Document qdO9ZEMvr1xEYRDG51Qm49nnx

Topics: PCB Chemical analysis Transformers Capacitors PCDF-PCDD Insulating oil EPRI EUEA-5440 Volume 3 Project 2028-10 Final Report June 1988 Analysis of Polychlorinated Dibenzofurans and Polychlorinated Dibenzo-p-Dioxins in Transformers and Capacitors Volume 3: Summary and Statistical Analysis of the Round-Robin Test Prepared by Research THangle Institute Research THangle Park, North Carolina GENP 011146 783957 REPORT SUMMARY SUBJECTS Hazardous and toxic waste management / Transmission substation design and operation / Distribution substations TOPICS PCB Chemical analysis Transformers Capacitors PCDF-PCDD Insulating oil AUDIENCE Environmental managers / Distribution engineers Analysis of Polychlorinated Dlbenzofurans and Polychlorinated Dibenzo-p-Dloxins in Itansform ers and Capacitors Volumes 1 -3 PCDF and PCDD-- by-products of PCB pyrolysis and combus tion-- apparently cause many of the health effects originally attributed to PCB. In parallel and round-robin studies, indepen dent laboratories found that normal equipment operation and arc ing failure cause little if any oxidation and do not significantly modify PCB fluids. BACKGROUND Partial oxidation of askarels-- mixtures of polychlorinated biphenyls (PCB) and tri- and tetrachlorobenzenes-- produces polychlorinated dibenzofurans (PCDF) and polychlorinated dibenzo-p-dioxins (PCDD). However, analytic techniques for separating the active components in these mixtures had not matured prior to this study. Information had also been insufficient on the quantity and types of PCDF and PCDD in as-manufactured PCB and on the formation of these by-products during normal operation or arcing failure in utility equipment. Because of these uncertainties, utilities and regulators have assumed the worst-case PCDF-PCDD formation when there is a FCBrelated accident, thereby overstating the potential toxicity. EPRI report EL/EA-4858 describes portions of the research performed in these projects. OBJECTIVE To improve techniques for measuring PCDF/PCDD in the presence of PCB. APPROACH Five laboratories (including two whose studies are discussed in report EL/EA-4858) performed round-robin gas chromatography-mass spectrome try analyses of PCDF-PCDD samples using specially prepared carbon-13 (13C) PCDF spiking compounds. After the first set of analyses, each labora tory modified its techniques and analyzed a larger group of samples from utility equipment. Another organization statistically analyzed all the labora tory results. RESULTS These three volumes report the second group of PC DF-PCDD analyses. Results from all participating laboratories correlated remarkably welt. Some of the results follow. EPRI EL/EA-5443a Vbls. 1-3 GENP 011147 783958  Use of the 13C compounds allowed successful separation of the com ponents in several closely eluting pairs of physiologically active PCDF and less-active compounds. Analyses of samples from utility equipment suggested that the PCDF-PCDD present did not result from high-load, high-temperature operation. Volumes 1 and 2 describe several analytic techniques, with the appen dix of Volume 2 including a report on the analysis by the University of Umea in Sweden, as well. Volume 3 summarizes all the laboratory techniques, presents comprehensive statistical analyses, and provides an executive summary. EPRI PERSPECTIVE This research is a landmark in PCDF-PCDD analysis. One important result is the synthesis of new PCDFs for use as spiking compounds in gas chromatography-mass spectrometry analysis. The ability to sepa rate many of the active PCDF compounds from more innocuous materi als is also of great value. The high correlation of results from several laboratories indicates that researchers can exercise great flexibility in selecting techniques and developing facilities. In addition, new work has started (EPRI project RP2028-21) in which the best features of each of the five analytic methods will be assembled to provide a single recommended method. PROJECTS RP2028-6, RP2028-7, RP2028-10 EPRI Project Managers: Gil Addis; Jacques Guertin Electrical Systems Division; Environment Division Contractors: New York State Department of Health; Battelle Columbus Laboratories; Research Triangle Institute For further information on EPRI research programs, call EPRI Technical Information Specialists (415) 855-2411. 783959 Analysis of Polychlorinated Dibenzofurans and Polychlorinated Dibenzo-p-Dioxins in Transformers and Capacitors Volume 3: Summary and Statistical Analysis of the Round-Robin Test EUEA-5443, Volume 3 Research Project 2028-10 Final Report, June 1988 Prepared by RESEARCH TRIANGLE INSTITUTE Analytical and Chemical Sciences Post Office Box 12194 Research Triangle Park, North Carolina 27709 Principal Investigators E. D. Pellizzari R. Perritt C. A. Clayton T D. Hartwell Prepared for Electric Power Research Institute 3412 Hillview Avenue Palo Alto, California 94304 EPRI Project Managers G. Addis Transmission Substations Program Electrical Systems Division J. Guertin Land and Water Quality Studies Program Environment Division 783960 ORDERING INFORMATION Requests for copies of this report should be directed to Research Reports Center (RRC). Box 50490. Palo Alta CA 94303. (415) 965-4081. There is no charge for reports requested by EPRI member utilities and affiliates. U.S. utility associations. U.S. government agencies (federal, state and local), media, and foreign organizations with which EPRI has an information exchange agreement. On request. RRC will send a catalog of EPRI reports. Electric P o **r Research institute and EPRI are registered service marks of Eleanc Power Researcn institute Inc Copyngnt 1980 Eleanc Power Researcn institute. Inc All ngnts reserved NOTICE This report was prepared Dy the organizations) named Deow as an account of work sponsored Dy the Eleanc Power Research institute Inc (EPRI). Neither EPRI. members of EPRI. the organizations) named Detow. nor any person acting on Denaif of any of them: (a) makes any warranty, express or implied, with resp e a to the use of any information, apparatus, method, or process disclosed in this report or that such use may not infringe privately owned ngnts: or (p) assumes any liabilities with resoea to the use of. or for damages resulting from the use of. any information, apparatus, method, or process disclosed in this report Preoared Dy Research Tnangie institute Researcn Tnangie Park, North Carolina 783961 GENP 011150 ABSTRACT An Interlaboratory study Involving five Independent laboratories was conducted to quantitatively assess accuracy and precision of polychlorinated d1benzo--d1ox1n (PCDD) and polychlorinated dlbenzofuran (PCDF) congener analysis In three spiked and seven In-service utility fluids. Data was obtained using capillary gas chromatography/mass spectrometry for 16 different PCDD and PCDF compounds. Five different ^C-labeled Internal standards were added to the samples and the samples were purified prior to analysis. Each laboratory's generated calibrations for the individual compounds were linear for concentrations over four orders of magnitude. Analyses were performed in triplicate on an extract to provide insight to instrumental reproducibility. The median and range of percent accuracy for four of the five laboratories was 102 and 56-149, respectively, across all chemicals measured 1n the two spiked samples. The fifth laboratory had a median and range of 107 and 8-3939, respectively. The median and range of percent reproducibility (coefficients of variation) for three of the four laboratories was 8.2 and 0-46, respectively, across all 10 synthetic and in-service samples and all compounds measured. The fourth laboratory had a median and range of 30 and 2-159, respectively. The fifth laboratory did not report replicate analyses on each extract and thus reproducibility for instrumental analysis could not be calculated. The largest variability (accuracy and precision) was attributed to the different types of samples (>90%) 1n three of the four laboratories while the fourth had as much as 63.1% variability from the sample type analyzed and as much as 36.9% due to Instrumental measurement error. At the approximate 95% prediction Intervals, the analysis of these 10 samples by any one of the five laboratories would be expected to produce variability 1n the results for each PCDD or PCDF congener measured ranging from a factor of two to 300. Of the five participating laboratories, the analytical method employed by Lab-D gave for the most part the best accuracy and precision while Lab-C was the poorest. GENP 011151 m 783962 ACKNOWLEDGMENTS The authors extend their sincere appreciation to the EPRI technical management team led by Dr. G. Addis and his colleagues, Dr. J. Guertln, and Dr. R. Komal who provided adroit guidance to this program and encouragement to all the participants. Dr. T. Rouse (General Electric Company) Is thanked for h1s contributions concerning utility use of dielectric liquids. We would like to also thank our research comrades who contributed the analytical data for this Interlaboratory collaborative study. They are Dr. S. Gordon (Illinois Institute of Technology Research Institute), Drs. G. Eadon and Dr. H H k e r (New York State Department of Health), Dr. M. Kilpatrick (Radian Corp.), Dr. M. Cooke (Battelle Columbus 01v.), and Dr. C. Rappe (U. of Umea). GENP 011152 V 783963 CONTENTS Section 1 INTRODUCTION 2 CHEMICAL METHODOLOGY Selection of Samples from Electric Equipment for PCDF and PCDD Analysis Acquisition of In-Service Samples Synthesis of Standards Preparation of PCDD and PCDF Standards Sample Extraction and Extract Preparations Capillary Gas Chromatography/Mass Spectrometry Analysis Quality Control 3 RESULTS AND DISCUSSION OF STATISTICAL ANALYSIS Evaluation of Calibration Model Results of Round Robin 4 REFERENCES - APPENDIX A EVALUATION OF CALIBRATION REGRESSION USED BY U. UMEA APPENDIX B DATA USED IN COMPARISONS BETWEEN LABS APPENDIX C PROJECT' SUMMARY Page 1-1 2-1 ` 2-1 2-1 2-1 2-10 2-11 2-16 2-38 3-1 3-1 3-5 4-1 A-l B-l C-l GE1SIP 011153 v11 783964 ILLUSTRATIONS 1 Overview of Analytical Protocol 2 Sample Clean-up Procedure: Round Robin Analysis 3 Summary of Major Elements of IITRI's GC/MS Analysis Protocol 4 Selected Ion Current Profile for Mass 322 From GC/MS Analysis of Performance Check Solution Containing TCDDs as Noted 1n The Plot, Using a 60 m SP-2330 Fused Silica Capillary Column 5 Selected Ion Current Profile for Mass 304 From GC/MS Analysis of Performance Check Solution Containing TCDFs as Noted In The Plot, Using a 60 m SP-2330 Fused Silica Capillary Column 6 TCDD Test Mixture Chromatogram on D8-5 7 TCDF Test Mixture Chromatogram on DB-5 8 PnCDF Test Mixture Chromatogram on DB-5 9 HxCDF Test Mixture Chromatogram on DB-5 10 TCDD Test Mixture Chromatogram on SP-2331 11 TCDF Test Mixture Chromatogram on SP-2331 12 PnCDF Test Mixture Chromatogram on SP-2331 13 HxCDF Test Mixture Chromatogram on SP-2331 1x 783965 TABLES Standard Solutions Baseline Fluid Samples - GC/HS Crosscheck Baseline Fluid Samples - Round Robin Analysis Analytical Sequence In GC/MS Determination of PCOFs andj'CDDs' HRGC-MS Analytical Conditions Conversion Factors Calibration Curves: LAB-A Calibration Curves: LAB-B Calibration Curves: LAB-0 Calibration Curves: LAB-E Coefficients of Variation by Compound, Lab, and Sample Minimum, Median, and Maximum CVs By Lab and Sample Over All Compounds Percent Accuracy for Spiked Samples By Compound, Sample, and Lab Minimum, Median, and Maximum Percent Found (Accuracy) For All PCDF and PCDD Congeners 1n Spiked Samples 1 and 2 Variance Components By Compound And Lab Variance Components By Compound For Lab-D Standard Deviations, Their Ranks, and Bartletts Test By Compound and Sample Levene's Test of Differences Among Labs By Compound Over Samples Approximate 95 Percent Prediction Intervals By Compound And Sample Page 2-3 2-7 2-8 2-22 2-29 3-2 3-2 3-3 3-3 3-4 3-6 3-10 3-11 3-15 3-16 3-21 3-22 3-26 3-27 GENP 011155 xi 783966 Table B-l 1 2 3 4 5 6 7 S 9 10 11 12 13 14 Data Used In Comparisons Between Labs Analytical Procedures Employed by Round Robin Labs Samples Analyzed 1n Round Robin Study Example of Measurements Performed for Dioxins/Furans By A Lab on One Sample (Lab-Bf Sample 14) - PPB Percent Accuracy for Spiked Samples for 2,3,7,8-TCDD By Sample and Laboratory Percent Accuracy for Spiked Samples for 2,3,4,6,7,8-HXCDF By Sample and Laboratory Median and Range of Percent Accuracy Across All Compounds For Spiked Samples Coefficients of Variation for 1,2,3,7,8 + 1,2,3,4,8-PNCDF By Compound and Laboratory Coefficients of Variation for 1,2,3,4,6,7,8-HPCDF By Compound and Laboratory Median and Range of Coefficients of Variation Across All Compounds By Laboratory and Sample Variance Components for 2,3,7,8-TCDD Across All Samples (1,2,3,11,12,13,14,17) Variance Components for 2,3,4,6,7,8-HXCDF Across All Samples (1,2,3,11,12,13,14,17) Variance Components By Compound For Lab-0 Levene's Test of Differences Among Labs By Compound Across Samples Approximate 95 Percent Prediction Intervals (PPB) Page B-2 C-3 C-5 C-5 C-7 C-7 C-8 C-8 C-10 c-xo C-ll C-ll C-12 C-14 C-15 783967 xi 1 GENP 011156 SUMMARY EPRI has sponsored a four-year research program to evaluate the analysis of selected Individual compounds of polychlorinated d1benzo-p-diox1ns (PCODs) and polychlorinated dlbenzofurans (PCDFs) 1n dielectric fluids. These are Impurities often associated with askarel mixtures of polychlorinated biphenyls (PCBs) and tr1/tetrachlorobenzene (TCB). They have been known to form either 1n the PCB manufacturing process or as a result of excessive heat (fires). This program included the synthesis of new natural (12c) and 1sotop1cally labeled (13C) PCDD and PCDF analytical standards. Standards and spiked test samples were prepared by one of the laboratories and distributed to the four other participating laboratories as part of a round robin study. The samples were then analyzed as unknowns. The impetus for the studies has been the wide range of toxicities (4-5 orders of magnitude) seen for the Individual compounds 1h animal tests and the difficulty of quantitatively analyzing for some of these chemically similar compounds of dissimilar toxicity, often in the presence of a great excess of PCBs. PCBs themselves are a group of compounds with chemical behavior relatively similar to PCOF/PCDD during sample cleanup stages. It should be remembered that the major object of the study was to evaluate and upgrade methods of analysis; analysis of field samples was an important secondary output. The study involving five participating laboratories, was conducted in two phases. Phase I was designed to located glaring discrepancies 1n the techniques, and Phase II to statistically Improve results of the different laboratories on a series of unknown samples. Phase I Involved the analysis of PCDFs and PCDDs in five spiked basic samples: Aroclor-1016, Aroclor-1242, Aroclor-1260, a tri- and tetrachlorobenzene mixture, and aged mineral' oil; and four In-service dielectric fluids. Each participating laboratory used its best available In-house extraction and analysis techniques (Gas Chromatography/Mass Spectrometry, GC/MS). Since validated analytical methods did not exist before this program was Initiated, this phase of the program was necessary to assess any deficiencies and optimize the analytical methodology. GENP 011157 S-l 783968 The'1sotop1ca1ly labeled standards were needed to calibrate the efficiency of cleanup and to determine the sensitivity (response) of the MS for the various . individual compounds of PCDF and PCOD being analyzed. After evaluating the results of Phase I and Implementing several Improvements to the methodology! Phase II was Initiated. Phase II was designed to provide quantitative data on PCDF and PCDD compounds In selected in-service dielectric fluids and to make a statistical Interlaboratory comparison of these data. The specific alms were 1) to determine the accuracy of analysis by each participating laboratory for a selected list of PCDFs and PCDDs 1n spiked samples, 2) to determine the precision of analysis by each laboratory for PCDFs and PCDDs 1n spiked and In-service utility samples, 3) to perform an interlaboratory comparison and thus a comparison of methods employed for accuracy and precision, and finally 4) to measure the levels of PCOFs and PCDDs 1n a .few selected in-service utility samples.' This study suggests that there 1s no strongly preferred analytical procedure (among those using a combination of extractlon/chemical cleanup, gas chromatography, and mass spectrometry) for measuring compound specific concentrations of PCDF and PCOD. For spiked'samples, originally free of PCDF and PCDD, five different methods (corresponding to five different participating laboratories) yielded concentrations accurate to within +20% (except one laboratory measured 2,3,4,6,7,8-hexachlorinated dlbenzofuran [2f3,4,6,7,8-H5-CDF] with an order of magnitude error). Since standards for all ISO PCDF and PCDD compounds were not available, it was not possible to know the accuracy of measurement on the unknown field samples. Only the precision of measurement could be ascertained. Considerable variation 1n measurement precision was observed between laboratory procedures. For example, measurement of very low concentrations (I to 10 ng/g range) showed an order of magnitude variation, whereas measurement 1n higher concentration range (100 to 10,000 ng/g) were considerably more precise. Again, a specific HgCDF measurement was an exception, having as much as four orders of magnitude variation. For total Individual congener classes, concentration measurements had only one order of magnitude variation. What range of concentrations were found In the field samples? The maximum concentration for 2,3,7,8-Tetra-CDD was 32 ng/g and the maximum concentration of 2,3,7,8-tetra-CDF + 2,3,4,8-tetra-CDF was 459 ng/g. The highest concentration found was for one sample of the much less critical octachlorlnated dlbenzofuran at 65,000 ng/g. These concentrations are reported with a statistical 95% confidence level. 783969 S-2 GENP 011158 SECTION 1 INTRODUCTION Because of the concerns about the health effects of polychlorinated biphenyls (PCBs) their production was banned and regulation Instituted 1n 1976 by the Toxic Substance Control Act (TSCA), PL 94-469. Certain totally enclosed electrical uses of PCBs are allowed for the remainder of the useful life of existing electrical equipment. Many of these adverse health effects appear to be attributable to trace contaminants such as polychlorinated dlbenzofurans (PCDF) and polychlorinated dibenzo--d1oxins (PCOD) in the PCBs. The Electric Power Research Institute (EPRI) 1nit1tated a program to investigate the chemistry of PCDF and PCDO as they may relate to use of Askarel, a dielectric liquid, which contains PCBs and potential trace contamination of PCDF/PCDD. This program had three primary objectives: (1) to develop and optimize analytical methods and quantitatively characterize these methods for measuring PCDF/PCDD in dielectric fluids, (2) to apply these methods to measurement of PCDF/PCDD levels in typical dielectric fluids used in utility operations, and (3) to investigate possible PCDF/PCDD formation 1n simulated transformer events. As part of the objective to evaluate methodology, the program included the synthesis of native and isotoplcally labeled PCDD and PCDF analytical standards (1) to allow for analysis of selected Individual congeners of PCDDS and PCDFs. Analytical standards and spiked matrix samples were prepared by a contract laboratory and then distributed to four other participating laboratories as part of a round robin study. These samples were analyzed blind by five Independent contract laboratories - Illinois Institute of Technology Research Institute (IITRI), New York State Department of Health (NYSOOH), Radian Corp., Battelle, Columbus Division, and U. Umea. The method evaluation study was conducted In two phases. Phase I Involved the analysis of PCDF's and PCDD's 1n five spiked baseline samples; for example, Aroclor-1016, Aroclor-1242, Aroclor-1260, tr1- and tetrachlorobenzene mixture, aged mineral oil, and four In-service dielectric fluids (1-4). Each participating laboratory used their best available 1n-house extraction and analysis techniques (1-4). Since validated methods for PCDD and PCDF analysis did not exist before this program was Initiated, this phase was necessary to assess any deficiencies and to optimize the analytical methodology. After Phase I results were evaluated and 1-1 783970 several Improvements were Implemented, then Phase II was undertaken. This report covers the effort under Phase II. Phase II was designed to yield quantitative data on PCDD and PCDF congeners 1n selected in-service dielectric fluids and to make a statistical Interlaboratory comparison on these data. The specific alms were (1) to determine the accuracy of analysis by each participating laboratory for a selected list of PCOD's and PCOF's 1n spiked samples, (2) to determine the precision of analysis by each laboratory for PCOF's and PCOD's 1n spiked and in-service utility samples, (3) to perform an interlaboratory comparison of methods for accuracy and precision, and finally (4) to measure the levels of PCOF's and PCOD's 1n a few selected In-service utility samples. Although this report primarily addresses the statistical analysis of PCDD and PCDF congener data from the five laboratories, a section on chemical methodology is included to allow the reader to examine the similarities and differences in the analytical procedures employed by each laboratory. These procedures are also described in more detail in the reports from the five individual laboratories (14). 783971 1-2 GENP 011160 SECTION 2 CHEMICAL METHOOOLOGY SELECTION OF SAMPLES FROM ELECTRICAL EQUIPMENT FOR PCDF AND PCDO ANALYSIS The rationale for the selection of transformers and capacitors to be sampled for PCDF and PCDD analysis was previously developed and reported elsewhere (5). Since the analysis 1s time consuming and difficult, the Intention was to choose a small number of samples, which would yield the greatest amount of information about the large population of equipment in use. Also, the samples were chosen to provide a potentially wide spectrum of challenge to the analytical methodology which was being statistically evaluated 1n the Round Robin Study. ACQUISITION OF IN-SERVICE SAMPLES The acquisition of in-service samples was performed by one of the participating laboratories and the procedures employed have been reported (4). The samples selected for use in the Round Robin Study are shown in Table 2. Two baseline samples were spiked with a selected number of congeners at known concentrations Into mineral oil and Aroclor 1016, respectively. A third sample, Aroclor 1260, was spiked at known levels with these same congeners but it also contained endogenous PCBF's and PCDD's. Samples 11-17 (Table 2) were in-service utility samples that were selected to represent the various types of utility uses (5). In some cases the dielectric was Aroclor 1260, or mineral oil containing PCBs. Dielectric fluids from transformers, capacitors, or precipitators were selected for this Round Robin Study. Some of the devices had been 1n use for many years (Table 2). SYNTHESIS OF STANDARDS Because *3C Isotope 1s present 1n a natural abundance of 1.1%, synthetically enriched compounds can be prepared for use as analytical Internal standards. Isotoplcally enriched compounds can be clearly distinguished from native substances by mass spectrometry. For this reason, Radian Corporation"prepared 13C-labeled and unlabeled (native) PCDO and PCDF congeners as analytical`standards for use by all laboratories participating in the Round Robin Study (1). Subsequently, analytical standard solutions were also prepared of these authentic PCDD and PCDF congeners and distributed by Radian Corp. GENP 011161 2-1 783972 Standard Materials Radian synthesized the following materials (1): 1.2.3.7.8- Pentachlorodlbenzofuran (PnCOF) 1,2,3 ,4-,7 r8-Hexachlorodlbenzofuran (HxCDF) 1',2,3,4,6,7,8-Heptachlorodlbenzofuran (HpCOF) 2.3.7.8- Tetrachlorodlbenzofuran-13Ci? (TCDF-13Cj 1.2.3.7.8- Pentachlorod1benzofuran-l3Ci2 (PnCDF-i3C) 1,2,3,4r7,8-Hexachlorodlbenzofuran-13Ci2 (HxCDF-13C) Several congeners 1n Radian's PCDF/PCDD Standards Inventory were also donated to the program. These congeners were: 2,3,7 r8-Tetrachlorodlbenzo-g-d1oxln 2.3.7.8- Tetrachlorod1benzo--diox1n-l3Ci2 2.3.7.8- Tetrachlorodlbenzofuran Octachlorodl benzo-jj-di oxi n-13Cx2 2.3.4.8- Tetrachlorodi benzofuran 2.3.4.7.8- Pentach1orodlbenzo furan 1 .2.3.6.7.8- Hexachlorodlbenzofuran 2.3.4.6.7.8- Hexachlorodlbenzofuran Octachlorodlbenzofuran A mixture of congeners was obtained from Professor Chrlstoffer Rappe of the University of Umea, Sweden. This semi-quantitative mixture, the "toxic cocktail", contained the following Isomers: 1,3,4-Trlchlorodlbenzofuran 2.3.4.8- Tetrach1orod1benzofuran 2.3.4.7.8- Pentachlorodlbenzofuran 1.2.3.4.8- Pentach1orodi benzofuran 1,2,3,4,7,9-Hexach1orodlbenzo furan 1.2.3.4.6.7.8- Heptachlorodibenzofuran Octachlorodlbenzofuran All the compounds prepared or obtained by Radian were made up as stock spiking solutions at 50 pg/mL or 100 /jg/mL. These solutions were further diluted to produce standard solutions for distribution to the participating analytical laboratories. Table 15 lists the standards provided by Radian. Baseline Liquids To determine the accuracy and precision of PCDD/PCDF measurements, a selection of baseline liquids were re-analyzed 1n the Round Robin Analysis portion of the program (1). The Baseline liquids (Table 16) were dissolved 1n hexane at levels of 0.18 - 0.20 grams/mL and spiked with measured levels of l3C-labeled congeners. In this Instance, however, a selection of native congeners were spiked "blind" at levels known only to the person performing the spiking. Additionally, a set of samples were prepared with no native congener spikes In order to determine background PC00/PC0F levels. Table 17 describes these samples and reveals the spiking levels. The seven "In-Service" Liquids were spiked with 13C-labeled 783973 2-2 GENP 011162 TABLE 15. STANDARD SOLUTIONS3 I.D. No. E729-14-02 E729-06-04 E729-15-02 E729-16-02 E729-05-04 E729-06-02 E729-05-01 E729-05-02 E729-05-03 E729-06-01 E729-07-02 E729-55-01 Description 2,3,7,8-Tetrachlorodlbenzofuran (TCDF)-13C 2,3,7,8-Tetrachlorodlbenzo-p-d1ox1n (TCDD)-13C 1,2,3,7,8-Pentachlorodlbenzofuran (PnCDF)-13C Octachlorodlbenzorp-d1ox1n (OCDO)-13C 2,3,7,8-Tetrach1orodlbenzofuran 2,3,7,8-Tetrach1orodlbenzo-p-d1ox1n 1,2,3,7,8-Pentachlorodlbenzofuran 1,2,3,4,7,8-Hexach1orodlbenzofuran (HxCDF) 1,2,3,4,6,7,8-Heptachlorodlbenzofuran (HpCDF) Radian PCDF/TCDD Mixture 2,3,7,8-TCDF 2,3,7,8-TCDD 1,2,3,7,8-PnCDF 1,2,3,4,7,8-HxCDF 1,2,3,4,6,7,8-HpCDF Professor Rappe's "Toxic Cocktail" 1,2,4-TrCDF 2,3,4,8-TCDF 2,3,4,7,8-PnCDF 1-,2,3,4,8-PnCDF 1,2,3,4,7,9-HxCDF 1,2,3,4,6,7,8-HpCDF OCDF Calibration Mixture I Native Congeners 2,3,7,8-TCDD 2,3,7,8-TCDF 1,2,3,7,8-PnCDF 2,3,4,7,8-PnCOF 1,2,3,4,7,8-HxCDF 2,3,4,6,7,8-HxCDF 1,2,3,4,6,7,8-HpCDF OCDF Labeled Congeners 2,3,7,8-TCD0-l3Ci2 2,3,7,8-TCDF-13Ci2 Concentration 100 ng/mL 100 ng/mL 100 ng/mL 40 ng/mL 100 ng/mL 100 ng/mL 100 ng/mL 100 ng/mL 100 ng/mL 100 ng/mL 100 ng/mL 100 ng/mL 100 -ng/mL 100 ng/mL ~116 ng/mL "100 ng/mL "100 ng/mL "100 ng/mL "100 ng/mL "100 ng/mL "160 ng/mL 7.5 /ig/mL 7.5 /ig/mL 7.5 ftg/mL 7.5 /ig/mL 7.5 /ig/mL 7.5 /ig/mL 7.5 /ig/mL 7.5 /tg/mL 500 ng/mL 500 ng/mL GENP011163 2-3 783974 I.D. No. E729-55-03 E729-55-04 TABLE 15 (cont'd.) Description l,2,3,7,8-PnCDF-13Ci2 1,2,3,4,7t8-HxCDF-13Ci2 0CD0-13Ci2 Calibration Mixture II Native Congeners 2,3,7,8-TCDD 2,3,7,8-TCDF .. 1,2,3,7,8-PnCDF 2,3,4,7,8-PnCDF 1,2,3,4,7,8-HxCDF 2,3,4,6,7,8-HxCDF 1,2,3,4,6,7,8-HpCDF OCDF Labeled Congeners 2,3,7,8-TCDD-13Ci2 2,3,7,8-TCDF-33Ci2 l,2,3,7,8-PnCDF-13c12 1,2,3,4,7,8-HxCDF-13c12 0CDD-13c 12 Calibration Mixture III Native Congeners 2,3,7,8-TCDD 2,3,7,8-TCDF 1,2,3,7,8-PnCDF 2,3,4,7,8-PnCOF 1,2,3,4,7,8-HxCDF 2,3,4,6,7,8-HxCDF 1,2,3,4,6,7,8-HpCDF OCDF Labeled Congeners 2,3,7,8-TCDD-13Ci2 2,3,7,8-TCDF-33Cj2 l,2,3,7,8-PnC0F-l3Ci2 1,2,3,4,7,8-Hx CDF-13C12 0C0D-l3Ci2 2-4 783975 Concentration 500 ng/mL 500 ng/mL 200 ng/mL 2.5 /ig/mL 2.5 /ig/mL 2.5 /ig/mL 2.5 /ig/mL 2.5 /ig/mL 2.5 /ig/mL 2.5 /ig/mL 2.5 /ig/mL 500 ng/mL 500 ng/mL 500 ng/mL 500 ng/mL 200 ng/mL 500 ng/mL 500 ng/mL 500 ng/mL 500 ng/mL 500 ng/mL 500 ng/mL 500 ng/mL 500 ng/mL 500 ng/mL 500 ng/mL 500 ng/mL 500 ng/mL 200 ng/mL GrENP 011164 l.D. No. E729-55-05 E729-55-06 E729-56-01 TABLE 15 (coni'd.) Description Calibration Mixture IV Native Congeners 2.3.7.8- TCDD 2.3.7.8- TCDF 1.2.3.7.8- PnCDF 2.3.4.7.8- PnCDF 1.2.3.4.7.8- HxCDF 2.3.4.6.7.8- HxCDF 1.2.3.4.6.7.8- HpCDF OCDF Labeled Congeners 2.3.7.8- TCDD-13c12 2.3.7.8- TCDF-13Ci2 1.2.3.7.8- PnC0F-l3Ci2 1,2,3,4,7, 8-HxCDF-l3Ci2 0CDD-13c 12 Calibration Mixture V Native Congeners 2.3.7.8- TCDD 2.3.7.8- TCDF 1.2.3.7.8- PnCDF 2.3.4.7.8- PnCDF 1.2.3.4.7.8- HxCDF 2.3.4.6.7.8- HxCDF 1.2.3.4.6.7.8- HpCDF OCDF Labeled Congeners 2,3L,7,8-TCDD-13c12 2.3.7.8- TCDF-13Ci2 1.2.3.7.8- PnCDF-l3Ci2 1.2.3.4.7.8- HxC0F-13Ci2 QCDD-13Ci2 Calibration Mixture VI Native Congeners 2.3.7.8- TCDD 2.3.7.8- TCDF GENP 011165 783976 Concentration 150 ng/mL 150 ng/mL 150 ng/mL 150 ng/mL 150 ng/mL 150 ng/mL 150 ng/mL 150 ng/mL 500 ng/mL 500 ng/mL 500 ng/mL 500 ng/mL 200 ng/mL 40 ng/mL 40 ng/mL 40 ng/mL 40 ng/mL 40 ng/mL 40 ng/mL 40 ng/mL 40 ng/mL 500 ng/mL 500 ng/mL 500 ng/mL 500 ng/mL 200 ng/mL 10 ng/mL 10 ng/mL I-D. No. TABLE 15 (cont'd.) Description 1,2,3,7,8-PnCDF 2,3,4,7,8-PnCDF 1,2,3,4,7,8-HxCDF 2,3,4,6,7,8-HxCDF 1,2,3,4,6,7,8-HpCDF OCDF Labeled Congeners 2 r3,7,8-TCDD-13c12 2,3,7,8-TCDF-13Ci2 l l2 , 3 l7 (8-PnCDF-13c^2 1,2,3,4,7,8-HxCDF-13C j2 O C DD -13C x 2 Calibration Mixture VII Native Congeners 2,3,7,8-TCDO 2,3,7,8-TCDF 1,2,3,7,8-PnCDF 2,3,4,7,8-PnCDF 1,2,3,4,7,8-HxCDF 2,3,4,6,7,8-HxCDF 1,2,3,4,6,7,8-HpCDF OCDF Labeled Congeners 2,3,7,8-TCDD-13Ci2 2,3,7,8-TCDF-l3Ci2 1,2,3,7,8-PnCDF-13Ci2 l^.S.M.S-HxCDF-lic^ OCDD-13C 12 aSee Ref. 1. for details. 2-6 Concentration 10 ng/mL 10 ng/mL 10 ng/mL 10 ng/mL 10 ng/mL 10 ng/mL 500 ng/mL 500 ng/mL 500 ng/mL 500 ng/mL 200 ng/mL 2.5 ng/mL 2.5 ng/mL 2.5 ng/mL 2.5 ng/mL 2.5 ng/mL 2.5 ng/mL 2.5 ng/mL 2.5 ng/mL 500 ng/mL 500 ng/mL 500 ng/mL 500 ng/mL 200 ng/mL 783977 GENPO11166 TABLE 16. BASELINE FLUID SAMPLES - GC/MS CROSSCHECK3 ID No. 729-09-01 E729-10-0 E729-11-01 E729-12-01 E729-08-01 E729-08-01 Description Spike Spiking Level Aroclor 1016 2.3.7.8- TCDF-l3C 1.2.3.7.8- PnCDF-l3C 2.3.7.8- TCDD-^C 0CDD-l3C 1.2.3.4.7.8- HxCDF 1.2.3.4.6.7.8- HpCDF 100 ng/g 100 ng/g 100 ng/g 40 ng/g 100 ng/g 100 ng/g Aroclor 1242 2,3,7,8-TC0F-l3C lf2 (3,7,8-PnCDF-13c 2.3.7.8- TC0D-13C 0CDD-13C 1.2.3.4.7.8- HxCDF 1.2.3.4.6.7.8- HpCDF 100 ng/g 100 ng/g 100 ng/g 40 ng/g 100 ng/g 100 ng/g Aroclor 1260 2.3.7.8- TCDF-13C 1.2.3.7.8- PnCDF-13C 2.3.7.8- TCDD-13C ocdd- 13c 1.2.3.4.7.8- HxCDF 1.2.3.4.6.7.8- HpCDF 100 ng/g 100 ng/g 100 ng/g 40 ng/g 100 ng/g 100 ng/g Tri-Tetrachlorobenzenes 2.3.7.8- TCDF-13C 1.2.3.7.8- PnCDF-13C 2.3.7.8- TCDD-l3C 0CDD-l3C 1.2.3.4.7.8- HxCDF 1.2.3.4.6.7.8- HpCDF 100 ng/g 100 ng/g 100 ng/g 40 ng/g 100 ng/g 100 ng/g Aged Mineral 011 2.3.7.8- TCDF-13C 1.2.3.7.8- PnCDF-13C 2.3.7.8- TCDD-13C 0CDD-13C 2.3.7.8- TCOD 2.3.7.8- TCOF 1.2.3.7.8- PnCDF 1.2.3.4.7.8- HxCDF 1.2.3.4.6.7.8- HpCDF 100 ng/g 100 ng/g 100 ng/g 40 ng/g 40 ng/g 20 ng/g 40 ng/g 120 ng/g 0 ng/g Professor Rappe1's "Toxic Cocktail" Aged Mineral 011 OCDF 1,2,3,4,7,9-HxCDF 2.3.4.7.8- PnCDF 1.2.3.4.8- PnCDF 2.3.4.8- TCDF 1,3,4-TrCDF *160 ng/g -100 ng/g -100 ng/g -100 ng/g -100 ng/g -116 ng/g Aroclor 1260 510 pq/q Tr1-Tetrachlorobenzenes 530 /g/g aSee Ref. 1 for details. GBNP011167 2-7 783978 TABLE 17. BASELINE FLUID SAMPLES - ROUND ROBIN ANALYSIS* ID No. E729-53-01 . E729-53-02 E729-53-03 Description Mineral Oil Aroelor 1016 Aroelor 1260 Spike Aroelor 1260 Tr1/Tetrachlorobenzenes 2,3,7t8-TCD0-13C 2.3.7.8- TC0F-13c ,, 1 12 13.7.8- PnC0F- 13r l.Z.3,4.7.8-HxCDF-13c 0CDD-" C 2.3 .7.8 - TCOD 2.3.7.8- TCOF 2.3.4.8- TCDF 1.2.3.7.8- PnCDF 2.3.4.7.8- PnCDF 1.2.3.4.7.8- Hx COF 1.2.3.6.7.8- HxCDF 2,3,4,6,7,8-HxCDF 1.2.3.4.6.7.8-HpCDF 0ZDf 2,3,7,8-TCDD-13C 2.3.7.8- TCOF-13c 1.2.3.7.8- PnCDF-13c 1.2.3.4.7.8- Hx CDF-13c 0CDD-13C 2.3.7.8- TCOD 2.3.7.8- TCDF 2.3.4.8- TCDF 1.2.3.7.8- PnCDF 2.3.4.7.8- PnCDF 1.2.3.4.7.8- HxCDF 1.2.3.6.7.8- HxCDF 2.3.4.6.7.8- HxCDF 1.2.3.4.6.7.8- HpCDF OCDF 2,3,7,8-TCDD-13C 2.3.7.8- TCDF-13C 1.2.3.7.8- PnCDF-13C 1.2.3.4.7.8- HxCDF-^3C 0CDD-13C 2.3.7.8- TCOD 2.3.7.8- TCDF 2.3.4.8- TCDF 1.2.3.7.8- PnCDF 2.3.4.7.8- PnCDF 1.2.3.4.7.8- HxCDF 1,2,3,6,7,3-HxCDF 2.3.4.6.7.8- HxCDF 1.2.3.4.6.7.8- HpCDF OCDF Aroelor 1260 Tr1/Tetrachlorobenzenes Spiking Level 640 ag/g 1250 naia loo ng/g 100 ng/g 100 ioo ng/g 40 ng/g 2S ng/g 100 ng/g 100 ng/g 150 ng/g 150 ng/g 250 ng/g 50 ng/g 175 ng/g 400 ng/g 0 ng/g 100 ng/g 100 ng/g 100 ng/g 100 ng/g 40 ng/g 53 ng/g 53 ng/g 0 ng/g 158 ng/g 158 ng/g 0 ng/g 53 ng/g 132 ng/g 526 ng/g 132 ng/g 100 ng/g 100 ng/g 100 ng/g 100 ng/g 40 ng/g 111 ng/g 833 ng/g 194 ng/g 111 ng/g 111 ng/g 500 ng/g 278 ng/g 56 ng/g 111 ng/g 222 ng/g 680 pq/q 1320 pq/q 783979 GENP011168 TABLE 17 (cont'd.) ID No. E729-54-02 E729-54-03 E729-54-04 Description Mineral Oil Aroclor 1260 Aroclor 1016 Spike 2.3.7.8- TCDD-l3C 2.3.7.8- TC0F-l3C 1.2.3.7.8- PnCDF-l3C 1.2.3.4.7.8-Hx CDF-13c 0CDD-13C 2.3.7.8- TCDD-13C 2,3t7,8-TCDF-13C 1.2.3.7.8- PnCDF-l3C 1.2.3.4.7.8-Hx CDF-13C o c d d - 13c 2.3.7.8- TC0D-13c 2,3,7(8-TCDF-13C 1.2.3.7.8- PnCDF-l3C 1.2.3.4.7.8- Hx CDF-13C 0CDD-13C aSee Ref. 1 for details. Spiking Level 100 ng/g 100 ng/g 100 ng/g 100 ng/g 40 ng/g 100 ng/g 100 ng/g 100 ng/g 100 ng/g 40 ng/g 100 ng/g 100 ng/g 100 ng/g 100 ng/g 40 ng/g GENP 011169 2-9 783980 congeners at Battelle Memorial Institute, Columbus, Ohio, using stock solutions provided by Radian Corporation (4). PREPARATION OF PCDD AND PCDF STANDARDS Calibration Standards) PCDD and PCOF standard compounds Including 13C-labelled standard compounds, dissolved In solvent, were received at each laboratory sealed In glass vials. The seven vials each contained all of the following compounds: 2.3.7.8- TCOD 2.3.7.8- TCDF 1.2.3.7.8- PnCDF 2.3.4.7.8- PnCDF 1.2.3.4.7.8- HxCDF 2.3.4.6.7.8- HxCOF 1 .2 .3 .4 .6 .7 .8 .9 - 0C0F 2.3.7.8- TCDD-" Ci2 2 .3 .7 .8- TCDF-13Ci2 l , 2 t3,7,8-PnC0F-13Ci2 1.2.3.4.7.8- HxCDF-13Ci2 1 .2 .3 .4 .6 .7 .8 .9 - 0CD0-I3C12 Each vial contained all of the native compounds at one of the following concentration levels: 7.500 ng/mL 2.500 ng/mL 500 ng/mL 150 ng/mL 40 ng/mL 10 ng/mL 2.5 ng/mL Each standard solution contained all of the ^ C compounds at 500 ng/mL each. Baseline Samples Three baseline samples (Aroclor 1016, Aroclor 1260, and aged mineral oil) were prepared at Radian Corporation as solutions in n-hexane at a concentration of about 0.2 g/mL. Each sample was spiked with 100 ng/g of each of the five 13C-compounds listed above. Each sample was also spiked with "blind11 amounts of the following unlabeled compounds: 2.3.7.8- TCOD 2.3.7.8- TCDF 2.3.4.8- TCDF 1.2.3.7.8- PnCDF 2.3.4.7.8- PnCDF 1.2.3.4.7.8- HxCDF 1.2.3.6.7.8- HxCDF 2.3.4.6.7.8- HxCOF 1.2.3.4.6.7.8- HpCDF OCDF In addition, the mineral oil sample was spiked with 625 pq/q Aroclor 1260 and 1250 pqfq trWtetrachlorobenzenes. 783981 2-10 GENP01U70 In-Service Samples Seven In-service samples were fluids consisting of five askarels and two mineral oils, all of which were prepared at a concentration of approximately 200 mg/mL 1n n-hexane. Each sample was spiked at Battelle with the five Isotoplcallylabeled Internal standards at concentrations of 100 ng per gram of oil. SAMPLE EXTRACTION AND EXTRACT PREPARATIONS For this Round Robin Study each participating laboratory employed their best available 1n-house extraction and analysis techniques which are shown 1n Table 1.' Because the methods varied between the laboratories, a parameter being statistically tested 1n the Round Robin Study, a detailed description of each is presented here. Procedures Employed by NYSDOH (3) The volume of a portion of samples of electric Insulating fluids and mineral oil was measured and extracted by adding 25 mL of acetone and mixing thoroughly. Hexane (25 ml) was added and the sample was stirred and extracted for a minimum of 1 h. To remove PCBs and nonplanar aromatics and to isolate one fraction containing all tetra- to OCDFs and OCDDs the sample extract was chromatographed on acid alumina, PX-21 carbon and neutral alumina. A preprogrammed Hamilton valve control system designed at NYSDOH was used to reprodudbly select solvents and chromato graphic columns (3). The recovery of all tetra- to OCDFs and OCDDs (as verified with available standards) from an Individual chromatographic column was >95%. The sample extract was applied to an activated acidic alumina column, followed by 30 mL of 3% CH2CI2 1n hexane. The CDF/CDD fraction was eluted with 70 mL of.50% CH2CI2 1n hexane onto a carbon column followed by 50 mL of 10% benzene in hexane.- The CDF/CDD fraction was eluted 1n the reverse direction with 30 mL of 50% xylene 1n hexane onto a neutral alumina column (replaces xylene with a volatile solvent) followed by 30 mL of 3% CH2CI2 1n hexane. The final purified CDF/CDD fraction was eluted with 70 mL of easily volatilized 50% CH2CI2.1n hexane and concentrated with first boiling water and then vacuum to 4-10 iL (3). Extracts were stored 1n the dark for 6C/MS In sealed 150 /iL boroslUcate capillary tubes. Procedures Employed by IITRI (2) The extraction and cleanup procedure which was employed was based on the general method described by Albro and coworkers (2) for the fractionation and class determination of complex mixtures of chlorinated aromatic compounds. An overview Is shown In Figure 1. Briefly, the procedure consisted of the following major steps. GENP 011171 2-11 783982 Figure 1. Overview of Analytical Protocol (Ref. 2) 783983 2-12 GENP oi 1172 1- To the sample (ca. 0.1 g)( spiked with a mixture of 13C-labe1ed PCDF and/or PCDO Internal standards, about 30 ng of 2,3,7-tri-CDD was added to serve as a "carrier" compound. 2. The sample was chromatographed on a gel permeation column (20 g Sephadex LH-20), eluting with 50% methylene chlor1de-1n-methanol and collecting the 60-150 mL fraction, to separate halogenated aromatic compounds from any alphatlcs present. 3. About 10 il of propylene glycol were added as a "keeper" compound, to prevent sample loss during roto-evaporat1on. 4. After roto-evaporat1on, the halogenated aromatic fraction was chromatographed on an A-540 basic alumina column (20 grams topped with a layer of anhydrous sodium sulfate), eluting with 180 mL of 2% methylene chloride-1n-hexane, to elute the PCB fraction, 5. Congeners eluted with 20% methylene chlor1de-1n-hexane (200-220 mL), to recover PCDFs, PCODs, polychlorinated quadrlphenyls, and less chlorinated PCBs. 6. About 10 fil of propylene glycol were added, then the fraction was concentrated to near dryness and loaded onto a 10 g EM-1078 acidic alumina column, eluting with 80 mL of 1% methylene chlor1de-1n-hexane( to purify the PCDF/PCOD fraction. The eluate was discarded. 7. The purified PCDF/PCDD fraction was eluted with 20% methylene chloride-in hexane (120 mL). 8. The extract was concentrated to 25 /*L immediately before analysis using toluene. 9. When necessary, the extract was purified further by loading 1t in hexane onto a small (1 cm) column of Carbopack C mixed with Cellte 545, eluting with 2 mL hexane, 1 mL of 50% methylene chlor1de-1n-cyclohexane and 1 mL of methylene chlor1de:methanolbenzene (75:20:5). The column flow was reversed and 8 mL of toluene were collected. Procedures Employed by Radian (1) The extraction technique used by Radian for the Round Robin analysis phase of this study was a modified Sm1th-Stall1ngs/EPA Region VII clean-up (1). Figure 2 describes the clean-up procedure used for these dielectric samples. An aliquot of the sample (approximately 0.25 g) was transferred to a column consisting of 2.5 cm sodium sulfate (on bottom), 5 g silica gel type 60, (70-230 mesh), 6 g basic silica gel (35% potassium hydroxide), and 10 g acid silica gel (40% sulfuric acid). This column was eluted with 170 mL 1:1 methylene chloride cyclohexane directly onto a 5.0 cm column of 1-9 ftm carbon (PX-2) on glass fibers. The silica gel was discarded and the carbon column washed with 25 mL 1:1 GENP 011173 2-13 783984 Part 1 Modified Smith-Stallings Clean-Up 150 vnL 1:1 Methylene chloride: cyclohexane Discard silica gel 25 mL 1:1 Methylene chloride: cyclohexane 50 mL 75:20:5 Methylene chloride: methanol: benzene Invert carbon column 50 mL Toluene (collect) M I Acid silica gel 10g Basic silica get 6g Silica gel 5g Sodium sulfate 1" ------Carbon/glass libers 0O 00 CD OCOl T Figure 2. Sample Clean-up Procedure G E N P 0 J]j74 Part 2 "Region 7" Clean-Up Acid silica gel 5g Silica gel 1g Acid alumina 6g Solvent exchange the toluene to hexane Transler hexane extract to silica gel column 60 mL hexane Discard silica gel 20 mL hexane 20 m L 20% m ethylene chloride: hexane (collect) Concentrate Round Robin Analysis (Ref. 1) methylene chloridecyclohexane and 50 mL 75:20:5 methylene chlor1de:methanol: benzene. The eluted solvents were discarded. The PCDDs and PCDFs were eluted from the carbon column with 50 mL toluene. The toluene was evaporated to dryness on a rotary evaporator. The dried extract was redlssolved 1n hexane and transferred to a column of 5 g ac1d-1mpregnated silica gel and 1 g silica gel (lower layer). This column was eluted with 60 mL hexane directly onto a 6 g acid alumina (AG 4, 100-200 mesh) column. The silica gel was discarded and the acid alumina was washed with another 20 mL hexane. The eluted hexane was discarded. The PCDDs and PCDFs were eluted off the alumina with 20 mL 20% methylene chlorlde/hexane. The volume of this eluate was reduced to 1-2 mL with a stream of nitrogen. The extract was transferred to a 1 mL conical vial and taken to dryness. Each sample was processed 1n triplicate (this was the only lab which performed sample processing in replicate). Procedures Employed by Battel!e (4) One milliliter aliquots (containing ~250 mg of the original fluid) of each sample was transferred to separatory funnels and diluted with 49 mL of hexane. The hexane solutions were washed with 10 mL portions of concentrated sulfuric acid until the washes were only slightly colored. A maximum of five acid washes were .used. Following the acid washes, each oil was washed with 20 mL of distilled water to remove any residual acid. The hexane solutions were transferred to glass collection tubes and concentrated to approximately 2 mL using gentle streams of nitrogen gas. The hexane solutions were transferred to multilayered silica gel columns containing activated silica gel, 44 percent concentrated sulfuric acid on silica gel, and 33 percent 1M sodium hydroxide on silica gel. The PCDD/PCDF Isomers were eluted from the columns using 70 mL of hexane and the entire eluates, Including the original sample volume, were collected. The eluates from the multilayered silica gel columns were concentrated to approximately 1 mL using gentle streams of nitrogen and transferred to the top of columns containing 5 g of activated basic alumina. These columns were eluted using hexane, hexane/methylene chloride (97:3, v/v) and hexane/methylene chloride (1:1, v/v) as elution solvents. The 1:1 hexane/methylene chloride eluates were collected, solvent exchanged into hexane, and transferred to the top of alumina columnns containing approximately 2 g of activated basic alumina. These columns were eluted with hexane, hexane/methylene chloride (97:3, v/v), and hexane/methylene chloride (1:1, v/v). The 1:1 hexane/methylene chloride eluates were collected, concentrated to near dryness with nitrogen gas, and spiked with 10 ng of l ^ ^ A - T C D D - i ^ c ^ * These solutions were stored at 0*C until they were analyzed. GENP 011175 2-15 783986 For samples still too contaminated with Interferences to be analyzed, they (41159-11-12,13,14,17 and E729-53-02,03) were chromatographed through florlsil columns to provide additional analyte enrichment. Each of the florlsil columns was prepared using 5 g of activated florlsil (60-100 mesh) which was slurry packed with hexane. The sample solutions were diluted with approximately 5 mL of hexane and transferred to the florlsil columns. The columns were eluted with hexane, hexane/ethyl ether (94:6, v/v), and hexane/methylene chloride (25:75, v/v). The 25:75 hexane/methylene chloride fraction was collected and concentrated to approximately 20 /tl. These solutions were stored at 0*C until they were analyzed. CAPILLARY GAS CHROMATOGRAPHY/MASS SPECTROMETRY ANALYSIS Procedures Used by NYSDOH (3) Low Resolution Mass Spectrometry-- All sample extracts were analyzed with a Hewlett-Packard 5970 Mass Selective Detector. A portion of each extract was Injected into the split-splitless Injection port (250*C) and eluted through a 50 m (.25 mm 1.d.; .33 ftm film thickness) Hewlett-Packard 53S phenyl methyl silicone fused silica capillary column. The following temperature program was used: H-P 5890 Temperature Program Level I 2 3 4 5 Initial Temp. 190 Initial Time 1.00 Rate (*C/m1n) 5.0 5.0 5.0 5.0 5.0 Final Temp. 220 235 250 270 300 Final Time 16.00 7.00 18.00 7.00 24.00 Total Time 23.00 33.00 54.00 65.00 95.00 783987 The gas chromatograph 1s Interfaced directly to the source of the mass selective detector and the transfer line and ion source are heated to 275*C and 150*C respectively. The mass selective detector was operated in the selective ion monitoring (SIM) mode. The m/z values characteristic of native TCDD, TCDF, PnCDF, HxCDF, 0CDF, and 0CD0 and TC0D-13Ci2, TCDF-13Ci2* PnCDF-13Ci2, HxCDF-l3C 12 and 0CDD- 13Ci2 were monitored at various times during the ninety-five (95) minute runs. The time window used for monitoring the different congener classes (TCDD, TCDF, etc.) were determined by the previous Injection of either window standards (TCDD) or a mixture of combusted PCS. Raw data was collected Into and processed by an H-P Quicksilver data system. The data system was used to generate electronic copies of 2'16 GBNP 01U76 reports of retention times and areas at each m/z value monitored. Additional custom software was used to identify GC peaks and quantitate analyte levels. This software was written.by NYSDOH staff and processed on an IBM PC or XT. The H-P reports were input 1n ASCII and transferred by using a customized H-P Pascal operating system (sender) and the program called Kermlt (receiver). Each sample extract was Instrumentally analyzed 1n triplicate. High Resolution Mass Spectrometry-- Some extracts were analyzed with a Kratos MS-50 High Resolution GC/MS/DS. In this system portions of the extracts were Injected Into the on-column injection port of a Carlo-Erba 4160 high resolution gas chromatograph and eluted through a 50 m (.20 mm 1.d,,; .33 m film thickness) Hewlett-Packard 5% methyl phenol silicone fused capillary column. The oven was temperature programmed as follows: Level 1 2 3 Initial Temp. 190*C H-P 5890 Temperature Program Initial Time 1.0 Rate (*C/m1n) 5.0 5.0 5.0 Final Temp. 220 235 300 Final Time 16.00 7.00 50.00 Total Time 23.00 33.00 86.00 The gas chromatograph 1s Interfaced to the MS-50 through a direct line heated to 250*C. The 1on source temperature was 250*C. The high resolution mass spectrometry was operated 1n either of two SIM type modes. In normal multiple peak monitoring peak (NMPM) the m/x values of a particular analyte and Its corresponding Internal standard were observed at 10,000 RP (10% valley) by Integrating all the signal detected.within 50 ppm of their exact mass values. In high resolution multiple peak monitoring mode (HRMPM) the mass spectrometry 1s operated at the same high resolving power but the analyte and Internal standard exact masses are profiled. The data system adjusts the accelerating and electric sector voltages to focus a mass slightly below the mass of Interest. The computer then triggers a 0.3 sec analog scan through a 300 ppm section of the mass range which Includes the exact mass of the 1on of Interest (3). In both 1on monitoring modes the raw data was acquired into and processed by the Kratos DS-55 data system. Quantitation was done manually by calculating concentrations using peak areas obtained from the DS55 output. 011177 2-17 783988 Procedures Used bv IITRI (2) Samples and standards were analyzed using combined capillary column gas chromatography/low-resolut1on mass spectrometry (HRGC/LRMS or GC/MS). Figure 3 summarizes the major elements of their analysis protocol. The gas chromatograph, a Varlan 3700, was equipped with a 60 m x 0.25 mm l.d. fused-s1Hca SP-2330 (Supelco) column. Samples were Injected 1n the splitless mode, with the column programmed limit for 60 minutes. The SP-2330 column had been shown to separate the 2,3,7,8-TCDD from all other tetra-Isomers, and to partially separate the 2,3,7,8-TCDF from the 2,3,4,8-TCDF (2). However, it was unable to resolve 1,2,3,7,8-PnCDF from 1,2,3,4,8-PnCOF and 1,2,3,4,7,8-HxCDF from 1,2,3,4,7,9-HxCDF. These Isomers can be separated to some extent on less polar columns, such as 0V-17 or DB-5 (2). An example of the chromatographic resolution obtained using the SP-2330 column 1s shown 1n Figure 4 for 2,3,7,8-TCDD and several closely-eluting tetradioxin Isomers. The plot clearly shows that this column was Isomer-specific for 2,3,7,8TCDD. Figure 5 shows the separation obtained between the 2,3,7,8- and 2,3,4,8-TCDF isomers. Here the resolution was not as good as In the previous case, but is nonetheless sufficient to permit quantification of the two isomers in most cases. All data were acquired by software-controlled multiple 1on detection using two 1on masses from the molecular ion cluster for each of the PCDF and PCDD levels of chlorination. Two Ions were also monitored for each internal standard which permitted the calculation and comparison of the Isotope ratios with their theoretical values to verify their identity. Several additional ion masses were monitored to Indicate possible Interferences from chlorinated diphenyl ethers (2). The presence of these compounds could give rise to fragment Ions with the same masses as those monitored for the PCFs. Therefore, the absence of a co-response for the diphenyl ether 1on mass when a PC0F compound shows a positive response was taken to indicate that the signal for the PCDF was "real". A listing of the masses that were monitored, along with the theoretical isotope ratios, Is given In Table 18. In order to obtain maximum sensitivity and selectivity during an experiment, the run was divided Into 4 adjoining time windows. Each window contained up to 16 1on masses (for specific native and labeled PCDFs and PCDDs, and chlorinated diphenyl ethers), and the sample dwell times and Interchannel delay times were chosen to give the most sensitive and rapid cycle time. Since the GC column used 1n this work did not yield complete separation of each chlorination level of PCDFs/PCDDs from the other groups in a given time windows, some of the 1on masses were Included 1n more than one window to ensure that all Isomers from a particular congener class were monitored. 783989 2-18 G E t iP 0 1 1 1 7 8 GC/HS ANALYSIS (IITRI) HRGC/LRHS - UNIT HASS RESOLUTION -1,000 - SP-2330, OV-17 COLUMNS SELECTED ION MONITORING - TWO ION HASSES MONITORED PER PCDF/PCOD CONGENER C U S S IDENTIFICATION CRITERIA - COINCIDENT GC RETENTION TIMES - CORRECT INTENSITY RATIOS FOR MONITORED IONS QUANTIFICATION - OIRECT COMPARISON WITH ADDED INTERNAL STANDARDS - MS RESPONSE FACTORS FROM STANDARDS Figure 3. Summary of Major Elements of IITRI's GC/MS Analysis Protocol (Ref. 2) 2-19 783990 ssx> ANALYSIS HOME. DR00'C003033JOPEM0002.MISj1 U03.0 MINOOH' 1 TITLE' TCOO PE MIX #2;1UL SPLITLESS INJ;SP2330< *1542 1. OKU OPERATOR' TCOO.WN;?, SCAN SPEED 600.0 MSEC SPC' 5 SAMPLE ID DATE 5-N0U-84 1023I5 COMMISSION' C08756 GENP 011180 Figure 4. Selected ion current profile for mass 322 from GC/MS analysis of performance check solution containing TCODs as noted in the plot, using a 60 m SP-2330 fused silica capillary column (Ref. 2). I-66S8Z ANALYSIS NAME' DR00-C003.0333CTM0100I.MIS#1 U03.0 HINDOU- 1 TITLE- CHROM TEST MIX 1-200PG/TJL 1UL S U SP2330<12278>183-230. ISO/'M OPERATOR* CYTH*Wi3200,H2-3167.W33?3e.U4-3730 SPC- 3 SAITLE 10* DATE* 2-JAN-86 08-34*45 COMMISSION* MASSMEHI>*CPRIPE.IWJ2 rrioo 783992 Figure 5. Selected ion current profile for mass 304 from GC/MS analysis of performance check solution containing TCDFs as noted on the plot, using a 60 m SP-2330 fused silica capillary column (Ref. 2). TABLE 18. ANALYTICAL SEQUENCE IN GC/MS DETERMINATION OF PCOFs AND PCOOs* Temp Tine S t a r t Stop GC Column Prog R ate MindQM ( in ) (n in ) Temp (C) (C/min) 1 14:00 26:00 05 15 3200 2 26:00 35:30 250 _ 3167 3 35:30 56:00 250 -- 3750 4 56:00 66:00 250 -- 3750 ---o Owell Tine (os) Ions Monitored (0/2) CoopoundS Monitored0 Mass 1 Mass 2 Isotope Ratio (Hass 1/Hass 2) 197 TC0F 303.9 305.9 13C-TC0F 3 1 5 .9 3 1 7 .9 TC00 319.9 321.9 13C-TC00 3 31.9 3 3 3 .9 PnCOF 337.9 339.9 13C-PnC0F 3 4 9 .9 3 5 1 .9 PnCOO 355.9 357.9 HxOPE* 373.8 -- HpOPE* 407.8 -- 0.77 0.77 0.77 0.77 0.61 0.61 1.54 -- 195 PnCOF 337.9 339.9 PnCDFb 274.9 -- PnCOO 355.9 357.9 PnC00b 290.9 -- HxCOF 373.8 375.8 13C-HXC0F 385.9 3 8 7 .9 HxCOF0 310.9 -- 13C-HxC0F 3 8 9 .8 391.8 HxCDFb 310.9 -- HxCOO 389.9 391.8 HxCDDb 326.9 ___ HpOPE* 407.8 -- OOPE* 443.8 -- 0.61 1.54 1-23 1.23 1.23 1.23 -- -- 372 HxCOF HxCOO HpCQF HpCOD OOPE* NOPE* 373.8 389.9 407.8 423.8 443.8 477.7 375.8 391.8 409.8 425.8 -- -- 1.23 1.23 1.03 1.03 -- -- 534 OCDF 441.7 443.7 OCOD 457.7 459.7 13C-OCOO 469.8 471.8 OOPE* 511.7 0.88 0.88 o .ss *See Ref, 2. bHxPE. HpOPE. DOPE, NOPE, DOPE d e s ig n a te h ex a-, h e p ta - , o c t a - , n on -, and decach loroD iph en yl e t h e r s ' rasp. cMass eo n ito re d to id e n tify fragm ent ion (M-C0C1). 783993 G E N P 011182 I The major criteria which were used for confirming the presence of specific PCDFs/PCDDs 1n the samples analyzed were the following: (1) correct retention times for each PCDF/PCDD of Interest relative to the appropriate 1sotop1cal1y- labeled Internal standard(s), (2) Intensity ratio for (M)+/(M+2)+ within 10% of theoretically-expected ratio (Table 18), and (3) signal-to-no1se response of each PCDF/PCDD of Interest greater than 2.5:1 for both 1on masses monitored. Each sample extract was Instrumentally analyzed 1n triplicate. Quantification-- Ideally, each of the 210 separate Isomers of PCDF and PCDD should be quantified using the Instrument response of the corresponding labeled internal standard as a reference. Since only a limited number of these standards were available 1n practice, the approach generally followed was to use an appropriate set of Internal standards which included representative Isomers from each chlorinated class of PCDFs and PCDDs, and further assumed that the data obtained for these was representative of all Isomers 1n each group. Relative response factors were determined from the analysis of standard solutions containing the Internal standrds and Isomers representative of each chlorination class. Response factors at IITRI were calculated from the following equation: RF = AxQ-jS/Ai sQx where Ax =* Ais = Q-fs = Qx = sum of the Integrated 1on abundances of the masses for the unlabeled compound sum of the Integrated 1on abundances of the masses for the appropriate labeled compound amount of the appropriate labeled compound (ng) amount of the labeled compound (ng) Quantification of PCDFs and PCDOs 1n the sample extract was achieved by calculating the ratios of the.mass spectral responses obtained for the 1ons characteristic of the labeled PCDFs/PCDDs to those of the appropriate internal standards, corrected for differences 1n response factor. The equation used for quantification was: Concentration (ng/$) (Ax x Qis)/Ais x W x Rp where W = weight of sample (g) Rp * response factor Because the labeled Internal standard was added prior to sample extraction and analysis, and the internal standard was quantified at the same time as the native components, any losses of PCDD/PCDF Incurred during the analysis were accounted for by the above approach. Detection limits, 1.e. the minimum detectable concentrations required to produce a signal 2.5 times the average background signal, were calculated for each GENP 011183 2-23 783994 PCDF/PCDD coingener class. For each class, the average width of the baseline noiseband for thej native compounds and the peak height for a known concentration of the associated IpC-labeled analog were measured manually. Measurements of the noise band were made 1n a region-of the selected 1on current profile that was free of Interferences and as close as practicable 1n the plot to the peak for the corres ponding l^C-labeled compound. The detection limit DL was calculated using the relationship1: DL (2.5) (Ax x Q i s ) M i s x Rp x W) where Ax height of the noise band of the selected mass' for the unlabeled compound Ais height of the peak corresponding to the labeled compound of the same congener class. Procedures Used by Radian (1) GC/MS analysis o f the sample extracts was performed on a Hewlett-Packard 5985/87 In the selected Ion monitoring mode for quantitation and Identification of PCDD/PCDFs. Each sample was extracted 1n triplicate and each extract was Instrumentally analyzed In triplicate yielding nine determinations on each sample. The first series of extracts were run on a Hewlett-Packard 50 m, Ultra-2 column (a methyl phenyl silicone coating equivalent to DB-5). On the Ultra-2 column, an Initial temperature of 80*C was used. The second and third determinations were performed using- a J&W 60 m, 0.32 mm I.D., 1 pm film, DB-5 column. The GC/MS conditions are shown on Table 19. Dried sample extracts were dissolved in 50 pL of i hexane and a 1.0 pL volume was Injected onto a cool SGE on-column Injector. Each of the three extracts was analyzed a fourth time using a Supelco SP-2331, 60 m, 0.32 mm I.D., 0.2 pm film capillary column to separate 2,3,7,8-TCOF and 2,3,4,8-TCDF.j For this separation, samples were evaporated to dryness and then reconstituted!with 40 pL of nonane. A two-stage temperature program of 170*-200* at 20#/m1n followed by 200*-275* at 4#/m1n was used. Chromatographic Test Mixture Part of this project Involved separating the more toxic PCDD/PCDF Isomers from coeluters. The chromatographic test mixture prepared at Battel!e Memorial Institute contained the following analytes at 200 ng/L each: 1.3.6.8- TCOO 1.2.8.9- TCOD 1,2,3,4-TC0Q 2.3.7.8- TCDD 1.3.7.8- TCDD 3.4.6.7- TCDF 2.3.4.8- TCDF 1,2,3;, 8-TCDF 2,3,7|,8-TCDF 1,2,3,4,8-PnCDF 1,2,4,6,8-PnCDF 2,3,4,6,7-PnCDF 1,2,3,7,8-PnCDF 2,3,4,7,8-PnCDF 1,2,4,6,7,8-HxCDF 2,3,4,6,7,8-HxCDF 1,2,3,4,7,8-HxCDF 1,2,3,6,7,8-HxCDF 783995 This test mixture was run dally to evaluate the chromatography. Chromatograms of all the analytes In the test mixture are shown on Figures 6 through 9 for DB-5 and 2-24 GEOE>0U184 rriioon ^1 00 CO Figure 6. TCDD Test Mixture Chromatogram on DB-5 (Ref. 1) CD COO) 783997 cIrinMo oo ON Figure*7. TCOF Test Mixture Chromatogram on DB-5 (Ref. 1) VwPIJO Figure 8 PnCDF Test Mixture Chromatogram on DB-5 (Ref. 1) 783998 G E N P 011188 Figure 9. HxCDF Test Mixture Chromatogram on DB-5 (Ref. 1) TABLE 19. HRGC-MS ANALYTICAL CONDITIONS3 Mass Spectrometer: Data System: Source Temperature: Internal Source Pressure: Mode: Electron Energy: Electron Multiplier Voltage: Resolution: Interface Temperature: Scan Mode: Cycle Time: Column: Carrier Gas: Carrier Linear Velocity: Injector: Interface to MS: GC Program: Injection Volume: Tailing Factor: Hewlett-Packard 5985 with 5987 Upgrade HP1000 RTE/6 250*C 3 x 10'5 Torr Electron Impact 70 eV 2650-2800 volts Unit 275*C Selected Ion Monitoring 0.5 sec 60 meters x 0.32 mm I.D., J&W 0B-5 fused silica capillary column, 1.0 pm film Hydrogen at 5 ps1 35 cm/sec Cool On-column (65*C) SGE Source Coupling at 275*C Initial hold of 0.5 min at 65*C then 65*-200* at 30*/m1n, then 200*-310* at 4*/m1n 1 ill 0.7-2.0 aRef. 1. Figures 10 through 13 for SP-2331. One analyte, 1,2,4,6,8-PnCDF, could not be found 1n the chromatographic tes't mixture. On DB-5, 2,3,7,8-TCDF and 2,3,4,8-TCDF were not resolved, but on SP-2331 this pair could be-separated with a 65% valley. On SP-2331, 1,2,3,4,8-PnCDF and 1,2,3,7,8-PnCDF could not be resolved, but on DB-5 this pair could be separated with a 55% valley. Actual samples were more difficult to resolve than the chromatographic test mixture. The analytes 1,2,3,4,8-PnCDF and 1,2,3,7,8-PnCDF could not be resolved 1 many cases, particularly when the 1,2,3,4,8-PnCDF concentration was much higher than the 1,2,3,7,8-PnCDF concentration. In one sample (41159-11-14), the G E N P 011189 2-29 784000 t rcoo Figure 10. TCDD Test Mixture Chromatogram on SP-2331 (Ref. 1) 04oo0^ 0 6 in O d K 3 riioo 00 otoo GENP011191 17000 IB000* I.I.M-Tcor i.,.l-TC D r 100 19000* 14000 13000* 12000 IIOOO*. IOOOO OOOO* aooo* I.U.ITCDF j.i.i.i- u tu oo (d o 70 oo . 50 7000* ODDO* 40 9000* 4000 l**S "**1WH*/ 3000 2000 IOOO o 12.0 SupeIco Sp-2311, 60 n, 0.32 mm III, 1 |in riltn. Ramp to 200", then 20Q*-275* 0 4*/mln. W^MI rze io 13.0 I 1 r r | ` . n < i i - i 14.0 19.0 . i I i > r r p I i |-i r t r | t . 11 to. o 17.0 IB. O Figure 11. TCDF Test Mixture Chromatogram on SP-2331 (Ref. 1) I Figure 12. PnCDF Test Mixture Chromatogram on SP-2331 (Ref. 1) G E N P 011192 ~nI oO043^ GENP 011193 Figure 13. HxCDF Test Mixture Chromatogram on SP-2331 (Ref. 1) 2,3,7,8-TCDF and 2,3,4,8-TCDF pair could not be resolved accurately due to a coeluting Interference. Several of the In-service liquid extracts were reported by Radian to be very complex and required the removal of 12 to 18 Inches of the front of the column after every Injection to give adequate resolution for subsequent samples. Column degradation for DB-5 and SP-2331 could be observed from the dally test mixture chromatograms over the course of this work. Calibration Four Cl2/Cl3 area ratio calibration curves were generated during this study, the first on Ultra-2, the second and third on DB-5 and the fourth on SP-2331. Concentrations of 40, 150, 500, 2500, and 7500 ng/mL were calibrated for the first three calibration curves with each calibration point being the average of three determinations. The fourth calibration curve on SP-2331 also was determined in triplicate for concentrations of 40, 150, 500, and 2500 ng/mL. Calibration curves were generated and statistical analysis performed for the following analyte/C13 internal standard sets: 2.3.7.8- TCDD 2.3.7.8- TCDF 2,3,7,8-TCDD-13C 2,3,7,8-TC0F-l3C 1.2.3.7.8- PnCDF 2.3.4.7.8- PnCOF l,2,3,7,8-PnCDF-l3C 1.2.3.4.7.8- HxCDF 2.3.4.6.7.8-HxCDF 1.2.3.4.6.7.8- HpCOF l,2,3,4,7,8-HxCDF-13C 0CDF 0C0D-13C Identification of PCDD/PCDFs Identification of native PCDFs and PCDDs was performed by measuring chlorine Isotope ratios and monitoring the presence of the fragment ions from loss of C0C1. For PCDFs, two chlorine Isotope peaks were monitored In addition to the quantitating Ion. Only the most abundant chlorine isotope peak was monitored for the first series of extracts. The acceptance criteria for PCDD/PCDF positive Identification required agreement better than +10% between the theoretical and experimentally determined chlorine Isotope ratios. The presence of either a second chlorine Isotope peak or a M-C0C1 peak also was necessary for positive Identification. In the few cases where there was substantial Interference 1n the Ion profile of the largest chlorine isotope peak, the second largest chlorine Isotope peak was used for evaluation of the Isotope ratio acceptance criteria. In a few cases, a +15% chlorine Isotope ratio was accepted. If a peak was not within +10% of the criteria on one determination but was within +10% on previous determinations, the determination was accepted. GENP 011194 784005 Quantitation Quantitation was performed by Isotope dilution using the 13C-labeled internal standards added before extraction. Concentrations were determined from C*2/cl3 ratios of sample analyte to Internal standard and a linear fit to the five-point calibration curves. A separate calibration curve was generated for 2,3,4,7,8PnCOF. All other PnCDFs were quantitated from the 1,2,3,7,8-PnCDF calibration curve. A separate calibration curve was produced for 2,3,4,6,7,8-HxCDF. All other HxCDFs and HpCDFs were quantitated from the 1,2,3,4,7,8-HxCDF calibration curve. OCDD and OCDF were quantitated from the OC DO -^ C calibration curve. Congener class total concentrations were calculated by summing concentrations of each peak which passed the Identification criteria for the congener class. Surrogate recoveries were determined using the mean 500 ng/mL response factor from the calibration checks made at the start and end of the analyses. One sample (41159-11-13), had a severe Interference which coeluted with the TCDF-13C peaks. TCDFs 1n this sample were quantitated verus the area of the PnCDF13C quantitating 1on with a normalization factor of 1.1 (the average ratio of TCDF*3C to PnCDF-l3C response). Detection Limits Instrumental detection limits were defined as the analyte concentration which gave a 3:1 signal-to-no1se ratio. Method detection limits for all analytes were approximately 3 ng/g except OCDF and OCDD which were 10 ng/g due to peak broadening at the later retention times. The lowest point on the calibration curves, 40 ng/mL, corresponds to a sample concentration of 8 ng/g. Procedures Used by Battel!e (4) Analysis-- The sample extracts were instrumentally analyzed 1n triplicate and quantified for PCDD/PCDF using combined capillary column gas chromatography/high resolution mass spectrometry (HRGC/HRMS). The HRGC/HRMS consisted of a Carlo Erba Model 4160 Gas Chromatograph interfaced directly Into the 1on source of a VG Model 7070 high resolution mass spectrometer. The chromatographic column was a 60 M DB-5 fused silica column using helium carrier gas at a flow velocity of 30 cm/sec. The mass spectrometer was operated In the electron Impact (El) Ionization mode at a mass resolution of 9000-12000 (M/AM, 10 percent valley definition). All HRGC/HRMS data were acquired by mu!t1ple-1on-detect1on (MID) using a VG Model 2035 Data System. Recovery of Internal Standards-- The recoveries of the Internal standards were calculated by comparison to an external standard, l,2,3,4-TCD0-13Ci2 which was added following the column cleanup steps. Relative response factors were determined from triplicate analyses of a GENP 011195 2-35 784006 standard mixture containing the labelled internal standards and the 1,2,3,4-TCDD- 13Ci2* The equation used to calculate the recoveries was: A 1s * Qrs x 100 Recovery (X) ^ ~ R~ Where: AjS Sum of integrated areas for Internal standard; Qrs * Quantity of recovery standard In ng; Q^s * Quantity of internal standard in ng; Ars Sum of Integrated areas for recovery standard; and Rf a Response factor. Quantification The PCDD/PCDF isomers and congener class concentrations were determined by comparing the sum of the two ion masses monitored for each class to the sum of the two Ion masses monitored for the corresponding Internal standard. It was assumed that the response factors for each of the Individual Isomers in a class were the same as the response factors calculated for the specific isomers present in the standard solutions. The response factors were calculated from six replicate analyses of standard solutions at four concentrations. The sum of the two most Intense ion masses 1n the molecular Ion region of each Isomer were used to generate the response factors. The formula used for quantifying the PCDD/PCDF isomers was: Where: x Q is .Quantity/Sample (ng/g) * Ajs x Rf x W Quantity * Total quantity in ng of target isomer or congener class Ac Sum of Integrated areas for the target Isomer or congener class Qls * Quantity of internal standard 1n ng Afs * Total Integrated areas for the internal standard Rf * Response factor. Each pair of resolved peaks In the selected-ion-current chromatograms was evaluated manually to determine 1f it met the criteria for a PCDD or PCDF isomer. By examining each pair of peaks separately, quantitative accuracy was improved over what 1s obtained when all of the peaks in a selected chromatographic window are averaged. When averaged data are used, it 1s possible for pairs of peaks with high and low chlorine isotope ratios to produce averaged data that meet the chlorine isotope ratio criterion. For example, two pairs of peaks having chlorine Isotope ratios of 0.56 and 0.96, both outside of the acceptable range, would have an average ratio of 0.76. By checking the Isotope ratio of these peaks separately, they would be eliminated from the class total. OENP 0111% 784007 The criterion that were used to Identify PCDD and PCDF Isomers were: 1. Simultaneous' responses at both 1on masses. 2. Chlorine Isotope ratio within +/-15X of the theoretical value. 3. Chromatographic retention times within windows determined from analyses of standard mixtures. 4. Signal-to-no1se ratio equal to or greater than 2.5 to 1.0. The Individual Isomers for which a 1sotop1cally labelled analog was available Included the additional criterion that they eluted within 2 seconds of the Internal standard. The limit of detection (LOO) was calculated for samples 1n which Isomers of a particular chlorine congener class were not detected. The formula used for calculating the LOO was: Hc x Q1s x 2.5 LOO/g (ng) His x Rf x W Where: LOO Single Isomer limits of detection for a congener class in ng; Hc 3 Height of congener class Isomer; Q-fs = Quantity of Internal standard 1n ng; His 3 peak height of Internal standard; Rf a Response factor; and W = Weight 1n grams of sample. Calibration Curve Calibration curves were generated for each of the nat1ve/1sotop1cally labelled isomer pairs. The calibration standards were analyzed at least five times at each of four concentration levels between 0.04 and 2.5 parts-per-mllllon (ppm). The repsonse factors were calculated by dividing the sum of the areas for the two most abundant 1on masses In the molecular 1on cluster of each native PCDD/PCDF Isomer by the sum of the areas for the corresponding 1on masses from the 1sotop1ca11y labelled standard. The Isomer pairs for which response factors were calculated 1ncluded: 2.3.7.8- TCDD vs 2,3,7,S-TCDD-l3Ci2 2.3.7.8- TCDF VS 2,3,7,8-TCDF-l3Ci2 OCDD vs OCDO-l3Ci2 OCDF vs 0CDD-13Ci2 1.2.3.7.8- Penta-CDF vs 1,2,3,7,8-Penta-CDF-13C12 2.3.4.7.8- Penta-COF vs 1,2,3,7,8-Penta-CDF-I3ci2 1.2.3.4.7.8- Hexa-CDF vs l,2,3,4,7,8-Hexa-CDF-13Ci2 2.3.4.6.7.8- Hexa-CDF vs 1,2,3,4,7,8-Hexa-CDF-13:^ 1.2.3.4.6.7.8- Hepta-COF vs 0CDD-13C12 GENP 011197 2-37 784008 It was assumed that each of the Isomers of a particular chlorine congener class had the same response factor. Detection Limits Detection limits were calculated for samples that did not contain PCDD/PCDF Isomers In a particular class. These detection limits were calculated using the quantification equation with the addition of a factor of 2.5 to account for the signal-to-no1se criterion. The noise level was measured as the average of the peak-to-peak noise that occurred 1n the baseline of the particular MID channel. If at all possible, the noise was measured at the retention time of the 1sotop1ca1ly labelled Internal standard. The height of the internal standard peak was measured to the average of the noise that occurred on the peak top. The reported limits of detection are for single PCDD/PCDF Isomers and have been corrected for recovery losses during sample workup. Baseline Analyses-- The three baseline samples, Aroclor 1260, Aroclor 1254-, and Aroclor 1016 were analyzed using the methodology described above. The Aroclor 1016 presented the greatest analytical problem due to the relatively high amount of the sample that passed through the column cleanup steps. Even after the four column cleanup steps, a noticeable polychlorinated biphenyl (PCB) background was evident in the single 1on current chromatograms. In-Service Analyses-- The seven in-service samples were also analyzed using the methodology described above. QUALITY CONTROL Each participating laboratory followed their own extensive quality control program which has been previously reported (1-4). GENP011198 2-38 . 784009 SECTION 3 RESULTS AND DISCUSSION OF STATISTICAL ANALYSIS EVALUATION OF CALIBRATION MODEL The calibration data from four of the five participating laboratories 1n the Round Robin Study was evaluated In terms of the regressions employed, l.e. their appropriateness and linearity over the range of Interest for PCDD and PCDF congener analysis (data from Lab C was not available for regression analysis). For each of the labs, the calibration model 1s: LOG jo (y) = a + b LOGiq (x ) Where x = concentration /"response ratio for LAB-A J peak area ratio for LAB-B y 3 \ area ratios for LAB-D (^normalized counts for LAB-E Exponentiating, the model becomes y a loa xb Since b 1s approximately equal to one for each lab and compound, the model can be simplified to y * CX where C = 10a Tables 20 to 24 give the estimates of C for each lab and compound. Where -a y c* x The Intercepts 1n Tables 21 through 24, except for 2,3,7,8-TCDF (Table 24) were all significantly different from zero at the 0.01 level. Thus, If laboratories employed regression models using a zero Intercept, considerable error 1n the calibration would result. For example, the regression model used by Lab-E was Inappropriate (See Appendix A). Discussions with the Lab-E Project Director at this laboratory Indicated'that the Isotopic response ratios of 13C/12c-PCDF and PCDD congeners were used for quantification, thus, the results reported did not employ the reported regressions. All other laboratories used regression models In which the regression was not forced through the zero Intercept. GENP 011199 3-i 784010 2,3,7,8-TCDD 2,3,7,8-TCDF 1,2,3,7,8-PNCDF 2,3,4,7,8-PNCOF 1,2,3,4,7,8-HXCDF 2,3,4,6,7,8-HXCDF 1,2,3,4,6,7,8-HPCDF OCDF TABLE 20. CONVERSION FACTORS LAB-A LAB-B LAB-0 0.002227 0.001916 0.001661 0.001713 0.001947 0.002217 0.002677 0.0009404 0.003117 0.003795 0.003915 0.003916 0.003657 0.006736 0.004121 0.000611 0.001994 0.002293 0.002513 0.002601 0.002211 0.002275 0.001679 0.001395 LAB-E 0.6901 1.0950 0.8904 0.7414 0.6701 0.4082 0.3281 0.0459 TABLE 21. CALIBRATION CURVES: LAB-A fAverage Response Factor x Concentration\ 500 X = LOGiq (Concentration) 2,3,7,8-TCDD 2,3,7,8-TCOF 1,2,3,7,8-PNCDF 2,3,4,7,8-PNCDF 1,2,3,4,7,8-HXCDF 2,3,4,6,7,8-HXCDF 1,2,3,4,6,7,8-HPCDF OCDF NUMBER OF OBSERVATIONS 6 6 6 6 6 6 6 6 INTERCEPT -2.704** -2.731** -2.819** -2.702** -2.769** -2.632** -2.600** -2.906** **Sign1f1cantly different from zero at .01 level. SLOPE 1.020** 1.005** 1.016** 0.973** 1.023** 0.991** 1.011** 0.949** RZ 0.9994 0.9997 0.9999 0.9996 0.9999 0.9998 0.9995 0.9980 Q 0 > '0 U 2 0 0 784011 TABLE 22. CALIBRATION CURVES: LAB-B Y * LOG10 (Averaqe ResPnse Factor x Concentration^ X * LOG0 (Concentration) NUMBER OF OBSERVATIONS INTERCEPT 2,3,7,8-TCDD 25 -2.475** 2,3,7,8-TCOF 25 -2.432** 1,2,3,7,8-PNCDF 25 -2.418** 2,3,4,7,8-PNCDF 25 -2.414** 1,2,3,4,7,8-HXCDF 25 -2.468** 2,3,4,6,7,8-HXCDF 25 -2.184** 1,2,3,4,6,7,8-HPCDF 25 -2.410** OCDF 25 -3.470** * Significantly different from zero at .05 level. ** Significantly different from zero at .01 level. SLOPE 0.985** 1.005** 1.005** 1.003** 1.014** 1.006** 1.011** 1.091** TABLE 23. CALIBRATION CURVES: LAB-D Y = LOGio (Average Area Ratio) X = LOGio (Concentration) NUMBER OF OBSERVATIONS INTERCEPT 2,3,7,8-TCDD 2,3,7,8-TCDF 1,2,3,7,8-PNCDF 2,3,4,7,8-PNCDF 1,2,3,4,7,8-HXCDF 2,3,4,6,7,8-HXCDF 1,2,3,4,6,7,8-HPCDF OCDF 4 4 4 4 4 4 4 3 -2.691** -2.609** -2.600** -2.636** -2.634** -2.617** -2.778** -2.762** * Significantly different from zero at .05 level. ** Significantly different from zero at .01 level. SLOPE 0.996** 0.988** 1.000** 1.020** 0.991** 0.989** 1.001** 0.966** R2 0.9991 0.9997 0.9994 0.9996 0.9996 0.9991 0.9994 0.9918 R2 0.9999 0.9999 0.9999* 0.9995 0.9999* 0.9999* 0.9999* 0.9999* GENP 011201 3-3 784012 TABLE 24. CALIBRATION CURVES: LAB-E Y LOGip (Raw Counts) X LOGio (Concentration) 2,3,7,8-TCDD 2,3,7,8-TCDF 1,2,3,7,8-PNCDF 2,3,4,7,8-PNCDF 1,2,3,4,7,8-HXCOF 2,3,4,6,7,8-HXCDF 1,2,3,4*, 6,7,8-HPCDF OCOF NUMBER OF OBSERVATIONS 18 18 18 18 18 18 18 18 INTERCEPT -0.286** 0.004 -0.144** -0.230** -0.259** -0.547** -0.668** -1.762** SLOPE 1.047** 1.014** 1.037** 1.038** 1.032** 1.058** 1.067** 1.142 * Significantly different from zero at .05 level. ** Significantly different from zero at .01 level. R2 0.9945 0.9994 0.9977 0.9960 0.9958 0.9893 0.9875 0.9779 GENP 011202 3-4 784013 The slopes were very nearly unity and the r2 of the regression Indicated excellent linearity over the three orders of magnitude concentration range used for calibrating the instrumentation.(Tables 21-24). RESULTS OF ROUND ROBIN Statistical analysis of the PCDD and PCOF concentration data first required the establishment of a standard reporting format for all of the laboratories. To satisfy this requirement, the following data conversions were performed: 1) For LAB-C 1,2,3,6,7,8-HXCDF was excluded because 1t was not reported separately but 1n combination with 1,2,3,4,7,8-HXCOF. 2) If measured amounts were not reported, then detection limits were substituted, If available. 3) For those labs which separately reported 2,3,7,8- and 2,3,4,8- TCDF, or 1,2,3,7,8- and 1,2,3,4,8-PCDF, or 1,2,3,4,7,8- and 1,2,3,4,7,9-HXCDF, the reported concentrations of the pairs, were added together (in order to conform with the reporting format of the other laboratories). Table B-l 1n Appendix B gives a listing of the data. Measured concentrations (ng/g) for each of the Individual Instrumental determinations of a single extract are shown, along with the means and standard deviations of the determinations, by laboratory, sample, and compound. Since LAB-E reported only one Instrumental determination of a single extract for each sample and compound, only that measured concentration Is given. LAB-D reported results for three extracts for each sample with three instrumental determinations for each extract. These extract results are listed separately. LAB-A reported six determinations for OCDD 1n sample 11 and 1,2,3,7,8,9-HXCDF in sample 12. Large ranges 1n concentrations are evident -- for example, by comparing the means among compounds within a sample 1 for LAB-A 1,2,3,4,6,7,8-HPCDF has an average concentration of 354 while OCDD has an average concentration of 2.08. Also, the standard deviations (precision) show great variation 1n magnitude. Because of this variation 1n standard deviations, equal variances could not be assumed. Since the standard deviations tended to Increase as the concentration levels Increased, basic data transformations of taking logarithms and square roots were tried. These, unfortunately, did not correct the problem. This limited the types of statistical tests that could be performed since many depend on the equality of variances. The coefficients of variations (CVs) which represents precision of the analytical methods for the labs, excluding LAB-E, are given 1n Table 25. Only those concentrations reported as measured amounts rather than detection limits were Included 1n the calculations. There appears to be little relation between the measured concentration and the CV, 1.e., the CV does not tend to Increase or decrease as the measured concentration Increases. These CVs are summarized 1n GENP 011203 3-5 784014 TABLE 25. COEFFICIENTS OF VARIATION BY COMPOUND, LAB, AND SAMPLE Sample LAB-A LAB-B ' COMPOUND = TOTAL TCDF 1 16.1 9.4 2 3.8 0.9 3 1.3 3.1 11 47.0 8.6 12 58.9 4.9 13 18.5 12.1 14 23.9 30.0 15 27.6 a 16 21.7 17 4.3 22.3 COMPOUND = TOTAL PNCDF 1 2 3 11 12 13 14 15 V 16 17 10.8 10.2 2.0 6.9 19.2 3.2 2.9 10.9 8.8* 3.1 3.1 0.7 5.8 9.7 4.7 10.8 23.7 61.6 22.5 COMPOUND - TOTAL HXCDF 1 13.1 2 5.8 3 3.5 11 2.6 12 4.6 13 14.6 14 1.9 15 36.2 16 .16.8 17 1.7 1.0 2.3 2.1 14.1 3.7 27.5 4.0 54.0 14.0 COMPOUND = TOTAL HPCDF 1 4.4 10.9 2 11.8 20.4 3 3.6 16.8 11 8.6 19.2 12 14.3 20.2 13 21.4 39.4 14 2.6 14.1 15 56.2 62.9 16 58.9 17 6.5 22.6 LAB-C 15.8 8.7 27.0 61.0 27.7 26.0 30.1 22.2 41.5 13.9 8.3 50.4 56.9 10.2 31.5 10.8 137.1 83.3 64.6 99.3 20.3 10.7 55.3 42.0 96.2 76.1 65.6 52.2 28.6 62.4 21.9 18.3 LA8-D-1 5.6 2.4 6.6 13.7 6.8 11.3 10.5 9 9 11.9 7.6 0.8 11.8 15.4 13.6 17.5 28.5 12.1 1.3 - 1.7 10.9 11.9 8.5 22.5 12.4 44.1 47.0 9.2 9.7 3.2 2.7 18.3 13.9 9 19.4 38.7 70.1 15.1 LAB-D-2 6.5 4.2 4.4 12.0 3.1 11.5 3.6 9 9 9.4 3.6 2.7 10.6 2.0 2.6 6.5 3.1 9 11.1 1.0 0.9 30.3 1.8 5.7 18.4 6.2 5.6 1.3 5.1 4.3 6.6 0.8 5.3 3.5 18.2 12.8 5.4 LAB-D-3 1.9 1.0 1.3 6.2 18.3 12.6 2.1 s 0 0.6 20.9 1.1 5.0 0.2 2.3 8.0 1.2 8.8 1.7 4.8 1.9 1.8 5.1 26.9 1.0 24.1 0.3 2.4 6.3 7.1 5.0 4.2 1.9 10.0 1.2 784015 GEKP 011204 TABLE 25 (cont'd.) Sample LAB-A COMPOUND 3 TOTAL OCDF LAB-B LAB-C 1 7.9 9 2 10.8 45.8 3 3.3 30.6 32.6 11 5.7 23.8 14.4 12 8.8 12.7 10.8 13 10.7 60.7 14 2.5 40.3 20.0 15 34.1 9 16 9 17 8.9 8.8 8.5 COMPOUND = TOTAL TCDD 1 11.8 17.0 2 16.0 2.2 2.5 3 3.3 9.2 15.8 14 56.6* 16 43.9* COMPOUND = OCDD 11 25.1** 14.5 91.7 12 18.1 13 19.0 6.9 14 3.0 46.6 15 11.5 24.4 16 12.2 9.9 11.8 17 20.5 28.5 30.1 COMPOUND = 2,3,7,8-TCDD 1 11.8 17.0 11.4 2 16.0 '2.2 2.5 3 3.3 9.2 15.8 COMPOUND 3 2,3,7,8 + 2,3,4,8-TCDF 1 15.5 9.4 15.8 2 4.5 0.9 8.7 3 0.8 1. 0 19.5 11 24.3 45.7 . 16.4 12 21.1 4.3 24.7 13 16.5 4.4 15.6 14 16.3 9.1 6.7 15 9.3 16 16.4 17 3.7 22.0 15.6 LAB-D-1 26.5* 1.6 2.9 13.6 12.2 39.0 4.5 4.8 47.8 12.3 2.3 3.0 1.7 9 9 9 9 2.3 3.0 1.7 5.6 2.4 1.1 9.6 11.5 9.9 9.7 9 9 14.5 LAB-D-2 4.1 6.0 8.5 2.4 4.4 11.8 2.6 27.4 3.6 4.0 2.3 5.3 # * 8.3 4.0 2.3 5.3 6.5 4.2 3.8 0.0 5.4 5.6 2.1 6.4 LAB-D-3 . 5.5 5.4 5.0 4.1 4.2 16.8 4.7 2.1 4.3 3.1 1.6 t 9 4.3 3.1 1.6 1.9 1.0 2.2 2.9 15.9 2.2 1.2 9 9 1.1 GENP 011205 3-7 784016 TABLE 25 (cont'd.) Sample LAB-A IAB-B LAB-C COMPOUND 1,2,3,7,8 + 1,2,3,4, 8-PNCDF 1 12.0 2.1 11.5 2 11.5 1.8 3.1 3 0.9 4.1 13.6 11 12.9 6.3 31.6 12 26.1 4.8 23.1 13 8.4 7.1 12.8 14 3.8 4.6 7.3 15 11.7 16 61.1 17 4.0 14.9 2.2 COMPOUND = 2,3,4,7,8-PNCDF 1 11.1 4.2 60.1 2 15.3 1.8 23.4 3 2.9 4.4 37.8 11 2.9 5.8 9.4 12 22.2 10.1 5.1 13 8.7 6.3 17.2 14 1.4 5.8 14.8 15 11.0 16 14.1* 60.1 17 7.2 13.7 7.3 COMPOUND => 1,2,3,4,7,8 + 1,2,3, 4,7,9-HXCDF 1 13.2 0.6 36.7 2 40.2 34.6 * 3 3.2 3.5 8.6 11 2.1 13.9 100.9 12 1.9 6.6 16.0 13 17.0 23.1 33.6 14 0.6 1.2 30.8 15 35.8 16 5.8 59.2 17 5.4 14.1 43.0 COMPOUND = 1,2,3,7,3,9-HXCOF 3 5.4 80.9 11 * v 82.3 12 27.1** 20.3 13 33.2 36.5 14 4.8 15.7 88.5 15 42.4 16 61.9 , 17 a 53;3 LAB-D-1 1.4 1.3 4.0 10.4 16.6 6.3 15.6 9 9 11.5 13.8 0.4 0.7 32.4 22.0 10.9 20.1 22.0 1.3 1.9 8.0 10.1 7.5 9.9 57.3 38.6 10.3 4.7 11.0 6.9 14.0 * 4.8 LAB-0-2 3.0 2.1 6.8 2.1 4.5 7.5 17.3 18.4 2.1 3.3 5.2 3.7 3.2 1.8 1.3 4.9 2.4 2.0 1.7 0.6 2.3 16.3 4.3 0.8 5.1 4.4 5.4 24.1 26.9 ft 2.3 LAB-D-3 0.7 1.0 3.7 2.2 2.9 0.0 5.7 # 4.5 2.1 1.4 2.8 2.7 5.3 2.5 1.9 4.9 3.7 3.0 1.7 4.7 8.4 3.4 27.2 19.2 0.9 1.5 2.5 1.5 15.7* 1.3 6.6 GENP0II206 3-8 784017 TABLE 25 (cont'd.) Sample LAB-A LAB-B COMPOUND = 1,2,3,6,7,8-HXCDF 1 14.9 0.0 2 3.4 2.5 3 4.0 3.2 11 1.0 19.6 12 2.7 5.3 13 12.3 27.2 14 1.7 2.6 15 34.9 16 64.9 17 3.0 2.0 COMPOUND = 2,3,4,5,7,8-HXCDF 1 6.2 1.5 2 7.3 4.2 3 4.8 4.7 11 12.7 13.4 12 25.0 19.5 13 8.8 31.5 14 5.2 13.7 16 51.3 17 12.3 19.2 COMPOUND = 1,2,3,4,6,7,8-HPCDF 1 * 4.0 2 11.7 3 4.1 11 10.3 12 9.9 13 21.1 14 1.6 15 16 17 3.4 10.9 20.4 12.1 17.8 18.7 31.6 11.5 60.6* 53.4 23.9 LAB-C e 9 159.1 86.9 80.9 121.3 30.6 41.2 125.1 52.7 96.2 76.1 144.9 58.9 36.9 60.9 39.9 29.2 LAB-0-1 4.1 5.3 5.2 7.0 8.1 7.6 7.9 2.2 0.4 1.2 15.3 21.7 11.2 5.8 9.7 3.2 1.4 17.6 14.9 . 20.4 21.7 57.4 14.9 LAB-D-2 5.9 2.7 9.3 20.8 7.9 10.2 14.7 22.5 1.6 0.4 11.4 4.0 3.2 10.2 26.2 1.6 5.1 4.3 5.4 0.7 5.5 7.5 17.6 6.7 4.8 LAB-D-3 8.1 7.6 2.0 1.3 27.6 16.0 8.7 1.3 3.7 2.1 0.5 5.4 23.6 2.0 5.5 2.4 6.3 5.7 4.0 4.9 1.2 10.3 1.6 * Based on two rather than three determinations. ** 8ased on six rather than three determinations. a. 3 No measurable value reported, 1.e., not detected, therefore no CV was calculated. GETiF 011207 3-9 784018 Table 26 showing the minimum, median, and maximum CVs over all compounds by laboratory and sample. Overall, Lab-C shows the largest median and maximum values for CVs, but there 1s also variation among the other laboratories. TABLE 26. MINIMUM, MEDIAN, AND MAXIMUM CVs BY LAB AND SAMPLE OVER ALL COMPOUNDS MINIMUM LAB-A LAB-B LAB-C LAB-D-1 LAB-D-2 LA8-D-3 MEDIAN LAB-A LAB-B LAB-C LAB-D-1 LAB-0-2 LA8-0-3 MAXIMUM LAB-A LAB-B LAB-C LAB-D-1 LAB-D-2 LAB-D-3 1 4.01 0.00 11.36 1.25 0.98 0.71 11.78 4.25 41.53 4.85 4.00 2.40 16.06 16.99 159.06 26.52 6.53 20.86 2 3.45 0.67 2.55 0.40 0.41 0.96 11.15 2.25 13.87 2.44 2.70 3.08 40.29 45.81 86.90 5.30 4.29 7.64 Sample 3 11 12 13 14 17 0.83 1.02 8.33 0.74 2.04 1.31 0.99 5.83 9.36 6.97 0.00 0.17 1.90 3.71 5.13 6.84 0.58 1.55 3.25 4.41 10.20 6.25 1.75 0.00 0.57 1.24 6.68 4.48 1.28 0.98 1.71 1.96 2.16 4.84 0.83 0.34 3.26 4.56 29.82 2.68 5.39 2.21 8.63 14.10 58.92 13.60 2.08 2.53 19.21 6.62 23.94 12.22 4.54 4.92 15.54 17.63 31.23 11.08 7.53 8.24 2.86 12.62 . 30.49 12.39 14.65 1.93 4.34 19.22 22.18 11.94 4.90 2.07 5.42 30.59 144.88 11.82 30.30 7.11 46.96 45.73 121.26 32.44 20.80 6.21 58.89 20.24 56.85 22.05 7.86 27.56 33.18 39.41 62.39 39.04 24.12 26.91 56.57 46.63 125.10 28.47 26.89 16.84 20.47 28.49 53.35 22.04 22.51 8.84 The spiked versus measured amounts 1n samples I and 2 are compared 1n Tables 27 and 28. The percent accuracy by compound, lab, and sample for the arithmetic mean, the geometric mean, the minimum, and the maximum are given 1n Table 27. The percent accuracy was determined by calculating the summary statistic, dividing by the spiked amount, then multiplying by 100. For example, % accuracy mean = mea-- x 100 spike With the exception of Total PNCDF; Total HXCDF; 2,3,4,7,8-PNCDF; 2,3,4,6,7,8-HXCDF, and I,2,3,4,6,7,8-HPCDF for LAB-C, these numbers generally fall In the 80 to 120 range. To summarize the percent found, Table 28 gives the minimum, median, and maximum percent found by laboratory. Percent found expresses each determination as * a percentage of the spiked amount similar to the formula above. LAB-C had by far 784019 3-10 GENP 011208 G E N P 011209 TABLE 2?. PERCENT ACCURACY FOR SPIKED SAMPLES BY COMPOUND, SAMPLE. AND LAB Saapla I Saapla 2 Lib ArMletahnaattc CaMoeaaentrlc HfnfMM Haxfaua ArMiethamnetic GeoMmeaentric Hlnfmua HaaiMM COMPOUND - TOTAL TCOF LLlabb--fAi Lab-C LLiibb--D0 EEaatt 21 LLaafba--EOEat J 791..09 9I5I..10 999522...017 97411...223 11.7 999225...570 7723..55 04..00 1981..00 95.0 9077..50 19140..00 999459...505 1II1S8.97 119 5 111128..62 111 1 117.0 111158..7 111111290...522 111173..00 111145..17 110157..15 111101737...025 1I7I3T..6I 112260..4 111157..10 117.0 COHPOUtO - TOTAL PNCDF lab-A LLaabb--C0 LLLaaabbb---000EEEaaattt 221 Lab-E 94.0 19249..41 190627...311 106.7 9939..67 119016621....4630 106.7 91763....3777 9606..37 106.7 104.0 111100311..330 11900956...773 1996004...766 1996003...465 9965..12 9905..12 19167..21 19167..12 19941475....1363 96.5 19114..19 106.1 1991599...499 19016..31 117.1 COMPOUM) - TOTAL HXCDF tio LLaabb--A0 Lab-C LLaabb--00 EEaatt 21 LLaabb--E0Eat 1 14.3 51309121...322 19091..12 103.2 3.1 21590312...273 19091..21 103.2 73.7 190721...135 19070..49 103.2 95.B 99.2 1310093..51 (11000213...123 114.4 1151102205272....9992 18015..12 111242..13 1152.3 110052..99 10015..11 100.6 120.0 41129001914....6659 11.1 121.6 2911i11002o1465a.....95946 l.l COMPOUND TOTAL HPCDF Lab-A Lab-B LLaabb--C0 Eat 1 LLLaaabbb---E00EEaatt 23 911..69 116005S2...314 19123..51 93119...229 110012..63 19132..15 1850..53 19012..103 91.5 112.5 93.0 100.0 111311324...155 19126..50 111090..25 11310454...124 1912090...730 111044..13 19199..98 19199..70 IDO.9 1126..41 11900492...735 119.0 135.4 16142..67 111010906...427 119.1 COMPOUND OCDF LLaabb--A1 LLaabb--C0 Eat 1 LLaabb--00 EEaatt 32 Lab-E _.a -- -- -- -- -- -- _ -- -- -- -- -- -- __ -- -- -- --_ _ -- -- -- -- -- -- 19--012..25 90.1 9--4.7 999627...090 42.4 96.0 9926..99 42.4 I5--5I..31 9049..74 9423..42 11--0409..12 1990771...070 42.4 04Oro0^o TABLE 27. (cont'd.l Sixpt* 1 Sixple 2 Lib AriHtehimnetic GeoMm*ei/tiric Hlntaux Hixtxua AriHtehimnetic GeHoeMilnrlc HIn1m m Hixlxux COMPOUND - TOTAL ICOO Lib-A LLiibb--CA LLiibb--00EExxtt 11 Ilxibb--E0 Ext 1 75.3 1120..41 159906101....000) 75.0 91101...439 99.9 5931..09 691160....4000 96.0 9565..00 956..20 11001040...400 5966..00 19001..39 19001..12 *5.5 5.5 999355...766 71.7 9955..66 9731..77 17040..10 13.0 9942..35 9701..67 110012..1 ' 8986..11 91.1 7916..72 COMPOUND - OCOO Lab-A LLiibb--fCl LLLiiibbb---D00EEExxxttt 321 Lib-E ---- ---- -- COMPOUND - 2.3.7.1-TCOO -- -- -- -- -- -- ---- ---- -- ---- ---- -- ---- -- -- ----- -- "" ---- ---- -- - -- ---- -- >rL--Isaj* LLLiiibbb---fCAl LLiibb--00 EExxtt 21 LLiibb--E0 Ext 3 11132...337 1990201...00T 56.0 7151..09 91..30 399391...909 6611..00 9766..00 591661...000 85.2 9926..00 110004..00 5966..00 19001..93 90.1 101.2 919555...566 989555...366 9731..77 9731..77 74.7 191904320....3300 90.6 71.7 102.B 19M0163.l..1A1 96.2 71.7 COMPOUND - 2,.3.7.1 2.3.,4.fl-TC0F LLiibb--fAl LLLiiibbb---CD0 EExxtt 21 LLiibb--E0 Ext 3 1.1 I751I..I03 9922..7 -94.5 711..13 9I4I..)2 999422...357 73.0 7034...300 1981..00 94.3 97.5 18170..00 99994941....5550 117.5 111195..57 111111112318....6852 117.4 115.7* 111198..22 111121..58 1IB.5 114.2 111011575...115 110173..52 111.5 111221367...640 I11I12T50...1O 110.5 COKPOUNO 1,.2.3.7.1 1.,2.3.4.1-PHCDf Lib-A LlLLLLiiiiiibbbbbb------CBE000EEExxxttt 321 90.3 119990007131.....03333 106.7 19.9 1909191...35S 9913..33 106.7 19903006....3700 1890926...377 110024..00 1991043...377 19064..70 100.9 1199019746....4000 19100..35 11190010946....4400 1992780...305 87.5 11190831...19 1992570...635 IOS.3 111200Q50...163 1992909...344 G E N P 011210 ItIIIII M IIIM ssasss mNN* ifiOo#a^apm 1 Is U P*O * Ofc%V QHOn9 wnrtOO9f1f0i1f9lMn Mitili I1 % s rQltOr*tOi9OQOit9n*n* i i it ii ii ii ii s* II O > O O O r* IrtOOPOPP*^ IM III! SSoo^S PPQPtflOP i ii ii ii ii ii ii S --o g m s iONtOCOl ^oAeoNiI ^0^00 Q9I NOOO PM009I ^Pl PfN*Oo--QN--n-P-i-I r^OttfOQ1 t IIIj !I NMppttd i0(W s * OtUOlOOl N QPMf0tIf0tOAOOI*A IlIM M PP^pOPP Is 9 101v n o o o N P O O m I M M I I * *P tffpl0ft P* 0 I O O I T4 4f 4 4? 4? 4? 4? ?4 44 4 4 4 4 ?? ?44<4?44?4 4444444 GENP 0JJ2JI 3-13 784022 Saapla I ArlthHtk Gaonatrlc Ufa Naan Haan Minima CQMPOUW - 1.2.3.4,4.7.6-HPCOF Uto~A Lato-4 Lab-C Lab-0 Ext 1 Lab-0 EkI 2 Lab-0 Imi 3 Lab-E 11.4 *1.4 SS.3 102.1 104.4 3.1 112.S 114 1.2 3.2 102. 101.3 3-1 112.9 IS. 3 10.5 1.3 1.0 102. 1.5 112.9 *B1om Hilt of datactlon. TABLE 27 (cool'd.} Nu (m Artthacllc Naan Salila 2 (Mirle Haan Minima 92.3 100.0 132.9 114.5 113. 96.0 112.5 111.9 100.9 35.2 104.4 114.1 . 99. 119.0 110.4 99.0 28 3 104.3 114.1 99.1 119. 109.3 is. 12.4 102.5 109.7 4.3 11. Mastma 134. 112.7 64.1 10.2 119.4 10.7 119.1 u -P" GENP 011212 TABLE 28. MINIMUM, MEDIAN, AND MAXIMUM PERCENT FOUND (ACCURACY) FOR ALL PCDF AND PCDD CONGENERS IN SPIKED SAMPLES 1 AND 2 LAB LAB-A LAB-B LAB-C LAB-0 LAB-D LAB-D LAB-E Minimum 68.0 68.0 8.2 80.7 88.0 66.7 56.0 Sample 1 Median 87.8 96.0 94.2 98.7 100.8 94.5 103.2 Maximum 105.3 124.0 3371.4 114.5 113.8 107.2 116.0 Minimum 74.7 55.3 12.4 90.5 89.4 90.6 42.4 Sample 2 Med1 an 106.8 112.0 120.3 100.8 104.9 98.7 113.9 Maximum 135.4 149.2 3939.4 120.8 119.4 120.8 120.2 the largest ranges. Among the other laboratories, LAB-B had the highest maximum values, while LAB-E had the lowest minimums. To provide a better comparison of the analytical method performance among the laboratories, the following steps were taken: 1) Samples 15 and 16 were dropped from further analysis because few compounds were found above the detection limit. 2) To eliminate the missing data for LABS-B and C caused by nonreportlng of detection limits, the following values, based upon detection limits reported by LAB-D, were Imputed: Compound_______ Limit of Detection (nq/q) Total TCDD 3 OCDD - 10 2,3,7,8-TCOD 3 2,3,4,7,8-PCDF 3 Analyses of variance were then run on the altered data to determine the sources of variation. Table 29 gives the total variance by compounds for each laboratory and then breaks 1t down Into sample (different types analyzed) and error (Instrumental measurement) variances. Since LAB-E reported only one observation per compound and sample, Its variance could not be broken down, but the total variance can be compared with the other labs. Generally, the TCDF, PNCOF, and OCDD compounds showed smaller variances, while OCDF showed the largest variances. There was some variability among laboratories 1n amount of total variance with LAB-E having a tendency to be higher, but no clear cut patterns emerge. Lab-C had the largest measurement error for almost every congener. The greatest percent of variation, however, was due to sample differences as reflected 1n comparing percent sample variance with percent measurement error variance.- on*3 3-15 7B4024 TOTAL TCDF LAB-A LAB-0 LAB-C LAB-D Ext.-l LAB-D Ext.-2 U B - D Ext.-3 UB-E TOTAL PNCDF LAB-A LAB-B LAB-C LAB-0 Ext.-l LAB-D Ext.-2 LAB-D Ext.-3 LABrE TOTAL HXCDF LAB-A LAB-B LAB-C U B - D Ext.-l U B - 0 Ext.-2 U B - D Ext.-3 UB-E TOTAL HPCDF UB-A UB-B UB-C U B - D Ext.-l U B - D Ext.-2 U B - 0 Ext.-3 UB-E OCDF LAB-A UB-B UB-C U B - D Ext.-l U B - D Ext.-2 U B - 0 Ext.-3 UB-E TABLE 29. VARIANCE COMPONENTS BY COMPOUND AND LAB Variance Components Measurement Total Sample Error As Percent of Total Measurement Sample Error 1.23 1.24 1.33 1.52 1.31 1.37 1.66 1.14 1.22 1.22 1.51 1.30 1.36 .a 0.09 0.02 0.12 0.01 0.01 0.01 - 93.0 98.3 91.1 99.4 99.5 99.5 - 7.0 1.7 8.9 0.6 0.5 0.5 - 1.55 1.37 1.57 1.71 1.55 1.60 1.90 1.54 1.36 1.42 1.68 1.55 1.59 - 0.01 0.02 0.15 0.02 0.00 0.01 - 99.4 98.9 90.3 98.6 99.7 99.5 - 0.6 1.1 9.7 1.4 0.3 0.5 - 3.34 2.96 3.50 4.42 3.18 3.38 4.20 3.33 2.94 2.86 4.41 3.16 3.36 - 0.01 0.01 0.65 0.01 0.02 0.01 - 99.8 99.5 81.6 .99.7 99.4 99.7 - 0.2 0.5 18.4 0.3 0.6 0.3 - 4.20 3.19 4.65 6.06 3.76 5.55 5.35 4.19 3.15 4.06 6.04 3.75 5.49 - 0.01 0.05 0.59 0.02 0.01 0.06 - 99.7 98.5 87.3 99.7 99.8 98.9 - 0.3 1.5 12.7 0.3 0.2 1.1 - 10.94 6.50 6.46 8.52 6.41 8.19 9.86 10.93 6.43 6.30 8.48 6.40 8.19 - 0.01 0.07 0.16 0.04 0.00 0.00 - 99.9 98.9 97.5 99.5 99.9 99.9 - 0.1 1.1 2.5 0.5 0.1 0.1 - 784025 GENP 011214 TABLE 29 (cont'd.) TOTAL TC0D LAB-A LAB-B LAB-C LAB-D Ext.-l LAB-D Ext.-2 LAB-D Ext.-3 LAB-E OCDD LAB-A LAB-B LAB-C LAB-D Ext.-l LAB-D Ext.-2 LAB-D Ext.-3 LAB-E 2,3,7,8-TCDD LAB-A LAB-B LAB-C LAB-D Ext.-l LAB-D Ext.-2 LAB-D Ext.-3 LAB-E 2,3,7.8 + 2,3,4.8-TDCF LAB-A LAB-B LAB-C LAB-D Ext.-l LAB-D Ext.-2 LAB-0 Ext.-3 LAB-E 1.2.3.7.8 + 1,2.3.4.8-PNCDF LAB-A i LAB-B LAB-C LAB-D Ext.-l LAB-D Ext.-2 LAB-D Ext.-3 LAB-E Variance Components Measurement Total Sample Error 5.61 2.22 2.24 2.25 2.30 2.19 7.52 4.34 2.22 2.24 2.25 2.30 2.19 - 1.27 0.00 0.00 0.00 0.00 0.00 - 2.73 0.24 0.59 - _ _ 0.34 2.15 0.15 0.53 _ _ - - 0.58 0.09 0.06 - - - 8.05 5.46 2.12 2.25 2.30 2.19 7.52 6.87 5.45 2.12 2.25 2.30 2.19 - 1.18 0.00 0.00 0.00 0.00 0.00 - 1.13. 1.32 1.70 1.57 1.43 1.43 1.65 l.il 1.29 1.67 1.56 1.42 1.43 - 0.02 0.03 0.03 0.01 0.00 0.00 - 1.70 1.54 2.10 1.94 1.92 1.65 2.02 1.69 1.54 2.08 1.93 1.91 1.65 - 0.02 0.00 0.03 0.01 0.01 0.00 - As Percent of Total Measurement Sample Error 77.4 99.8 99.8 99.9+ 99.9+ 99.9+ - 22.6 0.2 0.2 0.0 0.0 0.0 - 78.7 63.1 89.2 - 21.3 36.9 10.8 -- - 85.4 99.9 99.8 99.9+ 99.9+ 99.9+ - 14.6 0.1 0.2 0.0 0.0 0.0 - 98.0 97.6 98.4 99.5 99.8 99.8 - 2.0 2.4 1.6 '0.5 0.2 0,2 99.0 99.7 98.8 99.5 99.5 99.9 - 1.0 0.3 1.2 0.5 0.5 0.1 G E W 011215 784026 2,3,4,7,8-PNCDF LAB-A LAB-B LAB-C LAB-D Ext.-l LAB-0 Ext.-2 LAB-0 Ext.-3 LAB-E 1,2,3,4,7,8 + 1,2,3,4,7,9-HXCDF LAB-A LAB-B LAB-C LAB-0 Ext.-l LAB-0 Ext.-2 LAB-D Ext.-3 LAB-E 1,2,3,7,8,9-HXCDF LAB-A LAB-B LAB-C LAB-0 Ext.-l LAB-D Ext.-2 LAB-0 Ext.-3 LAB-E 1.2.3.6.7.8-HXCDF LAB-A LAB-B LAB-C LAB-0 Ext.-l LAB-0 Ext.-2 LAB-0 Ext.-3 LAB-E 2,3,4,6,7,8-HXCDF LAB-A LAB-B LAB-C LAB-0 Ext.-l LAB-0 Ext.-2 LAB-0 Ext.-3 LAB-E TABLE 29 (cont'd.) Variance Components Total Measurement Sample Error As Percent of Total Measurement Sample Error 1.27 1.20 1.65 1.30 1.32 1.29 1.76 1.26 1.20 1.58 1.26 1.32 1.29 - 0.01 0.01 0.08 0.04 0.00 0.00 - 98.9 99.5 95.4 97.0 99.9 99.9 - 1.1 0.5 4.6 3.0 0.1 0.1 - 6.47 7.25 4.48 5.93 5.69 5.78 8.71 6.44 7.22 4.24 5.93 5.69 5.78 - 7.86 3.50 4.68 4.88 4.46 4.86 6.06 7.48 3.50 4.04 4.86 4.44 4.85 - 2.99 2.48 _ 3.01 2.32 2.85 3.79 2.99 2.46 3.00 2.30 2.83 - 3.27 2.62 4.53 3.11 1.77 1.85 4.27 3.25 2.59 3.18 3.10 1.76 1.84 - 0.03 0.03 0.24 0.01 0.00 0.00 - 0.38 0.00 0.65 0.02 0.02 0.01 - 0.01 0.01 0.00 0.02 0.02 - 0.01 0.03 1.38 0.01 0.01 0.01 - 99.6 99.6 94.6 99.9 99.9 99.9+ - 95.2 99.9 86.2 99.7 99.6 99.8 - 99.8 99.4 99.9 99.2 99.3 - 99.5 99.0 69.6 99.6 99.2 99.5 _ 0.4 0.4 5.4 0.1 0.1 0.0 - 4.8 0.1 13.8 0.3 0.4 0.2 0.2 0.6 0.1 0.8 0.7 0.5 1.0 30.4 0.4 0.8 0.5 3-la - 784027 GENP 011216 1,2,3,4,6,7,8-HPCDF LAB- LAB-B LAB-C LAB-0 Ext.-l LAB-0 Ext.-2 LAB-D Ext.-3 LAB-E TABLE 29 (cont'd.) Variance Components Measurement Total Sample Error As Percent of Total Measurement Sample Error . 4.05 3.01 3.97 4.43 3.32 4.00 5.28 4.04 2.97 3.10 4.41 3.32 3.94 0.01 0.04 0.87* 0.02 0.01 0.06 99.8 98.7 78.1 99.6 99.8 98.5 0.2 1.3 21.9 0.4 0.2 1.5 aBelow limit of detection. GENP 011217 3-19 784028 For LAB-D, the variance component for repetitive extraction of the same sample could be included because data for three extracts was reported (see Table 30). Here as before, at least 95% of the variance was 1n the samples with extract and Instrumental measurement error contributing little. To gain further understanding of the differences 1n variance, Table 31 gives the standard deviations (the square root of the variances) by compound, sample, and laboratory (recalling that each lab used different analytical procedures as described 1n Section 2.0). The laboratories were then ranked by compound and sample according to the size of the standard deviation, l.e. for TCDF 1n Sample 1, LAB-D-3 had the smallest standard deviation while LA8-C had the largest. Generally, LAB-C had larger standard deviations (and therefore poorer accuracy and precision from their analytical method). While the other laboratories, excluding LAB-C, show no large differences, LAB-D tended to be lower (and therefore the best accuracy and precision from their analytical method). Also shown are the results of Bartlett's tests for homogeneity of variance among laboratories, by compound and sample (5). Here again, there 1s variability In the results. For some of the compound-sample combinations, the labs were significantly different at the .01 level while for other combinations the labs were not significantly different at the .05 level. Table 32 gives the results of Levene's test for homogeneity of variance among laboratories (over samples), by compound (6). For every compound, the laboratories were significantly different from one another at the .05 level, and for all but one compound (1,2,3,6,7,8-HXCDF), they were significantly different from one another at the .01 level. Table 33 provides approximate 95-percent prediction Intervals of levels reported from future analyses of the samples. The seven laboratories (LAB-0 Included three times) were treated as a random sample from a large population of laboratories, and a variance component, denoted as VL, of the log concentrations due to laboratories (and therefore analytical method used) was thus estimated. A measurement error variance component, VM, was also estimated. This component 1s the wlthln-extract variance of the log concentrations. Data from the LAB-E were not used 1n determining this component of variance, since there were no replicate observations. The two components were used to calculate the prediction intervals via the following formula: exp {x * 2V}, where x * mean of the log concentrations (across laboratories and extracts), and V * square root of (VL + VM). 784029 3-20 TABLE 30. 'VARIANCE COMPONENTS BY COMPOUND FOR LAB-0 Compounds Variance Components Total Sample Extract Error As Percent of Total Sample Extract Error Total TCDF Total PNCOF Total HXCDF Total HPCOF OCDF Total TCDD OCOD 2,3,7,8-TCDD 2,3,7,8 + 2,3,4,8-TCOF 1,2,3,7,8 + 1,2,3,4,8-PNCDF 2,3,4,7,8-PNCDF 1,2,3,4,7,8 + 1,2,3,4,7,9-HXCDF 1,2,3,7,8,9-HXCDF 1,2,3,6,7,8-HXCDF 2,3,4,6,7,8-HXCDF 1,2,3,4,6,7,8-HPCDF 1.40 1.62 3.66 5.13 7.73 2.25 0.00 2.25 . 1.48 1.38 1.61 3.59 5.00 7.41 2.25 0.00 2.25 1.47 1.84 1.82 1.31 5.80 1.30 5.80 4.73 2.73 2.25 3.92 4.71 2.69 2.16 3.87 0.01 0.00 0.05 0.09 0.30 0.00 0.00 0.00 0.00 0.01 0.00 0.00 0.01 0.02 0.08 0.02 0.01 0.01 0.01 0.03 0.02 0.00 0.00 0.00 0.00 98.9 99.2 98.3 97.6 95.9 99.9+ 99.9+ 99.9+ 99.5 0.01 99.2 0.01 0.00 98.9 99.9 0.01 0.01 0.01 0.03 99.5 98.6 96.1 98.7 0.6 0.0 1.3 1.8 3.9 0.0 0.0 0.0 0.2 0.4 0.0 0.0 0.2 0.8 3.4 0.5 0.5 0.8 0.4 0.6 0.2 0.0 0.0 0.0 0.3 0.4 1.1 0.1 0.3 0.5 0.5 0.7 iy p 01>219 3-21 784030 TABLE 3] STANOARO DEVIATIONS. THEIR RANKS. ANO BARTLETTS TEST BY COMPOUND ANO SAMPLE S a v i* LAS-A LAB-B CQMPOUNO TOTAL TCOF Standard Qaviations LAB-C 1LAB-O-r LAS-O-2 LAfl-O-3 1 28.31 14.60 2 2.36 0.58 3 20.82 49.32 11 101.01 8.33 12 182.17 6.35 13 143.93 60.92 14 261.08 216.61 17 23.39 78.17 COMPOUND TOTAL PNCQF 30.00 5.51 366.92 65.06 20.81 130.00 217.03 65.06 10.02 1.53 98.15 19.87 4.51 97.12 92.48 59.17 12.10 2.52 65.51 20.31 1.06 92.67 3D.75 39.55 3.51 0.56 20.55 10.44 19.60 109.61 18.45 3.21 1 30.31 9.29 155.03 21.94 11.14 54.93 2 31.73 2.0 72.11 2.52 1.39 3.46 3 26.45 66.37 100.00 166.96 134.52 67.80 11 123.42 166.86 860.31 312.68 41.07 3.52 12 74.61 15.26 177.30 42.45 7.09 7-23 13 11.53 27.10 25.17 46.48 20.64 26.51 M 300.00 1722.88 2542.31 2886.65 275.13 113.15 17 35.12 218.48 64.29 161.04 131.23 109.55 COHPOUNO TOTAL HXCOF 1 52.63 4.51 3504.6 2 12.22 5.29 2354.6 3 86.22 40.17 7689.6 11 172.14 913.30 24055.4 12 43.02 36.37 257.0 13 6.48 17.35 15.3 14 200.00 413.63 12165.5 17 30.55 2B3.28 66G0.8 COMPOUND TOTAL HPCOF 6.08 3.21 271.01 828.67 85.04 3.61 1217.54 206.65 4.73 1.73 622.22 126.26 49.90 9.45 590.93 30.29 7.94 9.29 42.15 131.05 44.66 11.93 96.11 8.08 I 15.57 2 73.90 3 85.44 11 793.87 12 295.01 13 2.16 14 52.92 17 107.86 COMPOUNO OCOF 40.00 107.98 175.84 1592.80 345.56 8.14 290.13 422.94 251.02 140.80 1172.40 4838.73 600.00 12.40 737.11 305.51 40.13 17.32 64.84 1855.90 332.58 0.00 387.58 322.10 22.03 25.77 147.14 92.60 125.01 0.58 370.24 139.41 9.02 32.96 146.19 546.43 95.04 4.04 40.70 28.04 1 0.27 0.00 4.58 0.56 0.00 2 13.01 62.00 2.08 5.03 7.00 3 166.53 829.67 1357.69 114.04 221.20 228.29 11 2260.33 8644.25 3464.10 3687.22 879.17 1420.07 12 640.86 1049.04 493.29 817.82 127.80 232.46 13 14 17 1.61 4.58 387.66 0.00 173.66 546.51 20.23 36.06 404.15 6.51 8.62 598.87 1.15 22.74 14.29 1.15 36.67 89.69 BT LAS-A LAS-S 5 4 ft* 2 6 ft* 6 6 *6 2 4 1 3 1 3 1 4 6 4 5 I ft* 3 ft* 5 1 ft 3 1 1 1 2 4 3 4 6 ft 5 ft 5 ft 3 I* 3 2 2 ft* 2 ft* 3 1 3 2 5 4 6 3 5 ft* 2 ft 4 ft* 2 ft* 3 ft 3 3 2 ft 2 4 5 5 4 5 5 3 6 ft 3 ft 4 H2 3 4 ftft 4 1 3 1.5 5.0 5.0 6.0 6.0 1.0 6.0 5.0 Rank! LAS-C LAS-0-1 62 63 65 53 32 53 53 54 63 62 46 65 64 36 56 25 63 62 64 64 65 51 65 64 65 61 6I 65 64 61 65 45 5 1 61 45 35 65 42 46 L4B-0-2 5 4 4 1 2 2 3 2 4 5 2 1 2 2 4 2 1 5 1 4 3 4 2 3 2 3 1 2 2 4 3 4.0 2.0 3.0 1.0 1.0 2.5 3.0 1.0 LAS-0-3 1.0 2 1 2 4 4 1 1 5 3 3 1 2 4 1 3 4 4 1 2 3 4 1 1 1 3 4 2 1 4 1 I 1.5 3.0 4.0 2.0 2.0 2.5 5.0 2.0 g e NP01J220 3-22 784031 TABLE 31 (eont'd.) Saa(i1 LA8-A LAB-0 C0KP0UN0 - TOTAL TCOO Standard OvviatIons LAS-C LAB-0-1 LAS-0-2 1 2.22 3.51 0.50 1.00 2 7.70 1.13 1.15 1.53 1.15 3 2.09 9.02 15.20 1.73 5.51 11 34.97 0.00 0.00 0.00 0.00 12 3,87 0.00 0.00 0.00 0.00 13 0.34 0.00 0.00 0.00 0.00 1* 0.47 0.00 0.00 0.00 0.00 17 0.24 0.00 0.00 0.00 0.00 COKPGUtt OCOO 1 2 .SI 0.00 00 2 2.14 0.00 00 3 4.G8 0.00 00 11 3.03 2 5 2 4.16 0 0 12 0.40 7.64 0.00 0 0 13 0.44 0.51 0.00 0 " 0 14 0.03 4.04 4.62 0 0 17 4.86 5.51 4.62 0 0 COMPOUND 2 .3 ,7 ,8-TCDO 1 2.22 3.51 2.31 0.58 1.00 2 7.70 1.15 1.15 1.53 1.15 3 2.09 9.02 15.28 1.73 5.51 11 34. 74 0.00 0.00 0.00 0.00 12 1.11 0.00 0.00 0.00 0.00 13 0.34 0.00 0.00 0.00 0.00 14 0.04 0.00 0.00 0.00 0.00 17 0.24 0.00 0.00 0.00 0.00 C0MP0UN0 2 .3 .7 . 0 * 2 .3 .4 ,0-TCOP 1 27.47 14.00 30.00 10.02 2 2.00 0.56 5.51 1.53 3 10.00 11.59 219.39 12.22 n 12.94 16.46 5.51 3.46 12 15.SO 1.73 5.77 3.06 13 30.35 4.50 41.63 25.01 M 66.46 25.1! 25.17 32.15 17 7.00 20.73 26.46 27.51 COMPOUND 1.2 .3. 7.0 1.2.3 4 .0-PNCDF 12.10 2.52 42.03 0.00 1.73 12.06 7.09 11.02 1 16.27 3.21 17.32 2.08 4.16 2 18.29 2.09 5.77 2.00 3.21 3 3.06 13.36 32.15 12.74 20.23 11 105.71 58.50 11.59 104.41 21.13 12 25.10 5.51 6.24 19.86 4.50 13 1.51 1.53 1.12 1.00 1.15 M 42.12 49.15 62.45 221.59 210.79 17 11.04 42.23 1.53 37.40 49.00 LAS-0-3 1.00 1.53 1.53 0.00 0.00 0.00 0.00 0.00 0 0 0 0 0 0 0 0 1.00 1.53 1.53 0.00 0.00 0.00 0.00 Q.OQ 3.51 0.50 25.42 1.15 5.51 5.77 4.00 2.00 1.00 1.53 12.17 22.01 3.00 0.00 56.04 14.47 BT LAS-A um -a 45 ft 6 3 ft 3 5 63 03 63 C3 3 5 5 5 5 ft ft ft 5 5 4 5 2.3 2.5 2.5 4.0 6.0 6.0 5.0 6.0 fftt 4 6 3 6 6 6 6 6 6 3 5 3 3 3 3 3 5 5 ft 1 5 6. 5 ft 6 2 4.0 1.0 2.0 6.0 1.5 1.0 3.0 6.0 ft 5 ft* 6 3 3 fftt 14 6 .4 63 56 12 25 Ranks LAB-C LAB-0-1 1 1.5 3 6.0 2 3.0 3 3.0 3 3.0 3 3.0 3 3.0 3 0 2.5 2.5 2.5 6.0 2.0 2.5 2.5 2.5 2.5 6.0 2.0 4.0 2.0 5.0 1 1.3 S 6.0 2 3.0 3 3.0 3 3.0 3 3.0 3 3.0 3 62 63 63 43 S3 64 45 45 62 52 63 13 45 32 46 14 LAS-0 -2 2.5 1.3 4.0 3.0 3.0 3.0 3.0 3.0 2.5 2.5 2.5 2.0 2.5 2.5 2.0 2.0 2.5 1.5 4.0 3.0 3.0 3.0 3.0 3.0 3.0 4.0 5.0 1.0 1.5 3.0 2.0 3.0 4 4 5 2 2 4 2 6 LAB-0-3 2.5 4.0 1.0 3.0 3.0 3.0 3.0 3.0 2.5 2.5 2.5 2.0 2.5 2.5 2.0 2.0 2.5 4.0 1.0 3.0 3.0 3.0 3.0 3.0 I 2 4 2 4 2 1 1 I 1 2 3 1 1 3 3 GENP O lm i 3-23 784032 T U L E 31 (eont'd.) Standard Oavlattons S u p la LAS-4 U l - f l LAB-C LAS0-1 LAS-O-2 COMPOUND 2.3.4.7.I-PMC0F 1 15.72 6.24 132.29 19.92 3.46 2 22.61 2.61 76.3 0.58 5.20 3 1.50 11.59 237.14 2.0 14.01 11 3.06 6.51 11.55 34.62 4.00 12 11.79 4.62 2.00 7.94 1.15 13 5.60 4.00 11.36 6.51 1.00 14 26.46 91.00 321.46 312.67 21.13 IT 11.93 22.61 20.62 44.66 9.64 COKPOUNO 1 .2 .3 .4 .7 .1 I 2.3.4,7.9-tOCDF I 30.12 1.53 95.4 3.4 5.11 2 1.71 0.56 0.00 0.00 3 39.45 33.72 152. 22.74 23.66 11 99.90 517.64 13519.3 394.71 10.19 12 11.59 35.80 120.5 67.98 3.06 13 3.12 4.93 10.1 1.53 0.5 14 23.02 46.74 1644.2 419.97 660.20 17 56.70 253.24 2294.2 145.23 11.53 COKPOUNO 1.2.3.7.8.9-HXCOF 1 0.27 2 0.24 3 2.05 558.64 11 1.36 0 647.08 12 6.11 0 9.45 13 0.09 0 6.93 14 8.14 31 3865.66 17 0.16 0 1024.30 COMPOUND I .2 . 3 .6 ,7.8-MXC0F 0.00 0.00 1.53 ' 28.62 2.52 1-15 173.14 2.89 0.00 0.00 1.73 11.68 2.08 1.53 289.83 1.53 1 6.17 0.00 2 1.95 1.53 3 13.32 12.06 11 3.61 100.72 12 1.18 4.36 13 0.46 2.06 14 17.32 29.69 17 1 9 4 2.31 COMPOUND 2 .3 .4 .6 .7 .S-HXCOF 2.08 3.06 15.72 22.28 3.00 0.00 63.66 4.51 2.89 1.53 27.30 64.49 3.51 0.56 110.39 18.61 I 10.02 2.52 3306.35 3.79 2.69 2 11.02 6361 2311.45 0.5 0.58 3 14.36 14.29 4659.61 3.79 39.58 11 5.67 10.36 909.45 30.05 8.08 12 3.05 3.08 11.72 5.00 MS 13 14 0.16 20.43 1.15 11.14 62.55 3461.99 0.00 105.72 1.15 227.21 17 7.27 13.45 1457.17 6.03 1.73 LAB-O-3 ' 3.21 2.0 6.08 3.0 2.06 1.53 33.05 9.29 9.54 0.00 35.23 3.93 25.67 2.08 148.33 13.32 0.00 0.00 0.58 6.81 0.58 1.00 15.72 4.51 3.46 4.58 6.00 4.04 8.08 0.51 110.53 6.03 2.31 4.93 6.43 1.00 2.01 2.52 19.73 6.00 0T LAB-A ffftttff*tt 4 5 3 ft* 1 6 4 * 2 3 tt 5 5 ffttft ft 5 3 2 4 ft* 1 ft* 3 4 4 ft 4 ft* 2 ft* 5 ftft ftft 2 1 2 5 3 ffttft 3 1 I 2 ft 11 ftft 5 5 ftfftt 4 2 3 ffttfftt 2 2 4 LAS-a 3 3 4 4 4 3 4 5 1 4 3 5 4 5 2 5 1 1 1 3 1 1 2 25 45 2 2 2 4 3 4 4 4 3 5 Ranks LAB-C LAS-O-1 65 6I 61 56 25 15 65 46 62 2 61 64 65 62 64 64 2 2 5 6 25 64 64 64 64 2 4 4 3 2 1 3 3 64 62 61 65 65 61 64 63 LAB-O-2 2 4 5a i i ! 2 3 2 2 I 1 1 5 1 2 2 3 4 3 55 3 3 51 4 3 3 4 5 3 I 5 3 1 ] 5 1 LAB-O-3 I 2 2 2 3 2 3 1 4 2 4 2 3 3 3 2 2 2 1 3 2 3 25 45 52] 54 4 1 3 2 1 2 51 2 GENP 011222 3-24 784033 TABLE 91 (cont'd.) Standard Deviations Saaple LAB-A LAB-0 LAB-C LAB-O-l LAB-O-2 CCWOUKO - 1.2,3 4 ,6 .7 .6-HPCOF 1 14.19 40.08 251.02 40.13 22.03 2 72.95 107.96 140.10 17.32 25.77 3 34.64 61.36 460.74 10.69 42.67 11 314.32 691.33 1708.60 672.92 29.02 12 60.62 116.97 200.33 101.35 37.00 13 0.11 3.06 3.33 0.00 0.51 n14 11.02 96.70 370.05 133.06 119.93 27.10 230.47 70.93 153.61 59.37 LAS-O-3 9.02 32.98 42.03 166.79 31.97 4.04 8.72 17.93 BT LAS-A * t* 2 4 *# 2 3 3 a* 3 2 2 LAB-6 4 5 3 3 3 4 6 6 Ranks LAB-C LAB-O-l 65 61 61 64 64 31 45 43 LAB-0 -2 3 2 4 1 2 2 6 3 LAB-O-3 1 3 3 2 1 6 1 1 B a rtle tt's tsst shows significant difference a t the .03 level, B a r tle tt's test shows sig n ifican t difference a t the .01 lev el. G N P 011223 3-25 784034 TABLE 32. LEVENE'S TEST OF DIFFERENCES AMONG LABS BY COMPOUND OVER SAMPLES Compound F-Value Degrees of Freedom Total TCDF Total PNCDF Total HXCDF Total HPCDF OCOF Total TCDD OCDD 2,3,7,8-TCDD 2,3,7,8 and 2,3,4,8-TCDF 1,2,3,7,8 and 1,2,3,4,8-PNCDF 2,3,4,7,8-PNCDF 1,2,3,4,7,8 and 1,2,3,4,7,9-HXCDF 1,2,3,7,8,9-HXCDF 1,2,3,6,7,8-HXCDF 2,3,4,6,7,8-HXCDF 1,2,3,4,6,7,8-HPCDF 8.89** 11.15** 23.12** 17.39** 6.97** 17.82** 12.81** 14.96** 6.60** 4.39** 9.76** 17.80** 16.10** 2.96* 39.93** 22.80** 5 5 5 5 5 5 5 5 5 5 5 5 5 4 5 5 * Significantly different at the .05 level. ** Significantly different at the .01 level. GENP 011224 3-26 784035 TABLE 33. APPROXIMATE 95 PERCENT PREDICITON INTERVALS BY COMPOUND AND SAMPLE3 Compound Total TCDF Total PNCDF Total HXCOF Total HPCDF OCDF Total TCDD OCDD 2,3,7,8-TCDD 2,3,7,8 + 1,2,3,4,8-PNCDF 1,2,3,7,8 + 1,2,3,4,8-PNCDF 2,3,4,7,8-PNCDF 1,2,3,4,7,8 + 1,2,3,4,7,9-HXCDF 1 ,2,3,7,8,9-HXCDF 1,2,3,6,7,8-HXCDF 2,3,4,6,7,8-HXCDF 1,2,3,4,6,7,8-HPCDF 1 144 222 206 431 155 1862 103 1105 2 89 6 60 1 59 15 30 144 221 123 168 99 248 182 347 <1 34 31 72 40 1041 103 1104 2 57 67 220 517 40 2118 137 1532 67 208 40 60 1 40 40 60 57 67 138 193 90 321 1 15 <1 25 45 72 24 1912 137 1530 3 1046 2321 1004 1682 798 10654 819 4341 2280 6436 63 132 2 32 63 132 977 1355 226 425 164 618 825 1838 2 965 254 412 55 3975 96 3258 Sample 11 12 57 37 362 321 1125 3079 143 590 3041 21535 726 1320 1581 41116 1467 3090 9641 3822 64743 10453 <1 <1 21 14 75 19 30 <1 25 24 61 41 7913 84 145 2138 12579 <1 20 15 81 27 264 28 59 447 809 1 8869 218 603 26 879 4 167 20 97 9 68 580 16405 418 976 13 14 389 ' 532 1247 1389 179 5846 442 13417 11 5897 198 21657 2 1391 52 3486 7 96 75 457 17 249 729 562 1999 610 13977 242 11063 442 24465 <1 <1 <1 15 20 10 <1 1 5 70 43 32 <1 <1 <1 13 32 11 90 251 119 459 459 241 8 742 30 1577 74 746 46 1324 80 2389 106 403 13 3144 486 36 5860 5023 <1 64 <1 61 7800 6136 2 561 10 1385 43 134 1 196 10 51 2882 2237 1 482 79 17 1130 5652 aValues are In ppb. 011225 3-27 784036 With about 95 percent confidence of being correct (1,e. 1 our of 20 measurements having a chance of being outside of this range), each of the reported Intervals 1s expected to cover a future laboratory's analytical results for the given compound and sample. This assumes that the concentrations are log normally distributed. Note that the data are not adequate for assessing this assumption; also the data for generating these Intervals are extremely limited and can be sensitive to spurious observations. Many of the Intervals are so wide that they are likely to be of little use. If the analytical method used by Lab-C was omitted, these intervals would Improve, but a quantltalve estimate of this improvement would be difficult to determine since the amount of data for statistical analysis would be Insufficient. 3-28 784037 SECTION 4 REFERENCES 1. A. W. Nichols, H. P. Kilpatrick, E. 0. Hardin and D. A. Hayes. Polychlorinated Dlbenzofurans (PCOF) and Polychlorinated 01benzo--D1ox1ns (PCDD) In Utility Transformers and Capacitors. Vol. 3, Palo Alto, CA: Electric Power Research Institute, February, 1987. EPRI EL/EA-4858,. 2. S. H. Gordon and H. Miller. Polychlorinated Dlbenzofurans (PCDF) and Polychlorinated D1benzo--D1ox1ns (PCDD) 1n Utility Transformers and Capacitors. Vol. 2. Palo Alto, CA: Electric Power Research Institute, December, 1986. EPRI EL/EA-4858. 3. D. R. Hllker, G. A. Eadon, K. M. Aldous, R. M. Smith, P. W. O'Keefe, H. Valente, S. Conner, and J. Jurislk. Analysis of Polychlorinated Dlbenzofurans and Polychlorinated D1benzo--01ox1ns 1n Transformers and Capacitors. Vol. 1. Palo Alto, CA: Electric Power Research Institute, October, 1987. EPRI EL/EA5443. 4. W. M. Cooke and F. L. DeRoos. Formation of PCDD and PCDF 1n Askarel and Contaminated Mineral Oil Equipment. Palo Alto, CA: Electric Power Research Institute, July, 1987. 5. T. 0, Rouse. Polychlorinated Dlbenzofurans (PCDF) and Polychlorinated Olbenzog-D1ox1ns (PCOD) In Utility Transformers and Capacitors. Vol. 1. Palo Alto, CA: Electric Power Research Institute, October, 1986. EPRI EL/EA-4858. 6. G. W. Snedecor and W. G. Cochran. Statistical Methods. The Iowa State University Press, 1980, p. 252-254. G E N P 0 227 4-1 784038 ii APPENDIX A EVALUATION OF CALIBRATION REGRESSION USED BY U. UMEA GENP 011228 784039 RESEARCH TRIANGLE INSTITUTE Center for Medcal.Environmental and Energy Statistics Scptotter 2, 1966 tOOWOH TO; Edo Felllrzari not: Andy Clayton SUBJECT: Problema with Calibration of Dioun/Furmn Data R m c t a a : Analys suite CTU-prodjefct 2028 (6/23/86) ft problems with the calibration procedure uaed by the Urea ttliversitet laboratory are noted. These are the following: 1. The regression oi x on y is performed, where x is the (injected) amount of dioxin/furan, and y is the (observed) normalized count. Although there has been rush controversy over the past 20-25 years concerning whether it is more appropriate to recess x on y or y on x, 3 believe that most statisticians agree that regressing y on x is the preferred approach. 2. Ordinary least squares (CCS) is used to fit the calibration curve (asssmnd to be a straight line). This`is a nueh more serious problem. CCS is appropriate when the magnitude of the measurement errors is approximately the came across the entire range of concentrations of interest. This is clearly not the situation here. As an example, consider the data for 2375-TOT. When x-7500, the y values range from 8593.7 to 9037.3, a difference of 443.6. On the other hand, when x-10, the range of observed counts is only 2.9. Clearly, the measurement error increases with increasing concentration, so that CCS is inappropriate. The severity of the problem is demonstrated by the estimates obtained for 2378-TOT. Using the values of a and b from the referenced assent, it is seen that a normalized count of 10 units will yield a predicted concentration of (.853)(10) 29.59 38.42 pg. Deamination of the calibration data indicates that the true concentration should be roughly 10 pg. Tor 2c*r concentrations (say, less than 200 pg), OLS has thus produced a curve that (in this case) results in a significant overestimation of the true concentration. It is precisely in this range where the transformer data occur. One solution to the above problems is to perform the calibration ourselves and to apply the new calibration to normalized counts obtained from the reported concentrations and the reported curve parameter estimates. Sore imprecision will occur due to the fact that the concentrations are reported to only two significant digits. Alternatively, we can request the raw (normalized count) data, perform, an appropriate calibration, and re-estimate the concentrations for the transforms. fbsiOfliceBos12194 ResearchTrianglePark,NorthCarolina27709 Telephone 9195-1-6??? GENP 011229 A-2 784040 APPENDIX B DATA USED IN COMPARISONS BETWEEN LABS GENP 011230 784041 Saapla Lab 1 NYSDOX 2 NVSOQH 3 NYSOOH TABLE a-1 DATA USED IN COMPARISONS BETWEEN LABS Coapound TOTAL TCOP TOTAL PNCDF . TOTAL HXCflP TOTAL HPCDF ocor total tcoo ocoo 2.3.7,3-TCOO 2.3.7, 2.3.4.8-TCOF 1 .2.s.7,a i.a.3.4.a-Mcar 2.3.4.T.B-PKCOF 1.2.3.4.7.a 1.2.3.4.7,a-8XC0F 1.2.3.T.I.B-HXC0P 1.2.3.6.T.a-HXCOP 2,3.4,fl.T.a-HXCOF 1.2.3.4.8.T.-HPCTF TOTAL TCOP TOTAL PNCSP TOTAL HXCflP TOTAL HPCDF ocor TOTAL TCSS OCOO 2.3,7,8-TCDD 2.3,7.S 2.3.4.8-TCDP 1,2.3.7.a 1.2.3.4.a-PNC0F 2,3.4,7,B-PNCOP 1.2,3.4,7.B * 1.2,3.4,7,9-HXCOP 1.2,3,7.a.9-KXCOF 1.2.3,6,7,0-HXCDF 2.3.4.S.7.8-HXCDP 1.2.3.4,S.7,a-HPCDP TOTAL TCOP TOTAL PNCOF TOTAL HXCOP TOTAL HPCOP OCOP TOTAL TCOO OCOO 2.3.7.8-TCOO I . 3,7.0 > 2,3.4,8-TCOP 1.2.3,7.a * 1.2.3.4,B-PNCOF 2.3,4,7,0-PHCOP 1.2.3.4,7.B * 1,2,3,4,7,9-HXCOP 1,2,3.7,t ,9-HXCOP 1.2.3.S.7.a-HXC0F 2.3.4.e.7.a-HXCOP 1.2.3.4,a,7,-KPCDF Oat. 1 143.00 231.00 390.00 334.00 3.40 17.00 3.09 17.00 141.00 121.00 127.DO 223.00 0.33 42.10 130.00 331.00 oo. ao 276.00 201.00 378.00 133.00 39. CD 1.44 39.60 80.ao 138.00 122.00 2.77 0.21 30.30 140.00 373.00 1330.00 1330.00 2490.00 2470.00 4830.00 63.10 3.73 83.10 1210.00 346.00 304.00 122B.00 37.80 339.00 313.00 083.00 Oat. 2 103.00 312.00 433.00 341.00 3.33 21.30 0.49 21.30 193.00 133.00 138.00 239.00 0.07 47.40 131.00 341.00 82.70 338.00 209.00 712.00 107.00 30.40 0.41 30.40 80.30 171.00 139.00 3.83 0.03 34.30 130.00 709.00 1820.00 1290.00 2380.00 2300.00 3090.00 63.30 0.72 65.30 1220.00 342.00 238.00 1209.00 33.80 321.00 288.00 003.00 Oat. 3 194.00 263.00 330.00 372.00 3.06 18.20 0.87 18.20 192.00 132.00 138.00 199.00 0.03 33.10 170.00 300.00 83.30 324.00 223.00 391.00 121.00 34.30 4.32 34.30 83.30 189.00 183.00 8.12 0.32 30.30 162.00 392.00 1390.00 1340.00 2530.00 2400.00 3170.00 SI .70 9.89 81,70 1200.00 348.00 291.00 1283.00 39.90 347.00 291.00 663.00 Naan Std. Dev. 178.00 282.00 400.00 336.00 3.37 18.80 2.03 ia. ao 17a.00 133.00 141.00 238.00 0.22 41.30 180.00 334.00 26.30 30.30 32.60 13.80 0.27 2.22 2.61 2.22 27.30 18.30 13.70 30.10 0.27 6.17 10.00 14.20 63.00 312.00 212.00 627.00 120.00 46.20 2.12 48.20 82.30 139.00 148.00 4.24 0.28 36.50 131.00 623.00 2.36 31.70 12.20 73.90 13.00 7.70 2.14 7.70 2.60 16.30 22.60 1.71 0.24 1.93 11.00 73.00 1813.00 1320.00 2473.00 2390.00 5037.00 84.10 4.78 64.10 1210.00 343.00 294.00 1241.00 37.80 336.00 297.00 845.00 20.60 28.30 60.20 83.40 187.00 2.09 4.68 2.09 10.00 3.06 8.30 39 40 2.03 13.30 14.40 34.60 oU ^ B-2 784042 TABLE B-l (cent'd.) 784043 oo oo oo oo oo oo no o < i r> i n ^ncn^i ii*l -mn ^r-n+mn<^*+n o ^ i od) ^^ ItoO o o n 90 9 - ^ o o o v ^ n n o o o o S o ooo ooo ooo foni o phi *o*oo ooooonwooor -oo onvotrih fici rivi nfiofhfiMfO Nonh ofn f i t* <* AH nh ooooanoNoooo o o oo a ^O M iN N o tta a n i o q n o i r n o 1 PooQo0oo9 tQ*VO9oOrO-OrO* oQOQOOooOO)Oa oo aooooooov to *AooQa aqfoOoH io noroionoa oo oo ooo^ ooBnMaQnOoNo^ oN oM Av if o^ an nh vo no o oon viir*c i roadnornii NeO<Q5 hAqAnf ii pm* on hr>ni nhhi iiOi(aoooontt*>-pt-i-qonHfl mp n a f ^ S8SSS3SSSS88S=83S H Q o O Q N iO N flit ifM lO HO h n O A . A h * * I fi f> OQQOOOOMOOOOOOOOO O O O O O V O v iv iO O O f-fiO n O N A n r)h On A A O f A Opo o Oo i M i i i h M AH) o c f l a n Ci f ir) tAf> oo oo o A On An QA AnOaOnOn oO tnt nO oO eO oO nM v> n>r t A O O O C h r i H d f N f 88888888888883833 Sh onA o Qf ioQ hi*O h f aanpota-n SAq h)qp3 Q Q O O Q iQ O iO O Q O O O Q O C IO O O O O O nvO AO AO AAO AO oonoootvooooviC^rtfiorf 4Nn an a* vo ao n v*i N A A f H Oa Oo oo ro- i i A i n o n o r a i H C o n n o o noo q h p< m J p E u s sUB! S n\ m' S 6 8 xg g,, A -( i t O? ?i?i h. sfeefee f* H h H O h O 0 0 v i v i -* v i f i v i ^g OO A A A it ANAnU n~8*S O Xi on *A A 888 S h o o i-o fcK A U -*. - h X X h DO hhf h 4 3 3 a. 3 a *: n - *. g 6 g s i^ " s : : - 88 Si m fv M h0 00f d in r i n r* M rt eK A* AI e :* v N k " S6Bx sees gssi s 8a "o - * `x8. "i-5l .iS-*n I-- a o i- e i* I- i out-r-vt-eo* < d ^ . a h M i o i n n - n EEB68682S:-:- nmi (N co <N w TABLE B-l (cant'd.) G E N P 011233 o--oio5tr*h^n" v--o p^ oi )-oo--vf i r> ^ nonO O O O O O O O rnoto^on)dt^NnN0NiOnon^OnOr)DootNa 0000*0n000000000 0000* 000000000o S' o o a c o H D i v o a * t 0 9 9 Mn o0m o0r*^nO 0 O OPnA^rtfO^pt i* O*Pdd O a o Hootfnvooo^oooo oooonocoooooooooo SPOOOO"*OppOOOOOO O OO OOa OOOOOOOOO 0 0 0 6 0O o o o *n o h o oo noto*qn o* o (oiDq n oh n n n F iO n r* MOON M H^nOvaOOOHnCOhbO fin V H H a H O d r t f h N H O O n no n n o o o n o o o o n o o o o nnn4QOOO(1*nOOOOn nO hM O nnannnO nnnN -o o o o n o r - o o o o o o o o o S ooqoftqpqQQQoooo O O O O VO O O O O O O O O O onoort O^HO V qqqoqqqo*<omnqq fta n M n O llO N fln q irtn n O qonnoooNOOvoofi n q q n ^ H ttn n ^ O h H n M n QO OO OO QO oO qO jO SO QO OO QO QO QO OO OO QO ssIS0" S l s ; i i i dn nMnhHivoOaOf Oi tN dOlhqnninMNdNHNa HMOOonoooo^oooo nn0^r0ot0nq0no0no oo0oq0co0nq0ov0ir 0l- ^^0nn0rr0fqo) n MfctihS see: EES! ri h ri 8 8 6 1 >* .S S S i m E a i>* t-a V ^ i < a ^ a S q ^ n. rt. ' I g E l i lC b f c t R b H 0 **************** 8A (" a( k wI m1 n ri n c. c. Ga see: 8S! V ri k>ri 8 8 8 X n -is S 5 5 *-a k 1 1 a a. a a k> a' ^ ^ ^ ( t ^ 0 i'<i - > a n f l r i a r i 1 1 b t Rt P Nn n N g68^ eP "a. ot bn (ic.c c,U a'atriSSSs n-8 Sili _ Har1***fCfct l.'-- o-Ia-VaI -*rI -> h S x 8 o r- i- a i- s a j ^ j J j r> t> n a n n n a bg gb gbgbgngb!h n* <i ^ a i ` a i i n ^or 0h0- -*1d- DO TABLE B -l (c o n f d .) GENF 011234 OD I en -040 O -CtJ1. rM rU Nr Mr Mr"MrCrIrM-PRHaDgma HaHaHa ..........u.....u.....d.....^....u.....g.....-o4 > r^* p r^ r ^t* H X X tlH .o.....-.j...-..j....o....-.j.....o...o *a gd *n5xn zd n ....................a n O odo^ tn o *xfInxMonxnoX*xdX***idXnft-a**u O o ^ ^ ^ u9 a * *t * 99 3x u H a9 <4 nU X o s ** r r*r r r r r r r s 3 g g 3 33 .u...A....u...u....u...k...u...M.....'..aO > n'< >r >r* r* p *0O-a*a<aAa H *9 ^ 4 - 4 9 - 4 9 9 * H O I --- n s i io ^ oaa t z) mo d X X s n Mu i XXX a - - O o x ndna nc ^ ^ M-^ M um o r *rr r r*r - 8 3 83 3 g 3 .u...*...u...u...t..f..*...u...-..4.. -- 4O 9 r-"* 9 r9 r> r9 r 4 f t (3 4 * 4 - 4 f t 1-1 B J t H :-***:-:-'t - * 3 S a l i NO -? St 5i gl * aXr --" D g a s a r 'p* ;5>'*N W - i: : ; atn gs 3 ___ UOH O OO o oo ___ _ _________ ____ h tt* *M U U U 4U *^ H 00000*4000000 oaoooooooooo _ M ft M O OO Ooo M *- w M Uft N M ftft(*ligw iN *M )4 * *N ^ ftftO ftftN ^ U Ik oooooooooooo oooooooooooo -* 9 m 9 U& * -4 9 o -a kiPMM 9 4 4 M (I U O 4 O M O O t l I-O M M o O O U O D 0 A O O O Q M 4** O O O O C 000000009 OoOOooG tM O9 O-JH9 OOO ooO ftA A U U U A U h A MftAMUAUAAMOW 00000*4000000 OOOOOOOOOOOO 009 O ft*X ooo ooo MM** 44 M M (J4ftftMMMftQ0)ft 0000404001*00 oooooooooooo oooooooooooo -4 O M m m ftM w H * u <4m a 9 U 9 U A 4 A kl aftuU HftftQMM4ftO 9 O O O O O M Q P M t t Q Q Q O t ft*9 O O O O C 0009000014 O M O O O O G 099 9 -4* OOO o oa M ** 4M u ftA A U ft4 O klm ft MAAuftftuyftkiftM 0DOOONOOOOOD oooooooooooo UM 09 9 -49 AJ OOO o oo W M M 9 9 UM 0ftft0MMjft4A-|ft 0ftA4MQHft4404 OOOOOOOOOOOO oooooooooooo ftM M M 4 4 O ttM (1 9 9 9 4 0 Q M U M klO ^ ftftMi*U4klO 9 O 9 O O O ft 004U000044400* oooapoooo*o*oo< 9 UImUV 1dMN0m0N WNma MLA 9 9 9 M ^ m w *O 44 Mf t 4f t QN On U*M p 4 O MMM O O D*aO O O I U 0 4 H ft ft A m Om i f t P O M I O M O 9 ftV A fta u o o v O *9 0 O MO *" 4 U O U O 4 0 9 A U fta M ftM U 4 h>o ^ O h> O o ^ I D 4 4 ^ O A ^ h 40409ft0ftU 784046 CID cn G 1 O tLop <-TI riTr.-.-r.-.-ggagasa U- U- M- M- M- -U -M -U -U gO ^> D*Hb HfcH> db U*U UU *U 44 r> r r r r A n ^ -i 4 a <40 A fiQD * N III 7 i mss Szs eXm r 'no *"*- **c* n ^ ^i M n Don3 o d o * z :a r.-rrr." r~" S383333 M U *M M .H....U....N....U...u- aH - O . H. ^. H. .H . u w u u u - 4 - 4 r r~f r* r* 4 *(a_^ >o**( * i h nO -Oo n n d Cl O O D 4 fi fi fi 'O * N O TSI SI IS * va r- yo SK I D D ' <4 M t> d -I 3* H a i r f r r r f r . " ." 83 83333 MOUbU i<NbM bU bH bU U D i ( d > f H H > I ukt>uu*w*-4 F FFFF SS $83333S3 I fifi I Ml III 7*1 T5SS *arr SBSar^yy n u U H fib fiO fi ou o Po o ooo fi * my o m 4 oo oooo oo oooo A-4 *f*i B f i M 4MBOA O O OO o o oo o o o M* M~ oop OOP 14 i- *- a o n 4 u o ) o o o oa o o Oo o Oo 4 4Of fi bi - n Mfi X m U OO QO O OOOo O uua 4 fi M o oo o o5 fi* cm uhm J b* M b* M m a y n ti 4 fiU fi4 fi4 fififi*0 8S888SSSSSSS i f i h 01 U fi 4 uM * A MU U O U fi O M N -ifi g M Oo O U M M fi M fi fi M a* fi 4 fi fiM fi ' U 4 fin ,u *a ** M* b- o fiDMMM*oO* C..i H 4 OOOM DO fi fif<0i Mm Wa S D a *fi M fi 4 O0O Mfifoifuiffii fi ffii NfiU*4 ffii WfifoifMifi fiffii ffiiMfi fOiffii fi U fi D 4 fi fi n N N B J fi< 4 fi4 fiO M U u ma O (ffii mW o -- 4 W n aa a g U fi4 4 o 4 m fi n fi 4 o A UUU o o fi 4 fifi ** o B)U b> i u a u y fi i O fi O U 4 fi 4I Cl * O O Ci fi fi ^ o 4 0o)gfihfig4 o * a a a aa oo oo *o oo ia o o ia a o m Oo f*ifoi Uo oO m fifiy oo oO oO I M F* b- * fil a b* OOO f i b> H m * a* fiM fi u 0 0 < a a a o o o o o M fii OOOOMOmOOOOi Std. 0v. Matti D*t. 3 0ml. 2 Dtt. 1 Coipound Ub Su p i 8t'CL L U d M d t ) 9 Z i- l u iw . ' * * - *CIvDl 0-o*f>P0cJ*k M U M N ~r "" rrr**.""gS 3? S5333?3232 "` - " hr hr ^v o o a T MD III z -o ? S 3 * S-" " 3 3 8 3 r 'V * .* O *D n *n m ga UH * aS -tj x z o n n no DODD rr .wr*rr rr r .-.r-..*-"" pftuuu*u^4 g8 2385 r1 s3 r 3s r r32sr3fa: 4 *1 9 -4A - * ** i n _fS_i_SaSPMn_M3v*_*XD8oMr cid n -u O- - "" t U9 36S *DH 0*3HS 3 3 OD D m o rr r" r r rrr.Mr r r*g 23 q8a3 323333 u * u u u * u 4 4 t- h r r r D -4<4 H P 4 HT p p d p d m 3 S 5rdtdno ozo nd* -T SI Sl Bl * xS r * O ssssr'-rr * 0 :3 I 3 3 TABLE B-l (cant'd.) M o M -- U M L) o O Ot o o o o o o o oooo oo p< 4(4*M M (JI^ O P oobooooouo oooooooooo h O0 aUutmtl att ltMudVo<4 ooooo ooooo oo oo oo A U U tf o oo oo oo oo oo oo *M M*4*Ui*P4 U<4 o oo pc o oo 5c 0 P -* mn u4 *U*|**) N P HM OV OP P PI p M ^ 4 O P ^ O H U CP S u u ^ L k i g a o u a a 4 >4 U -4 P O M o o m M U U u u oo oo oo oo oo oo oo oo oo oo oo oo oo S a g S ou o p OOO o -* O M ** Oo oO op ooooou oooo i-- U V ^ ppM N-4< 40M p O U U -I M M ' P D *4 U P P D oP Oo Oo Oo oD Oo oO oD oO oU Oo um a a u 4UO O 4 *p pMP oO oO oO oD oO n Ot DUO Oo oO oP m IIPM O 4 4 O P O 0 0 0 _0 _4 *_0 _ oooo P U PU P M a U M P* oo oo oo oo oo o ---- M"4 o A M P U p POP 0 0 0 0 0 0 0 0 O0oP*O4*4 UP n PaO Pup4pMMm-4 oO oO oO oO oO p u -i O<4O4IO<4 sssSo. ao4ouu oooooo M CP MP MO aU oU NQ Oo o4o0o0o0 ODOM oooo oa om 404ONM00 O OOOO OOOOOOOP ooooo fd*M p p M p o O omo iN4M<4Aa Sampie Lab Ccapound Hat. 1 0et. 2 Det. 3 Man Std, S Oo oO Oo nW mf t * t i 000001*000 OOhrtnnniD ooo*o*> v*o* i n Mh oo oOO NO OOO e a d r) r* h> *a a n ^ MN A Ho o oo o V)O I* N (0 o o o a nn n n * n i -- -o o *io o nno -- a N oooo aoo o o*)*)** o rt O A * A *noo Ao oo o Oo Oo oo hO oO AA OOOOOQOOOQ noor-noonr-o *4n v i n rt v o Hi n > 40 an h n o o o OO o o o oo o O fiN H nn Ah Qn oi-Nn o a o o oo n a o o oo ei o h n a o - *n Na "a Oa OI* O r h oa oo oo oo oo oo oo oo oo oo oo oo oo D*rt0f0t0h OA' OOOI Oo Oo Oo Oo Oo A O A Q -4 h - A O 0O*O1 O0 0O0O0O n r>a o H AAfl N a A 0O0O0O *40* 1 0 HA h r- oOaOoOoO 0000 OOOf* h n r) N --oo O-o--O--o--o-o o*o *oo oo OI gggg h rtti h ea nn gggggggggg AAAAAQAArtA r* -no h * 0gg000 00 00 A 1- Q A V *Oh * ON A NNtt ooo ooo 000000 rin O N A A A rAt nA h HP| 000 *10 0 0 * o 4h ** M 000 0000 O o o VA ooon nA o 00000 oooao oo o o o - A h t- A wm *4n n o o O A r A) SSS8 SSSSSS88SS wm * C* N pOO OO o oo oo A ci n A o 1o o o o o e O oooo ooo oo o *4 H r- A oo oo o oo o N rtA n o oooo ooooo oo oo gees bliJiJJ ^^^ sr * ha r a 8*: T H r> a ki 8 An * fc> A* A. 8 a r 1 a a a 8 Vnah GB82 rnarGSGs a rb.^ 5 u x s PB o--8 . N. -*i,I BIgt S B S 5 UO. ao - a .e**i 1H * A t i h A h h C a o* om >pa EHIS o g i- B x S a X li ga* a A OI A h H 9 O r - r - O *4 hh-AAlArtAA alJ 4 4 al h f ' r t A n o A A O aiIal-I J hhfl ArtA*i f)n n n n nnr) n ^ ^ ^ ^ o ^ o o n M i n r A t A f A o i AA Q * .......................................... * < < < U, < O .........................................................A 4 A ....................................................................... tr en <S 0 1 0t0aI 784048 1709.00 pn4f0< oo oo f) M in * m^ hn ht- ov * r t t O M ** n m h n w m io ooooo pto o o oo o ^ t- h q o a o o > o o o *4 *4^ O m v 9i n o t- rf> m n n h n N4 V * N w o__ h ch**Or<*0Oa0hr-0t*0* 00 oooooooo 1*0000000 h h t*f*- r- r-or-o ooooo *4 h w o o at* or>oo oooaoo n wo o a o r* n o N o n n ^ o i- 00.00 2900.00 io d o a o o w o o d "W '0 Q O w o o owa*w i*n h rt AN O C O oaoo a owoooo S S 8881 Oo oO oQ oO Oo oO oQ oO oQ d d d d 'd h o o OOOO <f O A N s: op o_ _ _ o O __ q oo o o mm b n r* o A ^ OOOOOOOO o o o -- oor-w hnn a t- ooooo oooo h * o a o N m OOOOOOO n n O (1 o *4HNS ss od o o o o o 8 0o o0o0o0*0 <pn a r*o h ^ h o H f- oo oooo oo oo oo oo OOOO oooo O O OO 0Q O O o ooooooo * 4O-0 0O^0O1Oo0 Oo0 o0O0I *i n A N Q oooo t* hn o0 o w *4000* 01M **4 8 *s a E* i a m f) N *^ u n h* 8 8 x a d*- 8 k 8j 5 4* MiA A p N L A h> 5 8 8 8 S H t-fl^ o' BSfi K ^HgHg4HgdHpgrHSQiiCiiwn-f*tr>fOrcpricit 5858 ESSi ip W d * a O N MlfeU aau v Nd* n*44t*SLhQ*-5A8-Q8ii-xhw *<0r>Ch0 HhKhahannNnNNnN B gm. o p a uxHuxO.uaz. x hl Ml he Me as hl 8h. nT VN||h 8 8 i Cl**4<OJ * Xx xKI a x i * a n l o a dh . Ad.Vd wot*t^r-w* hhnnn*n PnnnNN nN TABLE B*1 (cant'd.) 9 oe o Ch-O Q b CT I GO Staple Lab 12 BATTELLE 13 BATTEILE 14 8ATTELLE TABLE B-l (cant'd.} Coepound TOTAL TCDF ' TOTAL PNCDF TOTAL HXCOF TOTAL HPCOF OCOP TOTAL TCOO OCOD 2,3.7.8-TCDD 2 ,3 .7 .8 2 , 3 , 4 ,9-TCDF 1 ,2 .3 .7 ,2 1.2.3,4,8-PMCOF 2 .3 .4 ,7 ,9 -P N C O P 1 .2 ,3 ,4 .7 .8 1.2.3.4.7.9-fOCCDF 1,2.3 .7 .1 .9 -H X C D F 2.3,4.8.7,8-KXCO F 1.2.3.4.6,7,8-H PCDF TOTAL TCOF TOTAL PNCOF TOTAL HXCOF TOTAL HPCOF OCOF TOTAL TCOD OCOD 2.3.7.8-TCDD 2 .3 .7 ,8 * 2.3,4.8-TCOF 1 , 2 ,3 .7 ,8 1 . 2 . 3 , 4 ,8-PHCOF 2 , 3 ,4 ,7,6-PNCDF 1 . 2 .3 ,4 .7 .8 * 1 , 2 , 3 . 4 , 7 ,9-HXCOP 1 . 2 . 3 , 7 , 8 . 9-HXCDF 2 .3 ,4 . B. 7, B-HXCDF 1 . 2 . 3 , 4 . 8 . 7 ,S-HPCOF TOTAL TCOF TOTAL PNCDF TOTAL HXCOF TOTAL HPCOF OCOF TOTAL TCDO OCOD 2,3,7.8-TCOO 2 .3 ,7 .S * 2,3.4.8-TCDF 1 .2 ,3 .7 .8 1 , 2 , 3 , 4 . S-PNCOF 2 , 3 . 4 , 7 ,8-PNCDF 1 , 2 ,3 .4 .7 .8 * 1 . 2 . 3 , 4 . 7 ,9-HXCOP 1 .2 ,3 ,7 . B, 9 -HXCOF I . 3 , 4 . 6 . 7 . B-HXCDF 1 . 2 . 3 . 4 . 6 . 7 ,8-HPCDF Dat. 1 82.00 79.00 1300.00 1300.00 4000.00 Dat. 2 84.00 230.00 1300.00 2100.00 4800.00 Det. 3 99.00 310.00 990.00 3700.00 4900.00 Mean 73.00 208.00 1283.00 2100.00 4887.00 S td . Dev. 20.80 117.00 237.00 300.00 493.00 20.00 22.00 31,00 810.00 34.00 47.00 330.00 330.00 230.00 140.00 28.00 48.00 20.00 33.00 41.00 740.00 30.00 23.00 330.00 370.00 270.00 130.00 2B.00 44.00 30.00 34.00 37.00 840.00 38.00 43.00 730.00 330.00 220.00 180.00 3.80 10.00 23.30 27.00 39.00 733.00 48.70 38.30 343.00 500.00 247.00 143.00 19.90 33.30 3.77 3.24 2.00 121.00 9.43 11.70 200.00 130.00 23.20 13.30 12.40 20.20 300.00 10.00 , 71.00 37.00 13.00 23.00 7.10 280.00 3.30 74.00 18.00 13.00 17.00 8.30 220.00 7.90 33.00 31.00 27.00 39.00 1.80 970.00 11000.00 14000.00 3100.00 130.00 380.00 8700.00 38000.00 2800.00 170.00 810.00 8300.00 '16000.00 4200.00 220.00 18.00 400.00 870.00 2300.00 4700.00 1700.00 820.00 920.00 380.00 900.00 2400.00 7200.00 8800.00 8800.00 360.00 330.00 780.00 1800.00 4100.00 2600.00 930.00 1300.00 267.00 8.73 68.00 28.00 19.00 27.00 5 . BO 720.00 6067.00 22000.00 3367.00 180.00 18.00 377.00 830.00 2167.00 3333.00 4367.00 2783.00 927.00 41.60 1.12 11.40 10.80 6.93 11.10 3.53 217.00 2542.00 12166.00 737.00 38.10 23.20 62.40 32100 1644.00 3666.00 3462.00 370.00 ^84050 GENP011239 B-10 5tap! a 19 Lab BATTZLLE 18 9ATTELU 17 BATTELLE TABLE B-l (cant'd.) Coapound TOTAL TCOr TOTAL MCOF TOTAL HXCDF TOTAL HPCOF OCDF TOTAL TCOD OCDD 2.3.7.9-TCDO 2,3.7. 2.3,4,9-TCDF 1.2.3.7. 1.2.3.4.8-PnCOF 2.3.4,7,8-FNCOF 1.2.3.4.7. 1.2.3.4.T.9-KXCOF 1.2,3.7,1,9-HXCOF 2.3.4,8,7,1-HXCOF 1.2,3.4,4,7.1-HFCOF TOTAL TCOF TOTAL PKCDF TOTAL HXCOF TOTAL HPCOF OCDF TOTAL 7COO OCDO 2.3.7, 6-TCOD 2.3,7.8 2.3.4,8-TCOF 1.2.3.7,9 1.2.3.4,8-PNCDF 2.3.4,7.8-PNCDF 1,2,3.4.7.8 1.2.3.4,7,9-HXCDF 1.2.3,7,8.9-HXCOF 2.3.4,6.7.8-HXCOF 1.2.3.4,6.7.B-HPCOF TOTAL TCOF TOTAL PNCOF TOTAL HXCOF TOTAL HPCOF OCOF TOTAL TCOO OCOO 2.3,7.8-TCDO 2.3.7.8 2,3.4,8-TCOF I.2.3C7.B 1,2,3,4,8-PNCDF 2,3.4,7,8-PNCOF 1,2.3.4,7.8 1,2.3,4,7,9-MXCDF 1.2,3,7,8,9-HXCDF 2,3.4,8,7,8-HXCOF 1,2,3.4.6,7,8-HPCOF Oat. 1 Dat. 2 Oat. a Maan S td . Dav. 12.00 12.00 . 13.00 11.00 12.00 1.41 380.00 870.00 8100.00 1600.00 4300.00 la.oo 190.00 89.00 300.00 2700.00 780.00 1100.00 180.00 230.00 390.00 19000.00 1400.00 4800.00 18.00 140.00 72.00 280.00 8400.00 2300.00 3400.00 230.00 290.00 970.00 20000.00 2000.00 9100.00 10.00 180.00 71.00 290.00 8900.00 2700.00 3800.00 320.00 293.00 397.00 13700.00 1687.00 4733.00 13.30 170.00 70.70 283.00 8333.00 1920.00 27B7.00 243.00 65.10 64.30 6601.00 306.00 404.00 4.62 26.30 1.33 20.80 2294.00 1024.00 1437.00 70.90 784051 gbh? G E N P 011241 r--io mrO0Co3*. UM i i 51X**41 r.N....U."...fr...l.r..M..r..M.!. U..*..k..r.i..f..U..f....g.O..a.H...g.O..g.H...a.H...gM...g.H........rN....r...M.r...M.r...M.r..U..f..h..r..l..-.U...r.W...O. 3H8-Dt]3H* 3H> 3H 3H u * u u t * c i - 4 - 4 r r r r r* b i * u u u * u * ) r* r P r r * *.*** I Q M A *4* * -V ><ai tai a * 3O' ?5S3 *Srr H 3 3533 8no of i *o* III ?5S5 ia a : o U TCDD SS33 833? 5388r P n t i t*Ut .u u. i i a a i :a* T* 3 H r* o ** mjMllmMMMMhUNlmilNtMl * u ia *u m *ao4A44G - i ri IBP y* N O * i a o ? 5S S *8r r 5 3 8 B r <?.* Ua 1 U *H 31333 1r* r H S3 a 138 r fS 4U U ^ M U H * - O t i O (1 0 n m u * o o * o ^ o m c * oo M *l(JH M uOH O O OO M a(aU oooooooooooooooo oooooooooooooooo 4** au ^ ** M yn- uaui^cjitjaoaoa 0 (J^UOIM(IMm O M I O * oooooooopooooeog oooooooooooooooo ^V MMH M *M u -4( a u ^ v u M M ^ u c a Q opoopoooopupop ooooooooooooooo 4 M M M U M > M D * t U A A M v u y a y y o N O ^ N io u u ciA * * n M u < 4 o o o o o y o p OQOOOOOOOOOOOOO oooooooooooooooo 01 ^ N - M U WU u y *yy*uM yyoa oM gu a* 0 * tto aoC U * O O O O Q O O O O O O O aD oooooooooooooooo 0M U4 y i P* a v* M M U M po io fivcu ia i# u o vQ ttU pppOOaQOOOOOQO ooooooooooooooo 4 U N M M M uu* 0MU>)>4MOI0wyO9ONU M u a u 0iM y*4O ^ HU9M M oooooooooooooooo OOOOOOOOOOOOOOOO yU NH M (J> M u UUtf *Uat**UMuaoa 0uu *U 4H O M (oeiH (i oooooooooooooooo OOOOOOOOOOOOOOOO UM U * (I fi -IV QttA^NvNHftO o o u u y u < itto y y N u ooooooooooooooo ooooooooooooooo yy u M * * h My M 0y M * M * M h * 4 fi cN o 4uyaNa ^^ oy yo *wy iooi to ^y Ma uB N^ Oa OOOOOOOOOOOOOOOO OOOOOOOOOOOOOOOO a* *-* I- m Umu o* uu> u4 a ^ - 4u Na Ou Mo ou ^yi D tOpMo ya 00-JOOOO*-a-40*jOOOQ-4 oooooooooooooooo *m mmh m * y ^N -N yH i yi y* ** a a * om *y am M aa o OG '4QOOPQIQlOOOO O O OOOOOOOOOOOO M M M MM -4fc* N M ^a a <o oyy^ yyyyM OM ^ AM ya i-4iy-4o-4y*io-40aoo^ ooouoaoououaoooo M N* i o u o o o m * - * ^ o m m - i u m m u m o o o y a y iio y D u M y u aaaooaouuoyaoM N u * a* ** u o u m o u g m o o o o * > o a <- M y o o A D o o y o u u M ip ovaoaoaoaoyaoao 10.00 IT S.00 17T.00 141.00 104.00 TOTAL TCDF M0IAM EXT] GEN? 011242 cIo eO3nU co i g s 5 U M M M U M U U-|__|_|_ |^ |, |_ li k u u * U p ' P r P i ftft4 * 4 4 ftft S4ftSf4itSffattf4*tsf^atirft.u*M4O H90 03df3to3^aco5-fft8tH aft*8ftS*ftrm * *U - <f*ui0*3f.4a1*t : ^ : ": ^ p s 3 | 3 3 U *b )C *U ^ -l *O ft4ft44ftft 0 * 5 2 f t 4 4 f t 4 f t f t ft 4 - Ill X ft 0 0 oftbftfdtx4 9 f t O f t T M n nn QS?fSStfS5t5fOtrM*'8Vr1"4..Cw*M ft Uft ft43 r-::::" :-:n33333 04 -r s04as0i4 as00 *04 3Xa0 r0a.M"ft O5f8tf8tf3t.M- U ft ft *Uf.t f01t84ft 3 ss ;S S S 8 0 9U U U U M O U 4 U O U U U O UM VM UMI 44n 0 O M *-* MM o eQ u) N------- U- - O- -( i -CI 9 - 4 M I 0o 0o 0 0o 0o 0o 0o 0 o0 o0 o0 o0 o0 S9 0o O0 0O 0 0 0 0O 0O 0O 0O 0O 0O 0O 0O 0O 0O 0O *MM MO li M4 ff4tt UM mMO U M OMIMnM4* M O ft M4ftQft 4ft (i OUONUA4 fO iOoOoOooOooOo eO oO oQ QQ oQ oP Qo Qo M M ft h 4 U U U U U O ftft4 U O U U <0 m NM4 4 U U O ft 4N N *4 4 N ft U O U ftM ftU ftftftU N M ft ft ft ft (|ft0ftftl> N 0U 0U 40ftftft * OB N M U M U N ft NN U4ftN ft(i O O N U ftftftti ftH 4 U U U 4 Mm ft* Ml O O N O O U O ft O 4 N ft M ft (I H M N ft O U U U l 0 U 4 taU O U O U U O M Oo oO Do Oa oQ oO oO oO oO oO oQ oO oO oD til ft H ftH U UN M H ti M u W oOO MNft a_ a i Uftft044tllftuO UftftO ftU O Q O O Q O O O O O O O pO O O OOOOOOOOOOOOOOO _ Nf__t__MO___MM___ff_tt___a___N9___f_t______N_____________0____*_4._-_4 1 0 MM f tt 4 MM 0O f f t MM Ur 0^ u f -t---0--- *4 " QOOOOOOOQQQOQOQQ O O O O O O O O O O Q O aO O O N N f_t___f _t U U f t U O O f t M u O f t f t fUt f t f t f t 8QOftU4Q0bbfi 4 0- 4 000 00 00 00-4QOGGOOC3 O Q O o O f_t___M_ _ f t_ f t . . . 0Ok fi| ti ft u f t l) f t H f t a wu m o h o M ftM ftMOOft a OA m NN ff(fttt*f*to4*o4r oUI- i i N Uf t4 4f tf t M f t f t f t U O U O M f t M 4 4ftG 444ftU O U ftU ftN i Og 9O ?D ^4 9 2 ? ^ o o o p_ o_o_o_ _ OOQOQOOOOOOOOOOQ - _2_2O_ 2O_ 2Q_ 2O_ 2O2O9O92O0 2 2 2 2 ! OOO>MAMftoOOftOU04 O OO M ftftoO O O O ftO ftftN O O O O ft^ O O O O O N Q uA O 000 ^^ U * ft ft ft U U O M O4 U ftft D U O ft ft W N n ft f t f t U 0 0 0 4 f t U f t 0000O o ft ft 0o o0 o0 o* o0 o0 o0 o0 o0 o0 o0 o4 Sid Naan Dat. 3 Oat. 3 Ott. 1 CaapauBd U6 Swpl S upla 14 Lab UDIAN erti IS RADIAN EXT1 16 RADIAN ERTI TABLE B-l (cont'd.) Caapaud TOTAL TCOr TOTAL PNCDF TOTAL HXCOF TOTAL RFCOF OCDF TOTAL TCOO OCDD 2.3.T.B-TCOO 2.3.T.S 2.3.4,4-TCDF 1 .2 .3 .7 .8 1.2.3.4.9-PNCOP 2.3,4 ,6 .7 ,8 -K X C D F 1.2.3.4.8.7.I<H PC0P TOTAL TCDF TOTAL PNCDF TOTAL HXCOF TOTAL HPCDP OCDF TOTAL TCDO OCDD 2.3.7.8-TCDD 2 ,3 .7 .B 2,3.4.0-TCOF 1 .2 .3 .7 .S 1.2,3.4.8-PNCDF 2.3.4.T,6-PNCOF 1 .2 ,3 .4 ,T ,e 1.2.3.4,7.9-HXCOF 1 .2 .3 .7 ,8 .9 -H X C D F 1 .2 .3 .8 ,7 ,8 -K X C D F 2 . 3 ,4 , 8 .7 , HXCOF 1 . 2 . 3 , 4 , 6 , 7 , 6-HPCDF TOTAL TCDF TOTAL PNCDF TOTAL KXCDF TOTAL HPCDF OCDF TOTAL TCDO OCDD 2,3,7,6-TCDD 2 .3 .7 ,8 * 2.3.4.8-TCDF 1 .2 .3 ,7 .8 1.2.3,4.8-PNCDF 2 , 3 ,4 ,7.8-PNCOF 1 , 2 .3 . 4 .7 . 8 1 . 2 . 3 . 4 . 7 , 9-HXCDF 1 . 2 . 3 , 7 , B,9HXCDF 1.2 .3 .8 .7 .6 -H X C O r 2,3 ,4 ,4 ,7 ,8 -H X C D F 1.2.3.4.8,7,8-H PCDF Dat. 1 18.00 13100.00 9436.00 2010.00 200.00 3.00 10.00 3.00 308.00 1433.00 1803.00 4127.00 1166.00 822.00 903.00 671.00 50.00 30.00 34.00 47.00 106.00 3.00 10.00 3.00 20.00 20.00 20.00 SO. 00 20.00 20.00 20.00 30.00 3.00 3.00 21.00 SO.OO 131.00 3.00 10.00 3.00 3.00 3.00 3.00 24.00 3.00 3.00 3.00 18.00 ta t. 2 086.00 9983.00 11187.00 2384.00 194.00 3.00 10.00 3.00 368.00 1628.00 1659.00 4680.00 1432.00 912.00 1068.00 774.00 3.00 3.00 14.00 33.00 103.00 3.00 10.00 3.00 3.00 3.00 3.00 17.00 3.00 3.00 3.00 21.00 3.00 3.00 9.00 11.00 44.00 3.00 10.00 3.00 3.00 3.00 3.00 12.00 3.00 3.00 3.00 3.00 ta t. 3 833.00 7333.00 846.00 1609.00 183.00 3.00 10.00 3.00 318.00 1188.00 1204.00 3IS6.00 1107.00 789.00 871.00 810.00 3.00 3.00 21.00 21.00 113.00 3.00 10.00 3.00 3.00 3.00 3.00 24.00 3.00 3.00 3.00 21.00 3.00 3.Q0 11.00 24.00 107.00 3.00 10.00 3.00 3.00 3.00 3.00 14.00. 3.00 3.00 3.00 11.00 Mu 880.00 10139.00 9623.00 2001.00 192.00 3.00 10.00 3.00 331.00 1416.00 1555.00 4221.00 1235.00 841.00 947.00 682.00 18.70 18.70 23.00 33.70 107.00 3.00 10.00 3.00 8.67 8.87 8.87 30.30 S. 87 8.87 8.87 24.00 3.00 3.00 13.70 28.30 94.00 3.00 10.00 3.00 3.00 3.00 3.00 16.70 3.00 3.00 3.00 11.30 Std. DU. 92.50 2817.00 1218.00 388.00 8.82 0.00 0.00 0.00 32.10 222.00 313.00 420.00 173.00 83.70 108.00 133.00 27.10 27.10 10.10 13.00 5.13 0.00 0.00 0.00 9.8] 9.81 9.81 17.40 9.81 9.81 9.81 5.20 0.00 0.00 8.43 19.90 44.90 0.00 0.00 0.00 0.00 0.00 0.00 6.43 o.aa 0.00 O.GO 6.51 GENF 011243 B-H 784054 TABLE B -I (c a n t'd .) G EH P 011244 010 -4 C$occO-nn MUM UAUUU*U^>4 n >1 -4 A *4 at o <h '-4a< a Zv i -m o3 ? a a s a.- * a3 a' rr*rrrr*r " " 8383333 |u U M IJ m y M b M J O H O H H H H P^PPPi H X X *t)4 I Sisi 4O O3 D33Oa;5Q34 *4 *1 -lI lalid a2 - i 9 x x x I n-ci g^ SOM SDM SDX * "O TM--"--* 4 q D M U ffi *4 *u *4 *9*o-n4EI oH4n rrr rrr*r" ." ga 8 3 3 3 3 M U U M U U M U U B <H a H H H H i*UU(I^U44 "P^ PP PP a4* -4.9II- CI* i Sjit MlI( ? 5 S S 8r 5 8 8 3 r V" f aa ao sn 3n 3a a a o a *4 *- *8i4 i *4 3 8 m O ^ U 404U A U 4M 00044404 2o o9 o? o? o9 o? o0 o0 o0 o5 o0 o0 o0o oo oo HH *4 *4 WC(It 4A * ( 0WwM ^O^ WA ^MH^-44 o4 n4 oo oo oo oo oo oo oo oo oo oo oo oo oo oo oo aa 4i 44 44 4A ^A 04 OO 4A 4 0t*U 4U4 4M*ONAU oo oo oo oo oo oo oo oo oo oq oq oq oo oo op op i5 " ** ** - * * 4 4a vUat f a P UHAOA tN iOuMA MM a t4i 0u Ao Oo Oo Oo oO Oo Oo Qo Oo Oo Oo Oo Oo Oo Oo Oo Oo *M Uu o4 * (31O P O W ^ u m U A ^ A oo ao oo ao oo oo oo oa oo oo oo o_o HH H N M U * UK Uu N* Ht* V *o M>u4AU4 At ipaA4 g mo u 4m *a mA Q O * o - u ooooooooooooooo W- M M * wg m 00400004404000^4 oooooooooooooooo OO 4 0 0 0 0 0 0 0 0 0 0 0 0 0 oooooooooooooooa OOU-OOQOOOOOOOp ooooooooooooooo OwOO AUMm O m A uuiim U(HOOgMg^QMOO4UU OAgOOONMu)OA(JOU4N N NN oA U*I O A A O Mi I O4 4 O AA GftO g K> $ OgI r . - r r r . - r .-.-g 3 8 3 3 3 3 ..a4....^..i..o.4.l...t(..i.....^..i.a' ao.> .o .om.do4 o rdid ^n*aoS** n4* on 535 * S r" 5 8 3 S r * ,uM* * , i * r .3 5^ 1? 8 5- r.Nu.^fU.uMr.uMr.u.Nr.*Uf.uMr.-U..J.U^4(d8?H>3rO*8H*8>Hr>3r*3ir>rH3* *00'4> I4(I9 .a......*.4..........d......^......D..... * i h .--9...l.Xm..ia..X.tim...ti.X..t...t.*ia onzU -Mt U--(j DnO gTI O% O*m o! *M u -a -I Qd o X X^H nu on nX on o^ dn >o4 rM U rftf~tMrMUMmU*fPt>adS*3l3HdHH3 a a Uay*iu*i'4uVV1* ! ' f h f X f X 1P P h *^o *T^*^?W*T"**-a4*if***oh o3 dS Od ^ao T3 rsa* ar" SQ ST TS OT rU -*u^ T *H V i ia on ?3 3 " 8 4io- u U u (J (aIM4"i OMO ah u aMa cu nOO (O WMd u 4M4 4a4 m MmMMOu U a u O^ 4M U o MOM oI- V 4 M U p un u u a u o oo---oo- --oo -oo---oo- --oo --oo -oo 5o oo oo oo -o---o--o-- _, -IU M M mu Mm m 41 oU m M d 4UU4QUO *P40P0m m 0 ^ 4 n u U a A4 O4 8SS38SS383SS8SS8 d d I- HMON u 4u a iu 4c 0j 4u Mo uu u o u dMa uu *i nd 4 $ 2 ? 5 0 0 ?OO O OO O aO OOOOOOOOOOOOOOOO MO 4M M HU M 4M o M K)M A M 4Mt ut do da Mo uO oo du u o*-u r ou tot un v2 c4 OOOOOOOOOOOOOOOO OOOOOOOOOOOOOOOO UHM* M 4UM MMUOM MdMaU U 4 Ud Od uU cU ^H o4 oU oU^OoUo dSu oOddui oQ oO oO oP Po Po Oo Qo Qo Oo Oo Oo o Oo Oo Oo M *Mud^ud Mod uO d hO u uu p*u du oMa i oo oo oo oo oo oo oo oo oo oo oo oo oo oo oo oo 4U P 4ppU O U ftpvpQ oo oo oo oo oo eo oo oo oo oo ao oo oo oo Qo oo U MN M m ad*UMNUdMOH4OUM0d oq oo oo oo oo od oq oq oo oo oq oo oo oo oo oq OI M Mu d M UUAAdU**U(UJIo>u*M(*-JOgaUuM(4D4Mdtf 0o -o4o* 4o -o4 o0 o* 4o0 o0 o0 0o 0o0 o0 o0 0o *o4 m Mu M4 Me OM 4M UO OOM 4 b>M M iHlJNUM U MMMM MpNON* 4MUNUMAMQOOdUd4U 0M UOOMd400000(000 OdMdaaauoooootfoi ddd*OaMOODO(NdMO oot oo o o~4o<DoOo'-~oOoOoO Oo OoCoOofro* M U N NH M a N*NUa ooo TABLE B -l (c a n t'd .) O tO H D O O O O nvn Ha oo * i lOVOHOOOfl B)il it o mo h o 7t" o o o o oa oa aa on oo oo oo oo oo a ^0mOnh NNO D O t ' ^n*1nNOO} <<iAO| rHht iN'tOpN4 1o o t f v f t o a a o o o v o o o O O itO ilO O O O O O ilO O O 1 00^*^0000000000 c ot*nnonnOn 0hn o r t Nr^OlNMvv ftnOoiP*f iOfDtt i*v)OD n n *o* nn mh rtnnrtnwnn 8Q Q 8 f i 9 0 Q O o o o o o o < oooaooooooooO o o r- o o 1 O V) N 4 mS*4d0O 0^0d*rNt Q ( t p <hh*4nCNi P l PfC* nWh d i t i ) Dhdrton^ndr)dino O O Q Q O O PO O PO O O O O oaooaoooooooooo S S S S R ^S ^S S S m1 oppoooooooooopoo oooooooooooooooo nvchd C ri a N *i nttftM iH 1 d N h 9 h O O O pO O O PQ O pO O O O O oooooooooooooooo SS5 "2 SQ Q Q 9 8 9 S 9 S 0 Q 9 9 9 9 ooooooooooooooo t'NfbfiOfltpOfrpOOONt* S 3" B M 0 ^ 0 1 O Q O O O O O O O O O O O PPO oooooooooooooooo (]O i < t n o n o na 3 d` f *h~ n -d----d-----h- ---f---t--v-----d----M-- -l 0014* f)ftB*NVh 5M ooooopopoooopopQ OOOOOOOOOOOOOOOO nnoddrtQrtnnnn ----- - " BP*4. n^ rt r* to* to* to. V <i to h 6 5 8 L -8 * IfiE 5 i3 i ^* Of * r# i o 4 k --! ri --8 ! <4 axi. o i o a *- i ..8 8Kx 8xK *o 4t- BC4 j ^ ^ ^ ^ totonvnftnvn 6h hfteofteofOelH6O8 r"t 't o"- N'HHH"H!M6H6 t6ot8ot6otoO"t -o i'nHn'* *in" H " !* 8* r h4 n B 1 N to to toB 4M tort8 88x 2Sto. OO QO XX n - 8x * gcx i o* 87 ELBBCBto * B fl to ) to to B * to B B 4 j *j j j ^ j t o t o n ^ n n n ^ h tot o6t 6o t8o E8 t8o*O!*B!n'!M"!N"*HH* "i!n"!* 4" IoO 00 h- v o <N ft? o CIQ TMIE B-I (cant' d.) GENP0J1247 o o o o r - o - o o o o a o o o o pi n n p) ^ - * >-i * n dOt vo Ni oOoOo Oo oOiHA OOnN^iOs V fNnfol n^t*t>o^o*r)^notonNO S Oo Oa Oo On Oa Oa oO oO oQ oQ oO oO oO oO aP OOOOOOOOOOOOOOOO oD oooooooooonv-ao ' h.^ nNv.i n_a ___M. . . .. ^ . _ _. h rt .. o*,, nw_ OOOOpOOOOOOOOQOQ oooooooooooooooo in ^w ^ f io n o n o n nQ n- hrtn hw nwne^ 1 80o oQo0 o9 a0 o0 o0 oOaQoQoOoQoOoOoO non 0000000000000 -o4aooho0i o0 n ortn ccpt(eNOoMnc MnMo-n*o* n~ 00000 r*"o ~t - x o--n o--n---N- " *--i S 0O 0O 0O 0O 0O 0O 0Q 0P Q0 0O 0O 0O 0O 0O 0P Oo oO oO oO oO oO oO oO oO oO oO oO oO oO oO oO oo oo oo oo oo oo oo oo oo oo oo oo oo oo oo po o h fj 0 n H N 4 0 rt N x N n S OO OO OO OO OQ OQ OQ OO QQ OO OQ OO Op OO OO nnnrin oO oO Oo Oo Oo Oo Oo Oo Oo Oo Oo Oo Oo Oo Oo Oo n hh Oo oO oO oO Oo Qo Oo r t 0 <1 o o0 o0 o0 0o 0 0o 0o S ---n- o d ^o ~0 oh ntf 6H8e8g1: Psr nr vt<hi-4&883 OaSrf*eX+ oK.to-aOIi oI I** 'd' fe O O Q sigi on n k , fe. II. G *Nfc*4*938s aO Nn^ SX(k *Oi 0l iOih s . 8 S 0 H& S sXQ- fei I ". n ' n N i. fefe 0n'ft,rt888x 0 0ofaj nl*i^XiBIL*o- 5oi*o5l-nti- - 8 *rI - * a h h h a Ho t* r> 0 0 0 s s s m1 1 b t fl t f i w^ N^ 5 t b P^ ow* fi M n n fi h b b b P b P ^ n r i o r i r t f n 666iB6S22"2n"22" giSSSS: 594 5 00 I CO TABLE B -I (c a n t'd .) G E N P 011248 I *-- * ID 00 -U 0 0C1D rr>rrr.M r.k'." g 3 8 g g 3 3 M U M M k lU rjL ) U g p O H ^ H u u u `u * u 4 -j r F r r f *o *-4* >4 < >4*O ft *i4 *<4 0 C* *t m 3 l Il X O T355 *Or" 3x o0 *n n0*4n0%M*i ^ M* ** U H *4 TaI3c"4 T8 SSS3 ddot r rr r r ? r " .N 8 3 8 3 8 3 3 W O U U U U M O U d H d ilH H H u * > o u u O-J-I r "* r F F ? 0 - 4 -4 0 -4 9 H ** -- i n -a am axxi xox* o * a^orsif-rt --t*mi oq x-o oa no no ^* ^ N- *- ii a "4 9* *i nH ma o*4 rs so os dg os st S' r." rrr.w r." ." 8 3 8 3 3 3 3 M U U tlU U M U U D > )O H > 1 > l-) u u u u u <-i F '" F F P F I i i i la * i ^ a s a a * H kl<? T V i 3' THSB'Br"0 SBBBr"r* :a *a mI D u -m4 o h Mh Oa u * i o u o a a u U U M O <- SO O O O O O O O O O Q O O O O ooooooooooooooo -4 1 w M M * OOOOOQQOOOOOOOOQ oooooooooooooooo m m a u u ssssssssss O P Q P Q P P O P POQQOOO oooooooooooooooo O O -4 PJ Smm mo 0 n a u o< o w SP O O Q P O O O Q P O P O O O OOOOOOOOOOOOOOO aoU. _*oUau aaMuaN_Ma___uM_MM U^ ao UMMMM ou *a4at^Mf iu ^uMua Q O Q Q O PO O O Q O O Q O Q Q oooooooooooooooo a aa m ma iu* a a 8 2 SB S S 8 OPOO Q QO POO O OO PO O OOOOOOOOOOOOOOOO m-ag owm___ouu__Mn_i_^a_* imau_o*_oa p mo u a_uov__muo__auo-_i ma_o* m*u-* ooooooooooooo ooooooooooooo - _i a ,,i Mu u * o MP M aaaaaM H N ao*a*au O O O PO O O O O O O O Q O O Q OOOOOOOOOOOOOOOO U mp aHu a M o d*M< aMi Ha i MO oi So So SASOaM PO O O O O Q O O O Q O O PO O OOOOOOOOOOOOOOOO H^ oN oUaM* HH ^Op k u* oa u oa Ii OOOOOOOOOOOOOOOO * . M. . .M u nN aN.MU_ 4M M i pN* a ONa pMp HUa gM11 - - ------ ' - i a a o p p o Im* %oH*o uoo *aa*K *PM 0oa ^ M ^ a u u p H a o o Ho a* ^U a oM oooaaoaM ooooooa* 00*^ 0001*000000^ 0 _ _____p* a a o a a N P M O ^ a amM* * !a uo a*oaM O N *aoaooH U _ ' O U O U O O O-- O O uu * * OUOO TABLE B-l (cant'd.) pfOOO00O o m o o ciin o o " <rt f e t O9 *Pi < o o o o t f i * t O Q f f r0> v i > oo oo oot h- oo mn tof oO r 0j N # o** wf ) - o o n o Mn h t O M ) 0mo o o o o o o o o q q o o o o SOo oO aO oOoOo Oo Qr *O oOr O) oO nO nO rO> Oo eo h >* 4 tN* r9 ** o n ft h P ^o r )o po arti * 71 ^n a oo oo po po ntoo"'' o n p o o *n o noh n ft<o<M * opooooopeeppooop ooooooo oooaoooo Sm n o N Opfi t i d ( l f toOf 4 *n n ht iOnMHn)1 ooooooaooooepooo ooooooooooaoooao von*n ntnn 0 n' v d anr ) D- n h q nh on n n n- -o ooooooooeooooop OOOOOOOOOOOOOOO oo oo*n*t**no<oppnn o -n H ^ j nn oo o"n pv No * r Sp p o p e o o p a o o o p o p oooooooaoooaooo no no noo n o n ft o4 o oo*4 np n h oh ooooeoooaooooooo ooooooooaoooaooo hoOnfNnn n oNn ort n r t O M n n o n o o ooopooo oooo poo aoooopooaooooooo _ nnoononnoipoeoNii) h Nn a n14 i4 r t rn tnAo bof Hf to Ot A- ob 8883S8S8888S8SSS P P O p O O O O O Q O O p O O O oO oooooaoooooooo 2 S S S j j g - ' * S " S " S r t 3* o n o * a n a o r t t ' i r - Q r - T t ' i * O bonononoonoon OB N rt h N O N rt t* n aooooooooopoooop o o o o o o o o o o o o o o o fi n a fM h n o n n n N M N N B b r t | r t f t r t O H D N O h O p r t o) r t r t nabnN O ob AON r t ^ rt *. s 5i 8h-o. nr er 8-. r H n seep es *Nb'86G i -i-KB. ei sl sl i* n E a i> 88 *! O o n b. k. a. D * - n\ a" 8S8 x XX i I o -- u * x a: x a. x iii* a n a, s a e a t> au i ......................... g h O C h h f- O M oohr-hooo eigS \ -4*5 8 3 s Il No* t r f t* g * V0M Oil I SiB StO*ib M * ^J ^ J Qbbnnfinn ^J j j ^^ b^ n nn n n j j j j J t*bnnnnn o 0C O 1 a <N f C> fti O CVII CO S u p 1a 15 Lab RADIAN EXT3 16 RADIAN LXTO 17 RAOIAN EXT3 TABLE B-l (cont'd.) Coa pound TOTAL TCDF TOTAL PNCDF TOTAL KXCDF TOTAL MPCOF OCDF TOTAL TCDO OCDD a .J .T .B-TCOO 2 .3 .7 . 2 .3 . . -T C D F 1 .2 .3 .7 . 1 .2 .3 .4 .8 -PNCOF 2 .3 .4 .7 .8 - PNCDF 1 . 2 . 3 . 4 . 7 . 1 . 2 . 3 , 4 . 7 , - KXCDF 1 . 2 . 3 . 7 . 2 . -KXCDF 1 .2 .3 .6 .7 .8 - HXCDP 2 .3 .4 .8 .7 . - HXCDF 1 .2 .3 .4 .8 .7 .8 - HPCDF TOTAL TCDF TOTAL PNCDF TOTAL HXCDF TOTAL HFCDF OCDF TOTAL TCDO OCDD 2 .3 .7 .8 - TCDD 2 .3 .7 . 2 .3 .4 .8 -TCDF 1 .2 .3 .7 . 8 . 1 .2 .3 .4 .8 -PNCOF 2 .3 .4 .7 .8 - PNCDF 1 .2 .3 .4 .7 . 8 1 .2 .3 .4 .7 .8 -HXCDF 1 .2 .3 .7 .8 .9 - HXCDF 1 .2 .3 .8 .7 .8 - HXCDF 2 .3 .4 .6 .7 .8 - HXCDF 1 .2 .3 .4 .6 .7 .8 - HPCOF TOTAL TCDF TOTAL PNCDF TOTAL HXCDF TOTAL HPCDF OCDF TOTAL TCDO OCDD 2 .3 .7 .8 - TCOD 2 .3 .7 . 8 2 .3 .4 .8 -TCDF t .2 .3 .7 . 8 1 .2 .3 .4 .8 -PNCOF 2 .3 .4 .7 .8 - PNCDP 1 .2 .3 .4 .7 . 8 . 1 .2 .3 .4 .7 .9 - HXCDF 1 .2 .3 .7 .8 .9- HXCDF 1 .2 .3 .6 .7 .8 - HXCDF 2 .3 .4 .6 .7 .8 - HXCDF 1 .2 .3 .4 .6 .7 .8 - HPCDF D a t. 1 3.00 3.00 28.00 43.00 2.00 3.00 10.00 3.00 3.00 3.00 3.00 31.00 3.00 3.00 3.00 22.00 3.00 3.00 8.00 3.00 10.00 3.00 10.00 3.00 3.00 3.00 3.00 11.00 3.00 3.00 3.00 3.00 313.00 1283.00 2360.00 2412.00 4412.00 3.00 10.00 3.00 182.00 313.00 187.00 1406 00 64.00 70.00 109.00 1173.00 D a t. 2 3.00 3.00 20.00 31.00 100.00 3.00 10.00 3.00 3.00 3.00 3.00 18.00 3.00 3.00 3.00 24.00 7 00 3 00 3.00 3.00 10.00 3 00 10.00 3.00 7.00 3.00 3.00 a oo 3.00 3 00 3.00 3 00 312.00 1119 00 2374 00 2362 00 4334.00 3.00 10.00 3 00 184 00 313.00 182.00 1432 00 88 00 73 00 103.00 1143.00 D a t. 3 3.00 3.00 18.00 33.00 100.00 3.00 10.00 3.00 3.00 3.00 3.00 21.00 3.00 3.00 3.00 27.00 3.00 3.00 3.00 3.00 10.00 3.00 10.00 3 00 3.00 3.00 3.00 8.00 3.00 3.00 3 00 3.00 318.00 1334 00 2380.00 2363.00 4236.00 3.00 10.00 3.00 188.00 339.00 200.00 1424 00 73 00 63 00 113 00 1141 00 Naan 3.00 3.00 22.00 30 30 97 30 3.00 10.00 3.00 3.00 3.00 3.00 23.70 3.00 3.00 3.00 24.30 4.33 3.00 4.67 3.00 10 00 3 00 10 00 3.00 4.33 3 00 3 00 9 00 3 00 3 00 3 00 3 00 314 00 1239 00 2365 00 2380 00 4334 00 3 00 10.00 3 00 184 00 322 00 190 00 1421 00 88 30 89 30 109 00 1132.00 S td . Dav. 0.00 0 00 3.29 5.03 4.62 0.00 0.00 0.00 0.00 0.00 0.00 6 43 0.00 0.00 0.00 2.32 2 31 0.00 2.89 0 00 0 00 0.00 0 00 0 00 2.31 0 00 0 00 1.73 0 00 0 00 0 00 0 00 3 21 110 00 8 08 28 00 89 70 0 00 0 00 0 00 2.00 14 SO 29 13.30 4 31 6 03 8 00 17 90 GENP 011250 784061 APPENDIX C PROJECT SUMMARY GENP 011251 784062 PROJECT SUMMARY INTRODUCTION During the past four years Electric Power Research Institute (EPRI) has sponsored a research program to evaluate methodology for the analysis of selected Individual congeners of polychlorinated dibenzodloxlns (PCDDs) and polychlorinated dlbenzofurans (PCDFs) In dielectric fluids. This program Included the synthesis of native and 1sotop1ca11y labeled PCDD and PCDF analytical standards (I). Analytical standards and spiked matrix samples were prepared by a contract 1aboratory-and then distributed to four other participating laboratories as part of a round robin study. These samples were analyzed blind by five independent contract laboratories. The study was conducted In two phases. Phase I Involved the analysis of PCDF's and PCDD's in five spiked baseline samples; for example, Aroclor-1016, Aroclor-1242, Aroclor-1260, tr1- and tetrachlorobenzene mixture, aged mineral oil, and four In-service dielectric fluids (1-4). Since validated analytical methods did not exist before this program was Initiated, this phase of the program was necessary to assess any deficiencies and optimize the analytical methodology. Each participating laboratory used their best available in-house extraction and analysis techniques (1-4). After evaluating these results and Implementing several improvements to the methodology, then Phase II was initiated. Phase II was designed to provide quantitative data on PCDF and PCDD congeners 1n selected in-service dielectric fluids and to make a statistical interlaboratory comparison of the analytical methods. The specific aims were (1) to determine the accuracy of analysis by each participating laboratory for a selected 11st of PCDF's and PCDD's 1n spiked samples, (2) to determine the precision of analysis by each laboratory for PCDF's and PCDD's In spiked and in-service utility samples, (3) to perform an Interlaboratory comparison and thus a comparison of methods employed for accuracy and precision, and finally (4) to measure the levels of PCDF's and PCDD's 1n a few selected In-service utility samples. METHODS Five Independent laboratories participated 1n the Interlaboratory study. Each laboratory was provided with aliquots of the three baseline and seven In-service utility samples (4). Also, native and 1sotop1ca1ly labeled PCDD and PCDF analytical standards were prepared by one laboratory (4) and distributed to each participating laboratory. Subsequently, each laboratory employed their best GENP 011252 C-2 784063 available In-house extraction and analysis techniques (Table 1). Each laboratory added five carbon-13 labeled PCOD or PCDF congeners as Internal standards to the sample. Some laboratories employed a liquid-liquid extraction step, either hexane/acetone or hexane/sulfurlc acld/water, whereas others applied aliquots of the sample directly to a chromatographic column. All laboratories performed a purification step using open column chromatography (Table 1). For example, Lab-A used a d d alumina, PX-21 carbon and neutral alumina chromatography, respectively, for purification; Lab-B used LH20, basic alumina and a d d l e alumina; and Lab-C used a combination of sulfuric a d d s111ca/sU1ca sodium hydroxide followed by basic alumina and florlsll. Lab-D used a sandwich of a d d silica, basic silica and silica followed by PX-21 carbon, a d d s1l1ca/s1l1ca and then acidic alumina; Lab-E used sulfuric a d d Impregnated silica, PX-21 carbon and florlsll. Various solvent evaporation techniques were also used. Specifically, Lab-A used heat and vacuum; Lab-B used rotary evaporation, while Labs-C and -0 used nitrogen blowdown. During extract analysis, Lab-A used both high and low resolution MS, 1n the electron Impact mode. Lab-B, D and E used only low resolution mass spectrometry while Lab-C used exclusively high resolution mass spectrometry. All laboratories employed high resolution gas chromatography (HRGC), although different stationary phases and GC operating parameters were used. Calibration curves were generated over the linear response range of each gas chromatograph/mass spectrometry system at five concentrations. All labs used the same Initial stock calibration solution prepared by one laboratory containing eight native and five 1sotop1cally labeled PCOD's and PCDF's (1). The samples analyzed 1n the round robin study are shown 1n Table 2. Two samples were spiked with a selected number of congeners at known concentrations into mineral oil and Aroclor 1016, respectively. A third sample, Aroclor 1260, was spiked at known levels with these same congeners but It also contained endogenous PCDF's and PCDD's. Samples 11-17 (Table 2) were In-service utility samples that were selected to represent the various types of utility uses (5). In some cases the dielectric was Arochlor 1260, or mineral oil containing PCBs. Dielectric fluids from transformers, capacitors, or precipitators were selected. Some samples represented devices that had been In use for many years (Table 2). A large database was produced 1n this study. An example of the measurements that were performed for dioxins and furans by Lab-B on sample No. 14 Is shown In Table 3. All of the five participating laboratories reported either measurable values or limits of detection for each of these compounds In each of the ten samples. With the exception of one laboratory, each laboratory performed triplicate Instrumental analysis on a single extract from each sample. The mean and standard deviation were calculated (Table 3). Lab-E reported only one GENC-3 784064 P 011253 TABLE 1. ANALYTICAL PROCEDURES EMPLOYED BY ROUND ROBIN LABS Step .1 Add standards 2. Extraction 3. Chromatography Lab-A + Hex/acetone Acid alumina * PX-21 carbon Neutral alumina 4. Solvent evaporation 5. Analysis A and vacuum HRGC/LRMS (El) HRGC/HRMS (El) Lab-8 Lab-C + LH-20 Basic alumina Acidic alumina Roto-evaporat1on + Hex/H2S04; H2O S111ca/H2S04- silica/ NaOHsilica Basic alumina (2Xs) Fiori s1l N2 blow-down HRGC/LRMS (El) HRGC/HRMS (El) Step .1 Add standards 2. Extraction 3. Chromatography 4. Solvent evaporation 5. Analysis Lab-0 + - Acid s1l1ca/bas1c sillca/sIHca PX-21 carbon Acid s1l1ca/s1Hca Acid alumina * N2 blow-down HRGC/LRMS (EX) . Lab-E + - H2S04-s1l1ca PX-21 carbon Fiori s1l - HRGC/LRMS (El) GENP 011254 C-4 784065 Sample No. 1 2 3 11 12 13 14 15 16 ' 17 TABLE 2. SAMPLES ANALYZED IN ROUND ROBIN STUDY Dielectric Spiked mineral oil Spiked Aroclor 1016 Spiked Aroclor 1260 Aroclor 1260 Aroclor 1260 ? 7 Mineral oil (150 ppm PCB) Mineral oil (100 ppm PCB) Aroclor 1260 Utility Appliance - Transformer (LCN)a Transformer (LCN) Capacitor Precipitator Transformer (Arc furnace) Transformer (LCN) Transformer (LCN) Service Life (yr) - 31 28 ? 27 8 ? 20 aLCN = load circuit network. TABLE 3. EXAMPLE OF MEASUREMENTS PERFORMED FOR DIOXINS/FURANS BY A LAB ON ONE SAMPLE (LAB B. SAMPLE 14)- PPB Compound 2.3.7.8-TCDD Total TCDO OCOD 2.3.7.8 2.3.4.8-TCOF Total TCDF 1.2.3.7.8 * 1.2,3.4.8-PNCDF 2.3,4,7,8-PNCOF' Total PNCOF 1.2.3.4.7.8 1.2.3.4,7,9-HXCOF 1.2.3.7.8.9-HXCOF ].2,3,6.7.8-HXCOF 2.3.4.6.7.8-HXC0F Total HXCOF 1.2.3.4,6.7.8-HPCOF Total HPCOF OCOF Heasuraaant 12 3 (0.5)* (0.5) (0.5) (0.5) (0.5) (0.5) 8 13 5 297 246 282 970 . 617 576 1115 1039 1023 1171 1584 1726 9236 6589 6002 3802 3711 3775 174 187 233 1195 1155 1137 387 472 509 10.598 9918 10.666 728 90S 884 1725 2196 2254 293 626 374 Mean (0.5) (0.5) 8.7 276 721 1059 1694 7276 3763 198 1162 456 10.394 839 2058 431 S.D. (0.0) (0.0) 4.0 25 217 492 98 1723 47 31 30 63 414.0 97 290 174 * ( ) L1a1t of dataction. 784066 C-5 G EN P0I1255 measurement on each sample extract. Lab-0, on the otherhand, extracted each sample three times and then performed three Instrumental determinations on each of the extracts, thus producing nine measurements per sample. RESULTS AND DISCUSSION Accuracy of Measurement The accuracy of measurement for specific selected congeners, for each labora tory and by sample matrix was examined. To gain insight Into measurement accuracy, samples were spiked with known amounts of each congener and the percent accuracy of the mean of triplicate measurements was calculated and expressed as the mean divided by the spiked amount times one hundred. Table 4 shows an example of these data for just one chemical, 2,3,7,8-TCOO. These statistical data were obtained for each compound measured. The percent accuracy for spiked samples for 2,3,7,8-TCDD by sample and laboratory Is shown. Lab-D performed triplicate extract1on/pur1f1cation on each of the samples which also provided an estimate of precision for the entire analytical procedure. The mean value of the accuracy for 2,3,7,8-TCDD 1s given In Table 4 and the range for all determinations 1s shown In parenthesis. For these two samples the percent accuracy Is 1n general very good. Lab-E only performed a single determination and experienced a lower percent accuracy than the other four laboratories. Accuracy of analysis for 2,3,4,6,7,8-hexachlorodlbenzofuran congener 1s shown in Table 5. The percent accuracy of measurement for this PCDF, by laboratory and sample is given in Table 5. The accuracy of determination was quite good for four of the five laboratories. Lab-C reported more variable recoveries with a range of 80-3371 1n sample No. 1. Data such as these were obtained for each congener to determine the percent accuracy In two samples. With the exception noted above, acceptable accuracies were obtained. A third sample also was spiked at known concentrations with these congeners, but there were endogenous levels present 1n the Aroclor 1260. Thus, the percent accuracy could not be calculated. The median and range of percent accuracy across all. the chemicals that were spiked Into the two samples Is sunmarlzed 1n Table 6. As Indicated, the accuracy 1s reasonable 1n most cases with a few exceptions where the range was' quite large. The determination of the hexa CDF congeners by Lab-C yielded this ' large overall variability. Precision of Measurement Similarly to accuracy, the precision of measurement by congener, laboratory and sample matrix was also Investigated. The coefficients of variation for the determination of 1,2,3,7,8- plus 1,2,3,4,8-pentachlorodlbenzofuran (not chromatograph1cal1y resolved) by compound and laboratory are given 1n Table 7. In C-6 GENP 011256 784067 Laboratory A B C D-l D-2 D-3 E TABLE 4. PERCENT ACCURACY FOR SPIKED SAMPLES FOR 2,3,7,8-TCDD BY SAMPLE AND LABORATORY Sample 1 75 (68-85) 83 (68-96) 81 (76-92) 99 (96-100) 100 (96-104) 92 (88-96) 56a Sample 2 91 (75-103) 101 (100-104) 85 (83-87) 96 (92-98) 96 (94-98) 94 (91-96) 72a aS1ng1e determination. Laboratory A B C D-l D-2 D-3 E TABLE 5. PERCENT ACCURACY FOR SPIKED SAMPLES FOR 2,3,4,6,7,8-HXCDF BY SAMPLE AND LABORATORY Sample 1 92 (86-97) 94 (93-96) 1189 (80-3371) 98 (97-101) 101 (99-102) 99 (97-99) 97a Sample 2 114 (106-123) 123 (119-129) 2015 (515-3939) 100 (100-101) 106 (105-106) 102 (98-104) * 83 aS1ngle determination. 784068 c-7 GENP 011257 Laboratory A B C D-l D-2 D-3 E TABLE 6. MEDIAN AND RANGE OF PERCENT ACCURACY ACROSS ALL COMPOUNDS FOR SPIKED SAMPLES Sample 1 Sample 2 88 (68-105) 96 (68-124) 94 (8-3371) 99 (81-114) 101 (88-114) 94 (67-107) 103 (56-116) 107 (75-135) 112 (55-149) 120 (12-3939) 101 (90-121) 105 (89-119) 99 (91-121) 114 (42-120) Sample 1 2 3 11 12 13 14 15 16 17 TABLE 7. COEFFICIENTS OF VARIATION FOR 1,2,3,7,8 + 1.2, 3,4,8-PNCDF BY COMPOUND AND LABORATORY Laboratory No. A B C D-l D-2 12.0 11.5 0.9 12.9 26.1 8.4 3.8 11.7 4.0 2.1 1.8 4.1 6.3 4.8 7.1 4.6 61.1 14.9 11.5 3.1 13.6 3U6 23.1 12.8 7.3 2.2 1.4 1.3 4.0 10.4 16.6 6.3 15.6 11.5 3.0 2.1 6.8 2.1 4.5 7.5 17.3 18.4 D-3 0.7 1.0 3.7 2.2 2.9 0.0 5.7 4.5 GENEOU258 C-8 784069 general across all the samples, the coefficients of variation (1.e., precision) were quite low. Lab-D performed triplicate extractions on each baseline and utility sample and then performed triplicate instrumental determinations on each extract. Good precision for both extraction and measurement were observed. Another example 1s shown In Table 8 for 1,2,3,4,6,7,8-heptachlorodlbenzofuran by compound and laboratory. For this particular congener, a greater variability was observed amongst Labs-B and C. The median and range of coefficients of variation across al_[ compounds by laboratory and samples are given 1n Table 9. A small range 1n the coefficients of variation was observed except for Lab-C which had relatively poor precision. Components Contributing to Variability A more detailed statistical analysis was performed to determine the contribu tion of instrumental measurement, method extract1on/pur1flcatlon, and sample type to the total variability of a reported value for a PCDD or PCDF congener. The variance components for a set of measurements made on 2,3,7,8-TCDD across all except two samples, samples 15 and 16 (most data were non-measurable values), are listed 1n Table 10. The total variance and the percentage of the total variance apportioned to either the sample or the instrumental measurement step are shown. The greatest variance 1s attributed to the sample type as compared to the instru mental measurement step. Lab-A, as compared to the other laboratories, exhibited a higher percentage of variability for the measurement step (14.6%). The remain ing three laboratories exhibited the major variability (99+%) from the sample type. Table 11 gives the variance components for 2,3,4,5,6,7,8-hexachlorodibenzofuran across all sample types analyzed. Again, the total variance and the percent of the total variance for each of the laboratories are given for the sample and the measurement step. Except for Lab-C, the predominant variance 1s associated with the sample type. For Lab-C, 30% of the variability 1n these results was associated with the Instrumental measurement step. The variance components by congener for Lab-D are depicted 1n Table 12. Lab-0 performed triplicate extraction on the utility sample and triplicate Instrumental determinations on each sample extract. Thus, 1t was possible to apportion the variability due to sample type, extraction, and measurement. The largest contri bution to variation 1n the reported values was associated with the types of sample or sample matrix analyzed. Variability resulting from the extraction steps was smaller while an Insignificant variation was attributed to the instrumental deter mination step. The measured values between each lab were tested to determine whether they were significantly different from one to another. The Levene's test of GENP 011259 784070 Sample 1 2 3 11 12 13 14 15 16 17 TABLE 8. COEFFICIENTS OF VARIATION FOR 1,2,3,4,6,7,8-HPCOF BY COMPOUND AND LABORATORY Laboratory No. A B C D-l D-2 4.0 11.7 4.1 10.3 9.9 21.1 1.6 3.4 10.9 20.4 12.1 17.8 18.7 31.6 11.5 60.6 53.4 23.9 96.2 76.1 144.9 58.9 36.9 60.9 39.9 29.2 9.7 3.2 1.4 17.6 14.9 20.4 21.7 57.4 14.9 5.1 4.3 5.4 0.7 5.5 7.5 17.6 6.7 4.8 0-3 2.4 6.3 5.7 4.0 4.9 1.2 10.3 1.6 TABLE 9. MEDIAN ANO RANGE OF COEFFICIENTS OF VARIATION ACROSS ALL COHROUNOS BY LABORATORY ANO SAMPLE SanoIr L*O ortory 1 A 1 1 . 1 (4 - 1 6 ) B 4 . 2 (0 - 1 7 ) 23 1 1 . 2 (3 - 4 0 ) 3 . 3 ( 1 -5 ) 2 2 (1-4 6 ) 4 6 (1-3 0 ) 11 8 6 (1-4 7 ) 1 4 . 1 (6 - 4 6 1 12 13 1 9 . 2 (2 -5 9 ) 1 5 . 5 (3 - 3 3 ) 6 . 6 (4 - 2 0 ) 1 7 . 6 (4 - 3 9 ) 14 2 9 (1-5 7 ) 12 6 (1 -4 7 ) 17 4 3 (2 - 2 0 ) 19 2 (2 -2 8 ) C 4 1 . 5 ( 1 1 - 1 5 9 ) 1 3 . 9 (2 - 8 7 ) 2 9 . 1 (6 - 1 4 5 ) 4 6 9 (9 - 1 2 1 ) 2 3 9 (3 - 5 7 ) 3 1 . 2 ( 1 0 -6 2 ) 3 0 . 5 (7 - 1 2 3 ) 2 2 . 2 (2 -5 3 ) 0-1 4 . 1 (1- 2 7 ) 2 . 4 (0 - 5 ) 2 . 7 ( 1 - 1 2 ) 1 3 . 6 (7 - 3 2 ) 1 2 . 2 (7 - 2 2 ) 1 1 . 1 (6 - 3 9 ) 1 2 . 4 (4 - 2 8 ) 1 1 . 9 (5 - 2 2 ) 0 - 2 4 0 (1-7 ) 2 . 7 (0 .4 ) 5 . 4 (2 - 3 0 ) 2 . 1 (0 - 2 1 ) 4 5 (1 -8 ) 7 . 5 (2 - 2 4 ) 1 4 . 6 ( 1 - 2 7 ) 4 . 9 (1-2 3 ) 0-3 2 . 4 (1-2 1 ) 3 . 1 (1-- ) 2 . 2 (1- 7 ) 2 . 3 (0 -6 ) 4 9 (2 - 2 B ) 8 . 2 (0 - 2 7 ) 1 . 9 (1- 1 7 ) 2.1 (0 9 ) 78407i C-10 GENP 011260 TABLE 12. VARIANCE COMPONENTS BY COHPOUNO FOR LAB-0 Compound Total 2.3.7.8-TCOO 2.25 Tota l TC00 2.25 OCOO 0.00 2 . 3 . 7 . 8 2 . 3 . 4 . 6-TCOF 1.48 To tal TCOF 1.40 1 .2.3.7.8 1.2.3.4.8-PNCOF 1.84 2.3.4.7.8-PNC0F 1.31 Tota l PNCOF 1.62 1 . 2 . 3 . 4 . 7 , 8 l . 2 . 3 . 4 . 7 . 9-HXCOF 5.80 1 . 2 . 3 . 7 . 8 . 9-HXCDF 4.73 1 . 2 . 3 . 6 . 7 . 8-HXCDF 2.25 2 . 3 . 4 . 6 . 7 . 8-HXC0F 2.25 Total HXCOF 3.66 1.2.3.4.6.7.8-HPCOF 3.92 Tota l HPCOF 5.13 OCDF 7.73 Variane Conpontnts Saaola Extraction 2.25 0.00 2.25 0.00 0.00 0.00 1.47 0.00 1.38 0.01 1.82 0.01 1.30 0.00 1.61 0.00 5.80 0.00 4.71 0.01 2.16 0.08 2.16 0.08 3.59 0.05 3.87 0.02 5.00 0.09 7.41 0.30 Mas. 0.00 0.00 0.00 0.00 0.01 0.01 0.01 0.01 0.00 0.01 0.01 0.01 0.01 0.03 0.03 0.03 Saapla 99.9* 99.9+ 99.9 99.5 98.9 99.2 98.9 99.2 99.9 99.5 96.1 96.1 98.3 98.7 97.6 95.9 X of Total Extraction 0.0 0.0 0.0 0.2 0.6 0.4 0.0 0.0 0.0 0.2 3.4 3.4 1.3 0.5 1.8 3.9 Maas. 0.0 0.0 0.0 0.3 0.5 0.4 1.1 0.8 0.1 0.3 0.5 0.5 0.4 0.7 0.6 0.2 784072 GENP 011261 C -12 Laboratory A B C 0-1 D-2 0-3 E TABLE 10. VARIANCE COMPONENTS FOR 2,3,7,8-TCDD ACROSS ALL SAMPLES (1,2,3,11,12,13,14,17) Variance Components Total Sample Measurement 8.05 6.87 1.18 5.46 5.45 0.00 2.12 2.12 0.00 2.25 2.25 0.00 2.30 2.30 0.00 2.19 2.19 0.00 7.52 X of Total Sample Measurement 85.4 14.6 99.9 0.1 99.8 0.2 99.9+ 0.0 99.9+ 0.0 99.9+ 0.0 Laboratory A B C D-l 0-2 D-3 E TABLE 11. VARIANCE COMPONENTS FOR 2,3,4,6,7,8-HXCDF ACROSS ALL SAMPLES (1,2,3,11,12,13,14,17) Variance Components % of Total Total Sample Measurement Sample Measurement 3.27 3.25 * 0.01 99.5 0.5 2.62 2.59 0.03 99.0 1.0 4.53 3.15 1.38 69.6 30.4 3.11 3.10 0.01 99.6 0.4 1.77 1.76 0.01 99.2 0.8 1.85 1.84 0.01 99.5 0.5 4.27 - -- - 784073 C-ll G E N P 011262 differences amongst the laboratories by compound across the samples was applied to the data (6). For available data, the F value for these measurements and the degrees of freedom were calculated and are shown 1n Table 13. In every case, the values reported by each of the laboratories were significantly different from one to another at the 95% confidence level for each of the congeners except 1.2.3.6.7.8- hexachlorodlbenzofuran, which was significantly different at the 99% confidence level. This test Indicates that each laboratory reported statistically different quantitative values for each of these congeners 1n these samples. Prediction of Variability 1n Future Measurements In practical terms, what would be the predicted range of values 1f the analyses were repeated by a laboratory using one of the methods previously employed. Shown 1n Table 14 are the approximate 95% prediction Intervals for the laboratories if they were to repeat the analysis for each of these chemicals in the same utility samples. In other words, if one of the five laboratories were to repeat the analysis for 2,3,7,8-TCDD 1n sample No. 1, the reported value would fall within a range of 15-30 ppb. If 20 determinations were made, one value would be outside of this range. For 2,3,7,8-TCDD a factor of two variation would be anticipated. In some cases, the variability for other congeners may be as much as 20-fold or more. This variability does not appear to correlate to the congener concentrations actually present 1n the samples. For example, even though, in samples 11-17 the amount of 2,3,7,8-TCDD 1s quite low and the variability may be as much as a factor of 20-30 as Indicated here, the reported values for 1.2.3.7.8.9- hexachlorodibenzofuran could vary over the range of 1-8869. In sam ples 11-17, a large variability in levels for the individual isomers are predicted as well as for the total, for example, the total hexachlorodlbenzofuran. 784074 C-13 GENP 011263 TABLE 13. LEVENE'S TEST OF DIFFERENCES AMONG LABS BY COMPOUND ACROSS SAMPLES Compound 2,3,7,8-TCDD Total TCDD OCDD 2,3,7,8 + 2,3,4,8-TCDF Total TCDF 1,2,3,7,8 + 1,2,3,4,8-PNCDF 2,3,4,7,8-PNCDF Total PNCDF 1,2,3,4,7,8 + 1,2,3,4,7,9-HXCDF 1,2,3,7,8,9-HXCDF 1,2,3,5,78-HXCDF 2,3,4,6,7,8-HXCDF Total HXCDF 1,2,3,4,6,7,8-HPCDF Total HPCDF OCDF F-Value 14.96** 17.82** 12.81** 6.60** 8.89** 4.39** 9.76** 11.15** 17.80** 16.10** 2.96* 36.93** 23.12** 22.80** 17.39** 6.97** Degrees of Freedom 5 5 5 5 5 5 5 5 5 5 4 5 5 5 5 5 *S1gn1f1cant at O.Ol level (99%). **S1gn1f1cant at 0.05 level (95%). GENP 011264 C - 14 784075 TABLE 14. APPROXIMATE 95 PERCENT PREDICTION INTERVALS (PPB) Compound 1 2.3.7.8-TCDO OCOO 1530 159 2.3.7.8 + 2.3.4.8-TCDF 144221 1.2.3.7.8 1.2.3.4.8-PNCOF 123168 2.3.4.7.8-PNCOF 99.248 1.2.3.4.7.8 1.2.3.4,7,9-HXCOF 182347 1 .2.3.7.8.9-HXCDF 134 1.2.3.6.7-HXCOF 3172 2.3.4.6.7.8-HXCDF 401041 1.2.3.4.6.7.8-HPCDF 1031104 OCDF 289 Total TCDF 144222 Total PNCDF 206431 Total HXCOF 1551862 Total HPCDF 1031105 Total TCDO 660 2 4060 140 5767 138193 90321 115 125 4572 241912 1371530 67208 5767 220517 402118 1371532 4060 3 63132 232 9771355 226425 164618 8251838 2965 254412 553975 963.258 22806436 1046 2321 10041682 79810.654 8194341 63132 11 125 719 2461 417913 84145 213812.579 18869 218603 26879 58016.405 964164.743 57362 11253079 304121.535 158141.116 121 12 120 530 1581 27264 2859 447809 4167 2097 968 418976 382210.453 37321 143590 7261320 14673090 114 13 113 170 90459 830 4680 1336 161 2010 151 117 775 3891247 179442 11198 252 115 14 132 143 251459 7421577 13242389 31445860 647800 5611385 1962882 4821130 96457 532 1389 584613.471 589721.657 13913483 120 17 111 532 119241 74746 106403 4865023 16136 43131 102237 795652 44224.465 249 729 5621.999 61013.977 24211.063 110 C-15 784076 G E N p 01i265