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

-------