MIDWEST RESEARCH INSTITUTE
REPORT
COLLECTION, ANALYSIS AND CHARACTERIZATION OF VERMICULITE SAMPLES
FOR FIBER CONTENT AND ASBESTOS CONTAMINATION
TASK 32
FINAL REPORT
September 27, 1932
EPA Prime Contract No. 68-01-5915
MRI Project No. 4901-A32
Prepared for
U.S. Environmental Protection Agency
Office of Pesticides and Toxic Substances
Field Studies Branch
401 M Street, S.W.
Washington, D C. 20A60
Attn: Dr. Frederick Kutz, Project Officer
Mr. Thomas Dixon, Task Manager
MIDWEST RESEARCH INSTITUTE -'.?rj VOI.KE n lOULCVAHD. KANSAS CiTY. MISSOURI 641 "0 • BIG 753-7600
-------
E
0
0
0
COLLECTION, ANALYSIS AND CHARACTERIZATION OF VERMICULITE SAMPLES
FOR FIBER CONTENT AND ASBESTOS CONTAMINATION
by
Gaylord R. Atkinson
Donna Rose
Ken Thomas
David Jones
E. J. Chatfield
John E. Going
TASK 32
FINAL REPORT
September 27, 1982
EPA Prime Contract No. 68-01-5915
MRI Project No. 4901-A32
Prepared for
U.S. Environmental Protection Agency
Office 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
MIDWEST RESEARCH INSTITUTE 425 VOLKER BOULEVARD. KANSAS CITY. MISSOURI 64110 • 816753-7600
-------
DISCLAIMER
This docuoent has been reviewed and approved for publication
by the Office of Toxic Substances, Office of Pesticides and Toxic
Substances, U.S. Environmental Protection Agency, according to the
Agency's peer review aystea. The use of trade names or commercial
products does not constitute Agency endorsement or recommendation
for use.
-------
DISCLAIMER
4
This document has been reviewed and approved for publication
by the Office of Toxic Substances, Office of Pesticides and Toxic
Substances, U.S. Environoental Protection Agency, according to the
Agency's peer review systeal. The use of trade names or commercial
products does not constitute Agency endorsement or recommendation
for uae.
-------
d
ti
o
PREFACE
This final report presents the results obtained on MRI 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 No. 68-01-5915 with Midwest
Research Institute. Sample collection was conducted by MRI, 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
Manager, and IIT Research Institute, Mr. David Jones, Project Manager. This
report was prepared by Mr. Gaylord R. Atkinson, MRI Task Leader, with assis-
tance from Mr. Thomas, Dr. Chatfield, Mr. Jones, and Ms. Donna Rose (MRI).
MIDWEST.RESEARCH INSTITUTE
E. Going
Program Manager
Approved:
• Jl.
James L. Spigarelli, Director
Analytical Chemistry Department
ii
-------
E
6
0
CONTESTS
ii
Preface ......................... ....... iv
Figures ........................ '.'.'.'.I.... v
Abbreviations, Definition!, and Specifications ............. vl
1. Introduction ..................... '*'!'.. 3
2. Suanary ............................ 9
3. Experimental Protocol ..................... 9
Saople collection ....................
-------
FIGURES
Number p*o«
* O *a _
1 Wind rose pattern showing the direction and intensity of the
wind during the air sampling period at the Grace, Libby,
Montana, facility 16
2 Map of the Grace, Libby, Montana, facility showing the
stationary air sampling locations 17
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
4 Map of the Grace, Enoree, South Carolina, mill area showing
the stationary air sampling locations 23
5 Wind rose pattern showing the direction and intensity of the
wind during the air sampling period at the Grace, Enoree,
South Carolina, mine 24
6 Map of the Grace, Enoree, South Carolina, mine area showing
the stationary air sampling locations 25
7 Wind rose pattern showing the direction and intensity of the
wind during the air sampling period at the Patterson, Enoree,
South Carolina, facility 28
8 Map of the Patterson, Enoree, South Carolina, facility showing
the stationary air sampling locations 29
E
d
0
O
iv
-------
E
-------
I
0
TABLES (continued) c
Number
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Composition of Sanple 259-1 ,
Summary of Electron Microscopy Results
Grace, Grade 3 ,
Composition of Sample 282-1 ,
Summary of Electron Microscopy Results
Grace, Grade 4 ,
Composition of Sanple 264-1 ,
Summary of Electron Microscopy Results
Grace, Grade 5 ,
Composition of Sample 267-1 ,
Composition of Sanple 291-1 ,
Summary of Electron Microscopy Results
Grace, Head Feed ,
Composition of Sanple 294-1 ,
Summary of Electron Microscopy Results
Grace, Extractor ,
Composition of Sample 297-1
Sumaary of Electron Microscopy Results
Grace, Mill Dust
Composition of Sample 288-1
Summary of Electron Microscor/y Results
Grace, Screening Dust
Composition of Sample 430-7
Summary of Electron Microscopy Results
South Carolina, Grace, Grade 3. . . .
Composition of Sample 433-1
Summary of Electron Microscopy Results
for Sample Libby,
for Sample Libby,
for Sanple Libby,
for Sample Libby,
for Sample Libby,
for Sample Libby,
for Sample Libby,
for Sample Enoree,
for Sanrole Enoree.
Page
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
57
58
South Carolina, Grace, Grade 4 59
vi
-------
t)
TABLES (continued) °
Page
36 Composition of Saaple 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 Saaple Enoree.
South Carolina, Head Feed + 100 Mesh 63
40 Composition of Saaple 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 Samples
Collected at Three Vermiculit.- Sites 68
vii
-------
E
d
0
10
ABBREVIATIONS, DEFINITIONS, AND SPECIFICATIONS
The following are special terms or specifications that are used in this
report.
1. Venaiculite - 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 grades. 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.
Approximate
maximum Approximate
dimension number of Approximate weight/
("•»-' particles/g average particle
1 5-10 23 42 mg
\ 3'5 130 7.4 mg
I n J-3 1,700 0.58 mg
4 0.5-1 11,000 91 ug
5 0.2-0.5 130,000 7.6 pg
4. Asbestos - A general term for a number of naturally occurring fibrous
mineral silicates. Asbestos falls into two major classes, the serpentines and
amphiboles. 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 "anphiboles" 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 coaprised of a continuous varia-
tion with an arbitrary division between the two. The composition of tremolite
ranges from Ca2MgsSi022(OH)2 to Ca2Mg4FeSi8022(OH)2 and that of actinolite
ranges from Ca2Mg4FeSi8022(OH)2 to Ca2MgFe4Si8022(OH)2. Any composition within
this series is often reported simply as "tremolite/actinolite."
viii
-------
EM - Electron microscpy
TEM - Transmission electron microscopy
SAED - Selected area electron diffraction
JCPDS - Joint Committee on Powder Diffraction Standards
MRI - Midwest Research Institute, the prise contractor.
IITRI • IIT Research Institute
10 W. 35th Street
Chicago, Illinois 60616
ORF - Ontario Research Foundation
Sheridan Park Research Community
Mississauga, Ontario, Canada L5K 1B3
e
«
0-
ix
-------
. SECTION 1
INTRODUCTION
In December 1978, the veraiculir.e industry subnitted information to the
EPA regarding health problems experienced by employees who were processing
asbestos-contaminated veraiculite. The original subaission indicated that
bloody pleural effusions had been detected in 4 of 350 eaployees; symptom-
atology and clinical findings in the eaployees were siailar 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.
Veraiculite, mined in the United States since 1929, is a hydrated
magnesium-iron-aluminum silicate and is often contaminated with asbestifora
minerals. After mining, veraiculite is processed to remove impurities; how-
ever, some impurities, including asbestos, may remain in processed veraicu-
Although vermiculite may contain fibrous materials, the health effects
from vermiculite itself are unknown. A priority review of asbestos-
contaminated vermiculite, completed by the Office of Testing and Evaluation
in June 1980, suggested that the asbestos in veraiculite may be responsible
for the reported adverse health effects, and it concluded that cerUin infer-
maf t /\« AanriMAA.J.«*a*.AV^fJ11__i«_^^ , * . . w^^»*» **,u *u*. v&
in in-depth risk assessment on vermicu-
The available information on the composition of commercial vermiculite
indicated that asbestos contamination of venniculite does occur but that the
degree and kind of contamination might be difficult to assess and might varv
with the source of the vermiculite. Therefore, the objective of thif ulk*
was to sample and analyze vermiculite to deteraine the contaminants, particu-
anK °f "b?8ti5°™ »in«als present. The study nS 2'pS3d
to be used in the assessment of the risk to the population ex-
• vermicullt" at cacb of the
The original task objective was divided into two phases. The first chase
was to conduct an in-depth analysis of fibers present in, and associated wUh
vermiculite ore concentrates and beneficiated veraiculite from the major
££ >C«f %mineY?Kt5eiJnit^d States and be"fi«=iated venniculite from the
?Sp L^nH J7' Bothbuik and air saraPlefl wre 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-
Thelask J.^if'ijVf10^"",:1*11111 EPA' thC 8COpe °f thC ta8k «« "ducel.
The task was limited to the collection of air and bulk samples from three U S
I
4
0
0
-------
E
mines (W. R. Grace in Libby, Montana; W. R. Grace in Enoree, South Carolina: 0
and Patterson in Enoree, South Carolina). The air samples were analyzed only »
by phase contrast microscopy, and the originally planned electron microscopic
analysis was omitted.
With the reduction of effort, a set of the bulk samples that was con-
sidered to be representative of each nine was selected as "priority" samples
for i mediate 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 seven samples, was analyzed by various tech-
niques including electron microscopy for fiber content, with emphasis on as-
Destifonn minerals. The analysis was done by two independent laboratories.
It was considered possible that fibers could be bound between the veraiculite
plates and that fiberb 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 conoercial exfoliation in that
under the conditions of commercial exfoliation ouch of the fines and heavies
a"Jf!?OVe 5r0ffl thc vcrBic«lite- The laboratory exfoliation is done under
conditions that produce no sample fractionation. Thus, much of the asbestos
would be removed from the vermiculite during coamercial exfoliation, but none
would be removed during laboratory exfoliation.
»». jTh? *nalv8es of these samples were in various stages of completion when
the decision was made to reduce the scope of the task. Analytical results
rll^l f \{ e!i?eriin"tal ™*°4*> sampling, sample handling, analyt-
ical results, and appendix. Three additional volumes of appendices contain
the detailed analytical results.
-------
e
d
0
n
SECTION 2
SUMMARY
In December 1978, the vermiculite Industry suboitted iofonutlon to EPA
regarding the health problems experienced by employee! who were 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 be filled before an in-depth risk assessment of veraiculite could be ini-
tiated.
The objective of this task was to develop the protocol and to conduct
sampling and analysis to determine the composition of veroiculite with empha-
sis on the content of asbestiform minerals.
The original scope of the study included two phases. The first phase
was 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.
* < /nn 8amPles were analyzed by optical microscopy and x-ray dif f rac-
i!crL™)rS^t5e d?u8i
-------
TABLE 1. SUMMARY OP OPTICAL MICROSCOPY/XRD ANALYSIS RESULTS
e
tt
o
Sample3
Libby Grace
Grade 1, 270-1
Grade 2, 276-1
Grade 3, 259-1
Grade 4, 282-1
Grade 5, 264-1
Grade 5 (1-day), 267-1
Head feed, 291-1
Extract, 294-1
Baghouse mill, 297-1
Screen plant, 288-1
S.C. Grace
Grade 3, 430-1
Grade 4, 433-1
Grade 5, 427-1
Mill feed (+100 mesh),
436-1
Grade 3, expanded, 439-1
Grade 4, expanded, 442-1
S.C. Patterson
Ungraded, 573-1
Fibrous
Estimated
mass, %
4-6
4-7
2-4
0.3-1
2-4
2-5
21-26
1-4
8-12
2-5
t
< lb
t
< lb
v
< lb
< 1
v
< lb
v
< lb
< 1
phases
Mineral
types
Tren-actin
Trem-actin
Trem-actin
Trea-actin
Trea-actin
Trea-actin
Trem-actin
Trem-actin
Trem-actin
Trea-actin
Mixed
Anthopbyllite
Trem-actin
Mixed
Anthophyllite
Trem-actin
Mixed
Anthophyllite
Trem-actin
Mixed
Anthophyllite
Trem-actin
Mixed
Anthophyllite
Trem-actin
Mixed
Anthophyllite
Trem-actin
Mixed
Trem-actin
Anthophyllite
Nonfibrous
Estimated
oas*j^%
1-3
3-5
< 1
1-3
2-5
4-8
< 1
6-9
1-3
2-6
1-4
2-4
< 1
1-3
1-4
4-6
2-4
1-3
6-9
< 1
< 1
< 1
0.5-1
4-8
3-12
aophiboles
Mineral
types
Trea-actin
Trea-actin
Trem-actin
Trea-actin
Treo-actin
Trea-actin
Anthophyllite
Trea-actin
Trea-actin
Trea-actin
Trea-actin
Trem-actin
Anthophyllite
Anthophyllite
Trem-actin
Anthophyllite
Treo-actin
Anthophyllite
Trem-actin
Anthophyllite
Tren-actin
Anthophyllite
Trem-actin
Anthopbyllite
Trea-actin
a With the exception of Sample No. 267-1, all results are for composite
samples.
b Fiber bundles were mixed phase materials—both anthophyllite and
tremolite-actinolite were present.
-------
analyses of samples nearly completed. The analysis data are coaplete free °
Doth laboratories for the seven priority samples, but the data say or «ay 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 was interpreted by ORF to mean 100 asbestifora fibers, while IIIRI
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
samples with different fiber characteristics (the grade 5 samples fron Libby
Montana, and Enoree, South Carolina) were selected for each laboratory to re-
peat the analysis using the alternate procedure. Table 2 is a sioaary 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 asbestifora fibers associated
with the smaller size grades of venniculite than with the larger gradao. Both
dust samples collected at Libby were found to have a very high aamhibole con-
tent and indicate that considerable asbestos is removed froa the vemiculite
during beneficiation. The South Carolina vermiculite appears to contain aub-
stantially less asbestiform fibers than does that fron Libby, Montana.
Table 3 is a summary of the phase contrast results of the air sarnies.
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 lower than normal fiber counts.
Given the expected variability of the method, IITRI and ORF results ap-
pear to be in general agreement.
-------
TABLE 2. SUMMARY OF ELECTRON MICROSCOPY ANALYSIS
a Priority1"
Sample sample
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)
264-1
264-0
264-1(0)
264-0(1)
Head feed P
291-1
291-0
291-1
Extractor
294-1
Mill dust
297-0
297-1
Screening dust
288-0
288-1
Analysis,
exfoliated
no
X
X
X
X
X
X
X
A
X
X
X
X
ye«
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Asbestifora fibers, all
Amphibole
Fibers/g Mass
x 106 (ppo)
31.6
23.4
38.9
25
42.0
59
1
65
1.8
118
100
127
98
142
160
119
110
62.5
130
73.8
55.0
100
777
300
1,800
78
48.5
210
59
250
240
1
460
17
840
600
1,200
570
2,600
1,800
350
2,600
670
690
590
420
4,600
35,000
3,000
41,000
lengths
Chrysotile
Fibers/g Mass
x 10« (ppa)
0.9
0
0.9
< 2.1
0.4
< 1
0
< 0.4
—
< 1.4
—
m
< 1.6
< 1.6
1.4
1.2
0.7
•
-
< 1.6
3.5 x 10"3
0
0.01
6.1 x 10"3
0
.
—
_
m
-
0.13
< i
3.4 x 10"3
-
_
.
I
u
0
B
(continued)
-------
TABLE 2 (continued)
a Priorityb
Sample sample
S.C. Grace
Grade 3 P
430-1
430-0
430-1
430-0
Grade 4
433-1
433-0
433-1
433-0
Grade 5 P
427-1
427-0
427-1(0)
427-0(1)
427-1
427-0
427-1(0)
427-0(1)
Head feed P
436-1
436-0
436-1
Grade 3 exfoliated
Analysis,0
exfoliated
no
X
X
X
X
x
X
X
X
X
X
yes
X
X
X
X
X
X
X
X
X
Asbestifonn fibers, all lengths
Amphibole
Fibers/g
x 108
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
0.3
12
1.3
Mass
(PP«)
0.55
< 1
3.7
1
6 5
W • J
35
1.4
2
1 5
!> • J
37
4.8
130
4.1
120
7 3
1 • eficiated
Ungraded
573-1
573-0
573-1
573-0
X
X
X
X
11.7
0.03 3.7 x 10'4 0.03 1.4 x 10~4
1-7 27 < 0.3 - .
°-5 3 0.2 5.3 x 10"3
1.1 4 < 0.3
t«T ??i,T0" 5°™Swin8 the Banple nufflber indicates the analyzing labora-
tory, IITRI and ORF, respectively. The "(I)" and "(0)" indicate! the
the
h« W"e desi«Mted " Priority
the tine the program wag reduced in scope
for complete analysia at
?!" CODSc^d on,.the «««Ple» " received and following laboratory
c««ercial exfoliation, doea not caV ...pi?
-------
TABLE 3. RESULTS OF THE PHASE CONTRAST ANALYSIS OF AIR SAMPLES
COLLECTED AT THREE VERMICULITE SITES
Sample
Libby, Grace
106 Field blank8
133 Field blank8
131 Front loader
148 Pit haul driver
138 Mine analyst
141 Bottom operator
130 No. 2 operator
139 Dozer operator
101 Shuttle truck
104 Screening plant, DW
111 Screening plant, DW
108 Trailer court
136 No. 5 substation
South Carolina, Grace
312 Field blank8
346 Field blank8
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 blank8
533 Field blank8
508 Payload operator
520 Plant foreman
542 Bagger/forklift
513 (KE) downwind
506 Control off-site
515 (SE) crosswind
528 (SW) upwind
Saaple
vol. (£)
—
„
303
297
294
276
285
270
385
390
368
169
111.
-
340
478
240
314
285
287
80
291
154
264
257
354
255
252
249
188
274
299
147
Fibers/cc
ORF
< 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
< 0.02
< 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
< 0.02
< 0.02
< 0.01
0.01
< 0.01
< 0.01
< 0.01
0.01
0.02
i mi
0.04
0.05
0.04
0.01
1.9
• • 7
0.4
9.7
9 • 1
0 2
V • At
0 2
V • &
0 5
W • J
0.02
NDB
0.02
0.04
0 02
W • Wfc
0.03
o-2.
NDB
0 1
V • A
0 OS
W • V^
0.04
mr
fvV
0.02
0.02
0.01
0.3
0.14
< 0.01
0.02
0.04
01
.
-------
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
modifications were made following additional reviews. The protocol "Task 32
Study Protocol for the Collection and Analysis of Vermiculite and Related Sa*
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, and minor changes were made. This section dis-
cusses the general protocol briefly, with major emphasis on areas where modi-
fications were made.
SAMPLE COLLECTION
No najor changes were made in the saapling protocol. The following
items are noted:
1. W. R. Grace representatives would not allow MRI 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. W. 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 reo-
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. *
-------
Air Samples
The air sample filters were retained in the filter cartridges during
transport to MRI. The plugged cartridges were placed in a special container
to maintain the filters in a horizontal position with the collecting surface
up and hand-carried back to MRI. At MRI the filters were cut into three
equal portions and each portion individually taped to the bottoa surface of a
49 x 9 wm Milliporeft plastic petrie dish. A set of one-third of each air sta-
ple filter was hand-carried to the two laboratories for analysis.
SAMPLE ANALYSIS
Bulk Sample Analysis
Since no microscopy technique is capable of aeaiureaent 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
now the analytical results are to be applied.
The basic choice in the analysis of vermiculite was 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 material is 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 X-ray diffraction (XRD) measurements of the amphibole or ser-
pentine content of such a pulverized sample is of inadequate sensitivity (about
* 5?r ?mPhlb°les and possible 5% for serpentine). Moreover, XRD is incapable
of distinguishing the fibrous varieties from other amphiboles or serpentine
1979).
4. The crystallography of the fibers may be altered (Spumy et al.,
The procedure of Chatfield and Lewis (1980)2 was designed to retain the
size distribution of the material as it is noraally used, and to allow venT
(sensitive oeasureoent of asbestos fiber concentrations down to detection lim-
its in the parts per million (ppm) region.
10
-------
c
TJ
Essentially, their procedure was: °
1. To suspend the beneficiated venniculite 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 venniculite 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. Where large amounts of asbestos fibers are present throughout the
whole size 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)3 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 \im in diameter, whatever its length may be.
Hence, the size cut-off in the analytical method should exceed a unit density
equivalent spherical diameter of 10 \tm, which corresponds to a sphere of 5.6
Mm 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 = g'd2 (ps - pL)
18n
where V = terminal velocity
g = acceleration due to gravity
d = diameter of the sphere
ps = density of the sphere
pL = density of the liquid
H = coefficient of viscosity of the liquid
For a sphere of 5.6 \m diameter and density of 3.2 g/cm3 the terminal
velocity in water is calculated to be 0.0037 cm/sec, or 270 sec/cm. 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
-------
I
t)
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 n
this reason that the initial step of a low oagnlfication optical exaaination
was incorporated. In this way such stapled can be detected before effort is
expended on TEH fiber counts which nay be irrelevant. However, the TEH proce-
dure must still be used if determination of the respirable fiber concentration
is required.
Both analytical laboratorieo 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 toae of the results
were obtained. While to a degree the variations in procedures prevent the di-
rect comparison of results, the slightly different approaches coapleaent each
other and give a better overall understanding of the saaples 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 fibero. IITRI performed s»ore
complete qualitative and semiquantitative analysis of the bulk saople using
density fractional separation, followed by component identification by optical
microscopy and X-ray.
2. The following appeared in the protocol for the TEM analysis: "Make
fiber count - determine chrysotile or amphiboles. Count 100 fibers or 10
grids of 200-mesh screen. Determine the limits of detection and count aore
grids if necessary." ORF interpreted this statement to aean count 100 chryso-
tile or amphibole fibers; IITRI interpreted the statement to mean count 100
particle unite 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 saa-
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 TEM was not undertaken. The optical phase contrast analysis was
conducted according to the protocol.
12
-------
E
•d
•o
n
SECTION 4
SAMPLING
Sampling trips were made to the Grace mine and processing facilities near
Libby, Montana, during October 21-26, 1980, 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 was of two types, personal and stationary. For the personal sad-
pies, nine Dupont Model 4000 samplers were used. The flow rates were cali-
brated before and after sampling. Stationary air sampling was 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 were recorded using a Wang meteorological station.
Brief descriptions of sampling conditions and a list of samples collected fol-
low.
W. R. GRACE MINE, LIBBY, MONTANA
Mr. Fred Eaton of W. R. Grace, Cambridge, Massachusetts, was the company
representative for the coordination of sampling. Mr. Jim Salois of MSHA,
Helena, Montana, was present at the request of Ms. Diana M. Kraft, MSHA,
Arlington, Virginia. Mr. Salois was familiar with the Libby facilities, and
his presence and suggestions were very helpful.
At the request of Mr. Eaton, duplicate concurrent personnel air samples
were taken, one for this task and one for Grace. Thus each subject was 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
were determined to have flow rates ranging between 2.03 and 2.19 liters/min.
The flow rates for the stationary samplers were measured at the time the sam-
plers were set up, periodically during sampling, and at the end of sampling.
The specific flow rates, calibration data, and related information appear in
Appendix C of this report.
Sampling was scheduled and conducted on Thursday, October 23, 1980. For
several days prior to sampling the weather had been rainy, and sampling was
started in heavy fog with essentially no wind in the nine area. There was 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 were recorded during the sampling day.
13
-------
t
•a
The objective was to take a short (2-hr) sanple, followed by a longer o
(6-hr) sample that would complete the work shift. The actual tines varied »
somewhat from the intended times, but the actual times and voluaes for each
sample were recorded (Appendix D).
Grace has a routine bulk sampling procedure as part of their product
quality control program. At our request they had taken and retained bulk
samples of the five grades of product plus related head feed, tailings, 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 was 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 wind conditions and site positions for the stationary air samplers. The
bulk samples collected at Libby .ire given in Table S.
W. R. GRACE MINE AND PROCESSING MILL, ENOREE, SOUTH CAROLINA
Mr. Fred Eaton, who was the company representative at Libby, Montana,
was 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 was 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
bjout 1 mile south cf .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 Weihesday. However, because of rain during Tuesday morning,
the mine was closed and the schedule was reversed. A light rain fell Tuesday
morning; the remainder of the sampling period was clear and cool.
The air samples collected at the Grace Enoree operations are given in
Table 6 and the bulk samples are given in Table 7. The wind conditions and
air sampling site positions are shown in Figures 3 and 4 for the mill and in
Figures .S and 6 for the mine.
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 was underway on
the day of sampling, November 6, 1980, so sampling was 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 8 and the bulk samples in Table 9. The wind conditions and
air sampling site positions are shown in Figures 7 and 8.
MRI employees were excluded from all mill operations.
14
-------
TABLE 4. AIR SAMPLES COLLECTED AT THE GRACE MINE AND MILL. LIBBY. MONTAHA
Sample description
Pesonnel samplers
Front loader
Front loader
Pit driver
Pit driver
Mine analyst
Mine analyst
Mill operator, bottom
Mill operator, bottom
Mill operator No. 2
Mill operator No. 2
Bulldozer operator
Bulldozer operator
Shuttle truck driver
Shuttle truck driver
Stationary samplers
Station 7 screening plant D.W.
Station 7 screening plant D.W.
Station 7 screening plant D.W.
Station 7 screening plant D.W.
Station 6 screening plant U.W.
Station 6 screening plant U.W.
Station 2 perimeter D.W.
Station 2 perimeter D.W.
Station 2 perimeter D.W.
Station 2 perimeter D.W.
Station 4 perimeter C.W.
Station 4 perimeter C.W.
Station 5 lower meadow
Station 5 lower meadow
Station 9 trailer court
Station 9 trailer court
Station 8 car loading
Station 3 car loading
Station 3 "22" level dump
Station 3 "22" level dump
Station 1 substation No. 5
Station 1 substation No. 5
Field blanks
Filter
No.
131
135
148
126
138
129
141
146
130
125
139
128
101
121
104
111
112
120
116
124
109
113
145
147
103
149
119
115
108
102
123
122
107
134
136
132
114
137
106
133
110
Approz,
tisw
(hr)
2
5
2
5
2
5
2
5
2
5
2
5
2
7
2
2
6
6
2
6
2
2
6
6
2
2
2
6
2
6
2
3
2
6
2
6
Analysis
assignment
PC EM Hold
X
X
X
X
x
x
x
x
X
x
x
x
x
X
x
x
x
x
x
x
4\
V
A
x
A
x
A
X
X
X
X
X
a PC - phase contrast optical microscopy, EM
Hold - sample retained without analysis.
b Station numbers indicated in Figure 2.
15
- electron microscopy,
C
-o
0
10
-------
NV/
WNW
W
NNW
NNE
NE
SSE
:i£
SE
32<* Calm
'.Vlndi 0-3mph
Figure 1. Wind rose pattern showing the direction and intensity
of the wind during the air sampling period at the
Grace, Libby, Montana, facility.
E
«
10
16
-------
E
•a
10
• A. T 1-0 N_ A L
._— .xi- v « t ..v" '• t ; *£
W.R. GRACE
VERMICUUTE MINE
U8BY. MONTAh4A-
• StaHonory Air
Monitoring Srotlont
Figure 2. Map of the Grace, Libby, Montana, facility showing
the stationary air sampling locations.
17
-------
TABLE 5. BULK SAMPLES COLLECTED AT THE GRACE HIHE ASP MILL. LIBBY. MONTANA
Sample description No. of staples
Beneficiated Grade 1 vermiculite lla
Beneficiated Grade 2 vermiculite 11*
Beneficiated Grade 3 vermiculite 11*
Beneficiated Grade 4 vermiculite lla
Beneficiated Grade 5 vermiculite 11*
Dust from screening plant 8^
Dust from dryer 9C
Head feed 9C
Under 90 mesh 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. .ft,,,,
b Screening plant dusts collected: October 8, 9, 10, 13, 14, 15, 17, 23,
1980.
c Collected October 8, 9, 10, 13, 14, 15, 16, 17, 23, 1980.
XT
18
-------
TABLE 6. AIR SAMPLES COLLECTED AT GRACE, ENORBE, SOUTH CAROLINA
E
•d,
•0,
Sample description
Mine personnel sanies
Truck driver 1
Truck driver 1
Truck driver 2
Truck driver 2
Dragline operator 6
Dragline operator 6
Mine stationary saoples
Crosswind N. station 1
Crosswind N. station 1
Crosswind S. station 3
Crosswind S. station 3
Upwind V. station 4
Upwind W. 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
Mill personnel samples
Forklift operator
Forklift operator
Bagger
Bagger
Bagger
Bagger
Bagger
Bagger
Mill monitor
Mill monitor
Mill laboratory technician
Mill laboratory technician
Mill stationary samples
Mill office
Screening floor
Screening floor
Screening floor
Crosswind N. station 1
Crosswind N. station 1
Approx.
Filter tine
Mo. (hr)
315
324
310
320
301
306
351
318
307
353
316
338
323
352
334
350
331
354
305
339
314
322
330
349
347
337
336
340
308
321
304
300
332
341
345
343
2
2
2
5
2
6
5
4
2
5
4
2
2
6
2
5
2
5
2
4
2
4
2
4
3
4
3
3
3
4
7
2
3
1
2
5
Analysis
aisignaent
PC EM Hold
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
(continued)
19
-------
TABLE 6 (continued)
f
•d
•0
JO
Sample description
Approx. Analysis
Filter tine
No. (hr) PC
Mill stationary samples (continued)
Crosswind N. station 1
Crosswind N. station 1
Crosswind S. station 3
Crosswind S. station 3
Upwind W. station 4
Upwind W. station 4
Downwind E. station 2
Downwind E. station 2
Offsite control
Field blanks
328
313
342
326
344
309
335
302
329
312
346
319
348
327
2 X
5
5
2 (void)
5
2 X
5
X
X
X
X
X
X
X
X
X
a *'C - phase contrast optical microscopy, EM - electron microscopy,
Hold - sample retained without analysis.
b Station numbers indicated in Figure 4.
20
-------
V
id
10
TABLE 7. BULK SAMPLES COLLECTED AT THE PATTERSON OPERATIONS »
ENOREE, SOUTH CAROLINA
Sanple description No. of sanples
Raw ore from stockpile 1
Raw ore prescreening hopper 1
Postscreen ore 3a
Dried ore 4a
Exfoliated final product 4a
Waste from screening 4*
Waste from exfoliater 4*
a Samples were collected at approximately 2-hr intervals.
21
-------
E
•d
10
X
NNW
NNE
NW
WNW
ENE
WSW
Wlixb 4-7mph
Winds 0-3mph
SSW
SSE
23% Calm
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
-------
I y/
Figure 4. Map of the Grace, Enoree, South Carolina, mill area
showing the stationary air sampling locations.
23
-------
f
•d
to
NNW
Figure 5. Wind rose pattern showing the direction and
intensity of the wind during the air saopling period
at the Grace, Enoree, South Carolina, mine.
24
-------
t
rd
(0
^^•H
"~
Rood 50
/V
© ° if
-------
TABLE 8. AIR SAMPLES COLLECTED AT PATTERSON, ENORBE, SOUTH CAROLIHA
C
rd
10
Sample description
Personnel samplers
Bagger/forklift operator
Foreman
Payload operator
Stationary samplers
Crosswind S.E. station 2
Crosswind N.W. station 4
Upwind S.W. station 3
Downwind N.E. station 1
Remote, control
Field blanks
Filter
No.
542
504
517
520
521
511
508
519
516
515
502
503
518
531
540
528
525
513
523
506
527
505
522
533
538
Approx.
tine
(hr)
2
3
3
2
3
3
2
3
3
3
6
2
6
2
6
2
6
2
6
2
6
Analysis
assignment
PC EM Hold
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
a PC - phase contrast optical microscopy, EM - electron microscopy,
Hold - sample retained without analysis.
b Station numbers indicated in Figure 4.
26
-------
rd
io
TABLE 9. BULK SAMPLES COLLECTED AT THE GRACE MIME AND MILL »
ENOREE, SOUTH CAROLINA
Sample description
Eeneficiated Grade 3 vermiculite
Beneficiated grade 4 verniculite
Beneficiated grade 5 venniculite
Dryer composite
Mill feed +100 mesh
Mill feed -100 mesh
Wet 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 3 "coated"
Exfoliated Grade 3 plain
No. of sajpples
7a
7a
7a
7a
7a
7a
lb
1
1
1
lc
lc
lc
1
3 Da^8£ollected for 7'day samples: October 27, 28, 29, 30, 31, November 1,
b Composite collection for October 28, 29, 30, 1980.
c The exfoliated samples were saopled at the Kearney Expansion plant.
27
-------
rd
10
HNt
Figure 7. Wind rose pattern showing the wind direction and intensity
of the wind during the air sampling period at the
Patterson, Enoree, South Carolina, facility.
28
-------
rd
10
X
Mill and
Packaging
Operation
ff
I I
I I
•To Highway 221
Sketch of Pott«non V«rmlcullt«
(Sampled 11-6-80)
A Meteorological Station
Q Air Sampling Sratlon
Figure 8. Map of the Patterson, Enoree, South Carolina,
facility showing the stationary air sampling locations.
29
-------
rd
10
X
SECTION 5
SAMPLE HANDLING
1Thf tlle bulVaaple8 werc 8^PPed air freight to MRI from the sampling
The air samples were hand-carried and maintained in a horizontal,
sample-up position.
BULK SAMPLES
* *nan°Stf **•
**"*•
n°Stf **• the oulk 8amPles were collected as increment samples representing
Jh! **"*•.*»*•«** •«*!« were prepared for this analysis. To pre- *
increment 8afflPle was 'i«led to obtain a rep-
resentativrM K
If «rh T*J " he increment- Approximately equal weight fractions
of each of the increments were combined to make a composite sample. The com-
t£ 5LSF WaS . CD -?Xed and riffl£d t0 produce f°ur equal samples Sne of
anain rlffLTl aside,and, retained " • control. One of the fourths w^s
?fp« "ffjed 1° Produce two fractions, each one-eighth of the original sample
These two fractions were combined with the other two fourths to refult in the
composite division into 1/4, 3/8, and 3/8 of the original composite Tne
tlie othlr «vS«WfB "tained at H*1' one "3/8" fracton was sen? to IITRI, and
the other "3/8" fraction was sent to ORF for analysis.
glvenln
AIR SAMPLES
cumference of the filter but did not contact the sampling portion of the fil-
£SJ^?.^.TMWS=a?,S14L-.W
one-third of each air sample filter was hand-carried to IITRI in Chicaao
Illinois, and to ORF in Mississauga, Ontario. uucago,
30
-------
SECTION 6
ANALYTICAL RESULTS
analytical results of this program consist of findings from (a) the
h 11, Vy diffractio° (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
**** ** *** aPPendice8 are referenced for each sample
nt**"? da-3 obtain?d 5or 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.
(TBE)!' DeDSity 8reatCr than 2>9? 8/CC'
in TRF-c I1"- tha° 2'97 and 8reater thflo 2.76, floats on TBE, sinks
in TBE/isopropanol mixture with density of 2.76 g/cc.
3. Density less than 2.76, floaU 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 vSIlSlite
would be separated from most of the other materials, and that the materials
DercentDo? ? T6™1""*™" be separated into'two fraction's, ^weight
percent of materials in each density fraction of samples is given in Table
are llsteSln^hl^^8 idc?tifjed in each 8an>P^ 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 f ile card
numbers for the patterns which were used to identify the mineral species
The raw data obtained in the analyses appear in Appendix I 8Pecies-
I
rd
10
X
31
-------
TABLE 10. SPECIFIC GRAVITIES OF SELECTED MINERALS
rd
10"
__ Mineral
Venniculite
Biotite
Chrysotile
Serpentine
Talc
Anthophyllite
Actinolite
Tremolite
Fcrroactinolite
Curamingtonite
Grunerite
Diopside
Hornblende
Quartz
01 i vine
Cheaical fcmula
(Mg,Ca)0 3(Mg,Fe,Al)3 o(Al,Si)4010(OH)4
K(Mg,Fe)3(AlSi3010)(OH)2
Mg3Si206(OH)4
M83Si206(OH)4
Mg3Si4010(OH)2
(Mg,Fe)7Si8022(OH)2
Ca2(Mg,Fe)5Si8022(OH)2
Ca2Mg5Si8022(OH)2
Ca2FesSi8022(OH)2
(Mg,Fe)7Si8022(OH)2
Fe7Si8022(OH)2
CaMgSi206
(Ca,Na)2 3(Mg,Fe,Al)5Si6(Si,Al)2022(OH)2
Si02
(Mg,Fe)2Si04
Specific
travitv
.. . fll . HE — :
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
-
wt %
a hand-picked
Sample fibers
Libby Grace
Grade 1, 270-1 4.5
Grade 2, 276-1 4 5
Grade 3, 259-1 i.o
Grade 3, 259-1 duplicate
Grade 4, 282-1 o.3
Grade 5, 264-1
Grade 5 (1-day), 267-1
Head feed, 291-1
Extractor, 294-1 i.o
Baghouse mill, 297-1
Screen plant, 288 I
S.C. Grace
Grade 4, 433-1
Grade 5, 427-1
Mill feed (+100 mesh),
436-1
Grade 3, expanded, 439-1
Grade 4, expanded, 442-1
S.C. Patterson
Ungraded, 473-1
Wt %
tetrabromoethane
sinks
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
18.1
Wt %
2.76
sinks
5.1
5.6
22.6
14.1
11.4
25.6
6.1
27.3
17.6
25.3
48.9
4.6
23.6
0.4
•v 0.4
13.9
Wt %
2.76
floats
85.1
82.2
68.3
75.0
71.4
47.8
38.1
62.2
79.8
71.2
47.2
84.4
50.1
99.4
* 99.2
68.0
n)
Io
JO
With the exception of Sample No. 267-1, all results are for composite
samples. '
33
-------
TABLE 12. SUMMARY OF X-RAY DIFFRACTION ANALYSIS RESULTS
Sample
Fraction-Phase
Mineral phases identified fron XRD data
(exluding vermiculite)
Libby Grace
Grade 2, 276-1
Grade 3, 259-1
Grade 5, 264-1
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
Grade 5 (1-day), 267-1 TBE-SINK-fibers
TBE-SINK-total
Head feed, 291-1
TBE-SINK-total
2.76 SINK-total
2.76 FLOAT-total
Tremolite, talc
Tremolite, talc
Diopside, magnetite
Diopside, magnetite
Tremolite
Diopside, sphene, augite, fluorapatite
Tremolite, diopside, sphene, talc, magnetite
Diopside, treaolite, magnetite, fluorapatite,
sphene, hematite, rhodonite
Treaolite, diopside, talc, sphene, augite,
fluorapatite, quartz, magnetite
Diopside, sphene, treaolite, augite, quartz,
fluorapatite, magnetite, hematite
Diopside, treaolite, augite, fluorapatite,
sphene, magnetite, hematite, quartz
Biotite, treaolite, vermiculite-hydrobiotite,
diopside, quartz, talc, fluorapatite, sphene,
calcite, magnetite, hematite
Tremolite, diopside, quartz, vermiculite-
hydrobiotite, calcite, fluorapatite, talc,
antigorite
(continued)
a 3 a
-------
TABLE 12 (continued)
Sample
Fraction-Phase
Mineral phases identified from XRD data
(exluding veraiculite)
S.C. Grace
Grade 3, 430-1
Grade 5, 427-1
Mill feed, 436-1
S.C. Patterson
Ungraded, 571-1
2.76 SINK-total
2.76 FLOAT-nonmicaceous
TBE-SINK-fibers
TBE-SINK-total
TBE-SINK-fibers
TBE-SINK-green, glassy
TBE-SINK-green, milky
TBE-SINK-colorless, glassy
TBE-SINK-total
2.76 SINK-total
Sodium tremolite, hornblende, anthophyllite,
talc, vermiculite-hydrobiotite, fluorapatite,
sphene, calcite, quartz
Quartz, microcline, albite, sodium trenolite,
sphene, vermiculite-hydrobiotite
Tremolite, anthophyllite, sodium hornblende,
vermiculite-hydrobiotite, talc, sphene,
fluorapatite, albite, aagnetite
Sodium hornblende, treaolite, anthophyllite,
fluorapatite, sphene, vermiculite-hydrobiotite,
magnetite
Sodium treaolite, anthophyllite, talc,
hornblende
Hornblende, sodium tremolite, sphene
Sodium tremolite, hornblende, sphene,
fluorapatite
Fluorapatite, anthophyllite, sodiun treaolite,
hornblende
Trenolite, iron anthophyllite, sodium hornblende,
talc, fluorapatite, rutile, sphene, ugnetite
Talc, treaolite, anthophyllite, hornblende,
quartz, rutile, fluorapatite, veraiculite-
hydrobiotite
a With the exception of Sample No. 267-1, all results are for composite saaples.
8 3 S nt
-------
KEY TO TABLE 12
rd
10
X
Mineral name
Micaceous minerals
Verniculite
Biotite
Vermiculite-hydrobiotite
Amphiboles
Tremolite
Sodium tremolite
Anthophyllite
Iron anthophyllite
Hornblende
Sodium hornblende
Pyroxenes
Diopside
Acmite-augite
Serpentine
Antigorite
Iron oxides
Magnetite
Hematite
Others
Talc
Quartz
Microcline (feldspar)
Albite (feldspar)
Calcite
Fluorapatite
Sphene (titanite)
Rutile
Rhodonite
Chemical formula per JCPDS (file card
(Ng2.37Feo.37Xo.2e)(All.28Si2.72)09
-------
The XRD data generally confirmed the mineral identification nade micro-
scopically. The XRD data also confirmed the observation that partially al-
tered mineral phases-particularly altered biotite-vermvculite phases--were
present. Since venniculite is generally formed as an alteration product of
bi.otite mica, it was not surprising to find the intermediate phase, mixed
layer vermiculite-hydrcbiotite material in most samples.
The XRD analyses did provide some 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 (MRI Sample
No. 427-1). Microscopical examination of the ground material submitted for
XRD revealed that the vermiculite was intergrown with the amphibole fibers
and also did in fact occur in a pseudomorphically fibrous crystal habit.
Another rather interesting result of the XRD work was the identification of a
sodium-bearing fibrous tremolite phase in the South Carolina samples.
Interpretation of the XRD data was hampered by the peak intensity altera-
tions caused by crystal preferred orientations. Most of the mineral phases
had crystal morphologies with at least one exaggerated crvstallographic axis.
Thus, in preparing samples for XRD as thin films by filtration onto silver
membranes, the crystals tended to orient with the exaggerated crystal planes
parallel to the filter surface--!.e., venniculite plates and tremolite fiberc
landed on the filters lying flat, rather than on end. Even the pyroxenes,
such as diopside, which generally do not show the prominent prismatic mor-
phology because cleavage along the prism planes is not so perfect as it is
i:i 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 were 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
were hand-picked and were relatively pure phase materials. Diffraction peak
positions were consistent with the phases identified—tremolite and diopside—
but relative peak intensities were not consistent with 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 were due to crystal
orientation effects. That is, peaks showing higher relative intensities com-
pared to published data corresponded to crystal planes that were preferentially
placed in the x-ray path (e.g., the elongated axis of a prism or fiber) while
absent peaks or peaks with low intensities compared to published data corre-
sponded to crystal planes placed essentially out of view of the x-rays (e R
the end-on-view of a prism or fiber).
Individual Sample Results
The results for individual samples follow. The order of presentation
-------
Much of the detailed data and selected photographs are contained in ap-
pendices. The appendices appear in this volume and three suppleoentary vol-
uraos
The presentation for each sample include the sample type, code nuobers
aligned to the sample** and appendices page references, optical microscopy
diacription of the sample, and electron microscopy results.
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, Montana1 GrarP, r.rade 1. Composite
IITRI Code No. 129
Appendix references
Electron microscope 1-84-87
Macroscopically, this sample was composed of 1- to 20-nm chunks of gold
to green micaceous minerals, 1- to 15-mm bundles of white to pale green fibers,
and 1- to 8-mm chunks of nonmicaceous, nonfibrous minerals. The nonaicaceous
minerals ranged in color from deep green, to pale green, to white to colorless.
Prismatic as well as 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, as detenniqed 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 components. 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 so
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.
•E
rd
10
38
-------
TABLE 13. COMPOSITION OF SAMPLE 270-1
Estimated mass
Mineral phase concentration (%)
Tremolite-actinolite fibers 4-6
Tremolite-actinolite prisms 1-3
Sphene 1-2
Diopside 2-5
Augite < l
Hornblende < 1
Magnetite, hematite 1-2
Calcite 1-3
Quartz 3-5
Biotite 1-2
Talc < i
Vermiculite 72-82
Other minerals 1-3
Tremolite-actinolite was 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 was not as well-formed as it was 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 was 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 was composed of lamellated tremolite-
actinolite 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, calcite, 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
•e
rd
39
-------
rd
10
samples did. A higher proportion of the stacked vermiculite plates also ap- »
peared to be weathered in this sample compared to other samples. A summary
of the EM results for this sample appears in Table 14.
TUU 14. KIWI of IUCTM HionKorr qyATi n» utnt HUT, am. out i _
nu, ,-...,..1.. 'ii'""" - nu, ,-,„""
<„.) tw.t
HQ-t tifclut
*< )!.*
0.9
*.*-ii.l
1.J4-I1.1 0.9
19 M 1.3
l.J • I0"' 1
44 4 ^
r* r
o c
Sample 276, Libby, Montana, Grace, Grade 2, Composite
IITRI Code No. 128, ORF No.
Appendix references
Photographs E-14,15; XRD 1-157-160
Electron microscope 1-88-91
The sample was composed of 1- to 12-mm flakes of gold to deep green mica-
ceous 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.
-------
rd
10
X
TABLE 15. COMPOSITION OF SAMPLE 276-1
Estimated mass
Mineral phase concentration (%)
Tremolite-actinolite fibers 4-7
Tremolite-actinolite prisms 3-5
Sphene < 1
Diopside 4-7
Augite 1-2
Hornblende < 1
Magnetite, hematite 1-2
Calcite 1-2
Quartz 3-5
Biotite 2-4
Talc < 1
Vermiculite 66-72
Other minerals 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. Rather, the fiber bun-
dles were composed of short fibers intergrown at slight angles to each other.
All fiber bundles contained both the truly fibrous material as well 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
-------
.
«l Ca«M<«act
1*J^1« fWt« lot«r**t
r*«tik
L*4 •«•» to. •!
r«tl«a f i»*r» Ftfc»i
P») «M*tt4 irn
rt)
10
M
l lirillalrt J).^ 41. 1 1) I.)
Jl.l 0.4 It 1.1
00
Sample 259, Libby, Montana, Grace, Grade 3, Composite
IITRI Code No. 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 am) 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
Estimated mass
Mineral phase concentration (%)
Tremolite-actinolite fibers < 1
Tremolite-actinolite prisms 2-4
Sphene 1-3
Diopside 3-7
Augite 2-3
Magnetite, hematite < 1
Calcite 2-5
Quartz 2-5
Biotite 10-15
Vermiculite 65-72
Other minerals 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 pleo-
chroism.
Other mineral phases (including alteration products) were included in
each fiber bundle. In 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
fibirous 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
venniculite plates and growing in between layers of vermiculite. A summary of
the EM results for this sample appears in Table 18.
IE
rd
10
X
nt>«r* el ill ItBtlh
(CTTOT KICMUOT HJUITJ rt» Ur»U
•
<.„,. *,
«i-i )
151 Cwtfldvu* 1 fiber t c
« iftttrv.l >1»l»fi.<
. i n.i-4*.T
.a IVI-M.I 0.4
t
«.i.»i i.n
.1 • 1.01
.0 JI.I-^O.^
.1 V*. ••!*.* 0.*
.t
i4-l« 0.111
.0 • 0.11B
••cfntr.llv* fib*
118 '
0.01
» 1
»0 101
11
6.1 • ID*1
IW >
r
t 15
r4 N«in
11. »
II. 1
10
' 1.1
1>.>
14.1
It
< 1.0
LIIIT. CUC1. CUM ]
riWfl If.t.f IkM t.O * 1, !.»„»
»^«lv*lMt !• fltlMtt4 Mil
I.I-I7.I 118
t.l-ll.l 0.4
0-11 1.01 11
I.M
J. 1-10.1 140
>. 1-10.1 0.4
>.1-J1 O.MI IW
o.m
«. «f
to
0
0
)4
]4
0
II
0
;;-,
A
T
C
A
C
A
T
C
A
C
lltolt (SUO)i C • rhrriollU; i
Sample 282, Libby, Grace, Grade 4, Composite
IITRI Code No. 126, ORF No. 281
Appendix references
Electron microscope 1-92-97, 11-356-375
Macroscopically, the sample was observed to contain mostly 1- to 4-ran
goldish-brown vermiculite flakes. White fiber bundles up to 3 mm in length
were visible. Nonmicaceous, nonfibrous mineral phases were also observed; at
least three different mineral phases ranging in color from white to emerald
green were detected.
A3
-------
rd
10
The mineralogical composition determined for this sample by polarized x
light, microscopy is listed in Table 19.
TABLE 19. COMPOSITION OF SAMPLE 282-1
Estimated mass
Mineral phase concentration (%)
Tremolite-actinolite fibers (0.3)-1
Tremolite-actinolite prisms 1-3
Sphene < 1
Diopsidc 3-7
Augite 1-2
Hornblende < 1
Magnetite, hematite 1-3
Calcite < 1
Quartz 1-3
Biotite 1-3
Vermiculite 78-88
Other minerals 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
treraolite-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 pricm bundles, however, did tend to be composed of
crystals growing parallel to each other. Bundles composed of both the truly
fibrous material 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 anthophyllite, 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
-------
ro
JO
classifiable as fibers were composed primarily of the thicker, more brittle QQ
lamellated prisms.
The 2.76 sink fraction contained a significant fraction of biotite. The
other major constituents of this fraction were vermiculite flakes intergrown
with tremolite-actinolite, diopside, and iron oxides.
The 2.76 float fraction was relatively free of pyroxene and amphibole
mineral fragments. Few flakes of vermiculite intergrown with amphibole min-
eral phases were detected. A summary of the EM results for this sample ap-
pears in Table 20.
flb*f» «f ill Imiithi flbtri • -••Ur !••• 5.0 *» !• Itflith . . ..
&••»!•
111-0
Ill-l t-fvlUU'l
1.0 0>1.2 O.)40 1 1-0-4
' 0.4 - 0.3*0 • 0 < 0.4
61.0 440 11 1T.1
r:.n o.t toi 17. a
0
• Ub«r./t>
0.340
O.J40
0.7
,.M,^ «.. ... .t
(pp«> C«M>t«d (TF
0
o
440 IS
0
1.144 II 4 0.4 0-1.» 0.444
0.44ft • 0 < O.S - 0.4*4
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 was observed to be a fine, goldish-brown pow-
der composed of obviously flake-like and fibrous particles. The flake-like
particles were generally less than 2 mm in diameter. Fibers up to 3 mm in
length were present. At least two other nonvermiculite, nonfibrous mineral
phases were observed; one was green in color, while the other was colorless.
The mineralogical composition of the sample determined by PLM analysis of
the three density-separated fractions is listed in Table 21.
-------
TABLE 21. COMPOSITION OF SAMPLE 264-1
Estimated mass
. Mineral phase concentration (%)
Tremolite-actinolite fibers . 2-4
Tremolite-actinolite prisms 2-5
Sphene 1-3
Diopside 6-9
Augite 1-3
Hornblende < 1
Magnetite, hematite 1-3
Calcite 1-3
Quartz 1-3
Biotite 3-7
Venniculite 70-74
Other minerals 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.
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.
rd
10
00
46
-------
ILM9lt
:»*•>.
.-.4...
:»•-( to)*
:.4*o (||»
J44-I Ultimo
Ifcfc-0 ItUli*l«4
•ii.--.-i tot k
ti(olul«4
:**-o til
tldllljlld
thrylot.U.
"y^
Or**
III
11)
100
» 1.4
111
*>•
* 1*1
V
M
1M
lit
l»«
140
* 1 .»
lit
110
* 1 »
Ji
ltd
!•!».
11-1)0
1.4.1
0.4-11
11-11
IIO-I'O
r>».,. .1 .11 !«•<>.
TIO1 M.M/I!
.^-ji..lnt I. til
0..
1.1'
l.ll
1.1
1.42
1.41
1.41
1.41
1.1
1.))
1.11
4 1.1
1.10
1.10
1 10
1.10
1.10
LMl«-l ••••
tn*>
140
400
1.200
110
14
II
•M
1.400
1.100
•
!M
1.400
• 1
tl
41
1.400
U .1
11
107
O
r*
0
101
c
H
IS
II)
111
III
ttrt
10.'
n
i
4
It
1J
fiu, ..«...ffi|.. il^^;(ij.
itoAB lAt«rv«l -htttrt*-!
11. 1 1.3-11.1
11. ft 1.2-IOt 0.»
4) rl-M l.ll
c 1.4 - l.ll
41.2 „...,,., 1.1
II 11-11 l.->2
1., 0-1. 1 1.41
1.4 0-4. ft 1.42
41 11-44 1.42
14.2 41.1-121
11.1 41.4-121 1.2
11 41-11 1.1)
14.4 44.0-44.2 1.2
42
41
44
0
II
"
o
1
11
I]
tl
0
Jl
AH
o rt
1 <•»
1JL1
u r
rd
.0
00
if IITII uiinf the utF
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 ram
in length were visible. At least two green, nonoicaceous 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 inter-
grown anthophyllite--was definitely present. No serpentine fibers were de-
tected, however.
47
-------
re
10
00
TABLE 23. COMPOSITION OF SAMPLE 267-1
Estimated mass
Mineral phase concentration (%)
Tremolite-actinolite fibers 2-5
Tremolite-actinolite prisms 4-8
Anthophyllite (prisms and fibers) < 1
Sphene < 1
Diopside 10-15
Augite 1-3
Hornblende < 1
Magnetite, hematite 1-3
Calcite 1-2
Quartz 1-3
Biotite < 1
Talc 1-2
Vermiculite 65-70
Other minerals 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. Rarely 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 treraolite-
actinolite 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.
-------
re
rfl
.10
Unlike the 264-1 composited grade 5 sample from Libby, this uncomposited <#
sample did not appear to contain other mineral phases in pseudomorphically fi-
broun habits. None of the fibrous calcite detected in the 264-1 sample was
found in this sample.
Most of the nonraicaceous contaminant minerals present in this sample were
high density materials and were thus found in the TBE 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 XRD 1-130-134
Electron microscope 1-2-13, II-1-19
Macroscopically, this sample was quite variable in color, particle mor-
phology, and grain size. The overall color was a light brown. Relatively few
micaceous flakes were visible to the naked eye; the largest flakes were less
than 10 mm in size. Under the stereomicroscppe, most of the brownish, fine
powder (less than 1 mm) material present was observed to be micaceous. Sev-
eral large white to pale green elongated (and probably fibrous) rock chunks
greater than 20 mm were observed. Obvious mixed phase grains (i.e., mineral
phases partially altered) were also present as 1- to 15-mm grains.
Microscopically, the sample was 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
Estimated mass
Mineral phase concentration (%)
Tremolite-actinolite fibers 21-26
Tremolite-actinolite prisms 6-9
Sphene 1-3
Diopside 24-29
Augite : 2-5
Hornblende < 1
Magnetite, hematite 3-5
Calcite 3-5
Quartz 4-7
Biotite 1-2
Vermiculite 20-25
Other minerals 3-6
49
-------
ft
id
,10
Tremolite-actinolite was a major component of this sample and occurred in oo
both fibrous and prismatic crystal habits. The fibrous habit was found as
discrete fiber bundles, fiber bundles intergrown with prismatic amphibole and
pyroxene mineral phases, and as small fiber bundles protruding fron vermicu-
lite (or other micaceous mineral) plates. Fiber bundles were 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 wide range
of chemical compositions was present; i.e., some end member tremolite and
ferroactinolite phases as well as the intermediate actinolite were present.
The tremolite-actinolite fibrous phase was 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 lower density phases within each fiber
bundle as well as attachment of fiber bundles to lower density mineral grains
resulted in a lower than normal bulk density for the tremolite-actinolite.
Some unusual fibrous phases were present. Fiber bundles exhibiting the
anomalous blue and pink interference colors typical of crocidolite (riebeckite)
asbestos were observed. The refractive indices of these fibers as well as
their inclined extinction angles (and XRD data) ruled out crocidolite as the
mineral species. The fiber bundles were strongly pleochroic (yellow-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 was impossible. Hany inter-
mediate, partially altered phases were present in this sample. XRD data sug-
gested that antigorite is present in the low density fraction. A summary of
the EM results for this sample appears in Table 25.
___ t.iu n. latufi a IUCTK» nicmtopT man n»jwi_iniT. eua, MAO tup __
vo t» i« IMH
•*«l*ll«tt !• lltlMt*4 Mil to. •( M»lvtlfM t« tlttMlrt Mil ••. •!
1U CMflfeMt I flWf cMiralMIIM Uteri nlC««ll«tMI I IIMr cMCIIlrillM IIHrt
l»»r>il 4.1. . Hi In*) inatlt MM llt.r-ll faUclH c»»l«4
1-M-l
m-0
J'll-l [.lain
»1 i
M.4
1 4
IM
1.1
•t II. 1
41. 1-1). J
10- IW
00.1
iT.;*ioi.f
0.1
I.I!
I.IT
410
0.11
4*0
« 1
110
fl 2k. 1 ll.i-31.?
IOT ir* il.*>4i.«
1
111 II H-44
1 • 0.4
•> 10. i I-J1
O.T
0.111
0.141
t)0 U
40
0.0« 1
110 11
0
140 II
50
-------
i
rt
,10 •
Sample 294, Libby. Grace, Extractor Waste, Composite oo
IITRI Code No. 134
Appendix references
Photographs E-15
Electron microscope 1-98-102
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 1-4
Tremolite-actinolite prisms 1-3
Sphene 1-2
Diopside 3-7
Augite < 1
Hornblende < 1
Magnetite, hematite 1-2
Calcite 1-3
Quartz 4-10
Biotite 6-9
Talc < 1
Vermiculite 68-76
Other minerals 1-3
51
-------
re
in)
.10
The tremolite-actinolite occurred primarily as fibrous crystal bundles oo
and crystal bundles containing both fibrous and prismatic materials. Coarse
crystals composed only of bulky, prismatic materials were rare. Tremolite-
actinolite 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 small 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 vermicu-
lite was 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
pseudoraorphically fibrous calcite was 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 such low concentrations and could not 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 EM results for this
sample appears in Table 27.
ruu ;r sotuiT or menu* Hictojwrr JUUITI ft* auffti UMT. c**ci. tmucro* .
fl»tr» •( .It l
Flfcx t.xii«nt nt i»n [II
1i\ CMlU#*t» I llWc run
fteia tattrvil d«lftt«4
.
iaintiwi tiWfi *M CMflfeMi I lltaf CMCnlfitiM fltori ritor
(ffmt CMIIH NMi uicrol tfturlttf fn>> CMal»4 lr»*
111-1 til. II. I. < JVO >IO II II. •
• M 6.1 101 JS.S
0.) ).» i 10'' 1
A " •WMtali (SAID): C * rhrTiitltlt ••« T •
52
-------
re
ird
,10
Sample 297 Libby. Grace. Baghouse. Mill Dust. Composite w
IITRI Code No. 136, ORF No. 296
Appendix references
Photographs E-17
Electron microscope 1-103-108, 11-376-388
The sample was a brownish-green, very fine powdered mineral dust. Obvi-
ously micaceous flakes as large as 4 mm in diameter were observed. Obivously
fibrous crystals up to 2 ran 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 temperature-
induced 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
Estimated mass
Mineral phase concentration (%)
Tremolite-actinolite fibers 8-12
Tremolite-actinolite prisms 2-6
Sphene 1-3
Diopside 3-6
Augite 2-5
Hornblende . < 1
Magnetite, hematite 1-3
Calcite 1-3
Quartz 2-5
Biotite 1-3
Talc < 1
Venniculite 55-61
Other minerals 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 tremolite-
actinolite 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 latnellated prism types of tremolite-actin«lite 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 macro-
scopically definable as fibrous and fiber-like crystals produced from nineral
grains composed of thin, narrow, lamellated prisms, may be immaterial. It was
certainly not possible to define the origins of all the fiber-like tremolite-
actinolite 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 EH results for this sample appears in Table
29.
re
«j
10
oo
rcHri it iii imiti r
tor* iriitir tkM i.o im !• Itvclk
"TlUc iMtfiiiiim no- iiMff/u tim STiTvivin ii-.im[ii»4—
^Ti»in"" tllMt* Mil U. •! t«l«i«i'« bllMlM MM to. .1
m CM|I0 100 *]•
< 1.1
»>•! tlf.H.IM III
,U 1.10 4. WO « »» W-JI I.JO ».«>• «1
1.10 • 0 < 1.1
I). 00* 1M Ul
I.I III tot
0
1.10 • »
It. 000 10
I.I '1
I < > ttltltolt IIAIDIi C • TkrTMtllll >M T • Ulll.
Sample 288, Libby, Grace, Screening Plant Dust, Composite
IITRI Code No. 135, ORF No. 287
Appendix references
Photographs E-16
Electron microscope 1-108-113, 11-389-402
Macroscopically, 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.
-------
re
rt
.10
00
TABLE 30. COMPOSITION OF SAMPLE 288-1
Estimated mass
Mineral phase concentration (%)
Tremolite-actinolite fibers 2-5
Tremolite-actinolite prisms 1-4
Sphene < 1
Diopside 3-6
Augite 1-2
Hornblende < 1
Magnetite, hematite 1-3
Calcite 1-3
Quartz 4-7
Biotite 1-3
Talc < 1
Vermiculite 68-78
Other minerals 1-4
Identification of mineral phases was 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 \m 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 antho-
phyllite 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
-------
ft
ird
^,,u°ther fibr,ous Phases present included a morphology of hornblende that I1°
couia DC considered fibrous, and na»nrinmnmh<<»aii.. r^i...... ...._*._ i __i_^_ oo
could h r« v
The unfL«?°« Si™ Kiflbr°U8> and P8cud°n,orphically fibroui quartz and calcite.
If«n?i I ftable blue-green amphibole with the anouloue interference colors
similar to glaucophane-riebeckite series amphiboles was again detected.
in.*..1?6 2>?6 f.°at fraction contained a mineral phase with optical and morpho-
logical properties consistent with nonfibrous serpentine. Insufficient numbers
of P«ticle8 were available for further identification studies. A suamary of
the EM results for this sample appears in Table 31.
_r«»mii Pftxn or mem niacitorr m,
.... - -sr TS- =5* H .„ -as-
lll-l b(tlliu<
»0 IW-MO
1,100
1. 100
'•" >."5 III 10* 11. IW 1 II
I.M - 9 < 1 t . | {J
».( tl.OOO t| |,fM
0
• c
II t
Sample 430. Enoree. South Carolina. Grace, Grade 3. Composite
IITRI Code No. 121, ORF No. 429
Appendix references
Photographs E-7-8, XRD 1-135-138
Electron microscope 1-60-65
hroUnh-> the Sample was Ob8crved to contain 1- to 5-mm black to
brownish-gold nucaceous flakes, and 1- to 3-mm fragments of nonmicaceous min-
erals. The nonmcaceous minerals were white, pale green, or reddish-brown in
V "' '"
The composition of this sample as determined by PLM analyses of the
density-separated fractions is listed in Table 32. anaiyBes 01 tne
56
-------
TE
ird
no
TABLE 32. COMPOSITION OF SAMPLE 430-1 «o
Estimated mass
Mineral phase concentration (?
Anthophyllite < j
Tremolite-actinolite 2-4
Augite < j
Hornblende j_2
Apatite j_2
Magnetite, hematite 1-2
Calcite < j
Quartz, feldspars 4-5
Talc j.j
Vermiculite 80-90
Other minerals 1-2
As is evident from Table 32, the sample was composed primarily of vermic-
uUte Relatively little contaminant, nonvermiculit? mineral matter was
present.
Th. Botht!:emolite~?ctinolite and anthophyllite amphiboles were detected.
The tremolite-actinolite occurred almost exclusively in a very bulky wis-
fr ctu;e0rPfrom°tr However' '-"el-gated, parallel-sided £rUc& cou d
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 was present
Certainly, elongated fiber-like particles of anthophyllite were observed.
The 2.76 float density-separated fraction contained some unusual, not
iv character,^!,. ..^lti. irregular, light green particles closed
rather similar to serpentine minerals WP™ pre8-
258 '''"'"such- A iamtry °f th<
f|t.|
::: '•"'•• ••• ,.;... » - ... ••" ; •
-t! °"-" S:£ V I :::} : - : j j
.»-.«.,.„.,- ,., . ,, u •
« ,., ... >•• i .
tJO-O Uflll.lH 1.1 J.J.4 0 «fll ' I ... « C
•:- • S :::5 : 2:2! : • •
57
-------
re
"a
Sample 433, Enoree. South Carolina, GracRr 4 Comosite ho
—^————————• 100
IITRI Code No. 127, ORF No. 432
Appendix references
Electron microscope 1-114-119, 11-403-420
TABLE 34. COMPOSITION OF SAMPLE 433-1
Estimated mass
"meral phase concentration (
Fibrous mixed amphibole < j
Anthophyllite-prismatic 1-3
Tremolite-actinolite 1-4
Sphene, ilmenite < j
Augite < j
Apatite j_.j
Hornblende 2-5
Magnetite, hematite 1-2
Rhodonite, pyrolucite < i
Calcite < j
Quartz, feldspars 3.3
Talc 1-3
Vermiculite 7c 01
-., . • fJOi
Other minerals 2-5
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 inter-
grown 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 venniculite. In general, the vermiculite plates of this sam-
ple were more irregular, strained, and intergrown with other mineral phases
than were the venniculite 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 venniculite. A summary of the EM
results for this sample appears in Table 35.
rt
ird
no
1--I— fl
FlWt t»,..lr.H~ [I
iMflt I*J4« I4t«ml
411-1 1.4
4.1
411-0 1.1 0.1-1.)
< 0.1
411-1 llf.lllKl 1.1
1.1
411-0 t.t.ll.1.4 1.1 0.1-4.1
< 0.1
k»fl 4f 41 titilhB
'LI..ii!.'L
I llto
ttl«ll<
O.I
0.1U
0.24*
0.4
0.144
0.141
tlllMU4 MM •». •(
"•" '"""4
I.I II
4
0 0
U II
0
1.4 1
II
0
: 10
0
lit... .r
IVtitm i1' » JBJIUH
fiHf t»*cmt4ii«« I1U .. "f"'fj —
C4MM4r4tlM
•^ilnlffBI t* C4UM144 M4I ••.
til CMrifttae* 1 tlfe*r c«M4«lr4tlM (Ito
«... i.<.r.il
O.J
t.l
0.1 IHJ.I
« 0.1
1.4
0.4
< 0.)
< 1.)
••UCIH (*f41) C*M
>.S
• .1
«.iu n
0.14*
0.44
O.I
0.1M
0.144
1
44 «r»«
A
7
C
«
C
A
T
C
A
C
I 4 • M»hlk*lt IIAUll C • (lrrl«lllll *M 7 • ttl4l.
Sample 427, Enoree, South Carolina, Grace, Grade 5, Composite
IITRI Code No. 119, ORF 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 rnineralogical composition of the sample determined by PLM analyses of
the density-separated fractions is listed in Table 36.
59
-------
re
irfl
110
TABLE 36. COMPOSITION OF SAMPLE 427-1 m
Estimated mass
Mineral phase concentration (%)
Fibrous mixed aopbibole < 1
Anthophyllite 4-6
Tremolite-actinolite 2-4
Sphene, ilmenite 1-3
Hornblende 2-5
Apatite 1-2
Magnetite, hematite < 1
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.
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 (titanitel 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
-------
r»l«lt«l u mint* mil It. .f
Ml CMfl«nc> i liter mnMMilM fltaro
1.1.ml 4itKlt4 (m) nmlH
I (iMr
OlUtul
titMlM OM4 ••. «f
r»n>uitl« lltari n«,,
(»•» natM Irx
417-1
427-0
427-1
-------
MIDWEST RESEARCH INSTITUTE
425 Volfcer Boulevard
Kansas City. Missouri 04110
Telephone (816) 753-7600
September 30, 1982
Dr. E. T. Chatfield
Ontario Research Foundation
Sheridan Park Research Community
Mississauga, Ontario, Canada
L5K 1B3
Dear Dr. Chatfield:
Asbestos Contamination."
I wish to thank you for your contribution to this task and in the prepara-
tion of the report.
Sincerely,
Gaylord R. Atkinson
Task Leader
Approved:
5hn E. Going /
Task Manager
GRA:JEG:bm
End .
-------
MIDWEST RESEARCH INSTITUTE
425 Volker Boultvaro
Kansas City, MlMOurl 64110
Telephone (816) 753-7600
September 30, 1982
Mr. David R. Jones
I IT Research Institute
10 W. 35th Street
Chicago, IL 60616
Dear Mr. Jones:
Enclosed is one copy of the Vermiculite final report, "Collection, Analy-
sis and Characterization of Vermiculite Samples for Fiber Content and
Asbestos Contamination."
I wish to thank you for your contribution to this task and in the prepara-
tion of the report.
Sincerely,
Gaylord R. Atkinson
Task Leader
John E.,Going
Task Manager
GRA:JEr,:bm
End.
-------
TABU: 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 were detected, mostly in the tetrabromo-
ethane sinks fraction, but were less than 1% of the total sample. Both antho-
phyllite and tremolite-actinolite fibrous amphibole phases were detected. In
addition, it is likely i:hat fibrous hornblende was also incorporated within
the fiber bundles.
The three major amphibole types present, anthophyllite, tremolite-
actinolite, 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 were obviously
layered and cleavable to particles definable as fibers.
Talc was again rather abundant and was also found as fracture fragments
that might be classified as fibers.
The milky green, rough textured, irregular mineral grains were isolated
from the TBS fraction an.d 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)
were consistent with tremolite-actinolite, but refractive indices were slightly
lower 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 was a sodium tremol'ite. A summary of the
EM results for this sample appears in Table 39.
62
-------
T>IU n. B»a>»T » ivunef IIIOMCTT tuctTi PM uiru DCTC. torn oaeum.
tiwf« irm«c "" 'Q "• '• '«"'»
•ttr.ilrat i. UtUtlM mi •>. •( ««.ITIUU u UllliM «o to. •(
Ht CnfKMcf I liter CMtnlnllo llMn nl CMfldran I lltot c«*cwnll» tlWti
liuml ) 1«>IW Rtn I turn I
O.J o.tl II O.I ' •.» ) t
<•> ».)•). I o.o) n o.i o.i t r
0 0 C
" '•>•>• 0-lM II f> 1.0 O.l-l.l 0.1M II t >
»•! 0-0.1 O.IW < I I 0.1 0-4.1 O.IH < I 1C
|.| o.ll 1 ' 0 A
> ' (••» u o,» o.t i r
0 0 C
IUD)| C • ctryMtllii M t • I. 111.
Sample 439, Enoree, South Carolina/ Grade 3, Commercially Exfoliated
IITRI Code No. 133, ORF No.
Appendix references
Electron microscope 1-120-121
The sample was 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, stereomicroscopic 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
Estimated mass
Mineral phase concentration (%)
Fibrous mixed amphibole <
Anthophyllite-prismatic <
Tremolite-actinolite <
Sphene <
Augite <
Apatite 1-3
Hornblende < i
Magnetite, hematite 1-2
Rhodonite, pyrolucite < 1
Calcite < i
Quartz 1-3
Talc 1-2
Vermiculite 85-95
Other minerals 1-3
63
-------
Fluorapatite was the primary nonmicaceous 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 nonmicaceous
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 EM results for this sample appears in Table 41.
T"H «!• «^m * tmrgJ!lggt<»T Win P* F*H I***. *m cuauu. • tlllMtM XII N. .1
'•Mnlnlio litori 111 CatI4nci I fitaf
m CnllMui int _
Hum tiurval telMU4 Iffm) e*Ml*4 n«M ~uur»l * teuct«4 ' (ppaf" ctwtt* i'
M,lik.l. (UIOI. C '
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 ran.
Lengths of the stacked plates ranged from 1 to 5 mm. Particle colors were
white, tan, brown, and greenish brown. No nonmicaceous 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
-------
TABLE 42. COMPOSITION OF SAMPLE 442-1
Estimated mass
Mineral phase concentration (%)
Fibrous mixed amphibole < 1
Anthophyllite-prismatic < 1
Tremolite-actinolite 0.5-1
Sphene < 1
Augite < 1
Apatite 1-2
Hornblende < 1
Magnetite, hematite 1-2
Rhodonite, pyrolucite < 1
Calcite < 1
Quartz 1-2
Talc < 1
Venniculite 85-95
Other minerals 1-2
Tremolite-actinolite was the primary nonmicaceous mineral type in both
the tetrabrornoethane 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 small (less than 10 pm 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.
However, 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, XRD 1-148-152
Electron microscope 1-79-83, 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. Brownish-
gold, nonmicaceous grains ranged up to 10 mm in size, while the micaceous
65
-------
flakes were 7 to 8 an in ••ximua dimension. Fragments of nonmicaceous min-
erals up to 20 no in diameter were present. Colors of the nonmicaceous min-
erals were milky white, milky green, and glassy green. No obviously fibrous
phases were observed in either the unnagnified or stereomicroscopic examina-
tions.
The composition of the sample, as determined by PLM analyses of the
density-separated fractions, is listed in Table A3.
TABLE 43. COMPOSITION OF SAMPLE 573-1
Estimated mass
Mineral phase concentration (%)
Fibrous mixed amphiboles < 1
Anthophyllite 4-8
Tremolite-actinolite 8-12
Sphene, ilmenite, rutile 1-2
Hornblende 1-3
Apatite 1-2
Magnetite, hematite 1-2
Rhodonite, pyrolucite < 1
Calcite < 1
Quartz, feldspars 26-32
Talc 12-16
Venniculite 33-38
Other minerals 1-3
This sample appears to have been exposed to some type of heat treatment.
Glassy agglomerates were observed in the total sample and were, of course,
concentrated in the 2.76 floats fraction.
No obviously fibrous bundles were observed in the stereomicroscopic exam-
ination of the TBE fraction. Small, elongated, coarse fibrous to prismatic
white particles were observed in the TBE fraction; however, they were in
greater abundance in the 2.76 sink fraction. In the TBE, this particle type
was 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 were generally not teasable with a fine needle and thus are not in a
true fibrous habit. However, 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 were 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 were isolated and carefully examined. The particles were
found to be composed almost exclusively of talc and anthophyllite. Tremolite-
actinolite was 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 suggestive 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 was found in elongated, thin crystal habits. A summary of the EM re-
sults for this sample appears in Table 44.
T"g **• HWT °f ittgy* "lireyn *W* n» »«n* a**n. *m cuauu. utroai. nameo
-rre;
'""*'"*
tit*n tT9*\»t i>«< a.o •! i« i«>ttii
•^HlVtl*«t 1* t*lt*tlt
n\ CMM«MC* I (itor CMC MI i
l«uml 4«t*«t*4 (MI
|Wfi jpMtf UM> VO
.UM Ho- fl^Ti/J
C««c«atr*iT«i
ntraii
M^lWUM t» bllMt«4 MM ••. •(
I lltar CMCMtrciia flWrt
•i flUj
111-1
111-0
*
111-1. llf.lUIX
111-0. tif.ll.t.4
1 0.01 1.1 i !•"•
H
1 1.4 I 10 •
O.t-l.l O.IU 11
O.IU
1.0
O.I |
I.I 1 !• •
0.1-1.0 o.m 4
O.lftl
0.0)
O.I
< 0.)
O.I
O.I
0.1
> < O.I
0.0) |
o-i. i o.iu a ]
0.24* . 0
1 4 1
0-1 ' J
0
o^.t o.m 4 i
0.2U - 0 C
» • .^tltoli (UO)| C • OfTMIIIil
-------
TABLE 45. RESULTS OF THE PHASE CONTRAST ANALYSIS OF AIR SAMPLES
COLLECTED AT THREE VERMICULITE SITES
Sample
Libby, Grace
106 Field blank3
133 Field blank3
131 Front loader
148 Pit haul driver
138 Mine analyst
141 Bottom operator
130 No. 2 operator
139 Dozer operator
101 Shuttle truck
104 Screening plant, DW
111 Screening plant, DW
108 Trailer court
136 No. 5 substation
South Carolina, Grace
312 Field blank3
346 Field blank3
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 blank3
533 Field blank3
508 Payload operator
520 Plant foreman
542 Bagger/forklift
513 (NE) downwind
506 Control off-site
515 (SE) crosswind
528 (SW) upwind
Sample
vol. (£)
—
-
303
297
294
276
285
270
385
390
368
169
111
—
-
340
478
240
316
285
287
80
291
154
264
257
354
_
-
255
252
249
188
274
299
147
Fibers/cc
ORF
< 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
< 0.02
< 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
< 0.02
< 0.02
< 0.01
0.01
< 0.01
< 0.01
< 0.01
0.01
0.02
IITRI
0.04
0.05
0.04
0.01
1.9
0.4
9.7
0.2
0.2
0.5
0.02
NDB
0.02
0.04
0.02
0.03
°'g
NDB
0.1
0.05
0.04
NDB
0.02
0.02
0.01
0.3
0.14
< 0.01
0.02
0.04
0.3
0.1
NDv
NDb
0.01
NDB
a Vclues for blanks were calculated assuming a 100-liter sample.
b ND: No fibers detected (100 grids).
68
-------
REFERENCES
Spumy, K. R., W. 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).
Chatfield, E. J., and G. M. Lewis, "Development and Application of an
Analytical Technique for Measurement of Asbestos Fibers in Venniculite,"
In: Scanning Electron Microscopy/1980/I, SEM, Inc., AMF O'Hare
Chicago, Illinois, p. 329-340, 328.
Trimbrell, V., "The Inhalation of Fibrous Dusts," Annals of New York
Academy of Sciences, 132, 255-273 (1965).
Going, J., and J. Spigarelli, "Environmental Monitoring Near Industrial
Sites: Vinylidene Chloride," EPA-560/6-77-026 (1977).
Going, J., "Environmental Monitoring Near Industrial Sites: Acrylamide,"
EPA-560/6-78-001 (1978).
Going, J., P. Kuykendahl, S. Long, J. Onstot, and K. 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
AND RELATED SAMPLES FOR TriJT EVALUATION OF FIBER CONTEMt
WITH EMPHASIS ON ASBESTIFORM FIBERS
I. Background
In December 1978, the venniculite industry submitted information to
the EPA regarding health problems experienced by employees who were processing
asbestos-contaminated venniculite. 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 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 is a hydrated magnesium-iron-aluminum silicate which has
been mined in the United States since 1929. After mining, vermiculite 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 as a contaminant in processed
vermiculite.
Although vennicuiite may contain fibrous materials, the health ef-
fects from vermiculite itself are unknown at this time. A priority review of
asbestos-contaminated vermiculite, completed by the Office of Testing and Eval-
uation in June 1980, suggested that the asbestos in venniculite 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
vermiculite could be initiated.
Several projects have been initiated to fulfill the information gaps
nnd complete the preregulntory analysis on venniculite, A control options
analysis has been initialed to determine regulatory strategy to control asbes-
tiform mineral-contaminated vermiculite, and a substitute analysis is in prep-
aration to evaluate replacements for venniculite products. Work has also been
initiated on a materials balance to show the mass flow of vermiculite along
with the release of any associated asbestiform mineral, and the development of
a mineralogy profile with 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 vermiculite 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 venniculite
scroll.
A-l
-------
will provide the needed information on the risk to the population exposed to
asbestifonn minerals from vertniculite 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 asbestiforo fibers present in and associated
with vermiculite ore, ore concentrates, and beneficiated vermiculite from the
four major vermiculite mines in the United States and of beneficiated vermicu-
lite 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
by Exposure Evaluation Division (EED) to include all the major sources of
verraiculite.
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 Materials (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 (ANSD/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 will 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 will 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 arc representative of the operations or sites. It is an-
ticipated that properties of the materials of similar types will vary with
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 site. It is likely that a historical sample collec-
tion is maintained (by the mine company) from several operation points within
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, with a de-
cision made by an experienced sampler depending upon increment availability
and proper sampling procedures. All decisions will be documented with copies
sent to the EPA task manager.
a. Raw ore and ore concentrates (Phase I)
(1) A bulk sample of the raw vermiculite ore represent-
ing different parts of the mine.
A-3
-------
(2) A bulk sample of the concentrated ore before bene-
ficiation.
(3) A bulk sample of dust from the dust collection
equipment where such equipaent exists.
(A) A water sample from washings and dust control opera-
tions.
(5) A bulk sample of concentrated ore blend (beneficia-
tion feed) before beneficiation.
b. Beneficiated vermiculite (Phase I)
(1) A bulk sample of each of five grades of beneficiated
vermiculite.
(2) A bulk sample of material from one to three inter-
mediate beneficiation processing steps.
(3) A bulk sample of tailing from the beneficiation
process.
(4) A bulk sample of dust from dust collection equip-
ment.
(5) A water sample from washing and dust control opera-
tions.
c. Exfoliated plant samples (Phase II)
(1) A bulk sample of each of the five grades of vermicu-
lite before exfoliation.
(2) A bulk sample of each of five grades of exfoliated
vermiculite.
(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 method 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 0.4 urn Nuclepore* filters when possible, but allows for the use of
cellulose acetate (Milliporeti) filters.
There are advantages and disadvantages to the use of either Nucle-
poreg or Millipore® filters. The sample collected on the smooth polycarbonate
A-4
-------
(Nucleporeft) suface has poor retention efficiency and the sample may be lost
or redistributed during transport. The cellulose acetate (Millipore®) filters
requires an ashing procedure and reconstitution on a polycarbonate filter for
TEH analysis. Ashing and reconstitution is an extra step 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 some lots of
both Nucleporeg and MilliporeQ filters.
The current consensus of leaders in the field is to favor the use
of Millipore€> 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
will 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 100
filters, one from each box of 25, will be selected by random numbers as
filter blanks. The four filters will be quartered and one-fourth of each
combined as a composite sample.
b. Field blanks - One of every 10 filters will be a field
blank, subjected to all processing conducted with an actual air sample except
for the sampling itself.
3. Sampling procedure - All sampling, fixed and personal, will be
taken using 37 mm 0.45 (Jm Millipore§ filters backed with 5 (Jtm Millipore® filters
and a Millipore® support pad. The sampling rate will be approximately 2 liter/
min with the exact rate determined periodically throughout the sampling period
by the use of calibrated flow meters.
Sampling will be scheduled for 8 hr (or longer for ambient and back-
ground samples). However, if the flow is found to reduce during sampling,
indicating that the filter is loaded, sampling will be stopped and the time
and flow rate recorded.
4. Air samples will be collected - Personal air zone samples and
fixed samples located at targeted areas will be taken. The use of personal
samples will depend on the individual work patterns and on the cooperation of
the host company. When possible, individuals at each work station will be
equipped with personal samplers.
When possible, air samples will be taken which correspond to the
bulk samples (Section III-A-2). The following air samples will be taken.
Any variation from these sampling locations will be documented.
a. Air samples taken at operations before beneficiation
(Phase I).
A-5
-------
geographic region.
(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 downwind of dusty operations.
(5) Air samples along shipping lanes.
(6) An air sample as a background control within 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 work stations in
the beneficiation operation.
(2) Air samples around the beneficiation plant.
(3) Air samples at grading (screening) operation:;.
(4) Air samples downwind from dusty operations.
(5) Air samples along shipping lanes.
(6) An air sample as a background control within the
geographic region.
c. Air samples taken from the exfoliation operation. (Phase
ID
(1) An air sample from each of selected work stations in
the exfoliation operations.
(2) Air samples around the exfoliation plant.
(3) Air samples downwind 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
A-6
-------
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 vermiculite 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 example 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 vermiculite, 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 (PLM, XRD, etc.).
d. To isolate the fine fibers from vermiculite, 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
-------
Tall Form
Sample Grams IPA Beaker (ml)
Grades 1 and 2 40 400 1,000
Grades 3 and 4 20 200 400
Grade 5a 40 400 1,000
Unbeneficiated material and 40 400 1,000
other tailings, etc.
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., PLM) and to prepare a series of Nucleporeti filters for EM analysis.
Serial dilution may be required to obtain optimum filter loading.
e. The NucleporeQ filter with suspended fines will be used to
prepare a TEH grid by the EPA carbon-coated Nuclepore® filter technique.
f. Make fiber count - Determine chrysotile or amphiboles.
Count 100 fibers or 10 grid of 200 mesh screen. Determine the limit of de-
tection and count more grids if necessary.
g. Identify specific amphiboles using selected area electron
diffraction or zone axis selected area electron diffraction plus energy dis-
persive X-ray analysis.
h. Exfoliate a portion of the beneficiated vermiculite by
sprinkling no more than a one particle thick layer of sample, in a preheated
(800°C) shallow container and place the container back into a 800°C oven for
5 sec. Examine the exfoliated sample as described in No. 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
-------
Starting with a sample quantity depending on particle size, or grade,
Vermiculite Sample Water
Grade Weight (g) Voluae (ml)
1 and 2 40 2,000
3 and 4 20 1,000
5 40 2,000
and proceed as follows:
a. Float separation - Place the expanded veraiculite in a
2,000 ml plastic beaker and add water. Stir for 30 sec and skim off the
veraiculite and drain on a 50-mesh screen. Collect the drain water and re-
turn it to the beaker. Discard the veraiculite.2
b. Disperse the "sink" 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.
c. Examine preparation by PLM as a preliminary parallel exam-
ination.3
d. Prepare TEM grid by EPA carbon-coated Nuclepore® filter
technique.
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 PLM for the identification of gross
quantities of asbestiform fibers. If gross quantitier of fibers are identi-
fied, the quantities will be estimated and the analysis terminated.
7Selected samples will be examined to verify the absence of asbestos in
this fraction.
3 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 sample is not adequately characterized by PLM,
a portion of the sample will be dispersed in water and filtered for EM analy-
sis.
b. Wash water samples - The solids present will be dispersed
in the water and aliquots filtered for appropriate optical and EM analysis.
B. Air Samples
;
1. Portions of selected filters will be used to determine fiber
count by the standard NIOSH procedure using phase nicroicopy.
2. The major analysis of the air samples will 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
-------
PROCEDURES
BULK SAMPLE ANALYSIS
Quantitative analysis requires that valid and rigorous procedures be
used during all phases and steps of a procedure and that these procedures be
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 MRI,
• Discussions of the protocol with MRI, 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
work 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
- 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 Asbestiform Fibers (Revised November 13. 1980).
B-l
-------
Saapl* lUcalpc
•nd Lo((ln|
FUt
MffU
(Utatn B«l«ac«)
Prtllminiry
FUt
1PM
THOMAL
Hold for
Preliminary
FLH
IPA Sink
to Obtlln Finn
by Plp«tt«
Dllutt.
Fllttr.
Dry
To EM Protocol
Figure B-l. Flow chart for bulk saarple 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 extract a sample from the flowing powder streaa. This time-
averaged sample consisting of many small aliquott 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 beneficiation, thermal exfoliation, and a back-up sample. The target
size for each sample was:
• Twenty grams for vermiculite 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 asbestifortn 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-)quantitative
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
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 were made at this tine.
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 were obtained by coning and quarter-
ing* the sample fraction submitted for PLM analysis. The entire PLM sample
fraction was 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) was retained on the foil and the remainder of the PLM sample
was returned to its container. For those samples that received a duplicate
separation analysis, the coning and quartering was repeated.
Hand-Picking--
When the preliminary inspection revealed the presence of several bundles
of fibers at least 1 mm in diameter and 3 ran in length, hand-picking of the
fibrous phase(s) with a fine-pointed tweezers was the first separation step
performed. The separation subsample of the PLM sample was weighed on a piece
of tared foil and then spread to a monolayer of particles. While viewing
through the stereomicroscope, the fibers were tweezed from the subsample and
placed in a tared weighing pan. Mixed particle types containing at least 25%
fibrous material, as well as totally fibrous particles were tweezed. The pan
containing the fibrous phase as well as.the foil containing the nonfibrous
remainder of the subsample were then reweighed (to 0.1 nig) and the mass per-
cent of "pickable" fibrous material was calculated.
When smaller bundles were present, hand-picking was done either after
the first heavy liquid separation step or, for some samples, was 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 araphiboles and pyroxenes,
were 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, some of the amphibole minerals, and other
mineral contaminants commonly associated with vermiculite.
See Reference 2, page B-3.
B-4
-------
TABLE B-l. SPECIFIC GRAVITIES OF SELECTED MINERALS
Mineral
Chemical formula
Specific
gravity
Veroiculite
Biotite
Chrysotile
Serpentine
Talc
Anthophyllite
Actinolite
Tremolite
Ferroactinolite
Cummingtonite
Grunerite
Diopside
Hornblende
Quartz
Olivine
(Mg,Ca)0 3(Mg,Fe,Al)3 0(Al,Si)40,0(OH)4
K(Mg,Fe)3(AlSi3010)(OH)2
Mg3Si205(OH)4
Mg3Si2Os(OH)4
Mg3Si4010(OH)£
(Mg,Fe)7Si8022(OH)2
Ca2(Mg,Fe)5Si8022(OH)2
Ca2Mg5Si8022(OH)2
Ca2FesSi8022(OH)2
(Mg,Fe)7Si8022(OH)2
Fe7Si8022(OH)2
CaMgSi206
(Ca,Na)2 3(Mg,Fe,Al)5Si6(Si,Al)2022(OH2)
Si02
(Mg,Fe)2Si04
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
The density separation technique is based on the principal that parti-
cles with a greater density than the liquid they are suspended in will sink
while particles with densities equal to or less than the liquid density will
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 will improve
if grinding or sieving is performed prior to using the float-sink procedure.
While this would have been true for the vermiculite 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 (THE), specific gravity of 2.97; the
THE "sinks" were the first density separation fraction recovered. The TBE
"floats" were then suspended in a TBE/isopropanol 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 90°C; 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-nl 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 hi 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 \im pore size, MilliporeB membrane filter. The re-
covered mineral fractions were dried for 24 hr at 90°C 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 sample are based on the summed masses of recovered materials, rather
than on the subsanple 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 TBE 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 steps 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_ = 1.515) on a glass slide for PLM
analysis. Mineral phases present were identified, at least as to mineral
group, by morphological properties and by observation of optical properties;
including extinction angles, birefringence, refractive indices, color and
pleochroism. Individual particle types 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. Mass concentration ranges for individual mineral phases were described
as:
Primary > 25%
Major 5 to 25%
Minor 0.5 to 5%
Trace < 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-separated 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 sunning the
concentration determined for each density-separated fraction.
X-Ray Diffraction
Positive identification of mineral phases, particularly fibrous amphi-
boles, 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 lOfl milliamps. The CuK
x-ray lines, with an averaged wavelength of 1.54184 A, were thus generated.
Most 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:
• Beneficiation of vermiculite samples,
• Sample preparation, and
• Electron microscopy analysis.
3Electron 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.
B-7
-------
Beneficiation
Two beneficiation procedures were used for the saaplei; 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 less
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 tines 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 Nuclepore® 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 800°C 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 800°C oven for 5 sec. After the 5 sec ex-
pire, the dish is removed, permitted to cool and the exfoliated vermiculite
is stored 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 varmiculite 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
5Midwest Research Institute sample numbers.
B-8
-------
prepare filters is identical to that used in the isopropanol beneficiation,
except that filtered, distilled water is substituted for isopropanol.
Sample Preparation
The basic procedures IITRI used to analyze the samples are documented in
Electron Microscope Measurement of Airborne Asbestos Concentrations: A Pro-
visional Methodology Manual;3 however, four samples were analyzed using a mod-
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 tne 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 Nucleporeti 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 JEOL 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°.
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 fron 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 conparison 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 a« an
aid in the classification, particularly when nonasbestos phases capable of
providing false positive fiber identification are known (by PLM, XRD, or fron
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 TEM 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 (OKF).
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
>/K UppCr Umit fixcd by total P«rt"le loading) of, ideally, 10 as-
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
tiDers when the ORF procedure is followed; thus, the IITRI 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, verniculite 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 sample was prepared. It is computed
using Equation 1:
VA
where CN = Concentration, fibers per gram
Nf i = Number 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
Wg = Weight of sample, g
VT = Total volume of bonification fluid
VA = Ali(iuot volume used for filter preparation.
IITRI and ORF used similar but different grids. Those used by IITRI 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 openine
that was counted. r«=«j.u8
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 /»
is computed by substituting the total fiber mass estimate into Equation 1 8
B-ll
-------
n
Mf = Z ? x d2 x S. x e x 108 (2)
i=l
where: Mf = Estimated total fiber mass, f
d2 = Fiber diameter (projected width), \m
I = Fiber length, pm
e = Mineral density, g/cm3: chrysotile = 2.6 g/cm3
amphibole =3.0 g/cm3
Fiber detection limits for any sample can also be computed using Equation 1.
To do this, the value of 1 is substituted for Nf, equivalent to assuming 1
fiber is detected. All detection limits reported herein are in fibers per
gram.
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 by 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 Methodology, described in EPA-
600/2-77-178, revised June 1978. The NIOSH procedure is sunmarized 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 somple pro-
cedures.
Sample Preparation
Samples to be submitted for phase contrast enumeration are collected on
cellulose acetate (MilliporeS 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.5 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.
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 mininua of 20 areas and a
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—I fiber if length > 5 \tm
' Fiber has one end in counting area
Count—1/2 fiber if length > 5 pm
• 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 I 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:
F = Vf (3)
acVs
where: FC = Airborne fiber concentration, fibers/cc
f = Number of fibers enumerated
a = Total counting area, cm2
Af = Area of filter used for sample collection, cm2
V = Volume of air sampled, cc
s
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-nl 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 sanples
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.
C-l
-------
TABLE C-l. PERSONAL SAMPLER CALIBRATIONS (EACH VALUE REPRESENTS A MINIMUM OF THREE DETERMINATIONS)
Grace, Libby Montana
Sampler
ID
186172
186173
186174
186175
1-7328
o 2-7334
6-7329
7-7317
Pr?T a
sampling
2.13
2.12
2.09
2.12
2.07
2.16
2.12
2.19
Post-
sampling
1.91
1.92
1.96
1.93
1.89
2.13
1.99
1.99
Grace Mill
Pre-
sampling
2.02
1.96
1.99
1.98
2.02
2.00
2.02
2.01
Post-
sampling
2.07
1.94
2.00
2.01
2.04
2.04
2.07
2.00
Avga
2.04
1.95
2.00
2.00
2.03
2.02
2.04
2.01
Enoree, North Carolina
Grace Mine Patterson
Pre- Post- a Pre- Post-
sampling sampling Avg sampling sampling Avg
1.94 1.99 1.97
1.98 1.92 1.95 1.94 1.93 1.93
1.95 2.00 1.98 1.95 1.94 1.94
1.93 1.93 1.93
a Values used to calculate sample volume.
-------
TABLE C-2. DATA FOR VOLUME CALCULATIONS FOR PERSONAL SAMPLES
Sample
no.
101
121
125
126
128
129
130
131
135
138
139
141
146
148
300
301
304(S)a
305
306
308
310
314
315
320
321
322
324
330
322(S)a
336
337
339
340
34l(S)a
347
349
Sampler
ID
1-7328
186173
7-7317
186175
2-7334
186172
7-7317
186174
186174
186172
2-7334
6-7329
6-7329
186175
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
186175
186172
6-7329
2-7334
186175
1-7328
Flow
(£/nin)
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
TiM
stapled
(•in)
186
429
264
278
266
270
130
145
283
138
125
130
261
140
175
121
438
141
362
203
132
146
127
280
238
241
31
141
202
205
246
302
181
79
157
251
Voluae
W
385
909
578
589
575
575
285
303
591
294
270
276
555
297
354
240
876
288
717
408
257
285
250
546
478
470
61
285
408
418
492
616
369
160
314
507
(continued)
C-3
-------
TABLE C-2 continued
Tine
Sample Sampler Flow sampled Volume
no. ID (£/min) (min) (£)
504 2-7334 1.93 185 357
508 7-7317 1.93 132 255
511 6-7329 1.94 194 376
516 7-7317 1.93 188 363
517 2-7334 1.93 192 371
519 7-7371 1.93 177 342
520 6-7329 1.94 130 252
521 6-7329 1.94 184 357
542 2-7334 1.93 129 249
a 304(S), 332(S), and 34l(S) were stationary samples.
C-4
-------
TABLE C-3. VOLUME CALCULATIONS FOR STATIONARY SAMPLES
Sample
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
Average
flow
(A/rain)
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
1.85
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
Time
sampled
(•in)
270
182
178
109
146
177
301
143
321
174
128
299
179
166
343
362
404
114
386
387
369
321
222
321
329
314
320
100
121
387
128
124
94
136
327
328
84
122
331
332
352
322
324
Voluae
(4)
261
189
390
207
169
368
616
199
253
156
128
658
422
244
302
348
784
111
421
545
420
239
291
334
595
582
378
154
287
823
195
134
80
264
516
774
88
216
288
420
511
632
515
(continued)
C-5
-------
TABLE C-3 continued
Average Time
Sample flow sanpled Volume
no- U/min) («in) (£)
502 1.67 351 585
503 1.46 144 210
506 2.23 123 274
513 1.30 145 188
515 1.70 175 299
518 1.76 342 601
523 1.72 348 599
525 2.19 342 748
527 2.13 345 735
528 1.02 144 147
531 2.08 144 300
540 2.04 343 701
C-6
-------
APPENDIX D
INCREMENT BULK SAMPLES COLLECTED AND COMPOSITED
-------
Most 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
oanple. 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 conposite.
The two-eightho were combined with the two-fourths so that the composite was
divided into 1/4, 3/8, and 3/8 of the original. The "1/4" was retained at
MRI: one "3/8" was sent to IITRI; 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 fron the three collection lo-
cations that were processed into compositeo. Table D-4 lists the sacples that
were not processed.
D-l
-------
TABLE D-l. INCREMENT AND COMPOSITE SAMPLES FROM LIBBY, MT, GRACE
Sample
description
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
Date
collected
10/7/80
10/8/80
10/9/80
10/10/80
10/13/80
10/14/80
10/15/80
10/16/80
10/17/80
10/21/80
10/23/80
Composite
10/7/80
10/8/80
10/9/80
10/10/80
10/13/80
10/14/80
10/15/80
10/16/80
10/17/80
10/21/80
10/23/80
Composite
10/7/80
10/8/80
10/9/80
10/10/80
10/13/80
10/14/80
10/15/80
10/16/80
10/17/80
10/21/80
10/23/80
Composite
Sample
ID
151
152
153
154
155
156
157
158
159
160
161
270
162
163
164
165
166
167
168
169
170
171
172
276
173
174
175
176
177
178
179
180
181
182
183
259
Sample
weight
(8)
1,030
1,002
1,002
1,000
1,001
999
1,002
1,009
1,000
798
991
1,001
1,006
1,004
1,004
1,004
1,002
956
1,002
1,002
873
992
999
1,002
1,002
1,002
1,003
997
1,003
1,000
997
1,000
964
Weight
for composite
(g)
126
125
126
125
125
125
125
125
125
125
125
125
125
125
125
125
125
125
125
125
125
125
133
113
121
122
127
124
129
124
124
124
125
(continued)
D-2
-------
TABLE D-l continued
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 feed8
Head feed8
Head feed"
Head feed3
Head feed*
Head feed8
Head feed"
Head feed*
Head feed8
Head feed8
Date
collected
10/7/80
10/8/80
10/9/80
10/10/80
10/13/80
10/14/80
10/15/80
10/16/80
10/17/80
10/21/80
10/23/80
Composite
10/7/80
10/8/80
10/9/80
10/10/80
10/13/80
10/14/80
10/15/80
10/16/80
10/71/80
10/21/80
10/23/80
Composite
10/8/80
10/9/80
10/10/80
10/13/80
10/14/80
10/15/80
10/16/80
10/17/80
10/23/80
Composite
Sample
ID
184
185
186
187
188
189
190
191
192
193
194
282
195
196
197
198
199
200
201
202
203
204
205
264
223
224
225
226
227
228
229
230
231
291
Staple
weight
(g)
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 coaposite
(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
-------
TABLE D-l continued
Sample
description
Extractor
Extractor
Extractor
Extrictor
Extractor
Extractor
Extractor
Extractor
Extractor
Dust, screening plant
Dust, screening plant
Dust, screening plant
Dust, screening plant
Dust, screening plant
Dust, screening plant
Dust, screening plant
Dust, screening plant
Dust, mill baghouse
Dust, mill baghouse
Dust, mill baghouse
Dust, mill baghouse
Dust, mill baghouse
Dust, mill baghouse
Dust, mill baghouse
Dust, mill baghouse
Dust, mill baghouse
Date
collected
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
10/8/80
10/9/80
10/10/80
10/13/80
10/14/80
10/15/80
10/17/80
Composite
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
Staple
ID
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
Staple
weight
(8)
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 composite
(1)
125
125
125
125
125
125
125
125
125
125
125
125
125
125
125
125
125
125
125
125
125
125
125
All "head feed" samples contained rocks too large to pass through the
channels in the riffle box. These were separated out, weighed sep-
arately, and retained at MRI.
-------
TABLE D-2. INCREMENT AND COMPOSITE SAMPLES FROM EKOREE. SC. GRACE
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 S
Grade 5
Mill feed + 100 mesh
Mill feed + 100 mesh
Mill feed + 100 mesh
Mill feed + 100 mesh
Mill feed + 100 mesh
Mill feed + 100 mesh
Mill feed + 100 mesh
Mill feed + 100 mesh
Grade 3 exfoliated
Analysis ID
Grade 4 exfoliated
Analysis ID
Date
collected
10/27/80
10/28/80
10/29/80
10/30/80
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
10/27/80
10/28/80
10/29/80
10/30/80
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
Simple
weight
(8)
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
958
836
792
203
260
Weight
for composite
(8)
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 AND COMPOSITE SAMPLES FROM EMOREE. SC. PATTERSON
Sample
description
Ungraded
Ungraded
Ungraded
Ungraded
Ungraded
Date3
collected
11/6/80
11/6/80
11/6/80
11/6/80
Composite
Sample
ID
567 '
568
569
570
573
Sample
weight
(g)
1,770
1,700
1,186
1,780
Weight
for composite
(8)
250
250
250
250
a Increments taken at 2 hr intervals.
D-6
-------
TABLE D-4. BULK SAMPLES THAT WERE COLLECTED BUT NOT
SUBMITTED FOR ANALYSIS
Staple description Number 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, wet scrubber discharge 1
Grace, SC, composite total tails 1
Grace, SC, Lanford mine composite 1
Grace, SC, Foster mine composite 1
Grace, SC, No. 4 concrete aggregate, stabilized 1
Grace, SC, No. 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 waste pile 4
Patterson, SC, preexfoliated waste 4
Patterson, SC, bagged product 4
D-7
-------
APPENDIX E
PHOTOMICROGRAPHS AND TEM MICROGRAPHS OF SELECTED 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 samples
to illustrate the types of particles observed in samples of vermicu-
lite 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 diopside and
augite.
259-1-2.76 float; 84X. Large flakey particles are vermlculite with white
stress lines. Arrow points to a flake of talc intergrown with tremolite-
actinolite fibers.
E-2
-------
259-I-TBS-Flbers (ground); 208X. Large bundle in the center Is unquestionably
fibrous tremolite-actinolite. Note the (white) interference colors 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-lanellated prism morphology of
tremolite-actinolite was found within hand-picked fiber bundles.
E-3
-------
259-1-2.76 sink; 208X. Calcite (bright white) growing in a pseudoraorphically
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 lanellated quartz. The
flakey, lamellated morphology was caused by the quartz forming inbetween
vermiculite plates.
E-A
-------
264-I-TBS; 208X. Arrows point to tremolite-actlnolite In two different crys-
tal habits—the parallel, lamellated prisms and truly fibrous material. The
mottled coloring of the fiber bundle is due to the various angles at which
the individual fibers are intergrowing with each other.
264-I-TBS; 208X. The circle outlines one tremolite-actinolite particle
which contains 2 different crystal habits—the parallel, latnellated 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 fibroua tremolite-
actinolite with inclusions(arrows) of prismatic tremolite-actinolite. The
large flakey white particle (arrow) is a sheet of quartz adhering to vermi-
culite and was apparently growing inbetween vermiculite plates.
291-I{ 84X. The friability of the tremolite-actinolite 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 (minimal grinding); 208X. The tremolite-actinolite, hornblende and
anthophyllite ampttibole fragments are mostly Irregular to chunky prisms.
Some slender prisms 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 vermiculite and talc.
E-7
-------
430-1-2.76 float (ground); 208x. Although optical and morphological proper-
ties of the mottled phase depicted were consistent with serpentine minerals
XRD data ruled out serpentine as its identity. Mixed layer vermiculite-
hydrobiotite was identified in the XRD work.
427-I-TBS; 84X. Several large lamellated, parallel prisms of tremolite-
actinolite and anthophyllite are visible.
E-8
-------
427-I-TBS (ground); 208X. Crushing of the large amphibole prisms produced
prisms with splinter fragments that could morphologically be defined as
fibers.
427-I-TBS-Flbers (ground); 208X. Crushing of coarse "fibers" produced mostly
prismatic material. Arrows point to the numerous vermlculite 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 anthophyllite. Numerous flakes of
talc intergrown with anthophyllite were found in this fraction.
A36-1; 82X. The amphiboles are mostly chunky and prismatic in this sample,
Arrow points to tremollte composed of rather thick, lamellated prisms.
E-10
-------
436-I-Flbers (ground); 208X. Truly fibrous anphlbole an well as very chunky,
prismatic particles were produced when particles nacroscopically classifiable
as fiber bundles were ground.
436-I-TBS-green, glassy (ground); 208X. This phase was a mixture of horn-
blende and tretnolite-actinolite. The amphiboles' morphologies were
predominantly chunky prlsma.
E-ll
-------
436-I-TBS-milky green (ground); 208X. Mineral grains composed of this tretno-
lite-hornblend mixture were irregular in shape 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-verraiculite phases were mostly chunky, prismatic amphi-
bolea, iron oxides (black) and fluorapatite (rounded, gray particles).
573-I-TBS (ground); 208X. Grinding did result in fracture of some amphibole
grains into elongated fragments morphologically classifiable as fibers.
Most of the amphibole 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-1A
-------
276-I-TBS-dark, green glassy (ground); 208X. The pyroxene, diopslde, exhibits
some concholdal as well as prismatic fracture patterns.
'
294-I-TBS; 82X. Small amphibole fiber bundles and large amphlbole prisms are
present. Note the layered crystal growth of the fractured tremolite-
actinollte fragment (arrow).
E-15
-------
288-1; 52X. The mottled coloring of this large trenolite-actinolite fiber
bundle is due to the non-parallel, radiating growith pattern of the indivi-
dual fibers in the bundle.
•
288-1 (fines); 208X. Much abrasion and disintegration of large fiber bundles
has obviously occurred in the processing of the vermicullte, as indicated
by the numerous very fine fibers present.
E-16
-------
297-1; 208X. The fine, single fibers here were abraded from larger trenolite-
actinolite fiber bundles.
E-17
-------
' 1.0 Mm
Transmission electron micrograph showing
typical particulate matter found in vater
suspension after laboratory exfoliation
of Sample 264-0, Grace, Libby, Montana,
Grade S (composite).
E-18
-------
pm
Transmission electron micrograph showing
typical partlculate 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
-------
,r-v >rT>v •..•
'V--.- -•,>:..,• ••^•V
- •' "^ * • •
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-21
-------
1.0 um
Transmission electron micrograph showing
typical partlculate natter found In water
suspension after laboratory exfoliation
of Sample 427-0, Crace, Enoree, South
Carolina, Grade 5 (composite).
E-22
-------
iprs— r . »
, ,,,'•-,, A • /°, * ^ j= * «, <-
. 1.0 um
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
-------
•: ; ". .;.<'• •'''^/£'f.v*S& •'.•:•
'••"v»•>-.v-v- ;j^«K£;
.0 urn
Transmission electron micrograph showing
typical participate matter found In water
suspension after laboratory exfoliation
of Sample 573-0, Patterson, Enoree, South
Carolina, Ungraded, Dried Ore (composite).
E-24
-------
:*> v
* % •
• —- »•
• ••-•^ ;jfc
• v..
•. *
" V*- '
>5-
•1:
1 ". /V
',.•>»
:•>>
1.0 um
Transmission electron micrograph showing
typical particulate matter found in water
suspension after laboratory exfoliation
of Sample 573-0, Patterson, Enoree, South
Carolina, Ungraded, Dried Ore (composite).
E-25
------- |