Document NE6N6VMJzLq8KLqgYwNXd3kmQ

FILE NAME: Scotts (SCTS) DATE: 1982 Sept 27 DOC#: SCTS009 DOCUMENT DESCRIPTION: EPA Report - Analysis of Vermiculite Samples for Fiber Content & Asbestos l [ [ [ r i i i i i i f [ t i i i ^ * /~T~ z/ct/oe MIDWEST RESEARCH INSTITUTE COLLECTION, ANALYSIS AND CHARACTERIZATION OF VERMICULITE SAMPLES FOR FIBER CONTENT AND ASBESTOS CONTAMINATION TASK 32 FINAL REPORT September 27, 1982 EPA Prime Contract No. 68*01*5915 MRI Project No. 4901-A32 Prepared for U.S. Office Environmental Protection Agency of Pesticides and Toxic Substances Field Studies Branch 401 M Street, S.W. Washington, D.C. 20460 Attn: Dr. Frederick Kutz, Project Officer Mr. Thomas Dixon, Task Manager Received Office of Enforcement FEB 1 U 2000 Compliance & Env. Justice MIDWEST RESEARCH INSTITUTE 425 VOLKER bOUlFVARD. KANSAS CITY. MISSOURI 64110 616 753-7600 I I I I I 4 i I I f i I PREFACE This final report presents the results obtained on HRI Project No. 4901-A, Task 32, "Collection, Analysis and Characterization of Vermiculite Samples for Fiber Content and Asbestos Contamination." The task was undertaken for the Environmental Protection Agency under EPA Contract Ko. 66-01-5915 with Midwest Research Institute. Sample collection was conducted by KRI, Mr. Kenneth Thomas, sampling crew chief. The analytical portion of this task was conducted through subcontracts with Ontario Research Foundation, Dr. E. J. Chatfield, Project Hanager, and IIT Research Institute, Mr. David Jones, Project Manager. This report was prepared by Hr. Gaylord R. Atkinson, MRI Task Leader, with assis tance from Mr. Thomas, Dr. Chatfield, Mr. Jones, and Ms. Donna Rose (MRI). MIDWEST.RESEARCH INSTITUTE Approved: 'l James L. Spigarelli, Director Analytical Chemistry Department John E. Going Program Manager ii CONTENTS Preface................................................................... ii F i g u r e s ................ iv Tables ................................................................... v Abbreviations, Definitions, and Specifications.......................... viii 1. Introduction....................................... 1 2. S u m m a r y ........................................................ 3 3. Experimental P r o t o c o l ...................................... . . 9 Sample collection ........................................ 9 Sample handling .......................................... 9 Sample analysis ................................ . . . . . 10 4. Sampling. ....................................................... 13 V. R. Grace Mine, Libby,Montana................. 13 V. R. Grace Mine and Processing Mill, Enoree, South Carolina.................... 14 Patterson VermiculiteCompany, Enoree, South Carolina . . 14 5. Sample Handling . . . . . . . . ............................... 30 Bulk s a m p l e s . ............................................ 30 Air samples . . . . . ................................... 30 6. Analytical Results.......... -. . . ................. .. 31 R e f e r e n c e s ............................. 69 Appendices . A. Study protocol for the collection and analysis of vermiculite and related samples for the evaluation of fiber content vith emphasis on a6bestiform fibers........ .................... .. A-l B. Detailed analysis procedures as submitted by IITRI............ B-l C. Air samples flow rstes and sample v o l u m e s ..................... C-l D. Increment bulk samples collected and c o m p o s i t e d ............ .. D-l E. Photomicrographs and TEH micrographs of selected samples. . . . E-l iU . t 1 4 `h 'P ' i i Vl' Hi Number 1 2 3 4 5 6 7 8 FIGURES Page Wind rose pattern showing the direction and intensity of the wind during the air sampling period at the Grace, Libby, Montana, facility.............. 16 Map of the Grace, Libby, hontana, facility showing the stationary air sampling locations........................... 17 Wind rose' pattern showing the direction and intensity of the wind during the air sampling period at the Grace, Enoree, South Carolina, hill . . ................. .................. 22 Map of the Grace, Enoree, South Carolina, mill area shoving the stationary air sampling locations....................... 23 Wind rose pattern shoving the direction and intensity of the wind during the air sampling period at the Grace, Enoree, South Carolina, m i n e ....................................... 24 Map of the Grace, Enoree, South Carolina, mine area shoving the stationary air sampling locations....................... 25 Wind rose pattern shoving the direction and intensity of the wind during the air sampling period at the Patterson, Enoree, South Carolina, facility ................................... 28 Map of the Patterson, Enoree, South Carolina, facility shoving the stationary air sampling locations....................... 29 TABLES Humber Page 1 Summary of Optical Microscopy/XRD Analysis Results........... A % 2 Summary of Electron Microscopy A n a l y s i s ..................... 6 f * 1 3 Results of the Phase Contrast Analysis of Air Samples Collected at Three Venniculite Sites....................... 8 i 4 Air Samples Collected at the Grace Mine and Mill, Libby, M o n t a n a ..................... 15 5 Bulk Samples Collected at the Grace Mine and Mill, Libby, 1 M o n t a n a ..................................................... 18 6 Air Samples Collected at Grace, Enoree, South Carolina. . . . 19 7 Bulk Samples Collected at the Patterson Operations Enoree, South Carolina............................................... 21 ] 8 Air Samples Collected at Patterson, Enoree,South'`Carolina. . 26 1l 9 Bulk Samples Collected at the Grace Mine and Mill, Enoree, South Carolina............................................... 27 r 10 Specific Gravities of Selected Minerals ..................... , 32 i 11 Density-Separated (and Hand-Picked)Fractions Produced. . . . 33 * A 12 Summary of X-Ray Diffraction Analysis R e s u l t s ............... 34 13 Composition of Sample 270-1 .................................. ; 39 ' 1 14 Summary of Electron Microscopy Results for Sample Libby, Grace, Grade 1 ............................................... 40 # 15 Composition of Sample 276-1 ................. 41 * 16 Summary of Electron Microscopy Results for Sample Libby, Grace, Grade 2 ............................................... 42 v Number 17 18 19 20 21 22 23 24 25 26 27 ,, 28 29 30 31 32 33 34 35 TABLES (continued) Page Composition of Sample 2 5 9 * 1 ................................. 42 Summary of Electron Microscopy Results for Sample Libby, Grace, Grade 3 .............. .......................... .. 43 Composition of Sample 262*1 .................................. 44 Summary of Electron Microscopy Kesults for Sample Libby, Grace, Grade 4 .............................................. 45 Compos ition of Sample 264*1 ................................. 46 Summary of Electron Microscopy Results for Sample Libby, Grace, Grade 5 .............................................. 47 Composition of Sample 267*1 .................................. 48 Composition of Sample 2 9 1 * 1 ..................... 49 Summary of Electron Microscopy Results for Sample Libby, Grace, Head Feed................................... 50 Composition of Sample 294*1 ................................. 51 Summary of Electron Microscopy Results for Sample Libby, Graced Extractor. . ,, . . . ...... f . . . . 52 Composition of Sample 297*1 ................................. 53 Summary of Electron Microscopy Results for Sample Libby, Grace, Mill Dust........................................... 54 Composition of Sample 288-1 ....................... 55 Summary of Electron Microscopy Results for Sample Libby, Grace, Screening D u s t ..................................... 56 Composition of Sample 430*1 ................................. 57 Summary of Electron Microscopy Results for Sample Enoree, South Carolina, Grace, Grade 3 .......... . .............. 57 Composition of Sample 433*1 ....................... . . . . . 58 Summary of Electron Microscopy Results for Sample Enoree, South Carolina, Grace, Grade 4 ............................. 59 vi TABLES (continued) Number 36 Composition of Sample 427-1 .............................. 60 37 Summary of Electron Microscopy Results for Sample Enoree, South Carolina, Grace, Grade 5 ................. .. 61 38 Composition of Sample 436-1 .............................. 62 39 Summary of Electron Microscopy Results for Sample Enoree, South Carolina, Bead Peed + 100 M e s h ................... 63 40 Composition of Sample 439-1 .............................. 63 41 Summary of Electron Microscopy Results for Sample Enoree, South Carolina, Grade 3, Commercial Exfoliation . . . . 64 42 Composition of Sample 442-1 .............................. 65 43 Composition of Sample 573-1 .............................. 66 44 Summary of Electron Microscopy Results for Sample Enoree, South Carolina, Patterson, Ungraded................... 67 45 Results of the Phase Contrast Analysis of Air Semples Collected at Three Vermiculite Sites ................. 68 vii ABBREVIATIONS, DEFINITIONS, AND SPECIFICATIONS The following ere special terms or specifications that are used in this report. 1. Vermiculite - A naturally occurring hydrated laminar mineral silicate. Due to layers of water of hydration between the laminae, the vermiculite ex foliates or expands when heated. ' 2. Beneficiation - The process of obtaining vermiculite particles from the ore. 3. Vermiculite grades - W. R. Grace, the largest U.S. vermiculite pro . ducer, separates the beneficiated vermiculite into five size grhdes. Grade 1 is the largest; Grade 5 the smallest. Company grade specifications were not obtained, but the following size data were determined by examination of the five grades from Libby, Montana. The data are presented as an indication of differences among the grades. Grade No. Approximate maximum dimension (mm) Approximate number of particles/g Approximate weight/ average particle 1 5-10 23 2 3-5 130 3 1-3 1,700 4 0.5-1 11,000 5 0.2-0.5 130,000 42 mg 7.4 mg 0.5S mg 91 Mg 7.6 MS 4. Asbestos - A general term for a number of naturally occurring fibrous mineral silicates. Asbestos falls into two major classes, the serpentines and amphibol.es. Chrysotile is the generally encountered serpentine. The various amphiboles are not easily identified using the electron microscope, and the entire group is generally reported as "amphiboles" from the EM analysis. -(With greater effort the chemical composition, can be determined with the mi croprobe.) The major fibers identified by optical microscopy in this study were amphiboles of the tremolite/actinolite series. Tremolite and actinolite dif fer by the ratio of iron and magnesium in the molecule, which results in a range of refractive indices. The series is comprised of a continuous varia tion with an arbitrary division between the two. The composition of tremolite ranges from Ca2MggSi022(0H )2 to Ca2Mg<FeSis022(0H )2 and that of actinolite ranges from Ca2Mg<FeSig022(0H)2 to Ca2MgFe4Sig022(0H)2. Any composition within this series is often reported simply as "tremolite/actinolite." viii E M - Electron microscpy TEH - Transmission electron microscopy SAED - Selected area electron diffraction JCFDS * Joint Committee on Powder Diffraction Standards HRI - Midwest Research Institute, the prime contractor. IITRI - IIT Research Institute 20 V. 35th Street Chicago, Illinois 60626 OKF - Ontario Research Foundation Sheridan Park Research Community Mississauga, Ontario, Canada L5K 1B3 iz SECTION 1 INTRODUCTION In December 1978, the vermlculite industry submitted information to the EPA regarding health problems experienced by employees who were processing ashestos-contaminated vermlculite. The original submission indicated that bloody pleural effusions had been detected in A of 350 employees; symptom atology and clinical findings in the employees were similar to those found in individuals with asbestos-related diseases. Subsequent follow-up studies by the Occupational Safety and Health Administration (OSHA) revealed an even higher prevalence of health problems among the employees. Vermiculite, mined in the United States since 1929, is a hydrated magnesium-iron-aluminum silicate and is often contaminated with asbestiform minerals. After mining, vermiculite is processed to remove impurities; how ever, some impurities, including asbestos, may remain in processed vermicu lite. Although vermiculite may contain fibrous materials, the health effects from vermiculite itself are unknown. A priority review of asbestoscontaminated vermiculite, completed by the Office of Testing and Evaluation in June 1980, suggested that the asbestos in vermiculite may be responsible for the reported adverse health effects, and it concluded that certain infor mation gaps needed to be filled before an in-depth risk assessment on vermicu lite could be initiated. The available information on the composition of commercial vermiculite indicated that asbestos contamination of vermiculite does occur but that the degree and kind of contamination might be difficult to assess and might vary with the source of the vermiculite. Therefore, the objective of this task was to sample and analyze vermiculite to determine the contaminants, particu larly the amount of asbestiform minerals present. The study was to provide . information to be used in the assessment of the risk to the population ex posed to asbestiform minerals from vermiculite at each of the various stages of its commercial distribution. The original task objective was divided into two phases. The first phase was to conduct an in-depth analysis of fibers present in, and associated with, vermiculite ore concentrates and beneficiated vermiculite from the major vermiculite mines in the United States and beneficiated vermiculite from the ports of entry. Both bulk and air samples were to be collected and analyzed. The second phase was to have been a similar analysis of bulk and air samples from a representative number of exfoliation plants in the United States. Be cause of a shift of priorities within EPA, the scope of the task was reduced. The task was limited to the collection of air and bulk samples from three U.S. 1 mines (W. R. Grace in Libby, Montana; W. R. Grace in Enoree, South Carolina; and Patterson in Enoree, South Carolina). The air samples were analyzed only by phase contrast microscopy, and the originally planned electron microscopic analysis vas omitted. With the reduction of effort, a set of the bulk samples that vas con* sidered to be representative of each mine vas selected as "priority" samples for immediate analysis. This set included the head feed for the ore process ing mill and, where size grades were produced, the smallest and mid-size grades. This set, representing Eeven samples, vas analyzed by various tech niques including electron microscopy for fiber content, with emphasis on asbestiform minerals. The analysis vas done by two independent laboratories. It vas considered possible that fibers could be bound between the vemiculite plates and that fibers could be released with exfoliation. Therefore, analy ses were conducted both on the samples as received and after laboratory ex foliation. Laboratory exfoliation differs from commercial exfoliation in that under the conditions of commercial exfoliation much of the fines and heavies are removed from the vermiculite. The laboratory exfoliation is done under conditions that produce no sample fractionation. Thus, much of the asbestos would be removed from the vermiculite during commercial exfoliation, but none would be removed during laboratory exfoliation. The analyses of these samples were in various stages of completion when the decision vas made to reduce the scope of the task. Analytical results are available for most of this 6et, but for 6ome samples the complete set of results from both laboratories vas not obtained. This report includes material compiled from reports submitted by Ontario Research Foundation end IIT Research Institute. The remainder of this report presents the summary, experimental methods, sampling, sample handling, analyt ical results, and appendix. Three additional volumes of appendices contain the detailed analytical results. 2 SECTION 2 SUMMARY In December 1978, the v emiculite industry submitted information to EPA regarding the health problems experienced by employees vho vere processing asbestos-contaminated vermiculite. A priority review of asbestos-contaminated vermiculite, completed by the Office of Testing and Evaluation in June 1980, suggested that asbestos in vermiculite may be responsible for the reported adverse health effects, and it concluded that certain information gaps needed to he filled before an in-depth risk assessment of vermiculite could be ini tiated. . . The objective of this task was to develop the protocol and to conduct s a i l i n g and analysis to determine the composition of vermiculite with empha sis on the content of asbestifoim minerals. The original scope of the study included two phases. The first phase va6 for the collection and analysis of air and bulk samples associated with vermiculite ore and beneficiated vermiculite from the four major U.S. mines and ports of entry. The second phase was for a similar effort for a repre sentative number of exfoliation plants. Due to priority shifts within EPA, the second phase was not undertaken and the scope of the first phase was reduced. Three mines and benefication plants were sampled and the samples analyzed, but the scope of the analysis was reduced from the original protocol. The air sample analysis was limited to phase contrast optical microscopy for a selected set of the samples, and the electron microscopic analysis of the air samples was not performed. The bulk samples were analyzed by optical microscopy and x-ray diffrac tions (XRD) for the density separated fractions and by transmission electron microscopy (TEM) for the io&propanol suspended fractions of nonexfoliated vermiculite and the water suspended fractions of exfoliated vermiculite. The results of the optical microscopy and XRD analyses are summarized in Table 1. The bulk sample TEM analyses vere conducted in a semirandom sequence, but with the sequence ordered for the two analytical laboratories so that each of a set of priority samples would be analyzed by at least one of the laboratories early in the program. Vhen the program scope was reduced, a set of seven priority samples was selected to best represent the two largest mine operations. However, the sample for TEM analysis undergoes many steps of preparation for analysis, and when the program scope was reduced, several low priority samples had been pre pared for analysis. Authorization was later obtained for the completion of 3 TABLE 1. SUMMARY OF OPTICAL MICROSCOPY/XRD ANALYSIS RESULTS Sample8 Fibrous phases Estimated Mineral mass, % types Nonfibrous amphiboles Estimated Mineral mass, % types Libby Grace Grade 1, 270-1 Grade 2, 276-1 Grade 3, 259-1 Grade 4, 262-1 Grade 5, 264-1 Grade 5 (1-day), 267-1 Head feed, 291-1 Extract, 294-1 Baghouse mill, 297-1 -Screen plant, 268-1 S.C. Grace Grade 3, 430-1 4-6 4-7 2-4 0.3-1 2-4 2-5 21-26 1-4 8-12 2-5 < lb Grade 4, 433-1 < lb Grade 5, 427-1 < lb Mill feed (+100 mesh), 436-1 Grade 3, expanded, 439-1 < 1 <lb Grade 4, expanded, 442-1 < lb S.C. Patterson Ungraded, 573-1 < 1 Trem-actin Trem-actin Trem-actin Trem-actin Trem-actin Trem-actin Trem-actin Trem-actin Trem-actin Trem-actin 1-3 Trem-actin 3-5 Trem-actin < 1 Trem-actin 1-3 Trem-actin 2-5 Trem-actin 4-8 Trem-actin < 1 Anthophyllite^ ' * 6-9 Trem-actin 1-3 Trem-actin 2-6 Trem-actin 1-4 Trem-actin Mixed Anthophyllite Trem-actin Mixed Anthophyllite Trem-actin Mixed Anthophyllite Trem-actin Mixed Anthophyllite Trem-actin Mixed Anthophyllite Trem-actin Mixed Anthophyllite Trem-actin 2-4 Trem-actin < 1 Anthophyllite 1-3 Anthophyllite 1-4 Trem-actin 4-67 .) Anthophyllite 2-4 W * Trem-actin - 1-3 7 6-9 ^ Anthophyllite 'Trem-actin < 1 Anthophyllite < 1 Trem-actin < 1 0.5-1 Anthophyllite Trem-actin Mixed Trem-actin Anthophyllite' 4-8 7 -^Anthophyllite 8-12 Trem-actin a With the exception of Sample No. 267*1, all results are for composite samples. - b Fiber bundles were mixed phase materiaIs-*both anthophyllite and tremolite-actinolite were present. 4 analyses of camples nearly completed. The analysis data are complete from both laboratories for the seven priority samples, but the data may or may not be complete for the others. A difference in the Interpretation of the analytical protocol resulted in a variation in the counting procedure. The requirement to count 100 fibers vas interpreted by OBI to mean 100 aEbestiform fibers, while IITRI counted 100 particles, defined as fibers by their aspect ratio of equal or greater than 3. To check the significance of this counting variation, two camples with different fiber characteristics (the grade 5 samples from Libby, Montana, and Enoree, South Carolina) were selected for each laboratory to re* peat the analysis using the alternate procedure. Table 2 is a summary of the TEM analysis of the selected samples and includes the number and parts per million of fibers as determined by the two laboratories. The results suggest that there are more asbestiform fibers associated with the smaller size grades of vermiculite than with the larger grades. Both dust samples collected at Libby were found to have a very high amphibole con* tent and indicate that considerable asbestos is removed from the vermiculite during beneficiation. The South Carolina vermiculite appears to contain sub* stantially less asbestiform fibers than does that from Libby, Montana. Table 3 is a summary of the phase contrast results of the air samples. Only one of the analyzed air samples exceeded 2.0 fibers/cc. However, the rainy weather conditions at the time of sampling for all three locations might have resulted in lover than normal fiber counts. Given the expected variability of the method, IITRI and 0RF results ap* pear to be in general agreement. ' 5 Sample8 TABLE 2. SUMMARY OF ELECTRON MICROSCOPY ANALYSIS . Priority0 cample Analysis,c exfoliated no yes Asbestiform fibers, all lengths Amphibole Chrysotile Fibers/g Mass Fibers/g Mass z 106 (ppm) z 10 (ppm) Libby Grace Grade 1 270-1 Grade 2 276-1 Grade 3 P 259-1 259-0 259-1 259-0 Grade 4 282-0 282-1 282-0 Grade 5 P 264-1 264-0 264-1(0) 264-0(1) ?64-I 264-0 264-1(0) ' r 264-0(1) Head feed P 291-1 291-0 291-1 Eztractor 294-1 Mill dust 297-0 297-1 Screening dust 288-0 288-1 X 31.6 78 0.9 3.5 z 10 X 23.4 48.5 0 0 X X X X 38.9 25 42.0 59 210 0.9 0.01 59 < 2.1 250 0.4 6.1 x 10" 240 < 1 - X 1 1 X 65 460 0 0 X 1.8 17 < 0.4 - X 118 840- - - X 100 600 < 1.4 - X 127 1,200 - - X 98 570 - - X 142 2,600 - - X 160 1,800 < 1.6 - X 119 f 350 - '' - X 110 2,600 < 1.6 - X 62.5 670 1.4 X 130 690 1.2 X 73.8 590 - 0.13 < 1 - X 55.0 420 0.7 3.4 z 10 X 100 4,600 - - X 777 35,000 - - X 300 3,000 < 1.6 - X 1,800 41,000 - (continued) 6 Sample Priority sample TABLE 2 (continued) Analysis, exfoliated no yes Asbestiform fibers, all lengths Amphibole Chrysotile Fibers/g Hass Fibers/g Hass x 106 (ppm) x 106 (ppm) S.C. Grace Grade 3 P 430-1 X 430-0 430-1 430-0 X X X Grade 4 433-1 433-0 433-1 433-0 X X X X Grade 5 P 427-1 427-0 427-1(0) 427-0(1) 427-1 427-0 427-1(0) 427-0(1) X X X X X X X X Head feed P . 436-1, 436-0 436-1 . X X X Grade 3 exfoliated 439-1 S.C. Patterson beneficiated Ungraded P 573-1 573-0 573-1 573-0 X X X X 1.0 2.7 3.1 2.4 1.6 2.7 3.1 2.7 0.6 17 3.0 31 3.5 2.9 3.2 2.4 i 0.3 12 1.3 11.7 0.55 0.1 < 1 < 0.3 3.7 - 1 < 0.5 6.5 35 1.4 2 - < 0.3 - < 0.3 1.5 37 4.8 130 4.1 120 7.3 9 - 2.6 0.07 2.6 m < 0.3 - 0.9 0.49 22 0.81 1 0.3 - - - 5 x 10 4 - m - m 1 x 10 4 < 1 - - < 1 < 1 m m 0.03 1.7 0.5 1.1 L 3.7 x 10 'H 0.03 27 < 0.3 3 0.2 4 < 0.3 . -4 1.4 x 10 * 5 . 3 * 10'3 a The "I" and "0" following the sample number indicates the analyzing labora tory, IITRI and ORF, respectively. The "(I)" and "(0)" indicates the counting procedure, e.g., 264-1(0) are the results from IITRI using the ORF procedure. b Seven samples were designated as priority samples for complete analysis at the time the program was reduced in scope. c Analysis was conducted on the samples as received and following laboratory exfoliation, which unlike commercial exfoliation, does not cause sample fractionation. 7  f 1 1 1 ! If 11 i 1I j ir V I1 ir ? t t t f t i i TABLE 3. RESULTS OF THE PHASE CONTRAST ANALYSIS OF AIR SAMPLES COLLECTED AT THREE VERMICULITE SITES Sample Sample vol. () Fibers/cc ORF 1ITR1 Libby, Grace 106 Field blank8 m 133 Field blank8 - 131 Front loader 303 148 Pit haul driver 297 138 Mine analyst 294 141 Bottom operator 276 130 No. 2 operator 285 139 Dozer operator 270 101 Shuttle truck 385 104 Screening plant, DV 390 111 Screening plant, DV 368 108 Trailer court 169 136 No. 5 substation 111 South Carolina, Grace 312 Field blank8 346 Field blank8 - 340 Hill monitor 340 321 Hill lab technician 478 301 Dragline operator 240 347 No. 4 bagger - 314 330 No. 3 bagger 285 328 Hill (ENE) downwind 287 335 Hill (N) crossvind 80 307 Hine (N) crossvind 291 323 Hine (E) downwind 154 338 Hine (V) upwind 264 310 Truck driver 257 300 Screening plant floor 354 South Carolina, Patterson 505 Field blank8 533 Field blank8 508 Payload operator 520 Plant foreman 542 Bagger/forklift 513 (NE) downwind 506 Control off-site 515 (SE) crossvind 528 (SW) upwind - 255 252 ' 249 188 274 299 147 < 0.02 0.03 0.02 <0.01 1.5 1.2 3.1 0.02 0.1 0.08 0.1 0.03 0.03 0.04 0.05 0.04 0.01 1.9 0.4 9.7 0.2 0.2 0.5 0.g2 nr 0.02 < 0.02 < 0.02 0.03 0.07 < 0.01 0.06 0.1 0.05 0.04 < 0.01 0.01 0.03 < 0.01 0.06 0.04 0.02 0.03 0.2 Hr 0.1 0.05 0.04 Hr 0.02 0.02 0.01 0.3 0.14 < 0.02 < 0.02 < 0.01 0.01 < 0.01 < 0.01 < 0.01 0.01 0.02 < 0.01 0.02 0.04 0.3 H ND* ND O.gl ND a Values for blanks were calculated assuming a 100-liter sample, b ND: No fibers detected (100 grids). 8 SECTION 3 EXPERIMENTAL PROTOCOL A study protocol was prepared for the task and reviewed by EPA and the subcontractors before sampling and analyses were undertaken. Appropriate codifications were cade following additional reviews. The protocol "Task 32 Study Protocol for the Collection and Analysis of Vermiculite and Related Sam* pies for the Evaluation of Fiber Content with Emphasis on Asbestiform Fibers" appears in this volume as Appendix A. A detailed analytical procedure for bulk samples prepared by IITRI appears as Appendix B. During the program some modifications from the study protocol were found to be necessary or desirable, end minor changes were cade. This section dis cusses the general protocol briefly, with major emphasis on areas where modi fications were cade. SAMPLE COLLECTION No major changes were made in the sampling protocol. The following items are noted: . ' * i* ' * 1. W. R. Grace' representatives would not allow-BRI personnel into any of the processing facilities. All samples, both bulk and air, from within the processing facilities were collected by EPA personnel. 2. Ore processing at Libby, Montana, Involved a wet beneficiation pro cess, while both facilities in South Carolina used dry beneficiation processes. The differences in processes resulted in different types of waste materials. 3. V. R. Grace has automatic sampling equipment at various places in their processing for their QA program. Portions of these samples were ob tained that represented 7 to 10 days of operation before our sampling. One day of automatic sample collection was observed by the sampling crew EPA rep resentative. SAMPLE HANDLING Bulk Samples The bulk samples were packed in double sealed bags in the field and shipped by air freight to MRI. The increment samples were riffle divided, and approximately equal portions of each increment of the same sample type were combined, mixed, and riffle divided to obtain replicate composite samples for analysis. 9 Air Samples The air sample filters were retained in the filter cartridges during transport to HRI. The plugged cartridges vere placed in a special container to maintain the filters in a horizontal position with the collecting surface up and hand-carried back to HRI. At HRI the filters vere cut into three equal portions and each portion individually taped to the bottom surface of a 49 x 9 mm HilliporeG plastic petrie dish. A set of one-third of each air sam ple filter vas hand-carried to the two laboratories for analysis. SAMPLE ANALYSIS Bulk Sample Analysis Since no microscopy technique is capable of measurement over the whole size range of fragments present in vermculite samples, it is of extreme impor tance, prior to selection of the analysis procedure, to understand precisely hov the analytical results are to be applied. The basic choice in the analysis of vermiculite vas either to completely pulverize the material and reduce the particle sizes into a range suitable for a single analytical technique, or to retain the original size distribution and measure relevant parameters on the material as normally used. Using the lat ter approach, numerical fiber counts per unit mass of original material are meaningful and assist interpretation on the basis of current medical opinion that fiber numbers are the important exposure criterion. If the materiel 1e completely crushed, there are a number of disadvantages: 1. The fiber size distribution is not preserved, and any numerical fiber count is meaningless except that fiber volume can be considered as an indica tion of the mass percentage of fiber in the original -sample material. 2. Even the mass value thus obtained is not representative of that in the final product, since at exfoliation much of the massive material is sepa rated and discarded. 3. Simple E-ray diffraction (XRD) measurements of the amphibole or ser pentine content of such a pulverized sample i s of inadequate sensitivity (about 1% for amphiboles and possible 5% for serpentine). Moreover, XRD is incapable of distinguishing the fibrous varieties from other amphiboles or serpentine. 4. The crystallography of the fibers may be altered ( S p u m y et el., 1979).1 The procedure of Ghatfield and Lewis (I960)2 vas designed to retain the size distribution of the material as it i6 normally used, and to allow very . sensitive measurement of asbestos fiber concentrations down to detection lim its in the parts per million (ppm) region. 10 Essentially, their procedure was: 1. To suspend the beneficiated vermiculite in water and sample, for transmission electron microscope (TEM) analysis, only the range of particle sizes which would include all respirable fibers. 2. To simulate on a laboratory scale the industrial exfoliation proce dure, and to examine by TEM the fraction which does not float on water. The floating fraction would in fact be the final product. If fibers have been found in the earlier analyses, the floating fraction could also be examined to determine its fiber content. 3. To examine typical vermiculite flakes for the presence of intercalated fibers which may be released on exfoliation. If there are no very large fibers present, the assumption can be made that any fibrous component has been sampled representatively from the aqueous suspension, and the results can be interpreted as total fiber concentrations by weight. Vhere large amounts of asbestos fibers are present throughout the whole 6ize spectrum, the procedure introduces a size cut-off above which no particles are included in the analysis. Under these conditions the concentra tion by weight must be interpreted carefully, although concentrations by number will be almost unaffected. It is important when using this method that the size cut-off established does not restrict the representative sampling of the largest fibers considered to be respirable. Timbrell (1965)* has determined that the free falling speed of high aspect ratio fibers is proportional to the square of the diameter and only increases slowly with length. The largest compact particles normally found in lungs are about 10 micrometers (pm) in diameter (unit density), which as a first approximation was found to be equiv alent to a fiber of about 3.5 pm in diameter, whatever its length may be. Hence, the 6ize cut-off in the analytical method should exceed a unit density equivalent spherical diameter of 10 pm, which corresponds to a sphere of 5.6 pm diameter if the density is assumed to be 3.2 g/cm3 . The falling velocity of a sphere is obtained from the Stokes' relationship: v c *d2 (ps - pL) v 18n vhere V = terminal velocity g = acceleration due to gravity d = diameter of the sphere ps = density of the sphere . pL = density of the liquid r) s coefficient of viscosity of the liquid For a sphere of 5.6 pm diameter and density of 3.2 g/cm3 the terminal velocity in water is calculated to be 0.0037 cm/sec, or 270 secfa n . Under the agitation conditions in the ultrasonic bath, it is unlikely that particles of this low falling velocity will deposit during the period when representative samples of the dispersion are withdrawn for analysis. Accordingly, it can be stated that the method yields a fiber count which includes all fibers consid ered to be respirable. 11 The analytical procedure described above does not yield an actual total fiber content by weight where very large fibers are present, and it is for this reason that the initial step of a low magnification optical examination was incorporated. In this way such samples can be detected before effort is expended on TEM fiber counts which may be irrelevant. However, the TEH proce dure must still be used if determination of the respirable fiber concentration is required. Both analytical laboratories contributed to the preparation of the adopted protocol and both laboratories followed the protocol. However, there were variations in emphasis and interpretation of the protocol by the two lab oratories, and these differences were not recognized until some of the results were obtained. While to a degree the variations in procedures prevent the di rect comparison of results, the slightly different approaches complement each other and give a better overall understanding of the samples than would either single approach. The significant differences were as follows: 1. ORF examined the bulk samples by optical microscopy for the presence and qualitative identification of asbestiform fibers. IITRI performed more complete qualitative and semiquantitative analysis of the bulk sample using density fractional separation, followed by component identification by optical microscopy and X-ray. . 2. The following appeared in the protocol for the TEH analysis: "Hake fiber count - determine chrysotile or amphiboles. Count 100 fibers or 10 grids of 200-mesh screen. Determine the limitE of detection and count more grids if necessary." ORF interpreted this statement to mean count 100 chryso tile or amphibole fibers; IITRI interpreted the statement to mean count 100 particle units with an aspect ratio of equal to or greater than 3. To deter mine the effect of the difference of counting procedure between the two labor atories, two samples were selected for cross comparison. For these two sam ples each laboratory examined the sample by the other's procedure as well as their own. The two samples were selected to represent a high and low concen tration of asbestiform fibers (Grade 5 from Libby and Grade 5 from South Carolina Grace). Air Sample Analysis Due to a change in the scope of the task, the analysis of the air sample filters by TEH was not undertaken. The optical phase contrast analysis was conducted according to the protocol. 12 SECTION 4 SAMPLING Sampling trips were cade to the Grace mine and processing facilities near Libby, Montana, daring October 21-26, 1960, and to both the Grace and Patterson mines and processing facilities near Enoree, South Carolina, during November 3-6, 1980. Both air samples and bulk samples were collected at each location. Air sampling vas of two types, personal and stationary. For the personal sam ples, nine Dupont Hodel 4000 samplers were used. The flow rates vere cali brated before and after sampling. Stationary air sampling vas conducted us ing battery-powered stationary samplers designed by MRI. These samplers have proven to be effective in previous air sampling projects.4,6,6 Wind condi tions during sampling vere recorded using a Vang meteorological station. Brief descriptions of sampling conditions and a list of samples collected fol low. V. R. GRACE MINE, LIBBY, MONTANA ' Hr. Fred Eaton of V. K. Grace, Cambridge, Massachusetts, vas the company representative for the coordination of sampling. Mr. Jim Salois of MSEA, Helena, Montana, was present at the request of Ms. Diana M. Kraft, MSEA, Arlington, Virginia.:Mr. Salois vas familiar with the Libby facilities, and his presence and suggestions vere very helpful. At the request of.Mr. Eaton, duplicate concurrent personnel air samples vere taken, one for this task and one for Grace. Thus each subject vas fitted with two samplers. Mr. Eaton also requested that all personal air sampler pumps used for this task be recalibrated at the Libby facility even though they had been calibrated just before shipment from MRI. The nine pumps used vere determined to have flow rates ranging between 2.03 and 2.19 liters/min. The flow rates for the stationary samplers vere measured at the time the sam plers vere set up, periodically during campling, and at the end of sampling. . ..r The specific flow rates, calibration data, and related information appear in Appendix C of this report. Sampling vas scheduled and conducted on Thursday, October 23, 1980. For several days prior to sampling the weather had been rainy, and sampling vas started in heavy fog with essentially no wind in the cine area. There vas no evidence of dust in the mine, from either the mining operation or along the truck routes. The weather cleared shortly after noon. The wind direction and speed vere recorded during the sampling day. 13 The objective vas to take a short (2-hr) sample, followed by a longer (6-hr) sample that would complete the work shift. The actual times varied somewhat from the intended times, but the actual times and volumes for each sample were recorded (Appendix D). Grace has a routine bulk sampling procedure els part of their product quality control program. At our request they had taken and retained bulk sauries of the five grades of product plus related head feed, tailing6, and dusts for 7 to 10 days before air sampling. The samples taken on October 23 were comparable to the earlier samples, but their collection vas observed and verified by Tom Dixon of EPA.* The air samples collected at Libby, along with the approximate sampling duration, are given in Table 4. Figures 1 and 2 show the vind conditions and 6ite positions for the stationary air samplers. The bulk samples collected at Libby are given in Table 5. V. R. GRACE MINE AND PROCESSING MILL, ENOREE, SOUTH CAROLINA Mr. Fred Eaton, vho vas the company representative at Libby, Montana, vas also the company -representative at Enoree. Ore from two mines (Lanford and Foster) are hauled to the processing mill at a third location. During the sampling period only the Foster mine vas in operation. The Foster mine is located near the southwest corner of the junction of County Road 50 and Interstate 26 in Spartanburg County. The mill is located on Highway U.S. 221 about 1 mile 6outh of the junction with Highway 92, in Laurens County. The initial schedule was to sample at the mine on Tuesday, November 6, 1980, and at the mill on Wednesday. However, because of rain during Tuesday morning, the mine vas closed and the schedule vas reversed. A light rain fell Tuesday morning; the remainder of the sampling period vas clear and cool. The air samples collected at^the Grace Enoree Table 6 and the bulk samples are given in Table 7. air sampling site positions are shown in Figures 3 Figures 5 and 6 for the mine. operations are given in The vind conditions and and 4 for the mill and in PATTERSON VERMICULITE COMPANY, ENOREE, SOUTH CAROLINA * The Patterson mine and exfoliation/bagging operations are located approx** imately 7 miles northeast of the W. R. Grace mill. No mining vas underway on the day of sampling, November 6, 1960, 60 sampling vas only conducted around the processing plant. Patterson does not size their product and produces a single size grade. The air samples collected at the Patterson plant are listed in Table 6 end the bulk samples in Table 9. The vind conditions and air sampling site positions are shown in Figures 7 and 6. * MRI employees were excluded from all mill operations. 14 TABLE 4. AIR SAMPLES COLLECTED AT THE GRACE MIKE AND MILL. LIBBY. MONTANA Sample description Filter No. Approx. time (hr) Analysis assignment FC Em Hold Peconnel samplers Front loader Front loader Fit driver Pit driver Hine analyst Ifine analyst mil operator, bottom Hill operator, bottom Hill operator No. 2 mil operator No. 2 Bulldozer operator Bulldozer operator Shuttle truck driver Shuttle truck driver Stationary samplers** Station 7 screening plant D.V. Station 7 screening plant D.V. Station 7 screening plant D.V. Station 7 screening plant D.V. Station 6 screening plant D.V. Station 6 screening plant U.V. Station 2 perimeter D.V. Station 2 perimeter D.V. Station 2 perimeter D.V. Station 2 perimeter D.V. Station A perimeter C.V. Station A perimeter .V. Station 5 lover meadow Station 5 lover meadow Station 9 trailer court Station 9 trailer court Station S car loading Station 8 car loading Station 3 "22" level dump Station 3 "22" level dump Station 1 substation No. 5 Station 1 substation No. 5 Field blanks 131 2 X 135 5 X 1A8 2 X 126 5 X 138 2 X 129 5 X lAl 2 X 146 5 X 130 2 X 125 5 X 139 2 X 128 5 X 101 2 X 121 7 X 104 2 X 111 2 X 112 6 X 120 6 X 116 2 X 12A 6 X 109 2 X 113 2 X 1A5 6 X 1A7 6 X 103 2 X 1A9 2 X 119 2 X 115 6 X 106 2 X 102 6 X 123 2 X 122 3 X 107 2 X 134 6 X 136 2 X 132 6 X 114 X 137 X 106 X 133 X 110 X a PC - phase contrast optical microscopy, EM - electron microscopy, Hold - sample retained without analysis. b Station numbers indicated in Figure 2. 15 Figure 1. Wind rose pattern showing the direction and intensity of the wind during the air sampling period at the . Grace, Libby, Montana, facility. 16 1 I I I I I I t Figure 2 Hap of the Grace, Libby, Montana, facility shoving the stationary air sampling locations. i % 17 4 TABLE 5. BULK SAMPLES COLLECTED AT THE GRACE MINE AND MILL, LIBBY, MONTANA Sample description No. of samples Beneficiated Grade 1 vermiculite lla Beneficiated Grade 2 vermiculite lla Beneficiated Grade 3 vermiculite lla Beneficiated Grade 4 vermiculite lla Beneficiated Grade 5 vermiculite lla Dust from screening plant eb Dust from dryer 9C Head feed 9C Under 90 me&h 9C Coarse tails 9C Extractor 9C a Days collected for 11-day samples: October 7, 8, 9, 10, 13, 14, 15, 16, 17, 21, 23, 1980. b Screening plant dusts collected: October 6, 9, 10, 13, 14, 15, 17, 23, 1980. c Collected October 8, 9, 10, 13, 14, 15, 16, 17, 23, 1980. 18 TABLE 6. AIR SAMPLES COLLECTED AT GRACE, ENOREE, SOUTH CAROLINA Sample description Mine personnel samples Truck driver 1 Truck driver 1 Truck driver 2 Truck driver 2 Dragline operator 6 Dragline operator 6 Hine stationary samples^ Crosswind N. station 1 Crossvind N. station 1 Crossvind S. station 3 Crosswind S. station 3 Upwind W. station A Upwind V. station 4 Downwind E. station 2 Downwind E. station 2 1 mile offsite station 5 1 mile offsite station 5 Along haul route Along haul route Hill personnel samples Forklift operator Forklift operator Bagger Bagger ' Bagger Bagger Bagger Bagger Hill monitor Hill monitor Hill laboratory technician Hill laboratory technician Hill stationary samples^ Hill office Screening floor Screening floor Screening floor Crosswind N. station 1 Crosswind N. station 1 Filter No. Approx. time (hr) Analysis assignment PC EM Hold .. . 315 2 X 324 2 X 310 2 X 320 5 X 301 2 X 306 6 X 351 5 X 318 4 X 307 2 X 353 5 X 316 4 X 338 2 X 323 2 X 352 6 X 334 2 .X 350 5 X 331 2 X 354 5 X 305 2 X 339 4 X 314 2 X 322 4 X 330 2 X 349 4 X 347 3 X 337 4 X 336 3 X 340 3 X 308 3 321 4 X 304 7 X 300 2 X 332 3 X 341 1 X 345 2 X 343 5 X (continued) 19 TABLE 6 (continued) Sample description Filter Ho. Approx. time (hr) Analysis a fcc EM Hold Mill stationary samples Crossvind N. station 1 Crosswind N. station 1 Crosswind S. station 3 Crosswind S. station 3 Upwind V. station 4 Upwind U. station 4 Downwind E. station 2 Downwind E. station 2 (continued) Offsite control Field blanks 328 2 X 313 5 X 342 5 X 326 2 (void) 344 X 309 5 X 335 2 X 302 5 X 329 7 X 312 X 346 X 319 X 348 X 327 X a PC - phase contrast optical microscopy, EM - electron microscopy, Hold * sample retained without analysis. b Station numbers indicated in Figure 4. 20 TABLE 7. BULK SAMPLES COLLECTED AT TEE PATTERSON OPERATIONS ENOREE, SOUTH CAROLINA Sample description Raw ore from stockpile Raw ore prescreening hopper Postscreen ore Dried ore Exfoliated final product Waste from screening Waste from exfoliater No. of samples 1 1 3e e 4a 4e 4* a Samples were collected at approximately 2-hr intervals. 21 I 1 I I ( t I I I I 1 I I I I I I I i Figure 3. Wind rose pattern showing the direction and intensity of the wind during the air sampling period at the Grace, Enoree, South Carolina, mill. 22 Figure 4. Map of the Grace, Ecoree, South Carolina, mill area showing the stationary air sampling locations. 23 I I ! I f i* ! f 4 I i Figure 5. Wind rose pattern shoving the direction and intensity of the wind during the air sampling period at the Grace, Enoree, South Carolina, mine. 24 Rood 50 Figure 6. Hap of the Grace, Enoree, South Carolina mine area showing the stationary air sampling locations. TABLE 8. AIR SAMPLES COLLECTED AT PATTERSON, ENOREE, SOUTH CAROLINA Sample description Filter No. Approx. time (hr) Analysis assignment* PC EM Hold Personnel samplers Bagger/forklift operator Foreman Payload operator Stationary samplers** . Crosswind S.E. station 2 Crosswind N.W. station A Upwind S.W. station 3 Downwind N.E. station 1 Remote, control Field blanks 5A2 2 X 504 3 X 517 3 X 520 2 X 521 3 X 511 3 X 508 2 X 519 3 X 516 3 X 515 3 X 502 6 X 503 2 X 518 6 X 531 2 X 5A0 6 X 528 2 X 525 6 X 513 2 X 523 6 X 506 2 X 527 6 X 505 X 522 X 533 X 538 X a PC - phase contrast optical microscopy, EM - electron microscopy, Hold - sample retained without analysis. b Station numbers indicated in Figure A. 26 TABLE 9. BULK SAMPLES COLLECTED AT THE GRACE MIKE AMD MILL, ENOREE, SOUTH CAROLINA , Sample description_____________ Beneficiated Grade 3 vexmiculite Beneficiated grade 4 vermiculite Beneficiated grade 5 vermiculite Dryer composite Mill feed +100 mesh Mill feed *100 mesh Vet scrubber discharge composite Composite total tails, November 5 Lanford Mine composite, November 5 Foster Mine Composite, November 5 Exfoliated Grade 4 "stabilized" Exfoliated Grade 4 plain Exfoliated masonry insulation Grade Exfoliated Grade 3 plain ______________________ No. of samples 7a 7a 7a 7a 7a 7a lb 1 1 1 lc lc "coated" lc 1 a Days collected for 7-day samples: October 27, 26, 29, 30, 31, November 1, 1980. b Composite collection for October 28, 29, 30, 1980. c The exfoliated samples vere sampled at the Kearney Expansion plant. 27 I 1 I 1 ! i 1 1 Figure 7. Wind rose pattern shoving the wind direction and intensity of the wind during the air sampling period at the Patterson, Enoree, South Carolina, facility. 28  Mill end Packaging Opt ration / / // // // A / f I I i I J L <--To Highway 221 Sketch of Pcttemon Vermlcullte (Sampled U-6-fiO) A Meteorological Station Q Air Sampling Station CD- Figure 8. Hap of the Patterson, Enoree, South Carolina, facility shoving the stationary air sampling- locations. SECTION 5 SAMPLE HANDLING All of the bulk samples vere shipped air freight to MRI from the sampling sites. The air samples vere hand-carried and maintained in a horizontal, sample-up position. BULK SAMPLES Most of the bulk samples vere collected as increment samples representing a span of time. Composite samples vere prepared for this analysis. To pre pare the composite samples, each increment sample vas riffled to obtain a rep resentative fraction of the increment. Approximately equal veight fractions of each of the increments vere combined to make a composite sample. The com posite sample vas then mixed and riffled to produce four equal samples. One of the fourths vas set aside and retained as a control. One of the fourths vas again riffled to produce tvo fractions, each one-eighth of the original sample. These tvo fractions vere combined vith the other tvo fourths to result in the composite division into 1/4, 3/8, and 3/8 of the original composite. The " 1/4" fraction vas retained at MRI, one "3/8" fracton vas sent to IITRI, and the other "3/8" fraction vas 6ent to ORE for analysis. ; A list of the bulk samples and the increment veights used to prepare the composites are given in Appendix D. AIR SAMPLES The air sample filters vere retained in the filter cartridges during transport to MRI. The top retainer portion of the cartridge vas then removed and a cutting template positioned over the filter to allow the filter to be cut into three equal portions. The template held the filter around the cir cumference of the filter but did not contact the sampling portion of the fil ter. Each portion vas then removed from the sampling cartridge and taped to the bottom surface of a 49 x 9 mm Millipore plastic petrie dish. A 6et of one-third of each air sample filter vas hand-carried to IITRI in Chicago, Illinois, and to ORF in Mississauga, Ontario. 30 SECTION 6 ANALYTICAL RESULTS The analytical results of this program consist of findings from (a) the optical and x-ray diffraction (XRD) data and (b) electron microscopy data for the bulk samples and from the phase contrast optical microscopy data for the air samples. The detailed XRD and EM data are contained in three volumes of appendices. Specific data in the appendices are referenced for each sample in this section. The data obtained for the bulk samples by IITRI from the density sepa rated fractions provided a good overview of the composition and complexity of the samples. These data, including weight percent of three density fractions, and a listing of identified mineral phases in the various samples are pre sented in summary at the beginning of this section. This is followed by the complete results obtained for individual samples. In the IITRI examination the samples are separated into three fractions based on density. The fractions are: .1. Density greater than 2.97 g/cc, sinks in 1,1,2,2-tetrabromoethane (TBE). 2; Density less than 2.97 and greater than 2.76, floats on TBE, sinks in TBE/isopropanol mixture with density of 2.76 g/cc. 3. Density less than 2.76, floats on 2.76 g/cc liquid. , Table 10 is a list of selected related minerals and their densities (specific gravity). From these values it can be seen that the vennicullte would be separated from most of the other materials, and that the materials more dense than vermiculite would be separated into two fractions. The weight percent of materials in each density fraction of samples is given in Table ' 11. The mineral phases identified in each sample fraction analyzed by XRD are listed in Table 12. A key Is provided with this table which groups the minerals according to types, and lists chemical formulas and JCPDS file card numbers for the patterns which were used to identify the mineral species. The raw data obtained in the analyses appear in Appendix I. TABLE 10. SPECIFIC GRAVITIES OF SELECTED MINERALS Mineral Chemical formula Specific gravity Vermiculite Biotite Chrysotile Serpentine Talc Anthophyllite Actinolite Tremolite Ferroactinolite Cummingtonite Grunerite Diopside 1 Hornblende Quartz Olivine (Mg,Ca)o a(Mg,Fe,Al)3 o(Al,Si)<Oio(OH)4 KWgiFeJaCAlSiaOio) (OH) 2 MgsSi20s(0H)4 h g sSi20 (0H)< MgsSi<010(OH)2 (Mg,Fe)7SigO2 2 (0H)2 Ca2(Hg,Fe)Si80 22(0H)2 Ca2MgSifi02 2(OH)2 Ca2FeSig022(0H)2 (Mg,Fe)7Sig02 2 (0H)2 F e7Sig02 2 (0H)2 , CaMgSi206 , i . . . * (Ca,Na)2 S(M,Fe,Al)6Si6(Si,Al)2022(0H)2 Si02 (Mg,Fe)2SiO 2.4 2.8 - 3.2 2.5 - 2.6 2.3 - 2.6 2.7 - 2.8 2.85- 3.2 3.1 - 3.3 3.0 - 3.2 3.2 - 3.3 3.1 - 3.3 3.6 ; 3.2 3.0 - 3.4 2.65 3.27 - 4.37 32 TABLE 11. DENSITY-SEPARATED (AND HAND-PICKED) FRACTIONS PRODUCED Sample6 wt % hand-picked fibers Wt % tetrabromoethane sinks Wt % 2.76 6 inks Wt % 2.76 floats Libby Grace Grade 1, 270-1 Grade 2, 276-1 Grade 3, 259-1 Grade 3, 259-1 duplicate Grade 4, 282-1 Grade 5, 264-1 Grade 5 (1-day), 267-1 Head feed, 291-1 Extractor, 296-1 Eaghouse mill, 297-1 Screen plant, 288-1 S.C. Grace Grade 4, 433-1 Grade 5, 427-1 Hill feed (+100 mesh), 436-1 Grade 3, expanded, 439-1 Grade 4, expanded, 442-1 S.C. Patterson . Ungraded, 473-1 4.5 4.S 1.0 0.3 - - m 1.0 - . - - m - 9.8 12.2 9.1 8.7 10.9 17.2 26.7 55.8 10.5 .2-7 3.5 3.9 10.9 26.3 0.2 * 0.4 1 18.1 5.1 5.6 22.6 14.1 11.4 25.6 6.1 27.3 17.6 25.3 85.1 62.2 68.3 75.0 71.4 47.8 38.1 62.2 79.6 71.2 48.9 4.6 23.6 0.4 * 0.4 47.2 84.4 50.1 99.4 * 99.2 13.9 &.0 a With the exception of Sample Ko. 267-1, all results are for composite samples. 33 TABLE 12. SUMMARY OF X-RAY DIFFRACTION ANALYSIS RESULTS Sample Fraction-Phase Mineral phases identified from XRD data . (exluding vermicnlite) Libby Groce Grade 2, 276-1 . Grade 3, 259-1 Grade 5, 264-1 Grade 5 (1-day), 267-1 Head feed, 291-1 . TBE-SINK-fibers : TBE-SINK-milky, green' TBE-SINK-dk. green, glassy TBE-SINK-lt. green, glassy TBE-SINK-fibers TBE-SINK-total TBE-SINK-fibers * TBE-SINK-total " TBE-SINK-fibers * TBE-SINK-total TBE-SINK-total 2.76 SINK-total , 2.76 FLOAT-total Tremolite, talc Tremolite, talc Diopside, magnetite Diopside, magnetite Tremolite Diopside, sphene, angite, flnorapatite Tremolite, diopside, sphene, talc, magnetite Diopside, tremolite, magnetite, flnorapatite, sphene, hematite, rhodonite Tremolite, diopside, talc, sphene, angite, fluorapatite, quartz, magnetite Diopside, sphene, tremolite, augite, quartz, fluorapatite, magnetite, hematite Diopside, tremolite, augite, fluorapatite, sphene, magnetite, hematite, quartz Biotite, tremolite, vermiculite-hydrobiotite, diopside, quartz, talc, fluorapatite, sphene calcite, magnetite, hematite Tremolite, diopside, quartz, yermiculitehydrobiotite, calcite, fluorapatite, talc, antigorite (continued) Sample S.C. Grace Grade 3, 430-1 Grade 5, 427-1 Mill feed, 436-1 S.C. Patterson Ungraded, 571-1 TABLE 12 (continued) Fraction-Phase Mineral phases identified from XRD data (exluding vermiculite) 2.76 SINK-total :* 2.76 FLOAT-nonmicaceous TBE-SINK-fibers TBE-SINK-total /# T TBE-SINK-fibers TBE-SINK-green, glassy TBE-SINK-green, milky TBE-SINK-colorless, glassy Sodium tremolite, hornblende, anthophyllite, talc, vermiculite-hydrobiotite, fluorapatite, sphene, calcite, quartz Quartz, microcline, albite, sodium tremolite, sphene, vermiculite-hydrobiotite Tremolite, anthophyllite, sodium hornblende, vermiculite-hydrobiotite, talc, sphene, fluorapatite, albite, magnetite Sodium hornblende, tremolite, anthophyllite, fluorapatite, sphene, vermiculite-hydrobiotite, magnetite Sodium tremolite, anthophyllite, talc, hornblende Hornblende, sodium tremolite, sphene Sodium tremolite, hornblende, sphene, fluorapatite Fluorapatite, anthophyllite, sodium tremolite, hornblende TBE-SINK-total 2.76 SINK-total Tremolite, iron anthophyllite, sodium hornblende, talc, fluorapatite, rutile, sphene, magnetite Talc, tremolite, anthophyllite, hornblende, quartz, rutile, fluorapatite, vermiculitehydrobiotite a With the exception of Sample No. 267-1, all results are for composite samples. KEY TO TABLE 12 Mineral name Chemical formula per JCPDS (file card number) Micaceous minerals Vermiculite Elotlte Vermicul1te-hydroblot1te ( K g 2 #37^e 0 . S 7 ^ 0 . 6 ) ( ^ i . 2 8 ^ ^ 2 . T 2 ) 0 e ( 0 H ) s -4H20 K(Fe,Mg)sAlSi80 10(OH)2 K(Mg,Fe)8 (Si,Al)80 2 0 (OH)4 -nH20 (16-613) (2-45) (13-465) Amphiboles Tremolite Sodium tremolite Anthophyllite Iron anthophyllite Hornblende Sodium hornblende Ca2M g 5Si802 2 (0H)2 N a 0 28 (Caia7EKao>2E)Mg8Si80 22(OH)2 H g TSi8022(0H)2 (Mg,Fe)TSi8O 22(0H)2 Ca2 (HgtFe)E (Si,Al)80 22(0H)2 (Ca,Na)2a26(Mg,Fe,Al)EalE(Si,Al)80 22(0H)2 (13-437) (23-666) (16-401) (9-455) (21-149) (20-481) Pyroxenes Diopside Aanite-augite C a M g ( S i 0 8 )2 (Ca,Na)(Fe,Mn,Zn,Mg)(Si,Al)20 (11-654) . (19-1) Serpentine Antigorite Mg8Si20 8 (0H)2 (7-417) Iron oxides Magnetite Hematite (19-629) (13-534) Others Talc Quartz Microcline (feldspar) Albite (feldspar) Calcite Fluorapatite Sphene (titanite) Rutile Rhodonite HgsSi4O l0(OH)2 Si02 K A l S i 30 8 H a A l S i 80 8 CaC08 C a EF ( P 0 4)s CaTiSiOs Ti02 MnSi03 (19-770) (5-490) (19-932) (10-393) (5-586) (15-876) (11-142) (21-1276) (13-138) 36 The XRD data generally confirmed the mineral identification made micro scopically. The XRD data also confirmed the observation that partially al tered mineral phases-- particularly altered biotlte-vermicullte phases-- were present. Since vermiculite is generally formed as an alteration product of biotlte mica, it was not surprising to find the Intermediate phase, mired layer vermiculite-hydrobiotite material in most samples. The XRD analyses did provide Eome surprising results. The most inter esting result was the abundance of vermiculite in the hand-picked fiber frac tion of the Grade 5 composite from Grace's South Carolina mine (HRI Sample No. 427-1). Microscopical examination of the ground material submitted for XRD revealed that the vermiculite was intergrovn vith the amphibole fibers and alEO did in fact occur in a pseudomorphlcally fibrous crystal habit. Another rather interesting result of the XRD work vas the identification of a sodium-bearing fibrous tremolite phase in the South Carolina samples. Interpretation of the XRD data vas hampered by the peak intensity altera tions caused by crystal preferred orientations. Most of the mineral phases had crystal morphologies vith at least one exaggerated crystallographic axis. Thus, in preparing samples for XRD as thin films by filtration onto silver membranes, the crystals tended to orient vith the exaggerated crystal planes parallel to the filter surface-- i.e., vermiculite plates and tremolite fibers landed on the filters lying flat, rather than on end. Even the pyroxenes, such a6 diopside, which generally do not show the prominent prismatic mor- . phology because cleavage along the prism planes is not so perfect as it is in amphiboles, tended to orient themselves on the silver membrane and thus peak intensities in the XRD patterns did not correspond to published data. Quartz and feldspars vere practically the only mineral species detected that did not exhibit preferred orientation effects in the XRD data. ' The effects of crystal preferred orientations on diffraction peak in tensities are very clearly demonstrated by the XRD data obtained for the phases analyzed of Sample No. 276-1 (Libby, Grade 2). The phases analyzed vere hand-picked and vere relatively pure phase materials. Diffraction peak positions vere consistent vith the phases identified-- tremolite and diopside-- but relative peak intensities vere not consistent vith the published values. A careful review of the published crystal plane reflections corresponding to the peaks that demonstrated the greatest variation from published intensity values clearly indicated that the peak intensity variations vere due to crystal orientation effects. That is, peaks showing higher relative'intensities com-, pared to published data corresponded to crystal planes that vere preferentially placed in. the x-ray path (e.g., the elongated axis of a prism or fiber) vhile absent peaks or peaks vith low intensities compared to published data corre sponded to crystal planes placed essentially out of view of the x-rays (e.g., the end-on-view of a prism or fiber). Individual Sample Results The results for individual samples follow. The order of presentation is (a) graded samples from Libby; (b) other Libby samples; (c) graded samples from South Carolina, Grace; (d) other samples from South Carolina, Grace; and (e) samples from South Carolina, Patterson. 37 Much of the detailed data and selected photographs are contained in ap pendices. The appendices appear in this volume and three supplementary vol umes .* The presentation for each 6antple include the sample type, code numbers assigned to the sample** and appendices page references, optical microscopy discription of the sample, and electron microscopy re6ult6. The detailed optical microscopy examination results presented here are primarily those submitted by IITRI. The ORF optical microscopy examination provided qualitative information on the presence or absence of visible fibers and identification of those fibers observed. The ORF results were in agree ment with the more detailed results presented here. Sample 270, Libby, Montana, Grace, Grade 1, Composite IITRI Code Ho. 129 Appendix references Electron microscope 1-64-87 - Kacroscopically, this sample was composed of 1- to 20-mm chunks of gold to green micaceous minerals, 1- to 15-mm bundles of white to pale green fibers, and 1- to 6-mm chunks of nonmicaceous, nonfibrous minerals. The nonmicaceous minerals ranged in color from deep green, to pale green, to white to colorless. Prismatic as well a& conchoidally fractured chunks of nonmicaceous minerals were observed. Fibrous bundles were sufficiently large and numerous to allow hand-picking before the density separation was conducted. The mineralogical composition of the sample, a6 determined by polarized light microscopy analyses of the density-separated and hand-picked fractions, is presented in Table 13. Tremolite-actinolite fibers were identified as sig nificant sample component.6. No fibrous serpentine minerals were detected. * Photographs, this volume Appendix E, pp. E-2 to E-25. IITRI EM, Appendix I, pp. 1-1 to 1-121. IITRI XRD, Appendix I, pp. 1-122 to 1-160. . ORF EM, Appendix II (two volumes) pp. II-l to 11-203 and 11-204 to 11-420. * * Originally the three fractions of the sample composites were assigned unique ID numbers. For this report the designations were simplified 60 that portions of the same composite sample had the same ID number fol lowed by a letter indicating the analyzing laboratory. Most of the computer generated data in appendices I and II relate to the original ID numbers. 36 TABLE 13. COMPOSITION OF SAMPLE 270-1 Mineral phase Estimated mass concentration (%) Tremolite-actinolite Tremolite-actinolite Sphene Diopside Augite Hornblende Magnetite, hematite Calcite Quartz Biotite Talc Vermiculite Other minerals fibers prisms 4-6 1-3 1-2 2-5 < 1 < 1 1-2 1-3 3-5 1-2 < 1 72-82 1-3 Tremolite-actinolite vas identified as the primary fibrous phase present in the sample. Fiber color, refractive indices, and extinction angles were all consistent with a tremolite-actinolite amphibole. Both colorless (white, macroscopically) and green fiber bundles were evident; the green-colored fiber bundles were the more abundant. The fibrous phase of this sample vas not as well-formed as it vas in the other bulk samples analyzed from Libby. That is, fiber bundles contained higher proportions of materials that would be more correctly classified as prismatic rather than fibrous, than did other samples from Libby. Particles that could readily be classified as fibers tended to be much shorter in this sample compared to other samples analyzed from Libby. An unusual morphological particle type vas a significant component of the fibrous phase of this sample, and was noted as only a trace component of the tremolite-actinolite phase in other samples. The particle type vas composed of lamellated tremoliteactinolite prisms, intergrown at angles as great as 60 degrees to each other. In other samples, particles composed of the lamellated prisms were composed of entirely parallel crystals. It would appear that the nonparallel intergrown prisms represent an intermediate metamorphic state, between prismatic and fi brous tremolite-actinolite. Other mineral types such as diopside, hornblende, sphene, calclte, quartz, and even the vermiculite tended to mimic the nonparallel intergrown prism mor phology of the tremolite-actinolite, rather than the truly fibrous morphology. The degree of Intergrowth of amphibole and pyroxene mineral phases with the vermiculite appeared to be greater in this sample compared to other Libby samples. That is, a higher proportion of the vermiculite plates in this sam ple contained other mineral phases sandwiched between layers, than other Libby camples did. A higher proportion of the stacked vermiculite plates also ap peared to he weathered in this sample compared to other samples. A summary of the E M results for this sample appears in Table 14. faaru 171*1 b t o l t o i e f w n it. gotuiT t rucTto* utocwoft itmw rot wwt urn. o * a . cunt i fib e rs t f >11 U w iili_____________________ ______________ fib e r s treater thee l . l m i s 1--r t l_____________ lite r e-- eetratiee IIP * f ib e r t / e r ^ file r <10* H k n / i ) Cw o n r i i i H C M c m riU N falsaU se to b tto itto m m la . s i fu leaU ti to b t f a i u i b u s I s. a f t n fw fm rs 1 fiber SM M ttm lM fibers fSX C e e fld ie s t fiber reacaatrsUse fibers lu i toum l to u e ta i (ppe) * C M tto b u t bu rn t detected (^ e ) es--t s f *19* SVllii S4..tHo-UU.S . l.l %4 ft m i f _ 1f 4t SS.9) I M M i It ' e A Stobibela < U D ) ; C cbryestile; m i T t o u t . Sample 276t Libby, Montana, Grace, Grade 2, Composite IITRI Code Ko. 126, OBF No. Appendix references , Photographs E-14,15; XRD 1-157-160 Electron microscope 1-86-91 The sample was composed of 1- to 12-mm flakes of gold to deep green midaceous flakes. White to gray to green fiber bundles ranging in diameter from 1 to 5 mm and in lengths from 2 to 15 mm were relatively abundant. Other con stituents observed macroscopically were glassy, light to dark green mineral chunks; milky, pale green chunks; and colorless to milky white chunks. The abundance and large grain sizes of the fibrous bundles allowed for easy hand-separation of fibers for gravimetric determinations. The mineralogical composition of the sample, as determined by polarized light microscopy analyses of the various hand-picked and density-separated fractions, is listed in Table 15. The hand-picked fibrous material was readily recognized as the tremolite-actinolite amphibole. No serpentine fibrous mate rial was detected. 40 ( TABLE 15. COMPOSITION OF SAMPLE 276-1 Mineral phase Estimated mass concentration (%) Tremolite-actinolite Tremolite-actinolite Sphene Diopside Augite Hornblende Magnetite, hematite Calcite Quartz Biotite Talc Vermiculite Other minerals fibers prisms 4-7 3-5 < 1 4-7 1-2 < 1 1-2 1-2 3-5 2-4 < 1 66-72 1-4 Tremolite-actinolite was found to be the primary constituent of the fi brous phase of this sample. Although fiber bundles were up to 15 mm in length, no single fibers were found anywhere near this length. Bather, the fiber bun dles were composed of short fibers intergrovn at slight angles to each other. All fiber bundles contained both the truly fibrous material as veil as the more bulky, lamellated prisms. Inclusions such as diopside, hornblende, and calcite within fiber bundles tended to adopt a fibrous morphology. Optical properties of the tremolite-actinolite fibers again included in clined extinction angles and refractive indices slightly greater than the truly prismatic tremolite-actinolite fragments. Most of the fiber bundles ex hibited a slightly greenish color when mounted in immersion oil. The fibers were also pleochroic. ' The fiber bundles also probably contained some traces of anthophyllite; and talc was detected microscopically within several fiber bundles. There was an unusual fibrous phase found in the 2.76 sink fraction that .could not be identified. The fibers had refractive indices higher than the tremolite-actinolite fiber bundles and were a deep blue-green color. The strong coloration resulted in anomalous interference colors similar to those seen for glaucophane-riebeckite. This component was less than 0.1% of the total sample; therefore insufficient material was available for additional characterization studies. A summary of the EM results for this sample ap pears in Table 16. 41 tttti u. tatujrr <rracncwwrctcseonmum ret iwm titit. wua. m st t flte r * # a ll t f i i t l i H W f t o f t i t m l i w i i t w / n ~ toifWUHM* * I* bliMUi HI* t n C iif t i--ci t fu<r (M c n tn u it lu t lA U ltll IlM CUi h. ( Cik n IEEE flM w t f t i u r U * t t . i m w i H u t e o ^ w tm m IIP* liX n /tV CM cm ruiM q m ilu t te K id iiu i m m tJX CN filN Ct | liter ta tc n triilM u u n il itu c u i Iff*) . r . fiten fitef CM t K i m *H -t Id a liu d 0 .4 0 .1 4t.S B U 4.4 H U 4 4 O .t tt 4 4.4 T .. . C A t e U B U (1US); C B r p M lU ; m i T M tet. Sample 259. Libby. Montana. Grace, Grade 3, Composite IITRI Code Ko. 122, ORF No. 261* Appendix references - Photographs E-2-4, XRD 1-123-125 Electron microscope 1-14-25, 11-20-43 Microscopically, the sample was composed of large (1 to 7 mm) gold to black micaceous flakes; dark green, glassy fragments; white to pale green flexible fiber bundles up to 8 mm in length; and at least three other color less to pale green mineral phases. The mineralogical composition of the sample, as determined in the PLM analysis of the three density-separated fractions, is listed in Table 17. Amphibole (tremolite-actinolite) asbestos fibers were found in rather signifi cant concentrations. No serpentine minerals were detected, however. TABLE 17. COMPOSITION OF SAMPLE 259-1 Mineral phase Estimated mass ` concentration (%) Tremolite-actinolite Tremolite-actinolite Sphene* Diopside Augite Magnetite, hematite Calcite Quartz Biotite Vermiculite Other minerals fibers prisms <1 2-4 1-3 3-7 2-3 < 1 2-5 2-5 10-15 65-72 1-3 42 The tremolite-actinolite occurred almost exclusively in an unquestionably fibrous crystal habit with very little prismatic tremolite-actinolite present. The fiber bundles were found to be composed of very fine, teasable, flexible fibers. In most of the fiber bundles, the individual fibers were not perfectly parallel to each other and were not as long as the fiber bundles; numerous small groups of short fibers stacked at slight angles to each other both ver tically and horizontally comprised the "fiber bundle." Optical properties of the individual fibers included refractive indices greater than 1.610, extinc tion angles greater than 0 degrees, and pale green color with slight pleochroism. Other mineral phases (including alteration products) were included in each fiber bundle. Ic some cases, these other mineral phases were pseudo- morphically fibrous. Quartz, calcite, titanite, and diopside were all ob served in a fibrous habit. Calcite was the most abundant pseudomorphically fibrous mineral. Approximately 10% of the "2.76 sink" fraction was composed of white fiber bundles, which upon teasing and microscopic examination were found to be composed of 70 to 80% calcite overgrown on 20 to 30% tremolite- actinolite fibers. The tremolite-actinolite fibers were also observed to be tightly bound to vermiculite plates and growing in between layers of vermiculite. A summary of the EM results for this sample appears in Table 18. B 9-I H 0 asf-i tifo iu fi** KxfalUtad tm r ii. or tttcnaK mooscow to w n to* u m i im r, m e t, w m t fiber* of a ll iM rtSi fiber* r r n u r than S.9 m U 1earth lite r m w i U i t w ( I f fik tf.S.1 tlh t c N e m m in no* n te ri/il b a c m riiiM u tm triU M ifu iv ile a t u ta tU a to f i m i Bo. 4 f oquivalcat U ( J t l a t iH u i b . t m (a tflltm | CiWr m tittm m film tS t CMtfiOmt 1 liber i w m i n u n liberty lu a la u m l (tta c tH irr*> cawated Hew ia u m l detected (m i m .0 *s. . ts s.t 42.* 41.1 1.4 $ 1.0 f .1-42 8 .1 -3 0 .9 19.4-37.1 44-14 0.4 1.09 1.19 1.4 i.m .m 110 .*1 19 * ISO .1 IC** 140 ft 11.4 107 . 1 U 10 < l.t too 14.3 m 14.3 i if If t < 1.0 1,1-17,1 1.1*17.0 0*11 t.t-1 0 .1 9.0-30.7 f.0 -8 . S.Of t.o * 0.4 O .fM . 190 19 4 30 T 0 C 4| 1 4 C 340 34 4 34 T too * to 4 0 a As NfUktla (MS); C eUyMClUi m i T total. Sample 282, Libby, Grace, Grade 4, Composite IITRI Code Ho. 126, 0RF Ho. 281 Appendix references Electron microscope 1-92-97, 11-356-375 Macroscopically, the sample was observed to contain mostly 1- to 4-mm goldish-brown vermiculite flakes. White fiber bundles up to 3 mm in length were visible. Honmicaceous, nonfibrous mineral phases were also observed; at least three different mineral phases ranging in color from white to emerald green were detected. 43 The mineralogical composition determined for this sample by polarized light microscopy is listed in Table 19. TABLE 19. COMPOSITION OF SAMPLE 282-1 Mineral phase Estimated mass concentration (%) Tremolite-actinolite Tremolite-actinolite Sphene Diopside Augite Hornblende Magnetite, hematite Calcite Quartz Biotite Vermiculite Other minerals fibers prisms (0.3)-l 1-3 < 1 3-7 1-2 < 1 1-3 < 1 1-3 1-3 78-88 1-3 Tremolite-actinolite was again detected as a significant sample component and was again present in three distinct crystal habits. The truly prismatic tremolite-actinolite was most abundant. Most prismatic fragments had small bundles of fibers or bundles of the thin, lamellated prisms (that could readily fracture to produce particles definable as fibers) attached to them. Practi cally all bundles composed primarily of truly fibrous tremolite-actinolite contained thick, chunky prisms or the lamellated prisms. All truly fibrous tremolite-actinolite bundles were composed of intergrown fibers; i.e., no bun dles were composed of uniform length, parallel fibers. Groups of fibers grow ing at angles as large as 75 degrees to each other were observed in the same bundle. The lamellated prism bundles, however, did tend to be composed of crystals growing parallel to each other. Bundles composed of both the truly fibrous materiel and the thicker, lamellated prisms also tended to be composed of nonparallel crystal bundles. The tremolite-actinolite fiber bundles hand-picked from the sample ranged in color from pure white to deep green. Less than 10% of the fiber bundles were the pure white color; most fiber bundles were pale green and had refrac tive indices in the middle range reported for the tremolite-ferroactinolite solid solution mineral series. Extinction angles for the truly fibrous crys tals and crystal bundles were inclined at least 5 degrees. No fibrous enthophyllite, prismatic anthophyllite, or talc were detected in this sample. However, some hornblende in a morphology that could be clas sified as fibrous was detected. The hornblende bundles containing crystals 44 classifiable as fibers vere composed primarily of the thicker, more brittle lamellated prisms. The 2.76 sink fraction contained a significant fraction of biotite. The other major constituents of this fraction vere vermiculite flakes intergrovn with tremolite-actinolite, diopside, and iron oxides. The 2.76 float fraction vas relatively free of pyroxene and amphibole mineral fragments. Few flakes of vermiculite intergrovn with amphibole min* eral phases vere detected. A summary of the E M results for this sample ap pears in Table 20. l u St2*t lifilia ta i H2< U filb M CM l.t . ts .t n .ft .* M t a i l 20. tomutr or mcitow wiooscm m o w rot tum i ttttr . q u a , cuse 4 i Ibers ( all iM itli rw cm m iM 416* ItWre/e) b a c iitru iM K l CetfiAnce tu n ln t to 1 fiber b U M tii MM iN cm m laa iatam l ttoctod (m l a. ( fibers M td TIWt. mtWP O m 1.1 1. In ttk T i e r (m m iiiUm aio* fiM n m CaamtraitM C ^dnlH t to b llaatal s h i 3t CMfiOcwe 1 fiber aaecaatmiaa a t ia ta m l Itta e tH (m i *1.1 a .iu i > < i.4 * .>40 * .>40 C 1.4 .>40 440 fi IM 440 .v m tf.t I.M .4 IT 4 4.444 H .I .444 ii .444 a. of fibers fiber waaotod c SI >9 T f 1 A amphibole (SAED)i C to ry aatile ; m i T t o ta l. Sample 264, Libby, Grace, Grade 5, Composite IITRI Code No. 120, ORF No. 263 Appendix references Photographs E-5, 18-20, XRD 1-126-129 Electron microscope 1-26-52, 11-44-166 - _ ' Macroscopically, the sample vas observed to be a fine, goldish-brown pow der composed of obviously flake-like and fibrous particles. The flake-like particles vere generally less than 2 mm in diameter. Fibers up to 3 mm in length vere present. At least tvo other nonvermiculite, nonfibrous mineral phases vere observed; one vas green in color, while the other vas colorless. The mineralogical composition of the sample determined by Pin analysis of the three density-separated fractions is listed in Table 21. 45 I TABLE 21. COMPOSITION OF SAMPLE 264-1 Mineral phase Estimated mass concentration (%) Tremolite-actinolite Tremolite-actinolite Sphene Diopside Augite Hornblende Magnetite, hematite Calcite Quartz Biotite Vermiculite ' Other minerals fibers prisms 2-4 2-5 1-3 6-9 1-3 < 1 1-3 1-3 1-3 3-7 70-74 2-4 Tremolite-actinolite was present in a fibrous morphology. In this sam ple, however, a significant amount of prismatic tremolite-actinolite was also present. Both fibers and prisms exhibited inclined extinction. There ap peared to be some fibers close to the tremolite end member of the series, as refractive indices of some fibers were observed to be at or just sightly below 1.600. s /' - The tremolite-actinolite fiber bundles occurred as the nonparallel stranded bundles. A higher percentage of the fibers and fiber bundles was found intergrown with vermiculite, biotite, and the other low density minerals in this sample compared to the larger particle samples. Pseudomorphically fibrous quartz, calcite, diopside, and augite phases were again detected. A summary of the EM results appears in Table 22. 46 I ttw ***- U M l CM* K M CD* *64-1 b f t l l a u t 244-0 b l i U i u t 1*4-1 (0) It/ilU u T M M (I) ,, W tlU U T aam mu g. Bttmrr or titCTWiiawow fjk w tf ill lM ti n a m c r i t i M l ! f if c c n /.l m tm ftim n to u m l liiM lm u 1 lite r te u e u t b tia atW m u CMCMtlU M te. ( lite r s <-- U 4 w mwi liiw. qua. aupt 9 -n c 7 teaa flN n tw tU f |M t | . l M M M trU lM 71** U W ../.1 C M C M im iM IftiV llM t U 931 C u ilA o c * 1 lite r to u m l te to ru o -- ti Iw ttl b ito iu t m ii C M tM tm iM cw ) te. #r lite r s 4M S iW lite r 479* IU 123 100 1.* 127 W I.S 17 too HI 149 140 < 1.4 lit no 1.4 1.9 4 22 140 07.9-1)9 M I-149 93-240 90-1)0 19-42 20-91 120-210 101-177 lit-tto 1)0-200 42-174.4 1-1)0 0-4.9 4 .4 -U 12-19 110-170 0.4 i.r IJ > 1 .) 1.42 1.42 1.42 1.42 1.42 1.2 1.13 1.IS IJ 1.90 1.90 1.90 1.90 1.90 1.90 940 400 1,200 970 14 II 490 *. M i,too 290 *.*M 1 12 41 i. a a 99 11.1 107 IS . 74 < 1.4 Ml 41.2 9 17 1.9 19 2.9 23 1.4 11) 41 121 24.2 127 IS .1 109 n 0 1.4 102 H . 2 n 42 1.4 1 < 1.4 4 1.2 19 1.4 02 44 l.l-M .J 1.2-104 BMC * 144 43.1 17-97 0-7.1 0-4.4 11-44 4S.3-I23 U .4-122 42-17 - 44.0-44.2 27-97 0-4.9 t-f.l JH ] no 41 A 2.4 - - 4ft I 2 C i.u 1.97 IX it A - C I.J 1,100 21 A 2 C 1.42 t.42 1.42 1.42 1.42 900 24 A 1 C 4 2 r 91 t UR Hi 29 T 2,900 92 A 1.2 13 T I.U 1.93 1,700 - ii A 4 C 1.2 J00 42 A C 1.90 1J0 1.90 1.90 1.90 1.90 2,200 . 20 12 2,400 21 A ft 4 1 nr 2 MR 21 T 4 A a p fc ik U i C t t i y m l t e t T u u l ; O' M l n t U i M t t a r n l lite rs UR M w k i u * d iy M tili. # acral; m I 1 8 > M u lir asrptcUfy. M t M uU fiM U m a 2 Tte liu tr It tte itllu iii Uit tte CMtiltj feectewre I f IR ti U | tte m i y w n tm . lilt ttiu lly mH fey*te ttlur likonttiy it 1*4-1 (0) i n tte malts tkUiMt Sample 267, libby, Grace, Grade 5, 1 Day Appendix references Photographs E-14, XRD 1-153-156 Macroscopically, the sample was composed primarily of fine (0.5 to 3 mm) goldish-brown micaceous flakes. Pale green to white fiber bundles up to 2 mm in length were visible. At least two green, nonmicaceous mineral phases and one colorless, nonmicaceous mineral phase were observed. The mineralogical composition of the sample, as determined by polarized light microscopy analyses of the various density-separated fractions, is listed in Table 23. Fibrous amphibole-- mostly tremolite-actinolite, with some intergrown anthophyllite-- was definitely present. No serpentine fibers were de tected, however. 47 TABLE 23. COMPOSITION OF SAMPLE 267-1 Hineral phase Estimated mass concentration (%) Tremolite-actinolite fibers . Tremolite-actinolite prisms Anthophyllite (prisms and fibers) Sphene Diopside Augite Hornblende Magnetite* hematite Calcite Quartz Biotite Talc Vermiculite Other minerals 2-5 4-8 < 1 < 1 10-15 1-3 < 1 1-3 1-2 1-3 < 1 1-2 65-70 1-3 The tremolite-actinolite occurred in many different particles but occurred predominantly as irregularly fractured fragments of solid prisms. Prismatic fragments composed of numerous thin* stacked prisms which were capable of fracturing into elongated, parallel-sided fragments definable as fibers were observed but represented less than 1% of the total sample. Truly fibrous bun dles composed of very fine, teasable individual fibers, were significant sam ple components. Typically, the fibrous bundles were composed of short fibers that were not perfectly parallel. Barely were fiber bundles composed of indi vidual fibers that ran the entire length of the bundle. Obvious fibers were also found attached to (i.e., "growing from") chunky, prismatic fragments and from the lamellated prisms. Both prismatic and fibrous varieties of tremoliteactinolite were found bound to and intergrown with vermiculite plates to a minor extent. The extinction angles of both prismatic and fibrous varieties of the tremolite-actinolite were on the order of 7 to 18 degrees. Refractive indices of the prismatic variety tended to be greater than those of the fibrous habit. Prism colors were mostly green, while fiber bundle colors ranged from color less (white) to pale green, depending upon the types and amounts of inclusions present within the bundles. Anthophyllite fibers and prisms were detected. The fibrous form was found only intergrown with the tremolite-actinolite fibers. Free anthophyllite prisms as well as prisms intergrown with tremolite-actinolite prisms were present. Almost all the tremolite-actinolite fiber bundles that contained anthophyllite also contained talc. 46 Unlike the 266-1 composited grade 5 sample from Libby, this uncomposited sample did not appear to contain other mineral phases in pseudomorphically fi brous habits. None of the fibrous calcite detected in the 264-1 sample vas found in this sample. Host of the nonmieaceous contaminant minerals present in this sample were high density materials and were thus found in the THE sink fraction. The 2.76 density fractions were relatively free of nonvermiculite mineral phases, par ticularly the 2.76 float fraction. Sample 291, Libby, Grace. Head Feed Composite IITRI Code No. 130, ORF No. 290 Appendix references Photographs 3RD 1-130-134 Electron microscope 1-2-13, II-1-19 Hacroscopically, this sample vas quite variable in color, particle mor phology, and grain size. The overall color vas a light brown. Relatively few micaceous flakes were visible to the naked eye; the largest flakes vere less than 10 mm in size. Under the stereomicroscope, most of the brownish, fine powder (less than 1 mm) material present va& observed to be micaceous. Sev eral large white to pale green elongated (and probably fibrous) rock chunks greater than 20 mm vere observed. Obvious mixed phase grains (i.e., mineral phases partially altered) vere also present as 1- to 15-mm grains. Microscopically, the sample vas observed to be composed primarily of non micaceous, contaminant minerals. The overall sample composition determined by microscopical analyes of the density-separated fractions is presented in Table 24. ` TABLE 24. COMPOSITION OF SAMPLE 291-1 Mineral phase Estimated mass concentration (%) Tremolite-actinolite Tremolite-actinolite Sphene Diopside Augite Hornblende Magnetite, hematite Calcite Quartz Biotite Vermiculite Other minerals fibers prisms 21-26 6-9 1-3 24-29 2-5 < 1 3-5 3-5 4-7 1-2 20-25 3-6 49 Tremolite-actinolite was a major component of this sample and occurred in both fibrous and prismatic crystal habits. The fibrous habit was found as discrete fiber bundles, fiber bundles intergrown vith prismatic amphibole and pyroxene mineral phases, and as small fiber bundles protruding from vermiculite (or other micaceous mineral) plates. Fiber bundles vere composed mostly of smaller, shorter irregularly stacked bundles of fibers, rather than as bundles of perfectly parallel, uniform length fibers. Optical properties of fiber bundles identified as tremolite-actinolite indicated that a vide range of chemical compositions was present; i.e., some end member tremolite and ferroactinolite phases as veil as the intermediate actinolite vere present. The tremolite-actinolite fibrous phase vas a major component of each den sity fraction because of the multiphase nature of a large percentage of the particles in this sample. Inclusion of lover density phases vithin each fiber bundle as veil as attachment of fiber bundles to lover density mineral grains resulted in a lover than normal bulk density for the tremolite-actinolite. Some unusual fibrous phases vere present. Fiber bundles exhibiting the anomalous blue and pink interference colors typical of crocidolite (riebeckite) asbestos vere observed. The refractive indices of these fibers as veil as their inclined extinction angles (and 3RD data) ruled out crocidolite as the mineral species. The fiber bundles vere strongly pleochroic (yellov-green to blue-green), and this undoubtedly caused the anomalous interference colors. Further characterizations by electron microprobe, micro X-ray diffraction, and electron microscopy must be performed in order to fully identify this phase. Possible mineral identities include ferroactinolite, sodium tremolite, and glaucophane. Identification of all mineral phases present vas impossible. Many inter mediate, partially altered phases vere present in this sample. XRD data sug gested that antigorite is present in the lov density fraction. A summary of the EM results for this sample appears in Table 25. tfl-1 ttl-0 l f l - f Exfoliated tu ti is. totmr or Bier** mooscon ttwin rot t>wx uttr. q u a, m s rep Fiber* of a ll loertba ~ n c r C M entnU M (14* fibers/*) C M catruias f i l n l M t to t a t lotted moo Baas SS C u f l i M M liu m l 1 fiber ceecaetratlee decocted . 99*> Be. of fibers tOMtOd file r Rm i fibers c*aeaotrtioa K l C e tfid m to tem ! irs a u r t b t 4.4 ia ie iesrtb ----- 0Ceecestrotlee ivalest Co 1 fiber Estimated mass caocostrstloo Ittm a i Be. of fib e rs CSMted fib e r 12.1 13.4 1.4 UO 1.1 n .i n . 4 I.S -4 3 .S 4 1 .4 -tS .J 40-190 4-1.4 *7.7-109.9 44.1-111.1 4.7 1.17 1.17 4.4 470 4. IS 490 < | IK 1 *4.7 107 *7.4 1 US SI t 4.4 97 to .s I0S ao.s 4 11.1-77.9 I2 .H 1 .I 14-44 4-S3 4-SS 4.7 4.S42 4.443 4.4 430 n A 40 t 4 .H 1 C 120 IS 4 4 C *40 r A 17 T C 4 A * rn p H lh U ( U 0 ) i C c b ry se U lo ) mad T lo c a l. 50 Sample 294, Libby, Grace, Extractor Waste, Composite IITRI Code No. 134 Appendix references Photographs *15 Electron microscope 1*96*102 . This sample was quite broad in particle size. It was composed primarily of 0.2* to 15*mm gold to brown micaceous flakes. White fibrous bundles up to 9 mm in length were observed. The fibrous bundles were sufficiently abundant and large in size to allow hand*picking after density separations were con* ducted. Nonfibrous, nonmicaceous mineral fragments which were mostly green in color were present in diameters up to 6 mm. The mineralogical composition of the sample is listed in Table 26. Sam ple components were identified by polarized light microscopy analyses of the hand-picked and density-separated fractions. Fibrous amphibole mineral phases were detected in the analyses, but no fibrous serpentine mineral phases were detected. TABLE 26. COMPOSITION OF SAMPLE 294-1 Estimated mass Mineral phase_____ ;______concentration (%) Tremolite-actinolite fibers Tremolite-actinolite prisms Sphene Diopside Augite Hornblende Magnetite, hematite Calcite Quartz Biotite Talc Vermiculite Other minerals 1-4 1-3 1-2 3-7 < 1 < 1 1-2 1-3 4-10 6-9 < 1 68-76 1-3 The tremolite-actinolite occurred primarily as fibrous crystal bundles and crystal bundles containing both fibrous and prismatic materials. Coarse crystals composed only of bulky, prismatic materials were rare. Tremoliteactinolite fragments composed only of elongated, narrow, thin, lamellated prismatic crystals were also rare but were not as rare as the chunky prismatic crystals. Both prismatic crystal morphologies were observed mostly in con junction with the fibrous morphology; that is, most of the tremolite-actinolite mineral fragments that contained prismatic material also contained at least 25% truly fibrous material. Fibrous crystal bundles containing mostly fibrous crystals were again observed to contain 6mall bundles of short fibers that were stacked at slight angles to each other both longitudinally and laterally. Single fibers that ran the entire length of the fiber bundle were rarely seen and were only observed in rock fragments that contained at least 35% prismatic materials. The fibers grew parallel to and on top of the prismatic material. Refractive indices of the fibrous tremolite-actinolite tended to be greater'than those of the prismatic tremolite-actinolite crystals. Practically all fiber bundles exhibited a green coloration and pleochroism in plane polar ized light. Prismatic crystals tended to be more strongly colored than the fibrous crystals. Prismatic crystals also tended to be more blue-green than green. Extinction angles of both the prismatic and fibrous crystal habits were greater than zero. Intergrowth of tremolite-actinolite and pyroxene minerals with vermiculite vas greater in this Libby sample than it was in most of the other Libby samples included in this group of 10 lesser priority vermiculite samples. This is reflected by the relatively high weight percentage of the 2.76 sink fraction. Fibrous phases in addition to tremolite-actinolite were detected. Some pseudomorphically fibrous calcite vas found intergrown in tremolite-actinolite fiber bundles. Pseudomorphically fibrous diopside and (probable) hornblende were also detected. Traces of fibrous anthophyllite intergrown with talc within fibrous tremolite-actinolite bundles were also observed. There were two additional fibrous phases present that could not be identified because they were present in 6uch low concentrations and could cot be isolated for further studies. One fibrous type exhibited the anomalous Interference colors and higher refractive indices associated with the glaucophane-riebeckite series amphiboles. The other fibrous phase exhibited the lower refractive indices and anomalous interference colors this analyst has observed in chrysotile sam ples containing biotite and vermiculite. A summary of the EH results for this sample appears in Table 27. ftaafle m -f lifiliiu d . tA iti it . rwuiT or tucreor m aosart iis u iff ro> U ffii n u t, cua. txwnsm* i f a ll t-- ttbe llCrS e n tra ii-- 110* f i M n f t ) fiiicniritlM tn tMfii la te ra l fitv ilm u I liber detected l it iH t i d Mas iw c n tn tiM (ppa) IS .I 4)0 >2 14.0 7.7 0.1 . 101 tt.S 0.7 0.7 J.4 |0 * I a A aapbibaU ( U O ) . d i y i i l i l i ; sad T u u t . 400 V * U T 52 \ Sample 297 Libby, Grace, Baghouse, Mill Dust, Composite IITRI Code Ho. 136, OKF No. 296 Appendix references Photographs -17 ' Electron microscope 1-103-108, 11-376-388 The sample vas a brownish-green, very fine powdered, mineral dust. Obvi ously micaceous flakes as large as 4 m m in diameter were observed. Obivously fibrous crystals up to 2 mm in length were also detected. Density separations in heavy liquids did not result in clean separations of the various mineral phases. Fine particle sizes resulted in temperatureinduced turbulent motion of particles rather than strict density settling. The particles were also extremely agglomerated and thus behaved with the com posite densities of the component particles. The mineralogical composition of the sample, as determined by the polar- . ized light microscopy analyses of the various density-separated fractions, is presented in Table 28. TABLE 28. COMPOSITION OF SAMPLE 297-1 Mineral phase Estimated mass concentration (%) Tremolite-actinolite Tremolite-actinolite Sphene , Diopside Augite Hornblende Magnetite, hematite Calcite Quartz Biotite Talc Vermiculite Other minerals fibers prisms 8-12 2-6 1-3 3-6 2-5 < 1 1-3 1-3 2-5 1-3 < 1 55-61 2-5 Tremolite-actinolite was a major sample component. Bundles composed of truly fibrous crystals as well as lamellated prisms and chunky prismatic frag ments of tremolite-actinolite were detected. Single crystals of tremoliteactinolite in morphologies classifiable as fibers were not unexpectedly abun dant, as the sample obviously was fine particle material produced from frac ture and abrasion of larger mineral grains. Both the truly fibrous bundles 53 and lamellated prism types of tremolite-actinolite would be expected to give rise to single crystals morphologically definable as fibers. Again, the dis tinction between fiber-like amphibole crystals produced from particles macroscopically definable as fibrous and fiber-like crystals produced from mineral grains composed of thin, narrow, lamellated prisms, may be immaterial. It was certainly not possible to define the origins of all the fiber-like tremoliteactinolite single crystals present in the sample. Anthophyllite was again detected as a very trace (< 0.1%) sample compo nent. Fibrous hornblende and the unidentified fibrous blue-green amphibole were also detected again. Traces of a mineral with morphological and optical properties similar to nonfibrous serpentine were detected in the 2.76 float fraction. Concentration of this possible serpentine was estimated to be well below 0.1% of the sample. The low concentration precluded further isolation for verification of the pro posed identity. A summary of the EM results for this sample appears in Table 29. ` 4a*U ii> tr Im ra n it. oatiw or tuatm mcooscotT otsuiw wt um z tn n . qua, r u post , flMry w xM tn U H no* itiUi -- Cm o s u i u m K llf l lM t U CftttafUd MM VSX Ca*fl4am c* I flW r M c n tta U M la u r a l M a c tH (p f ) M . 9t fib e r* maaated flWw trtiuttti r^lWr wacm miwIlC* toK tru lilv ilw U tS t Cw fU w n Mm la u m l I fiber Ittm ai mtoI--tch b U aaiil MM m e n iritiM (ft*) >a. *f fiber* M aud fiber in HM M0 I.l 03*140 !I..S0 <TM fafelMud 977 fM 05 49 1.1 40-SI tee *4S tl! 404 I.M Moo I.U .* 14,000 4t1 ' C4 90 4 It 0C a i * i ^ l l l i k (UXS)} M r w c iU f md ? M tal. ; . . { r .* Sample 288, Libby. Grace, Screening Plant Dust, Composite IITEI Code No. 135, 0RF No. 287 ' Appendix references Photographs E-16 a Electron microscope 1-108-113, 11-389-402 Hacroscopically, the sample was observed to be a very fine, pale green powder. Obvious micaceous flakes up to 2 mm in diameter were present. From the bulk density of the sample, it appeared that micaceous type minerals were - the primary sample components. * Density separations produced a deeper green, powdery fraction (sinks in tetrabromoethane), a brownish-green fraction with some micaceous flakes (sinks in 2.76 density liquid), and a gold-colored fraction obviously composed pri marily of micaceous flakes (2.76 floats fraction). The mineralogical composi tion of the sample, as determined by polarized light microscopy analyses of the three density-separated fractions, is presented in Table 30. Separations of the density fractions were not very clean, in part due to the very small grain sizes, but mostly due to the intergrowth of high density phases with low density phases. 54 TABLE 30. COMPOSITION OF SAMPLE 268-1 Mineral phase Estimated mass concentration (%) Tremolite-actinolite Tremolite-actinolite Sphene Diopside Augite Hornblende Magnetite, hematite Calcite Quartz Biotlte Talc Vermiculite Other minerals fibers prisms 2-5 1-4 < 1 3-6 1-2 < 1 1-3 1-3 4-7 1-3 < 1 68-76 1-4 Identification of mineral phases vas somewhat hindered by the relatively small particle sizes of the fractured mineral fragments. Numerous shards of vermiculite were present and could easily be mistaken for fibrous mineral types on morphology only. 'The vermiculite shards could easily be distinguished from the fibrous amphiboles on the basis of refractive index, however. Although the mineral fragments in the sample were quite abraded and frac tured, three distinctly different morphologies of the tremolite-actinolite mineral phase were observed. The chunky, prismatic crystals that would not fracture to produce fragments definable as fibers were the least abundant tremolite-actinolite phase present. Fragments composed of elongated, narrow, thin, lamellated tremolite-actinolite prisms were as abundant as the bundles of nonparallel intergrown, truly fibrous tremolite-actinolite crystals. Mor phologies of the very small (less than 10 pn wide) single tremolite-actinolite crystals that could be classified as fibers on the basis of aspect ratios sug gested, that equal proportions of the larger lamellated prisms and true fiber bundles had been abraded to produce the single crystals. That is, many of the "fiber-like" single crystals were more platey than true fibers would be ex pected to be. However, at this small particle size, the origins of particles classifiable as fibers-- either from lamellated prisms or true fiber bundles-- are only speculation and may well be immaterial. In addition to the tremolite-actinolite fibrous amphibole, fibrous anthophyllite was detected by its parallel extinction angles and different refrac tive indices. Fibrous anthophyllite was well below 0.1% of the sample mass; it was most frequently found in association with talc. 55 Other fibrous phases present included a morphology of hornblende that could be considered fibrous, and pseudomorphically fibrous quartz and calcite. The unidentifiable blue-green amphibole with the anomalous interference colors similar to glaucophane-riebeckite series amphiboles vas again detected. The 2.76 float fraction contained a mineral phase with optical and morpho logical properties consistent vith nonfibrous serpentine. Insufficient numbers of particles vere available for further identification studies. A summary of the EM results for this sample appears in Table 31. tU 4 tU -I b fiU a U l Im i MS l.t l.*0 0 I.ISO t u u at. n ttv r t or u teritm n a to sa n nsm xt w m u x u r n . tam rr^c mart Bbeti of all IwtUi P* liW ri/t) MMUatiM C falnlM t n C itlM U i m u S t C u filM u la u m l t fiW r latacu i M M M UitiM <pH> la. if IlM n r-- t il fiber* arvitir thaa l.l i m to I m t l kibet mtMtritiu no* fttlV llM l U t lt lM U l M U S t t u fllu c i t fib e r M w u tw U u baa la u m l fe te c ta f (fft) M *)M IJ4 t t.f I .I N 41.M 0 m IM . < I.C f2 l,tM It7 I.1 U n *M * ui t,TOO I .U 11,100 . I f fib e rs m a tM 7 It U fib e r 1 A t a A a a tlflila ( liO ) ; c tifs e U le ; md T l i t i l . Sample 430, Enoree, South Carolina, Grace, Grade 3, Composite IITOI Code Ko. 121, ORF No. 429 Appendix references Photographs E-7-8, XRD 1-135-138 Electron microscope 1-60-65 Hacroscopically, the sample vas observed to contain 1- to 5-mm black to brownish-gold micaceous flakes, and 1- to 3-mm fragments of nonmicaceous min erals. The nonmicaceous minerals were white, pale green, or reddish-brown in color. No obviously fibrous phases were detected in the stereomicroscopic examination. The composition of this sample as determined by PLM analyses of the density-separated fractions is listed in Table 32. 56 TABLE 32. COMPOSITION OF SAMPLE 430-1 Mineral phase Estimated mass concentration (%) Anthophyllite Tremolite-actinolite Auglte Hornblende Apatite Magnetite, hematite Calcite Quartz, feldspars Talc Vermiculite Other minerals <1 2-4 < 1 1-2 1-2 1-2 < 1 4-6 1-2 80-90 1-2 As is evident from Table 32, the sample vas composed primarily of vermiculite. Relatively little contaminant, nonvenniculite mineral matter vas present. Both tremolite-actinolite and anthophyllite amphiboles were detected. The tremolite-actinolite occurred almost exclusively in a very bulky, pris matic morphology. However, small elongated, parallel-sided particles could fracture from these prisms to yield particles classifiable as "fibers." So few large particles of anthophyllite were observed that it is impossible to confidently state whether or not truly fibrous anthophyllite vas present. Certainly, elongated fiber-like particles of anthophyllite were observed. The 2.76 float density-separated fraction contained some unusual, not totally characterizable material. Irregular, light green particles composed of microcrystalline material rather similar to serpentine minerals were pres ent. This material vas obviously an alteration state intermediate between amphibole and later-stage products, and may veil include serpentine minerals, though none were identifiable as such. A summary of the EM results for this sample appears in Table 33. C^to Oft-t 4>-0 430*1 Kxfelieted 430-0 fafeUaced tAiti a . gmunr or m ere* tiooscorr tts a in ret i w t i poog. to m cootm . c u g . cum 3 lWn ef all im thi ttjv f M M I t n U M 114* liaryf> Cm c u k h i m u Catin u t a u i SI Cmfwr 1 fiber m m tm ia i baa iauwal litreui im> ill Tlton im tit Ui 3.0 m i leattt "n c r e t K M i r a U M 110- (iMrs/r) baetitaallM ataiealaat U b unted oaaa b . ef 51 C w ilbct 1 fiber ceaceatrattea fibers fiber baa laLarval detected tented t.O 4.1 I .H .I 0.3 0.1 0J) i to '4 0.2 JO 0.2 1 0.33 1 4 0.1 1 T i. 0.3-4.1 i.m * 1 9 < 0.3 o.m O 4 0.3 t.itf 0 < 0.3 * 0.1ft - i C 3.1 3.T 12 0.3 3.1 3 4 f.S 0.3 34 1.3 0.3 S T o C 3.4 0*3.4 0.402 1 0 0.3 0.402 0.3 0.402 0 4 0.3 0.402 0 4 4 c a ! ^ k ib slt C A ry se tlU : aa< T Meat. 57 Sample 433, Enoree, South Carolina, Grace, Grade 4, Composite IITRI Code No. 127, ORF No. 432 Appendix references Electron microscope 1-114-119, 11-403*420 Hacroscopically, the sample was observed to be composed of 0.5- to 4 -mm goldish-brown flakes of micaceous mineral and 0.2- to 2-mm white, brown, tan, and green nonmicaceous minerals. No obviously fibrous phases were detected in the macroscopic examination. The mineralogical composition of the sao>le, as determined by polarized light microscopy analyses of the density-separated fractions, is listed in Table 34. TABLE 34. COMPOSITION OF SAMPLE 433-1 Mineral phase Estimated mass concentration (%) Fibrous mixed amphibole Anthophyllite-prismatic Tremolite-actinolite Sphene, ilmenite Augite Apatite Hornblende Magnetite, hematite ' Rhodonite, pyrolucite Calcite Quartz, feldspars Talc . Vermiculite Other minerals < 1 1-3 1-4 < 1 < 1 1-3 2-5 1-2 < 1 < 1 3-8 1-3 75-81 2-5 Both tremolite-actinolite and anthophyllite amphibole mineral phases were detected in the sample. Neither amphibole was found in a bulk grain morphol ogy that could be considered fibrous. That is, the as-received mineral grains were definitely prismatic crystals which were, for the most part, stout, sin gle crystal prisms. Less than 10% of each amphibole occurred as grains com posed of the elongated, thin, narrow lamellated prisms that could readily fracture to produce particles definable as fibers. Lamellations tended to be imperfect; thus, fractured fragments, though elongated, did not have parallel sides and frequently had unevenly terminated ends. 58 Tremolite-actinolite prisms were generally pale green and pleochroic, and exhibited inclined extinction. The anthophyllite was colorles (in transmitted light), had refractive indices lower than the tremolite-actinolite, and exhib ited parallel extinction. Tremolite-actinolite was occasionally found intergrown with hornblende. The anthophyllite and tremolite-actinolite prisms were never found intergrown with each other in the same mineral fragment in this sample. The relatively high proportion of the 2.76 sink density-separated frac tion reflects the degree of intergrowth of the amphibole and pyroxene mineral phases with the vermiculite. In general, the vermicullte plates of this sam ple were more irregular, strained, and Intergrown with other mineral phases than were the vermiculite plates of most of the Libby, Montana, samples. At least two other micaceous mineral phases that could not be identified were present in addition to the vermiculite. As these phases were micaceous, they were included in the mass accounting for vermiculite. A summary of the EM results for this sample appears in Table 35. M afia 433*1 433-0 433-1 l i f r l U u i * 433-0 Exfoliated m u g . P t m r y tucwaf kictoscdpt ra u tn rot t w i t tnoax. w n ciuatw. qj>a. cucc ______ ' yiMiNi u f it e f Mm T lte t * a f >11 l i r t i t m m t n tJw tic* liiw ti/t| C M C M inU aa mm SI bafllM C i interval 1 liber detected m m m Uh (ffi) ti M. liter* w ti< fib er* treater r t it >.< m it Iw rtl liM r n x H tw tw C aaca iiriU M a^unUst m liiiMUim m t S l Cntfifontt I liter m m irm a t Mm interval itta e tr i 4ypn> fo. t( liter* -- ted rites type* 1.4 4.7 3.7 d I.J 3.1 M . 8.3 I.M J f.H .1 0 0.1 8.344 8.344 0.4 0.319 0.319 u 31 .I 4 1 .1 0 0 11 0.3 0 d 0.3 1.4 13 .* 0 w d 0.3 0 d 0.3 0-8.3 * - 0.1 8.344 8.144 0.4 0.319 0.319 . 31 8.04 3 4 1 t 8 C 1 4 8 C 1 4 1 f C 8 4 8 C e A * aapAitele (IU D )| t d r iN U U ] M l T u u l . Sample 427, Enoree, South Carolina, Grace, Grade 5, Composite IITRI Code No. 119, OKF No. 426 Appendix references Photographs E-8-10, 21-23, XRD 1-139-142 Electron microscope 1-66-78, 11-204-334 The sample was a goldish-brown powder composed of less than 2-mm flake like particles. No obviously fibrous phases were detected in the macroscopic examination. Submillimeter grains of nonmicaceous gold, green, colorless, and white minerals were observed in the stereomicroscopic examination. The mineralogical composition of the sample determined by PLM analyses of the density-separated fractions is listed in Table 36. 59 ( TABLE 36. COMPOSITION OF SAMPLE 427-1 Estimated mass : _____ Mineral phase___________ concentration (%) i Fibrous mixed amphibole < 1 Anthophyllite 4-6 f Tremolite-actinolite 2-4 Sphene, ilmenite 1-3 f Hornblende 2-5 i a Apatite 1-2 * Magnetite, hematite < 1 f Rhodonite, pyrolucite 1-2 Calcite < 1 Quartz, feldspars 3-6 Talc 1-3 Vermiculite 72-78 Other minerals 1-3 Definitely fibrous mineral phases were detected in the tetrabromoethane sinks fraction. However, the fibers were well below 10% of the TBS fraction and thus were less than 1% of the total sample. Analysis of selectively re moved fibers indicated that both anthophyllite and tremolite-actinolite were present within the fiber bundles. The anthophyllite was identified by its parallel extinction and its slightly lower refractive indices compared to.the tremolite-actinolite. r * The anthophyllite and tremolite-actinolite occurred primarily in very clearly prismatic crystal habits. However, grinding of the prisms did produce parallel-sided, elongated particles which could be classified as fibers. Talc was found both as free plates and incorporated within the fiber bun dles. The talc incorporated within the fiber bundles tended to be fibrous in morphology. Some pseudomorphically fibrous mineral phases were found in this sample.' Some of the sphene (titanite) and fluorapatite were present as fractured' frag ments morphologically characterizable as fibrous. A summary of the EM re sults for this sample appears in Table 37. ` 60 1 U t *7 . n m u r r or lu c n to o m a r e o r r m o m n > t v n t w o r n n u n CAitoLtVA. o u a . c u n 3 ] riW n of a ll I m t U _____r m n m a t e r U m l . t -- ia l m t k fiber (M CM tm iM lie* fiW n /t) n w r cM tM tn U M 110* fib e r*/a * 1 Carr u n t s -- C--ca atrett-- M lftlM t u litlM ta f t t Se. of M t i l m t to tetlaated a a u Ca. t f tSX f M f U - r . | (|W , m c m u iUh fitocn 931 Ceaftde*c t 1 (D u M c n t m l u ( D * n flW r c^u tom totam l duaetal -- ref a-- fteterrel touctal W *l -- tad type* t 437*1 4 1 34 14 4 3.1 3.1 31 t.9 3 4 1 4 4 f 3 437* 1.1*33 i.n t 3.4 4-4.4 147 427*1 m i* >.4 4-4.3 1 4 7 3.37 427-0 Cl)1 n KHt l. t.4 4-4.4 1 .0 f t.i 4*14 1.39 1 S3 tl-34 141 1 n u *to o 14! 1 44 130*140 1.31 437*1, b ftU a tW u 34 I 4174, b felleu d t.t 3 .9 *4 .m I 4 4.3 * 3.313 437-1 (0) x *.t b lalU U T 34 f ' 4 3 7 * (X) A 4.4-4.1 . 34H i to ie lle fo C t.f to-i.f 3.371 7.3 1.4*4.0 3.371 7.3 4.7*11 3.377 U 13*31 . 3.371 f 17 13 1.1 0 1 3 14 4.1 _ 43 34 1 M * 1 0-4.4 147 3-4.1 1.17 3.37 33 4 < - C 1.7 4 C DC 0 1 34 t.t 3 1.9 4*4 . . 1.97 3 * 0 1 4 1.9 1.97 m 3 13 1.3 1.91 vr 370 41 3.3 0*19 I4 t 340 CAB 100 131 13 0.3*33 1.25 440 f 4.1 * i.i 17 14 3 9.4 4 34 T C 130 3 14 0-3.3 3.973 120 4 3 < 3.3 3.933 C 7.1 13 14 . 44 4 34 t 3 C u 3 3 34 3*1.3 i.m < 1 3 0 34 3.977 3 4 C 7 14 i.i 3*1.3 3.971 1 cr 33 tf 34 0*1.9 3.977 M CAB If i.t 1.3-3.3 3.971 92 T 1 4 U b * U i C A r m t l i i i f u u l : 4 t o i t o a t i l M ito t n l fib e r ; C A B -- abnt-- ml-- ra t; ato TH t a k l i r w n baU rj. aet fd c a tifia b le a t dtoryaetiU. 4 fW Utter to U r M m U m i itolcaua th at tto tM tto i iwcitow m i U a t -- ally a n d tor ttoa --her letoeraten m 344*1 (0 ) ere tW -- l i e o tteia e4 i n u M toi U r fltr y im to w . 9 t Sample 436, Enoree, South Carolina, Grace , Head Feed, Composite IITRI Code No. 131, ORE No . 435 i Appendix references Photographs -10-12, XRD 1-143-147 a Electron microscope 1-53-59, 11-167-203 Macroscopically, the sample was observed to be quite varied in grain size, grain morphologies, and grain colors. Black to gold micaceous flakes up to 5 mm in diameter were major components. Dark to pale green glassy chunks up to 3 mm in diameter were present. Colorless to white irregular glassy frag ments, some with obvious iron-staining, were as large as 15 mm. Milky white to light green irregular chunks were also as large as 15 mm. The composition of the sample determined by PLM analyses of the densityseparated fractions is presented in Table 36. This sample was primarily non micaceous contaminant minerals, with less than 50% vermiculite. 61 V1 f ( I 4 I I 1 1 ii i i ! TABLE 38. COMPOSITION OF SAMPLE 436*1 Mineral phase Estimated mass concentration (%) Fibrous mixed amphiboles Anthophyllite-prismatic Tremolite-actinolite Sphene Hornblende , Apatite Magnetite, hematite , Rhodonite, pyrolucite Calcite Quartz, feldspars Talc Vermiculite Other minerals < 1 1-3 6-9 2-4 11-15 2-4 1-3 1-2 1-2 23-28 3-5 32-40 1-3 Fibrous amphibole mineral phases vere detected, mostly in the tetrabromoethane sinks fraction, but vere less than 1% of the total sample. Both entho* phyllite and tremolite*actlnolite fibrous amphibole phases vere detected. In addition, it is likely that fibrous hornblende vas also incorporated vithin the fiber bundles. The three major amphibole types present, anthophyllite, tremolite* ectinolite, and hornblende, occurred predominantly as prisms. Fracture of hornblende prisms to yield particles classifiable as fibers is unlikely. However, the prisms of anthophyllite and tremolite*actinolite vere obviously layered and cleavable to particles definable as fibers. Talc vas again rather abundant and vas also found as fracture fragments that might be classified as fibers. The milky green, rough textured, irregular mineral grains vere isolated from the TBS fraction and analyzed separately. Morphologies of the crushed fragments produced in grinding ranged from irregular to elongated prisms. ` Color and extinction characteristics (as observed on paralled*sided fragments) vere consistent with tremolite-actinolite, but refractive indices vere slightly lover than the indices of the glassy, obviously prismatic fragments of tremolite-actinolite observed in the sample. X-ray diffraction studies of this phase indicated this material vas a sodium tremolite. A summary of the EM results for this sample appears in Table 39. 62 taafla 00-1 04-0 Ot-l. CiMiMri t o n t. fftw ir t r B te n M gtqoscofT ttsu m rot u m i p c m . to cn a a o c m . pa p m s w ic n IlKt lua tlU n of all iMtthi mcHCritlM | W* i i U n h \ Caafaatract-- n taU toa kum l M inlnt m 1 fiber dltMUi t s t i m m t aaaa IMCMUItlH ( > a. t (than W tl< ?ib*ra t n t u r U" * m |Mt r im McmntiM 10' fWnfot mb 151 Canfideace laum l M inlnt m 1 fiber detected CatiMtad aau cu.M tm ln 14 t.a u i i.i ?.*W Hi 141 .WO .140 u *4t U t t .11 . U t.S 2 14 n i.i i j u 4.4 - I.H 4 H .I .S i.u c I.1M C.4 .u u 1 a. tf flbara B8Mlal < 4 1 rihr 4 t 4 C c A * ^ H h l i (SAD}; C cfcnrw tlli; m l f tatal. Sample 439, Enoree, South Carolina, Grade 3, Commercially Exfoliated i m i Code Ko. 133, ORF Ko. Appendix references . Electron microscope 1-120-121 The sample vas typical in appearance of expanded vermiculite used as packing material or soil conditioning material. Individual particles were obviously composed of multiple, stacked vermiculite plates. Colors of the stacks ranged from white to tan to brown to light green. Diameters of the plates ranged from 1 to 5 mm. Lengths of the expanded stacked plates were quite variable and ranged up to 15 mm. Non nonmicaceous mineral phases were detected in the gross, stereomicrqscopic inspection of the sample. . Density separations did not yield much higher density (greater than 2.76) material. Table 40 lists the mineralogical composition of the sample deter mined by the polarized light microscopy analyses. TABLE 40. COMPOSITION OF SAMPLE 439-1 Mineral phase Estimated mass concentration (%) Fibrous mixed amphlbole Anthophyllite-prismatic Tremolite-actinolite Sphene Augite Apatite Hornblende Magnetite, hematite Rhodonite, pyrolucite Calcite Quartz Talc Vermiculite Other minerals < 1 < 1 < 1 < 1 < 1 1-3 < 1 1-2 < 1 < 1 1-3 1-2 85-95 1-3 63 Fluorapatite was the primary nonmicaeeous mineral constituent of both the tetrabromoethane and 2.76 sinks density-separated fractions. Apatite crystals were significantly larger in size and more abundant than any other nonmicaeeous mineral phase detected. Particles classifiable as fibers on a morphological basis, upon high mag nification inspection, were found to be mostly vermiculite shards and scrolls. Refractive indices were a major characteristic observed to distinguish vermic ulite "fibers" from amphibole fibers since vermiculite refractive indices are significantly lower than tremolite-actinolite and anthophyllite refractive Indices. Tremolite-actinolite and anthophyllite were present as coarsely prismatic material and as fine fractured particles classifiable as fibers. The pris matic crystals each comprised less than 10% of each sink fraction and were thus each less than 0.1% of the total sample. The fiber-like crystals were present at a count rato of one per 1,000 particles; on a mass basis, therefore, their concentrations would have to be in the parts per million range. A sum mary of the EH results for this sample appears in Table 41. lam fU O hr tUtt 41. WWIT er IlggTBOWITQOSC8WIBPIWW Kfftf POttl. 0PT1CtfOttTU. OACCI, tttOClAl BWtlTlOi _ fiU n at all U w l _HWf wewtml--HO* ItNn^) , C oc--t n i l i a ffU n ln t u U U atui m m f i t f --fKanrw 1 liter w rm ritU t lu a (au ral Irtaciai if f ) la. f (iter* m m i*i ?itew t w m t U t I J .fiber ttaewtnUw UP* flWrt/t) tM cn tiiiiia u f i t Cn H I mcc I liter lu a toum l Irtattai w it U ttU titiaiu i m m --rn tra ti-- (f*e) l a . %l liter* w rte fib e r ifft* II .? 14 .9 I I i 1 .1 CT i M fbiteU (H O )i c *f T * * t i i* ; * ! f te u i. Sample 442, Enoree, South Carolina, Grade 4, Commercially Exfoliated This sample was a fine-grain, expanded vermiculite. The expanded, stacked vermiculite plates visible in this sample ranged in diameter from 0.5 to 3 mm. lengths of the stacked plates ranged from 1 to 5 mm. Particle colors were white, tan, brown, and greenish brown. No nonmicaeeous mineral phases were . detected in the macroscopic inspection. Density separations of this sample also produced relatively little high density (greater than 2.76) material. The mineralogical composition of the sample determined in the polarized light microscopy analyses of the various density fractions is presented in Table 42. 64 TABIE 42. COMPOSITION OF SAMPLE 442-1 Mineral phase Estimated mass concentration (%) Fibrous mixed amphibole Anthophyllite-prismatlc Tremolite-actinolite Sphene Augite Apatite Hornblende Magnetite, hematite Rhodonite, pyrolucite Calcite Quartz Talc Vermiculite Other minerals < 1 < 1 0.5-1 < 1 < 1 1-2 < 1 1-2 < 1 < 1 1-2 < 1 85-95 1-2 Tremolite-actinolite was the primary nonmicaceous mineral type in both the tetrabromoethane and 2.76 sinks fractions. It occurred primarily as coarse, chunky prisms. Up to 10% of the tremolite-actinolite occurred as overall prismatic fragments composed of elongated, thin, narrow lamellated prisms. This crystal form undoubtedly produced many of the small, fiber-like crystals observed. The larger anthophyllite fragments present occurred only as the. chunky prisms. The very 6mall (less than 10 (tm diameter) amphibole crystals present in morphologies definable as fibers were both anthophyllite and tremolite- actinolite. It was impossible to determine if these fine, fiber-like crystals were abraded from large bundles of truly fibrous material or were fractured from the lamellated prisms. Number concentrations of the fine, fiber-like amphibole crystals were greater in this sample compared to sample 439-1. Eovever, mass concentrations must be considered to be again in the parts per million range. / Sample 573, Enoree, South Carolina, Patterson, Ungraded, Composite IITRI Code 124, ORF No. 572 Appendix references Photographs E-13, 24-25, HU) Electron microscope 1-79-83, 1-148-152 11-335-375 Macroscopically, the sample was observed to have a wide size range and to be a brownish-gold material with some obviously micaceous flakes. Brownishgold, nonmicaceous grains ranged up to 10 mm in size, while the micaceous 65 flakes vere 7 to 6 am In maximum dimension. Fragments of nonmiceceous min erals up to 20 cm in diameter vere present* Colors of the nonmicaceous min erals vere milky white, milky green, and glassy green. Ko obviously fibrous phases vere observed in either the unmagnified or stereomicroscopic examina tions . The composition of the sample, as determined by PLM analyses of the density-separated fractions, is listed in Table 43. TABLE 43. COMPOSITION OF SAMPLE 573-1 Mineral phase Estimated mass concentration (%) Fibrous mixed amphiboles Anthophyllite Tremolite-actinolite Sphene, ilmenite, rutile Hornblende Apatite Magnetite, hematite Rhodonite, pyrolucite Calcite Quartz, feldspars Talc * Vermiculite Other minerals < 1 4-8 8-12 1-2 1-3 1-2 1-2 < 1 < 1 26-32 12-16 33-38 1-3 This sample appears to have been exposed to some type of heat treatment. Glassy agglomerates vere observed in the total sample and vere, of course, concentrated in the 2.76 floats fraction. No obviously fibrous bundles vere observed in the stereomicroscopic exam ination of the TBE fraction. Small, elongated, coarse fibrous to prismatic vhite particles vere observed in the TBE fraction; however, they were in greater abundance in the 2.76 sink fraction. In the TEE, this particle type vas less than 1% of the fraction, while in the 2.76 sink the prismatic to coarse fibrous phase represented 10 to 20% of the fraction mass. These par ticles vere generally not teasable with a fine needle and thus are not in a true fibrous habit. Eowever, gentle crushing and grinding produced long, thin parallel-sided particles which would be classifiable as fibers. The TBE fraction contained numerous pale green, prismatic amphibole min eral particles which vere determined to be tremolite-actinolite. Again, even these clearly prismatic particles could be fractured to yield particles defin able as fibers. Although the anthophyllite comprised 10 to 20% of this frac tion, practically all of it also occurred in an obviously prismatic crystal 66 habit. While fracture of the prismatic anthophyllite could yield fragments classifiable as fibers, this fracture was not readily accomplished; irregular, jagged fragments tended to be produced. The prismatic to coarse fibrous mineral phases found in abundance in the 2.76 sink fraction vere isolated and carefully examined. The particles were found to be composed almost exclusively of talc and anthophyllite. Tremoliteactinolite vas only a trace constituent of this fraction. Grinding of the particles resulted in ready fracture of both the talc, and anthophyllite into long, thin, parallel-sided fragments classifiable as fibers. Larger fragments showed splintered ends Euggestive of fiber bundles. Unlike the Grace samples from South Carolina, the Patterson sample con tained predominantly rutile rather than sphene titanium phases. Some of the rutile vas found in elongated, thin crystal habits. A summary of the EM re sults for this sample appears in Table 44. . t e a * La W t f u . t m a r r a t w rtcrtea w o o v e rr r n t s fe W M 1 * * . Kxmr CAiotiiu. f*T T t*3 ta. MOAOO Viter cw m tM tiM bu iatam l Fiber* o f a ll laatila 110* ( iW n / it (M CM tim aa f t f n l a t to 1 lik t ftatooted jit la a l ii aaas MOCMtettiSt 21] Viter Mao F ilar* rraeter that M C M n e tc iU K (10* ClWra/rt Cm m i u i u m otKiatloat to VS1 CasfM --na 1 UWr la u m l ftatecteJ mo is lean t C ftlaitai m ss Mscastn liw Crr-) 173*1 i.n i.< ] 1 .1 1 * ' 1 0 BM *. .* K *` 1 1134 i.i I.M .I .*u I.K 4 V ftt ft.S < IJ 1 .2 . B JU IJU tc |1 > I , IsfolUCOi u *. i .i M S U U 11 u 1 .* * 1 1.4 .* 1114, IxlaliaU i i.i ft-1-1.ft t.tu * .w ft fftt 1 .1 < ft.l H .t I.2&S BM i ft A * a*p41ble (UO); C AiyutiU; *M T * total. b . of riW rs oaeotiV 1 fSt 1 1ft f1t Fiber W f* ft tft fftt ft t ft Air Sample. Phase Contract Results The results of the examination of the air sample filters by phase con trast analysis are presented in Table 45. 67 TABLE 45. RESULTS OF THE PHASE CONTRAST ANALYSIS OF AIR SAMPLES _________ COLLECTED AT THREE VERMICULITE SITES__________ Sample Sample vol. () Fibers/cc ORF IITRI Libby* Grace 106 Field blank? 133 Field blank11 131 Front loader 148 Pit haul driver 138 Mine analyst 141 Bottom operator 130 Ko. 2 operator 139 Dozer operator 101 Shuttle truck 104 Screening plant, DV 111 Screening plant, DW 108 Trailer court 136 No. 5 substation South Carolina, Grace 312 Field blank? 346 Field blank 340 Mill monitor 321 Mill lab technician 301 Dragline operator 347 No. 4 bagger ' 330 No. 3 bagger 328 Mill (ENE) downwind 335 Mill (N) crosswind 307 Mine (N) crosswind 323 Mine (E) downwind 338 Mine (W) upwind 310 Truck driver 300 Screening plant floor . South Carolina, Patterson 505 Field blank? 533 Field blank 508 Payload operator 520 Plant foreman 542 Bagger/forklift 513 (NE) downwind 506 Control off-site 515 (SE) crosswind 528 (SW) upwind < 0.02 0.04 - 0.03 0.05 303 0.02 0.04 297 < 0.01 0.01 294 1.5 1.9 276 1.2 0.4 285 3.1 9.7 270 0.02 0.2 385 0.1 0.2 390 0.08 0.5 368 0.1 0.02 169 0.03 rar 111 0.03 0.02 < 0.02 0.04 - < 0.02 0.02 340 0.03 0.03 478 0.07 0.2 240 < 0.01 ND 314 0.06 0.1 285 0.1 0.05 287 0.05 0.04 80 0.04 ND 291 < 0.01 0.02 154 0.01 0.02 264 0.03 0.01 257 < 0.01 0.3 ..... 354 0.06 0.14 - 255 * 252 249 188 274 299 147 < 0.02 < 0.02 < 0.01 0.01 < 0.01 < 0.01 < 0.01 0.01 0.02 < 0.01 0.02 0.04 0.3 H ND ND 0.01 ND a Values for blanks were calculated assuming a 100-liter sample, b ND: No fibers detected (100 grids). 68 REFERENCES 1. S p u m y , E. R., V. Stober, H. Opiela, and G. Weiss, "Size-Selective Preparation of Inorganic Fibers for Biological Experiments," American Industrial Hygiene Association Journal, 40, 20-38 (January 1979). 2. Chatfield, E. J . , and G. H. Levis, "Development and Application of an Analytical Technique for Measurement of Asbestos Fibers in Vermiculite," In: Scanning Electron Microscopy/1980/I, SEM, Inc., AMF O'Hare Chicago, Illinois, p. 329-340, 328. 3. Trimbrell, V . , " T h e Inhalation of Fibrous Dusts," Annals of New York Academy of Sciences, 132, 255-273 (1965). 4. Going, J . , and J. Spigarelli, "Environmental Monitoring Near Industrial Sites: Vinylidene Chloride," EPA-560/6-77-026 (1977). 5. Going, J . , "Environmental Monitoring Near Industrial Sites: Acrylamide," EPA-560/6-78-001 (1978). 6. Going, J . , P. Euykendahl, S. Long, J. Onstot, and E. Thomas, "Environmental Monitoring Near Industrial Sites: Acrylonitrile," EPA-560/6-79-003 (1979). - * .. 69 APPENDIX A STUDY PROTOCOL FOR THE COLLECTION AND ANALYSIS OF VERMICULITE AND RELATED SAMPLES FOR THE EVALUATION OF FIBER CONTENT WITH EMPHASIS ON ASBESTIFORM FIBERS TASK 32 STUDY PROTOCOL FOR THE COLLECTION AND ANALYSIS OF VERMICULITE AMD RELATED SAMPLES FOR THE EVALUATION OF FIBER CONTENT WITH EMPHASIS ON ASBESTIFORM FIBERS I. Background In December 1978, the venniculite industry submitted information to the EFA regarding health problems experienced by employees who were processing asbestos-contaminated vermicullte. The original submission indicated that bloody pleural effusions had been detected in 4 of 350 employees; symptomatol ogy and clinical findings in the employees were similar to those found in in dividuals vith asbestos-related diseases. Subsequent follov-up studies by the Occupational Safety and Health Administration (OSKA) revealed an even higher prevalence of health problems among the employees. Venniculite is a hydrated magnesium-iron-aluminum silicate which has been mined in the United States since 1929. After mining, venniculite is pro cessed to remove impurities, including asbestiform minerals; however, all con taminants are not removed. Information suggests that the three major domestic deposits in Montana, South Carolina, and Virginia contain asbestiform minerals. Some impurities, including asbestos, may remain a6 a contaminant in processed vermicullte. . Although vermicullte may contain fibrous materials, the health ef fects from vermicullte itself are unknown at this time. A priority review of asbestos-contaminated vermicullte, completed by the Office of Testing and Eval uation in June 1960, suggested that the asbestos in vermicullte may be respon sible for the reported adverse health effects, and it concluded that certain information gaps needed to be filled before an in-depth risk assessment on vermicullte could be initiated. * Several projects have been initiated to fulfill the information gaps and complete the preregulatory analysis on vermicullte. A control options analysis has been initiated to determine regulatory strategy to control asbes tiform mineral-contaminated vermiculite, and a substitute analysis is in prep aration to evaluate replacements for vermiculite products. Work has also been initiated on a materials balance to show the mass flow of vermiculite along vith the release of any associated asbestiform mineral, and the development of a mineralogy profile vith a sampling and analysis protocol of vermiculite is underway to characterize the fibrous materials within vermiculite. From the available Information on the composition of vermiculite, it seems there is the possibility that asbestos contamination of vermicullte does occur, but that it may be difficult to assess the magnitude of the contamina tion. Therefore, the objective of this protocol is to specify the sampling and analysis procedure to determine the composition of vermiculite, particu larly the amount of asbestiform minerals1 present in the vermiculite. This* * For practical analysis purposes, the specific identification of asbestiform minerals will be limited to chrysotile, the amphiboles, and vermiculite scroll. A-I will provide the needed information on the risk to the population exposed to asbestifozm minerals from vermiculite at each of the various stages of its commercial distribution. The protocol will be conducted in two phases. The first phase will be an in-depth analysis of the asbestiform fibers present in and associated with vermiculite o ret ore concentrates, and beneficiated vermiculite from the four major vermiculite mines in the United States and of beneficiated vermiculite from the ports of entry. Both bulk and air samples will be collected and analyzed. The second phase will be a similar analysis of bulk and air samples from a representative number of exfoliation plants in the United States. The exfoliation plants where sampling will occur will be statistically chosen b y Exposure Evaluation Division (EED) to include all the major sources of vermiculite. II. Preparation for Sampling A. Inventory Supplies 1. All necessary equipment for sampling bulk vermiculite and air will be gathered and inventoried. 2. Filters to be used for the collection of airborne particles will be assembled and labeled before samples are obtained. 3. Calibration of the pumps will be performed prior to their ship ment to the sampling site and recalibrated in the field. 4. All supplies will be packed and shipped to the vermiculite sam pling site at least 2 days in advance of the arrival of the crew. B. Site Investigation 1. Survey site - Upon arrival at the site, the crew chief and other designated persons will survey the site to determine the location of the facility, its boundaries, and the locations of various operations within the facility. . 2. Select sampling points - The crew chief and other designated persons will select appropriate points for the collection of ore samples and airborne particulates. Officials of the host plant will be invited to par ticipate and assist in the survey and selection of sampling points. III. Sampling A. Bulk Material 1. Basis for selection of protocol - No American Society for Test ing Haterials (ASTM) method was found that is directly applicable to this situ ation. The following related methods are used for guidance. A-2 a. American National Standards Institute (ANSI)/ASTM D 75-71 (1978) Standard Methods of Sampling Aggregates. b. ASTM Designation: D 2234-72 Standard Methods for Collec tion of a Gross Sample of Coal. c. ANSI/ASTM E 105-58 (1975) Standard Recommended Practice for Probability Sampling of Materials. d. ASTM Designation: C 702-72 Standard Methods for Reducing Field Samples of Aggregate to Testing Size. e. ASTM Designation: C 516-75 Standard Specifications for Vermiculite Loose Fill Insulation. f . BS 812 (British Standards Institution) Methods for Sam pling and Testing of Mineral Aggregates, Sands and Fillers. g. Other Considerations - The minimum quantity of any sample should be 5 to 10 times the anticipated analytical needs. The analytical needs vill vary with particle size and range from approximately 40 g for fine material to 1,000 g for 25-mm particle size. Therefore, sample size minimums should range from 400 g to 10 kg. Each sample may consist of a composite of individual sampling increments representing different times and/or locations. Increments vill be sampled and stored separately with a composite made under laboratory condi tions by combining representative fractions of each increment. 2. Samples to be collected - The objective of sampling is to ob tain samples that are representative of the operations or sites. It is an ticipated that properties of the materials of similar types vill vary vith time, operation and specific mine site origin. Therefore, to obtain repre sentative samples, it is necessary that composite samples be prepared of a given sample type from individual sample increments, each representing a specific sample time or 6ite. It is likely that a historical sample collec tion is maintained (by the mine company) from several operation points vithin the facility. If these historical samples are available, it would be helpful to obtain selected increment samples for both the preparation of a time aver aged composite and a comparison of present to past conditions. The number of increment samples to be collected must depend on the variability of the sample and availability of increment sources, vith a de cision made by an experienced sampler depending upon increment availability and proper sampling procedures. All decisions vill be documented vith copies sent to the EFA task manager. ' a. Raw ore and ore concentrates (Phase I) (1) ing different parts of the mine. A bulk sample of the raw vermiculite ore rep A-3 ficiation. (2) A bulk sample of the concentrated ore before bene- (3) A bulk sample of dust from the dust collection equipment where such equipment exists. tions. (A) A water sample from washings and dust control opera* (5) A bulk sample of concentrated ore blend (beneficiation feed) before beneficiation. b. Beneficiated vermiculite (Phase I) vermiculite. (1) A bulk sample of each of five grades of beneficiated (2) A bulk sample of material from one to three inter mediate beneficiation processing steps. process. (3) A bulk sample of tailing from the beneficiation ment.- (A) A bulk sample of dust from dust collection equip- tions. (5) A water sample from washing and dust control opera- c. Exfoliated plant samples (Phase II) (1) A bulk sample of each of the five grades of vermicu lite before exfoliation. vermiculite. (2) A bulk sample of each of five grades of exfoliated (3) A bulk dust sample from dust collection equipment and other appropriate related material. B. Air Samples (Phases I and and II) 1. General considerations - Airborne particulate samples will be collected at designated points inside and outside the plant boundaries. The sampling will generally follow the EPA cethod described in Electron Microscope Measurement of Airborne Asbestos Concentrations - A Provisional Methodology Manual, EPA-600/2-77-178. Revised June 1978. This method recommends poly carbonate O.A pm Nuclepore filters when possible, but allows for the use of cellulose acetate (Millipore) filters. There are advantages and disadvantages to the use of either Nucle pore or Millipore filters. The sample collected on the smooth polycarbonate A-A (Nuclepore) suface has poor retention efficiency and the sample may he lost or redistributed during transport. The cellulose acetate (Millipore) filters requires an ashing procedure and reconstitution on a polycarbonate filter for T E H analysis. Ashing and reconstitution is an extra 6tep in the procedure but has the potential advantage of eliminating interfering organic particle analysis. This reduces the need of multiple time sampling to obtain a range of filter loadings. Asbestos contamination has been reported in 6ome lots of both Nuclepore and Hillipore filters. The current consensus of leaders in the field is to favor the use of Hillipore filters for the type of sampling and transport that will be re quired on this program. The potential presence of asbestos fibers in the filters themselves vill require careful attention to the selection and analysis of filter blanks and field blanks. 2. Control blanks a. Filter blanks - Four filters from each package of filters, one from each box of 25, vill be selected by random numbers filter blanks. The four filters vill be quartered and one-fourth of combined as a composite saople. 100 as each b. Field blanks - One of every 10 filters vill be a field blank, subjected to all processing conducted vith an actual air sample except for the sampling itself. 3. Sampling procedure - All sampling, fixed and personal, vill be taken using 37 mm 0.45 pm Hillipore filters backed vith 5 pm Hillipore filters and a Hillipore support pad. The sampling rate vill be approximately 2 liter/ min vith the exact rate determined periodically throughout the sampling period by the use of calibrated flow meters. Sampling vill be ground samples). However, indicating that the filter and flow rate recorded. scheduled for S hr (or longer for ambient and back if the flow is found to reduce during sampling, is loaded, sampling vill be stopped and the time 4. Air sanples vill be collected - Personal air zone samples and fixed samples located at targeted areas vill be taken. The use of personal samples vill depend on the individual work patterns and on the cooperation of the host company. When possible, individuals at each work station vill be equipped vith personal samplers. When possible, air samples vill be taken which correspond to the bulk samples (Section III-A-2). The following air samples vill be taken. Any variation from these sampling locations vill be documented. (Phase I). a. Air samples taken at operations before beneficiation A-5 (1) An air sample at the mine during mining. (2) An air sample at dumping or crushing operation. (3) Air samples around the mining facility. (4) Air samples dovnvind of dusty operations. (5) Air samples along shipping lanes. geographic region. (6) An air sample as a background control vithin the b. Air samples taken from the ore beneficiation operations. (Where mining and beneficiation are at the same general location, several of the samples may be the same.) (Phase I) (1) An air sample at each of selected vork stations in the beneficiation operation. (2) Air samples around the beneficiation plant. (3) Air samples at grading (screening) operations. (4) Air samples downwind from dusty operations. (5) Air samples along shipping lanes. geographic region. (6) An air sample as a background control vithin the c. Air samples taken from the exfoliation operation. (Phase n) (1) An air sample from each of selected vork stations in the exfoliation operations. (2) Air samples around the exfoliation plant. (3) Air samples dovnvind from dusty operations. d. A meteorological station will be installed on-site to col lect air speed and direction data throughout the sampling period. IV. Sample Handling A. Bulk Samples The increment samples will be shipped to a central laboratory. Each increment sample will be divided by appropriate procedures (riffle di vided or cone and quarter). Part of each increment will be retained and the remainder combined with other appropriate increments to form a composite sam ple. The composite will be mixed and split (riffle or cone and quartered) to provide appropriate split analytical samples. The composite samples will be properly designated and submitted for analyses. B. Air Samples 1. Special handling - When sampling is completed, the filter' cartridge will be turned to a position with the filter horizontal and -the collection surface up, the cartridge disconnected from the pump, the car tridge cover replaced, and the inlet and exit holes plugged. This horizontal filter position will be maintained during transport and storage. The car tridge will be placed in a special container for transport. 2. Each filter will be divided and each portion will be taped to the bottom surface of the petri dish and delivered to different laboratories to provide for replicate analysis. V. Sample Analysis A. Bulk Sample Analysis . The analysis protocol for the bulk vermicullte samples will include parallel approaches which, to a degree, support one another. However, because of the great differences in the detection limits of the different methods, the justification of some approaches is their simplicity as preliminary screening procedures rather than their sensitivity. X-ray diffraction of the unfractionated samples is an exasq>le of a simple procedure with limited sensi tivity. These methods may serve to identify some samples with gross quanti ties of asbestos and eliminate the need for continued analysis. 1. Unexfoliated vermicullte, before and after beneficiation a. Examine the sample as received with a low power (30X) stereomicroscope for quantities of visible fibers. b. If fibers are observed, estimate the weight (%) of fibrous material. If appropriate, remove (hand pick) the asbestos from the sample and weigh. c. Identify the isolated asbestiform mineral by appropriate means (FLU, XRD, etc.). d. To isolate the fine fibers from vermicullte, start with a sample quantity depending on particle size. Place the sample in a specified beaker size and add 10 times the sample weight of prefiltered isopropyl al cohol . The sample quantities, isopropyl alcohol volume and beaker sizes to be used are as follows: A-7 Sample Grades 1 and 2 Grades 3 and 4 Grade 5a Unbeneficiated material and other tailings, etc. Grams 40 20 40 40 Volume IPA 400 200 400 400 Tall Form Beaker (ml) 1,000 400 1,000 1,000 a Grade 5 is expected to have more variability than the other grades . Place the beaker in an ultrasonic bath, stir, allow the large particles to settle (during ultrasonic treatment) and withdraw aliquot por tions from near the center of the liquid for optical microscopic analysis (i.e., PIU) and to prepare a series of H u d e p o r e G filters for EM analysis. Serial dilution may be required to obtain optimum filter loading. . e. The Nuclepore filter with suspended fines will be used to prepare a TEM grid by the EPA carbon-coated Nuclepore filter technique. f. Hake fiber count - Determine Count 100 fibers or 10 grid of 200 mesh screen. tection and count more grids if necessary.. % ' g. Identify specific amphiboles diffraction or zone axis selected area electron persive X-ray analysis. chrysotile or amphiboles. Determine the limit of de . using selected area electron diffraction plus energy dis h. Exfoliate a portion of the beneficiated vermiculite by sprinkling no more than a one particle thick layer of sample, in a preheated (800cC) shallow container and place the container back into a 800C oven for 5 sec. Examine the exfoliated sample as described in No. 2.2 2. Exfoliate vermiculite (for laboratory expanded samples of Phase I and for Phase II) - An important feature of the analytical procedure to achieve high microfiber detection sensitivity is the fractionation of the sample to remove much of the interfering vermiculite, thereby greatly enriching whatever asbestiform fibers that may be present. The basis of fractionation is the floatation on water of the exfoliated vermiculite and the wetting and sinking of the asbestos and other fibers. This assumes that a proportionally high fraction of the fibers are not physically attached to the vermiculite parti cles. This is a reasonable assumption but one that will be verified by the examination of representative samples of the fraction that floats. A-8 I Starting with a sample quantity depending on particle size, or grade, i Vermiculite Grade Sample Weight (g) Water Volume (ml) 1 and 2 A0 2,000 3 and A 20 1,000 5 A0 2,000 and proceed as follows: a. Float separation - Place the expanded vermiculite in a 2,000 ml plastic beaker and add water. Stir for 30 sec and 6kim off the vermiculite and drain on a 50-mesh screen. Collect the drain water and re turn it to the beaker. Discard the vermiculite.2 ' b. Disperse the "6ink" material with ultrasonic treatment. Remove an aliquot during treatment for a preliminary PLM examination and for TEH analysis. Double dilution may be necessary to obtain proper grid load ing. ination.2 c. .Examine preparation b y PLM as a preliminary parallel exam . technique. ' d. Prepare TEM grid by EP A carbon-coated Nucleppre filter . - ' ; ' . ' . ' ' " . e. Fiber count - Determine chrysotile, total amphiboles and vermiculite scrolls. Count 100 fibers or 10 grids of 200-mesh screen. De termine limit of detection and count more grids if necessary. f . Identify specific amphiboles using SAED plus energy dis persive X-ray analysis. 3. Miscellaneous bulk samples a. Dust samples (1) A preliminary examination of the dust sample will be made by optical microscopy including PIM for the identification of gross quantities of asbestiform fibers. If gross quantities of fibers are identi fied, the quantities will be estimated and the analysis terminated.2 2 Selected samples will be examined to verify the absence of asbestos in this fraction. 2 If PLM examination reveals a gross quantity of identifiable asbestiform fibers, the quantity should be estimated and the analysis terminated. (Modified during the project to continue analyzing.) A-9 (2) If the cample ic not adequately characterized by PLM, a portion of the cample will he dlcperced in vater and filtered for E M analycic. b. Wash vater camples - The eolids present will be dispersed In the vater and aliquots filtered for appropriate optical and EM analysis. B. Air Samples - 1. Portions of selected filters vill be used to determine fiber count by the standard NIOSE procedure using phase microscopy. 2. The major analysis of the air samples vill basically follow that specified in the EPA document, "EPA-600/2-77-178, Revised June 1978, Electron Microscope Measurement of Airborne Asbestos Concentrations * A Provisional Methodology Manual. A-10 APPENDIX B DETAILED ANALYSIS PROCEDURES AS SUBMITTED BY IITRI I PROCEDURES BULK SAMPLE ANALYSIS t 1 Quantitative analysis requires that valid and rigorous procedures be used during all phases and steps of a procedure and that these procedures be 1 well-defined before work on the first sample is begun. The approach and logic used by IITRI was based on: . A study protocol prepared by KRI, * Discussions of the protocol with KRI, and Discussion on procedures with Ontario Research Foundation (ORF). The bulk sample procedure is presented in Figure B-l. Where we note that the first operation is, as with any sample, sample log-in. The bulk sample is then subdivided for two distinct series of sequential steps. The first begins * with analysis of the sample "as received" for gross (defined for this study as > 1% by weight) fiber contamination and characterization by polarized light microscopy. Samples which are not found to be grossly contaminated are moved into the steps of isopropyl alcohol beneficiation and electron microscopy analysis for fiber content. A parallel screening test for the second series of steps is whether the sample received is an exfoliated vermiculite or not. If it is exfoliated, no ffork is needed; if not, the sample is thermally exfoliated, the product is beneficiated, and the "sink" fraction analyzed by electron microscopy. In the subsections which follow, IITRI describes the procedures used for: Sample splitting Optical microscopy - Preliminary inspection - Sample separations - Polarized light microsocpy - X-ray diffraction Electron microscopy - Beneficiation a - Sample preparation - Electron microscopy analysis * Task 32-- Study Protocol for the Collection and Analysis of Vermiculite and Related Samples for the Evaluation of Fiber Content with Emphasis on Asbestifona Fibers (Revised November 13, 1980). ~ B-l S ta p le R e c e ip t a d lo ( { lo ( R iffle (R e tain l i l n c t ) n THERMAL m S in k Co O b tttin F in e i fcj r L peeee S lla te , n ite r , try To C l F ro to eo l C e a trlfe se o r F ilte r fo r n x EM A 1I q o o t ttte e S to re end In y en to ry C ry D ilu te , F ilt e r , end try T o EH P r o to c o l , % Figure B-l. Flow chart for bulk sample analysis. B-2 SAMPLE SPLITTING The samples were split into aliquots using a spinning riffler.2 The riffler Is a rotating tray containing sample receivers. As the tray rotates, the receivers ertract a sample from the flowing powder stream. This time- averaged sample consisting of many small aliquots produces a sample free of biases due to segregation variation in aerodynamic diameter and the powder's flow properties. . . Each fraction collected was placed in a clean, glass cream jar, labeled with the IITRI sample number and submitted for analysis or stored. IITRI riffled a minimum of four fractions from each sample-- one each for PLM, al cohol beneflciation, thermal exfoliation, and a back-up sample. The target size for each sample was: Twenty grams for vermlcullte Grades 3 and 4. * Forty grams for all others, including vermiculite Grades 1, 2, and 5. . OPTICAL MICROSCOPY The objectives of the polarized light microscopy analyses were to: * Determine if fibers were present, * Identify the fibrous phases detected, * Determine the concentrations of asbestiform phases in the bulk sample, ' and /. .* . " . ; . * Identify the prismatic mineral phases present that could fracture to yield "fibrous" particles. To achieve these objectives, several sample preparation steps and sup plementary analyses are used as integral parts of the polarized light micro scopy analysis. The sample separation steps enhance the (semi-)quantltative aspects of microscopical analyses which rely heavily on estimations of com ponent concentrations. The supplementary analyses, principally x-ray dif fraction, were conducted to establish irrefutable identities of phases-- especially those in a fibrous habit. Preliminary Inspections The bulk sample portions submitted for polarized light microscopy (PLM) were first inspected with a low power stereomicroscope to determine the number of different mineral phases present, the associations of the various phases, and the presence of fibrous phases. This preliminary inspection also served to determine which sample separation step (hand-picking of fibrous phases, or 2 ASTM C702 71T, Tentative Method for Reducing Field Samples of Aggregate to Testing Size. B-3 heavy liquid separation) should proceed first. Notations on sample color, texture and general particle size ranges vere made at this time. Sample Separations The objective of the sample separations was to concentration any fibrous phases to facilitate both the phase identification and quantitation tasks. Subsamples for the separation procedures vere obtained by coning and quarter ing* the sample fraction submitted for PLM analysis. The entire Fill sample fraction vas poured out of its container onto a clean piece 'of foil and quartered with a broad-bladed spatula. The desired subsample size (one or two quarters) vas retained on the foil and the remainder of the PLH sample vas returned to its container. For those samples that received a duplicate separation analysis, the coning and quartering vas repeated. Hand-Picking-- When the preliminary inspection revealed the presence of several bundles of fibers at least 1 mm in diameter and 3 mm in length, hand-picking of the fibrous phase(s) with a fine-pointed tweezers vas the first separation step performed. The separation subsample of the PLM sample vas weighed on a piece of tared foil and then spread to a monolayer of particles. While viewing through the 6tereomicroscope, the fibers vere tveezed from the 6ubsample and placed in a tared weighing pan. Hired particle types containing at least 25% fibrous material, a6 veil as totally fibrous particles were tveezed. The pan containing the fibrous phase as well as the foil containing the nonfibrous remainder of the subsample vere then reveighed (to 0.1 mg) and the mass per cent of "pickable" fibrous material vas calculated. When smaller bundles vere present, hand-picking vas done either after the first heavy liquid separation step or, for some samples, vas not feasi ble. The procedure for hand-picking fibers from the "sinks" fraction of the first heavy liquid separation is essentially the same as for the bulk sub samples . Heavy Liquid Separation-- Nonvermiculite mineral phases, particularly amphiboles and pyroxenes, vere separated from the vermiculite bulk samples on the basis of density us ing a simple sink-float method. The densities of the mineral phases must differ by at least 0.2 g/cm2 for this method to work. Table B-l lists the specific gravities of vermiculite, 6ome of the amphibole minerals, and other mineral contaminants commonly associated with vermiculite. ^ See Reference 2, page B-3. TABLE B-l. SPECIFIC GRAVITIES OF SELECTED MINERALS Mineral Chemical formula Specific gravity Venniculite Biotite Chrysotile Serpentine Talc Anthophyllite Actinolite Tremolite Ferroactinolite Cummingtonite Grunerite Diopside Hornblende Quartz Olivine (Hg,Ca)o (Mg,Fe,Al)s o(Al,Si)4O l0(OH)< K(Hg,Fe)|(AlSi20i6)(0H)2 M g 3Si20 (0E)4 HgsSi20 6 (0H)4 MgsSi40xo(OH)2 (Hg,Fe)7Si202 2 (0H)2 " 2(0H)2 2 C a 2F e s S i 20 2 2 (0H)2 (MgfFe)7Si202 2 (0H)2 Fe7Si2022(0H)2 _ * CaMgSi20e (Ca Ha)2-2(Mg,Fe ,A1) gSieCSi ,A1)2022(0B2) Si02 CMg,Fe)2Si04 2.4 2.B-3.2 2.5*2.6 2.3-2.6 2.7-2.B 2.85-3.2 3.1- 3.3 3.0- 3.2 3.2- 3.3 3.1- 3.3 3.6 3.2 3.0-3.4 2.65 3.27-4.37 The density separation technique is based on the principal that parti cles with a greater density than the liquid they are suspended in vill sink while particles with densities equal to or less than the liquid density vill float in the liquid. Mineral powder samples subjected to this float-sink procedure should be composed of single phase grains of fairly uniform size and less than 1 to 2 mm in diamter. As particle size approaches subsieve size (i.e., less than 37 pm), other forces including friction, particle shape, and thermal turbulence influence the settling characteristics of part icles as much as density. The separation of most bulk samples vill improve if grinding or sieving is performed prior to using the float-sink procedure. While this would have been true for the venniculite samples, these steps were omitted to preserve the integrity of the sample. Two density separations were performed on each subsample yielding three fractions based on density differences. The separation subsample was first suspended in 1,1,2,2-tetrabromoethane (TBE), specific gravity of 2.97; the TBE ''sinks" were the first density separation fraction recovered. The TBE "floats" were then suspended in a TBE/lsopropanol mixture (specific gravity of 2.76) yielding two other density-separated fractions, the "2.76 sinks" and the "2.76 floats." The procedure is described in detail below. The density subsamples were dried for 24 hr at 90C; then accurately weighed in tared 250-ml beakers. Approximately 200 ml of tetrabromoethane were added, and the mineral powder slurries were vigorously agitated with a spatula. The slurries were allowed to stand undisturbed for 12 to 24 hr. The top 160 to 175 ml of tetrabromoethane containing the "floats" were then B-5 - decanted into a 400-ml beaker and tetrabromoethane and isopropanol were added to produce 300 ml of 2.76 specific gravity liquid. This slurry was agitated and allowed to separate for 12 to 24 hr until migration of particles in the 2.76 sp. gr. liquid ceased. The top 200 to 250 ml of liquid containing the "2.76 floats" were decanted into a clean beaker. The separated fractions were recovered by filtering each suspension through a tared, 0.22 pore size, Mi 11ipore membrane filter. The re covered mineral fractions were dried for 24 hr at 90C and weighed. Since losses of materials were unavoidable during the decanting and handling, the calculations of the mass percent for each density fraction represented in the total saaq>le are based on the summed masses of recovered materials, rather than on the subsample mass determined at the start of the density separations. For samples from which fibers were hand-picked before the density separations, the mass of the hand-picked fibers was added to the mass of the TEE fraction for calculation of the mass percents of each density separation fraction represented in the total bulk sample. The actual losses occurring in the heavy liquid separation 6teps were determined to be 2.2 to 5.5%. Duplicate separations performed on two of the samples indicated the density fraction data were reproducible to within 5%. Polarized light Microscopy . A portion of the unseparated sample and of each separation fraction were dispersed in a standard immersion oil (n- c 1.515) on a glass slide for PLM analysis. Mineral phases present were identified, at least as to mineral group, b y morphological properties and by observation of optical properties; including extinction angles, birefringence, refractive.indices, color and pleochroism. Individual particle typet' removed from subsamples by hand picking were mounted in other standard refractive index liquids to allow pre cise determination of particle refractive indices. Comparison of the unknown mineral particle properties with those of known reference samples and pub lished handbook values allowed Identification of various mineral types. Individual phase concentrations were microscopically estimated in the density separated subsamples on the basis of relative particle sizes and fre quency. Miss concentration ranges for individual mineral phases were described as: Primary Major Minor Trace > 25% 5 to 25% 0.5 to 5% < 0.5% The concentrations of the individual mineral phases in the total bulk sample were obtained by multiplying the microscopically estimated concentration of the phase in the separation fraction by the mass fraction the subsample repre sented of the total sample. For many samples, individual mineral phases oc curred in small concentrations in all the density-6eparated fractions. Multiple-phase particles also contributed to the presence of high density B-6 mineral phases in the two lower density fractions. Thus, the concentrations of individual phases within the total sample were determined by summing the concentration determined for each density-separated fraction. X-Ray Diffraction Positive identification of mineral phases, particularly fibrous amphiboles, by polarized light microscopy, is difficult. Therefore, to clearly establish identities of major and fibrous mineral phases, x-ray diffraction (XRD) analyses were conducted on selected separation subsamples and individ ual phases hand-picked from separation subsamples. The major objective of the XRD analyses was identification of the fibrous phases. IITRI used the thin film technique, since this is most suitable for the very small quantities of materials available from hand-picking. The thin films were prepared by filtering the isopropanol suspended material through a silver membrane (25 mm diamter, 0.45 pm pore size). The XRD samples were ground to -325 mesh in a diamonite mortar prior to the filtration. As im plied above, this technique provides an XRD pattern for material quantities as low as 0.1 mg; its drawback is that quantitation is not practical due to preferred orientation of many minerals including amphibole fibers. . Diffraction patterns were obtained with a Rigaku brand, rotating copper anode diffractometer operated at 50 kilovolts and 100 milliamps. The CuK x-ray lines, with' an averaged wavelength of 1.54184 A, were thus generated. Host patterns were run at a scan rate of 2/min. After conversion of the 20 diffraction angles to d-spacings in angstroms, the sample diffraction pat terns were compared to standard diffraction patterns published by the Joint Committee on Powder Diffraction Standards for mineral identifications. ELECTRON MICROSCOPY The objectives of the electron microscopy analysis were to (a) determine if respirable asbestos fibers were present; (b) identify the fibers present within the limitations of the EPA Provisional Method;3 (c) determine the number concentration of respirable asbestos fibers in the bulk sample; and (d) estimate the mass concentration of "respirable" fibers in the bulk sam ple. The steps required to achieve these objectives are: Eeneficiation of vermiculite samples, Sample preparation, and Electron microscopy analysis. 3 Electron Microscope Measurement of Airborne Asbestos Concentrations: A Provisional Methodology Manual, EPA-600/2-77-178. Available from U.S. Environmental Protection Agency, Office of Research and Development Technical Information Staff, Cincinnati, Ohio 45268, Samudra, A., et al. Beneficiation Two beneficiation procedures were used for the samples; one used isopro panol and the other water as a working fluid. Each is described below. Isopropanol Beneficiation-- . The isopropanol beneficiation procedure was used for those samples whose asbestos fiber concentration, as determined during the PLM analysis, was le&6 than 1% by weight, and to samples 259-1, 264-1, and 291-1* at the specific re quest of MRI. This beneficiation required particle-free isopropanol, a sonic bath and tall form beakers. The weighed sample was placed in a tall form beaker of appropriate size; a volume approximately equal to 10 times the sample weight . of filtered isopropanol was added and the beaker and contents put in an ul trasonic bath. The ultrasonic bath was turned on, the mixture was stirred using a clean spatula, the large particles were allowed to settle, and aliquots were with drawn from the center of the liquid column. The aliquots were diluted with a sufficient quantity of prefiltered isopropanol to permit filtration (30 to 50 ml). Each aliquot was quantitatively transferred to a Xuclepore filter for electron microscopy analysis. Aqueous Beneficiation-- . One of the objectives of the study is to determine the fate of the as bestos which may be in the run-of-the-mine graded vermiculite. To determine the fate of the asbestos, IITRI exfoliated the graded vermiculite samples by exposing them in an 800eC oven, beneficiated the resulting exfoliated sam ples, and used electron microscopy to determine the asbestos content of the sink fractions. Thermal exfoliation is used to prepare expanded vermiculite products. This treatment can readily be simulated in the laboratory. A monolayer of unexfoliated vermiculite is sprinkled into a preheated shallow quartz con tainer which is inserted into an 800eC oven for 5 sec. After the 5 sec ex pire, the dish is removed, permitted to cool and the exfoliated vermiculite is 6tored for analysis. The process is repeated until all of the sample is exfoliated. The exfoliated sample still contains all of the components originally present; however, most of the vermiculite phase can now be removed by an aqueous float process. In the float process, the exfoliated vermiculite is placed in a container, 1 or 2 liters of filtered, distilled water is stirred for 30 sec and the "vermiculite floats" removed and drained on a 50 mesh, U.S. Standard sieve. The drain water is collected and returned to the original container, and the vermiculite floats are discarded. The water (in its container) is then placed in the sonic bath, soni cated, and aliquot samples are removed with a pipette. The procedure used to * Midwest Research Institute sample numbers. B-8 prepare filters is identical to that used in the isopropanol beneficiaticn, except that filtered, distilled water is substituted for isopropanol. Sample Preparation The basic procedures 1ITRI used to analyze the samples are documented in Electron Microscope Measurement of Airborne Asbestos Concentrations: A Pro visional Methodology Manual;8 however, four samples were analyzed using a pod- ified protocol. Both protocols use identical sample preparation procedures, which are discussed below followed by a description of the analytical proto* cols. ' The membrane on which the sample is deposited was air-dried overnight in a Class 100 clean workbench; a wedge-shaped portion of the filter was cut, carbon-coated, and 0.3 mm diameter circle was then removed with a punch from the wedge for transfer to a 200 mesh copper transmission electron microscopy grid. The Jaffe washer technique was used to transfer the sample to the grid. In the Jaffe washer technique, a stack of 40 clean, 4.4 cm diameter, paper filters is placed in a clean glass petrie dish. Spectroscopic grade chloroform is then added until the level is at the top of the stack of filter paper. Several small (but larger than the 0.3 mm diameter grid) pieces of 60 or 100 mesh stainless steel (SS) screen are placed on the stack of filter paper. An orientation mark is placed on the outside of the petri dish, and a "map" showing the location of each piece of SS mesh is drawn. A sample to be transferred is placed, carbon-coated side down, onto a transmission electron microscope grid and the pair is placed on the center of a piece of SS mesh. The sample identification is noted on the "map." The procedure is repeated until the washer is filled or all samples to be prepared are in place. The chloroform level is maintained at the top surface of the stack of filter paper for the 24 to 72 hr required to dissolve the Nuclepore membrane. The preparation is completed by allowing the residual chloroform to evaporate from the grid, then placing the grid in a labeled grid storage box. All of the procedures described are performed in a Class 100 clean workbench. Electron Microscopy Analysis EPA Provisional Method-- The prepared samples are examined using a JE0L 100C analytical electron microscope. The electron microscope (EM) is used in the transmission mode to screen the sample prior to the analysis and to perform the enumeration and sizing of each fiber located in the selected grid opening(s). The scanning transmission electron microscopy (STEM) mode is used to obtain nondispersive x-ray data at a tilt angle of 40. 3 See Reference 3, page B-7. B-9 Prior to analysis of the sample, it is inspected at low (500X) magnifi cation to assure that the majority (80%) or more) of the grid openings are clear and the carbon film intact. The particle loading is also determined at this time. Any sample found overloaded, damaged, or not adequately cleared is not used in the analytical process, but is reprepared from either the original suspension or the filter, as appropriate. The accepted preparations are immediately counted by randomly selecting a grid opening(s) and counting, sizing, and classifying each fiber found in the grid opening. The fibers are classified as chrysotile asbestos, amphibole asbestos, not asbestos, ambiguous, and no pattern based on morphology and the selected area diffraction (SAD) pattern. The SAD patterns are not recorded (except a few for documentation) or indexed. The comparison is visual and if the pattern is obviously compatible with one of the asbestos minerals, the fiber is so classified. A nondispersive x-ray pattern is often used as an aid in the classification, particularly when nonasbestos phases capable of providing false positive fiber identification are known (by PIM, SID, or from geological sources) to be present. Similarly, fibers whose SAD pattern characteristics differ significantly from asbestos pattern characteristics are classified as "not asbestos," and those providing no pattern* or indeterminate patterns are classified as "no pattern" or "ambiguous" fibers, respectively. These data are recorded, with the size data taken for each fiber, and is reported for each fiber observed. The fiber enumeration and sizing is continued until approximately 100 fibers have been completed or 10 grid openings have been examined. The data are taken from two different TEH grids whenever possible. Modified Provisional Method-- IITRI. analyzed four samples using a modified Provisional Method to count the fibers. The method was modified to facilitate obtaining statistically valid counts, which in turn permits estimation of confidence limits for the analysis. The modifications are based on discussions with Dr. E. Chatfield and one of our objectives is to provide data for comparison of our results with duplicate anayses at Ontario Research Foundation (ORF). As previously stated, the modifications affect the analysis of the sam ple during electron microscopy-- not during preparation. The specific changes are in the loading requirement and the fiber identification criteria. The difference in loading requirement is in the basis for selection of a grid. IITRI selects grids for analysis based on total loading of particles. We have found that this basis facilitates particle and fiber identification, although it can cause low fiber counts in samples containing asbestos at the part per million concentration level. ^ Fibers which are too close to another particle or to the edge of the grid frequently are "ambiguous." B-10 The ORF procedure bases grid selection of a fiber loading (with, of course, an upper limit fixed by total particle loading) of, ideally, 10 as bestos fibers per grid opening. Since the objective is to obtain statis tically valid counts, this procedure also calls for counting on four grids prepared from different areas of the filter. This enhances the statistical validity of the data. It is obvious that using the higher loading requires screening many more fibers when the ORF procedure is followed; thus, the I1TR1 method, which in volves classification.of each fiber encountered, must be modified. The ORF procedure allows screening on the basis of morphology and other information known about the samples. As an example, vermiculite scrolls and plates are common nonasbestos "fibers" in vermiculite samples. However, they have much lower contrast than do amphibole fibers and prisms and are readily distin guished by an experienced analyst scanning the vermiculite samples. Thus, the nonasbestos fibers are passed over, chrysotile is identified on The basis of its unique morphology and amphibole fibers are identified by morphology and the compatibility of the fibers' SAD pattern with known, amphibole patterns. The ORF procedure results in three classifications of fibers analyzed: chrysotile, amphibole, and unidentified fibers. Data Reduction-- The data from each procedure is reduced using similar procedures. The fiber concentration in fibers per gram is computed based on the original weight of vermiculite from which the saaq>le was prepared. It is computed using Equation 1: K f,i A f (1) W where C*.. = Concentration, fibers per gram . = Humber of fibers of type i counted n** = Number of grid openings counted A* = Area of one grid opening, cm2 A| = Area of filter from which grid was made, cm2 VT = Weight of &ample, g VI, = Total volume of bonification fluid c Aliquot volume used for filter preparation. IITRI and ORF used similar but different grids. Those used by IITR1 were extremely uniform in opening areas and not all openings were measured. ORF used a different grid which was not as uniform and measured each grid opening that was counted. * The total mass of fibers is estimated using a right circular prism model to compute a volume, then multiplying by a density and summing over each fiber type. This calculation is shown in Equation 2. The mass concentration, f /g, is computed by substituting the total fiber mass estimate into Equation 1, for Nj. B-ll  t i M. d2 z x e x io3 (2) i=l where: M f = Estimated total fiber mass, d2 = Fiber diameter (projected width), pm = Fiber length, pm e = Mineral density, g/cm3 : chrysotile = 2.6 g/cm3 amphibole = 3.0 g/cm3 Fiber To do fiber gram. detection limits for any sample can also be this, the value of 1 is substituted for N,, is detected. All detection limits reported computed using Equation 1. equivalent to assuming 1 herein are in fibers per When high fiber loadings are encountered, or when the ORF alternate pro cedure is used, the average fiber count per grid opening and the standard de viation (S) are determined. These values are computed using standard methods based on a normal distribution and the fiber count data are then given with 95% confidence limits. Airborne Fiber Analysis-- The enumeration and measurement of airborne fibers is accomplished using two well-defined and established techniques. The first is the NIOSH specified phase contrast fiber enumeration b y optical microscopy. The procedures are described in detail in DHEW (NIOSH) Publication No. 79-127. The second ana lytical method is the U.S. EPA's Provisional Hethodology, described in EPA600/2-77-178, revised June 1978. The NIOSH procedure is summarized below, since it was used for analyses on this program; the EPA electron microscope procedures are identical to the procedure described as Sample Preparation and Electron Microscopy Analysis (EPA Provisional Method) in the bulk sample pro cedures . Sample Preparation Samples to be submitted for phase contrast enumeration are collected on cellulose acetate (Millipore or equivalent) filters using an open face filter holder. The sample is prepared for enumeration using one of two techniques-- dissolving the filter in a solution of dimethyl phthalate and diethyl oxalate containing clean, dissolved membrane for viscosity control4 or by collapsing the membrane pore structure using a solution of hexane/1,2-dichloroethane/ p-dioxane and rendering the collapsed membrane transparent by exposure to'ace tone vapors.3 Both techniques end by covering the sample with a cover slip, with the latter technique requiring a 1.505 refractive index oil. IITRI uses the second procedure because it provides a permanent mount with no restric tions on the count-time frame. This preparation procedure is routinely used by IITRI for proficiency analytical testing (PAT) fiber enumeration.* * Contained in P&CAM 239, recommended by NIOSH. 5 Millipore Procedure, TS018. B-12 Fiber Enumeration The fibers on the filter are enumerated using the following protocol. A Porton Graticule is used to define counting areas. A minimum of 20 areas and maximum of 100 areas are counted. After 20 areas have been counted, the enumeration of fibers is topped when 100 fibers have been enumerated. Fibers are enumerated using the following rules: * Fiber is entirely within counting area - Count-- 1 fiber if length > 5 |Jm * Fiber has one end in counting area Count-- 1/2 fiber if length > 5 |m * Fiber crosses two sides of counting area Count-- no fiber * Fiber does not enter counting area Count-- no fiber Note: All fibers-- defined as particles having parallel sides, length to di ameter ratio 3 and length > 5 pm-- are enumerated by the NIOSH procedure. Thus, if the probability is high that nonasbestos fibers are present, the NIOSH procedure can overstate the actual asbestos fiber concentration. The reason for the possible over statement is the fact that phase contrast il lumination does not allow the analyst to identify the individual fibers. The fiber count data are converted to airborne fiber concentration using Equation 3: (3) where: F = Airborne fiber concentration, fibers/cc = Number of fibers enumerated a^ = Total counting area, cm2 Af = Area of filter used for sample collection, cur2 V* - Volume of air sampled, cc B-13 APPENDIX C AIR SAMPLES FLOW RATES AND SAMPLE VOLUMES The Dupont Model 4000 personal air samplers were calibrated before and after each day of sampling. The calibration was done with a 500-ml soap bubble meter and a stop watch. The sampler calibration values are given in Table C-l. The data for the volume calculation for the various personal air samples are given in Table C-2. The flow rates for the stationary samplers were measured at the beginning of sampling, periodically during sampling and at the end of sampling. The data for the stationary samplers are given in Table C-3. TABLE O l . PERSONAL SAMPLER CALIBRATIONS (EACH VALUE REPRESENTS A MINIMUM OF THREE DETERMINATIONS) Enoree, Worth Carolina Grace, Libby Montana Grace Mill Grace Mine Patterson Sampler Pre- Post Pre Post Pre Post Pre Post ID sampling0 sampling sampling sampling Avga sampling sampling Avga sampling sampling Avg 186172 186173 186176 186175 1-7328 2-7334 6-7329 7-7317 2.13 2.12 2.09 2.12 2.07 2.16 2.12 2.19 1.91 1.92 1.96 1.93 1.89 2.13 1.99 1.99 2.02 1.96 1.99 1.98 2.02 2.00 2.02 2.01 2.07 1.94 2.00 2.01 2.04 2.04 2.07 2.00 2.04 1.95 2.00 2.00 2.03 2.02 2.04 2.01 1.94 1.98 1.95 1.99 1.92 2.00 1.97 1.95 1.98 1.94 1.95 1.93 .1.93 1.94 1.93 1.93 1.94 1.93 a Values used to calculate sample volume < TABLE C-2. DATA FOR VOLUME CALCULATIONS FOR PERSONAL SAMPLES Sample no. Sampler ID Flow (/min) Time sampled (min) Volume <) 101 121 125 126 128 129 130 131 135 138 139 141 146 146 300 301 a 304(S)a 305 306 308 310 314 315 320 321 322 324 330 322(S)a 336 337 339 340 a 34l(S)a 347 349 1-7328 186173 7-7317 166175 2-7334 186172 7-7317 186174 166174 186172 2-7334 6-7329 6-7329 166175 2-7334 6-7329 186174 186172 6-7329 7-7317 2-7334 186173 1-7328 2-7334 7-7317 186173 1-7328 1-7328 2-7334 6-7329 166175 186172 6-7329 2-7334 186175 1-7328 2.07 2.12 2.19 2.12 2.16 2.13 2.19 2.09 2.09 2.13 2.16 2.12 2.12 2.12 2.02 1.98 2.00 2.04 1.98 2.01 1.95 1.95 1.97 1.95 2.01 1.95 1.97 2.03 2.02 2.04 2.00 2.04 2.04 2.02 2.00 2.03 186 385 429 909 264 578 278 589 266 575 270 575 130 285 145 303 283 591 138 294 125 270 130 276 261 555 140 297 175 354 121 240 438 876 141 288 362 717 203 408 132 257 146 285 127 250 260 546 238 478 241 470 31 61 141 285 202 408 205 ' 418 246 492 302 616 181 369 79 160 157 314 251 507 (continued) C-3 Sample no. 504 508 511 516 517 519 520 521 542 Sampler ID 2-7334 7-7317 6-7329 7-7317 2-7334 7-7371 6-7329 6-7329 2-7334 TABLE C-2 continued Flow (/min) 1.93 1.93 1.94 1.93 1.93 1.93 1.94 1.94 1.93 Time sampled (min) 185 132 194 188 192 177 130 184 129 a 304(S), 332(S), and 34l(S) were stationary samples. Volume () 357 255 376 363 371 342 252 357 249 C-4 no. 102 103 104 .107 108 111 112 113 115 116 119 120 122 123 124 132 134 136 145 147 149 302 307 309 313 316 318 323 328 329 331 334 335 338 342 343 344 345 350 351 352 353 354 TABLE C-3. VOLUME CALCULATIONS FOR STATIONARY SAMPLES Average flow (/min) Tine sampled (min) Volume () 0.97 1.04 2.19 1.90 1.16 2.08 2.05 1.39 0.79 0.90 1.00 2.20 2.36 1.47 0.88 0.96 1.94 0.97 1.09 1.41 1.14 0.74 1.31 1.04 1.81 1.65 1.18 1.54 2.37 2.13 1.52 1.08 0.85 1.94 1.58 2.36 1.05 1.77 0.87 1.26 1.45 1.96 1.59 270 261 182 189 178 390 109 207 146 169 177 368 301 616 143 199 321 253 174 156 128 128 299 658 179 422 166 244 343 302 362 348 404 784 114 111 386 421 387 545 369 420 321 239 222 291 321 334 329 595 314 582 320 378 100 154 121 287 387 823 128 195 124 134 94 80 136 264 327 516 328 774 84 88 122 216 331 288 332 420 352 511 322 632 324 515 (continued) C-5 Sample no. 502 503 506 513 515 518 523 525 527 528 531 540 TABLE 03 continued Average flow (/min) Time sampled (min) 1.67 351 1.46 144 2.23 123 1.30 145 1.70 175 1.76 342 1.72 348 2.19 342 2.13 345 1.02 144 2.08 144 2.04 343 Volume () 585 210 274 168 299 601 599 748 735 147 300 701 \ C-6 APPENDIX D INCREMENT BULK'SAMPLES COLLECTED AND COMPOSITED Host of the samples submitted for anaysis were composites of increment samples. To prepare the composite samples each increment sample was riffled to obtain a representative fraction of the increment. Approximately equal weight fractions of each increment fraction were combined to make a composite sample. The composite sample was then mixed and riffled to produce four equal samples. One of the fourths was set aside and retained as a control. One of the fourths was again riffled to produce two-eights of the original composite. The two-eighths were combined with the two-fourths so that the composite was divided into 1/4, 3/6, and 3/6 of the original. The 111/4" was retained at HRI: one "3/8" was sent to I1TRI; the other "3/8" was sent to Ontario (ORF) for analysis. Each of the increment samples was assigned a sample ID number. Tables D-l, D-2, and D-3 lists the increment samples from the three collection lo cations that were processed into composites. Table D-4 lists the samples that were not processed. D-l TABLE D-l. INCREMENT AMD COMPOSITE SAMPLES FROM LIBBY, HI, GRACE Sample description Date collected Sample ID Sample weight (8) Weight for compoj is) Grade 1 Grade 1 Grade 1 Grade 1 Grade 1 Grade 1 Grade 1 Grade 1 Grade 1 Grade 1 Grade 1 Grade 1 Grade 2 Grade 2 Grade 2 Grade 2 Grade 2 Grade 2 Grade 2 Grade 2 Grade 2 Grade 2 Grade 2 Grade 2 Grade 3 Grade 3 Grade 3 Grade 3 Grade 3 Grade 3 Grade 3 Grade 3 Grade 3 Grade 3 Grade 3 Grade 3 10/7/80 151 1,030 126 10/8/80 152 1,002 125 10/9/80 153 1,002 126 10/10/80 154 1,000 125 10/13/80 155 1,001 125 10/14/80 156 999 125 10/15/80 157 1,002 125 10/16/80 158 1,009 125 10/17/80 159 1,000 125 10/21/80 160 798 125 10/23/80 161 991 125 Composite 270 10/7/80 162 1,001 125 10/8/80 163 1,006 125 10/9/80 164 1,004 125 10/10/60 165 1,004 125 10/13/80 166 1,004 125 10/14/80 167 1,002 125 10/15/80 168 956 125 10/16/60 169 1,002 125 10/17/80 170 1,002 125 10/21/60 171 873 125 10/23/80 172 992 125 Composite 276 10/7/80 173 999 133 10/6/80 174 1,002 113 10/9/80 175 1,002 121 10/10/60 176 1,002 122 10/13/80 177 1,003 127 10/14/80 178 997 124 10/15/80 179 1,003 129 10/16/80 180 1,000 124 10/17/80 181 997 124 10/21/80 182 1,000 124 10/23/60 183 964 125 Composite 259 (continued) D-2 Sample # description ' Grade 4 ' Grade 4 Grade 4 Grade 4 Grade 4 Grade 4 Grade 4 Grade 4 Grade 4 Grade 4 Grade 4 Grade 4 - Grade 5 Grade 5 Grade 5 Grade 5 Grade 5 Grade 5 Grade 5 Grade 5 Grade 5 Grade 5 Grade 5 Grade 5 Head Head Head Head Head Head Head Head Head Head feeda feeda feedfi feeda feeda feeda feed* feeda feeda feeda TABLE D-l continued Date collected Sample ID Sample weight (8) 10/7/80 184 10/8/80 165 10/9/80 186 10/10/80 187 10/13/80 188 10/14/80 189 10/15/80 190 10/16/80 191 10/17/80 192 10/21/80 193 10/23/80 194 Composite 282 10/7/80 195 10/6/60 196 10/9/60 197 10/10/80 198 10/13/80 199 10/14/80 200 10/15/80 201 10/16/60 202 10/71/80 203 10/21/80 204 10/23/80 205 Composite 264 10/8/80 223 10/9/80 224 10/10/80 225 10/13/60 226 10/14/80 227 10/15/80 228 10/16/80 229 10/17/80 230 10/23/80 231 Composite 291 1,000 1,004 997 1,002 1,001 1,000 1,000 1,022 1,000 999 1,002 1,000 1,001 1,005 999 998 1,000 1,000 1,000 997 995 1,000 958 965 955 951 953 947 949 950 1,000 Weight for composite (8) 125 125 125 125 125 125 125 125 125 125 125 125 125 124 125 129 129 124 126 124 126 125 125 125 125 125 125 125 125 125 125 (continued) D-3 Sample description TABLE D-l continued Date collected Sample ID Sample weight (8) Extractor Extractor Extractor Extractor Extractor Extractor Extractor Extractor Extractor 10/8/80 10/9/80 10/10/80 10/13/80 10/14/80 10/15/80 10/16/80 10/17/80 Compos ite. Dust, Dust, Dust, Dust, Dust, Dust, Dust, Dust, screening plant screening plant screening plant screening plant screening plant screening plant screening plant screening plant 10/8/80 10/9/80 10/10/80 10/13/80 10/14/80 10/15/80 10/17/80 Composite Dust, Dust, Dust, Dust, Dust, Dust, Dust, Dust, Dust, Bill mill mill mill Bill mill Bill Bill Bill baghouse baghouse baghouse baghouse baghouse baghouse baghouse baghouse baghouse 10/8/80 10/9/80 10/10/80 10/13/80 10/14/80 10/15/80 10/16/80 10/17/80 Composite 250 251 252 253 254 255 256 257 294 206 207 208 209 210 211 212 288 215 216 217 214 218 219 220 221 297 870 860 852 860 860 893 887 890 1,008 1,000 998 1,230 996 965 980 1,006 998 1,003 1,009 977 989 1,002 1,004 Weight for compoi (g) 125. 125 125 125 125 125 125 125 125 125 125 125 125 125 125 125 125 125 125 125 125 125 125 a All "head feed" camples contained rocks too large to pass through the channels in the riffle box. These were separated out, weighed sep arately, and retained at URI. D-4 Sample description Grade 3 Grade 3 Grade 3 Grade 3 Grade 3 Grade 3 Grade 3 Grade 3 Grade 4 Grade 4 Grade 4 Grade 4 ' Grade 4 Grade 4 Grade 4 Grade 4 Grade 5 Grade 5 Grade 5 Grade 5 Grade 5 Grade 5 Grade 5 Grade 5 Hill feed + 100 mesh Hill feed + 100 mesh Hill feed + 100 mesh Hill feed + 100 mesh Hill feed + 100 mesh Hill feed + 100 mesh Hill feed + 100 mesh Hill feed + 100 mesh Grade 3 exfoliated Analysis ID Grade 4 exfoliated Analysis ID Date collected 10/27/80 10/28/80 10/29/80 10/30/60 10/31/80 11/1/80 11/2/80 Composite 10/27/80 10/28/60 10/29/60 10/30/60 10/31/60 11/1/80 11/2/80 Composite 10/27/80 10/26/60 10/29/80 10/30/60 10/31/80 11/1/80 11/2/80 Composite 10/27/80 10/28/80 10/29/80 10/30/80 10/31/80 11/1/80 11/2/80 Composite 11/5/80 11/5/80 Sample ID 389 390 391 392 393 394 395 430 396 397 398 399 400 401 402 433 403 404 405 406 407 408 409 427 375 378 379 381 384 386 388 436 424 439 422 442 Sample weight (6) 1,538 2,584 2,445 1,900 2,584 2,682 2,788 1,490 2,450 2,380 2,390 2,715 2,610 3,005 1,729 2,459 2,590 2,185 2,700 2,490 2,770 451 881 776 709 956 836 792 203 260 Weight for composite 120 120 125 125 125 125 126 125 125 125 125 125 125 125 123 126 119 125 125 125 129 225 225 225 225 225 225 225 D-5 TABLE D-3. INCREMENT ASP COMPOSITE SAMPLES FROM EKOREE, SC, PATTERSON Sample description Date* collected Sample ID Sample weight (8) Weight for composite (8) Ungraded 11/6/80 567 1,770 250 Ungraded 11/6/80 568 1,700 250 Ungraded 11/6/80 569 1,186 250 Ungraded 11/6/80 570 1,780 250 Ungraded Composite 573 a Increments taken at 2 hr intervals. D-6 TABLE D-4. BULK SAMPLES THAT WERE COLLECTED BUT HOT SUBMITTED FOR ANALYSIS Sample description Humber of increments Grace, Libby, under 90 mesh 9 Grace, Libby, coarse tails 9 Grace, SC, mill feed, under 100 mesh 7 Grace, SC, dryer composite 7 Grace, SC, vet scrubber discharge 1 Grace, SC, composite total tails 1 Grace, SC, Lanford mine composite 1 Grace, SC, Foster mine composite 1 Grace, SC, Ho. 4 concrete aggregate, stabilized 1 Grace, SC, Ho. 3 masonry insulation 1 Patterson, SC, raw ore prescreen 1 Patterson, SC, raw ore postscreen 3 Patterson, SC, raw ore multiple grab from main ore pile 1 Patterson, SC, main vaste pile 4 Patterson, SC, preexfoliated vaste 4 Patterson, SC, bagged product 4 APPENDIX E p h o t o m i c r o g r a p h s a n d TEM MICROGRAPHS OF SELECTFn SAMPLES 1. Photomicrographs (IITRI) (p. E-2 to E-17) of selected samples. All photomicrographs were taken with slightly uncrossed polars unless otherwise Indicated. 2. TEM micrographs (ORF) (p. E-18 to E-25) of selected sauries to Illustrate the types of particles observed in samples of vermicu- llte from the various locations represented In this study. It should be noted that these micrographs were taken using specimen grids pre pared for illustration purposes only and the particle loadings are considerably heavier than those required for TEM evaluation. * E-l 259-I-TBS; 52X. Unevenly fractured fragments of dark green dlopside and auglte. 259-1-2.76 float; 84X. Large flakey particles are vermiculite with white stress lines. Arrow points to a flake of talc intergrown with tremollteactinolite fibers. E-2 259-I-TBS-Fibers (ground){ 208X. Large bundle In the center Is unquestionably fibrous tremollte-actinollte. Note the (white) Interference dolors of fibers growing at angles to the main fiber bundle which has been rotated to an extinction position (and Is therefore.gray). 259-I-TBS-Fibers (ground); 208x. This parallel--lamellated prism morphology of tremollte-actinollte was found within hand-picked fiber bundles. E-3 259-1-2.76 6 ink; 208X. Calcite (bright white) growing in & pseudomorphlcally fibrous crystal habit within tremolite-actinolite fiber bundles. Some cal cite crystals (arrows) could be fractured to yield "fibers" . 259-1-2.76 float; 84X. Arrow points to a flake of lamellated quartz. The flakey, lamellated morphology was caused by the quartz forming Inbetween vermiculite plates. E-4 264-I-TBS; 208X. Arrows point to tremollte-actlnollte In two different crys tal habits-*-the parallel, lamellated prisms and truly fibrous material. The mottled coloring of the fiber bundle 16 due to the various angles at which the Individual fibers are intergrowing with each other. 264-1-TBS; 208X. The circle outlines one tremolite-actinollte particle which contains 2 different crystal habits-- the parallel, lamellated prisms on one end and matted, intergrown fibers (white portion) on the other end. E-5 264-1; 84X. The large elongated particle In the center Is fibrous tremollteactlnollte vith Inclusions(arrows) of prismatic tremollte-actlnollte. The large flakey white particle (arrow) Is a sheet of quartz adhering to vermicullte and was- apparently growing inbetween vermlcullte plates. 291-1; 84X. The friability of the tremollte-actlnollte fiber bundles is demonstrated here. Simple dispersion of the sample resulted in abrasion of numerous smaller fiber bundles from the large bundles present. E-6 430-I-TBS (minimaT~grindlng); 208X. The tremolite-actinolite, hornblende end anthophyllite emphlbole fragments are mostly Irregular to chunky prisms. Some slender pri&ms which could yield particles classifiable as fibers are present. 430-1-2.76 sink (after grinding); 208X. Most of the prismatic to near fibrous amphlbole separated into this density fraction because it was so Intimately intergrown with vermlcullte and talc. E-7 430-1-2.76 float (ground); 208x. Although optical and morphological proper ties of the mottled phase depicted vere consistent vith serpentine minerals XRD data ruled out serpentine as its identity. Mixed layer veraiculitehydrobiotlte was identified in the XRD work. 427-- I-TBS; 84X. Several large lamellated, parallel prisms of tremoliteactinolite and anthophyllite are visible. E-8 427-I-TBS (ground); 208X. Crushing of the large emphlbole prisms produced prisms vith splinter fragments that could morphologically be defined as fibers. 427-I-TBS-Fibers (ground); 208X. Crushing of coarse "fibers" produced mostly prismatic material. Arrows point to the numerous vemiculite and talc plates intergrown within "fiber" bundles. E-9 427-1-2.76 sink; 208X. The white portion of the circled particle is talc. The remainder is (lamellated) prismatic anthophylllte. Numerous flakes of talc intergrovn with anthophylllte were found in this fraction. 436-1; 82X. The amphiboles are mostly chunky and prismatic in this sample. Arrow points to tremolite composed of rather thick, lamellated prisms. E-10 436-1-Fibers (ground); 208X. Truly fibroxis amphibole as veil as very chunky, prismatic particles were produced when particles macroscopically classifiable as fiber bundles vere ground. 436-I-TBS-green, glassy (ground); 208X. This phase was a mixture of horn blende and tremolite-actinolite. The amphiboles' morphologies vere predominantly chunky prisms. E-ll 436-I-TES-millcy green (ground); 208X. Mineral grains composed of this tremolite-hornblend mixture were, irregular in 6hape and rough. Appearances of crushed fragments Indicated that the grains were composed of agglomerated smaller crystals which themselves were Irregularly grown. 436-1-TBS-colorless, glassy (ground); 208X. The predominant material is fluorapatite-- high contrast, conchoidally fractured particles. E-12 573-I-TBS; 82X. Non-vermiculite phases vere mostly chunky, prismatic amphiboles, iron oxides (black) and fluorapatite (rounded, gray particles). 573-I-TBS (ground); 208X. Grinding did result in fracture of some amphlbole grains into elongated fragments morphologically classifiable as fibers. Most of the amphlbole fragments retained chunky, prismatic morphologies. E-13 267-I-TBS; 82X. Fibrous and prismatic'amphiboles (arrows) are present with diopside. 276-I-TBS-milky green (ground); 208X. Like the milky green mineral grains of 436*1, the milky green grains of this sample were composed of multiple, poorly formed amphibole crystals. Unlike the 436-1 sample, the individual crystals in this sample exhibited mostly fibrous morphologies. E-14 * 276-I-TBS-dark. green glassy (ground); 208X. The pyroxene, dlopslde, exhibits some concholdal as veil as prismatic fracture patterns. 294-I-TBS; 82X. Small amphlbole fiber bundles and large amphlbole prisms are present. Note the layered crystal growth of the fractured tremollteactlnollte fragment (arrow). E-15 286-1; 52X. The mottled coloring of this large tremolite-actlnollte fiber bundle Is due to the non-parallel, radiating growlth pattern of the Indivi dual fibers In the bundle. 268-1 (fines); 208X. Much abrasion and disintegration of large fiber bundles has obviously occurred In the processing of the vermlcullte, as Indicated by the numerous very fine fibers present. E-16 r 297-1; 208X. The fine, single fibers here were abraded from larger tremoliteactinollte fiber bundles* * E-17 Wi-y j J.4 ' * ^ \! . ;v: s' ^ * ? & : ; V > ^ . - 3 ' ^ , 1O . ^ ____ i d i -11.0 wn Transmission electron micrograph shoving typical particulate matter found In water suspension after laboratory exfoliation of Sample 264-0, Grace, Libby, Montana, Grade 5 (composite). E-18 Transmission electron micrograph showing typical particulate matter found In water suspension after laboratory exfoliation of Sample 264-0, Grace, Libby, Montana, Grade 5 (composite). E-19 Transmission electron micrograph shoving typical particulate matter found In water suspension after laboratory exfoliation of Sample 264-0, Grace, Libby, Montana, Grade 5 (composite). E-20 f Transmission electron micrograph shoving typical particulate matter found in water suspension after laboratory exfoliation of Sample 427-0, Grace, Enoree, South Carolina, Grade 5 (composite). E-21 Transmission electron micrograph shoving typical particulate matter found in water suspension after laboratory exfoliation of Sample 427-0, Grace, Enoree, South Carolina, Grade 5 (composite). E-22 r Transmission electron micrograph showing typical particulate matter found In water suspension after laboratory exfoliation of Sample 427-0, Grace, Enoree, South Carolina, Grade 5 (composite). E-23 E-24 ' Transmission electron micrograph shoving typical particulate matter found in vater suspension after laboratory exfoliation of Sample 573-0, Patterson, Enoree, South Carolina, Ungraded, Dried Ore (composite). E-25