to have caused a water quality problem (40 CFR
Chemical-specific toxic permit limit violations must be re-
ported on the QNCR if two or more monthly average
measurements in a 6-month period exceed the limit by a
factor of 1 .2 for a Group I parameter or 1 .4 for a Group II
parameter as defined in the Regulation, or if four or more
monthly average measurements in a 6-month period ex-
ceed the limit by any amount (40 CFR 1 23.45(a)(2)(ii)(C)).
Any violation during the quarter of an interim monthly
average chemical-specific toxic limit established in an ad-
ministrative order or court order/consent decree must be
reported on the QNCR (40 CFR 1 23.45(a)(2)(ii)(A)). (Note:
Whole effluent toxicily is not characterized as a Group I or
Group II parameter, and as such, must be evaluated on a
professional judgement basis under 40 CFR
• Compliance schedule milestones that are not met within 90
days of the scheduled date must be reported on the QNCR
(40 CFR 123.45(a)(2)(ii)(B)).
• Failure to submit a report within 30 days of the due date
must be reported on the QNCR (40 CFR 1 23.45(a)(2)(ii)(D)).
126
-------
CHAPTERS
REFERENCES
1. Jensen, L.j. Permit Compliance System (PCS) Policy Statement. 5. Jensen, LJ. 1986. Clean Water Act Penalty Policy for Civil
^ .,„-,,-,.,-. ,. , . .,, ,., ,•„„„ Settlement Negotiations.
2. NPDES Compliance Inspection Manual, May 1988. y
, ' , . „ , „' , „ .. 6. Hanmer, R.W. 1986. Guidance for Preparation of Quarterly
3. Enforcement Management System for the Nat,onal Pollutant Qnd Sem,.Annua, Noncomp,iance Repo^
Discharge Elimination System, September 1986.
4. Hanmer, R.W. 1989. Whole Effluent Toxicity Basic Permitting
Principles and Enforcement Strategy.
127
-------
-------
7. CASE EXAMPLES
7.1 INTRODUCTION
This chapter presents examples of the development of water
quality-based discharge limits to illustrate the integration of the
guidance of the previous chapters. There are three examples: an
industrial discharge with ample dilution, a publicly owned treat-
ment works (POTW) with moderate dilution, and the combina-
tion of an industrial facility and a POTW discharge to the same
reach.
7.2 CASE1: INDUSTRIAL DISCHARGE
The first example is the Jaybird Corporation, a metal finishing
firm. The NPDES permit for the facility is about to expire, and the
corporation has submitted an application for a new permit. The
example shows the steps that a permitting authority would take
to determine if a water quality-based effluent limit is necessary
and then to establish such a limit. The example also illustrates
when best available technology (BAT) limits are applied instead of
water quality-based limits, the use of human health criteria, and
the variations in the limits derived by different wasteload alloca-
tion methods.
7.2.1 General Site Description and Information
The Jaybird Corporation facility discharges into the Locapunct
River. The river is approximately 60 miles long and its banks are
occupied by small towns separated by woodland and farmland.
The river is classified by the State in the water quality standards as
having designated uses of a fish habitat, primary contact recre-
ation, and a drinking water supply. For these uses, the State has
adopted the federal water quality criteria into the water quality
standards to protect aquatic life and human health. The State
standards also includes a narrative criterion of "no toxics in toxic
amounts" for other toxic materials.
Water quality monitoring indicates some infrequent excursions
above water quality criterion for copper and nickel. These pollut-
ants have been found in measurable quantities in the effluents of
several facilities.
The Jaybird Corporation is a metal finishing facility that specializes
in copper plating of lead shells for a nearby military installation.
As a metal finisher, the Jaybird Corporation is relatively small with
a discharge of 0.034 cfs (0.022 mgd). The effluent at the Jaybird
Corporation is treated by precipitation and settles before dis-
charge through a multiport diffuser. The corporation is subject to
BAT and best practicable technology (BPT) effluent limits for the
metal finishing industry.
7,2.2 Effluent Characterization for Specific Chemicals
The permitting authority has adopted a procedure in which pol-
lutants concentrations in each facility are evaluated for the poten-
tial to cause, have the reasonable potential to cause, or contribute
to an excursion of the water quality standards. The authority used
the effluent characterization process for specific chemicals de-
scribed in Chapter 3 in this evaluation. In general, the procedures
are designed to determine which pollutants are of concern and
which require effluent limits.
Step 1: Identify Pollutants of Concern
Data were obtained from a number of sources to identify and
quantify the pollutants of concern in the Jaybird Corporation
effluent:
• Effluent chemical concentrations were taken from trie Per-
mit Application Form 2C, Discharge Monitoring Reports
(DMRs), EPA's Permit Compliance System (PCS), and per-
mit files.
• EPA's STORET data base was used to obtain U.S. Geological
Survey flow data and ambient monitoring data for the river.
• BAT limits for the metal finishing industry were obtained
from 40 CFR 433 Subpart A.
The permitting authority noticed in review of these data that the
information in Form 2C replicated the information in the DMRs,
and therefore decided to use the DMR data as the primary basis
for characterizing the effluent. These data for toxicants DMRs are
shown in Table 7-1. For those parameters currently not covered
by the permit, Form 2C data indicated that pollutant concentra-
tions were below detection limits. The permitting authority re-
quested information from the facility showing the detection levels
used; these levels were consistent with the detection levels listed
in the National Pollutant Discharge Elimination System (NPDES)
regulations at 40 CFR 136.
The effluent from the Jaybird Corporation is regulated by the
Metal Finishing Point Source Category effluent guidelines at 40
CFR 433 Subpart A. These guidelines regulate the following toxic
pollutants: cadmium, chromium, copper, cyanide, lead, nickel,
silver, zinc, and total toxic organics.
Although these parameters were regulated at the Jaybird Corpo-
ration, the orily toxic pollutants evident in the discharge were
lead, copper, and nickel. The facility's treatment system reduced
concentrations of other pollutants to below detection.
Step 2: Determine the RAC, CMC, and CCC for Pollutants of
Concern
The State has adopted numeric water quality criteria for acute
toxicity (criterion maximum concentration [CMC]), chronic toxic-
ity (criterion continuous concentration [CCC]), and protection of
human health (reference ambient concentration [RAC]). The
water quality standards present the CMC and CCC criteria as
equations based on ambient hardness concentrations. The stan-
dards require that the 85th percentile lowest hardness be used.
This value is 100 mg/l as CaCO3 for the Locapunct River.
129
-------
Table 7-1. Effluent Data for the jaybird Corporation
n
1
2
3
4
S
6
7
8
9
10
11
12
Mean
SD
CV
Max
Min
N
Copper
US/1
1,317
1,092
1,073
1,059
1,072
1,677
2,664
1,058
3,439
6,596
1,211
1,082
1,945
1,650
0.8
6,596
1,058
12
Lead
W/l
187
230
258
423
227
275
364
170
259
264
267
175
258
74
0.3
423
170
12
Nickel
H9/I
223
261
464
341
369
1,058
199
259
437
773
300
356
420
252
0.6
1,058
199
12
Toxicity
TUC
5
10
5
20
10
7.1
0.7
20
5
4
Source: OMR data for chemicals; 308 request for whole effluent toxicity.
Notes:
Meuts reported as total recoverable metals; toxicity reported in chronic toxic
units (100/NOEq.
The permittee did not use a geometric dilution series for the toxicity tests. The
results are the highest toxic units for any of the test organisms used.
The aquatic toxicity criteria for metals in the standards are ex-
pressed as the acid soluble form of the metal. The State has
adopted a ratio to express the acid soluble form of metals as the
total recoverable form for the purposes of developing NPDES
permit limits. This ratio is based on historical data that the State
has collected for rivers in the basin where the Locapunct lies. The
values of the ratio are 0.35 for lead, 0.70 for copper, and 0.85 for
nickel. The standards consider the criteria for human health
protection to be in the total recoverable form of the metal.
Based on the hardness and acid soluble-to-total recoverable ra-
tios, the applicable state water quality criteria are the following:
Pollutant
Lead
Copper
Nickel
CCC
Oig/D
9.1
17.1
188
CMC
(ug/D
235
25.7
1,647
RAC
(ng/i)
50
NA
13.4
Step 3: Determine Dilution for Aquatic Life and Human Health
Impacts
The State water quality standards require that compliance with
water quality criteria be achieved at the edge of the mixing zone.
The standards specify the minimum dilution at which the criteria
apply. These are the 7Q10 flow for the CCC, the 1Q10 flow for
the CMC, and the harmonic mean flow for human health criteria
(RAC). The U.S. Geological Survey operates a gaging station on
the river; the flow statistics were calculated using the data from
this station:
Harmonic mean flow = 38.0 cfs
• 7Q10 flow =13.0 cfs
• 1Q10 flow =10.1 cfs.
The facility provided a study of the outfall that showed that the
multiport diffuser quickly achieved complete mixing across the
width of the river. Dilution at the edge of the mixing zone could
therefore be characterized by the complete mixing equation:
+ CsQs)/(Qe + Qs)
where
C = the receiving water concentration
i
Cem=the maximum effluent concentration
Qe = the effluent flow
Cs = the receiving water background concentration
ds = the appropriate receiving water flow.
Step
4: Determine Reasonable Potential for Excursions
To determine if the facility discharge was expected to cause or
have the reasonable potential to cause the CMC, CCC, or RAC to
be exceeded in the receiving water, the maximum receiving
water concentration of each pollutant was first compared to the
appropriate receiving water criterion. If the criteria were ex-
ceeded, then this was considered evidence that a water quality-
based limitation must be developed.
Maximum expected concentrations were calculated using the
average effluent flow, maximum effluent concentrations, back-
ground receiving water concentrations, and the relevant receiv-
ing water flow: the 1Q10 for the CMC, the 7Q10 for the CCC, or
the harmonic mean for the RAC. The background receiving water
concentrations for total recoverable metals were obtained from
ST6RET data:
Lead 1.6ug/l
Copper 4.8ug/l
Nickel 13.2u.g/l
The maximum effluent concentration was estimated using the
statistical approach in Chapter 3. There were 12 concentrations of
each metal reported in the DMRs. For lead, these concentrations
had a maximum value of 423 u.g/1, an arithmetic mean of 258 ug/
I, an arithmetic standard deviation of 74, and an arithmetic
coefficient of variation of 74/258, or 0.3. This coefficient of
variation and the number of observations determined which mul-
tiplier was selected from Table 3-1. In this case, the multiplier
value for 12 observations and a CV of 0.3 was interpolated from
the values for 12 observations and CVs of 0.2 and 0.4. The 99th
percentile multiplier was estimated to be 1.7. Similar calculations
were conducted for copper (multiplier of 2.8) and nickel (multi-
plier of 3.7).
130
-------
The receiving water concentration for lead for comparison with
the CCC was calculated using data from Table 7-1:
C = [(1.7 x 423 |ig/l x 0.034 cfs) + (1.6 ng/l x 13 cfs)1
(0.034 cfs + 13 cfs)
= 3.5ug/l
where
13 cfs = the receiving water flow at 7Q10
0.034 cfs = the mean effluent flow
423 (ig/l = the maximum effluent concentration
1.7 = the statistical effluent multiplier to estimate the
99th percentile concentration
1.6 jig/1 = the background receiving water concentration.
The value of the calculated receiving water concentration,
3.5 |ig/l, was less than the chronic water quality standard of 9.1
Hg/l for lead, and therefore there is no reasonable potential for the
CCC to be exceeded.
Using the effluent data presented in Table 7-1 , the receiving water
concentration is compared to the CMC as:
C= [(1 .7 x 423
x 0.034 cfs) + (1 .6 |i/l x 1 0.1 cfs)1
(0.034 cfs + 10.1 cfs)
= 4.0ng/i
where 1 0.1 is the receiving water 1 Q1 0 flow and the other values
are identical to those for the CCC comparison. The resulting
concentration of 4.0 p.g/1 was less than the acute standard of 234
Hg/l for lead. There is no reasonable potential for the CMC to be
exceeded. •
For human health criterion evaluation, the receiving water con-
centration for compared to the RAC was calculated as:
x 38 cfs)1
C= [(1 .7 x 423 n.g/1 x 0.034 cfs) + (1 .
(0.034 cfs + 38 cfs)
= 2.2ug/l
where 38 cfs is the harmonic mean flow and other values are the
same as above. This value was less than the human heath criteria
value of 50 |ig/l for lead, so there is no reasonable potential for the
RAC to be exceeded. :
Similar calculations were done for copper and nickel:
Criterion
Receiving Water
Concentration (ng/l)
Copper
CCC
CMC
Nickel
CCC
CMC
RAC
17.1
25.7
188
1^647
13.4
22.0
26.9
15.9
16.6
14.1
The effluent characterization showed the reasonable potential for
excursions above the CCC for copper and above the RAC for
nickel. Therefore, permit limits are necessary for these two pollut-
ants.
7.2.3 Effluent Characterization for Whole Effluent Toxicity
Whole effluent toxicity also was evaluated since there was a
potential for excursions above the narrative water quality criterion
due to the combination of effluent toxicants with other toxicants
in the receiving water and in the effluent but below the detection
level. The procedures used below follow those presented sche-
matically in Figure 3-2, Chapter 3.
Step 1: Dilution Determination
The initial dilution determination was used to establish the types
of toxicity tests that are conducted to characterize the effluent.
The dilution at the low-flow characteristics for the facility is the
following:
At the 7Q10, dilution = (0.034 cfs + 13 cfs)/0.034 cfs
= 383
At the 1Q10, dilution = (0.034 cfs +10.1 cfs)/0.034 cfs
= 298.
Step 2: Conduct Toxicity Testing
EPA recommends that a discharger having a dilution between
100 and 1,000 be required to conduct either chronic or acute
toxicity testing. The permitting authority decided to require
chronic testing but required the permittee to report the test
results at the 48-hour endpoint so that,acute toxicity could be
measured. One year before the permit was due to expire, the
permitting authority requested, under the authority of the Clean
Water Act (CWA) Section 308, that the permittee test his effluent
for toxicity to provide effluent information in order to write the
next NPDES permit. In this case, the permitting authority speci-
fied that the discharger submit quarterly chronic toxicity data for
1 year using the EPA toxicity tests for Selenastrum, Ceriodaphnia,
and Pimephales. The permitting authority also specified that up-
stream ambient water be used as the diluent in the tests so as to
allow the tests to measure additive effects from ambient toxics. In
response to the Section 308 request, the discharger submitted
the whole effluent toxicity data shown in Table 7-1.
Step 3: Determine Reasonable Potential for Excursions
The State interprets its narrative criteria for whole effluent toxicity
to require that the technical support document recommenda-
tions of 0.3 TUa and 1.0 TUC be used as numeric values for acute
and chronic toxicity, respectively. In accordance with the State
standards, the CMC applies under the 1Q10 flow and the CCC
applies under the 7Q10 flow.
The determination of exceedance of the CMC or the CCC was
simplified by the way in which the tests were conducted. Since
the upstream ambient water was used as a diluent, the test results
131
-------
already include an assessment of contributions from background
toxicity. Therefore, the upstream receiving water concentration
was set to zero.
The maximum effluent concentration was again estimated by
using the statistical approach in Chapter 3. As shown in Table 7-
1, there were four observations of whole effluent toxicity. Based
on the guidance of Box 3-4, these are insufficient to determine
the CV accurately; therefore, the default CV of 0.6 was used. The
effluent multiplier of 4.7 was obtained from Table 3-1 using the
number of observations, the CV, and the 99-percent probability
basis.
The receiving water concentration for chronic toxicity for com-
parison with the CCC was calculated using data from Table 7-1:
C= (4.7 x 20 TUC x 0.034 cfs) + (0 TUC x 13 cfs)]
(0.034 cfs +13 cfs)
« 0.25 TUC
where
13 cfs
0.034 cfs
4.7
20TUC
the receiving water flow at 7Q10
the mean effluent flow
the statistical effluent multiplier
the maximum effluent concentration.
The value of the calculated receiving water concentration, 0.25
TU& was less than the chronic water quality standard of 1.0 TUC,
and therefore there is no reasonable potential for the CCC to be
exceeded.
To calculate the receiving water concentration for acute toxicity,
the permitting authority first converted the chronic toxicity data
into equivalent acute toxicity units by applying the acute-to-
chronic ratio (ACR) of 5 obtained from the monitoring data. The
receiving water concentration for acute toxicity was then calcu-
lated:
C= [(4.7 x 20 TUC / 5 ACR x 0.034 cfs) + (0 TUC x 10.1 cfs)]
(0.034 cfs+ 10.1 cfs)
- 0.06 TUa
potential for excursions above the CMC or CCC for nickel, only
the WLA for human health was calculated.
To [determine WLAs, the numeric criteria in the water quality
standards and background concentrations were used to calcu-
late effluent concentrations that would result in compliance
with those standards. The calculation of WLAs used receiving
water flows that were appropriate to each standard: chronic
WLAs were calculated using the 7Q10 flow, acute WLAs were
calculated using the 1Q10 flow, and human health WLAs were
calculated using the harmonic mean flow. Since the effluent
was mixed rapidly by the multiport diffuser, the complete mix
equation was used:
WLA = [WQC x (Qe + Qs) - QsCs]/Qe
where
Qe = the effluent flow
Qs = the receiving water flow
Q = the background receiving water concentration
WQC = the water quality criterion.
The chronic and acute WLA for copper were calculated at the
7Q1 0 and 1 Q1 0 flows, respectively:
= [1 7.1 ug/l x (0.034 cfs + 1 3 cfs) - 1 3 cfs x
4.8 ug/l] / 0.034 cfs
= 4,720 ug/l
WLAa = [25.7 ug/l x (0.034 cfs + 1 0.1 cfs) - 1 0.1 cfs x
4.8 ug/l] / 0.034 cfs
i = 6,234 ug/l.
The human health WLA for nickel was calculated at the harmonic
mean flow:
WL\h = [1 3.4 ug/l x (0.034 cfs + 38 cfs) - 38 cfs x
' 13.2 ug/l/ 0.034 cfs,
; = 237 ug/l. ,
where 10.1 cfs is the receiving water flow at 1Q10, 5 is the acute
to chronic ratio, and the other values are the same as above. The
calculated value of 0.06 TUa is below the criterion of 0.3 TUa;
therefore, there is no reasonable potential for the CMC to be
exceeded. Since there was no reasonable potential for exceedances
above either the acute or chronic criterion, permit limits were not
developed for whole effluent toxicity.
7.2,4 Determine Wasteload Allocations
The wasteload allocation (WLA) was used to determine the level
of effluent concentration that would comply with water quality
standards in the receiving waters. A WLA will only be determined
for those parameters that have a reasonable potential to cause
exceedances of water quality standards. Therefore, WLAs were
determined for copper and nickel. Since there was no reasonable
7.2,5 Develop Permit Limits
Permit limits were developed using a steady-state, two-value WLA
model as described in Box 5-2, Chapter 5. Values for constants
were obtained from Table 5-2, Chapter 5.
Step 1: Calculate LTA (note: this is Step 2 in Box 5-2)
| • .
The| chronic long-term average (LTA) for copper was calculated
using the following formula: ,
LTAJ; = WLA x exp [0.5 cr2 - z a]
| •= 4,720 ug/l x 0.440
; = 2,077 ug/l
where values of exp [ 0.5 o2 - z oo ] are presented in Table 5-1
(see Chapter 5). The CV of 0.8 was used, and following the
132
-------
guidance of Section 5.5.4, the z value for the 99th occurrence
probability was used.
The acute LTA for copper was calculated, again using the 99th
percentile occurrence probability values from Table 5-1 as the
multiplier:
LTAa = 6,234 ng/lx 0.249
= 1,552|ig/l.
The LTA for nickel human health permitting is considered to be
the same as the WLA because the 70-year averaging period is
used for human health evaluations (see Section 5.4.4). The LTA is
calculated as:
LTAh = WLAh
= 237ng/l.
Step 2: Determine the More Limiting LTA
The limiting LTA for each pollutant was the minimum of the
chronic, acute and human health LTA. The limiting LTA value was
used in the next step to calculate maximum daily limits and
average monthly limits. The limiting LTA for copper was found to
be the acute LTA (1,552 ug/l) and the limiting LTA for nickel was
found to be the human health LTA (237 uxj/l).
Step 3: Calculate Maximum Daily and Average Monthly Limits
The maximum daily limit (MDL) for copper was calculated using
the expression:
MDL = LTA x exp [z o - 0.5 a2]
= 1,552 ng/lx 4.01
= 6,224 u.g/1
where the appropriate value for exp [ z a'- 0.5 <52] was taken from
Table 5-2 using the row with the CV for copper (0.8) and the
column for the 99th percentile probability basis.
The average monthly limit (AML) for copper was calculated using
the expression:
AML = LTA x exp [ z on - 0.5an2]
= 1,552 ng/lx 1.75 , ,
= 2,716ng/l
With a CV of 0.6, four samples per month for sampling, and a
99th percentile used for the MDL, the factor is 1.64:
MDL = AML x 1.64
= 237 ng/l x 1.64
= 389 |ig/l.
7.2.6 Determining and Expressing the Controlling Effluent
Limits
The NPDES regulations require that effluent limits require treat-
ment characteristic of the appropriate treatment technology and
also achieve water quality standards. If water quality-based limits
are more stringent than BAT limits, then the water quality-based
limits become the basis for the effluent limits. Conversely, if the
treatment technology (BAT) limits are more stringent, then they
become the basis of the limits.
The comparison between the water quality-based and technol-
ogy-based effluent limits are shown below. The more stringent
limits are different for different pollutants: for nickel, water qual-
ity-based limits are more stringent whereas for copper, the BAT
limits are the more stringent.
Copper
Nickel
Water quality MDL
AML
BAT
Limit to use
MDL
AML
MDL
AML
6,224
2,716
3,380
2,070
3,380
2,070
389
237
3,980
2,380
389
237
In accordance with NPDES regulations, the effluent limits were
expressed in the permit as mass (pounds per day) by multiplying
the concentrations above by the effluent flow of 0.034 cfs and the
conversion factor of 5.394:
Copper
(Ib/d)
Nickel
(Ib/d)
MDL
AML
0.62
0.38
0.071
0.043
where the value for exp [ z an - 0.5 oy,2] was taken from Table 5-
2 and, for this case, the number of samples per month was four.
Following the recommendations in Section 5.5.4, the z value for
the 95th percentile probability basis was used.
The effluent limits for nickel were determined by using the recom-
mendations in Section 5.4.4, Chapters. The AML was considered
to be identical to the WLA^ whereas the MDL was calculated from
the AML by using the appropriate multiplier factor in Table, 5-3.
7.2.7 Comparing Different Limit Development Methods
Permit limits for copper also were developed using a Monte Carlo
simulation in order to compare the results to the permit limits
derived from the two-value, steady-state model. A Monte Carlo
simulation was used to generate receiving water concentrations
to determine the effluent LTA for each of the pollutants such that
the water criteria are achieved at the required frequency in the
water quality standards.
133
-------
Monte Cario simulation used the same completely mixed dilution
equation as was used for the steady state calculation:
where C is the receiving water concentration (in ug/l); Ce and Cs
are the effluent concentration and the background concentration
of the receiving water, respectively (in u.g/1); and Qe and Qs and
effluent and receiving water flows, respectively (in cfs). Effluent
flows were held constant at the mean effluent flow, and river
flows were read from a computer file containing 60 years of daily
flow data provided by the U.S. Geological Survey. The effluent
concentrations were characterized by a lognormally distributed
random variate. The random variate had a coefficient of variation
that matched the CV of the pollutant in the effluent.
The Monte Carlo simulation was run using 22,276 iterations.
Once 22,276 receiving water concentrations had been calcu-
lated, receiving water concentrations were sorted, highest first.
The 20th value (corresponding to the maximum concentration
expected for 1 day in 3 years) was compared with the appropriate
criterion. The 1-day in 3-year return frequency is recommended
by EPA for criteria (see Chapter 2). If this value was higher than
the criterion, the effluent LTA was reduced, and a new set of
22,276 numbers was generated. When the receiving water con-
centration of the 20th value was just under the water quality
criterion (and the 1 9th value was just over the same value), then
the LTA effluent concentration generating these results was suffi-
cient to achieve the water quality criterion; this LTA was then used
in permit limit determinations.
For chronic criteria, 4-day average concentrations were generated
by taking the 4-day running average of modeled daily concentra-
tions. The recurrence concentration was calculated in the same
way as the 1-day calculations described in the previous para-
graph. Calculations were not made for the human health criterion.
The permit limits were calculated according to the procedures
given in Box 5-3. Each LTA was multiplied by the 99th percentile
multiplier from Table 5-3 for the MDL, and by the 95th percentile
multiplier from Table 5-3 for the AML. For the AML, the same
number of samples were used for the steady state and Monte
Cario permit limits (n=4). Thus, the resulting permit limits are
directly comparable. The results of the Monte Carlo simulation for
copper compared to the steady state calculations in units of
mtcrograms/liter are shown below:
Maximum
Daily
Average
Monthly
Monte Cario
Steady State
8,618
6,224
3,761
2,716
7.3 CASE 2: POTW DISCHARGE
The second example is of a fictitious POTW that discharges to the
same reach as the Jaybird Corporation. The NPDES permit for this
facility also is up for reissuance. The example highlights the use of
background receiving water concentrations, and demonstrates
the differences between industrial and POTW permit limits. In
developing permit limits for the POTW in this example, the
potential impacts from the Jaybird Corporation discharge were
considered in the use of background receiving water concentra-
tions. The interrelationships between the two facilities are dis-
cussed explicitly in Section 7.4.
7.3.1 General Site Description ant Information
The Locapunct River receives discharges from a POTW serving the
city of Auburn, a small city of about 10,000 people. The POTW
treats a mixture of household and industrial waste with an acti-
vated sludge process. The mean effluent flow from the POTW is
1.23 cfs. The POTW has no pretreatment program, but the
municipality generally is aware of the small industries that are
indirect dischargers because of research conducted by a local
university. Generally, the plant is well operated.
7.JL
2 Effluent Characterization for Specific Chemicals
Thb permitting authority's approach for determining which pol-
lutants cause, have the reasonable potential to cause, or contrib-
ute to excursions above water quality standards applies to POTWs
as well as industries. The authority used the procedures described
for the Jaybird Corporation in the evaluation of the Auburn POTW.
Stepl: Identify Pollutants of Concern
At the time of the last permit issuance, there was evidence of a
nufnber of toxic pollutants in the POTW's effluent, including
copper, chlorine, and ammonia. These pollutants had monitor-
ing requirements in the previous permit. Because there were
metals in the effluent and, due to the industries discharging into,
the POTW sewer system, the permitting authority requested the
POTW to conduct a complete priority pollutant scan of the efflu-
ent. The data received following the Section 308 letter request
indicated that the concentrations of all priority pollutants except
copper were below detection limits. The POTW's primary toxic
pollutants of concern were copper, chlorine, and ammonia (see
Table 7-2).
Step 2:
Determine RAC, CMC, or CCCfor Pollutants of Con-
cern
As described in the example of the industrial discharge, the water
quality standards include numeric criteria for copper. The State
alsp has adopted a numeric criterion for ammonia that is a
function of the river 85th percentile pH and temperature; these
values are 8.25°C and 25°C, respectively. Finally, the State inter-
prets its narrative criterion of "no toxics in toxic amounts" to
require use of the federal water quality criteria in the absence of a
numeric state criterion. As a result, the permitting authority uses
the federal criteria for chlorine. The applicable water quality
criteria for the river are as follows:
CCC
(ng/D
CMC
(Mg/i)
Copper
Chlorine
Ammonia
17.1
11
540
25.7
19
4,000
134
-------
Table 7-2. Effluent Data for the Auburn POTW
n
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Mean
SD
CV
Max
Min
Copper
M9/I
268
115
228
59
53 ,
213
68
200
262
519
53
474
115
259
404
57
101
187
103
76
198
265
60
112
185
133
0.7
519
52.6
Chlorine
M9/I
185
301
881
372
245
244
123
343
153
448
1,022
347
130
128
271
451
701
582
178
436
347
475
153
268
366
235
0.6
1,022
123
Ammonia
ng/i
11,009
1 3,025
12,201
24,548
9,700
15,645
21,358
3,976
22,307
7,427
11,834
8,430
4,382
9,330
6,1 37
6,448
37,772
14,307
16,848
28,205
12,119
11,778
3,109
4,474
13,182
8,491
0.6
37,772
3,109
Toxicity
TUC
2
1
1
2
1:5
0.6
0.4
.,- 2 '
1
Source: DMR data for chemicals; 308 request for whole effluent toxicty.
Notes: • -
Metals as total recoverable; toxicity in toxic units (100/NOEC).
The results are the highest toxic units for any of the test organisms used.
Step 3: Determine Dilution for Aquatic Life and Human Health
Impacts
The State water quality standards requires that compliance with
water quality criteria be achieved at the edge of the mixing zone.
The standards specify the minimum dilution at which the criteria
apply. These are the 7Q10 flow for the CGC, the 1Q10 flow for
the CMC, and the harmonic mean flow for human health criteria
(RAC). The U.S. Geological Survey operates a gaging station on
the river. The flow statistics were calculated-using the data from
this station:
• Harmonic mean flow = 38.0 cfs
• 7Q10 flow =13.0 cfs
• 1Q10flow=10.1 cfs.
The POTW is located at a bend of the river where mixing is rapid.
Therefore, the permitting authority used the complete mixing
equation to calculate the receiving water concentrations. This is
the same equation used for the industrial example.
Step 4: Determine Reasonable Potential for Excursions
The determination of possible exceedances in the CMC or CCC
was based on a calculation of the maximum receiving water con-
centration of each pollutant, followed by a comparison to the
appropriate receiving water criterion. The calculation of the maxi-
mum receiving water concentrations were made using the statisti-
cal estimate of the 99th. percentile concentration of each pollutant
in the effluent, the same flow used in the industrial example, and
considered background receiving water concentrations of:
Copper
Chlorine
Ammonia
4.8 ug/l
0 ug/l
120ng/l.
The maximum effluent concentration was estimated using the
statistical approach in Chapter 3. There were 24 concentrations
of each chemical reported in the DMRs. For copper, these
concentrations had a maximum value of 519 u.g/1, an arithmetic
mean of 185 u.g/1, an arithmetic standard deviation of 133, and
an arithmetic coefficient of variation of 133/185, or 0.7. The
multiplier was calculated to be 2.4 based on the CV of 0.7, 24
observations, and a 99-percent confidence level (see Section 3.3.2).
Similar calculations were conducted for chlorine (multiplier of
2.2) and ammonia (multiplier of 2.2).
The receiving water concentrations for each pollutant were calcu-
lated. An example calculation for the comparison of copper to
the CCC is shown below:
C= [(2.4 x 519 u.g/1 x 1.23 cfsl + (4.8 u.g/1 x 13 cfs)]
(1.23 cfs + 13 cfs)
= 112 ug/l
where
519 u.g/1 = the maximum measured effluent concentration
2.4 = the statistical nriultiplier
1.23 cfs = the average effluent flow
4.8 ug/l = the upstream receiving water concentration
13 cfs = theYOJOflow.
The maximum receiving water concentrations for comparison to
applicable standards for all pollutants were calculated to be:
Receiving Water
Criterion Concentration
(ug/l) (ug/l)
Copper
CCC
CMC
Chlorine
CCC
CMC
Ammonia
CCC
CMC
17.1
25.7
11
19
• •
540
4,000
112
140
194
244
7,292
9,128
135
-------
The effluent characterization showed the reasonable potential for
excursions above the CCC and CMC for copper, chlorine, and
ammonia. Therefore, permit limits were developed for these
pollutants.
7.3.3 Effluent Characterization for Whole Effluent Toxicity
wiere
13cfs =
1.23 cfs =
4.7
4TUr =
the receiving water flow at 7Q10
the mean effluent flow
the statistical effluent multiplier
the maximum effluent concentration.
Step 1: Dilution Determination
The Initial dilution determination was used to establish the types
of toxicity tests that must be conducted to characterize the efflu-
ent The dilution at the low flow characteristics for the facility is
the following:
At the 7Q10, dilution = (1.23 cfs + 13 cfs)/1.23 cfs
= 11.6
At the 1Q10, dilution = (1.23 cfs + 10.1 cfs)/1.23 cfs
= 9.2.
Step 2: Conduct Toxicity Testing
EPA recommends that a discharger having a dilution less than 100
be required to conduct chronic testing. The permitting authority
requested through a Section 308 letter that the POTW provide
quarterly chronic toxicity data for the year prior to permit
reissuance. Tests using Selenastrum, Ceriodaphnia, and Pimephales
were required. The permitting authority also required the permit-
tee to report the test results at the 48-hour endpoint so that acute
toxicity also could be measured. Table 7-2 summarizes the results
of the whole effluent toxicity testing.
Step 3: Determine Reasonable Potential for Excursions
As explained in the industrial example, the State interprets its
narrative criteria for whole effluent toxicity to require that the
technical support document recommendations of 0.3 TUa and 1.0
TUC be used as numeric values for acute and chronic toxicity,
respectively. In accordance with the State standards, the CMC
applies under the 1Q10 flow and the CCC applies under the
7Q10flow.
The reasonable potential determination of exceedance of the
CMC or the CCC was conducted in the same way as described in
the industrial example. Upstream ambient water was used as a
diluent to assess contributions directly from background toxicity;
therefore, the upstream receiving water concentration was set to
zero. The maximum effluent concentration was again estimated
by using the statistical approach in Chapter 3. For the same
reasons as were expressed in the industrial example, a multiplier
of 4.7 was used.
The receiving water concentration for chronic toxicity for com-
parison with the CCC was calculated using data from Table 7-2:
C= (4.7x2TUcx1.23cfs) + (OTUcx13cfs)
(1.23 cfs+13 cfs)
« 0.8 TUC
The value of the calculated receiving water concentration, 0.8
TUc, is less than the chronic water quality standard of 1.0 TUC, and
therefore there is no reasonable potential for the CCC to be
exceeded.
Tp calculate the receiving water concentration for acute toxicity,
the permitting authority first converted the chronic toxicity data
into equivalent acute toxicity units by applying the ACR of 2
obtained from the monitoring data. The receiving water concen-
tration for acute toxicity was then calculated:
C['= [(4.7 x 2 TUC / 2 ACR x 1.23 cfs) + (0 TUC x 10.1 cfs)]
| (1.23 cfs+ 10.1 cfs)
:= 0.5 TUa
F
where 10.1 cfs is the receiving water flow at 1Q10, 2 is the acute
to chronic ratio, and the other values are the same as above. The
calculated value of 0.5 TUa is greater than the criterion of 0.3 TUa-
Therefore, there is reasonable potential for the CMC to be ex-
ceeded and permit limits were developed for whole effluent
toxicity.
i
7.3.4 Determine Wasteload Allocations
WLAs for chemicals and whole effluent toxicity were determined
using information on the available dilution at the edge of the
mixing zone. The calculation of WLA using the steady-state model
was described in Section 7.2.4. The WLAs for the POTW using the
equation discussed in Section 7.2.4 are:
! WLAa
WLAC
Toxicity
(TU)
2.8
11.6
Copper
CMS/0
197
147
Chlorine
(ng/i)
175
127
Ammonia
Oig/0
35,860
4,979
7*3.5 Develop Permit Limits
The permit limit development process described in Box 5-2,
Chapter 5 was applied to all pollutants. This process is identical to
thfat explained in Section 7.2.5 except that (1) the WLA for acute
toxicity needs to be expressed in equivalent chronic toxic units by
multiplying by the ACR of 2, and (2) daily sampling of chlorine is
required in the permit. The calculated LTA and permit limits are:
Toxicity Copper Chlorine Ammonia
TUC (ng/1) Gig/1) (ng/1)
LTAa
| LTAc
1 MDL
! AML
1.8
6.1
5.6
2.8
55.4
70.7
197
91
56.2
66.9
175
87
11,511
2,625
8,162
4,067
136
-------
7.3.6 Determining and Expressing the Controlling Effluent
Limit
The treatment technology for POTWs is secondary treatment and
is characterized by effluent limits for biochemical oxygen de-
mand, total suspended solids, and pH. There are no BAT limits for
toxics for POTWs, so there was no need to compare these water
quality-based limits with other limits to determine which were
more stringent.
The permitting authority decided to use acute toxicity tests rather
than chronic tests to measure compliance with the toxicity efflu-
ent limits. The appropriate effluent limits in terms of TUa were
calculated by dividing the above calculation for TUC by the ACR of
2 that was obtained from effluent monitoring.
In accordance with NPDES regulations, the effluent limits for
chemicals were expressed in the permit as mass (pounds per day)
by multiplying the concentrations above by the effluent flow of
1.23 cfs and the conversion factor of 5.394. Because there is no
equivalent mass based unit for toxicity, toxicity mass limits are
impractical under the regulation.
MDL
AMI
Toxicity
TUa
2.8
1.4
Copper
(Ib/d)
1.31
0.64
Chlorine
(Ib/d)
1.16
0.58
Ammonia
(Ib/d)
54.2
27.0
7.3.7 Comparing Different Limit Development Methods
Permit limits also were developed using a Monte Carlo simulation
to compare the results to the steady-state permit limits. A Monte
Carlo simulation was used to generate receiving water concentra-
tions for determining the appropriate LTA for each of the pollut-
ants. The methodology for the Monte Carlo simulation is presented
in Section 7.2.7. The results for this case are presented below.
MDLs in TUa and |xg/l
Toxicity Copper Chlorine Ammonia
the degradation in water quality resulting from the combined
discharges, the development of total maximum daily loads (TM DLs)
for the river reach before generating WLAs, and the allocation of
discharges to each discharger. The following example describes
the permit development process when two dischargers release
effluent into the same reach of a river. The dischargers are the
jaybird manufacturing plant described in Case 1 and the Auburn
POTW described in Case 2. These facilities discharge into the
Locapunct River, whose flow characteristics previously were de-
scribed.
7.4.7 Effluent Characterization
The major differences in the effluent characterization for one
facility and for multiple facilities is to identify those pollutants that
are common to more than one facility, and to determine whether
the combined discharges cause or are likely to cause water quality
standards excursions.
Step 1: Identify Pollutants of Concern
Based on the data in Form 2C, the DMRs from the Jaybird
Corporation and the data in the DMRs and Section 308 request
from the Auburn POTW, the permitting authority found two
contaminants common to both discharges: copper and whole
effluent toxicity. Lead and nickel were found to be a problem at
the Jaybird Corporation, but since there were no complicating
discharges from the POTW, it was dealt with as a pollutant only at
the metal finishing facility. Similarly, chlorine and ammonia were
discharged solely by the POTW, so it was not necessary to provide
effluent limits for the metal finishing facility for these chemicals.
Step 2: Determine the CMC and CCC for Pollutants of Con-
cern
The numerical standards adopted by the State already have been
presented. The relevant values for copper and whole effluent
toxicity are:
CCC
CMC
Monte Carlo
Steady State
3.9
2.8
264
197
249
175
9,657
8,162
Copper
Toxicity
1 7.1 ng/l
1 .0 TUC
25.7 ng/l
0.3 TUa
Monte Carlo
Steady State
AMLs In TUa and
Toxicity Copper Chlorine Ammonia
2.7
1.4
171
91
170
87
6,614
4,067
7.4 CASE 3: MULTIPLE DISCHARGES INTO THE SAME
REACH
Permit development for water quality-based toxics control has
been illustrated for two single dischargers. This process increases
in complexity in cases of multiple dischargers into a reach. The
development of permit limits for multiple dischargers is based on
Step 3: Determine Dilution for Aquatic Life and Human Health
Impacts
Since this example is concerned with potential excursions above
standards resulting from the collective discharge of two discharg-
ers, the calculation of dilution includes the combined effluent flow
from both facilities. The combined dilution can be characterized
by the complete mixing equation:
C = (CelQe1 + Ce2Qe2 + CsQs)/(Qe1 + Qe2 + Qs)
where
Q61 and Qe2 = the flows of the two facilities
Cei and Ce2 = the effluent concentrations of the two facilities
Cs = the upstream receiving water concentration
Qs = the receiving water flow.
137
-------
Step 4: Determine Reasonable Potential for Excursions
To determine if the CMC or CCC were exceeded as a result of the
combined discharges into the river, the receiving water concen-
tration of each pollutant was calculated and compared to the
appropriate criterion. The receiving water concentration calcula-
tion was based on the maximum value of the effluent concentra-
tions (obtained from effluent data and multiplied by the appropri-
ate statistical factor), average effluent flows, background receiving
water concentrations, and appropriate river flows. All this infor-
mation has been presented previously in the separate examples.
The following results were obtained:
Copper
CCC
CMC
Criterion
(ng/0
17.1
25.7
Receiving Water
Concentration
(ng/D
156
194
CCC
CMC
1.0
0.3
,0.57
0.45
These calculations demonstrated exceedances of the copper CCC
and CMC criteria and the toxicity CMC criterion. Permit limits
were required.
7.4.2 TMDis and WLAs
WLAs were calculated to develop permit limits. WLAs for each
discharger and chemical were based on calculated TMDLs, the
total load to the Locapunct River that would not result in water
quality standards exceedances. TMDLs are comprised of a load
allocation for nonpoint sources, WLAs for point sources, and, if
required by the State, a reserve capacity. TMDLs are further
described in Section 4.2, Chapter 4.
Stepl: Calculate TMDL
The first step in developing individual WLAs for the two discharg-
ers was to develop TMDLs for each pollutant of concern. TMDLs
were developed in the same way as an individual WLA with the
total load of a pollutant from the two dischargers being consid-
ered as a single discharge.
The calculation of TMDLs used the following formula:
where
WQS « the water quality standard
Qt = the combined flow of both effluents
Qs = the appropriate receiving water flow.
The acute copper TMDL was calculated by using the data pre-
sented in the previous two examples as:
TMDL = 25.7 p.g/1 x (0.034 cfs + 1.23 cfs + 10.1 cfs)
= 292ng-cfs/l
where
J25.7|ig/l
t 0.034 cfs and
! 1.23 cfs
i 10.1
= the CMC
= the average effluent flows
= thelQIO.
Similar calculations were made for chronic copper and acute
toxicity. A TMDL was not calculated for chronic toxicity because
the information presented in Chapter 1 indicates that chronic
toxicity does not demonstrate additivity. The results are summa-
rized below. ,
Total Maximum Daily Loads
Chronic Acute
Copper (ug-cfs/l) 244 292
Toxicity (rua-cfs/l) NA 3.4
Step 2: Develop WLAs
The State had adopted an approach into the water quality man-
agement plan that described how WLAs were to be calculated.
The approach required that existing upstream concentrations be
Used to determine the load allocation part of the TMDL and that
10 percent of the TMDL had to be reserved and unavailable for
allocation. The remainder of the TMDL could be apportioned to
point sources in the WLA.
The permitting authority decided to allocate the wasteloads based
on the proportion of the existing load of each parameter that was
attributed to each of the existing discharges. Based on the
information shown in Tables 7-1 and 7-2 and the average effluent
flows, the pollutant loads from each facility are shown below.
Auburn POTW
Jaybird Corporation
Parameter
Copper
Oig-cfs/l)
Toxicity
(TUa-cfs)
Load
227.6
1.23
Proportion
0.77
0.90
Load
66.1
0.14
Proportion
0.23
0.10
Individual WLAs were then determined using the following
equation:
I WLA = (TMDL - LA -10% TMDL) x proportion/Qe
I
where the chronic TMDL was used to determine the chronic WLA,
and the acute TMDL was used to determine the acute WLA for
each facility. The WLAs for each pollutant and for each facility are
presented on the following page.
138
-------
Acute WLA
Chronic WLA
jaybird
Copper (|ig/l)
Toxicity (TUa)
134
2.2
1,450
9.0
98.4
NA
1,063
NA
7.4.3 Permit Limit Development
Once the WLAs had been determined, permit limit development
proceeded as in the previous examples. LTAs were calculated
from the WLAs, and the limiting LTA was selected for calculating
permit limits. For the metal finisher, where BAT limits were more
restrictive than the water quality-based limits, the BAT limits
applied. For the POTW, permit limits for toxic materials were
required only to prevent exceedances of water quality standards.
This process is summarized below.
Step 1: Calculate LTAs
The LTA was calculated for each discharger and pollutant as
described in Step 2, Box 5-2, Chapter 5; the LTAs are shown
below.
Parameter
Acute LTA
POTW jaybird
Chronic LTA
POTW jaybird
Copper (|ig/l)
Toxicity (TUa)
37.7
0.71
361
2.9
47.3
NA
468
NA
Step 2: Determine the More Lmiting LTA
The minimum LTA was used to calculate MDLs and AMLs. The
acute LTA was the lower LTA for both pollutants.
Step 3: Calculate the Maximum Daily and Average Monthly
Limits
The MDL and AML were calculated as described in Box 5-2,
Chapters:
Average Monthly Limit Maximum Daily Limit
Parameter POTW jaybird POTW jaybird
Copper (jj.g/1)
Toxicity (TUa)
62
1.1
632
4.5
134
2.2
1,448
9.0
Step 4: Express the Limits
The final step is to compare the water quality-based limits to the
BAT limits to ensure that the more restrictive of the two are used,
and to express the copper limits in terms of mass. The copper
water quality-based limits for Jaybird Corporation are lower than
the BAT ones (see Section 7.2.6). Therefore, the water quality-
based limits are required by the permit. In addition, the limits are
lower than those calculated when only one of the facilities were
considered. The final permit limits are listed below.
Parameter
Average Monthly Limit
POTW jaybird
Maximum Daily Limit
POTW jaybird
Copper (Ib/d)
Toxicity CTUa)
0.41
1.1
0.12
4.5
0.89
2.2
0.27
9.0
139
-------
-------
INDEX
Acute toxicity endpoints (ATEs)
lethal concentration (LC) 4
Acute toxicity testing 59
Acute-chronic ratio (ACR) 17
Additivity 24
Allowable Effluent Concentration Distribution 82
Ambient toxicity testing 61
Ambient-induced mixing 77
lateral dispersion coefficient 77
shear velocity 77
Amelia River 68
Ames test 25
Analytical considerations for chemicals 65
Antidegradation policy 29
Aquatic community 18
Aquatic Life Protection 34
ARM 84
BAF 38
Bioaccumulation 37, 38, 72
Bioaccumulation consideration 38
Bioconcentration 38, 64
Biocriteria 41,42
Biological assessment/bioassessment 18, 20
Biological criteria
biological integrity 1,18
Biological survey/biosurvey 18,19
Calculations
CCC for toxicity 85
CMC for toxicity 85
Concentration (multiple dischargers) 86
Concentration (nonconservative pollutant) 87
Harmonic mean flow 88
Lateral dispersion coefficient 77
Carcinogenicity 25
Carcinogens 68
Carcinogens, calculating RACs 40
CCC 34,85
CCC. See Criterion continuous concentration 79, 85
Center for Exposure Assessment Modeling (CEAM) 73
Chronic toxicity endpoints
effective concentration 4
lowest observed effect concentration 4
no observed effect concentration 4
Chronic toxicity testing 59
Clean Water Act (CWA) 1, 67
CMC 34, 71,85
CMC. 71
Coefficient of variation (CV) 95
in permit limit derivation 105
Completely mixed discharge-receiving water situations 78
Complex Effluent Toxicity Testing Program (CETTP) 7, 9
Compliance monitoring 123
discharge monitoring report/quality assurance 124
inspections 124
self-monitoring report 123
Compliance problems 53
Concentration
flow distance 77
Continuous Simulation 80
CORMIX1 76
Criterion continuous concentration (CCC) 48
Criterion maximum concentration (CMC) 48, 71, 85
Critical conditions 67
Critical design periods
estuaries and coastal bays 74
oceans 74
rivers and run-of-river reservoirs 73
lakes, reservoirs 73
Design periods. See Critical design periods
Designated use 54
Detection level
minimum level 111
practical quantitation limit 112
Detection levels 111
Determining the need for a limit
statistical approach 56
when is a chemical-specific limit sufficient? 62
with effluent data 55
without effluent data 50
Dilution 52-53,61
Dilution determination 58, 63
Discharge monitoring reports 54
Discharge-Induced Mixing 75
Duration 31-32
Duration for single chemicals and WET 35
Dye study 51,58
Dynamic modeling. See Modeling
Modeling techniques; Models 76, 78, 79
DYNHYD4 89
DYNTOX 83, 84
Effluent bioconcentration evaluation 42, 64-65
Effluent characterization 47
addressing uncertainty in 56
for aquatic life effects 48
for human health effects 48, 62
or multiple dischargers 60-61
for specific chemicals 61 -63
for whole effluent toxicity 56-60
process 53, 63
141
-------
Effluent characterization (cont.)
purpose 48
special considerations for marine and estuarine
systems 61
Effluent variability
basic principles 93
Enforcement 125
enforcement discretion 125
enforcement mechanisms 125
whole effluent toxicity enforcement 124,125
EXAMS-II 84
Excursion above CMC or CCC 60
Excursions above ambient criteria 50
Expressing limitations and developing monitoring require
ments 110
Expressing permit limits
mass-based limit 110
maximum daily 96
FCM2 85
FGETS 85
Fish consumption values 37
Fishable/swimmable 29
FLOSTAT 89
Food chain 87
Frequency 31, 32
Frequency for single chemicals and WET 36
GC/MS 64
Generating Effluent Data 56
Genotoxic pollutants 25
Health effects 25
nonthreshold effects 25
threshold effects 25
HPLC 65
HSPF 84
Human exposure
Background concentrations 37
drinking water ingestion 37
fish consumption 37
Human health criteria
Q1* 39,40
RfD 39, 40
updating 38
Human health protection 25
Human health protection (WQC) 36
Human health/Human exposure 37
Implementation methods for state antidegradation policies
Tier I 29
Tier II 30
Tier 111 (ONRWs) 30
Independent application 22, 31,49
Integrated approach 1
bioassessment approach 1, 22
chemical-specific approach 1, 20
whole effluent approach 1, 4, 21
Integrated Risk Information System (IRIS) 38
Isopleths, concentration 72
LA. See Load allocation 67 .
Lateral dispersion coefficient 77
LC50 4, 57, 58, 71
Limits for metals 111
Load allocation 67
Lognormal Probabilistic Dilution Model 82
Long term average (LTA) 95
-
Magnitude 31, 32
Magnitude for single chemicals 34
Magnitude for whole-effluent toxicity 35
Magnitude, duration, and frequency (Criteria) 31, 32
Margin of safety 67
Marine and estuarine discharges 61,
Marine and estuarine permitting 104
Maximum daily permit limits
chronic toxicity 96
Metals 111
MJNTEQA2 84,85
Mixing zone 58, 72
Mixing Zones 33
Modeling techniques
continuous simulation 81
lognormal probabilistic dilution model 82
Monte Carlo simulation 81
steady state 78
Models
CORMIX1 76
DYNHYD4 84
DYNTOX 84
EXAMS-II 84
FCM2 85
FGETS 85
FLOSTAT 89
HSPF 84
MINTEQA2 85
Mixing zone 70
PSY 78
SARAH2 85
Selection of 83
STORET 79
TOXI4 84
UDKHDEN 77
ULINE 77
UMERGE 77
UOUTPLM 77
UPLUME 76, 77
WASP4 84
142
-------
Monte Carlo simulation 80, 81
Monticello Ecological Research Station (MERS) 2
Multiple source toxicity testing procedures 60-61
Multiple-source discharge 60-61
Narrative criteria 29
Noncarcinogens 89
NPDES program 1
Number of samples
in permit limit derivation 105
Numeric criteria 29
Outfall design recommendations
in lakes and reservoirs 73
in oceans 74
in rivers 73
multiport submerged 73
single-port submerged 73
surface discharge 73
Outstanding national resource waters (ONRWs) 30
Penalties 124
Permit documentation 110
Permit limit derivation 98
aquatic life protection 98
average monthly 96
average weekly 96
basic objective 96
detection levels 111
direct application of both the acute and
chronic WLAs 104
dynamic 101
human health protection 104-105
maximum daily 96
other approaches 103
selection of monitoring frequencies 113
single value steady state 102
two value steady state 98
use of a WLA as a permit limit 96-97
Pollution prevention 111
Potential for excursion above CMC or CCC 58
POTW 53
Precision 2,11,12
coefficient of variation (CV) 12
inter-laboratory precision 2,11
intra-laboratory precision 11
variability 11
Priority toxic pollutants 30
Probability basis
average monthly 110
daily maximum 110
in permit limit derivation 110
LTAforWLA 110
Quality assurance 12,123
discharge monitoring report/quality assurance 124
Reasonable potential 48, 49, 50, 58
as a trigger 58
for multiple dischargers 60
for whole effluent toxicity 58
regulatory basis 49
with effluent data 50, 56
without effluent data 49 i
Receiving water concentration (RWC) 48 '
Reference ambient concentration (RAC) 48
Reference toxicants 12
Regulation requirements 48-49
Reporting violations 126
to the public 126
to the regulatory agency 124
Return period 82
Rhodamine WT. See Tracer (dye) studies 75
SARAH2 85
Screening protocol 53
Sediment 42, 67
Sediment criteria 42
Shayler Run, Ohio 2
Single value wasteload allocations
use in permit limit derivation 102
Species sensitivity 59
Species sensitivity differences 16
Statistical considerations of effluent limits 105
changes in CV on limits 106
changes in CV on LTA 105
changes in monthly samples 105,107
changes in probability basis on limits 105,110
Statistical considerations of WLAs 96
Statistical distributions of effluent data
lognormal (positively skewed) 95
normal (bell shaped) 95
STORET 79
TMDL. See Total maximum daily load
Total maximum daily load
calculation of 78
margin of safety 67
Total maximum daily load (TMDL) 1,67,78
TOXI4 84
Toxic units (TUs) 6
acute toxicity units (TUa) 7
chronic toxicity units (TUC) 7
Toxicity
acute toxicity 4
CCC for 85
chronic toxicity 4
CMC for 85
whole effluent toxicity 4, 71
143
-------
Toxfcity persistence 23
Toxicity reduction evaluations
308 letters 118-119
additional testing 117-118
administrative orders 114
approach 114
circumstances warranting a TRE 117
consent decrees 114
evaluation criteria 115
guidance documents 114
requiring TREs 118
toxicity identification evaluation (TIE) 116
TRE plans 115
Toxicily test endpoint concentrations
acute toxicity endpoints (ATEs) 4
chronic toxicity endpoints (CTEs) 4
Toxicity testing 11
composite sample 13
flow-through toxicity test 13
grab sample 13
off-site tests 11
on-site tests 11
static toxicity test 13
Toxicity testing procedures 58
Treatment plant performance 97
Triggers for permit limit development 58, 63-64
Tualatin River Basin 68
UDKHDEN 77
ULINE 77
UMERGE 77
UOUTPLM 77
UPLUME 76, 77
Variability
effluent variability 16
exposure variability 16
species sensitivity differences 16
Violation review 124
WASP4 84
Wasteload 67
Wasteload allocation
schemes 67, 69
Wasteload allocation (WLA) 1
Water quality criteria 1,29, 32
aquatic life 34
criterion continuous concentration (CCC) 32
criterion maximum concentration (CMC) 34, 35
RACs for non-carcinogens 39
reference ambient concentration (RAC) 36
reference tissue concentration (RTCs) 37
Water quality models
continuous simulation 81,98
dynamic 98
1 lognormal probabilistic 82
' Monte Carlo 81, 98
single value steady state 97
\ two value steady state 98
Water quality standards 1
biological criteria 1
' narrative water quality criterion 1
; numeric criteria 1
water quality criteria 1
Water quality standards regulatory considerations 30
Section 303(c)(2)(B) of CWA 30
I 40 CFR 122.44 & 40 CFR Part 131 31
Section 307(a) of CWA 30
| Section 131.11 Standards Regulation 2.1 29
When is a chemical-specific limit sufficient? 62
Whole effluent approach 4
Whole effluent toxicity
j "pass/fail" tests 104
[ acute endpoint sensitivity 112
chronic endpoint sensitivity 112
description of limits 112
detection levels 112
end-of-pipe approach 104
test methods 112
use of acute versus chronic tests 112
Whole effluent toxicity data generation 58
WlA 67
* This is not a comprehensive index. Only topics of importance
are highlighted.
144
-------
APPENDICES
A. Raw Data
B. EPA Regulations and Policies
C. Ambient Toxicity Testing and Data Analysis
D. Duration and Frequency
E. Lognormal Distribution and Permit Limit Derivations
F. Sampling
G. The Development of A Biological Indicator Approach to Water Quality-based Human Health Toxics Control
H. Reference Dose (RFD): Description and Use in Health Risk Assessment
1. Chemicals Available on IRIS
145
-------
APPENDIX A-l
TOXICITY TEST PRECISION DATA
-------
MARINE/ESTUARINE SHORT-TERM CHRONIC TOXBCITY
TESTS
-------
SHEEPSHEAD MINNOW (Cyprinodon variegatus)
Seven-day Larval Survival and Growth Test
Single Laboratory Precision Data
Table A-1-1. Single laboratory precision of test performed in
40 fathoms artificial seawater, using larvae from fish
maintained and spawned in 40 fathoms artificial seawater,
using copper as the reference toxicant [1].
Test
Number
1
2
3
4
5
6
7
8
n:
Mean:
CV(%):
NOEC
(mg/l)
0.05
<0.05*
<0.05*
0.05
<0.05*
0.05
0.05
0.05
5
0.05
NA
ICzs
(mg/l)
0.1133
0.0543
0.0418
0.0632
0.0577
0.0483
0.0796
0.1235
8
0.0727
41.82
ICso
(mg/l)
0.1523
0.0975
0.0714
0.0908
0.0998
0.1 325
0.1597
0.2364
8
0.1300
40.77
Most
Sensitive
Endpoint
S
G
G
S
S
G
G
G
* The lowest concentration tested was 0.05 mg./l
NOEC Range: >0.05* - 0.05 mg/l.
Copper concentrations in Tests 1-6 were 0.050, 0.10, 0.20, 0.40, and 0.80 mg/l
and Tests 7-8 were 0.025, 0.050, 0.10, 0.20, and 0.40 mg/l.
Prepared by Florence Kessler, TAI, Cincinnati, OH, January 11, 1990 (ICp
Program, version 1.1b).
Table A-1-2. Single laboratory precision of test performed in
40 fathoms artificial seawater, using larvae from fish
maintained and spawned in 40 fathoms artificial seawater,
using sodium dodecyl sulfate (SDS) as the reference toxicant
[1].
Test
Number
1
2
3
4
5
6
n:
Mean:
CV(%):
NOEC
(mg/l)
1.0
1.0
1.0
0.5
1.0
0.5
6
0.8
NA
IC2S
(mg/l)
1.2799
1 .4087
2.3051
1 .9855
1.1901
1.1041
6
1 .5456
31.44
ICso
(mg/l)
1 .5598
1 .8835
2.8367
2.6237
1 .4267
1 .4264
6
1.9595
31.82
Most
Sensitive
Endpoint
S
S
S
G
S
G
NOEC Range: 0.5 - 1.0 mg/l (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/l and in
Tests 3-6 were 0.20, 0.50, 1.0, 1.9, and 3.9 mg/l.
Prepared by Florence Kessler, TAI, Cincinnati, OH, January 11, 1990 (ICp
Program, version Lib).
Table A-1-3. Single laboratory precision of test performed in
natural seawater, using larvae from fish maintained and
spawned in natural seawater, using copper as the reference
toxicant [1].
Test
Number
1
2
3
4
5
n:
Mean:
CV(%):
NOEC
(mg/D
125
31
125
125
125
5
106.2
NA
ICzs
(mg/l)
320.3
182.3
333.4
228.4
437.5
5
300.4
33.0
ICso
(mg/l)
437.5
323.0
483.4
343.8
NC*
4
396.9
19.2
Most
Sensitive
Endpoint
S
G
S
S
S
* No linear interpolation estimate could be calculated from the data, since none
of the group response means were less than SO percent of the control response
mean.
NOEC Range: 31 -125 mg/l (this represents a difference of two exposure
concentrations).
Prepared by Elise Torello, SAIC, Narragansett, Rl, and Margarete Heber, EPA,
Washington, DC, February 1990 (ICp Program, version Lib).
Table A-1-4. Single laboratory precision of test performed in
natural seawater, using larvae from fish maintained and
spawned in natural seawater, using sodium dodecyl sulfate
(SDS) as the reference toxicant [1 ].
Test
Number
1
2
3
4
5
n:
Mean:
CV(%):
NOEC
(mg/l)
2.5
1.3
1.3
1.3
1.3
5
1.5
NA
(mg/l)
2.9
NC1
1.9
2.4
1.5
4
2.2
27.6
(mg/l)
3.6
NC2
2.4
NC2
1.8
3
2.6
35.3
Most
Sensitive
Endpoint
S
G
S
G
S
NOEC Range: 1.3 - 2.5 mg/l (this represents a difference of one exposure
concentration).
1 No linear interpolation estimate could be calculated from the data, since none
of the group response means were less than 75 percent of the control response
mean.
2No linear interpolation estimate could be calculated from the data, since none
of the group response means were less than 50 percent of the control response
mean.
Prepared by Elise Torello, SAIC, Narragansett, Rl, and Margarete Heber, EPA, and
Washington, DC, February 1990 (ICp Program, version Lib).
A-1-1
-------
SHEEPSHEAD MINNOW (Cyprinodon variegatus)
Seven-day Larval Survival and Growth Test
Interlaboratory Precision Data
Table A-1-5. Interlaboratory precision of test using an industrial effluent as
the reference toxicant [1 ].
Laboratory A
Laboratory B
Laboratory C
Laboratory D
n:
Mean:
CV(%):
Test
Number
1
2
1
2
1
1
2
Most Sensitive Endpoint
NOEC
(%)
3.2 (S,G)
3.2 (S,G)
j!
3.2 (S,G)
3.2 (S,G)
[.
1.0 (S)
i
3.2 (S,G)
1.0 (G)
7
2.6
NA
ICs
(%)
7.4 (S)
7.6 (S)
5.7 (G)
5.7 (G)
4.7 (S)
7.4 (G)
5.2 (S)
7
5.5
44.2
ICso
(%)
7.4 (G)
14.3 (G)
9.7 (G)
8.8 (G)
7.2 (S)
24.7 (G)
7.2 (S)
7
11.3
56.9
NOEC Range: 1.0 - 3.2 percent (this represents a difference of one exposure concentration).
Prepared by Elise Torello, SAIC, Narragansett, Rl, and Margarete Heber, EPA, and Washington, DC,
February (ICp Program, version 1.1b).
SHEEPSHEAD MINNOW (Cyprinodon variegatus)
Embryo-larval Survival and Teratogenicity Test
Single Laboratory Precision Data
Table A-1-6. Single laboratory precision of test performed in HW Marinemix
artificial seawater, using embryos from fish friaintained and spawned in HW
Marinemix artificial seawater, using copper as the reference toxicant [1].
Test
Number
1
2
3
4
5
6
n:
Mean:
CV(%):
EC,
(ug/i)
173
*
*
182
171
*
*
195
4
180
6.1
EC5
(ug/D
189
*
*
197
187
*
*
203
4
194
3.8
EC10
(ug/D
198
*
*
206
197
*
*
208
4
202
2.8
(ug/l)
234
*
*
240
234
*
*
226
4
233
2.5
NOEC
(ug/0
240
240
240
240
240
<200
220
220
7
234
NA
* Data do not fit the Probit model.
NOEC Range: 200 - 240 (this represents a difference of two exposure concentrations).
A-1-2
-------
SHEEPSHEAD MINNOW (Cyprinodon variegatus) (continued)
Embryo-larval Survival and Teratogenicity Test
Single Laboratory Precision Data
Table A-1-7. Single laboratory precision of 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 [1 ].
Test
Number
1
2
3
4
5
n:
Mean:
CV(%):
EC,
(mg/l)
1.7
*
0.4
1.9
1.3
4
1.3
51.2
EC5
(mg/l)
2.0
*
0.7
2.2
1.7
4
1.6
41.6
EC10
(mg/l)
2.2
*
0.9
2.4
1.9
4
1.9
35.0
EC50
(mg/l)
3.1
*
2.5
3.3
3.0
4
2.9
11.7
NOEC
(mg/l)
2.0
4.0
2.0
2.0
2.0
S
2.4
NA
* Data do not fit the Probit model.
NOEC Range: 2.0 - 4.0 ug/l (this represents a difference of one exposure concentration).
INLAND SILVERSIDE (Menidia beryllina)
Seven-day Larval Survival and Growth Test Single Laboratory Precision Data
Table A-1-8. Single laboratory precision of the inland
silverside (Menidia beryllina) 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 [1]-
Test
Number
1
2
3
4
5
n:
Mean:
CV(%):
NOEC
(ug/l)
63
125
63
125
31
5
81.4
NA
IC25
(ug/l)
96.2
207.2
218.9
177.5
350.1
5
209.9
43.7
ICso
(ug/l)
148.6
NC*
493.4
241.4
479.8
4
340.8
50.7
Most
Sensitive
Endpoint
S
S
G
S
G
* No linear interpolation estimate could be calculated from the data, since none
of the group response means were less than 50 percent of the control response
mean.
NOEC Range: 31 - 125 ug/l (this represents a difference of two exposure
concentrations).
Prepared by Elise Torello, SAIC, Narragansett, Rl, and Margarete Heber, EPA, and
Washington, DC, February 1990 (ICp Program, version 1.1 b).
Table A-1-9. Single laboratory precision of the inland
silverside (Menidia beryllina) larval survival and growth test
performed in natural seawater, using larvae from fish
maintained and spawned in natural seawater, using sodium
dodecyi sulfate (SDS) as the reference toxicant [1].
Test
Number
1
2
3
4
5
n:
Mean:
CV(%):
NOEC
(mg/l)
1.3
1.3
1.3
1.3
1.3
5
1.3
NA
IC25
(mg/l)
0.3
1.6
1.5
1.5
1.6
5
1.3
43.2
ICso
(mg/l)
1.7
1.9
1.9
1.9
2.2
5
1.9
9.4
Most
Sensitive
Endpoint
S
S
S
S
S
NOEC Range: 1.3 mg/l.
Prepared by Elise Torello, SAIC, Narragansett, Rl, and Margarete Heber, EPA, and
Washington, DC, February 1990 (ICp Program, version Lib).
A-1-3
-------
MYSID (Mysidcpsis bahia)
Seven-day Survival, Growth, and Fedundity Test Single
Laboratory Precision Data
Table A-1-10. Single laboratory precision of the mysid (Mysidopsis
bahid) 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 [1 ].
Test
Number
1
2
3
4
5
n:
Mean:
CV(%):
NOEC
(ug/l)
63
125
125
125
125
5
112.6
NA
IC25
(ug/D
96.1
138.3
156.3
143.0
157.7
5
138.3
18.0
ICso
(ug/l)
NC*
175.5
187.5
179.9
200.3
. 4
185.8
5.8
Most Sensitive
Endpoint
S
S
S
S
S
* No linear interpolation estimate could be calculated from the data, since none of the
group response means were less than 50 percent of the control response mean.
NOEC Range: 63-125 ug/l (this represents aidifference of two exposure concentrations).
Prepared by Elise Torello, SAIC, Narragansett, Rl, and Margarete Heber, EPA, and
Washington, DC, February 1990 (ICp Program] version 1.1 b).
Table A-1-11. Single laboratory precision of the mysid (Mysidopsis
bahid) 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 [1]. " • '
Test
Number
1
2
3
4
5
6
n:
Mean:
CV(%):
NOEC
(mg/l)
2.5
<0.3
<0.6
5.0
2.5
5.0
4
3.8
NA
IC25
(mg/l)
4.5
NC1
NC1
7.8
3.6
7.0
4
5.7
35.0
"Cso
(mg/l)
NC2
NC2
NC2
NC2
4.6
9.3
2
6.9
47.8
Most Sensitive
Endpoint
S
S
S
S
S
S
'No linear interpolation estimate could be calculated from the data, since none of the
group response means were less than 75 percent of the control response mean.
2No linear interpolation estimate could be calculated from the data, since none of the
group response means were less than 50 percent of the control response mean.
NOEC Range: <0.3 - 5.0 mg/l (this represents a difference of four exposure concentrations).
Prepared by Elise Torello, SAIC, Narragansett, Rl, and Margarete Heber, EPA, and
Washington, DC, February 1990 (ICp Program; version Lib).
A-1-4
-------
SEA URCHIN (Arbacia punctulata)
Fertilization test Single Laboratory Precision Data
Table A-1-12. 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 (40 Fathoms), using copper
as the reference toxicant [1].
Test
Number
1
2
3
4
5
n:
Mean:
CV(%):
NOEC
(ug/l)
5.0
12.5
<6.2
6.2
12.5
4
9.0
NA
IC25
(ug/l)
8.92
26.35
11.30
34.28
36.67
5
23.51
54.60
ICso
(ug/l)
29.07
38.96
23.93
61.75
75.14
5
45.77
47.87
NOEC Range: <5.0 - 12.5 ug/l (this represents a difference of one exposure
concentration).
Copper concentrations in Test 1 were 2.5, 5.0, 10.0, 20.0, and 40.0 ug/l and in
Tests 2-5 were 6.25, 12.5, 25.0, 50.0, and 100.0 ug/l.
Prepared by Florence Kessler, TAI, Cincinnati, OH, January 11, 1990 (ICp
Program, version Lib).
Table A-1-14. 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 [1].
Test
Number
1
2
3
4
5
n:
Mean:
CV(%):
NOEC
(ug/l)
12.2
12.2
24.4
<6.1
6.1
4
13.7
NA
IC25
(ug/0
14.2
32.4
30.3
26.2
11.2
5
22.8
41.9
"Cso
(ug/l)
18.4
50.8
46.3
34.1
17.2
5
29.9
48.2
NOEC Range: <6.1 - 24.4 ug/l (this represents a difference of two exposure
concentrations).
Prepared by Elise Torello, SAIC, Narragansett, Rl, and Margarete Heber, EPA, and
Washington, DC, February 1990 (ICp Program, version Lib).
Table A-1-13. 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 (40 Fathoms), using sodium
dodecyl sulfate (SDS) as the reference toxicant [1].
Test
Number
1
2
3
4
5
n:
Mean:
CV(%):
NOEC
(mg/0
<0.9
0.9
1.8
0.9
1.8
4
1.4
NA
IC2S
(mg/l)
1.11
1.27
2.26
, 1.90
2.11
5
1.73
29.7
ICso
(mg/l)
1.76
1.79
2.87
2.69
2.78
5
2.38
23.3
NOEC Range: 1.2 - 3.3 mg/l (this represents a difference of one exposure
concentration).
SDS concentrations for all tests were 0.9, 1.8, 3.6, 7.2, and 14.4 mg/l.
Prepared by Florence Kessler, TAI, Cincinnati, OH, January 11, 1990 (ICp
Program, version 1.1 b).
Table A-1-15. 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 [1 ].
Test
Number
1
2
3
4
5
n:
Mean:
CV(%):
NOEC
(ug/l)
1.8
1.8
1.8
0.9
1.8
5
1.6
NA
ICzs
(ug/0
2.3
3.9
2.3
2.1
2.3
5
2.58
28.7
ICso
(ug/D
2.7
5.1
2.9
2.6
2.7
5
3.2
33.3
NOEC Range: 0.9 -1.8 mg/l (this represents a difference of one exposure
concentration).
Prepared by Elise Torello, SAIC, Narragansett, Rl, and Margarete Heber, EPA, and
Washington, DC, February 1990 (ICp Program, version Lib).
A-1-5
-------
RED MACROALGAE (Champia parvula)
Reproduction Test Single Laboratory Precision Data
Table A-1-16. Single laboratory precision of the red
macroalga (Champia parvula) reproduction test performed in
SO/50 natural seawater and CP-2 artificial seawater. Copper is
the reference toxicant [1 ].
Test
Number
1
2
3
4
5
6
n:
Mean:
CV(%):
NOEC
(ug/D
1.0
1.0
1.0
1.0
0.5
0.5
6
0.83
NA
ICa
(ug/l)
1.67
1.50
0.69
0.98
0.38
0.38
6
0.93
59.6
"Cso
(ug/l)
2.35
1.99
1.53
1.78
0.76
0.75
6
1.5
43.7
Table A-1-17. 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 [1] (personal
communication with G. Thursby, SAIC, Narragansett, Rl).
NOEC Range: 0.5 - 1.0 ug/l (this represents a difference of one exposure
concentration).
Prepared by Elise Torello, SAIC, Narragansett, Rl, and Margarete Heber, EPA, and
Washington, DC, February 1990 (ICp Program, version Lib).
Test
Number
1
2
3
4
5
6
7
8
9
n:
Mean:
CV(%):
NOEC
(mg/l)
<0.80
0.48
<0.48
<0.48
0.26
0.09
0.16
0.09
<0.29
5
0.22
NA
IC25
(mg/l)
0.6
0.7
0.4
0.2
0.2
0.1
0.2
0.1
0.3
9
0.31
69.0
"Cso
(mg/l)
0.3
0.6
0.2
0.4
0.5
0.3
0.3
0.2
0.4
9
0.36
37.0
NOEC Range: 0.09 - 0.48 mg/l (this represents a difference of two exposure
concentrations).
Prepared by Elise Torello, SAIC, Narragansett, Rl, and Margarete Heber, EPA, and
,Washington, DC, February 1990 (ICp Program, version Lib).
Table A-l-18. Single laboratory precision testing of the red macroalga (Champia parvula) reproduction test in natural seawater (30
°/oo salinity). The reference toxicants used were copper sulfate (Cu)V2 and sodium dodecyl sulfate (SDS)2'3 [7].
Test
1
2
3
4
n:
Mean:
CV(%):
Cu (ug/l)
NOEC
1.00
0.50
0.50
0.50
4
0.63
NA
IC25
2.62
0.71
2.83
0.99
4
1.79
61.09
ICso
4.02
1.66
3.55
4.15
4
' 3.35
34.45
SDS (mg/l)
NOEC
0.60
0.60
0.30
0.15
4
0.41
NA
IC2J
0.05
0.48
0.69
0.60
4
0.46
62.29
ICso
0.50
0.81
0.89
0.81
4
0.75
22.92
,
'Copper concentrations were O.S, 1.0, 2.5, 5.0, 7.5, and 10 ug/l. Concentrations of Cu were made from a 100 ug/ml CuSO4 standard obtained from Inorganic Ventures,
Inc., Brick, N).
^All tests were conducted at 23 ± 1°C in natural seawater with irradiance set at 40 uE/m2/s.
concentrations were 0.0375, 0.075, 0.15, 0.30, 0.60, and 1 .20 mg/l. Concentrations of SDS were made from a 44.64 ± 3.33 mg/ml standard obtained from U.S. EPA-
EMSl, Cincinnati, OH.
Prepared by Steven H. Ward and Glen Thursby, EPA, Narragansett, Rl (ICp Program, version Lib).
A-1-6
-------
FRESH WATER SHORT-TERM CHRONIC TOXICITY TESTS
-------
FATHEAD MINNOW (Pimephales promelas)
Seven-day Larval Survival and Growth Test and
Embryo-larval Survival and Teratogenicity Test
Single Laboratory Precision Data
Table A-l-19. Single laboratory precision of the fathead
minnow (Pimephales promelas) embryo-larval survival and
teratogenicity test performed in using Diquat as the reference
toxicant [2].
Test
Number
1
2
3
4
5
n:
Mean:
CV(%):
L.CT
(mg/l)
0.58
2.31
1.50
1.71
1.43
5
1.51
41.3
Table A-1-20. Single laboratory precision of the fathead
minnow (Pimephales promelas) embryo-larval survival and
teratogenicity test performed in using cadmium chloride as the
reference toxicant [2].
Test
Number
1
2
3
4
5
n:
Mean:
CV(%):
LC,
(mg/0
0.014
0.006
0.005
0.003
0.006
5
0.0068
62
NOEC
(mg/l)
0.012
0.012
0.013
0.011
0.012
5
0.012
NA
NOEC Range: 0.011 - 0.013 mg/l (this represents a difference of one exposure
concentration).
A-1-7
-------
FATHEAD MINNOW (Pimephales promelas)
Seven-day Larval Survival and Growth Test
Single Laboratory Precision Data
Table A-1-21. Single laboratory precision of the fathead
minnow (Pimephales promelas) larval survival and growth test
performed in using NAPCP as the reference toxicant [2].
Test
Number
1
2
3
4
5
n:
Mean:
CV(%):
NOEC*
(ug/0
256
128
256
128
128
5
1 79.2
I NA
•Raw data unavailable, \C2S and IC^o values could not be calculated.
NOEC Range: 128 - 256 ug/l (this represents a [difference of one exposure
concentration).
[
Table A-1-22. Results of the performance evaluation for contract laboratories conducted for the California Regional Water Quality
Control Board. All tests were conducted using potassium chromate (expressed as Cr+^) and testing the fathead minnow (Pimephales
promelas) in the 7-day subchronic tests [3].
Lab
1
2
3
4
5
6
7
8
9
10
Mean:
CV(%):
Water
Tap'
MHRW
MHRW
Tap5
MHRW
MHRW
MHRW
Well7
MHRW
MHRW
Food
2X
2X
3X
3X
_6
3X
3X
2X
3X
3X
Age
<24
<24
<24
<24
<24
y<24
<24
<24
<24
<24
X Control
Weight
0.590
0.623
0.2744
0.670
0.773
0.635
0.390
0.346
0.415
0.255
Ctrl.
n
3
32
4
2
4
2
3
5
4
2
ICzs (CD
(mg/l as Cr+6)
3.7 (2.3-4.7)
1.63 (1.4-2.0)
i 2.23 (1.7-3.1)
4.1 (2.3-5.0)
1.33(1.2-1.5)
7.1 (2.0-8.2)
4.5 (3.5-5.4)
2.53 (1,9-3.3)
6.6 (5.3-7.6)
4.6 (4.1-5.9)
5.1
27
NOEC
Endpoint
3G
<3G
<3G
6G
<3G
6G
3G
<3G
6G
3G
IC50 (Cl)
(mg/l as Cr+6)
5.4 (4.5-8.3)
3.3 (2.8-4.0)
4.7 (3.9-5.6)
6.6 (5.0-8.4)
.63 (2.5-3.3)
9.9 (8.5-11)
7.4 (6.6-8.1)
8.1 (6.4-15)
9.2 (8.4-10)
7.88(5.2-12)
6.9
31
1
Moderate!/ hard tap water.
* Control with three replicates and all concentrations with two replicates.
3 Value is extrapolated and is not included in coefficient of variation calculation.
"* Weight measurements made with questionable techniques.
* Dcchlorinatcd Ukc Ontario tap water.
6 Not reported.
' Weil water mixed with spring water, moderately hard.
* Value may be skewed as middle concentration had 45 percent survival but no weights reported.
A-1-8:
-------
Seven-day Larval Survival and Growth Test
Intel-laboratory Precision Data
Table A-1-23. Intel-laboratory precision data of the fathead minnow (Pimephales promelas) 7-day larval survival and growth test.
Combined frequency distribution for survival NOECs for all participating laboratories [2].
Sample
Sodium Pentachlorophenate 1
Sodium Pentachlorophenate 2
Potassium Dichromate 1
Potassium Dichromate 2
Refinery Effluent 301
Refinery Effluent 401
Utility Waste 501
NOEC Frequency (%) Distribution
Tests with 2 Reps
Median
35
42
47
41
26
37
56
±1(a)
53
42
47
41
68
53
33
>2(b)
12
16
6
18
6
10
11
Tests with 4 Reps
Median
57
56
75
50
78
56
56
±l(a)
29
44
25
50
22
44
33
>2(b)
14
0
0
0
0
0
11
Percent of values with one concentration intervals of the median.
Percent of values within two or more concentrations intervals of the median.
Table A-1-24. Interlaboratory precision data of the fathead minnow (Pimephales promelas) 7-day larval survival and growth test.
Combined frequency distribution for weight NOECs for all participating laboratories [2].
Sample
Sodium Pentachlorophenate 1
Sodium Pentachlorophenate 2
Potassium Dichromate 1
Potassium Dichromate 2
Refinery Effluent 301
Refinery Effluent 401
Utility Waste 501
NOEC Frequency (%) Distribution
Tests with 2 Reps
Median
59
37
35
12
35
37
11
±1(a)
41
63
47
47
53
47
61
>2(b)
0
0
18
41
12
16
28
Tests with 4 Reps
Median
57
22
88
63
75
33
33
±1(a)
43
45
0
25
25
56
56
>2(b)
0
33
12
12
0
11
11
Percent of valuss with one concentration intervals of the median.
Percent of values within two or more concentrations intervals of the median.
A-1-9
-------
CERIODAPHNIA (Ceriadaphnia dubia)
Seven-day Reproduction Test
Single Laboratory Precision Data
Table A-1-25. Single laboratory precision of (Ceriodaphnia dubia)
reproduction test performed in using sodium pentachlorophenol as
the reference toxicant [2].
Test Number
19
46A
468
49
55
56
57
n:
Mean:
CV(%):
NOEC
(mg/l)
0.30
0.20
0.20
0.20
0.20
0.10
0.20
7
0.20
NA
K25
(mg/l)
0.3754
0.0938
0.221 3
0.2303
0.2306
0.1345
0.2241
7
0.2157
41.1
icso
(mg/l)
0.4508
0.2608
0.2897
0.2912
0.3177
0.1 744
0.2827
7
0.2953
27.9
NOEC Range: 0.25 - 0.30 mg/l (these values all fell within the same
concentration range).
Prepared by Florence Kessler, TAI, Cincinnati, OH, January 11, 1990 (ICp
Program, version 1.1 b).
Table A-1-26. Single laboratory precision, from six discrete laboratories, of the (Ceriodaphnia dubia) reproduction test performed using sodium
chloride (NaCI) as the reference toxicant. Tests were conducted in 1989 [4].
Laboratory
A
n:
Mean:
CV(%):
B
n:
Mean:
CV(%):
C
n:
Mean:
CV(%):
Test
Number
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
NOEC
(mg/l)
0.50R
1.00s
1.00R
1.00R
1.00R
0.50R
6
0.83
NA
1.00R
1.00s
0.50s
0.50s
1.00R
1.00R
6
0.83
NA
1.00s
0.50s
1.00s
0.50R
1.00s
1.00s
6
0.83
NA
IC25
(mg/l)
0.61
1.00
0.81
0.67
1.19
1.06
6
0.89
25.83
1.28
1.01
0.69
0.81
1.31
1.12
6
1.04
24.11
1.23
0.46
1.25
1.13
1.22
1.21
6
1.13
16.54
icso
(mg/l)
0.77
1.34
1.32
1.28
1.47
1.38
6
1.26
19.73
1.63
1.51
0.88
1.16
1.84
1.57
6
1.43
24.37
1.49
1.02
1.50
1.44
1.49
1.51
6
1.41
13.62
Laboratory
D
n:
Mean:
CV(%):
E
n:
Mean:
CV(%):
p
n:
Mean:
CV(%):
Test
Number
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
NOEC
(mg/l)
, 0.50R
0.25R
1.00s
1.00s
1.00s
0.50R
6
0.71
NA
1.00s
1.00s
1.00s
1.00s
1.00s
1.00s
6
1.00
NA
0.50R
0.50R
0.50s
0.50s
0.50R
0.50R
6
0.50
NA
«C25
(mg/l)
0.58
0.30
0.84
1.04
1.04
0.76
6
0.76
37.55
0.44
1.04
1.06
1.13
1.13
1.19
6
1.00
27.96
0.61
0.63
0.66
0.65
0.74
0.50
6
0.63
12.40
IC50
(mg/l)
0.84
0.60
1.22
1.38
1.37
1.14
6
1.08
28.56
0.74
1.37
1.37
1.42
1.42
1.46
6
1.30
21.20
1.13
1.20
0.83
0.81
1.04
0.73
6
0.95
19.32
R « Reproduction was the most sensitive endpoint.
S • Survival was the most sensitive endpoint f
Prepared by William Peltier, EPA, Region IV, November 28, 1990 (ICp Program, version fl.lb).
A-1-10
-------
CERIODAPHNIA (CerioHaphnia Hubia)
Seven-day Larval Reproduction Test Interlaboratory Precision Data
Table A-1-27. Interlaboratory precision of (Ceriodaphnia dubid)
reproduction test, using sodium chloride (NaCI) as the
reference toxicant. The single lab precision data are presented
in the preceding table [4].
Laboratory
A
B
C
D
E
F
n:
Mean:
CV(%):
NOEC
(mg/l)
0.83
0.83
0.83
0.71
1.00
0.50
6
0.80
NA
IC25
(mg/l)
0.89
1.04
1.13
0.76
1.00
0.63
6
0.91
20.53
ICso
(mg/l)
1.26
1 .43 -.
1.41
1.09
1.30
0.95
6
1.24
15.17
Prepared by William Peltier, EPA, Region IV, November 28,
1990 (ICp Program, version 1.1b).
Table A-1-28 Interlaboratory precision of (Ceriodaphnia dubia)
reproduction test, using an industrial effluent as tne reference
toxicant and sodium chloride (NaCI) as a reference toxicant.
Tests were conducted in May 1987 [3].
Lab
A
B
D
E
F
J
K
M
N
O
n:
Mean:
CV(%):
Effluent
IC50(%)
6.20
8.40
7.69
6.34
4.00
2.84
6.89
5.70
7.43
0.04*
9
6.17
29
IC25 (%)
4.9
6.2
5.8
5.0
1.2
1.9
5.3
1.9
5.9
0.02*
9
3.4
67
Reference Toxicant
ICso (%)l
33.0
38.8
36.3
36.6
8.1*
35.1
18.4
38.1
27.8
35.1
9
32.8
21
ICzs (%)
21.8
30.8
29.4
28.0
1.21*
25.2
13.2
31.0
10.4
27.3
9
24.1
31
Values were excluded from mean calculations because they fell outside of ± 2
standard deviations. For this reason, these values are considered statistical
outliers and, according to EPA's toxicity methods guidance [2] on reference
toxicant control charts, are excluded.
Table A-1-29. 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 7-day chronic
tests [3].
Lab
1
2
3
4
5
6
7
8
9
10
Mean:
CV(%):
Water
Tap1
Hard W3
DMW4
Tap5
HRW
Surface6
MHRW
MHRW
MHRW
DMW4
Food
YCT/Algae
TF/Algae
YCT/Algae
YCT
YCT
YCT/Algae
YCT/Algae
YCT
YCT
YAT/Algae
Age
0-4;<24
0-4
0-6
0-4
0-4;<24
0-6
4-8
<24
0-4
0-4
X Young/
Control
1 7.80.202
26.51.3
24.90.21 2
17.20.49
1 9.80.42
14.80.90
17.20.56
16.80.21 2
12.80.71
31.50.91
"Czs (Cl)
(g/l NaCI)
(0.14-0.35)
(0.78-1.7)
(0.17-0.54)
(0.35-1 .0)
(0.20-1.1)
(0.66-1.1)
(0.24-0.64)
(0.11-0.32)
(0.56-0.81)
(0.45-1.1)
0.76
40
NOEC
Endpt
<0.25 R
1.0R
<0.25 R
0.5 R
0.5 R
0.25 R
0.25 R
0.25 R
0.50 R
1.0R
Moderately hard tap water.
2 Dose response curve limited.
Hard well water.
^ Ten":percent diluted mineral water.
Dechlorinated Lake Ontario tap water.
6 Briones reservoir water.
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
A-1-11
-------
CERIODAPHNIA (Ceriodaphnia dubia) (continued)
Seven-day Larval Reproduction Test Intel-laboratory Precision Data
Table A-1-30. Interlaboratory precision data for\Ceriodaphnla dubia summarized for
eight materials, including reference toxicants apd effluents [5].
1
2
3
4
5
6
7
8
Test
Material
Sodium chloride
Industrial
Sodium chloride
Pulp & Paper
Potassium dichromate
Pulp & Paper
Potassium dichromate
Industrial
n:
Mean:
Standard Deviation:
Mean
ICso
1.34
3.6
0.96
60.0
35.8
70.2
53.2
69.8
CV%
29.9
83.3
57.4
28.3
30.8
7.5
25.9
37.0
8
37.5
23.0
Mean
ICs
1.00
3.2
0.90
47.3
23.4
55.7
29.3
67.3
CV%
34.3
78.1
44.4
27.0
32.7
12.2
46.8
36.7
8
39.0
19.1
SELENASTRUM CAPRICORNUTUM
Growth Test
Single Laboratory Precision Data
Table A-1-31. Single laboratory precision of (Selenastrum
capricornutum) growth test performed in using cadmium as
the reference toxicant [2].
Test
Number
1
2
3
4
5
6
7
8
9
10
11
12
n=10
Mean:
CV(%):
EC50
(g/0
2.3
2.4
2.3
2.8
2.6
2.1
2.1
2.1
2.6
2.4
2.7
2.4
I'
2.37
10.2
Control Variation
(%CV)
4.8
9.6
5.5
13.3
4.4
8.2
14.4
7.1
11.9
5.0
36.4*
77.8*
8.42
44.1
'Outlier values are excluded from mean because they fell outside of a QA
control table for these reference toxicants, j
Note: Sodium chloride concentrations were '\, 2, 4, 8, and 16 g/l in all tests.
Prepared by Dr. Cornelius Weber, EPA, Cinciriatti, OH, March 15, 1991.
A-1-12
-------
APPENDIX A-2
EFFLUENT VARIABILITY DATA
-------
Table A-2-1. Percent mortality in 100 percent 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 Reference 6.
Date
3/07/89
3/07/89
3/08/89
3/08/89
3/20/89
3/20/89
3/21/89
3/21/89
6/19/89
6/19/89
6/20/89
6/20/89
7/25/89
7/25/89
7/26/89
7/26/89
Time
1230
1830
0030
0630
1230
1830
0030
0630
1230
1830
0030
0630
1230
1830
0030
0630
% Mortality in 100% Effluent
P. promelas
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
D. pulex
15
85
65
30
0
100
95
70
5
40
100
100
0
100
100
55
C. dubia
100
100
100
80
100
100
100
100
100
100
100
100
100
100
100
100
Table A-2-3. LC50s for a POTW effluent over 7 months. All
tests were conducted using Ceriodaphnia dubia and tests were
run for 48 hours. All tests were conducted according to the
methods described in Reference 6. Dates with roman numeral
notation mean that more than one sample was collected at
different times over a short interval (1 to 2 days). (Data
source: [8].)
Table A-2-2. LC50s for a POTW effluent over 17 months. All
tests were conducted using Ceriodaphnia dubia and tests were
run for 48 hours. All tests were conducted according to the
methods described in Reference 6. Dates with roman numeral
notation mean that more than one sample was collected at
different times over a short interval (1 to 2 days). (Data
source: [8].)
Sample Date
Mean
CV(°/
n:
10/06/87-1
10/06/87-11
10/06/87-111
10/30/87-1
12/03/87-1
12/03/87-11
01/12/88-1
01/13/88-1
02/03/88-IX
02/03/88-X
03/03/88-111
03/03/88-IV
03/23/88-1
03/23/88-11
04/28/88-1
04/28/88-11
05/1 7/88-I
05/1 7/88-11
LCSO:
>):
48 hours LC50 (%)
71
71
71
87
61
35
61
58
50
50
87
81
25
35
50
55
61
35
58.0
31.4
18
Sample Date
Mean
CV(
-------
Table A-2-4. LCsos for a POTW effluent over 12 months.
Tests were conducted using either Ceriodaphnia dubia or
fathead minnows or both. Ceriodaphnia tests were conducted
for 48 hours while fathead minnow tests were 96 hours. Both
the 48-hour and 96-hour fathead minnow results are shown in
order to evaluate how the LCsos for the two species compare.
All tests were conducted according to the methods described
In Reference 6. Dates with roman numeral notation mean that
more than one sample was collected at different times over a
short interval (1 to 2 days). (Data source: [8].)
Sample Date
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
Mean LCSO:
CV (%):
n:
LC50(%)
C. dubia
48 hours
62
18
68
61
63
70
17
35
35
35
47
46
42
11
P. promelas
48 hours
35
*
>100
*
*
58
60
61
61
81
61
59.6
22.4
7
96 hours
25
*
>100
*
*
34
41
39
37
64
40
40.0
29.7
7
Data not available.
Note: Greater than (>) values were excluded from the mean LCso calculation.
| Table A-2-5. LC5QS for a POTW effluent over 4 months.
Tests were conducted using either Ceriodaphnia dubia or
fathead minnows or both. Ceriodaphnia tests were conducted
for 48 hours while fathead minnow tests were 96 hours. Both
the 48-hour and 96-hour fathead minnow results are shown in
order to evaluate how the LCsos for the two species compare.
All tests were conducted according to the methods described
I in Reference 6. Dates with roman numeral notation mean that
more than one sample was collected at different times over a
short interval (1 to 2 days). (Data source: [8].)
Sample Date
09/01/88-1
11/15/88-1
11/16/88-1
11/30/88-11
11/30/88-111
12/08/88-1
12/08/88-11
12/13/88-1
12/13/88-11
01/10/89-1
01/10/89-11
01/19/89-1
01/19/89-11
01/25/89-1
01/25/89-11
01/31/89-1
01/31/89-11
Mean:
CV (%):
n:
LCSO(%)
C. dubia
48 hours
2.1
92
61
>100
95
100
>100
90
87
75
61
100
87
>100
95
90
63
78.4
33.1
14
P. promelas
48 hours
>100
67
>100
>100
>100
87
87
>100
85
58
41
88
84
87
85
70
70
75.8
19.6
12
96 hours
77
37
100
33
>100
54
53
77
51
*
*
68
69
64
56
60
60
61.3
27.7
14
*Not obtained.
Note: Greater than (>) values were not included in the mean LCso calculation.
A-2-2
-------
Table A-2-6. NOECs for a POTW effluent conducted 20 times
over 1 year. All tests were conducted using Champia parvula
according to methods described in Reference 1. All effluent
samples were 24-hour composites collected post-chlorination.
(personnal communication—Glen Thursby, SAIC,
Narragansett, Rl).
Test Date
1 2/09/85
1 2/1 0/85
12/11/85
12/12/85
12/13/85
12/15/85
07/1 6/86
07/1 7/86
07/18/86
07/19/86
07/20/86
07/21/86
07/22/86
9/09/86
09/1 0/86
09/11/86
09/12/86
09/14/86
n:
Mean:
CV (%):
% Effluent
IC25
0.65
0.38
0.69
0.41
3.09
2.16
2.99
3.59
.44
.47
.24
.11
.84
.07
.17
.73
.57
.25
18
2.2
52.8
ICso
1.23
0.76
1.50
0.82
4.12
4.09
4.33
4.68
4.76
3.41
3.98
3.20
5.19
3.02
4.13.
3.62
1.89
1.76
18
3.1
46.8
NOEC
1.25
1.25
2.50
1.25
5.00
5.00
5.00
5.00
5.00
5.00
7.50
5.00
5.00
2.50
7.50
7.50
1.25
2.50
18
4.2
NA
Table A-2-8. NOECs for a POTW effluent over 1 year. All
tests used Mysidopsis bahia according to methods described in
Reference 1. All effluent samples were 24-hour composites
collected post-chlorination. (Data source: ERL-Narragansett,
Rl.)
Test Date
12/09- 12/16/85
07/16-07/23/86
09/09 - 09/1 6/86
11/11 -11/18/86
Mean:
CV (%):
n:
% Effluent
«C25
1.78(G)
2.75(R)
0.69(R)
0.66(R)
1.47
68.0
4
ICso
2.93(G)
6.3(S)
20.1(S)
0.99(R)
7.58
113.8
4
NOEC
1.0
3.2
10.0
3.2
4.4
NA
4
R-Reproductive endpoint
S-Survival endpoint
G-Growth endpoint
Table A-2-7. NOECs for a POTW effluent over 1 year. All
tests used Arbacia punctulata according to methods described
in Reference 1. All effluent samples were 24-hour composites
collected post-chlorination. (Data source: ERL-Narragansett,
Rl.)
Test Date
12/09/85
12/10/85
12/11/85
12/12/85
1 2/1 3/85
12/14/85
12/15/85
07/1 6/86
07/1 7/86
07/1 8/86
07/19/86
07/20/86
07/21/86
07/22/86
09/09/86
09/11/86
09/12/86
09/1 3/86
09/14/86
09/15/86
11/11 /86
11/13/86
11/14/86
11/15/86
Mean:
CV (%):
n:
% Effluent
IC25
1.09
1.41
0.75
3.28
2.65
1.11
1.29
0.17
0.21
0.63
1.09
0.54
0.40
0.40
0.31
0.47
0.21
3.30
0.23
0.10
0.27
0.88
0.82
0.34
0.91
101.3
24
ICso
1.71
2.84
1.09
4.06
5.32
1.60
1.84
0.35
0.46
0.86
1.68
1.13
0.58
0.56
0.41
0.79
0.48
5.42
0.35
0.15
0.54
1.48
1.61
0.56
1.49
96.9
24
NOEC
0.65
0.65
0.65
1.30
2.50
0.65
0.65
<0.30
<0.30
<0.30
<0.30
<0.30
<0.30
<0.30
<0.30
<0.60
<0.20
1.30
<0.20
<0.20
1.30
0.30
0.60
<0.30
0.95
NA
11
Note: Less than (<) values were excluded from CV and mean NOEC calculations.
A-2-3
-------
Table A-2-9. NOECs for a POTW effluent over 1 year. All
tests used Menldla beryllina according to methods described
In Reference 1. All effluent samples were 24-hour composites
collected post-chlorination. (Data source: ERL-Narragansett,
Rl.) ' '
Test Date
12/09-12/16/85
07/16-07/23/86
09/09-09/16/86
11/11 -11/18/86
Mean:
CV(%):
n:
% Effluent
ICzs
15.4
15.2
14.2
NC
14.9
4.3
3
ICso
21.3
21.0
20.1
NC
20.8
3.0 .
3
NOEC
10.0
10.0
10.0
10.0
10.0
NA
4
NC - Value Is not calculable.
Table A-2-10. LCsrjs for a refinery effluent over 14 months.
Tests were conducted using either Ceriodaphnia dubia or
fathead minnows Plmephales promelas or both. Ceriodaphnia
tests were conducted for 48 hours while fathead minnow tests
were 96 hours. Both the 48-hour and 96-hour fathead minnow
results are shown in order to evaluate how the LCsos for the
two species compare. All tests were conducted according to
the methods described in Reference 6. Dates with roman
numeral notation mean that more than one sample was
collected at different times over a short interval (1 to 2
days). (Data source: [8].)
Sample Date
12/01/87
01/05/88
02/09/88-1
02/09/88-1
03/02/88-1
03/02/88-11
03/24/88-1
05/06/88-1
07/14/88-1
07/28/88-1
07/28/88-11
09/29/88-1
12/01/88-1
12/07/88-1
01/27/89-1
01/27/89-11
03/23/89-1
Mean LCso:
CV(%):
n:
LCSO(%)
C. dubla
48 hours
15
35
35
35
17
<12
35
35
55
37
28
41
75
18
100
71
58
43
54
16
P. promelas
48 hours
35
36
35
35
*
38
35
*
61
35
31
39
56
67
61
60
54
45
28
15
96 hours
16
19
<12
<12
*
*
*
*
25
22
<25
25
34
13
37
25
20
24
32
10
Data not available.
Note: Less than values excluded from mean LCso calculations.
Table A-2-11. LCsrjs for a refinery effluent conducted over 6
months using fathead minnows (Pimephales promelas),
Ceriodaphnia dubia, and mysids (Mysidopsis bahia), according
to methods described in Reference 6. (Data source: Dorn,
1989.)
Test Date
1/24/86
2/26/86
3/05/86
3/1 2/86
3/19/86
4/02/86
4/09/86
4/1 7/86
4/23/86
5/14/86
5/28/86
6/11/86
Mean NOEC:
CV (%):
n:
(% Effluent)
C. dubla
-
65.0
50.9
39.3
66.5
65.4
69.8
71.2
71.8
82.0
65.4
82.0
66.3
18.7
11
P. promelas
26.6
24.5
-
36.6
40.5
32.8
34.2
37.2
35.9
38.7
22.0
-
32.9
19.5
10
M. bahla
-
•
-
-
-
-
-
-
38.0
35.8
-
24.7
32.8
21.6
3
Table A-2-12. NOECs for a refinery effluent conducted over 6
months using fathead minnows (Pimephales promelas),
Ceriodaphnia dubia, and mysids (Mysidopsis bahid), according
to methods described in References 1 and 2. (Data source:
Dorn, 1989.)
Test Date
1/24/86
2/26/86
3/05/86
3/12/86
3/19/86
4/02/86
4/09/86
,4/1 7/86
4/23/86
5/14/86
5/28/86
6/11/86
Mean NOEC:
CV (%):
n:
LCM
(% Effluent)
C. dubla
-
10.1
5.6
10.1
10.1
18.0
10.1
10.1
10.1
31.7
18.0
31.7
15.1
59.6
11
P. promelas
14.1
7.1
-
14.1
14.1
14.1
14.1
7.1
7.1
14.1
7.1
-
11.3
31.9
10
M. bahla
-
-
-
-
-
-
-
-
24.0
24.0
-
13.4
20.5
29.8
3
A-2-4
-------
Table A-2-13. LCsQS for a manufacturing effluent conducted over 2
ears. All tests were conducted using Daphnia magna according to
methods described in Reference 6. (Data source: [8].)
Test Date
1 982 (1 st quarter)
1 982 (4th quarter)
1 982 (4th quarter)
1983 (3rd quarter)
1983 (3rd quarter)
1983 (3rd quarter)
1 983 (3rd quarter)
1 983 (3rd quarter)
1 983 (3rd quarter)
1983 (3rd quarter)
1 983 (3rd quarter)
1983 (4th quarter)
1 983 (4th quarter)
1983 (4th quarter)
1 983 (4th quarter)
1 983 (4th quarter)
1 983 (4th quarter)
1 983 (4th quarter)
Mean LCso:
CV (%):
n:
LCso
(% Effluent)
56
90
70
69
36
36
32
<18
28
67
<10
46
75
78
24
26
32
19
45.1 + 24.3
53.9
18
Note: Less than (<) values were excluded from the mean
calculations.
Table A-2-14. LCsos for a manufacturing effluent conducted over 8
years. All tests were conducted using Pimephales promelas according
to methods described in Reference 6. (Data source: [8].)
Test Date
1 979 (1 st quarter)
1979 (1st quarter)
1979 (1st quarter)
1979 (3rd quarter)
1 981 (2nd quarter)
1 981 (4th quarter)
1 982 (2nd quarter)
1 982 (2nd quarter)
1 985 (1 st quarter)
1 985 (4th quarter)
1 986 (2nd quarter)
1 986 (2nd quarter)
1986 (2nd quarter)
1 986 (3rd quarter)
1 986 (3rd quarter)
1 986 (3rd quarter)
1986 (4th quarter)
Mean LCso:
CV (%):
n:
LCso
(% Effluent)
72.0
62.0
52.0
39.0
64.0
70.0
44.0
66.0
59.6
>100.0
49.2
63.8
50.0
75.7
80.0
79.0
71.0
64.5 + 15.1
23.5
17
Table A-2-15. LCsns for a manufacturing effluent conducted
over 5 years. All tests were conducted using Pimephales
promelas according to methods described in Reference 6.
(Data source: [8].)
Test Date
1980 (1st quarter)
1 980 (2nd quarter)
1 980 (3rd quarter)
1 980 (4th quarter)
1981 (1st quarter)
1 981 (2nd quarter)
1981 (3rd quarter)
1981 (4th quarter)
1982 (1st quarter)
1 982 (2nd quarter)
1 982 (3rd quarter)
1982 (4th quarter)
1983 (1st quarter)
1 983 (2nd quarter)
1983 (3rd quarter)
1 983 (4th quarter)
1984 (1st quarter)
1 984 (2nd quarter)
1 984 (3rd quarter)
1984 (4th quarter)
Mean LCSO:
CV (%);
n:
LCso
(% Effluent)
18.0
11.0
32.0
16.0
32.0
23.0
17.0
46.0
9.0
32.0
28.0
52.0
34.0
33.0
20.0
43.0
45.0
19.0
61.0
20.0
29.6 ±14.2
47.9
20
Table A-2-16. LCsos for a manufacturing effluent conducted
over 5 years. All tests were conducted using Daphnia magna
according to methods described in Reference 6. (Data source:
[8].)
Test Date
1981 (2nd quarter)
1 981 (3rd quarter)
1982 (3rd quarter)
1984 (4th quarter)
1 985 (1 st quarter)
1986 (1st quarter)
1 986 (2nd quarter)
1987 (1st quarter)
1987 (1st quarter)
1987 (1st quarter)
1 987 (1 st quarter)
Mean LCSO:
CV (%):
n:
LCso
(% Effluent)
100.0
>100.0
>100.0
80.0
75.0
25.0
82.0
75.0
24.0
>1 00.0
>100.0
65.9 + 29.5
44.8
11
Note: Greater than (>) values were excluded from the mean LCso calculations.
Note: Greater than (>) values were excluded from the mean LCso calculations.
A-2-5
-------
Table A-2-17. LCsrjs for a manufacturing effluent conducted
over 7 years. All tests were conducted using Daphnia pulex
according to methods described in Reference 6. (Data source:
[8].)
Test Date
1980 (1st quarter)
1980 (4th quarter)
1981 (1st quarter)
1981 (1st quarter)
1981 (1st quarter)
1981 (2nd quarter)
1981 (3rd quarter)
1982 (3rd quarter)
1986 (3rd quarter)
1986 (3rd quarter)
Mean LCs0:
CV (%):
n:
LCso
(% Effluent)
55.0
33.0
60.0
24.0
>100.0
>100.0
>100.0
>100.0
>100.0
>100.0
43.0 ± 1 7.3
40.2
10
Note: Greater than (>) values were excluded from the mean
calculations.
Table A-2-18. LC5QS for a manufacturing effluent conducted
I over 3 months. All tests were conducted using Daphnia
magna according to methods described in Reference 6. (Data
source: [8].)
Test Date
1 982 (4th quarter)
1 982 (4th quarter)
1 982 (4th quarter)
1 982 (4th quarter)
1 982 (4th quarter)
1982 (4th quarter)
1 982 (4th quarter)
1 982 (4th quarter)
1982 (4th quarter)
1 982 (4th quarter)
1 982 (4th quarter)
1982 (4th quarter)
Mean LC50:
CV (%):
n:
LCso
(% Effluent)
>100.0
81.0
57.0
61.0
87.0
90.0
90.0
>100.0
>100.0
54.0
74.0
>100.0
74.3 ±15.1
20.3
12
Note: Greater than (>) values were excluded from the mean
calculations.
A-2-6
-------
APPENDIX A-3
ACUTE-TO-CHRONIC RATIO DATA
-------
Table A-3-1. Example of Acute-to-Chronic Ratios
Oil Refinery1
Table A-3-2. Examples of Acute-to-chronic Ratios
Chemical Manufacturers
Mean ACR:
n:
Range:
Fathead
Minnow
1.89
3.47
2.60
2.87
2.33
2.43
5.26
5.08
2.74
3.11
5.1
3.3
11
1.89-5.26
Ceriodaphnia
9.09
3.89
6.58
3.63
6.91
7.05
7.11
3.63
2.59
5.5
4.4
>10.0
>3.3
>2.0
>3.0
2.8
5.42
5.3
13
2.59->10.0
Mysids
1.58
1.49
1.84
1.64
3
1 .49-1 .84
! Personal communication P. Dorn.
2 Personal communication M.L.C. Ramos and E. Bertoletti (Sao Paulo, Brazil).
Note: Greater than (>) values were excluded from mean calculations.
Mean ACR:
n:
Range:
Fathead
Minnow
0.17
0.07
8.4
7.6
>3.0
3.9
>3.0
1.8
3.7
6
0.07 - 8.4
Ceriodaphnia
>1.0
>1.0
>10.0
>50.0
>2.9
>1.4
1.4
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 1
14.01
4.31
2.51
1.81
5.51
5.41
3.72
20
1 .4 - >50
Personal communication M.L.L.C. and E. Bertoletti (Sao Paulo, Brazil).
' Greater than (>) values were excluded from the mean calculation.
A-3-1
-------
Table A-3-3. Example of ftcute-to-Chronic Ratios
POTWs
Mean ACR:
n:
Range:
Fathead
Minnow
2.9
6.1
1.5
13.0
1.8
2.6
9.3
> 1.0
>3.0
5.3
3.3
5.4
>3.0
3.0
•
I,
4.9
11
1.5-9.3
Ceriodaphnia
1.4
5.5
> 1.0
> 1.0
>1.0
1.8
1.4
2.0
2.4
3.0
3.0
5.5
4.9
>2.0
>8.0
>2.0
>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
3.8*
21
1.4-16.1
' Greater than (>) values were excluded from mean calculations.
A-3-2
-------
APPENDIX A
REFERENCES
1. Weber, C.I. et. al., eds. 1988. Short-Term Methods for Estimating the Chronic Toxicity of Effluents and Receiving Waters
to Marine and Estuarine Organisms. EPA 600/4-87/028. Office of Research and Development, Cincinnati, OH.
2. Weber, C.I. et. al., eds. 1989. Short-Term Methods for Estimating the Chronic Toxicity of Effluents and Receiving Waters
to Freshwater Organisms, 2d ed. EPA 600/4-89/001. Office of Research and Development, Cincinnati, OH.
3. Norberg-King, T.J. Memorandum: Summary of Effluent Toxicity Data for the Technical Support Document for Water
Quality-based Toxics Control, Chapters 2 and 3. April 11, 1989. U.S. EPA, Environmental Research Laboratory, Duluth,
MN, to M.A. Heber, U.S. EPA, Enforcement Division, Washington, DC. Memorandum: Technical Support Document for
Water Quality-based Toxics Control Data: Fathead Minnow Round Robin Data Analyzed to Obtain the IC25S. May 6,
1990. U.S. EPA, Environmental Research Laboratory, Duluth, MN, to M.A. Heber, U.S. EPA, Enforcement Division,
Washington, DC.
4. Peltier, W. January 23, 1991. Memorandum: Intra-and Interlaboratory Precision Data from Six Laboratories Conducting
Short-Term Chronic Toxicity Tests Using Ceriodaphnia dubia. U.S. EPA, Region IV, Environmental Services Division,
Athens, GA, to M.A. Heber, U.S. EPA, Enforcement Division, Washington, D.C.
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, OH (in preparation).
6. Peltier, W., and C.I. Weber. 1985. Methods for Measuring the Acute Toxicity of Effluents to Aquatic Organisms, 3d ed.
EPA 600/4-85/013. Office of Research and Development, Cincinnati, OH.
7. Ward, S.H. 1990. Technical Report: Procedures for Toxicant Testing and Culture of the Marine Macroalga, Champia
parvula. U.S. EPA, Region II.
8. Norberg-King, T.j. May 4, 1989. Memorandum: Additional Data for the Technical Support Document for Water Quality-
based Toxics Control, Chapters 2 and 3. U.S. EPA, Environmental Research Laboratory, Duluth, MN, to M.A. Heber, U.S.
EPA, Enforcement Division, Washington, DC.
A-3-3
-------
APPENDIX B-l
SUMMARY OF CLEAN WATER ACT PROVISIONS
-------
CLEAN WATER ACT (33 U.S.C. 1251 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 limitation in the
NPDES program is clearly authorized by the Clean Water Act (CWA).
Relevant provision of the CWA that provide the statutory authority for using toxicity testing and Whole
Effluent toxicity limitations include the following:
• Section 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 in Section 101(a)(3) the
national policy of prohibiting the "discharge of toxic pollutants in toxic amounts" (emphasis added).
• As defined at Section 502(15), biological monitoring means that "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."
• Section 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.
• Section 303(c)(2)(B) states, "Nothing in this section shall be 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 balance
population of shellfish, fish and wildlife, among other uses. The effluent limitations established
must reasonably be expected to contribute to attainment or maintenance of such water quality.
• Under Section 301(b)(1)(C) and Section 402, all NPDES permits must comply with any more stringent
limitations necessary to meet applicable water quality standards, whether numeric or narrative.
• CWA Section 308(a) and Section 402 provide authority to EPA or the Sate 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.
B-1-1
-------
APPENDIX B-2
POLICIES FOR Toxics CONTROL
-------
9016^ Federal Register / Vol. 49. No. 48 / Friday, March 9, 1984 / Notices
[OW-FRU-2533-D
Development of Water Quality-Based
Permit Limitations for 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.
B-2-1
-------
Federal Register / Vol. 49, No. 48 / Friday, March 9. 1984 / Notices
9617
SUPPLEMENTARY INFORMATION: As the
water pollution control effort in the
United States progresses and die
"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
thepolicy.
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 Vyater Act (the
Act), EPA or the-State may require
NPDES permit applicants ito provide
chemical, toxicity, and instream
biological data necessary ito assure
compliance with standards. Data
requirements may be determined on a
case-by-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 £nd 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 will 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 are 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 oy EPA. In
B-2-2
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, and set permit limitations to
achieve those standards. All States have
water quality standards which include
narrative 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, 1984 / Notices
compliance with water quality
standards.
Note.—Section 308 of the Act and
corresponding State statutes authorize EPA
and the States to require of the owner/
operator any information reasonably required
to determine permit limits and to determine
compliance with standards or permit limits.
Biological methods are specifically
mentioned. Toxicity permit limits are
authorized under Section 301 and 402 and
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 FR
52079, November 18,1982). 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-440/9-76-025J.
This policy is part of EPA's water
quality-based control program and does
not supersede other regulations, policy,
and guidance regarding use
attainability, site-specific criteria
modification, wasteload allocation, and
water quality management,
Implementation
State Role
The control of toxic substances to
protect water quality must be done in
the context of the Federal-State
partnership. EPA will work
cooperatively with the States in
identifying potential water quality
standards violations, assembling
relevant .data, developing appropriate
testing requirements, determining
whether standards are being violated,
and defining appropriate permit limits.
Note.—Under sections 303 and 401 of the
Act, States are given primary responsibility
for developing water quality standards and
limits to meet those standards. EPA's role is
to review the State standards and limits and
develop revised or additional standards or
limits as needed to meet the requirements of
the Act.
Integration of Approaches
The type of testing that is most
appropriate for assessing water quality
impacts depends on the type of effluent
and discharge situation. EPA
recommends that an integrated
approach, including both biological and
chemical techniques, be used to assess
and control water quality. The principal
advantages of chemical-specific.
techniques are that (1) chemical
analyses are usually less expensive than
biological measurements in simple
cases; (2) treatment systems are more
easily designed to meet chemical
requirements than toxicity requirements;
and (3) human health hazards and
bioaccumulative pollutants can best be
addressed at this time by chemical- .
specific analysis. The principal
advantages of biological techniques are
that (1) the effects of complex
discharges of many known and
unknown constituents can be measured
only by biological analyses; (2)
bioavailability of pollutants after
discharge is best measured by toxicity
testing; and (3) pollutants for which
there are inadequate chemical analytical
methods or criteria can be addressed.
Pollutant-specific chemical analysis
techniques should be used where
discharges contain few, well-quantified
pollutants and the interactions and
effects of the pollutants are known. In
addition, pollutant-specific techniques
should be used where health hazards
are a concern or bioaccumulation is
suspected. Biological techniques should
be used where effluents are complex or
where the combined effects of multiple
discharges are of concern. EPA
recognizes that in many cases both
types of analysis must be used.
Testing Requirements
Requirements for dischargers to
collect information to assess attainment
or nonattainment of State water quality
standards will be imposed only in
selected cases where the potential for
nonattainment of water quality
standards exists. Where water quality
problems are suspected but there is a
strong indication that complying with
BCT/BAT will sufficiently mitigate the
impacts, EPA recommends that
applicable permits include testing
requirements effective after BCT/BAT
compliance and reopener clauses
allowing reevaluation of the discharge.
The chemical, physical, and biological
testing to be conducted by individual
dischargers should be determined on a
case-by-case basis. In making this
determination, many factors must be
considered, including the degree of
impact, the complexity and variability of
the discharge, the water body type and
hydrology, the potential for human
health impact, the amount of existing
data, the level of certainty desired in the
water quality assessment, other sources
of pollutants, and the ecology of the
receiving water. The specific data
needed to measure the effect that a
discharger has on the receiving water
will vary according to these and other
factors,
B-2-3
An assessment of water quality
should, to the extent practicable, include
other point and nonpoint sources of
pollutants if the sources may be
contributing to the impacts. Special
attention should be focused on Publicly
Owned Treatment Works (POTW's)
with a significant contribution of
industrial waste-water. Recent studies
have indicated that such POTW's are
often significant sources of toxic
materials. When developing monitoring
requirements, interpreting data, and
determining limitations, permit
engineers should work closely with
water quality staff at both the State and
Federal levels.
A discharger may be required to
provide data upon request under section
308 of the Act, or such a requirement
may be included in its NPDES permit.
The development of a final assessment
may require several iterations of data
collection. Where potential problems are
identified, EPA or the State may require
monitoring to determine whether more
information is needed concerning water
quality effects.
Use of Data
Chemical, physical, and biological
data will be used to determine whether
after compliance with BCT/BAT
requirements, there will be violations of
State water quality standards resulting
from the discharge(s). The narrative
prohibition of toxic materials in toxic
amounts contained in all State
standards is the basis for this
determination taking into account the
designated use for the receiving water.
For example, discharges to waters
classified for propagation of cold water
fish should be evaluated in relation to
acute and chronic effects on cold water
organisms, potential spawning areas,
and effluent dispersion.
Setting Permit Limitations
Where violations of water, quality
standards exist or are projected, the
State and EPA will determine pollution
control requirements that will attain the
receiving water designated use. Where
effluent toxicity is an appropriate
control parameter, permit limits on
effluent toxicity should be developed. In
such cases, EPA may also fequire a
permittee to conduct a toxicity reduction
evaluation. A toxicity reduction
evaluation is an investigation conducted
within a plant or municipal system to
isolate the sources of effluent toxicity,
specific causative pollutants if possible,
and determine the effectiveness of
pollution control options in reducing the
effluent toxicity. If specific chemicals
are identified as the cause of the water
-------
Federal Register / Vol. 49, No. 48 / Friday, March 9, 1984 / 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
ore expressed in terms of a chemical or
lexicological 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 Jack E.
Ravan, Assistant administrator for Water.
Dated: February 16,19R4.
Jack E. Ravan,
Assistant Administrator for Water.
|FR Doc M-W4S Filfcd 3-S-W; SMS am]
BlUma CODE 8580-SO-H
B-2-4)
-------
APPENDIX B-3
REGULATIONS FOR Toxics CONTROL
-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON. D.C. 20460
AUG 2 I 1989 OFWF£EEROF
MEMORANDUM
SUBJECT: New Regulations Governing Water Quality-Based
Permitting in the NPDES Permitting Program
,^;^^-
FROM: J ame*^C/*ldeV;Direct or
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.
B-3-1
-------
1. §122.44(d)(l)(i)
i
This new paragraph provides t:hat 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.
i
2. §122.44(d)(l)(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 nuflieric water quality criteria;
subparagraphs (v) and (vi) deal wTth a state's narrative water
quaity criteria.
3. §122.44(d)(l)(iii)
This paragraph requires NPDES permits to include effluent
limitations for every individual pollutant that causes, has the
reasonable potential to cause, or pontributes 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)(l)(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
B-3-2
-------
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)(l)(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.
B-3-3
-------
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.P.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
B-3-4
-------
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 requirements,
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
B-3-5
-------
I
23868 Federal Register / Vol. 54, No. 105 / Friday, June 2, 1989 / Rules and Regulations
ENVIRONMENTAL PROTECTION
AGENCY
[FRL-3557-6]
40 CFR Parts 122,123 and 130
National Pollutant Discharge
Elimination System; Surface Water
Toxics Control Program
AGENCY: Environmental Protection
Agency.
ACTION; Final Rule.
SUMMARY: Today's action amends Parts
122,123, and 130 of EPA's regulations.
The regulations clarify EPA's surface
water toxics control program, and
incorporate section 308(a) of the Water
Quality Act of 1987 into EPA's toxics
control program. Section 308(a) of the
Water Quality Act added section 304(1)
to the Clean Water Act (hereafter
referred to as section 304(1]). Section
304(1) requires the states to identify
those waters that are adversely affected
by toxic, conventional, and
nonconventional pollutants, and
requires the states to prepare individual
control strategies that will control point
source discharges of toxic pollutants.
The states must submit lists of waters
and individual control strategies to EPA
for review, and if EPA disapproves a
slate's decision with respect to a list or
an individual control strategy, then EPA
must implement the requirements of
section 304(1) in cooperation with the
state. EPA and the states must
accomplish the tasks in section 304(1)
according to an ambitious series of
deadlines. Today's regulations will
strengthen State and Federal controls
over discharges to toxic pollutants, and
will assist EPA and the states in
satisfying the requirements of section
304(1) of the CWA.
EFFECTIVE DATE: These regulations shall
be effective on May 26,1989 at 1:00 p.m.
Eastern Daylight Savings Time. In
accordance with 40 CFR 23.2, EPA
hereby specifies that these regulations
shall be considered final agency action
for purposes of judicial review at 1:00
p.m. Eastern Daylight Savings Time on
May 26,1989.
FOR FURTHER INFORMATION CONTACT:
Paul Connor, Program Development
Branch, Office of Water Enforcement
and Permits, (EN-336), U.S.
Environmental Protection Agency, 401M
Street, SW., Washington, DC 20460,
(202) 475-9537, or Judith Leckrone,
Assessment and Watershed Protection
Division, Office of Water Regulations
and Standards, (WH-55.3), U.S.
Environmental Protection Agency, 401M
Street SW., Washington, DC 20460, (202)
382-7056. The Public record for this
regulation is available at the EPA
library, M2904, U.S. Environmental
Protection Agency, 401M Street SW.,
Washington, DC 20460.
B-3-6
-------
Federal Register / Vol. 54, No
1989 / Rules and Regulations 2SS95
PART 122—EPA ADMINISTERED
PERMIT PROGRAMS: THE NATIONAL
POLLUTANT DISCHARGE
ELIMINATION SYSTEM
1. The authority citation for Part 122
continues to read as follows:
Authority: The Clean Water Act, 33 U.S.C.
1251 et seq.
2. Section 122.2 is amended by adding
in alphabetical order a new definition as
follows:
§122.2 Definitions.
• * * * * *
Whole effluent toxicity means the
aggregate toxic effect of an effluent
measured directly by a toxicity test.
3. Paragraph (d)(l) of § 122.44 is
revised to read as follows:
§ 122.44 Establishing limitations,
standards, and other permit conditions
(applicable to State NPDES programs, see
§ 123.25).
*****
(d) * * *
(1) Achieve water quality standards
established under section 303 of the
CWA, including State narrative criteria
for water quality. ,
(i) Limitations must control all
pollutants or pollutant parameters
(either conventional, nonconventional,
or toxic pollutants) which the Director
determines are or may be discharged at
a level which will cause, have the
reasonable potential to cause, or
contribute to an excursion above any
State water quality standard, including
State narrative criteria for water quality.
(ii) When determining whether a
discharge 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 excursion 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
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 excursion above the numeric
criterion 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, or
contributes to an in-stream excursion
above a narrative criterion within an
applicable State water quality standard,
the permit must contain effluent limits
for whole effluent toxicity. Limits on
whole effluent toxicity are not necessary
where the permitting authority
demonstrates in the fact sheet or
statement of basis of the NPDES permit,
using the procedures in paragraph
(d)(l)(ii) of this section, that chemical-
specific limits for the effluent are
sufficient to attain and maintain
applicable numeric and narrative State
water quality standards.
(vi) Where 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
contributes to an excursion above a
narrative criterion within an applicable
State water quality standard, the
B-3-7
permitting authority must establish
effluent touts using 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 1983, 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
by the use of the effluent limitation;
(2) The fact sheet required by § 124.56
sets forth the basis for the limit,
including a finding that compliance with
the effluent limit on the indicator
parameter will result in controls on the
pollutant of concern which are sufficient
to attain and maintain applicable water
quality standards; •
(3) The permit requires all effluent and
ambient monitoring necessary to show
that during the term of the permit the
limit on the indicator parameter
continues to attain and maintain
applicable water quality standards; and
(4) The permit contains a reopener
clause allowing the permitting authority
to modify or revoke and reissue the
permit if the limits on the 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
ensure 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
-------
23896 Federal Register / Vol. 54, No. 105 / Friday, June 2. 1989 / Rules and Regulations
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:
§123.44 EPA review of and objections to
State permits.
*****
(c) * ' *
(8) The effluent limits of a permit fail
to satisfy the requirements of 40 CFR
122.44(d).
*****
3. In § 123.46 paragraph (a) is revised
and paragraphs (c), (d), (e) and (f) are
added, as follows:
§ 123.46 Individual control 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 pursuant 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 years 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 3040) 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
issue a final permit on or before
February 4,1989, an individual control
strategy may be a draft permit with an
attached schedule (provided the State
meets the schedule for issuing the final
permit) indicating that the permit will be
issued on or before February 4,1990. If a
point source is subject to (section
304(1)(1)(G) of the CWA and is also
subject to an on-site response action
under sections 104 or 108 of the
Comprehensive Environmental
Response, Compensation, and Liability
Act of 1980 (CERCLA), (421 U.S.C. 9601 et
seq.), an individual control strategy may
be the decision document (which
incorporates the applicable or relevant
and appropriate requirements under the
CWA) prepared under sections 104 or
106 of CERCLA to address the release or
threatened release of hazardous
substances to the environment.
(d) A petition submitted pursuant to
section 304(1)(3) of the CWA must be
submitted to the appropriate Regional
Administrator. Petitions must identify a
waterbody in sufficient detail so that
EPA is able to determine the location
and boundaries of the waterbody. The
petition must also identify the list or
lists for which the waterbody qualifies,
and the petition must explain why the
waterbody satisfies the criteria for
listing under CWA section 304(1) and 40
CFR130.10(d)(6).
(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 ok the ICSs,
then, not later than June 4,1989, 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:
(1) The notice of approval or
disapproval given under this paragraph
shall include the following:
(i) 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 ICSs disapproved under
this section and a finding that the ICSs
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 discretion, a list of the ICSs
approved under this section, and a
B-3-8
finding that the ICSs satisfy all
applicable review criteria.
(v) The location where interested
persons 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 shall also
provide written notice to each
discharger identified under section
304(1)(1)(C). that EPA has listed the
discharger under section 304(1)(1)(C).
(3) As soon as practicable but not
later than June 4,1990, the Regional
Offices shall issue 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 described in
paragraph (e) of this section except for
the following changes:
(i) The lists of ICSs reflecting any
changes made pursuant to comments or
petitions received.
(ii) A brief description of the
subsequent steps in the section 304(1)
process.
(f) EPA shall review, and approve or
disapprove, the individual control
strategies prepared under section 304(1)
of the CWA, using the applicable
criteria set forth in section 304(1) of the
CWA, and in 40 CFR Part 122, including
§ 122.44(d). At any time after the
Regional Administrator disapproves an
ICS (or conditionally aproves a draft
permit as an 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 passes to EPA.
4. Section 123.63 is amended by
adding paragraph (a)(5) to read as
follows:
§123,63 Criteria for withdrawal of state
programs.
(a) * * *
(5) Where the State fails to develop an
adequate regulatory program for
developing water quality-based effluent
limits in NPDES permits.
-------
APPENDIX B-4
WHOLE EFFLUENT TOXICITY PERMITTING PRINCIPLES
AND ENFORCEMENT STRATEGY
-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON. D.C. 20460
January 25, 1989
OFFICE OF
WATER
MEMORANDUM
SUBJECT: _Whole Effluent Toxicity Basic Permitting Principles and
( Enforcement Strategy
P HJ\a£^lL£-0\— HlXoo v^r^e^-
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
B-4-1
-------
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. i
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.
I1
If you have any questions on the attached materials, please
contact James Elder, Director of tlhe Office of Water Enforcement
and Permits, at (FTS/202) 475-8488.
i
Attachments
cc: ASWIPCA
Water Management Division Directors
B-4-2
-------
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
schedules 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.
B-4-3
-------
ATTACHMENT 1
I
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 pe'rmits 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, c-.d the~.=jfo,re
possible whole-effluent limits, as all majors and minors of
concern.
I,
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 monitoring 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.
B-4-4
-------
- 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.
B-4-5
-------
ATTACHMENT 2
Background to the Compliance Monitoring and Enforcement
for Toxics Control
The Compliance Monitoring and Enforcement Strategy for
Toxics 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 pf 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 tlhe 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.
B-4-6
-------
01/19/89
COMPLIANCE MONITORING AND ENFORCEMENT STRATEGY
FOR TOXICS CONTROL
I. Background
Issuance of NPDES permits now emphasizes 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
B-4-7
-------
- 2 -
chemical-specific water quality-basjed 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
™""™"™"^™^™"'"^^^^^""^™™^^~™^^"*^™™^^™™™^^™ i
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 pompliance through inspections,
audits, discharger data reviews, and other independent
monitoring or review activities.
I1
4) Regulators will enforce effluent limits and compliance
schedules to eliminate toxicity.
III. Primary Implementation Activities
In order to implement this Strategy fully, the following
activities are being initiated:
i-
A. Immediate development
1. The NPDES Compliance Inspection Manual was
revised ir. May 1988 to include procedures for
performing chronic toxicity tests and evaluating
toxicity reduction evaluations. An inspector
training module was also developed in August
1988 to support inspections for whole 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
B-4-8
-------
•"• J ~
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 provide 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.
B-4-9
-------
- 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 iipnediately 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.
I
B. Inspections
I
EPA/State compliance inspections of all major permittees
on an annual basis will be maintained. For all facilities
with water quality-based toxib 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.
B-4-10
-------
- 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.
B-4-11
-------
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 permittee laboratory performance on EPA
and/or State approved methods. 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.
B-4-12
-------
- 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.
D. 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.
B-4-13
-------
- 8 •
The following milestones are considered SNC when
90 days or more overdue: submit 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
monitoring 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
B-4-14
-------
- 9 -
administrative penalty should be considered.1
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 potential
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 plan2, 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.
2See 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 identif ication/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).
industrial permittees, the facility must be well-operated
to achieve all water quality-based, chemical specific, or BAT
limits, exhibit proper O & 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.
B-4-15
-------
- 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
I
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.
B-4-16
-------
- 11 -
V. Summary of Principal Activities and Products
A. Compliance Tracking and Review guidance
1. PXCS 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
" icting
- 2nd
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
FY 1989
B-4-17
-------
- 12 -
3. Methods for Aquatic Toxicity Indentification
Evaluations
Phase I.
b. Phase II,
Phase III
Tbxicity Characterization
Procedures, EPA-600/3-88/034-
September 1988
Toxicity Identification
Procedures, EPA-600/3-88/035-
2nd Quarter 1989
Toxicity Confirmation Procedures-
EPA-600/3-88/036 - 2nd Quarter
^Y 1989
B-4-18
-------
APPENDIX B-5
QUALITY CONTROL FACT SHEETS
-------
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
LCso/EC5o/NOEC/IC25 95 Percent Confidence Interval
Quality Control Summary
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 percent 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
B-5-1
-------
APPENDIX B-6
CASE DECISIONS ON WHOLE EFFLUENT TOXICITY
-------
CASE SUMMARY
Natural Resources Defense Council. Inc. v. EPA. 859 F.2ii 156 (D.C. Cir. 1988).
This consolidated case, which arose from EPA's promulgation of various National Pollutant Discharge
Elimination System regulations, addresses a multitude of issues. The following paragraphs note issues
particularly relevant to this document.
• The Court held that EPA has the authority to express permit limitations in terms of toxicity as long as
the limits reflect the appropriate requirements of the Clean Water Act (CWA), as provided in 40 CFR
125.3(c)(4). The Court concluded that although toxicity appears to be an attribute of pollutants
rather than a pollutant itself, the CWA (by means of the broad definition of "pollutant" in section
502(6)) authorizes the use of toxicity to regulate effluents.
• Industry asked the Court to address several other issues related to setting toxicity limitations (whether
EPA failed to demonstrate the existence of a reliable methodology for setting toxicity limits and
whether EPA's use of toxicity to set water quality-based limitations to meet "narrative" State water
quality standards represents an impermissible trespass on the State's right to set water quality
standards). However, the Court did not regard these issues to be adequately developed ("ripe") for
review in this case.
• 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 Federal Register
9016 [March 9, 1984]) and draft Technical Support Document ("TSD") were "rules" requiring notice and
comment under the Administrative Procedure Act, 5 DSC 553. The Court noted that informal
rulemaking regarding 40 CFR 125.3(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, and
the two documents did not have independent legal value. (In other words, the EPA national policy
and the TSD were not binding norms but general statements of policy/guidance.)
• Industry also challenged EPA's refusal to provide an affirmative upset defense to noncompliance with
water quality-based limits. The Court indicated that the CWA does not expressly allow such an upset
defense, and, upon considering the Act's structure and legislative history, it could discern no
congressional intent to provide for the defense in water quality permitting. Significantly, in
reaching this position, the Court relied heavily upon the language and legislative history of CWA
Section 301(b)(1)(C), by which Congress clearly did not relate compliance with water quality-based
limitations to the capabilities of technology. In the Court's view, "Congress had a deep respect for
the sanctity of water quality standards and a firm conviction of the need for technology-forcing
measures." 895 F.2d at 208-09. However, the Court concluded that EPA had acted arbitrarily in
dismissing the defense as impracticable, and directed EPA to conduct further proceedings on the issue.
• Finally, the Court rejected challenges to EPA's regulations governing State public participation
requirements and penalty levels. In deciding these issues, the Court noted Congressional desire for
nationally uniform effluent limitations as reflected in the legislative history of the 1972 CWA. The
Court stated:
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 (859 F.2d at
174 [footnotes omitted] and see accompanying footnotes 17-20 citing various provisions of the
legislative history of the 1972 CWA).
B-6-1
-------
APPENDIX C
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:
• Select instream sampling stations based on the mixing characteristics involved in the specific
discharge situation.
• Collect a daily grab sample or a daily composite sample of receiving water from each station.
• 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.
• Conduct testing at a low-flow period, although 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.
• Record the results of the testing in the format shown in Table C-1. The survival of the test organisms
and the effect on their growth or reproduction are used as endpoints. Figure C-1 plots the results in
graphic form so that the pattern of ambient toxicity can be observed.
C-1
-------
Table C-1. Young Production and Percent Survival of Ceriodaphnia in Ambient
Toxicity Tests at Ottawa River, Lima, Ohio
Station
1
2
3
3A
3B
4
4A
5
6
7
8
8A
9
Station Description
Above Lima
Above STP
Below STP
Midway between
STP and refinery
Above refinery
Above chemical plant
Below chemical plant
Shawnee Bridge
Route 1 1 7
Allentown
Rimer
"Boonie" Statibn
Kalida
River
Mile
46.0
37.7
37.4
37.3
37.1
36.9
36.3
36.4
32.5
28.8
16.0
8.0
1.0
Young
Female
15.5
14.1
0
0
0.4
7.5
11.1
5.7
12.6
16.8
17.4
25.0
25.6
Final
SD Survival 1
8.0 <
2.1
"
>0 100
0 100
0 100
0 100
0 90
3.6 10 100
4.6 i
4.0
10 100
0 90
3.8 10 100
6.1 100 100
9.5 100 90
3.3 100 100
5.5 100 100
2
100
100
100
100
90
100
100
90
TOO
100
90
100
100
Dailv Survival
3
100
100
10
10
40
100
100
90
100
100
90
100
100
4
90
100
0
0
0
100
100
90
100
100
90
100
100
5
90
90
0
0
0
100
100
90
100
100
90
100
100
6
90
10
0
0
0
50
40
60
100
100
80
100
100
7
90
0
0
0
0
10
30
0
10
100
80
100
100
25
£
a 15-
I .
•s
80
65
45
River Kilometers
32
16
Figure C-1. Ceriodaphnia Young Production in Water from Various Stream
Stations on the Ottawa River, Lima, Ohio
C-2
-------
Selecting Sampling Stations
The selection of sampling stations is determined by the characteristics of the site. When determining stations,
consider the following factors:
• 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 no
observed effect concentration is 10 percent, an instream station should be placed where dilution is
estimated to create a 10-percent 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
nonpoint source areas also should be used to set stations.
• 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.
• 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.
• 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:
Zone 1 —An upstream zone before the effluent enters
Zone 2—A zone of mixing
Zone 3—A zone after mixing and before additional dilution water enters
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:
• 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.
• 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.
• 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.
• 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.
C-3
-------
Analysis of Ambient Toxicity Measurement
• 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.
• Except when used for screening purposes, ambient toxicity measurements must be interpreted 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.
• 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.
• 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.
• Where the toxic effect decreases from station to station downstream in the absence of further dilution,
the effluent toxicity is degrading. If the decay rcite is rapid (e.g., no toxicity at the closest instream
station to the discharge point), the effluent has a nonpersistent 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. |
• 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.
• Trend analysis in the raw test data is important When interpreting 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 that can be linked to discharge events. For
example, in the table 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.
C-4
-------
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 criterion maximum concentration (CMC) and the criterion continuous concentration (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 TSD, the CMC and CCC for Whole Effluents can be specified generically in terms of toxic
units. Alternatively, for a particular effluent the CMC is specified in terms of an acute toxicity endpoint (ATE),
which is either an LC5Q or an EC5Q, and the CCC is specified in terms of a chronic toxicity endpoint (CTE),
which is either a no observed effect concentration (NOEC) or an ICxx/ if the LC5Q, EC5Q, 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 exceedances 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 exceedance 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 exceedances becoming increasingly rare [1, 2].
Although an exceedance 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 for exceedances, excursions that are
defined to be due to usual variation will be small, with larger excursions becoming increasingly rare. The
duration of the averaging period is intended to limit the impact of exceedances, whereas the average frequency
of allowed excursions is intended to limit the impact of excursions. (Note: The words "exceedance" and
"excursion" are used slightly differently here than in References 1 and 2.)
Although spills can impact aquatic communities, they are not considered exceedances or excursions because
they are not part of the usual variation in the concentrations of pollutants in receiving water. 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. Ambient toxicity measurements were taken over a 7-day period.
D-1
-------
During two of these studies [4, 5] spills of pollutants resujted 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.
•
i
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 brovided, 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 rationales that might be used as a
basis for selecting 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 exceedances above the CCC without allowing unacceptable effects. Thus, the averaging period
must appropriately limit the magnitude and duration of exceedances 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 7 days. If the duration of the averaging; period is too short, however, it will not allow any
meaningful exceedances 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:
• 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 Ceriodaphriia life-cycle test.
• The results of some chronic tests apparently are due to an acute effect on a sensitive life stage that
occurs at some time during the test, rather than jbeing 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 are probably the cause of at least some of the acute-chronic ratios that are not
much greater than unity.
• For both endrin and fenvalerate, Jarvinen et al. [| 8] found that a 72-hour exposure caused about the
same amount of effect on the growth of fathegd minnows in early life-stage tests as did a 30-day
exposure to the same concentration.
• 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 4 or 5 days [5, 19, 20].
• 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.
D-2
-------
Just as the concept of exceedances can be applied to the CCC, it also can 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-hour LCsQs are about the same as the 96-hour
LC5Q for some materials [21-27], the duration of the averaging period for the CMC should 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-hour LC5Q 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 1-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 CMC averaging would be acceptable
because the magnitudes and durations of exceedances above the CCC would be limited. A possible pollutant-
specific justification for a longer averaging period would be that the LC5Q decreases substantially as the
length of the exposure increases. For example, an 8-hour averaging period might be justified for the CMC if it
were shown that 24-hour exposures of a variety of sensitive species resulted in 96-hour LCsQs that were
substantially above the 96-hour LCSQs obtained from continuous exposure to a constant concentration for 96
hours.
In some situations the duration of the averaging period does not have to be stated explicitly because one can
be implicitly defined using an uptake rate and a depuration rate. For example, if it is known that a specific
concentration of a pollutant in the whole body or in a particular tissue of an important aquatic species will
result in an unacceptable effect on the survival, growth, and/or reproduction of that species, and if applicable
that species or tissue, the only additional information needed to allow calculation of an excessively high
estimate of the total maximum daily load from the record of daily flows is the allowed frequency of
exceedances of the concentration in the aquatic species. Thus, this approach can be used whenever the
following are available: '
• A record of daily flows of the body of water, preferably for more than 10 years
• A maximum acceptable concentration in the whole body or in a particular tissue of an aquatic species
• Uptake and depuration rates that are applicable to that pollutant in the whole body or tissue of that
species
• An allowed frequency of exceedances of the maximum acceptable concentration.
This approach is likely to be especially useful when an exposure causes delayed effects that are considered
unacceptable. For example, it might be found in a test that no fish die during a 2-day exposure of rainbow trout
to a pollutant but 50 percent of the fish die within 4 weeks of being transferred to clean water, whereas no
comparable control fish die. If values are available for the concentration of the pollutant in the fish at the
end of the 2-day exposure and for the uptake and depuration rates, these data could be used with a flow record
for a river to determine how often a specified constant daily input of the pollutant to the river would have
resulted in exceedances of this concentration and therefore the death of rainbow trout.
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 1-hour averaging period for the CMC would imply that the samples analyzed
should be 1-hour composites; the 4-day averaging period would imply that concentrations in all samples
obtained within any 4-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.
D-3
-------
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 segment 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 "rr)inor" 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 recover from the effect of an excursion and then 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, reservoirs, and estuaries.
General Considerations for Setting Frequency with Which Excursions of Criteria May Occur
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 affects 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 also are 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 a\. [38] reviewed the published literature and
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 percent of these systems were lotic, and the remainder were lentic (lakes and ponds). The
impacts were due to such disturbances as persistent and nonpersistent 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 (nonpersistent) disturbances, approximately 85 percent of all macroinvertebrate
endpoints indicated recovery in less than 2 years. Macroinvertebrate biomass, density, and taxonomic richness
recovered in less than 1 year for approximately 95 percent |of reported endpoints. Dipterans (flies, mosquitos,
midges, etc.), which generally have short generation times or high dispersal ability, recovered most rapidly,
whereas stoneflies and caddisflies recovered least rapidly. Fishes recovered in 2 years or less for over 85 percent
of reported endpoints. However, as discussed below, important exceptions did occur.
i
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 irtipacts than most criteria excursions are expected to
-------
cause, CMC excursions will result in death of some organisms. 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 every 3
years on the average. Excursions of the CCC are more difficult to evaluate because nonlethal excursions could
not be evaluated from the data used by Niemi et al. [38]. It is reasonable to expect, however, that cumulative
effects from too frequent excursion of the CCC also will 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 3 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 when criteria excursions also are 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 3 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, may be able to tolerate somewhat higher excursion
frequencies, 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 streams for at least two partially related reasons: (1)
segments 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 segments will therefore not be adapted to disturbances that are as frequent or severe
as those in lower-order segments. 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 major drought or spill.
Extreme cases in which recovery has taken much longer than 3 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
segment, will affect aquatic life over a large area. If maintenance of long-lived fish species in these segments
is desired, recovery periods up to 25 years may be necessary.
Based on the above considerations, recovery periods longer than 3 years may be necessary after multiple minor
excursions or after a single major excursion or spill during a low-flow period in medium-to-large rivers, and up
to 25 years where long-lived fish species are to be protected. Even longer times may be necessary as the size of
the affected area or the persistence of the pollutant increases.
D-5
-------
Calculation of Design Conditions
I
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 calculating the design flow. The
procedures outlined in the EPA document Technical Cuidand? Manual for Performing Waste Load Allocation, Book 6,
Design Conditions: Chapter 1, Steam Design Flow for Steady-State Modeling. (U.S. EPA 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 presently used by many States. It is based on selecting and
identifying 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 two 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, and (2) the U.S.
Geological Survey 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 3 years on the aVerage. In addition, it is difficult to use site-specific
durations and frequencies with this method. For toxic wasteload allocation 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 U.S. EPA Office of Research and
Development and directly uses the averaging periods and frequencies specified in the aquatic-life water quality
criteria for individual pollutants and Whole Effluents for determining design flows. The method is an
empirical iterative convergence procedure that includes tfie 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 1 day and 3 years for the CMC and 4 days and 3 years for the
CCC or site-specific durations and frequencies.
The two methods were used on approximately 60 different rivers to compare the hydrologically based 1Q10 and
7Q10 design flows with the biologically based l-day/3-year and 4-day/3-year design flows. For most of the
rivers the hydrologically based design flows resulted in more than the allowed number of excursions. For some
of the rivers, 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 [40]. The
hydrologically based design flows can be calculated using FLOSTAT or DFLOW; the latter uses a simplified
version of the log-Pearson Type III method. Both programs access the STORET Flow file that contains daily
flow records for U.S. Geological Survey 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. Thomas Pandolfi
U.S. Environmental Protection Agency
Office of Water Regulations and Standards (WH-553)
401 M Street, S.W.
Washington, D.C. 20460
(202) 382-7030
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 named DESCON that incorporate such supplemental water quality parameters as temperature, pH,
alkalinity, hardness, and dissolved oxygen to determine design conditions. Note that DESCON takes into
account such things as effluent variability, which DFLOW does not take into account. 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—Technical Guidance on Supplementary Stream Design Conditions for Steady
State Modeling [3] and DESCON Users Manual [40].
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 worst-case conditions necessary to protect water quality criteria might not occur
when the streamflow is low; 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), pentachlorophenol, 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.
D-7
-------
APPENDIX D
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 Perfprming Waste Load Allocation. Book 6. 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.
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. /. 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. /. Toxicol. Environ. Health 3:821-28.
8. Smith, L.L., Jr., S.J. 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 (5. clarki). Wat. Res.
15:911-17.
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 I. Effect on Growth, on Accumulation and
Distribution of Copper, and on Copper Tolerance. Comp. Biochem. Physiol. 72O15-19.
11. Ingersoll, C.G., and R.W. Winner. 1982. Effect on Dqphnia pulex (De Geer) of Daily Pulse Exposures to
Copper or Cadmium. Environ. Toxicol. Chem. 1:321-27.
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 |(So/mo gairdneri). Environ. Toxicol. Chem. 2:225-38.
13. Seim, W.K., L.R. Curtis, S.W. Glenn, and G.A. Chap|man. 1984. Growth and Survival of Developing
Steelhead Trout (Salmo gairdneri) Continuously or Intermittently Exposed to Copper. Can. J. Fish. Aquat.
Scl. 41:433-38.
14. Curtis, L.R., W.K. Seim, and G.A. Chapman. 1985. Toxicity of Fenvalerate to Developing Steelhead
Trout Following Continuous or Intermittent Exposure. /. Toxicol. Environ. Health 15:445-57.
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 fontinalis). Can. ]. Fish.
Aquat. Sci. 43:2036-40.
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 Ch/orine. Center for Great Lake studies, University of
Wisconsin-Milwaukee, Wl.
17. Horning, W.B., and T.W. Neiheisel. 1979. Chronic Effect of Copper on the Bluntnose Minnow,
Pimephales notatus (Rafinesque). Arch. Environ. Contam. Toxicol. 8:545-52.
18. Jarvinen, A.W., O.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.
D-8
-------
19. Mount, D.I., A.E. Steen, and T.J. 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.C. Foreman, T.R. Payne, and D.J. 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 (So/mo gairdneri) Resulting from Reduced Concentrations of Dissolved Oxygen. Can. ].
Fish. Aquat. Sd. 38:983-88.
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 1. Center for Lake Superior Environmental Studies,
University of Wisconsin-Superior, Superior, Wl.
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 2. Center for Lake Superior Environmental Studies,
University of Wisconsin-Superior, Superior, Wl.
25. Bailey, H.C., D.H.W. Liu, and H.A. javitz. 1985. "Time/Toxicity Relationships in Short-Term Static,
Dynamic, and Plug-Flow Bioassays." In Aquatic Toxicology and Hazard Assessment: Eighth Symposium.
ASTM STP 891. Ed. R.C. Bahner and D.J. Hansen. American Society for Testing and Materials,
Philadelphia, PA.
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 promelas), volume 3. Center for Lake Superior Environmental Studies,
University of Wisconsin-Superior, Superior, Wl.
27. Geiger, D.L., D.J. Call, and L.T. Brooke. 1988. Acute Toxicities of Organic Chemicals to Fathead Minnows
(Pimephales promelas), volume 4. Center for Lake Superior Environmental Studies, Univerisity of
Wisconsin-Superior, Superior, Wl.
28. Abel, P.O. 1980. Toxicity of Hexachlorocyclohexane (Lindane) to Cammarus pulex: Mortality in
Relation to Concentration and Duration of Exposure. Freshwater Biol. 10:251-59.
29. Abel, P.O. 1980. A New Method for Assessing the Lethal Impact of Short-Term, High-Level Discharges
of Pollutants on Aquatic Animals. Prog. Wat. Tech. 13:347-52.
30. Abel, P.O., and S.M. Garner. 1986. Comparisons of Median Survival Times and Median Lethal Exposure
Times for Cammarus pulex Exposed to Cadmium, Permethrin and Cyanide. Wat. Res. 20:579-82.
31. Heming, T.A., A. Sharma, and Y. Kumar. 1989. Time-Toxicity Relationships in Fish Exposed to the
Organochlroinepesticide Methoxychlor. Environ. Toxicol. Chem. 8:923-32.
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. ]. North Am. Benthol.
Soc. 7:433-55.
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." Ed. J.D. Yount
and G.J. Niemi. In 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. Ed. J.D. Yount and
G.j. Niemi. In 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." Ed. J.D. Yount and
G.J. Niemi. In 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." Ed. J.D. Yount and G.J.
Niemi. In 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." Ed. J.D. Yount and G.J. Niemi. In
D-9
-------
Recovery of Lotic Communities and Ecosystems from Disturbance: Theory and Applications. Environ.
Management (submitted).
38. Niemi, G.J., P. DeVore, N. Detenbeck, D. Taylor, ].Di Yount, A. Lima, J. Pastor, and R.J. Naiman. 1989.
"An Overview of Case Studies on Recovery of Aquatic Systems from Disturbance." Ed. J.D. Yount and
G.J. Niemi. in 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 ).A. Stanford. 1989. "The Role of Refugia in Recovery from
Disturbance: Modern Fragmented and Disconnected River Systems." Ed. J.D. Yount and G.J. Niemi. In
Recovery of Lotic Communties and Ecosystems from Disturbance: Theory and Applications. Environ.
Management (submitted).
40. Rossman, Lewis A. DFLOW Users Manual.
D-10
-------
APPENDIX E
LOGNORMAL DISTRIBUTION AND PERMIT LIMIT
DERIVATIONS
-------
LOGNORMAL DISTRIBUTION AND PERMIT LIMIT
DERIVATIONS
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 technology-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 E-1, E-2, and E-3 at the end of the appendix summarize
the permit limit calculations. This appendix describes the statistical methodology for three distributions that
are often used in determining permit limits. Other distributions can be used, and this topic is discussed in the
subsection, Other Distributions.
Section 1: Overview of Permit Limits
Two types of permit limits are contained in the effluent guidelines regulations: daily maximum limits and
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 monthly average permit limit is the maximum allowable value for the average of all daily samples obtained
during 1 month. Monthly average limits are in most cases based on the 95th percentile of the distribution of
averages of daily values.
The following two subsections discuss the derivation of water quality-based limits and the consistency among
different permit limits.
Derivation of Water Quality-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 Chapter 5.
Consistency Among Different Permit Limits
The current Technical Support Document for Water Quality-based Toxics Control (TSD) procedures provide
consistency among different permit limits. The stringency of permit limits is independent of monitoring
frequency and is determined entirely by the 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 based on 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 permittee performing according to the WLA
specifications will in fact be equally capable of meeting either of these monthly average limits when taking
the corresponding number of samples. The stringency of the TSD procedures, accordingly, is constant across
monitoring frequencies.
E-1
-------
Section 2: Statistical Methodology
The statistical procedures that 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 subsections: 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.
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 log-[ g(x) for ln(x) in the
formulas. i
I
Assumptions
The distribution fitting methods assume that the daily measurements are independent, uncorrelated
observations.
The fundamental assumptions underlying the discussion on calculating 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.
Recommendation of the use of the lognormal distribution for daily pollutant measurements is based on
practical rather than theoretical consideration. Usually environmental data sets possess the basic lognormal
characteristics of positive values and positive skewness. In addition, the lognormal distribution is flexible
enough to model a range of nearly symmetric data. Furthermore, in comparison to other positive valued,
positively skewed distributions that could be used to motdel environmental data, the lognormal is relatively
easy to use.
When lognormal data are log transformed, the properties of the normal distribution apply to the transformed
data. The section on statistical methodology describes the properties of the normal distribution and its
relationship to the lognormal distribution. The delta-lognormal distribution is a generalization of the
lognormal distribution and may be used to model data that are a mixture of non-detect measurements with
measurements that are lognormally distributed. In delta-lognormal procedures, nondetect values are weighted
in proportion to their occurrence in the data.
In determining permit limits based on averages (e.g., monthly average permit limits), a distribution should be
used that approximates the distribution of an average of pollutant measurements. The lognormal distribution
can be used for approximating the distribution of averages for small sample sizes where the individual
measurements are approximately lognormally distributed. (For larger sample sizes, 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. The section on statistical methodology provides procedures and guidance for
calculating averages for both small and large samples sizes where the individual measurements are lognormally
distributed.
E-2
-------
The shape of the observed data is the key factor in evaluating a distributional model. For environmental data
the lognormal distribution is usually appropriate. The critical question in a given situation is how well a
particular distribution models the shape of observed data. Although the lognormal does not provide an exact
fit in all cases, it usually provides an appropriate and functional fit to observed environmental data. Graphical
displays and goodness-of-fit tests, as described in the subsection, Other Distributions, may be used as a guide in
verifying assumptions and selecting a distribution.
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 data sets 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. The CV is a measure of
the relative variation in observed data. 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 that 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 the Environmental Protection Agency (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
occasions 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 seven daily observations from a plant, with at least three of the seven 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 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. In most cases, the
normal distribution is not an appropriate model for individual pollutant measurements; however, the normal
distribution is related to the lognormal distribution that is used to establish many permit limits. In most cases,
the simple logarithmic transformation of effluent and water quality data results in data distributions that are
normally distributed. Such data are referred to as being lognormally distributed. When lognormal data are log
transformed, the properties of the normal distribution apply to the transformed data. Since the normal and
lognormal distributions are related in a straightforward manner, the methods of analysis for normal and
lognormal data also are easily related. The normal distribution is described below and is followed by a
discussion of the lognormal distribution and its relationship to the normal distribution.
E-3
-------
Figure E-1. Normal Probability Distribution
The normal probability distribution is encountered in a number of applications. The bell-shaped curve of the
normal distribution is shown above in Figure E-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.
A sample of independent observations, denoted by XT, X2,...,X|<, from a normally distributed population can be
used to estimate the mean, u., and variance, o2' according tb the formulas below:
fi. = estimated mean
= S[xj] / k, 1 £ i < k
Cf2 _ estimated variance
a = estimated standard deviation
cv
= estimated coefficient of variation
A , A
= a/u.
The characteristics of the normal distribution are the range is defined for positive and negative values, and the
frequency curve is bell-shaped and symmetric about the mean. In most cases, the normal distribution is not an
appropriate model for the distribution of individual pollutant measurements. Environmental data rarely are
symmetric, which is a fundamental property of the normal distribution. In addition, the normal distribution is
defined over a range that includes negative values while pollutant measurements are restricted to nonnegative
values. Thus, fitting a data set to a normal distribution allows for the possibility, however small, of observing
negative values. The lognormal distribution, or any positive valued distribution, is not defined for negative
values and thus avoids assigning any probability to negative values.
Dally Maximum Permit Limits Based on the Normal Distribution
For data sets which have the characteristics of the normal distribution, the daily maximum permit limits can be
calculated. The upper percentile daily maximum permit limits for the normal distribution are calculated using
the quantity Zp, the standardized Z-score for the pth percentile of the standardized normal distribution (i.e.,
normal distribution with mean = 0, and variance = 1). For example, the Z-score for the 95th percentile is 1.645.
Z-scores are listed in tables for the normal distribution (in most statistical textbooks and references). The pth
percentile daily maximum limit is estimated by:
.
Xn s pth percentile daily maximum limit
""
E-4
-------
For example:
X 95 = 95th percentile daily maximum limit -~-~.*•,.-. •:••-•
= ft + 1.645
& 99 = 99th percentile daily maximum limit
= ft+ 2.326
Note:
Z95 = 1.645
299 = 2.326
The daily variability factors (denoted by VF-j) are estimated by:
Daily maximum 95th percentile VF-) = £ 95 / ft
Daily maximum 99th percentile VF-j = & 99 / ft
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 usually is not an appropriate model for individual pollutant
measurements, the averages of those individual measurements can often be modeled by the normal distribution.
This subsection 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
recommended to be sufficiently "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 p.n and on 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
ftn and on, respectively.
ft = estimated mean of distribution of X
a^ = estimated variance of distribution of X
ftn = mean of distribution of the n-day monthly average
A
= \i •
an = variance of distribution of the n-day monthly average
A O .
an = standard deviation
•)
cvn = coefficient of variation
A i A
= °n / Hn.
E-5
-------
The upper percentile limits are:
ft p = pth percentile n-day monthly average limit
where zp is the pth percentage point of the standard normal distribution.
•
|
For example:
ft 95 = 95th percentile n-day monthly average limit
= An + 1.645 cn
ft.99 = 99th percentile n-day monthly average limit
= An + 2.326 cn
Note:
295 = 1.645
= 2.326.
The monthly average variability factors (denoted by VFn) are estimated by:
Monthly average 95th percentile VFn = £.95 / jl
Monthly average 99th percentile VFn = ft 99 / jl
The above discussion of the normal distribution can be modified for data from the lognormal distribution. The
next subsection 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 also are often
lognormally distributed. Figure E-2 displays the positively skewed shape of the lognormal distribution.
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 E-3 shows the lognormal distribution applied to data used in the
development of the OCPSF effluent guidelines regulation [1].
E-6
-------
Long-term average
CV=0.2
CV=0.4
CV=0.6
CV=0.8
CV=1.0
0.5
3.5
150
100
50
Figure E-2. Examples of Lognormal Densities
9 15 21 27 33 39 45 51 57 63 69
Concentration in ug/l
Figure E-3. BOD Frequency Distribution - Plant C
E-7
-------
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 av
and ay respectively. These parameters can be estimated by:
and
where
y
a
Kyi) / k
- A)2] / (k - 1), respectively
yj = In(Xj) for i=1,2,...k.
When data are lognormally distributed, these values from the normal distribution can then be used to calculate
the mean, variance, and coefficient of variation for the random variable X that is lognormally distributed. The
mean, variance, and coefficient of variation of the random variable X may be estimated by E(X), fy(X), and
cv(X), respectively.
I(X) = daily average
= exp( Ay + &y / 2)
= variance
= exp(2Ay + a>
cv(X) = coefficient of variation
Dally Maximum Permit Limits Based on the Lognormal Distribution
The upper percentile limits for the random variable X (which is lognormally distributed) are:
& p = pth percentile daily maximum limit
where zp is the pth percentage point of the standard normal distribution.
For example:
^.95 - 95th percentile daily maximum limit
= exp[ Ay + 1.645 &y]
^99 = 99th percentile daily maximum limit
= exp[ Ay + 2.326 CTy].
Note:
Z95 = 1.645
Z99 = 2.326.
E-8
-------
The daily maximum variability factors (denoted by VF-|) are estimated by:
Daily maximum 95th percentile VF-j = ^.95 / E(X)
Daily maximum 99th percentile VF-| = ^.99 / E(X).
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 flognormally distributed values with another lognormal distribution. Although, the Central Limit Theorem
h^lds 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 [3]. For sample sizes larger than 10 when the data are lognormally distributed, it is recommended
that the calculations given in Table E-3 should be used. For the purposes of permit writing, monitoring sample
sizes of 10 or less are recommended 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 ftn, on, and cv, estimated by:
an = variance
= ln{n
on = standard deviation
cvn = coefficient of variation
where
I(X) = 6XP( jly + CTy / 2)
The upper percentile limits of the maximum n-day monthly average are:
ft p = pth percentile n-day monthly average limit
= exp[ jln + zp c>n]
where Zp is the pth percentage point of the standard normal distribution.
For example:
ft 95 = 95th percentile n-day monthly average limit
X 99 = 99th percentile n-day monthly average limit
= exp[ An + 2.326 an]
E-9
-------
Note:
Z95 = 1.645
299 = 2.326.
The variability factors are:
Monthly average 95th percentile VFn = & 95 / jj.n
Monthly average 99th percentile VFn = £.99 / (
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 nondetect values and values above the detection
limit and can be used to model nondetects in water quality-based limits. In delta-lognormal procedures,
nondetect values are weighted in proportion to their occurrence in the data. 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 monthly averages with the recommended number of monitoring
samples being 10 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 nondetect 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 E-4 contains a number of observations that were
reported as "nondetect." 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-lognorma) distribution often provides an appropriate and
computationally convenient model for analyzing such data:
The estimation procedure for the delta-lognormal 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 detailed in Reference 4.
These values set to D are observations that can only by quantified as nondetect (ND) at some minimum level.
This minimum level is the detection limit as established by the laboratory performing the chemical analysis.
Let Xi,X2,...,xr,xr+-j,.../X|< denote a random sample of size k, with r observations recorded as nondetects, 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-lognormal distribution with censoring point
2
equal to the detection limit D. Let jly and CTV be the sample mean and variance of the distribution of the
logarithmic transformation Y = ln(X) of the observations greater than the detection limit. Let & be the sample
proportion of nondetects. Then the estimates of the mean and variance of the delta-lognormal distribution are
estimated by: '
')
= daily average
= to + (1 - & ) exp( p.y + 0.5 c>y )
= variance
-o-» •' '
CTy)[exp(Gy)-(1 -
cv(X )
= coefficient of variation
= CvYX*)]1/2/£(X*)
- & ) D [ D - 2 exp( Ay + 0.5 Oy )]
E-10
-------
I
-------
where
k = number of samples
D = detection limit
r s number of nondetect values in sample
k-r at number of values greater than the detection limit
yj = In(Xj) r+1 < i £ k, r < k
= S(yj) / (k - r) r+1 < i < k, r < k
y - I(yi-Ay)2/(k-r-1) r+10.99
where
[max [D, exp( Ay + z*CTy)] $ < 0.99
Zw = *-l[(0.99-&)/(!-&)].
"1 [ ] is the mathematical notation for Z-scores. For example, when & = 0, then the corresponding value is
= Zpj = 2.326. Values of *" [ ] 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 = ft 95 /
Daily maximum 99th percentile VF = X 99
Delta-Lognormal Distribution of Averages
The derivation of the formulas for the averages computationally is difficult and beyond the scope of this
appendix. However, the formulas for n-day averages are included in Table E-2. 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]. For the purpose of permit writing, it is recommended that data sets of greater than
10 samples be assumed to fit the normal distribution and the averages be calculated using the formulas given in
Table E-3.
E-12
-------
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 data set 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 [5]. Consider an independent sample of size k, labeled Xi,x2,<.vX|c Let u1>u2,...,U|( be the
ordered sample of x-values in ascending order in which u1^u2S,...,
-------
1-
0-
1-1-
S
-3-
Lead
-3.0
-1.5
0.0
Z Score
1.5
3.0
Figure E-5. Example of a Log-Probability
Plot with a Normal Distribution
E-14
-------
In the case of the monthly average limit derivation, the assumption that observed pollutant levels are
independent can be quite important. 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 tend to be correlated with each other 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 4 to 8 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.
Significant correlation between observed pollutant levels, when present, should be factored into monthly
average permit limits.
Several different methods can be used to account for correlation in determining limits. One general approach
involves time series modeling. Another possible approach is to use a direct computation of the covariance
among the observed data to adjust the variance of the average used in determining the limit. Help in adjusting
the sample size for correlation is available from the OW Statistics Section (phone number [202] 382-5397).
Table E-1. Daily Maximum Permit Limit Calculations
The daily maximum permit limit is usually the 99th upper percentile value of the pollutant distribution. In
certain cases the 95th percentile value may be allowable. The following gives the formulas:
WITH ALL MEASUREMENTS > DETECTION LIMIT (based on lognormal distribution)
X 95 = 95tn percentile daily maximum limit
= exp[£y+ 1.645 Oy]
X 99 = 99th percentile daily maximum limit
= exp[(iy + 2.326 ay]
where
Xj = daily pollutant measurement i
k = sample size of data set
Ay = £(yj) / k 1 < i < k
= exp( Ay + 0.5 cty
cv(X) =
E-15
-------
Table E-1. Daily Maximum Permit limit Calculations (continued)
WITH SOME MEASUREMENTS < DETECTION LIMIT (based on delta-lognormal distribution)
95th percentile daily maximum limit
To S 2:0.95
\95
L max [D, exp( (L + z* oy)] & < 0.95
with z* = &-1 [(0.95 - & ) / (1 - S)]
X 99 = 99th percentile daily maximum limit
To &;>0.99
/\99 s I
Lmax [D, exp$y + z*oy)] S < 0.99
with z* « O"1 [(0.99 - S ) / (1 - & )]
where
k
D
r
k-r
VI
= daily pollutant measurement i
= sample size of data set
= detection limit (as established by the laboratory)
= number of nondetects (x-j ,X2,...,x r are < D)
~ number of detects (xr+Vxr+2'—'^k are> D)
= In(xj) for r+1 s i ^ k
= r/k
= 2(yj) / (k - r) r+1 ^ i ^ k (exclude values < D from sum)
I(X*)
2[(y,--Ay)2]/(k-r-1) r+1
SD + (1-S)exp(Ay + 0.5Ay)
*) = (1 - S )exp(2 Ay + 6y) [exp(0y) - (1 - & )] + S (1 - & )D[D - 2 exp( Ay + 0.5
E-16
-------
Table E-2. Monthly Average Permit Limit Calculations for
Ten Samples or Less
The monthly average permit limit is usually based on the estimates of the 95th percentjle of the distribution of
the average of the daily 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 includes nondetects).
All MEASUREMENTS > DETECTION LIMIT (based on lognormal distribution)
£ 95 = 95th percentile n-day monthly average limit
= exp[£n + 1.645 on]
X 99 = 99th percentile n-day monthly average limit
= exp[ £n + 2.326 on]
where
cvn
Xj = daily pollutant measurement i
Yi = In(Xj)
k = sample size of data set
= !(yj) / k 1 1 i < k
= I[(yi-Ay)2]/(k-D
I(X) = exp( Ay + 0.5 n
E-17
-------
Table E-2. Monthly Average Permit Limit Calculations for
Ten Samples or Less (continued)
SOME MEASUREMENTS < DETECTION LIMIT (based on delta-lognormal distribution)
X^95 = 95th percentile n-day monthly average limit
[D & >0.95
* 95 ='
[max [D, exp( £n + z*on)] 8 < 0.95
With Z* r:*'1 [ (0.95 - S ) / (1 - 8)].
X 99 = 99th percentile n-day monthly average limit
[D 6" >0.99
*'" = '
L max [D, exp( jin + z*on)] S < 0.99
with z* = -1 [(0.99 - S ) / (1 - 8 )]
where
= daily pollutant measurement i
= sample size of data set
k
D = detection limit (as established by the laboratory)
r = number of nondetects (xi,X2,..*,xrareS D)
k-r = number of detects (xr+1'*r-(-2'— 'xk are>
y; = In(xj) for r+1 < i < k
& = r/k
with
/ (k - r) r+1 ^ i
k (exclude values
-------
Table E-3. Monthly Average Permit Limit Calculations for More Than Ten Samples
The monthly average permit limit usually is based on the estimates of the 95th percentile of the distribution
of the average of the daily effluent values. These daily values are assumed to be lognormally distributed. For
sample sizes larger than 10, the averages (represented by the random variable Xn) are assumed to be normally
distributed.
X 95 = 95th percentile n-day monthly average limit
.99
where
xi
y\
99th percentile n-day monthly average limit
E(Xn)+ 2.326
daily pollutant measurement i
In (Xj)
sample size of data set
/ k, 1 < i < k
= H(yj-My)2]/(k-D 1
-------
1.
2.
3.
4.
5.
6.
APPENDIX E
REFERENCES
U.S. EPA. 1987a. Development Document for Effluent Limitations Guidelines and Standards for the
Organic Chemicals, Plastics and Synthetic Fibers. EPA 440/1-87/009.
Mendenhall, W., R. Scheaffer, and D. Wackerly. 1981. Mathematical Statistics with Applications. 2d
ed. Wadsworth, MA.
Barakat, R. 1976. Sums of Independent Lognormally Distributed Random Variables. /. Optical Soc. Am.
66:211-16.
Aitchison, J., and J. Brown. 1963. The Lognormal Distribution. Cambridge University Press.
Johnson, R., and D. Wichern. 1982. Applied Multivariate Statistical Analysis. New Jersey: Prentice-
Hall.
Hollander, M., and D. Wolfe. 1973. Nonparametric Statistical Methods. New York: Wiley.
ADDITIONAL REFERENCES
Kahn, H. 1989. Memorandum: Response to Memorandum from Dr. Don Mount of December 22, 1988.
August 30, 1989. U.S. EPA, Washington, DC, to J. Taft, U.S. EPA, Permits Division, Washington, DC.
Kahn, H., and M. Rubin. 1989. Use of Statistical Methods in Industrial Water Pollution Control
Regulations in the United States. Environmental Monitoring and Assessment 12:129-48.
E-20
-------
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 parameters of interest can be obtained. Unfortunately, many of the industrial and
municipal National Pollutant Discharge Elimination System 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 (and 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 to test chamber volume can be scaled to simulate the time-varying concentration at the mixing
zone boundary.
Although flowthrough methods can provide a realistic simulation of time-varying exposure, they are relatively
expensive and are usually conducted on site. Therefore, flowthrough 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 flowthrough 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
F-1
-------
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.
I
Sampling Frequency
Nonrandom 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 versus 52 for the weekly single sample approach. Assuming
that daily toxic events of environmentally significant intensity and duration would not be masked by short-
term composites, it might be more efficient to collect eight samples each composited over a 3-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 (see Section 3, Table 3-1).
i
Box F-1 presents EPA's recommendations on sampling methods.
F-2
-------
Box F-1. Recommendations
The initial sampling design step should involve stratification of sampling periods to account
for nonrandom 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.
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 flowthrough sample of serially diluted effluent. If no systematic
annual variations (e.g., batch processing) are known or suspected, flowthrough testing can be
conducted at a minimum of quarterly intervals for at least 1 year.
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 onsite or at a remote lab. The design of a toxics grab-sampling
program must take into account the tradeoff 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 (4-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 (e.g., quarterly) an uncertainty factor
for variability should be used in the toxicity evaluation (see Section 3).
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.
F-3
-------
APPENDIX F
REFERENCES
i
1. Nemetz, P.N., and H.D. Dreschler. 1978. The Role of Effluent Monitoring in Environmental Control. ].
Water, Air, and Pollution 10:477-97.
I:
2. Bender, E.S. 1984. "Sources of Variations in Effluent Toxicity Tests." \r\EnvironmentalHazardAssessmentof
Effluents. Ed. H. Bergman, R. Kimerle, and A.W. Maki. Proceedings of the Fifth Pellston
Environmental Workshop, Cody, WY, August 1982.
F-4
-------
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
further 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.
G-1
-------
Genotoxicity Tests for Non-Threshold Chemicals
Genotoxicity is the ability of a substance to damage an organism's genetic material (its DMA). Certain
positively-charged compounds tend to bind to DNA ancl 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 DN/\ 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 m vitro
transformation provide supportive evidence of carcinogehicity [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].
i
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 mamrnalian 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
G-2
-------
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/liter) 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 changes 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 dose response (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 Genotoxicity 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
G-3
-------
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 ihe need for limited testing to ensure that the low is
really indicative of the risk. Low risk in both exposure knd 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 (TOQ 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.
i
A process evaluation is a study in which components in the wastewater are determined from an analysis of
feedstocks, manufacturing processes, products, by-products> and pollution control in place. The result is a list
of compounds or classes of compounds with a high probability of being 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 typhimurium 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 witlh 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 an^l specificity are documented [Ashby and Tennant,
1988].
G-4
-------
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 Chromotesti [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 documented [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.
Cost: ??
• 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
G-5
-------
Disadvantages: Carcinogen levels in unconcentrated effluent may not be high enough to produce
tumors at a detectable frequency in exposed Ipopulations, 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: S20.000 Flohnson. 19891.
i
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.
i
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.
I
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. r
G6
-------
APPENDIX G
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 ]. Puerner. 1985. Toxicity Determined in 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: Groom-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 in 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 )., 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.
G-7
-------
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 Caarcinogenesis 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 (RrD): DESCRIPTION AND USE IN
HEALTH RISK ASSESSMENTS
-------
REFERENCE DOSE (RrD) DESCRIPTION AND USE IN
HEALTH RISK ASSESSMENTS (REVISED O2/1O/88)
Principal Author:
Donald Barnes, Ph.D. (OPTS)
RfD Work Group:
Donald Barnes, Ph.D. (OPTS)
Judith Bellin, Ph.D. (OSWER)
Christopher DeRosa, Ph.D. (ORD)
Michael Dourson, Ph.D. (ORD), Co-Chair
Reto Engler, Ph.D. (OPTS)
Linda Erdreich, Ph.D. (ORD)
Theodore Farber, Ph.D. (OPTS)
Penny Fenner-Crisp, Ph.D. (OW)
Elaine Francis, Ph.D. (OPTS)
George Ghali, Ph.D. (OPTS)
Richard Hill, M.D., Ph.D. (OPTS)
Stephanie Irene, Ph.D. (OPTS)
William Marcus, Ph.D. (OW)
David Patrick, P.E., B.S. (OAR)
Susan Perlin, Ph.D. (OPPE)
Peter Preuss, Ph.D. (ORD), Co-Chair
Aggie Revesz, B.S. (OPTS)
Reva Rubenstein, Ph.D. (OSWER)
Jerry Stara, D.V.M., Ph.D. (ORD)
Jeanette Wiltse, Ph.D. (OPTS)
Larry Zaragosa, Ph.D. (OAR)
H-1
-------
REFERENCE DOSE (RFD): DESCRIPTION AND USE IN
HEALTH RISK ASSESSMENTS
Introduction !
This concept paper describes the U.S. Environmental Protection Agency's (U.S. EPA) principal approach to and
rationale for assessing risk 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
paper complements the new risk assessment guidelines (U.S. EPA, 1987), which describe the Agency's approach
to risk assessment in other areas, specifically carcinogenicity, mutagenicity, developmental toxicity, exposure,
and chemical mixtures. (In this document the term ^'systemic toxicity" refers to an effect other than
carcinogenicity or mutagenicity induced by a toxic chemical.)
Background and Summary
Chemicals that give rise to toxic endpoints other than cancer and gene mutations are often referred to as
"systemic toxicants" because of their effects on the function of various organ systems. In addition, chemicals
that cause cancer and gene mutations also commonly evoke other toxic effects (i.e., systemic toxicity). 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 populations) below which there are no
observable adverse effects. This characteristic distinguishes systemic endpoints from carcinogenic and
mutagenic endpoints, which are often treated as nonthreshold processes.
Systemic effects have traditionally been evaluated using sqch terms as "acceptable daily intake (ADI)," "safety
factor (SF)," and "margin of safety (MOS)," concepts that are associated with certain limitations described
below. The U.S. EPA established the Reference Dose (RfD) Work Group to address these concerns.
•
In preparing this report, the RfD Work Group has drawn on traditional report on risk assessment (NRC, 1983), to
more fully articulate the use of noncancer, nonmutagenic experimental data in reaching regulatory decisions
about the significance of exposures to chemicals. In the process, the Work Group has coined less value-laden
terminology - "reference dose (RfD)," "uncertainty factor (UF)"; "margin of exposure (MOE)"; and "regulatory
dose (RgD)" -- to clarify and distinguish between aspects of risk assessment and risk management. These
concepts are currently in general use in many parts of U.S. EPA.
Traditional Approach to Assessing Systemic Toxicity
The U.S. EPA's approach to assessing the risks associated With systemic toxicity is different from its approach
to assessing the risks associated with carcinogenicity, because of the different mechanisms of action 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 theoretically 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, however, organic homeostatic, compensating, and adaptive mechanisms exist that must be
overcome before a toxic endpoint 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.
H-2
-------
Description of 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 (ADI)" derived from an experimentally determined "no-observed-adverse-effect level
(NOAEL)." The ADI is commonly defined as the amount of a chemical to which a person can be exposed on a
daily basis over an extended period of time (usually a lifetime) without suffering a deleterious effect. The ADI
concept has often been used as a tool in reaching risk management decisions (e.g., establishing allowable
levels of contaminants in foodstuffs and water.)
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 NOAELs, 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 a statistically or biologically significant adverse effect. The NOAEL for the critical
toxic effect is sometimes referred to simply as the NOEL. This usage, however, invites ambiguity in that there
may be observable effects that are not of toxicological significance (i.e., they are not "adverse"). For the sake
of precision, this document uses the term NOAEL to mean the highest NOAEL in an experiment. In cases in
which a NOAEL has not been demonstrated experimentally, the term "lowest-observed-adverse-effect level
(LOAEL)" is used.
Once the critical study demonstrating the toxic effect of concern has been identified, the selection of the
NOAEL results from an objective examination of the data available on the chemical in question. The ADI is
then derived by dividing the appropriate NOAEL by a safety factor (SF), as follows:
ADI (human dose) = NOAEL (experimental dose)/SF . (Equation 1)
Generally, the SF consists of multiples of 10, each factor representing a specific area of uncertainty inherent in
the available data. For example, a factor of 10 may be introduced to account for the possible differences in
responsiveness between humans and animals in prolonged exposure studies. A second factor of 10 may be used
to account for variation in susceptibility among individuals in the human population. The resultant SF of 100
has been judged to be appropriate for many chemicals. For other chemicals, with databases that are less
complete (for example, those for which only the results of subchronic studies are available), an additional
factor of 10 (leading to a SF of 1000) might be judged to be more appropriate. For certain other chemicals,
based on well-characterized responses in sensitive humans (as in the effect of fluoride on human teeth), an SF as
small as 1 might be selected.
While the original selection of SFs appears to have been rather arbitrary (Lehman and Fitzhugh, 1954),
subsequent analysis of data (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.
Some Difficulties in Utilizing the Traditional Approach
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:
a. Too narrow a focus on the NOAEL means that information on the shape of the dose-response curve is
ignored. Such data could be important in estimating levels of concern for public safety.
b. As scientific knowledge increases and the correlation of precursor effects (e.g., enzyme induction)
with toxicity becomes known, questions about the selection of the appropriate "adverse effect"
arise.
c. Guidelines have not been developed to take into account the fact that some studies have used larger
(smaller) numbers of animals and, hence, are generally more (less) reliable than other studies.
H-3
-------
These and other "scientific issues" are not susceptible to immediate resolution, since the database needed is not
yet sufficiently developed or analyzed. U.S. EPA work groups are presently considering these issues.
Management-related Issues
The use of the term "safety factor" - 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.
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 marly (including risk managers) 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
which is not likely to result in adverse effects to humans. The ADI is generally viewed by risk assessors 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 all risk to all people cannot be assured at this level.
Possible limitations imposed on risk management decisions - Awareness of the "softness" of the ADI
estimate, as discussed above, argues for careful case-by-caSe consideration of the toxicological implications of
individual situation, so that ADIs are not given a degree of significance that is scientifically unwarranted. In
addition, the ADI is only one factor in a risk management decision and should not be used to the exclusion of
other relevant factors. j
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 ADI by
adopting factors different from the standard factors listed in Table 1. For example, if the toxic endpoint 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 practices, as employed by a number of scientists in different programs/agencies, exercising their best
scientific judgment, have in some cases resulted in different ADIs for the same chemical. The fact that
different ADIs were generated (for example, by adopting different SFs) can be a source of considerable
confusion when the ADIs are used exclusively in risk management decisionmaking. The existence of different
ADIs need not imply that any of them is more "wrong"-or "right"~than the rest. It is more nearly a reflection of
the honest difference in scientific judgment.
However, on occasion, these differences in judgment of the [scientific data, can be interpreted as differences in
the management of the risk. As a result, scientists may be inappropriately impugned, and/or perfectly
justifiable 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 ADI as an absolute measure of safety.
EPA Assessment of Risks Associated with Systemic Toxicity
The U.S. EPA approach to analyzing systemic toxicity data 'follow the general format set forth by NRC in its
description of the risk assessment process (NRC, 1983). The determination of the presence of risk and its
potential magnitude is made during the risk assessment profess, 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 considers control options available under existing statutes and
other relevant non-risk factors (e.g., benefits to be gained and costs to be incurred). All of these considerations
go into the determination of the regulatory decision (Figure 1).
H-4
-------
Table 1. Guidelines for the Use of Uncertainty Factors in Deriving Reference
Doses and Modifying Factors
Standard Uncertainty Factors (UFs):
Use a 10-fold factor when extrapolating from valid experimental results in 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 and is referenced as
"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 involved in
extrapolating from animal data to humans and is referenced as "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 involved in extrapolating from less than chronic
NOAELs to chronic NOAELs and is referenced as "1 OS".
Use an additional 10-fold factor when deriving an RfD from a LOAEL, instead of a NOAEL.
This factor is intended to account for the uncertainty involved in extrapolating from
LOAELs to NOAELs and is referenced as "10L".
Mollifying Factor (MF):
Use professional judgment to determine the MF, which is an additional uncertainty factor
that 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 data base 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 and Stara, 1983
Hazard Identification
Evidence
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. In assessments of the risk posed by a chemical, the toxic endpoints from all available
studies are considered, although primary attention usually is given to the effect (the "critical effect")
exhibiting the lowest NOAEL. In the case of chemicals with limited data bases, additional toxicity testing
may be necessary before an assessment can be made.
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
H-5
-------
Figure 1.
Conceptual Framework for Risk Assessment and Risk Management*
Risk Assessment
Hazard Identification
Dose-Response Assessment
(e.g., RfD)
Exposure Assessment
v v
Risk Characterization-
(e.g., Criterion)
Risk Management
Control Options
Non-risk Analyses
V v
->Regulatory Decision
(e.g., RgD, Standard)
*Source: Adapted from NRC, 1983
used in the quantitative dose-response assessment phase of the risk assessment. These studies are of two types:
studies of human populations (epidemiologic investigations) and studies using laboratory animals.
1. Epidemiologic studies - Human data are often Useful in qualitatively establishing the presence of
an adverse effect in exposed human populations. When there is information on the exposure level
associated with an appropriate endpoint, epidepiiologic studies can also provide the basis for a
quantitative dose-response assessment. The presence of such data obviates the necessity of
extrapolating from animals to humans; 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 a^id 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 - For most chemicals, there is ia lack of appropriate information on effects in
humans. In such cases, the principal studies are frawn from experiments conducted on nonhuman
mammals, most often the rat, mouse, rabbit, guinea pig, hamster, dog, or monkey.
Supporting studies - These studies provide supportive, rather than definitive, information and can include
data from a wide variety of sources. For example, metabolic and other pharmacokinetics studies can provide
insights into the mechanism of action of a particular compound. By comparing the metabolism of the chemical
exhibiting the toxic effect in the animal with the metabolism found in humans, it may be possible to assess the
potential for toxicity in humans or to estimate the equitoxic dose in humans.
Similarly, in vitro studies can provide insights into the chemical's potential for biological activity; and under
certain circumstances, consideration of structure-activity relationships between a chemical and other
H-6
-------
structurally related compounds can provide clues to the test chemical's possible toxicity. More reliable in vitro
tests are presently being developed to minimize the need for live-animal testing. There is also increased
emphasis on generating mechanism-of-action and pharmacokinetics information as a means of increasing
understanding of toxic processes in humans and nonhumans.
Route of exposure - The U.S. EPA 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 an inhalation exposure for an air
contaminant). In most cases, the toxicologic data base does not include detailed testing on all possible
routes of administration, with their possibly significant differences in factors such as mechanism-of-action and
bioavailability. In general, the U.S. EPA's position is that the potential for toxicity manifested via one route
of exposure is relevant to considerations of any other route of exposure, unless convincing evidence exists to
the contrary. Consideration is 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 delineated.
Length of exposure - The U.S. EPA 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
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 low-dose chronic studies. Special attention is given to studies involving low-dose, chronic exposures,
since such exposures can elicit effects absent in higher dose, shorter exposures, through mechanisms such as
accumulation of toxicants in the organisms.
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.
Weight-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 principal and supportive studies. Emphasis 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.
Dose-Response Assessment
Concepts and Problems
Empirical observations have generally revealed 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
H-7
-------
with dose; and in continuous responses, in which changes in a biological parameter (e.g., body or organ
weight) vary with dose.
i'
In evaluating a dose-response relationship, certain difficulties arise. For example, one must decide on the
critical endpoint 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.
Selection of the Critical Data
Critical study - Data from experimental studies in laboratory animals are often selected as the governing
information when performing quantitative risk assessments, since available human data are usually 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
which 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 (e.g., for instance using comparative pharmacokinetics data).
In the absence of a clearly most relevant species, the most sensitive species (i.e., the species showing a toxic
effect at the lowest administered dose) is used by risk assessors at U.S. EPA, since there is no assurance that
humans are not at least as innately sensitive as the most sensitive species tested. This selection process is more
difficult when the routes of exposure in the animal tests are different from those involved in the human
situation under investigation. In order to use data from controlled studies of genetically homogeneous
animals, the risk assessor must also extrapolate from animals to humans and from high experimental doses to
comparatively low environmental exposures, and must account for human heterogeneity and possible concurrent
human exposures to other chemicals.
Although for most chemicals there is a lack of well-controlled cohort studies investigating noncancer
endpoints, in some cases an epidemiologic study 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 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).
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 severity of an adverse effect in
individuals in an exposed group when compared with individuals in an appropriate control group. 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.
Critical endpoint - As noted under "Traditional Approach to Assessing Systemic Toxicity", a chemical may
elicit more than one toxic effect (endpoint), 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 endpoint used in the dose- response assessment is the effect exhibiting the lowest NOAEL.
H-8
-------
Reference Dose (RfD)
The reference dose (RfD) and uncertainty factor (UF) concepts have been developed by the RfD Work Group in
response to many of the problems associated with ADIs and SFs, as outlined under "Traditional Approach to
Assessing Systemic Toxicity" above. 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 sets used to estimate RfDs. For example, a valid chronic animal NOAEL is normally divided by an UF of
100. In addition, a modifying factor (MF), is sometimes used which is based on a professional judgment of
the entire data base of the chemical. These factors and their rationales are presented in Table 1.
The RfD is determined by use of the following equation:
RfD = NOAEL / (UF x MF)
which is the functional equivalent of Equation 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
generally expressed in units of milligrams per kilogram of body weight per day (mg/kg/day).
The RfD is useful as a reference point from which to gauge the potential effects of the chemical at other doses.
Usually, doses less than the RfD are not likely to be associated with adverse health risks, and are therefore less
likely to be of regulatory concern. As the frequency and/or magnitude of the exposures exceeding the RfD
increase, the probability of adverse effects in a human population increases. However, it should not be
categorically concluded that all doses below the RfD are "acceptable" (or will be risk-free) and that all doses
in excess of the RfD are "unacceptable" (or will result in adverse effects).
The U.S. EPA 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 (high,
medium, or low) the evaluators have in the stability of the RfD. High confidence indicates the judgment that
the RfD is unlikely to change in the future because there is consistency among the toxic responses observed in
different sexes, species, study designs, or in dose-response relationships, or that the reasons for existing
differences are well understood. High confidence is often given to RfDs that are based on human data for the
exposure route of concern, since in such cases the problems of interspecies extrapolation have been avoided.
Low confidence indicates the judgment that the data supporting the RfD may be of limited quality and/or
quantity and that additional information could result in a change in the RfD.
Exposure Assessment
The third step in the risk assessment process focuses on exposure issues. For a full discussion of exposure
assessment, consult U.S.. EPA's guidelines on the subject (U.S. EPA, 1987). In brief, the exposure assessment
includes consideration of the size and nature of the populations exposed and the magnitude, frequency,
duration and routes of exposure, as well as evaluation of the nature of the exposed populations.
Risk Characterization
Risk characterization is the final step in the risk assessment process and the first input to the risk management
(regulatory action) process. The purpose of risk characterization is to present the risk manager with a synopsis
and synthesis of all the data that should contribute to a conclusion with regard to the nature and extent of the
risk, including:
/
a. The qualitative ("weight-of-evidence") conclusions as to the likelihood that the chemical may pose
a hazard to human health.
b. A discussion of the dose-response information considered in deriving the RfD, including the UFs
and MFs used.
c. Data on the shapes and slopes of the dose-response curves for the various toxic endpoints,
toxicodynamics (absorption and metabolism), structure-activity correlations, and the nature and
severity of the observed effects.
H-9
-------
d. Estimates of the nature and extent of the exposure and the number and types of people exposed.
e. Discussion of the overall uncertainty in the analysis, including the major assumptions made,
scientific judgments employed, and an estimate of the degree of conservatism involved.
In the risk characterization process, a comparison is made between the RfD and the estimated (calculated or
measured) exposure dose (EED). The EED should include all sources and routes of exposure involved. 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), 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).
When the MOE is equal to or greater than UF x MF, the need for regulatory concern is likely to be small.
"Hypothetical, Simplified Example of Determining and Using RfD" contains an example of the use of the
concepts of NOAEL, UF, MF, RfD, EED, and MOE.
Application In Risk Management
Once the risk characterization is completed, the focus turns to risk management. In reaching decisions, the risk
manager utilizes the results of risk assessment, other technological factors, and legal, economic and social
considerations in reaching a regulatory decision. These additional factors include efficiency, timeliness,
equity, administrative simplicity, consistency, public acceptability, technological feasibility, and nature of
the legislative mandate.
t
Because of the way these risk management factors may impact different cases, consistent - but not 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
(SDWA) calls for standards which protect the public "to the extent feasible." Consequently, it is entirely
possible and appropriate that a chemical with a specific Rr"D may be regulated under different statutes and
situations through the use of different "regulatory doses (RgDs)."
That is, after carefully considering the various risk and nonrisk factors, regulatory options, and statutory
mandates in a given case (i), the risk manager selects the appropriate statutory alternative for arriving at an
"ample" or "adequate" margin of exposure [MOE(i)]. As shown in Equation 2 below, this procedure establishes
the regulatory dose, RgD(i) (e.g., a tolerance under FIFRA, or a maximum contaminant level under SDWA),
applicable to the case in question:
RgD(i) = NOAEL / MOE(i)
(Equation 2)
Note that different RgDs are possible for a given chemical wi£h a single RfD. Note also 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) / (UF x MF).
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 compiled by the risk assessors, as discussed under "Risk
Characterization". In some cases, additional explanation and interpretation, may be required from the risk
assessors in order to arrive at a responsible and clearly articulated final decision on the RgD.
It Is generally useful to the risk manager to have information Regarding the contribution to the RfD from various
environmental media (e.g., air, water and food). Such information can provide insights that are helpful in
H-10
-------
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.
Work is underway in the U.S. EPA to apportion the RfD among the various environmental media. For example,
consider the case of a food-use pesticide which is a contaminant in drinking water. In selecting among risk
management actions under the Safe Drinking Water Act, it might be prudent to assume an RfD for drinking water
purposes which is some fraction of the total RfD. Such an apportionment would explicitly acknowledge the
possible additional exposure from ingestion of treated crops. The apportionment of the RfD would, in general,
provide additional guidance for risk managers of the various media- specific programs.
Other Directions
In addition to the development of reference doses, the U.S. EPA is pursuing other lines of investigation for
systemic toxicity. For example, the Office of Air Quality Planning and Standards is using probabilistic risk
assessment procedures for criteria pollutants. In this procedure, the population at risk is characterized, and the
likelihood of the occurrence of various effects is predicted through the use of available scientific literature and
of scientific experts' rendering their judgments concerning dose-response relationships. This dose-response
information is then combined with the results of the exposure analysis to generate population risk estimates for
alternative standards. Through the use of these procedures, decisionmakers are presented with ranges of risk
estimates in which uncertainties associated with both the toxicity and exposure information are explicitly
considered. 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 designed to increase the reliability of risk assessments. They are investigating the use of
extrapolation models as a means of estimating RfDs, taking into account the statistical variability of the
NOAEL. and underlying UFs. ORD is also exploring procedures for conducting health risk assessments that
involve less- than-lifetime exposures. Finally, they are working on approaches to ranking the severity of
different toxic effects.
Hypothetical, Simplified Example of Determining and Using RfD
Experimental Results
Suppose the U.S. EPA had a sound 90-day subchronic gavage study in rats with the data in Table 2.
Analysis
Determination of the Reference Dose (RfD)
Using the NOAEL - Because the study is on animals and of subchronic duration,
UF = 10Hx10Ax 105 = 1000 (Table 1)
In addition, there is a subjective adjustment (MF) based on the high number of animals (250) per dose group:
MF = 0.8. These factors then give UF x MF = 800, so that
RfD = NOAEL/(UF x MF) = 5/800 = 0.006 (mg/kg/day).
H-11
-------
Table 2. Hypothetical Data to Illustrate the Reference Dose Concept
Dose
mg/kg/day
Observation
Effect Level
0 Control—no adverse effects observed
1 No statistically or biologically significant differences
between treated and control animals
5 % decrease* in body weight gain (not considered to be
of biological significance); increased ratio of liver
weight to body weight; histopathology indistinguishable
from controls; evaluated liver enzyme levels
25 20% decrease* in body weight gain; increased* ratio of
liver weight to body weight; enlarged, fatty liver with
vacuole formation; increased* liver enzyme levels
•Statistically significant compared to controls.
NOEL
NOAEL
LOAEL
Using the LOAEL - If the NOAEL is not available, and if 25 mg/kg/day had been the lowest dose tested that
showed adverse effects,
UF * 10H x IDA x 10S x 10L = 10,000 (Table 1).
Using again the subjective adjustment of MF = 0.8, one obtains
RfD - LOAEL/(UF x MF) = 25/8000 = 0.003 (mg/kg/day).
Risk Characterization Considerations
Suppose the estimated exposure dose (EED) for humans exposed to the chemical under the proposed use pattern
were 0.01 mg/kg/day (i.e., the EED is greater than the RfD). Viewed alternatively, the MOE is:
MOE = NOAEL/EED = 5 (mg/kg/day) / 0.01 (mg/kg/day) = 500.
Because the EED 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 arid the exposure estimates, and the comments of the
risk assessors. In addition, the risk manager will need to wejgh the benefits associated with the case, and other
non-risk factors, in reaching a decision on the regulatory dose (RgD).
H-12
-------
APPENDIX H
REFERENCES
1. Dourson, M.L. and J.F. Stara. 1983. Regulatory Toxicology and Pharmacology. 3:224-238.
2. Lehman, A.|. and O.G. Fitihugh. 1954. Association of Food Drug Officials. USQ Bull. 18:33-35.
3. NRC (National Research Council). 1983. Risk Assessment in the Federal Government: MalriaglHg tWfe
Process. NAS Press, Washington, DC.
4. U.S. EPA. 1987. The Risk Assessment Guidelines of 1986. Office of Health and Environmental
Assessment, Washington, DC. EPA/600/8-87/045.
H-13
-------
APPENDIX I
CHEMICALS AVAILABLE IN IRIS
-------
Page No.
09/28/90
CAS Chemical
number Name
USE OMLY FOR SCREENING Values from IRIS 9/1/90
Consult IRIS for Update and Whenever Possible, Site Specific
Lipid, Consumption and Bioaccumulation Factors Should be
Used in Application of RAC in Regulatory Action.
RfD q1*
mg/kg/day /mg/kg/day
Estimated RAC (mg/l) RAC (mg/l)
BCF RL: 10E-6 RL: 10E-6
3% Lipid 0.0065 kg/day 0.020 kg/day
Consumption Consumption
50000 Formaldehyde
50293 p.p'-DDT
50328 Benzo[a]pyrene
51285 2,4-Dinitrophenol
55185 N-Nitrosodiethylamine
56235 Carbon tetrachloride
56359 Tributyltin oxide
56382 Parathion
57125 Cyanide, free
57249 Strychnine
57749 Chlordane
58899 gamma-Hexachlorocyclohexane
58902 2.3,4.6-Tetrachlorophenol
60297 Ethyl ether
60515 Dimethoate
60571 Dieldrin
62384 Phenylmercuric acetate
62533 Aniline
62737 Dichlorvos
62759 N-Nitrosodimethylamine
63252 Carbaryl
64186 Formic acid
65850 Benzoic acid
67561 MethaneI
67641 Acetone
67663 Chloroform
67721 Hexachloroethane
70304 Hexachlorophene
71363 n-Butanol
0.2 **
0.0005
** **
0.002
** 150
0.0007 0.13
0.00003
** **
0.02
0.0003
0.00006 1.3
0.0003
0.03
0.2
0.0002
0.00005 16
0.00008
** 0.0057
0.0008 0.29
** 51
0.1
2
4
0.5
0.1 **
0.01
0.001 0.014
0.0003
0.1
40000
10960
5.2
0.40
29.72
87.6
0.4
3804
146.8
416
0.77
0.4
32.04
0.58
0.84
0.4
0.4
12.2
4.92
0.4
0.4
5.56
700
40000
0.71
0.0001
4
0.0002
0.003
8
0.000002
0.00002
0.8
3000
6
0.00002
2
2
0.1
0.0005
90
9000
14000
20
0.001
0.00008
2000
0.00004
1
0.00006
0.0009
3
0.0000007
0.000007
0.3
900
2
0.000007
0.5
0.7
0.3
0.0002
30
3000
4000
6
0.0004
0.00002
500
-------
Page No.
09/28/90
CAS Chemical
number Name
USE ONLY FOR SCREENING Values from IRIS 9/1/90
Consult IRIS for Update and Whenever Possible, Site Specific
Lipid, Consumption and Bioaccumulation Factors Should be
Used in Application of RAC in Regulatory Action.
RfD q1*
rng/kg/day /mg/kg/day
Estimated RAC (mg/l) RAC (mg/l)
BCF RL: 10E-6 RL: 10E-6
3X Lipid 0.0065 kg/day 0.020 kg/day
Consumption Consumption
71432 Benzene
71556 1,1,1-Trichloroethane
72208 Endrin
72435 Hethoxychlor
72548 p,p'-DDD
72559 p.p'-DDE
74839 Bromomethane
74908 Hydrogen cyanide
75070
75092
75150
75252
75274
75354
75694
75718
75876
75990
76131
76448
77474
77781
78002
78488
78591
78831
78864
78933
Acetaldehyde
Dichloromethane
Carbon disulf ide
Bromoform
Bromodichloromethane
1,1-bichloroethylene
Trichloromonof luoromethane
Oichlorodif luoromethane
Chloral
Da Upon, sodium salt
CFC-113
Heptachlor
Hexachlorocyclopentadiene
Dimethyl sulfate
Tetraethyl lead
Herphos oxide
Isophorone
Isobutyl alcohol
2-Chlorobutane
Methyl ethyl ketone
**
0.09
0.0003
0.005
**
**
0.0014
0.02
u.vw6
**
0.06
0.1
0.02
0.02
0.009
0.3
0.2
0.002
0.03
30
0.0005
0.007
**
0.0000001
0.00003
0.2
0.3
**
0.05
0.029
**
**
0.24
0.34
**
0.0075
0.0079
0.6
4.5
**
**
0.0041
**
**
7.84
14.52
32.04
2564
12840
40000
1.13
Of
.4
0.4
1.54
11.92
7.16
7 44
13.36
6
3.23
2.31
63.2
692
744
0.4
9
0.56
15.5
0.4
0.05
70
0.1
0.02
0.000004
0.0000008
10
inn
cuo
0.9
0.1
30
0.002
200
400
7
100
5000
0.000004
0.1
0.3
6000
1000
0.01
20
0.03
0.007
0.000001
0.0000002
4
m
3u
0.3
0.04
10
0.0008
80
100
2
50
2000
0.000001
0.03
0.09
2000
400
-------
Page No.
09/28/90
CAS Chemical
number Name
USE ONLY FOR SCREENING Values from IRIS 9/1/90
Consult IRIS for Update and Whenever Possible, Site Specific
Lipid, Consumption and Bioaccumulation Factors Should be
Used in Application of RAC in Regulatory Action.
RfD q1*
mg/kg/day /mg/kg/day
Estimated RAC
-------
Page No.
09/28/90
USE ONLY FOR SCREENING Values from IRIS 9/1/90
Consult IRIS for Update and Whenever Possible, Site Specific
Lipid, Consumption and Bioaccumulation Factors Should be
Used in Application of RAC in Regulatory Action.
CAS Chemical
number Name ,
RfD q1*
rng/kg/day /mg/kg/day
Estimated RAC (ng/l) RAC (ng/l)
BCF RL: 10E-6 RL: 10E-6
3X Lipid 0.0065 kg/day 0.020 kg/day
Consumption Consumption
94746 HCPA 0.0005
94757 2,4-Dichlorophenoxyacetic acid 0.01
94815 HCPB 0.01
94826 4-(2,4-Dichlorophenoxybutyric acid 0.008
95487 o-Cresot 0.05
95498 o-Chlorotoluene 0.02
95501 1,2-Dichlorobenzene 0.09
95578 2-Chlorophenol 0.005
95658 3,.4-Dimethylphenol 0.001—
95943 1,2,4.5-Tetrachlorobenzene 0.0003
95954 2,4,5-Trichlorophenol 0.1
96184 1,2,3-Trichloropropane 0.006
98011 Furfural 0.003
98077 Benzotrichloride **
98828 Cumene 0.04
98862 Acetophenone 0.1
98953 Nitrobenzene 0.0005
99354 1,3,5-Trinitrobenzene 0.00005
99650 m-Dinitrobenzene 0.0001
100414 Ethylbenzene 0.1
100425 Styrene 0.2
100447 Benzyl chloride **
100527 Benzaldehyde 0.1
101213 Chlorpropham (CIPC) 0.2
101553 p-Bromodiphenyl ether **
101611 44'Hethylenebis(NN'dimethyl)aniline **
103231 Di-(2-ethylhexyl)adipate 0.7
103333 Azobenzene **
105602 Caprolactam 0.5
0.17
0.046
0.11
27.32
20.96
50.4
38.56
7.6
88
103
8.8
24.72
1404
176.4
5.84
0.54
283
138
2.82
4.92
1.93
3.08
66.8
29.24
21.5
2.42
96.4
2179
888
40000
175.2
0.4
0.2
5
2
2
70
2
9
6
Q.4
0.002
6
10
60
3
400
1
0.3
0.4
20
80
0.003
500
20
0.0003
0.2
0.0006
10000
0.06
2
0.7
0.7
20
0.8
3
2
0.1
0.0007
2
4
20
1
100
0.4
0.09
0.1
5
20
0.001
100
7
0.00009
0.06
0.0002
4000
-------
Page No.
09/28/90
CAS Chemical
number Name
USE ONLY FOR SCREENING Values from IRIS 9/1/90
Consult IRIS for Update and Whenever Possible, Site Specific
Lipid, Consumption and Bioaccumulation Factors Should be
Used in Application of RAC in Regulatory Action.
RfD q1*
mg/kg/day /mg/kg/day
Estimated RAC (mg/l) RAC (mg/l)
BCF RL: 10E-6 RL: 10E-6
3% Lipid 0.0065 kg/day 0.020 kg/day
Consumption Consumption
106376 1,4-Dibromobenzene
106445 p-Cresot
106478 p-Chloroaniline
106898 Epichlorohydrin
106934 1,2-Dibromoethane
106990 1,3-Butadiene
107028 Acrolein
107051 Allyl chloride
107062 1,2-Dichloroethane
107131 Acrylonitrile
107186 Allyl alcohol
107211 Ethylene glycol
107302 Chloromethyl methyl ether
108101 Methyl isobutyl ketone
108316 Haleic anhydride
108394 m-Cresol
108452 m-Phenylenediamine
108883 Toluene
108907 Chlorobenzene
108918 Cyclohexylamine
108941 Cyclohexanone
108952 Phenol
109693 1-Chlorobutane
110009 Furan
110543 n-Hexane
110861 Pyridine
111444 Bis(chloroethyl)ether
114261 Baygon
115297 Endosulfan
0.01
0.05 **
0.004
0.002 0.0099
** 85
**
** **
** **
** 910
** 0.54
0.005
2
** **
0.05
0.1
0.05 **
0.006
0.2 **
0.02 **
0.2
5
0.6
** **
0.001
** **
0.001
** 1.1
0.004
0.00005
181.2
7.6
5.32
0.4
3.76
5.08
0.4
2.20
2.26
0.4
0.4
0.4
0.4
1.38
7.6
0.4
25.52
28
1.92
0.69
2.33
15.5
1.75
179
0.53
0.98
2.81
0.6
70
8
3
0.00003
0.000005
0.05
100
60000
400
70
200
80
8
1000
80000
3000
6
20
0.01
20
0.2
20
2
0.9
0.00001
0.000002
0.02
40
20000
100
20
50
30
3
400
30000
900
2
7
0.003
5
-------
Page Ho.
09/28/90
CAS Chemical
number Name
USE ONLY FOR SCREENING Values from IRIS 9/1/90
Consult IRIS for Update and Whenever Possible, Site Specific
tipid, Consumption and Bioaccumulation Factors Should be
Used in Application of RAC in Regulatory Action.
RfD q1*
mg/kg/day /ntg/kg/day
Estimated RAC (mg/l) RAC (rog/l)
BCF RL: 10E-6 RL: 10E-6
3% Lipid 0.0065 kg/day 0.020 kg/day
Consumption Consumption
115322 Dicofol
116063 Aldicarb
117817 Bis(2-ethylhexyl)phthalate
118741 Hexachlorobenzene
118967 2,4,6-Trinitrotoluene
120127 Anthracene
120616 Dimethyl terephthalate
120821 1,2.4-Trichlorobenzene
120832 2,4-Dichlorophenol
121142 2,4/2,6-Dini trotbtuene mixture^
121697 N-N-Dimethylaniline
121755 Halathion
121824 RDX
122349 Simazine
122394 Diphenylamine
122429 Propham
122667 1,2-Diphenylhydrazine
123331 Haleic hydrazide
123911 1,4-Dioxane
124403 Dimethylamine
124481 Dibromochloromethane
126987 Methacrylonitrite
127184 Tetrachloroethylene
129000 Pyrene
131113 Dimethyl phthalate
131895 4.6-Dinitro-o-cyclohexyl phenol
133062 Captan
133073 Folpet
133904 Chloramben
**
**
0.02
0.0008
0.0005
0.3
0.1
**
0.003
**
0.002
0.02
0.003
0.002
0.025
0.02
**
0.5
**
**
0.02
0.0001
0.01
0.03
**
0.002
0.013
0.1
0.015
0.44
**
0.014
**
**
0.68
0.11
**
0.80
0.011
**
**
**
0.0035
9680
1.21
40000
18800
2.27
550
11.56
383.6
42
5.96
11.16
3.83
15.36
115.2
13
35.36
0.4
0.4
0.4
9.24
0.42
38.72
1280
2.51
800
8.24
0.000003
0.00002
0.0005
2
6
100
0.8
0.002
2
60
1
2
10
0.0004
10000
3
20
3
3
0.3
0.03
20
0.0000001
0.000006
0.0001
0.8
2
30
0.3
0.0009
0.6
20
0.5
0.8
4
0.0001
4000
0.8
8
0.8
0.9
0.08
0.009
6
-------
Page No.
09/28/90
CAS Chemical
number Name
USE ONLY FOR SCREENING Values from IRIS 9/1/90
Consult IRIS for Update and Uhenever Possible, Site Specific
Lipid, Consumption and Bioaccumutation Factors Should be
Used in Application of RAC in Regulatory Action.
RfD q1*
mg/kg/day /mg/kg/day
Estimated RAC (mg/l) RAC (mg/l)
BCF RL: 10E-6 RL: 10E-6
3% Lipid 0.0065 kg/day 0.020 kg/day
Consumption Consumption
137268 Thiram
139402 Propazine
141662 Bidrin
141786 Ethyl acetate
143339 Sodium cyanide
145733 Endothall
148185 Sodium diethyldithiocarbamate
150505 Nerphos
151508 Potassium cyanide
156605 trans-1,2-Dichloroethylene
206440 Fluoranthene
298000 Methyl parathion
298044 Disulfoton
300765 Naled
302012 Hydrazine/Hydrazine sulfate
309002 Aldrin
319846 alpha-Hexachlorocyclohexane
319857 beta-Hexachlorocyclohexane
319868 delta-Hexachlorocyclohexane
330541 Diuron
330552 Linuron
460195 Cyanogen
504245 4-Aminopyridine
506616 Potassium silver cyanide
506649 Silver cyanide
506683 Cyanogen bromide
506774 Chlorine cyanide
507200 t-Butylchloride
510156 Chlorobenzilate
0.005
0.02
0.0001
0.9
0.04
0.02
0.03
0.00003
0.05
0.02
0.04 **
0.00025
0.00004
0.002
** 3.0
0.00003 17
** 6.3
** 1.8
** **
0.002
0.002 **
0.04
** **
•0.2
0.1
0.09
0.05
** **
0.02
47.2
0.4
0.54
0.4
2.5
1280
25.48
60
65.9
1638
146.8
146.8
146.8
23.56
51.6
0.4
12.2
434
5
3
20000
600
90
0.3
0.1
0.007
0.3
0.0000004
0.00001
0.00004
0.9
0.4
0.5
1
0.9
6000
200
30
0.1
0.03
0.002
0.1
0.0000001
0.000004
0.00001
0.3
0.1
0.2
-------
Page Ho.
09/28/90
CAS Chemical
nutter Name
USE ONLY FOR SCREENING Values front IRIS 9/1/90
Consult IRIS for Update and Whenever Possible, Site Specific
Lipid, Consumption and Bioaccumulation Factors Should be
Used in Application of RAC in Regulatory Action.
RfD ql*
wg/kg/day /mg/kg/day
Estimated RAC (ng/l) RAC (mg/l)
BCF RL: 10E-6 RL: 10E-6
3% Lipid 0.0065 kg/day 0.020 kg/day
Consumption Consumption
oo
541731 1,3-Dichlorobenzene
542621 Barium cyanide
542756 1,3-Oichloropropene
542881 Bis(chloromethyl)ether
544923 Copper cyanide
556887 Nitroguanidine
556887 Nitroguanidine
557211 Zinc cyanide
563122 Ethion
563688 The It iurBcetate —
576261 2,6-Dimethylphenol
592018 Calcium Cyanide
598776 1,1,2-Trichloropropane
608731 tech-Hexachlorocyclohexane
608935 Pentachlorobenzene
615543 1,2,4-Tribromobenzene
621647 N-Nitrosodi-N-propylamine
630104 Selenourea
630206 1.1,1,2-Tetrachloroethane
709988 Propanil
732116 Phostnet
759944 S-Ethyl dipropylthiocarbamate
765344 Glycidyaldehyde
834128 Ametryn
886500 Terbutryn
924163 N-Nitroso-di-n-butylamine
930552 N-Nitrosopyrrolidine
944229 Fonofos
950378 Hethidathion
** **
0.07
0.0003 **
** 220
0.005
0.1
0.1 **
0.05
0.0005
0.00009 **
0.0006
0.04
0.005
0.003 1.8
0.0008
0.005
** 7.0
0.005
0.03
0.005
0.02
0.025
0.0004
0.009
0.001
** 5.4
** 2.1
0.002
0.001
103
2.94
1.05
191
24.72
17.4
146.8
5120
524
1.85
39.72
105.2
15.2
68.4
40.4
83.6
12.68
0.4
193.6
1
0.00005
0.03
0.3
3
0.00004
0.002
0.1
0.0009
8
0.5
10
4
2
0.1
0.0002
0.01
0.1
0.4
0.00002
0.009
0.08
1
0.00001
0.0005
0.03
0.0003
3
0.2
4
1
0.8
0.04
0.00005
0.004
0.04
-------
Page No.
09/28/90
USE ONLY FOR SCREENING Values from IRIS 9/1/90
Consult IRIS for Update and Whenever Possible, Site Specific
Lipid, Consumption and Bioaccumulation Factors Should be
Used in Application of RAC in Regulatory Action.
CAS Chemical
number Name
RfD q1*
mg/kg/day /mg/kg/day
Estimated RAC (mg/l) RAC (mg/l)
BCF RL: 10E-6 RL: 10E-6
3% Lipid 0.0065 kg/day 0.020 kg/day
Consumption Consumption
vo
957517 Diphenamid
961115 Tetrachlorovinphos
1024573 Heptachlor epoxide
1071836 Glyphosate
1116547 N-Nitrosodiethanolamine
1163195 Decabromodiphenyl ether
1314325 Thallic oxide
1314621 Vanadium pentoxide
1314847 Zinc phosphide
1330207 Xylenes
1332214 Asbestos
1336363 Polychlorinated biphenyls
1445756 Diisopropyl methyl phosphonate
1563662 Carbofuran
1582098 Trifluralin
1596845 Alar
1610180 Prometon
1646884 Aldicarb sulfone
1689845 Bromoxynil
1689992 Bromoxynit octanoate
1861321 Dacthal
1861401 Benefin
1897456 Chlorothalonil
1910425 Paraquat
1912249 Atrazine
1918009 Dicamba
1918021 Picloram
1918167 Propachlor
1929777 Vernam
0.03
0.03
0.000013
0.1
**
0.01
**
0.009
0.0003
2
**
**
0.08
0.005
0.0075
0.15
0.015
0.0003
0.02
0.02
0.5
0.3
0.015
0.0045
0.005
0.03
0.07
0.013
0.001
9.1
**
2.8
**
**
**
**
7.7
0.0077
**
10.72
64.8
13.72
0.4
40000
0.4
14.32
1784
0.4
40.8
0.4
36.36
5800
327.6
1784
178.4
26.96
14.08
11.56
16.72
179.2
30
5
0.00009
0.01
0.003
2000
4
0.0008
4000
4
8
6
0.04
20
2
0.9
2
20
70
9
0.06
10
2
0.00003
0.003
0.0009
700
1
0.0003
1000
1
3
2
0.01
5
0.6
0.3
0.6
7
20
3
0.02
-------
Page Ho.
09/28/90
10
CAS Chemical
number Name
USE ONLY FOR SCREENING Values from IRIS 9/1/90
Consult IRIS for Update and Whenever Possible, Site Specific
Lipid, Consumption and Biosccunulation Factors Should be
Used in Application of RAC in Regulatory Action.
RfD ql*
mg/kg/day /rog/kg/day
Estimated RAC (mg/l) RAC (mg/l)
BCF RL: 10E-6 RL: 10E-6
3% Lipid 0.0065 kg/day 0.020 kg/day
Consumption Consumption
1929824 Nitrapyrin
2008415 Butylate
2050477 p.p'-Dibromodiphenyl ether
2104645 EPN
2164172 Fluometuron
2212671 Holinate
2303175 Trial late
2310170 Phosalone
2312358 Propargite
2385855,Mirex
2425061 Captafol
2439103 Dodine
2691410 Octahydro-1.3,5,7-tetranitro-1,3.5.
2921882 Chlorpyrifos
3337711 Asulam
3689245 Tetraethyldithiopyrophosphate
5234684 Carboxin
5902512 TerbaciI
6108107 epsilon-Hexachlorocyclohexane
6533739 Thallium carbonate
7287196^ Rrometryn
7439921 Lead and compounds
7439965 Manganese
7439976 Mercury, (inorganic)
7440020 Nickel, soluble salts
7440144 Radium 226 and 228
7440144 Radium 228 (and 226)
7440224 Silver
7440360 Antimony
0.0015
0.1
** **
0.00001 "
0.013
0.002
0.013
0.0025
0.02
0.000002
0.002
0.004
0.05
0.003
0.05
0.0005
0.1
0.013
** **
0.00008 **
0.004
** **
0.1 **
**
0.02 **
** **
** 2.6E-5/pCi/L
0.003
0.0004
79.2
292
11882
648
10.08
35.28
388
752
800
0.4
3.37
5.4
146.8
71.2
0.2
4
0.0002
10
0.6
0.4
0.00003
0.04
1000
300
30
0.6
0.07
1
0.00005
5
0.2
0.1
0.000009
0.01
500
100
8
0.2
-------
Page No.
09/28/90
11
CAS Chemical
number Name
USE ONLY FOR SCREENING Values from IRIS 9/1/90
Consult IRIS for Update and Whenever Possible, Site Specific
Lipid, Consumption and Bioaccumulation Factors Should be
Used in Application of RAC in Regulatory Action.
RfD q1*
mg/kg/day /mg/kg/day
Estimated RAC (mg/l) RAC (mg/l)
BCF RL: 10E-6 RL: 10E-6
3% Lipid 0.0065 kg/day 0.020 kg/day
Consumption Consumption
7440382 Arsenic, inorganic
7440393 Barium
7440417 Beryllium
7440428 Boron (Boron and Borates only)
7440439 Cadmium
7440473 Chromium(VI)
7440508 Copper
7440611 Uranium, natural
7446186 Thallium(I) sulfate
7723140 White phosphorus
7773060 Ammonium sulfamate
7782414 Fluorine (soluble fluoride)
7783008 Selenious acid
7783064 Hydrogen sulfide
7791120 Thallium chloride
7803512 Phosphine
8001352 Toxaphene
8001589 Creosote
8007452 Coke oven emissions
8065483 Demeton
10102439 Nitric oxide
10102440 Nitrogen dioxide
10102451 Thallium nitrate
10265926 Hethamidophos
10453868 Resrnethrin
10595956 N-Nitroso-N-methylethylanrine
12035722 Nickel subsulfide
12039520 Thallium selenite
12122677 Zineb
**
0.07
0.005
0.09
**
0.005
**
**
0.00008
0.00002
0.25
0.06
0.003
0.003
0.00008
0.0003
**
**
**
0.00004
0.1
1
0.00009
0.00005
0.03
**
**
0.00009
0.05
*»
4.3
**
**
**
**
**
**
**
1.1
**
414
0.00002
0.000008
22
**
0.4
11900
0.4
1
0.03
0.001
0.4
0.009
0.0004
-------
Page No.
09/28/90
12
USE ONLY FOR SCREENING Values from IRIS 9/1/90
Consult IRIS for Update and Whenever Possible, Site Specific
Lipid, Consumption and Bioaccumulation Factors Should be
Used in Application of RAC in Regulatory Action.
CAS Chemical
number Name
RfD q1*
mg/kg/day /mg/kg/day
Estimated RAC (mg/l) RAC (mg/l)
BCF RL: 10E-6 RL: 10E-6
3% Lipid 0.0065 kg/day 0.020 kg/day
Consumption Consumption
12427382 Maneb
13463393 Nickel carbonyl
13S93038 Quinalphos
13684634 Phenmedipham
14797558 Nitrate
14797650 Nitrite
14859677 Radon 222
15299997 Napropamide
15972608 Alachlor
16065831 Chromium(III)
16672870 Ethephon
16752775 Methomyl
17804352 Benonryl
19044883 Oryzalin
19408743 Hexachlorodibenzo-p-dioxin mixture
19666309 Oxadiazon
20859738 Aluminum Phosphide
21087649 Hetribuzin
21725462 Cyanazine
22224926 Fenamiphos
22967926 Methyl mercury
23135220 Oxamyl
23564058 Thiophanate-methyl
23950585 Pronamide
24307264 Mepiquat chloride
25057890 Bentazon
25329355 Pentachlorocyclopentadiene
26628228 Sodium azide
27314132 Norflurazon
0.005
**
0.0005
0.25
**
0.1
**
0.1
0.01
1
0.005
0.025
0.05
**
**
0.005
0.0004
0.025
0.002
0.00025
0.0003
0.025
0.08
0.075
0.03
0.0025
**
0.004
0.04
1.8E-6/pCi/L
6200
42.8
124
156
226
31.1
3.64
21.5
2.92
33.92
0.1
20
7
0.5
6
0.1
300
20
0.04
7
2
0.2
2
0.04
100
8
-------
Page No.
09/28/90
13
CAS Chemical
mutter Name
USE ONLY FOR SCREENING Values from IRIS 9/1/90
Consult IRIS for Update and Whenever Possible, Site Specific
Lipid, Consumption and Bioaccumulation Factors Should be
Used in Application of RAC in Regulatory Action.
RfD q1*
mg/kg/day /rng/kg/day
Estimated RAC (mg/l) RAC (mg/l)
BCF RL: 10E-6 RL: 10E-6
3X Lipid 0.0065 kg/day 0.020 kg/day
Consumption Consumption
28249776 Thiobencarb
29232937 Pirimiphos-methyl
30560191 Acephate
32534819 Pentabromodiphenyl ether
32536520 Octabromodiphenyl ether
33089611 Amitraz
33820530 Isopropalin
34014181 Tebuthiuron
35367385 Oiflubenzuron
35554440 Imazelit
36483600 Hexabromodiphenyl ether
36734197 Iprodione (Rovral)
39148248 Fosetyl-al
39515418 Oanitol
39638329 Bis(2-chloroisopropyl) ether
40088479 Tetrabromodiphenyl ether
40487421 Pendimethalin (Prowl)
41851507 Chlorocyclopentadiene
42874033 Oxyfluorfen
43121433 Bayleton
43222486 Difenzoquat
49690940 Tribromodiphenyl ether
50471448 Vinclozolin
51218452 Metolachlor
51235042 Hexazinone
51630581 Pydrin
52315078 Cypermethrin
52645531 Permethrin
55285148 Carbosulfan
0.01
0.01
**
0.002
0.003
0.0025
0.015
0.07
0.02
0.013
**
0.04
3
0.0005
0.04
•*
0.04
**
0.003
0.03
0.08
**
0.025
0.1
0.033
0.025
0.01
0.05
0.01
0.0087
**
«*
71.6
66.7
0.4
5520
0.4
44.8
19.84
5170
21.9
3416
11040
39.28
252.4
348.4
35200
10320
40000
2
2
3
0.03
2000
5
7
0.001
20
0.1
0.003
8
4
1
0.008
0.01
0.01
0.5
0.5
1
0.01
600
2
2
0.0003
6
0.04
0.001
3
1
0.3
0.002
0.003
0.004
-------
Page No.
09/28/90
CAS Chemical
number Name
USE ONLY FOR SCREENING Values from IRIS 9/1/90
Consult IRIS for Update and Whenever Possible, Site Specific
Lipid, Consumption and Bioaccumulation Factors Should be
Used in Application of RAC in Regulatory Action.
RfD ql*
mg/kg/day /mg/kg/day
Estimated RAC (mg/l) RAC (mg/l)
BCF RL: 10E-6 RL: 10E-6
3X Lipid 0.0065 kg/day 0.020 kg/day
Consumption Consumption
55290647 Dimethipin
57837191 Metalaxyl
58138082 Tridiphane
59756604 FI undone
60207901 Propiconazole
60568050 Furmecyclox
62476599 Sodium acifluorfen
63936561 Nonabromodiphenyl ether
64902723 Chlorsulfuron
65195553 Avermectin B1
66215278 Cyromazine
66332965 Flutolanil
66841256 Tralomethrin
67485294 Amdro
67747095 Prochloraz
68085858 Cyhalothrin/Karate
68359375 Baythroid
69409945 Fluvalinate
69806402 Haloxyfop-methyl
72128020 Fomesafen
72128020 Fomesafen
74051802 Sethoxydim
74115245 Apollo
74223646 Ally
76578148 Assure
76738620 Paclobutrazol
77182822 Glufosinate-amnonium
77323843 Trichlorocyclopentadiene
77323854 Tetrachlorocyclopentadiene
0.02
0.06
0.003
0.08
0.013
**
0.013
**
0.05
0.004
0.0075
0.06
0.0075
0.0003
0.009
0.005
0.025
0.01
0.00005
**
»*
0.09
0.013
0.25
0.009
0.013
0.0004
**
**
**
0.030
**
0.15
0.19
0.19
**
**
24.1
1856
30
0.5
9
0.2
19.0
0.54
40000
10700
40000
40000
30
200
0.002
0.005
0.007
0.003
9
50
0.0007
0.002
0.002
0.0009
-------
Page No.
09/28/90
15
CAS Chemical
nunber Name
USE ONLY FOR SCREENING Values from IRIS 9/1/90
Consult IRIS for Update and Whenever Possible, Site Specific
Lipid, Consumption and Bioaccumulation Factors Should be
Used in Application of RAC in Regulatory Action.
RfD ql*
ing/kg/day /mg/kg/day
Estimated RAC (mg/l) RAC (mg/l)
BCF RL: 10E-6 RL: 10E-6
3X Lipid 0.006S kg/day 0.020 kg/day
Consumption Consumption
77501634
78587050
79277273
81335377
81335775
82558507
82657043
83055996
85509199
88671890
90982324
101200480
Lactofen
Savey
Harmony
Imazaquin
Pursuit
Isoxaben
Biphenthrin
Londax
NuStar
Systhane
Chlorimuron-ethyl
Express
0.002
0.025
0.013
0.25
0.25
0.05
0.015
0.2
0.0007
0.025
0.02
0.008
40000
0.004
0.001
------- |