EPA-600/2-81-0555
April !.F,81
EVALUATION OF PCS DESTRUCTION
EFFICIENCY IN AN INDUSTRIAL BOILER;
AUDIT REPORT
by
P. f. Collins and G. F. Hunt
Research Triangle Institute
P.O. Box 12194 .
Research Triangle Park, NC 27709
EPA Contract No. 63-02-3146
Task No. 129
EPA Task Officer
0. Sanchez
Industrial Environmental Research Laboratory
Office of Environmental Engineering and Technology
Research Triangle Park., Worth Carolina 27721
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH ANO DEVELOPMENT
WASHINGTON, O.C. 20460
-------
ABSTRACT
Systems audits and an evaluation of the quality of data obtained by GM
and GCA in the analysis of a test burn oil for PC3 were conducted ty two
members of the Environr, ental Quality Assurance Department OT the Research
Triangle Institute during the week of August 1&, 1S80. Audits were conducted
Dy inspection of available documentation and records, discussion of analytical
methodology and data with personnel of the organization being judited, and
independent data reduction. The analytical data reported by 'JM and GCA were
subsequently confirmed by separate analyses performed by the Analytical Chem-
ist -y Branch of EPA/HERI.-RTP. The results are reported in AppendU A.
This report was submitted in fulfillment of Task Directive No. 129 on EPA
Contract No. 68-02-3146 by the Resea.vh Triangle Institute. This report
covers the period August 6, 1900, to September 30, 1980, and work //as completed
as of October 1980.
ii
-------
CONTENTS
Abstract ii
Figures iv
Tables iv
1 INTRODUCTION 1
2 SUMMARY 2
3 AUDIT PROCEDURLS 4
4 AUDIT RESULTS 9
4.1 Audit of GM Data 9
4.1.1 Sampling 9
4.1.2 Sample Custody 9
4.1.3 Sample Analysis 9
4.1.4 Preparation of Test Burn Oil 10
4.2 Audit of GCA Data 12
4.2.1 Sampling 12
4.2.2 Sample Custody 12
4.2.3 Sample Analysis 12
5 CONCLUSIONS AND RECOMMENDATIONS 16
5.1 Conclusions 16
5.2 Recommendations 17
APPENDIX A: Memorandum 18
iii
-------
FIGURES
NuTber Page
1 Checklist used at GM 5
2 Checklist used at GCA 7
TABLES
Page
Summary of PCB Concentrations Found in
Test Burn Oil by GM and GCA 3
2 Results Obtained by GM for PCB in Test Burn Oil 11
3 GM Data Related to Preparation of Test Burn Oil 13
4 Results Obtained by GCA for PCB in Test Burn Oil 15
iv
-------
SECTION 1
INTRODUCTION
In May 1980, a verification test burn for destruction of polychlorinated
biphenyls (PCB), was cone! icted to evaluate the PCB destruction efficiency of
the No. 3 industrial boiler at the Chevrolet plant in Bay City, Michigan. The
verification burn was conducted by General Motors Corporation, with GCA
Corporation serving as a contractor to the U.S. Environmental Protection
Agency (USEPA) to provide and operate the flue gas and ambient air sampling
equipment and to perform the necessary analytical work.
The verification burn fired a reclaimed oil that had been prepared (spiked)
by GM to contain approximately liOO ppm PCB as Aroclor 1242. Analysis of this
test burn oil for PCB was condurtud by both GM and GCA, and the results of
these analyses were in questionable agreement. Consequently, the Research
Triangle Institute (RTI) was renuosted by USEPA to perform systems audits and
to establish the quality of date, obtained by GM and GCA in the analysis of ti.e
test burn oil for PCB.
During the week of August 18. 198C, two auditors from the Environmental
Quality Assurance Department of RTI visited the Central Office, Chevrolet
Motor Division, General Motors Corporation, Warren, Michigan, and the Technology
Division of GCA Corporation in Bedford, Massachusetts, to perform the systems
audits. This report includes a description of the audit procedures used, the
audit results, and conclusions and recommendations.
Subsequent to RTI's systems audit and in response to the audit's conclu-
sions and recommendations, analyses of the PCB reference standard material and
the split samples of the test burn oil were performed by the Applied Chemistry
Branch of thu Health Effects Research Laboratory, EPA-RTP. The results of
these analyses are presented in Appendix A.
The assistance and cooperation of the personnel at GM and GCA are acknowl-
edged with appreciation.
-------
SECTION 2
SUMMARY
Systems audits and an evaluation of the Quality nf data obtained by GK
and GCA in the analysis of a test burn oil for P-D were conducted by two
members of the Environmental Quality Assurance Department of the Research
Triangle Institute during the week of August 18, 1980. Audits were conducted
by inspection of available: documentation and records, discussion of analytical
methodology and data with personnel of the organization being audited, and
independent data reduction. A checklist was used as a guide in conducting the
audits.
An examination of the available- documentation regarding sampling and
sample custody indicated no records were available at either facility for the
test burn oil sample f«"oni the time it was taken at the Bay City plant until it
was received in the laboratories.
The analytical results obtained by GM and GCA are summarized in Table 1,
along with the Aroclor 1242 used for calibration or as a reference standard
for each quantisation. In some cases, the analytical result was calculated by
the RTI auditors, using data supplied by the organizations. II was concluded
that the analytical methodology used at each organization is acceptable and is
not primarily responsible for the interlaboratory difference in the results.
Rather, the differences are most likely attributable to the different Aroclor
1242 materials used for calibration or as reference standards. This can only
be confirmed by examination by one laboratory of all the Aroclor 1242 refer-
ence materials used in the analyses at GM and GCA.
Subsequent to the audit, the reference materials and sample of test burn
oil from each organization were reanalyzed by EPA/HERL-RTP, as described in
Appendix A. Those results for the test burn oil were intermediate between
results obtained by GM and GCA, suggesting that the difference between results
by GM and GCA is due to analytical variability, which is often encountered in
the analysis of a complex mixture such as Aroclor 1242.
-------
P381 -137270
Evaluation of PCB Destruction Efficiency in an
Industrial Boiler: Audit Report
Research Triangle Inst.
Research Triangle Park, NC
Prepared for
Industrial Environmental Research Lab
Research Triangle Par«c, NC
Apr 81
vsrrv^-jw^. i^>...-r^CMV».-»-Kt- 7ite—» »a «-^iarVlrm'*y
BECK
U.S. Dspartment of Commerce
National Technical Information Service
-------
United States
Environmental Protection
Agency
EPA-600/2-81-055b
April 1981
EVALUATION OF PCB DESTRUCTION
EFFICIENCY IN AN INDUSTRIAL BOILER:
AUDIT REPORT
Prepared for
Office of Pesticides and Toxic Substances
Prepared by
Industrial Environmental Research
Laboratory
Research Triangle Park NC 27711
"NATIONAL TECHNICAL
, INFORMATION SERVICfc
US H»*H«« Of COHKtlCl
am
^ •
-------
TECHNICAL REPORT DATA
/Plcesf rrad lassniciiuns on the retftc before completing)
RE°ORT NO
EPA-600/2-81-055b
TITLE AND SUBTITLE
Evaluation of PCB Destruction Efficiency in an Indus-
trial Boiler: Audit Report
3 RECIPIENTS ACCESSION NO
6 REPORT DATE
April 1981
6 PERFORMING ORGANIZATION CODE
AUTHOR S)
P.F. Collins and G. F. Hunt
8 PERFORMING ORGANIZATION REPORT NO
PERFORMING ORGANIZATION NAME AND ADDRESS
Research Triangle Institute
P.O. Box 12184
Research Triangle Park, North Carolina 27709
10 PROGRAM ELEMENT NO
C1YLIB
11 CONTRACT/GRANT NO
68-02-3146, Task 129
12 SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND, PEI
Task FL.al; 8-9/80
RIOD COVERED
14 SPONSORING AGENCY CODE
EPA/600/13
is SUPPLEMENTARY NOTES ffiRL-RTP project officer is David C. Sanchez, Mail Drop 62,
919/541-2547.
16 ABSTRACT
The report gives resuics of systems audits and an evaluation of the quality
of data obtained by GM and GCA in the analysis of a test burn oil for PCB conducted
by Research Triangle Institute. Audits included inspection of documentation and
records, discussion of analytical methodology and data with personnel of the organ-
ization being audited, and independent data reduction. The analytical data reported
by GM and GCA were subsequently confirmed by separate analyses by EPA's Health
Effects Research Laboratory (RTP) and are reported in Appendix A.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
ti IDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
Pollution
Chlorine Aromatic Compounds
Biphenyl
Fuel Oil
Combustion
Auditing
Analyzing
Pollution Control
Stationary Sources
Poly chlorinated Bi-
phenyls
13B
07C
21D,11H
21B
05A
14 B
13 DlSTOigijTION STATEMENT
Release to Public
19 SECURITY CLASS tThil Reporl)
Unclassified
21 NO OF PAGES
32
20 SECURITY CLASS (This page)
Unclassified
22 PRICE
EPA Form 2220-1 (9-73)
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development. U S Environmental
Protection Agency, have been grouped into nine series These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields
The nine series arc:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research ar.d Development
8 "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
EPA REVIEW NOTICE
This report has been reviewed by the U S. Environmental Protection Agency, and
approved for publication. Approval does not signify that the contents necessarily
reflect the views and policy of the Agency, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service. Springfield. Virginia 22161.
-------
EPA-600/2-8l-055b
April IfiBl
EVALUATION OF PCB DESTRUCTION
EFFICIENCY IN AN INDUSTRIAL BOILER:
AUDIT REPORT
by
P. F. Collins and G. F. Hunt
Research Triangle Institute
P.O. Box 12194
Research Triangle Park, NC 27709
EPA Contract No. 63-02-3146
Task No. 12S
EPA Task Officer
0. Sanchez
Industrial Environmental Research Laboratory
Office of Environmental Engineering and Technology
Research Triangle Park, North Carolina 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND DEVELOPMENT
WASHINGTON, O.C. 20460
-------
ABSTRACT
Systems audits and an evaluation of the quality of data obtained by GM
and GCA in the analysis of a test burn oil for PCB were conducted t-y two
members of the Environmental Quality Assurance Department ot the Research
Triangle Institute during the week of August 1C, 1980. Audits were conducted
by inspection of available documentation and records, discussion of analytical
methodology and data with personnel of the organization being judited, and
independent data reduction The analytical data reported by 'JM and GCA were
subsequently confirmed by separate analyses performed by the Analytical Chem-
istry Branch of EPA/HERL-RTP. The results are reported in Appendix A.
This report was submitted in fulfillment of Task Directive No. 129 on EPA
Contract No. 68-02-3146 by the Research Triangle Institute. This report
covers the period August 6, 1900, to September 30, 1980, and work was completed
as of October 1980.
ii
-------
CONTENTS
Abstract ii
Figures iv
Tables iv
1 INTRODUCTION 1
2 SUMMARY 2
3 AUDIT PROCEDURES 4
4 AUDIT RESULTS 9
4.1 Audit of GM Data 9
4.1.1 Sampling 9
4.1.2 Sample Custody 9
4.1.3 Sample Analysis 9
4.1.4 Preparation of Test Burn Oil 10
4.2 Audit of GCA D*ta 12
4.2.1 Sampling 12
4.2.2 Sample Custody 12
4.2.3 Sample Analysis 12
5 CONCLUSIONS AND RECOMMENDATIONS 16
5.1 Conclusions 16
5.2 Recommendations 17
APPENDIX A: Memorandum 18
in
-------
FIGURES
Page
Checklist used at GM 5
Checklist used at GCA 7
TABLES
Page
Summary of PCB Concentrations Found in
Test Burn Oi 1 by GM and GCA 3
2 Results Obtained by GM for PCB in Test Burn Oil 11
3 CM Data Related to Preparation of Test Burn Oil 13
4 Results Obtained by GCA for PCB in Test Burn Oil 15
IV
-------
SECTION 1
INTRODUCTION
In May 1980, a verification test burn for destruction of polychlorinated
biphenyls (PCB), was cond'icted to evaluate the PCB destruction efficiency of
the No. 3 industrial boiler at the Chevrolet plant in Bay City, Michigan. The
verification burn was conducted by General Motors Corporation, with GCA
Corporation serving as a contractor to the U.S. Environmental Protection
Agency (USEPA) to provide and operate the flue gas and ambient air sampling
equipment and to perform the necessary analytical work.
The verification burn fired a reclaimed oil that had been prepared (spiked)
by GM to contain approximately liOO ppm PCB as Aroclor 1242. Analysis of this
test burn oil for PCB was conducted by both GM and GCA, and the results of
these analyses were in questionable agreement. Consequently, the Research
Triangle Institute (RTI) was reoucsted by USEPA to perform systems audits and
to establish the quality of dato obtained by GM and GCA in the analysis of tlie
test burn oil for PCB.
During the week of August 18, 198C, two auditors from the Environmental
Quality Assurance Department of RTI visited the Central Officp, Chevrolet
Motor Division, General Motors Corporation, Warren, Michigan, and the Technology
Division of GCA Corporation in Bedford, Massachusetts, to perform the systems
audits. This report includes a description of the audit procedures used, the
audit results, and conclusions and recommendations.
Subsequent to RTI's systems audit and in response to the audit's conclu-
sions and recommendations, analyses of the PCB reference standard material and
the split samples of the test burn oil were performed by the Applied Chemistry
Branch of the Health Effects Research Laboratory, EPA-RTP. The results of
these analyses are presented in Appendix A.
The assistance and cooperation of the personnel at GM and GCA are acknowl-
edged with appreciatic-i.
-------
SECTION 2
SUMMARY
Systems audits and an evaluation of the quality o-f data obtained by GM
and GCA in the analysis of a test burn oil for PLB were conducted by twc
members of the Environmental Quality Assurance Department of the Research
Triangle Institute during the week of August 18, 1980. Audits were conducted
by inspection of available documentation and records, discussion of analytical
methodology and data with personnel of the organization being audited, and
independent data reduction. A checklist was used as a guide in conducting the
audits.
An examination of the available documentation regarding sampling and
sample custody indicated no records were available at either facility for the
test burn oil sample f«-oni the time it was taken at the Bay City plant until it
was received in the laboratories.
The analytical results obtained by GM and GCA are summarized in Table 1,
along with the Aroclor 1242 used for calibration or as a reference standard
for each quantisation. In some cases, the analytical result was calculated by
the RTI auditors, using data supplied by the organizations. It was concluded
that the analytical methodology used at each organization is acceptable and is
not primarily responsible for the interlaboratory difference in the results.
Rather, the differences are most likely attributable to the different Aroclor
1242 materials used for calibration or as reference standards. This can only
be confirmed by examination by one laboratory of all the Aroclor 1242 refer-
ence materials used in the analyses at GM and GCA.
Subsequent to the audit, the reference materials and sample of test burn
oil from each organization were reanalyzed by EPA/HERL-RTP, as described in
Appendix A. Those results for the test burn oil were intermediate between
results obtained by GM and GCA, suggesting that the difference between results
by GM and GCA is duo to analytical variability, which is often encountered in
the analysis of a complex mixture such as Aroclor 1242.
-------
TABLE 1. SUMMARY OF PCB CONCENTRATIONS FOUND IN
TEST BURN OIL BY GM AND GCA*t
Organization and
analysis method
GM-
GM-
GC/EC
GC/MS
GM: GC/MS, standaM
addition
GCA-
GCA:
GC/EC
GC/EC
Aroclor 1242
reference material
Monsanto Aroclor 1242
Monsanto Aroclor 1242
Monsanto Aroclor 1242
Supelco C/N 4-4803
EPA/RTP Reference,
Aroclor 1242 found,
(jg/g (ppm)
496+ (493)
529, 539t
510t (486)
-- (471)
750 (738)
GCA: GC/EC
GCA: GC/EC
Lots 8337, 8758
(Monsanto Aroclor 1242)
Applied Sc^snce,
Lot #478
EPA/EMSL-Ci
QC Samples, WP679,
Cone. 7 and 8
750 (731)
-- (538-590)
*Following the audit, the sample of test burn oil that each organization had
analyzed was reanalyzed for Aroclor 1242 by EPA/HERL-RTP, as described in
Appendix A. Results were- GM sample, 614 ppm; and GCA sample, 640 ppm.
rvalues in parentheses were calculated by RTI auditors from data supplied by
the applicable organization.
fData provided by GM indicate that results would be 8 percent higher if
referenced to EPA Reference Aroclor, Lot 5041, from EPA/HERL-RTP.
-------
SECTION 3
AUDIT PROCEDURES
The procedures used for the audits of the analytical data obtained by GM
and GCA for PCB in the test burn oil were designed to evaluate all aspects of
the data gathering process, including sampling, sample custody, sample analysis,
and data reduction. As a guide in conducting the audit, the checklists shown
in Figure 1 (GM) and Figure 2 (GCA) were completed (as applicable). Each
audit included inspection of the documentation, discussion of analytical
methodology and data, and independent data reduction, which sprved to verify
the calculations. Each analytical laboratory was also visited. At the conclu-
sion of each audit, a debriefing session was held with representatives of the
organization.
-------
COMPOUND:
•*
i
let *«
(2) Analysis Procedure:
(a) Here sample preparation steps described? X
(b) Were instrument parameters documenteo? /
ff/e/. *r"«c*.^M,*-e.- J»9r&t~.t,'(t,"3 r ••**!'
*»•*.&. S*sl. 4+J ,'<.ff
-------
Yes No
(3) Standardization Procedures:
(a) Was purity of standard compound verified?
Ff^. f fA -£<.£ jfj. -
(b) Has a fi*e-point calibration curve prepared? / _
(c) Was instrument, calibration checked daily? i/ _
If not, at what interval?
(d) Did calibration curve bracket sample concen-
tration? -^ _
(4) Recording Procedures:
(a) Were data sheets and notebooks signed and
dated? _ •/
(b) Were data entered properly? _ t/
(c) Were sample calculations shown with proper
unit designations? / _ ^
(1) Injection volumi2 corrections made? >/ _
(2) Dilution and/or concentration
corrections made? S _
(d) Was data reviewed? _ i/
If so, how frequently? _ _
COMMENTS
Figure 1 (continued)
-------
COMPOUND: "TVri flux*. O> I
Plant/City/State: T*
Sample Point Location:
Date Sampled: #»t/c.*«viu'ix
Date Analyzed: £ - i*t!3 - BQ
Yes No
(1) Was chain of custody procedure followed:
(2) Analysis Procedure:
(a) Were sample preparation steps described? J[ _
(b) Were instrument parameters documented? _/ _
(c) Was compound identified on chromatogram? j/ _
(d) How was data compiled?
(1) Manually? _
(2) Electronically? /
(e) How was data quantitated?
(1) Manually? i/
(2) Electronically? _
(f) Did retention times of sample, standard and
spiked sample for compound agree?
-------
Yes No
(3) Standardization Procedures:
(a) Was purity of standard compound verified? _
(b) Was a -five-point calibration curve prepared? '
(c) Was instrument calibration checked daily? J
If not, at what interval?
(d) Did calibration curve bracket sample concen-
tration?
-------
SECTION 4
AJDlf RESULTS
The results of the systems .vjdits for data quality evaluation conducted
at GM on August 18-19, 1S80, and at GCA on August 20-21, 1980, are presented
in this section.
4.1 AUDIT OF GM DATA
4.1.1 Sampling
Documented information concerning the sampling of the test burn oil was
not available to the auditors, r.-om discussions with GM and GCA personnel, it
appears that the oil v»as sample-1 by GM-Bay City, with one portion of the
sample given to the onsite GCA fie id crew and one portion transported to the
GM-Warren Laboratory.
4.1.2 Sample Custody
Documentation concerning custody of the oil sample prior to receipt in
the GM laboratory was not available to the auditors. Once the sample was
received in the laboratory, it was logged into a notebook and then maintained
under tho custody of the analyst. A work order number was assigned to the
sample that referenced a computer file containing the receiving date. It was
stated that the sample would be retained for 6 months before disposal.
4.1.3 Sample Analysis
4.1.3.1 Analysis by GC/EC--
The test burn oil sample was initially analyzed by GM using gas chroma-
tography with an electron capture detector (GC/EC). The documentation of the
procedure given to the auditors was a hand-written outline of GC operating
conditions and instructions for standard preparation. No sample cleanup was
used; the oil sample was simply weighed and diluted with hexane prior to
injection into the chromatograph. A 0.5 ug/mL solution of Monsanto Aroclor
1242 in hexane was used as the standard.
-------
The sample was quantitated using the peak height of the major peak in the
chromatogram; this w^s i.easured using a valley baseline correction. Instead
of using a calibration curve for data calculation, the intent was to adjust
the injection volumes of the sample and standard to obtain approximately the
same peak heights and then to calculate the sample concentration by ratio.
This wab achieved in the analyses of the test burn oil where the peak heights
were 6.7 and 6.75 cm for the standard and sample, as measurer! oy the auditors.
When the peak height of the standard is considerably different fron *~hat of
the samples error may be introduced by using the ratio method of calculation.
To evaluate this, the auditors requested that a calibration curve be prepared.
It was found that the peak heights are proportional to the amount of Aroclor 1242
injected over the range of 0.5 to 2.5 ng, hut the calibration curve does not
pass through the origin, as is assumed in the ratio method of calculation.
However, this error is not present in the analytical data for the test burn
cil.
As noted above, the standard was prepared from Aroclor 1242. At the
request of the auditors, a standard so prepared was compared to a standard
prepared from EPA Reference Aroclor 1242, Lot No. 5041, obtained on February 10,
197G. Using the chromatograms, the auditors calculated that the 0.50-ppm
stdndard used by GM was 0.54 ppm relative to the EPA reference material.
The results obtained for the test burn oil sample using the Monsanto
standard are shown in Table 2. The result obtained bv GC/EC is from a single
analysis.
4.1.3.2 Analysis by GC/MS—
The test burn oil was also analyzed by GM using gas chromatography/mass
spectrometry (GC/MS), again using a standard prepared frcm Monsanto Aroclor
1242. Mass 256 was used to monitor Aroclor 1242 and mass 172 used to monitor
the 2-fluorobiphenyl, which was added as an internal standard. A standard
addition experiment was conducted in which 250 and 500 ppm of Aroclor 1242 (on
the initial oil basis) was added to the sample. The results of the GC/MS
analysis are included in Table 2.
4.1.4 Preparation of Test Burn Oil
The data on the preparation of the test burn oil were examined to deter-
mine if the analytical results were consistent with the amount of PCB added.
10
-------
TABLE 2. RESULTS OliTAIHED BY LM FOR
PCS IN TEST BURN OIL*
Method Aroclor 1242 Found (pg/gH
GC/EC 496 (493)
GC/MS 529, 339
GC/MS, standard addition 510 (486)
"Values in parentheses were calculated by RTI auditors from
data supplied by GM.
TValues were obtained using Monsanto Aroclor 1242 as the
standard. Based on a comparison of standards prepared f^om
Monsanto and EPA Reference Aroclor 1242, Lot No. 5041, the
results would be 8 percent higher if the EPA material had been
used as the standard.
11
-------
The preparation of the test burn oil involved initial analysis of a let of
reclaimed oil for PCB and then addition, in two stages, of capacitor fluid
that contained 93 percent Aroclor 1242. based on GM analysis. This mixture
was analyzed using GC/EC. The data are summarized in Table 3.
4.2 AUDIT OF GCA DATA
4.2.1 Sampling
No documented infotmat.ion was made available concerning the sampling of
the test burn oil. After examining the sample container, GCA personnel con-
cludet! that the sample had been t-,ken by GM and then transferred to the on-
site GCA field crew.
4.2.2 Sample Custody
No records relating to the custody of the sample prior to receipt at the
GCA facility in Bedford were tiv,iilable. Upon receipt at Beofcrd, the sample
was logged in, and from this po'nt, sample custody appears satisfactory with
adequate documantation.
4.2.3 Sample Analysis
The test burn sample was analyzed t>y GCA using gas chromatography with
electron capture detection. Prior to analysis, a portion of the sample was
weighed, diluted with hexane, and treated using a sulfuric acid cleanup pro-
cedure. The cleaned, diluted sample was th^n injected into the chromatograph.
Adequate documentation of the analysis procedure was provided. The standards
used initially were prepared from an Aroclor 1242 stock solution obtained from
Supelco, Inc.
Quantitation of the chromatograms was done by summing the areas of five
major peaks. A calibration curve was prepared daily relating total area to
concentration of Aroclor 1242 injected. Data examined by the auditors yielded
calibration curves of satisfactory linearity and the sample was bracketed by
standards.
Quality control measures used in this aiiilysis included spiking of No. 6
fuel oil with Aroclor 1242, blind checks using Aroclor 1242 reference materials
from EPA/KERL-RTF, Lots 8758 and 8937, and the use of quality control check
samples procured from EPA/EHSL-Ci. A stock solution of Aroclor 1242 from
12
-------
TABLE 3. GM DAIA RELATED TO PREPARATION
OF TEST BURN OIL
Aroclor 1242, ug/g
Sample Found Expected
Initial reclaimed oil, 132.5 —
(430 qal. 1,489 kg)
Oil after initial spike
with 408 g of Aroclor
1242* 370 407
Oil after second spike
with 103 g of Aroclor
1242* 496 476
*Weight of added Aroclor 1242 calculated using a specific
gravity of 1.3695 and a value of 98 percent Aroclor 1242 in
the capacitor fluid.
13
-------
Applied Science Labs, Inc., in addition to th
-------
TABLE 4. RESULTS OBTAINED BY CCA
FOR PCB IN TEST BURN OIL*
Aroclor 12*2 found
Reference material in test burn oil
Supelco, Inc., C/N 4-4803 — (471)
(Kit PCB-A-21)
EPA/H2RL-RTP Reference
Aroclor 1242, Lots 8937,
8758 750 (738)
Applied Science Labs, Inc.
Lot #478 750 (731)
EPA/EMSL-Ci
QC Samples WP679.
Cone. 7 and 8 --- (538-590)
"Values in parentheses were calculated by RTI auditors
from data supplied by GCA.
15
-------
SECTION 5
CONCLUSIONS AND RECOMMENDATIONS
5.1 CONCLUSIONS
1. The recordkeeping and procedural documentation related to the sampl-
ing of the test burn oil and to sample custody prior to receipt of
the samples in the laboratories was inadequate, thus jeopardizing
the analytical data.
2. The analytical methodologies used by each of tte laboratorias,
though different, appear acceptable and comparable in the case of
th
-------
dor 1242 obtained from Monsanto. Subsequent to the audit, EPA/HERL-
RTP analyzed a sample of the test burn oil from each laboratory and
also reference materials that liad been used (GM, Monsanto Aroclor 1242
and GCA, a solution of EPA/HERL-RTP reference 1242, Lot 8937). The
results indicated that there was no significant difference between
the standards, and the results for 1242 in the test burn oil were
614 ppm for the GM sample and 640 ppm for the GCA sample. Thus, the
differences among the laboratories appear to represent the analytical
variability that is inherent ir. the analysis for complex mixtures
such as Aroclor 124?.
5. The analytical data obtained at GM and GCA are consistent with the
data obtained in the preparation of the test burn oil.
5.2 RECOMMENDATIONS
1. It is recommended that in any future projects involving PCB analysis,
the analytical procedures and standards to be used be established
before analysis is begun. Also, it is desirable to analyze critical
samples more than once.
2. It is recommended that in projects involving sampling and analysis,
all related activities should be fully documented and that valid
sample chain-of-custody procedures be instituted
17
-------
APPENDIX A
18
-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
September 10, 1980
SUBJECT Results of Analyses of PCB Standards
F«OM Merrill Jackson, Research Chemist TTU7^ i**"*
ACB/ETD/HERL (MD-69) V '
TO David Sanchez, IERL (MD-62)
THRU: Dr. Robert G. Lewis, Chief
Analytical Chemistry Branch/ETD/HERL (MD-69)
This is to report to you the results of our analyses of PCB stand-
ards and PCB contaminated oil used in the General Motors PCB boiler
burn.
On Sept. 3, 1980, Dr. Peter Collins of RTI delivered to me an
Arochor 1242 standard which he had received from GM and a sample of the
diluted 1242 standard used by GCA in their analyses. 'I requested a new
1242 standard from the E!'A repository and ona from lot 9224 was delivered
on this date. On Sept. 8, 1980, Dr. Collins delivered two PCB contami-
nated fuel samples, one each from those analyzed by GM and GCA.
I prepared stock solutions of the EPA and GM 1242 standards. The
GCA 1242 standard, was labeled as 6.7 ng/ul. Final dilutions of all
three standards were prepared so that the concentration was 1 ng/ul.
(This was the concentration needed for my gas chromatographic conditions.)
I did not perform any cleanup on the fuel oils. They were simply
diluted with hexane to the final concentration of 1 ng/ul of 1242 based
on a 500 ppm initial concentration.
Quantitation of the chromatograms was by totaling the peak heights
of all peaks present. Under our conditions Arochor 1242 has 14 major
peaks.
However, the diluted fuel had a major interference which blocked
out the last 4 peaks, therefore on the fuel analyses only the first 10
peaks were measured and compared to the first 10 peaks of the standards.
We found that the GM Arochor 1242 standard was 1.03 ng/ul and the
GCA was 0.97 ng/ul when compared to the EPA standard. Repeated GC
analyses (injections) of the EPA standard solution gave a result of
1.00^0.03 ng/ul. therefore there were no analytical differences be-
tween any of the 1242 standards. We found 614 ppm of 1242 in the GM oil
end 640 ppm in the GCA oil. Again we believe that the differences are
analytical and that an average of the two is valid.
I have attached copies of the chromatograms in the order in which
they were ran.
CC: Peter Collins, RTI
D. E. Gardner
19
EPA Fo>n 1J204 (Re,. 3-16}
-------
I I I I I I II I II I I I I I II I I I II I
Figure A-l. Chromatogram of 3.1 uL of EPA-1242, 1
20
-------
I
*
1^—T
-4
7&
lift 3
Lill''' " i ;| ;) i ' •
" pjftaitrii
\\\\ ;:
-;
-f
v= {"'•
\ I
Illllllllllllfllllllllllll
Figure A-2. Chroma tog ram of 3.1 vl of EPA-1242, 1 vg/ml.
21
-------
Illllllllllllllllllllll
Figure A-3. Chromatogram of 3.1 pL of GCA standard, 1 vg/ml.
22
-------
H:-f
! I
•
•
,.
--
-
; U;
- 4
- — t-
1 -
"-"i:
1
-
i u 111 MI 1111 ei 1111111111
Figure A-4. Chromatogram of 3.1 vl of GM standard, 1 pg/mL.
23
-------
iiiiiiiiiiiiiiiiiiiiiiiin
Figure A-5. Chromatogram of 3.1 vL of EPA-1242, 1 yg/mL.
24
-------
IIII6IIIIIIIIIMIIIIIMIIIIIC!
Figure A-6. Chromatogram of 2.1 yL of diluted GCA sample of test burn oil
25
-------
Ilillllllllll
Figure A-7. Chromatogram of 3.1 pL of EPA-1242, 1 yg/mL.
26
-------
111111111111111111111111111111 (
Figure A-8. Chromatogram of 2.1 uL of diluted GM sample of test burn oil,
27
------- |