United State*
            Environmental Protection
            Agency
            Office of
            Water
            Washington DC 20460
Otllet at feMirch
md Dtvtiopnwnt
DRAFT
March 1991
            Water
vvEPA
Assessment and Control of
Bioconcentratable
Contaminants in Surface
Waters
                    OOOD91100

-------
e

•*«
B**

fc
 I

-------
                UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
                                WASHINGTON, D.C. 20460
                                                                             OFFICE OF
                                                                              WATER
Dear Colleague,


       Enclosed for your information and review is a copy of EPA's draft guidance entitled "Assessment
and Control of Biocohcentratable Contaminants in Surface Waters."  Also widbsed is the Federal Renter
Notice that made this document available for public review and comment  A summary of the draft guidance
document is provided in the Federal Register Notice.

       The overall purpose of this draft guidance document is to provide guidance for identifying, and
where necessary, controlling bioconcentratable organic compounds that may be present in effluents, nonpoint
source runoff, receiving waters, bedded sediments, dredged material,  and the tissues of aquatic organisms.
Such compounds can cause adverse human health and aquatic life impacts.  Identification of
bioconcentratable and bioaccumulative compounds in water, tissue and sediment can serve as the basis for
regulatory decisions concerning all sources of these pollutants.

       EPA solicits comment on each chapter in this document relating to th« oveialUpplicablity of the
procedures described and the practical utility of the approach for addressing concerns about
bioconcentratable pollutants in the Nation's waterways. In particular, EPA asks for public comment on two
areas that are known to be of considerable interest to the public and where additional comment would be
particularly helpful.  First, comments on the selection of dischargers for assessment would be useful in
determining the number and types of facilities which may be required to conduct this analysis.  Second,
comments are solicited on the capability and capacity of laboratories, both current and potential, which may
be involved in applying the analytical methods described in this draft  guidance.
       Tin ill iiitjiiiii nt i if Iliii iliunaiiil lim 1m 11 i runii i itiio i (Tint irfrtr Fin in nail nld Bin •( li
Laboratory in Duhith, Office of Health and Environmental Assessment in Ctncsmati, Office of Water
Regulations and Standards, and the Permits Division of the Office of Water Enforcement and Permits.
Should you have any questions regarding the document please contact me at 202/475-9545, or have your staff
contact William Morrow at 202/475-9531.
                                                           Sincerely,
                                                               lia C Dougherty, Director
                                                                 i Division (EN-336)
Enclosures
                                                                                  PriMid at RtcycUd Paptr

-------
                        GUIDANCE ON

                 ASSESSMENT AND CONTROL OF
               BIOCONCENTRATABLE  CONTAMINANTS
                      IN  SURFACE WATERS
                         March 1991

                           DRAFT
       Note:

       This draft document  contains  procedures
       which are currently  advisory  and subject to
       validation.  These include:   1)  Appendix B,
       "Laboratory Procedures  for Determining
       Bioconcentratable Chemicals in Aqueous
       Samples" and 2) the  inclusion of
       bioaccumulation factors (B^Fs) in the
       Chapter 3 formulas for.  calculating
       reference ambient concentrations (RACs) .
                              .
                              ,r: •
                     .
      United States Environmental .^Protection Agency
                    •  v •:     »,•  S •*
       National Effluqnt Toxicit^y Assessment Center,
        Environmental Research Laboratory  -  Duluth;
          Office of Watef"  Enforcement and Permits
        Office of Water Regulations and Standards
Office of Health and Environmental Assessment -' Cincinnati
                                    H.3. rrHre!


                                    Chicago jL' 606C-i-,

-------
                            Foreword
     Human consumption of fish and shellfish contaminated by
exposure to industrial and municipal discharges is a potential
toxic chemical exposure route of serious concern.  Protection of
human health is part of the full complement of water quality-
based controls for toxic pollutants.  This guidance document,
"Assessment and Control of Bioconcentratable Contaminants in
Surface Waters" (EPA 600/x-xx-xxx),  provides guidance to State
and Regional regulatory agencies in developing reference ambient
criteria and determining necessary controls to protect against
human health impacts due to consumption of contaminated fish and
shellfish.  It also contains the specific analytical procedures
for use in determining the presence, identity and concentrations
of bioconcentratable and bioaccumulative compounds in aqueous,
fish tissue and sediment samples.  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 guidance will be revised periodically to reflect
advances in the area of bioconcentratable pollutant control.
Comments from users will be welcomed.  They should be sent to the
Office of Water Enforcement and Permits (EN-338),
U.S. Environmental Protection Agency, 401 M. St. S.W.,
Washington, D.C. 20460.

-------
                         Acknowledgements


     This document results from the cooperative efforts of
individuals at the Environmental Research Laboratory, Duluth
(ERL-D); the Office of Water Enforcement and Permits  (OWEP); the
Office of Water Regulations and Standards (OWRS); and the Office
of Health and Environmental Assessment  (OHEA).  The principle
authors were Lawrence P. Burkhard (ERL-D), John R. Cannell
(OWEP), Katharine Wilson Dowell (OWEP), William J. Morrow  (OWEP),
Barbara Riedel Sheedy (AScI Corporation, ERL-D), and Rick Brandes
(OWEP).

The contributions and assistance of the following individuals in
the preparation of this document is gratefully acknowledged:

Donald I. Mount (AScI Corporation, ERL-D)
Hiranmay Biswas and Charles Delos (OWRS)
Linda R. Papa and Randall Bruins (OHEA)
Gilman D. Veith (ERL-D), Technical Advisor
Nelson A. Thomas (ERL-D), Project Leader
Robert Wood (OWEP), Validation Study Work Assignment Manager
Mildred Thomas (OWEP), Document Preparation
Jim Pendergast (OWEP)

The assistance of the Office of Marine and Estuarine Protection
and all reviewers is greatly appreciated.
                                11

-------
                        Table of Contents

     Foreword	i
     Acknowledgements	ii
     Table of Contents	iii
     Glossary	ix
     Abbreviations	xiii
     Introduction	xiv

1.   Approach to Assessment and Control of Bioconcentratable
     Contaminants

          1.1  Scope of Approach to Bioconcentration
               Assessment and Control	1-3
          1.2  Selection of Dischargers for Assessment....1-4
          1. 3  Tissue Residue Option	1-4
          1.4  Effluent Option	1-4
          1.5  Selection of Assessment Option	1-5
          1.6  Timing and Mechanisms for Assessment	1-9
          1.7  Field Validation of Bioconcentration
               Protocol	1-10
          1.8  Evaluation of Contaminants in Sediments....1-10

2.   Principles of Bioconcentration Control

          2.1  Concept of Bioconcentration and
               Bioaccumulation	II-l
          2.2  Concern for Bioconcentration and
               Bioaccumulation	II-2
          2.3  Bioconcentration Factors	II-2
          2.4  Bioaccumulation Factors	II-3
          2 . 5  Log P-Log BCF Relationship	II-3
          2.6  Measured versus Calculated
               Bioconcentration Factors	II-5
          2.7  Analytical Chemistry and
               Bioconcentration Control	II-5
          2.8  Chemicals of Highest Concern	II-7

3.   Assessment Options for Bioconcentratable Contaminants

          3.1  Tissue Residue Option	III-l
               3.1.1  Sampling Considerations	III-l
               3.1.2  Compositing of Organism Samples
                      for Analytical Procedure	III-4
               3.1.3  Extraction of Composite
                      Tissue Samples	III-5
               3.1.4  Gel Permeation Chromatography/Silica
                      Gel Cleanup of Tissue Extract	III-6
               3.1.5  Analyze via GC/MS	III-6
               3.1.6  Library Searching using the
                      Chemicals of Highest Concern
                      Mass Spectral Library	III-7

                               iii

-------
     3.1.7  5 Mg/kg Tissue Concentration
            Decision Point	III-7
     3.1.8  Library Searching using the
            EPA/NIH/NBS Mass Spectral Library.III-8
     3.1.9  Analytical Summary	III-8
     3.1.10 RfD, ql*, and/or Water
            Quality Standard?	III-9
     3.1.11 (Tissue Concentration x 4)
              > RTC?	III-9
     3.1.12 Issue Confirmation Request	111-10
     3.1,13 Outputs From Tissue Residue
            Assessment	III-ll
     3.1.14 Caveats for the Tissue
            Option	III-ll
3.2  Effluent Option	III-ll
     3.2.1  Sampling Considerations	III-ll
     3.2.2  Extraction of Effluent Sample	111-12
     3.2.3  Sample Cleanup	111-12
     3.2.4  Fractionation via HPLC	111-13
     3.2.5  Analyze via GC/MS	111-15
     3.2.6  Library Searching using the
            Chemicals of Highest Concern
            Mass Spectral Library	111-16
     3.2.7  100 ng/L Effluent Concentration
            Decision Point	111-17
     3.2.8  Library Searching using the
            EPA/NIH/NBS Mass
            Spectral Library	111-17
     3.2.9  (Fraction BCF x Concentration
              x Dilution) >  1 Mg/kg	111-18
     3.2.10  Analytical Summary	111-19
     3.2.11  RfD, ql*, and/or Water
             Quality Standard?	111-21
     3.2.12  (FM x BCF X concentration
               x dilution) > 1 Mg/kg	111-21
     3.2.13  (Concentration x dilution
                X 4) > RAC?	111-21
     3.2.14  Issue Confirmation Request	111-22
     3.2.15  Outputs From Effluent
             Assessment	111-22
     3.2.16  Caveats for the Effluent
             Option	111-23
3.3  Confirmation	111-23
     3.3.1   General Considerations	111-23
     3.3.2   Confirmation Process	111-24
3.4  Sediment Assessment	111-25
     3.4.1   General Considerations	111-25
     3.4.2   Sampling	111-25
     3.4.3   Sample Extraction	111-27
     3.4.4   Sample Cleanup.	111-28
     3.4.5   HPLC and GC/MS Analysis	111-28

                      iv

-------
               3.4.6   Library Searching using CHC
                       Mass Spectral Library	111-28
               3.4.7   5 Mg/kg Sediment Concentration
                       Decision Point	111-28
               3.4.8   Library Searching using the
                       EPA/NIH/NBS Mass Spectral
                       Library	111-29
               3.4.9   Analytical Summary	111-29
               3.4.10  Caveats for the Sediment
                       Analytical Procedure	111-29
          3.5  Quality Assurance/Quality Control for
               Analytical Procedures	111-29

4.   Reference Concentration Derivation

          4.1  Establishing RACs for Bioconcentratable
               Contaminants	IV-1
          4.2  Human Exposure Considerations	IV-2
          4.3  Fish Consumption Values	IV-2
          4 .4  BCF Evaluation/Selection	IV-3
          4.5  Bioaccumulation Considerations:
               Food Chain Multiplier Selection	IV-5
          4.6  Updating Criteria and Generating RACs
               and RTCs Using IRIS	IV-5
          4.7  Calculating RACs for Non-Carcinogens	IV-9
          4.8  Calculating RACs for Carcinogens	IV-10
          4.9  Calculating RTCs for Non-Carcinogens	IV-11
          4.10 Calculating RTCs for Carcinogens	IV-12
          4.11 Calculating RTCs from State Numeric
               Water Quality Standards	IV-12
          4.12 Deriving Quantitative Risk Assessment in
               the Absence of IRIS Values	IV-12

5.   Exposure Assessment and Wasteload Allocations

          5.1  Total Maximum Daily Loads	V-2
          5.2  Mixing Zones	V-2
          5.3  Point of Application of the Criteria	V-4
          5.4  Determining the Wasteload Allocation	V-4
               5.4.1  Steady-State Dilution Models	V-5
               5.4.2  Dynamic Models for Wasteload
                      Allocations	V-6
               5.4.3  Monte Carlo Simulation Models	V-6
          5.5  Averaging Periods for WLA' s	V-7
          5.6  Dilution Design Conditions for Freshwater
               Receiving Waters	V-8
               5.6.1  Carcinogens	V-8
               5.6.2  Noncarcinogens	V-12
          5.7  Lake, Marine and Estuarine Receiving
               Waters	V-12
               5.7.1  Marine and Estuarine Mixing	V-12

-------
               5.7.2  Approaches for Marine and
                      Estuarine Nearfield Models	V-14
          5.8  Exposure from Contaminants in Sediments....V-16

6.    Permit Limits for Control of Bioconcentratable
     Contaminants

          6.1  Basis for Effluent Characterization
               for Human Health	VI-1
               6.1.1  Recommended Approach for Effluent
                      Characterization	VI-2
               6.1.2  Determination of "Reasonable
                      Potential" for Excursions Above
                      RACs	VI-5
          6.2  The Wasteload Allocation as Basis for
               Permit Limit Derivation	VI-6
          6.3  Procedures for Developing Limits	VI-6
          6.4  Water Quality-based Permit Limit
               Derivation for Human Health Protection....VI-8
          6.5  Detection Limits for Compliance
               Monitoring	  VI-9
          6.6  Compliance Monitoring	VI-10
          6.7  Tissue Residue Monitoring	VI-12
          6.8  Data Generation Mechanisms	VI-12
                               VI

-------
                       Appendices

A.   Laboratory Procedures for Determining Bioconcentratable
     Chemicals in Tissue Samples

B.   Laboratory Procedures for Determining Bioconcentratable
     Chemicals in Aqueous Samples

C.   Laboratory Procedures for Determining Bioconcentratable
     Chemicals in Sediment Samples

D.   Chemicals Available on IRIS

E.   IRIS Values:  1) Reference Dose Description and Use in
     Health Risk Assessments and 2) EPA Approach for
     Assessing the Risk Associated with Exposure to
     Environmental Carcinogens

F.   Procedures for Determining the Harmonic Mean Flow

G.   Sample Permit Language

H.   Overview of Selected Available Tools
              t
I.   Field Validation Studies

J.   Example Results and Reports

K.   References
                          VI1

-------
Figure 1.1


Figure 3.1


Figure 3.2


Figure 3.3

Figure 3.4


Figure 4.1
               List of Figures

Approach for Assessment and Control of Biocon-
centratable Contaminants in Surface Waters	1-2

Identification of Bioconcentratable Pollutants
in Fish Tissue	III-2

Identification of Bioconcentratable Contaminants
in Effluents	111-14

RTCs for IRIS Chemicals with  log P > 3.5	111-20

Identification of Bioconcentratable Contaminants
in Sediments	111-26

Procedure for Revising an EPA Human Health
Criterion or Developing a Reference
Ambient Concentration	IV-8
Table 2.1

Table 4.1


Table 6.1


Table 6.2


Table 6.3
                List of Tables

Chemicals of Highest Concern	II-9

Estimated Food Chain Multipliers (FMs) for
Various log P Values	IV-6

Reasonable Potential Multiplying Factors:
95% Confidence above the mean	Vl-7

Reasonable Potential Multiplying Factors:
99% Confidence above the mean	VI-7

Multipliers for Calculating Maximum Daily Permit
Limits from Average Monthly Permit Limits	Vl-11
                                Vlll

-------
                              Glossary
aggregate risk
analyte
bioaccumulation
bioaccumulation
factor (BAF)
A methodology that estimates the total
number of both cancer and non-cancer cases
developed by populations exposed to
pollutants from fish or shellfish.

Chemical of interest for a specific
analytical procedure.

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

Measure of a chemical's tendency to
bioaccumulate.
bioconcentration
Uptake of substances by an organism from
the surrounding medium through gill
membranes or other external body surfaces.
bioconcentration
factor (BCF)
biomagnification
blank correction
carcinogenic potency
factor (ql*)
The measure of a chemical's tendency
to bioconcentrate.  The BCF is calculated
by dividing the concentration of the
chemical in the exposed organism's tissues
by the concentration of the chemical in
the exposure medium.

the process by which the concentration  of
a compound increases in different
organisms, occupying successive trophic
levels.

Subtraction of the signal,  output, or
response observed for the method blank
from that observed for the sample.  A
method blank is created by performing an
analysis by using all the glassware,
reagents, etc., that would be used with a
sample.

A factor indicative of a chemical's
human cancer-causing potential (upper 95%
confidence limit of the slope of a linear
dose-response curve).  ql*'s (pronounced
"que-one-star") are based on extrapolation
of high dose levels to low dose levels and
a lifetime exposure period.  The ql* is
expressed in units of reciprocal mg/kg/day
                                 IX

-------
chromatography
corrected retention
time
depuration
eluate
elution
food chain
harmonic mean flow
lowest observed effect
concentration
no observed effect
concentration

no observed adverse
effect level
and is also called the cancer potency
slope factor and the oral slope factor.

A method of separating and analyzing
mixtures of chemical substances by
preferential adsorption of chemical
components in ascending molecular sequence
onto a solid absorbance material.

The corrected retention time for an
analyte eluting from a chromatographic
column is the retention time of that
analyte minus the retention time of an
unretained analyte.  The corrected
retention time is expressed in units of
minutes.

The elimination of a chemical from an
organism via a complex interaction of
exchange processes across gills and
excretion via kidneys or bile.

The chromatographic fluid (normally, an
organic solvent) after passing through the
chromatography column.

The process of removing an absorbed
chemical by means of a solvent in a
chromatography column.

The scheme of feeding relationships by
trophic levels which unites the member
species of a biological community.

A long term mean flow value calculated by
retrieving several years of daily flow
records, taking the reciprocal of each
value, calculating the average, and taking
the reciprocal of the average.

The lowest dose that results in a level
statistically significant effect in the
test population.

The highest dose tested at which no level
effects are observed.

The highest dose tested at which there
level is no statistically or biologically
significant increase in adverse effects.

-------
n-octanol/water
partition coefficient
reference ambient
concentration (RAG)
reference tissue
concentration (RTC)
reference dose (RfD)
reverse-search
signal-to-noise ratio
The ratio at equilibrium, in a two-phase
system of a chemical in the n-octanol
phase to that in the water phase.

The concentration of a bioconcentratable
chemical in water which will not cause
adverse impacts to human health.  The RAG
is expressed in units of mg/L.

The concentration of a chemical in
edible fish or shellfish tissue which will
not cause adverse impacts to human health
when ingested.  The RTC is expressed in
units of mg/kg.

An estimate (with uncertainty spanning
perhaps an order of magnitude) of the
daily exposure during a lifetime to the
human population (including members of
sensitive subgroups) that is likely to be
without appreciable risk of deleterious
effect.  The RfD is expressed in units of
mg/kg/day.

A library searching technique for
identifying components detected during a
GC/MS analysis.  Identification is
performed by comparing a mass spectrum of
the unknown component to mass spectra
stored in reference libraries on the GC/MS
data system.  With reverse searching,
similarity between the unknown and
reference spectra is measured by how well
the unknown is included in the reference
mass spectrum.

The ratio between the amplitude of the
analyte to the amplitude of the background
noise signal at the same point.
trophic level
wasteload allocation
One of the successive levels of
nourishment in a pyramid of numbers, food
web, or food chain;  plant producers
constitute the first (lowest) trophic
level, and dominant carnivores constitute
the last (highest)  trophic level.

The portion of a receiving water's total
maximum daily pollutant load that is
allocated to one of its existing or future
point sources of pollution.
                                 xi

-------
xenobiotic                 Synthetic organic and organometallic
                           chemicals foreign to living organisms.
                                 xii

-------
                            Abbreviations
BAF
BCF
CAS #
CHC
FM
GC/MS
HEAST
HPLC
IRIS
LOAEL

log 10
log P

LOEC
MCL
NOEC
PHN
NOAEL

NPDES
QSAR
ql*
RAC
ETC
RfD
RT
RL
RWC
WLA
7Q10
30Q5
bioaccumulation factor
bioconcentration factor
Chemical Abstract Services Registry Number
Chemicals of Highest Concern
food chain multiplier
gas chromatography/mass spectrometry
Health Effects Assessment Summary Tables
high pressure liquid chromatography
Integrated Risk Information System
lowest observed adverse effect level (human/animal
toxicology)
logarithm base 10
log of octanol/water partition coefficient (also known as
log Kow)
lowest observed effect concentration
maximum containment level
no observed effect concentration
public health network
no observed adverse effect level (human/animal
toxicology)
National Pollutant Discharge Elimination System
Quantitative Structure Activity Relationship
carcinogenic potency factor
reference ambient concentration
reference tissue concentration
reference dose
retention time
risk level
receiving water concentration
wasteload allocation
seven day average low flow with a ten year return period
thirty day average low flow with a five year return
period
                                 xiii

-------
                           INTRODUCTION

     The EPA surface water toxics control program is designed to
provide the necessary regulatory controls to assure the
protection of aquatic life, wildlife, and human health.  To
implement this program, EPA has developed whole effluent
toxicity, chemical specific criteria, and bioassessment
procedures to protect aquatic life.  EPA has also initiated and
continues to develop procedures to protect wildlife.

     EPA has also developed criteria and control procedures to
protect human health from adverse exposure to contaminants in
drinking water and consumed fish or shellfish.  It is this latter
exposure from consumption of contaminated fish and shellfish
which is the focus of this document.

     During the past three decades, problems with chemical
residues in fish, shellfish, and wildlife have been prominent
throughout the country.  Chemicals such as DDT, PCB's, mercury,
tributyltin and 2,3,7,8-dibenzo-p-dioxin have received
substantial attention and have resulted in fish consumption
advisories and/or banning of commercial fish and shellfish
harvesting.

     Many studies have indicated that fish and shellfish
consumption may be a major human exposure route to some
bioconcentratable chemicals [1].  Ingestion of contaminated fish
and shellfish poses serious health risks to the general public
and this route of exposure currently has little if any specific
controls.

     Historically, water quality-based toxics control has relied
on chemical-specific effluent limitations to protect human
health.  However, only those chemical-specific limits based upon
human health water quality criteria for fish consumption can be
usually preventive of chemical residue formation in fish or
shellfish tissue.  Procedures to amend EPA's 1980 human health
water quality criteria for fish consumption have been developed
and the criteria amended.  The number of water quality criteria
is small relative to the large number of toxic pollutants being
discharged to surface waters.


Approach Description

     This document provides guidance for the control of
bioconcentratable pollutants in effluents including those not
presently controlled by water quality criteria.  This objective
is accomplished by presenting analytical procedures to identify
and quantify bioconcentratable pollutants in environmental
samples, by presenting procedures for deriving criteria for
aquatic organisms and criteria for receiving waters, and by

                               xiv

-------
presenting permitting guidance for the control of these
pollutants from point sources.  Two options are described.  For
one, the effluent is screened for the presence of chemicals that
bioconcentrate.  If present, a determination is made as to
whether the concentrations are sufficiently high to cause residue
problems.  Alternatively, residues in aguatic organisms can be
measured and if they are above the permissible concentrations,
the sources are traced using target analysis and then limited as
for the above approach.  If no residues of concern are found, no
further action is needed except for periodic checks for emerging
problems.  By integrating these separate aspects, an unified
approach has been developed for assessment and control of
bioconcentratable chemicals in surface waters.

     Detailed analytical procedures are provided for three
environmental matrices, water, sediment, and tissue.  Therefore
environmental media such as effluents, biological tissues,
sediments, dredge materials, stormwater, leachates, groundwater,
and non-point runoff samples can be analyzed for the presence of
bioconcentratable chemicals.  These procedures are presented in a
form suitable for dispersement to analytical laboratories for
performing the analysis.


Approach Objectives

     Application of this approach provides information on the
presence of bioconcentratable chemicals which are present at
concentrations which may cause problems from the consumption of
fish by humans.  Specifically, the results of this approach
include the identity and concentrations of bioconcentratable
chemicals in a sample for which human health criteria or risk
information is available enabling the implementation of controls.
Other bioconcentratable chemicals, for which human health risk
information is not readily available, will also be identified.
This information can be used to focus the development of human
health risk information on chemicals shown to be present in the
environment at potentially hazardous levels.  The approach will
also identify any chemicals present in a sample at very high
concentrations, which represent a potential for bioconcentration,
as well as any other chemicals which are present but which do not
represent a significant potential for bioconcentration.  This
information will be useful for water guality regulatory
authorities for better understanding the environmental risks from
bioconcentratable contaminants and for determining the need for
regulatory controls for these chemicals.
                               xv

-------
Approach Merits

     The approach in this document complements the existing
chemical-specific criteria approach and is recommended for the
following reasons:

•    A broad screen approach is achieved in a cost effective
     manner that does not require a prior knowledge of the
     presence of specific chemicals.  It thus avoids over
     reliance on chemical by chemical screening for targeted
     pollutants which cannot provide as complete a
     characterization of bioconcentratable contaminants in an
     effluent.

•    The approach focuses regulatory attention on those chemicals
     which bioconcentrate in the environment.  In using this
     approach, patterns and trends may be discerned and new or
     previously unknown problem toxicants can be identified and
     controlled.

•    The approach provides a standardized assessment methodology
     with broad applicability which offers the same sample
     handling, detection levels, and consistency of results for
     different applications (effluents, fish tissues, and
     sediments).

•    The approach can be used to assess bioconcentration hazards
     for wildlife and for the protection of wildlife for which
     criteria are available.

•    The approach provides a means to better prioritize
     permitting resources.  If contaminants are not found at
     levels of concern, then that particular source can be
     considered a lower priority for establishing controls.

•    The approach provides procedures for assessing and
     controlling complex mixtures.  The identification of those
     chemicals present in a complex mixture such as an effluent,
     which pose a threat to human health,  enables the additive
     risk from those chemicals to be addressed in the development
     of reference ambient concentrations.

     Depending on the application, the analytical procedures
which are presented may be conducted by either regulatory
authorities or by the regulated community.   The results will be
useful to regulatory authorities in making decisions on the
selection of control methods for both the remediation and the
prevention of environmental impacts from bioconcentratable
pollutants.  For example, a dredged materials sample might be
analyzed to determine if bioconcentratable contaminants are
present which would pose a problem for a particular disposal
                               xvi

-------
practice.  This information would be useful in deciding on an
appropriate regulatory action for those dredged materials.


Limitations of Approach

     Although these analytical methodologies can be used for
assessing a wide range of samples from different media and
different sources, it was developed primarily for the assessment
and control of bioconcentratable pollutants discharged in
effluents of NPDES permittees.  Guidance is provided for NPDES
permitting authorities on selecting dischargers which may be
required to conduct this analysis, for requiring and interpreting
the results of the effluent and tissue residue procedures and for
implementing permit requirements for the control of
bioconcentratable contaminants.  The document provides permittees
and laboratories required to conduct these evaluations with the
analytical procedures to identify, quantitate and confirm
individual bioconcentratable pollutants.  These procedures are
appropriate for use in freshwater, estuarine and marine
applications.

     The analytical procedures contained in this document will
not detect all bioconcentratable chemicals which form residues in
aquatic organisms.  The analytical procedures will only detect
nonpolar organic chemicals which can be successfully analyzed
using GC/MS.  Some of the general limitations of the approach are
the sensitivity of the analytical method selected and the ability
to identify unknown GC/MS components.  Analytical limitations
unique to each option; tissue, effluent, and sediment; are
described in each of the corresponding sections of Chapter 3
(Sections 3.1.14, 3.2.16, and 3.4.10).

     This document does not provide for an entire human health
risk assessment for bioconcentratable pollutants, rather it
focuses only on the fish consumption exposure route.  The
procedures to address human health effects of contaminated fish
consumption that are described within this document are
consistent with existing EPA risk assessment methodology.

     This guidance document is intended to supplement EPA's water
quality based toxics control program.  As such/ the guidance and
recommendations found in this document are not mandatory.


Contents of Document

     Chapter 1, Approach to Assessment and Control of
Bioconcentratable Contaminants, defines the scope of this
guidance document and discusses the assessment options.
                              xvii

-------
     Chapter 2, Principles of Bioconcentration Control, provides
background information and introduces bioconcentration,
terminology, processes, and methods.

     Chapter 3, Assessment Options for Bioconcentratable
Contaminants, provides a detailed description of tissue and
effluent options and their analytical procedures.  Chapter 3 also
provides discussions on generating chemical analytical data,
quality assurance, and sediment analysis.

     Chapter 4, Reference Concentration Derivation, provides
procedures for deriving RTCs and RACs for bioconcentratable
pollutants identified by the analytical procedures used for the
assessment.  RTCs and RACs are used to determine the need for
controls and for the derivation of effluent limitations.

     Chapter 5, Exposure Assessment and Wasteload Allocations,
discusses approaches to wasteload allocation development under a
variety of discharge situations.  Discussions focus on receiving
water characteristics such as available dilution and mixing, type
of pollutant(s) being discharged, data availability, and data
generation.

     Chapter 6, Permit Limits for Control of Bioconcentratable
Chemicals, provides guidance on determining reasonable potential
for exceedence of RACs, and the development of NPDES permit
limits specific to the control of bioconcentratable chemicals and
the protection of human health.

     The appendices contain the technical specifications for the
laboratory procedures along with other supporting information.
                              xvi 11

-------
                            Chapter 1

              Approach to Assessment and Control of
                  Bioconcentratable Contaminants
     A generalized flowchart for this approach to the assessment
and the control of bioconcentratable contaminants in surface
waters is presented in Figure 1.1.  This flowchart presents a
conceptual overview of the major steps and decision points
contained in the approach described in this document.  Each of
the components of this overall process are described in detail in
the corresponding sections of the document.

     The approach illustrated in Figure l.l is a seven step
procedure.  These steps are:  1) selection of dischargers or
receiving waters for assessment, 2) selection of the appropriate
assessment option, effluent bioconcentration or tissue residue
option, 3) analysis of tissue or effluent samples for
bioconcentratable chemicals, 4)  calculation of reference tissue
concentrations (RTCs)  and/or reference ambient concentrations
(RACs) for the identified bioconcentratable contaminants, 5)
development of wasteload allocations, 6) determination if
concentrations are present which have the reasonable potential to
pose health risks for human consumers of fish and shellfish, and
if so, 7) permit limit development.

     Depending on the application of this approach,  the
regulatory authority may require a discharger to conduct step 3,
the effluent or tissue residue assessment options,  or these
assessment options may be utilized by the regulatory authority.
An analytical chemistry laboratory with residue chemistry and
GC/MS capability will be needed to conduct the analytical methods
for effluent and tissue bioconcentratable chemical identification
and the confirmation of the identified chemicals.   The specific
step-by-step laboratory method instructions are contained in the
appendices to this document.

     The recommended data interpretation procedures to be
followed by the regulatory authority in reviewing the reported
chemical analytical results are contained in the discussion of
the assessment options in Chapter 3.   In requiring a discharger
to conduct these assessments the regulatory authority should
specify what information and results the discharger needs to
generate and report.   This should include information on sampling
and sample handling as well as the other QA/QC information that
is specified in the methods appendices.
                               1-1

-------
                                                              Figure 1.1
           Procedure for Assessment and Control of Bioconcentratable Contaminants in Surface Waters
Chapter 3
Ch.ipier 4
Chapter 5
Chapter 6

                                                    Select Dischargers and/or Receiving Waters

                                                           Select Assessment Option
                           Effluent Bioconcentraiioa Option
                                     Tissue Residue Option
                                                                                                           No Further Regulatory Action
Bioconcentraubie ConUmirunU Present?
                                                             Develop RAC/RTC
                                                          Develop Waste Load Allocation
                                                                                                            No Further Regulator \.
                                                        Does Reasonable Potential Exist?
                                                             Develop Permit Limits
                                                              Conduct Monitoring

-------
     Once compounds that bioconcentrate are identified, several
pieces of information may need to be determined if water quality
standards for those compounds are not in place.  For the tissue
residue option the Reference Tissue Concentration (RTC) for each
contaminant must be developed to determine if unacceptable
concentrations are present.  For the effluent option, a Reference
Ambient Concentration (RAC) is developed.  The RAC is similar to
a water quality criterion for human health.  The RTCs and RACs
are based on each specific bioconcentratable chemical's toxicity
and risk to humans.  Chapter 4 provides a discussion of the
information needed, and the procedures for development of RTCs
and RACs.

     If the tissue residue option rather than the effluent
screening option has been used, the RTC is first used to screen
for the presence of potentially hazardous concentrations of the
bioconcentratable chemicals in fish tissue.  (For the sake of
simplicity, the remainder of this document the term fish is
generally used to mean both fish and shellfish).   If this proves
to be the case, then target chemical analysis for those chemicals
must be done on effluent samples and the RAC calculated and
utilized as described above.

     Following the development of RTCs and/or RACs,  the approach
described in this document proceeds through the wasteload
allocation and, if needed, permit limit development.  These more
traditional pollutant control procedures follow the guidance
provided in the Technical Support Document and are discussed in
the context of this approach to the control of bioconcentratable
contaminants in Chapters 5 and 6.


1.1  Scope of Approach to Bioconcentration Assessment and
     Control

     The approach in this document identifies and controls
contaminants in effluents, and contaminants in other aqueous
samples, capable of forming fish tissue residues based upon the
tendency of the compound to bioconcentrate.  Chemicals that
bioconcentrate include organic compounds, and a small number of
metals and organometals.  With the tissue residue option,  the
approach described in this document is limited to nonpolar
organic chemicals which produce measurable chemical residues in
aquatic organisms.  With the effluent option,  the approach is
limited to nonpolar organic chemicals with characteristics which
cause these compounds to bioconcentrate, i.e.  log P values
greater that 3.5.  This threshold value of log P > 3.5 is
discussed in Section 3.2.4.  This approach does not address other
types of chemicals known to bioconcentrate, such as metals (e.g.
mercury, selenium) and organometals (e.g. tributlytin).  Also,

                               1-3

-------
this approach may not detect the presence of some compounds, such
as dioxin, which can form unacceptable residues at very low
exposure concentrations (i.e. below the method detection level,
see discussion in Section 2.7).


1.2  Selection of Dischargers for Assessment

          Guidelines are necessary to help NPDES permitting
     authorities prioritize dischargers for assessment.  At
     this time, the EPA is soliciting comments on the
     selection of point source dischargers for assessment.
     The final document will provide recommendations for the
     selection process.


1-3  Tissue Residue Option

     The tissue residue option measures the concentrations 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 recommends that the
identity of those contaminants then be confirmed prior to taking
subsequent action.

     In order for a tissue residue analysis to accurately assess
the effects of a given discharge of bioconcentratable
contaminants in an effluent it is essential for the tissue sample
analyzed to be representative of a long term exposure to the
effluent.  For this reason the ambient sampling for this option
must be carefully designed and the tissue residue option also
recommends target chemical analyses of the associated effluents
for the specific residue chemicals identified in the tissue
samples from the receiving water.  The tissue residue option may
be applied to measure residues in organisms which arise from
other sources of the chemical to the receiving water.  These
sources may include nonpoint sources, sediments, and any other
upstream point source dischargers.


1.4  Effluent Option

     The effluent option measures the concentrations of organic
bioconcentratable chemicals in effluent samples from point source
dischargers.  This analysis involves the collection of effluent

                               1-4

-------
samples, the extraction of the organic chemicals from the
effluent sample, and the separation of the chemicals which have
characteristics known to result in bioconcentration from the
other chemical components of the effluent sample.  This
separation is achieved by way of an analytical chemistry
methodology called high pressure liquid chromatography (HPLC).
The use of HPLC also enables the fractionation of the effluent
sample into three sub-samples or "fractions".  These three
fractions would 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 bioconcentratable contaminants.  The
effluent procedure also provides recommendations to sort the
results of the initial screening analysis in order 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.

     It is important to recognize that these effluent
bioconcentration analysis procedures are subject to a number of
basic principles and assumptions.  These principles and
assumptions, described in Chapter 2, provide a number of
constraints on the application of the analytical procedure and
should be recognized and understood in order to appropriately
conduct and interpret the results of the procedure.  These
underlying principles also hold for the application of this
approach to other sources (i.e. dredged materials)  from which
aqueous samples can be extracted.  It is also important to note
that the collection of effluent samples is subject to the effects
of effluent variability.  In order to accurately assess an
effluent with high variability, it may be necessary to collect
and perform this analysis on a greater number of samples.


1.5  Selection of Assessment Option

     While either of the assessment options described above may
be utilized for a given discharger, generally one of these
options will be preferred by the regulatory authority for an
initial assessment.  The regulatory authority should select the
assessment approach based on the available site and facility
specific information and the objectives of each application.

     In general, EPA recommends that a discharger be required to
conduct the effluent option if existing fish tissue and/or
facility information suggests the potential presence of
bioconcentratable contaminants.  Examples of this are waters
under a fishing ban due to bioconcentratable pollutants,  or an
organic chemical facilities known to manufacture

                               1-5

-------
bioconcentratable chemicals.  In these cases, there exists a
strong possibility for the bioconcentration of pollutants in fish
tissues to unsafe levels and the effluent option might be used to
determine if a point source discharger is in fact a contributing
source of these types of pollutants.

     EPA recommends that the tissue residue option be required if
the objective of the regulatory authority is to assess existing
ambient bioconcentration or bioaccumulation problems in the
absence of existing water body or facility information on the
presence of these contaminants.   In these cases, an overall
assessment of ambient exposure is needed.  The tissue residue
option allows for a direct assessment of the ambient conditions
which may include the effects from multiple sources.  For
example, for certain waterbodies one species of fish may be of
predominant concern (e.g. salmon) and this option might be
selected to determine the identities of any bioconcentratable
contaminants which may be present.  It may also be used for trend
analysis in determining the effectiveness of any previous
controls.

     The selection by the regulatory authority of an assessment
option for a given discharger will, to a large extent, be
determined by the site specific circumstances of each application
and the specific objectives or questions which the assessment is
being required to address.  The selection of the appropriate
option will greatly increase the utility of the analytical data
generated.  The trade offs inherent in the options must be
understood in order to make this selection.   The following
discussion compares these options and is intended to assist in
this selection.

     The tissue residue option tends to assess problems due to
bioconcentration on a receiving water basis and the effluent
option on a discharge by discharge basis.  The tissue residue
option measures existing residues in indigenous organisms, while
the effluent option examines effluents for chemicals with the
known potential to bioconcentrate.  Both approaches will provide
information on the presence and identity of bioconcentratable
chemicals and may be used to base controls on these contaminants.


     The tissue residue option measures existing chemical
residues in indigenous organisms sampled from the receiving water
for an effluent discharge.  The residues measured in these
organisms may arise as a result of some or all of the sources of
a particular chemical to the receiving water.  This could include
loadings from multiple point source discharges, any nonpoint
sources of the chemical and sediments.  Consequently, an existing
residue found in the tissue of the indigenous organism might have
no relationship to a given discharger or this discharger may be

                               1-6

-------
only partially responsible for the presence of the contaminant in
the tissue sampled.  In order to tie a specific discharger to
those chemical residues found, the tissue residue option includes
the recommendation to conduct follow up target analyses of
effluent samples for those specific chemicals.

     The effluent option begins with a selected discharger and
directly determines the presence and concentrations of
bioconcentratable chemicals in the effluent.  This assessment
option does not integrate multiple point sources discharges, nor
does it incorporate nonpoint sources and sediments.  If the
regulatory authority's primary objective is to assess the
cumulative effects of these sources then the tissue residue
option is the more appropriate initial approach.  In this way the
total amount of the contaminants from these sources which result
in tissue residues can be determined and the total loading can be
controlled by allocation among these multiple sources.

     The effluent option may also be used to assess multiple
point source discharges by requiring each discharger to conduct
the analyses.  The results of these assessments could then be
used in setting controls, either through the traditional single
source wasteload allocation process (which may not adequately
account for the multiple source loadings) or by developing a
multiple source wasteload allocation for those selected
dischargers.  This approach would not directly incorporate
loadings from nonpoint sources or sediments (unless these
assessments are performed separately)  and therefore in some
cases, may not result in controls which are stringent enough to
totally prevent the formation of tissue residues.  However, this
is not to say that this approach would not be effective in
developing controls for the selected discharges, only that the
level of control which is set may not factor in the other sources
mentioned.

     Another distinction between the two assessment options
concerns whether the objective is primarily to determine if there
are existing problems in a waterbody or if a specific discharger
is causing, or may in the future cause such a problem.  The
tissue residue option is limited to those contaminants already
existing in indigenous organisms which are sampled and which can
be identified in the target chemical effluent analyses.  The
tissue residue option cannot prevent residue problems due to new
chemicals, either new to the receiving water or new to the
organism sampled, because the option can only detect chemicals
which have had time to form a residue.  For most chemicals, a
continuous laboratory exposure of 28 days is used to determine
measured bioconcentration factors.  The effluent option may
identify these compounds as well as any additional chemicals in
the effluent with the potential to bioconcentrate.  Because of
this, the effluent option may prevent tissue contamination from

                               1-7

-------
occurring as well as assessing existing problems.  Whichever
option is selected, setting controls on point source discharges
will reguire the calculation of an RAC based on the chemical's
BCF and a food chain multiplier which are described in Chapters 2
and 4.

     The tissue residue option may provide greater sensitivity
than the effluent option for those chemicals with large BCFs and
which are present at very low concentrations in a given effluent.
This enhanced sensitivity for the residue option exists due to
the organism concentrating those chemicals over time from the
receiving water.  Of course this increased concentration will
only occur in organisms which have been exposed to the chemicals
from a discharge and requires the development of a sampling
requirements with this point in mind.  For example, for
discharges confined to small streams and rivers, a time period of
one to two months may be necessary for the residue concentration
in the organism to reach equilibrium.  This time period could be
much greater for discharges of a chemical to larger bodies of
water.

     The tissue residue option may detect a wider range of
residue forming chemicals than the effluent option.  This is due
to the analytical techniques required in the effluent option to
simplify the sample and remove the non residue forming chemicals
from the effluent extract.  Unfortunately, these procedures may
also cause some chemicals which do form residues in organisms to
decompose.  This clean up of the sample extract is not required
for the tissue option since the organism itself, via the uptake,
depuration and metabolic processes, will have eliminated the non-
residue forming chemicals from the tissue prior to extraction.
For this reason the effluent option may detect a narrower range
of residue forming chemicals.

     Another limitation of the effluent option also arises as a
result of the analytical methods used.  Hydrocarbons, such as
those found in lubricants, oils and gasoline, are not removed by
the aforementioned clean up step.  These chemicals rarely form
residues in aquatic organisms but do cause interferences in the
analyses.  Specifically, these types of compound prevent
successful GC/MS analysis of the third fraction of the effluent
extracts.  For this reason, application of this option to
discharges expected to contain very large numbers of
hydrocarbons, such as refineries, is not recommended.  However,
since this type of chemical does not form residues, the tissue
residue option is not subject to this analytical interference and
may be applied.

       A final consideration in the selection of the assessment
option is the complexity for implementation of the two options.
The analytical procedures used in the tissue residue option are

                               1-8

-------
somewhat less extensive than,those for the effluent option since
the extraction method is simpler and the use of HPLC
fractionation is not required.  However, this is somewhat offset
by the more elaborate field sampling design and implementation
which may be required for the tissue residue option in comparison
to the collection of effluent samples for the effluent option.


1.6  Timing and Mechanisms for Assessment

     EPA recommends that for an initial assessment the effluent
bioconcentration evaluation and/or fish tissue evaluation be
conducted by the selected permittees from one to four times over
a period of a year.  If the effects of seasonality or effluent
variability are of relatively low concern, then a sampling
frequency of once per year would be appropriate.  On the other
hand, if seasonal or effluent variability are of concern, these
assessments should be scheduled accordingly more frequently, four
times per year, to address this variability.  The sampling
results should be recorded and used for the effluent
characterization step of the permitting process (described in
Chapter 6).  Since average concentrations are of most concern,
composite rather than grab samples should be used in the
assessment.

     In order for the regulatory authority to make a
determination on the need to develop permit limits for
bioconcentratable contaminants for a given facility at the time
of permit reissuance, the permittee would need to be required to
conduct these assessments one year in advance of permit
reissuance.  This would allow time for the required samples and
analyses to be conducted and the results submitted to the
regulatory authority prior to the time of permit reissuance.

     Alternatively, the requirement to conduct these assessments
may be placed in the permit at the time of reissuance and if
limits are determined to be needed, then the permit may be
reopened or the limits may be placed in the permit at the next
reissuance.  Effluent or fish tissue evaluations may also be
required in permits annually if the regulatory authority has
reason to believe a change in process or discharge may occur
which would result in the appearance of new chemicals not found
in the initial screening.

     The regulatory authority should determine which of these
time frames is most appropriate for a given facility based on the
site specific information available for that discharge.  For
dischargers that are considered of high priority for this
assessment, EPA recommends dischargers be required to begin to
conduct these analyses in advance of permit reissuance and
provide the results for review at the time of permit reissuance.

                               1-9

-------
1.7  Field Validation of Bioconcentration Protocol

     Because the regulatory application of this assessment
procedure will direct regulatory decisions on the control of
bioconcentratable pollutants,  EPA has designed and implemented a
series of field applications to establish the validity of this
approach.  The field validation study is designed to show that
the bioconcentration procedures are correlated to the
bioconcentratable contaminants identified in the effluent
discharges, and with the approximate concentrations in organisms
collected at the associated discharge sites.  The validation
studies will be carried out at a series of sites with both
saltwater and freshwater receiving waters.  A more detailed
description of the study designs and of the results of these
field validations is contained in Appendix I.

     EPA initiated these studies to show that this methodology
can predict, with reasonable accuracy, the concentrations of
bioconcentratable pollutants in fish tissues when organisms are
exposed to these pollutants in the environment.  The reasonable
demonstration of accurate predictions in several situations will
be considered to establish the correlation between effluent
release of bioconcentratable contaminants and tissue
contamination.


1.8  Evaluation of Contaminants in Sediments

     The assessment of sediment for bioaccumulative contaminants,
described in Chapter 3, can determine the presence, identity, and
concentrations of pollutants in sediment samples subjected to
contamination from different sources.  Since sediments can
accumulate these types of pollutants over relatively long periods
of time, the bioaccumulative chemicals may be present in greater
concentrations in sediment than in a given effluent sample.  In
some cases, this may facilitate detection of contaminants which
are present in an effluent or other sources at very low
concentrations or are only released periodically.  For point and
non-point sources, the results of the sediment evaluation can
help influence the investigation of potential problem areas.
Data from sediment evaluations may also be used to determine the
spatial extent of a remediation area, monitor the benefits
derived from remediation activities, help pinpoint responsible
parties, evaluate the impacts of depositing contaminated
sediments in aquatic environments, and evaluate the success of
remediation activities.
                               1-10

-------
                            CHAPTER 2

              Principles of Bioconcentration Control
2.1  Concept of Bioconcentration and Bioaccumulation

     Fish, shellfish, and wildlife act, in a sense, like magnets
for certain types of chemicals.  Like the attraction of iron
filings to a magnet, organisms, when exposed to certain types of
chemicals, will collect and retain these chemicals in their
bodies.  The amount of chemical collected in an organism can
become very high and on a concentration basis the tissues of an
organism can achieve concentrations which are orders of magnitude
larger than those for the chemical in the environment.

     The accumulation process, i.e., the collection and retention
of the chemical in the organism, occurs with all concentrations
of the chemical in the environment.  For aquatic organisms, this
accumulation process is referred to as either bioconcentration
and/or bioaccumulation.  Chemicals which have the propensity to
accumulate in aquatic organisms are, in general, called
bioconcentratable.

     In this document, the definitions relating to
bioconcentratable chemicals, as proposed by Brungs and Mount [3]
and summarized by Murty [4], are used.  These definitions are:

     "Bioconcentration is the process by which a compound is
     absorbed from water through gills or epithelial tissues
     and is concentrated in the body; bioaccumulation is the
     process by which a compound is taken up by an aquatic
     organism, both from water and through food; and
     biomagnification denotes the process by which the
     concentration of a compound increases in different
     organisms, occupying successive trophic levels."

In this document, these terms will always be used according to
this definition.  In the literature, these terms are often used
interchangeably and may cause some confusion.

     In comparing the bioconcentration and bioaccumulation
processes, concentrations of chemicals in aquatic organisms
resulting from bioaccumulation will always be equal to or greater
than the tissue concentrations caused by the bioconcentration
process above.  For some predatory fishes, the difference in
tissue concentrations can approach two orders of magnitude.  The
structure of the food chain for the organism and n-octanol/water
partition coefficient of the residue forming chemical
significantly influence the level of bioaccumulation.   Further
information about the bioconcentration and bioaccumulation is
available in the literature [4-9].
                               II-l

-------
2.2  Concern for Bioconcentration and Bioaccumulation

     Chemical residues caused by bioconcentration and
bioaccumulation processes in fish and shellfish can cause serious
health problems for their predators, i.e., humans and wildlife.
These processes occur at exposure concentrations that are not by
themselves toxic to the aquatic organisms.  Thus, ingestion of
contaminated fish by humans and wildlife can result in toxic
doses of the residue forming chemicals even though perfectly
healthy looking fish are consumed.  This route of exposure is
direct and cannot be controlled for wildlife after a chemical is
released into the environment.  For human consumers, this
exposure can be limited by banning commercial fishing and issuing
fish advisories.  Currently, the issuance of such bans and
advisories by States is increasing significantly.


2.3  Bioconcentration Factors

     The potential for a chemical to bioconcentrate in aquatic
organisms is quantitatively expressed using the bioconcentration
factor (BCF).   The BCF is defined as the ratio of the
concentration of the chemical in the organism to the
concentration in water  surrounding the organism.

     BCFs can be calculated from experimental measures by
dividing the measured concentration of the chemical in the
exposed tissue by the measured concentration of the chemical in
the exposure water, after a steady-state condition is reached
[10].  In equation form:

          BCF  =      Concentration in Tissue	
                      Concentration in Water

     Bioconcentration factors can also be calculated by dividing
the uptake rate, k1,  by the  elimination  rate,  kg  [11].  In
equation form:

          BCF  =

     BCFs can also be estimated using structure-activity
relationships based upon the relationship between the BCF and the
n-octanol/water partition coefficient (log P)  for organic
chemicals [10,12-14].

     BCFs for organic chemicals cover a wide range of values,
depending upon the characteristics of the individual chemicals.
Some chemicals have BCFs of one million or greater.  BCFs for
most compounds have been found to be constant over a wide range
of exposure concentrations [15].  The BCFs of non-metabolized,
highly persistent, lipophilic organic chemicals are well-
correlated with their n-octanol/water partition coefficients [10,
12-14].  Compounds with low BCFs reach steady-state residue
concentrations relatively quickly [16],  whereas compounds with
high BCFs may never reach steady-state.   Compounds with low BCFs

                               II-2

-------
are more water  soluble  and have  shorter retention  times  on  a
reverse phase high performance liquid chromatography  (HPLC)
column than compounds with higher BCFs.

2.4  Bioaccumulation Factors

     The potential for  a chemical to bioaccumulate in  aquatic
organisms  is quantitatively expressed using the bioaccumulation
factor (BAF).   The BAF  can be calculated from experimental
measures by dividing the total uptake rate from water  and food,.
k.,,  by the elimination rate of the chemical,  kg  [11].   In
equation form:

           BAF   =  k1 /kg

     The BAF is dependent upon the structure of the food chain
for the organism of concern and  the log P value of the chemical.
For ecosystems  with different food chains, the same organism may
have substantially different BAFs due to differences in  feeding
habits of  the organism, the feeding habits of their prey, the
feeding habits  of prey  that their prey eats, etc.  [17-19].

     For chemicals with log P values below 5.0, BAFs and BCFs are
equal regardless of the ecosystem structure.  For  these
chemicals, the  bioconcentration  process is more important than
the bioaccumulation process from food.  For chemicals  with  log P
values ranging  from 5.0 to 7.0,  bioaccumulation from food becomes
more important  with increasing log P value and complexity of the
food chain [17,18].  For chemicals with log P values greater than
about 7.0, there is some uncertainty regarding the degree of
bioaccumulation, but generally,  food chain structure appears to
become less important due to slow uptake rates, low
bioavailability, and "dilution"  by growth for these types of
chemicals.

     In this document,  rather than attempting to define  BAFs,
bioaccumulation is accounted for by "adjusting" the BCF  using a
food chain multiplier (FM) for the organism of concern.  The
bioaccumulation and bioconcentration factors for a chemical are
related as follows [17,18]:

                BAF  =   FM * BCF

By incorporating the FM and BCF terms into the equations for
development of  reference concentrations, bioaccumulation is
included.  FMs  are provided in tabular form as a function of log
P and food chain position (trophic level)  of the organism.


2.5  Log P-Log  BCF Relationship

     For organic chemicals, bioconcentration is a  partitioning
process between the lipids of the organisms and the surrounding
water.  This mechanism,  proposed by Hamelink et al. [20], has
gained general  acceptance because the BCF and the  n-octanol/water

                              II-3

-------
partition coefficient  (P) are strongly correlated [10,12-14,21-
23].  The general form of this correlation is:

     Equation 1.1)  log BCF = A log P + B

where, A and B are constants derived using measured experimental
data.

     However, for chemicals with log P values higher than
approximately 6.0, the measured BCFs are often lower than those
predicted.  Gobas et al. [24] have attributed this over-
estimation of the BCF to violations of the conditions required
for a BCF determination.  These violations are caused by slow
uptake rate, low bioavailability, and "dilution" by growth for
the chemical of interest.

     Numerous log BCF-log P correlations have been developed and
reported in the literature for small groups of chemicals for many
species of aquatic organisms [25].  In this guidance, a
correlation based on 122 BCF values for 13 species of freshwater
and saltwater species is used [22].  Zaroogian et al. [26] have
shown that the correlation is the same for both freshwater and
saltwater species.  This correlation predicts BCFs for tissues
with 7.6% lipid content.  The equation expressing the
relationship is:

     Equation 1.2)  log BCF =0.79 log P - 0.40      (r2 = 0.86)

     Since the BCF is in part dependent on the lipid content, a
correction for lipid content is needed for different species or
for different edible portions.   Equation 1.3 incorporates this
correction for organisms and tissues with a 3.0% lipid content:

     Equation 1.3)  log BCF = 0.79 log P - 0.40 - log (7.6/3.0)

     In this guidance document,  BCF values will be presented and
discussed on a 3.0% lipid content, typical of fillets,  unless
otherwise noted.  Equation 1.3  can be used for prediction of BCF
values for other lipid contents by replacing the 3.0% with the
desired value lipid content (in percent).

     The equation derived by Veith et al [22] has 95% confidence
limits for the prediction of an individual BCF of approximately
one order of magnitude and has 95% confidence limits for the
predicted mean BCF value of approximately 5%.  Thus, for a
chemical with an estimated BCF of 100, the 95% confidence limits
for this value would range from approximately 10 to 1000.  For
BCFs of extremely hydrophobic chemicals, i.e., chemicals with log
Ps greater than 6.5, over estimation of the BCF value by log P
regression equations will be greater as the log P increases above
6.5 [24].
                               II-4

-------
2.6  Measured versus Calculated Bioconcentration Factors

     EPA recommends that BCF values calculated from the log P -
log BCF relationship be used in the calculation of the reference
tissue and ambient concentrations.  Use of calculated BCF values
will be necessary in most cases because carefully measured values
will not be available and the cost to measure these properly will
be high.  However, since the methods for calculating BCF values
do not include metabolism (which will reduce the BCF), these
values will be conservative and measured values may be necessary
to get more precise values for chemicals that metabolize.

     When measured BCF values are used, the utmost caution is
necessary when selecting an appropriate BCF value.  For most
chemicals great variation in measured BCF values exists in the
literature.  This variability arises from inappropriate
experimental conditions and/or poor analytical measurements.
Questionable BCF values exist when either of these conditions
exist during the BCF determination.  Many of the literature BCF
values will be inappropriate for use in the guidance procedures
due to the above problems.  Unfortunately, detection of incorrect
BCF values is made difficult because experimental conditions are
often incomplete.  Methods used should follow ASTM's  "Standard
Practice for Conducting Bioconcentration Tests with Fishes and
Saltwater Bivalve Mollusks,  1022-84" [27].  Experimental
measurements should include:  control residues, measured exposure
concentrations, analytical recoveries for both tissue and
exposure water quantification methods, wet weight tissue
concentrations, lipid content of the tissues, use of flow-through
exposures, and demonstrated attainment of steady-state
conditions.  The ASTM method recommends that the exposure
duration continue for 28 days or until apparent steady-state is
reached.  Because steady-state can depend on the species, life
stage, physiological condition, test conditions, etc., it is
difficult to set exposure time to a uniform length.  The ASTM
method also recommends that all organisms be of uniform size and
age.  Use of a juvenile or older life stage organisms is
recommended.
2.7  Analytical Chemistry and Bioconcentration Control

     The analytical methods provided in this document have a
fundamental difference from other EPA methods.  The methods
described in this document look for a certain type of chemical in
the sample and when a component with the proper characteristics
is detected by the GC/MS, it is identified and quantified.  In
essence, these methods survey/screen/inspect the sample and
provide a listing of the "bioconcentratable" chemicals in the
sample.  In contrast, other EPA methods are chemical specific and
these methods are designed to quantify a specific predetermined
chemical.  Chemical specific or target chemical analyses will
only provide information about the individual chemicals of
interest.
                               II-5

-------
     This fundamental difference requires that the data generated
by the assessment methods be viewed in a different light than the
data generated by target chemical analysis.  With target chemical
analyses, the identity of the chemical is known and concentration
of the chemical is measured accurately.  With the assessment
methods described herein, the reported identity and concentration
of a chemical are less certain.  This occurs because model
compounds are used to quantify the identified chemical and
because mass spectral algorithms for identifying unknown
chemicals are, currently, imprecise.

     The use of model compounds for quantifying the identified
chemicals is required since we do not know a priori what
chemicals are in the sample.  Quantifications based upon the
model compounds assume that analytical recoveries and mass
spectral responses are the same for the model and identified
chemicals.  These assumptions can be expected to cause error in
the quantification of no worse than one order of magnitude.  The
largest part of the overall error in quantification is caused by
the wide differences in mass spectral responses among the
individual compounds [23].

     These uncertainties in quantification and identification of
the GC/MS components are eliminated in later steps in the
guidance approach.  With the tissue and effluent assessment
options, confirmation analyses are required before development of
a RAC, wasteload allocation and when necessary, permit limits for
a chemical.  Confirmation analyses provide conclusive
identification and substantially more accurate quantification for
the GC/Ms component of interest.  In addition, with the tissue
option target chemical analyses on the effluent will be required
for the chemical of interest prior to developing wasteload
allocations and permit limits.  In general, target chemical
analysis techniques have much smaller quantification errors than
the analytical procedures included in this guidance.  For
example, EPA method 1625 has initial method quantification
accuracy requirements for bioconcentratable chemicals which are
typically no worse than a factor of 2.

     Mass spectral library searching algorithms are used to
assign tentative identifications to components detected in the
GC/MS analysis of the prepared sample extracts.  Two libraries of
mass spectral data are used in the assessment methods, the
Chemicals of Highest Concern (CHC) and the EPA/NIH/NBS mass
spectral libraries.  These algorithms compare the mass spectra of
the GC/MS component to those in the libraries and the ten best
fitting/matching tentative identifications with fits/matches of
70% and greater are reported.  These identifications are
considered tentative because a mass spectra by itself is not
enough information to conclusively identify a GC/MS
component/peak.  Multiple tentative identifications are provided
for each component because the correct identification is often
not the best matching tentative identification.  This imprecision
in the searching algorithms has important implications for
evaluation of the reported data.

                               II-6

-------
     Computer algorithms for identifying unknown mass spectra via
library searching are often categorized as either forward or
reverse searching.  In general, reverse searching algorithms have
demonstrated advantages for identifying  unknown mass spectra
when the unknown is not chemically pure [38].  With GC/MS
analyses, mass spectral data can never assumed to be pure and
thus, the use of reverse searching algorithms is recommended when
available.  Unfortunately, some GC/MS systems do not have reverse
searching algorithms.  In these cases, library searching should
be performed using the default algorithm provided by the
manufacturer of the GC/MS system.

     To evaluate the data generated by the assessment methods,
all tentative identifications must be evaluated for each
component.  This requirement is absolutely necessary since the
best matching (fitting) tentative identification is often not the
correct identification for the component.   Analyses to confirm
the true identity of the chemical are performed after evaluation
of the analytical data.  A chemical would be considered confirmed
when the retention time on the GC/MS column and mass spectra of
the component are identical between the sample and a standard
that is made from the pure chemical.

     The analytical methods provided in this document have been
designed to achieve low levels of detection.  Minimum levels of
detection are assured in these methods by the use of surrogate
compounds.  These chemicals are placed into the sample at low
concentrations at the start of the analysis, 100 ng/1 and 5 ng/g
for the effluent and tissue procedures, and detection of these
chemicals in the GC/MS analysis of the prepared extracts ensured
that these levels of detection are achieved.  Detection limits
for the methods are estimated to be approximately 10 ng/1 and
1 ng/g, respectively.  These levels of detection will require
substantially better analytical technique than currently used by
many contract laboratories which perform standard EPA methods.
These methods can be performed successfully, on a routine basis,
with the use of good low level residue techniques.


2.8  Chemicals of Highest Concern

     The analytical methods for the residue and effluent options
determine the presence of bioconcentratable chemicals in tissues
and effluents.  To identify compounds, GC/MS analyses are
performed on sample extracts and all peaks/components in the data
are compared to two libraries of mass spectral data.  These
libraries are the Chemicals of Highest Concern (CHC) and
EPA/NIH/NBS mass spectral libraries.

     The CHC library consists of approximately 30 chemicals which
pose serious risks to human health due to high toxicities and
high potential to bioconcentrate.  These characteristics cause
residues in fish and shellfish which are of concern even when
these chemicals are present at very low concentrations in the
receiving water.  With either assessment option,  detection of

                               II-7

-------
these chemicals will be difficult.  To increase the chance of
detecting these chemicals, all components in the GC/MS data are
compared to the CHC library to determine if any of these
chemicals are present.  If any of the chemicals in the CHC
library are found, efforts to evaluate and control these
chemicals should be of the highest priority.  The CHC library
(Table 2.1) was compiled by selecting chemicals which produce
residues of concern at very low ambient concentrations.
                              II-8

-------
                            TABLE  2-1

                CHEMICALS OF HIGHEST CONCERN LIST
CAS number
                    chemical name
   50-
   57-
   58-
   60-
   70-
   72-
   72-
   76-
   91-
   95-
  101-
  115-
  117-
  118-
  309-
  319-
  319-
  608-
  608-
  924-
 1024-
 1746-
 2104-
 2385-
 8001-
39515-
11096-
11097-
11104-
11141-
12672-
12674-
53469-
•29-3
•74-9
•89-9
•57-1
•30-4
•54-8
•55-9
•44-8
•94-1
•94-3
•61-1
•32-2
•81-7
•74-1
•00-2
•84-6
•85-7
•73-1
•93-5
•16-3
•57-3
•01-6
•64-5
85-5
35-2
41-8
82-5
69-1
28-2
16-5
29-6
11-2
21-9
p,p'-dichlorodiphenyltrichloroethane (DDT)
chlordane
hexachlorocyclohexane  (lindane)
dieldrin
hexachlorophene
p,p'-dichlorodiphenyldichloroethane  (ODD)
p,p'-dichlorodiphenyldichloroethylene  (DDE)
heptachlor
3,3'-dichlorobenzidine
1,2,4,5-tetrachlorobenzene
4,4'-methylene bis(N,N'-dimethyl)  aniline
dicofol
bis(2-ethylhexyl)phthalate  (BEHP)
hexachlorobenzene
aldrin
alpha-hexachlorocyclohexane  (alpha-HCH)
beta-hexachlorocyclohexane  (beta-HCH)
technical-hexachlorocyclohexane  (t-HCH)
pentachlorobenzene
N-nitroso-di-n-butylamine
heptachlor epoxide
dioxin (2,3,7,8-TCDD)
ethylp-nitrophenylphenylphosphorothioate(EPN)
mirex
toxaphene
danitol
polychlorinated biphenyl 1260
polychlorinated biphenyl 1254
polychlorinated biphenyl 1221
polychlorinated biphenyl 1232
polychlorinated biphenyl 1248
polychlorinated biphenyl 1016
polychlorinated biphenyl 1242
                               II-9

-------
                            CHAPTER 3

      Assessment Options for Bioconcentratable contaminants
     In Figures 3.1 and 3.2, detailed flowcharts for the
assessment process are presented for the tissue and effluent
options.  These flowcharts diagram the data generation, decision
points, and evaluation logic for each option.  In this chapter,
the discussion presented will start with sample collection and
proceed as if one was performing the assessment process.  The
outputs from this process for either option are lists of
identified bioconcentratable chemicals which require reference
concentration development (Chapter 4), waste load allocation
(Chapter 5), and when necessary, permit limits (Chapter 6).

     The discussion presented in this chapter will provide
detailed information on the decision points, analytical
procedures, and underlying rational used by the assessment
options.  Some of the section titles in this chapter are the same
as the blocks in the flowcharts.  This one to one correspondence
is intentional and is intended to aid readers of this document to
quickly find discussion relevant to the flowcharts.  Also
included in this chapter is an abbreviated discussion on the
analytical procedure for assessing the presence of
bioconcentratable chemicals in sediments.   The detailed
analytical procedures for the tissue, water, and sediment
matrices are provided in Appendices A, B,  and C.


3.1  Tissue Residue Option

3.1.1  Sampling Considerations

     For successful implementation of the tissue option, samples
which are representative of the organisms in the receiving water
are necessary.  Analyses performed on representative samples
allows the determination of the mean concentrations for residue
forming chemicals in the receiving water population.  Due to
natural variation, haphazard sample collection can result in
effluent permitting situations which are not protective for
bioconcentratable chemicals because the assessment was performed
on unrepresentative tissue samples.

     Sampling designs which result in representative tissue
samples must consider:  1)  when to sample (time of year and
frequency), 2) how to sample (number of species,  age,  size,  type
of species, number of organisms per sample), and 3)  where to
sample (number of sampling stations and to consider exposure to
                              Ill-l

-------
                                                  Figure 3.1
                                          TISSUE ANALYSIS
        Collection & extraction
           of tissue sampl
          Analyze vit GC/MS
      I No further tcQoo
  Reverse-search all chromatographic peaks
   against the CHC miss spectral library
          identified fit >70%?
          Concentration
          >5ugykg?
                                                        Reverse-search EPA/NIH/NBS
                                                           mass spectral library
                                                            identified ft >70%?
   (concentration x dilution x 4) > RAC7
RSD. q* 1, ind/or Water Quality Standard*?
        Issue confirmation requat
                                                          Issue confinnauoD requetf
No further action  I
                                                         Region or state develop or request
                                                       development of RED and/or q*l values
                                                                                                   c
                                                   REPORT: list 3
J
                                             III-2

-------
sources and effects of dilutions and mixing zones).  Complete
descriptions of individual programs for tissue residue monitoring
programs are available from many States and other groups [44].

     When to Sample

     Chemicals that form residues in aquatic organisms have, in
general, a longer half-life in the organisms than in effluents
and receiving waters.  This difference occurs in part due to the
relative time scales associated with bioconcentration and
bioaccumulation processes, i.e., weeks to years, in comparison to
hydrologic processes associated with effluents and receiving
waters, i.e., hours to days.  Consequently, residue measurements
on indigenous organisms from a receiving water provides
information on chemical exposure for a longer time frame, i.e.,
weeks to months, than does chemical analyses on effluent samples
which are typically 24 hour composites.

     Since the objective of this guidance is to protect humans
from unacceptable residue consumption, a prime consideration for
time of sampling is the period when most organisms are harvested
for consumption.  In the absence of other evidence, sampling
should be done after periods of normal or sub-normal dilution of
effluents and if possible, include at least two seasons of the
year.  Seasonal variation in residues concentrations may occur
from changes in food type, feeding level, changes in metabolism,
and reproductive stages for the species.  There should be no need
to sample organism residues more than monthly; one sample each
quarter is probably more inclusive than four samples taken in
consecutive months.

     How to Sample

     Two species of organisms which are consumed by humans should
be collected from a receiving water and these species should have
different feeding habits, occupy different habitat niches and
represent different physiology.  One species at least should be a
predator since bioaccumulation via the food chain may be
important for the receiving water.  When possible, a vertebrate
and invertebrate species is advised since major differences in
metabolic activity exist between these groups [30].

     In general, small but fully adult organisms are most likely
to contain the highest residue for a given exposure (if a
correction is made for lipid content) and thus,  are recommended.
It is not necessary to sample organisms actually consumed by
humans; however, it is important to include information for
species consumed by humans in the subsequent assessments of the
residue data.  In some situations, the problems of collecting or
availability of organisms will dictate the species to be
collected.  When possible, 10 organisms should be collected per
sample.  By using 10 organisms per sample, a better estimate of
the average residue concentrations in the receiving water
organisms can be obtained than with samples based upon 1 or 2
organisms.

                              III-3

-------
     Where to Sample

     The location of sampling is a site-specific decision and
there are trade-offs to consider with regards to proximity to the
sources and complexity of the receiving water.  In general, one
would expect the highest residues closest to the outfall.
However, if bioaccumulation via the food chain is important,
depositional zones downstream of discharge might produce higher
residues in the indigenous organisms.  Collection of organisms in
the mixing zone is not recommended since these organisms might
have unusually low residue concentrations.  These organisms may
avoid the mixing zone or reside in it for very short time
periods.  Additionally, if food chain accumulation is an
important source for the residue forming chemical, sufficient
time for bioaccumulation may not occur there.  On the other hand,
if locations too distant from the discharge are sampled,
dilution, degradation, or sorption may cause residues to be
lower.

     The ideal sampling design would consist of sites where the
organisms are confined by the geography of the receiving water,
where the exposure of the organisms to the discharge is well
established, and where the site is close to the discharge.
Geographical features which could confine organisms include dams
and waterfalls on streams and rivers and salinity gradients in
marine estuaries.  The sampling of sessile organisms can also be
used to define exposure as well as limit mobility of organisms.

     With well defined sampling sites, one well placed site is
probably enough for a successful assessment with the tissue
option.  As the complexity of the receiving water increases,
i.e., more dischargers and mixing in the receiving water becomes
less defined, additional sites might be necessary.  These
additional sites should, possibly, be placed downstream of the
first site.  If additional sites are required, increased costs
for performing this option will occur.  A possible alternative to
evaluating these additional sites at the same time would be to
perform assessments on a rotating basis, i.e., each quarter a
different site is assessed.
3.1.2  Compositing of Organism Samples for Analytical Procedure

     Concentrations of the bioconcentratable chemicals in the
receiving water organisms will not be exactly the same among
individual organisms due to natural variation.  To obtain the
best estimate of the average residue concentration for an
individual species, it is recommended that a tissue sample
consist of 10 organisms and that equal portions of each organism
be used in the tissue composite.  If more than one species of
organisms is collected for a sampling site, each species should
be composited separately.
                              III-4

-------
     If fish fillets are to be composited, equal portions by mass
from each fillet would be combined.  For whole fish, two options
are available.  One of these approaches would consist of grinding
each fish separately and then, subsampling equal portions by mass
of the ground tissues to obtain the composite sample.  The other
approach would consist of selecting fish of equal mass and
compositing them.  For this approach, the coefficient of
variation for the masses of the whole fish should not exceed 10%.
With whole fish, removal of the guts and possibly the head before
grinding and/or subsampling may be desirable since human
consumers, in general, rarely ingest these parts of the
organisms.

     For composites from the fillets and whole fish, thorough
grinding of the tissue to yield homogenous tissue samples is
essential.  In general, homogeneous grinding of tissue samples is
obtained by passing the tissue composite through a grinder three
times.  Ground tissues are stored at -10°C in solvent rinsed
glass jars sealed with aluminum foil or Teflon lined lids until
analysis.

     If less than 10 individuals per sample were collected,
compositing would be performed as described except with fewer
individuals per composite sample.


3.1.3     Extraction of Composite Tissue Samples

     The tissue sample is extracted using standard residue
chemistry techniques.  This process consists of mixing 20 grams
of the tissue after thawing with enough anhydrous sodium sulfate,
100 to 140 grams, to dry the sample.  Half of this mixture is
placed into a Soxhlet extraction thimble and three surrogate
chemicals are spiked onto the tissue in an extraction thimble.
The remaining sodium sulfate/tissue mixture is then placed into
the extraction thimble and the extraction thimble is placed into
a Soxhlet extractor body.  The tissue is extracted for a minimum
of 18 hours using a 50:50 mixture of methylene chloride and
hexane.  The extract is concentrated using a Kuderna-Danish
concentrator and its volume is adjusted to 10.0 mis.  A portion
of the extract, i.e., 0.50 or 1.0 ml, is removed from the extract
and is placed into a tared weighing pan.  After evaporation of
the solvent in the weighing pan, the pan is placed into a 105°C
oven to dry the sample.  The dried sample is reweighed and
percent lipid content of the tissue is determined.
                                           j O
     The surrogate chemicals, d10-biphenyl,  C6-l,2,4,5-
tetrachlorobenzene,  and 13C6-hexachlorobenzene,  are added to the
tissue at a 5.0 /zg/kg concentration prior to extraction of the
sample.  These chemicals  are referred to as surrogates because
their behavior mimics that of bioconcentratable chemicals in the
analytical procedure.  The surrogate chemicals are used for
quality control and quantification in this analytical procedure.
This procedure is presented in detail for laboratory use in
Appendix A.

                              III-5

-------
3.1.4     Gel Permeation Chromatoqraphv/Silica Gel Cleanup of
          Tissue Extract

     The remaining volume of the extract, i.e., 9.0 or 9.5 mis,
is adjusted to a suitable volume for gel permeation
chromatography (GPC),  e.g. 5.0 ml.  The extract is injected onto
the GPC column and lipids are removed from the extract.  The GPC
process might have to be performed multiple times using aliquots
of the extract because some GPC columns can handle limited
amounts of lipid before coming overloaded.  GPC is typically
performed using automated analytical instrumentation.  This
instrumentation injects the extract onto the GPC column, performs
the chromatography process, and collects the proper eluate from
the GPC column which contains the chemicals of interest.

     After the lipids are removed from the extract, the extract
is concentrated to 0.5 ml.  Silica gel chromatography is then
performed on the extract and this process removes cholesterol
like chemicals from the extract.  Silica gel chromatography is
performed on a 2.1 gram column of 1% deactivated silica gel.  The
column, after transfer of the extract to the top of the column,
is eluted using a 15:85 methylene chloriderhexane mixture.  The
solvent eluting from the column is collected and concentrated to
0.10 ml.
     The discussion presented to this point, Sections 3.1.1
through 3.1.4, refers to the "Collection & Extraction of Tissue
Sample" box in the tissue residue option flowchart, Figure 3.1.
The sampling and analytical procedures, in most cases, will be
performed by the discharger or contractor.

     The above procedure is presented in detail for laboratory
use in Appendix A.


3.1.5     Analyze via GC/MS

     Gas chromatography/mass spectrometry (GC/MS)  analyses are
performed on the prepared tissue extracts using standard residue
techniques.  Prior to GC/MS analysis, an internal standard, d12-
chrysene, is added to the sample extract to calibrate the
response of the mass spectrometer.  Sample analyses are performed
on a 30 m capillary column with a temperature program, e.g. 50-
175°C at 10°C/min and then 175-275°C at 5°C/min, 275°C for 20
min.  Mass spectral data are collected using full scan electron
impact ionization mass spectrometry.  After GC/MS analysis, the
extract is saved for confirmation analysis.

     After analysis of the sample extracts and standards,
standard curves are calculated for the three surrogates; d10-
biphenyl, 13C6-1,2,4,5-tetrachlorobenzene, and  C6-hexachloro-
benzene; using an internal standard method.   These curves are
used to quantify the surrogates in the sample extracts.  These
quantifications are used to determine the percent recovery for

                              III-6

-------
each of the surrogate chemicals.  All other peak/components in
the GC/MS data are quantified using the standard curve for the
13C6-hexachlorobenzene.
The above procedure is presented in detail for laboratory use in
Appendix A.


3.1.6     Library Searching using the Chemicals of Highest
          Concern Mass Spectral Library

     All chromatographic peaks in the GC/MS data are compared
with the Chemicals of Highest Concern (CHC) mass spectral library
(see Table 3.1).  Peaks with fits/matches of 70% and greater are
considered tentatively identified.  For each tentatively
identified component, a list of the best mass spectral library
identifications (up to a total of ten identifications) is
reported along with the percent fit values, CAS number of the
tentative identification, GC retention time, and the
concentration for the GC/MS component.  This report is identified
as Report 1 in the tissue residue option flowchart, Figure 3.1.

     The CHC mass spectral library can,  in nearly all cases, be
derived from the EPA/NIH/NBS mass spectral library using the
software on the data system of the GC/MS system.  The recommended
70% library searching fit criterion was chosen based upon best
scientific judgement and is subject to modification by the
regulatory authority.

     Computer algorithms for identifying unknown mass spectra via
library searching are often categorized as either forward or
reverse searching.  In general, reverse searching algorithms have
demonstrated advantages for identifying unknown mass spectra when
the unknown is not chemically pure [31].  with GC/MS analyses,
mass spectral data can never be assumed to be pure and thus, in
the procedure, the use of reverse searching algorithms is
required when available.  Unfortunately, some GC/MS systems do
not have reverse searching algorithms.  In these cases, library
searching should be performed using the default algorithm
provided by the manufacturer of the GC/MS system.


3.1.7     5 jug/kg Tissue Concentration Decision Point

     If a GC/MS component is not identified with the CHC mass
spectral library search (the no arrow in the flowchart), the
chemicals present at concentrations below 5 M9/kg are dropped
from further consideration.  The concentration of 5 jig/kg is used
as the cut-off for further investigation because 5 /ig/kg is the
concentration of the surrogate compounds added to the sample
prior to extraction.  The surrogate concentrations represent the
minimum level at which adequate quantitation can occur.  The
amount of uncertainty associated with a value which lies below
the surrogate concentrations increases greatly with decreasing
concentration.  The purpose of this cutoff level is to prioritize
unknown/non-CHC peaks for further investigation.

                              III-7

-------
     GC/MS components with concentrations greater than 5
which are not identified with CHC mass spectral library are then
library searched with the EPA/NIH/NBS mass spectral library.


3.1.8     Library Searching using the EPA/NIH/NBS Mass Spectral
          Library

     GC/MS components not identified using the CHC library search
and with concentrations greater than 5 p-gfkg are library searched
against the EPA/NIH/NBS mass spectral database.  Peaks with
fits/matches of 70% and greater are considered tentatively
identified.  For each tentatively identified component, a list of
the best mass spectral library identifications (up to a total of
ten identifications) is reported along with the percent fit
values, CAS number of the tentative identification, GC retention
time, and the concentration for the GC/MS component.  This report
is identified as Report 2 in Figure 3.1.

     For those components with fits/matches less than 70% but
greater than 25%, the two best mass spectral library
identifications along with the percent fit values, the CAS number
of the tentative identifications, GC retention time, and the
concentration are reported for each the GC/MS component.  For
GC/MS components with fits/matches less than 25%, the
concentrations and GC retention times for these components should
be reported and the components labeled as being "unknown".  This
report is identified as Report 3 in Figure 3.1.  The tentative
identifications listed in Report 3 are of less certain
reliability and these results are provided for informational
purposes only.

     The EPA/NIH/NBS mass spectral library is available from the
U.S. Government Printing Office (941 North Capitol St. N.E.,
Washington, D.C. 20401).  In addition, most GC/MS manufacturers
have this library available in a form suitable for their
respective instruments.

     The recommended 70% library searching fit criterion was
chosen based upon best scientific judgement and is subject to
modification by the regulatory authority.  As with the CHC
library searching process, the use of reverse searching
algorithms is strongly suggested when available.  In cases where
reverse searching algorithms are not available, library searching
should be performed using the default algorithm provided by the
manufacturer of the GC/MS system.


3.1.9     Analytical Summary

     With the generation of the three reports, Reports 1, 2, and
3, the procedures presented in Appendix A have been performed for
the tissue assessment option.  The discussion presented above in
Section 3.1.1 through 3.1.8 outline this procedure and the

                              III-8

-------
decision points for processing the GC/MS data collected on the
tissue samples.  Example data and reports are provided in
Appendix J.

     Reports l, 2, and 3 as well as the QA/QC report for a tissue
analysis will be sent to the regulatory authority.  The
regulatory authority will evaluate the data and decide which
tentative identified chemicals will need confirmation by
performing the tasks in the two subsequent decision rectangles
labeled, "RfD, ql*, and/or Water Quality Standard?" and
"(concentration x 4) > RTC".


3.1.10    RfD, ql*f and/or Water Quality Standard?

     This decision rectangle in the tissue assessment option,
Figure 3.1, simply asks the question, "Is there a Water Quality
Standard or sufficient information (RfD, Ql*) to develop a RAC
available for a tentative identification listed in Report 2?".
The purpose of this decision is to identify the chemicals for
which there is information to later develop NPDES limits.  For
each GC/MS component, up to 10 tentative identifications might be
reported.  For each of the tentative identifications, the above
question must be answered.  If any of the tentative
identifications for a GC/MS component have a RfD, ql*, and/or
Water Quality Standard available, the evaluation process for that
component would proceed to the "(concentration x 4) > RTC"
decision rectangle, the yes arrow in Figure 3.1.  Further
information concerning the Rfd, ql*,  and water quality standards
are presented in Chapter 4 of this guidance.

     If none the tentative identifications have the required
information for a single GC/MS component, the regulatory
authority would issue a request for confirmation of that GC/MS
component to the discharger (the no arrow leaving this decision
rectangle in Figure 3.1).

     This evaluation process is performed on all tentatively
identified GC/MS components listed in Report 2.   Furthermore, for
each tentatively identified GC/MS component, each tentative
identification must be evaluated.
3.1.11    (Tissue Concentration x 4)  > RTC?

     This decision rectangle in the tissue residue alternative,
Figure 3.1,  evaluates whether the measured concentration of a
tentatively identified GC/MS component in the tissue sample is
likely to be larger than the reference tissue concentration
(RTC).  This screening step compares, for tentatively identified
chemicals, the residue concentration times 4 to the RTC.  The
factor of 4  was developed by using the statistical analysis for

                              III-9

-------
reasonable potential as presented in the Technical Support
Document [32].  The table below was derived by using the TSD
approach with the only change here being the use of the
population mean rather than the population extreme value as the
target concentration.  The table shows the maximum expected
difference between one sample and the mean effluent concentration
based on the expected effluent variability as measured by the
coefficient of variation (CV):

                              Percent Confidence
                CV            95%            99%
               0.2            1.4            1.6
               0.4            1.9            2.5
               0.6            2.5            3.6
               0.8            3.2            5.1
               1.0            3.9            6.9

     EPA's treatability database suggests that variability in
many effluents can be characterized by a CV of 0.6 or less.  At
this CV and with 99% confidence, the maximum ratio of the
effluent population mean to one sample is 3.6.  At 95%
confidence, the ratio is less (2.5).  Therefore, the multiplier
of 4 provides a reasonable way to estimate the mean concentration
based on one sample for this screening comparison to the RTC.

     For each GC/MS component,  this comparison is performed for
each of the tentative identifications which have Rfd, ql*, and/or
water quality standards.  If this comparison is true for any of
the tentative identifications for a GC/MS component, a
confirmation request is issued by the regulatory authority to the
discharger for that component (the yes arrow leaving this
decision point).   If this comparison is false for all of the
tentative identifications for a GC/MS component, no further
action (the no arrow leaving this decision point) is required for
this component.

     All tentatively identified GC/MS components in Report 1 are
evaluated at this decision point.  For tentatively identified
GC/MS components in Report 2, only those tentative
identifications with RfD, ql*,  and/or water quality standards
information are evaluated.

     The calculation of the reference tissue concentrations (RTC)
is described in detail in Chapter 4 of this document.  RTCs can
be calculated using human dose information, Rfd and/or ql*, and
fish consumption rates by humans.

3.1.12    Issue Confirmation Request

     In the tissue option,  confirmation requests by the
regulatory authority can be issued from two different decision
points, see Figure 3.1.  In either case, the discharger is
required to confirm the identity of the GC/MS component
tentatively identified in the tissue extract.  The confirmation
process is discussed in further detail in Section 3.3.  The

                             111-10

-------
results of the confirmation process are reported back to the
regulatory authority.  These results include the confirmed
identity of the GC/MS component and its concentration.  Since
confirmation may, for certain complex samples, involve
considerable analytical chemistry work and the associated costs
may be high, the requirement for confirmation was placed in the
option logic sequence so that only the chemicals most significant
concern would require confirmation.


3.1.13    Outputs From Tissue Residue Assessment

     The outputs from the tissue assessment process are two lists
of identified bioconcentratable chemicals.  The first list
consists of chemicals from Reports 1 and 2 which require
reference concentration development (Chapter 4),  waste load
allocation (Chapter 5), and when necessary, permit limits
(Chapter 6).  The second list consists of chemicals which arises
from Report 2 and do not have RfD's, ql*'s, and/or water quality
standards.  In addition to this, the actual GC/MS Chromatograms
should be included in the reported information.  Development of
Rfd, ql* and/or water quality standards are required or requested
to determine if current residue levels pose serious risk to human
consumers of fish and shellfish.  Monitoring of these chemicals
in effluents is suggested in order to establish discharge
information.

     It must be noted that chemicals in both of these lists form
residues in aquatic organisms.  These chemicals would have been
found in indigenous organisms from the receiving water under
consideration.


3.1.14    Caveats for the Tissue Option

     The tissue assessment option will not detect all
bioconcentratable chemicals which form residues in aquatic
organisms.  The analytical procedures outlined above will only
detect nonpolar organic chemicals which can be successfully
analyzed using GC/MS.  These procedures are fairly robust;
however, they do have limitations.  Some of these limitations
include sensitivity of the analytical method, our ability to
identify unknown GC/MS components, and the lack of RfD and ql*
information.
3.2  Effluent Option

3.2.1     Sampling Considerations

     For successful implementation of the effluent option,
samples which are representative of the discharger's effluent are
necessary.  Analyses performed on unrepresentative samples can
lead to permitting situations which are not protective for
bioconcentratable chemicals.

                              111-11

-------
     Sampling should be performed when the facility operations
are typical.  Sampling during unusual facility operations, i.e,
storm events, plant shutdowns, production and treatment facility
changes, should be avoided.  24 hour composite samples are
recommended and grab effluent samples should be avoided.  Volume
of effluent collected per sample should be at a minimum of 10
liters; 12 liters is preferable.  Samples should be stored in the
dark at 4°C and be extracted within 7 days after collection.


3.2.2     Extraction of Effluent Sample

     The extraction process consists of spiking a 10 liter
effluent sample with three surrogate chemicals.  After thorough
mixing, the effluent sample is extracted using liquid-liquid
extraction with hexane three times.  The extract is dried using
anhydrous sodium sulfate and concentrated using a Kuderna-Danish
concentrator to approximately 10 mis.

     The apparatus for performing the extraction of 10 liters of
effluent will be laboratory specific.  Depending upon the
available equipment, different bottles, vessels, shakers,
tumblers, etc. will be used by different laboratories for the
extraction process.

     The surrogate chemicals, d10-biphenyl, 13C6-1,2,4,5-
tetrachlorobenzene, and 13C6-hexachlorobenzene,  are added to the
effluent at a 100 ng/1 concentration prior to extraction of the
sample.  These chemicals are referred to as surrogates because
their behavior mimics that of bioconcentratable chemicals in the
analytical procedure.  The surrogate chemicals are used for
quality control and quantification in this analytical procedure.


3.2.3     Sample Cleanup

     To remove biologically-derived and some easily metabolized
organic chemicals commonly found in effluents,  an acid clean-up
of the effluent extract has been included in the analytical
procedure.  These materials, e.g., fatty acids, fatty acid
esters, sterols, phthlates, and phenolic plant materials, if not
removed, cause serious interferences in common chemical residue
analysis procedures.     This procedure consists of constructing
a column containing (bottom to top) glass wool, silica gel (2 g),
sodium sulfate (2 g), 70% sulfuric acid solution (5 ml) on Celite
(10 g), and sodium sulfate (2 g).  After placing the sample
extract on top of the column, the column is eluted with hexane.
The eluate from the column is collected and then, concentrated
using a Kuderna-Danish concentrator to approximately 10 ml.  The
extract is further reduced to 0.5 ml using a gentle stream of
clean air.

     This procedure which successful removes these interfering
materials from the extract will also remove bioconcentratable

                              111-12

-------
chemicals which are unstable in acidic conditions.  Consequently/
bioconcentratable chemicals which are unstable in acidic
conditions will not be detected with the effluent assessment
procedure.

     The acid treatment of the sample causes reactions like
hydrolysis of esters, cleavage of ethers, additions to olefins
which changes the biologically-derived materials to polar organic
materials.  These materials are removed from the sample extract
by passing the extract though a silica gel column.  This column
retains the polar materials and allows the nonpolar materials to
pass through the column.  This commonly used acid clean-up
procedure [33] yields sample extracts which predominantly contain
the chemicals of interest (i.e., nonpolar organic chemicals).  In
addition, this treatment allows significantly lower detection
limits for the method since substantially fewer components are in
the sample extracts during the subsequent steps of the procedure
which utilize instrumental analysis (e.g. HPLC, GC/MS).


3.2.4     Fractionation via HPLC

     The effluent extract after acid clean-up contains numerous
chemicals with low potential to bioconcentrate.  Since these
chemicals are of lower concern, these chemicals are removed from
the effluent extract using a well documented high performance
liquid chromatography (HPLC) technique.

     Chemicals with low potential for bioconcentration are
defined in this guidance document as chemicals with log P values
below 3.5.  This bioconcentration threshold was derived by
comparing the relative exposure for humans to bioconcentratable
chemicals from the consumption of contaminated drinking water and
fish.  As the BCF increases above 100,  the intake from fish will
become proportionately greater than from water.  This finding was
derived using the average consumption amounts for drinking water
of 2 liters/day, and for fish 20 g/day.  EPA recognizes that this
fish consumption rate is higher than that assumed in deriving
§304(a) water quality criteria.  This higher rate is used as a
conservative approach to screen for toxicants which have the
potential to be a problem.  Pollutants screens for further
analysis need not necessarily require further controls to meet
water quality standards.

     Therefore,  the effluent assessment procedure in this
guidance document was developed and optimised for organic
chemicals with log P values of 3.5 and greater.  This threshold
corresponds to a BCF of approximately 100 on a 3% mean lipid
content using the recommended equation for estimating a BCF from
log P (Equation 1.3).  This threshold value was selected to
target those organic chemicals of greatest concern.
                             111-13

-------
                                              Figure 3.2
                                     EFFLUENT ANALYSIS
                               Collect & aorta effluent simple
                                  Fractionate vii HPLC
       Analyze via GC/MS
Reverse-search all chromatographi
                                                               Concentration
                                                                > lOOng/t?
        dcntified fit >70%?
                                                         Reverse-search EPA/MH/NBS
                                                             ma« spectral library
                                                             identified fit >70%?
                                                      (fraction BCF x concentnticn x dilution)
 (concentration x dilutioa x 4) > RAC7
                                                     KID. q*l, and/or Water Quality Standards?
                                                       (FM x BCP x concentration x dilution)
     luue confirmation requcxt
                                                           Issue confumation requect

                                                         Region or state develop or request
                                                       development of RfD and/or q*l value*

                                     iri-14

-------
     To remove chemicals below the bioconcentration threshold, a
relationship between log P and the retention time of chemicals on
a reverse phase HPLC column is used to fractionate the effluent
extract [34,35].  Fractionation of the effluent extract allows
chemicals with log P values less than the bioconcentration
threshold to be removed and allows the rest of the chemicals to
be subdivided into 3 separate fractions.  By subdividing the
bioconcentratable portion of the effluent extract into sub-
portions (fractions), an initial estimate for a chemical's BCF
can be provided by using the average log P value of the fraction
containing the chemical and the log P/log BCF relationship
previously discussed.

     The log P/retention time relationship is usually expressed
in the following form [35-38]:

               log P = C + D * log TR

where C and D are constants derived for the HPLC system used and
TR is the corrected retention time for a chemical.   By using the
appropriate HPLC conditions (which are derived using the
procedures specified in Appendix B), fractionation is performed
by collecting the solvent eluting from the HPLC column during
specified timed intervals after injection of the effluent extract
onto the HPLC column.

     Three fractions are collected from the HPLC column.  These
fractions have log P ranges of 3.5 to 4.5, 4.5 to 5.7, and 5.7 to
8.2.  These log P ranges result in BCF value ranges of 91 to 560,
560 to 5000, and 5000 to 470,000 (3% lipid content)  and have
average BCF values of 230, 1,700, and 49,000, respectively.

     The effluent extract (which now is divided into three
fractions)  after the HPLC fractionation, contains primarily those
chemicals with high potential to bioconcentrate, i.e., chemicals
with higher log P values.  These fractions are prepared for GC/MS
analysis by diluting the HPLC fractions with water and
extracting the bioconcentratable chemicals using liquid-liquid
extraction or solid phase extraction techniques.  The fractions
after extraction are reduced to 0.10 ml and are stored at -10°C
until GC/MS analysis.


3.2.5     Analyze via GC/MS

     Gas chromatography/mass spectrometry (GC/MS) analyses are
performed on the three HPLC fractions using standard residue
techniques.  Prior to GC/MS analysis, an internal standard, d12-
chrysene, is added to each of the HPLC fractions to calibrate the
response of the mass spectrometer.  Sample analyses are performed
on a 30 m capillary column with a temperature program, e.g. 50-
175°C at 10°C/min and then 175-275°C at 5°C/min, 275°C for 20
min.  Mass spectral data are collected using full scan electron
impact ionization mass spectrometry.  After GC/MS analysis, all
of the extracts are saved for confirmation analysis.

                              111-15

-------
     After analysis of all three fractions and standards,
standard curves are calculated for the three surrogates; d10~
biphenyl, 13C6-l,2,4,5-tetrachlorobenzene, and 13C6-hexachloro-
benzene; using an internal standard method.  These curves are
used to quantify the surrogates in the sample extracts.  These
quantifications are used to determine the percent recovery for
each of the surrogate chemicals.

      The three surrogate chemicals were chosen so that each of
the HPLC fractions will contain only one of the surrogates.
These chemicals, d10-biphenyl, 13C6-l,2,4,5-tetrachlorobenzene, and
13C6-hexachlorobenzene,  will be in the first,  second,  and third
fractions, respectively.  If these chemicals are present in
different fractions, the HPLC fractionation procedure was
performed incorrectly or with improper HPLC conditions.  Quality
control procedures would require that corrective actions be taken
and that the new effluent sample be extracted.

     Quantification of the remaining GC/MS components are
performed by using the responses of the surrogates in their
respective fractions.  For the first fraction, all of the GC/MS
components are quantified using the standard curve for d10-
biphenyl.  For the second fraction, all of the GC/MS components
are quantified using the standard curve for  C--1,2,4,5-
tetrachlorobenzene.  For the third fraction,  all of the GC/MS
components are quantified using the standard curve for 13C6-
hexachlorobenzene.  This procedure is presented in detail for
laboratory use in Appendix B.


3.2.6     Library Searching using the Chemicals of Highest
          Concern Mass Spectral Library

     All chromatographic peaks in the GC/MS data are compared
with the Chemicals of Highest Concern (CHC) mass spectral library
(see Table 2.1).  Peaks with fits/matches of 70% and greater are
considered tentatively identified.  For each tentatively
identified component, a list of the best mass spectral library
identifications (up to a total of ten identifications) is
reported along with the percent fit values, CAS number of the
tentative identification, HPLC fraction number, GC retention
time, and the concentration for the GC/MS component.   This report
is identified as Report 1 in the Effluent option flowchart,
Figure 3.2.

     The CHC mass spectral library can,  in nearly all cases, be
derived from the EPA/NIH/NBS mass spectral library using the
software on the data system of the GC/MS system.  The recommended
70% library searching fit criterion was chosen based upon best
scientific judgement and is subject to modification by the
regulatory authority.

     Computer algorithms for identifying unknown mass spectra via
library searching are often categorized as either forward or

                              111-16

-------
reverse searching.  In general, reverse searching algorithms have
demonstrated advantages for identifying unknown mass spectra when
the unknown is not chemically pure [31].  With GC/MS analyses,
mass spectral data can never be assumed to be pure and thus, in
the procedure, the use of reverse searching algorithms is
strongly recommended when available.  Unfortunately, some GC/MS
systems do not have reverse searching algorithms.  In these
cases, library searching should be performed using the default
algorithm provided by the manufacturer of the GC/MS system.


3.2.7     100 na/L Effluent Concentration Decision Point

     If a GC/MS component is not identified with the CHC mass
spectral library search (the no arrow in the flowchart), the
chemicals present at concentrations below 100 ng/1 are dropped
from further consideration.  The concentration of 100 ng/1 is
used as the cut-off for further investigation because 100 ng/1 is
the concentration of the surrogate compounds added to the sample
prior to extraction.  The surrogate concentrations represent the
minimum level at which adequate quantitation can occur.  The
amount of uncertainty associated with a value which lies below
the surrogate concentrations increases greatly with decreasing
concentration.  The purpose of this cutoff level is to prioritize
unknown/non-CHC peaks for further investigation.

     GC/MS components with concentrations greater than 100 ng/1
which are not identified with CHC mass spectral library are then
library searched with the EPA/NIH/NBS mass spectral library.


3.2.8     Library Searching using the EPA/NIH/NBS Mass Spectral
          Library

     GC/MS components not identified using the CHC library search
and with concentrations greater than 100 ng/1 are library
searched against the EPA/NIH/NBS mass spectral database.   Peaks
with fits/matches of 70% and greater are considered tentatively
identified.  These components are then evaluated in the
"(fraction BCF x concentration x dilution) > 1 /ig/kg"  rectangle
in Figure 3.2 (the yes arrow leaving the EPA/NIH/NBS mass
spectral library search rectangle).

     For those components with fits/matches less than 70% but
greater than 25%, the two best mass spectral library
identifications along with the percent fit values, the CAS number
of the tentative identifications,  HPLC fraction number, GC
retention time,  and the concentration are reported for each the
GC/MS component.  For GC/MS components with fits/matches less
than 25%, the concentrations,  HPLC fraction number,  and GC
retention times for these components should be reported and the
components labeled as being "unknown".  This report is identified
as Report 3 in the effluent option flowchart,  Figure 3.2.  The
tentative identifications listed in Report 3 are of less certain
                              111-17

-------
reliability and these results are provided for informational
purposes only.

     The EPA/NIH/NBS mass spectral library is available from the
U.S. Government Printing Office (941 North Capitol St. N.E.,
Washington, D.C. 20401).  In addition, most GC/MS manufacturers
have this library available in a form suitable for their
respective instruments.

     The recommended 70% library searching fit criterion was
chosen based upon best scientific judgement and is subject to
modification by the regulatory authority.  As with the CHC
library searching process, the use of reverse searching
algorithms is strongly recommended when available.  In cases
where reverse searching algorithms are not available, library
searching should be performed using the default algorithm
provided by the manufacturer of the GC/MS system.


3-2.9     (Fraction BCF x Concentration x Dilution) >  1 ttg/kcr

     For each of the tentatively identified components in the
EPA/NIH/NBS library search rectangle, the product of the fraction
BCF, effluent concentration, and effluent dilution is calculated
and compared to a 1 M<3/kg tissue residue concentration.  For
those GC/MS components with a product exceeding 1 ug/kg, a list
of the best mass spectral library identifications  (up to a total
of ten identifications) is reported along with the percent fit
values, CAS number of the tentative identifications, HPLC
fraction number, GC retention time, and the concentration for the
GC/MS component.  This report is identified as Report 2 in Figure
3.2.

     For those GC/MS components with a product less than 1 /ig/kg,
a list of the best mass spectral library identifications (up to a
total of ten identifications)  is reported along with the percent
fit values, CAS number of the tentative identifications, HPLC
fraction number, GC retention time, and the concentration for the
GC/MS component.  This report is identified as Report 4 in Figure
3.2.

     This decision point removes chemicals which are present at
concentrations too low to form residues in aquatic organisms
which pose risks to human consumers.  The l jug/kg cutoff for
residue concentrations in aquatic organisms was selected by
determining the reference tissue concentrations (RTCs) for all
chemicals in the EPA's IRIS database with log P values of 3.5 and
greater.  Using a fish consumption rate of 6.5 g/day and a risk
level of 10"6, RTC values were calculated for 83 nonpolar organic
chemicals (Appendix D) and these values are summarized in
Figure 3.3.

     Two chemicals, aldrin and alpha-hexachlorocyclohexane, have
RTCs below 1 Mg/fcg and for both chemicals, production and use in
the United States has been discontinued.  Both of these chemicals

                              111-18

-------
are included in the CHC mass spectral library and thus, GC/MS
components tentatively identified as aldrin or alpha-
hexachlorocyclohexane will be reported in Report 1.  These
components, those tentatively identified as aldrin and alpha-
hexachlorocyclohexane, will never be evaluated in the "(fraction
BCF x concentration x dilution) > 1
     Based upon the above considerations, i.e., no chemicals in
current production have RTCs below 1 M9/kg, and that the tissue
assessment option has detection limits approaching 1 /xg/kg for
tissue analyses, the 1 M RAC".
                              111-19

-------
Number of Chemicals
                                      CO
                                       •^
                                       (D
                                       CO
                                       CO
                                       DO

                                       O
                                       en
                                       O
                                       CO
                                       i
                                       o'
                                       o
                                      CO
                                       V
                                       CO
                                       en
         111-20

-------
3.2.11    RfD. al*. and/or Water Quality Standard?

     This decision rectangle in the effluent assessment option,
Figure 3.2, simply asks the question, "Is there a
Water Quality Standard available or sufficient information  (RfD
or ql*) to develop a RAC for a tentative identification reported
in Report 2?".  The purpose of this decision criterion is the
same as for Section 3.1.10.  For each GC/MS component, up to 10
tentative identifications might be reported.  For each of the
tentative identifications, the above question must be answered.
If any of the tentative identifications for a GC/MS component
have a RfD, ql*, and/or Water Quality Standard available, the
evaluation process for that component would proceed to the
"(concentration x dilution x 4) > RAC" decision rectangle, the
yes arrow in Figure 3.2.  Further information concerning the RfD,
ql*, reference concentrations, and BCF are presented in Chapter 4
of this guidance.

     If none the tentative identifications have the required
information for a given GC/MS component (i.e. peak), the
regulatory authority may evaluate the tentative identifications
further by obtaining improved BCF (and FM) data for each
tentative identification.

     This evaluation process is performed on all tentatively
identified GC/MS components listed in Report 2.  Furthermore, for
each tentatively identified GC/MS component, each tentative
identification should be evaluated.


3.2.12    (FM x BCF X concentration x dilution) > 1 ug/kg

     For those GC/MS components with no RfD, ql*, and/or water
quality standards available for any of their tentative
identifications (the no arrow leaving the "RfD, ql*, and/or Water
Quality Standards?" decision point), BCFs for each tentative
identification may be derived.  With these values and the FM for
the species of interest, the product of the FM, BCF,
concentration, and dilution is determined and compared to a
1 ^g/kg residue concentration.  If the product of any of the
tentative identifications for a GC/MS component are above the 1
Mg/kg value, the regulatory authority will issue a request for
confirmation of the identity of that component.  If the product
is below the 1 /zg/kg value for all tentative identifications, no
further action is taken on that component.  This evaluation and
decision is performed by the regulatory authority in order to
focus on those compounds of greatest potential hazard and to
reduce the potential for requiring confirmation for chemicals of
lesser importance for this assessment.


3.2.13    (Concentration x dilution x 4)  > RAC?

     This decision rectangle in the effluent option, Figure 3.2,
evaluates whether the measured concentration of a tentatively

                              111-21

-------
identified GC/MS component in the effluent is likely to be larger
than the reference ambient concentration (RAC).   The RAC is the
highest receiving water concentration for chemical that doesn't
result in a residue which poses health risks to humans consuming
fish and shell.  This comparison is based on the same principles
described in Section 3.1.11.

     For each GC/MS component, this comparison is performed for
each of the tentative identifications which have available Rfd,
ql*, and/or water quality standards.  If this comparison is true
for any of the tentative identifications for a GC/MS component, a
confirmation request is issued by the regulatory authority to the
discharger for that component (the yes arrow leaving this
decision point).  If this comparison is false for all of the
tentative identifications for a GC/MS component, no further
action (no arrow leaving this decision point) is recommended for
that component.

     All tentatively identified GC/MS components in Report 1, CHC
chemicals, are evaluated at this decision point.  For tentatively
identified GC/MS components in Report 2, only those tentative
identifications with RfD, ql*, and/or water quality standards
information available are evaluated.

     The calculation of the reference ambient concentrations
(RAC) is described in detail in Chapter 4 of this document.  RACs
are calculated using human dose information, Rfd and/or ql*,  BCF,
and fish consumption rates by humans.


3.2.14    Issue Confirmation Request

     In the effluent option, confirmation requests by the
regulatory authority can be issued from two different decision
points, see Figure 3.2.  In either case, the discharger is
required to confirm the identity of the GC/MS component
tentatively identified in the HPLC fractions.  This confirmation
process is discussed in further detail in Section 3.3.  The
results of the confirmation process are reported back to the
regulatory authority.  These results include the confirmed
identity of the GC/MS component and its concentration.
Confirmation of the identity of the chemicals is conducted at
this point in the logic sequence in order to focus on those
contaminants of greatest concern and to for complex samples
reduce the potential number of chemicals requiring confirmation.


3.2.15    Outputs From Effluent Assessment

     The outputs from the effluent assessment process are two
lists of identified bioconcentratable chemicals.  The first list
consists of chemicals from Reports 1 and 2 which require
reference concentration development (Chapter 4), wasteload
allocation (Chapter 5), and when necessary, permit limits
(Chapter 6).  The second list consists of chemicals which also

                              111-22

-------
arises from Report 2 and do not have RfD's, ql*'s, and/or water
quality standards.  Development of Rfd, ql* information for that
chemical may be requested to determine if current residue levels
pose serious risk to human consumers of fish and shellfish (the
yes arrow leaving the "(FM x BCF x concentration x dilution)  > 1
/ig/kg" decision point) .  Monitoring of these chemicals in
effluents is recommended in order to establish discharge
information.

     It must be noted that chemicals in both of these lists have
high potential to form residues in aquatic organisms.


3.2.16    Caveats for the Effluent Option

     The effluent assessment option will not detect all
bioconcentratable chemicals which form residues in aquatic
organisms.  The analytical procedures outlined above will detect
acid stable nonpolar organic chemicals with log P values of 3.5
and greater which can be successfully analyzed using GC/MS.
These procedures are fairly robust; however, they do have
limitations.  Some of these limitations include interferences
from hydrocarbons, sensitivity of the analytical method, the lack
of reliably measured BCFs and/or BAFs, and the possible absence
of RfD and ql* information for a given contaminant.  (Note:
hydrocarbon interference will only be a concern for facilities
with a predominance of this type of chemical,  such as
refineries.)


3.3  Confirmation

3.3.1     General Considerations

     When a discharger is issued a request for confirmation by
the regulatory authority, the primary objective for this process
is to attain an accurate identification of the GC/MS component.
Determining the exact concentration of a chemical is of secondary
concern for two reasons.  First, the confirmation of the
concentration of the chemical in the sample extract does not
directly affect attaining a BCF value for that chemical.  Second,
the procedures for the evaluation have an acceptable recovery
range of from 25-120%.  This means that there is a very low
likelihood of over estimating the actual concentration of a
bioconcentratable contaminant in an effluent or tissue sample.
In fact, it is more likely that the quantitation will be lower
than the actual effluent concentration.  It can be expected that
some of the contaminant will be lost in the analytical procedures
and not all of the contaminant will be recovered in the GC/MS
quantification.

     This potential underestimation is not a direct concern at
this stage of the procedure since the concentration is not used
for attaining the BCF value, the derivation of the RAG,  nor
developing control requirements such as NPDES limits.  The limit

                              111-23

-------
is unaffected by effluent or tissue concentration but rather
dependent on the standard or reference concentration and the
available dilution.  Underestimation does not become a concern
until that step of the procedure where the regulatory authority
will determine the need for a permit limit by comparing the RAG
with the receiving water concentration (RWC).   The discussion of
characterization of the effluent for human health effects in
Chapter 6 describes how this potential underestimation may be
addressed in that step of the procedure.


3.3.2     Confirmation Process

          Confirmation is performed by obtaining relative
retention times and mass spectral data for standards made from
pure reference chemicals.  When confirming a tentatively
identified chemical, chemicals analyzed on the GC/MS should
probably include the tentative identifications reported for the
GC/MS component and perhaps other logical chemicals.  The
discharger should utilize any available facility specific
information to assist in the confirmation since this information
will often decrease the cost and time required for confirmation.
Such data might include process information, chemical use, and/or
MSDS information.

     In some cases, none of the tentative identifications
assigned to a GC/MS component will be the correct identification.
Additional confirmation analyses beyond the initial tentative
identifications and other logical choices will require increased
efforts and costs, and it must be recognized that there is no
single analytical approach that will work with all effluents
and/or tissue extracts.  The analytical methods for identifying
such a chemical will in all likelihood be site specific and will
depend upon the complexity of the sample, the knowledge of the
processes which generate the discharger's effluent, and the
complexity of the mass spectrum for the unknown.  Some of the
analytical approaches which may be utilized in conjunction with
available information on the facility are:  1) concentrating the
extract and reanalyzing the fraction on the GC/MS, 2)
inventorying and evaluating the chemicals used and produced by
the manufacturing process, 3) evaluating the reaction pathways in
the manufacturing process for reaction by-products, 4) analyzing
the extract using different GC conditions, GC columns, MS
conditions, and/or MS systems, 5) manual interpretation of the
mass spectrum for the unknown, and 6) performing more exotic MS
analyses such as GC/MS/MS or GC/FTIR/MS with manual
interpretation of the spectral data.

     A chemical would be considered confirmed when its relative
retention time and mass spectral data from the tissue or effluent
extract match the GC/MS data for the standard made from the pure
reference chemical.  This procedure should be conducted by
following the method described in EPA method 1625 [39]„
Quantification of the confirmed chemical in the extract should


                              111-24

-------
also be performed using  EPA method  1625 with d12-chrysene as the
internal standard [39],

     The discharger should report the results of the  confirmation
analyses including the confirmed "true" identification  and
quantification as well as the results for each of the other
tentative identifications which were analyzed but found to  be
false identifications.   Additional  QA/QC information  should also
be reported as described in the EPA method utilized for the
confirmation.
3.4  Sediment Assessment

3.4.1     General Considerations

     In some receiving waters, sediments may be a significant
source for bioconcentratable chemicals.  In this section, an
analytical procedure for identifying bioconcentratable chemicals
in sediments is outlined.  A flowchart outlining the sediment
analytical procedure is given in Figure 3.4.  A detailed
procedure, in a form suitable for dispersement to contract
laboratories, is provided in Appendix C.  This procedure is
combines various portions of the tissue and effluent analytical
procedures in Appendices A and B to derive this methodology for
sediments.
3.4.2     Sampling

     As with the tissue and effluent assessment options,
collection of representative samples is important for accurate
assessment of the bioconcentratable chemicals present in the
sediments.  Collection of representative samples will involve the
following site-specific decisions:

     When to Sample

     In some receiving waters, sediments are the ultimate sinJc
for bioconcentratable chemicals since sedimentation processes
will gradually bury the sediments containing the residue forming
chemicals.   Sedimentation processes occur on a much slower time
scale than the processes associated with the water column and
biotic compartments of the receiving water.  Consequently,
sediment sampling can be performed almost anytime during the year
and result in representative samples.
                             111-25

-------
                                    Figure 3.4

                         SEDIMENT ANALYSIS
                            Collect & extract sediment sample
                                Fractionate via HPLC
     I  Analyze via GC/MS
No further action  I
Reverse-search all chromatographic peaks
 against the CHC mass spectral library
       identified fit >70%?
                                                     Reverse-search EPA/NIH/NBS
                                                        mass spectral library
                                                         identified fit >70%?
       REPORT: U*l
                                                         REPORT: list 2
                                                                                     REPORT: list 3
                                             HI-26

-------
     In the absence of other evidence, sampling should be done
after periods of normal or sub-normal dilution of the effluents.
Sampling is not recommended during and after major storm events,
i.e., floods and hurricanes, and possibly, during spring runoff
since sediment resuspension during these events can dramatically
alter the composition of the sediments.  Generally, sampling once
or possibly, twice a year will be sufficient for an initial
assessment.  Additional sampling, i.e., monthly or weekly, may be
necessary for a more comprehensive sediment assessment.

     How to Sample

     Surface sediments, those in the benthic mixing zone, should
be collected using a Ponar dredge, Eckman dredge and/or coring
sampler.  The coring sampler type is preferred since better
control is provide so that surface sediments can be collected.
Sediments with low organic carbon content, i.e., those with high
sand and gravel content, should not be collected since organic
bioconcentratable chemicals reside in the organic carbon fraction
of the sediment.  Collected sediments should be placed into clean
solvent rinsed glass jars and sealed with aluminum foil or teflon
lined lid.  Samples should be stored in the dark at 4°C.

     Where to sample

     The location of sampling is a site-specific decision.  In
general, samples should be collected in the depositional zone
close to the outfall of interest.  Collection of samples in the
mixing zone of the effluent may not be representative since
deposition of particulate matter with the bioconcentratable
chemicals from that discharge may not occur there.  Collection of
samples too far from the outfall is not recommmended since
dilution and degradation may diminish the residue concentrations
in the sediments.

     Sampling sites with well defined depositional zones and
receiving water hydraulics may be assessed with one well placed
site.  As the complexity of the receiving water and deposition
zones increase, additional sampling sites along a gradient may be
necessary.


3.4.3     Sample Extraction

     The extraction procedure consists of mixing 20 grams of
ground, air dried, sediment with anhydrous sodium sulfate.  Half
of this mixture is placed into a Soxhlet extraction thimble and
three surrogate chemicals are spiked onto the sediment in
extraction thimble.  The remaining sodium sulfate/sediment
mixture is then placed into the extraction thimble and the
extraction thimble is placed into a Soxhlet extractor body.  The
sediment is extracted using acetone for 4 hours, and then,
Soxhlet extracted using 1:3 toluenermethanol for a minimum of 12
hours.   The extract,  acetone and toluene:methanol, is

                              111-27

-------
concentrated using a Kuderna-Danish concentrator to approximately
10 mis.  Percent moisture and organic carbon of the sediment are
measured using standard techniques.

     The surrogate chemicals, d10-biphenyl, 13C6-1,2,4,5-
tetrachlorobenzene, and 13C6-hexachlorobenzene,  are added to the
sediment at a 5 jug/kg concentration prior to extraction of the
sample.  These chemicals are referred to as surrogates because
their behavior mimics that of bioconcentratable chemicals in the
analytical procedure.  The surrogate chemicals are used for
quality control and quantification in this analytical procedure.


3.4.4     Sample Cleanup

     The sediment extract is cleaned up using the procedures
outlined in the effluent option, see Section 3.2.3.  Elemental
sulfur is removed from the sediment extract by passing the
extract through a column containing activated copper filings.
The extract after removal of the sulfur is concentrated to 0.5
mis.
3.4.5     HPLC and GC/MS Analysis

     The sediment extract is fractionated using the procedures
outlined in the effluent option, see Section 3.2.4.  The three
HPLC fractions are analyzed using the procedures outlined in the
effluent option, see Section 3.2.5


3.4.6     Library Searching using CHC Mass Spectral Library

     The GC/MS data for the HPLC fractions are library searched
using the procedures outlined in the effluent option, see Section
3.2.6.  Peaks with fits/matches of 70% and greater are considered
tentatively identified.  For each tentatively identified
component, a list of the best mass spectral library
identifications (up to a total of ten identifications) is
reported along with the percent fit values, CAS number of the
tentative identification, HPLC fraction number, GC retention
time, and the concentration for the GC/MS component.  This report
is identified as Report 1 in the sediment analysis flowchart,
Figure 2.3.


3.4.7     5 ug/kg Sediment Concentration Decision Point

     For GC/MS components not tentatively identified in the CHC
mass library search, those present at concentrations less than 5
Mg/kg are dropped from further consideration.  For further
information, see the discussion in the tissue option, Section
3.1.7.
                              111-28

-------
3.4.8     Library Searching using the EPA/NIH/NBS Mass Spectral
          Library

     The GC/MS components not tentatively identified by the CHC
mass spectral library search and having concentrations greater
than 5 M9/kg are library searched against the EPA/NIH/NBS mass
spectral library.  Two reports are generated by this procedure,
Reports 2 and 3  (see Figure 3.4). For further information, see
the discussion in the tissue option section 3.1.8.


3.4.9     Analytical Summary

     Three reports are generated by the sediment analytical
procedure outlined above.  These reports, Reports 1, 2, and 3,
are identical in format to those created in the tissue option.
The actual GC/MS chromatogram should also be submitted as part of
the reporting requirements.  The detailed analytical procedure is
presented in Appendix C.


3.4.10    Caveats for the Sediment Analytical Procedure

     The sediment analytical assessment procedure will not detect
all bioconcentratable chemicals which form residues in aquatic
organisms.  The analytical procedures outlined above will detect
acid stable nonpolar organic chemicals with log P values of 3.5
and greater which can be successfully analyzed using GC/MS.
These procedures are fairly robust; however, they do have
limitations.  Some of these limitations include interferences
from hydrocarbons, sensitivity of the analytical method, and our
ability to identify unknown GC/MS components.  (Note:
hydrocarbon interference will only be a concern for facilities
with a predominated with this type of chemical,  such as
refineries.)


3.5  Quality Assurance/Quality Control for Analytical Procedures

     Data obtained as a result of the tissue, effluent, and
sediment analysis procedures must pass their quality control (QC)
requirements.  In Appendices A, B, and C, these requirements are
presented with the detailed procedure.

     For all of the procedures, recoveries for the surrogates
must be greater than 25% but not more than 120%.   If this
criteria is not met,  the analytical data is not of sufficient
quality and the analysis must be repeated.

     For the effluent and sediment procedures which use HPLC
fractionation,  the HPLC fractions 1,  2,  and 3 should contain the
surrogates d10-biphenyl, 13Cp-l,2,4,5-tetrachlorobenzene, and 13C6-
hexachlorobenzene, respectively.  If surrogates are in the wrong
HPLC fractions,  the analytical data is not of sufficient quality
and the analysis should be repeated.

                             111-29

-------
     QC information must also be provided for HPLC, GC, and MS
performance as specified and this information must be evaluated.
If the HPLC, GC, and MS performance criteria are not met, the
analyses should also be repeated.  For samples which do not meet
the QC criteria, the discharger should be required to analyze a
new sample.  QC information should be included in the reported
information to the regulatory authority.
                              111-30

-------
                            CHAPTER 4

                Reference Concentration Derivation
     The assessment options described in the preceding chapter
produce two list of identified bioconcentratable chemicals.  The
first list consists of chemicals from Reports 1 and 2 which
require reference concentration development and the second list
consists of those chemicals from Report 2 which lack RfD's,
glut's, and/or water guality standards.  The identity of these
chemicals, from both lists, will have been confirmed.

     This chapter presents procedures for developing reference
concentrations if and when no State numeric water quality
standard exists for a given contaminant to protect human health
based on the consumption of fish.  These procedures require
knowledge of fish consumption rates, lipid content of the species
consumed, human dose information (RfD and/or ql*). and the BCF
(or BAF) for each bioconcentratable chemical.

     The reference ambient concentrations (RACs) and reference
tissue concentrations (RTCs) are used in the effluent and tissue
assessment options respectively, to derive the first list of
bioconcentratable chemicals discussed above.  The RAG is also
used in both the tissue and effluent options for waste allocation
and permit development.   The RACs and RTCs are derived to protect
human health from contaminants in consumed fish only.  RACs and
RTCs are not derived to protect aquatic organisms from direct
toxicity or human health from drinking water.


4.l  Establishing RACs for Bioconcentratable Contaminants

     This section describes procedures to avoid unacceptable
tissue residues in organisms used for human consumption.  The
procedures for setting levels of exposure or concentrations for
human consumption are not a part of this document.   A complete
human health effects discussion is included in the (draft)
"Guidelines and Methodology Used in the Preparation of Health
Effects Assessment Chapters of the Consent Decree Water
Documents." by EPA's Environmental Criteria and Assessment
Office, Cincinnati (ECAO-Cin)  [40].

     The procedures contained in the above ECAO-Cin document are
used in the development and updating of EPA water quality
criteria and may be used in developing RACs for those pollutants
lacking State or EPA human health criteria.   Although the same
procedures are used to develop criteria and RACs,  only those
values which are subjected to the regulatory process of Regional,
State, and public comment can be considered "criteria."  RACs may
be applied as site-specific interpretation of narrative standards

                              IV-1

-------
and as a basis for permit limitations under 40 CFR
122.44(d)(1)(vi).  Developing Reference Ambient Concentrations
(RACs) using the guidelines established in this document does not
absolve States from developing criteria under 303(c)(2)(B).

4.2  Human Exposure Considerations

     The RAC value for protecting human health should include all
exposures.   These include consumption of water, consumption of
fish, and combined consumption of both water and fish on a per-
person per-day basis.  Levels of actual human exposures from
consuming contaminated fish vary depending upon a number of case-
specific consumption factors.  These factors include type of fish
species consumed, type of fish tissue consumed, tissue lipid
content, consumption rate and pattern, and food preparation
practices.   In addition, depending on the spatial variability in
the fishery area, the behavior of the fish species, and the point
of application of the RAC or criterion, the average exposure of
landed fish may be only a small fraction of the expected exposure
at the point of application of the criterion, or if an effluent
attracts fish, it might be greater than the expected exposure.
Another consumption factor is the percentage of fish consumed
from the fishery area of concern, which could vary from 0% to
100%, depending on size, character and the value of the fishery.

     With shellfish, such as oysters, snails, and mussels, the
entire body tissue is commonly eaten, whereas with fish,  only
muscle tissue and roe are commonly eaten.  This difference in the
types of tissues consumed has implications for the amount of
bioconcentratable contaminants likely to be ingested.   Whole body
shellfish consumption presumably means ingestion of the entire
burden of bioconcentratable contaminants.  However, with most
fish, selective cleaning and removal of internal organs,  and
sometimes body fat as well, from edible tissues may result in
removal of the lipid rich tissues that contain the majority of
bioconcentratable contaminants.

     This document focuses primarily on fish consumption and only
incidentally on the drinking water exposure route.   A complete
human exposure evaluation for toxic pollutants of
bioconcentration concern would not only encompass estimates of
exposures due to fish consumption, but also exposure due to
background concentrations and other exposure routes, including
recreational contact, occupational, dietary intake from other
than fish,  inhalation of air, etc.  However, other exposure route
information should be considered to the extent it is available.
4.3  Fish Consumption Values

     EPA's 1980 human health criteria assumed a human body weight
of 70 kg and the consumption of 6.5 g/day of fish and shellfish.
The national fish consumption value for freshwater and estuarine
fish and shellfish, calculated by the US EPA based on data from a

                               IV-2

-------
1974 food survey, was 6.5 g/day with the 95th percentile for all
seafood at 42 g/day  [41].  The mean lipid content of fish tissue
found to be consumed was 3.0% [42].

     EPA recommends 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.  For
example, some States have adopted their own fish consumption
estimates, ranging from 20 g/day in Wisconsin, Louisiana,
Illinois and Arizona to 37 g/day in Delaware for salt water
species [43].

     Currently, four levels of fish consumption are provided in
the EPA guidance manual "Assessing Human Health Risk from
Chemically Contaminated Fish and Shellfish [44]"  These are:

     6.5 g/day to represent a low estimate of average consumption
     of fish and shellfish from estuarine and fresh waters by the
     entire U.S. population [45].  This fish consumption level is
     based on the average of both consumers and non-consumers of
     fish.

     20 g/day to represent a high estimate of the average
     consumption of fish and shellfish from marine, estuarine,
     and fresh waters by the entire U.S. population [46].  This
     average also includes both consumers and non-consumers of
     fish.

     165 g/day to represent average consumption of fish and
     shellfish from marine, estuarine, and fresh waters by the
     99.9th percentile of the U.S. population consuming the most
     fish or seafood [47],

     180 g/day to represent a "reasonable worst case" based on
     the assumption that some individuals would consume fish at a
     rate equal to the combined consumption of red meat, poultry,
     fish, and shellfish in the U.S. (EPA Risk Assessment Council
     assumption based on data from the USDA Nationwide Food
     Consumption Survey of 1977-1978; see Appendix H [44]).

     EPA is currently updating the national estuarine and
freshwater fish and shellfish consumption values to provide a
range of recommended national consumption values.   This range
will include:  1) mean values appropriate to the population at
large,  and 2) values appropriate for those fishermen who consume
a relatively large proportion of fish in their diets (maximally
exposed individuals).


4.4  BCF Evaluation/Selection

     In order to derive a reference concentration the regulatory
authority will need to determine a BCF for each identified
bioconcentratable compound using one of the following methods:

                              IV-3

-------
1)   Use the estimated BCFs from the Quantitative Structure
     Activity Relationship (QSAR) [48] data base.  BCF values in
     QSAR are based on 7.6% lipid content and may need to be
     normalized to 3% by multiplying by 0.395 (3%/7.6%).

2)   For compounds with laboratory-derived and/or estimated log P
     values (laboratory-derived is preferred [49]), but unknown
     BCFs, use the following relationship to estimate the BCF at
     a 3% lipid content [22]:

          log BCF = 0.79 log P - 0.40 - log (7.6/3.0)

3)   For compounds for which BCF or log P values cannot be
     attained using 1 or 2, use the average BCF for the fraction
     in which the compound was found.  These averages were
     calculated by applying the equation above in 2) to the
     average log P in each of the fractions in the HPLC
     separation procedure described in Chapter 3.

                    Average  BCF  for  Fraction

          Fraction            (3% lipid content)

              I                      230
              2                    1,700
              3                   49,000


     In situations where metabolism of the chemical is known or
suspected to be of importance or where information on measured
BCF values are desired, the following sources of information may
be consulted:

1)   US-EPA Ambient Water Quality Criteria documents issued in
     1980 or later.
2)   AQUIRE on-line data base.
3)   Published scientific literature.
4)   Reports issued by US-EPA or other sources.
5)   Unpublished data which meets minimum ASTM test requirements.

     For BCFs selected using 2 through 5 the the minimum test
requirements as specified in Section 2.6 must be met.  When more
than one BCF value meets the minimum data requirement, the
geometric mean of those BCFs, after normalization for lipid
content, should be used.  In order to calculate the geometric
mean BCF value, the acceptable BCF values should be normalized to
3.0% lipid content using the following relationship:

          BCF at 3% lipid content =
                     (BCF @ L% lipid content)  x (3.0/L)
                               IV-4

-------
4.5  Bioaccumulation Considerations;  Food Chain Multiplier
     Selection

     In this document, bioaccumulation considerations are
incorporated in the calculation of the RTCs and RACs by use of
the food chain multiplier  (FM) term along with the BCF value.  In
Table 4.1, FM values derived from the work of Thomann [17,18] are
listed according to log P value and trophic level of the
organism.  Trophic level 4 organisms are generally the most
desirable species for sport fishing and therefore, FMs for
trophic level 4 are recommended for use in the equations for
calculating RTCs and RACs.  In those situations where only lower
trophic level organisms are found, e.g., oyster beds, a FM for a
lower trophic level might be used in calculating the RTCs and
RACs.

     Experimentally measured BAFs reported in the literature may
be used when available (this may be most desirable for those
chemicals with log P values above 6.5).  To use experimentally
measured BAFs in calculating the RAG or RTC, the "(FM x BCF)"
term, is replaced by the BAF in the equations in Sections 4.7-
4.10.  Relatively few BAFs have been measured accurately and
reported, and their application to sites other than the specific
ecosystem where they were developed is problematic and subject to
uncertainty.  The BAF may have to be corrected for trophic level
if the measured BAF is for a lower trophic species and the
species of interest is a higher predatory species.  Of course,
the option is also available to develop BAFs experimentally,  but
this will be extremely resource intensive if done on a chemical
and site specific basis with all the necessary experimental and
quality controls.


4.6  Updating Criteria and Generating RACs and RTCs Using IRIS

     EPA recommends that the process of updating criteria and
generating RACs and RTCs use the most current risk information.
The Integrated Risk Information System (IRIS)  is an electronic on
line data base of the EPA that provides chemical-specific risk
information on the relationship between chemical exposure and
estimated human health effects.  The health assessment
information contained in IRIS, except as specifically noted,  has
been reviewed and agreed upon by an interdisciplinary group of
scientists representing various Program Offices within the Agency
and represent an Agency-wide consensus.  Risk assessment
information and values are updated on a monthly basis and are
approved for Agency-wide use [50].

     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
management decisions.   IRIS is available to Federal and some
State and local environmental agencies through the EPA's


                               IV-5

-------
                       Table 4.1
        Estimated Food Chain Multipliers (FMs)
                           Trophic Levels


               Log P      2      3       t
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
>6.5
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
10.1
12.5
15.4
19. 2^
19.2*
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
12.7
16.5
21.4
25.2
29.4
34.1
39.3
44.9
44.9*
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
11.4
16.2
23.2
33.3
47.2
66.5
75.2
84.1
91.8
98.4
103.8
103.8*
These 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.
                          IV-6

-------
 electronic MAIL system and is also available to the public
through the Public Health Network  (PHN) and TOXNET.  Since IRIS
is designed to be a publicly available database, interested
parties may submit studies or documents for consideration by the
appropriate interdisciplinary review group for chemicals
currently on 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, IRIS
contains a series of sections identified by service codes which
serve as a user's guide as well as provide background
documentation on methodology.  Additional information is
available from IRIS USERS SUPPORT: 513/FTS 684-7254.

     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 non-
carcinogenic (target organ) effects.  The carcinogenic potency
estimate (formerly known as ql )  represents the  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 E contains additional information
on the derivation of RfDs and ql*s.

     EPA periodically updates risk assessment information
including RfDs, cancer potency estimates,  and related information
on contaminant effects, and reports the current information on
IRIS.  A list of the BCF, RfD and carcinogenic potency estimates
values current at the time of publication of this document is
included in Appendix D.  The inclusion of this list in this
document is intended for use in initial screening only.     Since
IRIS contains the Agency's most recent quantitative risk
assessment values, current IRIS values should always be used in
developing new RACs.  This also applies to updating the 1980 EPA
human health criteria.  The procedure for deriving an updated
human health water quality criterion would require inserting the
current Rfd or Carcinogenic potency estimate on IRIS into the
appropriate equation in Sections 4.7 - 4.10.   In the absence of a
promulgated State numeric standard, the RAC may be calculated
using these formulas.  Appendix D should not be used to revise
any criteria or derive RACs without checking IRIS for the most
current values.

     Figure 4.1 shows the procedure for updating a criterion or
deriving a RAC using IRIS data.   If a chemical has both
carcinogenic and non-carcinogenic effects; i.e./ both a cancer
potency estimate and RfD/ the carcinogen RAC formula in Section
4.8 should be used as it will typically result in the more
stringent RAC of the two.
                               IV-7

-------
                               Figure 4.1
         Procedure for Revising an EPA Human Health Criterion
           or Developing a Reference Ambient Concentration
                             EPA's water quality
                              criterion available?

[ IRIS
upd.

data ^| ^
ated?
YES




NO C
\ Data
exists 1
1 in IRIS? 1

( Calculate RAC ]
NO V J

YES

{ Calculate RAC 1

NO
Use current criterion

                                                      Evaluate other sources of data:
                                                      HEAST. Rlsk»Asslstant, drinking
                                                      water MCL's, fish consumption
                                                      advisory levels, FDA action levels, cti

-------
4.7  Calculating RACs for Non-Carcinogens
     The RfD is an estimate of the daily exposure to the human
population (including sensitive subgroups) that is likely to be
without appreciable risk of causing deleterious effects during a
lifetime.  The RfD is expressed in units of mg toxicant/kg human
body weight/day.

     RfDs are derived from the "no observed adverse effect
level"(NOAEL) or in the absence of a NOAEL the "lowest observed
adverse effect level" (LOAEL) identified from chronic or
subchronic human epidemiology studies with clearly defined
exposure levels or from experimental animal studies.  LOAEL and
NOAEL refer to animal and human toxicology and are therefore
distinct from the aquatic toxicity terms "no observed effect
concentration (NOEC) and the "lowest observed effect
concentration" (LOEC).  Uncertainty factors are then applied to
the NOAEL or LOAEL to account for uncertainties in the data
associated with variability among individuals, extrapolation from
nonhuman test species to humans, data on other than long-term
exposures and the use of a LOAEL [44].  An additional uncertainty
factor may be applied to account for significant weakness or gaps
in the database.

     The RfD is a threshold below which adverse effects are
unlikely to occur.  While exposures above the RfD increase the
probability of adverse effects, they do not produce a certainty
of adverse effects.  Similarly, while exposure at or below the
RfD reduces the probability, it does not guarantee the absence of
effects in all persons.   The RfDs contained in IRIS are values
that represent EPA's consensus judgement.  These values may have
uncertainty spanning perhaps an order of magnitude.

     For non-carcinogenic effects,  an updated criterion or an RAC
can be derived using the following equation:

Equation 4.1:

          C or RAC (mg/1)   =  (RfD x WT) - (DT + IN)  x WT
where:
  C
RAC
RfD

 WT
 DT

 IN

 WI
            WI + [FC X L X (FM X BCF)]

updated water quality criterion (mg/1)
reference ambient concentration (mg/1)
reference dose (mg toxicant/kg human body
weight/day)
weight of an average human adult [70 kg]
dietary exposure [other than fish]
(mg toxicant/kg human body weight/day)
inhalation exposure
(mg toxicant/kg human body weight/day)
average human adult water intake
[2 liters/day]
                               IV-9

-------
           FC =     daily fish consumption (kg fish/day)
            L =     ratio of lipid fraction of fish tissue
                      consumed to 3%
           FM =     food chain multiplier [from Table 4.1]
          BCF =     bioconcentration factor (mg toxicant/
                    kg fish divided by mg toxicant/L water) for
                    fish with 3% lipid

     If the receiving water body is not used as a drinking water
source, the factor WI can be deleted.  Where dietary and/or
inhalation exposure values are unknown, these factors may be
deleted from the above calculation.  For identified non-
carcinogenic chemicals without known RfDs, State or EPA
procedures can be used to estimate the RfD (see Appendix E).


4.8  Calculating RACs for Carcinogens

     Any human health criterion for a carcinogen is based on at
least three inter-related considerations:  potency, exposure, and
risk characterization.  States may make their own judgments on
each of these factors within reasonable scientific bounds, but
documentation to support their judgments should be clear.

     Maximum protection of human health from the potential
effects of exposure to carcinogens via contaminated fish would
require an RAG of zero.  The zero level is based upon the
assumption of non-threshold 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 the zero level may never be attainable, a numerical
estimate of risk (in Mg/1) which 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 (i.e.,  10"5 risk level = 1 additional cancer in a
population of 100,000).  Cancer risk is calculated by multiplying
the experimentally derived cancer potency estimate by the
concentration of the chemical in the fish and the average daily
human consumption of contaminated fish.  The risk for a specified
population (i.e. 100,000 people or 10"5) is then calculated by
dividing the risk level by the specific cancer risk.  EPA's
ambient water quality criteria documents provide risk levels
ranging from 10  to 10"7 as examples.

     When the cancer potency estimate, or slope factor (formerly
known as the ql*),  is derived using animal studies, high dose
exposures are extrapolated to low dose concentrations and
adjusted 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 which
the model postulates to occur at low doses.   When based on human
(epidemiological) data, the slope factor is based on the observed
increase in cancer risk, and is not extrapolated.  For deriving

                              IV-lo

-------
RACs for carcinogens, the oral cancer potency estimates or  slope
factors from IRIS are used.

     It is important to note that cancer potency factors may
overestimate 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 vs. animal data) and 2) limited
information regarding the mechanism of cancer causation.  The
actual risk may be much lower, perhaps as low as zero,
particularly for those chemicals for which human carcinogenicity
information is lacking.  Risk levels of 10"5, 10"6, and 10  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 chemicals is not necessarily the overall risk
from exposure to water.  For example an individual risk level of
10"6 may yiel a higher overall risk level if multiple carcinogenic
chemicals are present.

     For carcinogenic effects, the RAC can be determined by using
the following equation:
Equation 4.2:
           c or RAC  (mg/1)
                             (RL x WT1
                              ql* [WI + FC x L X  (FM x BCF)]
where:
  C
RAC
 RL
 WT
ql*
 WI

 FC
  L

 FM
BCF
updated water quality criterion (mg/1)
reference ambient concentration (mg/1)
risk level (10~x)
weight of an average human adult [70 kg]
carcinogenic potency factor (kg day/mg)
average human adult water intake
[2 liters/day]
daily fish consumption (kg fish/day)
ratio of lipid fraction of fish tissue
consumed to 3%
food chain multiplier (from Table 4.1)
bioconcentration factor (mg toxicant/
kg fish divided by mg toxicant/L water) for
fish with 3% lipid
     If the receiving water body is not used as a drinking water
source, the factor WI can be deleted.  For identified
carcinogenic chemicals without known cancer potency estimate
values, extrapolation procedures can be used to estimate the
cancer potency (see Appendix E).

4.9  Calculating RTCs for Non-Carcinogens

     The following formula may be used to calculate a RTC for a
non-carcinogens.  Readers may also consult EPA's "Assessing Human
Health Risks from Chemically Contaminated Fish and Shellfish"
[44].
                              IV-11

-------
     The basic equations for deriving RTCs uses the same
variables as in Equations 4.1 where the BCF is normalized at 3.0%
lipid:

Equation 4.3:  RTC  (mg/kg)  =   (RFD x WT) - (DT + IN) x WT
                                  WI /(BCF X FM X L) + FC

     The above equation should be corrected for site specific
lipid content and bioaccumulation factors where data are
available.
4.10  Calculating RTCs for Carcinogens

     The following formula may be used to calculate a RTC for a
carcinogens.  Again, the basic equations for deriving RTCs use
the same variables as in Equations 4.2 where BCF is normalized at
3.0% lipid:

Equation 4.4:  RTC (mg/kg)  = 	RL x WT	
                              ql* [WI/(BCF X FM X L) + FC]

     The above equation should also be corrected for site
specific lipid content and bioaccumulation factors where data are
available.

4.11 Calculating RTCs from State Numeric Water Quality Standards

     When a State numeric water quality standard is available for
a chemical, a RTC may be needed for the evaluation of tissue
residue assessment results.  In these cases an RTC can be
calculated from the numeric standard by using the BCF
relationship.  In equation form:

Equation 4.5:

     RTC (mg/kg)  = (BCF x FM x L) x (State numeric standard)

     The above equation should be corrected for site specific
lipid content (L) and bioaccumulation factors (BCF x FM) where
data are available.
4.12 Deriving Quantitative Risk Assessment in the Absence of
     IRIS Values

     The RfDs or cancer potency factors provide the existing
quantitative risk assessments for developing RACs.  However, as
indicated above, effluents and tissues may contain identified for
which no risk value is currently available on IRIS.  When IRIS
data are unavailable, quantitative risk level information may be
determined according to a State's own procedures.  Some States
have established procedures whereby risk factors can be developed
based upon extrapolation of acute and/or chronic animal data to

                              IV-12

-------
concentrations of exposure protective of fish consumption by
humans (See Appendix H).  Where no State procedure exists, risk
assessment values may be based upon extrapolation from mammalian
or other data using the IRIS documentation in Appendix E or
information available through other EPA risk data bases such as
HEAST or Risk*Assistant (Appendix H).  Also, where no other
information or procedure exists, drinking water maximum
contaminant levels (MCLs)  or FDA Action Levels may be used as
guidance in developing numerical estimates.

     As a general matter,  some of the assumptions made in
deriving FDA action levels for fish and shellfish (as well as FDA
levels of concern and tolerance levels) are inappropriate for use
in regulating water quality [44],  In particular, FDA exposure
assumptions, in accordance with its legislative mandate, reflect
expected consumption by buyers of fish in interstate commerce.
FDA generally assumes, for example, that contaminated fish would
not constitute a high proportion of such a consumer's diet.  In
contrast, to adequately protect the user, EPA and States must
consider the individual who frequently fishes at the site being
regulated or who regularly eats fish from the area.   Thus,
without a separate analysis and determination that at a
particular site the FDA level of concern sufficiently meets the
Clean Water Act's objective, the FDA level of concern is not an
appropriate basis for a water quality criterion or RAC.
                              IV-13

-------
                            CHAPTER 5

          Exposure Assessment and Wasteload Allocations
     The purpose of this chapter is to provide procedures for
developing wasteload allocations (WLAs) for bioconcentratable
contaminants.  Procedures for deriving bioconcentration-based
WLAs differ somewhat from derivation of aquatic toxicity-based
WLAs because bioconcentration-based WLAs involve human health
protection and thus, can involve different exposure
considerations.  Exposure considerations and design conditions
for control of bioconcentratable pollutants involve longer time
frames than for pollutants causing acute or chronic aquatic
because the total lifetime exposure to humans must be considered.

     At this step in the control of bioconcentratable pollutants,
chemicals of concern have been identified, and RACs have been
developed or State water quality standards identified.  Where
State numeric criteria for the protection of human health from
bioconcentration in consumed fish are not available, an RAC
should be developed, in accordance with 40 CFR 122.44(d)(1)(vi)
using the procedures described by the State for interpretation of
the narrative standards or if none exist, the procedures
described in Chapter 4 should be used.  The next step is to
calculate the WLA to determine if permit limits are necessary.

     The WLA condition is that which attains the RAC at the
critical design conditions.  In discharge situations where
nonpoint source contributions are significant, controls of point
sources alone may be insufficient to attain State water quality
standards.  Critical conditions at alternate higher flow
conditions may be necessary in these instances.  After developing
a human health based WLA for a given chemical, the two WLAs for
protection of aquatic life should be compared with the human
health WLA, and the most stringent of these should be used in
permit limit development.  The permit limit is the effluent
concentration and/or mass limitation that attains the WLA for the
required exposure periods.  Where the human health WLA is
determined to be most stringent, the regulatory authority should
follow the procedures in Chapter 6 to apply this WLA in permit
limit development.  The reader is referred to the TSD for a more
detailed discussion of exposure assessment and wasteload
allocations.  The discussion in this document is intended to
provide a brief introduction to this subject, focusing on those
aspects of direct concern for the control of bioconcentratable
pollutants.
                               V-l

-------
5.1  Total Maximum Daily Loads

     The development of a wasteload allocation is based on the
existing total maximum daily load (TMDL)  for the receiving water.
A TMDL is the sum of the individual wasteload allocations (WLAs)
for point sources and the load allocations (LAs) for nonpoint
sources of pollution and natural background sources, tributaries,
or adjacent segments.  A TMDL is the amount of a pollutant from
point sources and nonpoint and natural background sources,
including a margin of safety, that may be discharged to a water
quality-limited waterbody without exceeding the applicable water
quality criterion.  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
wastewater pollutant loads are set and controls on nonpoint
source loadings are established so that predicted receiving water
concentrations do not exceed water quality criteria or RACs.
TMDLs and WLAs/LAs should be established at levels necessary to
attain and maintain the applicable narrative and numerical water
quality standard  (WQS)  with seasonal variations and a margin of
safety that takes into account any lack of knowledge concerning
the relationship between point and nonpoint source loadings and
water quality.  Determination of WLAs and TMDLs should take into
account critical conditions for stream flow,  loading, and water
quality parameters.  WLAs and Permit limitations should be issued
based on TMDLs where available.  A more detailed discussion of
TMDLs can be found in the Technical Support Document [32].


5.2  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 concentrations above the criteria in small areas near
outfalls.  These areas are called mixing zones.  Since these
areas of impact, if disproportionally large,  could potentially
adversely impact the productivity of the waterbody, and have
unanticipated ecological consequences, they should be carefully
evaluated  and appropriately limited in size.  As 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 that, considering likely
pathways of exposure, there are no significant human health

                               V-2

-------
risks.  One means to achieve these objectives is to limit the
size of the area affected by the mixing zones.

     In the general case, where a State has both acute and
chronic aquatic life criteria, as well as human health criteria,
independently 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 sized to protect the ecology of the waterbody as a
whole, and the health criteria mixing zone sized to prevent
significant human risks.  The acute mixing zone need not be
smaller than the other two.  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
requirements.  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 Clean Water
Act goals of maintaining and improving water quality.  EPA
recommendations for allowances for mixing zones, and appropriate
cautions about their use, are contained in this section.


     The Clean Water Act allows mixing zones at the discretion of
the State.  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.

     In all cases, the size of 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 combined 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 multi-step procedure for
evaluating the overall acceptability of mixing zones [32].

     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
residues are a concern (either because of measured or predicted

                               V-3

-------
residues), mixing zones should not be projected to result in
significant health risks to average consumers of fish and
shellfish, after considering exposure duration of the affected
aquatic organisms 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 conditions such as the following:  (A) 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.  (B) Mixing zones might be denied where such denial is
used as a device to compensate for uncertainties in the
protectiveness of the water quality criteria or uncertainties in
the assimilative capacity of the waterbody.


5.3  Point of Application of the Criteria

     The point at which the human health criteria or RACs 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 edge of the mixing zone, the criterion would be
     applied at that point.

•    Where State standards allow considerations of spatial
     averaging, the criterion may be met as an average within a
     specified area.

•    Where State standards apply to aggregate risk, the criterion
     for each specific chemical must be reduced such that the sum
     of the exposures does not exceed the maximum level
     permissible under the standards.


5.4  Determining the Wasteload Allocation

     If a State does not allow a mixing zone, then the criterion,
or RAC, is applied at the end-of-pipe and no wasteload allocation
(WLA)  determination is necessary unless a facility can achieve
compliance with water quality standards using a diffuser.  In

                               V-4

-------
some instances, a "complete mix" situation will exist, especially
in an effluent-dominated scenario.  In those situations, the
criterion or RAC is applied after the dilution determination
using the appropriate critical flow condition (as described in
Sections 5.6.1 and 5.6.2).  For the purpose of this document,
WLAs are only necessary or required where instream dilution is
allowed for bioconcentratable pollutants.

     Where dilution is allowed, a WLA should be calculated to
achieve the RAC selected above [29,51].  The human health WLA
should be compared with the acute and chronic aquatic life WLAs
for a given chemical and the most stringent WLA selected for use
to determine the permit limit.


5.4.1  Steady State Dilution Models

     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, where intermittent
nonpoint sources contributions are significant,  and where
additional factors need to be considered.  In some cases, fate
and transport models, where sufficient input data and information
are available, are useful tools for accounting for an array of
variables which may have an impact on the fate of
bioconcentratable pollutants in the environment.

     In simple situations, the WLA is determined from the ambient
criterion and the dilution 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 the State water quality standards requirements which include
reasonable assumptions about the behavior of aquatic organisms
and the target human population.

     In some cases,  it may be necessary to apply the RAC within
this spatial averaging scale mixing zone, if it is reasonable to
assume that (a) the bioconcentrating aquatic organisms have
little mobility, thus spending most of their time within the
mixing zone, and (b)  the target human population consumes a
lifetime exposure to pollutants in fish from the mixing zone.
The lifetime exposure consists of the sum of all fish tissue
concentrations consumed over a 70 year time period.  It includes
infrequent high concentrations as well as more frequent lower
concentrations.
                               V-5

-------
5.4.2  Dynamic Models for Wasteload Allocations

     Steady-state modeling considers only a single condition; the
effluent flow and loading are assumed .to be constant for the
design condition.  The impact of variability in receiving water
flow on the exceedence of the criteria is implicitly included in
the critical design conditions; effluents are considered
constant.  Dynamic modeling techniques explicitly predict the
effects of receiving water and effluent flow and concentration
variability.  The three dynamic modeling techniques recommended
by EPA for wasteload allocations are continuous simulation, Monte
Carlo simulation, and log-normal probability modeling.  These
methods calculate a probability distribution for receiving water
concentrations rather than a single, worst-case concentration
based on critical conditions.

     The dynamic modeling techniques have an additional advantage
over steady-state modeling in that they determine the entire
effluent concentration frequency distribution required to produce
the desired frequency of criteria compliance.  Maximum and
monthly average permit limits can be obtained directly from this
distribution.  For the purposes of this discussion only the Monte
Carlo simulation, dynamic modeling technique will be described.
The other methods use slightly different means to obtain a
similar result.
5.4.3  Monte Carlo Simulation Models

     Monte Carlo simulation combines probabilistic and
deterministic analyses since it uses a fate and transport
mathematical model with statistically described inputs.  The
probability distributions of effluent flow, effluent
concentration, and other model input must be defined using the
appropriate duration for comparison to the criteria.  If 90-day
average receiving water concentrations must be predicted for RAC
comparisons, probability distributions of 90 day model input data
are needed for Monte Carlo simulation.  If 70 year average
concentrations must be predicted for RAC comparisons, the
probability distributions of 70 year average input data are
required.  The computer selects input values from these
distributions using a random generating function.  The fate and
transport model is repetitively run for a large number of
randomly selected input data sets.  The result is a simulated
sequence of receiving water concentrations.  These concentrations
do not follow the temporal sequence 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].
                               V-6

-------
     The approach for calculating the allowable effluent
concentration distribution using Monte Carlo simulation is the
same as that for continuous simulation models.  The advantages of
Monte Carlo simulation are the following:

     •    it can predict the frequency and duration of toxicant
          concentrations in a receiving water;

     •    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; and

     «    it can incorporate the cross-correlation and
          interaction of time-varying flow and pollutant
          discharges if the analysis is developed separately for
          each season and the results are combined.

The primary disadvantage of Monte Carlo simulation is that it
requires the availability of data for effluent variability and
receiving water flows.


5.5  Averaging Periods for WLAs

     The duration of exposure assumed in deriving the reference
ambient concentration (RAC), should affect the selection of the
design conditions and may affect the point of application of a
criterion.  This duration of exposure or averaging period
underlying the derivation of the water quality criterion or RAC
is the time period required to obtain an adverse effect or an
unacceptable risk from marginally exceeding the RAC or criterion.
The duration should have a logical connection with the design
flow.  Often, however, the flow averaging period only
approximates the criterion averaging period, as for example when
the 7Q10 is coupled with the 4-day average aquatic life
criterion.
                               V-7

-------
     The duration of the averaging period for determining the WLA
should be consistent with the assumptions used to derive the
criterion or RAC being applied.  For example, the human health
criteria for carcinogens are based on the assumed risks
associated with a lifetime exposure which can consist of a series
of short term exposures or a continuous exposure over a 70-year
period.

     The human health criteria for non-carcinogens is pollutant
specific; health effects could develop from one short-term
exposure for some compounds.  The criteria are based on various
exposure periods ranging from a few days to many years depending
upon the pollutant.

     Therefore, the averaging period underlying the RAC is 70
years for carcinogens, but may be pollutant specific or unknown
for noncarcinogens.  The duration of exposure assumed in deriving
criteria for non-carcinogens is more complicated due to a wide
variety of endpoints; some developmental (and thus age-specific
and perhaps sex-specific), some lifetime as with carcinogens, and
some, such as organoleptic effects, not age-related at all.


5.6  Dilution Design conditions for Freshwater
     Receiving Waters

     The appropriate dilution design conditions to use in
calculating a WLA depend on the averaging period of the exposure.
For averaging periods greater than one year, as is the case for
carcinogens/ the harmonic mean flow should be used as the
critical design flow.  For averaging periods of one month to one
year/ the 30-day flow with an appropriate recurrence interval
(e.g. five or ten years) should be used; for averaging periods
less than one month/ the 7Q10 as the design flow should be used.


5.6.1  Carcinogens

     In well mixed situations, the receiving water concentration
is determined by the pollutant load, and the combined receiving
water plus effluent flow, such that:

Equation 5.1             C = W/Q

where C is the instream concentration, W is the effluent load,
and Q is the combined effluent and river flow.

     The design flow for calculating average receiving water
concentrations over a several-year period is selected as the flow
that results in the average dilution obtained over that period.
Since the dilution obtained by a given flow is proportional to

                               V-8

-------
the reciprocal of the flow  (the instream concentration  is equal
to the discharge load divided by stream flow) , a simple mean  of
the daily flows does not characterize the average dilution
obtained.  The harmonic mean flow, is the flow at which average
dilution is obtained, and this flow should be used as the design
flow for multiple-year averaging periods as is the case for
carcinogens.

     The recommendation of the mean flow is the best estimator of
the lifetime exposure to carcinogenic pollutants.  The  adverse
impact of carcinogenic pollutants is estimated in terms of
receptors' (human) lifetime intakes.  To be within the  acceptable
level of a lifetime exposure of any carcinogen, such intakes
should not exceed the RAC during "the lifetime" of the  receptor.
A lifetime for carcinogenic pollutants is defined by EPA as 70
years or approximately 365  (days/year) x 70 years.  In  estimating
the lifetime exposure, EPA assumes that people will catch and
consume fish at random times spaced over a year.  Consumption
over 70 years of random fishing can be best estimated by
calculating the exposure of fish to the average concentration in
the receiving water which in turn is best calculated by using the
harmonic mean flow.  Once fish tissue concentrations approach an
equilibrium with the average receiving water concentrations,
variations in the tissue concentrations will reflect variations
in the receiving water, but the averages will not change.

     The exposure to carcinogenic pollutants is numerically
expressed as:

Equation 5.2        CLT  =  ( C, + C2 +  --- +  Cn ) / n

where CLT is the averaged lifetime exposure, C is the yearly
exposure to fish tissue, and n is (365 days/year) x 70  years.

     Based on an assumed constant daily load from a treatment
facility, the fully mixed in-stream concentration will  go up  or
down inversely with the ups and downs of receiving water flows.
Therefore, in-stream concentration is inversely proportional  to,
the streamflow downstream of the discharge:

Equation 5.3        1/QLT =  ( 1/Q1 + 1/Q2 + — +  l/Qn )  / n

     The inverse of Equation 5.3 is the definition of the
harmonic mean:

Equation 5.4        Qhm  = n /  (  1/Q1 + 1/Q2 + --- + l/Qn  )

                           i=n
Equation 5.5:       Qhm  = n/2   (1/Qj)
                               V-9

-------
where n = the number of recorded flows, and Q is the observed
streamflow plus effluent flow on day i.  A five-year minimum
period of record is recommended to assess the average flow.

     The harmonic mean flow represents the long-term average
based flow of a stream minimizing the rainfall-induced spikes of
high flow which can skew the annual arithmetic average flow.
This average base flow is the flow which aquatic life is
subjected to for the majority of the year and is more appropriate
for relating to bioconcentration.  The difference between the
arithmetic and harmonic mean flow is greatest in rivers
characterized by long periods of a low base flow followed by
short periods of high flows.

     With bioaccumulative pollutants, there may be some concern
that use of a long term harmonic mean flow may undercalculate the
actual lifetime exposure if humans consume an annual amount of
fish in a short period (i.e. warm weather months amenable to
recreation) or for pollutants which cannot be depurated or
metabolized quickly.  However, there is a distinct possibility in
some instances that people fish only during the periods when a
river is at a low flow condition.  An example of this is in
western streams where low flows occur during the summer months.
In these instances, use of the harmonic mean flow based on annual
flows can undercalculate the exposure.  The exposure can be
adjusted to reflect this in two ways.  First, the harmonic mean
flow can be recalculated based on the flows from only the months
most likely for fishing.  Actual data is necessary to do this
because the relationships shown below only hold for annual data.
Second, the RAC can be adjusted to reflect the fish consumption
during the fishing months.

     In order to establish a maximum allowable discharge load,
one would set the average receiving water concentration to the
level established by risk-based exposure analysis (the RAC) and
multiply it by the harmonic mean flow.  Thus from this
perspective, the harmonic mean flow serves as a design flow for
performing waste load allocations for carcinogens.

     EPA recommends that the harmonic mean flow be calculated
directly from the historical flow record if possible.  In the
absence of adequate data (5 years or more), the harmonic mean
flow, Qhm,  may be estimated by any of several methods described
in Appendix F, if flows are approximately log-normally
distributed.  One method that assumes a log normal distribution
of flows relates the harmonic mean to the geometric and
arithmetic mean flows:

                          Q  2
                    Qhm=^
                          Qam

                               V-10

-------
where Q   is the geometric mean flow,  and Qam  is the arithmetic
mean flow.  Kurtosis and skewness coefficients of the log
transformed data should be computed before using harmonic mean
flows (which assumes log normality of the distribution).

     Normally, if there are at least 5 years of data to compute
these coefficients, then it would be more appropriate to compute
harmonic mean flow directly using HHDFLOW below.  However, in
some instances there may only be summaries of the statistical
parameters in USGS flow records.  In cases where at least 5 years
of flow data are not available and a 7Q10 has been estimated, EPA
recommends the use of multiplication factors as described below
to estimate the Qhm.

     The two software packages which are available for
computation of harmonic mean flow are:  WQAB FLOW (a description
on how to use this software is in Appendix F) and HHDFLOW.  The
HHDFLOW program  can be used regardless if data are  log-normally
distributed or not.  HHDFLOW has been incorporated into the WQAB
system as a computational tool within the existing DFLOW software
package (see the TSD [32] for more information on these exposure
assessment tools).

     In order to develop some quantitative sense of how a
long-term harmonic mean flow of any stream compares with its
7-day 10-year low flow, EPA's Assessment and Watershed Protection
Division and the Risk Reduction Engineering Laboratory at
Cincinnati, Ohio analyzed flow records of 60 streams.   These are
the same stream flow records which had been analyzed for stream
design flow condition for aquatic life protection criteria as
listed in Book VI. Design Conditions;   Chapter 1,  Stream Design
Flow for Steady-State Modeling [53].  Based on the long-term
harmonic flow and 7-day 10-year low flow estimates for these
sixty streams, the long-term harmonic mean flows of all 60
streams were equal to or greater than 2 times the 7Q10 low flow.
Fifty-four of the streams' harmonic mean flows were equal to or
greater than 2.5 times their 7Q10 low flows.  Finally, 40 of the
60 streams' harmonic mean flows were equal to or greater than 3.5
times the 7Q10.

     Based on the above observations,  permit authorities may
choose a multiplication factor of 3 x 7Q10 to estimate stream
design flow for human health protection for carcinogenic
pollutants, except for regulated streams where HHDFLOW should be
used instead.  This provides an upper bound estimate of the
harmonic mean flow based on the data from the 60 streams
referenced above.  For those States with 7Q2 design flow, the 7Q2
is generally 2 times greater than 7Q10.
                              V-ll

-------
     Alternatively, in situations where long-term daily flow data
do not exist but estimates of common flow statistics are
available from limited data, for example in ungauged discharge
sites, the following equation may be used to estimate harmonic
mean flow [54]:

             Qhm = [1.194 * (Qam)°'473] *  [(7Q10)0'552]

     The 7Q10 low flow and annual average flow (Qam)  for most
parts of the U.S. are readily available.  The 7Q10 and Qam can be
estimated using the USGS computer program, FLOSTAT.


5.6.2  Non-carcinogens

     The choice of return interval represents a level-of-
protection consideration inherent in the risk management decision
to be made by the regulatory authority.  If a short-term duration
of exposure is chosen (i.e., 90 days or less), design flows may
be appropriately based on critical low flow conditions.  In that
the effects from non-carcinogens can be associated with short
term exposures, the Technical Support Document [32] recommends
the use of the design flow 30Q5.  However, in the comparisons of
flows for smaller rivers (i.e., low flow 50 cfs)  the 30Q5 flow
was only 1.1 times that of the 7Q10.  For larger rivers (i.e.,
low flow of 600 cfs) the factor was 1.4 times.  Therefore,
multiplication of these factors by 7Q10 can be used as an
approximation.


5.7  Lake. Marine and Estuarine Receiving Waters

     Estuarine and lake areas in particular are often
characterized by long retention times,  low potential for quick
initial mixing, and concentrated numbers of dischargers.  For
these reasons, pollutants tend to reside within these areas
rather than become quickly washed out.   Impacts from individual
dischargers can overlap.  Therefore, cumulative impacts should be
considered, and a model to incorporate such impacts may need to
be used.  The following discussion summarizes the characteristics
of marine and estuarine discharges and presents a few of the
different types of approaches available for developing WLAs for
lake, marine and estuarine discharges.


5.7.1  Lake. Marine and Estuarine Mixing

     The first stage of mixing that occurs as a wastewater is
discharged to the lake, marine and estuarine environment is
caused by the properties of the discharge itself.  The models
which describe discharge-induced mixing are termed nearfield

                               V-12

-------
models.  Discharge-induced mixing is caused by two influences:
jets and plumes.  Jets are caused by discharge velocity, as the
difference in velocity between the discharge and the ambient
environment tends to entrain ambient water into the discharge and
cause dilution.  Plumes are caused by effluents which are buoyant
with respect to the ambient environment.  Effluents discharged to
the environment are typically less dense than the ambient water
and tend to rise upon discharge; this movement also serves to
entrain ambient water and cause dilution.  Often times, a
discharge will contain both buoyancy and momentum.  This
situation is termed a buoyant jet, and the two types of mixing
work in concert to dilute the effluent.

     As the distance from the outfall increases, mixing caused by
ambient turbulence in the receiving water becomes important.  The
models which describe ambient induced mixing are typically
referred to as farfield models.  The point where ambient-induced
mixing dominates discharge-induced mixing is a function of the
ambient turbulence and the discharge's buoyancy and momentum.
This point is not easily defined.  Ambient mixing in marine and
estuarine systems can be caused by many factors; tidal currents,
Coriolis effects, freshwater inflow and  wind-driven currents are
a few examples.

     Generally, for the development of WLAs for bioconcentratable
contaminants discharged to estuarine and marine receiving waters,
the use of steady-state, nearfield models for mixing zone
applications, is recommended.  The inputs to the nearfield mixing
zone models depend primarily on discharge characteristics and can
be realistically assumed to remain constant over time.  Also, the
response time of the waterbody to changes in inputs is also
relatively fast.  Consequently, steady-state models are
appropriate for use in assessing nearfield mixing zone issues.
Cumulative impacts, however, depend upon ambient conditions and
may require an extension to steady-state modeling called tidal
averaging (see TSD [32] for more information on this subject).

     Water quality models described in this section can be
applied either deterministically or stochastically to assess
compliance under these conditions.  A deterministic model
application predicts a single environmental response to a single
set of model inputs.  Stochastic model application,  such as Monte
Carlo Simulation (Section 5.4.3), predicts the entire
distribution of water quality conditions which can occur in
response to the expected range of environmental and loading
conditions.
                              V-13

-------
5.7.2  Approaches for Marine and Estuarine Nearfield Models

     The following discussion describes the approaches which may
be taken to determine the bioconcentratable pollutant
concentrations at the edge of the mixing zone.  Four approaches
are addressed:

               1.   Desktop calculations
               2.   UPLUME model
               3.   UMERGE model
               4.   COREMIX model
1.  Desktop Calculations

     The most direct recommended approach for predicting
pollutant concentrations at the edge of a mixing zone consists of
an equation which can be solved using a hand calculator.  The
desktop equation recommended for use considers dilution caused by
discharge momentum only.  It may be appropriately applied to
cases where the effluent is neutrally buoyant in the ambient
water, such as freshwater outfalls.  Effluents discharged into
estuarine and coastal waters tend to be positively buoyant, and
are rarely completely neutrally buoyant.  The equation, taken
from the Technical Support Document for Water Quality-based
Toxics Control [32] is:

                                          x
                                 S = 0.3  -
                                          d

     where:    S = average dilution (dimensionless)
               x = distance from outlet to edge of mixing zone
               d = diameter of outlet

     The equation provides an estimate of mixing zone dilution
only, since it assumes that ambient-induced mixing is zero.
Ambient-induced mixing begins to dominate jet-induced mixing as
the distance from the outfall increases.  Consequently,
predictions using this equation are most appropriate for
predicting dilution close to the outfall.  Predictions become
increasingly conservative as distance from the outlet to the edge
of the mixing zone increases, and ambient-induced mixing becomes
more important.

2.  UPLUME Model

     EPA supports a series of nearfield mixing models designed to
describe the dilution of effluents discharged to marine
environments which are described in detail in the TSD [32].
UPLUME, considers a buoyant plume issuing at an arbitrary angle

                              V-14

-------
into a stagnant environment.  UPLUME is designed to assess a
single port discharge, although the model will predict dilution
for multiple port discharges up to the point where adjacent
plumes merge.  UPLUME is recommended for use for single port
discharges only.  The model UMERGE discussed subsequently is
recommended for multi-port diffusers.

     UPLUME will predict average dilution in the effluent plume
at increasing distances from the discharge.  These dilution
calculations can then be used to determine allowable effluent
concentrations.  Calculations continue until the plume stops
rising and its trajectory becomes horizontal.  This endpoint will
be caused by either: 1) the density of the plume becoming equal
to the density of the ambient environment, or 2) the plume
reaching the water surface.  UPLUME can be applied to either
stratified or non-stratified environments.

3.  UMERGE Model

     UMERGE extends upon UPLUME's capabilities in two ways.
First, UMERGE accounts for merging among adjacent plumes and is
therefore appropriately applied to multi-port diffusers.  Second,
UMERGE allows specification of ambient current velocities at
various depths in the water column.  Similar to UPLUME, UMERGE
will provide calculations of average dilution within the plume at
various depths until the plume vertical velocity becomes zero.

4.  COREMIX Model

     EPA has supported development of an expert system for the
analysis of submerged discharges, entitled CORMIX (EPA, 1989).
The user supplies CORMIX with information about the discharge and
ambient environment.  CORMIX then provides information regarding
the dilution and geometric configuration of the plume in the
ambient water.  CORMIX also explicitly considers regulatory
mixing zone dimensions and requirements.  CORMIX contains two key
elements.  The first is a rigorous analytical scheme that
classifies any given discharge/environment situation into one of
several categories with distinct hydrodynamic features.  The
second element is a collection of predictive models appropriate
for different classification schemes.  CORMIX automatically
selects and applies the model framework appropriate for the
discharge/environmental conditions supplied by the user,
providing dilution and mixing zone compliance information.

     In summary, for the four approaches for determining WLAs for
marine and estuarine discharges,  the desktop method is used to
assess dilution of neutrally buoyant jet discharges, UPLUME for
single-port buoyant jets, and UMERGE for multi-port buoyant jets.
The COREMIX model can be used for all applications.
                               V-15

-------
5.8  Exposure from Contaminants in Sediments

     Sediments may act as sinks for bioconcentratable pollutants
discharged from effluents or from other sources.  Bioaccumulative
organic pollutants will adsorb to particulates in an effluent or
receiving water, the particulates may later be deposited as
sediments.  Dissolved organic compounds may also partition from
the water column to the particulates forming a sediment deposit.
High log P chemicals (above log P of 6.5)  are known to have a
greater tendency to concentrate in sediments.  Sediments also act
as sources of bioconcentratable and bioaccumulative organic
chemicals to aquatic life.  The chemicals adsorbed to sediments
can partition from the sediments back into the water column and
the sediments themselves may be ingested by benthic organisms and
fish species, both of which may contribute food chain effects and
bioaccumulation.

     The assessment of sediment samples for bioconcentratable
contaminants, described in Chapter 3, can determine the presence,
identity, and concentrations of these pollutants.  Since
sediments can accumulate these chemicals over relatively long
periods of time the bioaccumulative chemicals may be present in
greater concentrations in sediment than in a given effluent
sample.  In some cases, this may facilitate detection of
contaminants which are present in an effluent or other sources at
very low concentrations or which are only released periodically.
For assessments of point source discharges, the results of the
sediment assessment may be used to specify chemical specific
tests of effluent samples from a given discharger in order to
confirm the presence and concentrations of the bioconcentratable
chemicals in the effluent.

     The relationship between bioconcentratable chemicals and
sediments is, in many ways, parallel to that between these
chemicals and biota.  That is, sediments provide a chemical
environment that is energetically favorable for hydrophobia
chemicals relative to water.  It is for this reason that
sediments are a potentially a major reservoir for hydrophobic
chemicals.  The organic carbon fraction of the sediment acts like
the lipid fraction of the biota:  it is the phase into which the
chemicals partition.  It is commonly observed that the
concentration of hydrophobic chemicals in sediment dwelling
organisms are in the same order of magnitude as the surrounding
sediment.  This is the result of the similar concentrations of
lipid in organisms and organic carbon in sediments.  Hence the
sediments provide another environmental compartment in which
bioconcentratable chemicals can be found.
                              V-16

-------
     When an effluent containing bioconcentratable chemicals is
discharged into a receiving water, a fraction of the chemical is
adsorbed by the organic carbon fraction of the suspended
particles.  The fate of the sorbed fraction of the chemical is
determined by the fate of these particles.  During quiescent
periods a portion of these particles settle thereby transporting
the chemical to the bottom sediment and removing it from the
water column.  However, during more turbulent periods, periods of
high flows for example, particles from the sediment can be
scoured into the overlying water, thereby providing a source of
bioconcentratable chemicals.  The chemical that remains in the
sediment can also affect the overlying water concentration via
another mechanism.  Chemicals sorbed to sediment particles can
desorb into the interstitial water of the sediment ( surface
sediments are typically 90% water and only 10% particles by
weight),  Depending on the concentration in the overlying water,
chemical can either diffuse into or out of the sediment
interstitial water.

     As a consequence of these mechanisms, the sediments of a
body of water receiving a discharge of bioconcentratable chemical
can accumulate a large mass of the chemical.  However/ the
quantity of chemical stored is not equivalent to the removal rate
of the chemical.  If the net burial rate of sediments is large,
then a fraction of the sorbed chemical can be lost to the deep
sediments by burial.  However, in most flowing waters, the
sediments act as a temporary storage, accumulating chemicals
during quiescent periods and releasing chemicals during turbulent
periods.  Hence, when computing the exposure concentration for
fish, it is imprudent to assume that a significant rate of loss
of chemical occurs via sedimentation and burial.  A detailed mass
balance analysis of the suspended solids and the chemical of
concern is required to substantiate this loss.  Any net release
of a chemical from the sediments would need to be accounted for
in the LA portion of the TMDL.  In the absence of such an
analysis/ EPA recommends that the TMDL consider no significant
loss of chemical to the sediments.  That is, the chemical is
assumed to behave as a conservative substance in the water column
and the receiving water concentration is computed using only the
effluent chemical concentration and the dilution that occurs,  as
was described above.
                              V-17

-------
                            CHAPTER 6

    Permit Limits for Control of Bioconcentratable Pollutants
     Once the pollutants of concern, the Reference Ambient
Concentration or State water quality criteria for those
pollutants, and the wasteload allocation have been determined, as
were described in Chapters 3, 4 and 5 respectively, the next step
is to characterize the effluent for human health effects.  The
purpose of effluent characterization is to determine whether or
not the discharge of the identified bioconcentratable pollutants
causes, has the reasonable potential to cause, or contributes to
an adverse impact to human health.

     The effects of the discharge of bioconcentratable pollutants
can be characterized by utilizing the measured effluent
concentrations of those pollutants of concern to determine the
expected exposure concentrations in the receiving water and then
by comparing this exposure concentration to the criteria or RAC
for that pollutant.  The receiving water concentration (RWC) is
the calculated exposure concentration of a pollutant in the
receiving water at the critical flow condition and after mixing.
An appropriate effluent flow based on the exposure period should
be used in this calculation.  Chapter 5 provides a detailed
discussion of the appropriate critical flow conditions and mixing
zone considerations to be used in developing wasteload
allocations (WLAs).  The same basic components described in that
discussion for wasteload allocations are used for the purposes of
calculating the RWC for effluent characterization.  In brief, the
recommendations for RWCs are to use the harmonic mean flow for
carcinogenic pollutants and the 30Q5 for non-carcinogenic
pollutants.


6.1  Basis for Effluent Characterization for Human Health

     For individual pollutants, the potential for adverse impact
in the receiving water is minimized where the RWC is less than
the RAC, and becomes maximized where the RWC exceeds the RAC (the
term, RAC is used throughout this section interchangeably with
EPA human health criteria).   Therefore, to prevent adverse
impacts, the RWC of the pollutant (based on allowable dilution
for the discharge and pollutant)  must be less than the applicable
criterion.  Protection of human health will be achieved where the
RWC is less than the EPA criterion or the RAC:

     RWC < RAC  (RAC is equivalent to human health criterion)


                              VI-1

-------
Where no mixing zone is allowed, the RAC would be applied at the
end-of-pipe:

     Effluent Concentration < RAC

      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 effluent characterization and
wasteload allocation development albeit from different
perspectives.  In the case of effluent characterization, the
objective is to project receiving water concentrations based upon
existing effluent quality to determine whether or not an
excursion above ambient criteria occurs, or has the reasonable
potential to occur.  In developing wasteload allocations on the
other hand, the objective is to fix the RWC at the desired
criteria level and determine an allowable effluent loading which
will not cause excursions above the criteria.

     Recommendations for projecting the RWC are described within
this chapter.  Chapter 5, Exposure Assessment and Wasteload
Allocation, provides recommendations for determining allowable
effluent loadings to achieve established ambient criteria and for
calculating wasteload allocations for establishing permit limits.
The procedures described within Chapter 5 can also be used to
calculate the dilution for the purpose of effluent
characterization.  The Permit Limits section of this chapter
describes the actual calculation of permit limits after effluent
characterization and loadings and wasteload allocations are
complete.


6.1.1  Recommended Approach for Effluent Characterization

     The following four step approach for effluent
characterization is adapted from the recommendations provided in
the Technical Support Document for Water Quality-based Toxics
Control [32].  The objective of this approach is to determine
whether or not permit limits need to be developed based upon the
results of the effluent bioconcentration evaluation.

Step l - Identify the Pollutants of Concern

     The pollutants which are of concern are identified in report
no. 4 of the analytical procedures described in Chapter 2.

Step 2 - Determine the Basis for Establishing RACs for the
         Pollutants of Concern

     If a State has a numeric water quality criterion for
bioconcentration for the pollutant of concern, this should be

                              VI-2

-------
used to characterize the effluent and develop permit limits (40
CFR 122.44(d)(1)(iii).  If a State has not adopted a numeric
water quality criterion for the pollutant of concern, then one of
three options for using the narrative criterion may be used (40
CFR 122.44(d)(1)(vi)) to determine whether a discharge causes,
has the reasonable potential to cause, or contributes to an
excursion above a narrative standard due to 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 may be applied in
characterizing effluents in order to determine whether limits are
necessary.

     Option A allows the regulatory authority to establish
     limitations using a "calculated numeric water quality
     criterion" which the regulatory authority demonstrates
     will attain and maintain applicable water quality
     standards.  This option allows the regulatory authority
     to use any criterion that protects aquatic life and
     human health.  This option would also allow the use of
     site specific factors, including local human
     consumption rates of aquatic foods, the State's
     determination of an appropriate risk level, and any
     current data that may be available to calculate an RAC.

     Option B allows the regulatory authority to establish
     effluent limits using EPA's Water Quality Criteria
     guidance documents, if EPA has published a criteria
     document for the pollutant (supplemented, where
     necessary, by other relevant information, such as
     recent information from IRIS).

     Option C may be used to develop limits based on an
     indicator parameter under limited circumstances.   An
     example of an indicator parameter is total toxic
     organics (TTO); effluent limits on TTO are useful where
     an effluent contains organic compounds.   However, use
     of this option must be justified to show that controls
     on one pollutant control one or more other pollutants.
     Where such data are available, this option may be used
     provided several conditions are met.  (See 40 CFR
     122.44(d)(1)(vi)(C)).   Using this option presents
     complications.   When trying to determine whether or not
     a pollutant of concern has a reasonable potential to
     cause an excursion above the narrative standard,
     development of specific information on the pollutant of
     concern and comparison with an RAC (using option A or
     option B)  will normally be necessary.
                              VI-3

-------
step 3 - Dilution Determination

     Once a basis for comparison has been established, the margin
between projected RWC and the RAC after any allowable mixing may
be compared.  If the Water Qualtiy Standards so allow the water
quality analyst will apply the appropriate mixing zone
considerations for human health, which are distinctive from those
generally used for aquatic life (see Chapter 4 for a discussion
of mixing zones).  These comparisons will lead to one of the
three outcomes discussed below.

     There are two options for comparing the RWC to RAC, both
based on the margin between measured chemical concentration and
the RAC.  The first level is to use simple fate models based on a
dilution analysis and comparison with the RAC.  The second level
of analysis is to use more complex fate models, including dynamic
models.  These may be applied to lakes, rivers, estuaries, and
coastal systems using a desktop calculator or microcomputer [32].

Step 4 - Decision Criteria for Permit Limit Development

     After this dilution analysis has been performed, the
projected RWC is compared to the RAC for each pollutant of
concern.  This step should be performed using the highest
concentration in each data set for a pollutant.  The three
possible outcomes discussed above in step 3 are:

     1.  Excursion above the RAC.

     2.  Potential for an excursion above the RAC.

     3.  No potential excursion above the RAC.

If these evaluations project excursions or the reasonable
potential to cause or contribute to excursions above the RAC,
then a permit limit is required (40 CFR 122.44(d)(1)(ii)).  The
statistical approach described in Section 6.1.2 or an analogous
approach developed by a regulatory authority can be used to
determine the reasonable potential.  Effluents that are shown not
to cause or have a reasonable potential to cause or contribute to
an excursion above an RAC should be re-evaluated at permit
reissuance.

     Where these test results do not show a reasonable potential
but indicate a basis for concern after consideration of other
facility specific factors discussed Chapters 1 and 3, or if there
were inadequate information to make a decision, the permit should
contain chemical testing requirements and a reopener clause.
This clause would require reopening of the permit and
establishment of a limit based upon any test results which show
chemical concentrations at levels which cause or have a

                               Vl-4

-------
reasonable potential to cause or contribute to an excursion above
the RAG.
6.1.2  Determination of "Reasonable Potential" for Excursions
       Above RACs

     The procedure below or a similar method required by a
permitting authority, can be used to determine if a reasonable
potential to exceed an RAC exists using the results of the
effluent bioconcentration evaluation.  A permittee does not have
reasonable potential to exceed a Water Quality Standard or RAC if
it can be demonstrated with high confidence that the mean of the
log-normally distributed effluent concentrations is below the RAC
at the specified flow condition.  If the permittee cannot make
this demonstration, then there is a reasonable potential to
exceed the receiving water standard, and the discharge must be
limited.

     This recommended procedure consists of the following 5
steps:

Step 1         Determine the number of total observations ("n")
               for a particular set of effluent data
               (concentrations or TUs), and determine the highest
               value from that data set.

Step 2:        Determine the coefficient of variation for the
               data set (actual or estimated) and probability
               basis (95% or 99%).

Step 3         Determine the appropriate ratio from Table 6-1 or
               6-2.

Step 4         Multiply the highest value from a data set by the
               value from Table 6-1 or 6-2.  Use this value with
               the appropriate dilution to project a maximum RWC.

Step 5         Compare the projected maximum RWC to the RAC.
               Reasonable potential occurs when the projected RWC
               is greater than the RAC.

     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 mean
concentration for the effluent.  The estimated mean concentration
is calculated from the log-normal 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

                               VI-5

-------
for an effluent limit.  This procedure is discussed in more
detail in the Technical Support Document with the difference that
the mean effluent concentration is estimated rather than an upper
bound concentration [32].


6.2  The Wasteload Allocation as Basis for Permit Limit
     Derivation

     A WLA value must be applied in a regulatory context by
translation into daily maximum and monthly average permit limits.
Compliance monitoring associated with permit limitations allows
the regulatory authority to determine if the permit limitations
are violated and if the wasteload allocation is being achieved.

     WLAs for the protection of human health are typically based
upon a single criterion, and, in contrast to WLAs for protection
of aquatic organisms,  do not specify two levels of protection for
both acute and chronic effects.  The human health WLAs provide a
level of protection which in a broad sense can be considered a
long term, or chronic, exposure.  This is appropriate for the
protection from bioconcentration effects on human health exposure
durations of up to 70 years.  As part of the water quality-based
approach to the control of toxics, the human health WLA should be
compared with the acute and chronic aquatic life WLAs for a given
chemical, and the most stringent WLA must be selected for use to
determine the permit limit.  This comparison should be made for
each chemical specific water quality-based permit limit which is
developed.

     Steady-state analyses or models used to develop these
wasteload allocations assume that the effluent is constant, and
therefore, that the WLA value will never be exceeded.  This
presents a problem in deriving permit limits because permit
limits must address effluent variability.  The proper application
of a steady-state WLA depends on the exposure period for the
criteria or RAG for the pollutant which is to be limited.  The
impact associated with bioconcentratable contaminants is time
dependent as reflected in the averaging periods for the RACs for
carcinogens and non-carcinogens.   Because of these factors the
WLA outputs have specific implications for the subsequent permit
limit development process.


6.3  Procedures for Developing Limits

     Since compliance with permit limitations is by regulation
determined on a daily and monthly basis, it is necessary to set
permit limitations expressed in these contexts that meet a given
                              VI-6

-------
                          Table 6-1
Ratio of Maximum Sample to Mean for  the 95   Percent Confidence Interval

           Effluent Coefficient of Variation
N
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0.1
1.2
1.1
1.0
1 0
1.0
1.0
1.0
1.0
0.9
0.9
0.9
0.9
0.9
0.9
0.9
0.9
0.9
0.9
0.9
0.9
0.2
1.4
1.2
1.1
1.0
1.0
0.9
0.9
0.9
0.9
0.9
0.9
0.9
0.8
0.8
0.8
0.8
0.8
0.8
0.8
0.8
flzl
1.6
1.2
1.1
1.0
1.0
0.9
0.9
0.9
0.8
0.8
0.8
0.8
0.8
0.8
0.8
0.8
0.7
0.7
0.7
0.7
0.4
1.9
1.3
1.2
1.0
1.0
0.9
0.9
0.8
0.8
0.8
0.8
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.5
2.2
1.4
1.2
1.0
0.9
0.9
0.8
0.8
0.8
0.7
0.7
0.7
0.7
0.7
0.7
0.6
0.6
0.6
0.6
0.6
0.6
2.5
1.5
1.2
1.0
0.9
0.9
0.8
0.8
0.7
0.7
0.7
0.7
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.5
0.7
2.8
1.6
1.3
1.0
0.9
0.8
0.8
0.7
0.7
0.7
0.6
0.6
0.6
0.6
0.6
0.5
0.5
0.5
0.5
0.5
0.8
3.2
1.7
1.3
1.0
0.9
0.8
0.8
0.7
0.7
0.6
0.6
0.6
0.6
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.9
3.6
1.8
1.3
1.1
0.9
0.8
0.7
0.7
0.6
0.6
0.6
0.6
0.5
0.5
0.5
0.5
0.5
0.5
0.4
0.4
1.0
3.9
1.9
1.4
1.1
0.9
0.8
0.7
0.7
0.6
0.6
0.6
0.5
0.5
0.5
0.5
0.5
0.4
0.4
0.4
0.4
1.1
4.3
2.0
1.4
1.1
0.9
0.8
0.7
0.6
0.6
0.6
0.5
0.5
0.5
0.5
0.4
0.4
0.4
0.4
0.4
0.4
1.2
4.7
2.0
1.4
1.1
0.9
0.8
0.7
0.6
0.6
0.5
0.5
0.5
0.5
0.4
0.4
0.4
0.4
0.4
0.4
0.4
1.3
5.1
2.1
1.4
1.1
0.9
0.8
0.7
0.6
0.6
0.5
0.5
0.5
0.4
0.4
0.4
0.4
0.4
0.4
0.4
0.3
1.4
5.5
2.2
1.5
1.1
0.9
0.8
0.7
0.6
0.6
0.5
0.5
0.5
0.4
0.4
0.4
0.4
0.4
0.3
0.3
0.3
1.5
6.0
2.3
1.5
1.1
0.9
0.7
0.7
0.6
0.5
0.5
0.5
0.4
0.4
0.4
0.4
0.4
0.3
0.3
0.3
0.3
L*
6.4
2.4
1.5
1.1
0.9
0.7
0.6
0.6
0.5
0.5
0.4
0.4
0.4
0.4
0.4
0.3
0.3
0.3
0.3
0.3
1.7
6.8
2.4
1.5
1.1
0.9
0.7
0.6
0.6
0.5
0.5
0.4
0.4
0.4
0.4
0.3
0.3
0.3
0.3
0.3
0.3
1.8
7.2
2.5
1.6
1.1
0.9
0.7
0.6
0.6
0.5
0.5
0.4
0.4
0.4
0.4
0.3
0.3
0.3
0.3
0.3
0.3
1.9
7.6
2.6
1.6
1.1
0.9
0.7
0.6
0.5
0.5
0.4
0.4
0.4
0.4
0.3
0.3
0.3
0.3
0.3
0.3
0.3
2^0
8.1
2.6
1.6
1 1
0.9
0.7
0.6
0.5
0.5
0.4
0.4
0.4
0.4
0.3
0.3
0.3
0.3
0.3
0.3
0.3
                          Table 6-2
Ratio of Maximum Sample  to Mean for the 99   Percent Confidence  Interval

           Effluent Coefficient of Variation
JL !
1
2
3
4
5
6
-r
8
o
10
11
12
13 (
14 C
15 (
16 (
17 C
18 (
19 C
20 C
).1 0
.3 T
.1 1
.1 1
.0 1
.0 1
.0 1
.0 1
.0 1
.0 1
.0 0
.0 0
.0 0
).9 0
J.9 0
).9 0
).9 0
).9 0
).9 0
1.9 0
.9 0
^2
.6
.3
.2
.1
.1
.0
.0
.0
.0
.9
.9
.9
.9
.9
.9
.9
.9
.9
.9
.8
0.3
2.0
1.5
1.3
1.2
1.1
1.0
1.0
1.0
0.9
0.9
0.9
0.9
0.9
0.8
0.8
0.8
0.8
0.8
0.8
0.8
0.4
2.5
1.6
1.4
1.2
1.1
1.0
1.0
0.9
0.9
0.9
0.9
0.8
0.8
0.8
0.8
0.8
0.8
0.7
0.7
0.7
0.5
3.0
1.8
1.5
1.3
1.1
1.0
1.0
0.9
0.9
0.9
0.8
0.8
0.8
0.8
0.7
0.7
0.7
0.7
0.7
0.7
9^6
3.6
2.0
1.5
1.3
1.2
1.1
1.0
0.9
0.9
0.8
0.8
0.8
0.7
0.7
0.7
0.7
0.7
0.7
0.6
0.6
0.7
4.3
2.2
1.6
1.4
1.2
1.1
1.0
0.9
0.9
0.8
0.8
0.7
0.7
0.7
0.7
0.7
0.6
0.6
0.6
0.6
0.8
5.1
2.5
1.7
1.4
1.2
1.1
1.0
0.9
0.8
0.8
0.8
0.7
0.7
0.7
0.6
0.6
0.6
0.6
0.6
0.6
0.9
6.0
2.7
1.8
1.4
1.2
1.1
1.0
0.9
0.8
0.8
0.7
0.7
0.7
0.6
0.6
0.6
0.6
0.6
0.5
0.5
1.0
6.9
2.9
1.9
1.5
1.2
1.1
1.0
0.9
0.8
0.8
0.7
0.7
0.6
0.6
0.6
0.6
0.6
0.5
0.5
0.5
LI
7.9
3.1
2.0
1.5
1.3
1.1
1.0
0.9
0.8
0.7
0.7
0.7
0.6
0.6
0.6
0.5
0.5
0.5
0.5
0.5
1.2
9.0
3.4
2.1
1.6
1.3
1.1
1.0
0.9
0.8
0.7
0.7
0.6
0.6
0.6
0.6
0.5
0.5
0.5
0.5
0.5
1.3
10.1
3.6
2.2
1.6
1.3
1.1
1.0
0.9
0.8
0.7
0.7
0.6
0.6
0.6
0.5
0.5
0.5
0.5
0.5
0.4
1.4
11.3
3.8
2.3
1.6
1.3
1.1
1.0
0.8
0.8
0.7
0.7
0.6
0.6
0.5
0.5
0.5
0.5
0.5
0.4
0.4
LI
12.5
4.0
2.4
1.7
1.3
1.1
1.0
0.8
0.8
0.7
0.6
0.6
0.6
0.5
0.5
0.5
0.5
0.4
0.4
0.4
L6
13.7
4.2
2.4
1.7
1.3
1.1
1.0
0.8
0.8
0.7
0.6
0.6
0.6
0.5
0.5
0.5
0.4
0.4
0.4
0.4
LI
15.0
4.5
2.5
1.7
1.4
1.1
0.9
0.8
0.7
0.7
0.6
0.6
0.5
0.5
0.5
0.5
0.4
0.4
0.4
0.4
JL8
16.4
4.7
2.6
1.8
1.4
1.1
0.9
0.8
0.7
0.7
0.6
0.6
0.5
0.5
0.5
0.4
0.4
0.4
0.4
0.4
1.9
17.7
4.9
2.6
1.8
1.4
1.1
0.9
0.8
0.7
0.7
0.6
0.6
0.5
0.5
0.5
0.4
0.4
0.4
0.4
0.4
LO
19.1
5.1
2.7
1.8
1.4
1.1
0.9
0.8
0.7
0.7
0.6
0.6
0.5
0.5
0.4
0.4
0.4
0.4
0.4
0.4
                           VI-7

-------
WLA every month.  The statistical procedures for permit limit
derivation in the TSD are designed to accomplish this for aquatic
life protection where the use of shorter term averaging periods
is consistent with two number aquatic life criteria [32].

     However, if the TSD procedures were directly used for
setting permit limits on bioconcentratable pollutants, both
maximum daily and average monthly permit limits could exceed the
wasteload allocation necessary to meet the criterion [32].  These
permit limits would assure that the long term average effluent
discharge would comply with the human health derived WLA only if
the assessment of the effluent variability was precise.  With
bioconcentratable pollutants where exposure duration ranges up to
70 years, EPA believes that effluent variability cannot be
reliably estimated from existing data for exposure periods of
over 30 days.  If the effluent variability was over-estimated
when establishing the permit limits, then a facility could be
discharging in compliance with the permit limits but would be
exceeding the wasteload allocation for human health protection.
This approach is clearly unacceptable.

     This problem does not arise when using the TSD statistical
procedure for setting permit limits for protecting against
aquatic toxicity [32].  In this case, the monthly average and
daily maximum permit limits are more closely related to the four
day average and one hour maximum used as exposure periods for the
criteria.  Any imprecision in assessing effluent variability
would therefore not have as great an effect on the permit limits.

     EPA recommends that for setting water quality-based permit
limitations for human health protection the average monthly limit
must be set equal to the wasteload allocation, and the maximum
daily permit limit must be calculated based on effluent
variability and the number of samples per month (see Table 6.3).
This approach ensures that the water quality standards will be
met over the long term regardless of the uncertainties in
estimating effluent variability and provides a defensible method
for calculating a maximum daily permit limit.


6.4  Water Quality-based Permit Limit Derivation for Human Health
    Protection

     The recommended approach for setting water quality-based
limitations for human health protection with statistical
procedures is as follows:

•    Set the monthly average limit equal to the WLA.
                              Vl-8

-------
•    Calculate the daily maximum limit based on effluent
     variability and the number of samples per month using the
     multipliers provided in Table 6.3.

This approach ensures that the State numerical water quality
standard (or RAG in the absence of a numerical standard) will be
met over the long term and provides a defensible method for
calculating a maximum daily permit limit.  Appendix G contains
example permit language for requiring the effluent
bioconcentration assessment and for the incorporation of permit
limits for bioconcentratable pollutants.


6.5  Detection Limits for Compliance Monitoring

     A commonly encountered problem is the expression of
calculated limits for specific chemicals where the concentration
of the limit is below the analytical detection level for the
pollutant of concern.  This is particularly true for pollutants
which are toxic in extremely low concentrations (e.g., dioxin).
The recommended approach for these situations is to include, in
Part I of the permit, the appropriate permit limitation derived
from the water quality model and the wasteload allocation for the
parameter of concern, regardless of the proximity of the limit to
the analytical detection level.

     However, the limit should also contain an accompanying
notation indicating the specific analytical method which should
be used for purposes of compliance monitoring.  The note should
indicate that any sample which is analyzed in accordance with the
specified method and found to be below the detection level will
be deemed to be in compliance with the permit limit, unless other
monitoring information (as discussed below) indicates a problem.

     The detection level for the analytical method cited in the
permit should be clearly defined and quantified.  For most KPDES
permitting situations, EPA recommends that the detection level be
defined in the permit as the minimum level (i.e.,  the level at
which the entire analytical system gives recognizable mass
spectra and acceptable calibration points).  The minimum level is
developed based on inter-laboratory analyses of the analyte in
the matrix of concern (i.e., wastewater effluents).  The minimum
level should not be confused with the method detection level,
which is based on a single laboratory analysis of the analyte in
a given matrix, usually reagent water.

     Where water quality-based limitations below analytical
detection levels are placed in permits, it is recommended that
permit special conditions also be included in the permit to help
ensure that the limitations are being met and that excursion
above water quality standards are not occurring.  Examples of

                               VI-9

-------
such special conditions include:  fish tissue collection and
analysis; limitations and/or monitoring requirements on internal
waste streams; and limitations and/or monitoring for surrogate
parameters.

     To summarize, where the final calculated limitation will be
below the current level of detection the permit should contain:

     a)   The calculated water guality-based permit limitation
          for the pollutant.

     b)   A statement in the permit that the minimum level is the
          threshold for compliance/non-compliance determinations.

     c)   Requirements for additional monitoring.


6.6  Compliance Monitoring

     Monitoring to determine the effectiveness of and compliance
with permit limits for bioconcentratable pollutants is essential.
Once the bioconcentratable chemical is identified and limited, a
chemical-specific monitoring frequency of once a week is
recommended.  Monthly monitoring is the recommended absolute
minimum.  These frequencies are recommended because of the need
to precisely measure the discharge to assure that short intense
discharges do not lead to long term ambient problems.  Since the
bioconcentratable pollutants are very persistent in the
environment, any discharged mass will tend to reside for long
periods.  The permit writer will need to determine a
representative monitoring program which assures that spike
loadings are addressed.  If monitoring costs are of a concern due
to the size of the facility, the permit writer may consider
compositing over multiple days as a way to balance the need for
more precise monitoring against the costs.

     The effluent limit should also contain an accompanying
notation indicating the specific analytical method which should
be used for purposes of compliance monitoring.  This is
particularly important for those pollutants for which there are
no promulgated analytical methods under 40 CFR 136.  In these
instances, the permit writer may need to consult with EPA
chemists to specify an appropriate method that provides
sufficient resolution at the calculated permit limit.

     Additional annual effluent analysis according to the
procedures in Appendix B is recommended for selected dischargers
due to the potentially hazardous nature and persistence of
bioconcentratable compounds.  Where the contaminants in a
discharge can be expected to change due to process changes, etc.,
repeating the effluent analysis is also recommended.

                              VI-10

-------
                                                 Table 6.3

                              Multipliers for Calculating Maximum Daily Permit
                                 Limits  from Average Monthly Permit Limits
To obtain the maximum daily permit limit for  a  bioconcentratable pollutant, multiply the average monthly
permit limit (the uasteload allocation)  by the  appropriate value in the following table.

Each value in the table is the ratio of  the maximum daily permit limit, MDL, to the average monthly permit
limit, AML,  as calculated by the following relationship derived from step 4 of the statistically-based
permit limit calculation procedure in the Technical Support Document for Water Quality-based Toxics Control
[321.
MDL       exp [zmff  - 0.5<7!3

AML
                                           exp [zaan -  0.5an']
where a '  = In (CWn + 1)
      «7P = In (CV  + 1)
      CV = the coefficient of  variation  of  the effluent concentration
      n = the number of samples per month
      z   =  the percentile exceedance probability for the maximum daily limit
      zfl = the percentile exceedance probability  for  the average monthly limit



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 daily ant
Maximum = 99th percenti le
Average = 95th percent ile
n=1 n=2 n=4 n=8 n=30
1.07 1.12 1.16 1.18 1.22
1.14 1.25 1.33 1.39 1.46
1.22 .37 1.50 1.60 1.74
1.30 .50 1.67 1.82 2.02
1.38 .62 1.84 2.04 2.32
1.46 .73 2.01 2.25 2.62
1.54 .84 2.16 2.45 2.91
1.61 1.94 2.29 2.64 3.19
1.69 2.03 2.41 2.81 3.45
1.76 2.11 2.52 2.96 3.70
1.83 2.18 2.62 3.09 3.93
1.90 2.25 2.70 3.20 4.13
1.97 2.31 2.77 3.30 4.31
2.03 2.37 2.83 3.39 4.47
2.09 2.42 2.89 3.46 4.62
2.15 2.47 2.93 3.52 4.74
2.21 2.52 2.98 3.57 4.85
2.27 2.56 3.01 3.61 4.94
2.32 2.60 3.05 3.65 5.02
2.37 2.64 3.07 3.67 5.09
1 average monthly permit limits
Maximum = 99th percenti le
Average = 99th percenti le
n=1 n=2 n=4 n=8 n>=30
1.00 1.07 1.12 1.16 1.20
1.00 1.13 1.24 1.32 1.43
1.00 1.19 1.36 1.49 1.67
1.00 1.24 1.46 1.66 1.92
1.00 1.28 1.56 1.81 2.18
1.00 1.31 1.64 1.95 2.43
1.00 1.34 1.71 2.08 2.67
1.00 1.35 1.76 2.19 2.89
1.00 1.36 1.80 2.27 3.09
1.00 1.37 1.83 2.34 3.27
1.00 1.37 1.84 2.39 3.43
1.00 1.36 1.85 2.43 3.56
1.00 1.36 1.85 2.45 3.68
1.00 1.35 1.84 2.46 3.77
1.00 1.34 1.83 2.46 3.84
1.00 1.33 1.82 2.46 3.90
1.00 1.32 1.80 2.45 3.94
1.00 1.31 1.78 2.43 3.97
1.00 1.30 1.76 2.41 3.99
1.00 1.29 1.74 2.38 4.00
                                                 VI-11

-------
6.7  Tissue Residue Monitoring

     Discharger monitoring may also include periodic
(e.g., annual for those dischargers with chemicals on the CHC
List in their effluent) analysis of aquatic organisms exposed to
the effluent in the receiving water.  This is particularly
important for those pollutants where the calculated permit limit
is below the analytical minimum level.  Tissue residue monitoring
should also be repeated periodically to double check receiving
waters where facilities have previously conducted the assessment
procedures described in this document.  The regulatory authority
may perform this periodic assessment or require the permittee to
repeat the tissue residue analysis described in Appendix A.
Where only the specific chemicals identified by the effluent
bioconcentration option are of concern, chemical specific
analyses of the tissue for those pollutants may be required in
the permit.  Tissue concentrations found in the exposed fish
should then be compared with RTCs derived from the equations in
Chapter 4.  Where fish tissue concentrations exceed an RTC,
analysis of the effluent for concentrations exceeding the RAG
should be made to evaluate the need for revised limits in a
modified or reissued permit.


6.8  Data Generation Mechanisms

     Two mechanisms for data generation are available to the
regulatory authority:  1) a Section 308 letter, or 2) the Special
Conditions section of the permit.  With either option, data
generation may be required at the same time various effluent
parameters are limited.  For example,  the permit writer could
both set requirements for additional effluent bioconcentration
analysis or tissue residue monitoring and,  at the same time, set
limits for human health in the permit.  The data requirements
recommended for both the effluent bioconcentration analysis and
for tissue analysis are discussed below along with the options
available to the regulatory authority after receiving the data
generated.  The results of either of these data generation
analyses should subsequently result in modification under 40 CFR
§ 122.62(a)(2)  or reissuance of the NPDES permit if the data
generated show the presence of bioconcentratable contaminants
which may exceed a State standard or RAC.

Option 1:   The Section 308 letter

     Section 308 of the Clean Water Act authorizes EPA and the
     States to impose monitoring requirements on any point source
     discharge so long as the data generation conforms to the
     criteria of reasonableness.   Specifically, biological
     monitoring (including tissue residue monitoring) is listed
     in Section 308.

                              VI-12

-------
     Example language is provided in Appendix G.  The example
     provided requires effluent bioconcentration evaluation (the
     procedures listed are not necessarily recommended, but are
     only examples).

Option 2:  NPDES Permit Special Conditions

     Permits can be issued with data generation requirements
     described in the Special Conditions section of the permit
     itself.  The special conditions are written to augment the
     limits imposed on other parameters.  Testing procedures
     require permittees to generate data on their effluents so
     that the permit writer can determine if additional permit
     limits should be imposed.  An example of a bioconcentration
     analysis monitoring requirement that could be placed in the
     Special Conditions section of an NPDES permit is provided in
     Appendix G.

     These two possible mechanisms for requiring a discharger to
conduct an effluent bioconcentration evaluation also apply for
requiring additional follow up analyses or confirmation if
indicated by the results of the initial assessment options.  The
appropriate mechanism to be used for a given facility will be
determined by the timing for requiring this assessment as
discussed above.  The regulatory mechanisms which may be used to
require this assessment are:

•    Via a Section 308 letter prior to permit issuance or after
     permit issuance; if needed limits may be developed for the
     permit at reissuance or the permit may be reopened.

•    Via permit requirement placed in the special conditions
     section of the permit; if needed limits may be placed in
     reopened permit or in permit at next reissuance.

     An example Section 308 letter and example permit language
for requiring additional analyses or confirmation based on the
results of the initial screening assessment are provided in
Appendix G.  Following notification of the requirement for this
additional analyses,  the selected discharger will repeat the
assessment option or conduct confirmation procedures as is
described in Chapter 3 and Appendix A and B.  Appendix G also
contains example permit language for requiring fish tissue and
sediment evaluations for bi©accumulative pollutants.  The
specific laboratory procedures for fish tissues and for sediments
are contained in Appendices A and C of this document.
                              VI-13

-------
                            APPENDIX A






LABORATORY PROCEDURES FOR DETERMINING BIOCONCENTRATABLE CHEMICALS




                        IN TISSUE SAMPLES

-------
                            Appendix A


Fish and Shellfish Tissue Analysis Procedure - Revision 1.0


1.  Scope and Application

     This method provides procedures for preparing fish and
shellfish samples for GC/MS analysis using Soxhlet extraction,
gel permeation chromatography, and silica gel chromatography, and
procedures for identifying and quantifying bioconcentratable
chemicals in the prepared samples with GC/MS.  This method is
applicable to organic chemicals which can be chromatographed
using gas chromatography and detected using mass spectrometry.

2.  Summary of Method

     20 grams of a ground fish or shellfish sample is mixed with
anhydrous sodium sulfate, spiked with a surrogate standard
mixture, and then Soxhlet extracted using methylene
chloride/hexane for a minimum of 12 hours.  Gel Permeation
Chromatography (GPC) followed up with silica gel chromatography
are performed to remove lipids and cholesterol-like materials
from the extract, respectively.  Extract is then concentrated to
0.10 ml and is spiked with an internal standard.  The extract is
analyzed using capillary column gas chromatography with full scan
electron impact ionization mass spectrometry (GC/MS).  After
GC/MS analysis, the extract is saved for confirmation analysis
and the peaks in the GC/MS data are identified and quantitated.
Also included is a procedure for measuring lipid content of the
fish or shellfish sample.

     The three surrogate compounds, added to each sample before
extraction, are d10-biphenyl, 13C6-1,2,4,5-tetrachlorobenzene, and
13C6-hexachlorobenzene and the internal standard is d12-chrysene.
Standard curves are calculated using an internal standard method
for each surrogate and subsequently, percent recovery for each
surrogate is determined.  All other GC/MS components are
quantified using the response factor of 13C6-hexachlorobenzene
with d12-chrysene as the internal standard.

     All chromatographic peaks are reverse-searched against
(compared with) the Chemicals of Highest Concern (CHC) mass
spectral library (see Table A-l).  Those GC/MS components with
fits of 70% and greater are considered tentatively identified.
For each tentatively identified component, a list of the best
mass spectral library identifications (up to a total of ten
identifications)  is reported along with the percent fit values,
CAS number of each tentative identification, GC retention time,
                               A-2

-------
and the concentration for the GC/MS component.  This report is
called Report 1.

     For those GC/MS components not identified with the CHC
search with tissue concentrations greater than or equal to 5
ug/kg, these components are reversed-searched against the
EPA/NIH/NBS mass spectral library.  Those GC/MS components with
fits of 70% and greater are considered tentatively identified.
For each tentatively identified component, a list of the best
mass spectral library identifications (up to a total of ten
identifications) is reported along with the percent fit values,
CAS number of each tentative identification, GC retention time,
and the concentration for the GC/MS component.  This report is
called Report 2.

     For those components with fits/matches less than 70% but
greater than 25%, the two best mass spectral library
identifications along with the percent fit values, the CAS number
of each tentative identification, GC retention time, and the
concentration are reported for each the GC/MS component.  For
GC/MS components with fits/matches less than 25%, the
concentrations and GC retention times for these components are
reported and the components labeled as being "unknown".  This
report is identified as Report 3.

     This procedure yields four reports which will be sent to the
regulatory authority.  These reports are:

     1)   Report 1, components tentatively identified using the
          CHC mass spectral library.

     2)   Report 2, components tentatively identified with tissue
          concentrations greater than or equal to 5 ng/g using
          the EPA/NIH/NBS mass spectral library.

     3)   Report 3, components with tissue concentrations greater
          than or equal to 5 ng/g, and fits less than 70% using
          the EPA/NIH/NBS mass spectral library.

     4)   QA/QC Report, recoveries of the three surrogate
          chemicals in the sample and blank, percent lipid
          content of the tissue, GC/MS chromatograms for the
          sample and blank, GPC performace data, silica gel
          performance data, GC/MS performace data, and precision
          data.
3.   Definitions
                               A-3

-------
     3.1  Bioconcentration Factor (BCF).   Ratio of the
          concentration in the tissue of the organism to that in
          water for an individual chemical.  In equation form,

                          BCF = CF/CW

          where CF and Cw are the concentrations  in the tissue
          and aqueous phase.

     3.2  Surrogate Compound.  A pure compound added to a sample
          before extraction.

     3.3  Internal Standard.  A pure compound added to a sample
          extract prior to GC/MS analysis.

     3.4  CHC Mass Spectral Library.  This library is a subset of
          the EPA/NIH/NBS Mass Spectral Library which contains
          the chemicals in Table A-l.

     3.5  EPA/NIH/NBS Mass Spectral Library.  A library of
          reference mass spectra published by National Bureau of
          Standards, U.S. Government Printing Office, Washington,
          D.C.

     3.6  Procedural Blank.  A sample analysis performed in the
          laboratory with no sample tissue that is treated as a
          sample including exposure to all glassware, equipment,
          solvents, reagents, internal standards, and surrogates
          that are used with other samples.

     3.7  Laboratory Duplicate.  Two sample aliquots taken in the
          analytical laboratory and analyzed separately with
          identical procedures.

4.    Interferences

     4.1  Interferences may be caused by contaminants in
          solvents, reagents, glassware,  and other sample
          processing equipment.  Procedural blanks are analyzed
          routinely to demonstrate that these materials are free
          of interferences under the analytical conditions used
          for samples.

     4.2  To minimize interferences, glassware (including sample
          bottles) should be meticulously cleaned.  As soon as
          possible after use, rinse glassware with the last
          solvent used.  Then wash with detergent in hot water
          and rinse with tap water followed by distilled water.
          Drain dry and heat in a muffle furnace at 450°C for a
          few hours.  After cooling, store glassware inverted or
          covered with aluminum foil.  Before using, rinse each

                               A-4

-------
          piece with an appropriate solvent.  Volumetric
          glassware should not be heated in a muffle furnace.

5.   Safety

     5.1  The toxicity or carcinogenicity of each chemical used
          in this method has not been precisely defined.
          Therefore, each should be treated as a potential health
          hazard, and exposure should be reduced to the lowest
          feasible concentration.  Each laboratory is responsible
          for safely disposing materials and for maintaining
          awareness of OSHA regulations regarding safe handling
          of the chemicals used in this method.  A reference file
          of material data handling sheets should be made
          available to all personnel involved in analyses.
          Additional information on laboratory safety is
          available [11.1,11.2,11.3].

     5.2  The following method analytes have been classified as
          known or suspected human or mammalian carcinogens:
          d10-biphenyl, 13C6-l,2,4,5-tetrachlorobenzene, d12-
          chrysene, and  C6-hexachlorobenzene.   Primary standards
          of these compounds should be prepared in a hood.  A
          toxic gas respirator should be worn when the analyst
          handles solutions containing high concentrations of
          these compounds.

6.   Apparatus and Equipment (All specification are suggested.
     Catalog numbers are included for illustration only.)

     6.1  Glassware

          6.1.1  Soxhlet extractors —  200 ml capacity, 500 ml
                 flask, coarse fritted glass Soxhlet extraction
                 thimble.

          6.1.2  Chromatography Column — glass column
                 approximately 190 mm long X 9 mm ID with Teflon
                 stopcock.

          6.1.3  Concentrator Tube — 10 mL graduated Kuderna-
                 Danish design with ground-glass stopper.

          6.1.4  Evaporative Flask — 500 mL Kuderna-Danish
                 design that is attached to concentrator tube
                 with springs.

          6.1.5  Snyder Column — three-ball macro Kuderna-Danish
                 design.
                               A-5

-------
     6.1.6  200 ^L autosampler microvials and/or microvial
            inserts.

6.2  Gel Permeation Chromatography (GPC)  System.

     The GPC system must be capable of injecting large
     sample volumes, e.g.,  1 to 10 mL, UV detection at 254
     nm, pumping capacity of 3 to 10 ml/min,  and capacity to
     collect fractions.

6.3  GPC Column

     2.5 x 50 cm glass column packed with neutral, porous
     styrene-divinylbenzene copolymer beads with molecular
     weight exclusion limit of 2000,  e.g. Bio-Beads SX-3.
     Note, a larger diameter column might be desirable, see
     9.2.

6.4  GC/MS System

     6.4.1  The GC must be capable of temperature
            programming, splitless or on-column injection,
            and have a designed capillary column injector.

     6.4.2  The MS must be capable of full scan mass
            spectral analysis using electron ionization at a
            electron energy of 70 ev.  The required MS scan
            rate should be >0.5 s and <1.5 s.

     6.4.3  An interfaced data system (DS)  is required to
            acquire, store, reduce, and output mass spectral
            data.  The DS must be capable of performing
            typical mass spectral data manipulations; i.e.,
            creating and plotting total and selected ion
            current profiles, integrating chromatographic
            peak areas,  perform quantifications using an
            internal standard method, etc.

     6.4.4  The data system must be capable of library
            searching detected chromatographic peaks.

     6.4.5  The DS must have the latest release of the
            EPA/NIH/NBS mass spectral library.  This library
            is available for most GC/MS systems from their
            manufacture.

     6.4.6  The DS must have a mass spectral library
            consisting of the chemicals in Table A-l.  This
            library, the CHC library, is a subset of the
            EPA/NIH/NBS mass spectral library.  This library
                          A-6

-------
            can be constructed on most GC/MS systems without
            running GC/MS analyses on standard solutions!

6.6  GC Column

     A 30 m X 0.32 mm or 30 m X 0.25 mm ID fused silica
capillary column coated/bonded with a 0.25 pm or thicker
film crosslinked 5% phenyl methyl or methyl silicone.

6.7  Miscellaneous Equipment

     6.7.1  Microsyringes — various standard sizes.

     6.7.2  Boiling chips — approximately 10/40 mesh.  Heat
            at 400 C for 30  min.  or extract with methylene
            chloride in a Soxhlet apparatus.

     6.7.3  Water bath — heated, with concentric ring
            cover, capable of temperature control within
            ±2°C.

     6.7.4  Analytical balance — capable of accurately
            weighing to 0.0001 g.

     6.7.5  Beakers —  250 ml.

     6.7.6  Disposable aluminum weighing pans

     6.7.8  Desiccator

     6.7.9  Soxhlet extractor heating mantle

Reagents and Consumable Materials

7.1  Solvents.  High purity, distilled-in-glass hexane and
     methylene chloride.  For precise injections with
     splitless injectors and capillary columns, all samples
     and standards should be contained in the same solvent.
     Effects of minor variations in solvent composition
     (i.e., small percentage of another solvent remaining in
     hexane extracts) are minimized with the use of internal
     standards.  (External standard calibration is not
     acceptable).

7.2  Sodium Sulfate.  ACS, granular, anhydrous.  Purify by
     heating at 400°C for 4  h in  a  shallow tray.

7.3  Silica Gel.   60-200 Mesh, Soxhlet extracted with
     hexane/methylene chloride (1:1).

7.4  Glass Wool.   Purify by heating at 400°C for 24  h.

                          A-7

-------
7.5  Nitrogen Gas.  High purity, dry.

7.6  GPC Performance Solution.  5 mg/ml Dacthal, 0.2 mg/ml
     pyrene, and 4 mg/ml di-2-ethylhexylphthalate in
     methylene chloride.

7.7  Silica-Gel Performance Solution.  2 mg/ml dieldrin and
     10 mg/ml cholesterol in hexane.

7.8  Internal Standard Spiking Solution.  200 ppm solution
     (in hexane) containing d12-chrysene.
7.9  Surrogate Standard Spiking Solution.  1 ppm solution
     (acetone) containing d10-biphenyl, 13C6-1,2,4,5-
     tetrachlorobenzene, and 13C6-hexachlorobenzene.
7.10 MS Performance Check Solution.  10 ppm solution  (in
     hexane) containing decafluorotriphenyl-phosphine
     (DFTPP).

7.11 GC/MS Calibration Solutions.  Five hexane solutions are
     required.  These solutions contain constant
     concentrations of the internal standard, d12-chrysene
     and varying concentrations of the surrogates.
     Composition and approximate concentrations are given in
     Table A-2.

7.12 GC Performance Solution.  10 ppm solution (in hexane)
     containing /?-BHC,  -BHC, d12-chrysene, and endrin ketone
     or a 10 ppm solution (in hexane) containing anthracene,
     phenanthrene, benz[a]anthracene, and chrysene.

Calibration

8.1  GPC Chromatography Conditions

     Chromatography Conditions — Fill solvent reservoir with
     HPLC grade methylene chloride, set the flow rate to 5.0
     mL/min, and set the UV detector at 254 nm.

8.2  GPC Performance Criteria

     8.2.1  Inject 350 /iL of the GPC performance solution and
            record GPC chromatogram.

     8.2.2  Baseline resolution should exist among Dacthal, di-
            2-ethylhexylphthalate and pyrene.

     8.2.3  If conditions stated in 8.2.2 are not met, modify
            Chromatography conditions or replace the
            Chromatography column.  Repeat 8.2.2 until its

                          A-8

-------
            conditions are met before performing GPC on
            samples.

8.3  GPC Fraction Time (Collection Time) Identification.
     8.3.1  Inject 350 ML of the GPC performance solution and
            record the retention times (RTs) of Dacthal, di-2-
            ethylhexylphthalate, and pyrene.  Calculate the
            starting and ending times for collection of the
            eluate from the GPC column.  The starting time is
            equal to the average of the retention times of di-
            2-ethylhexylphthalate and Dacthal.  The ending time
            is equal to 1.7 times the average of the retention
            times of pyrene and di-2-ethylhexylphthalate.

     8.3.2  RT reproducibility — For each compound in the GPC
            performance solution, the absolute RTs should not
            vary by more than ±0.40 minutes from one analysis
            to the next.

8.4  Silica Gel Chromatography Conditions

     Clean silica gel activated at 225°C  for 18  hours  is
     allowed to come to room temperature.  100 grams of silica
     gel and 1 ml of clean water are placed into a clean
     scalable container.  The container is sealed, shaken to
     disperse the water, and then, allowed equilibrate for 18
     hours.  After equilibration, a silica gel column is
     prepared for use.

8.5  Silica Gel Chromatography Criteria

     8.5.1. A silica gel column is prepared and 1 ml of the
            silica gel performance solution is placed on to the
            top of the column.  The column is eluted with
            methylene chloride/hexane (15%, v:v, 60 mL) .  The
            eluant, analyzed by GC/MS or flame ionization
            detector/gas Chromatography (FID/GC) , must not
            contain more than 10% of the cholesterol while at
            least 90% of the dieldrin must be recovered.

     8.5.2  If conditions in 8.5.1 are not met,  modify silica
            gel Chromatography conditions to obtain desired
            separation.  These conditions must be met before
            performing silica gel Chromatography on samples.


8.6  GC/MS Conditions.

     8.6.1  Recommended gas Chromatography conditions
                          A-9

-------
            Column Type:        DB-5
            Film Thickness:     0.25
            Column Dimensions:  30 m X 0.32 mm or 30 m X
                                0.25 mm
            Helium Linear Velocity:
                     30 cm/ sec @ 250°C

            Temperature Program:
                     Inject 50°C,  hold 4  mins.,
                     increase to 175°C at 10°C/min,
                     increase to 275°C at 5°C/m,
                     hold at 275°C for 20 mins.

            Injection Volume:   l or 2 /iL

     8.6.2  Recommended acquisition conditions for mass
            spectrometer .

            Mass Range:         45-545 m/z
            Total Cycle Time
              per Scan:         0.5 < cycle time < 1.5 seconds

     8.6.3  Mass Spectrometer Calibration

            8.6.3.1  Calibrate and tune MS with standards and
                     procedures prescribed by the manufacturer.

            8.6.3.2  Inject 1 juL or 2 ML aliquot of MS
                     performance check solution.  If spectrum
                     does not meet criteria for DFTPP (Table
                     A-3 ) ; return to 8.6.3.1 and
                     recalibrate/tune MS.

8 . 7  GC/MS Performance Criteria

     8.7.1  GC Performance — Inject 1.0 ML of the GC
            performance solution.  Baseline separation between
            /3-BHC and  -BHC and between endrin ketone and d12-
            chrysene should exist.  Alternatively, anthracene
            and phenanthrene should be separated by baseline
            and benz[ a] anthracene and chrysene should be
            separated by a valley whose height is less than 25%
            of the average peak height of these two compounds.
     8.7.2  MS Sensitivity — Inject 1.0 /xL of the 0.5 ppm
            GC/MS calibration solution.  Using the total ion
            chromatogram, a signal to noise ratio of greater
            than 3 should be observed for each surrogate.

     8.7.3  MS Calibration — Inject 1.0 ML of the 0.5 ppm
            GC/MS calibration solution.  For d12-chrysene,

                         A-10

-------
            abundance of m/z 241 relative to that of m/z 240
            should be >15% and <25%.

     8.7.4  GC Stability — Perform multiple GC/MS analyses on
            the same GC/MS calibration solution.  RTs should
            not vary by more than t seconds.  Calculate the
            value of t with the equation, t = (RT)1'3, where RT
            is the observed average RT (in seconds).

8.8  Response Factor Calculation for MS

     8.8.1  Inject 1.0 /iL of each GC/MS calibration solution
            and acquire GC/MS data.

     8.8.2  Calculate response factors (RF) for each surrogate
            relative to d12-chrysene

                      RF = ASQC/ACQS

            where As = integrated total ion abundance for the
                        surrogate
                   Ac =  integrated  total  ion abundance for the
                       internal standard, d12-chrysene
                  Qs  = injected quantity  of surrogate
                  Q0  = injected quantity  of d12-chrysene

     8.8.3  RF Reproducibility — For each surrogate, calculate
            the mean RF.  When the relative standard deviation
            (RSD) exceeds 30%, analyze additional aliquots of
            GC/MS calibration solutions to obtain acceptable
            RSD for the RF, or take action to improve GC/MS
            performance.

8.9  Continuing Calibration Check

     8.9.1  GPC

            8.9.1.1  With the following procedures, verify GPC
                     column performance at the beginning and
                     end of each 12 h period during which
                     analyses are performed.

            8.9.1.2  Demonstrate acceptable performance for
                     criteria described in Section 8.2.3.

            8.9.1.3  Demonstrate acceptable performance for
                     criteria described in Section 8.3.2.

     8.9.2  GC/MS
                         A-ll

-------
       8.9.2.1  With the following procedures,  verify
                initial calibration at beginning and end
                of each 12 h period during which analyses
                are performed.

       8.9.2.2  Inject 1 or 2 /nL aliquot of MS performance
                check solution.   Ensure acceptable MS
                calibration and  performance.

       8.9.2.3  Demonstrate acceptable performance for
                8.7.

       8.9.2.4  Determine the area for d12-chrysene has
                not changed by more than 30% from most
                recent analyses  of the GC/MS calibration
                solutions.

       8.9.2.5  For an acceptable continuing calibration
                check, the measured RF for each surrogate
                must be within 30% of the mean value
                calculated during initial calibration.

8.9.3 Silica Gel

       8.9.3.1  With the following procedure, verify
                silica gel chromatography performance for
                each lot of material and/or every 2
                months, which ever comes first.

       8.9.3.2  Demonstrate acceptable performance for
                criteria described in Section 8.5

8.9.4  Remedial Actions

       Remedial actions must be  taken if criteria are not
       met; possible remedies are:

       8.9.4.1  Check and adjust operating conditions.

       8.9.4.2  Clean or replace injector liner on GC.

       8.9.4.3  Flush column with solvent according to
                manufacturers instructions.

       8.9.4.4  Break off a short portion (approximately
                0.33 m) of the GC column; check column
                performance by analysis of performance
                check solution for GC.

       8.9.4.5  Replace column;  performance of all initial
                calibration procedures then required.

                    A-12

-------
                 8.9.4.6  Adjust MS for greater or lesser
                          resolution.

                 8.9.4.7  Calibrate MS mass scale.

                 8.9.4.8  Prepare and analyze new concentration
                          calibration/performance solutions.

                 8.9.4.9  Prepare new concentration calibration
                          curve(s).
9.    Procedures

     9.1  Extraction
          9.1.1  Ground fish or shellfish tissue (20 g) is blended
                 with enough anhydrous sodium sulfate  (100 to 140 g)
                 in a 250 ml beaker to completely dry the sample.

          9.1.2  Place two-thirds of the mixture into a coarse
                 fritted glass Soxhlet extraction thimble and then
                 add 0.100 mL of the surrogate solution.  The
                 remaining sample is added to the thimble and the
                 extraction thimble is placed into a clean Soxhlet
                 extractor body.

          9.1.3  300 ml of methylene chloride/hexane (50:50 v:v)  and
                 one or two boiling chips are placed into the
                 Soxhlet extractor flask.  The flask along with
                 Soxhlet extractor body are placed on to the
                 extraction rack.

          9.1.4  The heaters on the extraction rack are turned on
                 and the sample is extracted for at least 12 hours.

          9.1.5  After allowing the extractor to cool,
                 quantitatively transfer the extract to a 500 mL
                 Kuderna-Danish (KD)  apparatus fitted with a 10 ml
                 lower tube.

          9.1.6  Add one or two clean boiling chips to the flask and
                 attach a 3-ball Snyder column to the KD apparatus.
                 Concentrate the extract on a steam bath to a volume
                 of approximately 8 ml.

          9.1.7  Place a weighing pan in a 105±5°C  oven for  15
                 minutes, let cool for 15 minutes in desiccator,  and
                 then weigh to 5 places.

          9.1.8  Quantitatively transfer extract to 10 ml volumetric
                 with hexane and adjust volume to 10 ml.

                              A-13

-------
     9.1.9  Take tared weighing pan from desiccator and place
            1.0 ml of the extract into the weighing pan.

     9.1.10 Let solvent evaporate from weighing pan, place pan
            into 105±5°C  oven  for 15 minutes,  let  cool  in
            desiccator, and reweigh to 5 places.

     9.1.11 Using a gentle stream of dry air or nitrogen
            concentrate remaining sample, i.e., the remaining 9
            ml, to approximately 1.0 ml.

     9.1.12 Transfer extract to vial suitable for use with GPC
            system using methylene chloride.  Cap vial and
            store in freezer.

9.2  GPC Cleanup

     9.2.1  Remove sample from freezer and allow it to warm to
            ambient temperature.  Dilute extract with methylene
            chloride to volume needed for GPC analysis, e.g. 2
            ml.  The size of the injection loop on the GPC
            systems vary according to manufacture.  Analyst
            judgement/experience with their GPC system is
            required to determine this volume.

     9.2.2  Using the GPC conditions determined previously (see
            Section 9.1), inject all of the sample into the GPC
            column.  Note, the capacity of the GPC column
            (described in 6.3) is approximately 0.5 g of lipid.
            If the total amount of lipid exceeds 0.5 g, steps
            9.2.2 (injection)  and 9.2.3 (collection of purified
            extract)  should be repeated using the appropriate
            injection volumes so that all of the extract has
            been fractionated on the GPC.  Analyst judgement
            and experience are required for this procedure.

     9.2.3  At the starting time determined in 8.3.1, a clean
            flask under the waste tube from the UV detector and
            collect the column eluate.  Collect eluate from the
            UV detector until the ending time determined in
            8.3.1.

     9.2.4  Quantitatively transfer all of the collected eluate
            for a sample to a 500 ml KD with a 10 ml lower
            tube.

     9.2.5  Add one or two clean boiling chips to the KD and
            attach a 3 ball Snyder column to Kuderna-Danish
            apparatus.  Add approximately 30 ml of hexane to
            KD.  Concentrate the extract on a steam bath until
            volume of the extract is less than 8 mL.  Allow

                          A-14

-------
            Kuderna-Danish apparatus to cool and detach the
            lower tube, rinsing the joint with hexane into the
            sample.

     9.2.6  Use a nitrogen stream and a warm (<40°C)  water bath
            to concentrate the extract to a volume of 1.0 ml.

     9.2.7  Quantitatively transfer sample to vial using
            hexane.  Cap vial and store in freezer.

9.3  Silica Gel Chromatography

     9.3.1  Remove extract from freezer and allow it to come to
            room temperature.

     9.3.2  Adjust extract volume to 0.5 mL.

     9.3.2  Prepare silica gel column using 19 cm X 9 mm ID
            glass column containing from bottom to top, glass
            wool, 0.5 cm of sodium sulfate, 2.1 g of
            deactivated silica gel (see 8.4), and 0.5 cm sodium
            sulfate.  Wash column with 50 ml of hexane.

     9.3.3  Quantitatively transfer extract to column using
            hexane.

     9.3.4  Using elution conditions determine in 8.5, elute
            column and collect eluate in receiving container.

     9.3.5  Use a nitrogen stream and a warm (<40°C)  water bath
            to concentrate the eluate to a volume of 0.50 mL.

     9.3.6  Quantitatively transfer sample to GC microvial and
            concentrate to 0.100 ml.   Cap vial and store in
            freezer.

9.4  GC/MS Analysis

     9.4.1  Remove sample from freezer and allow it to warm to
            ambient temperature.  Adjust volume of sample to
            100 /zL and then spike sample with 5 /xL of the
            internal standard spiking solution, 1000 ng of d12-
            chrysene.  Mix sample after spiking sample with
            d12-chrysene.

     9.4.2  Inject a 1 juL or 2 jj.l> aliquot of blank or sample
            extract into GC operated under conditions used to
            produce acceptable results during calibration.

     9.4.3  Acquire MS data using full scan conditions (see
            Section 8.6).

                         A-15

-------
     9.4.4  Recap samples and store extracts in freezer.  Note,
            autosampler crimp caps after being punctured by the
            syringe used for sample injection must be replaced.

9.5  Percent Lipid Calculation

     9.5.1  Calculate difference between the before and after
            oven drying of the tared weighing pan.  This
            difference is the residue weight.

     9.5.2  Calculate percent lipid with the following formula:

            Percent lipid =

                     =    Residue weight * 100	
                          0.10 * mass of tissue extracted

9.6  GC/MS Data Analysis

     Data analysis for this analytical procedure can be divided
into three tasks, 1) quantification of surrogates and other
GC/MS components, 2) library searching with the CHC mass
spectral library, and 3) library searching with the EPA/NIH/NBS
mass spectral library.


     9.6.1  Quantification of Surrogates and Other GC/MS
            Components

            9.6.1.1  Surrogate and Internal Standard
                     Identification

            Identify surrogates and internal standards by
            comparison of their mass spectrum (after background
            subtraction) to reference spectrums in the user-
            created data base.  The GC retention time of the
            sample component should be within 10 sec of the
            time observed for that same compound when a
            calibration solution was analyzed.  In general, all
            ions that are present above 10% relative abundance
            in the mass spectrum of the standard should be
            present in the mass spectrum of the sample
            component and should agree within absolute 20%.
            For example, if an ion has a relative abundance of
            30% in the standard spectrum, its abundance in the
            sample spectrum should be in the range of 10 to
            50%.  Some ions, particularly the molecular ion,
            are of special importance, and should be evaluated
            even if they are below 10% relative abundance.
                         A-16

-------
9.6.1.2  Peak Integration

Use the GC/MS peak detection and  integration
software to obtain areas for all  chromatographic
components from the total ion chromatogram with a
signal to noise of 3 and greater.   (Note, the
solvent front need not be examined.)  Verify that
each surrogate was integrated.  If  not, use a more
sensitive peak detection and integration settings
to obtain peak areas for the surrogates.

9.6.1.3  Calculate Surrogate Concentrations

         cx   =    (AX-Q|S)/(A|S-RF-M)

Where    Cx   =    Concentration  of surrogate

         Ax   =    Area of surrogate  in total ion
                   chromatogram

         QIS  =     Quantity of internal standard
                   added to extract before GC/MS
                   analysis

         A|S  =     Area of internal standard in
                   total ion chromatogram added to
                   the extract before GC/MS
                   analysis

         RF   =    Mean response  factor of
                   surrogate from initial
                   calibration analysis and/or from
                   GC/MS calibration  solutions run
                   with sample analyses.

         M    =    Mass of extracted  tissue, i.e.,
                   20 g.

Alternatively, use the GC/MS system software or
other available proven software to  compute the
concentrations of surrogate using linear, second,
or third order regression or using  piecewise
calibration curves.

9.6.1.4  Calculate Surrogate Recoveries

Surrogate Recovery = (Cx •  100)/SSC

Where    Cx   =    Concentration  of surrogate
         SSC  =    Surrogate spiking concentration,
                   i.e., 5 ng/g

              A-17

-------
       9.6.1.5  Calculate concentrations of all
                chromatographic components

                Cx = (A,, • Q|S • 100) / (A|S • RF • M • REC)

       Where    Cx  =   Concentration of chromatographic
                        component

                Ax  =   Area of chromatographic component
                        in total ion chromatogram

                QIS =    Quantity of internal standard
                        added to extract before GC/MS
                        analysis

                A|S =    Area of internal standard in
                        total ion chromatogram

                M   =   Mass of extracted tissue, i.e.  20
                        g

                RF  =   Mean response factor of surrogate
                        from initial calibration analyses
                        and/or from GC/MS calibration
                        solutions, run with sample
                        analyses.  Use RF for 13C6-
                        hexachlorobenzene.

                REC =   Surrogate recovery of 13C6-
                        hexachlorobenzene.

       Alternatively, use the GC/MS system software or
       other available proven software to compute the
       concentration of all chromatographic components
       using linear, second,  or third order regression or
       using piecewise calibration curves.   These
       calculations must use the response curve for the
       13C6-hexachlorobenzene  surrogate and  the reported
       concentrations must be corrected for recovery of
       the 13C6-hexachlorobenzene surrogate.

9.6.2  Library searching with the CHC mass spectral
       library

       9.6.2.1  Algorithm Selection

       A reverse searching algorithm is required when
       available.  If GC/MS system does not have a reverse
       searching algorithm, library searching should be
       performed using the default algorithm supplied by
       the manufacture of the instrument.

                     A-18

-------
       9.6.2.2  Searching

       All chromatographic components detected in 9.6.1.2
       (Peak Integration) are searched against the CHC
       mass spectral library.  Those GC/MS components with
       fits of 70% and greater are considered tentatively
       identified.  For each tentatively identified
       component, a list of the best mass spectral library
       identifications (up to a total of ten
       identifications) is reported along with the percent
       fit values, CAS number of each tentative
       identification, GC retention time, and the
       concentration for the GC/MS component.  This report
       is called Report 1.

9.6.3  Library Searching with EPA/NIH/NBS Mass Spectral
       Library.

       9.6.3.1  Algorithm Selection

       A reverse searching algorithm is required when
       available.  If GC/MS system does not have a reverse
       searching algorithm, library searching should be
       performed using the default algorithm supplied by
       the manufacture of the instrument.

       9.6.3.2  Eliminating of Compounds Below 5 ng/g

       All unidentified components from the CHC library
       search with concentrations less than 5 ng/g are
       eliminated from further data processing.  This
       elimination can be performed by comparing peak
       areas or heights of each chromatographic peak to
       the peak area or height of the 13C6-
       hexachlorobenzene surrogate.  This elimination may
       also be performed by comparing the recovery
       corrected concentration of each chromatographic
       component to 5 ng/g.

       9.6.3.3  Searching

       The remaining components are searched against the
       EPA/NIH/NBS mass spectral library.  Those GC/MS
       components with fits of 70% and greater are
       considered tentatively identified.  For each
       tentatively identified component, a list of the
       best mass spectral library identifications (up to a
       total of ten identifications) is reported along
       with the percent fit values, CAS number of each
       tentative identification, GC retention time,  and
                    A-19

-------
            the concentration for the GC/MS component.  This
            report is called Report 2.

            For those components with fits/matches less than
            70% but greater than 25%, the two best mass
            spectral library identifications along with the
            percent fit values, the CAS number of each
            tentative identification, GC retention time, and
            the concentration are reported for each the GC/MS
            component.  For GC/MS components with fits/matches
            less than 25%,  the concentrations and GC retention
            times for these components are reported and the
            components labeled as being "unknown".  This report
            is identified as Report 3.

     9.6.4  Elimination of chromatographic components common to
            the spiked blank and tissue GC/MS data.

            Chromatographic components with retention times
            within ten seconds between the spiked blank and
            tissue and with the same mass spectrums should be
            removed from the analysis.

9.7  Reporting of Data

     9.7.1  Report l:  CMC mass spectral identifications.

            For each chromatographic component tentatively
            identified using the CHC search (fits greater than
            or equal to 70%),  a list of the best mass spectral
            library identifications (up to a total of ten
            identifications) is reported along with the percent
            fit values, CAS number of each tentative
            identification, GC retention time, and the
            concentration for the GC/MS component.

     9.7.2  Report 2:  EPA/NIH/NBS mass spectral tentative
            identifications (fits > 70% and > 5 ug/kg).

            GC/MS components tentatively identified in the
            EPA/NIH/NBS mass spectral search.  For each
            tentatively identified component, a list of the
            best mass spectral library identifications (up to a
            total of ten identifications)  is reported along
            with the percent fit values, CAS number of each
            tentative identification, GC retention time, and
            the concentration for the GC/MS component.

     9.7.3  Report 3:  EPA/NIH/NBS mass spectral tentative
            identifications (fits < 70% and > 5 ug/kg).
                         A-20

-------
                 GC/MS components not tentatively identified in the
                 EPA/NIH/NBS mass spectral search.  For those
                 components with fits/matches less than 70% but
                 greater than 25%, the two best mass spectral
                 library identifications along with the percent fit
                 values, the CAS number of each tentative
                 identification, GC retention time, and the
                 concentration are reported for each the GC/MS
                 component.  For GC/MS components with fits/matches
                 less than 25%, the concentrations and GC retention
                 times for these components are reported and the
                 components labeled as being "unknown".

          9.7.4  QA/QC Report

                 9.7.4.1  Percent lipid content of the sample.

                 9.7.4.2  Recoveries.

                 For d10-biphenyl, 13C6-l,2,4,5-tetrachlorobenzene,
                 and 13C6-hexachlorobenzene,  recoveries will be
                 reported.

                 9.7.4.3  GC/MS Chromatograms.

                 For the sample and its corresponding blank, total
                 ion Chromatograms must be provided.

                 9.7.4.4  QA/QC.

                 Data demonstrating GPC Resolution, see
                      Section 8.2.2
                 Data demonstrating silica gel performance,
                      see Section 8.5.
                 Data demonstrating GC resolution, see
                      Section 8.7.1
                 Data demonstrating MS sensitivity, see
                      Section 8.7.2
                 Data demonstrating MS calibration, see
                      Section 8.7.3
                 Data demonstrating DFTPP performance,
                      see Section 8.6.3.2
                 Data demonstrating precision, see
                      Section 10.4

10.   Quality Control

     10.1 Recoveries of Surrogates.  Method:

           % Recovery = measured surrogate amount  X 100
                         spiked surrogate amount

                              A-21

-------
     Quality Assurance Requirement

                25% < % recovery < 120%


     Quality Control Action.  If percent recovery is out of
     range, re-extract and re-analyze sample.

10.2 Calibration Stability

     10.2.1 Continuing Calibration Checks

            Quality Assurance.  See Sections 8.9

            Quality Control Action.  Re-extract and re-analyze
            samples.  For corrective actions see Section 8.9.

     10.2.2 GPC Stability.

            Quality Assurance.  Immediately before and after
            fractionation solution must be analyzed.  See
            Section 8.3 for stability requirements.

            Quality Control Action.  If RT stability
            requirements are not met,  re-extract and re-analyze
            the sample.

     10.2.3 GC/MS Stability

            Quality Control.  During the GC/MS analysis run, a
            GC/MS calibration solution should be analyzed
            twice; once at the beginning and once at the end of
            the GC/MS analysis sample sequence.  See Sections
            8.7 and 8.9 for stability requirements.

            Quality Control Action.  If RT and/or MS
            sensitivity stability requirements are not met,
            repeat GC/MS analysis on samples after correction
            of instrumental problems.

10.3 Blanks

     10.3.1 Procedural Blanks

            Quality Control.  One procedural blank should be
            performed with every set of tissue samples
            analyzed.

            An acceptable procedural blank:

              a)   Meets Section 10.1 requirements.

                          A-22

-------
              b)  Contains no compound with elution
                   characteristics and mass spectral features
                  that would interfere with identification and
                  quantification of the surrogates.
              c)  Contains few chromatographic peaks in the
                  GC/MS total ion chromatograms.

            Quality Control Action.  Locate and eliminate the
            source of contamination.  Re-extract and re-analyze
            the entire batch of sample.

     10.4.2 Unspiked Procedural Blanks

            Quality Control.  One procedural blank not spiked
            with the surrogate spiking solution should be
            performed with every 20 tissue samples.

            An acceptable unspiked blank.  The sample should
            not contain detectable amounts of the surrogates.

            Quality Control Action.  Locate and eliminate
            source of contamination, re-extract and re-analyze
            entire batch of sample.

10.4 Precision

     Quality Control.  A duplicate set of tissue samples will
     be analyzed with every set of 10 samples.  To measure the
     precision, the relative percent difference between the lab
     duplicate will be determined for each surrogate.

          Relative % difference =

         surrogate amount        surrogate amount
           duplicate 1	-	duplicate 2	   X 100
                  Average surrogate amount

     The relative percent difference should be less than 150%.

     Quality Control Action.  If relative percent difference is
     out-range for any of three surrogates, re-extract and re-
     analyze the sample.

10.5 Sample Sets.

     Sample sets are defined as a group of samples that are
     carried through the analytical procedure at the same time.
     Each sample set will include a minimum of two QC samples,
     one of them being a spiked blank.  QC samples include
     spiked blanks, unspiked blanks, and replicates.
                         A-23

-------
          A set will normally contain 10 samples plus the required
          QC samples.

     10.6 GC/MS Analysis Sets.

          A GC/MS analysis set is defined as a group of prepared
          samples (tissues and QC samples)  and GC/MS calibration
          solutions analyzed during one GC/MS run.   A set will
          normally contain 12 prepared samples, all five GC/MS
          calibration solutions,  and 1 or 2 duplicates of the GC/MS
          calibration solutions.   The duplicate GC/MS calibration
          solutions will be analyzed in the beginning and at the end
          of GC/MS sample sequence.

11.   Appendix A References.

     11.1 "Carcinogens — Working with Carcinogens", Department of
          Health Service, Center for Disease Control, National
          Institute for Occupational Safety and Health, Publication
          No. 77-206, August 1977.

     11.2 "OSHA Safety and Health Standards, General Industry", 29
          CFR 1910,  Occupational Safety and Health Administration,
          OSHA 2206, Revised January 1976.

     11.3 "Safety in Academic Chemistry Laboratories", American
          Chemical Society Publication, Committee on Chemical
          Safety, 3rd Edition, 1979.
                              A-24

-------
                              TABLE A-l

                  CHEMICALS OF HIGHEST CONCERN LIST
CAS number
                    chemical name
   50-
   57-
   58-
   60-
   70-
   72-
   72-
   76-
   91-
   95-
  101-
  115-
  117-
  118-
  309-
  319-
  319-
  608-
  608-
  924-
 1024-
 1746-
 2104-
 2385-
 8001-
39515-
11096-
11097-
11104-
11141-
12672-
12674-
53469-
•29-3
•74-9
•89-9
•57-1
•30-4
•54-8
•55-9
•44-8
•94-1
•94-3
•61-1
•32-2
•81-7
•74-1
•00-2
•84-6
•85-7
•73-1
•93-5
•16-3
•57-3
•01-6
•64-5
•85-5
35-2
41-8
82-5
69-1
28-2
16-5
29-6
11-2
21-9
p,p'-dichlorodiphenyltrichloroethane (DDT)
chlordane
hexachlorocyclohexane  (lindane)
dieldrin
hexachlorophene
p,p'-dichlorodiphenyldichloroethane (DDD)
p,p'-dichlorodiphenyldichloroethylene (DDE)
heptachlor
3,3'-dichlorobenzidine
1,2,4,5-tetrachlorobenzene
4,4'-methylene bis(N,N'-dimethyl)  aniline
dicofol
bis(2-ethylhexyl)phthalate  (BEHP)
hexachlorobenzene
aldrin
aIpha-hexach1orocyc1ohexane  (aIpha-HCH)
beta-hexachlorocyclohexane  (beta-HCH)
technical-hexachlorocyclohexane  (t-HCH)
pentachlorobenzene
N-nitroso-di-n-butylamine
heptachlor epoxide
dioxin (2,3,7,8-TCDD)
ethyIp-nitrophenyIphenyIphosphorothioate(EPN)
mirex
toxaphene
danitol
polychlorinated biphenyl 1260
polychlorinated biphenyl 1254
polychlorinated biphenyl 1221
polychlorinated biphenyl 1232
polychlorinated biphenyl 1248
polychlorinated biphenyl 1016
polychlorinated biphenyl 1242
                               A-25

-------
Table A-2. GC/MS Calibration Solutions, Concentrations of
Surrogates and Internal Standards.
13C6-l,2,4,5-tetra-
chlorobenzene d10-biphenyl
ppm ppm
0.5 0.5
1 1
10 10
50 50
100 100
C6-hexachloro-
d12-chrysene benzene
ppm ppm
10 0.5
10 1
10 10
10 50
10 100

Table A-3. DFTPP Ion Abundance Criteria.
m/z
51
68
70
127
197
198
199
275
365
441
442
443
Criteria
10-80% of the base peak
<2% of m/z 69
<2% of m/z 69
10-80% of the base peak
<2% of m/z 198
base peak or >50% of 442
5-9% of m/z 198
10-60% of the base peak
>1% of base peak
present and 50% of m/z 198
15-24% of m/z 442
A-2 6

-------
                            APPENDIX B






LABORATORY PROCEDURES FOR DETERMINING BIOCONCENTRATABLE CHEMICALS




                        IN AQUEOUS  SAMPLES

-------
                            Appendix B


Effluent Analysis Procedure - Revision 1.0


1.  Scope and Application

     This method provides procedures for fractionating effluents
using HPLC and procedures for identification and quantification
of bioconcentratable chemicals in these fractions using GC/MS.
This method is applicable to organic chemicals which are stable
under acidic conditions, are hexane extractable, and can be
chromatographed using gas chromatography.

2.  Summary of Method

     A 10 L undiluted effluent sample is spiked with three
surrogate chemicals, i.e., d10-biphenyl, 13C6-1,2,4,5-
tetrachlorobenzene, and 13C6-hexachlorobenzene, and extracted with
hexane.  The hexane extract is subsequently cleaned up using
sulfuric acid, and concentrated to a volume of 0.50 mL.  The
extract is chromatographed using reverse phase HPLC, and three
fractions are collected.  These fractions have nominal BCF ranges
of approximately 91 to 560,  560 to 5,000, and 5,000 to 470,000.
The fractions are extracted and concentrated to 0.10 mL, and are
subsequently spiked with the internal standard, d12-chrysene.
The fractions are analyzed using capillary gas chromatography
with full scan electron impact ionization mass spectrometry
(GC/MS).  After GC/MS analysis, the fractions are saved for
confirmation analysis and the peaks in the GC/MS data are
identified and quantitated.

     Standard curves are calculated using an internal standard
method for each surrogate and subsequently, percent recovery for
each surrogate is determined.  All other GC/MS components are
quantified using the response factor appropriate for each HPLC
fraction:  d10-biphenyl/d12-chrysene  for the first  fraction,
13C6-1,2,4,5-tetrachlorobenzene/d12-chrysene for the second
fraction, and 13C6-hexachlorobenzene/d12-chrysene for the third
fraction.

     For each HPLC fraction, all chromatographic peaks are
reverse-searched against (compared with)  the Chemicals of Highest
Concern (CHC) mass spectral library (see Table B-l).  Those
chemicals with fits of 70% and greater are considered tentatively
identified.  For each tentatively identified component, a list of
the best mass spectral library identifications (up to a total of
ten identifications) is reported along with the percent fit
values, CAS number of each tentative identification, HPLC


                               B-2

-------
fraction number, GC retention time, and the concentration for the
GC/MS component.  This report is called Report 1.

     For those GC/MS components not identified with the CHC
search with effluent concentrations greater than or equal to 100
ng/1, these components are reversed-searched against the
EPA/NIH/NBS mass spectral library.  Those chemicals with fits of
70% and greater are considered tentatively identified and these
components are, then, further evaluated.

     This evaluation process consists of calculating a predicted
tissue concentration (the product of fraction BCF, effluent
concentration, and effluent dilution) and subsequently, comparing
this product to 1 ug/kg.  For those GC/MS components with fits of
70% and greater and a product > 1 ug/kg, a list of the best mass
spectral library identifications (up to a total of ten
identifications) is reported along with the percent fit values,
CAS number of each tentative identification, HPLC fraction
number, GC retention time, and the concentration for the GC/MS
component.  This report is called Report 2.

     For those GC/MS components with fits of 70% and greater and
a product less than 1 ug/kg, a list of the best mass spectral
library identifications (up to a total of ten identifications) is
reported along with the percent fit values, CAS number of each
tentative identification,  HPLC fraction number, HPLC fraction
number, GC retention time, and the concentration for the GC/MS
component.  This report is identified as Report 4

     For those components with fits less than 70% and greater
than 25% and present at concentrations of greater than or equal
to 100 ng/1, the two best mass spectral library identifications
along with the percent fit values,  the CAS number of each
tentative identification,  GC retention time, and the
concentration are reported for each the GC/MS component.  For
GC/MS components with fits less than 25%, the concentrations,
HPLC fraction numbers,  and GC retention times for these
components are reported and the components labeled as being
"unknown".  This report is identified as Report 3.

     This procedure yields five reports which will be sent to the
regulatory authority.  These reports are:

     1)   Report 1, components tentatively identified using the
          CHC mass spectral library.

     2)   Report 2, components tentatively identified with
          effluent concentrations greater than or equal to 100
          ng/1 and a prediction tissue concentration > 1 ug/kg
          using the EPA/NIH/NBS mass spectral library.


                               B-3

-------
3)   Report 3, components with effluent concentrations
     greater than or equal to 100 ng/1, and fits less than
     70% using the EPA/NIH/NBS mass spectral library.

4)   Report 4, components tentatively identified with
     effluent concentrations greater than or equal to 100
     ng/1 and a predicted tissue concentration < 1 ug/kg
     using the EPA/NIH/NBS mass spectral library.

4)   QA/QC Report, recoveries of the three surrogate
     chemicals in the six sample and blank HPLC fractions,
     GC/MS chromatograms for the sample and blank, HPLC
     performance data, GC/MS performance data, and precision
     data.

Definitions

3.1  Bioconcentration Factor (BCF).  Ratio of the
     concentration in the tissue of the organism to that in
     water for an individual chemical.  In equation form,

                     BCF = CF/CW

     where CF  and Cw are the concentrations  in the tissue
     and aqueous phase.

3.2  Surrogate Compound.  A pure compound added to a sample
     before extraction.

3.3  Internal Standard.  A pure compound added to a sample
     extract prior to GC/MS analysis.

3.4  CHC Mass Spectral Library.  This library is a subset of
     the EPA/NIH/NBS Mass Spectral Library which contains
     the chemicals in Table B-l.

3.5  EPA/NIH/NBS Mass Spectral Library.  A library of
     reference mass spectra published by National Bureau of
     Standards, U.S. Government Printing Office, Washington,
     D.C.

3.6  Laboratory Reagent Blank.   An aliquot of reagent water
     that is treated as a sample.

Interferences

4.1  Interferences may be caused by contaminants in
     solvents, reagents, glassware, and other sample
     processing equipment.  Laboratory reagent blanks (LRBs)
     are analyzed routinely to demonstrate that these
                          B-4

-------
          materials are free of interferences under the
          analytical conditions used for samples.

     4.2  To minimize interferences, glassware (including sample
          bottles) should be meticulously cleaned.  As soon as
          possible after use, rinse glassware with the last
          solvent used.  Then wash with detergent in hot water
          and rinse with tap water followed by distilled water.
          Drain dry and heat in a muffle furnace at 450°C for a
          few hours.  After cooling, store glassware inverted or
          covered with aluminum foil.  Before using, rinse each
          piece with an appropriate solvent.  Volumetric
          glassware should not be heated in a muffle furnace.

5.    Safety

     5.1  The toxicity or carcinogenicity of each chemical used
          in this method has not been precisely defined.
          Therefore, each should be treated as a potential health
          hazard, and exposure should be reduced to the lowest
          feasible concentration.  Each laboratory is responsible
          for safely disposing materials and for maintaining
          awareness of OSHA regulations regarding safe handling
          of the chemicals used in this method.  A reference file
          of material data handling sheets should be made
          available to all personnel involved in analyses.
          Additional information on laboratory safety is
          available [12.1,12.2,12.3].

     5.2  The following method analytes have been classified as
          known or suspected human or mammalian carcinogens:
          d1Q-biphenyl, 1 C6-l ,2,4, 5-tetrachlorobenzene,  d12-
          chrysene, and 13C6-hexachlorobenzene.   Primary standards
          of these compounds should be prepared in a hood.  A
          toxic gas respirator should be worn when the analyst
          handles solutions containing high concentrations of
          these compounds.

6.    Apparatus and Equipment (All specification are suggested.
     Catalog numbers are included for illustration only.)

     6.1  Sampling Equipment

          6.1.1  Water Sample Bottles — Meticulously cleaned
                 (Section 4.2)  glass bottles fitted with a
                 Teflon,  aluminum foil, or polypropylene lined
                 screw cap.  A 10 L aliquot of sample is required
                 for one analysis and thus, water sampling
                 bottles could consist of one 10 L, two 5 L, or
                 three 1 gallon bottles.  (Bottles in which high
                 purity solvents were received can be used as

                               B-5

-------
            sample bottles without additional cleaning if
            they have been handled carefully to avoid
            contamination during and after use of original
            contents . )

6.2  Glassware

     6.2.1  Extraction Glassware — six 2 L separatory
            funnels, one 10 L bottle, two 5 L bottles, six 2
            L volumetrics, one 5 gallon carboy, or three 1
            gallon solvent bottles.

     6.2.2  Chromatography Column — glass column
            approximately 400 mm long X 19 mm ID with Teflon
            stopcock and 300 mL reservoir.

     6.2.3  Concentrator Tube — 10 mL graduated Kuderna-
            Danish design with ground-glass stopper.

     6.2.4  Evaporative Flask — 500 mL Kuderna-Danish
            design that is attached to concentrator tube
            with springs.

     6.2.5  Snyder Column — three-ball macro Kuderna-Danish
            design.
     6.2.6  200 ML autosampler microvials and/or microvial
            inserts.

6.3  High Performance Liquid Chromatography System.

     6.3.1  The HPLC must be capable of solvent programming,
            be capable of injecting large sample volumes,
            i.e., 175 ul, and be capable of UV detection at
            200 nm.

     6.3.2  Stop watch/ timing device (required if 6.3.3 not
            available) for manual collection of HPLC
            fractions.

     6.3.3  Fraction collector (optional) if used, should be
            capable of taking time window fractions from
            HPLC.

6.4  HPLC Column

     A25cmX4.6mmi.d., 5 micron C18 column.

6.5  GC/MS System
                          B-6

-------
     6.5.1  The GC must be capable of temperature
            programming, splitless or on-column injection,
            and have a designed capillary column injector.

     6.5.2  The MS must be capable of full scan mass
            spectral analysis using electron ionization at a
            electron energy of 70 ev.  The required MS scan
            rate should be >0.5 s and <1.5 s.

     6.5.3  An interfaced data system (DS) is required to
            acquire, store, reduce, and output mass spectral
            data.  The DS must be capable of performing
            typical mass spectral data manipulations; i.e.,
            creating and plotting total and selected ion
            current profiles, integrating chromatographic
            peak areas, perform quantifications using an
            internal standard method, etc.

     6.5.4  The data system must be capable of library
            searching detected chromatographic peaks.

     6.5.5  The DS must have the latest release of the
            EPA/NIH/NBS mass spectral library.  This library
            is available for most GC/MS systems from their
            manufacture.

     6.5.6  The DS must have a mass spectral library
            consisting of the chemicals in Table B-l.  This
            library, the CHC library, is a subset of the
            EPA/NIH/NBS mass spectral library.  This library
            can be constructed on most GC/MS systems without
            running GC/MS analyses on standard solutions!

6.6  GC Column

     A 30 m X 0.32 mm or 30 m X 0.25 mm ID fused silica
capillary column coated/bonded with a 0.25 Mm °r thicker
film crosslinked 5% phenyl methyl or methyl silicone.

6.7  Miscellaneous Equipment

     6.7.1  Volumetric flasks — 2 mL, 5 mL,  10 mL, 25 mL,
            50 mL, 100 mL, 200 mL, and 1 L with ground glass
            stoppers.

     6.7.2  Microsyringes — various standard sizes.

     6.7.3  Boiling chips — approximately 10/40 mesh.   Heat
            at 400 C for  30 min.  or extract with methylene
            chloride in a Soxhlet apparatus.
                          B-7

-------
     6.7.4  Water bath — heated, with concentric ring
            cover, capable of temperature control within
            ±2°C.

     6.7.5  Analytical balance — capable of accurately
            weighing to 0.0001 g.

     6.7.6  Mixing device — Magnetic stir plates and stir
            bars, stirrers, tumbler, or separatory funnel
            shake (optional).

Reagents and Consumable Materials

7.1  Solvents.  High purity, distilled-in-glass hexane and
     methylene chloride.  For precise injections with
     splitless injectors and capillary columns, all samples
     and standards should be contained in the same solvent.
     Effects of minor variations in solvent composition
     (i.e., small percentage of another solvent remaining in
     hexane extracts) are minimized with the use of internal
     standards.   (External standard calibration is not
     acceptable).

7.2  Sodium Sulfate.  ACS, granular, anhydrous.  Purify by
     heating at 400°C for 4 h in a  shallow tray.

7.3  Sulfuric Acid.   AR Select, Mallinckrodt or equivalent
     guality.

7.4  Celite 545.  Fisher Scientific.

7.5  Silica Gel.  60-200 Mesh, Soxhlet extracted with
     hexane/methylene chloride (1:1), dried, stored at 110°C
     for 48 h before use.

7.6  Glass Wool.  Purify by heating at 400°C for 24 h.

7.7  Solid Phase Extraction Columns.  1 mL, 3 mL low density
     (LD), 6 mL high capacity (HC), J.T. Baker.

7.8  Internal Standard Spiking Solution.  200 ppm solution
     (in hexane) containing d12-chrysene.

7.9  Nitrogen Gas.  High purity, dry.

7.10 Surrogate Standard Spiking Solution.  1 ppm solution
     (acetone) containing d10-biphenyl, 13C6-1,2,4,5-
     tetrachlorobenzene, and 13C6-hexachlorobenzene.
                          B-8

-------
     7.11 HPLC Calibration Solution.  10 ppm solution (in hexane)
          containing benzene, bromobenzene, biphenyl, bibenzyl,
          p,p'-DDE, and 2,2',4,5,5'-pentachlorobiphenyl.

     7.12 HPLC Fractionation Standard.  10 ppm solution (in
          hexane) containing 1,4-dichlorobenzene, 1,3-
          diethylbenzene, p,p'-DDE,  and decachlorobiphenyl.

     7.13 MS Performance Check Solution.  10 ppm solution (in
          hexane) containing decafluorotriphenyl-phosphine
          (DFTPP).

     7.14 GC/MS Calibration Solutions.  Five hexane solutions are
          required.  These solutions contain constant
          concentrations of the internal standard, d12-chrysene
          and varying concentrations of the surrogates.
          Composition and approximate concentrations are given in
          Table B-2.

     7.15 GC Performance Solution.  10 ppm solution (in hexane)
          containing 0-BHC,  -BHC, d12-chrysene, and endrin ketone,
          or a 10 ppm solution (in hexane) containing anthracene,
          phenanthrene,  benz[a]anthracene, and chrysene.

     7.16 HPLC Performance Solution.  10 ppm solution (in hexane)
          containing biphenyl, 1,3-diethylbenzene, and bibenzyl.

8.   Sample Collection,  Preservation and Storage

     8.1  Water Samples

          8.1.1  Samples must be collected in clean (Section 4.2)
                 glass containers.

          8.1.2  Samples must be iced or refrigerated at 4°C  from
                 time of collection until extraction.  If samples
                 will not be extracted within 72 h after collection,
                 use either sodium hydroxide or sulfuric acid to
                 adjust sample pH to within a range of 5 to 9.
                 Record the volume of acid or base used.

          8.1.3  Samples should be extracted within 7 days after
                 collection and analyzed within 40 days after
                 extraction.

9.   Calibration

     9.1  HPLC Chromatography Conditions

          9.1.1  Initial HPLC Chromatography Conditions — Fill
                 solvent reservoirs with HPLC grade water and

                               B-9

-------
            acetonitrile, set the flow rate to 1.0 mL/rnin, and
            set the UV detector at 200 run.  Solvent conditions
            should be isocratic for 1.0 minutes initially with
            80% acetonitrile/ 20% water, linearly programmed to
            100% acetonitrile at a rate of 1.5%/min., and then,
            isocratic for 30 minutes with 100% acetonitrile.  A
            10 minute recycle time should be used between runs
            with the initial isocratic solvent conditions.
     9.1.2  Inject 10 nL of the HPLC calibration solution.  RTs
            should be within 20% of those shown in Table B-3.
            If RTs are not within 20%, adjust either or both
            the gradient program and/or flow rates to obtain
            the desired RTs.

9.2  HPLC Performance Criteria

     9.2.1  RT — log P conditions

            9.2.1.1  Using finalized HPLC conditions (see
                     9.1.2), inject 10 jUL of the HPLC
                     calibration solution and record RTs of
                     each compound .

            9.2.1.2  Using finalized HPLC conditions (see
                     9.1.2), inject 10 /LtL of acetone and record
                     its RT.

            9.2.1.3  Determine corrected RTs (i.e., RT of each
                     compound minus the RT of acetone) .

            9.2.1.4  Perform a regression analysis using the
                     log P values in Table B-3 and the
                     corrected RTs.  An equation of the form:
                     log P = A * log tc +  B   where A and B are
                     regression coefficients and tc is the
                     corrected RT, should be used.

            9.2.1.5  The correlation coefficient for the
                     regression, r2,  should  be  greater than
                     0.95.  If this condition is not met, HPLC
                     chromatography conditions are not correct.
                     Return to 9.1.2 and determine improved
                     conditions.

     9.2.2  Resolution

            9.2.2.1  Using the finalized HPLC conditions  (see
                     9.1.2), inject 10 /xL of the HPLC
                     performance solution.  Baseline separation
                         B-10

-------
                     between biphenyl, 1,3-diethylbenzene, and
                     bibenzyl should exist.

            9.2.2.2  .If baseline separation between all three
                     chemicals does not exist, either replace
                     the chromatpgraphy column and return to
                     9 . 1 or return to 9.1.2 and determine
                     improved conditions.

9.3  HPLC Fraction Time (Collection Time) Identification.

     9.3.1  Inject 10 /iL of the HPLC fractionation standard
            using the finalized chromatographic conditions and
            record RTs of 1,4-dichlorobenzene, 1,3-
            diethylbenzene, p,p'-DDE, and decachlorobiphenyl .
            These times are used for fraction collection from
            the HPLC.

     9.3.2  RT reproducibility — For each compound in the HPLC
            fractionation standard, the absolute RTs should not
            vary by more than ±0.10 minutes from one analysis
            to the next.

9.4  GC/MS Conditions.

     9.4.1  Recommended gas chromatography conditions

            Column Type:        DB-5
            Film Thickness:     0.25 /im
            Column Dimensions:  30 m X 0.32 mm or 30 m X
                                0.25 mm
            Helium Linear
              Velocity:         30 cm/sec @ 250°C
            Temperature
              Program:          Inject 50°C, hold 4  mins.,
                                increase to 175°C at 10°C/min,
                                increase to 275°C at 5°C/m,
                                hold at 275°C  for 20 mins.
            Injection Volume:    1 or 2
     9.4.2  Recommended acquisition conditions for mass
            spectrometer .

            Mass Range:         45-545 m/z
            Total Cycle Time
              per Scan:         0.5 < cycle time < 1.5 seconds

     9.4.3  Mass Spectrometer Calibration

            9.4.3.1  Calibrate and tune MS with standards and
                     procedures prescribed by the manufacturer.

                         B-ll

-------
            9.4.3.2  Inject 1 /iL or 2 juL aliquot of MS
                     performance check solution.  If spectrum
                     does not meet criteria for DFTPP  (Table
                     B-4); return to 9.4.3.1 and
                     recalibrate/tune MS.

9.5  GC/MS Performance Criteria

     9.5.1  GC Performance — Inject 1.0 /itL of the GC
            performance solution.  Baseline separation between
            /3-BHC and  -BHC and between endrin ketone and d12-
            chrysene should exist.  Alternatively, anthracene
            and phenanthrene should be separated by baseline
            and benz[a]anthracene and chrysene should be
            separated by a valley whose height is less than 25%
            of the average peak height of these two components.

     9.5.2  MS Sensitivity — Inject 1.0 juL of the 0.5 ppm
            GC/MS calibration solution.  Using the total ion
            chromatogram, a signal to noise ratio of greater
            than 3 should be observed for each surrogate.

     9.5.3  MS Calibration — Inject 1.0 /^L of the 0.50 ppm
            GC/MS calibration solution.  For d12-chrysene,
            abundance of m/z 241 relative to that of m/z 240
            should be >15% and <25%.

     9.5.4  GC Stability — Perform multiple GC/MS analyses on
            the same GC/MS calibration solution.  RTs should
            not vary by more than t seconds.  Calculate the
            value of t with the equation, t = (RT)1'3,  where RT
            is the observed average RT (in seconds).

9.6  Response Factor Calculation for MS

     9.6.1  Inject 1.0 /uL of each GC/MS calibration solution
            and acquire GC/MS data.

     9.6.2  Calculate response factors (RF) for each surrogate
            relative to d12-chrysene

                      RF = ASQC/ACQS

            where  As = integrated total  ion abundance  for  the
                         surrogate
                   Ac = integrated total  ion abundance  for  the
                         internal standard, d12-chrysene
                   Qs = injected  quantity of surrogate
                   Qc = injected  quantity of d12-chrysene
                          B-12

-------
     9.6.3  RF Reproducibility — For each surrogate, calculate
            the mean RF.  When the relative standard deviation
            (RSD) exceeds 30%, analyze additional aliquots of
            GC/MS calibration solutions to obtain acceptable
            RSD for the RF, or take action to improve GC/MS
            performance.

9.7  Continuing Calibration Check

     9.7.1  HPLC

            9.7.1.1  With the following procedures, verify HPLC
                     column performance at the beginning and
                     end of each 12 h period during which
                     analyses are performed.

            9.7.1.2  Demonstrate acceptable performance for
                     criteria described in Section 9.2.2.

            9.7.1.3  Demonstrate acceptable performance for
                     criteria described in Section 9.3.2.

     9.7.2  GC/MS

            9.7.2.1  With the following procedures, verify
                     initial calibration at beginning and end
                     of each 12 h period during which analyses
                     are performed.

            9.7.2.2  Inject 1 or 2 jiL aliquot of MS performance
                     check solution.  Ensure acceptable MS
                     calibration and performance.

            9.7.2.3  Demonstrate acceptable performance for
                     9.5.

            9.7.2.4  Determine the area for chrysene-d12 has
                     not changed by more than 30% from most
                     recent analyses of the GC/MS calibration
                     solutions.

            9.7.2.5  For an acceptable continuing calibration
                     check, the measured RF for each surrogate
                     must be within 30% of the mean value
                     calculated during initial calibration.

     9.7.3 Remedial Actions

            Remedial actions must be taken if criteria are not
     met; possible remedies are:
                         B-13

-------
                 9.7.3.1  Check and adjust instrumental operating
                          conditions.

                 9.7.3.2  Clean or replace injector liner on GC.

                 9.7.3.3  Flush column with solvent according to
                          manufacturers instructions.

                 9.7.3.4  Break off a  short portion (approximately
                          0.33 m)  of the column; check column
                          performance  by analysis of performance
                          check solution for GC.

                 9.7.3.5  Replace column; performance of all initial
                          calibration  procedures then required.

                 9.7.3.6  Adjust MS for greater or lesser
                          resolution.

                 9.7.3.7  Calibrate MS mass scale.

                 9.7.3.8  Prepare and  analyze new concentration
                          calibration/performance check solution.

                 9.7.3.9  Prepare new  concentration calibration
                          curve(s).

10.   Procedures

     10.1 Extraction (Semi-automated)

          10.1.1 Mix effluent.  Place  10 L of effluent into
                 extraction bottle(s).  Depending upon your
                 laboratory setup,  one 10 L bottle, two 5 L bottles,
                 or five 2 L bottles will be required.

          10.1.2 Add 1 mL of surrogate solution to the effluent.
                 This 1 mL volume of solution must be equally
                 divided among all bottles.  Seal bottles and mix
                 effluent for 15 minutes.

          10.1.3 Open bottles and add  60 mL of hexane per liter of
                 effluent to each extraction bottle.  Seal
                 bottle(s).

          10.1.4 Place bottles onto mixing device, i.e., shaker,
                 tumbler, or stirring  apparatus.  Shake, tumble, or
                 mix the sample for a  minimum of 30 minutes.  Longer
                 mixing times might be required for your apparatus
                 to ensure that quantitative extraction is obtained.


                              B-14

-------
     10.1.5 Remove sample containers from mixing device.  Allow
            hexane layer to separate from water phase for 10-15
            minutes.  If emulsions are greater than one third
            of the solvent layer, use mechanical technigues to
            complete the phase separation.  Optimum technigue
            depends upon sample, but may include sonication,
            centrifugation, filtration through glass wool, or
            physical methods.  Collect the hexane in a clean
            flask.

     10.1.6 Repeat steps 10.1.3-5 two additional times and
            combine hexane extraction solvent.

     10.1.7 Dry extract by passing it through a drying column
            containing about 10 cm of anhydrous sodium sulfate.
            Collect the dried extract in a Kuderna-Danish
            concentrator flask.  Rinse flask which contained
            the solvent extract with 20-30 ml of hexane and add
            it to the column to complete the guantitative
            transfer.  Optional, see 10.1.9.

     10.1.8 Add one or two clean boiling chips to the flask and
            attach a three ball Snyder column to Kuderna-Danish
            apparatus.  Concentrate the extract on a steam bath
            until volume of the extract is less than 8 mL.
            Allow Kuderna-Danish apparatus to cool and detach
            the lower tube, rinsing the joint with hexane into
            the sample.  Optional, see 10.1.9.

     10.1.9 Note, steps 10.1.7, sample drying, and 10.1.8,
            sample concentration, are optional.  The hexane
            extracts resulting from 10.1.6 may be taken
            directly to 10.2, sample cleanup.  If 10.1.7 and
            10.1.8 are not performed, flasks for collecting and
            concentrating the eluate from the acid/Celite
            columns must be larger than what is written in 10.2
            to accommodate the increased volume of hexane.

10.2 Sample Cleanup

     10.2.1 Construct column containing (bottom to top)  glass
            wool, silica gel (2 g),  sodium sulfate (2 g), 70%
            sulfuric acid solution (5 ml)  on Celite (10 g),  and
            sodium sulfate (2 g).  The column is washed with
            100 mL of hexane is not allowed to go dry.

     10.2.2 Quantitatively transfer the hexane extract from
            10.1.7 to the column using hexane solvent.  Collect
            eluate into a 500 mL Kuderna-Danish evaporation
            flask with a 10 mL lower tube.
                         B-15

-------
     10.2.3 Wash the acid/Celite column with 100 mL hexane and
            collect the eluate in the same Kuderna-Danish
            flask.

     10.2.4 Add one or two clean boiling chips to the flask and
            attach a three ball Snyder column to Kuderna-Danish
            apparatus.  Concentrate the extract on a steam bath
            until volume of the extract is less than 8 mL.
            Allow Kuderna-Danish apparatus to cool and detach
            the lower tube, rinsing the joint with hexane into
            the sample.

     10.2.5 Use a nitrogen stream and a warm (<40°C)  water bath
            to concentrate the extract to a volume of 0.50 mL.

     10.2.6 Quantitatively transfer extract to a HPLC
            autosampler vial.  Cap and store extract in freezer.

10.3  Fractionation of Effluent Extract

     10.3.1 Remove sample from freezer and allow it to warm to
            ambient temperature.  Uncap the vial and allow the
            sample to evaporate naturally to the volume of
            500 /nL.  Recap the vial and mix well.

     10.3.2 Using the HPLC conditions determined previously
            (see Section 9.1), inject 175 juL of the sample into
            the HPLC column.

     10.3.3 At the RT of 1,4-dichlorobenzene (determined in
            Section 9.3), place a clean flask under the waste
            tube from the UV detector and collect the column
            eluate.  Label this fraction as Fraction 1.

     10.3.4 At the RT of 1,3-diethylbenzene, place a clean
            flask under the waste tube from the UV detector and
            collect the column eluate.  Label this fraction as
            Fraction 2.

     10.3.5 At the RT of p,p'-DDE,  place a clean flask under
            the waste tube from the UV detector and collect the
            column eluate.  Collection should continue until
            the RT of decachlorobiphenyl.  Label this fraction
            as Fraction 3.

     10.3.6 Repeat steps 10.3.2 to 10.3.5 two more times so
            that all the sample has been injected onto the HPLC
            column.  Combine the like fractions into the
            flasks.
                         B-16

-------
10.4 Concentration of Effluent Extracts (Liquid/Liquid
     Extraction)

     10.4.1 Quantitatively transfer each HPLC fraction into its
            own extraction bottle, e.g., separatory funnels,
            bottles, etc.  Dilute fractions a minimum of 10
            fold with HPLC grade water.  Seal bottle and mix
            diluted fractions thoroughly.

     10.4.2 Open bottles and add 60 mL of hexane to each
            diluted fraction.  Seal bottles.

     10.4.3 Place bottles onto mixing device, i.e., shaker,
            tumbler, or stirring apparatus.  Shake, tumble, or
            mix the sample for a minimum of 30 minutes.
            Alternatively, separatory funnel extraction by hand
            with periodic venting may be used; shake for
            minimum of 2 minutes

     10.4.4 Remove bottles containing the diluted fractions
            from mixing device.  Allow hexane layer to separate
            from water phase to 10-15 minutes.  Collect the
            hexane in a clean flask.  Each HPLC fraction must
            have its own flask.

     10.4.5 Repeat steps 10.4.2-4 two additional times and
            combine the hexane from these extractions with the
            hexane from the first extraction.

     10.4.6 Dry each extract by passing it through a drying
            column containing about 10 cm of anhydrous sodium
            sulfate.  Collect the dried extract in a Kuderna-
            Danish concentrator flask.  Rinse flask which
            contained the solvent extract with 20-30 mL of
            hexane and add it to the column to complete the
            quantitative transfer.

     10.4.7 Add one or two clean boiling chips to the flask and
            attach a three ball Snyder column to Kuderna-Danish
            apparatus.  Concentrate the extract on a steam bath
            until volume of the extract is less than 8 mL.
            Allow Kuderna-Danish apparatus to cool and detach
            the lower tube, rinsing the joint with hexane into
            the sample.  Three extracts for each sample will
            results, i.e., fractions l, 2, and 3.

     10.4.8 Using a stream of dry clean air, evaporate extracts
            to approximately 1 mL.
                         B-17

-------
     10.4.9 Quantitatively transfer and concentrate extracts
            for each fraction to 100 p.L in microvials.  Cap and
            store in freezer.

10.5 Concentration of Effluent Extracts (C18SPE)

     10.5.1 Dilute fractions a minimum of 10 fold with HPLC
            grade water and mix diluted fractions thoroughly.

     10.5.2 Activate three C18 solid  phase  extraction (SPE)
            columns by passing methanol and then HPLC grade
            water through the column as specified by the
            manufacture.  The size of the column required is
            dependent upon the volume of diluted HPLC
            fractions; use manufacturer's recommendations.  In
            general, 1 mL, 3 mL LD,  and 3 mL columns will be
            needed.

     10.5.3 Pass diluted extracts through the C18 SPE columns.
            Do not exceed maximum flowrates recommended by
            manufacture .

     10.5.4 Remove excess water for the column.  This can be
            done using a slight positive pressure of clean air
            or nitrogen gas, a slight vacuum, or (spun out
            using) a centrifuge.  Do not let the columns go
            dry.
     10.5.5 Elute C18SPE  columns with  three  500  ML aliquots of
            methylene chloride and then one 500 j^L aliquot of
            methylene chloride.

     10.5.6 Dry extracts by passing C18SPE eluent  through micro
            Na2SO4 columns.  Micro drying columns may be
            prepared by placing 1 cm of Na2SO4 into an empty
            1 mL SPE column and passing approximately 2 ml of
            hexane through the column.

     10.5.7 Concentrate eluate from drying column to
            approximately 100 /xL and quantitatively transfer
            concentrate to microvials.  Cap and store in
            freezer.  Three extracts for each sample will be
            obtained.

10.6 GC/MS Analysis

     10.6.1 Remove sample from freezer and allow it to warm to
            ambient temperature.   Adjust volume of sample to
            100 j^L and then spike sample with 5 /uL of the
            internal standard spiking solution,  1000 ng of d12-


                          B-18

-------
            chrysene.  Mix sample after spiking sample with
            d12-chrysene.

     10.6.2 Inject a 1 juL or 2 juL aliquot of blank or sample
            extract into GC operated under conditions used to
            produce acceptable results during calibration.

     10.6.3 Acquire MS data using full scan conditions (see
            Section 9.4).

     10.6.4 Recap samples and store extracts in freezer.   Note,
            autosampler crimp caps after being punctured by the
            syringe used for sample injection must be replaced.

10.7 GC/MS Data Analysis

     Data analysis for this analytical procedure can be divided
into three tasks, l) quantification of unknowns and surrogates,
2) library searching with the CHC mass spectral library,  and 3)
library searching with the EPA/NIH/NBS mass spectral library.
This process is performed on the sample and blank extracts.


     10.7.1 Quantification of Surrogates and Other GC/MS
            Components

            10.7.1.1 Surrogate and Internal Standard
                     Identification

            Identify surrogates and internal standards by
            comparison of their mass spectrum (after background
            subtraction) to reference spectrums in the user-
            created data base.  The GC retention time of the
            sample component should be within 10 sec of the
            time observed for that same compound when a
            calibration solution was analyzed.  In general, all
            ions that are present above 10% relative abundance
            in the mass spectrum of the standard should be
            present in the mass spectrum of the sample
            component and should agree within absolute 20%.
            For example, if an ion has a relative abundance of
            30% in the standard spectrum,  its abundance in the
            sample spectrum should be in the range of 10 to
            50%.  Some ions, particularly the molecular ion,
            are of special importance, and should be evaluated
            even if they are below 10% relative abundance.

            10.7.1.2 Peak Integration

            For each fraction, use the GC/MS peak detection and
            integration software to obtain areas for all

                         B-19

-------
chromatographic components from the total ion
chromatogram with a signal to noise of 3 and
greater.  (Note, the solvent front need not be
examined.)  Verify that the surrogate in each
fraction was integrated.  If surrogate was not
integrated use more sensitive peak detection and
integration settings to obtain peak areas for the
surrogates.

10.7.1.3 Calculate Surrogate Concentrations

    cx = (AX-QIS)/(AIS-RF-V)

Where   Cx  = Concentration of  surrogate

        Ax  = Area  of  surrogate in total
             ion chromatogram

        QIS = Quantity of internal standard added
             to extract before GC/MS analysis

        A|S = Area of internal standard in total
             ion chromatogram added to extract
             before GC/MS analysis

        RF  = Mean response factor of surrogate
             from initial calibration analysis
             and/or from GC/MS calibration
             solutions run with sample analyses.

        V   = Volume of extracted water, 10 L.

Alternatively, use the GC/MS system software or
other available proven software to complete the
concentrations of surrogate using linear, second,
or third order regression or using piecewise
calibration curves.

10.7.1.4 Calculate Surrogate Recoveries

Surrogate Recovery = (Cx •  100)/SSC

Where    Cx       = Concentration  of surrogate
         SSC     = Surrogate spiking
                   concentration,  i.e.,  100 ng/L

10.7.1.5 Calculate concentrations of all
         chromatographic components

         cx  = (AX * QIS * 10°)/(Ais  • RF • V • REC)
              B-20

-------
Where    Cx  =   Concentration  of  chromatographic
                 component

         Ax  =   Area of  chromatographic component
                 in total ion chromatogram

         QIS =    Quantity of internal  standard
                 added to extract  before GC/MS
                 analysis

         A|S =    Area of  internal  standard in
                 total ion  chromatogram

         V   =   Volume of  extracted water,  i.e.
                 10 L

         RF  =   Mean response  factor  of surrogate
                 from initial calibration analyses
                 and/or from GC/MS calibration
                 solutions, run with sample
                 analyses.  For fraction #1,  use
                 RF for d10-biphenyl.   For fraction
                 #2, use  RF for 13C6-1,2,4,5-
                 tetrachlorobenzene.   For fraction
                 #3, use  RF for 13C6-
                 hexachlorobenzene.

         REC =   Surrogate  recovery calculated in
                 Section  10.4.1.4.  For fraction
                 #1, use  recovery  of d10-biphenyl.
                 For fraction #2,  use  recovery of
                 13C6-1,2,4,5-tetrachlorobenzene.
                 For fraction #3,  use  recovery of
                 13C6-hexachlorobenzene.

Alternatively, use the GC/MS system software  or
other available proven software to compute the
concentration of all chromatographic components
using linear, second, or  third order regression or
using piecewise calibration curves.  These
calculations must use the response curve for  the
surrogate appropriate for the fraction of interest
and the reported concentrations must be corrected
for recovery of the surrogate.  For fraction  #1,
use response curve and recovery of d10-biphenyl.
For fraction #2, use response curve and recovery  of
13C6-1,2,4,5-tetrachlorobenzene.   For fraction  #3,
use response curve and recovery of 13C6-
hexachlorobenzenes.
              B-21

-------
10.7.2 Library searching with the CHC mass spectral
       library

       10.7.2.1 Algorithm Selection

       A reverse searching algorithm is required when
       available.   If GC/MS system does not have a reverse
       searching algorithm, library searching should be
       performed using the default algorithm supplied by
       the manufacture of the instrument.

       10.7.2.2 Searching

       All chromatographic components detected in 10.7.1.2
       (Peak Integration)  are searched against the CHC
       mass spectral library.  Those chemicals with fits
       of 70% and greater are considered tentatively
       identified.  For each tentatively identified
       component,  a list of the best mass spectral library
       identifications (up to a total of ten
       identifications)  is reported along with the percent
       fit values, CAS number of each tentative
       identification, HPLC fraction number,  GC retention
       time, and the concentration for the GC/MS
       component.   This report is called Report 1.

10.7.3 Library Searching with EPA/NIH/NBS Mass Spectral
       Library.

       10.7.3.1 Algorithm Selection

       A reverse searching algorithm is required when
       available.   If GC/MS system does not have a reverse
       searching algorithm, library searching should be
       performed using the default algorithm supplied by
       the manufacture of the instrument.

       10.7.3.2 Eliminating of Compounds Below 100 ng/L

       All unidentified components from the CHC library
       search with concentrations less than 100 ng/L are
       eliminated from further data processing.  This
       elimination can be performed by comparing peak
       areas or heights of each chromatographic peak to
       the peak area or height of the surrogate
       appropriate for that fraction.  This elimination
       may also be performed by comparing the recovery
       corrected concentration of each chromatographic
       component to 100 ng/L.
                    B-22

-------
10.7.3.3 Searching

For those GC/MS components not identified with the
CHC search with effluent concentrations greater
than or equal to 100 ng/1, these components are
reversed-searched against the EPA/NIH/NBS mass
spectral library.  Those chemicals with fits of 70%
and greater are considered tentatively identified.
These components are then further evaluated, see
10.7.3.4.

For those components with fits/matches less than
70% but greater than 25%, the two best mass
spectral library identifications along with the
percent fit values, the CAS number of each
tentative identification, HPLC fraction number, GC
retention time, and the concentration are reported
for each the GC/MS component.  For GC/MS components
with fits/matches less than 25%, the concentration,
HPLC fraction number and GC retention time for each
component is reported and the components labeled as
being "unknown".  This report is identified as
Report 3.

10.7.3.4 Segregating GC/MS Components Tentatively
         Identified Components Using the
         EPA/NIH/NBS Mass Spectral Library Search.

For HPLC fractions 1, 2, and 3, predicted tissue
concentrations are calculated for the effluent.
Predicted tissue concentrations are determined by
calculating the product of fraction BCF, effluent
concentration, and effluent dilution.  For
fractions 1, 2, and 3, fraction BCF values of 560,
5000, and 470000 are used.  The effluent dilution
value is provided by the regulatory authority and
the effluent concentration calculated in section
10.7.1.3.

For those GC/MS components with fits of 70% and
greater and a predicted tissue concentration > 1
ug/kg, a list of the best mass spectral library
identifications (up to a total of ten
identifications)  is reported along with the percent
fit values,  CAS number of each tentative
identification, HPLC fraction number, GC retention
time, and the concentration for the GC/MS
component.   This report is called Report 2.

For those GC/MS components with fits of 70% and
greater and a predicted tissue concentration less

              B-23

-------
            than 1 ug/kg, a list of the best mass spectral
            library identifications (up to a total of ten
            identifications)  is reported along with the percent
            fit values, CAS number of each tentative
            identification, HPLC fraction number, GC retention
            time, and the concentration for the GC/MS
            component.  This report is identified as Report 4.

     10.7.4 Elimination of chromatographic components common to
            the spiked blank and effluent GC/MS data.

            Chromatographic components with retention times
            within ten seconds between the spiked blank and
            effluent and with the same mass spectrums should be
            removed from the analysis.

10.8 Reporting of Data

     10.8.1 Report 1:  CHC mass spectral identifications.

            For each chromatographic component tentatively
            identified using the CHC search (fits greater to
            and greater than 70%),  a list of the best mass
            spectral library identifications (up to a total of
            ten identifications)  is reported along with the
            percent fit values, CAS number of each tentative
            identification, HPLC fraction number, GC retention
            time, and the concentration for the GC/MS
            component.

     10.8.2 Report 2:  EPA/NIH/NBS mass spectral tentative
            identifications (fits > 70%, > 100 ng/1, and
            predicted tissue concentration > 1 ug/kg).

            GC/MS components tentatively identified in the
            EPA/NIH/NBS mass spectral search and a predicted
            tissue concentration > 1 ug/kg.  For each
            tentatively identified component,  a list of the
            best mass spectral library identifications (up to a
            total of ten identifications)  is reported along
            with the percent fit values, CAS number of each
            tentative identification,  HPLC fraction number, GC
            retention time, and the concentration for the GC/MS
            component.

     10.8.3 Report 3:  EPA/NIH/NBS mass spectral tentative
            identifications (fits < 70% and > 100 ng/1).

            GC/MS components not tentatively identified in the
            EPA/NIH/NBS mass spectral search.   For those
            components with fits less than 70% and greater than

                         B-24

-------
       25%, the two best mass spectral library
       identifications along with the percent fit values,
       the CAS number of each tentative identification,
       HPLC fraction number, GC retention time, and the
       concentration are reported for each the GC/MS
       component.  For GC/MS components with fits/matches
       less than 25%, the concentration, HPLC fraction
       number, and GC retention time for each component is
       reported and the components labeled as being
       "unknown".

10.8.4 Report 4:  EPA/NIH/NBS mass spectral tentative
       identifications (fits > 70%, > 100 ng/1, and
       predicted tissue concentration < I ug/kg).

       GC/MS components tentatively identified in the
       EPA/NIH/NBS mass spectral search and a predicted
       tissue concentration < 1 ug/kg.  For each
       tentatively identified component, a list of the
       best mass spectral library identifications (up to a
       total of ten identifications) is reported along
       with the percent fit values, CAS number of each
       tentative identification, HPLC fraction number, GC
       retention time, and the concentration for the GC/MS
       component.

10.8.5 QA/QC Report

       10.8.5.1 Recoveries.

       For d10-biphenyl, 13C6-l,2,4,5-tetrachlorobenzene,
       and 13C6-hexachlorobenzene,  recoveries are reported
       for each sample and blank HPLC fraction.

       10.8.5.2 GC/MS Chromatograms.

       For the sample and its corresponding blank, total
       ion Chromatograms must be provided for each sample
       and blank HPLC fraction.

       10.8.5.4 QA/QC.

       Chromatogram demonstrating HPLC resolution,
            see Section 9.2.2.1
       Data demonstrating GC resolution, see
            Section 9.5.1
       Data demonstrating MS sensitivity, see
            Section 9.5.2
       Data demonstrating MS calibration, see
            Section 9.5.3
                    B-25

-------
                 Data demonstrating DFTPP performance,
                      see Section 9.4.3.2
                 Data demonstrating precision, see
                      see Section 11.5.

11.  Quality Control

     11.1 Recoveries of Surrogates.  Method:

           % Recovery = measured surrogate amount  X 100
                         spiked surrogate amount

          Quality Assurance Requirement

                     25% < % recovery < 120%
          Quality Control Action.  If percent recovery is out of
          range, re-extract and re-analyze sample.

     11.2 Surrogate Fraction Location.

          Quality Assurance Requirement

          Fraction 1      should contain d10-biphenyl and should not
                          contain 13C6-1,2,4,5-tetrachlorobenzene or
                          13C6-hexachlorobenzene

          Fraction 2      should contain 13C6-1,2,4,5-
                          tetrachlorobenzene and should not contain
                          d10-biphenyl or 13C6-hexachlorobenzene

          Fraction 3      should contain 13C6-hexachlorobenzene and
                          should not contain 13C6-l,2,4,5-
                          tetrachlorobenzene or d10-biphenyl

          Quality Control Action.  If surrogate is in wrong
          fraction, take corrective action for improper HPLC
          conditions.  Re-extract and re-analyze the sample.

     11.3 Calibration Stability

          11.3.1 Continuing Calibration Checks

                 Quality Assurance.  See Sections 9.7

                 Quality Control Action.  Re-extract and re-analyze
                 samples.  For corrective actions see Section 9.7.
                               B-26

-------
     11.3.2 HPLC Stability.

            Quality Assurance.  Immediately before and after
            fractionation solution must be analyzed.  See
            Section 9.3 for stability requirements.

            Quality Control Action.  If RT stability
            requirements are not met, re-extract and re-analyze
            the sample.

     11.3.3 GC/MS Stability

            Quality Assurance.  During the GC/MS analysis run,
            a GC/MS calibration solution should be analyzed
            twice; once at the beginning and once at the end of
            the GC/MS analysis sample sequence.  See Sections
            9.5 and 9.7 for stability requirements.

            Quality Control Action.  If RT and/or MS
            sensitivity stability requirements are not met,
            repeat GC/MS analysis on samples after correction
            of instrumental problems.

11.4 Blanks

     11.4.1 Spiked Blanks

            Quality Control.  A 10 L (preferred) or a 2 L
            sample of reagent water is analyzed using the
            analytical procedure (Section 10).   The sample is
            spiked with the surrogates using the same amount of
            compound, i.e., 1000 ngs, as the effluent samples.

            One spiked blank should be performed with every set
            of effluent samples analyzed.

            An acceptable spiked blank:

              a)   Meets Sections 11.1 and 11.2  requirements.
              b)   Contains no compound with elution charac-
                    teristics and mass spectral features that
                    would interfere with identification and
                    quantification of the surrogates.
              c)   Contains few chromatographic peaks in the
                    GC/MS total ion chromatograms for each
                    fraction.

            Quality Control Action.  Locate and eliminate the
            source of contamination.  Re-extract and re-analyze
            the entire batch of sample.
                         B-27

-------
     11.4.2 Unspiked Blanks

            Quality Control.  A 10 L (preferred)  or a 2 L
            sample of reagent water is analyzed using the
            analytical procedure (Section 10).   The sample is
            NOT spiked with the surrogates!  It is spiked with
            the internal standards.  One unspiked blank should
            be performed with every 20 effluent samples.

            An acceptable unspiked blank.  The sample should
            not contain detectable amounts of the surrogates.

            Quality Control Action.  Locate and eliminate
            source of contamination, re-extract and re-analyze
            entire batch of sample.

11.5 Precision

     Quality Control.  A duplicate set of effluent samples will
     be analyzed with every set of 10 samples.   To measure the
     precision, the relative percent difference between the lab
     duplicate will be determined for each surrogate.

          Relative % difference =

         surrogate amount        surrogate amount
           duplicate 1	-	duplicate 2	   X 100
                  Average surrogate amount

     The relative percent difference should be less than 150%.

     Quality Control Action.  If relative percent difference is
     out-range for any of three surrogates, re-extract and re-
     analyze the sample.

11.6 Sample Sets.

     Sample sets are defined as a group of samples that are
     carried through the analytical procedure at the same time.
     Each sample set will include a minimum of two QC samples,
     one of them being a spiked blank.   QC samples include
     spiked blanks, unspiked blanks, and replicates.

     A set will normally contain 10 samples plus the required
     QC samples.

11.7 GC/MS Analysis Sets.

     A GC/MS analysis set is defined as a group of prepared
     samples (effluents and QC samples) and GC/MS calibration
                         B-28

-------
          solutions analyzed during one GC/MS run.  A set will
          normally contain 12 prepared samples,  all five GC/MS
          calibration solutions,  and 1 or 2 duplicates of the GC/MS
          calibration solutions.   The duplicate GC/MS calibration
          solutions will be analyzed in the beginning and at the end
          of GC/MS sample sequence.

12.   Appendix B References.

     12.1 "Carcinogens — Working with Carcinogens", Department of
          Health Service, Center for Disease Control, National
          Institute for Occupational Safety and Health, Publication
          No. 77-206, August 1977.

     12.2 "OSHA Safety and Health Standards, General Industry", 29
          CFR 1910, Occupational Safety and Health Administration,
          OSHA 2206, Revised January 1976.

     12.3 "Safety in Academic Chemistry Laboratories", American
          Chemical Society Publication, Committee on Chemical
          Safety, 3rd Edition, 1979.

     12.4 Robert A. Hughes, Gilman D. Veith, and G. Fred Lee, "Gas
          Chromatographic Analysis of Toxaphene in Natural Waters,
          Fish and Lake Sediments."
                              B-29

-------
            TABLE B-l




CHEMICALS OF HIGHEST CONCERN LIST
CAS number
50-29-3
57-74-9
58-89-9
60-57-1
70-30-4
72-54-8
72-55-9
76-44-8
91-94-1
95-94-3
101-61-1
115-32-2
117-81-7
118-74-1
309-00-2
319-84-6
319-85-7
608-73-1
608-93-5
924-16-3
1024-57-3
1746-01-6
2104-64-5
2385-85-5
8001-35-2
39515-41-8
11096-82-5
11097-69-1
11104-28-2
11141-16-5
12672-29-6
12674-11-2
53469-21-9
chemical name
p,p'-dichlorodiphenyltrichloroethane (DDT)
chlordane
hexachlorocyclohexane (lindane)
dieldrin
hexachlorophene
p,p'-dichlorodiphenyldichloroethane (ODD)
p,p ' -dichlorodiphenyldichloroethylene (DDE)
heptachlor
3,3' -dichlorobenzidine
1,2,4, 5-tetrachlorobenzene
4 , 4 ' -methylene bis (N,N' -dimethyl) aniline
dicofol
bis ( 2 -ethy Ihexy 1 ) phtha late ( BEHP)
hexachlorobenzene
aldrin
alpha-hexachlorocyclohexane (alpha-HCH)
beta-hexachlorocyclohexane (beta-HCH)
technical-hexachlorocyclohexane (t-HCH)
pentachlorobenzene
N-nitroso-di-n-butylamine
heptachlor epoxide
dioxin (2, 3 , 7, 8-TCDD)
ethylp-nitrophenylpheny Iphosphorothioate ( EPN)
mirex
toxaphene
danitol
polychlorinated biphenyl 1260
polychlorinated biphenyl 1254
polychlorinated biphenyl 1221
polychlorinated biphenyl 1232
polychlorinated biphenyl 1248
polychlorinated biphenyl 1016
polychlorinated biphenyl 1242
             B-30

-------
    Table B-2.
GC/MS Calibration Solutions, Concentrations of
Surrogates and Internal Standards.
13
 Cfi-l,2,4,5-tetra-
                             13
                              C6-hexachloro-
chlorobenzene
ppm
0.5
1
10
50
100
d10-biphenyl
ppm
0.5
1
10
50
100
d12-chrysene
ppm
10
10
10
10
10
benzene
ppm
0.5
1
10
50
100

    Table B-3.
Suggested Retention Times (RTs) for Compounds in
the HPLC Calibration Solution.
Chemical
Benzene
Bromobenzene
Biphenyl
Bibenzyl
p , p ' -DDE
2 , 2 ' , 4 , 5 , 5 ' -Pentachlorobiphenyl
Retention
Time
4.3
5.5
6.8
8.1
13.4
15.1
Log P
2.13
2.99
3.76
4.81
5.69
6.11
                              B-31

-------
Table B-4.     DFTPP Ion Abundance Criteria.





     m/z            Criteria



      51            10-80% of the base peak



      68            <2% of m/z 69



      70            <2% of m/z 69



     127            10-80% of the base peak



     197            <2% of m/z 198



     198            base peak or >50% of 442



     199            5-9% of m/z 198



     275            10-60% of the base peak



     365            >1% of base peak



     441            present and 50% of m/z 198



     443            15-24% of m/Z 442
               B-32

-------
                           APPENDIX C






LABORATORY PROCEDURES FOR DETERMINING BIOCONCENTRATABLE CHEMICALS




                       IN SEDIMENT SAMPLES

-------
                            Appendix C


Sediment Analysis Procedure - Revision 1.0


1.  Scope and Application

     This method provides procedures for sample preparation using
reverse phase HPLC and procedures for identification and
quantification of bioconcentratable chemicals using GC/MS for
sediments.  This method is applicable to organic chemicals which
are stable under acidic conditions and can be chromatographed
using gas chromatography.

2.  Summary of Method

     20 grams of ground, air dried, sediment is mixed with
anhydrous sodium sulfate, spiked with a surrogate standard
mixture, Soxhlet extracted using acetone for 4 hours, and then,
Soxhlet extracted using 1:3 toluene:methanol for a minimum of 12
hours.  The extract is subsequently cleaned up using copper and
sulfuric acid, and then, concentrated to a volume of 0.50 mL.
The extract is chromatographed using reverse phase HPLC, and
three fractions are collected.  These fractions have nominal BCF
ranges of approximately 91 to 560, 560 to 5,000, and 5,000
to 470,000.  The fractions are extracted and concentrated to 0.10
mL, and are subsequently spiked with the internal standard.  The
fractions are analyzed using capillary gas chromatography with
full scan electron impact ionization mass spectrometry  (GC/MS).
After GC/MS analysis, the fractions are saved for confirmation
analysis and the peaks in the GC/MS data are identified and
quantified.  Also included is a procedure for determining the
percent moisture of the sediment.  However, a method for
determining the organic carbon content of the sediment not
provided even though reporting of this parameter is required.

     The three surrogate compounds, added to each sample before
extraction, are d10-biphenyl, 13C6-l,2,4,5-tetrachlorobenzene, and
1  C6-hexachlorobenzene and the internal standard is d12-chrysene.
Standard curves are calculated using an internal standard method
for each surrogate and subsequently, percent recovery for each
surrogate is determined.  All other GC/MS components are
quantified using the response factor appropriate for each HPLC
fraction:  d10-biphenyl/d12-chrysene  for the first  fraction,
  C6-l,2,4,5-tetrachlorobenzene/d12-chrysene for the second
fraction, and 13C6-hexachlorobenzene/d12-chrysene for the third
fraction.
     For each HPLC fraction, all chromatographic peaks are
reverse-searched against (compared with) the Chemicals of Highest

                               C-2

-------
Concern (CMC) mass spectral library (see Table J-l).  Those GC/MS
components with fits of 70% and greater are considered
tentatively identified.  For each tentatively identified
component, a list of the best mass spectral library
identifications (up to a total of ten identifications) is
reported along with the percent fit values, CAS number of each
tentative identification, HPLC fraction number, GC retention
time, and the concentration for the GC/MS component.  This report
is called Report 1.

     For those GC/MS components not identified with the CHC
search with sediment concentrations greater than or equal to 5
ug/kg, these components are reversed-searched against the
EPA/NIH/NBS mass spectral library.  Those GC/MS components with
fits of 70% and greater are considered tentatively identified.
For each tentatively identified component, a list of the best
mass spectral library identifications (up to a total of ten
identifications) is reported along with the percent fit values,
CAS number of each tentative identifications, HPLC fraction
number, GC retention time, and the concentration for the GC/MS
component.  This report is called Report 2.

     For those components with fits/matches less than 70% and
greater than 25%,  the two best mass spectral library
identifications along with the percent fit values, the CAS number
of each tentative identification, HPLC fraction number, GC
retention time, and the concentration are reported for each the
GC/MS component.  For GC/MS components with fits/matches less
than 25%,  the concentrations and GC retention times for these
components are reported and the components labeled as being
"unknown".  This report is identified as Report 3.

     This procedure yields four reports which will be sent to the
regulatory authority.  These reports are:

     1)   Report 1, components tentatively identified using the
          CHC mass spectral library.

     2)   Report 2, components tentatively identified with
          sediment concentrations greater than or equal to 5 ng/g
          using the EPA/NIH/NBS mass spectral library.

     3)   Report 3, components with sediment concentrations
          greater than or equal to 5 ng/g and fits less than 70%
          using the EPA/NIH/NBS mass spectral library.

     4)   QA/QC Report, recoveries of the three surrogate
          chemicals in the six sample and blank fractions,
          percent organic carbon, percent moisture, GC/MS
          chromatograms for the sample and blank, HPLC performace
          data, GC/MS performace data, and precision data.

                               C-3

-------
3.    Definitions

     3.1  Bioconcentration Factor (BCF).   Ratio of the
          concentration in the tissue of the organism to that in
          water for an individual chemical.  In equation form,

                          BCF = CF/CW

          where CF and Cw are the concentrations  in the  tissue
          and aqueous phase.

     3.2  Surrogate Compound.  A pure compound added to a sample
          before extraction.

     3.3  Internal Standard.  A pure compound added to a sample
          extract prior to GC/MS analysis.

     3.4  CHC Mass Spectral Library.  This library is a subset of
          the EPA/NIH/NBS Mass Spectral  Library which contains
          the chemicals in Table C-l.

     3.5  EPA/NIH/NBS Mass Spectral Library.  A library of
          reference mass spectra published by National Bureau of
          Standards, U.S. Government Printing Office, Washington,
          B.C.

     3.6  Procedural Blank.  A sample analysis performed in the
          laboratory with no sediment sample that is treated as a
          sample including exposure to all glassware, equipment,
          solvents, reagents, internal standards, and surrogates
          that are used with other samples.

     3.7  Laboratory Duplicate.  Two sample aliquots taken in the
          analytical laboratory and analyzed separately with
          identical procedures.

4.    Interferences

     4.1  Interferences may be caused by contaminants in
          solvents, reagents, glassware,  and other sample
          processing equipment.  Procedural blanks are analyzed
          routinely to demonstrate that  these materials are free
          of interferences under the analytical conditions used
          for samples.

     4.2  To minimize interferences, glassware (including sample
          bottles) should be meticulously cleaned.  As soon as
          possible after use, rinse glassware with the last
          solvent used.  Then wash with  detergent in hot water
          and rinse with tap water followed by distilled water.
          Drain dry and heat in a muffle furnace at 450°C for a

                               C-4

-------
          few hours.  After cooling, store glassware inverted or
          covered with aluminum foil.  Before using, rinse each
          piece with an appropriate solvent.  Volumetric
          glassware should not be heated in a muffle furnace.

5.   Safety

     5.1  The toxicity or carcinogenicity of each chemical used
          in this method has not been precisely defined.
          Therefore, each should be treated as a potential health
          hazard, and exposure should be reduced to the lowest
          feasible concentration.  Each laboratory is responsible
          for safely disposing materials and for maintaining
          awareness of OSHA regulations regarding safe handling
          of the chemicals used in this method.  A reference file
          of material data handling sheets should be made
          available to all personnel involved in analyses.
          Additional information on laboratory safety is
          available [11.1,11.2,11.3].

     5.2  The following method analytes have been classified as
          known or suspected human or mammalian carcinogens:
          d10-biphenyl, 13C6-l,2,4,5-tetrachlorobenzene, d12-
          chrysene, and 13C6-hexachlorobenzene.   Primary standards
          of these compounds should be prepared in a hood.  A
          toxic gas respirator should be worn when the analyst
          handles solutions containing high concentrations of
          these compounds.

6.   Apparatus and Equipment (All specification are suggested.
     Catalog numbers are included for illustration only.)

     6.1  Glassware

          6.1.1  Soxhlet extractors —  200 ml capacity, 500 ml
                 flask, coarse fritted glass Soxhlet extraction
                 thimble.

          6.1.2  Concentrator Tube — 10 mL graduated Kuderna-
                 Danish design with ground-glass stopper.

          6.1.3  Evaporative Flask — 500 mL Kuderna-Danish
                 design that is attached to concentrator tube
                 with springs.

          6.1.4  Snyder Column — three-ball macro Kuderna-Danish
                 design.

          6.1.5  200 juL autosampler microvials and/or microvial
                 inserts.
                               C-5

-------
     6.1.6  Chromatography Column — glass column
            approximately 400 mm long X 19 mm ID with Teflon
            stopcock and 300 mL reservoir.

6.2  High Performance Liquid Chromatography System.

     6.2.1  The HPLC must be capable of solvent programming,
            be capable of injecting large sample volumes,
            i.e., 1 mL and greater, and be capable of UV
            detection at 200 mm.

     6.2.2  Stop watch/timing device (required if 6.2.3 not
            available) for manual collection of HPLC
            fractions.

     6.2.3  Fraction collector (optional) if used, should be
            capable of taking time window fractions from
            HPLC.

6.3  HPLC Column

     A 25 cm X 4.6 mm i.d. ,  5 micron C18 column.


6.4  GC/MS System

     6.4.1  The GC must be capable of temperature
            programming, splitless or on-column injection,
            and have a designed capillary column injector.

     6.4.2  The MS must be capable of full scan mass
            spectral analysis using electron ionization at a
            electron energy of 70 ev.  The required MS scan
            rate should be >0.5 s and <1.5 s.

     6.4.3  An interfaced data system (DS) is required to
            acquire, store,  reduce, and output mass spectral
            data.  The DS must be capable of performing
            typical mass spectral data manipulations; i.e.,
            creating and plotting total and selected ion
            current profiles, integrating chromatographic
            peak areas, perform quantifications using an
            internal standard method, etc.

     6.4.4  The data system must be capable of library
            searching detected chromatographic peaks.

     6.4.5  The DS must have the latest release of the
            EPA/NIH/NBS mass spectral library.  This library
            is available for most GC/MS systems from their
            manufacture.

                          C-6

-------
     6.4.6  The DS must have a mass spectral library
            consisting of the chemicals in Table C-l.  This
            library, the CHC library, is a subset of the
            EPA/NIH/NBS mass spectral library.  This library
            can be constructed on most GC/MS systems without
            running GC/MS analyses on standard solutions 1

6.6  GC Column

     A 30 m X 0.32 mm or 30 m X 0.25 mm ID fused silica
capillary column coated/bonded with a 0.25 /xm or thicker
film crosslinked 5% phenyl methyl or methyl silicone.

6.7  Miscellaneous Equipment

     6.7.1  Microsyringes — various standard sizes.

     6.7.2  Boiling chips — approximately 10/40 mesh.  Heat
            at 400 C for  30  min.  or  extract with methylene
            chloride in a Soxhlet apparatus.

     6.7.3  Water bath — heated, with concentric ring
            cover, capable of temperature control within
            ±2°C.

     6.7.4  Analytical balance — capable of accurately
            weighing to 0.0001 g.

     6.7.5  Beakers —  250 ml.

     6.7.6  Disposable aluminum weighing pans

     6.7.8  Desiccator

     6.7.9  Soxhlet extractor heating mantle

     6.7.10 Mortar and pestle

     6.7.11 Spatulas

     6.7.12 Drying oven

     6.7.13 Mixing device — Magnetic stir plates and stir
            bars,  stirrers,  tumbler, or separatory funnel
            shake (optional).

     6.7.14 Extraction flasks — separatory funnels: 250,
            500,  1000 ml; flasks:  100,  250, 500, and 1000
            ml volumetrics;  etc.

     6.7.15 Disposable Pasteur pipets

                          C-7

-------
7.    Reagents and Consumable Materials

     7.1  Solvents.  High purity, distilled-in-glass acetone,
          hexane, toluene, and methanol.  For precise injections
          with splitless injectors and capillary columns, all
          samples and standards should be contained in the same
          solvent.  Effects of minor variations in solvent
          composition (i.e., small percentage of another solvent
          remaining in hexane extracts)  are minimized with the
          use of internal standards.  (External standard
          calibration is not acceptable).

     7.2  Sodium Sulfate.  ACS, granular,  anhydrous.  Purify by
          heating at 400°C for 4  h in a  shallow tray.

     7.3  Silica Gel.  60-200 Mesh, Soxhlet extracted with
          hexane/methylene chloride  (1:1), dried stored at 110°C
          for 48 h before use.

     7.4  Glass Wool.  Purify by heating at 400°C  for  24  h.

     7.5  Nitrogen Gas.   High purity, dry.

     7.6  Sulfuric Acid.  AR Select, Mallinckrodt or equivalent
          quality.

     7.7  Celite 545.  Fisher Scientific.

     7.8  Solid Phase Extraction Columns.   1 mL, 3 mL low density
          (LD), 6 mL high capacity (HC), J.T. Baker.

     7.9  HPLC Calibration Solution.  10 ppm solution (in hexane)
          containing benzene, bromobenzene, biphenyl,  bibenzyl,
          p,p'-DDE, and 2,2',4,5,5'-pentachlorobiphenyl.

     7.10 HPLC Fractionation Standard.  10 ppm solution (in
          hexane) containing 1,4-dichlorobenzene,  1,3-
          diethylbenzene, p,p'-DDE, and decachlorobiphenyl.

     7.11 HPLC Performance Solution.  10 ppm solution (in hexane)
          containing biphenyl, 1,3-diethylbenzene, and bibenzyl.

     7.12 Internal Standard Spiking Solution.  200 ppm solution
          (in hexane) containing d12-chrysene.

     7.13 Surrogate Standard Spiking Solution.  1 ppm solution
          (acetone) containing d10-biphenyl, 13C6-1,2,4,5-
          tetrachlorobenzene, and 13C6-hexachlorobenzene.
                               C-8

-------
     7.14 MS Performance Check Solution.  10 ppm solution (in
          hexane)  containing decafluorotriphenyl-phosphine
          (DFTPP).

     7.15 GC/MS Calibration Solutions.  Five hexane solutions are
          required.  These solutions contain constant
          concentrations of the internal standard, d12-chrysene
          and varying concentrations of the surrogates.
          Composition and approximate concentrations are given in
          Table C-2.

     7.16 GC Performance Solution.  10 ppm solution (in hexane)
          containing 0-BHC,  -BHC, d12-chrysene, and endrin ketone
          or a 10 ppm solution (in hexane) containing anthracene,
          phenanthrene, benz[a]anthracene, and chrysene.

8.    Calibration

     8.1  HPLC Chromatography Conditions

          8.1.1  Initial HPLC Chromatography Conditions — Fill
                 solvent reservoirs with HPLC grade water and
                 acetonitrile, set the flow rate to 1.0 mL/min, and
                 set the UV detector at 200 nm.  Solvent conditions
                 should be isocratic for 1.0 minutes initially with
                 80% acetonitrile/20% water, linearly programmed to
                 100% acetonitrile at a rate of 1.5%/min., and then,
                 isocratic for 30 minutes with 100% acetonitrile.  A
                 10 minute recycle time should be used between runs
                 with the initial isocratic solvent conditions.

          8.1.2  Inject 10 ML of the HPLC calibration solution.  RTs
                 should be within 20% of those shown in Table C-3.
                 If RTs are not within 20%, adjust either or both
                 the gradient program and/or flow rates to obtain
                 the desired RTs.

     8.2  HPLC Performance Criteria

          8.2.1  RT — log P conditions

                 8.2.1.1  Using finalized HPLC conditions (see
                          8.1.2), inject 10 juL of the HPLC
                          calibration solution and record RTs of
                          each compound.

                 8.2.1.2  Using finalized HPLC conditions (see
                          8.1.2), inject 10 /iL of acetone and record
                          its retention time (RT).
                               C-9

-------
            8.2.1.3  Determine corrected RTs (i.e., RT of each
                     compound minus the RT of acetone).

            8.2.1.4  Perform a regression analysis using the
                     log P values in Table C-3 and the
                     corrected RTs.  An equation of the form:
                     log P = A * log tc +  B   where A and  B  are
                     regression coefficients and tc is the
                     corrected RT, should be used.

            8.2.1.5  The correlation coefficient for the
                     regression, r2,  should  be  greater than
                     0.95.  If this condition is not met,  HPLC
                     chromatography conditions are not correct.
                     Return to 8.1.2 and determine improved
                     conditions.

     8.2.2  Resolution

            8.2.2.1  Using the finalized HPLC conditions  (see
                     8.1.2), inject 10 juL of the HPLC
                     performance solution.  Baseline separation
                     between biphenyl, 1,3-diethylbenzene, and
                     bibenzyl should exist.

            8.2.2.2  If baseline separation between all three
                     chemicals does not exist,  either replace
                     the chromatography column and return to
                     8.1 or return to 8.1.2  and determine
                     improved conditions.

8.3  HPLC Fraction Time (Collection Time)  Identification.

     8.3.1  Inject 10 /iL of the HPLC fractionation standard
            using the finalized chromatographic conditions and
            record RTs of 1,4-dichlorobenzene,  1,3--
            diethylbenzene, p,p'-DDE, and decachlorobiphenyl.
            These times are used for fraction collection from
            the HPLC.

     8.3.2  RT reproducibility — For each compound in the HPLC
            fractionation standard, the absolute RTs should not
            vary by more than ±0.10 minutes from one analysis
            to the next.

8.4  GC/MS Conditions.

     8.4.1  Recommended gas chromatography conditions

            Column Type:        DB-5
            Film Thickness:     0.25 /im

                          C-10

-------
            Column Dimensions:  30 m X 0.32 mm or 30 m X
                                0.25 mm
            Helium Linear
              Velocity:         30 cm/ sec @ 250°C
            Temperature Program:
                     Inject 50°C,  hold 4  mins. ,
                     increase to 175°C at 10°C/min,
                     increase to 275°C at 5°C/m,
                     hold at 275°C for 20 mins.
            Injection Volume:   1 or 2
     8.4.2  Recommended acquisition conditions for mass
            spectrometer .

            Mass Range:         45-545 m/z
            Total Cycle Time
              per Scan:         0.5 < cycle time < 1.5 seconds

     8.4.3  Mass Spectrometer Calibration

            8.4.3.1  Calibrate and tune MS with standards and
                     procedures prescribed by the manufacturer.
            8.4.3.2  Inject 1 ^iL or 2 juL aliquot of MS
                     performance check solution.  If spectrum
                     does not meet criteria for DFTPP (Table
                     C-4 ) ; return to 8.4.3.1 and
                     recalibrate/tune MS.

8 . 5  GC/MS Performance Criteria

     8.5.1  GC Performance — Inject 1.0 p>L of the GC
            performance solution.  Baseline separation between
            /3-BHC and  -BHC and between endrin ketone and d12-
            chrysene should exist.  Alternatively, anthracene
            and phenanthrene should be separated by baseline
            and benz [a] anthracene and chrysene should be
            separated by a valley whose height is less than 25%
            of the average peak height of these two compounds.

     8.5.2  MS Sensitivity — Inject 1.0 /iL of the 0.5 ppm
            GC/MS calibration solution.  Using the total ion
            chr omatogram , a signal to noise ratio of greater
            than 3 should be observed for each surrogate.

     8.5.3  MS Calibration — Inject 1.0 /uL of the 0.5 ppm
            GC/MS calibration solution.  For d12-chrysene,
            abundance of m/z 241 relative to that of m/z 240
            should be >15% and <25%.
                         C-ll

-------
     8.5.4  GC Stability — Perform multiple GC/MS analyses on
            the same GC/MS calibration solution.  RTs should
            not vary by more than t seconds.  Calculate the
            value of t with the equation, t = (RT)1  ,  where RT
            is the observed average RT (in seconds).

8.6  Response Factor Calculation for MS

     8.6.1  Inject 1.0 IJ.L of each GC/MS calibration solution
            and acquire GC/MS data.

     8.6.2  Calculate response factors (RF) for each surrogate
            relative to d12-chrysene

                      RF = ASQC/ACQS

            where  As = integrated  total  ion abundance for the
                         surrogate
                   Ac = integrated  total  ion abundance for the
                         internal standard, d12-chrysene
                   Qs = injected  quantity of surrogate
                   Qc = injected  quantity of d12-chrysene

     8.6.3  RF Reproducibility — For each surrogate,  calculate
            the mean RF.  When the relative standard deviation
            (RSD)  exceeds 30%, analyze additional aliquots of
            GC/MS calibration solutions to obtain acceptable
            RSD for the RF, or take action to improve GC/MS
            performance.

8.7  Continuing Calibration Check

     8.7.1  HPLC

            8.7.1.1  With the following procedures,  verify HPLC
                     column performance at the beginning and
                     end of each 12 h period during which
                     analyses are performed.

            8.7.1.2  Demonstrate acceptable performance for
                     criteria described in Section 8.2.2.

            8.7.1.3  Demonstrate acceptable performance for
                     criteria described in Section 8.3.2.

     8.7.2  GC/MS

            8.7.2.1  With the following procedures,  verify
                     initial calibration at beginning and end
                     of each 12 h period during which analyses
                     are performed.

                         C-12

-------
                 8.7.2.2  Inject 1 or 2 ML aliquot of MS performance
                          check solution.  Ensure acceptable MS
                          calibration and performance.

                 8.7.2.3  Demonstrate acceptable performance for
                          8.5.

                 8.7.2.4  Determine the area for chrysene-d12 has
                          not changed by more than 30% from most
                          recent analyses of the GC/MS calibration
                          solutions.

                 8.7.2.5  For an acceptable continuing calibration
                          check, the measured RF for each surrogate
                          must be within 30% of the mean value
                          calculated during initial calibration.

          8.7.3  Remedial Actions

                 Remedial actions must be taken if criteria are not
          met; possible remedies are:

                 8.7.3.1  Check and adjust operating conditions.

                 8.7.3.2  Clean or replace injector liner on GC.

                 8.7.3.3  Flush column with solvent according to
                          manufacturers instructions.

                 8.7.3.4  Break off a short portion (approximately
                          0.33 m) of the GC column; check column
                          performance by analysis of performance
                          check solution for GC.

                 8.7.3.5  Replace column; performance of all initial
                          calibration procedures then required.

                 8.7.3.6  Adjust MS for greater or lesser
                          resolution.

                 8.7.3.7  Calibrate MS mass scale.

                 8.7.3.8  Prepare and analyze new concentration
                          calibration/performance solutions.

                 8.7.3.9  Prepare new concentration calibration
                          curve(s).

9.    Procedures

     9.1  Sample Preparation for Extraction

                               C-13

-------
     9.1.1  Homogenize wet sediment sample.

     9.1.2  Place enough wet sediment on solvent rinsed
            aluminum foil to obtain approximately 25 g of
            sample after drying.

     9.1.3  Place aluminum foil with sample in hood and allow
            to air dry.

     9.1.4  Grind air dried sample in mortar and pestle until
            uniform and stored the ground material in solvent
            rinsed jar and cover with a foil lined lid.

9.2  Percent Moisture

     9.2.1  Place a weighing pan in a 105±5°C  oven for 15
            minutes, let cool for 15 minutes in desiccator, and
            then weigh to 5 places.

     9.2.2  At the same time as 9.1.2 is performed, place 1 or
            2 grams of wet sediment into tared weighing pan.

     9.2.3  Measure weight of tared weighing pan with wet
            sediment to 5 places.

     9.2.4  Place tared weighing pan with wet sample in an oven
            at 105±5°C  and dry sample for a minimum of 12
            hours.

     9.2.5  Let weighing pan cool in desiccator, and reweigh to
            5 places.

     9.2.6  Repeat 9.2.4 and 9.2.5 until a constant weight is
            obtained, i.e., ±1.0%.  Note, if a forced-draft
            oven is used, 10 h usually is considered
            sufficient.  If a convection oven is used, samples
            should be dried for a least 24 h.

     9.2.7  Calculate percent moisture with the following
            equation:

               =  (Wet weight - Dry weight)*100/(Wet weight)

9.3  Percent Organic Carbon

     9.3.1  Measure organic carbon content of ground sediment.

     9.3.2  Calculated percent organic carbon of the sediment.

9.4  Extraction of Dried Sediment
                         C-14

-------
9.4.1  Place 20 g of ground sediment into a 250 ml beaker.
       Add 20 g of coarse sodium sulfate and mix.

9.4.2  Place 1/4" of silica gel into a coarse fritted
       glass Soxhlet extraction thimble.

9.4.3  Add 30 g of sodium sulfate to the thimble

9.4.4  Place two-thirds of the mixture into a coarse
       fritted glass Soxhlet extraction thimble and then
       add 0.100 inL of the surrogate solution.    The
       remaining sample is added to the thimble.

9.4.5  Add 30 g of sodium sulfate to extraction thimble
       and then a layer of glass wool

9.4.6  Place the extraction thimble into a clean Soxhlet
       extractor body.

9.4.7  Add 200 mL of acetone and one or two boiling chips
       to Soxhlet extraction flask.

9.4.8  Rinse 250 ml beaker with acetone 3 times and pour
       acetone into Soxhlet body.

9.4.9  Assemble Soxhlet extractor and place onto
       extraction rack.  Turn heaters on and extract
       sediment for 4 hours.

9.4.10 Turn off heaters and after cooling, quantitatively
       transfer sample to a Kuderna-Danish concentrator.
       Cap Kuderna-Danish concentrator.

9.4.11 200 ml of tolueneimethanol (1:3) solution and one
       or two boiling chips are placed into the emptied
       Soxhlet extractor flask.  The flask along with
       Soxhlet extractor body are placed on to the
       extraction rack.

9.4.12 The heaters on the extraction rack are turned on
       and the sample is extracted for at least 12 hours.

9.4.13 Turn off extraction rack and allow toluene:methanol
       extract to cool.

9.4.14 Add one or two clean boiling chips and attach a 3-
       ball Snyder column to the KD apparatus.
       Concentrate acetone extract on a steam bath to a
       volume of approximately 20 ml.
                    C-15

-------
     9.4.15 After allowing toluenermethanol to cool,
            quantitatively transfer this extract to the KD with
            the concentrated acetone extract.

     9.4.16 Add one or two clean boiling chips and attach a 3-
            ball Snyder column to the KD apparatus.
            Concentrate the extract on a steam bath to a volume
            of approximately 8 ml.

     9.4.17 Transfer extract to vial suitable for storage with
            hexane.  Cap vial and store in freezer.

9.5  Sample Cleanup

     9.5.1  Construct column containing (bottom to top) glass
            wool, silica gel (2 g),  sodium sulfate (2 g),  70%
            sulfuric acid solution (5 ml)  on Celite (10 g), and
            sodium sulfate (2 g).   The column is washed with
            100 mL of hexane is not allowed to go dry.

     9.5.2  Quantitatively transfer the hexane extract from
            9.4.17 to the column using hexane solvent.  Collect
            eluate into a 500 mL Kuderna-Danish evaporation
            flask with a 10 mL lower tube.

     9.5.3  Wash the acid/Celite column with 100 mL hexane and
            collect the eluate in the same Kuderna-Danish
            flask.

     9.5.4  Add one or two clean boiling chips to the flask and
            attach a three ball Snyder column to the Kuderna-
            Danish apparatus.  Concentrate the extract on a
            steam bath until volume of the extract is less than
            8 mL.  Allow Kuderna-Danish apparatus to cool and
            detach the lower tube, rinsing the joint with
            hexane into the sample.

     9.5.5  Use a nitrogen stream and a warm (<40°C) water bath
            to concentrate the extract to a volume of 2.0 mL.

     9.5.6  Cap extract and store in freezer.

9.6  Removal of Sulfur

     9.6.1  A column consisting of a disposable Pasteur pipet
            packed from bottom to top with glass wool, 5 cm of
            activated copper filings, and glass wool is made
            and the column is washed sequentially with 20 mL of
            acetone, 20 mL toluene,  and 2X 20 mL hexane.
                         C-16

-------
     9.6.2  Remove extract from freezer and allow it to come to
            room temperature.  Quantitatively transfer the
            extract to the column using hexane and collect
            eluate.

     9.6.3  Wash column with three 10 ml aliquots of hexane or
            methylene chloride and collect eluate in the same
            receiving vessel.

     9.6.4  Use a nitrogen stream and a warm water bath to
            concentrate the extract to a volume of 0.3 mL.

     9.6.5  Quantitatively transfer extract to a HPLC
            autosampler vial.  Cap vial and store vial in
            freezer.

9.7  Fractionation of Sediment Extract

     9.7.1  Remove sample from freezer and allow it to warm to
            ambient temperature.  Uncap the vial and allow the
            sample to evaporate naturally to the volume of
            500 jLtL.  Mix extract and cap vial.

     9.7.2  Using the HPLC conditions determined previously
            (see Section 8.2), inject 175 ML of the sample into
            the HPLC column.

     9.7.3  At the RT of 1,4-dichlorobenzene (determined in
            Section 8.3), place a clean flask under the waste
            tube from the UV detector and collect the column
            eluate.  Label this fraction as Fraction 1.

     9.7.4  At the RT of 1,3-diethylbenzene, place a clean
            flask under the waste tube from the UV detector and
            collect the column eluate.  Label this fraction as
            Fraction 2.

     9.7.5  At the RT of p,p'-DDE, place a clean flask under
            the waste tube from the UV detector and collect the
            column eluate.  Collection should continue until
            the RT of decachlorobiphenyl.  Label this fraction
            as Fraction 3.

     9.7.6  Repeat steps 9.7.2 to 9.7.5 two more times so that
            all the sample has been injected onto the HPLC
            column.  Combine the like fractions into the same
            flasks.

9.8  Concentration of Sediment Extracts (Liquid/Liquid
     Extraction)
                         C-17

-------
9.8.1  Quantitatively transfer each HPLC fraction into its
       own extraction bottle, e.g., separatory funnels,
       bottles, etc.  Dilute fractions a minimum of 10
       fold with HPLC grade water.  Seal bottle and mix
       diluted fractions thoroughly.

9.8.2  Open bottles and add 60 mL of hexane to each
       diluted fraction.  Seal bottles.

9.8.3  Place bottles onto mixing device, i.e., shaker,
       tumbler, or stirring apparatus.  Shake, tumble, or
       mix the sample for a minimum of 30 minutes. (Note,
       longer extraction times might be necessary to
       obtain proper extraction with your apparatus.)
       Alternatively, separatory funnel extraction by hand
       with periodic venting may be used; shake for
       minimum of 2 minutes

9.8.4  Remove bottles containing the diluted fractions
       from mixing device.  Allow hexane layer to separate
       from water phase to 10-15 minutes.  Collect the
       hexane in a clean flask.  Each HPLC fraction must
       have its own flask.

9.8.5  Repeat steps 9.8.2-4 two additional times and
       combine the hexane from these extractions with the
       hexane from the first extraction.

9.8.6  Dry each extract by passing it through a drying
       column containing about 10 cm of anhydrous sodium
       sulfate.  Collect the dried extract in a Kuderna-
       Danish concentrator flask.  Rinse flask which
       contained the solvent extract with 20-30 mL of
       hexane and add it to the column to complete the
       quantitative transfer.

9.8.7  Add one or two clean boiling chips to the flask and
       attach a three ball Snyder column to Kuderna-Danish
       apparatus.  Concentrate the extract on a steam bath
       until volume of the extract is less than 8 mL.
       Allow Kuderna-Danish apparatus to cool and detach
       the lower tube, rinsing the joint with hexane into
       the sample.  Three extracts for each sample will
       results, i.e., fractions 1, 2, and 3.

9.8.8  Using a stream of dry clean air, evaporate extracts
       to approximately 1 mL.

9.8.9  Quantitatively transfer and concentrate extracts
       for each fraction to 100 juL in microvials.  Cap and
       store in freezer.

                     C-18

-------
9.9  Concentration of Sediment Extracts  (C18SPE)

     9.9.1  Dilute fractions a minimum of 10 fold with HPLC
            grade water and mix diluted  fractions thoroughly.

     9.9.2  Activate three C1fl solid phase extraction (SPE)
            columns by passing methanol  and then HPLC grade
            water through the column as  specified by the
            manufacture.  The size of the column required is
            dependent upon the volume of diluted HPLC
            fractions; use manufacturer's recommendations.  In
            general, 1 mL, 3 mL LD, and  3 mL columns will be
            needed.

     9.9.3  Pass diluted extracts through the C18SPE columns.
            Do not exceed maximum flowrates recommended by
            manufacture .

     9.9.4  Remove excess water for the  column.  This can be
            done using a slight positive pressure of clean air
            or nitrogen gas, a slight vacuum, or (spun out
            using) a centrifuge.  Do not let the columns go
            dry.
     9.9.5  Elute C18SPE columns with three 500 /nL aliquots of
            methylene chloride and then one 500 /iL aliquot of
            hexane.

     9.9.6  Dry extracts by passing C18SPE eluent through micro
            Na2S04 columns.  Micro drying  columns may be
            prepared by placing 1 cm of Na2804  into an empty
            1 mL SPE column and passing approximately 2 ml of
            hexane through the column.

     9.9.7  Concentrate eluate from drying column to
            approximately 100 /uL and quantitatively transfer
            concentrate to microvials.  Cap and store in
            freezer.  Three extracts for each sample will be
            obtained.

9.10 GC/MS Analysis

     9.10.1 Remove sample from freezer and allow it to warm to
            ambient temperature.  Adjust volume of sample to
            100 nil and then spike sample with 5 /uL of the
            internal standard spiking solution, 1000 ng of d12-
            chrysene.  Mix sample after spiking sample with
            d12-chrysene .
                         C-19

-------
     9.10.2 Inject a l /*L or 2 juL aliquot of blank or sample
            extract into GC operated under conditions used to
            produce acceptable results during calibration.

     9.10.3 Acquire MS data using full scan conditions (see
            Section 8.5).

     9.10.4 Recap samples and store extracts in freezer.   Note,
            autosampler crimp caps after being punctured by the
            syringe used for sample injection must be replaced.

9.11 GC/MS Data Analysis

     Data analysis for this analytical procedure can be divided
into three tasks, 1) quantification of surrogates and other
GC/MS components, 2) library searching with the CHC mass
spectral library, and 3) library searching with the EPA/NIH/NBS
mass spectral library.  These procedures are performed on all
HPLC fractions for the the sample and blank.


     9.11.1 Quantification of Surrogates and Other GC/MS
            Components

            9.11.1.1 Surrogate and Internal Standard
                     Identification

            Identify surrogates and internal standards by
            comparison of their mass spectrum (after background
            subtraction) to reference spectrums in the user-
            created data base.  The GC retention time of the
            sample component should be within 10 sec of the
            time observed for that same compound when a
            calibration solution was analyzed.  In general, all
            ions that are present above 10% relative abundance
            in the mass spectrum of the standard should be
            present in the mass spectrum of the sample
            component and should agree within absolute 20%.
            For example, if an ion has a relative abundance of
            30% in the standard spectrum, its abundance in the
            sample spectrum should be in the range of 10 to
            50%.  Some ions, particularly the molecular ion,
            are of special importance, and should be evaluated
            even if they are below 10% relative abundance.

            9.11.1.2 Peak Integration

            For each fraction, use the GC/MS peak detection and
            integration software to obtain areas for all
            chromatographic components from the total ion
            chromatogram with a signal to noise of 3 and

                          C-20

-------
greater.  (Note, the solvent front need not be
examined.)  Verify that the surrogate in each
fraction was integrated.  If surrogate was not
integrated use more sensitive peak detection and
integration settings to obtain peak areas for the
surrogates.

9.11.1.3 Calculate Surrogate Concentrations

         cx = (VQis)/(Ais'RF'M)

Where    Cx  =  Concentration of surrogate

         Ax     =  Area of surrogate in total ion
                   chromatogram

         QIS     =  Quantity of internal standard
                   added to extract before GC/MS
                   analysis

         A,s     =  Area of internal standard in
                   total ion chromatogram added to
                   extract before GC/MS analysis

         RF     =  Mean response factor of
                   surrogate from initial
                   calibration analysis and/or
                   from GC/MS calibration solutions
                   run with sample analyses.  For
                   fraction #1, use RF for d10-
                   biphenyl.  For fraction #2, use
                   RF for 13C6-1,2,4,5-
                   tetrachlorobenzene.  For
                   fraction #3, use RF for 13C6-
                   hexachlorobenzene.

         M      =  Mass of extracted sediment,
                   i.e., 20 g.

Alternatively,  use the GC/MS system software or
other available proven software to compute the
concentration of the surrogates using linear,
second, or third order regression or using
piecewise calibration curves.

9.11.1.4 Calculate Surrogate Recoveries

         Surrogate Recovery =  (Cx •  100)/SSC

Where    Cx      =  Concentration of surrogate
              C-21

-------
         SSC   =   Surrogate spiking concentration,
                   i.e.,  5 ng/g

9.11.1.5 Calculate concentrations of all
         chromatographic components

         Cx = (Ax • Q|S • 100)/(A|S • RF • M • REG)

Where    Cx    =   Concentration of chromatographic
                   component

         AX    =   Area of chromatographic
                   component in total ion
                   chromatogram

         QIS    =   Quantity of internal standard
                   added to extract before GC/MS
                   analysis

         A)S    =   Area of internal standard in
                   total ion chromatogram

         M     =   Mass of extracted sediment, i.e.
                   20 g

         RF    =   Mean response factor of
                   surrogate from initial
                   calibration analyses and/or from
                   GC/MS calibration solutions, run
                   with sample analyses.  For
                   fraction #1, use RF for d10-
                   biphenyl.  For fraction #2, use
                   RF for 13C6-1,2,4,5-
                   tetrachlorobenzene.  For
                   fraction #3, use RF for 13C6-
                   hexachlorobenzene.

         REC   =   Surrogate recovery calculated in
                   Section 9.11.1.4.  For fraction
                   #1, use recovery of d10-
                   biphenyl.  For fraction #2, use
                   recovery of 13C6-1,2,4,5-
                   tetrachlorobenzene.  For
                   fraction #3, use recovery of
                   13C6-hexachlorobenzene.

Alternatively, use the GC/MS system software or
other available proven software to compute the
concentration of all chromatographic components
using linear, second, or third order regression or
using piecewise calibration curves.  These

              C-22

-------
       calculations must use the response curve for the
       surrogate appropriate for the fraction of interest
       and the reported concentrations must be corrected
       for recovery of the surrogate. For fraction #1, use
       response curve and recovery of d10-biphenyl.  For
       fraction #2, use response curve and recovery of
       13C6-l,2,4,5-tetrachlorobenzene.   For fraction #3,
       use response curve and recovery of 13C6-
       hexachlorobenzene.

9.11.2 Library searching with the CHC mass spectral
       library

       9.11.2.1 Algorithm Selection

       A reverse searching algorithm is required when
       available.  If GC/MS system does not have a reverse
       searching algorithm, library searching should be
       performed using the default algorithm supplied by
       the manufacture of the instrument.

       9.11.2.2 Searching

       All chromatographic components detected in 9.11.1.2
       (Peak Integration) are searched against the CHC
       mass spectral library.  Those GC/MS components with
       of 70% and greater are considered tentatively
       identified.  For each tentatively identified
       component, a list of the best mass spectral library
       identifications (up to a total of ten
       identifications) is reported along with the percent
       fit values, CAS number of each tentative
       identification, HPLC fraction number, GC retention
       time, and the concentration for the GC/MS
       component.  This report is called Report 1.

9.11.3 Library Searching with EPA/NIH/NBS Mass Spectral
       Library.

       9.11.3.1 Algorithm Selection

       A reverse searching algorithm is required when
       available.  If GC/MS system does not have a reverse
       searching algorithm, library searching should be
       performed using the default algorithm supplied by
       the manufacture of the instrument.

       9.11.3.2 Eliminating of Compounds Below 5 ng/g

       All unidentified components from the CHC library
       search with concentrations less than 5 ng/g are

                    C-23

-------
            eliminated from further data processing.  This
            elimination can be performed by comparing peak
            areas or heights of each chromatographic peak to
            the peak area or height of the surrogate
            appropriate for that fraction.  This elimination
            may also be performed by comparing the recovery
            corrected concentration of each chromatographic
            component to 5 ng/g.

            9.11.3.3 Searching

            The remaining components are searched against the
            EPA/NIH/NBS mass spectral library.  Those GC/MS
            components with fits of 70% and greater are
            considered tentatively identified.  For each
            tentatively identified component, a list of the
            best mass spectral library identifications (up to a
            total of ten identifications) is reported along
            with the percent fit values, CAS number of each
            tentative identification, HPLC fraction number, GC
            retention time, and the concentration for the GC/MS
            component.  This report is called Report 2.

            For those components with fits/matches less than
            70% and greater than 25%, the two best mass
            spectral library identifications along with the
            percent fit values, the CAS number of each
            tentative identification, HPLC fraction number, GC
            retention time, and the concentration are reported
            for each the GC/MS component.  For GC/MS components
            with fits/matches less than 25%, the concentration,
            HPLC fraction number, and GC retention time for
            each component is reported and the components
            labeled as being "unknown".  This report is
            identified as Report 3.

     9.11.4 Elimination of chromatographic components common to
            the spiked blank and sediment GC/MS data.

            Chromatographic components with retention times
            within ten seconds between the spiked blank and
            sediment and with the same mass spectrums should be
            removed from the analysis.

9.12 Reporting of Data

     9.12.1 Report 1:  CHC mass spectral identifications.

            For each chromatographic component tentatively
            identified using the CHC search (fits greater than
            or equal to 70%), a list of the best mass spectral

                         C-24

-------
       library identifications (up to a total of ten
       identifications) is reported along with the percent
       fit values, CAS number of each tentative
       identification, HPLC fraction number, GC retention
       time, and the concentration for the GC/MS
       component.

9.12.2 Report 2:  EPA/NIH/NBS mass spectral tentative
       identifications (fits > 70% and > 5 ug/kg).

       GC/MS components tentatively identified in the
       EPA/NIH/NBS mass spectral search.  For each
       tentatively identified component, a list of the
       best mass spectral library identifications (up to a
       total of ten identifications) is reported along
       with the percent fit values, CAS number of each
       tentative identification, HPLC fraction number, GC
       retention time, and the concentration for the GC/MS
       component.

9.12.3 Report 3:  EPA/NIH/NBS mass spectral tentative
       identifications (fits < 70% and > 5 ug/kg).

       GC/MS components not tentatively identified in the
       EPA/NIH/NBS mass spectral search.  For those
       components with fits/matches less than 70% but
       greater than 25%,  the two best mass spectral
       library identifications along with the percent fit
       values, the CAS number of each tentative
       identification, HPLC fraction number, GC retention
       time, and the concentration are reported for each
       the GC/MS component.  For GC/MS components with
       fits/matches less than 25%, the concentration, HPLC
       fraction number, and GC retention time for each
       component is reported and the components labeled as
       being "unknown".

9.12.4 QA/QC Report

       9.7.4.1  Percent moisture of the sample.

       9.7.4.2  Percent organic carbon content of the
                sediment.

       9.7.4.3  Recoveries.

       For d10-biphenyl, 13C6-l,2,4,5-tetrachlorobenzene,
       and 13C6-hexachlorobenzene,  recoveries are reported
       for each sample and blank HPLC fraction.
                     C-25

-------
                 9.7.4.4  GC/MS Chromatograms.

                 For the sample and its corresponding blank, total
                 ion chromatograms must be provided for each sample
                 and blank HPLC fraction.

                 9.7.4.5  QA/QC.

                 Chromatogram demonstrating HPLC resolution,
                      see Section 8.2.2
                 Data demonstrating GC resolution, see
                      Section 8.5.1
                 Data demonstrating MS sensitivity, see
                      Section 8.5.2
                 Data demonstrating MS calibration, see
                      Section 8.5.3
                 Data demonstrating DFTPP performance,
                      see Section 8.4.3.2
                 Data demonstrating precision, see
                      Section 10.5

10.   Quality Control

     10.1 Recoveries of Surrogates.  Method:

           % Recovery = measured surrogate amount  X 100
                         spiked surrogate amount

          Quality Assurance Requirement

                     25% < % recovery < 120%


          Quality Control Action.  If percent recovery is out of
          range, re-extract and re-analyze sample.

     10.2 Surrogate Fraction Location.

          Quality Assurance Requirement

          Fraction 1      should contain d10-biphenyl and should not
                          contain 13C6-1,2,4,5-tetrachlorobenzene or
                          13C6- hexachlorobenzene

          Fraction 2      should contain 13C6-1,2,4,5-tetra-
                          chlorobenzene and should not contain d10-
                          biphenyl or 13C6-hexachlorobenzene

          Fraction 3      should contain 13Cg-hexachlorobenzene and
                          should not contain 13C6-1,2,4,5-
                          tetrachlorobenzene or d10-biphenyl

                               C-26

-------
     Quality Control Action.  If surrogate is in wrong
     fraction, take corrective action for improper HPLC
     conditions.  Re-extract and re-analyze the sample.

10.3 Calibration Stability

     10.3.1 Continuing Calibration Checks

            Quality Assurance.  See Sections 8.7

            Quality Control Action.  Re-extract and re-analyze
            samples.  For corrective actions see Section 8.7.

     10.3.2 HPLC Stability.

            Quality Assurance.  Immediately before and after
            fractionation solution must be analyzed.  See
            Section 8.7 for stability requirements.

            Quality Control Action.  If RT stability
            requirements are not met, re-extract and re-analyze
            the sample.

     10.3.3 GC/MS Stability

            Quality Assurance.  During the GC/MS analysis run,
            a GC/MS calibration solution should be analyzed
            twice; once at the beginning and once at the end of
            the GC/MS analysis sample sequence.  See Sections
            8.5 and 8.7 for stability requirements.

            Quality Control Action.  If RT and/or MS
            sensitivity stability requirements are not met,
            repeat GC/MS analysis on samples after correction
            of instrumental problems.

10.4 Blanks

     10.4.1 Procedural Blanks

            Quality Control.  One procedural blank should be
            performed with every set of sediment samples
            analyzed.

            An acceptable procedural blank:

            a)  Meets Section 10.1 requirements.
            b)  Contains no compound with elution
                 characteristics and mass spectral features
                 that would interfere with identification and
                 quantification of the surrogates.

                         C-27

-------
            c)  Contains few chromatographic peaks in the
                 GC/MS total ion chromatograms for each
                 fraction.

            Quality Control Action.   Locate and eliminate the
            source of contamination.   Re-extract and re-analyze
            the entire batch of sample.

     10.4.2 Unspiked Procedural Blanks

            Quality Control.  One procedural blank not spiked
            with the surrogate spiking solution should be
            performed with every 20  sediment samples.

            An acceptable unspiked blank.   The sample should
            not contain detectable amounts of the surrogates.

            Quality Control Action.   Locate and eliminate
            source of contamination,  re-extract and re-analyze
            entire batch of sample.

10.5 Precision

     Quality Control.  A duplicate set of  sediment samples will
     be analyzed with every set of 10 samples.  To measure the
     precision, the relative percent difference between the lab
     duplicate will be determined for each surrogate.

          Relative % difference =

         surrogate amount        surrogate amount
           duplicate 1	-	duplicate 2	   X 100
                  Average surrogate  amount

     The relative percent difference should be less than 150%.

     Quality Control Action.  If relative  percent difference is
     out-range for any of three surrogates, re-extract and re-
     analyze the sample.

10.6 Sample Sets.

     Sample sets are defined as a group of samples that are
     carried through the analytical  procedure at the same time.
     Each sample set will include a  minimum of two QC samples,
     one of them being a spiked blank.  QC samples include
     spiked blanks, unspiked blanks,  and replicates.

     A set will normally contain 10  samples plus the required
     QC samples.
                         C-28

-------
     10.7 GC/MS Analysis Sets.

          A GC/MS analysis set is defined as a group of prepared
          samples (sediments and QC samples) and GC/MS calibration
          solutions analyzed during one GC/MS run.  A set will
          normally contain 12 prepared samples, all five GC/MS
          calibration solutions, and 1 or 2 duplicates of the GC/MS
          calibration solutions.  The duplicate GC/MS calibration
          solutions will be analyzed in the beginning and at the end
          of GC/MS sample sequence.

11.   Appendix C References.

     11.1 "Carcinogens — Working with Carcinogens", Department of
          Health Service, Center for Disease Control, National
          Institute for Occupational Safety and Health, Publication
          NO. 77-206, August 1977.

     11.2 "OSHA Safety and Health Standards, General Industry", 29
          CFR 1910,  Occupational Safety and Health Administration,
          OSHA 2206, Revised January 1976.

     11.3 "Safety in Academic Chemistry Laboratories", American
          Chemical Society Publication, Committee on Chemical
          Safety, 3rd Edition, 1979.
                              c-29

-------
                              TABLE C-l

                  CHEMICALS OF HIGHEST CONCERN LIST
CAS number
                   chemical  name
   50-
   57-
   58-
   60-
   70-
   72-
   72-
   76-
   91-
   95-
  101-
  115-
  117-
  118-
  309-
  319-
  319-
  603-
  608-
  924-
 1024-
 1746-
 2104-
 2385-
 8001-
39515-
11096-
11097-
11104-
11141-
12672-
12674-
53469-
29-3
•74-9
89-9
57-1
30-4
54-8
55-9
•44-8
94-1
94-3
61-1
32-2
81-7
74-1
•00-2
•84-6
•85-7
73-1
•93-5
16-3
57-3
01-6
•64-5
85-5
35-2
41-8
82-5
69-1
28-2
16-5
29-6
11-2
21-9
p,p'-dichlorodiphenyltrichloroethane  (DDT)
chlordane
hexachlorocyclohexane  (lindane)
dieldrin
hexachlorophene
p,p'-dichlorodiphenyldichloroethane  (ODD)
p,p'-dichlorodiphenyldichloroethylene  (DDE)
heptachlor
3,3'-dichlorobenzidine
1,2,4,5-tetrachlorobenzene
4,4'-methylene bis(N,N!-dimethyl) aniline
dicofol
bis(2-ethylhexyl)phthalate  (BEHP)
hexachlorobenzene
aldrin
alpha-hexachlorocyclohexane  (alpha-HCH)
beta-hexachlorocyclohexane  (beta-HCH)
technical-hexachlorocyclohexane  (t-HCH)
pentachlorobenzene
N-nitroso-di-n-butylamine
heptachlor epoxide
dioxin (2,3,7,8-TCDD)
ethylp-nitrophenylphenylphosphorothioate(EPN)
mirex
toxaphene
danitol
polychlorinated biphenyl 1260
polychlorinated biphenyl 1254
polychlorinated biphenyl 1221
polychlorinated biphenyl 1232
polychlorinated biphenyl 1248
polychlorinated biphenyl 1016
polychlorinated biphenyl 1242
                               C-30

-------
    Table C-2.
GC/MS Calibration Solutions, Concentrations of
Surrogates and Internal Standards.
13
 Cfi-l,2,4,5-tetra-
                             13
                              C6-hexachloro-
chlorobenzene
ppm
0.5
1
10
50
100
d10-biphenyl
pom
0.5
1
10
50
100
d12-chrysene
ppm
10
10
10
10
10
benzene
ppm
0.5
1
10
50
100

    Table C-3.
Suggested Retention Times (RTs) for Compounds in
the HPLC Calibration Solution.
Chemical
Benzene
Bromobenzene
Biphenyl
Bibenzyl
p , p ' -DDE
2 , 2 ' , 4 , 5 , 5 ' -Pentachlorobiphenyl
Retention
Time
4.3
5.5
6.8
8.1
13.4
15.1
LOO P
2.13
2.99
3.76
4.81
5.69
6.11
                              C-31

-------
Table C-4.     DFTPP Ion Abundance Criteria.





     m/z            Criteria



      51            10-80% of the base peak



      68            <2% of m/z 69



      70            <2% of m/z 69



     127            10-80% of the base peak



     197            <2% of m/z 198



     198            base peak or >50% of 442



     199            5-9% of m/z 198



     275            10-60% of the base peak



     365            >1% of base peak



     441            present and 50% of m/z 198



     443            15-24% Of m/Z 442
               C-32

-------
        APPENDIX  D






CHEMICALS AVAILABLE IN IRIS

-------
-  - C X

- - * !
E — u. C


i S g I
  a. •- v


* si c
         -   Is
         "* »*
         rjg- I
            1
         — ^o o» —
          fr ry
                                                          §     Mi
                                                          *     e 9 ^r
                                                              »dddedd
               - I " *
               o — c -c

              k C £ Z «
              c •£ •• u o
              C B. f •

              >. L. •£ i •-  »
              5. ~ e t> ~ c
              I-
              x O


              U
              SL ?
              S £

              f 1
                   O *t<  *   ftuttX*»>£

                   **. U.CD'NiXU^a.Ul^U
                                 t 8.5 I


                                 III]
                                 u. o o a.
                    ^1 B W^
                    a K «


                    o o —
                          •
           c,

           \
           1
  ^p
  — L>
  u c

^.   o "5
L. U - C
5 — o a

f ! £ f

5 f £ *
                                                                H
                                                              E 6 &
                                                    322
                     t. L> w **
                     c • e 3
                     — « x a
                     £ b L
                     u z z c



                   5 5 ~ S 3
                   ^ i N. f*"i *^i

                   « !e « S C

-------
   o fc*

   I -8 g

   ^ * * -z
   O -  u
   — — u O

   I5 * •

   i i i 1
   — -D — O
   * C CO
   uj c _ C


   dc t» C •*-
   o ~ « «-


      fc u

      1 =

   - t a &
-;,!l
O> W M ft



S « d u



cjl



t*S*
u .. 1 g
2 £ d S

I  S

I  5
 •8
          *  f
          o- •£
   ^ U     1
   > e S <-i   TI
       •-  e
                       K.
                   ~-  Kl O
                   e  o 5

           o   •  ~ o  • e §
d K> o o d o •*  m  d  d  -S
                                     00   o
                                     «-O«--rgir
                                               OO
 8e
 6  S
 • O  O
O rvi  v»
                   e  -888
                   d 5 d d d d •-
          <^» a  *»» v
         jr 1^ O •« 4 5 Kl


         Z. ^ rO S S I -.
                                     •fi ^
                                     u^  • ^r

                                     d ^ d
         a
                       • o o
              >c  R

             . 8  8  «
             * o  d  o
                       §m

                       g
                 g^
                   06  o^eooKiruoe   o o  ooroKi  o

                   d d « d d d d o* d d d d KI d o « d o d d « d
   u. ex u T>
   ^ -  £


    - -o" =



    y -


    g -
          II
c 8
          f
          u
         *;.<
         M  * L.

         E-.?
         • *- u



         c c c
                             8 II  ±
                             t,
                •e  * ^ — L.

                   E *• •*• s
               - «*•  B c> c 5
         r e t t x —  i- o £ —  T>
         EX— •B»-3  Oi-6»-  8
         S(J— X^Bl  — OE—  I/I  t.
                                                 &£ r
                                                 — « 5
                                 §« *•   B t c 3   •-    u*-c>t>
                             .*^x-   tee-   -o    X---0
                                                        i 2 x
                                                        e 3 JE
          lilll^i^^l'sictll-isl^l-
          OO«— OOBOJC^^-O.^'CU^'CiiS— X
          U*-~x£gE •  UJtO«- ^«fri-n.Cl£££

            UU*** CUL. * t- -~ JC * **• O fe •— ft* fc ir (/:  • C





    ^cSsscccccccccetcs^tlliill
                        a a eo

-------
    8
    S
             S K - 5-
             • e _ •
         St*l
         Qt uy   Q,


         g]i!i
  S      I   -o
  !t H    IB   'a
  S'    ~   =
  a — •   v u.
* <" 5 C   *
^ «, J .S   •"
                           2
                           5
                                 .   ™
                                 o' — d
                                  I.
lid
                                               •c  e>

                                               d o- o
o   e o >>. e iv rsi

d   e e e e e e
                                 o
                                 5  N.
                                     . o *n o
                                     '  •  • o
                           O r- -C

                         IM d d d
                                          .
                        oo-oo   — ~^-o--^
           4   IM       fM  IM
         o     '25    *
         «-^oi*tXXeKi~^  • «s
         «-Kli^Kl^^^-fW^IMfV<«
K »>" £ u" X
- 2 * £ -8
i '« "" * ™
£ « e "3 J
-- 6 8 ft >»
!'*! *^
- 1 i - x
> c § u "8
f v < ^
c 5 « w
« 5 * •*
* o •- o •v
: e • ° 2 r
u • c
LU -g 8
K *i e —
u ** •> w
*" -S e S
g |.S 'Z
* ! a &
i £ § *
5  — v
= - .-5 =

3 a
i-
u

Is
u
1*
U «K
!S e ir £
e o V e
rsi ro
o 5 B ^ S o
o o o o o o
o « o o « « eee




i il |

1 , JI I
O J ex 1-
— JE" T5 * * • 'E
JE *• — o «• < 8
U tt If u ~ 41 u
— o •? * JE d — o
i O 'I U . C C —
IMJC — ^T-IM*6.C! •
•UXX£ •j£^.*>'
^**<*x^*fipsa.
• «
IM
S IM
O O Kl O O IM


%<
IB
^
i
CD
">• i
* * * -e
_• «/ • b
75 — To x
ill J
8 T- •£. ^. •- —
t f T- ^ § J
CC O 4U O O CL
§ S
O O
« d d

S
t\l O
o *- « e



a
"o
* ^ t
is !r> •-
1518
l^S?
&l|i
"x -x •- c
33*5

tS 0
« d IM «

SKI O Ig O O O
o o o o o o
o o « e« o o o o




_ |

§ *
J 1 f £
& t w 4
s 1 2 !_
C O JE OX
£ L. u — C
JO— JE S t
1 — i- _ u i c a
| f r -8 5 £5 T
jo • «O in • ec •- ^
c*- *O* > KO.
K C-* C»n^ ca>»
«^ii ••— ««O>LJ »
XaiMOm^CBlCIM


o
o

•D
u

a>
|

1
g
|
t-
IX
*l
  CO
                                                                       (/\ fw «O
                                                                       < K. tv

                                                                       5 S 5

-------
                          u  ••  e   c
                          «t  _j    •   o
                          *  at  o   o
                                                             K.  N.      «5
                                                                                                                                               8
                                                                                                                 S
                                                                                            s. 8
                                                                                            > «-  K.       O  O  O  ~r
                           *  at  O   o

                           "8      TJ

                           i      5-
     S  St  a
                  ?C       HUM
                  O       U  03  Kl
          fc  X   •-
          «-  I/I   ••
                           11
                          X   8
                                                                                                                              SKI  sr          S  C           o  r\i  o
                                                                                                                                «    •  o  O    *  O  O        •    •   *
                                                                                                                 Ki<*<-oorviSotnrM      o  o  o
                                                                                                                                                             ,   S-       r
                                                                                                                                                             «   O       O
                                                         ooooooo5o^-oo      o<*
                                                    ooooooooooooo*   ooooooo*   oo«   «   o  •   o
    i      I
     U       X
     "       I
             i
    I  8
—   L.  1>
                                                                                                                                                                  >.
                                                                                                                                                                 £

                                                                                                                                                                 i   i
                                                                                                                                                                 -o   •-
                                                                                                                                                                      •
                                                         *       0       4C       t^OO-D               *C           i       u—t«
                                                         f       U       C   4"  —  £   O  L.   L.       •—                Of           T)       ^   X ••-
                                                         &       O       5  J  0   O.  S  O   O       i-                UM           '-            C  ^
                                                         ?      -       5^c-i.	o       i»  I;  ^   P   ft       *:.*€**
                                                         O       w        O   O
                                                                                            ft*  U   U       X
     u        •    O  t.   u   i_  E  u-\  *-   »-   «   u      4V*
    ••-       -*j   tf.  O  ••"   O  •••    *   '    '    u   ^   t1   u.  iD
    O         *  fr  ^  O  ^  O  ^  ^   '^   S   O   C   O  T5  b    _
«   .    OD  rsj   u  X   >X   -     -   «    .^Nfc^-L.    . —    Xi-K^Oi-Sw^t
o.**a^'uurwo>j0ttuCL<^O^(-J
 v
8

-------
            ~  I g
            ^

            f «
            w ^

            u  .. §  9
                       rg     O- O

                       d S rg d d
                 Ss
                                            I  s s s *>

                                                                     8
                                                                   rg   f» e m
  §
k "°

-1 §
  — 4. _
  ^ •- trt *-

  O- -   u
       •— ^ * S
       u .. S E
       2 id S

       I   2

       i   3
               521?   8-
5   £S8«£
                                                                         d
 •   f^   O IP  O 'O         B


•- *.' in d KI in d rg rg d d d d •-'
    *O ^ •   ^*GVO  • ^» p* u*\ O* B
          K. •  •  • m  . s.
          (V«-O"V"-'-'-OOIVJ
  —  -ox

  2 ± Z o
  _ -o « ~
  i |.f
  5?'i
  uu •
         I





        'l

       sf
                             o        o m

                             d ic   * * o o
o o

d d
                                          S
                                                O •- O O INI O IV
                                            s   s
                                            OOOOOOOiAO* O
  •" I e S

  8 l-s =

  " t a &
  >- o E <


  8 i^ c •-
    li £
  y^l ^ ^ ^


  = »• r" =

    5 c.
           11
           u  z
           BI
                                          o
                                          X
hy
                                            o

                                          o ^

                                          * »
                                            rs S-r *
                                                  ?
                                                  i
                                            4^ L.  —
                                          '^!  1
                                          w Hi C   ^
                                            n:
                                      r I
                                      * -

                                      c o
ic anhyd

esol
ene

am< ne
                                                                     c
                                                                     X
                                                                     X
                                                                     X
                           £ <

                           ii
obut
                                                                             g
          c "x e  i
          fc X X  7
        V i »i O   I.
        C 7 x x — o
        fctooo«c..___  -
        So — — Cx «  t( — ».  »  b
        — —'Uu&'ukxutrX'O
        oxxxx<  3   x —  e  C
        — LJ U l_l O. •- u.  C O. CD  CC  UJ
                                                                 g 8 f   r
                                                                 . ,- X g =
.i

d
                         -r « O-
                         «r IM INI
                         »- ^ in
0

-------


§
^
V.
.—
O-
U>
oc

i
u
*•-
|
«
>


X
I
ac
u

*
g
L^


NO


o- >v
e t>
0. 0

u
u
SL
1/J
c
-
c/?
^
s
(A
8
a.
$

f


"O
1
e
!
«
(X
3









i
1
X

£
£
o
«
"•
§

3
u

0
i
I
r
|
y
g
u
•c
c.







RAC  u
u 4
1



d dddrviKi d d d r\j dd^d«a*o co
Kl
P (V
o' e d IM ^"»|o «-*-*^r£ooot>






** •-
** «^ tfi •*- Q ^
•* o 55 — So
• « « « > • « - • «
0*0 * « o o* oo«



§0 i*n ^ t"*^ ^ v\
o o O rsj o O c\/ i\i rw
C) t*} *~ O O O O O O O l/^ O
« « OOOOO* O« OOOOOO4 O* * O

£
i s
5 'i
J i J J
Q. i "e c — — £ N a
*^flj— £CjOO'^ * £
— CO *^JJC*^— ' ^4^ *"
KKlO'fi.l.£°E X— I
«j c L. i7 o e. *" « * £N t1 "
£t«- u — O — — C — « Cl.
x 1 e •. i t £ i 1 'i fr -fe S i £
£ u — e — — o — c « t x e e £
£*'Ol»4)_i.£< OvO Cy-^ ££K*-'V
— _ D _ «-ox»-uv' • G£.^-' 1>> fjlb.
.-2.- £"**' 6 » E XI
O«CBZ(Ng«o^"f^(SiryrgrgKiKi>f>f



0 »
d d
Kl Kl

(M
0 S











o o
o o




,
ethacryloni tri le
etruchloroethyle
x *-
i*
ss



s
d
Kl
d

§







:




o
o





1
a.
§
R





in
(SI






J





•





imethyl phthalat
o
Kt
Kl



d
£
d


S










rg
§
o

1
1
"x
K
1
,6 Oini tro-o-cyc
•»
g













fcT
fO
§
d



Kl
o »-
o o





c ~
* 4
ll
U "-
OJ Kl
Kl Kl
Kl Kl



•O
O
M

••»
CO











O
o





i
o
D
S
Kl
Kl




-------
6- v> ; c   «i u X
^   So   uj « m
                            •* d o   d
                                                 K.


                                             t* ~- o n


                                             d d d d   odd   o' d
                                                                             in

                                                                             d
                                               5   »
                    (Minom  ;<> •» ^ *r Ki •-
                    ^rg*<  »-»-^^-(Min
                                                     * * *
                                                     •o -c
                                                     <•» •*
    —  » v >•
    ac c *- u     X

    ~ S i £     -8
    £•->-•     -v
    C «   —     01

    i S 8 p.     £
      a. — S   *  BI



    iifl   *5

    5 J 5 u     -8
    "" • — S
    lu   ^ O
    ot «> e —
    u *> « ••

    w -S c S

    8 I| =

    " i- a £
    >• o t «
    -j »- 3
    z   v C

        S "


    !itJJ
      •-•03

      3 E.
c' g
2 ~-
  GC
              11
                        §IV
                        e
                                               «
 . O 6 O; S S S


o o o d o d d
i
                                                               §§
                                         oooooo* o « *
                                                               o o o « o o o o « o
                                                                               gs

-------
   i S 8
            ±.|!
            Is * I
         u ••
         •C -I
            -   Is
            - -C W ~
            i  ' •* *
            rs » I
            «- ^ wj S
                           •» o

                           o o
    8    sf 58
    o  '-dodo
                                           6 rg  e
                                  KI        See
                                  o   KI    S 5 — 8     in
                                  o   o  KI o o o e   to o —
                                                                  e e o o
                                                                  «- o o «-_
                                                             •*  
                         £
                         ^>
                         c
                       H
                       ^ ^
                       ^f.
                       fe I S^I*
                       - % X C  C C
                              i ^
                                5
                       £ §
                       £ O
                       U —'

                       Sf
                      % Kl w E ~
                       -' £ -' *
  2 § «  * 8 -
  ft|  fe fej
  t x iZ^ I
  « x X  u y L.
    «. u  — • "


il*i*J
             y«
      %
      jo

      1
                4=5
                  ^ S
         It

         !!
                                                — U IA L.  . —
                                                £•- O ? rg — ». _
    O — rv  '
.- — .  u  • £
X « •« — •-  y
« £  • •  .  i>
Ul ^ ^ U ^-  ^
                "* ** E " «
                rg i — •"•" O
                 • i  i»  • u
                — m v< — a.
                                                                    I
                                                                    i
                                                                    x

                                                                    I
                                                                    c
                 H
                 i-
                 t *
                                                                        
                       ^••-'O^mN.rwt-.rx
                       tT'fXt1B~S* — ~
                       ~#t'S'ei?5?|M'-
                       — rwiNifsi>r*3S>Ci*i
                       ^ »
                       if< in
                           s
                           SIP
    *r
           .- C

i^iric'-o^SSS
l^miTl^lAlAl^
-------
         -of1
         f V* I
           • e u
         at « o u
    g g   £ SS
O- -   u

  in  <
« * £ r    x
— — u o    •
  .O • «-    "O
E — u. •    ^
0 in   —    Oi


-u II !  .1
• H "     1
> e S o     xl

           n

S   8 ••   e *
_ -n ._ O   •• B>
» £ »    « 1



5 S I Z

                                             g
                        O- • rg



                        Kl
                    **.*".•
                    o e> e e •
         (C     Kl          "^



ry«  « odooooooociooooooo
" u S. S

- ° I *

8 » § -£
UJ oc u T3
IA —   4
  " T3

  5 E

-------
































o
• o
£ £
CC
C IX
V ^









c-
r i
a '
CO.— •
2: " 3 S
"- »/ X -
x *- en *-
O- - w
tfj «" «
— -ox
at t< *• t-
— — - o o
£ C, ~
S "5 - '
"- •> C 5
- i? J r
? 2 «
C 4. •
5 t> — c
*h«
C o *
o i * *
5 , « -8
a C os
u « c
u; ^ o
K C C "~
U *• • w
vt c a
1 S S
8*| =
" i. a £
S £ 5 «
X " C
S « g -
uj at o "p
v> — c
^ » y
*- "O 3
5 a
(^ -
c -<
o
u







-•86
c? iz
? 8- 1
w^ S 2
S ^ 8
ac ac O u
X
- *g
- 0 B> -
7s ' * *
f S ^ I
*• »-  m


IP
X)
^
D>
J<
^!
^ X
1
•s.
01
^
O ^
«- Ol
K e










c
u
1 1
u z
in t-
« b
u j




_
e 5r»— o ooo M
* " • * • O O *
O •- Oinoo o O in •- «j o



s 8
"J S. a*.*. °. ^|go ^
o * a ~ o o o o"SS e

~ g s s ^ «J rg
eS-s^^S F 8 ". T». S -
K. ^i — .£ •- m rt K. 100 wi^i.i s.

w
•^
•^
^
^
W
£ *«•«««•
* ««*«««r\j
r\/
S E^^MRErg^r^SKiS^ KiS
8 - 4 8 5 8 5 8 S I g § £ § S § ~ 5 §§ „ S §1
oo« ooodddddo'cjdoodd* do* d* d* * do

IA
**C
1 i s
u t « «
c. .r -g. j
•e C tf i
f BOO
•- * -
- t S- x ~ " -
^ V o X ^ ** *^
F" *• t y ° — *
r ' x OCIBI— e«orj
S ^ O. t-fTJCuiiMnj
* -OOIBCBISI
•- ^ '^ TE201*'_'S
•n k£ '99fc--~'^C
c 1 g :>§ i H 1 is**
.Z P £ a . *• • *- — tu w wooeo
* fe ^ &IO**-' O'— X X C Al (JrVifVl
II! 2 s « 1 •», o -fex I.E=t§^£x.R!RI r
&|"? 1 ?r -* & « -s gi- e g s s k Jr c g I si § § t I
iiiEiiiifiiiisiimilifiiiiii
iil!l^lili!!l = ligll!ll!i!lll
S|SS*^s?;sS3(3SSRgRS2SS333§III§
-rgrvjrgM^^^^^^^^jj^i^jvi^^^^^jrj.jjj.jjfj.irjj



£  g

-------
s from IRIS 9/1/90
Possible, Site Specific
tion Factors Should be
Regulatory Action.
1* Estimated RAC (mg/l) RAC (mg/l)
mg/lcg/day BCF RL: 10E-6 Rl: 1W-6
3X lipid 0.0065 kg/day 0.020 kg/day
Consumption Consumption
•
1C V -» •
I: 1
U « "V
UK 9
Si M h.
0 - 0 ^ 0
•• *O ••• O *• 9 * *
• | • acl • 0
" 1 R 5
5 *l |
lr O E <
* K g £
uj at u *p
£ — *
zi %
"5 a.
1 •; .S
u i
01
"e J
1 1 « S
t £ ML.
x B k •
__ U * « CD
•- in i. rv r>
« fc « 0-
O 4 Kl K1
1 ° "?
3 •* «»
c 5 rJ r£
II



is
o o •

£
\
a>
m
M
Beryl 1 ium
Boron (Boron and B
Cadmium
7440417
7440428
7440439




§
o






>
I
i
e
7440473



s~
I 1 S S 8
« * o e o o o


~
|0
]
t> C
• 1 s i 1 ,
3 » o •• — —
*• _^ £_ — O 0
!-i4§83
l|i|]i|
lllllll

o
d
« • *
« •" *
„ *
ill i
O O 0 « * « O





V
Hydrogen sulfide
Thai 1 ium chloride
Phosphine
Toxaphene
Creosote
Coke oven emission
Demeton
iliilli




,_
o






«
b.
+*
ar
10102439





^*






Nitrogen dioxide
10102440


•
g
§
O






u
*^
"•
£
*-
10102451

o o'
~ o
•* o-
o =

^
i s
e o






Nethamidophos
Resmethrin
102*5926
10453868

e
o'
o
rv *


« «


J
]
"x
t<
N-Nitroso-N-methyl
Nickel subsulfide
I0M5956
12035722

*
*
g
i
o






I
"tl
VI
"e
£
12039520




£
0






t
12122677


-------


1
«~
<>

i-
ac
™"
S

V)
$
TJ
>

u
z

u
I
(/>
u
^
I/I
.
Cy
2

1

^
|

W
I


1
X
trt
tf!
o
*-
*
u.
§
*>•
•
u

o
IB


g

Tj
4t
>
(_
2
*
1
•V
oc

s

^

-.!.§
•^ i 01 *•
i*:i
a^»i * ~. ^§
ac ac o u OK- IM o rvo •- «o
= .li
««. i J< *;
rs^ I
w ^ S
 * O1 •
IAUM fW(M IT\ iNl *- ••— - »**i
UJ«DK> s»«- w-fy pn f^rsj iMfO



xl
>. •••
*S ^L
"Jj ^
*** w o
« 9 m o

a ^ • — ic «
X
"O u^

e>< § § rC - -5 §00 SiooSioSooo o o
(KB o'*dd«
8 I
  CD
zi

ne
                                                             S t
                                                                 x
                                                                 u
                                                                 i
                •
                c
                SL
                M g
HIM
> cv tv K j: L.
U *- JC O N- O.
rv  KI K»
M ru M 
-------








u
i *

c " S S
x! *• i/i Z>
O- •- u
- -ox
ft; 4* ** t
" 3 K £
E •- i*. «
w ** oe

* *• *, A
Slf 6
LW Tj O
Oc 4* C •"
(J •* c fr"
*" "8 5 *
o f .£ -
•*. *»> Q,
* fe f £
S £ S •-
u, ai u 7
in — C
D • eft
3 '5.
I ""



= -{«
««. i o> *•
IS J 1
•X w • O
m m o u
- Is
—  •-
*» ' * *
ISs 1

K « d u
1 -0
I -f

« LJ »*
UJ • K>



£"
5
-J
a N.
e ^
•E I








Chemical
Name
u jj


•l ^ 5. 8 1
CO— O S r\; l>j d



5 § i
fVfVK) O S3 »TN N. O

^c K. e to 2 o
• • «9 fy *4» • • K.
r ^ . *r • *» > ^
K> ^c o in o ^ ^- in





la.
d « • « <
§K> 1C in ft o
Beeeeo 2 8
d e « d d d d d d d « o KI e




t ,.
£ f 0
*z I
ill 1 •
| tl c I I
* • ? - cc£ S«
•5 II -k2 !*«
u«»B| S S F - 1 c '
C&*^k OM a. '~ (•" D 0 — —
V— • £ k. • O £ i — i. — >.O
4s£«ii.j.t:'e-fi'B«'*'
O<» u*^ C^> aS«> M • O 4>*^
•-i-FC4-.-6i». ex t. ire
£•- U t V i «l T — ft> Q. O •
•~a«a.6«--0-*-"0
^o^v^^Ov^o^-inoo^-nn
Kpj>^rg-SBic35o.5-
^P»^'Win^^^Kt'*«*-fv^
»~O»r<5o>»f~>»>^~»»u^
fVftjPnKIKIKlKlfOKl^KSKIKlK!

3 i
>O O & t*l



-. §
iv d d n

o e
o> ^ *j oj

r^ ^ ^ &





* *

o o i s §
o • o « o o o



c
£
^ b.
f
" £
— ~ ^ L>
1:1!
Ht3
•5 * £ |
L. T — 0 C *•
o 5 « — »> «
Z I f X t C |
V | I 8 § S £
e c 7 fee * £
IT. » C — X X f
— tl t £ K « •-
CD •- &. LJ O «B O
s 2 s S i 5 *
in i5 W ITS N. rg INI
O* O O ^ ^ '^ ^^
W «* •»* N»» **-»**

«lil
^1 O O w
»- d d d d



8 5 5
•» •- d d d

•* •* p o o
fvi K i IIO.UC.LJ
OWCylNj—B— «0
•«^m««9K.'^-'
2^>»eino*n —
S^«ipS>C££
$:f'Mfv«i**'«rg
^o^v^^ivfuin
•*>nmw«ir»ininin

-------










LI
£ i
*Jis
— « .e —
^ •• IA v
O — U
l/> W «
V) u
— - O >
t( f »• k
- S £ £
1 - * f
* 1 1 !
Sis:
~ 5 ^ »
5 Ho
C u «
£ U K
S » 8 -
B-BS :
uj • _ C
UJ y O
K «i e •-
u »- • «-
VI • «
•ecu
§^!i
. & |l
8 j § •-
u. B O Tl
v> — b
3 « V
^ T =
~s 'a.
- Is
— « -X —
>» T Ol *.
IS* !
« lj • B
at m e u
^g
£«o ••!
r« " 1
" i S
< li . i
m m e u
T r
1 !
2 §g






X
•s
•v
e>
M
±1
I
O>
e <
-i r







6
                                   rg

                           O-      0
S      O
                                                                                 e  o  e
                                                                                     §K.  t*l

                                                                                     8  §
                           a     i
                                                                                                °~ z
                                                                                                d e
                                                              1       4
                               Kt      Kl     K>         ^r»     N.efrtAI^

                       SS8S5     5     SI  8  £ 8 §  §  8S5

-------
               ~   -8 8
               — -C v. —
                                             5
               w «- « S
                    e IK
               u •• 5 c

               2 £ d 5


               *   -D



               I   5
               ^ ik
               m u X
               w • PO
                  I
   i  S 8  g
     0. —  l>
   i -5  .S  o


   * «  '  g

   S *  ? .2
   c> ^  c +*
   trt «     «

     tf  C  «•>

   g 1.2 r

   "• L.  S  &

   I £  I *
   u^ oe  u "D
   t/> —    i

   D     - v
     -  -O 3
               U
                  1
s  ?
               in  k.

               si
M in KI         in    o in   to
OfNJ^'tnm^^'    QfNJfwp
OOPiMrwOOrv&pop


dddddo'dddddd
                                                       X

                                                       £
               s    *. I *.  i  I       |i«
               •t^Csr— -OCKI.C--I*
               £rl2S;|^-ffej;

               vibi^sll^SLzfr
               ->«)X — d —  CD_J»(«>JUJ

               SgCtCfe5*8;2f58
               «&rvimh.7gg;?£p;s
               — s:r'SO''1«oKt^fr«~'v5
               Sf2t,E'*1^«^ic<£^*>?5
o
o

-------
                           Appendix E

                           IRIS Values:

1)  Reference Dose Description and Use in Health Risk Assessments

                              and

     2)  EPA Approach for Assessing the Risk Associated with
             Exposure to Environmental  Carcinogens

-------
                                REFERENCE DOSE (RfD):
                  DESCRIPTION AND USE IN HEALTH RISK ASSESSMENTS
PRINCIPAL AUTHOR:

Donald Barnes, Ph. D. (OPTS)

RfD WORK GROUP:

Donald Barnes, Ph.D. (OPTS)
Judith Bellin, Ph.D. (OSWER)
Christopher IDeRosa, Ph.D. (ORD)
Michael Dourson, Ph.D. (ORD)*
Reto Engler, Ph.D. (OPTS)
Linda Erdreich, Ph.D (ORD)
Theodore Farber, Ph.D. (OPTS)
Penny Fenner-Crisp.Ph.D. (OOW)
Elaine Francis, Ph.D. (OPTS)
George Ghali, Ph.D. (OPTS)
Richard Hill, M.D., Ph.D. (OPTS)

*Co-Chair
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)*
Aggie Revesz, 8.5. (OPTS)
RevaRubenstein.Ph.D. (OSWER)
Jerry Stara, D.V.M., Ph. D (ORD)
JeanetteWittse,Ph.D  (OPTS)
Larry Zaragosa, Ph.D. (OAR)
                                         E-2

-------
I. INTRODUCTION

This concept paper describes the U.S. Environmental Protection Agency's principal approach to and
rationale for assessing risks for health effects other than cancer and gene mutations from chronic chemical
Exposure. By outlining principles and concepts that guide EPA risk assessmentfor such systemic"'effects,
the report complementsthe new risk assessmentguidelines.which describethe Agency's approach to risk
assessmentin other areas (carcinogenicity, mutagenicity,developmentaltoxicity, exposure, and chemical
mixtures). See the IRIS glossary for a description and citation of each guideline.

A. Background

Chemicals that give rise to toxic end points other than cancer and gene mutations are often referred to as
"systemictoxicants" because of their effects on the function of various organ systems. It should be noted,
however, that chemicals which cause cancer and gene mutations also commonly evoke other toxic effects
(systemic toxicity).  Generally, based on our understanding of homeostatic and adaptive mechanisms,
systemic toxicity is treated as if there is an identifiable exposure threshold (both for the individual and for
the population) below which effects are not observable. This characteristicdistinguishessystemicendpoints
from carcinogenic and mutagenic end points, which are often treated as nonthreshold processes.

Systemic effects have traditionally been evaluated in terms of concepts such as  "acceptable daily intake"
and "margin of safety." The scientific community has identified certain limits on some of these approaches,
and these limits have been borne out in EPA's experience. Nonetheless.EPA is called upon to apply these
concepts in making and explaining decisions about the significance for human health of certain chemicals
in the environment.

To  meet these  needs, the  RfD  Work  Group  has drawn on  traditional concepts, as well as on
recommendationsin the 1983 National Academy of Sciences (NAS) report on risk assessment.to more fully
articulate the use of noncancer, nonmutagenicexperimentaldata in reaching decisions on the significance
of exposuresto chemicals. In the process, the Agency has coined new terminologyto clarify and distinguish
between aspects of risk  assessment and  risk management. EPA has tested and  implemented these
innovations in developing consistent information for several recent regulatory needs, for instance under
RCRA.

B. Overview

This Appendix consists of four parts in addition to this introduction.  In Section II, much of the traditional
information on assessing risks of  systemic toxicity  is presented, with the focus  on the concepts of
"acceptable daily intake (ADI)" and "safety factor (SF)."  Issues associated with these approaches are
identified and discussed.  In Section III, the Agency's approach to assessing the risks of systemic toxicity
is presentee! in the context of the NAS scheme of risk assessment and risk management in regulatory
decision-making.This approach includes recasting earlier ADI and SF concepts into the less value-laden
terms "referencedose (RfD)" and "uncertaintyfactor (UF)." A new term, "marginof exposure,"as utilized in
the EPA regulatory context,  is introduced to avoid some 'of the issues associated with  the traditional
approach.

Section IV examines how these new concepts can be applied in reaching risk managementdecisions.while
Section V briefly mentions some of the additional approaches the Agency is using and exploring to address
this issue. Section VI provides a sample RfD calculation.
*      In this document the term systemic refers to an effect otherthan carcinogenicity or mutagenicity induced by toxic ch emical

**     In this Appendix, the ratio of the NOAEL to the estimated exposure (often referred to as "margin of safety") is referred to
       as the "margin of exposure (MOE)" in order to avoid confusion with the original use of the term "margin of safety" in
       pharmacology (i.e., the ratio of the toxic dose to the theraputic dose) and to avoid the use of the value-laden term " safety



                                              E-3

-------
II. TRADITIONAL APPROACH TO ASSESSING SYSTEMIC (NONCARCINOGENIC) TOXICfTY

The Agency's approach to assessingthe risks associated with systemic toxicity is different from that for the
risks associated with carcinogenicity.  This is because different mechanisms of action arethought to be
involved in the two cases.  In the case of carcinogens, the Agency assumes that a small number of
moleculareiiientscan evoke changes in a singlecell that can lead to uncontrolledcellularproliferation. This
mechanismfor carcinogenesisis referredto as "nonthreshold.'since there is essentiallyno level of exposure
for such a chemical that does not pose a small, but finite, probability of generatinga carcinogenic response.
In the case of systemic toxicity, organic homeostatic, compensating, and adaptive mechanisms exist that
must be overcome before the toxic end point is manifested. For example, there could be a large number
of cells performingthe 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 individualthreshold hypothesis holds that
a range of exposures from zero to some finite value can be tolerated by the organism with essentiallyno
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 individualswrthin a population.

A. The Traditional Approach

In many cases, risk decisions on systemic toxicity have been made by the Agency using the concept of the
"acceptable daily intake (ADI)." This quantity is derived by dividing the appropriate "no-observed-adverse-
effect level (NOAEL)" by a "safetyfactor (SF)" as follows:*

                  ADI (human dose) = NOAEL (experimental dose) I SF       (1)

The AOI is often viewed as the amount of a chemical to which one can be exposed on a daily basis over
an extended period of time (usually a lifetime) without suffering a deleteriouseffect. Often, the AOI has been
used as a tool in reaching risk management decisions; e.g., establishing allowable levels of contaminants
in foodstuffs and water.

Once the critical study demonstrating the toxic effect of concern has been identified, the selection of the
NOAEL derives from an essentiallyobjective, scientific examination of the data available on the chemical in
question.

Generally, the SF consists of multiples of 10, each factor representinga specific area of uncertainty inherent
in the available data. For example, an SF may be developedby taking into account the expected differences
in responsivenessbetween humans and animals in prolonged exposure studies; i.e., a 10- fold factor.
In addition, a second factor of 10 may be introduced to account for variability among individuals with in the
human population. For many chemicals, the resultant SF of 100 has been judged to be appropriate. For
other chemicals, with a less complete data base (e.g., those for which only the results of chronic studies are
available), an additional factor of 10 (leading to an SF of 1,000) might be judged to be more  appropriate.
On the other hand, for some chemicals, based on well-characterized responses in sensitive humans (e.g.,
effect of fluoride on human teeth), an SF as small as 1 might be selected.
       A NOAFL 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 statistical or
       adverse biological effect. In some treatments, the NOAEL for the critical toxic effect is simply referred to as the NOEL
       This latter term, however, Invites ambiguity In that there may be observable effects which are of toxic- significance,  i e
       they are not "adverse." In order to be explicit, this Appendix uses the term NOAEL and it refers to the highest NOAEL m
       an experiment. Further, in cases in which a NOAEL has not been demonstrated experimentally, the formulation  calls for
       use of the "lowest-observed-adverse-effectleve! (LOAEL)". In order to focus on the major concepts, however, we will  use
       NOAEL as a general example.
                                              E-4

-------
While the original selection of SFs appears to have been rather arbitrary (Lehman and Fitzhugh, 1954)*,
subsequent analysis of data as reviewed by Dourson and Stara (1983) lends theoretical (and in some
instances experimental) support for their selection. Further, some scientists, but not all, within the EPA
interpretthe absence of widespread effects in the exposed human populations as evidenceof the adequacy
of the SFs traditionally employed.

B.  Some Difficulties in Utilizing the Traditional Approach

1. Scientific issues

While the traditional approach has performed well over the years and the Agency has sought to  be
consistent in its application, observers have identified scientific shortcomings of the approach. Examples
include the following:

o  By focusing on the NOAEL, information on the shape of the dose-responsecurve is ignored. Such data
   could be important in estimating levels of concern for public safety.

o   As scientific knowledge is increased and the correlation of precursor effects (e.g., enzyme induction)
   with frank toxicity becomes known, questionsabout the selectionof the appropriate "adverseeffect"arise.

o   Guidelineshave not been developed to take into account the fact that some studies have used larger
   numbers of animals and, hence, are generally more reliable than other studies.

Theseand other "genericissues"are not susceptibleto immediateresolirtion,because the data base needed
is not yet sufficiently developedor analyzed.  Therefore, these issues are beyond the scope of this Appendix.
However, the Agency has established a work group to consider them.

2. Management-relatedssues

a.  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.

b.  The implication that any exposure in excess of the ADI is "unacceptable"and that any exposure less than
   the ADI is "acceptable"or "safe"

In practice, the ADI is viewed by many as an "acceptable'level of exposure, and, by inference.any exposure
greater than the ADI is seen as  "unacceptable." This strict demarcation between what is "acceptable"and
what is "unacceptable"is contrary to  the views of most toxicologists, who typically interpretthe ADI as a
relatively crude estimate of a level of chronic exposure not likely to  result in adverse effects to humans. The
ADI is generally viewed as a "soft'estimate, whose bounds of uncertainty can span an order of magnitude.
That is, within reasonable limits, while exposures somewhat higher than the  ADI are associated with
increased probability of adverse effects, that probability is not a certainty. Similarly, while the ADI is seen as
a level at which the probability of adverse effects is low, the absence of risk to all people cannot be assured
at  this level.

c.  Possible limitations imposed on risk management decisions

Awarenessof the "softness'bf the ADI estimate (see b. above) arguesfor carefulcase-by-caseconsideration
of  the implications of the toxicological analysis as it applies to any  particular situation. To the degree that
ADIs generated by the  traditional approach are the determiningfactors in risk

*   Lehman, A. J. and Fitzhugh, O.G. (1954) Association of Food Drug Officials. USQ Bulletin 18:33-35.


                                             E-5

-------
managementdecisions.they can take a significance beyond that intended by the toxicologist or merited by
the underlying scientific support.

Further, in administeringrisk/benefit or cost/benefit statutes, the risk manager is requiredto considerfactors
otherthan risk (e.g., estimatedexposures compared to the ADI) in reaching a decision. The AID) is only one
factor in a managementdecision and should not prevent the risk manager from weighing the full range of
factors.

d. Development of different ADIs by different programs

In addition to occasionally selectingdifferentcritical toxic effects, Agency scientists have reflected their best
scientific judgments in the final ADI by adopting factors differentfrom the standard factors listed in Table A-
I.  For example, if the toxic end point for a chemical in experimental animals is the same as that which has
been establishedfor 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 differentprograms, exercisingtheir best scientific
judgment, have in many cases resulted in different ADIs for the same chemical. The fact that different ADs
were generated (e.g., by adopting different SFs) can be a source of considerable confusion when the ADIs
are applied in risk managementdecisionmaking(see c. above). For example, although they generally agree
on the experimental data base for 2,3,7,8-TCDD, regulatory agencies within the United States and around
the world have generated different ADIs by selecting different "safety factors"; specifically, 1000,500, 250,
and 100. These different ADIs have been used to justify different regulatory decisions. The existence of
different ADIs need not imply that any of them is more "wrong"-or "right"-thanthe rest.  It is more nearly
a reflection of the honest difference in scientific judgment.

These differences, which may reflect differences in the interpretation of the scientific data, can also be
characterized as differences in the management of the risk. As a result, scientists may be inappropriately
impugned, and/or perfectly 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.

III. EPA ASSESSMENT OF RISKS ASSOCIATED WITH SYSTEMIC TOXICITY

In 1983, the National Academy of Sciences published a report* which discussesthe conceptual framework
within which regulatory decisions on toxic chemicals are made; see Figure C*1. The determination of the
presence of risk and its potential magnitude is made during the risk assessmentprocess, which consists of
hazard identification, dose-response assessment, exposure assessment, and risk characterization.  Having
been apprised by the risk assessor that a potential risk exists, the risk manager answers the question:
"What, if anything, are we going to do about it?"

A.  Hazard Identification

1. Evidence
a. Type of effect

Exposure to a given chemical, depending on the dose employed, may result in a variety of toxic  effects.
These may range from gross effects, such as death, to more subtle biochemical, physiologic, or pathologic
changes. The risk assessor considers each of the toxic end points from all studies evaluated in assessing
the risk posed by a chemical, although primary attention usually is given to the effect exhibiting the lowest
NOAEL, often referred to as the critical effect.  For chemicals with a limited data base, there may be a need
for more toxictty testing.

*    NAS Risk Assessment in the Federal Goverment: Managing the Process (NAS Press, 1963).


                                             E-6

-------
                                         FIGURE C-1
                     Dose-response
                     Assessment (e.g. RfD)
        Hazard
        Identification
                                      X
                 Exposure
                 Assessment
Risk
Characterization
(e.g. criterion)
                                                Control
                                                Options
Regulatory
Decision
(e.g. RgD, Standard)
                                                                      '
                         Non-risk
                         Analyses
b. Principal studies

Principal studies are those that contribute most significantly to the qualitative assessmentof whether or not
a particular chemical is potentially a systemic toxicant in humans. In addition, they may be used in the
quantitative dose-response assessment phase of the risk assessment These studies are of two types:

(1) Human studies

Human data are often useful in qualitatively estaWishingthepresenceof an adverse effect in exposed human
populations. Further, when there is information on the exposure level associated with an appropriate end
point, epidemiologicstudiescan also provide the basis for a quantitative dose-responseassessment. Use
of these latter data avoids the necessity of extrapolating from animals to humans, and therefore, human
studies, when available, are given first priority, with animal toxicrty studies serving to complement them.

In epidemiologicstudies, confounding factors that are recognized can be controlled and measured, within
limits. Case reports and acute exposures resulting in severe effects provide support for the choice of critical
toxic  effect,  but they are often of limited  utility in  establishing a  quantitative relationship  between
environmental exposures and anticipated effects. Available human studies on ingestion are usually of this
nature. Cohort studies and clinical studies may contain exposure-response information that can be used
in estimating effect levels, but the method of establishing exposure must be evaluated for validity and
applicability.

(2) Animal studies

Usually, the data base on a given chemical lacks appropriate information on effects in humans.  In such
cases, the principal studies are drawn from experiments conducted on non-human mammals, most often
the rat, mouse, rabbit, guinea pig, hamster, dog, or monkey.

c. Supporting studies

Supporting studies include information from a wide variety of sources. For example, metabolic and other
pharmacokinetic studies can provide insights into the mechanism of action of a particular compound. By
comparing the metabolism of the compound exhibiting the toxic effect in the animal with the metabolism
found in humans, some light may be cast on the potential for the toxic manifestation in humans or for
estimating the equitoxic dose in humans.

Similarly, in vitro studies can provide insights into the compound's potential for biological activity, although
                                              E-7

-------
a  definite connection to the  human experience cannot be  drawn.   Under certain  circumstances,
considerationof structure-activityrelationshipsbetweenthe chemicalundertest and the effects of structurally
related agents can provide a clue to the biological activity of the former.

At the present time, these data  are supportive, not definitive, in assessing risk. However, there i focused
activity aimed at developing more reliable in vitro tests to minimizethe need for live-animaltesting. Similarly,
there is increased emphasis on generating mechanism-of-action and  pharmacokinetic information as a
means of increasing the fundamental understanding of toxic processes in humans and nonhumans.  It is
expected that in the future these considerations will play a larger role in our determination of toxicity of
chemicals.

d. Route of exposure

The Agency often approaches the investigation of a chemical with a particular route  of exposure in mind;
e.g., an oral exposure for a driving water contaminant or a residue in food.  Although the route of exposure
is oral  in both cases, specific considerations may differ, for example, the bioavailability of the chemical
administered in food may differ from that when administered in water or inhaled. Usually, the toxicologic
data base on the compound does not include detailed testing on all possible routes  of administration.

In general, it is the Agency's view that the poiential for toxicity  manifested by one  route of exposure is
relevant to any other route of exposure, unless convincing evidence exists to the contrary Consideration is
always given to potential differencesin absorption or metabolism resultingfrom differentroutesof exposure,
and wheneverappropriatedata (e.g., comparativemetabolismstudies)are available.thequantrtatrveimpacts
of these differences on the risk assessment are fully delineated.

e. Length of exposure

The Agency is concerned about the potential toxic effects in humans associated with all possible exposures
to chemicals. The magnitude, frequency, and duration of exposure may vary considerably in different
situations. Animal studies are conducted using a variety of exposure durations (e.g., acute, subchronic, and
chronic) and schedules(e.g.,single,intermitterrt,orcontinuousdosing). Information from all of thesestudies
is useful in the hazard identification phase of  risk assessment. For example, overt neurological problems
identified in high-dose acute studies tend to reinforce the observation of subtle neurological changes seen
in a low-dose chronic study. Special concern exists for low-dose, chronic exposures, however, since such
exposures can elicit  effects absent in higher-dose, shorter exposures, through mechanisms such  as
accumulation of toxicants in the organisms.

f.  Quality of the study

Evaluation of individual  studies in humans  and  animals  requires  the consideration of several factors
associated with a study's hypothesis, design, execution, and interpretation. An ideal study addresses a
clearly delineated hypothesis, follows a carefully prescribed protocol, and includes sufficient subsequent
analysis to support its conclusions convincingly.

In evaluatingthe results from such studies, consideration is given to many other factors, indudingchemical
characterization of  the compound(s) under study, the type of test species, similarities and differences
betweenthe testspeciesand humans (e.g., chemical absorption and metabolism),the number of individuals
in the study groups, the number of study groups, the spacing and choice of dose levels tested, the types
of observations and methods of analysis, the nature  of pathologic changes, the  alteration in metabolic
responses, the sex  and age of test animals, and the route and duration of exposure.

2. Weight-of-Evidence Determination

As the culmination of the hazard identification step, a discussion of the weight-of-evidencesummarizes the
highlightsof the information  gleaned from the entire range of principal and supporting studies.  Emphasis
in the analysis is given to examining the results from different studies to determine the extent to which a


                                              E-8

-------
consistent, plausible picture of toxicity emerges. For example, the following factors add to the weight of the
evidencethat 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
betweenthe chemical and evidenceof the effect of concern in humans. The greaterthe weight-of-evidence,
the greater one's confidence in the conclusions drawn.

B. Dose-Response Assessment

1. Concepts and Problems

Empirical observation generally reveals that as the dosage of a toxicant is increased, the toxic response (in
terms of severity and/or incidenceof effect) also increases. Thisdose-responserelationshipis well-founded
in the theory and practice of toxicology and pharmacology.  Such behavior is observed in the following
instances: in quanta) responses in which the proportion of responding individuals in a population increases
with dose; in graded responses, in  which the severity of the toxic response within an individual increases
with dose; and in continuous responses, in which changes in a biological parameter (e.g., body or organ
weight) vary with dose.

However, in evaluating a dose-responserelationship, certain dirficultiesarise. For example, one must decide
on the critical end point to measure as the "response."One must also decide on the correct measure of
"dose." In addition to the interspeciesextrapolation aspects of the question of the appropriate units for dose,
the more fundamentalquestionofadministereddose versus absorbed dose versus target organ dose should
be considered. These questions are the subject of much current research.

2. Selection of the Critical Data

a. Critical study

Often animal data are selectedas the governing information for quantitative risk assessments^ince available
human data are generally insufficient for this purpose. These animal studies typically reflect situations in
which exposure to the toxicant has been carefully controlled and the problems of heterogeneity of the
exposed population and concurrent exposures to other toxicants have been minimized. In evaluatinganimal
data, a series of professional judgments are made that involve, among others, consideration of the scientific
quality of the studies. Presented with data from several animal studies, the risk assessor first seeks to
identifythe animal model that is most relevant to humans, based on the most defensiblebiological rationale,
for instance using comparative pharmacokinetic data.  In the absence of a clearly most relevant species,
however, the most sensitive species (i.e., the species showing a toxic effect at the lowest admininistered
dose) is adopted as a matter of scientific policy at EPA, since no assurance exists that humans are not
innately more sensitivethan any speciestested. This selection process is made more difficult if animal tests
have been conducted using different routes of exposure, particularly if the routes are different from those
involved in the  human situation under investigation.

In any event, the use of data from carefully controlled studies of  genetically homogeneous animals
inescapably confronts the risk assessor with the problems of extrapolaing between species and the need
to account for human heterogenertyand concurrent human exposuresto other chemicals,which may modify
the human risk.

While there is usually a lack of well-controlled cohort studies that investigate non-cancer end points and
human exposure to chemicals of interest, in some cases human data may be selectedas the Tcritical  data
(e.g., in cases of cholinesteraseinhibition). Risk assessments based on  human data have the advantage
of avoiding the problems inherent in interspeciesextrapolation. In many instances, use of such studies, as
is the case with the animal investigations, involves extrapolation from relatively high doses (such as those
found in occupational settings) to the low doses found in the environmentalsituationsto which the general
population is more likely to be exposed. In some cases, a well-designedand well-conductedepidemiologic


                                             E-9

-------
study that shows no association between known exposures and toxictty can be used to directly project an
RfD (as has been done in the case of fluoride).

b. Critical data

In the simplest terms, an experimental exposure level is selectedfrom the critical study that represents the
highest level tested in which "no adverse effect" was demonstrated. This "no-observed-adverse-effedevel"
(NOAEL) is the key datum gleanedfrom the study of the dose-responserelationshipand, 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 biologicallysignificantincreasein frequency or severity of an adverse effect
between individuals in an exposed group and those in its appropriate control. (See also discussion in the
footnote on page A-4). As noted above, there may be sound professional differencesof opinion in judging
whether or not  a particular response is adverse. In addition, the NOAEL is a function of the size of the
population understudy. Studies with a small number of subjects are less likely to detect low-dose effects
than studies using larger numbers of subjects. Also, if the interval between doses in an experiment is large,
it is possible that the experimentally determined NOAEL is lower than that which would be observed in a
study using intervening doses.

c. Critical end point

A chemical may elicit more than one toxic effect (end point), even in one test animal, or in tests of the same
or differentduration (acute, subchronic, and chronic exposure studies). In general,NOAELs for theseeffects
will differ. The critical end point used in the dose-response assessment is the one at the lowest NOAEL

3. Reference Dose (RfD)

In response to many of the problems associated with AIDIs and SFs, which were outlined in Section II, the
concept of the "reference dose (RflD)" and "uncertainty factor (UP)* is recommended.  The RflO  is a
benchmark dose operationally derived from the NOAEL by consistent application of generally  order of
magnitude uncertainty factors (UFs) that reflect various types of data used to estimate RfDs (for example,
a valid chronic human NOAEL normally is divided by an UF of 10) and an additional modifying factor (MF),
which is based  on a professional judgment of the entire data base of the chemical.* See Table A-l.

The RfO is determined by use of the following equation:

                               RfD= NOAEL/(UFxMF)         (2)

which is the functional equivalent of Eq. (1).  In general, the RfD is an estimate (with uncertainty spanning
perhaps an order of magnitude)of a daily exposure to the human population (includingsensrtrvesubgroups)
that  is likely to be without an appreciable risk of deleterious effects during a  lifetime.   The RfD is
appropriately expressed in units of mg/kg-bw/day.

The RfD is useful a referencepoint for gauging the potential effects of other doses.  Usually, doses that are
less than the RflD are not likely to be associated with any health risks, and are therefore less likely to be of
regulatory concern. However, as the frequency of exposures exceedingthe RfO increases.and as the size
of the excess increases, the probability increases that adverse effects may be  observed in a human
population. Nonetheless.a clear conclusion cannot categorically drawn that all doses below the RfD are
"acceptable'and that all doses in excess of the RfD are "unacceptable."

*       "Uncertainty factor" is the new description applied to the term "safetytactor" see Page A-4). This new name is more
        descriptive in that these factors represent scientific uncertainties, and avoids the risk management connotation of "safety
        The "modifying factor" can range from  greater than zero to 10, and reflects qualitative professional judgements regardin g
        scientific  uncertainties not covered under the standard UF, such as the completeness of the overall data base and the
        number of animals in the study.

                                             E-10

-------
                                         TABLE A-l.
GUIDELINES FOR THE USE OF UNCERTAINTY FACTORS IN DERIVING REFERENCE DOSE (RfD)


Standard Uncertainty Factors (UFs)

Use a  10-fold factor when extrapolating from valid experimental results from studies using prolonged
exposureto average healthy humans. This factor is intendedto account for the variation in sensitivityamong
the members of the human population. [1 OH]

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
intendedto account for the uncertainty in extrapolating animal data to the case of humans. [10A]

Use an additional 10-fold factor when extrapolating from less than chronic results on experimental animals
when there are no useful long-term human data. This factor is intendedto account for the uncertainty in
extrapolating from less than chronic NOAELs to chronic NOAELs.  [10S]

Use an additional  10-fold factor when deriving a RfD from a LOAEL, instead of a NOAEL This factor is
intended to account for the uncertainty in extrapolating from LOAELs to NOAELs.  [1OL]

Modifying Factor (MF)

Use professional judgmentto determineanother uncertainty factor (MF) which is greaterthan zero and less
than or equal to 10.  The magnitude of the MF depends upon the professional assessment of scientific
uncertainties of the study and database not explicitly treated above; e.g., the completeness of the overall
data base and the number of species tested. The default value for the MFisl.

SOURCE:     Adapted from Dourson, M.L; and Stara, J.F. (1983)
              Regulatary Toxicology and Pharmacology 3:224-238.


(This is a consequence of the inability of either the traditional or the RfD approach to  completely address
the question of dose-response extrapolation.)

The Agency is attempting to standardize its approach to determining RfDs.  The RfD Work Group has
developed a systematic approach to summarizing its evaluations, conclusions, and reservations regarding
RfDs in a "cover sheet" of a few pages in length. The cover sheet includes a statement on the confidence
the evaluators have in the stability of the RfD: high, medium, or low.  High confidence indicates that the RfD
is unlikely to change in the future because there is  consistency among the toxic responses observed in
different sexes, species, study designs, or in dose-responserelationships,or the reasons for differences, if
any, are well understood. Often, high confidence is given to RfDs that are based on  human data for the
exposure route of concern, because in such cases the problems of  interspeciesextrapolation are avoided.
Low confidence indicates that the RfD may be especially vulnerable to change if additional chronic toxicity
data are published on the chemical, because the data supporting the estimation of the RfD are of limited
quality and/or quantity.

C. Exposure Assesment

The third step in the risk assessmentprocess focuses on exposure issues. For a full discussion of exposure
assessment.the reader is referredto EPA's recently published guidelineson the subject  ($1 Federal Register
34042-340S4,Sept. 24,1986). There is no substantive difference in the conceptual approach to exposure
assessmentin the case of systemic toxicants and of carcinogens.

In brief, the exposure assessment includes consideration of the populations exposed  and the magnitude,
frequency, duration and routes of exposure, as well as evaluation of the nature of the exposed populations.


                                            E-11

-------
D. Risk Characterization

Risk characterization is the final step in the  risk assessment process and the first step in the  risk
managementprocess. Its purpose is to present to the risk managera synopsis and synthesis of all the data
that contribute to a conclusion on the risk, including:

o      The qualitative ("weight-of-evidence") conclusions about the likelihood that the chemical
       may pose a hazard to human health.

o      A discussion of dose-responseand how this information, through the use of particular uncertainty
       and modifying factors, was used to determinethe RfD.

o      Data such as the shapes and slopes of the dose-responsecurves for the various toxic end points
       toxicodynamics (absorption and metabolism), structure-activity correlations, and the nature and
       severity of the observed effect. These data should be clearly discussed by the risk assessor, since
       they may influence the final decision of the risk manager (see below).

o      The estimates of exposure, the nature of the exposure, and the number and types of people
       exposed, together with a discussion of the uncertainties involved.

o      A discussion of the sources of uncertainty, major assumptions, areas of scientific judgment, and,
       to the extent possible, estimates of the uncertainties embodied in the assessment.

In the risk characterization process, comparison is made between the RfD and the estimated (calculated or
measured) exposure dose (EED), which should consider exposure by all sources and routes of exposure.
The risk assessmentshould contain a discussion of the assumptions underlying the estimation of the RfD
(nature of the critical end point, nature of other toxic end points, degree of confidence in the data base,
etc.), and the degreeof conservatism in its derivation. The assumptions used to derive the EED should also
be discussed. If the EED is less than the RfD, the need for regulatory concern is likely to be small.

An alternative measure that may be useful to some risk managers is the "margin of exposure (MOE)" (see
footnote on p. A-3), which is the magnitude by which the NOAEL of the critical toxic effect exceeds the
estimated exposure dose (EED), where both are expressed in the same units:

                MOE = NOAEL (experimental dose)/EED (human dose)         (3)

In parallel to the statements above on EED and RfD, the risk assessmentshould contain a discussion of the
assumptions undertyingthe estimatesof the RfD and the degreeof possible conservatism of the UF and MF.
It can be noted that when the MOE is equal to or greater than UF x MF, the need for regulatory concern is
likely to be small.

Section VI contains an example of the use of the concepts of NOAEL, UF, MF, RfD, and MOE.

IV.  APPLICATION IN RISK MANAGEMENT

Once the risk characterization is completed, the focus turns to risk management. In reaching decisions, the
risk manager must consider a number of risk factors, nonrisk factors, and regulatory options that influence
the final judgment. It is generally useful to the risk managerto have information regarding the contribution
to the RfD from various environmental media.  Such information can provide insights that are helpful in
choosing among available control options. However, in cases in which site-specific criteria are being
considered,local exposures through various media can often be detenminedmore 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 ingestionof treated crops.
                                            E-12

-------
RfOs should be apportioned by route of exposure. Where specific exposure analysis can be made, such
apportionment is readily performed. If exposure information is not available, assumptions must be made
concerningtherelativecontributionsfromdifferentroutes of exposure. At present,differentEPA offices use
assumptions that differ to some degree. These assumptions are being reviewed by an Agency risk
assessment group.

As illustrated in Figure A-1, the risk manager utilizes the results of risk characterization, other technological
factors, and nontechnical social and economic considerations in reaching a regulatory decision. Some of
these factors include efficiency,
timeliness, equity, administrative simplicity, consistency, public acceptability, technological feasibility, and
legislative mandate.

Because of the way these risk management factors may impact different cases, consistent-but not
necessarilyidentical-riskmanagementdecisionsmust 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 (FIF~)  calls for "an ample margin of safety," taking benefits into account; and the Safe
Drinking Water Act (SOWA) calls for standards that protect the public 'to the extent feasible."Consequently,
it is entirely possible and appropriate that a chemical with a specific RfD may be regulated under different
statutes and situations through the use of different "regulatory doses (RgOs)".

Expressed in general terms, after carefully considering the various risk and nonrisk factors, regulatory
options, and statutory mandatesin a given case (i), the risk managerdecidesupon the appropriate statutory
alternatives to arrive at  an "ample"or "adequate"margin of exposure [MOE(i)], thereby establishing the
regulatory dose, RgD(i) (e.g.,  a tolerance under FIFRA or a maximum contaminant level under SDWA),
applicable to that case:

                             RgD(i)=  NOAELVMOE(i)            (4)

Note that, for the same  chemical (with a single RfD), the risk manager(s) can develop different regulatory
doses  for different situations that may involve different exposures, available control options, alternative
chemicals, benefits, and statutory mandates. Also note that comparing the RfD to a particular RgD(i) is
equivalentto comparing the MOE(i) with the UF x MF:

                          RfD/RgD(i) = MOE(i)/UF x MF           (5)

In assessingthe significance of a case in which the RgD is greater (or less) than the RfD, the risk manager
should carefully consider the case-specific data laid out by the risk assessors, as discussed in in Section
III. D. 4. In some cases this may require additional explanation and insight from the risk assessor. In any
event, the risk manager has the responsibility to clearly articulate the reasoning leading to the final RgD
decision.

V. OTHER DIRECTIONS

Whilethe Agency is in the process of systematizingthe approach outlined in this Appendix, risk assessment
research for systemic toxicity is also being conducted along entirely separate lines. For example, the Office
of Air  Quality  Planning and Standards is using probabilistic risk assessment procedures for  criteria
pollutants.  This procedure characterizes the population at risk,  and the likelihood of  various  effects
occurring, through the use of available scientific literature and elicitation of expert judgment concerning
dose-responserelationships. The dose-responseinfomnation is combinedwith exposure analysis modeling
to generatepopulation risk estimatesfor altemativestandards. Theseprocedurespresentthe decisionmaker
with ranges of risk estimates,and explicitly considerthe uncertainties associated with both the toxicity atid
exposure information. The Office of Policy, Planning, and Evaluation is investigating similar procedures in
order to balance health risk and cost. In addition, scientists in the Office of Research and Development
have initiated a series of studies that should lead to future improvements in risk estimation. First, they are
investigatingthe use of extrapolation models as well as the statistical variability of the NOAEL and underlying
IJFs as means of estimating RfDs. Second, they are exploring procedures for less-than-lifetimehealth risk


                                             E-13

-------
assessment. Finally, they are working on ranking the severity of toxic effects as a way to further refine
EPA's health risk assessments. While these procedures are promising, they cannot be expected at this time
to serve as a foundation of a generalized health risk assessmentfor systemic toxicrty in the Agency.
VI.  HYPOTHETICAL, SIMPLIFIED EXAMPLE OF DETERMINING AND USING RfD
Suppose the Agency had a sound clay subchronic gavage study in rats with the following data:
A. Experimental Results
                                     Observation
 Dose
(mg/kg-day)
  0
  1
                             Control - no adverse effects observed
                             No statistical or biological significant
                            differences between treated and control animals
                             2% decrease* in body weight gain (not
                            considered to  be of biological significance)
                            Increased ratio of liver weight to body weight
                            Histopathology indistinguishabfefrom controls
                            Elevated liver enzyme levels
                                                                         Effect Level


                                                                            NOEL

                                                                             NOAEL
   25
                                                                             LOAEL
                             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.
B. Analysis
1. Determination of the Reference Dose (RfD)

a. From the NOAEL
   UF= 10Hx10Ax105=  1000
   MF = 0.8, a subjective adjustment based on the fact that the experiment involved an astonishing 250
             animals per dose group
   Therefore UFxMF=  800, so that
   RfD =  NOAEL/(UF x MF) = 5 mg/kg-day/800 = 0.006 mg/kg-day
b. From the LOAEL (i.e. if a NOAEL is not available)
   If 25 mg/kg-day had been the lowest dose tested,
                                            E-14

-------
   UF= 10Hx10Ax105x10L= 10,000

   MF = 0.8

   Therefore UF x MF = 8,000, so that

   RfD = LOAEL7(UF x MF) = 25 (mg/kg-day)/8000  = .003 mg/kg-day)

2. Risk Characterization Considerations

Suppose the estimated exposure dose (EED) for humans exposed to the chemical under the proposed use
pattern were .01 mg/kg-day;  i.e.,

   EED>RfD

   Viewed alternatively, the MOE is:

   MOE =  NOAEL/EED =  5 mg/kg-day /0.01 mg/kg-day = 500

Because the EEO exceeds the RfD (and the MOE is less than the UF x MF), the risk manager will need to
look carefully at the data set, the assumptions for both the RfD  and the exposure estimates, and the
comments of the risk assessors. In addition, the risk manager will need to weigh the benefits associated
with the case, and other nonrisk factors, in reaching a decision on  the regulatory dose (RgD).
                                          E-15

-------
       EPA APPROACH FOR ASSESSING THE RISK ASSOCIATED WITH EXPOSURE TO
                           ENVIRONMENTAL CARCINOGENS
PRINCIPAL AUTHOR:

Robert E. McGaughy, Ph.D. (ORD)

CARCINOGEN RISK ASSESSMENT VERIFICATION ENDEAVOR (CRAVE) WORK GROUP:

Larry D. Anderson, Ph.D. (OW)              Elizabeth H. Margosches, Ph.D (OPTS)
Diane D. Seal, Ph.D. (OPTS)                Robert E. McGaughy, Ph.D. (ORD)
Judith Bellin, Ph.D. (ORD)                  Stephen Nesnow, Ph.D. (ORD)
Chao W. Chen, Ph.D. (ORD)*               Peter Preuss, Ph.D. (ORD)*
Ha Coty, Ph.D. (ORD)                     John A. Quest, Ph.D.(OPPE)
Theodore Farber, Ph.D. (ORD)              Reva Rubenstein.Ph.D. (OSWER)
William Farland, Ph.D. (ORD)                Rita Schoeny, Ph.D. (ORD)*
Herman J. Gibb, B.D.,M.P.H.(ORD)           Dot G. Wellington, Ph.D. (OPPE)
Richard Hill. M.D., Ph.D. (OPTS)

* Co-Chair
                                       E-16

-------
I.  INTRODUCTION

In the analysis of data regarding the potential human carcinogenicity of chemicals, the Agency uses the
approach described in the document entitled Guidelinesfor Carcinogen Risk Assessment(51 PR 33992-
34003, Sept. 24, 1986).  This approach had its origins in the 1976 Interim Guidelinesfor Health Risk and
Economic Impact Assessments of Suspected Carcinogens (41 FR  21402-21405), which describes the
conceptual basis of carcinogen risk assessment.  The approach is  consistent with the broad scientific
principles of carcinogen risk assessmentdevelopedby the Office of Science and Technology Policy (OSTP)
(50 FR 10372-10442),and the EPA guidelinesquote the OSTP principles extensively. Detailed applications
of the procedures currently used by the Agency are described in two  documents: (1) Health Assessment
Document for Epichlorohydrin, p.  7-32 to  7-48 (EPA  600/8-83-032F, December, 1984);  and (2) Off
Assessment of  Health  Risk of Garment Workers  and Certain Home  Residents from  Exposure to
Formaldehyde, Appendix 4 (April, 1986).

The Agency approach follows the general format of the National Academy of Sciences (NAS) description
of the risk assessment process (see Risk Assessmentin  the Federal Government: Managing the Process
[NAS Press, 1983]). In that report, the four elements of the risk assessment
process are defined as follows:

(1)    Hazard identification, in which a determination is made of whether human exposure to the agent in
       question has the potential to increase the incidence of cancer.

(2)    Dose-response assessment, in which a quantitative relationship is derived between the dose, or
       more generally the-human exposure, and the probability of induction of a carcinogenic effect.

(3)    Exposure assessment, in which an  evaluation is made of the human exposure to the agent.
       Exposure assessments identify the exposed  population, describe its composition and size, and
       present the type, magnitude, frequency, and duration of exposure.

(4)    Risk characterization, in which the exposure and dose-response assessments are combined to
       produce  a  quantitative risk estimate, and  in  which  the strengths and weaknesses, major
       assumptions, judgments, and estimates of uncertainties are discussed.

The carcinogen summary sheets included in  the IRIS system are designedto supply concise information
about the hazard identification and dose-responseassessmentsteps in this overall process. In orderto use
this information, individuals who wish to  estimate geographic site-specific risks must be able to do an
exposure evaluation based on the  information available, and must  be able to  combine the first three
elements into a  comprehensive risk characterization which can support  regulatory decision.  The risk
assessmentprocess is an activity independentof the process of formulating regulatory control options being
considered and  independentof economic and political factors influencing the regulatory process.  The
Agency  recognizes the distinction between these regulatory concerns (referred to as "risk management
considerations "in the 1983 NAS report) and the risk assessmentprocess.

II. ELEMENTS OF CARCINOGEN RISK ASSESSMENT

A. Hazard Identification

The purpose of this evaluation  is to arrive at some conclusions as  to whether or not the agent poses a
carcinogenic hazard in exposed populations. The main types of evidence bearing on this question are: (1)
human studiesof the association between cancer incidenceand exposure; and (2) long-term animal studies
under controlled laboratory conditions. Other evidencedsuch as short-term tests for genotoxicity, metabolic
and pharmacokineticproperties.toxicological effects other than cancer, structure-activity relationships, and
physical/chemical properties of the agent, is ancillary to the primary evidence.

The question of the likelihood that the agent is a  human carcinogen is answered by considering all of the
available information re levant to  carcinogenicity.by judgingthe quality of the studies available, by attempting


                                            E-17

-------
to reconcile any differences found between studies an coming to an overall evaluation.  This process is
termed the weight-of-evidenceapproach, and the results are expressed in terms of an EPA stratification
system for the weight of this evidence. The system, which is a modification of the approach taken by the
International Agency for Research on Cancer (IARC),* classifies the likelihood that the agent is a human
carcinogen into the following five categories:

       Group        Description

        A           Human Carcinogen

       B1 or 62      Probable Human Carcinogen

        B1           indicates that limited human data are available

        82          indicates sufficient evidence in animals and inadequate or no evidence in humans

        C           Possible Human Carcinogen

        D           Not Classifiable as to human carcinogenictty

        E           Evidence of Non-Carcinogenicityfor Humans

In making this classification for an agent, a two-stage procedure is followed. In the first stage, a provisional
classification is made based on the degree of human and animal evidence.  The degree of evidence is
characterized separately for both human studies and animal studies as sufficient, limited, inadequate, no
data, or evidence of no effect.  The guidelines broadly define the meaning of these terms, which are
basically the same as the IARC definitions. In the second stage, EPA scientists adjust these provisional
classifications upwards or downwards, based  on the supporting evidence of carcinogenicfty described
earlier, using judgments about the degree of adjustmentwarranted in each case. For further description of
the role of supporting evidence, see the EPA Guidelines.

B. Dose-Response

The purpose of the dose-response assessment is to define the relationship between the dose of an agent
and the likelihood of a carcinogenic effect, on the assumption that the agent is a human carcinogen. After
the dose-response assessment is made, it is combined with the exposure evaluation to yield a numerical
estimate of risk. Numerical estimates can be presented in one or more of the following four ways: 1) unit
risk, 2) the concentration corresponding to a given level of risk, 3)  individual, and 4) population risk.  The
summary sheets include only unit risk and risk-related air and water concentrations. The numerical risk
estimation activity is not dependenton the likelihood of human carcinogenicrty.as categorized in the hazard
identification process. Instead, it is an independentpiece of information which is to be combined with the
hazard identification in making  regulatory decisions.

As the Guidelinesobserve, dose-responseassessment "usuallyentails an extrapolation from the generally
high doses administered to  experimental animals or exposures noted in epidemiologic studies to the
exposure levels expected from  human  contact with the agent in the environment; it also  includes
considerations of the validity of these extrapolations." Extrapolation is ordinarily carried out first by fitting
a mathematical model to the observed data and then by extending either the model or a bound on the risks
it predicts in the observed range down toward risks expected at low exposure.

The main elementsof a dose-responseassessmentare: (1) the selection of the appropriate data sets to  use;
(2) the derivation of estimates at  low doses from experimental data at high doses using an extrapolation
model; and (3) the choice of an equivalent human dose corresponding to the animal dose used.

*      IARC. (1982) lARCMonographson the Evaluation of the Carcinogenic Risk of Chemicals to Humans, Supplement 4lyon.
       France.


                                            E-18

-------
1. Choice of Data Sets

In choosing the appropriate data sets to use, the main principles are as follows:

(a)     Human data are preferable to animal data, provided that quality is adequate.

(b)     Data from a species which responds biologically most like humans (with respect to factors such as
       metabolism, physiology, and pharmacokinetics) are used. When no clear choice is possible on this
       basis, data corresponding to the most sensitive animal species/strain/sex combination are given
       the greatest emphasis.

(c)     The  route of administration which is the route of human exposure is used. When this is not
       possible, the route differences are noted as a source of uncertainty.

(d)     When the incidence of tumors is significantly elevated at more than one site by the agent, risk
       estimates are made by determiningthe number of animals with one or more of these tumor sites.

(e)     Benign tumors are generally combined with malignant tumors, unless the benigh tumors are not
       considered to have potential to progress to the associated malignancies of the same histogenic
       origin. See Guidelinesfor Combining Neoplasms for Evaluation of Rodent CarcinogenesisStudies
       (1986). McConnell, E.E., Solleveld.HA, Swenberg.JA, Boorman, G.A.  JNCI 86:283-289.

2. Choice of Extrapolation Model

Since risk at low exposure levels cannot be  measured directly either  by animal  experiments or by
epidemiologic studies, a number of mathematical models and  procedures  have been developed to
extrapolate from high to low dose. Different extrapolation methods may give reasonable fit to the observed
data but may lead to large differencesin the projected risk at low doses. In keeping with the Guidelinesand
the OSTP principles, the choice of low-dose extrapolation method is governed by consistency with current
understandingof the mechanism of carcinogenesisand not solely on goodness-of-fitto the observed tumor
data. When data are limited, and when uncertainty exists regardingthe mechanismsof carcinogenicaction,
the OSTP principles suggest that  models or procedures which incorporate low-dose linearity are preferred
when compatible with the limited information available.  The Guidelines recommend that the linearized
multistage procedure be employed in the absence of adequate information to the contrary.

The first step of the linearized multistage procedure, abbreviated by LM on the summary sheets, calls for
the fitting of a multistage model to the data. This is an exponential model approaching 100% risk at high
doses with a shape at low doses described by a polynomial function. When the polynomial is of first
degree, the model is equivalentto a one-hit or linear model, so called because at low doses it produces an
approximately linear relationship between dose and cancer risk.

The second step of the procedure estimatesan upper bound for risk by incorporating an appropriate linear
term into the statistical bound for the polynomial. At sufficiently small exposures, any higher-orderterms
in the polynomial will contribute negligibly, and the graph of the upper bound will look like a  straight line.
The slope of this line is called the slope factor on the summary sheets. Since the slope at higherexposures
could be differentthan at low exposures r some chemicals, this slope factor is generallynot valid when the
exposures are  sufficiently high. In the summary sheets the exposure corresponding to a risk of 1 /100 is
arbitrarily chosen as sufficiently high that the slope factor and the unit risks derived from  it should not be
used.

Other models that could be used-are the Weibull (W), Probit (P), Logit (LO) one-hit (OH),  and  gamma
multihit(GM) models. These models are defined in the IRIS Glossary.  Except for the one-hit model, these
models all tend to give the characteristic 5-shapes of many biological experiments,with varying curvature
and tail length. Their upper bounds tend to parallel the curvature of the models themselves unless a
procedure has been devised to provide otherwise, as is the case for the linearized multistage procedure.
The slope factor designatedon the summary sheets for these models is the slope of the straight line from


                                            E-19

-------
the upper bound at zero dose to the dose producing an upper bound of 1%.

Two alternative approaches have been used for dealing with the spontaneous background rate of tumor
occurrence in risk estimation. Both approaches are summarized by a slope factor.

One approach defines "added risk" as the difference between the total response rate under an exposure
condition and the background incidencein the absence of exposure. The corresponding equation is AR =
P(d) - P(O).  The other approach, called "extra risk", can be described as the "added risk" applied to that
portion of the population that did not show background tumors. The correspondingequation is ER =  [P(d) -
 P(0)]/[1 -P(O)].  "Extra risk" is the most commonly used approach, but the alternative approach, "added
risk", is being explored by the Agency for its utility in certain circumstances and has been used in several
cases. When the background response is sizable, "extra risk" is larger than "added risk", and when the
background is small, both types of risk are essentially equal.

3. Determination of Human Equivalent Doses

The human dose that is equivalentto the dose in an animal study is calculated using the assumption that
different species are equally sensitive to the effects of a toxin if they absorb the same dose per unit body
surface area. This assumption is made only in the absence of specific information relevant to equivalent
dose for that agent. Since surface area is approximately proportional to the two-thirds power of body
weight, the equivalentdose is milligramsper (body weight raised to the two-thirds power) per day.  It follows
that if the animal dose is expressed in units of mg/kg'day, the equivalent human dose, in the same units,
is smaller than the animal dose by a factor equal to the cube root of the ratio  of human weight to animal
weight. Since the Agency generally assumes a human weight of 70 kilograms, this factor becomes 13 for
mice with a weight of 30 grams, and 5.8 for rats with a weight of 350 grams. In the calculation of human
equivalent doses, the actual animal weight in the bioassay is used wheneverthat information is available;
otherwise, standard species weights are used.

In using animal inhalation experimentsto estimate lifetime human risks for partially-solublevapott or gases,
the air concentration is generally considered to be the equivalent d~ between species based on equivalent
exposure times; i.e., a lifetime exposure to a 1-ppm concentration in humans is assumed to produce the
same effect as a lifetime exposure to a  l-ppm concentration in animals. In the inhalation of particulates or
completely-absorbed gases, the amount absorbed per unit of body surface area is considered to be the
equivalentdose between species.

In order to evaluate human risks for both air and water contamination when only one route has been tested
in animals, additional assumptions with corresponding additional uncertaintiesmust be introduced. For this
reason, the summary sheets specify the route of exposure that was used for the calculation of air  and
drinking water unit risks.

4. Summary of Dose-ResponseParameters

Quantitative risk estimates have several uses, and the expression of the results should tailored to each use.
For comparing the carcinogenic characteristics of several agents, the cancer risk per unit absorbed dose
is a useful parameter. It could be expressed on a weight basis (e.g., milligrams of the substance absorbed
per kilogram body weight per day, mg/kg/day) or on a molar basis (e.g., m moles/kg/day). The low-dose
slope factor described on page B-3 is used for this purpose in the IRIS summary.

For determining the concentrations of air or water at certain designated levels of lifetime risk, the ratio of
that level of risk/unit risk for water or air is calculated. For example, if the water unit risk is 0.4 E-4 per ug/L,
the water concentration corresponding to an upper bound of E-5/(0.4  E-4)=0.25 ug/L

For evaluating risks to environmentalagents, the concentrations of the agent in the medium where human
contact occurs is the measure of exposure used. Therefore, the appropriate measure of doseresponseis
risk per concentration unit, with standardized conventions of exposure durations and of intake of each
medium being understood. These measures are called the unit risk for air and the unit risk for drinking


                                            E-20

-------
water.  The standardized duration assumption is understood to be continuous lifetime exposure. The
concentration units for air and drinking water are usually micrograms per cubic meter (ug/cu m) of air and
micrograms per liter (ug/L) of water, respectively. For food, the agents are usually identified in specific
foods (e.g., fish or com) which constitute characteristic fractions of the daily diet, so the amount of the
agents consumed per day in all food known or expected to contain  residues of the agent is the most
appropriate measure of exposure. For this use, the summary sheets provide the slope factor in units
adjusted for body weight (e.g., mg/kg/day). If a different fraction of the agent is absorbed in humans from
the human diet than is absorbed from the animal diet, an appropriate correction is needed when applying
the animal-derived value to humans.

In summary, the quantities appropriate for calculating upper bound risks for air, drinking water, and food
are, respectively, the air unit risk (risk per yg/cu.  m of air), the drinking water unit risk (risk per H~1  of
drinking water), and slope factor (risk per mg/kg/day of the agent). However, a smaller dose unit (e.g.,
yg/kg/day for dietary intake risk) is often used if the risk corresponding to the dose unit (e.g., mg/kg/day)
exceeds 10-2.

5. Statement of Confidence in Dose-ResponseParameters

A judgment about  the degree of confidence the Agency has in the accuracy of the risks derived from the
data is given in the summary sheets as a high, medium, or low rating. The factors increasing the Agency's
confidence in the accuracy of these risk bounds includes the following:

(1)     The existenceof experimental data to replace default assumptions.

(2)     Close agreement in the risk parameters derived from experiments in different animal species.

(3)     Similarity in the  route of exposure between the tested species and route of interest in humans.

(4)     The existence of experimental data on the effective dose for the exposure route of interest.

(5)     A large number of animals or people in the studies used.

(6)     A large number of dose groups or a large range of doses in the studies being used.

(7)     Sufficient purity  of the test agent so that contamination is not a factor in interpretation of results.

(8)     Similarities between the animal strain and humans as to metabolism and pharmacokineticsof the
       agent.

(9)     For human occupational studies, determinationof exposure or differentworksit.es as opposed to an
       average exposure for the entire workplace.

(10)    For epidemiologicstudies,exposure measurementsconcuirentto the period being evaluated (e.g.,
       time period of employment).

(11)    Lack of concurrent exposuresin epidemiologicstudies which would reasonably have been expected
       to modify the dose-response.

(12)    The ranking of epidemiologicstudy designs according to their usefulness in deriving accurate risk
       assessments:cohort case-control ecologic studies.

(13)    The epidemiologicstudies provided sufficient information on dose, duration of exposure, and age
       to permit one to separate the effects-of each on the dose-response relationship.

(14)    An adequate time period was allowed in epidomiologicstudies for a cancer latency period.
                                            E-21

-------
(15)    Time regimens of animal exposure are similar to those of human exposure.

The factors decreasingthe Agency's confidence, in addition to factors contrary to the points above, are as
follows:

(1)     The  use of non-continuous dosing when we  have reason to believe that there is an effective
       continuous dose but pharmacokinetic information is inadequate to esti'nate'rt.

(2)     The use of vehicles, such as com oil,  which may confound or interact wish the agent under study
       in producing tumors at specific sites.

(3)     Situations in which special test systems (such as mouse skin painting, strain A mouse pulmonary
       adenomas, and in vitro tests) are not similar enough to human systems to justify their use as a basis
       for human quantitative risk estimates.

(4)     Lack of concurrent control groups.


(5)     Poor animal husbandry.
                                            E-22

-------
        APPENDIX F






PROCEDURES FOR DETERMINING




  THE HARMONIC MEAN FLOW

-------
                           Appendix  F

         PROCEDURES FOR OBTAINING THE HARMONIC MEAN FLOW


Direct Calculation of the Harmonic Mean

     Direct calculation of the harmonic mean flow, hm[x], can be
done by retrieving several years of daily flow records at an
appropriate USGS gage, taking the reciprocal of each value,
calculating the average, and taking the reciprocal of the
average.  That is,

      hm[x] = N/S (1/x)                     (1)

where x is daily flow and N is the number of daily values.

     It is important to note that the harmonic mean must be
determined for the flow downstream (not upstream)  of the
discharger.  Unlike arithmetic means, harmonic means are not
additive.  Where the effluent flow may be a significant portion
of the streamflow, the downstream harmonic mean is not the
upstream harmonic mean plus the effluent flow.  Rather it is the
harmonic mean of the combined daily upstream plus effluent flows:

      hm[downstream] = N /Z {!/(upstream + effluent)}        (2)


     It might also be noted that the harmonic mean is zero if any
value in the distribution is zero.  However, the dilution can
never actually be zero unless the pollutant load were discharged
without an accompanying effluent flow,  or if the downstream flow
were evaporated to dryness (as opposed to disappearing beneath
the stream bed, or temporarily accumulating behind a controlled
outlet dam during periods of zero release from the impoundment).

     As the analyst may not wish to undertake the computer
programming needed to obtain the reciprocals of daily flows for
the period of record, it is useful to consider simpler
approximations.  Such approximations can be made using data from
the "duration table" generated by the USGS FLOSTAT program.  This
data is very easy to obtain.   STORET users simply (a) log on to
the NCC computer system, (b)  issue the command WQAB FLOW, and in
response to prompts, (c) supply the USGS gage number and years to
be retrieved,  (d) select option 2 (data analysis), and (e)
specify the remote terminal where the data is to be printed.  The
WQAB FLOW procedure is one of several WQAB procedures intended to
be of help to permit writers.  Other procedures can assist in
finding a USGS gage near a discharger having a particular NPDES
number.  STORET users may issue the command WQAB HELP to obtain a
summary of available procedures.


                               F2

-------
     An example of duration table output and accompanying
statistics is attached.  Summarizing the entire period retrieved,
the table indicates the number of occurrences of each of 34
classes of flow magnitude.  It shows, for example,  four
occurrences of flows of approximately 51 cfs, 31 occurrences  of
flows of approximately 60 cfs, and so forth up to flows of 9000
cfs.  If n  is the number of occurrences of flows having
magnitude x , then the harmonic mean can be approximated by:

          hm[x] = N/S  (n,  /X,)                 (3)

This procedure should generally yield a result close to
Equation 1.


Indirect Methods for Approximating the Harmonic Mean

     Even simpler methods for estimating the harmonic mean are
derived below.  These methods, however, hinge on the flows being
log-normally distributed, and are thus subject to greater error.

     Given a log-normally distributed variable, x,  the following
derivation shows the relationship between its harmonic mean,
hm[x]; arithmetic mean, am[x]' geometric mean, gm[x]; and median,
x(50).  Since x is log-normally distributed, a variable y = In x
is normally distributed with mean and median p. and  standard
deviation a, as noted on p. E-4 of the Technical Support Document
for Water Quality Based Toxics Control.  Furthermore, if x is
log-normally distributed, 1/x must also be log-normally
distributed; a variable z = In(1/x) = -In x is then normally
distributed with mean -fj. and standard deviation o.

     The geometric mean of x corresponds to the arithmetic mean
of In x, as follows:

          gm[x] = exp n                      (4)

And, as noted on p. E-4 of the TSD:

          am[x] = exp(/i + a2/2)               (5)

It follows that:

          am[l/x] = exp(-M + cr2/2)            (6)

Since by definition, am[l/x]={S (l/x)}/N, and hm[x] = N/S (1/x),
where N is the number of observations in any distribution, it
follows that:

          am[l/x] = l/hm[x]                  (7)
                                F3

-------
Therefore, combining Equation 6 and 7:

          hm[x] = l/exp(-M + a2/2)

                = exp(M - a2/2)                     (8)

It might be noted that Equations 5 and 8 indicate  that on  a  log
scale, the geometric mean of a log-normal distribution lies
midway between the harmonic mean and arithmetic mean.

     If the duration table is obtained for a USGS  gage, then
statistics accompanying the retrieval will provide the mean  and
standard deviation of the base 10 logarithms.  Then one may
calculate    equal to 2.3026 times the mean of the base 10 logs,
and   equal to 2.3026 times the standard deviation of the  base 10
logs, and use Equation 8 to calculate the harmonic mean.

     Alternatively, Equation 8 can be rewritten as follows:

          hm[x] = exp M/exp (o"2/2)  = (exp /i)2/{exp  /x  • exp(cr2/2)}
               =  (exp M)2  /  exp(M + a2/2)         (9)
Equations 4 and 5 can now be substituted into Equation 9:

          hm[x] =  gm2[x]/am[x]                     (10)

Information on the geometric and arithmetic means  is available  in
the statistics accompanying the duration table.  Alternatively,
since for a log-normal distribution, the geometric mean, gm[x] ,
equals the median, x(50) , also provided in the duration table
statistics:

          hm[x]  =  {x(50) }2/am[x]                  (11)

Finally, since as noted on p E-4 of the TSD, the coefficient of
variation is given by CV[x] = [exp (a2)  - 1}   , it also follows
that:

          hm[x] = gm[x] / {CV^x]  +  1}"'             (12)

While Equations 8, 10, 11, and 12 are equivalent for perfectly
log-normal distributions with very large N, their results can
ordinarily be expected to differ slightly from each other for
actual flow distributions.

     The usual pattern of flow distributions is that the lowest
flows tend to be higher than expected for a log-normal
distribution.  In this case, Equations 8, 10, 11, and 12 will
tend to underestimate the harmonic mean flow.
                                F4

-------
     Again, whether using Equation 1, 3, 8, 10, 11, or 12, the
harmonic mean must be determined for the combined upstream and
effluent flow.
                               F5

-------
SUn == 	
• • n M n m
m
S M ••» — — ••
m ^•9^Vl"^V^^vi

•••••^
*
».»>*MMMft>
JV IN * -4 S M « M • £ M ^
S3 "* «>«£*«<« — MMMM
UK 9 £83**S23S» — m8
M • —
ii £ 2 8 8 8 * 45 it 5 2 8 8
*? *< •* a * 2 " * 2 * S * *
• •

M g M : --------

•>*M«*»4»:±J2 = 3S!=M:a




iiiiiiiiiii

— w
If
N:sxi:StHi

• ; 5 3 5 3 j; * • * !•

Ui U

<
^
s . :

o

S* * * M M
o — — **«^mSil
i « w* -*5S ««S f •



SM M M — — M » M — — ^ M — •••> —
M • » « — »« M • M ^ •• M«4*»> 4

n M » M If — * M M » M « M M M •* — «*M M« — «W •> • * <4 » •> ^ ^ 5 * 1 4


O

g 	 _8a
« • • » » ••«<*(» M M M Ml « » * — * * •
X — — — ^ M M M ——••«. —

O



J= M»«*MW «««»« W-MM » «M
a
— — « *M«MM — M— W W —
l» — M**»MM» M M »
— li
— * » M « M — — M — •
X » - -- >
X
S
• — » UI
I Z X n
1 -I
a S
< S
m -
> *
S S
» a

- 2 "
w -«
" 3 S
^ S -
* S

»
1 »
M«
; i
* V
•
' S
• am
• » -
M »
f s i
• r- Z
S 3
M 9
8
! •< ?
S S
» e
; 5 °
m
^
I
m
*
O





S
«
w
1
•
•
t
                                                       X
                                                       a.
                                                       O
                                                       VI
                                                       1
                                                       a>
                                                       o
                                                       3
F-6

-------
<
•
a
M
e
«


5 *
*
m
3 a
S
-4
1 e
M
*s
R *
3 £
J -
§
. :
E
M
m -g
I
3 8
8
m
" S
m 5
m *
m
$
s
«
«
-c
VI
*
» V
• «


















•
•
1 C


















DISCHARGE IN CFS
tLOC SCAtEl















•


•
1 I






•
*
*
•
»
*
•
•
*
m
i
i



•
- <
i i
•
•
•
•
8
•
•
>
•
•
1












DURATION PLUI Ut UA1
3TAMOM ID: 0<«l«9000
inAAA
c
>
*
«(
;
n ^ .
^ .
M M '
I 5;
r- » »
m « X
<
^ m n
2 »?
5 s;
8*
•^ n
O T
Ul -
•4 n
m 3
a a
" «
x |c
It • •
? i
^
»rt •
^ ^
m .
2
-i e
w <

o
a
»
z
»
m
»
*
m
it
9
0
e
£
3
F-7

-------
3
                                                                    "I
                                                                    » S
                                                                    •• »i
                                                                    < m
                                                                    »  n

                                                                    fi  s
                                                       » « HI ••
                                                                        >  s
                                                                i!

                                                                «l
                                                                »»
                                                                38
                                        F-J

-------
n  r>  M  3
22-*

S  S  I  *

551*

i  i  s  a

«  «  |  :

|  i  s  £

   55?

   I  i  !
      M
it o 5 • wow***''1*1**
II II II II II It II II II II II II II II II II II II II


                             Ij
               II II II II II II II II II II II II II II II II H II
      M
      m

      X
                                           S
                                            s
                                            If
Z > »     o  a M e

juo     i  £ » B

ma-.     _  s — -
9  •*     —  » g —

  S|   55  f?39
 i  S   ^    m M
       fc a  -en


i»o   » »  z « <
C S Z   m     * m
m M 9     <  n —
«* - -   Z»  C * u»
  s ••   8 i-  9 « -<
M3»   -«e  •••»
<^ 3 g     r*  n • -t



si,

^i
••      MM
       n z

       p- m    •
       " »  wi  »
       * 3  m  fl
         O  (i  •
       Si» r8  a ^» n
       M »  Z I- H
      (/•<««  B M m

,*S  isS   52
Z O m  m e      ui
m     Z  ^   • «
  5—    -. — s   « -
  •«     o g   < n

N <    Z >     3 ^


3c    -5i   5i

ss    2!:   s.
•     m M M
  m    a n Q   •» .


      i:!   §

      >  0   fl
      -* w» f   m
      ** •* m   «

      25-   3
                                                 i
                                          e
                                          M
                                          M





                                          |




                                          S
                     F-9

-------
      APPENDIX G




SAMPLE PERMIT LANGUAGE

-------
                           Appendix  G

                      Sample Permit Language

A.   Bioconcentration Evaluation Reguirement for Characterizing
     Effluent Bioconcentration Potential

          The permittee shall perform effluent bioconcentration
     evaluation, according to methods in "Assessment, Reference
     Concentration Development and Control of Bioconcentratable
     Contaminants in Surface Waters"  (EPA 600/x-xx-xxx),  on the
     discharge(s) from outfall(s) 	, as described
     below:

     1.   The permittee shall initiate effluent bioconcentration
          evaluation within 90 days of the effective date of this
          Part to determine effluent  bioconcentration potential.
          Such testing will determine if an effluent sample
          contains compounds with a log P greater than 3.5
          ("bioconcentratable compounds")  in an amount sufficient
          to present a bioconcentration hazard in the receiving
          water, (i.e., in exceedance of relevant reference
          ambient concentration(s)).

     2.   The effluent bioconcentration evaluation will:

          a) consist of the procedures in Appendix B of
          "Assessment and Control of  Bioconcentratable
          Contaminants in Surface Waters" (EPA 600/x-xx-xxx).
          All tests will be conducted on 24-hour composite
          samples.   A minimum of 2 replicates will be used in the
          tests.

          b) be conducted every four  months for a period of one
          year following the effective date of this permit.  If
          no identifiable bioconcentratable compounds are found,
          no further analysis is necessary unless a change in
          process or discharge occurs.   Data will be reported
          according to Appendix B (EPA 600/x-xx-xxx), Section
          10.8  "Reporting of Data,"  and shall be submitted to
          the permitting authority in the appropriate monthly
          Discharge Monitoring Report.   Following review of these
          results,  the permitting authority may require the
          permittee to confirm the identity and quantity of any
          compounds for which the permitting authority determines
          this additional analysis is necessary.

     3.   Upon determination by the permitting authority that the
          potential exists for discharge of bioconcentratable
          pollutants in exceedance of reference ambient
          concentrations, the permit  may be reopened and modified

                               G-2

-------
          to incorporate necessary effluent limitations for the
          individual bioconcentratable pollutants.


                            FACT SHEET

A.   Effluent Limitations on Bioconcentratable Compounds

     1.   No bioconcentratable pollutant shall be discharged in
          an amount which would cause exceedance in the receiving
          water of its reference ambient concentration as
          determined by the procedures in Chapter 3 of
          "Assessment, Reference Concentration Development and
          Control of Bioconcentratable Contaminants in Surface
          Waters" (EPA 600/x-xx-xxx).

     2.   Where they are necessary, effluent limitations will be
          derived according to Chapter 6 of the "Assessment,
          Reference Concentration Development and Control of
          Bioconcentratable Contaminants in Surface Waters"
          (EPA 600/x-xx-xxx), for bioconcentratable compounds
          which have been identified in the effluent by the
          bioconcentration evaluation procedure of Appendix B of
          the above document.

     3.   Where final calculated limitations for individual
          pollutants fall below the current level of detection,
          the limit for the pollutant will be the one calculated;
          the analytical method to be used for monitoring will be
          specified in the permit; and the minimum level for that
          method (if one is available) will be specified in the
          permit as the threshold for compliance/non-compliance
          determinations.  Where a minimum level is not available
          for the specified analytical method, an alternative
          threshold for compliance/non-compliance determinations
          will be specified in the permit (e.g., the method
          detection limit).

B.   Effluent Limitations Compliance Monitoring Requirements

     Effluent analysis for the purpose of determining compliance
with limits on individual bioconcentratable compounds shall be
conducted by the permittee once per week and reported on the
Discharge Monitoring Report using EPA method 1625.  Levels of
detection and quantitation shall be reported.
                               G-3

-------
                    Sample Section 308 Letter

CERTIFIED MAIL NO.
RETURN RECEIPT REQUESTED

Ms.  Ann Powell
Plant Manager
Chemico Corporation
Anytown, USA
                                          RE:  NPDES No. XX000123
                                               Chemico Anytown
Dear Ms. Powell,

     The U.S. Environmental Protection Agency (EPA) is preparing
to reissue the NPDES permit for the Chemico Anytown facility.
Preliminary evaluations conducted by EPA and water quality
monitoring conducted by the (State Water Control Authority)
indicate that the discharges from the Chemico Anytown facility
may be violating State water quality standards.   In order for EPA
to fulfill its responsibilities under the Clean Water act, 33
U.S.C. 1251, et seq..  additional information regarding the nature
of the discharges from the Chemico facility is required.

     Therefore, you are hereby required (1) to perform the
sampling and analysis programs described below,  (2) to maintain
for possible use by EPA all records regarding plant operations
during the sampling and analysis programs within the time limits
specified.  These requirements are imposed pursuant to the
authority provided in Section 308 of the Clean Water act, 33
U.S.C. Section 1318.  Failure to comply with this request may
result in enforcement proceedings under Section 309 of the Clean
Water act, 33 U.S.C. Section 1319, which could result in the
judicial imposition of civil or criminal penalties.

Effluent Bioconcentration Evaluation -

     In order to identify potentially bioconcentratable
pollutants discharged from Outfall 001, Chemico shall analyze
effluent discharges as described below.  Wastewater samples for
analysis shall be composited over 24 hours once per quarter for a
period of one year commencing within 30 days of receipt of this
letter.

      The procedures to be used shall conform to the methods in
Appendix B of  "Assessment, Reference Concentration Development
and Control of Bioconcentratable Contaminants in Surface Waters"
(EPA 600/x-xx-xxx).  All tests will be conducted on 24-hour
composite samples.  A minimum of 2 replicates will be used in the
tests.  A reasonable effort shall be made to identify and confirm

                               G-4

-------
those chemicals designated as chemicals of highest concern and
those present in concentrations likely to cause excursions above
water quality standards.  Following review of the results of this
evaluation, the permitting authority may direct the permittee to
conduct additional confirmation on chemicals reported as present
in the effluent.

     Data will be reported according to Appendix B
(EPA 600/x-xx-xxx), Section 10.8  "Reporting of Data," and shall
be provided with a description of the analyses to the permitting
authority within 45 days of sample collection.

     If you have any questions please contact 	.   Thank
you for your cooperation.

Sincerely,


Director,
Water Division
                               G-5

-------
                 Sample Permit Language for Fish
                 Tissue and Sediment Evaluations

Part III Special Conditions

A.  Fish Tissue and Sediment Evaluations

1.  The permittee shall conduct the Tissue residue evaluation and
Sediment Evaluations as described below.  The purpose of these
evaluations is to determine compliance with applicable State
water quality standards, and to determine if pollutants
accumulate in fish tissue or sediments to levels which would
cause harmful effects to aquatic life and/or unacceptable risks
to human health.  The primary objectives of these analyses are
to:  1) evaluate the discharge related bioaccumulation of organic
chemical pollutants to the aquatic life in the receiving water,
2) evaluate the discharge related accumulation of organic
chemical contaminants in the sediments of the receiving water, 3)
provide information for evaluating the re-issuance of this
permit.  Following review of the results from the annual
evaluations submitted by the permittee, the permitting authority
may adjust the frequency and extent of the data collection and
analysis if such adjustments in the sampling design are necessary
to meet the objectives of the tissue residue and sediment
evaluations.

2.  Tissue Residue Evaluation - The objective of this evaluation
is to identify and confirm the presence and concentrations of
bioaccumulative pollutants in aquatic organisms in the receiving
water.  Monitoring will be required annually during the month of
	.  Sampling stations are to be located at 	, as
illustrated in Figure 1.  Sampling methods, tissue residue
evaluation procedures and data reporting should follow those
specified in  Appendix A of  "Assessment, Reference Concentration
Development and Control of Bioconcentratable Contaminants in
Surface Waters" (EPA 600/x-xx-xxx).  Species to be collected
shall include ten individuals of the benthic organism species
	, and a minimum of ten individuals from at least two of
the following fish species, 	.  These samples
shall be collected from at least two sampling stations (including
a reference and effect station).  One station must be in the
proximity of the 	 facility mixing zone.

     The results of this evaluation shall be submitted no later
than 45 days after the sample collection date.  Upon review of
the results of this evaluation the permitting authority may
require additional chemical specific effluent monitoring for any
of the bioaccumulative contaminants identified in the results of
the fish tissue evaluation.
                               G-6

-------
3.  Sediment Evaluation - The objective of this evaluation is to
identify and confirm the presence and concentrations of
bioaccumulative pollutants in the receiving water sediments
associated with the discharge from outfall 001.  Monitoring will
be required annually during the month of 	.  Sediment
sampling stations are to be located at 	, with at least one
station to be located in the 	 facility mixing zone, as
illustrated in Figure 2.  Sampling methods, sediment evaluation
procedures and data reporting should follow those specified in
Appendix C of  "Assessment, Reference Concentration Development
and Control of Bioconcentratable Contaminants in Surface Waters"
(EPA 600/x-xx-xxx).

     The results of this sediment evaluation shall be submitted
no later than 45 days after the sample collection date.  Upon
review of the results of this evaluation the permitting authority
may require additional chemical specific effluent monitoring for
any of the bioaccumulative organic chemical contaminants
identified in the results of the sediment evaluation.
                               G-7

-------
             APPENDIX H






OVERVIEW OF SELECTED AVAILABLE TOOLS

-------
                           Appendix  H

               Overview of  Selected Available  Tools



A.   Chemical-Specific Toxicity Data Bases

     1.   QSAR System  [48]

          Reference:   "QSAR System User Manual", a joint project
          of the Institute for Biological and Chemical Process
          Analysis, Montana State University,  Bozeman, Montana
          and the United States Environmental Protection
          Agency/Environmental Research Laboratory-Duluth
          (U.S.E.P.A./E.R.L.-D.), Duluth, Minnesota (October
          1986).

          To obtain:
                     Christine L. Russom
                     U.S.E.P.A./E.R.L.-D.
                     6201 Congdon Blvd.
                     Duluth,  Minnesota  55804
                     (218)-720-5709
                     (FTS)-780-5709


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

          Cost:  No charges are applied for Federal, State or
          Local governments.

          Release Date:  January 1991
                               H-2

-------
Integrated Risk Information System  (IRIS) [50]

Reference:  "Integrated Risk Information System",
Volume 1.  "Supportive Documentation", Volume 11.
Chemical Files (EPA-600/8-86-032a and b), Office of
Health and Environmental Assessment, Office of Research
and Development, U.S. EPA, Washington, D.C., March,
1987.

To obtain:  IRIS is available on EPA electronic
mailbox.  IRIS User's Guide obtained through IRIS User
Support at (FTS) 684-7254.  Those outside EPA can
obtain an IRIS account by contacting Mike McLaughlin,
DIALCOM, Inc., Federal Systems Division, 600 Maryland
Avenue S.W.,  Washington, D.C.  IRIS is also available
through the Public Health Network (Paul Johnson, (202)
898-5600) of the Public Health Foundation.  IRIS will
be made available on the NIH National Library of
Medicine's TOXNET system ((301) 496-6531) sometime in
late summer or fall of 1988.

Use:  IRIS is a computer-housed, electronically
communicated catalogue of Agency risk assessment and
risk management information of chemical substances.
The IRIS system is designed especially for federal,
state, and local environmental health agencies as a
source of the latest information about Agency health
assessments and regulatory decisions for specific
chemicals.  The risk assessment information contained
in IRIS, except as specifically noted, has been
reviewed and agreed upon by intra-agency review groups,
representing an Agency consensus.  An intra-agency work
group has been responsible for the development of IRIS
[22].  There are currently 260 chemicals in IRIS.  New
chemicals are added regularly and existing chemicals
are revised as warranted by new scientific findings.

Cost:  No charges are applied for EPA users and anyone
having EPA-paid-for electronic mail accounts.  All
other users are subject to charges applied by the
particular system they use to access IRIS.

Caution:  IRIS data is subject to update on a frequent
basis.
                     H-3

-------
3.   Health Effects Assessment Summary Tables (HEAST)

     Reference:  "Health Effects Assessment Summary Tables",
     United States Environmental Protection Agency/Office of
     Solid Waste (U.S.E.P. A. /O.S.W.) , Washington D.C.,
     (released quarterly)

     To obtain:  The U.S.E.P.A./O.S.W. requests that their
     users (i.e. O.S.W. staff, contractors, and State solid
     waste programs) call Susan Griffin of the Office of
     Solid Waste at (202)-382-6392 to obtain copies.
     Regional O.S.W. staff are reminded that copies of HEAST
     are sent to all Regional libraries.  All others must
     purchase the document from:

          National Technical Information Service (N.T.I.S.)
          5285 Port Royal Road
          Springfield, Virginia   22161
          (703)-487-4650

     Use:  The Health Effects Assessment Summary Tables have
     been developed to provide information on chemicals
     commonly found at Superfund and RCRA sites.  The
     reports are intended as pointer systems to identify
     current literature or changes in assessment criteria
     for many chemicals of interest to the Superfund
     program.  The information in HEAST should be used as a
     secondary source to the information contained in the
     IRIS database.
     Cost:  There is a charge to receive this from N.T.I.S.

4.   Risk*Assistant Prototype

     Reference:  "Risk*Assistant Prototype, An Overview of
     the Software System", The Hampshire Research Institute,
     Alexandria, Virginia, May 1990.

     To Obtain:
               RISK*ASSISTANT
               Hampshire Research Institute
               1600 Cameron Street, Suite 100
               Alexandria, Virginia  22314
               (703)-683-6695

     Use:   RISK*ASSISTANT is a microcomputer-based software
     system that provides an array of analytical tools,
     databases, and information-handling capabilities for
     individuals who must assess the health risks posed by
     chemicals.  Requiring only estimates of the nature and

                          H-4

-------
          concentrations of hazardous chemicals present in a
          sample the system can provide information on
          carcinogenic and other toxic hazards of chemicals.  The
          RISK*ASSISTANT system is designed especially for use by
          federal, state, and local agencies as a source of the
          information for evaluating the health risks posed by
          specific chemicals.  The system can be used to retrieve
          information on carcinogenic and other toxic hazards of
          chemicals and to help direct additional data collection
          or set priorities for subsequent regulatory action.

          Cost:  The Risk Assistant Prototype is available
          without charge at this time.


B.   Chemical-Specific Bioconcentration Tests

     1.   American Society of Testing and Materials [27]

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

          To obtain:  ASTM, Customer Service, 1916 Race St.,
          Philadelphia,  PA  19103, (215) 299-5400, (FTS) 299-
          5400.  It is not necessary to order the entire Annual
          Book of Standards.  A copy of Standard Practice E 1022-
          84 costs $8.00.

          Use:  Where protection against bioaccumulating
          compounds is warranted and individual, potentially
          bioaccumulative compounds in the effluent can be
          identified, the permit writer should consider requiring
          the permittee to perform bioconcentration tests on
          these compounds using the methods described in Standard
          Practice E 1022-84.

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

          Cautions:  Some techniques described in the method were
          developed for tests on non-ionizable organic chemicals
          and may not apply to ionizable or inorganic compounds.
          The bioaccumulation potential of many non-ionizable
          organic compounds can be predicted much more cheaply
          using physical-chemical properties such as the log P.
          For example, QSAR (see A.I.) uses log P to calculate
          BCFs.
                               H-5

-------
     2.   log P values [49]

          Reference:  Hansch, C. and A.J. Leo.  "Substituent
          Constants for Correlation Analysis in Chemistry and
          Biology."  John Wiley and Sons:  New York, 1979.
C.   Models
     1.   Food and Gill Exchange of Toxic Substances (FGETS,
          GETS)

          Reference:  "FGETS (Food and Gill Exchange of Toxic
          Substances:  A Simulation Model for Predicting
          Bioaccumulation of Nonpolar Organic Pollutants by Fish"
          by M.C. Barber, L.A.  Suarez, and R.R. Lassiter. (EPA-
          600/3-87-038) [25].  Also, "GETS, A Simulation Model
          for Dynamic Bioaccumulation of Nonpolar Organics by
          Gill Exchange:  A User's Guide," by L.A. Suarez, M.C.
          Barber, and R.R. Lassiter.  1987 (EPA-600/S3-86-057)
          [26].

          To obtain:  Environmental Research Laboratory, Office
          of Research and Development, U.S. EPA, Athens, GA
          30613.

          Use:  This model is for the bioaccumulation of
          nonpolar, non-metabolized organic chemicals by fish.
          The FGETS model simulates thermodynamically-driven
          chemical exchange by fish assuming either aqueous
          exposure only or joint aqueous and food chain exposure.

          Cost:  Single copies free of charge.

          Caution:  The FGETS model is still in development.
          Initial analysis indicates that FGETS can simulate
          quite well observed patterns of bioaccumulation and
          depuration of a single, nonpolar, slowly or non-
          metabolized organic pollutant by individual fish.
          Extensions of the model to incorporate simulation of
          mixtures and to account for other variables are
          planned.
D.   Supporting Information and Guidance

     1.   Health Risk Assessment Procedure for Consumption of
          Contaminated Fish and Shellfish [44]
                               H-6

-------
     Reference:  "Assessing Human Health Risks from
     Chemically Contaminated Fish and Shellfish:  A Guidance
     Manual", United States Environmental Protection
     Agency/Office of Marine and Estuarine Protection
     (U.S.E.P.A./O.M.E.P.)/ Washington D.C., September 1989.

     To obtain:

               Paula Monroe
               U.S.E.P.A./O.M.E.P.
               WH-556-F
               401 M Street S.W.
               Washington D.C.  20460
               (FTS/202)-475-6182

     Use: This guidance manual provides step-by-step
     assistance for assessing health risks from exposure
     through consumption of chemically contaminated aquatic
     organisms.  The guidance is applicable to freshwater,
     brackish water, and saltwater fish and shellfish.
     Guidance is provided on mathematical models used to
     estimate chemical exposure and risk.  Information on
     sampling design is provided.  Sources of information on
     toxic chemicals and model variables are noted.
     Additionally, suggestions for presentation of risk
     assessment results are provided and uncertainties are
     summarized.   The guidance provided in this manual is
     directed primarily at risk assessment related to
     recreational fisheries.

     Cost:  No charges are applied for Federal, State or
     Local governments.


2.    State Procedures for Calculating RACs in the Absence of
     IRIS Values

     Reference:

          1)    Department of Natural Resources, "Guidelines
               for Rule 57(a)  (5 and 6), Michigan Department
               of Natural Resources, Environmental
               Protection Bureau, 1987 [51].

          2)    Department of Natural Resources, "Proposed
               Chapter NR10S.08", Wisconsin Department of
               Natural Resources, Bureau of Water Resources
               Management, 1987 [52].
                          H-7

-------
Use:  Michigan's procedures for deriving a dose factor
depend upon the type and quality of the toxicity data
base used [51].  Wisconsin uses toxicity data in
conjunction with feeding habits for mammalian and avian
test species to estimate RTCs for fish consumption by
wildlife.  Where data for multiple species are lacking,
Wisconsin applies an uncertainty factor of 0.1 to
account for differences in species sensitivity [52].
These procedures have not been formally reviewed or
approved by the EPA and the inclusion of these
procedures, pending such review, is intended to be
advisory only.
                     H-8

-------
                       APPENDIX I



                FIELD VALIDATION STUDIES






The Field Validation Report Will be Added When Completed

-------
       APPENDIX J




EXAMPLE DATA AND REPORTS

-------
                            Appendix  J

Illustrative Data and Reports  for  Effluent,  Tissue,  and Sediment
Analytical Procedures.

1.0  General Considerations

     In  this  appendix,   example  data  and  discussion  on  data
evaluation are  provided for the tissue,  effluent,  and sediment
analytical procedures.   The reports generated by the different
analytical procedures are similar.   The formats for reporting the
data  and  some  of  the  reported  data  in  this  document  are
illustrative.  Reporting formats for laboratories and consultants
performing these procedures may differ.

2.0  Tissue Alternative

     In evaluating any report,  the  QA/QC report should be examined
first since this information will provide a good indication of the
quality of the analytical data.  The QC  requirement  for recovery
of  the surrogates  is 25%  <  %  recovery  <  120%  and all  three
surrogates, d10-biphenyl, 13C6-l,2,4,5-tetrachlorobenzene, and 13C6-
hexachlorobenzene, should have recoveries within  this range.   If
recovery data is out of this range, the analytical data should not
be used.   Provided  acceptable  recoveries are obtained, the GC/MS
chromatograms and the QA/QC data for GPC resolution,  silica gel
performance,  etc.  should be examined.  This examination should take
5 to 15 minutes and will provide a good indication of the overall
quality of the analysis.

     In Table J-l, an illustrative report  for  the tissue analytical
procedure is provided.  After examining the recovery data for the
surrogate  chemicals,  the  GC/MS chromatograms for  the procedural
blank and the sample extract should  be examined.   The procedural
blank  should contain few peaks as  illustrated  in  the  report.
Procedural blanks should always contain fewer peaks than the sample
extract.  Peaks in the GC/MS chromatograms should be narrow, e.g.,
10 to 20 seconds base widths, and when large fat peaks exist, e.g.,
base widths exceeding 1 minute,  overloading  of the capillary column
on  the GC/MS  has  occurred.    Efforts  should  be  directed  at
determining  why  these large peaks exist  in the  sample  extract.
Sometimes these large peaks exist because  the  cleanup procedure for
the  tissue extract  was  performed poorly,   the  capacity  of  the
cleanup procedure was exceeded,  and/or severe  blank problems exist.
In these situations, additional  cleanup  or  extraction of another
portion of the tissue composite should be requested.

     The evaluation of the QC data for GPC resolution, silica gel
performance,  GC resolution,  MS  sensitivity,  MS calibration, DFTPP
performance,   and  precision  are  all   straight  forward.     The
evaluation criteria for these QC procedures are:

a) GPC Resolution:  Baseline resolution between the three
                               J-2

-------
     performance   chemicals,    Dacthal,    pyrene,    and   di-2-
     ethylhexylphthalate,  must  exist.    The data  in  Table  J-l
     presents excellent resolution.

b) Silica Gel Performance:  The eluate from the column using the
     silica performance solution must not contain more than 10% of
     the cholesterol  while at least 90%  of  the  dieldrin must be
     recovered.   This is  performed for  each  lot of silica  gel
     and/or every two months,  which ever comes first.  This data
     can be presented using control charts or in tabular form.

c) GC Resolution:  One of two solutions can be used to evaluate GC
resolution.   For  the solution  containing  /3-BHC,   -BHC,  endrin
ketone and d12-chrysene,  baseline resolution between the j8-BHC/ -
BHC and  endrin ketone/d12-chrysene  pairs should exist.   For  the
solution containing  anthracene,  phenanthrene,  benz[a]anthracene,
and  chrysene,   anthracene  and  phenanthrene  should  be  baseline
resolved and benz[a]anthracene and chrysene should be  separated by
a valley whose height is less than 25% of the average peak height
of these two compounds.   The data in Table  J-l  is  for the first
solution  above.   Note,   when using  the  first solution,  fresh
solution is required  since endrin ketone degrades.

d) MS Sensitivity:  For the 0.5 ppm calibration solution, the
     GC/MS peaks for  the three surrogate chemicals  should have a
     signal to noise  ratio of 3 or more.

e) MS Calibration:  For d12-chrysene, the  ratio of the abundance
     of the 241 mass/charge relative to that  of 240 should be >15%
     and <25%.   In the example data, this ratio is 20%.

f) DFTPP Performance:  Comparison of the mass spectrum for DFTPP
     to  the performance  criteria  is required.    A DFTPP  mass
     spectrum is reported in Table J-l.

g) Precision:  A duplicate analysis should be performed with
     every 10 sample  analyses.  This data can be presented by using
     control charts or tabular reports.   Precision data should not
     exceed 150%.

     If any of  the QC/QA procedures are  not met,  the analytical
data  is  of  questionable  quality and  use  of  the  data is  not
recommended.   Depending  upon which  factors are  out  of  range,
reanalysis of the sample extract (c,d,e,f) or  extraction of another
portion of the  tissue composite  (a,b,g)  will be  required.   Some
judgement will  be required in this evaluation.

     If the QA/QC Report is acceptable,  the data in Reports I, 2,
and 3 can be used.

     For Tissue Report 1,   GC/MS components tentatively identified
using the CHC  mass  spectral library search  are reported.   Those
GC/MS  components with fits of  70% and  greater are considered


                               J-3

-------
tentatively  identified.   For  each tentatively  identified GC/MS
component, the following information is require:

     1)   GC retention time of the component, minutes
     2)   Amount of the component, ug/kg
     3)   A list of the best mass  spectral library identifications
          (up to a total of ten identifications)
     4)   For each tentative identification:
               Library searching fit, %
               CAS number

     This report often contains none or few tentatively identified
components.

     For Tissue Report 2, GC/MS components tentatively identified
using the EPA/NIH/NBS  mass  spectral  library search are reported.
Those GC/MS components with fits of 70% and greater are considered
tentatively identified. This report contains only those components
with concentrations of 5 ug/kg and greater.  For each tentatively
identified GC/MS component, the following information is require:

     1)   GC retention time of the component, minutes
     2)   Amount of the component, ug/kg
     3)   A list of the best mass  spectral library identifications
          (up to a total of ten identifications)
     4)   For each tentative identification:
               Library searching fit, %
               CAS number

     For  Tissue  Report  3,  GC/MS  components  not  tentatively
identified,  i.e.,  fits  <70%, using the  EPA/NIH/NBS mass spectral
library  search  are reported.   This  report contains  only those
components with concentrations  of  5 ug/kg  and  greater.   For each
GC/MS  component   not  tentatively  identified,   the  following
information is require:

     1)   GC retention time of the component, minutes
     2)   Amount of the component, ug/kg
     3)   If the fit parameter  is  >25%, the two best mass spectral
          library   identifications   and    for   each   tentative
          identification,  their  library  searching  fit and  CAS
          number.
     4)   If the  fit parameter is <25%,  compound is  labeled as
          unknown.

3.0  Effluent Alternative

     In evaluating any report,  the QA/QC report should be examined
first since this information will  provide a good  indication of the
quality of the analytical data.   The  QC requirement for recovery
of  the surrogates  is  25%  <  % recovery  < 120%  and  all  three
surrogates,  d10-biphenyl, 13C6-1,2,4,5-tetrachlorobenzene, and   C6-
hexachlorobenzene, should have  recoveries within this range.   In
addition, HPLC fraction 1 should contain d10-biphenyl and should not

                               J-4

-------
contain  13C6-1,2,4,5-tetrachlorobenzene  or 13C6-hexachlorobenzene.
HPLC fraction  2 should contain 13C6-1,2,4, 5-tetrachlorobenzene and
should  not contain  d10-biphenyl  or   C6-hexachlorobenzene.   HPLC
fraction  3 should  contain 13CR-hexachlorobenzene and  should not
                              6
contain  C6-l,2,4,5-tetrachlorobenzene or d10-biphenyl.

     If recovery data is out of range and/or the  surrogates are in
the  wrong  fraction,  the  analytical  data   should  not  be  used.
Provided acceptable recoveries and surrogate location  in the HPLC
fractions are obtained, the GC/MS chromatograms and  the QA/QC data
for HPLC  resolution, GC resolution, etc. should be examined.  This
examination should  take  5 to 15 minutes  and will provide a good
indication of the overall quality of the analysis.

     In  Table  J-2,  an  illustrative  report  for  the  effluent
analytical procedure  is  provided.   After  examining the recovery
data for the surrogate chemicals, the GC/MS  chromatograms for the
procedural blank and the sample extract should  be examined.  There
will be six GC/MS chromatograms, three for the procedural blank and
three for the  effluent  sample as a result of  the HPLC fractionation
procedure.  The GC/MS chromatograms for the  HPLC  fractions for the
procedural blank should  contain few peaks  as  illustrated in the
report.  Procedural blanks should always contain  fewer peaks than
the sample  extract.   Peaks in the  GC/MS chromatograms should be
narrow, e.g.,  10  to 20  seconds  base widths,  and when large fat
peaks exist, e.g.,  base widths exceeding 1 minute,  overloading of
the capillary column on the GC/MS has occurred.   Efforts should be
directed at determining why these large peaks exist in the sample
extract.   Sometimes these large peaks  exist because the cleanup
procedure  for  the  effluent  extract was  performed  poorly,  the
capacity of the cleanup procedure was exceeded,  and/or  severe blank
problems  exist.    In  these  situations, additional   cleanup  or
extraction  of  another portion of  the effluent  sample should be
requested.

     The  evaluation  of   the  QC  data  for   HPLC resolution,  GC
resolution, MS sensitivity,  MS calibration,  DFTPP performance, and
precision are all straight  forward.   The evaluation criteria for
these QC procedures are:

a) HPLC Resolution:   Baseline resolution between the three
     performance  chemicals,  biphenyl,  l,3-diethylbenzene,  and
     bibenzyl, must  exist.  In Table J-2, chromatographic data with
     excellent resolution is shown.

b) GC Resolution:  One of two  solutions can  be  used  to evaluate GC
resolution.   For the  solution  containing  /3-BHC,   -BHC,  endrin
ketone and d1?-chrysene,  baseline resolution between the /J-BHC/ -
BHC and  endrin ketone/d12-chrysene  pairs should   exist.   For the
solution containing  anthracene,  phenanthrene,  benz[a]anthracene,
and  chrysene,   anthracene and  phenanthrene should  be  baseline
resolved and benz[a]anthracene and chrysene  should be separated by
a valley whose height is less than 25% of the average peak height


                               J-5

-------
of these two compounds.  Note, when using the first solution, fresh
solution is required since endrin ketone degrades.

c) MS Sensitivity:  For the 0.5 ppm calibration solution, the
     GC/MS peaks  for the  three  surrogate chemicals  should have a
     signal to noise ratio of 3 or more.

d) MS Calibration:  For d12-chrysene, the ratio of the abundance
     of the 241 mass/charge relative to  that of 240 should be >15%
     and <25%.

e) DFTPP Performance:  Comparison of the mass spectrum for DFTPP
     to  the  performance  criteria  is   required.    A DFTPP  mass
     spectrum is presented in Table J-l.

f) Precision:  A duplicate analysis should be performed with
     every 10 sample analyses.  This data can be presented by using
     control charts or tabular reports.   Precision data should not
     exceed 150%.

     If any  of  the QC/QA procedures are not  met,  the analytical
data is of  questionable quality and its use  is  not recommended.
Depending upon which factors  are  out of range,  reanalysis of the
HPLC fractions  (b,c,d,e)  or extraction  of  another portion of the
effluent sample  (a,f)  will be required.   Some  judgement will be
required in this evaluation.

     If the QA/QC Report is acceptable,  the data in Reports 1, 2,
3, and 4 can be used.

     For Effluent Report 1, GC/MS components tentatively identified
using the CHC mass spectral library search are  reported.   Those
GC/MS  components with  fits  of  70% and  greater are  considered
tentatively  identified.    For  each tentatively  identified  GC/MS
component, the following information is require:

     1)   GC retention time of the component,  minutes
     2)   Amount of the component, ng/1
     3)   HPLC fraction number of the component
     4)   A list of the best mass spectral library identifications
          (up to a total of ten identifications)
     5)   For each tentative identification:
               Library searching fit, %
               CAS number

     This report often contains  none or  few tentatively identified
components.

     For Effluent Report 2, GC/MS components tentatively identified
using the EPA/NIH/NBS  mass spectral library  search  are reported.
Those GC/MS components with fits of 70%  and greater are considered
tentatively identified. This report contains only those components
with effluent concentrations of  100 ng/1 and  greater  and  with
predicted tissue concentrations of 1 ug/Xg and greater.   For each

                               J-6

-------
tentatively identified GC/MS component, the following information
is require:

     1)   GC retention time of the component, minutes
     2)   Amount of the component, ng/1
     3)   HPLC fraction number of the component
     4)   A list of the best mass  spectral library identifications
          (up to a total of ten identifications)
     5)   For each tentative identification:
               Library searching fit, %
               CAS number

     For  Effluent  Report   3,  GC/MS  components not  tentatively
identified, i.e.,  fits  <70%,  using  the EPA/NIH/NBS mass spectral
library  search  are reported.   This  report contains  only those
components with concentrations of 100 ng/1 and greater.  For each
GC/MS  component   not  tentatively   identified,   the  following
information is require:

     1)   GC retention time of the component, minutes
     2)   Amount of the component, ng/1
     3)   HPLC fraction number of the component
     4)   If the fit parameter is  >25%,  the two best mass spectral
          library   identifications   and  for   each   tentative
          identification, their library searching fit, %, and CAS
          number.
     5)   If the  fit parameter is  <25%,  compound is  labeled as
          unknown.

     For Effluent Report 4, GC/MS components tentatively identified
using the EPA/NIH/NBS mass  spectral  library  search are reported.
Those GC/MS components with fits of 70%  and greater are considered
tentatively identified.  This report contains only those components
with effluent  concentrations of  100 ng/1  and  greater  and with
predicted tissue concentrations of  less than 1  ug/Jtg.   For each
tentatively identified GC/MS component, the following information
is require:

     1)   GC retention time of the component, minutes
     2)   Amount of the component, ng/1
     3)   HPLC fraction number of the component
     4)   A list of the  best mass spectral library identifications
          (up to a total of ten identifications)
     5)   For each tentative identification:
               Library searching fit, %
               CAS number

4.0  Sediment Analytical Procedure

     In evaluating any report, the QA/QC report should be examined
first since this information will  provide a good  indication of the
quality of the analytical data.   The QC requirement  for recovery
of  the surrogates  is  25%  < %  recovery  <  120%  and all  three
surrogates, d10-biphenyl, 13C6-l,2,4,5-tetrachlorobenzene, and 13C6-

                               J-7

-------
hexachlorobenzene, should have  recoveries  within this range.  In
addition, HPLC fraction 1 should contain d10-biphenyl and should not
contain  13C6-l, 2,4,5-tetrachlorobenzene  or 13C6-hexachlorobenzene.
HPLC fraction  2 should contain 13C6-1,2 ,4, 5-tetrachlorobenzene and
should not  contain d10-biphenyl or   C6-hexachlorobenzene.   HPLC
fraction  3  should contain  13C6-hexachlorobenzene and  should not
contain 13C6-1,2,4,5-tetrachlorobenzene or d10-biphenyl.

     If recovery data is  out of  range and/or the  surrogates are in
the  wrong  fraction,  the  analytical data should  not  be  used.
Provided acceptable recoveries and surrogate location in the HPLC
fractions are obtained, the GC/MS  chromatograms and the QA/QC data
for HPLC resolution,  GC resolution,  etc.  should be examined.  This
examination should take  5  to 15 minutes and will  provide a good
indication of the overall quality of the analysis.

     In  Table  J-3,   an  illustrative  report  for  the  effluent
analytical procedure  is  provided.   After  examining the recovery
data for the surrogate chemicals,  the GC/MS chromatograms for the
procedural blank and the  sample  extract  should be examined.  There
will be six  GC/MS  chromatograms, three for the procedural  blank and
three for the effluent sample as a  result of the HPLC fractionation
procedure.  The GC/MS chromatograms  for  the HPLC  fractions for the
procedural blank  should  contain few peaks as  illustrated in the
report.  Procedural blanks should always contain fewer peaks than
the sample  extract.   Peaks in the  GC/MS chromatograms should be
narrow,  e.g.,  10  to  20  seconds base widths,  and when large fat
peaks exist, e.g., base widths exceeding 1 minute, overloading of
the capillary column on the GC/MS  has occurred.   Efforts  should be
directed at determining why these large peaks exist in the sample
extract.  Sometimes  these  large peaks  exist because  the cleanup
procedure for the  extract was performed poorly,  the capacity of the
cleanup procedure  was  exceeded, and/or severe blank problems exist.
In these situations,  additional cleanup or extraction of another
portion of the sediment sample should be requested.

     The  evaluation  of  the  QC  data  for HPLC resolution,  GC
resolution,  MS sensitivity,  MS calibration, DFTPP performance, and
precision are  all straight  forward.   The evaluation criteria for
these QC procedures are:

a) HPLC Resolution:   Baseline resolution between the three
     performance  chemicals,  biphenyl,  1,3-diethylbenzene,  and
     bibenzyl,  must  exist.    The  data  in  Table  J-2  presents
     excellent resolution.

b) GC Resolution:  One of two solutions  can be used to evaluate GC
resolution.    For the solution containing /9-BHC,  -BHC,  endrin
ketone and d1?-chrysene,  baseline  resolution  between the 0-BHC/ -
BHC and  endrin ketone/d12-chrysene  pairs should exist.   For the
solution containing  anthracene, phenanthrene,  benz[a]anthracene,
and  chrysene, anthracene  and  phenanthrene  should  be  baseline
resolved and benz[a]anthracene and chrysene should be separated by
a valley whose height is less than 25% of the average peak height

                               J-8

-------
of these two compounds.  Note, when using the first solution, fresh
solution is required since endrin ketone degrades.

c) MS Sensitivity:  For the 0.5 ppm calibration solution, the
     GC/MS peaks  for the  three surrogate chemicals should have a
     signal to noise ratio of  3 or more.

d) MS Calibration:  For d12-chrysene, the ratio of the  abundance
     of the 241 mass/charge relative to that of 240 should be >15%
     and <25%.

e) DFTPP Performance:  Comparison of the mass spectrum for DFTPP
     to  the  performance  criteria  is  required.    A  DFTPP  mass
     spectrum is presented in Table J-l.

f) Precision:  A duplicate analysis should be performed with
     every 10 sample analyses.  This data can be presented by using
     control charts or tabular reports.  Precision data should not
     exceed 150%.

     If any  of  the QC/QA procedures are not met, the analytical
data is of  questionable quality and its use is  not recommended.
Depending upon which factors  are  out  of range,  reanalysis of the
sample extract  (b,c,d,e)  or extraction  of  another portion of the
sediment sample  (a,f)  will be required.   Some  judgement  will be
required in this evaluation.

     If the QA/QC Report  is acceptable, the data  in Reports 1, 2,
and 3 can be used.

     For Sediment Report l, GC/MS components tentatively identified
using the CHC mass spectral library search  are  reported.   Those
GC/MS  components with  fits of  70% and  greater  are  considered
tentatively  identified.   For  each tentatively  identified  GC/MS
component,  the following  information is require:

     1)   GC retention time of the component, minutes
     2)   Amount of the component, ug/kg
     3)   HPLC fraction number for the component
     4)   A list of the best mass spectral library  identifications
          (up to a total of ten identifications)
     5)   For each tentative identification:
               Library searching fit,  %
               CAS number

     This report often  contains none or  few tentatively identified
components.

     For Sediment Report 2, GC/MS components tentatively identified
using the EPA/NIH/NBS mass  spectral  library search are reported.
Those GC/MS  components  with fits of 70%  and greater are considered
tentatively identified.  This report contains only  those components
with concentrations of 5 ug/kg and greater.  For each tentatively
identified GC/MS component, the following information is require:

                               J-9

-------
     1)   GC retention time of the component, minutes
     2)   Amount of the component, ug/kg
     3)   HPLC fraction number for the component
     4)   A list of the best mass  spectral library identifications
          (up to a total of ten identifications)
     5)   For each tentative identification:
               Library searching fit, %
               CAS number

     For  Sediment  Report  3,  GC/MS  components not  tentatively
identified,  i.e.,  fits  <70%, using  the EPA/NIH/NBS mass spectral
library search  are reported.   This  report contains  only those
components with concentrations of 5  ug/kg  and  greater.  For each
GC/MS  component   not  tentatively   identified,   the  following
information is require:

     1)   GC retention time of the component, minutes
     2)   Amount of the component, ug/kg
     3)   HPLC fraction number for the component
     4)   If the fit parameter  is  >25%, the two best mass spectral
          library   identifications    and   for   each   tentative
          identification, their library searching fit, %, and CAS
          number.
     5)   If the  fit  parameter is  <25%,  compound is  labeled as
          unknown.
                              J-10

-------
Table J-l.  Example Data for the Tissue Analysis  Procedures
Tissue QA/QC Report	

Percent lipid content of tissue sample

               Tissue mass    20.0 g
               Residue         0.200 g  (10% of total  extract)
               Percent lipid = 10.0%

Recoveries

     d10-biphenyl                             67%
     13C6-1,2,4,5-tetrachlorobenzene          50%
     13C6-hexachlorobenzene	46%	
                               J-ll

-------
GC/MS Chromatogram
Procedural Blank
 4NK

 :eflee*
GC/MS Chromatogram
Sample Extract
  t too ooo-j
  KOODOO
  uooaoo
  120O300
  100000*
  tooaoo
  too ooo
  40OOOO J
  200000
         10
                        J-12

-------
GPC Resolution Check
u
0)
Q
             ^
                             a
                             I
                  I v

                          -A-


                  —t	->
                  .J-	
                       ••3
                          --v
                 J-13

-------
Silica Gel Performance Check
Date of   Silica Gel  Analyst   Chemical  Recovery
Analysis  Lot #                              %
                                           Elution
                                           Solvent
                                           (MECL:HEX)
2/3/89


4/4/89


6/5/89


7/8/89
Baker
A27345

Baker
A27345

Baker
A27345

Baker
A27345
JJ
LB
JJ
CJ
Cholesterol     8
Dieldrin       95

Cholesterol     5
Dieldrin       90

Cholesterol     9
Dieldrin       97

Cholesterol     4
Dieldrin       92
15:35


15:35


15:35


15:35
                            J-14

-------
o
-
o •
f-
® .
*4 '
••r!

"10
•cu
•
o
0
o
o
"9"
1-t •
r? «!


-^
• n
! to
"vO
• n
! o
U
IS

&

£
 C/3
CO
U
o
 o


"?
        10
            o-^
            <**
o
10
to
 I
o
        »    J
    CD-
                                             vO


                                        K   2


                                        in
                                        !«*
                                        10
                                    o


                                    01
                                    (U
       
                                        *,2
                      I   o


                      *   Si
                      O   '
                       *

                      o
                      10
                      to
                      I
                      o


                      to   o-j
                                                     (U
sO

(U
N.
fw

ct
                                                                                   • n
                                                             3UOW)l
                                                                                    o
                                                                                    '10

                                                                                    n
                                                                                    n
                                                 •n
                                                  n




                                                  o


                                                 'P>
                                                  o




                                                  10


                                                 '01
^
 o
 o
 o
^
 o
 o
 o
"o^
 o
 o
 o
 (M
                                  J-15

-------
                  GC/MS Sensitivity Check
TOTflL  ION CHROnfiTOGRfln
Fil« >fl7708 45.0-550.0 a»u. 0.5 PPfl SURR080TES 44593-61-15
TIC
70000-
60000-
50000-
40000-
30000-
20000-
10000-
400 800 1200 1600 2000
, , , , , , 1 , , , 1 , , , ' , , , 1 , , , 1 ! , , I , , , 1 , , , 1 , , , 1 , <






. :_ 	 ..JL, tMl\ 	 ,___







II 1 ' 1 ' 1 ' 1 1 ' 1 1 1 1 1 ' 1 1 * 1 ' 1 ' 1 ' 1 ' | ' 1 ' 1 ' 1 ' 1 1 . 1
8 12 16 20 24 28 32 36 40 44 48 52 56
  Data
  Nam«:
>A7708::DO
   SURROGATES
Quant Output  Fil«: AA7708::DO
                           J-16

-------
GC/MS  Calibration  Check
                  SCMI 1787 (31.401 mn) of 90c 730401004.4 SCALfO

       tnc«
   10000 -



   9000 -i



   SOOO -



   7000-1



   • 000 1



   SOOO -I


      i
   4000 ->


      J
   3000 -j

      I

   2000 -
             i 20


   1000-I  /    I


      t ' i    • ' ;             M       l80

     10O      120      140      100      KO     200     220     240
                             J-17

-------
DFTPP Performance Check
3FTPP Ion Abundance Criteria.
    m/z

     51
     68
     70
    127
    197
    198
    199
    275
    365
    441
    442
    443
     ia
10-80% of the base peak
S2%  of m/z 69
<2%  of m/z 69
10-80% of the base peak
£2%  of m/z 198
base peak or >50% of 442
5-9% of m/z 198
10-60% of the base peak
>1%  of base peak
present and 50% of m/z
15-24% Of m/z 442
198
Measured

    10%
    0%
    0%
    30%
    0%
    100%
    6%
    17%
    2%
    16%
    54%
    21%
Abundance

120000-
100000-
80000-
60000-


40000-


20000-


0-








Scan 1264






,j







r















127

69


5


,


1

i





1

N/Z -> 50
110



,,lii
• • i









l-










136
167
, L
i i .
100 150
i
1
3












i

(22.753 min) : 0101001. D
















255





244
|l
' 'II

200








t -





275


1
i
i • '
250


. 32?34 365 - —
i ! IT l
2










1 ' i • ' ' • i ' ' ' ' ;
300 350 4CO
                          J-18

-------
Precision Data
Sample ID     Date of    Surrogate     Surrogate   Precision
	Analysis   Chemical	Concentration	

                                         ng/I           %

99C3015        2/8/89    Biphenyl      85     56       41
                         TCB           67     28       82
                         HCB           47     34       32

89C4115        3/1/89    Biphenyl      68     63       11
                         TCB           50     46        8
                         HCB           50     35       35

89C4185        3/25/89    Biphenyl      53     41       25
                         TCB           46     52       12
                         HCB           40     34       25
                        J-19

-------
Table J-l continued:
     Example Reports 1, 2, and 3 for the Tissue
     Analytical Procedures
Tissue Report 1:    CHC    Mass   Spectral    Library   Tentative
                    Identifications

Peak RT   Amount    Fit  Tentative Identification  (CAS #)
          (ug/kg)    %
12.602    31
99   Dieldrin (60-57-1)
Tissue Report 2:
Peak RT   Amount
(minutes) (ug/kg)
EPA/NIH/NBS Mass  Spectral  Library Tentative
Identifications

Fit  Tentative Identification (CAS #)
11.748    39



14.889    32



19.756    25
99
90
87

95
95
94

76
1,3-Butadiene,
(87-68-3)
1,1,2,3,4,4-hexachloro-
1,1'-biphenyl (92-52-4)
Naphthalene, 2-ethenyl- (827-54-3)

Phenanthrene (85-01-8)
9H-Fluorene, 9-methylene (2523-37-7)
Benzene, 1,!'-(!,2-ethynediyl)bis-  (501
65-5)
Anthracene  (120-12-7)
Tissue Report 3:    GC/MS Components not Identified by EPA/NIH/NBS
                    Mass Spectral Library Search
Peak RT   Amount
(minutes) (ug/kg)
Fit  Tentative Identification (CAS #)
 8.621    60
 8.680    30
19.756    25
64   2-Pentene, 4,4-dimethyl-,  (E)- (690-08-4)
64   Pyridine, 2,3,4,5-tetrahydro- (505-18-0)

64   Benzene, 1,2,3-trichloro- (87-61-6)
42   Benzene, 1,2,4-trichloro- (120-82-1)

     Unknown
                               J-20

-------
Table J-2.  Example Data for the Effluent Analytical Procedures
Effluent QA/QC Report	

Recoveries                              HPLC Fraction Number
                                              123

     d10-biphenyl                             67%   nda  nd
     13C6-l,2,4,5-tetrachlorobenzene          nd   50%   nd
	13C6-hexachlorobenzene	nd    nd  46%

and - not detected
     For the  effluent example,  illustrative copies of  the HPLC
performance check and Reports 1, 2, 3, and 4 are presented on the
following pages.   The remaining portions of the  effluent QA/QC
report have previously  been presented in Table  J-l.   Consultant
Table J-l  for copies  of the GC Resolution, MS  Sensitivity,  MS
Calibration, DFTPP  Performance,  and Precision checks  as well as
the GC/MS chromatograms.
                              J-21

-------
n
     Detector response               ,^v"
      . ? , ,  2,0. .  4,0 ,  ?,D  ,  gp  ,190

                                         0
                 J-22

-------
Table J-2 continued:
               Example Reports  1,  2,  3, and  4  for the
               Effluent Analytical Procedure
Effluent Report l;
          CHC   Mass    Spectral    Library
          Identifications
                               Tentative
Peak RT   Amount
          (ng/1)
          HPLC Fit  Tentative Identification (CAS #)
          Fr #  %
12.602
31
99   Dieldrin (60-57-1)
Effluent Report 2:  EPA/NIH/NBS  Mass Spectral  Library Tentative
                    Identifications

Peak RT   Amount  HPLC   > 1  Fit  Tentative Id. (CAS #)
(minutes) (ng/1)  Fr #   ppb   %
11.748 391

14.889 132


19.756 250





2 Yes 99

1 Yes 90
87

2 Yes 95
95

94

76
1,3 -Butadiene, 1,1,2,3,4,4-
hexachloro- (87-68-3)
l,l'-biphenyl (92-52-4)
Naphthalene, 2-ethenyl- (827-
54-3)
Phenanthrene (85-01-8)
9H-Fluorene, 9-methylene (2523-
37-7)
Benzene, 1 , 1 ' - ( 1 , 2 -
ethynediyl)bis- (501-65-5)
Anthracene (120-12-7)
Dilution = 25%
Effluent Report 3:
            GC/MS   Components   not    Identified   by
            EPA/NIH/NBS Mass Spectral Library Search
Peak RT   Amount  HPLC   Fit  Tentative Identification (CAS #)
(minutes) (ng/1)  Fr #    %
8.621 602



8.680 303

10 7Rfi 5Rn
1



1
1
^
64

64

64
42
-—
2-Pentene,
08-4)
Pyridine,
18-0)
Benzene, 1
4,4-dimethyl-, (E)- (690-

2,3,4,5-tetrahydro- (505-

,2,3-trichloro- (87-61-6)
Benzene, 1,2,4-trichloro- (120-82-1)
TTnlf nnwn

                               J-23

-------
Table J-2 continued:     Example Reports  for  Effluent Analytical
                         Procedure
Effluent Report 4:  EPA/NIH/NBS Mass  Spectral  Library Tentative
                    Identifications

Peak RT   Amount  HPLC   > 1  Fit  Tentative Id. (CAS #)
(minutes) (ng/1)  Fr #   ppb   %	

     No data with predicted tissue residues of < 1 ug/kg

Dilution = 25%
                              J-24

-------
Table J-3.  Example Data for the Sediment Analytical Procedure



QA/QC Report	

Percent Organic Carbon        15%   (dry mass basis)

Percent Moisture              33%

Recoveries                              HPLC Fraction Number
                                              123
     d10-biphenyl                             67%   nda  nd
     13C6-l,2,4,5-tetrachlorobenzene      nd  50%   nd
	13C6-hexachloroben2ene	nd   nd  46%	

and = not detected


     For the sediment example, illustrative copies of  the remaining
portions of the sediment QA/QC report and the Reports 1, 2, and 3
have previously  been presented in  Tables  J-l and J-2.   Consult
Table J-l  for copies  of the GC  Resolution, MS  Sensitivity, MS
Calibration, DFTPP Performance, and Precision checks as well  as the
GC/MS chromatograms and Reports 1, 2, and 3.  Consultant Table J-
2 for a copy of the HPLC resolution check.
                              J-25

-------
APPENDIX K




REFERENCES

-------
                           APPENDIX K

                           References
 1.  Connor, M.S., Comparison of the Carcinogenic Risks from Fish
     vs.   Groundwater   Contamination   by   Organic   Compounds,
     Environmental Science Technology,  18:628-631

 2.  United States Environmental Protection Agency, Office of Water
     Regulations  and  Standards;   Bioaccumulation   of  Selected
     Pollutants in Fish,  Draft Report; United States Environmental
     Protection Agency, Washington D.C., May 1990.

 3.  Brungs, W.A.  and D.I. Mount,  "Introduction to a Discussion of
     the Use of Aqutic Toxicity Tests for Evaluation of the Effects
     of Toxic Substances",  Estimating the hazard of Chemical
     Substances to Aquatic Life, ASTM STP 657, J. Cairns, Jr., K.L.
     Dickson, and A.W.  Maki,  Eds., pp. 15-26, American Society for
     Testing and Materials,   (1978).

 4.  Murty, A.S.,  Ed., Toxicity of Pesticides to Fish,  Vol. 1,
     69-116, CRC Publications,  (1986).

 5.  Lyman, Reehl and Rosenblatt,  Ed.,  Handbook of Chemical
     Properties.,  p. (5-1)-(5-30), McGraw Hill (1982).

 6.  Esser, H.O.  and P. Moser,  An Appraisal of Problems Related
     to the Measurement and Evaluation of Bioaccumulation,
     Ecotoxicology and Environmental Safety, 6:  131-148 (1982).

 7.  Ernst, W., Accumulation in Aquatic Organisms,  Appraisals of
     Tests to Predict Environmental Behaviour of Chemicals, P.
     Sheehan, F.  Korte, W.  Klein,  and P.H. Bourdeau,  pp. 243-255,
     J. Wiley and Sons Ltd.,  (1985).

 8.  Spacie, A. and J.L.  Hamelink, Bioaccumulation Fundamentals
     of Aquatic Toxicology,  Rand G.M. and S.R.  Petrocelli,  pp.
     495-525, Hemisphere Publishing Corporation, (1985).

 9.  Barron,  M.G.,  Bioconcentration,  Environmental  Science  and
     Technology,  vol.24,  no.11,  pp.  1612-1618,  (1980).

10.  Veith, G.D.,  D.L. DeFoe and B.V. Bergstedt, Measuring and
     estimating the bioconcentration factor of chemicals in fish,
     J. Fish. Res. Board Ca., 36:  1040-1048 (1979).

11.  Spacie,  A.   and  J.L.  Hamelink,  Alternative  models  for
     Describing the  Bioconcentration of Organics  in Fish., Environ.
     Toxicol. and Chem.,  1:309-320 (1982).

                               K-2

-------
12.  Chiou, C.T., Partition Coefficients  of Organic Compounds in
     Lipid-Water    Systems    and    Correlations    with    Fish
     Bioconcentration  Factors,  Environ.  Sci.  Technol.  19:57-
     62(1985).

13.  Davies,   R.P.   and   A.J.   Dobbs,   The   Prediction   of
     Bioconcentration in  Fish,  J. Water  Res.,  18(10): 1253-1262
     (1984).

14.  Neely,  W.B.,  D.R.  Branson,  and G.E.  Blau,   Partition  Co-
     efficient to  Measure Bioconcentration Potential  of Organic
     hemicals in Fish, Environ.  Sci. and Technol., 6(13):1113-1115
     (1974).

15.  Veith,  G.  D.; Bioconcentration  Tests for  Effluents;  paper
     submitted for publication.

16.  McKim, J., P.K. Schmieder, and G.D. Veith, Absorption Dynamics
     f Organic Chemical  Transport across  Trout Gills as related to
     Octanol-Water  Partition  Coefficient.,  Toxicol. and  Applied
     Pharm., 77:  1-10 (1985).

17.  Thomann, R.V.    Bioaccumulation  Model  of  Organic  Chemical
     Distribution in Aquatic Food Chains, Environ.  Sci. Technol.,
     23:699-707 (1989)

18.  Thomann,  R.V.     A  Statistical  Model   of   Environmental
     Contaminants  Using Variance  Spectrum  Analysis.    Report to
     National Science Foundation.  August  1987.   NTIS  #:  PB88-
     235130/A09.

19.  Rasmussen,  J.B.,  D.J. Rowan, D.R.S. Lean,  and J.H. Carey; Food
     Chain Structure in  Ontario Lakes Determines PCB Levels in Lake
     Trout (Salvelinus namaycush) and  Other Pelagic  Fish; Canadian
     Journal of Fish Science;  47:2030-2038; (1990).

20.  Hamelink,  J.L., R.C. Wybrant and R.C. Bal, A Proposal:
     Exchange Equilibria Control the Degree Chlorinated hydro-
     carbons Are Biologically Magnified in Lentic Environments,
     Trans. Am.  Fish. Soc., 100(2): 207-214 (1971).

21.  E.E.  Kenaga,   Correlation  of Bioconcentration  Factors  of
     Chemicals in  Aquatic and  Terrestrial Organisms  with  their
     Physical and Chemical Properties.  Environmental Science and
     Technology,  vol. 14, no.  5, 553-556  (1980).

22.  Veith, G.D.  and P.  Kosian, Estimating bioconcentration
     potential from octanol/water partition coefficients, in
     Physical Behaviour of PCBs in the Great Lakes. MacKay,
     Paterson,  Eisenreich, and Simmons, Eds.,  Chapter 15, Ann
     Arbor Science, Ann Arbor,  (1982).

                               K-3

-------
23.  Veith, G.D., K.J. Macek, S.R. Petrocelli, and J. Carroll,
     An evaluation of using partition coefficients and water
     solubility to estimate bioconcentration factors for organic
     chemicals in fish, in Aquatic Toxicology, J.G. Eaton, P.R.
     Parish, and A.C. Hendricks, Eds. ASTM STP 707, 116-129
     (1980).

24.  Gobas, F.A.P.C., K.E. Clark, W.Y. Shiu and D. Mackay,
     Bioconcentration of Polybrominated Benzenes and Biphenyls
     and Related Superhydrophobic Chemicals in fish:  Role of
     Bioavailability and Elimination into the Feces, Environ.
     Toxicol. and Chem., 8: 231  (1989).

25.  Schuurmann, G. and W. Klein, Advances in Bioconcentration
     Prediction, Chemosphere, 17(8): 1551-1574 (1988).

26.  Zaroogian, G.E., J.F. Heltshe and M. Johnson, Estimation of
     Bioconcentration in Marine Species using Structure-Activity
     Models, Environ. Tox. and Chem., 4: 3-12 (1985).

27.  American Society of Testing and Materials. "Standard Practice
     for  Conducting  Bioconcentration   Tests with  Fishes  and
     Saltwater Bivalve Mollusks."  Designation E 1022-84,
     1986 Annual Book of ASTM Standards, vol. 11-04, Publication
     Code Number (PCN):  01-110485-48, pp. 702-724, (1985).

28.  Sauter A.O.,  L.D.  Betowski and J.M.  Ballard;  Comparison of
     Priority  Polluant Response  Factors  for Triple and  single
     Quadrupole Mass Spectrometers; Analytical Chemistry, 55:116-
     119 (1983).

29.  U.S. Environmental Protection Agency, Office of Water Enforce-
     ment and Permits, Permit Writers Guide to Water Quality-
     based Permitting for Toxic Pollutants (EPA 440/4-87-005);
     July 1987.

30.  James, M.O.;  "Biotransformation and Disposition of  PAHs in
     Aquatic  Invertebrates"  in  chapter three  of Metabolism  of
     Polycyclic Aromatic Hydrocarbons in the Aquatic Environment;
     Edited by  U.  Varanasi; Published  by CRC Press,  BocaRaton,
     Florida; (1989).

31.  Abramson, F.P., Automated Identification of Mass Spectra by
     the Reverse Search, Anal.  Chem., 47(1):  45-49 (1975).

32.  Draft  Technical Support  Document  for   Water  Quality-based
     Toxics Control, U.S.E.P.A., Washington D.C.  (April 1990)
                               K-4

-------
33.  Hughes, R.A., G.D. Veith, and G.F. Lee.  Gas Chromatographic
     Analysis of Toxaphene in Natural Waters, Fish and Lake Sedi-
     ments.  Water Research, Vol. 4, pp. 547-558, Pergamon Press
     (1970).

34.  Burkhard, L.P., D.W. Kuehl and G.D. Veith, Evaluation of
     reverse phase liquid chromatography/mass spectrometry for
     estimation of n-octanol/water partition coefficients for
     organic chemicals, Chemosphere, 14(10): 1551-1560 (1985).

35.  Veith, G.D., N.M. Austin and R.T. Morris, A rapid method
     for estimating log P for organic chemicals, Water Res.
     J., 13: 43-47 (1978).

36.  Eadsforth, C.V., and P. Moser, Assessment of Reverse-Phase
     Chromatographic Methods for Determining Partition Coefficients
     Chemosphere, 12: 1459-1475 (1983).

37.  McDuffie,   B.,    Estimation   of   Octanol-Water   Partition
     Coefficients for Organic Pollutants Using Reverse-Phase
     HPLC, Chemosphere, 10: 73-83 (1981).

38.  Klein, W., W. Kordel, M.  Weiss and H.J. Poremski, Updating
     of the OECD Test Guideline 107 "Partition Coefficient N-
     Octanol/Water":  OECD Laboratory Intercomparison Test on
     the HPLC Method, Chemosphere, 17: 361-386  (1988).

39.  Code  of  Federal Regulations;  40  CFR Part  136,  Appendix A,
     Method 1625,  Revision B: Semivolatile  Organic  Compounds by
     Isotope Dilution GC/MS.

40.  United  States Environmental  Protection  Agency,  Office  of
     Research   and  Development,   Environmental  Criteria   and
     Assessment   Office,   Cincinnati,   Ohio,    (USEPA/ORD/ECAO);
     Guidelines and Methodology Used in the Preparation of Health
     Effects  Assessment  Chapters  of  the  Consent  Decree  Water
     Documents.

41.  U.S. Environmental Protection Agency, Office of Water
     Regulations and Standards; Seafood Consumption Data
     Analysis:  Final Report,  Submitted by SRI International,
     September 24, 1980.

42.  Memorandum  from  C.E.  Stephan  to  J.  Stara;   Per  Capita
     Consumption of Non-Marine Fish and Shellfish;  United States
     Environmental Protection Agency; July 3, 1980.

43.  Status Report, State Numerical Water Quality Criteria for
     Toxics as of August, 1989, Office of Water Regulations
     and Standards, U.S. EPA,  1989.
                               K-5

-------
44.  United  States Environmental  Protection  Agency,  Office  of
     Marine and Estuarine Protection; Guidance Manual for Assessing
     Human  Health  Risks  from Chemically  Contaminated Fish  and
     Shellfish, EPA-503/8-89-002, September, 1989.

45.  United  States Environmental  Protection Agency,  Washington
     D.C.; Water  Quality Criteria Documents: Availability; Federal
     Register, vol.45, no. 231, part 5,  79318-79379.

46.  United   States   Department  of  Agriculture;   Agricultural
     Statistics,  page  506; United States Department of Agriculture,
     Washington D.C.   (1984).

47.  Finch, R.; The MECCA Project: Effects of Regulatory Guidelines
     on  the  Intake  of Mercury  from Fish; Fisheries  Bulletin,
     71:615-626 (1973).

48.  Hunter R.S.  and F.D.  Culver; QSAR System User Manual; United
     States Environmental  Protection Agency/Environmental Research
     Laboratory-Duluth, Duluth, Minnesota,  August 1988.

49.  Hansch,  C.   and  A.J.  Leo.    "Substituent  Constants  for
     Correlation Analysis  in  Chemistry and  Biology."   John Wiley
     and Sons:  New York,  1979.

50.  Integrated Risk Information System Volume I.  Supportive
     Documentation, Volume II.  Chemical Files (EPA-600/8-86-
     032a and b), Office of Health and Environmental  Assessment,
     Office of Research and Development,  U.S. EPA, Washington,
     D.C.,  (1987)

51.  United States  Environmental Protection Agency, Office of Water
     Enforcement and Permits,  Technical Support Document for Water
     Quality-based Permitting for Toxic  Pollutants (EPA-440/4-85-
     032); July 1987.

52.  Thornton, R.W.; D.E.  Ford and J.L. Norton; Use of Monte Carlo
     Water Quality Simulations to  Evaluate  Reservoir Operational
     and   Management   Alternatives,   Presented   at    the   Third
     International  Conference   on   State-of-the-Art  Ecological
     Modeling; Colorado State University (1982).

53.  United States  Environmental Protection Agency, Office of Water
     Regulations and Standards; Chapter  1:  Stream Design Flow for
     Steady-State Modeling of Book VI. Design Conditions;  August,
     1986.

54.  Rossman, L.A., Design  Stream Flows  Based  on Harmonic Means,
     (Submitted for Publication, 1989).
                               K-6

-------