xvEPA
United States
Environmental Protection
Agency
Environmental Sciences Researe*fc,
Laboratory '^4,
Research Triangle Park NC 2771 l£r
f i
Research and Development
EPA-600/S3-82-080 Dec. 1982
Project Summary
Method for Measuring
Carbon Fiber Emissions
from Stationary Sources
W. M. Henry, C. M. Melton, and E. W. Schmidt
Carbon fibers are small, highly
conductive, and lightweight. When
released as emissions from produc-
tion, manufacturing, processing and
disposal sources, they are readily
airborne and can be dispersed over
wide areas. Because of their high
electrical conductivity, carbon fibers
settling out of the atmosphere onto
electronic or electrical components
can cause malfunctions and damage.
This research program was initiated to
develop techniques to identify and
measure carbon fiber concentrations
emitted from manufacturing,
processing, fabricating and incinerat-
ing sources. The overall objective of
the research was to develop a
measurement method to meet
possible emission source regulations.
Because fiber count, fiber size range,
and total fiber mass concentration
were required, a method based on
light microscopy was selected for test
and development.
Experimental work, carried out on
laboratory and field site samples,
showed that carbon fibers could be
readily recognized and counted
without extensive separation from
other particulate "debris" collected
during source sampling. Although
fibers were reduced in diameter during
oxidative and thermal processes such
as incineration, the size reductions
were not beyond the resolution power
of light microscopy.
The method was tested on samples
collected from various stationary
emission sources and on samples
collected from a pilot-scale in-house
incineration facility and the results of
these emission measurements are
reported.
This Project Summary was developed
by EPA's Environmental Sciences
Research Laboratory, Research
Triangle Park, NC, to announce key
findings of the research project that is
fully documented in a separate report
of the same title (see Project Report
ordering information at back).
Introduction
The rapidly developing use of carbon
or graphite* fibers incorporated into
resin matrix composites is a potential
environmental problem. Present
generation fibers are small in diameter
(typically 5 to 15 ^im), and very light-
weight (1.5 to 2 g/crn3); they are
chemically inert and are excellent elec-
trical conductors. Therefore, when the
fiber filaments are separated from their
composite forms, they are readily
airborne and can be transported long
distances by normal atmospheric
motion. The principal environmental
concern is the potential of released
fibers to settle out of the atmosphere
onto electrical and electronic
*The term "carbon fibers" is used in this report to
include both carbon and graphite fibers resulting
from pyrolysis of fibrous materials which have
been heat treated to temperatures higher than
the decomposition temperature of the precursor
polymer (typically 1,000 to 1,200°C for carbon
fibers and above 2,000°C for graphite fibers)
-------
components, causing short circuits and
arcing and damagirj or destroying
equipment. Manufacturers, processors,
and fabricators of carbon fibers and
composites are aware of these potential
hazards and take measures to minimize
fiber emissions. Despite these control
measures, fibers are released during
fiber manufacture, composite formation
machining and grinding, and other
fabrication and processing operations.
Additional fiber contamination hazards
can occur during waste disposal of
consumer goods by incineration, which
can thermally decompose the matrix
material and release fibers into the
atmosphere via stack emissions.
While current production and use of
carbon fiber composites (estimated as
less than 500 T in 1980) may present
only a limited hazard, large increases in
demand are foreseen for these
lightweight, high-strength materials,
especially in the automotive and aircraft
industries to achieve fuel savings via
weight reductions. For this reason, the
Federal government has instituted
various programs to study all aspects of
carbon fiber production, usage, and
disposal. The research project described
herein is one of several initiated by the
U.S. Environmental Protection Agency
(EPA). The risk potential for carbon
fibers released to the atmosphere is
relatable to the size range, number, and
mass of fibers emitted. In this study,
fibers with a length-to-diameter ratio of
5:1 were of primary concern, although
the research showed that particles of
lower length-to-diameter ratios could
be identified and counted readily.
Conclusions and
Recommendations
Studies of available analytical tech-
niques for measuring carbon fiber emis-
sions from stationary sources indicate
that light microscopy will provide the
necessary fiber mass, number, and size
range data.
Since extraneous particulate matter
collected with carbon fibers does not
interfere with the carbon fiber
identification measurements, rigorous
separation procedures are not required.
Still, separating the carbon fibers from
other particulate matter might reduce
the time required for analysis and
improve the accuracy and precision of
the method; for that reason, the use of
cyclones preceding a filter backup
sampling system warrants further
investigation.
Although sampling methodology was
not explored in depth, the use of
membrane filters as a collection
medium may be warranted. The
membrane filter can be readily removed
from the collected particulate prior to
microscopy with the use of simple nitric
acid digestion treatment. For sampling
at incineration sites, where a
membrane filter is not thermally stable,
the use of cyclones preceding a quartz
or glass filter results in most of the
carbon fiber emissions being collected
in the cyclones and in the probe portion
of the sampling systems, thus
simplifying the measurement
procedure.
If necessary to support regulatory
actions on carbon fiber emissions, the
methodology is compatible with the
personnel skills and equipment
available at most Federal and local
regulatory and industrial testing
laboratories. As with any newanalytical
technique based on microscopy (even
light microscopy), the skill and
experience of the operator is paramount
in obtaining reliable data. To establish
precision and accuracy data, the
method should be evaluated by
intercomparisons among several
laboratories.
The field site and incinerator
emission data in the final report are
among the first such data available, and
should be useful in guiding control
practices and risk evaluations of carbon
fiber emissions.
Procedure and Results
The field sampling procedure involved
drawing the fiber-laden air through
high-efficiency filters with a vacuum
pump. A cellulose acetate (Millipore)
filter (pore size = 5.0 urn; diameter = 47
mm) was used.
Light microscopic analyses were
performed over a selected radial filter
area; typically, one-eighth of the47-mm
filter area was analyzed. All carbon
fibers observed in the radial section
being analyzed were counted and
individual lengths and widths were
measured. These data were combined
with the known sampling times and
flow rates to calculate fiber number
concentrations. The corresponding
mass concentrations were calculated as
follows: mass concentration = (number
concentration) • (average fiber volume
[determined by length and width meas-
urements]) • (an assumed density of
1.85g/cm3).
Table 1 summarizesthedata obtained
from the various field samples. The
carbon fiber concentrations ranged
from 100 ng/m3 to 5500 ng/m3.
In all samples the fiber width was
relatively constant, but the fiber length
varied greatly from process-to-process
and was widely distributed within a
given sample.
For the field samples analyzed, no
treatment to remove extraneous
particulates was required, a result
which can be expected for most
emission samples from noncombustion
processes. Treating the specimen in
boiling HN03 (or another appropriate
treatment) would be necessary only
when the NIOSH clearing method does
not produce a specimen suitable for
light microscopy. The NIOSH clearing
method is a standard procedure for
asbestos analysis and involves
dissolving the Millipore filter in a 1:1
mixture of diethyl oxalate and dimethyl
phthalate.
Table 1 Carbon Fiber Analyses of Field Collections
Operation
Winding™
Prepregging ""
Shuttle Loom Weaving
Machining1^
Machining
Rapier Weaving™
Average
Length
(fjim)
69.5
213.1
749.4
30.9
32.8
706.0
Average
Width
(»m)
6.5
6.1
6.7
6.6
6.2
3.9
Number
Concentration
(fibers/m3)
46
24
88
823
1.212
340
Mass
Concentration
(ng/m3)
482
356
5,497
2.062
2,812
6,831
(a) A verage of four samples taken in work area.
"" Unless otherwise noted, all samples are of outside emissions.
lcl Work area sample.
ldl Outside ambient, downstream of baghouse. Some fibers appeared damaged.
-------
A stoker-boiler designed and
constructed to study combustion
processes was used to simulate
combustion conditions likely to occur in
a municipal incinerator. After reaching
a steady-state combustion rate of 75
Ib/h of refuse, 1 Ib of carbon fiber
composite was fed onto the burning
hearth over a 1-min period and the
combustion continued for 60 min.
Sampling employed two simultane-
ous modes of collection: a Source
Assessment Sampling System (SASS)
train using three cyclones preceding a
filter, and a Method 17 (in-stack) filter
sampling system. The Method 17 filter
was changed every 15 min in order to
provide information on the rate of
carbon fiber emissions with time. Care
was taken to maintain an isokinetic
sampling period.
The samples consisted of three types:
(1) bulk paniculate brushed from the
cyclone catches, (2) cyclone and probe
washings deposited on DM-450
(Gelman) filters, and (3) Method 17
samples and the SASS train collections
on quartz (2500 OAST) filters. For
analyses of the incineration samples,
the basic sample preparation consisted
of sonicating the weighed bulk sample
and/or filter in a measured volume
(~10 ml) of filtered deionized water for
15 min. Twenty-five milliliters of hot
(180°F) concentrated nitric acid were
added to destroy extraneous organic
materials in the sample that might
interfere with the microscopic analysis.
After cooling to room temperature
(25°C), the nitric acid mixture was
diluted to 1000 ml and redeposited in
aliquots of 100, 200, and 700 ml on SM
Millipore filters (pore size = 5.0 /urn;
diameter = 47 mm). When thoroughly
dry, each deposit-bearing filter was
transferred to a glass slide (2x3 in). The
sample was then cleared using the
standard NIOSH solution (dimethyl
phthalate and diethyl oxalate at 1:1 plus
0.5 g/ml Millipore filter) and covered
with either glass cover slips or slides.
Analysis was performed by screening
samples and original quartz filters at
250 to 400X magnification on a Leitz
Orthoplan microscope with transmission
objectives.
The analysis showed that fibers are
oxidized during incineration and that
the widths are reduced, resulting in
width variations. When they are not
oxidized (as in the aforementioned field
samples), fiber widths are quite
uniform. Oxidation of carbon fibers
during incineration occurs nonuniformly.
700
I
producing dimples or notches in the
fibers. In the laboratory, it was noted
that as oxidation proceeds, each
notched fiber is cut in two at the site of
the notch, leaving fibers of relatively
short length and pointed ends. Thus,
most oxidized fibers of small diameter
also are relatively short.
Table 2 shows the results of the
incineration study in terms of emission
rates. The total emission rates p /o
measured by the SASS train (55.6 u
mg/dscm particulate; 0.998 mg/dscm ;§
carbon fiber) agree quite well with those
measured by Method 17 (73.0
mg/dscm particulate; 0.840 mg/dscm 4>
carbon fiber), given the nature of the ,£
experiment. Inspection of the Method c
17 emission rate data also reveals a -2
decay in fiber emission rate with time as -S3
would be expected considering the ^ 1.0
single-point (with respect to time)
introduction of the fiber composite into
the stoker. This phenomenon is shown
graphically in Figure 1. The carbon fiber
emission rates display a nearly
exponential decay in time, whereas the
total particulate emission rate is
constant within normal operating
bounds.
Table 3 is a breakdown of the results 0.1
by individual sample. It is interesting to -15
note that nearly all of the carbon fibers
(99.6%) were captured in the first two
cyclones, whereas only a relatively Figure 1.
small portion (17.1%) of the total
particulate matter was collected by
these cyclones. This pattern is not
surprising, given the aerodynamic
diameter of these carbon fibers
(approximately twice the physical
diameter given in Table 1, or 10 /urn).
This observation suggests that cyclones
may be used to separate the carbon
fibers from extraneous particulate
material, thus facilitating analysis.
Table 2. Incineration Test Emissions Rates
Relatively Constant
Rate of Total Emissions
60 ± 20%
///,
Decreasing Rate of\
Fiber Emissions I
0
Fibers Introduced
15 30 45 60
Time, minutes
Rates of carbon fiber and
total particulate emissions.
Sample
SASS Train
Method 1 7
Pretest
0-15 min
15-30 min
30-45 min
45-60 min
Nozzle Rinse
Total Method 1 7
Sample
Volume
(dscm)
7.88
0.295
0.293
0.291
0.287
0.286
1.452
1.452
Total
Particulate
(mg/dscm)
55.6
57.6
63.8
70.1
52.6
47.7
14.6
73.0
Carbon Fiber
(mg/dscm)
0.998
0.003
1.40
0.62
0.45
0.24
0.30
0.84
(fibers/dscm)
4310
—
—
—
—
—
-------
Table 3. Results of Incineration Stack Sampling
Sample
SASS Train
10 urn Cyclone
3 fjm Cyclone
1 iim Cyclone
Probe
Filter
Total
Paniculate
fmg)
58.3
16.7
0.4
13.3
391.8
Carbon
Fiber
(mg)
6.33
1.50
0
0.03
0.0005
%
Carbon
Fiber
10.9
9.0
0
0.23
0
% of Total
Carbon Fiber
Sample
80.5
19. 1
0
0.4
0
Carbon
Average
Length
(fjm)
5534
1320
—
108
55
Fiber
A verage
Width
(fjm)
4.98
3.45
—
2.94
5.6
Total SASS 438.2
Method 17
Pretest 17.0
0-15 min 18.7
15-30min 20.4
30-45 min 15.1
45-60 min 13.6
Nozzle Rinse 21.2
Total Method 17 106.0
7.86
0.001
0.41
0.18
0.13
0.07
0.43
1.22
1.79
2.19
0.88
0.86
0.51
0.20
1.15
0.1
33.6
14.8
10.7
5.7
35.2
W. M. Henry, C. M. Melton, and E. W. Schmidt are with Battelle Columbus
Laboratories. Columbus, OH 43201.
Roy L. Bennett is the EPA Project Officer (see below).
The complete report, entitled "Method for Measuring Carbon Fiber Emissions
from Stationary Sources," (Order No. PB 83-118 760; Cost: $8.50, subject to
change) will be available only from:
National Technical Information Service
5285 Po£ Royal Road
SpringfieHf, V'A 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Environmental Sciences Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
U.S. GOVERNMENT PRINTING OFFICE: 1983 6S9-O17/O8E
Postage and
Fees Paid
Environmental
Protection
Agency
EPA 335
Official Business
Penalty for Private Use $300
------- |