xvEPA
            United States
            Environmental Protection

            Agency
            Office of    EPA-833-D94-001
            Water         Offte* of n*M«rch
            Washington DC 20460
DRAFT
March 1991
            Water
Assessment and Control of
Bioconcentratable
Contaminants in Surface
Waters
                                      Printed on Recycled Paper

-------

-------
                       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 (BAFs) in the
       Chapter 3 formulas for calculating
       reference ambient concentrations (RACs).
      United States Environmental Protection Agency

       National Effluent Toxicity Assessment Center,
        Environmental Research Laboratory - Duluth
          Office  of  Water Enforcement and Permits
         Office of Water  Regulations  and Standards
Office of Health and Environmental Assessment - Cincinnati

-------
                             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, Dxiluth
       fERL-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)
       Oilman 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.
€
                                        ii

-------
                       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

I.   Field Validation Studies

J.   Example  Results and Reports

K.   References
                           vii
                                                                 €

-------
                           List of Figures

Figure 1.1  Approach for Assessment and Control of Biocon-
            centratable Contaminants in Surface Waters	1-2

Figure 3.1  Identification of Bioconcentratable Pollutants
            in Fish Tissue	III-2

Figure 3.2  Identification of Bioconcentratable Contaminants
            in Effluents	.111-14

Figure 3.3  RTCs for IRIS Chemicals with log P > 3.5	...111-20

Figure 3.4  Identification of Bioconcentratable Contaminants
            in Sediments	111-26

Figure 4.1  Procedure for Revising an EPA Human Health
            Criterion or Developing a Reference
            Ambient Concentration	IV-8


                            List of Tables

Table 2.1   Chemicals of Highest Concern.	II-9

Table 4.1   Estimated Food Chain Multipliers (FMs) for
            Various log P Values	IV-6

Table 6.1   Reasonable Potential Multiplying Factors:
            95% Confidence above the mean	 .VI-7

Table 6.2   Reasonable Potential Multiplying Factors:
            99% Confidence above the mean	VI-7

Table 6.3   Multipliers for Calculating Maximum Daily Permit
            Limits from Average Monthly Permit Limits	VI-ll
                                 Vlll

-------
aggregate risk
analyte
bioaccumulation
bioaccumulation
factor (BAF)
   Glossary

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- esponse curve).  ql*'s (pronounced
"que- rie-star") are based on extrapolation
of h  "i dose levels to low dose levels and
a li   ime exposure period.  The ql* is
expr   ed 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 1'brary searching technique for
idfc  ifying components detected during a
GC/i4S 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

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

log 10
log P

LOEC
MCL
NOEC
PHN
NOAEL

NPDES
QSAR
ql*
RAG
RTC
RfD
RT
RL
RWC
WLA
7Q10
30Q5
                            Abbreviations
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
                                 Xlll

-------
                          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

-------
€

-------
                                   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 1.1 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 tne
       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

-------
this approach may not detecr   he presence of some compounds, such
as dioxin, which can form u   oeptable residues at very low
exposure concentrations (i..   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 contaminar-s.  The procedure provides
recommendations to sort the res  ~s of this screening analysis in
order to determine which of the .;ontaminants pose a hazard and
require regulatory action.  The approar  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 ti,,-»
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 re.idue 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. '.  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 cmables 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 eind 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 require the calculation of an RAG 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 watir.   Of  course this increased concentration will
only occur  in organises 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
nonresidue  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 th~rd 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
timeframes 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

-------
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, k,,  by the elimination rate,  kg  [11].  In
equation form:

          BCF  »  k1/k2

     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,
defending upon the characteristics of the individual chemicals.
Sv.,ae chemicals have BCFs of one million or greater.  BCFs for
most compounds have been found to be constant over a wide •  ^nge
of exposure concentrations  [15].  The BCFs of non-metabol    i,
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 performace 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.
k1f  by the elimination rate of the chemical,  kg  [11].  In
equation form:

          BAF  =  It, 7*2

     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-Loa BCF Relationship

     For organic chemcials, 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      (r* = 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%  ipid 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, typic - 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 ASTM1s  "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 flowthrough
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,
lifestage, 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 lifestage 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 [28].

     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 RAG, 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

-------
                                                  Figure 3.1
                                           TISSUE ANALYSIS
                                                          Region or stale develop or request
                                                        development of RfD and/or q* 1 values

                                                                                                                                *
        Collection & extraction
           of lluue simple
          Analyze via GQMS
I No further action
  Reverse-Kirch til chromaiofraphic peaks
   agitnst the CHC matt spectral library
          identified fit >70%?
                                                         Reverse-search EPA/NIH/NBS
                                                            mass spectral library
                                                             identified fit >70%?
     oncentrslioo x dilution x 4) > RAC7
  and/or Water Quality Standards?
        liiue coonrouboa request
Issue confirmation request
No further action  I
                                                                                                    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

-------
     Whe_ . 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 occar 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 c:nfined 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 samplir.  of sessile organisms can also be
used to define expose e 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
-iditional sites should, possibly, be placed downstream of the
nrst 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 c
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.

     The  surrogate chemicals, d10-biphenyl,  13C6-1,2,4,5-
tetrachlorobenzene, and 13C6-hexachlorobenzene, are added to the
tissue at a 5.0  M9/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 Chromatoaraphv/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 chlorideihexane 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,   C6-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

                               III-6

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


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 /ig/ka 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 /ig/kg  are  dropped
from further  consideration. The  concentration  of  5 MgAg  is used
as the cut-off  for  fxirther investigation  because 5 /xg/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  /^g/kg
 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  vg/kg 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 repor-,3, 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 1, 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*. 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
                              95%            99%
                              1.4            1.6
                              1.9            2.5
                              2.5            3.6
                              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 t*"" RTC.

     For each GC/MS component, this comparison is perf   ^d for
each of the tentative identifications which have Rfd,      and/or
water quality standards.  If this comparison is true f   any of
the tentative identifications for a GC/MS component, a
confirmation request is issued by the regulatory authority to the
discharger for that con-onent (the yes arrow leaving this
decision point).  If th^s 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 Recuest

     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 ^n 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 inluded 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.

                              III-ll

-------
     Sampling should be performed when the fac:'ity 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  C6-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 b« 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
 S304(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 optimized 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
                               CoHect 4 extract rtn«rrt sample
                                  Fractionate via HPLC
      I  Analyze via GC/MS
Reverie-search all cbjomatographic peaks
 against the CHC nun ipectral library
        identified fit >70*7
                                                                               Concentration
                                                                               >100ng/L?
                                                          Reverse    i EPA/NIH/NBS
                                                             m       'id library
                                                             ic      si >70%?
                                                                    amtriUon x dilution)
                                                                    ug/kg?
  (concentration x dilision x 4) > RAC7
                                                      RfD, q*l , andAr '  
-------
     To remove chemicals below the bioconcentrative 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,  C6-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 13C6-1,2,4,5-
tetrachlorobenzene.  For the third fraction, all of the GC/MS
components are quantified using the standard curve for  C6-
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 [31J.  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
11 (fraction BCF x concentration x dilution) > 1 Mg/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

-------
Number of Chemicals
                                      c
                                      3
                                      CO
                                      CO
                                      33
                                      H
                                      O
                                      C/5
                                      co
                                      O
                                      OD
                                      o
                                      co
                                      •u
                                      V
                                      CO
                                      en
                                                   c
        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 RAG 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
11 (concentration x dilution x 4) > RAG" 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) > l 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 /ig/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 /xg/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)  > RAG?

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

                              111-21

-------
identified GC/MS co-—onent  in the effluent  is  likely to be  larger
than the reference   ..nient  concentration  (RAG) .  The RAG is the
highest receiving wa-er 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
(RAG) 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 twc
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 cind 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* informaition 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  Conf irmation           .

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 accureite 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 che
available dilution.  Underestimation does not become a concern
until that step of the procedure where the regulatory authority
wil3 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 t  itative 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)
inventor- '.ng and evaluating the chemicals used and produced by
the manu icturing 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 manv 1
interpretation  f the spectral data.

     A chemical would be considered confirmed wh   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 sink
for bioconcentratable"chemicals since sedimentation processes
will gradually bury the sediments containing the residue forming
chemicals.   Sedimentaition 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
       Analyze via GC/MS
                                                           No further action
Reverse-search all cfaromatographic peaks
 against the CHC mass spectral library
       identified fit >70%?
                                                      Reverse-search EPA/NIH/NBS
                                                         mass spectral library
                                                          identified fit >70%?
       REPORT: Vat I
                                                          REPORT: list 2
                                                                                     REPORT: list 3
                                             III-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.
                     i

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 anc 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 activa"ad 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 uq/kq Sediment Concentration Decision Point

     For GC/MS components not tentatively identified in the CHC
mass library search, those present at concentrations less than  5
jtg/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  jug/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
n ;W 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,
ql*'s, and/or water quality 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.


4.1  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 RAG 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 RAG 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 manucil "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 meirine, 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 estimat  i 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 l 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)

              1                      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, FHs 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 RAC 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       ^
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
3.2
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 guantitative 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 RAG.  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 queility 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 RAG may be calculated
using these formulas.  Appendix D should not be used to revise
any criteria or derivo 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
C
        IRIS data
        updated?
            NO
                                  EPA's water quality
                                   criterion available?
                           YES
                     YES
                       C
                      Calculate RAC
Use current criterion
                                                         NO
                                                                  Data exists
                                                                  In IRIS?
                                                                YES
j       C
                                               Calculate RAC
NO
                     Evaluate other sources of data:
                     HEAST, Rtsk*Assistant. drinking
                     water MCL's, fish consumption
                     advisory levels, FDA action levels, etc-.
                                           IV-8

-------
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 RAG
can.be derived using the following equation:

Equation 4.1:

          C or RAC (mg/1)    =  (RfD x WT) - CDT + INI  x WT
                                WI + [FC X L X (FM X BCF) ]

where:      C =     updated water quality criterion (mg/1)
          RAC =     reference ambient concentration (mg/1)
          RfD =     reference dose (mg toxicant/kg human body
                    weight/day)
           WT =     weight of an average human adult [70 kg]
           DT =     dietary exposure [other than fish]
                    (mg toxicant/kg human body weight/day)
           IN =     inhalation exposure
                    (mg toxicant/kg human .body weight/day)
           WI =     average human adult water intake
                    [2 liters/day]
                               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 tc>r.icant/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 judgre^ts should be clear.

     Maximum protection of human he.   n from the potential
effects of exposure to carcinogens  ^.a 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 p.g/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"5 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 career risk, and is not extrapolated.  For deriving

                              IV-10

-------
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 RAG can be determined by using
the following equation:
Equation 4.2:
           O or RAC  (mg/1)  =
                             (RL x
                              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)  =   fRFD x WT^ - fPT + 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 ava. .able 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
timeframes 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 RAG
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 RAG 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 recieving 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 withot 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 qualtiy 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
qualtiy 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.  As these areas
of impact, if controlled, may significantly reduce the
productivity of the waterbody, and may have unanticipated
ecological consequences, they should be strictly limited in size.
As understanding of pollutant impacts on ecological systems
evolves, there may be cases identified where no mixing zone is
appropriate.

     EPA recommendations on mixing zones have four goals:  (1) to
minimize the size of the area affected by mixing zones,  (2) to
prevent mixing zones from impairing the integrity of the water
body as a whole, (3) to prevent lethality to organisms passing

                               V-2

-------
through the mixing zone, and  (4) to prevent mixing zones from
causing significant human health risks, considering likely
pathways of exposure.

     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 water body 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 governs the level of pollution control
required.

     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.


     Allowance for mixing zones is 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 water body.  If
the total area affected by elevated concentrations within all
mixing zones combined is small compared to the total area of a
water body (such as a river segment), then mixing zones are
likely to have little effect on the integrity of the water body
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
cannot result in significant health risks, when evaluated using
reasonable assumptions about exposure pathways.  Thus, where
drinking water contaminants are a concern, mixing zones should

                               V-3

-------
not encroach on dr;,-•  i.ng water intakes.  Where fish tissue
residues are a conct  •>. (either because of measured or predicted
residues), mixing zc  3 should not be allowed to result in
significant health r_^ks to 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 water bodies 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
by the public for fishing, and particularly where stationary
species such as shellfish are harvested.  (B) Mixing zones may
well 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 water body.


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.
                               V-4

-------
 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  diffuseir.   In
 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

                               V-5

-------
infrequent high concentrations as well as more frequent lower
concentrations.


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 lognormal 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
efflue-J  concentration frequency distribution required to produce
the de.  red 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
rand  iy selected input data sets.  The result is a  simulated
sequr je of receiving water concentrations.  These concentrations
do noc follow the temporal sequence but they can be  ranked in
order of magnitude anf used to form a frequency distribution.

                               V-6

-------
Monte Carlo analyses can be used with steady-state or continuous
simulation models  [52].

     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


                               V-7

-------
the 7Q10 is coupled with the 4-day average aquatic life
criterion.

     The duration of the averaging period for determining the WLA
should be consistent with the assumptions used to derive the
criterion or RAG 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 tha 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.
                               v-8

-------
     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
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 RAG 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  =  ( C1 + 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/Q|T = (  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  )


                               V-9

-------
Equation 5.5:       Qhm = n/S  (1/Qj)
                           i=1

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 t'-at 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 RAG 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 RAG) 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, C,m,  may be estimated by any of several methods described
in App  idix 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:

                               V-10

-------
                           am
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

                               V-ll

-------
referenced above.  For those States with 7Q2 design flow, the 7Q2
is generally 2 times greater than 7Q10.

     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)a473] *  [(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.
                               V-12

-------
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
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 reco>mmended.  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 of 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


                               V-13

-------
response to the expected range of environmental and loading
conditions.


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.
                               V-14

-------
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
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.


                              V-15

-------
     In summary, for the four approaches for determining WLAs for
marine and estuarine discharges, the deskt p 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.


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 an  bioaccumulative organic
chemicals to aquatic life.  The ch<  icals 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 hydrophobia
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 ore ir of magnitude as the surrounding
sediment.  This is the result of the similar concentrations of

                               V-16

-------
lipid in organisms and organic carbon in sediments.  Hence the.
sediments provide another environmental compartment in which
bioconcentratable chemicals can be found.

     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 wasteloeid 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 mciximized 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 iss equivalent to human health criterion)


                               VI-1

-------
Where nc -nixing zone is allowed, the RAG would be applied at the
end-of-p_pe:

     Effluent Concentration < RAG

      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 <.ffluent 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 estak  'shing permit limits.
The procedures described within Chapter 5 c   also be used to
calculate the dilution for the purpose of e__luent
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 1  - 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 guality
     criterion" which the regulatory authority demonstrates
     will attain and maintain applicable water guality
     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 RAG.

     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 RAG 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 I22.44(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

                               VI-4

-------
reasonable potential to cause or contribute to an excursion above
the RAC.


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  "fluent limit.  This procedure is discussed in more
detail    the Technical Support Document with the difference that
the me   effl ;"nt concentration is estimated rather than an upper
bound concentration [32].


6.2  T.1, e 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
                                         th
Ratio of Maximum Sample to Mean for the 95  Percent Confidence Interval
Effluent
_N
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
fiJL
1.2
1.1
1.0
1 n
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
0.3
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
a^
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
OJ
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
Coefficient of
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
Variation
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
1^6
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
LI
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=2
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^
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
                                        ,th
Ratio of Maximum Sample to Mean for the 99  Percent Confidence Interval
N
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 1
1.3
1.1
1.1
1.0
1.0
1.0
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
0.2
1.6
1.3
1.2
1.1
1.1
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
0.8
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
0.6
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
Effluent
0.7 f-«
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
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
Coefficient of
0.9 i.n 1.1
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
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
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
Variation
1.2 1 -^
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
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
1.5
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
1.6
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
1.7
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
1.8
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
? 0
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 poll cants 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 maximvn 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.
                               VI-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 RAC 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 NPDES
permitting situations, EPA recommends that the detection level be
defined in the permit as the minimum level (i.e.,  the level at
which the entire analytical system gives recognizable mass
spectra and acceptable calibration points).   The minimum level is
developed based on 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 quality-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 wasteload allocation) by the appropriate value in the following table.

Each value in the table is the ratio of the maximum daily permit limit, MDL, to the average monthly permit
limit, AML, as calculated by the following relationship derived from step 4  of the statistically-based
permit limit calculation procedure in the Technical Support Document for Water Quality-based Toxics Control
[32].
                                 MDL       exp [zma -  0.5a']
                                 i . ...I   =		
                                 AML       exp [za<7_ - 0.5(7'
where a* = In (CV'/n +1)
      o* = In (CVZ + 1)
      CV = the coefficient of variation of the effluent concentration
      n = the number of samples per month
      zm  =  the percentile exceedance probability for the maximum daily limit
      z  = 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
.0
.1
.2
.3
.4
.5
.6
.7
.8
.9
2.0
Ratio between maximum daily anc
Maximum = 99th percentile
Average = 95th percentile
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 1.37 1.50 1.60 1.74
1.30 1.50 1.67 1.82 2.02
1.38 1.62 1.84 2.04 2.32
1.46 1.73 2.01 2.25 2.62
1.54 1.84 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
i average monthly permit limits
Maximum = 99th percentile
Average = 99th percentile
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 .24 1.46 1.66 1.92
1.00 .28 1.56 1.81 2.18
1.00 .31 1.64 1.95 2.43
1.00 .34 1.71 2.08 2.67
1.00 .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 .85 2.45 3.68
1.00 1.35 .84 2.46 3.77
1.00 1.34 .83 2.46 3.84
1.00 1.33 .82 2.46 3.90
1.00 1.32 .80 2.45 3.94
1.00 1.31 .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 tp
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 biocqncentration evaluation  (the
     procedures listed are not necessarily recomiaended, 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 bioaccumulative pollutants.   The
specific laboratory procedures for fish tissues and for sediments
are contained in Appendices A and C of this document.
                              VI-13

-------
:   i

-------
                            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-l,2,4,5-tetrachlorobenzene,  and
 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  of  13C6-hexachlorobenzene
 with d12-chrysene as the internal standard.

     All chromatographic peaks are  reverse-searched against
 (compared with)  the Chemicals  of  Highest Concern  (CMC) mass
 spectral library (see Table  A-i).   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
 tits 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
 CAqn,™          K^         al°ng with the Per<=ent 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
 i?6"^1^^3^0^ al°2g ^ thS Percent fit values, the CAS number
 of each tentative  identification,  GC retention time, and the
 ^Sf   atl°n are  reP°rted f°r each the GC/MS component.  For
 GC/MS components with fits/matches less than 25%  the
 f^06!?^1^*1^ GC Detention times for these components are
 reported and the components labeled as being "unknown".   This
 report is identified as  Report 3.
 ™  T      Pro=edure yields  four reports which will be sent to the
 regulatory authority.   These  reports  are:

      1)    Report  1,  components  tentatively  identified usinq the
           CHC mass  spejctral 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
           ^anTr^/fT?^;Lt0  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
          QclTlcl •


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,
           U * d* •

      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, d19-
           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 ML 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 /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


      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.  Purifv bv
     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
      	-bi
                              13,
(acetone) containing d10-biphenyl, 13C6-l,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 0-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 ML 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 ^L 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 retentign
             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
      sealable  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
             contciin 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 ^m
             Column Dimensions:  30 mx 0.32 mm or 30 mX
                                 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  /uL

      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  ML or 2 fj,L  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 i.o /XL of the GC
            performance solution.   Baseline separation between
            /3-BHC and  -BHC and between endrin ketone and d1P-
            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 ML 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 nL 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/ACQ
                                  'S
            where As= integrated total  ion abundance  for  the
                        surrogate
                   Ac = ^ntegrated total ion abundance for the
                        internal standard, d12-chrysene
                  Qs = injected quantity of surrogate
                  Qc = 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 ML 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 l.o 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 usina
             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 /iL and then spike sample with 5 /xL of the
            internal standard spiking solution,  1000 ng of d -
            chrysene.  Mix sample after spiking sample with
            d12-chrysene .
     9.4.2  Inject a 1 /*L or 2 ML 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-QIS)/(A1S-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  =  (Ax • Q|S • 100)/(A|S • RF • M • REG)
        Where    Cx  =    Concentration of chromatographic
                         component

                 \  —    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
        C6-hexachlorobenzene surrogate and the reported
       concentrations must be corrected for  recovery of
       the  C6-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 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
            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

-------
                       TT           n0t 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
10.
             9.7.4.1  Percent lipid content of the sample.

             9.7.4.2  Recoveries .

             F°5 &o~bj-Phenyl, 13C6-1, 2, 4 , 5-tetrachlorobenzene,
             and  C6-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

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'-dichlorodiphenyIdichloroethane (DDD)
p,p'-dichlorodiphenyIdichloroethylene (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-nitrophenyIphenylphosphorothioate(EPN)
mirex
toxaphene
daiiitol
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.
13C6-1 ,2,4, 5-tetra- 13C6-hexachloro-
chlorobenzene d10-biphenyl d12-chrysene benzene
ptOtl TJTJHl PPm PTDHl
0.5 0.5
1 1
10 10
50 50
100 100
10 0.5
10 1
10 10
10 50
10 100

Table A-3 .
m/z
51
68
70
127
197
198
199
275
365
441
442
443
.DFTPP Ion Abundance Criteria.
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-l,2,4,5-
 tetrachlorobenzene,  and  C6-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,
 Ce"1'2/4/ 5~tetrachlorobenzene/d12-chrysene  for  the  second
fraction,  and  C6-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-s  and  present at concentrations  of  greater than or  equal
 to  loo 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/kq
          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, arid precision
          data.

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 — <_F/ v,w

          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:
          d10-biphenyl,   C6-l,2,4,5-tetrachlorobenzene, d1?-
          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  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 /zL 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

     A 25 cm X 4.6 mm i.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 ioA ''
             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 E)S  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 /xm or 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
      quality.

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  fin
           hexane) containing 1,4-dichlorobenzene,  13-
           diethylbenssene, p,p'-DDE, and decachlorobiphenyl.

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


      7.14 GC/MS Calibration Solutions.  Five hexane solutions are
           required.   These solutions contain constant
           concentrations of the internal standard,  d19-chrysene
           and varying concentrations of the surrogates.
           Composition and approximate concentrations are given in
           TclJDXG B~2 «

      7.15 GC Performance Solution.   10 ppm solution (in hexane)
           containing  /3-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/min,  and
            set the UV detector  at 200  rim.   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 /iL 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 /*L 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 (J.L 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 fj.L  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 chromatography 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 ML 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 /zL

     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 ML or 2  /zL 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 /zL of the GC
             performance solution.   Baseline separation between
             £-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 /iL 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 l.p juL 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 /iL 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 Reprcducibility ~ 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 /iL 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 inL 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 techniques to
             complete the phase separation.  Optimum technique
             depends upon sample, but may include scnication
             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 quantitative
             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 bath1
            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 /iL.  Recap the vial and mix well.

     10.3.2 Using the HPLC conditions determined previously
            (see Section 9.1), inject 175 ML 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/Licmid
     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 1, 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 pL 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 C18solid 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 C18SPE 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 juL 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 /iL and then spike sample with  5 pL of the
            internal standard spiking solution, 1000 ng  of d,2-


                          B-18

-------
             chrysene.   Mix sample after spiking sample with
             d12-chrysene.
      10.6.2  Inject  a  1  ML  or  2  ML  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, 1) 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., l.l 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 -  (VQ|S)/(A,s-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  =
                   Q|S * 100) /(A|S • RF • V • REG)
              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

          AIS =     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 "Cg-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~bipher,vl.
                  For  fraction #2,  use recovery"of
                   C6-l,2,4,5-tetrachlorobenzene.
                  For  fraction #3,  use recovery of
                   C6-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 #l,
use response curve and recovery of d10-biphenyl.
For fraction #2, use response curve and recovery of
 C6-l,2,4,5-tetrachlorobenzene.  For fraction #3,
use response curve and recovery of 13CR-
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 i.

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 beina
        "unknown" .                                  y

10.8.4  Report  4:  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 .

10.8.5 QA/QC Report

       10.8.5.1 Recoveries.
           d10-biphenyl, 13C6-l,2,4,5-tetrachlorobenzene,
       and  C6-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  C6-l,2,4,5-tetrachlorobenzene or
                           C6-hexachlorobenzene

          Fraction 2      should contain 13C6-l, 2 , 4, 5-
                          tetrachlorobenzene and should not contain
                          d10-biphenyl or  C6-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

                                       re-extract and re-analyze
     11.3.3 GC/MS  Stability
rJ/M^ Assu^ance-  During the GC/MS analysis run,
GC/MS calibration solution should be analyzed

1CCe a  the b€*
                   /o      the b€*Tinning 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

            5S?i21?iHI?SCedUre (Section 10> •   The sample is
            e£tn I^H   *   hS surr°9ates 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.
             K>)   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
           JS«rf^°f containination.  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
                    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)
hexachlorobenz ene
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
                               B-30

-------
    Table B-2.
 GC/MS Calibration Solutions, Concentrations of
 Surrogates and Internal Standards.
13
 C6-l,2,4,5-tetra-
  chlorobenzene
      ppm
                                                13
d10-biphenyl
    ppm
d12-chrysene
C6-hexachloro-
  benzene
0.5
1
10
50
100
0.5
1
10
50
100
10
10
10
10
10
fJUIH
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
SJP'-DDE
2 , 2 ' , 4 , 5 , 5 ' -Pentachlorobiphenyl
Retention
Time
4.3
5.5
6.8
8.1
13.4
15.1
Loa 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            >i% 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
 C6-hexachlorobenzene and the internal standard is d12-chrysene.
 Standard  curves are  calculated using  an internal  standard method
 for  each  «urrogate 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:  d^-biphenyl/d^-chrysene  for the first fraction,
 C6-l,2,4,5-tetrachlorobenzene/d12-chrysene for the second
 fraction, and  C6-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 (CHC) 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
          ??/5S i™™^ ant ^concentration are reported  for each the
          JC/MS component.  For GC/MS  components with fits/matches less
          man 25-s, 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 usina the
                   CHC mass spectral library.

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

              3)   JSeport 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,
           D.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-
           djp-biphenyl,   C6-1,2,4,5-tetrachlorobenzene, d,,-
           chrysene, and  3C6-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 only7)

     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.

          4.1.3  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 piL 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 soluti^i!

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 M^ or thJcker
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


     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  C6-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, d19-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 /3-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
                 ?°* 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 ML 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 ML 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 t  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,  HPLC1
                      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  nL  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
    j.       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 pm

                          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 /iL

     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  nL of the GC
            performance solution.   Baseline separation between
            /3-BHC and  -BHC and between endrin ketone and d19-
            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
     v      GC/MS calibration solution.  Using the total ion
     -[.      chromatogram, a signal to noise ratio of greater
     &      than 3  should be observed for each surrogate.

     8.5.3  MS Calibration — Inject 1.0 /iL 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

-------
8.6
     8.6.2
      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?3,  where RT
             is the observed average RT (in seconds).

      Response Factor Calculation for MS

      8.6.1  Inject l.o (J.L of each GC/MS calibration solution
             and acquire GC/MS data.

             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

             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
     8.6.3
    8.7.1
    8.7.2
            HPLC

            8.7.1.1





            8.7.1.2



            8.7.1.3



            GC/MS

            8.7.2.1
                    With the  following procedures,  verify  HPLC
                    column performance at the  beginning  and
                    end of each  12 h period during  which
                    analyses  are performed.

                    Demonstrate  acceptable performance for
                    criteria  described in Section 8.2.2.

                    Demonstrate  acceptable performance for
                    criteria  described in Section 8.3.2.
                    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-d19 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 •'?.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 mL 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  toluene:methanol  (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.

ff.-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.
     W
9.6  fckmoval 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 ML.   Mix extract and  cap vial.

      9.7.2  Using the HPLC conditions determined previously
             (see Section 8.2),  inject 175  juL 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/Liauid
     Extraction)                                  H
                         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 ML 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 mi* diluted  fractions thoroughly.

      9.9.2  Activate three C18solid 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 nztrogen gas,  a slight vacuum,  or (spun out
             using)  a  centrifuge.   Do not let the columns  qo
             dry.                                          y

      9.9.5  Elute C18SPE columns with three 500 /xL aliquots of
             methylene chloride- and then  one  500  uL aliauot  of
             hexane.                                    M

      9.9.6  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.

     9.9.7   Concentrate  eluate from drying column to
             approximately 100 /LtL and quantitatively transfer
             concentrate to microvials.  Cap and store in
             freezer.  Three extracts for  each sample will be
             obtained.

9.10 GC/MS Analysis

          l Remove sample from freezer and allow it to warm to
            ambient temperature.  Adjust volume of sample to
            100 /iL and then spike sample with 5 /uL of the
            internal standard spiking solution, 1000 ng of d -
            chrysene.   Mix sample after spiking sample with 12
            d12-chrysene.
                         C-19

-------
      9.10.2  Inject  a  1  juL or  2  ^L 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 =
 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-l,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
Where
         4is
         M
         RF
                    QIS * 10°)/(A,S • RF  • M  • REC)

                    Concentration of chromatographic
                    component

                    Area of chromatographic
                    component in total ion
                    chromatogram

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

                    Area  of internal standard in
                    total ion chromatogram

                    Mass  of extracted sediment,  i.e.
                    20 g

                    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
                    hexachlorobenzene.
                                          '13,
        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-l, 2,4,5-
                   tetrachlorobenzene.  For
                   fraction #3, use recovery of
                    C6-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 #l  use
         response curve and recovery of  d10-biphenyl.  For
         fraction #2,  use response curve and recovery of
          C6-l,2,4,5-tetrachlorobenzene.   For fraction #3
         use response curve and recovery of 13C -          '
         hexachlorobenzene.                    6

 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'
        T™!PeCiral library-  ^ose 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 ^lo-bipheny1 '  13C6-l,2,4,5-tetrachlorobenzene,
       and  C6-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   C6-l,2,4,5-tetrachlorobenzene or
                           C6- hexachlorobenzene

          Fraction 2      should contain 13C6-i,2,4,5-tetra-
                          chlorobenzene  and should not contain din-
                          biphenyl  or 13C6-hexachlorobenzene

          Fraction 3      should contain 13C6-hexachlorobenzene and
                          should not contain 13C6-l, 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.
            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-
  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'-dichlorodiphenyIdichloroethane (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-ethyIhexy1)phthalate  (BEHP)
hexachlorobenzene
aldrin
alpha-hexachlorocyclohexane (alpha-HCH)
beta-hexachlorocyclohexane  (beta-HCH)
technical-hexachlorocyclohexane (t-HCH)
pentachlorobenz ene
N-nitroso-di-n-butylamine
heptachlor epoxide
dioxin  (2,3,7,8-TCDD)
ethylp-nitrophenylphenyIphosphorothioate(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.
13C6-l,2,4,5-tetra-
chlorobenzene
ppm
0.5
1
10
50
100
Table C-3 .
Chemical
GC/MS Calibration Solutions, Concentrations of
Surrogates and Internal Standards.
, , . . . 13C6-hexachloro-
a.10~biphenyl d12-chrysene benzene
ciom ppm „„„
0.5 10
1 10 -
10 10
50 10
100 10
r*r"" 	
0.5
1
10
50
100
?ISgwo^d^R^?nti°n Times (RTs) for Compounds in
the HPLC Calibration Solution.
Retention
Time T.nrr D
Benzene 4 ^ 3
Bromobenzene 5 m 5
Biphenyl 6 m B
Bibenzyl 8 ^ 1
^p'-DDE 13>4
2,2' ,4, 5,5" -Pentachlorobiphenyl 15.1
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            >i% 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

-------
-«J  -*J O»
S  S 3
a  2 a
o>  o>

S2
o»  o.
a  "
^
=  !3
£  2 S!
   §3- 3-

   3T O
—  -» -1

  il
   (9 3-

   i i
     z  o > x
o  I

O  3
—*  O
P
^  rj -s -s
a>  vji wi w
fr  (\J -O -N|

-«  n z o
O"  O"
r>j  r\j
ui  LM
U*
..  s s
J?  2 S
-•4 fVJ -*
*•» *- ru

•O •O W
                     If
          — 3- — 3 — J5
          - — —-<• a -
       -< O  1
        — M  <
          9  °
          a »
                       •<

                       S
                       3'
          5  1
             o -o
             M -5

          •n  /sj ia  n vi n
          —  -  3  3- -> •<
          :r  w i  — -t a

          1  V 1  $ •* 3.
                •  a =r a
          a  o s at 3 it
          ~  i  It  3 — -
          3-  - J«  i» 3
                                 1 =
                                 — o
                                 9  n
V^  ^ "^  V^


1*4  1*4 (NJ  _»
co  vi w  —
fNJ  -O Wl


•o  -« n
                                                        Sea  ro «o
                                                        ut  os w
                                 i?
                                         *— »
                                           .?
                       S f ?  - 6
                       2:^? 3  S 5'
                         §— ^  yi —.
                         - r»  g -
                         — •••  Q, "i

                         o  S  2 "3^
                         X  3T  3- »

                         a o  — o


                         *l\~
OOOOOO^MO  »  O  »OOOOOOOOO  »
                                                        11»p»pp

*_>oao'->v/i      1*  'o'ooersjoooeo'oe    b   bru

   « ° -                 «    § §.§    ^iip^    oo    S   §
                               § s

     •^wioox^   -»oooowoo*^-*wo   oo   ruoin^c*
        §•*•*    rj''*-r\)'«-«*-a3'    -^   ^..po
        V* *^ *"• -O   •  **  ** 09  vJi •   ** ^ O O* o  **   •     •   r* r\j ^ o
        <>      (\j   r«      i^  oa p   -*j   •   S*      OI-NOOO
                                 K        Q)             rO      o o
     p

     s
                       OOMIMOOWOOOCO
                                                                       n o » so
                                                                       O •   f™ ^



                                                                       I " II
                                                                       S >r i  x.
                                                                       — ID O> —
                                                                                                      > *





                                                                                                      I?








»N, O
K"

i
O
"^
i*

1C
X,
a
c
i
3
f



S


n
n

3'
33
m
c
e
o
_a.'
n
jt
1
i'

9
i
03
o'
a
n
g
p
M
»
**
i'

a
n
"
30

?
f
ffl

i


7
3
j
*
•i
o
tn
V)
£

rn
i
"•*
i

n
rn
rn
z


<
2.
c
V)
^
i

U1OOO>    t* *M   W
O  f  •        o o   ^)
ooo      99
                                    IVI-OOOOW
                                                                       I S " S
                                                                       I * II
                                                                       f|-a

-------

                                                                            § ?
i -< -i
s a. - g
x" 5. £. °
^— tu o
^ ii §.
ra o
o "~
3
"










p I p p
O O4 fvj
Wl


i : P
§







O -• O <3
JIN ^" U!


S S P
§ § w
71 .-» -< — 3- £
0 ^ ~ S 0
X — M o -I
a - £4?
(D (* -*> O
u (b —
a ^

QJ
Q


3







O O » O O UJ
• * * . • o
0 O O O
0 O O 0
§ g •* s
tsi O
o
* » **
wn







O -J O> U.
• ^ *Q O4
fvj


O O Ul
- b g
8 °
loromonof tuo
orodif luorom
al
on, sodium s
a_ n ->

3 -»

0)
a






o o o o
S 3 ~ -
rvj











w r\j
- « s


•-» ->i *> i\»
o o o
0 00
n
3-
o"
-)
O
re
3"
•<_
i»"
3










o
8
•o


p
»







fc


o
i
iform
idichlorometh
03
H










O O
b b
M I\J


o
8







°* iS


gP
3 O^3tQ3 OOv n
-j a -. o ^ oo-< 3
Q.O ft *+ 3 »^ m o o
•"' 3 J" 1 3" ^
W (S -< — O O ~-
— =r §" <-T a H n
-*> 0 3
a S a
(^ ro












p p » p p o * * o o
-» o o o o bo
^ O M O 0 O
O* -* Wl O


o * o o » *
' * • • » *
o 04 ro
0 S K
Ul







— oo -»'-r--.njw
• • • • o PO wi PU
W-**1^ — » ^ Q3 O .
•**• w o -F js o
O O .P-


p g — p p p o
-o o b b b —
O 0 f\i
*— 4 ^-^
ene
1-Trichloroe

3"
01
3
0









O »
S *


p
b
o







S 8
ro


SP
s
o
o


c
Crt
s.

3'
j
n
' 0


§'
| 3 S.
. >
t
^ J3 ='
J =0
— ia
"• c
ia —
^. 0
a S
ffi O
-<



w O3 m o'
X n. w 3
r- _.
"• 3
TJ 5
51 i
1 8 " S

f—
•5'
a

n
S
w
1
i"

a>

03
o'
a
n
n

f
i'

3
O
O
-1
V)

Should
jT


n
o
3
,n
£
~

n
V)

O*
-1
I
9

£
3
§
A
£
£
O
w
Q-
—
»

C/)

Itt
—
o





c
C/l
o
2
^
i
CO
n
n
m
m
^
ac

0)
c"
(9
T
i

—
M
t/j

-o
S
o



*~
g
    is
3 o :. o


1S=I
"O 5*" ni ifi

-------


s a
Wl -^
rj r\j
r* *•«
-^ — i
o
zr
|
o
§
X
n

o
R
a


o o
O 0
- §



*
»









O9 — *

0* S**
M


— O
CK




O O
*» f\J




•o < -o o
U f\3 M -1
o* oft wt *o
>J1 *^ FVJ **
IM U1 «- _•
X a -» w
i | r ?
a — a.
5' * "'
i i"
8-
s
3
M
CL
i'







o o o *
o o o
O O Wl
W M



nj o
a *•>
U1







*» (S) f\J -*

01 S" "s|
N


0 0 M 0
""' g §
§ §
IN)


O O O O
M S -S g

S -s


ui <>
-a PJ
o rvj
g >
I t
ichlor
i
1








o »
§



*
*









fNJ — »

•
r\j


o
s




0
o




CO CO CO
\J\ ^ l^i
T3 2C X
moo
3 X X
~ CO 03
111
|||
111
1








o o o
s § i
ru ru



0

a>







^4^01

S o -*1
ro


o o o
~ § §
W, 0



000


_^


1
O^
z

-t
o
w
8.
i
re
i
re'







»




o

S
'O






1/1





o
§




0
o




CO CO CO
—* ~*i ^3
co a -a
c c J-
•<. •< 3-
S- o- =
^» S o
? 3 o,
5»l
^* ? a
-< o n>
•<" o
o
o
0)
ffi




— * o rvj
INJ



*
»









OJ -^

(M fNJ
• O



g ^
^




•0 p
Ot




v/l-f*JT*JSo>J
oomo»-ocoo
^' ff 3- «' S 3 ™ w'
C C*< ^fo^^a-
"1¥^i 3-2. |
•g. f ?. " o 3 2.
Sx 5 § ?"
Q) .— (B — .
— ^ r+

~ 3
o m
o
5"
• *









O3 *S ^ *O -^ 4*> -^
O • -*4 _. _» . Q>
CD -O • * O ^4 *-
rv INJ f\» o O*1* •
CO


-* J- «-n o o o w
o o In o '-g
O M




p -* rvj p p p —
*•* O O fVJ O (NJ
o o
o


i
wi
—
>
7
Tetrac
3-
cT
-i
o
re
3-
§
re






*




o
PVJ
O







-^

In
rv>


P
o
0
w



o
o
o



33333
M O O^ -* O

X * * -* — »
ft n r> -i »
? •< •< n' »~*
- o 1 f ~
rt —.-!-)
=r » g. o -•
o — *•* sr
-i a 3- —
0 -< 0
o *— n
a no
5- S S
3 ?
re







» o o » o
§ § §
o *•
M


*• O O
Ln o b
-ft U1
-* *>j







p O O O

***"§?



rvj p o o
§° b '•- b
a w




-4 O O O
o b b b
O Od Q
M «


§• n
re ^
-T (/)
X O
ffi 3-
* i.
o
0}










J "
yr
(O

a.
s.
3 -
& »
•^
3^
^
Q.
CO
*<



O4 ai m
X n v>
•n ^*
-. a'
•o s
s 1
P ° ? 2
O • ' ^
" 11 "

^ jr T1 ^
s 1*-
•<
8 ° s s
3 o « o
M ro
1 °II








n
• r- |
•5' c_
c a ~
i o s
— S CO
3 ifl
J^,
'-;
o S I
o a
- 3 re
o' S.
0 "I
°§ |
jo n ?


3 J
9 1
ID -. TJ
•§ s s
— V)
O -n —
«-» O CT
00 —

•< o -
n —
^ tn ^
o' o "

S "S
Q
sr =;
o










o ~a
o »
•^. ta
"U -5
CO
'O O
0 •

~*



•
— .
CO
m
i

•*
i
en
n
m
m
z
i


0
1

^>
§



!/i
«

^

o














-------
                  Q.

                  ™
                            ?
     ^£
'— •— -I  — 3  o

S  S S  ¥ 2.  7

-§i-i    3
2  5- o  3     n

                                       o         3


                                       > A    3  a

                                       It    J 7.
                                       "*X      (Q  O

                                       ff      £ S
                                       >      a>  -n
                                       CL      -»  o>
                                       o>      on
       S
       (A

       -*>


       i
— po

o o ki
                 o
                 s
                                                                                   -
                                       (VI  *> l\> O O fj
  s
                                                                           S  b r: s

-------
S
•^. -
JNJ
sfjjs||illlll|ls§ia§22|flfllSS§
3 » — * •< *C • 3" •< •< 3* o • • CD o 3* ••» — r» » •— n -"*•"* -o.i »~*
Q, •< yi i x "i r> re rt o — — TJ n — r* — 3- — n M — *•> w M — o c* *••
S'tn^— 003-3'- — OC3»--im3-OX»<-< ' -< O ' -03--1 .
C 3 3* **• SI O ^~ 3* ST Q 3 3 M n *•" O O O •* • O C "•• *~ O trt — •
•" '•"33 "I -t> O CT O *< O 33ffl3OO—'**CTOTOO*
•*• O Q O O. X X re — — OJ— So— — 3"3"3g-i-iO^-i
a T cr tn •< 3 re 3^*^ n**^*- n.ooo» g
™ O 333 S. Q. C — ^<" O — O -• 3OQ.— n"
3- 3 rt>^-»(ft — • -j-^o^-oreo. o^*~i-" o
f 1^
(D ^


oo o o ^ njor\joo-*o o oooo
*S22 IN» O Ul . VB O OOU1-*
\S1

— »» »»» » 0-0»»0)0»
* »» »*» * •_*»» VJ1 . »
** 2










rvioo-*-'-*njo->i>jroo-s -*oooor\JruowiLxowt-g-* w
SSS^biuiMSiS In ** ^ g'4NJN'eN*KoJo*S§of'Csr\j'*'? r^
ro ru —
•o
a
ivjoro o. u CD -• oa cn ro -N4 «> o>-*oo o w co -4 o no
0*0 ooo ooo o oo<* > o* o*
0 000000000 0 0> 50
— * ooo ouio o tnp
o ooo C *
o o 3 w
S w U sr
_. (Q
u
-------

124481 Dibromochloromethane
126987 Methacrylonitrile
27184 letrnchloroethytene
29000 Pyrene
31113 Dimethyl phthalate
31895 4,6-Dinitro-o-cyclohexyl phenol
33062 Captan
33073 Folpet
33904 Chloramben

oooo » o o o o
Ul C*J f\J o

b : *
Ul









03 g PJ j— Lg O O
M o vsi OB • .r* r\j
** -• o ^*j r\j r»

O • • o
2 ^




O p O O O 00
b o *->-> M-'UI^UI->Ul!M->
OJ >U4 M^O. *U1 -NJO
O^ f\» O^ O^ O* O* O

w-— o-»rw-* o*r*joo -»O*PJO
o * o o • • • o •
O O o 09 O O
•^ Ul



O£O«*OO IN) o O O WMOO
• o- •• O'.' o ••
OD O O OB Ul O» O M OB O
00 0 0
2 S 2


117817 Bis(2-etnythexyl)phthalate


o
M

O
b









«.
o

o
b
o
o



0.000006


115322 Dicofol
116063 Aldicarb


> »

: ?










_ «
-• o

o
o
0
U4


O
b
g
O
09




















n

i
»•»

3
Consumpt
— •

















w
X
r-
'•5'
a
o
b
§
yr
.

1
z
a

3
(2
"V
Q.
0}
^.
i
ia
^^
a
0
•<





03
O

3O
^

I



f
O
m
o

CHS Chemical

n


j
*









m
i'
i
*>
o
I
•^

*^
n
I


c
V)
2.
3
|
n
m

i'
o
-*.
JQ
O
3'
1
D

O
-I



n













a'
n
o
p
3
D
s.
09
o'
n
0*
i'

9
n

0



(A

_
a.










b
3
tfs
CO
O
-i
f
a

9
|
1
T
•o
s
-.





crt


"
t/l
o
n








O TJ
•o a
>. a
fNJ ^
O3
« 3
0 ?
o>
c
00
m
O
r—
i
n
n
m
T)
z
z
CJ
<
s
-1
1




00

«

^
-o
0










-------
Ul
o
Chlofobcnz
s"
r»
fa












p
o
r\i








U1 \J1 \JI \JI \jt tfl t*
O O O ^9 Q O ^*-
"*^ Sc 5c K ** ®
O *• W -O 3* Wl  > 3
C O 3 < 0 g 3
£ i- 1 * i i' 1
g.* 3 o g 3 3
— • ii CT a T
- 3 1 | = |
* £ §•"* 3 *

n
3
3
Q.'
 O
VI O i«


» »
» ••





ix w -•
O O 'O
|if
o
X
Qt
n
3"
o
o
n

n
£
o


i





o o *
s §
f\j nj


* »
4 *





i
VI
•-J
03
z
fo
X
&
n
3-
o"
O
n

o

3^
m
X
m
3
(t>





*



^
CO





< ^
™ o
& >
"£ S
3" 1
O" —
' 3
z
<»
X
QJ
n
3-
o
"1
0
o

n
i
n

CO
3





» O
O
s
0
M

O —
W ^





ru
O
r\j
Hydrazine/
x
i
0
N

3
Fo
&
£L

o
^*
n





*



U4
O




W PO M
§11
Wl X* O
S? ° 3F
9 •*• n
2 C 3"
si
a
3-
i'












o o o
§ § §
r\j o o
O M
** ^








•T* O
O Ul
trans- 1,2-
Fluoranthe
»!
o"


1^
3-


n





o o
So
rvj


•
»





-* o
o o
O9 VI
Herphos
Potassium
r>
i
a
(9











0 0
S 1
§
U4








09 VI W -k -* O ^
-* *g w -NJ ». *v ro
09 O4 W ffl ^ o O.
Wl 1>1 • IN) (\J 08
1 i a i 1 1 t
-• 0 -• «< -i •6 n
i 5 i - 5" S i
IS 0 -.
Q. — 0 0 3
? is
•<_ a i?
9:

3"
O
O
0
-1


i
IT
™




o o o o o o o
o o o -o o o o
w ru ** o nj o
O V









1 2
n trt
X. 0
S 3*
?3 •§
n
0)





C
w
A
Q.

3'

~~
QJ

o'
a -
t s a
?r
(Q V
1 *
M; — .
-..a 3
It i?
'**. (D
a *
a o
•< -i







"§
a
n

w
i
s


s.
O3
I'
D
o


S"
**
s
o"
n
o




o


c_
"


00
?
c
31

m
a>
a.
m
i
<
n
^
s
?

."
o -o
•o a
^ J3
'vj T
1 =
O '
^J




1
c


i
r~
i

n
m
m
X
c?
<

i
w
1
3

2
in
b>


o
**


0
—




ro -.-.-.
In ^ ?* ?"
^ OB 0 OS

o o o
•o o o
o o
o a
o a

0 0 O
w 11
S 0



s
as


o
o
o
o
g
0
1
g



ui o wi r\j •
• • O3 v^t
•o *» o
a>

o o o o -o
W 0 1* W °
-«1

o e o o w
-00-°
ru




o o o *•»
V* \ji ** .


Srvj w wi
o
o o
§

s s ° -
o o •
O 0 -O
o








o
i
tn
1
i'
o
i
•D
r*

S
M
X
r-
•a*
a
o
'o
vyi
€
0
i
7T

»
i^
o
m

<>

m
V)
^*
i'
i
90
s
I
^
3D
^
O
<^
j
»s,

*^

-------
S S 8 3 ! * f* * 1C 1-3 I 1 * 2; g f = 8    U 2 i 2
?  ir  It  ~mi°ii-ll!  IJSif^z
3  IT    =•-   "  ~'al---i?S  £ * S-' ° I ° § S
   *                           *


              ->  -S 5 3 o

              I  JS  »l
              ?  3    (9 fB
              to  .*.    ^
              i  S    §
P .° : : ? P ? ? ? p p p * p p p.p p p p p
 0 0 0 0 M

 -00-
 pppoo  —*>oi O (%
                                             1   " i i
                                 ,-u f
                                   ffl -i eS

                                 <2 g' ? -»
                                 C_ 3 w -i

                                 ^ a CT
                                 on— —
                                                 O O  ^
                                                 3 C 0-1 -o
                                                 '   ~

-------
^CaQ»ggf\JQ«sj^-*vior\j'OjF-
ra -!—«—« <•» ffl 3" IB Q) 1 1
5 n -i S* — c o — 01 xx
3-C9e3-5~3'--<<
— 3 re -» 3- 33
o o — —
§ 0 ""
1
01
O










ooooooooooo
ooooooouJv/ioo
O -* -•j c* o o -* t\t t\j
—* O4 Ul X* Ul



»
*










• ^J Lrt O *O • 5> • O W
r\j PO o co o J> o o


o «o -«j ro ru o INJ ru o o
0 ° ° :<> ° b
5 *-



o w rvj ^i o OOUIOIM
. o • • • •
ru -•
Illllll
— -i — T" a — o
Q. o a — -i — •—
— I. -) •* g- en •<
O re ^ o Q n
a ~ c •* -6 3-
•— 5 -o T
s. 5' D - i'
? s °
•? o-
_^ "5-
rr re

"i- ^~
o
i
re




O O O O 0 O »
o o -• o o o
O — V/l O O CO



O -M
1

"X








is ? is oj f i,
rj


03 ** o ? *" o
o o o
00 0



M _. -. 0 — ^1
l§ §
M M — * -fc -* &•
PO O ** ** ** OJ
M «J 00 O W —
-* o ** IN) na *o
** -Nj *J -• s/l V^l
> X M < -< O
« *< — o> 3- re
cabromodiphenyl
allic oxide
nadium pentoxide
nc phosphide
lenes
best os
re
3-
re
-i







» M O O » O
§ § 2
o -o



» » » »
» » » *










i
o
o


o
§



o
0
1116547
z
Nitrosodiethanol
1.

3
re







»




M
09










^


O
0



o
8
o o ~o -o
-s fvj O» v/i
— • *» _• ->i
CD vn -• i^
L*J "0 — > -b
O* W Wl ->4
O Z — « O
^ rt> (S — .
ph en am id
'trachlorovinphos
ptachtor epoxide
yphosate











o o o o
-• o o o
OWL*
o
o

w
»  -M


O Ul U
8 °
o


O nj -*
_ 0
o* r)
(? >
1 trt
z n
9? 3°
i i
n'
£.











I 2
«
(C

a
o
^
t "
^
CO
£L
O






X O W)
tl *^
f— ^.
— 3
•a 5
£ g_
S P ? 5
5 o •• o
•fl v/l S <1
-•  »-
s | "
D O 3D 3D
o • r- >
1 ° - "




c
irt
S.
_fi
3
|
n'
a

f

o

0

3"
JO
c
*~
r+
O
-1



O
",
§










r-
•o
CL

n
^
yi
I
I'

03
i
03
i'
n
n

3
o
S'


(9
n

o



VI
1
?









n
o
3
C^
^^

n
V)
_^
0
1
^

^*
0}
5
a.
§

3
j
«
j
V)
Cfl
^

ffl


V)

n
t/i
!

o












t/J
m
o
z
-n
6
55

o
m
m

z
CT

<
0
i"
V)
•^
i
3



tfl

•o

•o
o








1:11

-------
          P J.°.°?<=<=opppppoooo  »oo

5   -   I    3-s§ass22'-
       •N  -^
                                    IjJ _ O -« I\J

                                    VI O ff» -• -O
                                    •  •  CO 00 PJ
                                    ru o   5
                                    oo oa   rv
                                    P o P   ** P
                                    o   o     r\»
                                                     •o'  1

                                                     a
                                                   vj o



                                                   "O   rn ffl
               o
               O
g p
o o
O WI

ru
y ^ ? s
1 g •• "

I ° II

-------
12035722
12039520
12122677
/sj -< Z
3 o n'
fS — K*
cr — a
•I
a 



8001589
Creosote










8001352
Toxaphene










7803512
5
O
VI
3










779H20
Thallium c
3-
ST


Q.
re




7783064
Hydrogen s
c_

•MI
a
re





-j
CD
Selenious
CD
0

a







7782414
Fluorine (
V)
o

£
er

u
••»
f

f
7723140
7773060
3 re*
Bf
V) VI
£*
-*» O
§T
C
«





r^
i
Thalliumd
""
VI
c_


s
re



7440611
Uranium, n
a
C
-t







7440473
7440508
Chromium(V
Copper
~









7440439
Cadmium










7440428
03
O
S
CD
O
O

CO


09
O
a
re
Cn
0
7440417
Beryl! ium










7440382
7440393
Arsenic, i
Barium
g

(O
0>
3
O




                                                                  o
                                                                r- D
                                                                — V)
                                                                •a c
                                                                o -*
                                                              — 3 w

                                                              3 s -

                                                              a < °
                                                              2. -. c
                                                              ^^ (8 ^*
                                                              — 3 *»
  o o o

  o o o
  04 O O
§
o
S
         p p
         kj o
o
S
                                                        in

                                                        !
                 -*   »
                                                        it
                                                        (O

                                                        5:
                                           n

                                           = N

                                           ill
                                                              -. 3-


                                                              D C

                                                                 "
                                             n

                                             <
                                                      •o
                                                      a
                                             If =
s
                                                    1
                                                    — ta O —
                                                    S 3:   -
                 §
                 §
                 CO
                                                    a °
                                                    3 O
                                                    v) rvj
                                 I

-------
266282
273K1
Pen
Sod
lorocyc
azide
     §ffl  2.

     **  a.
II * 5'
n> Q Q

|










c |
_.
3 -0
=r
o
rji
•3-
fr









a» j
^ —
§Q
^

f
O
•o
a.
o
X
3"
3
X
c
o
                                - 4
p p

b b
» o o o o  o
              oooooo
                                »  o o
                                 SSg
                                    U, u,
                                      p -pp


                                              -
                                                             00  •

                                                             ~ b  »
                                                             «1 g
CAS
ber
3  5

'N.  O

 *
                                                                                     -  s ,
                                                                                     —  40
                                                                                     •o  c


                                                                                   S •£  " c
                                                                                   2    — CO
                                                                                   a. n  TO m
                                                                                  3 —
                                                                          I
                                                                                       * s
                                                                                         yi VI

                                                                                       o    —• -C
                                                                       •D

                                                                       a
     " S      o
                                          0 M
                                                        g
                                                                          3!

                                                                          § b ." S


                                                                         I s ?!

-------

o» D n *5 O •*•«•• o rtt *"• o p o •— Q. C O —> QL Q w> ~n
a s- -.33.- 3 3 = f • 3 f i i 3 3 i' 1 i "•
(B ft) 3 u^ — * O J 3 ^ q to
3 3 3- W < Hi -^ <^ i"" 7 3" ^*

tr> i -n n»
?
o
-1




0000000*000 * o * o o w o »ooooooo»oo g
o o o o o — o ooo o 02 o ooooooo oo "^
•^V/l~ ggo'oS g»o ^ov i
oi
ooo^f* ooo o f*o o *s<^ir\jo UJNJNI now
. . * i.o- o« 2L^r"
OOO O -* O OO 3O*<
g s 2 oo4 jas
TJ m
ftf jf t
-. 10 O
s 1
o o o o -* O4O o 0*0 r\jroc*p -*pp ripao
§*OOUI OO O OO W1W1 3 O «•
oo o**o -» 2K
s |ei

CAS Chemical










JO
o


•£»

#











m
(A
i'
a.
3D
O
I
_
~
3D
n
*-»
i



C

i
3

o
0)

§

5
n

3'


t

0

o
^


n


§'









r—
•5'
a

n
§
V)
I
i'

0
3.
OJ
I'
o
0

a
i~


§


a
n
O
-i
V)

V)
a-
1
y








o
o
3
ift
C
^

j
in
°
f
s

o
o
g.

3
J
*


g
(A


•~
»

C/)


o
(/I
n
o






3 2
--. U3
"vj T
5 ^
^x Z
-O O
o •


C
m
i
=
i
«
D
m
m
X
Z
CD
^
B
S
V)

^



»
„•
01




i









-------
•*'  "H "•<  2<  2* 'y
a  a 5  s  a s
rvj  iv g»  M  ^ rsj


wi  3s ro  fo  ** 3£
•c*  w ftj  o  a o*
                       -* o  -i -i a  f-
                                                                             ?o-
                                                                             L*
                          a  g S
 -<  -.  O  -O  >  >
 f5  -)  —  D  tf)  ^~
 r*  —  c   n  w  —
 -i  n  -**  *-  c •<
 3)  3"  O   O  "J
 n  —  u»   a*  (9

 Z$  5-S
 O  O  O   1
 -i  n  ~  o>
 O  •<  («   N
 n  n   i   o
•<  —  o   —
 nog

 o"  S  g
•333
 ?5  ^  —
 3  a  c

 j  a  §
 a  m

 o  o

 i
                  ' — 73  ft
                         t/>  -n  -n
                                 3  -8


                                    '
        Ol 00  *^
     _  -O v/J  -g

    -O  W CO  O

    **•  -«>J v/i  -O
    v/i  ui oo  «-n


    -n  as n  -a
    ^  a)  •<_  i

    C X  3-  O
    <  rt Q>  n

    ffl  3- —  3-

    —  -i  O  —

    —  o  r*  q


    o
                                        rt>
o  c> c>  c*  c^
-^  & CT'  fr  VJ1
Xs  CD M  rvj  —• ~-"•"«>


!  •  «M  3

°   i" °   1
    S  0}   N
        a  T  a

           —  N




           3"!
       -  g  3


              03
                  c   a.

                  if
^  o o
MOO
t~  ijl I\J

S  S °


-O  VI O
•35  o -.

en  -n -Q


8.  S  3

c  I  "2.
3  n  n

. T5  S
O  ~- o
—  ON

—  x  £.

i"     *
                                    Cn  04


                                    S  g
       U1  VJ1
       •S  v/t
       oa  rvj
       M  -o
       •N  O


_  _  5  *
**  nj  _»  ^g


•n  -i  X  O
~  -I  o  —.

C   —  «  3
-i   O.  a  re
                                                                                                                                                 O  "3

                                                                                                                                                 •O  3

                                                                                                                                                 ~- 10


                                                                                                                                                 S  «
                                                                                                                                  c  a
                                                                                                                                  » •       c
                                                                                                                                  15     —  tfl
                                                                                                                                  Q.  n *3  m

                                                                                                                                  —  S  trt  O
                                                                                                                                  •    -
    »  O  O p  0  0  p


                      2.3
pppppppopoo   » o   »  o o  o o

oooboo
                                                          §§pS     2     2§1°S
                                                                                                                       I  5
                                                                                                                       •^  o
                                                                                    o

                                                                                    o
                                                                                    OJ
                                                                                    o
                                       §
                             §•
                             W1
                      o      ««
                      o
                                                                         •o

                                                                         o

                                                                                                                                  TJ  -8  1


                                                                                                                                  -  -• =  i
                                                                                                                                  o   §  1  *
                                                                                                                                  QJ      O  V)
                                                                                                                                  **   a  **  n
                                                                                                                                  —  3  i»  »
                                                                                                                                      §O.     rn
                                                                                                                                         o  m
                                                                                                                                  *,  S  5  «


                                                                                                                                  "  S  I  <
                                                                                                                                  -.132-
                                                                                                                                  3   —  S   C

                                                                                                                                  »  s  ••   s
                                                                                                                                  (i   —  ^
                                                                                                                                  (Q   O  O   -*»
                                                                                                                                  C   3  VJ   T
                                                                                                                                  —     w   Q
                                                                                                                                  o   -n  ~ '  3

                                                                                                                                  S   S  S  -
                                       o  o  o
                                       W  ^(  U1
                                                                                                               J  "I
                                                                                                                    1/1 °J
                                                                                                                    7^-  1   ^
                                                                                                                   
-------
S
CO
                          CO   _  _..    .


         _   _    -    _-  ,-  no  \ji  ui
"•J  O  vyi  V/l  W1  1^4  U4  •~v|  OJ  o

-*55U?^4COVJ1--^  —
CD  00  CO  09  CO  CO  "S
wi  OJ  r\j  rvj  —•  —•  <3
. _  0  —.   ._....  j*r
    *5  -o  *•*
o  -o  o  w

Crt  Z  f-  03

tf)  VI  3  ~o]

3"  ffi   Q)   ft
ffl  -i   x   3


3           3-
                                      a
                              3  $
                              >-n  M
                              o   -
                                                                                                  33
                                                                                                  -*^

                                                                                                  •O

oooooooooooo
fooorjoorvjrjooo
rvjrxjo -^uiLnwi-^fVO
v^i o wi M wi rvj




^
»







g


°

o n
b 3
2 i
— ,
§
0 n
'- I

_,,
S
















«

I—
•5'
a
o
0
S
u3
&
o
§
(O

&

3
•*^
w
ta

a.
0)
-<

I

ia
}




as
n
•n



30
,,
1
»

»
I

o-


X
o




^
»







m
w


I'
S
JO
o
I
*—
~
T,
n
I
•*^

*^
Used in Application

30

n


3
?
t
21
O


n
^,
S














lonsult IRIS for Update a
lipid. Consumption and
» a
Si
o tt
^ 3


9 "1
— -o
SO
"

S Z
-> *
o •
•n
yi 
-------
                           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)
ChristopherlDeRosa, 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)
JeanetteWiltse.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
 rationalefor 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 describe the 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

 Chemicalsthat 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 characteristicdistinguishessystemicend points
 from carcinogenic and mutagenic end points, which are often treated as nonthreshold processes.

 Systemic effects have traditionally been evaluated in terms of concepts such as  "acceptable daily intake"
 and "margin of safety." The scientific community has identified certain limits on some of these approaches,
 and these limits have been borne out in EPA's experience.Nonetheless,EPA is called upon to apply these
 concepts in making and explaining decisions about the significancefor 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 decisionson the significance
 of exposuresto chemicals. In the process, the Agency has coined new terminology to clarify and distinguish
 between aspects of risk  assessment and risk management. EPA  has tested and  implemented these
 innovations in developing consistent information for several recent regulatory needs, for instance under
 RCRA.

B. Overview

This Appendix consists of four parts in  addition to this introduction.  In Section II, much of the traditional
information on assessing risks of  systemic toxicity is presented, with the focus on the concepts of
 "acceptable daily intake (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 presented 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 "uricertaintyfactor (UF)." A newterm, "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 mentionssome 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 refersto an effect otherthan carcinogenicity ormutagenicity induced by toxic chemical.

**     In this Appendix, the ratio of the NOAEL to the estimated exposure (often referred to as "margin of safety") is referred to
       as the "margin of exposure (MOE)" in order to avoid confusion with the original use of the term "margin of safety" 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) TOXICITY

The Agency's approach to assessingthe risks associated with systemictoxicity is differentfrom 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.'sincethere is essentiallyno level of exposure
for such a chemical that does not pose a small, but finite, probability of generatinga carcinogenicresponse.
In the case of systemictoxicity, organic homeostatic, compensating, and adaptive mechanisms exist that
must be overcome before the toxic end point is manifested. For example, there could be a large number
of cells performing the same or similarfunction 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 individuals within a population.

A. The Traditional Approach

In many cases, risk decisionson systemictoxicity 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 ADI 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 (usuallya lifetime)withoutsufferinga deleteriouseffect.  Often, the ADI 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 multiplesof 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 within the
human population.  For many chemicals, the resultant SF of 100 has been judged to be appropriate. For
other chemicals, with a less complete data base (e.g., those for which only the results of 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-characterizedresponsesin sensitive humans (e.g.,
effect of fluoride on human teeth), an SF as small as 1 might be selected.
        A NOAFLis 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 in
        an experiment.  Further, in cases in which a NOAEL has not been demonstrated experimentally,  the formulation calls for
        use of the "lowest-observed-adverse-effectlevel (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 becomesknown, questionsabouttheselectionof the appropriate "adverseeffecf'arise.

 o  Guidelineshave not been developedto take into account the fact that some studies have used larger
   numbers of animals and, hence, are generally more reliable than other studies.

 These and other "genericissues "are riot susceptibleto immediateresolution, because the data base needed
 is not yet sufficientlydevelopedor analyzed. Therefore.these issuesare beyond the scope of this Appendix.
 However, the Agency has established a work group to consider them.

 2.  Management-relatedlssues

 a. The use of the term "safety factor"

 The term "safetyfactor" suggests, perhaps inadvertently,the notion of absolute safety, i.e., absence of risk.
 While there is a conceptual basis for believing in the existence of a threshold and "absolute safety"
 associated with certain chemicals, in the majority of cases a firm experimental basis for this notion does not
 exist.

 b. The implication that any exposure in excess of the ADI is "unacceptable"and that any exposure less than
   the ADI is "acceptable"or "safe"

 In practice, the ADI is viewed by many as an "acceptable"level of exposure, and, by inference.any exposure
 greater than the ADI is seen as "unacceptable." This strict demarcation between what is "acceptable"and
what is "unacceptable"is contrary to the views of most toxicologists, who typically interpret the ADI as a
 relatively crude estimate of a level of chronic exposure not likely to result in adverse effects to humans. The
ADI is generally viewed as a "soft" estimate, whose bounds of uncertainty can span an order of magnitude.
That is,  within reasonable limits, while exposures somewhat higher than  the ADI are associated with
increased probability of adverse effects, that probability is not a certainty. Similarly, while the ADI is seen as
a level at which the probability of adverse effects is low, the absence of risk to all people cannot be assured
at this level.

c. Possible limitations imposed on risk managementdecisions

Awarenessof the "softness'bf the ADI estimate (see b. above) argues for careful case-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 manageris requiredto considerfactors
otherthan risk (e.g., estimated exposures compared to the ADI) in reaching a decision. The AIDI is only one
factor in a managementdecision and should not prevent the risk manager from weighingthe full range of
factors.

d. Development of different ADIs by different programs

In additionto occasionallyselectingdifferentcritical toxic effects, Agency scientists have reflectedtheirbest
sclentificjudgmentsin 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 incentiveto others
to generate the appropriate data.

Such practices, as employed by a number of scientistsin 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 generallyagree
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 differencesin the managementof the risk. As a result, scientists may be inappropriately
impugned, and/or perfectly justifiable risk managementdecisionsmay 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-responseassessment,exposureassessment,and risk characterization. Having
been apprised by the risk assessor that a potential risk exists, the risk manager answers the question:
"What, if anything, are we going to do about it?"

A.  Hazard Identification

1. Evidence
a. Type of effect

Exposure to a given chemical, depending on the dose employed, may result in a variety of toxic effects.
These may range from gross effects, such as death, to more subtle biochemical, physiologic, or pathologic
changes.The risk assessor considers each of the toxic end points from all studies evaluated in assessing
the risk posed by a chemical, although primary attention usually is given to the effect exhibiting the lowest
NOAEL, often referred to as the critical effect. For chemicals with a limited data base, there may be a need
for more toxicity testing.

*    NAS Risk Assessments the Federal Goverment: Managing the Process (NAS Press, 1983).


                                              E-6

-------
                                         FIGURE C-1
                     Dose-response
                     Assessment (e.g. RfD)
        Hazard
        Identification
                 Exposure
                 Assessment
Risk
Characterization
(e.g. criterion)
                                                Control
                                                Options
Regulatory
Decision
(e.g. RgD, Standard)
                                                                     t
                         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-responseassessmentphase of the risk assessment. These studies are of two types:

(1) Human studies

Human data are often usefulin qualitativelyestablishingthepresenceof 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 toxicity studies serving to complementthem.

In epidemiologicstudies, confounding factors  that are recognized can be controlled and measured, within
limits. Case reports and acute exposures resulting in severeeffects 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 tie 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,
 considerationofstaicture-activityrelationshipsbetweenthechemicalundertest 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 developingmorereliablein vitro tests to minimizetheneedforlive-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
 relevantto any other route of exposure, unless convincing evidence exists to  the contrary Consideration is
 always givento potentialdifferencesin absorption or metabolismresultingfrom differentroutes of exposure,
 and wheneverappropriate data (e.g., comparative metabolismstudies) are available.the quantitative impacts
 of these differences on the risk assessment are fully delineated.

 e. Length of exposure

The Agency is concerned about the potential toxic effects in humans associated with all possible exposures
to chemicals.  The magnitude, frequency, and duration of exposure may vary considerably in different
situations. Animal studiesare conducted using a variety of exposure durations (e.g., acute, subchronic, and
 chronic) and schedules(e.g.,single,intermittent,orcontinuousdosing). Information from all of these studies
 is useful in the hazard identification phase of risk assessment.  For example, overt neurological problems
 identifledin 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
•*posures 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 givento many other factors, includingchemical
characterization of the compound(s) under study, the type of test species, similarities and differences
betweenthetestspeciesand humans (e.g., chemical absorption and metabolism),the numberof individuals
in the study groups, the numberof study groups, the spacing and choice of  dose levels tested, the types
of observations and methods of analysis, the nature of pathologic changes, the alteration in metabolic
responses, the sex and age of test animals, and the route and duration of exposure.

2. Welght-of-EvIdence Determination

As the culmination of the hazard identification step, a discussion of the weight-of-evidencesummarizesthe
highlightsof the information gleaned from the entire range of principal and supporting studies.  Emphasis
in the analysis is givento 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
evidencethatthe 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
                                                                                            in
 Empirical observation generally reveals that as the dosage of a toxicant is increased.the toxic response (i. i
 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 individualsin 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^ dose-responserelationship,certaindifficultiesarise. 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 fundamentalquestionof administereddose 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 governinginformationfor quantitative risk assessments^ince available
 human data are generally insufficientfor 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 concurrentexposuresto othertoxicants have been minimized. In evaluatinganimal
 data, a series of professional judgmentsare made that involve, among others, consideration of the scientific
 quality of the studies. Presented with data from several animal studies, the risk assessor first seeks to
 identify the animal model that is most relevantto 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 undesr investigation.

 In any event,  the use of data from carefully controlled studies of genetically homogeneous animals
 inescapably confronts the risk assessorwith the problems of extrapolaing between species and the need
to account for human heterogeneityaind concurrenthumanexposuresto otherchemicals.which may modify
the human risk.

While there is usually a lack of well-controlled cohort studies that investigate non-cancer end points and
human exposure to chemicals of interest, in some cases human data may be selected as the Tcritical data
 (e.g., in  cases of cholinesteraseinhibition). Risk assessments based on human data have the advantage
of avoiding the problems inherentin 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 environmental situations to which the general
population is more likelyto be exposed.  In some cases, a well-designedand well-conductedepidemiologic


                                             E-9

-------
study that shows no association between known exposures and toxicity can be used to directly project an
RfO (as has been done in the case of fluoride).

b. Critical data

In the simplestterms, an experimental exposure level is selectedfrom the critical study that representsthe
highestleveltestedin which "no adverse effect"was demonstrated. This "no-observed-adverse-effecievel"
(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 systemictoxicants. 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 statisticallyor biologically significant increase in frequency or severityof an adverse effect
between individualsin 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 exposurestudies). In general, NOAELs for these effects
will differ. The critical end point used in the dose-responseassessmentis 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 (RfID)" and  "uncertainty factor (UF)"  is recommended.  The RfID 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 RfD 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 exposureto the human population (includingsensitivesubgroups)
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 "safetyfactor" see  Page  A-4). This new name  is more
       descriptive (n that thesefactors represent scientific uncertainties, and avoids the risk management connotation of "safety."
       The "modifying factor" can range from greater than zero to 10, and reflects qualitative professional judgements regarding
       scientific uncertainties not covered under the standard UF, such as the completeness of the overall data base and the
       number of animals in the study.

                                              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
 exposure to 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.  [1 OS]

 Use an additional 10-fold factor when deriving a RfD from a LOAEL, instead of a NOAEL This factor  is
 intendedto account for  the uncertainty in extrapolating from LOAELs to NOAELs. [1OL]

 Modifying Factor (MF)

 Use professionaljudgmentto determineanotheruncertaintyfactor (MF) which is greaterthan zero and less
 than or equal to  10. The  magnitude of the MF depends upon the professional assessment of scientific
 uncertaintiesof 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 MFisI.

 SOURCE:     Adapted from Dourson, M.L; and Stara, J.F. (1983)
              Regulatary Toxicology and Pharmacology 3:224-238.


 (This is a consequenceof the inability of eitherthe 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 "coversheef'of a few pages in length. The cover sheet includes a statementon 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 vulnerableto 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 referred to 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 presentto 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 influencethe 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 underlyingthe 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 greaterthan 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 managermust considera number of risk factors, nonrisk factors, and regulatory options that influence
the final judgment  It is generally useful to the risk manager to have information regarding the contribution
to the RfD from various environmental media. Such information can provide insights that are helpful in
choosing among available  control options.  However, in cases in which site-specific criteria are being
considered.localexposuresthrough various media can often be determinedmore accurately than exposure
estimates based upon generic approaches.  In such cases, the exposure assessor's role is particularly
Important. For instance, at a givensite, 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

-------
 RfDs should be apportioned by route of exposure. Where specific exposure analysis can be made, such
 apportionment is readily performed. If exposure information is not available, assumptions must be made
 concerningtherelativecontributionsfrom differentroutes 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
 necessarily identical-riskmanagementdecisions must be made on a case-by-case basis. For example, the
 Clean Water Act calls for decisions with "an ample margin of safety";the Federal Insecticide, Fungicide and
 RodenticideAct  (FIF~)  calls for "an ample margin of safety," taking benefits into account; and the Safe
 Drinking Water Act (SDWA) calls for standards that protect the public 'to the extent feasible."Consequently,
 it is entirely possible and appropriate that a chemical with a specific RfD may be regulated under different
 statutes and situations through the use of different "regulatory doses (RgDs)".

 Expressed in general terms, after carefully considering the various risk and nonrisk factors, regulatory
 options, and statutory 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)= NOAEL/MOE(i)           (4)

 Note that, for the same chemical (with a single RfD), the risk manager(s) can develop different regulatory
 doses for different situations that may involve different exposures, available control options, alternative
 chemicals, benefits, and statutory mandates. Also note that comparing the RfD to a particular  RgD(i)  is
 equivalent to  comparing the MOE(i) with the UF x  MF:

                           RfD/RgD(i) =  MOE(i)/UF x MF           (5)

 In assessingthe significanceof a case in which the RgD is  greater (or less) than the RfD, the risk  manager
 should carefully considerthe 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 insightfrom the risk assessor.  In any
 event, the risk manager has the responsibility to clearly articulate the reasoning leading to the final RgD
 decision.

 V. OTHER DIRECTIONS

While the Agency is in the process of systematizingthe approach outlined in this Appendix, risk assessment
 research for systemic toxicity is also being conducted along entirelyseparate 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-responseinformation is combined with exposure analysis modeling
to generatepopulation risk estimatesfor alternativestandards. 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 initiateda series of studies that should lead to future improvements in risk estimation. First, they are
 investigatingthe use of extrapolation modelsas 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. Whilethese procedures are promising, they cannot be expected at this time
to serve as a foundation of a generalized health risk assessmentfor system ictoxicity 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 indistinguishablefrom 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 UF x MF =  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= 10Hx10Ax 105x101=  10,000

   MF= 0.8

   Therefore UF x MF =  8,000, so that

   RfD =  LOAEL7(UF x MF) = 25 (rng/kg-day)/8000 = .003 mg/kg-day)

2. Risk Characterization Considerations

Suppose the estimated exposure dose (BED) for humans exposed to the chemical underthe 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 EED exceeds the RfD (and the MOE is less than the UF x MF), the risk manager will need to
look carefully at the data set, the  assumptions for both the RfD 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)
Diane D. Beat, Ph.D. (OPTS)
Judith Bellin, Ph.D. (ORD)
Chao W. Chen, Ph.D. (ORD)*
lla Coty, Ph.D. (ORD)
Theodore Farber, Ph.D. (ORD)
William Farland, Ph.D. (ORD)
Herman J. Gibb, B.D.,M.P.H.(ORD)
Richard Hill, M.D., Ph.D. (OPTS)

* Co-Chair
Elizabeth H. Margosches, Ph.D (OPTS)
Robert E. McGaughy, Ph.D. (ORD)
Stephen Nesnow, Ph.D. (ORD)
Peter Preuss, Ph.D. (ORD)*
John A. Quest, Ph.D.(OPPE)
RevaRubenstein.Ph.D. (OSWER)
Rita Schoeny, Ph.D. (ORD)*
Dot G. Wellington, Ph.D. (OPPE)
                                        E-16

-------
 I. INTRODUCTION

 In the analysis of data regarding the potential human carcihogenicity of chemicals, the Agency uses the
 approach described in the document entitled Guidelinesfor Carcinogen Risk Assessment (51 FR 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 assessrnentdevelopedby the Office of Science and Technology Policy (OSTP)
 (50 FR 10372-10442),and the EPA guidelinesquotethe 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 elementsof the risk assessment
 process are defined as follows:

 (1)    Hazard identification,in which a determinationis made of whetherhuman 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 designed to  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"^ 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 evidencebearing on this question are: (1)
human studiesof the association between cancer incidenceand exposure; and (2)' long-term animal studies
undercontrolled laboratory conditions. Other evidencedsuch as short-term tests for genotoxicity, metabolic
and pharmacokineticproperties,lexicological effects otherthan cancer, structure-activityrelationships.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 relevantto carcinogenicity.by judgingthe quality of the studiesavailable, 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 B2      Probable Human Carcinogen

        B1          indicates that limited human data are available

        B2          indicates sufficient evidence in animals and inadequateor no evidencein humans

        C           Possible Human Carcinogen

        D           Not Classifiable as to human carcinogenicity

        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 carcinogenicity described
earlier, using judgments about the degree of adjustment warranted in each case. For further description of
the role of supporting evidence, see the EPA Guidelines.

B. Dose-Response

The purpose of  the dose-responseassessmentis 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-responseassessmentis  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
estimationactivity is not dependenton the likelihoodof human carcinogenicity.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 extendingeitherthe 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 4 Lyon,
       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 specieswhich responds biologically most like humans (with respectto 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 CombiningNeoplasmsfor Evaluation of Rodent CarcinogenesisStudies
        (1986). McConnell, E.E., Solleveld, H.A., Swenberg, J.A., 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 reasonablefit 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 suggestthat 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 generally not 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 incidence in 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 essentiallyequal.

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 sensitiveto 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 raisedto 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  generallyconsideredto be the equivalentd~ betweenspecies 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 betweenspecies.

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 expressedon a weight basis (e.g., milligramsof 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 determiningthe 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 environmental agents, 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 corn) 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-derivedvalue 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 10r2.

 5. Statement of Confidence in Dose-Response Parameters

 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 sheetsas a high, medium.or low rating. The factors increasingthe Agency's
 confidence in the accuracy of these risk bounds includesthe following:

 (1)    The existence of 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 pharmacokinetics of the
       agent.

 (9)    For human occupational studies, determinationof exposure or different worksites as opposed to an
       average exposure for the entire workplace.

 (10)    For epidemiologicstudies, exposure measurementsconcurrentto 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 usefulnessin 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 epidomiologicstudiesfor 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'nateit.

(2)     The use of vehicles, such as corn oil, which may confound or interact wish the agent understudy
       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 enoughto human systemsto 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 (i/x)                      (i)

 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  /S {!/(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 v«ry 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 (nj /Xj)                 (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 fj. 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  a.

      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(M  + a2/2)                (5)

It  followgfithat:

                   =  exp(-/i + <7*/2)            (6)
Since by definition, cim[l/x]={S  (i/x)}/N, and hm[x] = N/Z  (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(-ji  + a2/ 2)
                        i - (72/2)                    (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 (a2/2) =  (exp  /i)2/{exp  M - exp(a2/2)}

                = (exp M)2  /  exp(/i + cr2/2)         (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) - l}* ,  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^jjistributions with very  large N, their results  can
ordinarily* b« 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

-------
2S5(Is£

H H it n H m
                                  n

                               — 0»
                                  M
                                  IB
03335  53355  32S2S  s55*«™"
o«0.*»  w*uiS2  3S0S?  m5SS2S5
                                        W)
DISC
MEAN
FLIN
       X
       n
    W «J O
                           «j o- w
RG
                                                                                   o   a


                                                                                   -   S
e M

II M
             gjOw»jjM*rs»in«o
                                oS
                                  M
                                  w
                          • •  •  • a
                          i ui •- in •« o
              WWWMUiNMNMMM
                      ' r* M ui * •> a m
Ul
— — 0 —

— * Ul Ul
Ul » At W —
^ « 0 * <•
» f» Ul Ul Ul
ui » — o e
0 * * 0 M
O W M M 0




-- -u,




M » M «• Ul
* * •«
•" * »
»
M
Ul Ul



J> M M O> M
•  5>

w - - - -

--*"
M M n>

M

_ _
Ul « W » Ul
M M *4 M Ul
• K» « Ul Ul
* * 5 M *
Ul •> T0 ~ »
N M
*» M
•o «•
— * w w



$M W l» —
Ul K> » Ul
fS» IV» X- — —

•• * Ul K» «•



i— ». ••
...» tai
•• » M •.
^^ ^
.. ..

•• Ul •> Ul M
» M Ul
«• * *
•• M —
M
* 0 .
•" N
>
fvy n
— -0 — N
^1
Ul * O
Ul «g Ul O> ui 5
» ui N m
—  U)
N
» — Ul Ul Ul
Ul N Ul "
IU
M
* S
w * e!
MP n>
tal
n
i
H4
n
^!
rn
m
^
^
m
«
CD
m
n
|
5
I
o ui

< S
§ i
^i a
i
•< X1
rn *
a 5
e
ni o
9 °
k4
S
Ul
fn
•o
m
i
m
a
w




m
3
T;
0

o
— H
o
c
cu

ZJ
a/
*
O
c
c


                                                 F-6

-------
o
t*
e

N
3 *
s
m
5 S
*
^
fi 5
ft*
6 M
0 •
5
S S
IS
S
f s
u
m M
fi S
s
s s
8
m
s *
s s
a
m
O
s







e
e
e
•


« '
*
•
*
• *
'«













DURA) ION PIOT OF DA
STAIION ID: 0414900
10000
• ~?
^
O
>
>
T*
8
11 ^i .
^ *
55 ' m
S 25
B Z
5 z5
^ m 5
Ss
Ul
•* m
e -a
Ul ~t
•* m
m 3
S ffl
• a
x SS
i —
i '1
! "
M ^
? $
n ••
'
5 i

i
|
«
m
»
a
>
. it
9
O
e
S
3
r*
F-7

-------
                                                            " 3 O

                                                            iis
2
                                                            r§

                                 	                  5 S  n


                                 Issslslslliiisliiii  *i a  5
                                                         m ^ i*  "*


                                                         B* >  g
                                                         Mi 55


                                                         S »    o


                                                         y    §






                                 §§2SSSSSg°SSSggSS2S  »|

                                 -j-o - ~ 3 5 S 3 * ; B S;; „,». M _ „, r   s
                                                          n






                                                                  UI
                                                                  •^
                                                                  »
                                                                  •M



                                                               s  i




                                                               r*  m
                                                                  XI




                                                               I  I
                                                               w»  ^
                                                                  o
                                                                  o
                                                                  e









                               *-*******wt—^^^"»^»—^^^^^B—»—^  MZ



                                                        m

                                                        55


                                                        a5
                                                        M <




                               22S £ £ Si 5* S *"«• •«! *tn-~»*  35

                               2i5^S.5S»iSSSSS.S?:s  S5


                               issssgssgsssssssssss  II

                                •"  ••      — •• —     »IU      *"
                               • WvlvN«M|M>.l||4g|9,9>2«o»  WZ
                                                        o S
                                 F-8

-------
n  n  us  3

§  8  s  1
5  5  8  z
              " " " " II 1 ....... II II II II II II II
z  z  o  5

f  :  <  8
fi  e  M
~  ^  >  n

*  S  8  M
m  a  Z  .
                                             c ffl
                                             3 a

                                             ss
                                             m
                                             e <

                                             si
                                      [25     i  2iJS
                                       2 o     3  u> » S
                                      I 3 ~     *  "-!-,»
2 - * £ S 2 S » £ » - M M M "*" ~ ~    *=    ^^S  u"*  Z?S
? • ? ^ * r4 * • r»« * * S « 5 5 S w 3  ™JS    3S~  S2>  c*w
eeeeeoeeeeeeeeeeooe    n    «§2  !§Sc  5 S 3
                                 ••       2Skn
              H II II II U II II II II II || || „ |, tl „ „ „ „
      I

      i
      5

      X
                                     2*2  •fa
                                i    ssi  S2™
                                3    ° •• S  •« n 2
                                I    ,25  8?2  •«  S
                                *    a-*»  ^r*  u>  »
                                              S
                                                               -<  rt

                                                               T)  *
                                                          ^ »  ui  »
                                                         > •< -n  m  5
                                                         w  p  n  S
                                             ^> i-  o n n
                                             •» »  Z r- -i
                                             -" -H  B M m
                                                   S   >  ui •< m   z a
                                                   zss  sis0   5s
                                                   m   ..   2  2   a iS
                                                    <
                                                    >
                                              C    O M    S ^ C

                                              («     •"    a n o
                                                   N a    S e z
                                              M    mo    a ?

                                                   §S    352
                                              «>     "8    >  5

                                                   5W    -""IS
                                                   < n
                                                   m ui
                                                   a


                                                   il
                                                                O
                                                                U)
                                                        8 >-
                                                        *3     2
                                                          w     o
                                                    ut
                                                          CD
                     F-9

-------

-------
      APPENDIX G




SAMPLE PERMIT LANGUAGE

-------
                            Appendix G

                      Sample Permit Language

A.   Bioconcentration Evaluation Requirement for Characte^izinq
     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

-------
                       a*e necessary effluent limitations for the
                       bioconcentratable pollutants.
                            FACT SHEET

A.
       Effluent  Limitations  on Bioconcentratable Compounds
           «n b™nCO?Ce^ratable P°llutant  shall  be discharged in
           an amount which would  cause  exceedance  in the receiving
           water of its reference ambient  concentration    eceivln, £ /X~XX~XX?} '  for bioconcentratable compounds
           which have been. identified in the effluent by the
                               luation procedure °f
           Where final calculated limitations for individual
           pollutants fall below the current level of detection
           Si limtt+.?°rnthe Pollutant will be the one calculated •'
           SnL??a W^aj nethod to be used for monitoring will b4
           SS?SiJ ^ln  ^ Per*it; and the minimum level for that
           method (if one is available)  will be specified in the
                           Where  a minimum  level  is  not  available
            resolorc   f?*1^*1 method, an alSernaSve   ^
          will be snf ^ComPllance/n°n-?omPliance determinations
                          d
                                                 the

B.   Effluent Limitations Compliance Monitoring Requirements
detection and quantisation shall be reported.
                               G-3

-------
                      Sample Section 308 Letter

  CERTIFIED MAIL NO.
  RETURN RECEIPT REQUESTED

  Ms.  Ann  Powell
  Plant  Manager
  Chemico Corporation
  Any town,  USA

                                           RE:  NPDES No.  xx000123
  Dear Ms.  Powell,                               Chemico Any town
      The U.S. Environmental Protection Agency

             hS f DJ? Permit f°r the Chemico A         c
             evaluations conducted by EPA and water quality
              nCt^ by the (State Water Control Authority)
                   discharges from the Chemico Anytown facility
                  State «ater. quality standards.  in order tor EPA
 ii   r  ,*,    responsibxlities under the Clean Water act, 33
 o^ ;£l J?  £ et Seg' '  additional information regarding the nature
 of the discharges from the Chemico facility is required.

      Therefore,  you are hereby required (1)  to perform the
              analvsis  Programs described below, (2)  to maintain
              US?-bY EPA a11 records regarding plant operation's
             a*pllng and analysis programs within the time liSts
  ,,          ^ffe^^^ements are imposed pursuant to the
 authority provided in  Section  308 of the Clean Water act,  33
 ?eSu?; ?n S5      \  Failure  to comply with this request may
 Waler aS enf°TrTcem^nt  Proceedings under Section 309  of the Clean
 ?»S?  •  ?  •     V:?'c- Section 1319,  which could result in  the
 judicial  imposition of civil or criminal penalties.

 Effluent  Bioconcentration  Evaluation -
              fcu  identify Potentially bioconcentratable
         S«  S  arged  fr°m Outfa11  °01' Chemico shall analyze
         Charges as described below.  Wastewater samples  for
         ^      composited over 24 hours once per quarter  for a
              ^^ conmencing witnin 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
a£2* ™,   °f Bloconcentratable 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
   nal  ^""nation on  chemicals reporteS  as prelen?
     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 analyles to the pJrmiJtini
authority within 45 days of  sample, collection.        permitting

     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 in Special Conditions

 A,  Fish Tissue and Sediment Evaluations




 water quality standards,  and to determine if pollStants
 accumulate in fish tissue or sediments to levels whiSh would
            the discharge related accumulation of organic
                                      s-2^ ssnks
 permit.  Following review of the results from the annual
shall mclude ten individuals of the bSnthic organi
                                                          of
               sp«
                                           sa.-
                                            in the
                             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.  Monitorina 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.  Russman
                     U.S.E.P.A./E.R.L.-D.
                     6201 Congdon  Blvd.
                     Duluth,  Minnesota  55804
                     (218J-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,  bioaccumulation, 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
                     compounds.
            it:  No charges are applied for Federal, State or
         Local governments.

         Release Date:  January 1991
                              H-2

-------
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,
     19 o / *

     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.

1.   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 to'xic 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 coJlection
          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 bioaccuiaulation 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'   ?^eT7Pfocedures for Calculating RACs in the Absence of
     IKJ.S 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 NRios.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


                                    Effluent'
1.0  General Considerations

     In  this  appendix, ..example  data  and  discussion  on  data
evaluation  are  provided for  the tissue,  effluent,  and  sedimJn?
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  dement  a
                                                  s      ment   are
 illustrative   Reporting formats for laboratories and SSSSltaSS '
 performing these procedures may differ.               ^wnsuitants

 2.0  Tissue Alternative
 firi n              reP°rt'.the QVQC report should be examined
 first since this information will provide a good indication of the
 quality of the  analytical data.   The QC requirement for recover?
 of  the  surrogates  is 25%  <  %  recovery  <  120%  and  all  three
 S£°2?   A dio-biP^yl ' 13C6-1 ,2,4, 5-tetrachlorobenzene  and *C -
 hexachlorobenzene, should have recoveries within  this  ranae   if
 recovery data is out of this  range, the analytical data should not
 be used.  Provided acceptable  recoveries are obtained   the
 chromatograms and the.  QA/QC  data for GPC resolution   silic
                                           examinatiSn should
                                                    of the overall
      In Table J-l, an illustrative report for the tissue analvtical
 procedure  is provided,.   After examining the recover^laJa for Se
 SSn^JnS  ^emica1^' the GC/MS  chromatograms for the  procedural
 blSk should6  cSo^P ? e? raCt t°Uld be examine<*.  The  procedural
 tolank should  contain few  peaks  as  illustrated  in  the  reoort
 Procedural blanks should always contain fewer peaks than the S££le-
 extract.  Peaks in the GC/MS chromatograms should be narrow "5
 10 to 20 seconds base widths,  and when large fat peaks  sSSt' -I' a"
 base widths exceeding 1 minute, overloading of th5 cMilSJy'cSiuiA
 on  the  GC/MS  has  occurred.    Efforts  should  be   directed a?
 determining why these large  peaks exist in  the staple  Strac?
 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
        o^^rV^3'  additional  cle^uP or -extraction of  anotSr
        of the tissue composite  should be requested.
     The evaluation of the QC data for GPC resolution, silica ael
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-l

-------
      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 d1?-chrysene, baseline resolution  between  the  /3-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  o£  questionable  quality  and use  of the  data  is   not
 recommended.   Depending upon which factors  are  out  of range,
 reanalysia 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 1, 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-2

-------
  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.                                           * J-ue".l-J-t J-ea_


       For Tissue Report  2, GC/MS  components  tentatively identified
 Thing ^?McEPA/NIH/NBS  ^SS  sPectral l^rary 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  tentative!?
 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
 identi                     3'   GC/MS  comP°nents  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 eaS
                                      identified,   the   following
      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
          'JXtunber.
      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-l ,2,4, 5-tetrachlorobenzene,  and 13c -
hexachlorobenzene, should  have  recoveries  within this range    in
addition,  HPLC fraction 1 should contain d10-biphenyl and should not

                               J-3

-------
 contain   C6-l,2,4,5-tetrachlorobenzene  or 13C6-hexachlorobenzene.
 HPLC fraction 2 should contain  C6-l,2,4,5-tetrachlorobenzene and
 should not  contain d10-biphenyl  or  13CR-hexachlorobenzene.   HPLC
 fraction  3  should contain   C6-hexachlorobenzene and  should  not
 contain  C8-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,   1,3-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 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-4

-------
  sfnn^on *W° comP.ounfs-.  Note' wh*n using the first solution, fresh
  solution is required since endrin ketone degrades.
  C)  MS™£fitivityi   For the °'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)  MSo?att£r24i°»»V,  f°h  ^-chrysene, the ratio of the abundance
       and <25%   mass/char9e relative to that of 240 should be >15%



  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-i.


  f)  Precision:  A duplicate analysis  should be  performed with
       every 10 sample analyses.  This data can be presented  by using

       SxSeed 150?    °r    Ular reP°rts'   Precision data should
   _   If any  of  the QC/QA procedures are  not met,  the analytical
 data is of  questionable quality and its  use is not recommended

 SlES^rSctS ^ich,fa?to«  -e  «* of range,  re^naly'sTs JfSe
 ««? fractions  (b,c,d,e) or extraction  of another  portion of the


            •     <*• re                          °
                      >                  tha da^ in Reports i, 2,
 3,  an  4


      For Effluent Report l, GC/MS components tentatively identified

 r?ml thS CHC maSS  sPectral  libraryP search are reported    Those
 GC/MS components  with fits  of 70%  and  greater  are  considered
 tentatively  identified.   For  each  tentatively idStified  cS/Ss
 component, the  following  information  is require: iaentiriecl  GC/MS


      1)   GC  retention time of the component,  minutes
      2)   Amount of  the component, ng/1
      3)   HPLC  fraction number of  the  component

      4)   ^i1?* of t^best mass spectral library  identifications
          (up to a total  of ten identifications)
      5)   For each tentative  identification:
               Library searching fit,  %
               CAS number
*
componen?s!eP°rt ^^ C°ntains none or few tentatively  identified


     For Effluent Report 2, GC/MS components tentatively  identified
T^ng ^eMCEPA/NIH/NBS  mass spectral library search T are rlporSS
Those GC/MS components with fits of 70% and greater  are  conSTdSJ^d

SIS^V?!17 ^dentified- T^is reP°rt contains5 only ?hose  exponents
with effluent  concentrations  of  100  ng/1 and areater anrt «?^
predicted tissue concentrations of i ug/9£ and glea?ir! For each
                               J-5

-------
 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/kg.   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)    Jk  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-6

-------
  hJJ?fhlorobenzene, should have  recoveries  within this
                                                                 m
                                                                no?
         *°onco
  should not  contain d10-biphenyl  <5r l3c,-hexaohlorobenlene    HPLC
  contain" V3 -fffs l^l , ' V^chlorobenzene and" should "no?
  conrain  C6-1, 2 , 4, 5-tetrachlorobenzene or d10-biphenyl.

       If recovery data  is out of range and/or the surroaates are in
         d^r^V^i011'   ^  anal^tical  ^ta  should  not be  Jled?
       ded acceptable recoveries and. surrogate location in the HPLC
  ?n?CHPT°?S are. gained, the GC/MS chromatograms arid the SA/QC data
  for HPLC resolution, GC resolution, etc. should be examined   This
  examination  should take 5 to  15  minutes and will provide  a gSol
  indication of  the overall quality of the analysis.            9
      In  Table  J-3 ,  an  illustrative  report  for  the  eu^n
 X?ayf £a Jh Pr°cedure, 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   Therl
 Xji b? slxhGC/MS chromatograms, three for the procedural blank an*
 three for the effluent sample as a result of the HPLC  fractionatiSn
 procedure   The GC/MS chromatograms  for the HPLC fractions rSrtSe
 procedural blank should contain few peaks  as illustrated in  thl
 SS°«' 1Proc
-------
 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 d^-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 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, 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)   Eor 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-8

-------
      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
i** +.***  Sediment  Report 3,  GC/MS  components  not tentatively
identified, i e. ,  flts  <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 eSch
GC/MS  component  not  tentatively   identified,   the  following
information is require:                                 j-^ixuwj.ng


     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
          *i rf^.  ldentifications   and   for   each   tentative
          identification,  their library searching fit, %,  and CAS
          number .
                                       each   tentative
            on,   er    rary searchin

        fit  arameter  i   <
unknown .
     5)    If the  fit  parameter is  <25%,  compound is  labeled as
                              J-9

-------
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


                                             67%
      ,
      Cg-1, 2 , 4 , 5-tetrachlorobenzene     50%
      C6-hexachloroben z ene                   4 6 %
                              J-10

-------
 GC/MS Chromatogram
 Procedural Blank
 .290655
  citfet?;
 "ffilMto
GC/MS  Chromatogram
Sample Extract
                           vain./
                                          413
 20OOOO

-------
GPC Resolution Check
phthalate
               O
               rd
               Q

    	1.1	.__.[.I _:	l\

1
I
\ 1
\ r
\ i
:- v / .
r~~r
\ . /
\ * i
. ^ .'. ^
\ "
\
\ •
T 	 	
"\
        C 3
                         -o
                          3

-------
2/8/89


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        g'o

 Cholesterol      9
 Dieldrin        97

Cholesterol     4
Dieldrin       92
15:85


15:85


15:65


15:85

-------
  SJ
  »-™
 u
 CJ

CO
2
CJ
o

-------
                  GC/MS Sensitivity Check
TOTQL
         CHROnflTOGRQn
     >fi7708  45.0-550.0

                400
                                    URROSflTES44,593-61-''
 7000<
 6000(
 5000C
 4000(
3000C
2000C
1000
Data
Name:
            >A7708::DO
           PPM SURROGATES
Quant Output File: AA7708::
DO

-------
GC/MS Calibration  Check
               SC*(1 1787 (31.601 min) of 90c/30501005.d SCALED

-------
 Precision Data
 Sample ID
89C3015
89C4115
89C4185
Date of
Analysis
               2/8/89
               3/1/89
3/25/89
                          Surrogate
                          Chemical
  Surrogate  Precision
Concentration

    ng/1
•• Biphenyl
TCB
HCB
Biphenyl
TCB
HCB
Biphenyl
TCB
HCB
	 —
85
67
47
68
50
50
53
46
40
— _
56
28
34
63
46
35
41
52
34
4 I
~ J.
82
32
1 1
X J.
35
25
12
25

-------
 DFTPP Performance Check
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
Measured


    6%"
    0%
    30%
    0%

    5%
    17%
    2%
    16%
    54%
    21%
120000
100000
80000-
60000-
40000-

20000-
o-i
1



]
69
110
51
I

VZ -> 50 100
                 Scan 1264  (22.753 mm) :  0101001.D
              127
                                                         442

-------
  Table J-l continued:
                          Example Reports 1, 2, and 3 for the Tissue
                          Analytical  Procedures
 Tissue  Report  l:     CHC   Mass .   Spectral   Library    Tentative
 __ . _  Identifications
Peak RT
 12.602    31
                      Fit   Tentative  Identification (CAS /)
                     99   Dieldrin  (60-57-1)
 Tissue Report 2:
 Peak RT   Amount
 (minutes)  (ug/kg)

 11.748    39
 14.889    32


 19.756    25
                    EPA/NIH/NBS Mass  Spectral  Library Tentative
                    Identifications
                         --                        	

                    Fit  Tentative Identification (CAS #)
                    90
                    87

                    95
                    95
                    94

                    76
                   •    — r»^/w/*/*
      (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, l,l'-(l,2-ethynediyl)bis-  (501-

      Anthracene  (120-12-7)
Tissue Report  3:
Peak RT   Amount
(minutes) (ug/kg)
 8.621
         60
 8.680    30
19.756    25
                   GC/MS Components not Identified by EPA/NIH/NBS
                   Mass Spectral Library Search
                   Fit  Tentative Identification  (CAS #)
64
64

64
42
                        2-Pentene, 4,4-dimethyl-,  (E)- (690-08-4)
                        Pyridine, 2,3,4,5-tetrahydro- (505-18-0)

                        Benzene, 1,2,3-trichloro- (87-61-6)
                        Benzene, 1,2,4-trichloro- (120-82-1)

                        Unknown
                              J-19

-------
Table  J-2.   Example  Data  for  the  Effluent Analytical Procedures



Effluent QA/QC Report	

Recoveries                              HPLC Fraction Number
                                              123
d10-biphenyl
Cg-l, 2 , 4 , 5-tetrachlorobenzene
C6-hexachlorobenz ene

nd

67%
50%
nd
nda
.nd
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-20

-------
Detector respons
       "      4,0
    , rtv'' 7
100
                                    0

                                    TJ
                                    0

                                    '.1-

-------
   Table  E-2  continued:
  Effluent Report  l:
                     CHC Mass Spectral Library Tentative
                     Identifications
                     *—^—^^^^^^—- -
  Peak RT   Amount
            (ng/1)
  12.602
           31
HPLC Fit  Tentative Identification  (CAS
       .—.


1    99   Dieldrin (60-57-1)
  Effluent
  Peak RT
  (minutes)
          Report 2:   EPA/NIH/NBS Mass Spectral Library Tentative
          	Identifications
          Amount   HPLC
           (ng/1)   Fr #
     >  1   Fit   Tentative id.  (CAS =?)
     ppb    %                        '
 19.756
          250
 Dilution = 25%
                          Yes  95
                               95

                               94

                               76
                                     1,3-Butadiene,  1,1,2,3,4,4-
                                     hexachloro- (87-68-3)

                           Yes   90    l,l'-biphenyl  (92-52-4)
                                87    Naphthalene, 2-ethenyl-  (3-?-
                                     54-3)
                                   Phenanthrene  (85-01-8)
                                   9H-Fluorene,  9-methylene
                                   (2523-37-7)
                                   Benzene, l,l'-(i,2-
                                   ethynediyl)bis-   (501-65-5'
                                   Anthracene  (120-12-7)
 Effluent Report 3:
 8.621     602
8.680    30T
19.756    250
                      GC/MS Components not Identified by
                      EPA/NIH/NBS Mass Spectral Library Search
                                         ^ntification
                      1
                      1
                          64

                          64
                         64
                         42
         2-Pentene,  4,4-dimethyl-,  (E)-
         (690-08-4)
         Pyridine,  2,3,4,5-tetrahydro-  (505-


         Benzene,  1,2,3-trichloro-  (87-61-6;
         Benzene,  1,2,4-trichloro-  (120-32-


         Unknown

-------
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 < i ug/kg

Dilution = 25%
                             J-2 3

-------
 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.

      djio-kipk^Y1                             67%   nda   nd
       C6-l,2,4,5-tetrachlorobenzene       nd   50%   nd
	C6-hexachlorobenzene	      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-i 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-24

-------
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,  (^.980).

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 Meassure 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  Coeficient.,  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  Aqueitic 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.D.,  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,  Steite 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

-------

-------

-------