FINAL REPORT
TOXIC POINT SOURCE ASSESSMENT OF INDUSTRIAL
DISCHARGES TO THE CHESAPEAKE BAY BASIN. PHASE
PROTOCOL VERIFICATION STUDY
VOLUME I
AND APPENDIX A
Contract 68-02-3161
August 1982
MONSANTO RESEARCH CORPORATION
A SUBSIDIARY OF MONSANTO COMPANY
DAYTON
LABORATORY
DAYTON, OHIO 454OT
-------
903R811O4
TOXIC POINT SOURCE ASSESSMENT OF INDUSTRIAL DISCHARGES TO THE
CHESAPEAKE BAY BASIN. PHASE III: PROTOCOL VERIFICATION STUDY
VOLUME I
AND APPENDIX A
by
S. C. Wilson
B. M. Hughes
G. D. Rawlings
Monsanto Research Corporation
Dayton, Ohio 45418
Contract 68-02-3161
August 1982
Project Officer
Mark Alderson
Chesapeake Bay Program
U.S. Environmental Protection Agency
Annapolis, Maryland 21401
CHESAPEAKE BAY PROGRAM OFFICE
TOXICS PROGRAM
U.S. ENVIRONMENTAL PROTECTION AGENCY
ANNAPOLIS, MARYLAND 21401
-------
ABSTRACT
The Monsanto Research Corporation conducted a three phased pro-
gram in the Chesapeake Bay region to identify toxic substances
entering the system. The overall objectives were two-fold.
1. To characterize the industrial effluents of a broad
range of industries discharging to Chesapeake Bay or
its tributaries in order to assess the impact of these
discharges upon the Bay ecosystem.
2. To develop a protocol for characterization of such
effluents which may be implemented by the States to
support discharge control decisions.
The effleunts of twenty-eight industries in the basin were charac-
terized using the protocol developed during the study. Industries
were selected to participate based on a toxicity ranking of the
major dischargers, along with input by the state agencies in
Maryland and Virginia. The screening protocol consists of five
major components:
1. Bioassays;
2. Organics (chemical) analysis;
3. Bioaccumulation potential;
4. Metals analysis, and
5. NPDES parameters.
The comprehensive characterization scheme is designed to be used
by managers to make better decisions with regard to controlling
toxic substances. The basic approach starts with several simplistic
111
-------
analyses and becomes progressively more detailed as the source of
the problem is investigated. The scheme is designed to first eval-
uate whether the effluent is toxic, then to pinpoint the substance
causing the toxicity.
IV
-------
CONTENTS
Abstract iii
Figures vi
Tables vii
Acknowledgements viii
1. Introduction . ..... 1-1
1.1 Background 1-1
1.2 Purpose and Scope 1-2
2. Summary 2-1
2.1 Introduction 2-1
2.2 Phase I: Screening Study 2-2
2.3 Phase II: Protocol Development Study 2-13
2.4 Phase III: Protocol Verification Study. . . . 2-19
2.4.1 Approach 2-19
2.4.2 Site specific toxicity identification
program: -recommended final
protocol . . 2-21
2.4.3 Results of application of the scheme. 2-28
3. Program Approach 3-1
3.1 Introduction 3-1
3.2 Plant Selection 3-2
3.3 Process Engineering Analysis 3-7
3.4 Field Sampling Methodology 3-18
3.5 Phase I: Screening Study 3-21
3.5.1 Phase I analysis scheme ....... 3-21
3.5.2 Phase I screening study results . . . 3-24
3.6 Phase II: Protocol Development Study 3-36
3.6.1 Phase II analysis scheme 3-36
3.6.2 Phase II results 3-37
3.7 Phase III: Protocol Verification Study. . . . 3-42
3.7.1 Phase III analysis scheme 3-42
3.7.2 Phase III results 3-53
3.8 References 3-55
4. Site Specific Toxicity Identification Program 4-1
4.1 Basic Philosophy and Intent 4-1
4.2 Elements of TIP 4-2
4.3 Practical Application of Scheme 4-10
4.3.1 Plant B141S 4-12
4.3.2 Plant B111D 4-13
4.4 References 4-14
Appendix A: Data Correlations from Toxic Point Source
Program A~l.
Volume II contains Appendices B and C
Volume III contains Appendices D and E
Volume IV contains Appendices F through J
-------
FIGURES
Number Page
2-1 Phase I analytical protocol 2-3
2-2 Bioassay protocol used for Phase I 2-4
2-3 Recommended Phase II chemical characterization
protocol 2-9
2-4 Recommended bioassay testing protocol 2-10
2-5 Actual chemical analysis scheme used in Phase II ... 2-14
2-6 Site specific toxicity identification program (TIP)
designed to evaluate effluent toxicity 2r22
2-7 Actual site specific toxicity identification program
used in Phase II 2-24
3-1 Fifty-gallon composite sample collection system pro-
posed for use in Phase II. 3-20
3-2 Phase I chemical analysis protocol 3-22
3-3 Bioassay protocol used for Phase I 3-23
3-4 Phase II chemical analysis protocol 3-38
3-5 Phase III extractable organics analysis scheme .... 3-46
3-6 Order of progression through an analysis of extract-
able organics 3-47
3-7 Sediment organics analysis scheme 3-51
4-1 Site specific toxicity identification program (TIP)
designed to evaluate effluent toxicity 4-3
4-2 Site specific toxicity identification program actually
used in Phase III 4-7
VI
-------
TABLES
Number Page
2-1 Decision Criteria for Phase III Samples 2-29
2-2 Effectiveness of Decision Analysis for Phase III
Effluents 2-30
3-1 Data Bases Used to Evaluate Chesapeake Bay Outfalls. . 3-4
3-2 Itemization of Effluent and Sediment Samples Collected 3-6
3-3 Sample Grouping by Industry Type 3-7
3-4 Process Analysis Data Sheets 3-10
3-5 Data Logs on GC/MS Samples 3-50
3-6 Checklist for Extractable Organics Analysis 3-52
4-1 Recommended Stage I - Basic Analysis 4-4
4-2 Potential Biological Tests for Assessing Environmental
Impact of Discharges to the Chesapeake Bay 4-5
4-3 Decision Criteria for Phase III Samples 4-11
VI1
-------
ACKNOWLEDGEMENTS
The authors wish to thank the Central Regional Laboratory of EPA
(Annapolis, Maryland) for their performance of NPDES analyses dur-
ing the program. The contribution of the following laboratories
is also gratefully acknowledged: EG&G Bionomics, Wareham, MA;
Commercial Testing Company, Chicago, IL; Battelle Memorial Labor-
atory, Columbus, OH; Howard Laboratory, Dayton, OH; Pollution
Control Science, Dayton, OH; Litton Bionetics, Kensington, MD;
Departent of Botany, University of Montana, Missoula, MT; U.S.
EPA, Corvallis, OR; and U.S. EPA, Research Triangle Park, NC.
The contributions of the three EPA project officers, Dr. Donald E.
Francisco, Dr. Richard E. Purdy, and Mr. Mark Alderson, and of the
Maryland and Virginia staff representatives, Mr. Charles Bostater,
Mr. Mike Haire (Maryland), and Mr. John Roland (Virginia) is
greatly appreciated.
In addition, the contribution to this program by the following
MRC staff members is greatly appreciated:
M. Arnold
S. Barksdale
J. Brooks
L. Campbell
S. Cheng
L. Cornett
D. Day
B. Desai
K. Dunlap
D. Dunn
M. Ehlenbach
C. Feldstein
P. Fletcher
J. Fullenkamp
C. Geiger
D. Glasgow
J. Graham
J. Gridley
J. Hammond
G. Hess
M. Hershey
B. Hickmott
W. Hillan
L. Holmes-
L. Kasten
B. Kennerly
F. Kulik
A. Lang
B. Lauper
J. Lavoie
R. Lewis
L. Metealfe
J. Miller
S. Mitrosky
R. Neal
S. Paterchak
B. Peters
J. Peters
J. Pustinger
D. Reinhardt
J. Reynolds
G. Rinaldi
W. Ross
D. Saunders
D. Sharp
T. Shell
C. Smith
J. Spillman
K. Tackett
G. Thomas
D. Trapp
B. Warner
J. Weaver
M. Wininger
A. Wright
R. Yelton
Vlll
-------
SECTION 1
INTRODUCTION
1.1 BACKGROUND
The Chesapeake Bay Program was established by Congress to deter-
mine the causes of the deterioration of the Bay ecosystem and
recommend strategies for restoration and protection of the Bay
resources. The Chesapeake Bay Program is administered by the
U.S. Environmental Protection Agency in cooperation with the
States of Virginia and Maryland as well as a Citizen's Committee
established for the program.
The Chesapeake Bay Program has four areas of emphasis:
1. Toxics Accumulation in Food Chain Study
2. Submerged Aquatic Vegetation Study
3. Eutrophication Study
4. Management Study
The Toxics Accumulation in Food Chain Study includes the identi-
fication, transport, and effect of a variety of toxic compounds
in the Bay. The Toxic Point Source Assessment of Industrial Dis-
charges to the Chesapeake Bay Basin covered by this report is one
task within this study.
The submerged Aquatic Vegetation Study is designed to define the
magnitude of the deterioration, and to recommend guidelines for
restoration and protection of the communities of submerged aquatic
1-1
-------
vegetation. One aspect of this study is to investigate the effects
of potentially toxic materials on the community.
The Eutrophication Study emphasizes data acquisition and simula-
tion modeling of the eutrophication process in the Bay. The ul-
timate goal is to recommend a control strategy.
The Management Study is designed to determine the best management
strategy or system for the future protection of the Bay. Since
the goals of each of the other studies is to recommend control
strategies, the Management Study is to integrate control strate-
gies with the political or management entities active in the Bay
Region.
1.2 PURPOSE AND SCOPE
Through competitive bids, Monsanto Research Corporation (MRC) was
contracted by the U.S. Environmental Protection Agency (contract
68-02-3161) to conduct a three-phased, 27-month study with the
following two initial objectives:
To characterize 80 industrial effluents from a broad range
of industries discharging into the Chesapeake Bay or its
tributaries in order to assess the impact of these dis-
charges upon the Bay ecosystem.
To develop a protocol for characterization of such efflu-
ents which can be implemented by the States of Maryland and
Virginia and the U.S. Environmental Protection Agency to
support discharge control decisions.
It was recognized by EPA early in the formulation of this project
that these objectives were very agressive and that there were
countless chemical and biological analyses which could be used.
Therefore, EPA in the Request for Proposal (DU-79-A213), set forth
1-2
-------
an initial set and sequence of chemical and biological analyses
to use in Phase I of the project. The project was then conducted
in three phases, with each phase building on the success of the
preceding phase. As the program matured, emphasis frequently
shifted between these two basic objectives and the approach to
take for each objective.
The EPA also recognized that a massive amount of data would be
collected in this project and that there would not be adequate
time or funds available at this time to conduct detailed data in-
terpretation and correlation studies for all the outfalls sampled.
Therefore, the basic intent of this project was 1) to present
major results and conclusions, and 2) to document and archive all
the raw data so they could be consulted by future researchers for
their specific project needs.
This report presents an overview of the results of Phases I and
II, presents all the raw data from the 28 effluent and 22 sediment
samples collected in Phase III, and presents the Site Specific
Toxicity Identification Program (TIP).
This four-volume report is organized as follows:
Volume I
1. Introduction
2. Summary
3. Program Approach
4. Recommended Decision Analysis Scheme for Toxicity
Assessment
Appendix
A. Data Correlations from Toxic Point Source Program
Volume II: Appendices B and C
B. Phase III Plant Presurvey Reports
C. Phase III Sampling and Analytical Methods
1-3
-------
Volume III: Appendices D and E
D. Phase III Plant-by-Plant Analysis Results for Effluents
E. Phase III Plant-by-plant Analysis Results for Sediments
Volume IV: Appendices F through J
F. Relative Retention Indices of Phase III Effluents
G. Total Ion Chromatograms for Phase III Effluent Samples
H. Relative Retention Indices of Phase III Sediments
I. Total Ion Chromatograms for Phase III Sediment Samples
J. Analysis of Blanks and Standards Associated with the
Analysis of Plant Effluents and Sediments
1-4
-------
SECTION 2
SUMMARY
2.1 INTRODUCTION
I
EPA research efforts on the Chesapeake Bay range from determining
what is .being discharged into the Bay to studying the ultimate
fate/disposition of pollutants, including accumulation in the
food chain. In support of these efforts, Monsanto Research Cor-
poration (MRC) was awarded EPA contract 68-02-3161 to conduct a
three-phased, 27-month study with the following two initial
objectives:
To characterize 80 industrial effluents from a broad range
of industries discharging into the Chesapeake Bay or its
tributaries in order to assess the impact of these dis-
charges upon the Bay ecosystem.
To develop a protocol for characterization of such efflu-
ents which can be implemented by the States of Maryland and
Virginia and the U.S. Environmental Protection Agency to
support discharge control decisions.
The EPA designed the program in three phases. In general, in
Phase I a chemical analytical and bioassay test scheme was devised
and implemented using effluent samples from 10 industrial sites.
Data from this research effort were then evaluated by the Chesapeake
Bay Program Office,~U.S. EPA, Maryland and Virginia state water
2-1
-------
pollution control authorities, and other Bay researchers to deter-
mine how the type and quality of data met their needs. Results
of this evaluation were then incorporated into the analytical
scheme in Phase II and implemented on effluent samples from an
additional 10 industrial sites. In addition, sediment samples
were collected near 8 of 10 sites and subjected to the analytical
scheme. At the end of Phase II, another data evaluation was per-
formed to finalize the analytical scheme from additional sites in
Phase III.
The following two subsections briefly outline and summarize results
from Phase I and II. The purpose of these subsections is to put
the program in perspective so that the approach, objectives, and
interpretation of results from Phase III are understood.
2.2 PHASE I: SCREENING STUDY
The main purpose of Phase I was to collect final effluent samples
from 10 different industries to assess the type, quantity, and
usefulness of the data generated by subjecting samples to the
chemical and biological analyses shown in Figures 2-1 and 2-2.
•
Initially, through the results of a previous contractor (EPA con-
tract 68-02-2607), EPA was to supply MRC with a list of 80 indus-
trial outfalls to sample; evenly distributed between Maryland and
Virginia within the Chesapeake Bay Basin. Results of that con-
tract were not as successful as planned, so MRC worked closely
with the wastewater control authorities in Maryland and Virginia
and with the EPA to quickly develop a candidate list of industries.
In order to maintain Phase I on schedule, the following 10 out-
falls were selected from those which best challenged the analytical
protocol and from companies most interested in cooperating with
this project:
2-2
-------
1
SHEEPSHEAD
MINNOW
ASSAY
DAPHNIA'
ASSAY
INDUS
irru
1
AQUATIC BIOASSAYS
1
, 1 a 1 'a 1
TAT ^Ly SP M££ TOTAL ORGANIC
ASSAY ASSAY ASSAY ASSAY CARBON UOniL)
TRIAL
JENT
(3.800mL)
CHO CYCTOTOXICITYON
NEAT SAMPLE USING
ANTIBIOTICS
(DOSE RESPONSE)
|
RCRA
BIOACCUMIIIATKiN
POTENTIAL
FILTRATION
STERILIZATION 0.2pm (30mLI \ PARTICULATES > 5pm(1.000mLl
FILTRATE) 100 mL EACH)
AMES
MUTAGENICITY
(SPOT ASSAY.
TOXICITY ASSAY.
PLATE INCORPORATION-
DOSE RESPONSE)
ACUTE CYTOTOXICITY
ICHO CLONAL-
DOSE RESPONSE)
SOXHLET EXTRACTION OF FILTERS
(125 mL OF ACETONE)
CONCENTRATE
I
EXTRACT (20 mL)
r
AMES
MUTAGENICITY
(THREE CONCENTRATION
DOSE RESPONSE)
t
CHO ACUTE
CYTOTOXICITY
(SPOT TESV)
8 TEST PERFORMED ON ONLY TWO SAMPLES
Figure 2-1. Phase I analytical protocol.
-------
K>
I
CONCENTTIAT1
»
CC/MS
DIP-MS
SPECIFIC
TOXICS
s
N
COllir.l SAMI'lt 1 ._
INHfOR BOO. ml). 1 SI NO
I'DISPARAMIIIRS
1 ' — - — -i
SINDIOMKC
Tnr ^MI
IIOLITERSI
1 1
PERFORM ORGANIC FXTRASAMPIE
ANALYSIS (Iqall
1
rr/M«-^_ VOLATILE* BY
GC"VIS PURGE AND TRAP
ORGAN! C-
PHASE
r* ORGANIC-
PHASE
CONCENTRATE
COMBINE f
1 GC/MS
rrn OIPMS
/.OA,, SPECIFIC
GR.AV TOXICS
1
NONVOLATILES
1
- — CH.CI EXTRACTION
Z ZATpH?
1
- — CH.CI. EXTRACTION L
' 'AIPH« FIITR*
AQUEOUS PHASE TQC
SPARGE WITH N; iminiK I
TOC ANALYSIS2 J^J15 '
HqBYAA
SAVE AT <°C
FOR BIOTTSIS
HYSIS
(ILITTRI
I
PERFORM INORGANIC
ANALYSIS
_FIlTRATIpN
NITROCaiULOSC
FILTERS
i
TE [ FILTER
DRY AND -»
IY 1C
r~
DIGEST SSMS
1
Hg ANALYSIS
BYAA
I I I Tl I I
TCO AND GRAV; GC/MS AND DIP - MS
ANALYSIS OF EACH FRACTION
IR OF FRACTIONS 5.6* 7
Figure 2-2. Bioassay protocol used for Phase I
-------
Number of
Industry type outfalls
Pulp and paper 3
Plastic and synthetics 2
Textiles 2
Inorganic and organic chemicals 1
Leather tanning 1
Oil refining 1
As a result of the above two criteria, these 10 outfalls were all
located in Virginia and were principally located on tributaries to
the Chesapeake Bay and not directly discharging into the Bay.
As it turned out, the 10 plants who were most cooperative also
had, for the most part, well operated wastewater treatment plants.
Therefore, in Phase I these samples were relatively clean in
terms of harmful chemical compounds and acute aquatic toxicity
and did not adequately challenge the analytical scheme.
The intent of the chemical analytical scheme in Phase I (Figure 2-1)
was to analyze samples for the plant specific NPDES permit param-
eters, metals (soluble and insoluble), and major (>1 ug/L) organic
compounds which were methylene chloride extractable. For this
screening study, the intent of the organic analysis portion of the
scheme was to identify major chemical compounds which appeared
above the background. For this project the objective was not to
spend time trying to identify all trace (<1 ug/L) compounds, but
to store the GC/MS data on magnetic tapes and send them to the
EPA for archiving for potential future use in other Bay research
programs. For example, the complete GC/MS chromatogram could po-
tentially be used as a "fingerprint" of specific unknown compounds
to be used by R. J. Huggett at VIMS for identifying possible
sources of unknown organic compounds found in oyster tissue and
bay sediment samples.
Of special interest, in Phase I all metals samples were analyzed
by spark source mass spectrometry (SSMS) for 76 elements. SSMS
2-5
-------
was specifically used in this phase to screen samples for the
presence of elements which are not normally analyzed in waste-
water samples, but could be present due to specific industrial
processes (such as rare earth elements leached from catalysts).
It was recognized that SSMS would be a little more expensive and
that the quantitation was not as good as more common metals analy-
sis procedures.
However, SSMS is an excellent tool for screening samples for ele-
ments which could be present. Results of Phase I indicated, how-
ever, that no elements beyond the 26 metals normally detected by
the ICAP analytical technique were present in concentrations sig-
nificantly above their detection limits. Therefore, it was rec-
ommended that metals analyses for the remainder of the program be
performed by ICAP.
MRC also incorporated into the analytical scheme separate metals
analyses of effluent filtrate samples and the corresponsing fil-
tered particulates. These analyses were included to see if spe-
cific metals were selectively associated with the suspended solids
or remained in solution. This information could have a very sig-
nificant impact on fate and deposition of metals in the bay, on
selection of wastewater treatment control options, and on effluent
toxicity. Results from Phase I did indicate that for the few
samples analyzed, there were indications that selected metals
from specific types of industries appeared predominately either
in solution or in the particulates. Therefore, this analysis
was continued throughout the project.
The bioanalytical testing protocol used in Phase I is shown in
Figure 2-2. The EPA recognized in the formulation of this project
that there was a countless number of bioassay tests which could
be used. Issues which were discussed included: acute vs. chronic
tests; static vs. flow-through tests; standardized test species
vs. species indigenous to the Chesapeake Bay; type of species
2-6
-------
(fish, algae, shrimp, daphnia, etc.); whether to incorporate
benthic species (oysters, worms, etc.); larvae growth studies,
mutagenic tests (e.g., Ames test), cytotoxicity tests (e.g., CHO,
RAM, etc.), and bioaccumulation tests.
The EPA decided only to do acute toxicity tests instead of incor-
porating chronic tests due to 1) the much higher costs associated
with chronic tests, 2) uncertainties in how to interpret and use
results from chronic tests, 3) acute toxicity tests are consistent
with the philosophy of the chemical characterization scheme, and
4) uncertainties in which species to test. In terms of whether
to use standardized test species or species indigenous to the
area, EPA decided, for this program, to use a standardized series
of test species. By using the same test species for all outfalls,
the relative toxicity of each outfall could be prepared which was
the purpose behind program objective two. Also, for the species
selected, there were well established EPA and ASTM procedures and
a considerable amount of effects data available in the literature.
Flow-through acute toxicity tests were not selected principally
because of the much higher costs of performing the test and be-
cause static acute tests on a single sample do give data consis-
tent with the project's needs.
For Phase I, EPA did include tests for mutagenicity (Ames test),
cytotoxicity (CHO test), and bioaccumulation. The purposes for
including these tests were to see what type of data would result,
and to evaluate how the information could be used by this project
and by EPA permit authorities. Of specific interest in Phase I,
MRC recommended conducting Ames and CHO tests on the aqueous efflu-
ent sample and on the suspended solids in the effluent to evaluate
if harmful compounds were preferentially adsorbed onto the solids
or remain in solution. Since the 10 plants selected in Phase I
had relatively clean effluents, results of these two tests did
not yield enough data to adequately evaluate the concept. For
2-7
-------
Phases II and III, EPA decided to exclude these two tests (Ames
and CHO), principally because the permitting authorities had
trouble incorporating this type of data into their permitting
process.
Results from the bioaccumulation potential test were very promis-
ing and did give data which could be used to screen samples for
potential toxicity problems and data that could be used by the
Chesapeake Bay researchers. Therefore, this test was continued
throughout the remainder of the project and did mature into an
integral part of a total effluent toxicity assessment protocol.
As one might expect, based on the comprehensive objectives of
Phase I and based on the complexity of the chemical and biological
characterization schemes used, a massive amount of data was gen-
erated which resulted in numerous results, conclusions, and recom-
mendations. This information was compiled into a 1,200-page,
two-volume report issued to the EPA in May 1980. The improved
chemical characterization and bioassay protocols which were rec-
ommended for Phase II are shown in Figures 2-3 and 2-4.
Figure 2-3 presents the chemical characterization scheme recom-
mended for Phase II of the program. The basic Phase II philosophy
was to improve the identification of those organic compounds which
were either nonextractable or extractable but not chromatographable.
The following items highlight the significant features of the
recommended Phase II chemical analysis protocol:
Inductively coupled argon plasma (ICAP) spectroscopy will be
used for metals analysis. The ICAP method is more precise,
more cost-effective, and provides the same data as did the
spark source mass spectrometry (SSMS) analyses in Phase I.
2-8
-------
SAMPlt 1
SIM) 111 AIDFOK
NPDtS fARAMHIHS.
SELECTED ANIONS.
ANDMHAIS
--•f~
[ SEND TO MRC
IIOLITCRS)
t
PERFORM ORGANIC
ANALYSIS
f
^
SEND FOR AQUA! 1C
BIOUS1S
110011
(^LITERS!
(lliriRI
I
EXTRA SAMPIE
14 LITERS'
1 PERFORM ELEMENTAL
1 ANALYSIS
MIJTACW:i|Y
CYTUTOXICITV &
BIOACC'JM'ilATION
i
vO
ORGANIC-*—CH.CI, EXTRACTION
PHASE ZATpH2
CONCENTRATE
r
ORGANIC-*-CH,CI, EXTRACTION
PHASE
22
COMBINE
CONCENTRATE
ATpHIZ
SAVEAT4"C
- DIRECT WASTE INJECTABLES
- ALDEHYDE ANALYSIS
- NITROGEN - PHOSPHOROUS
DETECTOR FIDfGC
OERIVITIZATION
TCOGRAV
GC/EC
w/CAPILLARY COLUMN
LC FRACTIONATION
(CAP
FOR 24 METALS
As&HgBYAA
SULFUR BY PEROXIDE
DIGESTION
DRY AND
WEIGH TSS
DIGEST
I CAP FOR 24 METALS
As&HgBYAA
SULFUR BY PEROXIDE
DIGESTION
DIP-MS IHPLC w///V DETECTOR AND*
DIP-MS ANALYSIS OPTIONAL FOR
MAJOR UNKNOWN COMPOUNDS)
TCO.GRAV. IR. CAPILLARY COLUMNGC/MS
-IGC/FTIR,OPTIONAL FOR MAJOR UNKNOWN COMPOUNDS!*
BASED ON EPA PROJECT OFFICER APPROVAL
Figure 2-3. Recommended Phase II chemical characterization protocol.
-------
INIHISTKIAI IIIIIIINI
linn 11
(WO ml I
® IN VIIRU HtAITH EFFI.CIS ASSAYS
K)
I
MARINE DISCHARGES FRESHWATER DISCHARGES
[• SHEEPSHFAD MINNOWS FRESHWATER ALGA -1
}• MYSID SHRIMP DAPHNIA 4
L MARINE ALGA FATHEAD MINNOWS J
© FILTER STERILIZE 110 itt)
0.45pm FILTER
(9.10) mil POTENTIAL TES1
30 ml ®
" CHO CLONAl CYTOTOXICITY ASSAY
(STFRILI71D VIA ANTIBIOTICS ADDITION)
1 MAX. CONC. 1350 Ml/ml.)
(IF TOXIC. DOSE RESPONSE TEST FOR ECW)
© CENTRIFUGATION, FILTRATION. AND CONCENTRATION
W/5um FILTER AND XAD RESIN 175(10 mil
1
FILTRATE 10 ml
t
/^ FILTERS W/
W CENTRIFUGED PARTICULATES
1
XAD RESIN
1
0 AMES MUTAGENICITY ASSAY
WITH STRAINS
TA98, TA100. TAIS3S. TA1537
WANDW/DS -9 ACTIVATION
I.OODtll MAX. DOSE TEST
(IF POSITIVE. A 3 CONC. DOSE RESPONSE
TEST IS REQUIRED)
©CHOCLONAL CYTOTOXICITY ASSAY
1 MAX. CONC. 1250 id/ml I
(IF TOXIC, DOSE RESPONSE TEST FOR EC,,,)
@ SOXHLET EXTRACT
W/ACETONE
(FINAL VOLUME 10 ml)
t
I
® AMES MUTAGENICITY
ASSAY USING STRAINS
TA98. TA100. TAI535. TA1537
WANDW/DS-9 ACTIVATION
50.10. AND I ||L/PIATE
I
® CHO CLONAl CYTOTOXICITY
1 MAX. CONC. (BO Ml/mil
(IF TOXIC. DOSE RESPONSE TEST FOR EC^I
@ SOXHI.ET EXTRACT
W/ ACnONE
(FINAL VOLUME 10 mil
i) AMFS MDTAGENICITY ASSAY
WITH STRAINS
TA98. TAlflO. TA1535. TAI53?
W ANOVV/OS -9 ACTIVATION
50. IV V 1. 0.2. AND 0.05 Ul
Figure 2-4. Recommended bioassay testing protocol.
-------
Continue to conduct metals analyses on filtrate and fil-
tered particulates to define partitioning of these species
between aqueous and solid phases. These data are useful to
researchers studying sediments in the bay.
Continue to have EPA Annapolis Field Office (AFO) perform
analyses for the NPDES permit parameters, selected anions,
and metals because of AFO's close proximity to the plants.
This ensures rapid analysis for those parameters which must
be analyzed within 24 hours of collection.
Use capillary column gas chromatography/mass spectrometry
(GC/MS) for improved organic compound separation and
identification.
Use capillary column gas chromatography with an electron
capture detector to screen samples"5 for potentially toxic
halogenated organic compounds.
Since capillary column GC/MS gives better resolution com-
pared to packed column GC/MS, the seven liquid chromatography
fractions will be combined into EPA's recommended four frac-
tions for more cost-effective GC/MS analyses.
More specific chemical characterization approaches will be
employed for determination of those compounds known or
suspected to be present in the effluent (based on the site
presurvey) which are not amenable to analysis by the basic
Phase II protocol.
For those samples which contain a high percentage of non-
chromatographable organic content, high pressure liquid
chromatography fractionation and DIP-MS analysis will be
used.
2-11
-------
The Phase II bioanalytical protocol was developed based upon a
critique of Phase I results. The recommended Phase II bioassay
protocol is shown in Figure/2-4.
The emphasis of the recommended Phase II bioassay scheme was to
provide sufficient data, both using ecological effects and health
effects tests, to determine the toxicity of the effluent to the
receiving stream and to those species which use the water. There-
fore, it was necessary to continue to use a battery of tests,
since each test measures a different mechanism of toxicity.
The following items highlight the significant features of the
recommended Phase II bioassay protocol:
Continue to use static acute aquatic tests as a screening
tool to assess potential toxicity to a range of aquatic
vertebrates, invertebrates, and plants as recommended by
EPA.
Conduct the EPA RCRA bioaccumulation potential assay.
The bioaccumulative organic compounds would be identified
by 1) comparison of retention times to reference data of
known compounds, or 2) further instrumental analytical
characterization of the fractions from the HPLC measurement.
Continue to conduct Ames mutagenicity tests since 1) it is
the only chronic health affects test in the Phase II proto-
col, and 2) it is the most completely validated mutagenicity
test available.
• Conduct Ames tests on concentrated filtrate (resin accumu-
lators) in cases where mutagenicity is not observed on the
as-received wastewater effluents.
2-12
-------
Continue to conduct the CHO cytotoxicity assay for acute
toxic effects since it was the most sensitive measure of
health effects in the Phase I studies.
2.3 PHASE II: PROTOCOL DEVELOPMENT STUDY
Figure 2-5 shows the chemical characterization scheme actually
used in Phase II based on EPA's review of the scheme recommended
from Phase I, the program objectives, and cost and time constraints
Only a few modifications were made to the scheme used in Phase I.
For example, in the Phase I scheme, the effluents had been ex-
tracted to give acid and base/neutral fractions, but the extracts
had been recombined prior to GC/MS analysis. Based on results of
Phase I, the chemical analysis scheme was modified for Phase II
to improve chromatographic separation and identification of
organic species in the effluents. This involved separating ex-
tractable organics into base/neutral and acid fractions before
GC/MS analysis, and performing all GC analyses of nonpurgeable
organics with capillary columns.
Secondly, we included an evaluation step; "Is TCO complex?" .The
use of the capillary column GC for both the total chromatograph-
able organics (TCO) and the GC/MS analyses had the advantage that
the chromatograms obtained in the TCO analysis could be inspected
to see whether a liquid chromatographic separation of the sample
would lead to substantial improvement in compound identification
in the GC/MS analyses. The chromatograms from the TCO and GC/MS
analyses were sufficiently similar that one could predict what
dilutions to use for the GC/MS samples, thereby minimizing ana-
lytical costs- Also, the relative retention indices (a calcula-
tion performed for data support of other CBP researchers) could
be calculated from the data obtained in the TCO analysis.
2-13
-------
SAMPLES RECEIVED
FOR
ANALYSIS AT MRC
|5CCr.L" FILTER
4L
TOC. TSS
PURGEABLEORGANICSBY
PURGE AND TRAP AND
CC'MS ANALYSIS
[VARIABLE
SPECIAL TESTS
DETERMINED BY
FIELD PRESURVEY
BIOACCUMUUTIOM
HHL
ORGANIC ANALYSES
80 mL
10L
EXTRACTABLE ORGANICS
ORGANIC PHASE
CH.CL EXTRACT AT pH12
CONCENTRATE TO 10 mL
AQUEOUS PHASE
CH2CI? EXTRACT AT pH2 '
€RAV, CAPILLARY: TCO.
TCG, GC/EC
ORGANIC PHASE
CONCENTRATE TO 10 mL
GRAV. CAPILLARY:TCO,
TCG, GC/EC
DIRECT AQUEOUS
INJECT! ON GC/MS
GRAV, CAPILLARY:TCO,
TCG. GC/MS ON EACH
FRACTION
AQUEOUS PHASE
Figure 2-5. Actual chemical analysis scheme
used in Phase II.
2-14
-------
A brief study was included in Phase II to see whether major non-
purgeable and nonextractable species were present in the effluents
which could be detected by direct injection into the GC/MS of the
aqueous effluent remaining after basic and acidic extraction. The
detection limits for this procedure proved to be quite high and,
except for one effluent where a high concentration of acetic
acid wa.s found using the direct injection technique, little use-
ful information was obtained.
After reviewing results from Phase I and based on review comments
from other CBP researchers, the EPA decided to include sediment
analyses into the Phase II analytical scheme. Therefore, in order
to keep the program within cost and time constraints, EPA decided
to exclude all bioassay tests (as recommended in Figure 2-4) from
Phase I. This decision was also based in part on the fact that
1) state agencies already knew what type of data could be gen-
erated from acute aquatic toxicity tests and knew how to use it,
and 2) although the Ames and CHO tests gave useful information,
the states were uncertain how to interpret the results and effec-
tively use the data. Also, effluent samples from 8 of the 10
effluents sampled in Phase II were subjected to mysid bioassay
testing by Battelle Columbus Laboratories for EPA contract 68-02-
2686. Therefore, some acute aquatic toxicity data were to be
available. While eliminating the bioassay tests did result in a
greater level of chemical characterization, results from the
Battelle study were not available in time to study the potential
environmental significance because cause/effect relationships
could not be evaluated.
For Phase II, 10 outfalls from industries discharging directly
into the Bay were selected to provide a broad range of industry
types and effluent compositions which would best challenge the
analytical protocol. The industry types and number of outfalls
selected were as follows:
2-15
-------
Number of
Industry type outfalls
Iron and Steel 2
Coke Production 1
Organics and Surfactants 1
Soap and Detergent 1
Inorganic Chemicals 2
Waste Recovery 1
Municipal/Industrial POTW 1
Pesticides and Plasticizers 1
In addition, Bay sediment samples were collected adjacent to 8 of
the 10 outfalls. The two outfalls not sampled, for sediment dis-
charged into separate storm sewers, and the end discharge points
of these storm sewers were unknown.
In terms of conclusions, approximately 40% of the NPDES parameters
measured for the Phase II effluents were above the respective
plant discharge limits and normal monthly averages. Most efflu-
ents had at least one parameter with values which exceeded their
discharge limits. These data indicated that the 3-hr composited
grab samples, while sufficient for testing the protocol, were not
always representative of normal operation, and any conclusions
about impacts to the Bay drawn from these data must be considered
in this light.
Direct aqueous injection of effluent samples to identify organic
components in the aqueous phase after methylerie chloride extrac-
tion has very limited utility due to high detection limits. In
the case of effluent B142S, however, it proved to be very infor-
mative. This sample had a very high TOC value (960 ppm) which
was not accounted for by GC analyses. This implied the presence
of a high concentration of nonpurgeable and nonextractable organic
species. Approximately 2,000 ppm of acetic acid was found in
2-16
-------
this sample when the effluent, after extraction and adjustment of
the pH to 2, was analyzed by direct aqueous injection.
As stated earlier, the purpose of the bioaccumulation test is to
determine how many organic compounds in a methylene chloride ex-
tract of the effluent are potentially accumulative in fatty tissue.
Potentially bioaccumulative compounds are those which elute from
the HPLC with a log P £3.5 and response £25% full-scale deflection.
In Phase I, several samples contained several compounds which were
classified as bioaccumulative. So, in Phase II we made a greater
effort to identify the exact compounds which caused the positive
response. This identification effort was conducted using the fol-
lowing information: 1) log P values determined for each efflu-
ent sample, 2) organic compounds identified in each sample by
GC/MS, 3) published log P values for the compounds identified by
GC/MS, and those identified by the process analysis, and 4) com-
pounds published in the literature with log P values similar
(within ±0.05) to the values determined in the sample. By com-
paring all the above information for each sample, it was possible
in Phase II to make several tentative identifications of compounds
which caused a positive bioaccumulation response. In Phase III,
further improvements to this analysis were implemented.
While a correlation between several GC/MS compounds and log P data
was found for each sample, the HPLC data contained more compounds
(several which were bioaccumulative) than could be accounted for
by GC/MS. Two factors may account for this difference in log P
and GC/MS results. First, some of the compounds detected by HPLC
may not be chromatographable with a capillary column, yet may
elute through an LC column. Second, while the concentration
factors for the extracts injected into the two instruments are
essentially the same, the HPLC received 10 times more extract
(10 pL vs. 1 ML) than the GC/MS. This may explain why some com-
pounds were either not seen by the GC/MS, or were seen but not
2-17
-------
identified due to their low level and to the project time and
budget constraints.
The final data correlation in Phase II revolved around the attempt
to determine if the respective industrial outfalls were contrib-
uting to an increase in toxic organics in the bay sediment, and
thus posing a threat to bottom-dwelling species such as oysters
and clams. In attempting to correlate effluent data and sedi-
ment data from samples collected near the effluent outfall, both
the history of each site and recent meteorological events must
be considered. Since little is known about the dynamics of sedi-
ment movement in the bay, identification of organics being con-
centrated in sediments from industrial effluents may never proceed
beyond the tentative stage. However, even if a correlation is
not evident, the analysis of the sediment for GC/MS organics and
bioaccumulation potential will determine if an environmental prob-
lem exists, and long-term monitoring will determine if the problem
is increasing or decreasing in magnitude. The sediment/effluent
comparisons in Phase II did reveal some possible correlations,
especially with respect to potentially bioaccumulative species,
several of which were present at a concentration 1,000 times
greater than that of the corresponding effluent.
As a result of Phase II, several recommendations were made for
the Phase III effort and the final protocol development:
Capillary column chromatography should be continued; its
mass spectral data are far superior to those from packed
column chromatography.
The provisions for adding special analytical tests based
on presurvey information should be continued, because all
plants should not be assumed to discharge the same
compounds.
2-18
-------
Correlations between GC/MS organics and bioaccumulative or-
ganics should be developed further by collecting the positive
bioaccumulation organics as they elute from the HPLC and
injecting them into the GC/MS.
Aquatic toxicity testing should be included, as there is no
other effective means for determining if a negative impact
is imposed on the receiving stream. This will place empha-
sis on the real environmental significance of the organics
analysis data.
2.4 PHASE III: PROTOCOL VERIFICATION STUDY
2.4.1 Approach
In Phase II, the principal thrust was directed towards collecting
considerable chemical characterization data on both effluent and
sediment samples. These data were then used by other researchers
studying effects in the Bay. For Phase III, EPA shifted the prin-
cipal emphasis back to the second program objective; namely, to
develop a protocol for the characterization of effluents which
could be used by the States of Maryland and Virginia and the U.S.
EPA to support discharge control decisions.
The chemical analytical scheme used in Phase II (Figure 2-5) was
basically the scheme used in Phase III. Improvements to the
scheme implemented in Phase III mainly centered around how to ef-
ficiently handle a large number of samples in a multi-step analy-
tical scheme. As a result, MRC developed standardized check
sheets which followed each sample through the analytical scheme.
As was requested by EPA, all biological effects tests were elim-
inated from Phase II to allow more effort on chemical characteri-
zation of the sample. This lack of data made it very difficult
to interpret the potential impact on the Bay due to the effluents
2-19
-------
tested. Therefore, in Phase III, biological testing was again
included into the testing protocol, although the scope was much
less than that used in Phase I. For Phase III acute static bio-
assays were conducted on 18 of 28 effluent samples using the mar-
ine species Mysidopsis bahia. The State of Virginia Water Control
Board conducted acute static bioassays using the fathead minnow
Pimephales promelas on the remaining 10 effluents; they also
tested 2 of the above 18 samples to provide limited companion
data for different species.
The industry types and number of outfalls and corresponding sedi-
ment samples collected in Phase III were as follows:
Number of Number of
Industry type outfalls sediment
Organic and Inorganic Chemicals Manufacture 3 2
Wood Treating and Processing 1 1
Plastics Manufacture ' 1 1
Municipal Sewage Treatment Plant (POTW) 8 6
Nylon Resins and Fiber Manufacture 3 3
Industrial Waste Treatment and Recovery 1 1
Stainless Steel Manufacture 2 0
i
Tobacco Processing 2 1
Commercial Electric Power Generation 1 1
Chemical Fertilizer Production 2 2
Pork Slaughter, Processing, and Packaging 1 1
Rubber Product Manufacture 1 1
Inorganic Pigment Manufacture 2 2
Again, as one might expect, the amount of chemical characteriza-
tion and bioassay data resulting from 28 effluent and 22 sediment
samples was massive. These data and their interpretation are
therefore presented in detail in Appendices A through J. The
remainder of this summary section will discuss the protocol which
2-20
-------
was developed and tested in Phase III, and the success of using
it to identify causes of toxicity.
2.4.2 Site Specific Toxicity Identification Program (TIP);
Recommended Final Protocol i
This project was designed with the principal goal of developing
an analytical scheme which used data, both from chemical analyses
and bioassays conducted in a logical sequence, which could be
used to determine if industrial effluents were toxic and, if so,
identify the cause of the toxicity. The approach MRC developed
centers around a decision analysis that starts with several simple
analyses and becomes progressively more detailed as the cause of
toxicity is investigated. The actual chemical analyses, bio-
assays, and decision criteria used in the protocol are "totally
the user's choice." The intent of Phase III was to test and, some-
what verify, that the "sequence" of chemical analyses and bioassays
recommended in Phase II, and the "type and placement of the deci-
sion points" in the protocol, met the program's objective.
The basic philosophy and sequence of steps for the final Site
Specific Toxicity Identification Program (TIP) resulting from
Phase III is presented in Figure 2-6.
"Up front planning" is the key to the success of using this pro-
tocol. First, the user of the protocol must decide why he wants
to use it and what type of data he needs to make the types of
decisions necessary. For example, a state water control agency
may want to use the protocol to screen numerous effluents to .
identify those which are most toxic so they can prioritize their.
resources and effort. Therefore, the type of chemical and bio-
assay analyses they select may be different than if the personnel
at a single industrial plant want to use the protocol to study
the potential toxic effects of their effluent.
2-21
-------
Site Specific
Process Evaluation
1
Sample Collection
I
Stage I
Basic Analyses
No /SampleVYes
Toxic
Stage II
Intermediate Analyses
Yes
No
Stage III
Advanced Analyses
1
Toxic Reduction
Program
Figure 2-6.
Site Specific Toxicity Identification Program
(TIP) designed to evaluate effluent toxicity.
2-22
-------
The types of chemical analyses used will depend on the capability
of the user's laboratory facilities and/or funds available for an
outside laboratory. The type of bioassays selected also depends
on the same criteria, and other factors such as does the effluent
discharge into a marine or freshwater environment, are there par-
ticular aquatic or benthic tests which should be conducted due
to the location of the outfall, etc.
While all of these factors require some thought and planning, the
effort is not that extensive and the detailed guidance for making
the decisions is presented in Section 4 of this report.
The actual protocol MRC used in Phase III is shown in Figure 2-7.
This figure shows which specific chemical analyses, bioassays,
and decision criteria were used to meet the other goals of this
program. Once again, planning is the key to successfully imple-
menting the protocol. The first essential step, then, is to know
as much about the plant site and its effluent as possible before
you spend a dollar to collect or analyze samples. Due to the
large number and variety of industries involved, MRC conducted a
brief "site-specific process evaluation" in order to learn more
about each effluent and what might be present. The purpose here
was to go through a methodical process of data gathering which
resulted in a comprehensive discharge assessment. A part of this
data gathering consisted of determining if any "special" analytical
techniques were necessary to characterize the water sample.
When the evaluation was completed, a field sampling team was out-
fitted with the necessary bottles, labels, preservatives, and
packing materials. A 24-hr (minimum) effluent composite was col-
lected. Samples of shorter duration (e.g., grabs) could give
misleading results, especially if a toxic slug or a particularly
clean slug of effluent is discharged at the time of sampling.
2-23
-------
,'..»' .1. K* (•<«*: U-. X
-.: . i,'\ »SiVl»t> !/•
STAGE 1
BASIC
ANALYSES
M*. 1- S'll'Illi
nikii! IhJirilM
Ul'IiL MIULS
?t Mil if)
-- I -
U~
«••
^^
NO
X^
IrilPJ 1C
I»:TICMTKK
»
Y!SNN^
ICO ICC «[> CMV
C'dUB. COUJMl,
ccws «Mmis
L-L"<
I • ^ TOXIC S
, ^s^IM"TI
STAGE II
INTERMEDIATE
ANALYSES
POTENTIAL
r*ir >.??_* y-iTSii
FOSUU ANAIYSCS:
RiO-iM KOl.l1t»|r.C
»'CTICV>T10t,BlMSS«y
FICft tt^Vi^MHOASSAY
. [IrPlM? STREAKS
mviSTion (icvcu/iiusi
AhO lIUlMiM KCDiriCAIlOlS
10 HOUCUKMOVt lOIICllr
Figure 2-7. Actual Site Specific Toxicity Identification
Program used in Phase III.
2-24
-------
For Phase III of this project the Stage I Basic Analyses included
the NPDES permit parameters, anions, metals, purgeable organics,
TOC, bioaccumulation potential, and static acute aquatic toxicity
tests. Data from the NPDES analyses were mainly used to determine
if the sample collected was representative of the plant's average
discharge. Representativeness was determined by comparing results
of these analyses with the last several months' "discharge moni-
toring reports" submitted by each plant to the State. This type
of information was necessary so that other Chesapeake Bay research-
ers could decide how to use the data when studying potential ef-
fects to the Bay from that discharge.
The philosophy of the decision analysis at this point was to col-
lect just enough data in order to 1) determine if the effluent was
toxic to aquatic species, 2) determine if there were organic com-
pounds present in the effluent which were bioaccumulative (and
thus enter the food chain), and 3) determine if there were signi-
ficant concentrations of organic compounds present. No "direct"
human health effects tests were included in this program because
no one drinks water from the Chesapeake Bay. Therefore, the only
way human health can be directly effected by discharges to the
Bay is through the food chain.
As a result of data collected in Phases I and II, and based on
MRC's experience with data from similar programs, a series of
decision points have been established and are shown in Figure 2-7.
In order to determine "is the effluent acutely toxic to aquatic
species," an EC50 value of 50% was selected. EC50 means the con-
centration of the effluent at which 50% of the test species have
the desired effect. For invertebrate and vertebrate species, the
measured "effect" was death of the species tested. An EC50 of
50% means half of the species tested died in a 50% (v/v) effluent/
diluent solution. This value was selected as a "starting point"
2-25
-------
for decision analysis and means if the effluent is diluted one-to-
one with the receiving stream, the acute aquatic toxic effects will
be significantly reduced.
The second primary decision point was "are there bioaccumulative
compounds present?" If bioaccumulative compounds were present,
then the goal was to determine what the compounds were, are they
toxic, and do they cause the aquatic toxicity (if EC50 <50%)? If
the number or concentration of these bioaccumulative compounds was
significant and could not be identified, then further, more sophis-
ticated analyses may be necessary. If, on the other hand, one
could identify the compounds as indicated in Figure 2-7, then no
further analyses were required.
The third primary decision point was "are there significant con-
centrations of organic compounds present in the effluent?" For
this decision, the TOC (total organic carbon) analysis was used.
Again, based on data from this program and other references, a
decision point value of TOC £50 mg/L was selected as a "starting
point."
It is critical to remember that the values selected for these
three decision points totally depend on the "user's needs." For
example, for second round NPDES permit issuance, many people
within EPA are conducting single, static, acute bioassays where
the species are placed in 100% effluent. If more than 80% of the
test species die, then they decide the effluent is toxic and rec-
ommend to proceed with further analyses.
If the cause of the acute aquatic toxicity could not be identified,
and/or the bioaccumulative compounds could not be identified, and/
or the TOC was above the criterion level, then there was a need
for further sample characterization. This Stage II set of inter-
mediate analyses involved wide-scan organics analysis by GC/MS
techniques. The goal of this effort was to identify as many of
2-26
-------
the organic compounds as possible within time and funding con-
straints, and determine if the compounds identified were the cause
of aquatic toxicity, compounds which were bioaccumulative, or tox-
ic to humans based on published toxicity data.
If the causes of the toxicity were then identified, a toxicity
reduction program could be recommended. However, if the above
effort could not identify the cause of the toxicity problem, a
third level of more sophisticated analyses must be considered.
Examples of the type of analyses which may be required include:
Flow-through aquatic bioassay - to determine acute toxicity
on samples which are more representative of the plant's
effluent, which include the volatile organics which are
lost in static tests, and which maintain a constant oxygen
level for samples that have a high oxygen demand.
Fractionation-Static aquatic bioassay - by fractionating
the effluent one can determine if the toxic response is
organic or inorganic in nature and determine if the tox-
icity falls into one of three organic fractions based on
polarity.
Microtox® analysis of selected process streams - the anal-
ysis of specific in-plant water streams before entry to the
treatment system by inexpensive techniques such as Microtox®
may determine the source of the toxicity within the plant.
An analysis of the process may then reveal the identify of
the toxicant.
Sediment sampling and analysis - analysis of sediments
collected near the outfall may make the identification of
low level toxicants easier due to the sediment acting as a
chemical concentrator.
2-27
-------
When the source and identity of the toxicity is determined, dis-
charge control decisions can be made and either production proc-
ess modifications and/or additional end-of-pipe treatment can be
implemented to reduce the toxic effect of the discharge to an
acceptable level.
2.4.3 Results of Application of the Scheme
At the conclusion of Phase II of this program, the decision anal-
ysis scheme (as shown in Figure 2-7) was proposed. Phase III
samples were then collected and the Stage I Basic Analyses were
performed. From these data the need for further sample analysis
was determined. Applying the decision criteria to the data re-
sulted in the recommendations given in Table 2-1. The metals
analysis data were not available from EPA at the time Table 2-1
was compiled, so the effects of toxic metals could not be incor-
porated into any decisions at that point. The specific details
of the decisions taken for each sample, and the resulting conclu-
sion from implementing the decision analysis is contained for
each plant in Appendix D under Data Interpretation.
Table 2-2 summarizes the results of applying the decision analy-
sis. From examining the table it can be seen that the decision
analysis was quite successful in narrowing down the potential
causes for toxicity and in providing clues about the types of
compounds which were potentially bi©accumulative. Information
gathered at the end of the Stage I and Stage II analyses used in
Phase III are adequate to indicate what type of control technol-
ogy modifications are necessary to influence effluent quality.
It appears that the only significant gap in the protocol is the
lack of a test to analyze the nonchromatographable organic por-
tion of some samples.
2-28
-------
TABLE 2-1. DECISION CRITERIA FOR PHASE III SAMPLES
Acute aquatic
toxlcity
96-hr
Plant code LC5(,
rO
1
rO
vO
Maryland:
B126S
B133S
B141S
B142S
B143S
B147S
B149S
C169S
Virginia:
A101
A109
B111D
B112D
B113D
B119D
B124D
C150D
C153D
C1S6D
C151D
C154D
C155D
C157D
C1S8D
C1S9D
C160D
C161D
C169D
C1640
u
<3b
41
<3
4
22
54b
<3b
12
>100
24
>100
BO
>100
>100C
59
48
>100
24
56
5
15
77
89
42
MOO
10
59
7
Rat ing
High
Moderate
High
High
Moderate
Low
High
Moderate
None
Moderate
None
Low
None
None
Low
Moderate
None
High
Low
High
High
Low
Low
High
None
High
Moderate
High
TOC.
mg/L
6
65
131
57.
15
20
50
5
2
46
56
IB
8
43
64
55
7
43
2
7
35
65
16.
18
2
60
65
29
GRAV.
mcj/L
0.6
12.4
30.1
5.9
0.5
4.2
31.7
1.2
1.8
1.8
4.3
10.7
2.1
6.5
4.9
6.3
0.5
4.1
1.8
0.5
5.0
1.1
4.6
2.7
0.9
20.4
9.4
4.6
Effluent
TSS. mg/l.
101
29
114
24
101
7.3
106
416
13
221
127
IS
42
19
9
42
231
31
19
359
25.3
6
23
7
5
69
33
12
nioac:cumulat inn
No. of
positives Comments
6
5
13
8
1
7
23
2
2
8
3
19
4
14
1
13
2
9
5
10
10
3
7
4
0
9
10
7
All <10 |.ig/L
1 Value 290 |ig/L
All <10 ug/L
11 >100 ug/L
1 Value 230 ug/L
11 >100 ug/L
2 >200 ug/L
1 Value »360 ug/L
(due to S8)
1 Value »360 ug/L
(due to S8)
All <20 ug/L
All <60 ug/L
All <10 ug/L
1 Value »360 ug/L
(due to SR)
2 >100 ug/L
other relevant analytical datn
TKN =
S0.,<
BODr, =
(N02 +
Cl =
Cl =
Cl =
so200 mg/L, TSS =114 mg/L
N03) = 565 mg/L. Cl = 2,140 mg/L
1,875 mg/L
7,600 mg/L
6,000 mg/L, S0«2 = 3.200 mg/L
1,230 mg/L
= 1,139 ug/L
92 mg/L
11,200 mg/L, S0«2 = 3,000 mg/L
= 125 ug/L
13,400 mg/L, S0«2 = 2.200 mg/L
72 mg/L
490 mg/L
141 mg/L
420 mg/L
RecommfMi'lat ions
Toxicity due to ions-Stop
GC/MS
GC/HS. Bioaccumulation
fractions' ion
GC/HS
Toxicity due to Cl -Stop
Stop
GC/MS, Bioacciimulation
fractional ion
Stop
Stop
GC/MS, Bioaccumulation
fractionation
Stop
Bioaccumulation
fractionation
Stop
GC/MS, Bioaccumulation
fractionation
Stop
GC/MS, Bioaccumulation
fractionation
Stop
GC/HS. Bioaccumulation
fractionation
Stop
Toxicity due to anions-Stop
GC/MS
Stop
Stop
Toxicity to freshwater
species due to Cl -Stop
Stop
GC/MS, Bioaccumulation
fractionation
See C159D, GC/MS
GC/MS
"Toxicity rating based on trend in LC50 value from 24-hr through 96-hr test and on final 96-hr LC5n.
100% mortality occurred in all test concentr.itions.
°Could not be determined because <50% mortality occurred in 100% effluent concentration.
^ioaccunulation fractionation means to collect fractions from the HPLC and analyze them by GC/MS to identify the compound.
-------
TABLE 2-2. EFFECTIVENESS OF DECISION ANALYSIS FOR PHASE III EFFLUENTS
II. ml.
oirlr
Aim
AlO'i
Bl 111-1
Bl I2D
BJ 1 3D
BJ 1 'JO
B124D
B126S
| B133S
U)
O
B141S
B142S
B143S
B147S
IU49S
C150D
C151D
C153L)
Rat i ng
NOIH-
Mn...•<:,
NM
Y<::;
No
Yes
No
Yes
No
No
No
Yes
No
No
No
Yes
Yes
No
No
lli
-------
TABLE 2-2 (continued)
I
CO
Pl.int
code
C154D
C1T>5D
C156D
C157D
C158D
C159D
C160D
C161D
C164D
C169D
C169S
• • Toxicit.y
Rnl.ing
High
High
Moderate
Low
Low
Moderate
None
High
High
Moderate
High
PosnibJe caiisn
Chromium, lead
Chlorine, ammo-
nia
Chlorine, metals
Not identified
Chlorine, metals
Ammonia, metals
_c
Chlorine, metals.
organics
Chlorine, acry-
lonitrile
Ammonia
Chromium, chlo-
rinated organ-
ics
High otgonic content"
Ycr./no
No
Y«r,
Yen
Yes
No
No
No
Yes
Yes
Yes
No
I'osr. i b 1 *• '.:
ommended for iden
tifying unknown
mass spectra
_c
The decision point for high organic content was 50 mg/L or greater.
The decision point for high bioaccumulative content was 20 M9/L or greater.
°Not applicable.
Further analysis not performed due to nontoxic nature of effluent.
eEIfluent was nontoxic to mysid shrimp but moderately toxic to fathead minnows.
-------
SECTION 3
PROGRAM APPROACH
The purpose of this section is to orient the reader to the over-
all approach taken in Phases I, II, and III of this Chesapeake
Bay toxicity assessment program. The results and conclusions
which ultimately produced the site-specific effluent screening
protocol (described in detail in Section 4) are also highlighted.
In addition, Phase III effluent and sediment analysis data, not
previously reported, are referenced in Section 3.3 (and pre-
sented in Volumes III and IV).
3.1 INTRODUCTION
At the beginning of this three-phase program, MRC was given two
objectives by the EPA. These were: (1) to comprehensively char-
acterize- industrial effluents in terms of chemical species pre-
sent at concentrations greater than 1 pg/L and their potential
toxic effect on the Bay ecosystem and to man, and (2) to develop
a chemical and biological analysis scheme which state water pol-
lution control authorities could use to collect sufficient data,
in a cost-effective manner, to screen discharges to the Chesapeake
Bay for the presence of toxic chemical species and gross toxicity.
To accomplish the above objectives, the program was designed as a
three-phased data feedback system. In Phase I, a chemical analyt-
ical and bioassay test scheme was implemented using 11 effluent
samples collected at 10 industrial sites (one site was sampled in
duplicate). Data from this research effort were then evaluated by
3-1
-------
the EPA Chesapeake Bay Program Office, state water pollution con-
trol authorities, and other researchers to determine how the type
and quality of data met the various program needs. The results
of this data analysis were then fed back to MRC, and an updated
chemical analytical and bioassay test program was designed and
implemented in Phase II for effluent samples collected at nine
additional industrial sites. In Phase III, the final data assess-
ment scheme was tested at 24 more sites, and MRC trained state
water pollution control authorities to set up and implement such
a toxicity assessment program. This broad-based approach was
designed to provide data which would enable a determination of
the most timely and cost-effective analytical path.
The sequence of principal activities in Phases I, II and III was
plant selection, process engineering analysis, field sampling,
chemical analysis and bioanalytical testing, and data interpre-
tation. The general philosophy and approach to plant selection
and process engineering analysis, consistent throughout the pro-
gram, are discussed in Sections 3.2 and 3.3, respectively. Field
sampling methodology, including minor variations from phase to
phase, is described in Section 3.4. Sections 3.5 through 3.7
present the chemical and biological analysis protocols employed
in Phases I through III, respectively. Data interpretation is
the subject of Appendix A and Subsection 3 of Sections D.I through
D.28 in Appendix D.
3.2 PLANT SELECTION
The first step in selecting the outfalls for effluent and sediment
sampling and analysis under this study was to determine the num-
ber, location, and identity of all wastewater dischargers to the
Chesapeake Bay. To accomplish this, EPA's Industrial Environmental
Research Laboratory in Research Triangle Park, North Carolina,
awarded EPA contract 68-02-2607 to GCA to provide an inven-
tory of dischargers in Maryland and Virginia. This inventory
3-2
-------
identified both major and minor dischargers and arranged the data
by individual river basins. A list of potentially toxic or haz-
ardous constituents known or presumed to be present in the indi-
vidual outfalls at each major source was also developed. These
chemical constituents were ranked on the basis of toxicity. An
indicator of the degree of hazard associated with each constitu-
ent was computed, based on flow, concentration, and toxicity data,
and then summed for each outfall. All major facilities having
similar effluent characteristics were listed, and the individual
outfalls within each category were prioritized on the basis of
this degree of hazard. This prioritized list provided the ration-
ale for GCA's selection of sources for future toxic/hazardous
waste effluent sampling studies [1].
For the purposes of MRC's Toxic Point Source Assessment contract,
each outfall was considered as a separate facility. Upon receipt
of the GCA report, MRC consulted its industry description data
bases (Table 3-1) to determine the outfalls that best satisfied
the following selection criteria: (a) representation of a broad
range of industry types; (b) effluent and sediment characteristics
that would challenge the analytical protocol; and (c) in most cases,
distance to the Bay of less than 50 miles. Table 3-2 is the re-
sulting coded listing (in alphanumeric order) of 44 industrial
facilities from which effluent and sediment samples were collected
in Phases I, II, and/or III. Table 3-3 is a summary listing of the
13 industry types from which 51 effluent samples and 30 sediment
samples were taken. ' . '
[1] Hopper, T. G., et al. Inventory and Toxicity Prioritization
of Industrial Facilities Discharging into the Chesapeake Bay
Basin, two volumes. Submitted to EPA by GCA/Technology Divi-
sion, GCA Corporation, EPA Contract 68-02-2607, Work Assign-
ment No. 30, August 29, 1979.
3-3
-------
TABLE 3-1. DATA BASES USED TO EVALUATE CHESAPEAKE BAY OUTFALLS
w
Data source title
Source Assessment Data Base
Prioritization of Stationary
Water Pollution Sources
Petrochemical Data Base
Status Assessment Data Base
Organic Chemical Producers
Data Base
Industrial Hastewater
Treatability Manual
Chesapeake Bay Program
Toxics Data Base
Computerized Data Base
developed by CCA
Industrial Process Profiles
for Environmental Use
Review of Organic Chemical
Processes for Potentially
Toxic Materials
Chemistry, Production, and
Toxicity of Selected Syn-
thetic Organic Chemicals
Assessment: Treatment and
Control of Wastes, Manu-
facture and Use of Toxic
and Hazardous Organic
Chemicals
Plant
location Compounds
data produced
X X
X X
X
X X
X
X X
X X
X X
X
X
X
Types of ri.it* contained
Toxicity,
Production caicinogpntcity,
process Chemistry mutagmicity Bioaccumulation
data data data data
XX X X
X
X X
XXX X
X
XX X
XXX X
XX X
X X X •
XXX X
XXX X
Source
prioritization Emission
data or methods data
X X
X
X
X
X X
X
X
X
X
X
(continued)
-------
TABLE 3-1 (continued)
_Tyj>er._of_djjto contained
10
Data source title
Toxicity.
Plant Production carcinogenirity. Source
location Compounds process Chemistry mutagenicity Bioaccumulalion prioritizntion Emission
•*-*- —-•«•••—-1 J-'- cl;ita data data data or methods dais
data
produced
data
File of NPDES Dischargers
Chemicals Which Have Been
Tested for Nontoxic
Effects
Registry of Toxic Effects
of Chemical Substances
Handling and Transport
Procedures for Industrial
Chemicals
MRC Microfiche File of
Government Reports
Computerized Toxicological
Data Systems
NIH-EPA Substructure
Search System
Handbook of Chemistry and
Physics
Hazardous Properties of
Industrial Chemicals
Merck Index
Kirk-Othmer Encyclopedia
Use Diagrams for Industrial
Organic Chemicals
Report on Municipal and
Industrial WWT Systems:
A Statistical Compendium
X
-------
TABLE 3-2. ITEMIZATION OF EFFLUENT AND SEDIMENT SAMPLES COLLECTED
U)
I
Plant
code
A100
A101
A102
A103
A104
A105
A106
A107
A108
A109
B111D
B112D
B113D
B119D
B124D
61255
B126S
B127S
B129S
B130S
B131S
B133S
B136S
B137S
B141S
B142S
B143S
B146S
B147S
B149S
C150D
C151D
C153D
C154b
C155&
C1S6D
C157D
C158D
C159D
C160D
C161D
C163D
C164t>
C169S
Industry
type
Pulp and paper
Synthetic fibers
Textile dyeing
Petroleum refining
Textile dyeing
Pulp and paper
Pulp and paper
Synthetic resins and fibers
Leather tanning
inorganic and organic chemicals
Tobacco processing
Wood preserving
Synthetic resins and fibers
Polyester film and plastics
Tobacco products
Synthetic detergent
Inorganic pigments
Iron and steel
Pesticides and plasticizers
Iron and steel
Publicly owned treatment works
Plasticizers
Coke production
Organic chemicals and surfactants
Publicly owned treatment works
Waste neutralization
Inorganic pigments
Catalyst and support manufacturing
Stainless steel manufacturing
Stainless steel manufacturing
Publicly owned treatment works
Electric power generation
Phosphate fertilizer
Ammonia fertilizer
Publicly owned treatment works
Publicly owned treatment works
Synthetic resins and fibers
Publicly owned treatment works
Meat processing
Fabricated rubber products
Publicly owned treatment works
Publicly owned treatment works
Publicly owned treatment works
Industrial inorganic chemicals
manufacturing
TOTALS
SUindrird industii.ii
cjnssj f jcfitinn
(SIC) code(s)
2621, 2631
2824, 2299
221
2911
221
2631
2611, 2621, 2631
2824, 2299
3111
2811, 2819, 2869
2621
2491
2821, 2824
3079
2111
2841
2816
3312
2821, 2879
3312
4952
2821
3312
2819, 2865, 2911
4952
4963
2816
2819
3312
3312
4952
4911
2874
2873
4952
4952
2821, 2824
4952
2011
3069
4952
4952
4952
2819
-
Phase 1 Phase II Phase
effluent'' Effluent Sediment Effluent
X
X X
Xb
Xb
X
xb
XD
X
X
X X
X
X
X
X
X
X
X
X X
X X
X X
X
X
XXX
X X
X
XXX
XXX
X X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
13 10 8 28
III
Sediment
X
X
X
X
X
X
X
X
X
X
X
X
\ „
X
X
X
X
X
X
X
X
X
X
22
No sediment samples collected during Phase I.
duplicate samples collected.
-------
TABLE 3-3. SAMPLE GROUPING BY INDUSTRY TYPE
Number of
samples collected
Industry type
Coal use .
Inorganic chemicals
Iron and steel
Leather tanning
Meat processing
Organic chemicals
Petroleum refining ,
Plastics and fibers
Publicly owned treatment works
Pulp and paper
Textiles
Tobacco processing
Wood preserving
Totals
Effluent
3
8
4
1
1
6
2
8
9
4
2
2
1
51
Sediment
2
8
2
0
1
4
0
5
6
0
0
1
1
30
alncludes coke and boiler ash quench water from elec-
tric power generation.
Includes inorganic fertilizers and waste neutrali-
zation.
clncludes detergents, pesticides, and combined
organic/inorganic chemical production.
Includes rubber plants.
3.3 PROCESS ENGINEERING ANALYSIS
One result which the EPA expressly desired, from MRC's efforts was
development of a technique for desk-top plant process engineering
analyses as a means of making logical decisions concerning prob-
able chemical species present in wastewater discharge or sediment
samples. Water pollution control authorities in Virginia and
Maryland would then be able to focus quantitative analytical ex-
periments on specific compounds of interest, rather than having
to conduct only a generic search for unknown compounds.
3-7
-------
Before actually embarking on visits to field sites, MRC performed
such an analysis of each production process to be sampled in order
to determine the possible pollutant species present at each spe-
cific plant. In Phase I, the data sources listed in Table 3-1, as
well as the engineering experience of personnel assigned to the
project, were used to determine processes employed, process descrip-
tions and configurations, raw materials used, intermediate and final
products, discharge sources and their treatment, and previous waste
stream characterization. These data sources also provided data on
toxic, mutagenic, carcinogenic, and teratogenic chemical species
suspected to be present in the effluent. In most cases, this ex-
ercise involved starting with the EPA Effluent Guidelines Division
(EGD) list of pollutants found during sampling of the general
industry category in question. This list was then supplemented
by the other data sources and evaluated with respect to specific
process details for individual plants.
Possible sources of process wastewater pollutants which were
considered included water formed as a product of combustion or
other process reactions, direct-contact cooling water, product
wash water, reactor washout wastes, condenser and scrubber water
that had contacted either products or reactants, noncontact cool-
ing water that may have been contaminated due to process leaks,
and secondary pollutants resulting from the wastewater treatment
operations. Possible background contamination due to the composi-
tion of plant intake waters was not considered; the program was
concerned only with the contribution of each point source to pol-
lutant loadings in the Chesapeake Bay as a result of wastewater
discharges.
The next step in the process engineering analysis was to deter-
mine the toxicity of the organic compounds in the "candidate
list" finally generated for each of the plants studied. Those
compounds suspected to be particularly toxic were to be determined
3-8
-------
semiquantitatiyely whenever possible. The toxicity evaluation
criteria used in Phases I and II were as follows:
Any lethality rating less than 500 mg/kg
Any identified carcinogenicity
Any identified mutagenicity
Any identified teratogenicity
Any known toxic decomposition products.
The remaining compounds on each list were to be identified by
mass spectrometry during subsequent sample analysis, but not
quantified.
After completing the desk-top evaluation of the plant, the next
step was to visit the site. The purpose of the site visit was to
verify the data and conclusions from the desk-top study, update
information for accuracy, plan field sampling logistics, and
define safety requirements.
Site visits were conducted by a two-man team from MRC with assis-
tance from state personnel responsible for wastewater permits for
the plant in question. The team generally met with key plant
personnel for a two-hour meeting followed by a tour of the plant.
At these meetings, process analysis data sheets such as those
presented in Table 3-4 were completed. These data sheets specified
the information needed by the sampling and analytical staff to
accurately and safely perform their activities at each individual
plant.
The data sheets are divided into seven sections, each of which
provides for specific items of necessary information. Section I
records the name, address, and phone number of the plant, and the
names and phone numbers of others attending the meeting; these
are recorded to facilitate communication between MRC and the plant.
(Note: All plant identification information on the actual form
was replaced in the published reports by a plant code number to
3-9
-------
TABLE 3-4. PROCESS ANALYSIS DATA SHEETS
I. NAME OF COMPANY_
ADDRESS
DATE OF
SUMMARY
PHONE
NAME OF CONTACTS
MRC PERSONNEL
EPA PERSONNEL
STATE PERSONNEL
INDUSTRY TYPE
PHONE
PHONE
PHONE
PHONE
PHONE
PHONE
PORTION OF PROCESS TO BE SAMPLED
II. PROCESS DESCRIPTION
3-10
-------
TABLE 3-4 (continued)
II. Con't.
Raw materials and amounts
Fuels
Products and amounts
Operating Cycle:
Check: Batch Continuous Cyclic_
Timing of batch or cycle
Best time to sample
Length of Operating day
Length of operating week
Scheduled shutdowns
Other
III. WASTEWATER TREATMENT PLANT DESCRIPTION:
Chemicals added and amounts
Handles rainfall runoff?
Includes sanitary waste, flow
Source of plant intake water
Hydraulic retention time: Thru plant
Thru treatment
unit operations
Recent treatment plant performance
3-11
-------
TABLE 3-4 (continued)
III. Con't.
NPDES permit parameters and limits
Final effluent flow rate
List of potential pollutants
Recent analyses available?
{Sampling point description
' Use automatic sampler?
Electricity available
Extension cord and type of outlet?
IV. Safety Checklist
A. Personnel Protection Equipment (check If required)
Item
Safety glasses
Goggles
Side shields
Face shields
Hard hats
Ear plugs
Safety shoes
Life belt
Ladder climbing
device
Plant
MRC
/
Item
Dust masks
Vapor masks
Air purifying
Air supply
Air packs
Chem. res't clothes
Heat res't clothes
Chem. res't gloves
Heat res't gloves
First aid
Plant
MRC
3-12
-------
TABLE 3-4 (continued)
B. SAMPLE SITE
1. Smoking restrictions
2. Vehicle traffic rules
3- Possible set-up/clean-up facilities?_
4 . Evacuation procedures
5. Alarms
6. Hospital location
7. Hospital Phone
Emergency Numbers
V. Plant Entry
A. Plant Requirements
Special time constraints:
E. KRC Agreement
C. Potential Problems
3-13
-------
TABLE 3-4 (continued)
VI. SAMPLING HANDLING
A. Ice availability
B. Sample splitting requested
Describe
C. Nearest airport:
D. Chemical available: H-SO*
HN03 _
NaOH
3-14
-------
TABLE 3-4 (continued)
VIL Field Test Schedule
^X. Time
Day ^v
Sunday
Monday
Tuesday
Wednesday
Thursday
Friday
Saturday
AM
.
PM
3-15
-------
protect the plant's confidentiality for this research effort.)
Other information recorded in this section included the general
industry type and the sampling point. The industry type normally
consisted of a Standard Industrial Code (SIC) description which
was further clarified to identify the specific industry the plant
site is associated with. The portion of the process to be sampled
was normally an NPDES discharge number for that plant site.
Section II further identified the specific industry type by de- .
scribing the major unit operations or processes used. This sec-
tion allowed MRC and CBP personnel to compare similar types of
plants while also noting why each plant site is different. Gen-
erally, a step-by-step description was not necessary because it
tended to confuse the true differences and could potentially have
violated plant confidentiality by calling attention to peculiar
operations.
To assist in potential pollutant identification, a water flow
diagram was requested which noted each general process using
water in some manner, the type of exposure (i.e., contact or non-
contact), and the points at which water was discharged to the
treatment facility. Water quantities were not necessary except
in cases where the noncontact water greatly exceeded the contact
water quantities. Also collected in this section was information
on raw materials, fuels, products, and plant operating cycles.
Raw materials, fuels, and products were requested to assist in
identifying potential pollutants that might enter the wastewater
stream. Information on intermediates and secondary products were
also requested. The operating cycle was requested to assist in
determining necessary sampling deviations and logistical planning.
Section III was used to identify and generally describe the type
of wastewater treatment facility used at the selected plants.
These descriptions were obtained to allow a later analysis to
determine whether current treatment practices were generally
3-16
-------
effective on toxic pollutants and to determine the effectiveness
of BAT and BPT treatment technologies in protecting the bay area.
Because of the nature of these evaluations, fairly detailed treat-
ment descriptions were necessary. Information such as types of
wastewater treated, chemicals added, and whether or not the treat-
ment plant handled runoff and sanitary waste was necessary to
evaluate potential pollutants and the quantity of solids present.
Retention time was requested to permit an evaluation by MRC's
engineering staff of the probability that process upsets would
affect the treatment system and thereby affect the analytical
results. These upsets were determined by comparison with the
NPDES permit parameters. The final effluent flow rate was esti-
mated during the presurvey to ensure that the flow was sufficient
for sampling. The sampling point was described in detail here to
inform the sampling crew of the exact point for taking the com-
posite samples. The remaining information in Section III was to
assist the sampling crew in determining the equipment needed for
the sampling procedure.
Section IV dealt with the safety checklist and other safety fea-
tures necessary under all MRC sampling jobs. Plant personnel
instructed the presurvey and sampling crew of the safety equip-
ment necessary and the actions to be taken in case of an
emergency.
Section V informed the presurvey and sampling teams of plant
requirements for plant entry during normal business hours and
after hours. Necessary security agreements were recorded here
and any potential problems were noted. If the plant requested
constant observation by plant personnel, this fact was noted
here, and the sampling crew was informed later whom to report to
upon arrival at the plant.
3-17
-------
Section VI discussed sample handling by the sampling crew, be-
cause of the necessity to adequately preserve the samples and to
ship them quickly. Requested sample splitting was recorded here
to inform the sampling crew of any increase in sample volume
needed.
Section VII contained the tentative field test schedule agreed to.
The time for the sampling crew to arrive at the plant and the
approximate time for the night equipment check were recorded
along with any specific instruction regarding the schedule.
All information on the presurvey data sheets was requested at the
presurvey meetings. Although no information was mandatory, a
spirit of cooperation between plant, CBP, and MRC personnel was
suggested in an effort to get maximum participation. If a facil-
ity felt that information was company confidential or might vio-
late the plant's confidentiality, this was noted by the presurvey
team for use on the program but the information was not written
on the presurvey data sheets.
3.4 FIELD SAMPLING METHODOLOGY
Effluent sampling techniques during Phase I were designed to col-
lect "representative" samples; that is, samples typical of those
ordinarily obtained during routine NPDES monitoring. Sampling
during Phase I was conducted by a pair of two-man sampling crews.
Each crew visited five plant sites during a 2-week period in
December 1979. The crews collected two 24-hour composite sam-
ples and six 40-mL grab samples from each final effluent stream.
Sampling was performed in accordance with EPA-approved proced-
ures [2].
[2] Draft Final Report: Sampling and Analysis Procedures for
Screening of Industrial Effluents for Priority Pollutants.
U.S. Environmental Protection Agency, Cincinnati, Ohio,
April 1977.
3-18
-------
The two 24-hour composite samples were collected by Isco auto-
matic samplers, modified to sample intermittently for an extended
period of time. Manning automatic samplers were also used at
some sites. Two separate composite samples were taken because of
the nature of the collection vessels required and the types of
analyses to be performed. One sample was collected in a 190-liter
(50-gallon) plastic container for use in the bioanalytical work,
which required a large amount of sample. Tests run on this com-
posited sample included the ecological and health effects tests,
as well as the monitored NPDES parameters and the inorganic anal-
yses. The second composited sample was collected in a 19-liter
(5-gallon) glass container and was used for organic compound,
total organic carbon (TOC), and total phenol analyses. Segrega-
tion of the samples in different types of containers minimized
potential contamination (such as plasticizers or leached metals)
from the collection vessels. The glass vessel was generally packed
in ice during the sampling period to reduce the possibility of
loss of volatile components. The composite sampling was started
by mid-morning and extended for 23 to 25 hours.
Six 40-mL grab samples were used for determination of volatile
organic compounds (VOC). These samples were collected at various
times during the 24-hour period. A typical schedule for collec-
tion of VOC samples was as follows:
On arrival at the sampling site (8-10 a.m., first day)
After lunch (12-2 p.m., first day)
Before leaving for supper (4-6 p.m., first day)
During the night equipment check (8-10 p.m., first day)
On arrival at the sampling site (7-9 a.m., second day)
Before or during splitting and packing of samples
(10-12 a.m., second day).
Grab samples were taken in several ways, depending on the specif-
ic circumstances at each sample site. The most common method was
to collect a sample in a Teflon-lined bucket and then fill a 40 mL
vial by completely immersing it in the bucket. Other methods
3-19
-------
used include direct immersion in the effluent stream and filling
a vial at a plant wastewater sampling port. Samples were hermet-
ically sealed immediately after sampling, labeled, and stored at
4°C until shipment. The vials were shipped in ice to maintain
this temperature.
In general, the field sampling procedures in Phase I satisfied
the program objectives in terms of being both straightforward and
readily adaptable to a variety of circumstances. At one site,
however, the composite sample was exposed to an unexpectedly high
ambient temperature (^18°C in December) and to direct sunlight.
This incident had only a minor adverse impact on evaluation of
the wastewater analysis protocol. Nevertheless, to prevent vol-
atilization or photodegradation, MRC fabricated the equipment
shown in Figure 3-1 to collect effluent samples while shielding
out sunlight and maintaining a temperature of 4°C. This appa-
ratus was proposed, but never used, for Phase III collection of
24-hour composite samples. Furthermore, as a cost-saving measure
at EPA's request, Phase II samples from the 10 outfalls were com-
posited over periods of 3 rather than 24 hours. Sediment sam-
ples were collected adjacent to eight of these outfalls during
Phase II (the two sites not sampled for sediment discharged into
separate storm sewers whose final outlet locations were unknown).
100-6AOON TANK
SAMPLE
DISTRIBUTION
WASTE
WATER
Figure 3-1.
Fifty-gallon composite sample collection
system proposed for use in Phase II.
3-2O
-------
The principal thrust of Phase III was to collect samples that
would test the ability of the analytical protocol resulting from
Phase II to characterize wastewater discharges. Grab samples
were satisfactory for this purpose, as opposed to the 24-hour
composites taken in Phase I, to assure "representativeness." As
part of their'overall training in site-specific toxicity assess-
ment, state personnel from Maryland and Virginia assisted MRC in
the collection of the Phase III effluent and sediment samples.
i
3.5 PHASE I: SCREENING STUDY
3.5.1 Phase I Analysis Scheme
3.5.1.1 Chemical Analysis Protocol—
The chemical analytical scheme used throughout Phase I for each
of the effluent samples is shown in Figure 3-2. This approach
was patterned after EPA's Level 1 analytical scheme [3], supple-
mented with techniques to identify specific organic compounds and
to provide mass spectral "fingerprints" of unknown compounds. The
"fingerprint" information may help to identify possible sources of
unknown organic compounds found in the tissues of Chesapeake Bay
oysters.
3.5.1.2 Biological Analysis Protocol—
The bioanalytical testing protocol used throughout Phase I for
effluent samples is presented in Figure 3-3. The purpose of the
testing program was to: (1) determine potential toxic effects to
aquatic organisms due to exposure to the effluent, (2) measure
the presence of chemical mutagens (presumptive carcinogens) either
in the liquid or solid phase of the effluent, (3) measure the
toxicity to mammalian cells of chemicals in the liquid or solid
[3] Lentzen, D. E., et al. IERL/RTP Procedures Manual: Level 1
Environmental Assessment (Second Edition). EPA-600/7-78-201,
U.S. Environmental Protection Agency, Research Triangle Park,
North Carolina, February 1979.
3-21
-------
PERFORM ORGANIC
ANALYSIS
GC/MS-
i.
PERFORM INORGANIC
ANALYSIS
U)
N>
N>
ORGANI C-*— CH.CI, EXTRACTION
PHASE
.,
Z Z
ATpH2
f* ORGANIC ——CH.CI, EXTRACTION
* *ATpH12
CONCENTRATE
*_
COMBINE
PHASE
CONCENTRATE
*
GC/MS
DIP-MS
SPECIFIC
TOXICS
TCO
GRAV
1C FRACTIONATION
I
GC/MS
DIP-MS
SPECIFIC
TOXICS
AQUEOUS PHASE
SPARGE WITH N2
TOC ANALYSIS
SAVE AT ««C
I 1 I | I I I
TCO AND GRAV: GC/MS AND DIP - MS
ANALYSIS OF EACH FRACTION
IR OF FRACTIONS 5.61.1
TOC
ANIONS BY 1C
SSMS
HgBYAA
DIGEST SSMS
I
Hj ANALYSIS
BYAA
Figure 3-2. Phase I chemical analysis protocol
-------
1
SHEEPSHIAD
MINNOW
ASSAY
1 a
DAPHNIA
ASSAY
1
AQUATIC BIOASSAYS
1
MARINE8 FRESHWATER
ALGAL ALGAL
ASSAY ASSAY
1 .
MYSIO
SHRIMP
ASSAY
iNnns
[fin
1
[*™J*jJ TOTAL ORGANIC
ACCAV CARBON 110 n.L)
AbbAY )TOO
TRIAL
II NT
(3.800mL)
CHO CYCTOTOXICITYON
NEAT SAMPLE USING
ANTIBIOTICS
(DOSE RESPONSEI
RCRA
BIOACCUMUI.ATION
POTENTIAL
FILTRATION
STERILISATION 0.2(jm(30mU i PARTICULATES > 5|im(1.000fnL)
10
I
to
CO
FILTRATE 1100 mL EACH)
I
AMES
MUTAGENICITY
(SPOT ASSAY.
TOXICITY ASSAY.
PLATE INCORPORATION-
DOSE RESPONSEI
ACUTE CYTOTOXICITY
(CHO CLONAL-
DOSE RESPONSEI
SOXHLET EXTRACTION OF FILTERS
(125 mL OF ACETONE!
CONCENTRATE
I
EXTRACT (20 mL)
t
AMES
MUTAGENICITY
(THREE CONCENTRATION
DOSE RESPONSEI
CHO ACUTE
CYTOTOXICITY
(SPOT ILST)
aTEST PERFORMED ON ONLY TWO SAMPLES
Figure 3-3. Bioassay protocol used for Phase I.
-------
phase, and (4) determine the presence of chemical compounds which
have a high potential for bioaccumulation in fatty animal tissue.
All tests were performed in accordance with EPA's Level 1 bioassay
procedures [4].
3.5.2 Phase I; Screening Study Results
Part of the objective of the analytical protocol development was
to conduct an engineering evaluation of the process technology
used at specific industrial sites to determine what chemical spe-
cies are known to be or suspected to be present in the effluent.
Based on this engineering assessment a more comprehensive, cost-
effective chemical analytical and bioanalytical testing program
could then be implemented. Once the Phase I plants were approved
by EPA, MRC conducted such an engineering evaluation of the spe-
cific plants selected. The chemical species known to be present
or suspected to be present in each outfall were listed, and those
on the list which were known to be toxic or suspected to be toxic
were specifically highlighted. The chemical analytical scheme
used in Phase I was structured, in part, to specifically search
for this list of chemicals.
In summary, the engineering assessment of the plant specific proc-
ess technology did result in a list of 5 to 20 chemical species
known or suspected to be present in the effluents. This site-
specific list contained many raw materials and process chemicals
which were not included in the list of EPA 129 priority pollut-
ants . The analytical scheme was not sufficiently flexible to in-
corporate additional methods (such as direct water injectables or
derivatization) needed to better identify some of the compounds
[4] Duke, K. M., M. E. Davis, and A. J. Dennis. IERL/RTP Proce-
dures Manual: Level 1 Environmental Assessment Biological
Test for Pilot Studies. EPA-600/7-77-043, U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina,
April 1977.
3-24
-------
on the list of chemicals. The recommended Phase II protocol was
modified to include these changes.
In the field, 24-hr composited samples were collected at 9 of the
10 Phase I sites according to EPA approved methods [5]. A 12-hr
composite sample was collected at Plant A108 because the plant
chose not to cooperate in this program, and the state water pollu-
tion control authority elected to collect the sample at a logis-
tically difficult site at the point where the effluent entered
the receiving stream.
All samples in Phase I were collected in appropriate containers,
preserved when necessary, and shipped to the appropriate labora-
tories for analysis.
At one site the composite sample was exposed to direct sunlight
and an unexpected higher ambient temperature in December of ^18°C.
In terms of evaluating an analytical protocol for a variety of
wastewater matrices, this error in sampling had a minor, adverse
impact. In Phase II, however, an improved field sampling protocol
was developed to insure that the samples were shielded from the
direct sunlight and kept at 4°C during the 24-hr compositing period.
Aliguots of each composited sample were sent to EPA's Region III
Annapolis Field Office (AFO) for analysis of the NPDES permit
parameters (such as BOD5, COD, TSS, and metals), total cyanide
and total phenol. TThese analyses were performed to help evaluate
the representativeness of the sample collected with respect to
the plant's normal operating conditions. Results of this series
of analyses did provide good information for evaluating the cur-.
rent operating status at each plant sampled.
[5] Draft Final Report: Sampling and Analysis Procedures for
Screening of Industrial Effluents for Priority Pollutants.
U.S. Environmental Protection Agency, Cincinnati, Ohio,
April 1977.
3-25
-------
A 1-L aliquot was collected at each plant and filtered at MRC
through a 0.45 pm nitrocellulose filter. This type of filter was
selected to minimize background interferences during the metals
analysis. The resulting solid and liquid phases were analyzed
for (1) 76 elements by spark source mass spectrometry (SSMS),
(2) mercury by atomic absorption, and (3) selected anions (chlo-
ride, fluoride, nitrite, nitrate, sulfite, sulfate, and phosphate)
by ion chromatography. Results of the ion chromatography analyses
indicated that high concentrations (>10 mg/L) of chloride and
sulfite anions in the 10 effluent samples prevented the detection
of the remaining anions.
SSMS analyses were performed in order to screen the samples for
the presence of elements which are not normally analyzed in waste-
water samples, but could be present due to the specific industrial
process (such as rare earth elements used in catalysts). The
results of Phase I indicated, however, that no elements beyond
the usual 26 heavy metals detected by ICAP analyses were present
in concentrations significantly above their detection limits.
Therefore, in Phase II metals analyses were performed by ICAP
instead of SSMS.
Of particular significance, performing metals analysis of both
the solid and liquid phases provided very valuable data. It was
observed at many plants that metals (such as beryllium, cadmium,
mercury, and silver) were predominately associated with the solid
phase. These data are important because other research efforts
associated with the Chesapeake Bay Program are measuring sediment
composition, migration, and toxicity.
The organic compound analysis scheme used in Phase I was struc-
tured to identify the major (>10 pg/L) compounds present in each
effluent. Those organic compounds which were volatile were ana-
lyzed by the typical Bellar purge and trap technique. A 5-mL
aliquot of the sample was purged with helium and the volatile
3-26
-------
compounds adsorbed on a Tenax-GC/silica gel packed column. The
sample on the column was then thermally desorbed, and the vapor
went into a Hewlett-Packard 5985-A coupled gas chromatograph/
mass spectrometer (GC/MS) with a 5934 computer data system.
In the Phase I samples, the principal (>1 pg/L) volatile organic
compounds detected were benzene, chloroform, ethylbenzene, tetra-
chlorobenzene, toluene, trichlorobenzene, and xylene.
A 10-L aliquot of each effluent sample was analyzed for those
organic compounds which were methylene chloride extractable. The
10-L neat sample was first acidified to pH 2 with 12M HCl and ex-
tracted twice with pesticide grade methylene chloride. The organic
phase was retained, and the aqueous phase was basified to pH 12
with concentrated NaOH and extracted twice with methylene chlo-
ride. Each organic phase extract was then concentrated to 10 mL
using a Kuderna-Danish evaporator with a 3-ball Snyder column.
In Phase I, an attempt was made to monitor the organic pollutants
due to methylene chloride extraction, by using total organic car-
bon (TOC) analyses of the neat sample and of the aqueous phase
after extraction. Even after sparging the aqueous phase with
nitrogen for 30 min, there was still enough soluble methylene
chloride present to give a positive interference. Therefore, it
was not possible to directly determine what portion of the organ-
ics in the effluent were methylene chloride extractable.
An aliquot of each concentrated methylene chloride extract was
analyzed by GC/MS for those specific compounds identified in the
plant presurvey and to collect data to determine if selected or-
ganic compounds would be lost in a subsequent fractionation scheme.
Results of this GC/MS analysis indicated that 30% to 40% of the
compounds listed from the presurvey were detected in each sample.
Data also indicated that several of the organic compounds present
at <1 M9/L were not detected after sample fractionation.
3-27
-------
The total mass of methylene chloride extractable organics was
determined by measuring the amount of total chromatographable
organics (TCO) and nonchromatographable organics (as measured by
a gravimetric (GRAV) analysis [3]. Of particular significance,
results of this analysis indicated that 50% to 90% of the methylene
chloride extractable organic mass was in the GRAV fraction, which
means it could not be analyzed by GC/MS.
According to EPA Level 1 procedures, if the sum of the TCO and
GRAV analysis is greater than or equal to 15 mg of organics, the
sample may be sufficiently complex that a 7-step fractionation
should be performed prior to chemical analysis [3]. Of the 10
Phase I effluent samples, 6 had a total organic mass £15 mg and
thus required fractionation.
In the 7-step liquid chromatographic fractionation scheme, the
extract was charged onto a glass column (200 mm x 10.5 mm I.D.)
packed with activated silica gel. Seven solvents ranging in po-
larity from 100% pentane to 50% (v/v) methanol in methylene chlo-
ride were passed separately through the column, and the resulting
7 solutions were collected. Each fraction was then subjected to
TCO and GRAV analysis to determine organic recovery efficiency.
In summary, recovery efficiencies for each effluent sample ranged
from 50% to 100%.
Each of the 7 fractions (or the single sample from the unfraction-
ated samples) was spiked with an internal standard (d10-anthracene)
and analyzed by GC/MS for the chromatographable organics) and di-
rect introduction probe (DIP) - MS (for the nonchromatographable
organics).
MRC's Hewlett-Packard 5985 GC/MS system with a dual EI/CI source
and interfaced with a HP 5934 data system was used. The mass
spectrometer was operated in the electron impact (El) mode, and
spectra were obtained by continuously scanning under control of
3-28
-------
the data system for both chromatographable and nonchromatographable
organic analyses.
Chromatographable organics (i.e., ethylbenzene to dibenzopyrenes)
were studied by GC/MS using a glass column (2 mm I.D. x 180 cm)
packed with 3% SP-2100 (a methyl silicone fluid) on 80/100 mesh
Supelcoport and programmed at 0°C for 4 min, then increased at
16°C/min and held at 300°C for up to 20 min. Approximate quanti-
tation for certain toxic substances of particular interest was
based on external standards made with the compounds of interest.
Nonchromatographable organics were determined using direct intro-
duction probe-MS. Aliguots of sample, usually 3 jjL to 5 pL, were
placed in the capillary of the probe and allowed to evaporate. The
probe was then inserted into the mass spectrometer where an effort
to achieve satisfactory fractionation was made by employing a tem-
perature programmed heating. Nonchromatographable organics were
identified using the mass spectra produced, deleting the compounds
previously identified by the GC/MS technique. In many cases no
resolution of individual species could be provided; therefore,
only a general characterization of individual species may have
been possible.
All mass spectra generated for both samples and quality control
were stored on 9 track IBM compatible magnetic tape through
Hewlett-Packard 5934 Data System Programs DUMP/LOAD/UTILITY,
and DPUNCH.
These analyses resulted in the accumulation of thousands of spec-
tra. These were then compared to those published in the Eight
Peak Index [6].
[6] Eight Peak Index of Mass Spectra, Vol. Ill, Second Edition,
Mass Spectrometry Data Center, AWRE, Aldermaston, Reading,
United Kingdom, 1974.
3-29
-------
In summary, results of the 10 Phase I samples showed low concen-
trations of total organics and that no specific chromatographable
organic compound was present in concentrations greater than 1 pg/L.
Several organic compounds were detected by GC/MS in concentrations .
less than 1 pg/L. The Level 1 GRAV analysis, in turn, showed that
the majority of the organics in the Phase I samples had a boiling
point higher than 300°C. These high molecular weight species were
generally nonchromatographable. This observation thus points out
the need for more specific analysis for methylene chloride extract-
able nonchromatographable organic species, if total characteriza-
tion of organics is desired.
Of the organic species that were identified, none was found in
concentrations that would be responsible for the toxic effects
detected in the bioassay tests.
The bioanalytical testing protocol used throughout Phase I for
the 10 effluent samples is presented in Figure 3-3. For the eight
industrial sites which discharged into a freshwater receiving
stream, the following test species were used for static bioassays:
fathead minnows (Pimephales promelas), daphnide (Daphnia magna),
and algae (Selenastrun capriconatum). For the one industrial
site which discharged into a marine environment, the following
test species were used: sheepshead minnow (Cyprinodon variegatus),
mysid shrimp (Mysidopsis bahia), and algae (Skeletonema costatum).
And for the one site discharging into an esturine system, both
freshwater and marine static aquatic toxicity tests were performed.
All of the static aquatic toxicology tests were performed in
accordance with EPA's Level 1 bioassay procedures and were per-
formed for MRC by EG&G Bionomics Laboratories in Wareham, MA, and
Pensacola, FL [4]. Of the 10 effluents, 8 showed toxicity or
inhibitory/stimulatory effects in at least one test system.
3-30
-------
In order to interpret all the aquatic toxicity data, the level of
toxicity observed was compared to a predicted instream waste con-
centration or IWC. The IWC is a mathematical expression (used
by the EPA in the NPDES Bi©monitoring Compliance Manual) which
estimates the amount of dilution that an effluent discharge may
be expected to undergo under the condition of continuous release
into a receiving stream.
IWC % = Qr gWQw x 100 x 0.01
Where Qw = volume of the discharge in MGD, Qr = the 7-day, 10-year
low flow volume of the receiving water in MGD, and the 0.01 is a
safety factor currently being used by the EPA in all IWC calcula-
tions for issuance of NPDES permits. This percent dilution value
can then be compared to the percent effluent solution that caused
the measured toxic effect in the aquatic toxicity tests.
For the 10 effluent samples, the IWC was used to calculate the
receiving stream flow rate, Qr, that would result in an instream
concentration equal to or greater than the toxicity concentration
of each test species. For example, with plant A101, the growth
of freshwater algae was inhibited at an effluent solution of 11%.
Therefore, if the IWC is to be greater than 11% and its plant has
a discharge rate of 0.48 MGD, then Qr =400 MGD. Hence, if the
receiving stream that plant A101 discharges into has less than a
400 MGD flow, then the effluent concentration in the river has a
potential for inhibiting algal growth.
Larger Qr's mean more water is required to dilute the potential
adverse impact on the receiving stream. In summary, the Qr val-
ues, using the most toxic response or algal stimulation (eutroph-
ication) of the 3 aquatic test species, were calculated for the
10 plants and are listed below in descending order of magnitude:
3-31
-------
Plant
107
103
109
101
105
104
102
108
100
106
Lowest toxicity
value, %
8a
7a
6
11
61
32a
39a
13a
>100
>100
Discharge
(Qw), MGD
5.9
1.0
0.7
0.5
5.0
1.0
1.2
0.2
22.7
10.7
Qr, MGD
6,800
1,300
1,100
400
320
210
190
130
-
—
aBased on algal stimulation.
Therefore, this relationship takes into account toxicity and vol-
ume of discharge. This relationship will be used to determine
which plants have the highest potential for adversely impacting
the receiving stream.
The Ames mutagenicity bioassay was performed on the filtrate and
suspended solids phases of each effluent to determine if a chemi-
cal mutagen (presumptive carcinogen) was present. One liter of
each sample was filtered through a polyester drain disc and then
a 5.0 urn Teflon filter to separate solids from the aqueous phase.
The filtrate was then filter sterilized through a 0.45 pm filter
to remove further microbial activity. The Ames/Salmonella tests
were performed using MRC's modified Level 1 protocol, which in-
cluded toxicity/viability tests, spot tests, and the plate incor-
poration tests using triplicate platings at six filtrate concen-
trations with four strains of Salmonella bacteria (TA98, TA100,
TA1535, and TA1537), with and without a microsomal activation
system. For quality assurance, negative controls included three
replicate solvent controls and one high volume water control
(1,000 uL) per batch of samples. Five chemical compounds known
to be mutagenic in the test system were used as positive controls
3-32
-------
The particulate phase collected on the filters was Soxhlet ex-
tractated in acetone and the resulting solution concentrated to
10 mL. This solution was then submitted for an Ames mutagenicity
test using only the plate incorporation assay.
Results of the Ames tests indicated one repetitive positive muta-
genicity response. The neat aqueous phase of Plant A108 (leather
tannery) was mutagenic to Salmonella strain TA1535 and did not
require metabolic activation.
The above aqueous and extracted particulate phases were also
tested for acute cytotoxicity using the Chinese Hamster Ovary
(CHO-K1) clonal assay procedure. The clonal assay test is the
only health effects test for acute cytotoxicity used in the
Chesapeake Bay Phase I evaluation and therefore extends the tox-
icity data from ecological tests using nonmammalian organisms to
a mammalian cell test system. The test is valuable for the eval-
uation of aqueous effluent, samples, and suspended particulate
extracts because of the systems high sensitivity to low concen-
trations of known toxic compounds, good dose response character-
istics, good response to complex environmental mixtures, and good
repeatability.
Approximately 300 cells were cultured in 25-mL culture flasks un-
til they were attached to the bottom of the flast (^24 hr). Trip-
licate culture flasks of cells at each dose level were exposed to
the test sample for 24 to 48 hours. The surviving viable cells
were permitted to replicate for 6 to 8 days into discernible
clones (colonies) for counting. The medium was removed and the
clones counted by an automatic counter. The percentage of sur-
viving exposed cells was determined by comparing the number of
test cells with control cultures.
Initial range finding tests were conducted by using six concen-
trations of the test material to expose the cultured cells to at
3-33
-------
least three orders of magnitude change in concentration. After a
gross toxicity range was found, a narrow range of concentrations
was used for the definitive determination of an EC50; i.e., the
concentration at which 50% of the cells survive. The ECSO data
were compared with the response of known standard materials
tested by the clonal assay and live animal tests.
Results of the CHO cytotoxicity tests were particularly signifi-
cant because the test system was very sensitive to acute toxic
effects of the samples and because the toxicity was present in the
liquid and suspended solids phase, depending on the sample. Data
that can show specific toxicity associated with particulates are
useful to the other researchers in the program studying sediments
in the bay.
Numerous refinements to the testing protocol for Ames mutagenicity
and CHO clonal assay for both solid and liquid phases were defined
in Phase I, but are too numerous to mention in the summary. The
most significant improvement recommended for inclusion in Phase II
deals with the Ames test.
In general, the Ames test can only detect a chemical mutagen if
it is present at greater than 100 pg/L in the effluent. From a
health effects point of view, it is necessary to detect chemical
mutagens at least as low as 1 pg/L. Therefore, it was recommended
that each effluent sample be concentrated in order to increase
the detection limits of the Ames test. The CHO assay is suffi-
ciently sensitive that sample concentration is not necessary to
detect mammalian acute toxic effects in the neat/effluent sample.
Aliquots of each industrial effluent were also tested to deter-
mine the boaccumulation potential of organic compounds. This
analysis is particularly useful because it is the only test which
provides data on compounds which may be accumulated in the food
3-34
-------
chain. The resulting data are useful to Dr. Huggett who is ana-
lyzing compounds in oyster tissues.
The bioaccumulation potential test method was defined by the EPA,
and it estimates the octanol/water partition coefficient (meas-
ured as log P) of compounds present in the sample. Samples were
analyzed with a high-pressure liquid chromatograph (HPLC) in
which the organic compounds elute in order of hydrophilicity in
proportion to their hydrocarbon/water partition coefficient. The
EPA defines a positive response as any compound which produces a
log P £3.5 at an instrumental response £25% full-scale deflection.
Following EPA's bioaccumulation potential protocol, the 10 Phase I
samples produced data ranging from no positive responses to 11
positive responses, depending on the sample tested. An average of
two compounds per effluent were detected as having a log P £3.5.
In addition to using the positive/negative criteria, this test,
coupled with the chemical analytical data, can produce signifi-
cantly more information about potential bioaccumulative materials.
Since the strip chart output from the HPLC is simply a chromato-
gram, the evolution order is a function of the compound's polarity.
Using response data from the six standards used to calibrate the
instrument, published log P values, and results of the chemical
analyses, it is possible to tentatively identify the compound .pro-
ducing the positive response. From this identification one can
then run standards with the compound to accurately determine if
in fact, that is the compound causing the positive response. It
was recommended that this data analysis and verification of com-
pound identification be performed in Phase II of the program.
3-35
-------
3.6 PHASE II: PROTOCOL DEVELOPMENT STUDY
3.6.1 Phase II Analysis Scheme
The chemical analysis scheme implemented in Phase II was designed
to collect sufficient data to screen the effluent and sediment
samples for the presence of chemicals known or suspected to be
present, and to identify as many of the other compounds as pos-
sible within time and budget constraints. All aquatic and health
effects bioassay testing was deleted from Phase II in order that
more funds could be devoted to analysis of the chemistry of
wastewaters being discharged into the Bay. Capillary column
chromatography was used in Phase II to enhance the identification
of unknown compounds. It provides better mass spectra than were
obtained in Phase I.
The objectives of the Phase II chemical analysis scheme were as
follows:
(1) Quantitative analysis of NPDES parameters, anions and metals;
(2) Semiquantitative analysis of organic compounds known or
suspected to be present in the samples (based on an engi-
neering evaluation of the plant production process), iden-
tified as being potentially toxic;
(3) Qualitative analysis of other organic compounds suspected to
be present in the sample, but not particularly toxic;
(4) Qualitative analysis of other unknown organic compounds,
detected in the sample by gas chromatography/mass spec-
trometry (GC/MS); and
(5) Determination of the potential for organic compounds in the
samples to accumulate in the food chain.
3-36
-------
The Phase II chemical analysis scheme is presented in Figure 3-4.
Because capillary-column GC/MS analyses give greater resolution
of components than the packed-column type, more of the mass
spectra obtained were of single compounds, which greatly facili-
tated identification of unknown organics. Also, the chromato-
grams obtained by capillary column GC for TCO analysis could be
inspected to see whether a liquid chromatographic separation of
the sample would lead to substantial improvement in compound
identification by GC/MS analyses.
In Phase I, the process analysis indicated the probable presence
of several organic acids which were not detected by the Phase I
analytical scheme. To account for such species, a provision for
derivatization of these species was added to the Phase II proto-
col, to be used only if the plant presurvey revealed the likeli-
hood of organic acids being present in the effluent. The low
recovery of organics in Phase I also indicated the possible need
for an analytical test to account for low levels of pesticides
and other chlorinated species. Therefore, capillary column elec-
tron capture/gas chromatography (EC/GC) was added to Phase II,
with the intention of determining a total mass of these types of
compounds. A final addition to the organics analysis procedures
involved injection of the remaining aqueous phase (after acid and
base/neutral extraction) into the GC/MS. The purpose of this test
was to determine if any high levels (>50 mg/L) of organics were
escaping the extraction scheme.
3.6.2 Phase II Results
Approximately 40% of the NPDES parameters measured in the Phase II
effluent samples were above the respective plant discharge limits
and monthly averages. Most effluents had at least one parameter
with values which exceeded discharge limits. The concentrations
of anions not covered by NPDES permits varied by as much as three
orders of magnitude from one effluent to another. However, none
3-37
-------
SAMPLES RECEIVED
FOR
ANALYSIS AT MRC
1500 ml* FILTER
TOC. TSS
PURGE ABLE ORGANICS BY
PURGE AND TRAP AND
CC-'MS ANALYSIS
VARIABLE
SPECIAL TESTS
DETERMINED BY
FIELD PRESURVEY
BIOACCUMULATION
104 L
ORGANIC ANALYSES
80 ml
10L
EXTRACTABLE ORGANICS
ORGANIC PHASE
CH2CI2EXTRAaATpH12
CONCENTRATE TO 10 mL
AQUEOUS PHASE
CH2CI2 EXTRACT AT pH2
GRAV. CAPILLARY: TCO.
TCG, GC/EC
ORGANIC PHASE
CONCENTRATE TO 10 mL
CRAV. CAPILLARY:TCO.
TOG. GC/EC
DIRECT AQUEOUS
INJECTION GC/MS
AQUEOUS PHASE
GRAV. CAPILLARY:TCO.
TCG. GC/MS ON EACH
FRACTION
Figure 3-4. Phase II chemical analysis protocol,
3-38
-------
of these anions was observed in a concentration that would be
deleterious to the Chesapeake Bay ecosystem. These data indicate
that the grab samples, while sufficient for testing the protocol,
were not always representative of normal operation, and any con-
clusions about impacts to the Bay must be considered in this light.
The use of capillary versus packed-column GC/MS resulted in greater
peak resolution (cleaner mass spectra), making the extra inter-
pretive time devoted during Phase II worthwhile. In Phase I, an
average of 4 extractable organics per sample were specifically
identified, while the average in Phase II was 30. This differ-
ence is more pronounced when considering the levels of identified
compounds. In Phase I, few species were identified below the
10 yg/L (ppb) level. However, in Phase II, the availability of
cleaner mass spectra allowed identification in the 0.1 ppb range.
In Phase ,1, the GC/MS analyses only identified a small amount
("-10%) of the organics as compared to the TOC value. One possi-
ble explanation was low extraction efficiency. So, in Phase II,
eight deuterated organic compounds were spiked into the 10-liter
effluent sample at concentrations ranging from about 5 yg/L to
200 M9/L. The sample was then extracted and analyzed. Percent
recoveries of the deuterated spike compounds were then calculated.
Results indicate recovery efficiencies ranging from zero to 140%,
with an average recovery efficiency of about 80% to 90%. This
superior recovery efficiency of the Phase II extraction scheme
was further demonstrated by the GC/MS identification of three
times (that is, nearly 30% on the average) the mass of organics
which contributed to the TOC.
Lower recovery efficiencies were observed in Phase II analyses
when small concentration (~5 jjg/L) of spiked compounds were used.
In other words, it was difficult to guantitate organic compounds
in trace levels (1 pg/L to 10 ug/L), and the accuracy achieved was
about ± 100%. For compounds in higher concentrations (>100 pg/L),
3-39
-------
the recoveries were much greater, and accuracy was about ±10%.
Data obtained in Phase II using capillary columns for both GC/FID
(flame ionization detection) and GC/MS analyses demonstrated the
superiority of capillary over packed columns (based on the
Phase I packed-column data). The resolution afforded by these
columns was excellent, allowing separation of large numbers of
components without fractionation, and thereby improving identi-
fication capabilities. The retention time stability of compo-
nents eluting from capillary columns was very good. Over a
period of about a week, the components in mixtures injected for
TCO/TCG quantitation and retention time calibration varied by
only about ±0.02 min.
Direct aqueous injection of effluent samples to identify organic
components in the aqueous phase after methylene chloride extrac-
tion had very limited utility due to high detection limits. In
the case of effluent B142S, however, it proved to be very informa-
tive. This sample had a relatively high TOC value (960 ppm),
which was not accounted for by the sum of the GRAV, TCO, and TCG
fractions. This implied the presence of a high concentration of
nonpurgeable and nonextractable organic species. Approximately
2,000 ppm of acetic acid was found in this sample when the effluent,
after extraction and adjustment of the pH to 2, was analyzed by
direct aqueous injection.
One of the data correlations in Phase II revolved around attempts
to determine if the respective industrial outfalls were contribut-
ing to an increase in toxic organics in the Bay sediment which can
pose a threat to bottom-dwelling species such as oysters and clams,
accumulate potentially toxic substances in the food chain, cause
ecosystem upsets, etc. In attempting to correlate effluent data
and sediment data from samples collected near an effluent outfall,
both the history of each site and recent meteorological events
must be considered. Because little is known about the dynamics of
sediment movement in the Bay, identification of organics being
3-40
-------
concentrated in sediments from industrial effluents may never pro-
ceed beyond the tentative stage. However, even if a correlation
is not evident, analysis of sediment for organics and bioaccumula-
tion potential will determine if an environmental problem exists,
and long-term monitoring will determine if the problem is increas-
ing or decreasing in magnitude. The sediment/effluent comparisons
in Phase II did reveal some possible correlations in terms of simi-
lar compounds detected and relative quantities present. This was
particularly true in the case of potentially bioaccumulative species,
several of which were present at concentrations 1,000 times greater
in the sediment than those in the corresponding effluent.
The purpose of the bioaccumulation tests during Phase II was to
determine how many organic compounds in the methylene chloride
extract of an effluent sample are potentially accumulative in
fatty tissue. Potentially bioaccumulative compounds are those
which elute from the HPLC with a log P £3.5 and response £25%
full-scale deflection. In Phase I, several samples contained
several compounds which were classified as bioaccumulative. In
Phase II, MRC made a greater effort to identify which compounds
caused the positive response. This identification effort was
conducted using the following: (1) log P values determined for
each effluent sample by the HPLC method, (2) organic compounds
identified in each sample by GC/MS, (3) published log P values
for the compounds identified by GC/MS, and by the process analy-
sis, and (4) additional compounds in the literature with log P
values similar (within ±0.05) to the values determined in the
sample by the HPLC method.
By comparing all the above information for each sample, it was
possible in Phase II to make several tentative identifications of
compounds that caused a positive bioaccumulation response. The
number of potentially bioaccumulating compounds ranged from 5 to
17 per plant in the effluent samples and from 11 to 23 per plant
in the sediment samples. The key Phase II effort was the attempt
3-41
-------
to specifically identify the organic compound which corresponded
to each bioaccumulative positive response. This was accomplished
by comparing Phase II data, including GC/MS, with published bio-
accumulation data. From this effort, it was possible to tenta-
tively identify about 25% of the bioaccumulative compounds.
Correlations of bioaccumulative responses in the effluents with
those in the corresponding sediments were not straightforward.
Differences in sample work up procedures for effluents and sedi-
ments and matrix interferences made data correlations difficult.
3.7 PHASE III: PROTOCOL VERIFICATION STUDY
3.7.1 Phase III Analysis Scheme
3.7.1.1 Improvements over Phase II—
The problems MRC experienced with implementation of the Phase II
analytical protocol may be divided into three categories:
(1) problems arising from handling large volumes of samples and
data, (2) problems generated by the complexity of the analytical
scheme, and (3) problems derived from sample workup and analysis.
Improvements for Phase III are given below.
3.7.1.1.1 Sample and data handling—The analytical scheme for
Phase II, particularly for extractable organics, generated many
samples from each effluent received, due to division for different
types of analyses. If an extracted organics fraction from a par-
ticular sample required LC separation of the acid and base/neutral
fractions, one sample generated six individual fractions to be
analyzed. In addition, problems with a particular analysis neces-
sitated reanalyzing the sample one or more times. The resulting
proliferation of analytical records in such a case becomes over-
whelming unless one develops a standardized sequence of analyses
and a cross-referenced recbrdkeeping system to provide ready access
to the data. The system that evolved at MRC during Phase II, and
which was implemented for Phase III, is described in Section 3.7.1.21
3-42
-------
Extracting information from raw data files also becomes extremely
time consuming when huge volumes of data are generated, as in the
case of capillary GC/MS analysis of organics. To reduce data in
a time-efficient manner, one must learn to utilize the equipment
at hand with the utmost efficiency, and even modify it, if neces-
sary, to achieve the desired time for data reduction. Based on
Phase II experience, MRC developed a data reduction procedure
that substantially reduced the time required to process the data
generated in TCO and GC/MS analyses in Phase III.
3.7.1.1.2 Analytical scheme complexity—The analytical scheme
prepared for the extractable organics portion of Phase II arose
in response to concerns about a variety of potential problems.
For example, a detailed scheme was prepared to handle those cases
in which a precipitate would form during the extraction-concentration
of the effluents. Because no precipitate formation was observed in
effluent workups, this aspect of the scheme was eliminated in
Phase III, greatly simplifying it. Parts of the scheme, especially
the LC fractionation portion, were designed with reference to the
EPA Level 1 procedures, with the volumes of eluents chosen to
approximate those in the more complex Level 1 system.
The complicated chemical analytical scheme used in Phase II had
one major disadvantage. It required a wide variety of concen-
tration/dilution factors to convert the concentrations of com-
ponents of a particular fraction measured at the instrument to
the concentration in the original sample. In such a case, cal-
culating the necessary correction factors and being careful to
apply them to the proper fractions becomes a major activity.
This situation was eliminated in the Phase III scheme by making
small adjustments in the volumes used at various points in the
extractable organics scheme, so that, except for the GRAV's,
every sample analyzed experienced a 1,000-fold concentration
factor from the effluent to the instrument.
3-43
-------
3.7.1.1.3 Sample workup/analysis—A problem evident in the Phase II
workup procedure for effluents, which was resolved in Phase III,
was the low percentage recovery of the deuterated recovery stand-
ards that were added to the samples before extraction. In other
cases, the percent recovery of the spikes was greater than 100%
for most of the deuterated compounds in the mixture, suggesting
that some of the solvent had evaporated before the GC/MS analysis
was performed. In Phase III, screw-cap vials (rather than crimp-
ons) were used to give positive assurance of a good seal and to
permit easy seal replacement after puncturing by a syringe.
In Phase II analyses, the use of derivatization by a trimethyl-
silylating reagent was investigated. Because of the limited
number of trimethylsilyl esters in mass spectral libraries,
derivatization of an effluent extract followed by GC/MS analysis
was not very productive. Few compounds were identified after
much effort, and in most cases they would have been identified
without derivatization. In Phase III, derivatization was used
only in seeking particular substances (such as dicarboxylic
acids) which are on the presurvey list for a given effluent and
which did not pass through the GC column readily in underivatized
form.
Several problems encountered during sediment analysis in Phase II
stemmed from the fractionation of the sediment extract by Bio-
Beads SX-3. An O-ring in a column connection contributed several
prominent alkanes (C14-C17) to the sediment sample, and the
published procedure failed to exclude sulfur from the fraction
containing the compounds of interest. These problems were elimi-
nated in Phase III by the use of different column fittings and .by
adjusting the elution profile used.
The effect of high total organics on the mass spectrometer source
sensitivity was studied in Phase III by analyzing samples using
progressive addition of standards. The sediment workup procedure
3-44
-------
was also modified to avoid the high level of organics seen at the
instrument in some of the Phase II samples. This was accomplished
by extracting smaller quantities of sediment.
The Phase II, GC/EC analyses of the sediments were complicated by
incomplete removal of CH2C12 by the concentration step after the
Bio-Beads fractionation. This problem can be remedied by an addi-
tional solvent exchange step. The Bio-Beads fractionation scheme
can be altered to allow removal of S6, S7, and S8 by changing the
mobile phase to increase the retention of sulfur and/or decrease
the retention of organics in the column.
3.7.1.2 Chemical Analysis Protocol—
3.7.1.2.1 Effluents—The Phase III scheme for analysis of the
extractable organics portion of an effluent sample is shown in
Figure 3-5. This scheme has been simplified compared to Phases I
and II by removing considerations of precipitate formation during
the concentration steps. The volumes of extract have been adjust-
ed so that each sample analyzed for TCO/TCG or by GC/MS is concen-
trated exactly 1,000-fold over the original effluent, whether the
analysis is prior to or subsequent to the LC fractionation. A
substantial reduction in effort and minimization of calculation
errors was realized by this change.
The extent to which a given analysis proceeded through this
scheme depended on the decision tree described in Section 4. One
sample may have required only GRAV and TCO/TCG analyses; another
may have required GC/MS analysis; or the entire scheme including
the LC separations and subsequent analyses might have been neces-
sary to adequately characterize the sample.
The detailed order of analyses for the Phase III samples is given
in Figure 3-6. Obviously, many of the details of any such list
will need to be modified.to fit the operations and instrumenta-
tion in a particular laboratory. The data logs (Table 3-5)
3-45
-------
OJ
IF THE INITIAL CONCENTRATION PROCEEDS TO 10 ml FOR AIL TCO. CMC AND CC/MS
ANALYSES. 1 Mg/ml MEASURED AT THE INSTRUMENT -1 MS* IN THE EFFIUEN1. FOR GRAV*
MULTIPLY THE MEASURED WEIGHT IN mg BY 0.25 TO GET mj/L. FOR CRAVb
THE MEASURED WEIGHT IN mg=mg/l IN THE EFRUENT.
SAMPLE FOR
DIRECT AQUEOUS
INJECTION
nut
J ml - PUDGE 13 mln.
WITH HELIUM
GC/MS DIR. AO. IN).
(CONCENTRATE TO 2 ml)
ICONCtMTRATE TO 2 mil
ICONONKAIt TO 2 mil
(CONCENTRATE TO 7 ml)
(CONCENTRATE TO 2 mil
•IF PRECIPITATION OCCURS. DISCONTINUE CONCENTRATION
AND PROCEED WITH ANALYSES.
Figure 3-5. Phase III extractable organics analysis scheme.
-------
TCO/TCG
WITHQC SAMPLES1
I
CALCULATE TCOrtCG
IMMEDIATELY
ISLC
FRACTIONATION
NECESSARY
YES
FRACTIONATE INTO
3 FRACTIONS
1
NOTIFY
WORKUP AREA
ALL SAMPLES:
RUNGC/MS3>C USING
DILUTION FACTORS BASED
ON TCO/TCG. USE SIMULTANEOUS
QC SAMPLESd
KEEP DATA LOGS
CURRENT !
1
CALCULATE SPIKE RECOVERIES
IMMEDIATELY WITH QUANTID
REANALYZE SAMPLE BY GC/MS
USE THE BETTER RESULT
ARE SPIK
RECOVERIES
EASONABLE
CORRECT TCO/TCG VALUES AND
PREPARE FINAL TABLE
T
(CONT.)
Figure 3-6.
Order of progression through an
analysis of extractable organics.
3-47
-------
OBTAIN HARD COPIES OF
CHROMATOCRAMS FROM GC/MS RUNS
1
BATCH GC/MS DATA AT M * LEVa
TOGIVE<_35PEAKS
PEAK DETECTED FILES.
I
LABEL PEAK NUMBERS NEATLY ON
CHROMATOGRAM
i
IDENTIFY MISSING MAJOR PEAKS MISSING FROM
AUTOMATED BATCH OUTPUT AND OBTAIN THEIR MASS
SPECTRA AND AREAS MANUALLY USING "SPEED" PROGRAM
1
XEROX CHROMATOGRAMS AND PLACE COPIES
IN 3-RING BINDERS IMMEDIATELY. WITH LABELED
DIVIDERS. LABEL ORIGINALS FOR REPORTS
AND FILE THEM.
1
USE AUTOMATIC COMPUTER SEARCH
OF PEAK DETECTED FILES vs. MASS SPECTRAL LIBRARY
TO GIVE 1ST LEVEL IDENTIFICATION OF MASS SPECTRA.
I
EXAMINE COMPUTER RESULTS AND SUPPLEMENT
WITH MANUAL SEARCH TO GIVE 2ND LEVEL
IDENTIFICATION.
1
I
COMPLETE QUANTITATION
ON 3-6 SAMPLES AT A TIME.9
AVOID MANY SAMPLES IN
PROGRESS AT ONCE,
1
COMPLETE DATA INTERPRETATION
ON 3-6 SAMPLES AT ATI ME.h
AVOID MANY SAMPLES IN
PROGRESS AT ONCE.
t
(CONT.)
Figure 3-6 (continued)
3-48
-------
USE CHECKLIST ' TO SEE THAT ALL
DATA ARE COMPLETED.
i
MOVE RAW DATA TO TAPES.
J
i
COMPILE GRAV, TCO/TCG. PURGEABLES.
AND GC/MS RESULTS INTO COMPLETE
REPORT FOR EACH SAMPLE.
I
PERFORM NEXT SET OF GC/MS ANALYSES.
a Check levels marked on vials to
verify no evaporation.
b Blanks
C, - C,, Retention markers
Cg - C,. Stds at several concns
Reagent blanks
Spikes (PNA markers)
Watch out for solvent variations..
c Run batch of ~6 acid fraction samples
one time; ~6 B/N fraction samples the
next.
d
Approximate sequence
(All samples spiked with anth-d.J :
(1) CH2CI2
(2) Spike Solution
(3) Reagent blanks
(4 - 9) Samples
(10 -12) Standards
(13) Alkane mix
(14) Column performance standard
See Table 3-4.
f
Label with percent threshold and whether
background subtraction was used.
9 Person quantitating enters results in table
prepared by interpreter.
Person doing the interpretation should fill in
the tables with t name for compound, and
4 major masses.
J
.See Table 3-5.
Place tape numbers in data log.
Figure 3-6 (continued)
3-49
-------
TABLE 3-5. DATA LOGS ON GC/MS SAMPLES
1. Chronological log 2. Sample log
(a) Date (a) Sample ID
(b) FRN number (b) FRN numbers for all
(c) Sample JD . analyses of that sample
(d) CRN(AR)D (c) Comments6
(e) Tape number (file) .
(f) Anthraceoe-djo area
(g) Comments
File reference number, generated when GC/MS analysis is per-
formed. To be numbered sequentially, insofar as possible.
Cartridge reference number and area where raw data are stored.
Identification of tape (permanent) storage location of raw
data.
Area of standard added at the instrument; checked routinely to
ascertain that injection was proper and instrument response
was normal.
eComments should include, for example, which standards and
blanks should be associated with which samples, why a sample
is being rerun, peculiarities of the analysis, etc.
associated with this order of analysis may be generally helpful
insofar as they reveal the details that have been found necessary
at MRC to keep proper track of the data generated. Likewise, the
checklist (Table 3-6) may indicate which items will impede the
progress of the analysis if they are overlooked or not completed
in a timely manner.
3.7.1.2.2 Sediments—In Phase III, the necessity for and the
type of sediment analysis was determined from use of the decision
tree mentioned above. When an analysis of organics in a sediment
was necessary, the scheme in Figure 3-7 was used. Organics ex-
tracted from the sediments were much less concentrated than was
the case in Phase II (most of the latter samples, too concentrated
for direct analysis, had to be diluted prior to injection into the
instrument).
3-50
-------
FREEZE-DRY I
(GRIND)
(CONE AND QUARTO I
/SOXHIET EXTRACTION \
1 CHA
V 48 nr /
(CONCENTRATE TO 10 ml)
(CONCENTRATE TO 10 (H)
3ml
TCO
1ml
1ml
BIOACCUMULATION
1ml
GUMS
(SOLVENT EXCHANGE)
" Sims procedure as tor effluent.
Figure 3-7. Sediment organics analysis scheme
3-51
-------
TABLE 3-6. CHECKLIST FOR EXTRACTABLE ORGANICS ANALYSIS
Sample # FRNa
D Chromatogram labeled for report and Xeroxed copy available
a QUANTID on spikes completed
D All spikes accounted for
D BATCH at % with /without background subtraction
D Peaks labeled and no major ones missed
D FRN data logs filled out
D Presurvey list compounds searched
D All mass spectra "identified"
o Estimated concentrations calculated
D All guantitations completed
D Table completed
D Special comments on sample: .
aFile reference number, generated when GC/MS analysis is
performed.
The sediment samples were spiked with higher levels of deuterated
compounds to improve recovery guantitation. Closer attention was
paid to instrument response to generate accurate percent recoveries.
The use of automated data reduction procedures and adherence to
the checklist in Phase III resulted in major time and manpower
savings in these analyses, similar to those experienced in the
effluent analyses.
3.7.1.3 Biological Analysis Protocol—
At the request of the EPA, all biological effects testing (except
the bioaccumulation potential test) was eliminated from Phase II,
to allow more effort on chemical characterization of the samples.
In Phase III, biological testing was again a part of the proto-
col. However, the scope was far less than that in Phase I. In
Phase III, MRC's subcontractor, EG&G Bionomics, conducted acute,
static bioassays on 18 of 28 effluent samples using the marine
species Mysidopsis bahia. The State of Virginia Water Control
3-52
-------
Board conducted bioassays using the fathead minnow (Pimephales
prontelas) on the remaining 10 effluents plus 2 samples which
overlapped with EG&G to provide a limited comparison of species
response.
3.7.2 Phase III Results
Due to the amount of analytical data generated, the results of
the Phase III effluent and sediment sample characterizations are
tabulated separately in Volume III of this report. Appendices D
through J present the Phase III analytical results for the appro-
priate plants in the following order, with those for which GC/MS
analyses were performed appearing first:
GC/MS Analyses performed No GC/MS analyses
A109 A101
B112D B111D
B119D B113D
C150D B124D
C155D C151D
C156D C153D
C157D C154D
C161D C158D
C164D C159D
C169D C160D
B133S B126S
B141S B143S
B142S B147S
B149S C169S
Appendix D, in Sections D.I through D.28, presents the plant-by-
plant Phase III effluent analysis data in the above order. Each
plant description includes the process engineering analysis list
of chemical compounds suspected to be present in its effluent.
The following results are provided:
• Wet chemistry, ioni chromatography, and NPDES parameters
• Purgeable organics
• Acid fraction organics (for samples analyzed by GC/MS)
3-53
-------
• Base/neutral fraction organics (for samples analyzed by
GC/MS)
• Organic carbon distribution
• ICAP metals
• Bioaccumulation potential test
• Bioaccumulation - GC/MS correlations (for samples analyzed
by GC/MS)
• Aquatic toxicity results
• Data interpretation
The following analytical results for the Phase III sediment
samples are given in Appendix E:
• Extractable organics
• Carbon balance
• ICAP metals
The remaining appendices, comprising Volume IV of this report,
contain the following information:
Appendix Contents
F Relative retention indices and TCO/TCG chromatograms
of Phase III effluents
G Total ion chromatograms for Phase III effluents
H Relative retention indices and TCO/TCG chromatograms
of Phase III sediments
I Total ion chromatograms for Phase III sediments
J Analysis of Blanks and Standards Associated with
the Analyis of Plant Effluents and Sediments
3-54
-------
3.8 REFERENCES
1. Hopper, T. G., et al. Inventory and Toxicity Prioritization
of Industrial Facilities Discharging into the Chesapeake Bay
Basin, two volumes. Submitted to EPA by GCA/Technology Divi-
sion, GCA Corporation, EPA Contract 68-02-2607, Work Assign-
ment No. 30, August 29, 1979.
2. Draft Final Report: Sampling and Analysis Procedures for
Screening of Industrial Effluents for Priority Pollutants.
U.S. Environmental Protection Agency, Cincinnati, Ohio,
April 1977,
3. Dentzen, D. E., et al. IERL/RTP Prodecures Manual: Level 1
Environmental Assessment (Second Edition). EPA-600/7-78-201,
U.S. Environmental Protection Agency, Research Triangle Park,
North Carolina, February 1979.
4. Duke, K. M., M. E. Davis, and A. J. Dennis, IERL/RTP Proce-
dures Manual: Level 1 Environmental Assessment Biological
Test for Pilot Studies. EPA-600/7-77-032, U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina,
April 1977.
5. Draft Final Report: Sampling and Analysis Procedures for
Screening of Industrial Effluents for Priority Pollutants.
U.S. Environmental Protection Agency, Cincinnati, Ohio,
April 1977.
6. Eight Peak Index of Mass Spectra, Vol. Ill, Second Edition,
Mass Spectrometry Data Center, AWRE, Aldermaston, Reading,
United Kingdom, 1974.
3-55
-------
SECTION 4
SITE SPECIFIC TOXICITY IDENTIFICATION PROGRAM
4.1 BASIC PHILOSOPHY AND INTENT
The National Pollutant Discharge Elimination System (NPDES) per-
mit program was established by the U.S. EPA to control and monitor
wastewater discharges to our national waterways. Under this pro-
gram several generalized chemical analyses were defined for dis-
chargers to use to monitor the quality of their effluent. However,
over the years these chemical analyses were not adequate to moni-
tor chemicals being discharged which were toxic to the environment.
As a result, through the second round of NPDES permit issuance, the
EPA is trying to incorporate effluent toxicity testing.
The particular problem is in selecting which chemical analyses and
biological tests to use and in what sequence. For example, chemi-
cal analyses alone .cannot adequately characterize the potential
hazardous nature of a complex wastewater effluent because of the
lack of information on the biological effects of most compounds
detected and on the synergistic and antagonistic effects of the
mixture. On the other hand, biological analyses alone are not
adequate because if a toxicity problem is discovered no data, would
be available to determine what caused the toxicity or, more impor-
tantly, how to reduce the toxicity through wastewater control
technology.
The Site Specific Toxicity Identification Program (TIP) presented
in this section was developed as an approach to the solution of
this problem.
4-1
-------
The approach MRC developed centers around a decision analysis that
starts with several simple inexpensive analyses and becomes pro-
gressively more detailed and expensive as the cause of the toxi-
city is investigated. Figure 4-1 shows the basic philosophy and
sequence of steps for this protocol. The actual chemical analy-
ses, bioassays, and decision criteria used in the protocol are
"totally the user's choice." The user must decide why he wants
to use the approach and what type of data he needs to make the
types of decisions necessary. The types of chemical and biolog-
ical analyses used will depend on the capability of the user's
laboratory facilities and/or funds available for an outside
laboratory.
4.2 ELEMENTS OF TIP
As was pointed out earlier, the key to the success of this approach
is up-front planning and familarization with the site being studied.
This section provides some guidance and information to help plan
the program.
In terms of site familarization, Section 3.3 provided detailed in-
formation on the type of data necessary to properly structure the
program. Also, Section 3.4 gave detailed instructions on tech-
niques to use for field sampling. In most cases it is recommended
that 24-hour composited samples be collected.
The purpose of the Stage I - Basic Analyses is to assess, in a cost-
effective manner, if the effluent is toxic or potentially harmful
to the environment or man, and to provide data to indicate what
steps should be taken next. The term "cost-effective" does not
necessarily mean low cost because everyone has different concepts
of "low." It does mean, however, the most amount of "useful"
data for the lowest cost.
4-2
-------
Site Specific
Process Evaluation
I
Sample Collection
1
Stage I
Basic Analyses
No /Sample\Yes
Toxic
Stage II
Intermediate Analyses
Yes
No
Stage III
Advanced Analyses
1
Toxic Reduction
Program
Figure 4-1. Site Specific Toxicity Identification Program
(TIP) designed to evaluate effluent toxicity.
4-3
-------
Based on the results from all three phases of this project and
other similar projects, the chemical and biological tests recom-
mended for Stage I are presented in Table 4-1. Using commercial
laboratories the estimated costs for each of the types of analy-
ses are: $300 to $500 for the chemical analyses; $800 to $1,200
for the bioaccumulation potential analysis; and $2,500 to $3,000
for the three static acute bioassays.
TABLE 4-1. RECOMMENDED STAGE I - BASIC ANALYSES
Type
Analysis
Chemical analyses Site specific NPDES permitted parameters
Anion/cations of at least: SO^2 , SO^2 , NH3 ,
Cl~, NC-2,
Trace metals in filtrate and suspended solids
Total organic carbon (TOC), BOD5 , and COD
GC/FID including relative retention indices
Bioaccumulation
potential
HPLC method for potential bioaccumulative
compounds
Biological tests
Static acute vertebrates (e.g., fathead min-
now or sheepshead minnow)
Static acute invertebrates (e.g., daphnia or
mysid)
Static acute algal assay
Therefore, for about $4,000 one will have enough data to very ef-
fectively assess whether his outfall will have an adverse effect
on the environment and some information to determine what might
be causing the problem. Table 4-2 presents other bioassay tests
which might be selected for Stage I depending on the type of data
needed.
4-4
-------
TABLE 4-2.
POTENTIAL BIOLOGICAL TESTS FOR ASSESSING ENVIRONMENTAL
IMPACT OF DISCHARGES TO THE CHESAPEAKE BAY
Test
Effect
Microbial Mutagenesis
(Ames Test)
Cytotoxicity
Presumptive
carcinoqenicity
Mammalian cellular
toxicity
Rodent Acute Toxicity Whole animal toxicity
(RAT Test)
Bioaccumulation Potential Accumulation in food
chain
pp'-ci ipjtion
Test Outputs_
Health Effects Tests
Genetically sensitive strains of micro-
organisms are exposed to various
doses of sample with and without
metabolic activation.
Selected cells (RAH, CHO, or WI-38)
are exposed to various doses of
sample, then various endpoints are
measured.
Rats or other rodents are fed a quan-
tity of sample, then observed daily
for adverse symptoms over a 14-day
period. The experiment is terminated
with a necropsy exam.
An instrumental technique which pilots
a biological response.
Mutagenic response is measured relative
to controls.
An index of functional impairment,
toxicity, and metabolic change is
established relative to controls.
Inventory of pharmacological and gross
physiological effects in a whole
animal system.
An indication of the number and quantity
of compounds in the sample which may
accumulate in fish and oyster tissue.
Algal Growth Response
Algal growth inhibition
or promotion
Aquatic Animal Exposure Animal toxicity
(Static Acute Bioassay)
Terrestrial
Benthic Studies
Toxicity to plants,
insects, and soil
microbiota
Toxicity to bottom
dwelling species
Ecological Effects Tests
Cultures of selected marine and/or
freshwater algae are used to gauge
reaction to sample or dilution thereof.
Select marine and/or freshwater verte-
brates and invertebrates are exposed
to a graded dilution series of
samples.
Stress to species is measured as a
function of inreased or decreased
growth and procreativity.
Cultures of selected marine and/or
freshwater benthic species are ex-
posed to a graded dilution series of
samples.
Growth response measure-stimulation or
inhibition.
Gross index of toxic potential to repre-
sentative animals.
Effects on plants, insects, and soil of
complex samples.
Effect of toxic compound accumulation
in sediments on bottom dwelling,
"filter feeding" species.
-------
With the data collected from the Stage I analyses one can deter-
mine the following:
Was the sample collected representative of the site's
typical effluent quality?
Was the effluent acutely toxic to any of the species
tested?
Are there any organic compounds which produced a positive
response on the bioaccumulation test?
Are there significant amounts of unknown organic compounds
present?
Figure 4-2 shows the TIP scheme and decision criteria actually
used in Phase III. The acute toxicity of effluents is generally
measured by a value known as an "EC50." For this program, where
vertebrate and invertebrate species were testsd, EC50 means the
concentration of the effluent in dilution water at which half of
the test species die. An EC50 of 50% means half of the species
tested died in a 50% (v/v) effluent/diluent solution. For
Phase III, an EC50 value of 50% was selected as a preliminary
decision point for further investigation of toxicity. At an EC50
value of 50% or greater, the acute toxicity to species in the re-
ceiving stream where the flow rate was equal to or greater than
the discharge would be significantly reduced. For effluents with
an aquatic species EC50 of 50% or lower, further data would be
necessary to determine the source within the plant and identity
of the toxic compounds.
The second primary decision point was "are there bioaccumulative
compounds present?" If bioaccumulative compounds are present,
then the goal was to determine what the compounds were, are they
toxic, and did they cause the aquatic toxicity (if EC50 <50%)?
4-6
-------
«coMtiwiin a nt faiaiu*
«1QL!M MOKIT08INC
lUniOMTICk IICKSSAT
rm iiaacii no«ss»
•MIS/CMC wcioim
Oklft-IUNl STUAMS
Figure 4-2.
Site-specific Toxicity Identification
Program actually used in Phase III.
4-7
-------
If the number or concentration of these bioaccumulative compounds
was significant and could not be identified, then further, more
sophisticated chemical analyses may be necessary. If, on the other
hand, one could identify the compounds causing the positive re-
sponse, then no further analyses were required.
The third primary decision point was "are there significant con-
centrations of organic compounds present in the effluent?" For
this decision, the TOC (total organic carbon) analysis was used.
Again, based on data from the program and other references [1-4],
a decision point value of TOC £50 mg/L was selected.
If the cause of the acute aquatic toxicity cannot be identified,
and/or the bioaccumulative compounds cannot be identified, and/or
the TOC is above the criteria level, then there is a need for fur-
ther sample characterization. Stage II - Intermediate Analyses
involves organics analysis by capillary column GC/MS techniques.
The goal of this effort is to identify as many of the organic com-
pounds as possible within the time and funding constraints, and
determine if the compounds identified are the cause of aquatic
toxicity, the bioaccumulative compounds, or toxic to humans based
on published toxicity data.
[1] Rawlings, G. D., and J. R. Klieve. Source Assessment: Tex-
tile Plant Wastewater Toxics Study, Phase I. EPA-600/2-79-
004, U.S. Environmental Protection Agency, March 1978.
[2] Rawlings, G. D., and J. R. Klieve. Source Assessment: Tex-
tile Plant Wastewater Toxics Study, Phase II. EPA-600/2-79-
019i, U.S. Environmental Protection Agency, December 1978.
[3] Wilson, S. C., et al. Toxic Point Source Assessment of
Industrial Discharges to the Chesapeake Bay Basin, Phase I
Draft Report. EPA Contract 68-02-3161, U.S. Environmental
Protection Agency, May 1980.
[4] Wilson, S. C., et al. Toxic Point Source Assessment of In-
dustrial Discharges to the Chesapeake Bay Basin, Phase II
Draft Report. EPA Contract 68-02-3161, U.S. Environmental
Protection Agency, July 1981.
4-8
-------
The methods employed in Stage II are designed to give the analyst
a feel for the distribution of the organic carbon in the sample
and provide identification for the major organic species. The
detailed methodology of each of these techniques is given in
Appendix C of this report.
If the causes of the toxicity are identified, then a toxicity
reduction program can be implemented. However, if the above
effort cannot identify the cause of the toxicity problem, then a
third level of more sophisticated analyses must be considered.
Examples of the types of analyses which could be used include:
Flow-through aquatic bioassay - to determine acute toxicity
on samples which are more representative of the plant's ef-
fluent, which include the volatile organics that are lost in
static tests, and which maintain a constant oxygen level for
samples that have a high oxygen demand.
Fractionation-Static aquatic bioassay - by fractionating the
effluent, one can determine if the toxic response is organic
or inorganic in nature and determine if the toxicity falls
into one of three organic fractions based on polarity [5-8].
[5] Browning, S. E., M. A. Eischen, and K. M. Duke. Fractionation
Bioassay of Eleven Selected Discharges to the Chesapeake Bay
Basin, Draft Report. EPA Contract 68-02-2686, Task 119, U.S.
Environmental Protection Agency, June 3, 1980.
[6] Bean, D. M., and K. M. Duke. Fractionation Bioassay of
Selected Chesapeake Bay Discharges, Draft Report. EPA Con-
tract 68-02-2686, Task 119, U.S. Environmental Protection
Agency, August 1981.
[7] Leenheer, J. A., and E. W. D. Huffman, Jr. Journal of Re-
search of the United States Geological Survey, 4:737-751,
1976.
[8] Huffman, E. W. D., Jr. Isolation of Organic Materials from
In Situ Oil Shale Retort Water Using Macroreticular Resins,
Ion Exchange Resins, and Activated Carbon. Measurement of
Organic Pollutants in Water and Wastewater, ASTM STP 686,
C. E. Van Hall, ed. American Society for Testing and Mate-
rials, Philadelphia, Pennsylvania, 1979. pp. 275-290.
4-9
-------
Microtox® analysis of selected process streams - the analy-
sis of specific in-plant water streams before entry to the
treatment system by inexpensive techniques such as Microtox®
may determine the source within the plant of the toxicity.
An analysis of the process may then reveal the identity of
the toxicant.
Sediment sampling and analysis - analysis of sediments col-
lected near the outfall may make the identification of low
level toxicants easier due to the sediment acting as a
chemical concentrator.
When the source and identity of the toxicity is determined, dis-
charge control decisions can be made and either production proc-
ess modifications and/or additional end-of-pipe treatment can be
implemented to reduce the toxic effect of the discharge to an
acceptable level.
4.3 PRACTICAL APPLICATION OF SCHEME
At the conclusion of Phase II of this program, the preliminary
decision analysis scheme was proposed and discussed in the Phase II
report [4]. The Phase III samples were then collected and basic
and primary effects analyses were performed. In addition, the
TCO/TCG/GRAV series of tests were performed to expedite further
analyses. From these data the need for further sample analysis
was determined. Applying the decision criteria to the data re-
sulted in the recommendations given in Table 4-3. The metals
analysis data were not available from the EPA at the time Table 4-3
was compiled, so the affects of toxic metals could not be incor-
porated into any decisions at that point. The specific details
of the decisions taken for each sample, and the resulting conclu-
sions from implementing the decision analysis is contained for
each plant in Appendix D under Data Interpretation. To assist
the reader in understanding the thought processes necessary for
4-10
-------
TABLE 4-3. DECISION CRITERIA FOR PHASE III SAMPLES
Acute atjiiat ic
toxic ity
96-hr
Plant code LCr,(l
Maryland:
B126S
B133S
B141S
B142S
B143S
B147S
B149S
C169S
Virginia:
A101
A109
B111D
£* B112D
1
£ B113D
^ B119D
B124D
C1SOD
C153D
C156D
C1510
C154D
C155D
C157D
C158D
C159D
C160D
C161D
C169D
C164D
K
<3b
41
<3
4
22
54b
<3b
12
>100
24
>100
80
>100
>100
59
48
>100
24
56
5
15
77
89
42
>100
10
59
7
Ratinq
High
Moderate
High
High
Moderate
Low
High
Moderate
None
Moderate
None
Low
None
None
Low
Moderate
None
High
Low
High
High
Low
Low
High
None
High
Moderate
High
TOC,
mg/L
6
65
131
57
15
20
50
5
2
46
56
18
8
43
64
55
7
43
2
7
35
65
16.
18
2
60
65
29
<;RAV ,
my/I.
0.6
12.4
30.1
5.9
0.5
4.2
31.7
1.2
1.8
1.8
4.3
10.7
2.1
6.5
4.9
6.3
0.5
4.1
1.8
0.5
5.0
1.1
4.6
2.7
0.9
20.4
9.4
4.6
Effluent
TSS, mg/l.
101
29
114
24
101
7.3
106
416
13
221
127
15
42
19
9
42
• 231
31
19
359
25.3
6
23
7
5
69
33
12
Bioaccumulalton
No. of
positives Comments
6
5
13
8
1
7
23
2
2
8
3
19
4
14
1
13
2
9
5
10
10
3
7
4
0
9
10
7
All .'10 pg/L
1 Value 290 pg/L
All <10 pg/L
11 >100 |ig/t.
1 Value 230 pg/L
11 >100 pg/L
2 >200 pg/L
1 Value »360 pg/L
(due to Sa)
1 Value »360 pg/L
(due to Sfl)
All <20 pg/L
All <60 pg/L
All <10 pg/L
1 Value »360 pg/L
(due to Ss)
2 >100 pg/L
Oilier relevant analytical data
Ti;H = 300 mg/L, NH3 = 348 mg/L
•M\' = 1,050 mg/L
BOD... = >200 mg/L, TSS = 114 mg/L
(NOj + NOa) = 565 mg/L, Cl" = 2.140 mg/L
Cl" = 1.875 mg/L
Cl" = 7,600 mg/L
Cl" = 6,000 mg/L, SO,2 = 3,200 mg/L
SO^2 = 1,230 mg/L
Phenol = 1,139 pg/L
BODS = 92 mg/L
Cl" = 11,200 mg/L, SO^2 = 3,000 mg/L
Phenol - 125 pg/L
Cl" = 13,400 ng/L, SO^2 = 2,200 mg/L
BODS = 72 mg/L
Cl" = 490 mg/L
BOD5 = 141 mg/L
Cl" = 420 mg/L
R> i
-------
the decision analysis, a discussion of two examples from the
Phase III data—one toxic and one nontoxic (Plants B141S and
B111D, respectively)—follow.
4.3.1 Plant B141S
Evaluation of Protocol Performance
The Effluent from Plant B141S (a combined municipal/industrial
waste treatment plant) was found to be highly toxic to mysids
(LC50 = <3% at 96 hr). The TOC value for this plant was the high-
est for all plants sampled in Phase III (131 mg/L). The trace
metal and ionic species analyses indicated a potential toxicolog-
ical problem with ammonia (19.8 mg/L), chromium (93.3 mg/L), and
copper (123 pg/L). Due to the extremely high TOC, the decision
analysis dictated that further characterization of the organic
content was required and GC/MS and bi©accumulative species frac-
tionation and identification were implemented.
The subsequent organic analyses identified the presence of 90
organic compounds in the purgeable acid and base-neutral frac-
tions, most of which could not be specifically identified. A
number of the identified compounds were chlorinated and/or aro-
matic structures. In total, 15 potentially bioaccumulative com-
pounds (log P £3.5) were observed. Approximately 2/3 of the total
organic material was nonchromatographable.
For this plant the analytical protocol succeeded only in confirm-
ing the existence of a large number of organic compounds, but
failed to provide an unambiguous identification of the specific
source of the observed toxicity.
Possible Source of Chemical Toxicity
The primary source of toxicity in the effluent of Plant B141S is
presumed to be organic chemical species, and especially the chlo-
rinated organics to which mysids are particularly susceptible.
4-12
-------
The major portion of the organic content in the effluent can be
ascribed to fatty acids and steroid-based compounds that are asso-
ciated with human waste and public owned treatment facilities.
Other secondary causes for the observed toxicity could be the
relatively high copper content (123 pg/L) and free dissolved
chlorine from the final plant unit operation.
4.3.2 Plant B111D
Evaluation of Protocol Performance
Effluent from Plant B111D was nontoxic to mysids at all effluent
concentrations. The protocol dictated screening for bioaccumula-
tive compounds, of which four peaks were evaluated as positive.^
Further sample analysis was terminated at this point due to the
relatively low estimated concentrations of the peaks detected.
In a case such as this, the protocol did provide efficiency, in
that unnecessary, costly, analytical characterization was not pur-
sued due to the relatively low indicated toxicity and the absence
of large concentrations of bioaccumulative species.
4-13
-------
4.4 REFERENCES
1. Rawlings, G. D., and J. R. Klieve. Source Assessment: Tex-
tile Plant Wastewater Toxics Study, Phase 1. EPA-600/2-79-
004, U.S. Environmental Protection Agency, March 1978.
2. Rawlings, G. D., and J. R. Klieve. Source Assessment: Tex-
tile Plant Wastewater Toxics Study, Phase II. EPA-600/2-79-
019i, U.S. Environmental Protection Agency, December 1978.
3. Wilson, S. C., et al. Toxic Point Source Assessment of
Industrial Discharges to the Chesapeake Bay Basin, Phase I
Draft Report. EPA Contract 68-02-3161, U.S. Environmental
Protection Agency, May 1980.
4. Wilson, S. C., et al. Toxic Point Source Assessment of
Industrial Discharges to the Chesapeake Bay Basin, Phase II
Draft Report. EPA Contract 68-02-3161, U.S. Environmental
Protection Agency, July 1981.
5. Browning, S. E., M. A. Eischen, and K. M. Duke. Fractionation
Bioassay of-Eleven Selected Discharges to the Chesapeake Bay
Basin, Draft Report. EPA Contract 68-02-2686, Task 119, U.S.
Environmental Protection Agency, June 3, 1980.
6. Bean, D. J., and K. M. Duke. Fractionation Bioassay of
Selected Chesapeake Bay Discharges, Draft Report. EPA Con-
tract 68-02-2686, Task 119, U.S. Environmental Protection
Agency, August 1981.
7. Leenheer, J. A., and E. W. D. Huffman, Jr. Journal of Re-
search of the United States Geological Survey, 4:737-751,
1976.
8. Huffman, E. W. D., Jr. Isolation of Organic Materials from
In Situ Oil Shale Retort Water Using Macroreticular Resins,
Ion Exchange Resins, and Activated Carbon. Measurement of
Organic Pollutants in Water and Wastewater, ASTM STP 686,
C. E. Van Hall, ed. American Society for Testing and
Materials, Philadelphia, Pennsylvania, 1979. pp. 275-290.
4-14
-------
APPENDIX A
DATA CORRELATIONS FROM TOXIC POINT SOURCE PROGRAM
The purpose of this appendix is to discuss some of the key issues
which have surfaced during the course of this 27-month effort.
A.I. RESULTS OF PLANTS SAMPLED TWICE
Several plants were sampled and the effluents analyzed in more
than one phase of the program to determine the variability of
effluent quality and differences in the characterization due to
differences in protocol. Plants A101 and A109 were sampled both
in Phase I and Phase III. Plant B136S was sampled in the interim
between Phases I and II and again during Phase II. Plants B142S
and B143S were sampled in both Phase II and Phase III. The
results and comparisons follow.
A.1.1 Comparison of Final Effluent Analyses for Plant A101
Plant A101 final effluent was sampled and analyzed for various
parameters during Phases I (November 1979) and III (April 1981).
During Phase I, 24-hour composite samples were collected; in
Phase III, grab samples were collected. Table A-l presents data
comparisons for the parameters analyzed in both phases. There is
a significant increase (^700%) in flow during Phase III. Although
the concentrations of species are nearly the same in both the phases,
the loadings to the Bay will be significantly different. One log P
value was a match between the two samples, indicating the sample
contained a potentially bioaccumulative compound, which was not
identified by GC/MS. The effluent was nontoxic to fathead minnows
in both phases.
A-l
-------
TABLE A-l. PLANT A101 FINAL EFFLUENT DATA
COMPARISON PHASE I & III
Parameter
Flow, m3/d (MGD)
pH, standard units
Total cyanide
Total phenols
Total suspended solids
Chloride
Sulfate
Fluoride
Nitrate and nitrite nitrogen
Phosphate
Sulfite
TOC
GRAV
TCO
Total GC/MS mass, mg
Phase I
concentration
1,900 (0.48)
7.6
<0.010
<0.050
3.4
990
280
45
4.57
0.5
0.2
Phase III
concentr ati on
15,520 (4.1)
7.6
<0.02X>
<0.010
12.7
34
130
2.4
<0.095
0.024
<25
2
1.75
Note: Blanks indicate species not detected.
Concentration in mg/L, unless otherwise noted.
A.1.2 Comparison of Final Effluent Analyses for Plant A109
Plant A109 final effluent was sampled and analyzed for various
parameters during Phases I (November 1979) and III (April 1981).
During Phase I, 24-hour composite samples were collected, whereas
in Phase III grab samples were collected. Table A-2 presents
data comparison for the parameters analyzed in both the phases.
As flow data are not available, only concentration comparisons
can be made. Comparisons will indicate variation with time and
calendar day due to two different types of sampling and different
days of sampling in the two phases. The major differences appear
to be a 14-fold increase in TSS and a 2-fold increase in TOC.
Two compounds were identified as bioaccumulative that appeared in
both samples. The effluent was non-toxic to fathead minnows in
both phases, but was highly toxic to mysid shrimp in Phase I and
moderately toxic in Phase III.
A-2
-------
TABLE A-2. PLANT A109 FINAL EFFLUENT DATA
COMPARISON PHASE I AND III
Parameter
pH, standard units
Temperature, °C (°F)
Total cyanide
Total phenols
Total suspended solids
Fluoride
Chloride
Nitrate and nitrite nitrogen
Phosphate
Sulfite
Sulfate
TOC
GRAV
TCO
Total GC/MS mass, mg
Phase I
concentration
7.6
24 (76)
<0.010
0.110
15.7
5,300
3,000
20
1.55
1.90
0.093
Phase III
concentration
8.3
17.2 (63)
<0.020
<0.010
221
<0.25
6,000
<0.04
0.267
<2.5
3,200
46
1.85
6.10
<3.63
Note: Blanks indicate species
a_ . . .
not detected.
• A. j
A.1.3 Comparison of Final Effluent Analyses for Plant B136S
Plant B136S final effluent was sampled and analyzed for various
parameters during the interim between Phases I and II (May 1980)
and during Phase II (November 1980). Grab samples were collected
during both phases. Table A-3 presents data comparisons for the
group of parameters analyzed in both phases. As flow data are not
available, only concentration comparisons can be made. A large
increase in purgeable organic content can be seen as well as an
increase in total mass of extractable organics.
TABLE A-3. PLANT B136S FINAL EFFLUENT DATA COMPARISON
INTERIM PHASE AND PHASE II
Interim phasePhase II
concentration, concentration,
Parameter ng/L pg/L
Purgeable organics 20 4,280
Extractable organics 170 400
A-3
-------
A.1.4 Comparison of Analyses of Final Effluent for Plant B142S
Plant B142S final effluent was sampled and analyzed for various
parameters during Phase II (November 1980) and Phase III (April
1981). Grab samples were collected during both phases. Table A-4
presents data comparisons for the parameters analyzed in both phases.
For this plant, significant decreases are observed for all spe-
cies except chloride. In Phase II it was noted that a major proc-
ess upset occurred on the day of sampling. This accounts for the
difference. Three potentially bioaccumulative species were found
in both samples. One was tentatively identified as a
C 3-naphtha1ene.
TABLE A-4. PLANT B142S FINAL EFFLUENT DATA
COMPARISON PHASE II AND III
Parameter
Total cyanide
Total phenols
Total suspended solids
Total mercury (pg/L)
Fluoride
Chloride
Nitrite and nitrate nitrogen
Ortho phosphate
Sulfate
TOC
GRAV
TCO
Purgeable organics
Identified organics
Phase II
concentration
2.12
0.04
148
0.7
30
256
909
<0.006
4,300
960
12.7
3.91
0.121
2.4
Phase III
concentration
1.278
0.037
24
<0.2
2.4
2,140
565
<0.007
860
57
5.9
2.1
<0.107
0.817
Concentration in mg/L, unless otherwise stated.
A.1.5 Comparison of Sediment Analyses for Plant B142S
Grab sediment samples were collected and analyzed for organics
during Phase II (November 1980) and Phase III (April 1981).
A-4
-------
Total amount of organics detected'in the sediment during Phase II
was 236 H9/9/ compared to 70.7 ug/g for Phase III. A considerably
higher amount of organics was present in the Phase II sample com-
pared to the Phase III sample. This can be expected in sediment
sampling and analysis due to the difficulty in obtaining samples
for the same area on two different dates.
A.1.6 Comparison of Analyses of Final Effluent for Plant B143S
Plant B143S final effluent was sampled and analyzed for various
parameters during Phase II (November 1980) and Phase III (April
1981). Grab samples were collected during both phases. Table A-5
presents data comparisons for the parameters analyzed in both
phases. Most of the parameters for this plant are nearly the same,
indicating little variability in effluent quality. Two potentially
bioaccumulative species were identified as being present in both
samples.
TABLE A-5. PLANT B143S FINAL EFFLUENT DATA
COMPARISON PHASE II AND III
Parameter
Total cyanide
Total phenols
Total suspended solids
Total mercury, ug/L
Fluoride
Chloride
Ortho phosphate
Sulfite
Sulfate
TOC
GRAV
TCO
Total purgeables organics
Phase II
concentration
<0.02
<0.02
67.4
0.4
6.4
1,580 .
<0.006
<2.5
1,860
14
1.57
0.16
0.012
Phase III
concentration
<0.02
0.061
101
<0.20
0.4
1,880
<0.007
<2.5
382
15
0.50
0.07
<0.095
Concentration in mg/L, unless otherwise stated.
A-5
-------
A.1.7 Comparison of Sediment Analyses for Plant B143S
Grab sediment samples were collected and analyzed for organics
during Phase II (November 1980) and Phase III (April 1981).
Total amount of organics detected during Phase II was 9.7 \ig/g
compared to 4,872.5 pg/g for Phase III. A considerably higher
amount of organics was present in the Phase III sample compared
to the Phase II sample.
A.2 TOXIC METALS PARTITIONING
In Phase I, spark source mass spectrometry (SSMS) was performed
to measure elemental concentrations in both the filtrate and par-
ticulate phases after filtration of 1 liter of wastewater efflu-
ent as received.
Table A-6 summarizes the SSMS results for some particular ele-
mental concentrations in the filtrate and filtered solids from
the plant effluents examined in the Phase I study.
The major purpose for conducting the elemental analyses in the
Phase I screening effort was to obtain information on how the
elements are partitioned between the particulate and aqueous
phases of the effluent. The distribution of metals of toxicolog-
ical significance from the priority pollutant list are presented
in Table A-7. Except for beryllium, cadmium, silver, and mer-
cury, these metals are found primarily in the filtrate of the
effluent wastewater.
The results of Phase I indicated that no elements other than
sulfur beyond the usual 26 metals detected by ICAP analyses
were present in concentrations significantly above their detec-
tion limits. Therefore, in Phases II and III metals analyses
were performed by ICAP instead of SSMS. The results of the
analysis of 8 elements by ICAP for the wastewater sample filtrate
A-6
-------
TABLE A-6. SUMMARY OF SSMS RESULTS FOR PLANT EFFLUENTS IN PHASE I (pg/L)
A100
Element
Pb
Hg
Sb
Tl
Cd
Se
As
Zn
Cu
Ni
Cr
Be
Ag
Filtrate
6
0.3
2
<7
10
40
7
50
4
Suspended
solids
2.11
<0.1
<0.16
<0.11
<52.6
15.8
1.05
31.6
<0.01
A101
Filtrate
0.5
<10
<2
100
30
30
100
Suspended
solids
6.85
9.8
0.01
0.03
1.96
0.01
<9.78
0.88
0.10
0.78
0.06
A102
Filtrate
20
0.3
<30
2
800
10
20
4
Suspended
solids
0.76
3.8
0.02
0.95
15.21
13.31
0.17
11.4
<0.01
3.8
A103
Filtrate
0.3
<40
8
100
20
20
1
Suspended
solids
1.2
68.6
0.12
0.03
9.63
4.81
0.24
0.72
(continued)
-------
TABLE A-6 (continued)
00
A104
Element
Pb
Hg
Sb
Tl
Cd
Se
As
Zn
Cu
Ni
Cr
Be
Ag
Filtrate
9
0.3
70
<3
4
60
100
30
1
Suspended
solids
0.42
8.3
0.02
0.01
<0.01
<0.01
8.33
<20.8
0.04
0.83
1.3
A105
Filtrate
5
0.2
<7
4
80
20
20
.10
Suspended
solids
18.2
<0.1
0.41
0.14
<0.09
40.9
31.8
<45.5
2.73
<0.05
A106
Filtrate
70
0.3
10
80
30
20
10
Suspended
solids
3.0
42.6
0.03
0.09
<0.09
<42.6
8.51
0.17
1.70
0.09
A107
Filtrate
0.4
<10
8
300
60
20
5
Suspended
solids
0.90
32.1
0.02
0.27
0.89
<8.91
6.24
0.01
<8.91
<0.02
(continued)
-------
TABLE A-6 (continued)
A108
Element
Pb
Hg
Sb
Tl
Cd
Se
As
Zn
Cu
Ni
Cr
Be
Ag
filtrate
7
0.3
3
<5
<5
100
400
5
Suspended
solids
3.33
38.3
0.03
0.50
0.15
0.24
5
<16.7
0.17
<16.7
A109
Filtrate
0.1
7
2
10
20
100
10
600
1
Suspended
solids
3.7
106
<0.01
0.19
0.07
0.01
<1,850
0.93
0.06
0.37
<0.01
0.04
A110
Filtrate
20
0.3
40
10
90
50
20
4
Suspended
solids
17.7
44.2
0.31
0.04
17.7
4.42
0.44
3.98
<0.01
<0.09
Note: Blanks indicate not detected.
-------
TABLE A-7. DISTRIBUTION OF TOXIC METALS IN PHASE I SAMPLES
EPA toxic metalsa Major component phase
Antimony Liquid
Arsenic Liquid
Beryllium Parti cul ate
Cadmium Particulate
Chromium Liquid, particulate
Copper Liquid, particulate
Lead Liquid
Mercury Particulate
Nickel Liquid
Selenium, Liquid
Thallium
Zinc Liquid, particulate
aMetals listed as toxic, based on 1976
consent decree.
Found only at one plant at concentration
gO. 009
and filtered solids from plant effluents are presented in Tables
A-8 and A-9, respectively.
In Phases II and III, MRC filtered a portion of the effluent
through an acid-washed, tared, glass-fiber filter. Both the neat
and filtered effluent were preserved with acid. The filter was
retained for TSS determination and the metals loading in the
solid phase was determined by difference.
Of particular significance, performing metals analyses of both
the solid and liquid phases provided very valuable data. It was
observed in Phase III at many plants that metals (such as copper,
cadmium, nickel, and chromium) were predominantly associated with
the solid phase. These data are important because other research
efforts associated with the Chesapeake Bay Program are measuring
sediment composition, migration, and toxicity. In addition, this
partitioning of metals between solid and liquid phases presents
A-10
-------
TABLE A-8. SUMMARY OF ICAP RESULTS FOR PLANT EFFLUENTS IN PHASE II (|jg/L)
B137S '
Elements
Be
Cd
Cr
.Cu
Ni
Pb
Zn
Ag
Filtrate
<1.0
2.06
57.3
7.05
25.2
<70.0
<40
<3.0
Suspended
solids
1.29
B143S
Filtrate
<1.0
4.71
34.9
15.5
176
286
133
25.9
B131S
Elements
Be
Cd
Cr
Cu
Ni
Pb
Zn
Ag
Filtrate
<1.0
<2.0
40.0
123
56.7
<70.0
171
4.25
Suspended
solids
38.6
31
21
6
0.83
Suspended
solids
110.1
10.7
21.0
74
4.7
B136S
Filtrate
<5.0
<10.0
126
<30.0
<75.0
<350
<200
<15.0
B127S
Filtrate
<1.0
<2.0
92.2
11.0
<15.0
<70.0
71.0
<3.0
Suspended
solids
5.0
15.0
0.26
Suspended
solids
22.2
B130S
Filtrate
<1.0
<2.0
163
26.3
58.8
<70.0
186
<3.0
Suspended
solids
1.28
110
11.6
8.4
135
B125S
Filtrate
<1.0
<2.0
<8.0
<6
16.8
<70.0 "
<40
<3.0
Suspended
solids
0.34
2.13
Note: Blanks indicate not detected.
-------
TABLE A-9. SUMMARY OF ICAP RESULTS FOR PLANT EFFLUENTS IN PHASE III
H
ts>
B147S
Elements
Be
Cd
Cr
Cu
Ni
Pb
Zn
Ag
Filtrate
<1.0
13.6
42.7
12.2
71.6
<70
<40
. 4.52
Suspended
solids
19.8
24.2
3
B133S
Filtrate
<1.0
2.0
25.3
<6
<15
<70
<40
<3
B141S
Elements
Be
Cd
Cr
Cu
Ni
Pb
Zn
Ag
Filtrate
<1.0
3.20
93.3
123
<15
<70
156
4.38
Suspended
solids
1.46
106
248
80.8
66
11
3.17
Suspended
solids
25.7
B143S
Filtrate
<1.0
2
<8
<6
138
<70
65
6.92
C169S
Filtrate
<5
<10
355
<30
<75
386
<200
<15
Suspended
solids
174
Suspended
solids
26.5
47.1
29
B126S
Filtrate
<1.0
2,290
30.9
7.9
<15
<70
332
4.19
Suspended
solids
420
62.2
27.2
338
3.32
B142S
Filtrate
<5
<10
<40
70.8
<75
<350
<200
<15
Suspended
solids
33.6
173
48.2
614
1.4
Note: Blanks indicate not detected.
-------
the bay program and state management staffs with a control op-
tion; e.g., further clarification of wastewaters containing high
levels of toxic metals may reduce the discharge toxicity.
Other readings on this subject are given below:
Hoffman, M. R., et al. "Characterization of Soluble and
Colloidal-phase Metal Complexes in River Water by Ultra-
filtration. A Mass Balance Approach." ES&T, 15(6), 1981.
Nelson, P. 0., et al. "Factors Affecting the Fate of Heavy
Metals in the Activated Sludge Process." JWPCF, 53(8), 1981.
• Sposito, G. "Trace Metals." ES&T, 15(4), 1981.
Oakley, S. M., et al. "Model of Trace-Metal Partitioning
in Marine Sediments." ES&T, 15(4), 1981.
A.3 BIOACCUMULATION POTENTIAL TESTING
The purpose of performing the bioaccumulation potential test is
to determine if there are any organic compounds in a methylene
chloride extract of the effluent sample are potentially accumu-
lative in fatty tissues. Compounds having bioaccumulation poten-
tial are those which elute from the HPLC with a log P S3.5 and a
response £25% of full-scale deflection (FSD).
A.3.1 Phase II Study
During completion of Phase II an effort was made to identify spe-
cific compounds which caused the positive bioaccumulation potential
response as measured by the HPLC. The following information was
used to conduct this identification effort:
Log P values determined by HPLC for each effluent sample,
Organic compounds identified in each sample by GC/MS analyses,
A-13
-------
Published log P values for the GC/MS identified compounds
and also for those compounds tentatively identified by
process engineering analysis, and
Compounds identified in the literature having log P values
similar (within ±0.05 units) to the sample values measured
by HPLC.
For each effluent sample these data were tabulated in a manner
which permitted comparison of the HPLC log P values with those
for the presurvey compounds and the GC/MS identified compounds.
Using the results of this comparison, it was possible to tenta-
tively identify several compounds which caused a positive bio-
accumulation response. The number of potentially bioaccumulative
compounds in the effluent samples ranged from 5 to 17.
Based on the results of this effort, it was concluded that this
procedure was viable as a means of tentatively identifying com-
pounds having bioaccumulation potential, and that correlation of
sample analytical data with published information should continue
during Phae III of the study. Also, at the conclusion of Phase II
it was recommended that correlations between GC/MS-identified
organics and bioaccumulative organics should be developed further
by collecting the positive bioaccumulation organics as they elute
from the HPLC and subsequently analyzing them by GC/MS.
A.3.2 Phase III Study
During Phase III of the study, the Phase II recommendations with
respect to continued bioaccumulation potential testing were imple-
mented. As previously described, this involved continued measure-
ment of sample log P values by HPLC, identification of organic
compounds present in the sample by GC/MS, and also GC/MS analyses
of several HPLC fractions showing positive bioaccumulation response.
A-14
-------
In general, evaluation of the HPLC and GC/MS data proceeded as in
Phase II. Data were tabulated in a manner which permitted com-
parison of the HPLC log P data with those GC/MS-identified organic
compounds for which literature log P values could be found.
Where these log P values (established by different methods) cor-
related within ±0.05 units, the GC/MS compound was tentatively
identified as a potentially bioaccumulative compound. In addi-
tion, an effort was made to correlate HPLC data with literature
log P values and their corresponding compounds based on the
presence of structurally similar compounds in the sample. These
efforts were hindered by the relatively small body of log P data
which is readily available in the literature.
To illustrate the data analysis which resulted in tentative iden-
tification of potentially bioaccumulative compounds, the HPLC and
GC/MS data for the sample from Plant C150D are discussed herein
as an example. Table A-10 lists the data which is used in making
the correlations and tentative identifications. Similar tabular
data for the remaining samples are given in Appendix D.
As shown in Table A-10, a literature log P value was found for
only a single GC/MS-identified compound (dichlorobenzene). One
of the reported log P values for this compound is 3.62, which
correlates quite closely with a value of 3.61 which was identi-
fied by HPLC. On the basis of this correlation, which falls well
within the ±0.05 interval, dichlorobenzene is tentatively identi-
fied as a potentially bioaccumulative compound. Since literature
log P values were not available for the other 11 GC/MS-identified
compounds, no correlations could be drawn between these and the
remaining HPLC data.
The next step in the data analysis was to compare the HPLC data
with the listing of literature log P values in increasing order.
Each of the HPLC values (±0.05) was compared to the log P listing,
and those listed compounds falling within the interval and having
A-15
-------
TABLE A-10.
CORRELATION3 OF BI ©ACCUMULATION POTENTIAL** AND
GC/MS ANALYSIS TEST RESULTS: PLANT C150D
Organic compounds iden-
tified by GC/MS analysis
Dichlorobenzene
( Pentachlorocyclohexene ) e
(1,2, 3-Trichlorobenzene ) e
(2 ,4-Dichlorotoluene)
(1,2,3, 4-Te trachlorobenzene )e
(Pentachlorophenol)
(Hexachlorobenzene )
Molecular
formula
C6H4C12
C6H5C15
CeHsCls
C7H6C12
C6H2C14
C6HC150
C6C16
Literature
log P value [1]
3.62C'd
3.85
4.11
4.24
4.64
5.12
6.27
Actual sample
log P values
by HPLC method
3.61d
3.85
4.12
4.21
4.61
4.76
4.87
5.09
5.22
5.51
5.70
6.28
6.80
7.10
2-chloro-6-methyl-phenol
Benzeneacetic acid
Benzenepropanoic acid
l-H-Purine-2,6-dione, 3,7-
dihydro-1,3,7-trimethyl-
Hexanedioic acid, dioctyl
ester
C7H7C10
C9H10°2
C22H42°4
(continued)
A-16
-------
TABLE A-10 (continued)
Organic compounds iden- Molecular
tified by GC/MS analysis formula
Literature
log P value [1]
Actual sample
log P values
by HPLC method
l-methyl-(l-methylethyl)-ben-
zene C10H14
Ethyl-dimethyl-bezene CioH14
Benzeneethanol CgH100
1,7,7-Trimethyl-, bicyclo
[2.2.1]heptan-2-ol C10H180
Cyclohexanol, 5-methyl-2-(l-
methylethyl)-, la,20,50)- C10H200
3-Cyclohexene-l-methanol, a,a,
4-trimethyl-, (S)- C10H180
Potential correlation is defined as log P values agreeing ±0.05 units.
Data in this table are only for compounds which may bioaccumulate (log P
£3.5 and £25% FSD).
C0ther values reported are 3.38, 3.39, 3.55, 3.60.
i>otential correlation.
tentative identification based solely on log P values and similar compounds
found in sample.
No literature log P found.
[1] Hausch, C., Computerized Printout of Log P Values by Increasing Molecular
Carbon Content and Increasing Log P, 1980.
A-17
-------
a structure similar to GC/MS-i
-------
TABLE A-ll. CORRELATION OF POSITIVE RESPONSE HPLC
FRACTIONS AND GC/MS ANALYSES3
HPLC data
GC/MS identified organics
Fraction
Log P
Compound
Literature
log P [1]
Sample B149S
3.63 (Naphthalene)
3.79C Biphenyl
1- and 2-methylnaphthalene
Tetramethyl benzene
4.25, 4.49° Dimethyl naphthalene
4.93 Tetramethyl naphthalene
5.02 Dimethyl tetrahydronaphthalene
5.75 n-Dodecane
Sample B112D
4.21 Acenaphthalene
Fluorene
4.32° Dimethyl naphthalene
4.49C Phenanthrene
3.76"
3.86, 3.87
4.0
4.26 -,4.44
d
"d
"d
4.18"
4.18C e
4.26 - 4.44
4.46°
Potential correlation defined as log P values agreeing ±0.05 units.
Tentative identification based solely on log P value and similar
compounds found in sample.
Potential correlation.
N
literature log P found.
eOnly fraction one contained significant amounts of chromatographable
compounds .
Taken as a whole, these data demonstrate that HPLC and GC/MS
analyses used in conjunction with published information can be
utilized to tentatively identify specific compounds which corre-
spond to bi ©accumulative (potential) positive responses. In the
effluent samples tested, the number of potentially bioaccumula-
tive compounds ranged from zero to nine, and on the average it
was possible to identify approximately 33% of those species which
had bioaccumulation potential.
A-19
-------
A.4 REFERENCES
1. Hausch, C., Computerized Printout of Log P Values by Increas-
ing Molecular Carbon Content and Increasing Log P, 1980.
A-20
------- |