&EPA
i i-i'i.'i] States Imiir.ir M Environmei ' i. !< :
EnvirI'imu'iital Protecti Laboraior\
nc\ Resean h fnangle P.ni- N( 2771 '
Electrostatic
Effects in Fabric
Filtration:
Volume II.
Triboelectric
Measurements and
Bag Performance
(Annotated Data)
Interagency
Energy/Environment
R&D Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of. and development of, control technologies for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Approval does not signify that the contents necessarily reflect
the views and policies of the Government, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service. Springfield. Virginia 22161.
-------
EPA-600/7-78-142b
July 1978
Electrostatic Effects in Fabric Filtration:
Volume II. Triboelectric Measurements
and Bag Performance
(Annotated Data)
by
E.R. Frederick
Carnegie-Mellon University
Schenley Park
Pittsburgh, Pennsylvania 15213
Grant No. R803020
Program Element No. EHE624
EPA Project Officer: James H. Turner
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
ABSTRACT
The construction and application details of a bench scale,
single bag, experimental filter unit are described. Also consid-
ered at length are a variety of supporting and complementary
evaluation procedures, together with the data that they provide.
Of special significance among the latter are the methods for, and
results of, the electrical determinations that are not normally
applied to filter media and particulates. The effect of these
electrical parameters on the collection process is considered and
used to explain performance variations.
The results of a number of separate filtration studies,
carried out on a variety of industrially generated particulates,
including those from a power plant and from metallurgical and
chemical processes, are reviewed in detail and explained, as
permitted, on the basis of electrostatic properties. The collec-
tion of a flyash, for example, was favored by the use of mid-
triboelectric position media and not by the highly electropositive
or electronegative fabrics that are employed for their high temp-
erature properties. Three different electric furnace dusts also
tended to respond best filtrationwise with mid-triboelectric
position fabrics, using appropriate constructions for different
cleaning practices. Steel grinding and burning dusts were shown
to offer very critical filtration characteristics that demanded
control of aerosol flow and particulate loading, as well as spec-
ial care in the selection of the filter media. A ferromolybdenum
by-product dust was collected best by very electropositive fabrics,
but three different, high resistivity polymeric dusts performed
best with mid-triboelectric position fabrics of suitable construc-
tion. Silica was found to be collected most efficiently by high
cover electropositive fabrics with variations in construction
dependent upon the adopted method of cleaning.
ii
-------
PREFACE
It is generally conceded that electrostatics may play a role
in the capture and retention of many kinds of dust particles
during the filtration process. While this feature of particle-
to-filter medium interaction is relatively easy to accept, the
influence of electrostatics on collection efficiency, fabric
cleanability and especially on particulate agglomeration is not
always readily acknowledged nor easily justified. Opposition to
the concept of significant overall electrostatic involvement in
the filtration process persists despite the fact that electrical
forces have already been shown to be much stronger than gravi-
tational, thermal, adhesion, and often inertial forces for
particles in the 0.1 to 1 um range.1
If the collected particles and the collection medium are
both subject to charging, and these processes are not disputed,
and if these charges are of high magnitude and particularly of
opposite polarity, it seems apparent that a high level of attrac-
tion can occur between the two materials. What tends to be more
difficult to compreh'end, though, is the influence of these
electrostatic characteristics on the agglomerating tendencies of
the particles being collected.
In this report, the obvious practical effects of electro-
statics are recognized but the real key to superior performance
is relegated to the agglomeration phenomon. As small particles
are brouaht into close oroximitv bv and/or on suitably charged
fibers , a kind of aggregation is considered to occur that results
in a porous deposit. A cake formation such as this then is
indicated to be less resistant to gas flow than that of a compact
cust layer occurring without such electrostatic interaction.
Recent activities devoted to studies of electrostatic augment-
ation in the mechanical separation, wet scrubbina or filtration
of fine particles have reaffirmed the usefulness of this extra
input for providing superior performance, in terms of both better
operating parameters and collection efficiency- Despite this
effort, however, little attention seems to have been directed, and
Whitby, K. T. and B. Y. H. Liu, The Electrical Behavior of
Aerosols, In Aerosol Science. Davies. CN.(Ed.). New York, NY,
Academic Press, 1966.
iii
-------
no direct evidence has been obtained to identify the fundamental
principles that lead to such improvements. The work of Gaylord
W. Penney reported in Volume I of this two-part series on
Electrostatic Effects In Fabric Filtration, however, considered
the basic principles and provides evidence to indicate that
electrostatics truly does contribute to the formation of a
porous deposit. If electrostatics by augmentation is a factor
in this kind of change, then surely the same features accomplished
by natural charging processes (contact/impact, etc.) should also
be effective. Accordingly, in this report (Volume II of the
series), the importance of, in fact, the critical need for,
suitable electrostatic charge balancing (between medium and
particulate) and obtainable naturally without augmentation, is
stressed as the mechanism by which optimal filtration performance
may be achieved. Ideal collectability is considered possible
under normal baghouse conditions when, but only when, the
appropriate electrostatic characteristics are established even
without the need for - or restrictions imposed by - augmentation.
Selection of the most favorable fabric then is made on the basis
of its triboelectric properties commensurate with those of
the particulate, while at the same time, giving appropriate
consideration to the constraints of fabric construction.
iv
-------
CONTENTS
Abstract ii
Preface iii
Figures vi
Tables vii
Acknowledgments Till
Definitions & Abbreviations ix
Conclusions 1
Recommendations 2
Experimental Filtration Facilities 3
Experimental Filtration Procedure 5
Supporting Instrumentation 6
Determinations of Electrical Properties 6
Electrical Resistivity 6
Surface Resistivity of Fabrics 10
Electrostatic Properties of Fabrics 12
Particulate Resistivity 27
Electrostatic Properties of Particulates 27
Particulate Agglomeration of Porous Cake Forming Tendencies. . 27
Visual Examinations of Particulates (In the Course of
Filter Tests) 29
Representative and Experimental Filter Fabrics 29
Experimental Filtration Studies 29
The Collection of Flyash, WPP-S 32
The Collection of Some Metallurgical Dusts 34
The Collection of Electric Furnace Dust, U-D .... 34
The Collection of Electric Furnace Dust, C-T .... 38
The Collection of Electric Furnace (Stainless
Steel) Dust, U-J 40
The Collection of Stainless Steel Burning Dust,
U-H 41
The Collection of a Ferromolybdenum By-Product
Dust, CM-L 45
The Collection of a Steel Grinding Dust, R-C .... 48
The Collection of Polymeric Dusts 51
The Collection of Polymeric Dusts P-K-3085 and
P-6140 51
The Collection of a Fine Polymeric Dust, RH-P ... 55
The Collection of Silica, PG-C 57
Summary 62
Index 66
-------
FIGURES
Number Page
1 Experimental Filtration Equipment . 4
2 The Permeometer 7
3 Fabric Triboelectrification and Charge Measurement ... 8
4 Dust Resistivity Equipment ..... 9
5 Fabric Resistivity Equipment .... 11
6 Particulates Before and After Fabric Contact 31
7 Experimental Filtration of Electric Furnace Dust,
U-D (1) 36
8 Experimental Filtration of Electric Furnace Dust,
TJ-D (11) 37
9 Steel Grinding Dust, X 500 52
10 Aggregated Deposit of Precharged Particulate on a
Filter Fiber 63
-------
TABLES
Number Paoe
1 Fabric Resistivity vs Anti-Static Rating 12
2 Estimated Triboelectric Position of Some Filter
Fabrics 14
3 Electrical Variations In Some Fabric Filter Media ... 16
4 Some Properties of Various Filter Media 18
5 Electrical Resistivity of Some Particulates 25
6 Agglomeration and Apparent Density Changes of Some
Particulates by Rolling 30
7 Experimental Filtration of Fly Ash, WPP-S 33
8 Experimental Filtration of Electric Furnace Dust, C-T. . 39
9 Experimental Filtration of Stainless Steel Burning
Dust, U-H 43
10 Experimental Filtration of a Ferromolybdenura By-
Product Dust, CM-L (1) 47
11 Experimental Filtration of a Ferromolybdenum By-
Product Dust, CM-L (11) 47
12 Experimental Filtration of Steel Grinder Dust, R-C ... 50
13 Experimental Filtration of Polymeric Dusts P-K-3085
. and P-6140 54
14 Triboelectric Position of Silica, PG-C 59
15 Experimental Filtration of Silica, PG-C 60
16 Metrication of Some Filter Parameters 65
irii
-------
ACKNOWLEDGMENTS
The co-investigators are particularly grateful to Dr. James
H. Turner, the EPA Project Officer, for his thoughtful considera-
tions and technical support during the course of the investigative
program. Others, too numerous to mention here, have contributed
significantly to the success of the study through their cooperation
in providing commercial and experimental filter media and in
supplying problem dusts. The extraordinary assistance given by
Messrs. John W. Brooks and Robert W. MacWilliams is acknowledged
with sincere appreciation.
The project could not have functioned effectively without
the dedication and independent contributions of the graduate
student Robert Lembach and the technicians Ron Feigel, Kirk Lud-
ington, David Richey, John Wieczorkowski and Brent van Zandt.
viii
-------
LIST OF DEFINITIONS AND ABBREVIATIONS
DEFINITIONS
aerosol
agglomerate
antistat
augmentation
cake
charge
...polarity
...intensity
...dissipation rate
cleaning
Dacron
Darlan
Dralon T
dust
efficiency
electrostatic
fabric
fiber
fiberglass
filter
Kevlar
Kodel
Microtain
Nomex
nylon
particle
permeability
plug
PVA
resistance
triboelectrification
weave
yarn
yarn count
particulate dispersion in gas
- consolidation of particles
- charge bleed-off agent
- increase (of charge) artificially
- collected solids, removable by cleaning
method
electrical quality
electropositive or electronegative property
-- magnitude
loss rate
particulate removal process
-- a polyester fiber
- a dinitrile based fiber, discontinued
a polyacrylic fiber
- particles or particulates
- control effectiveness
-- electricity at rest
-- cloth structure
basic element of filter
-- fine glass fiber
-- gas-particulate separator
-- aramid fiber
polyester fiber
proprietary acrylic name
-- aramid fiber
-- polyamide fiber
-- particulates or dusts
air flow-through rate
-- particles held in or on fabric and not
removed by cleaning
polyvinylacetate
-- electrical resistance
-- rub-induced electrification
-- fabric construction
construction element of woven fabric
- warp and filling yarns/unit length
ix
-------
ABBREVIATIONS
1
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
AATCC American Assn. of Textile
Chemists & Colorists
AF Albany International or
Globe Albany Corp.
AFI Air Filters, Inc.
as antistatic
cal calendered
Carb The Carborundum Co.
cond conductive
cs cotton system
DFTw double faced twill
Eld E.I. du Pont de Nemours
and Co., Inc.
F filling yarn
°F degrees Fahrenheit
fil filament
fin finish
g gram
Gore W.L. Gore & Assoc., Inc.
h hour
H-as Hyamine-antistatic agent
Hb T.J. Heimbach Gmbh & Co.
(Duren, Germany)
h cover high cover
Horn Homestead Mills
JPS J.P. Stevens & Co., Inc.
knit weaving variation by
looping threads
lam laminate
MSA Mine Safety Appliance Co.
nap high concentration of
fiber ends
nat natural
nd not determined
os one side
AP pressure drop (in. w.c.)
across filter
APp
pe
Pi
PP
P & S
ptfe
Resp
Si, Sil
sp
ss, s steel
st
Tas
Tef
tex
Troy
unt
USF
V
W
W-Ac
ws
ZnR
C-T \
CM-L
PG-C
R-C
RH-P
U-D
U-H
0-J
WPP-S /
pressure drop (in.
w.c.) across plugged
filter
polyester
plain
polypropylene
P & S Textiles Inc.
polytetrafluoro-
ethylene (Teflon)
respirator felt
silicone
spun
stainless steel
staple
Taslan (bulked
filament)
Teflon
texturized (bulked
yarn)
Troy Mills, Inc.
untreated
United States Filter
Corporation
volts
warp
Western Acadia, Inc.
woolen system
Zinc resinate
code letters identi-
fying dust origin
-------
VOLUME II
EXPERIMENTAL FILTRATION PROGRAM
CONCLUSIONS
A single bag, bench scale experimental fabric filter unit has
been designed, built, modified, and used effectively to demonstrate
how different fabrics produce very different filtration character-
istics. These results, together with data from additional facil-
ities for determining the electrical as well as other parameters
of fabrics and particulates, indicate that filtration properties:
differ with changes in only the fiber make-up of similarly
constructed fabrics,
differ with changes in only the construction of fabrics made
from the same kind of fibers,
differ with changes in the surface properties of the same
fabric,
differ with changes in the surface properties of the same
particulate,
seem to be influenced critically by the electrical properties
of both particulate and filter fabric, and
tend to be enhanced most significantly by the electrostatic
conditions of the fabric filter that produce an agglomerated,
porous-type filter cake.
-------
REC OMMENDATIONS
The hit or miss tactics used so commonly in prescribing fil-
ter media, the questionable quality and the inconsistencies among
media once specified all need to be eliminated if baghouses are
to achieve their full potential for optimal overall performance.
One of the first steps considered necessary to utilize fully the
electrostatic properties is that of establishing the fabric's
ability to develop an electrostatic charge. Secondly, the elec-
trostatic properties of all industrial filter media as well as
those of the particulates should be known. Thirdly, the preferred
fabric filter media-collected particulate charge relationship for
optimum collectability, efficiency and energy conservation should
be established for filtration processes of concern.
In order to achieve these goals, the following specific re-
commendations are offered:
include electrical resistivity among the specifications for
fabric filter media;
include the electrostatic properties of charge intensity,
charge decay rate and triboelectric position in the list of
properties of filter media;
develop a procedure for determining the electrostatic prop-
erties of particulates;
determine the influence of industrial environmental conditions
on the charging properties of media and particulates; and
... establish an acceptable procedure for - and carry out, reg-
ularly, experimental evaluations of - the collectability of
particulates for the purpose of providing accurate prescrip-
tions for media and to relate the electrical as well as the
other physical properties of media with filtration parameters.
-------
EXPERIMENTAL FILTRATION FACILITIES
The bench scale fabric filtration system shown in Figure 1
was devised and fashioned to simulate the performance of one bag
in a commercial baghouse and to supply information on every char-
acteristic of its operation. This testing equipment allows for
considerable versatility in evaluation procedures with the necess-
ary instrumentation for reliable and complete delineation of elec-
trical and other physical properties as well as filtration para-
meters. Most varieties of particulates may be processed by either
the shaking, reverse air or pulse collector method. The more im-
portant characteristics of the test facility are described as
follows:
(Please note that the letters referred to in the text identify the
parts so noted in Figure 1.)
The particulate feeder (A), a critically important part of
the filter unit, was changed in the course of the investigations,
from the original vibrating-auger system to a vibrating and rota-
ting gear type of dust dispenser. The latter was developed and
used with agitation also applied to the dust in the hopper to in-
troduce even the sticky and difficult-to-feed particulates into
the air injector (B) continuously at a uniform rate. The quantity
of dust occluded in the space between two teeth of the gear can be
made to be quite consistent through the aid of the gentle settling
afforded by vibration and the constant movement imparted by gentle
stirring in the hopper. The separate parcels of dust are delivered
at a rate determined by the gear's speed of rotation. These par-
cels, essentially of constant size for the gear now in use, are
dispensed into the injector by a controlled (C) air stream where
the dust is dispersed by moderately high (100 psi) air pressure.
The resulting aerosol passes through the flow measuring venturi
(D) and the heating section (E). Further on in the system, the
aerosol reaches the sampling station (F) and the popper (G) vent
(H) before the reheater (I) and header (J) under the test bag (K).
Dust fallout from the aerosol before it reaches the bag as well as
dust shaken from the test filter bag is collected in the jar (L).
Incidentally, a determination of the ratio of the dust fallout
into the header to that collected on the bag during repeated runs,
serves to indicate the condition of the dust (i.e., the amount of
aggregation and the relative concentration of coarse particles).
In operation, the test bag (K), suitably fitted between the
two outside stitch lines with a fine, 30 gauge wire in order to
-------
Figure 1 Experimental Filtration Equipment
-------
allow electrical measurements during the filtration process, is
enclosed in the cabinet (N). The cabinet is kept under a slightly
negative pressure in order to exhaust any leaked dust to a large
filter bag behind the panel. Appropriate taps are included, one
in the header for pressure drop recording continuously on the
chart (0) and for direct reading on the manometer (P). The flow
rate of the aerosol stream is determined by means of a calibrated
venturi (D) at the manometer (Q). In order to restrict the humid-
ity of the aerosol and as needed to conform with commercial condi-
tions, heat is applied to the air-dust stream at the regions (D)
and (I) and controlled by the thermostat in the header (J).
Cleaning of the bag may be accomplished by the horizontal
shake action imparted by the spring return-solenoid activated unit
(U) or by means of the "popper" (G) system that applies a vacuum
type, pull-back action.
EXPERIMENTAL FILTRATION PROCEDURE
The performance, filtrationwise, of different media in col-
lecting various particulates is determined as a function of the
weight of collected solids, the weight and air flow resistance of
the plugged solids held within the fabric and the effectiveness of
the fabric in retaining the particulates. The data needed for
making appropriate filtration comparisons is obtained under uniform
and reproducible conditions of essentially equivalent dust loadings,
aerosol flow rate and, in these studies, to an arbitrarily, pre-
selected and fixed pressure drop limit. Physical changes in the
particulate from its condition in the initial dispersion to that
found in the collected form from the collecting bag are also deter-
mined and the changes in appearance are recorded photographically.
Following preliminary determinations of the filter bag weight,
clean bag pressure drop, and resistivity, filtration is commenced
and the changes in pressure drop across the filter fabric as a
function of deposited cake are recorded on the chart (O) and read
on the manometer (P). Periodically, in the course of the evalua-
tion, when the pressure drop across the filter flow is halted to
stop aerosol flow, the terminal pressure is recorded, the bag is
removed from the header and weighed to determine collected solids
(cake and plug). After reinstallation, the bag is cleaned by the
selected method (shake, reverse air, or simulated pulse jet) and
the plugged bag pressure drop is recorded. The cleaned bag is now
removed from the header and weighed to determine the amount of
accumulated plug. Simulation of the commercial reverse air clean-
ing process on woven fabric bags or simulation of the pulse air
jet practice applied to felted fabrics is accomplished by means of
the "popper." Its activation is accomplished by means of the switch
and timer (V) to impart a controlled number of air pulses with
fixed intermittent dormant periods, usually of one second.
The number of filtration cycles needed to reach equilibrium
depends upon a number of material and operating conditions.
Usually, at least ten separate filtration and cleaning operations,
-------
with at least five cake and plug weight determinations during the
final cycles are required to establish these parameters.
SUPPORTING INSTRUMENTATION
A number of instruments, gauges and other supporting analytical
facilities are required to determine dust loading, aerosol flow
rate, pressure drop, clean, caked and plug bag weights, pressure
drop, and permeability as well as the electrical properties of the
filter fabric of the caked and plugged bags and of the dusts being
filtered.
A balance (W) mounted for convenience at side of the bag en-
closure serves to obtain clean, caked and plugged bag weights.
Suitable balances are also available for determinations of dust
loadings, particulate accumulations in the collector jar and leaked
dust. Such other instruments or equipment as manometers, ammeters,
voltmeters, timers, variable drives, vibrators, stirrers are also
evident in Figure 1. In addition, back-up facilities are available
such as a fabric permeometer (Figure 2), a fabric electrostatic
generator-evaluator (Figure 3), dust resistivity measurement equip-
ment (Figure 4) and such others as fabric sewing, washing and dry-
ing machines together with photographic facilities.
DETERMINATIONS OF ELECTRICAL PROPERTIES
Electrical Resistivity
The electrical resistivity of dusts and fabrics influences
chargeability mostly by increasing the rate of charge dissipation.
Antistatic finishes do not eliminate static, but they are effect-
ive in bleeding-off charges rapidly. Because these agents invari-
ably reduce electrical resistance and they are used so extensively
in the processing and/or production of fibers and fabrics, fabric
resistivity data can serve to indicate the cleanliness of materials
offered for filter use.
The electrical resistivity of a substance may be defined in
two ways; through it and referred to as volume resistivity or over
its surface and, of course, referred to as surface resistivity.
Volume Resistivity
Volume resistivity refers to the electrical resistance meas-
ured through the bulk of the substance under noted conditions of
temperature and humidity and is an especially important parameter
of both particulates to be collected and filter fabrics. The
guarded electrode method (ASTM D 257-61) is useful for determining
either volume or surface resistivity, but other proceedures were
adopted for making these measurements.
Test Procedure
Consider, for example, a cube of material to be evaluated.
-------
manometer
pressure drop across sample
Figure 2 The Permeometer
-------
. 3A
Static Generation and Evalu
ion Equipment '" Static Charge
Generation. (7)Test Fabric; (£)
Test Fabric Tensioning Weight;
(3)Test Fabric Support Frame;
(5)-Test Fabric Frame Tensioning
Weight; ©~ Reference Fabric
(contacting test fabric): (?) Ref-
erence Wheel Drive Motor; (?)
Voltage Probe (retracted)
3E
AFC Static Generation and Evalua-
tion Equipment j Static Charge
Measurement. (7)Test Fabric;
(z)Test Fabric Tensioning
Weight; (T)-Test Fabric Support
Frame; (4)Test Fabric Frame
Tensioning Weight; (^Refer-
ence Fabric (removed from test
fabric); @Reference Wheel
Drive Motor; (7)Voltage Probe
in Measuring Position.
FIGURE 3. FABRIC TRIBOELECTRIFICATION AND CHARGE MEASUREMENT
*U.S. Patent 3487296
8
-------
OVERHEAD. CLOSE-UP, INSIDE VIEU Opl
SAMPLE ENCLOSL'RE
Figure A Dust Resistivity Equipment
-------
With a conductive plate or a naturally conductive surface on oppo-
site faces of the cube, a DC voltage is impressed in stages from
low potential (^ 200 V) to high values (^ 2000 V) until a reproduc-
ible measure of current is obtained. From Ohm's Law, then, the
value of resistance, R is calculated. Using the equation R=pL/A,
volume resistivity (p) in n-cm is determined from the distance
L(cm) between the faces and the area A(cm2) of the smallest oppos-
ing face.
The procedure adopted in this laboratory to determine the
resistivity of particulates is considered on page 28 with the
apparatus shown in Figure 4.
Surface Resistivity of Fabrics--
The surface resistivity of filter fabrics as produced or as
caked or just plugged may be determined by any of several proced-
ures including the ASTM guarded electrode method1 referred to above,
the AATCC Test Method 76-1972l (approved as the ANSI L14.112-1973
method), or by the square method1 described below and shown in
Figure 5. Surface resistivity by this practice is defined as the
resistance in ohms per square (fi/p) of the test specimen.
It will be apparent that this is a relatively easy to carry
out test employing no special weighted and shaped electrodes but
only the 3 in. wide clamps to hold the specimen at a separation
of 3 in. In operation, the equilibrated test specimen of suitable
size (3"+ x 3"+), is clamped securely and the DC voltage is applied
in stages beginning at a moderately low value of 200 V and gradually
increased as required, usually up to 2000 V or until a reproducible
value is provided for the measured current. Ordinarily, values of
current below 10 ll or above 10~5 need only to be noted, not meas-
ured, as will become evident from the following discussion.
The current obtained by impressing a known voltage on the
specimen and measured by means of the electrometer is used in the
Ohm's Law relationship to calculate resistivity as follows:
At voltage E(V), resistivity R (n/n) = g (applied voltage) V
J LJ I (measured current) amps
Where E is usually 500 V but may be increased to at least 2000 V
as required;
I is the current in amperes; and
R is the calculated resistance in ohms per square, n/-i.
[It will be apparent that because the path of the impressed voltage
covers both the top and bottom surface of a fabric, the value of
the resistance determined under these conditions will actually be
one half of the real value, and the true surface resistivity will
be twice that measured. Ideally, then, especially since filter
}Data obtained by these methods are essentially equivalent and
well within the limits needed to determine whether or not a fabric
carries an antistatic finish.
10
-------
Figure 5 Fabric Resistivity Equipment
-------
media are so often made with different surfaces, both of the clamp
edges on one side, but on the same side, should be insulated.
Reversing the clamps with insulated surfaces on a different side
each time provides current values that allow calculation of R (ft/pi]
directly for each surface.]
Resistivity values of about 10}1 fl/Q and less, indicate anti-
static qualities, more or less significant, as shown for similar
evaluations in the following Table:
TABLE 1. FABRIC RESISTIVITYa VS ANTI-STATIC RATING
Volume/vertical Surface/horizontal Anti-static
resistivity resistivity0 rating
(ohm-cm) (ohm)
[Material equilibrated at 70°F]
>1011 >1013 Nil
lO^-lO11 1012-1013 Poor
109-1010 lO^-lO12 Moderate
108-109 lO^-lO11 Fairly good
<108 <1010 Good
?ASTM D257-61
E. R. Frederick/ Report to Mellon Institute, December 1, 1964
May 31, 1965
D. Wilson, J. Textile Inst.f 54: T97 (1963), as measured
between opposite edges of a 1-in. square or, for high resist-
ance materials, between 2-in. electrodes, 0.2 in. apart.
It should be evident that since moisture can influence re-
sistivity quite significantly, especially the resistivity of mate-
rials that sorb water, the measurements should be conducted under
conditions of controlled temperature and humidity, preferably at
moderate to low values (not over 50% RH).
Electrostatic Properties of Fabrics
Whenever two dissimilar materials (usually, at least one of
which is an insulator) are rubbed together, one becomes electro-
positive, the other electronegative. This is one way particles
and fabrics become charged.
Polarity variations among fabrics reflect inherent fiber
differences that may be demonstrated from rubbing tests. The
testing unit that we use is shown in Figure 3. By the rubbing
12
-------
method, a triboelectric series is obtained in which all fabrics
may be listed from those that are very electropositive like wool,
glass, and nylon, to those that are quite electronegative like the
polyolefins, and especially, "Teflon." It will be apparent that
by repeated rubbing trials, the series may be expanded to include
any number of fabrics (Table 2). A variety of materials, including
particulates, may be located in the same series. Other factors
being equal, the greater the spread between materials in the series,
the greater the interaction between the two materials.
The intensity of the electrostatic charge (together with any
polarity differences) is considered to have a marked influence on
the collection and is a function of surface (fiber, yarn, or fab-
ric) roughness as well as of the inherent properties of the poly-
mer. Other factors equal, the rougher the surface, the higher the
generated charge. Rough fibers like wool, rough yarns like low
twist spun yarns, and high cover fabrics like those that are nap-
ped, tend to develop higher charges than smooth surfaced materials
such as those made from smooth (melt extruded) continuous filament
fibers in low twist, pressed, calendered, or otherwise smooth
fabrics.
Evaluation Procedure (Figure 3)
Triboelectrification and Measurement
1) Mount a clean and conditioned reference fabric on the 3 in.
diameter wheel of the motor using the wedge to fix the over-
lapped ends in the slot of the wheel. (Do not permit fabric
ends or wedge to extend above surface.)
2) Move the motor and other parts of the platform to the rear
position.
3) Without soiling the conditioned test fabric, cut the sample
to 3 in. width and 12 in. length and insert one end of the
length in the center of the clamp at the base of the sample
support.
4) Place the test fabric over both rods in the support frame
and attach the clamp weight to the free end on the left-hand
side.
5) ' Pull the motor support platform to the front position so that
the reference fabric wheel is centered on the test fabric.
6) Zero the electrometer and ground off charges from the test
fabric.
7) Extend the tensioning string from the sample support frame
over the pulley rod of the frame at the far right of the test
unit base and allow the weight to hang free. This will cause
the test fabric to engage the reference fabric covered wheel.
13
-------
TABLE 2. ESTIMATED* TRIBOELECTRIC POSITION OF SOME FILTER FABRICS
4" 8
21 & 97 WOOL/NYLON, 21
104 WOOL, HOM 8
78 WOOL/NYLON
APPROXIMATE LOCATIONS
112 DACRON
-J- £.. NYLON 800 B
102 WOOL. HOM 7 f85%~ 23 WOOL/COTTON [J00*c
4.5. 122A DRALON T CDYED~) NAP
*
'
^_ 3
4- 1 .
11
ft
V
-2
-3
,5 DACRON
103 WOOL. HOM 6
1 8 POLYESTER
1 17 ACRYLIC, ZC [>>*« j
NYLON
122B
110 NYLON, NAP
.114 POLYESTER/GLASS
116 POLYESTER
_ » 18 POLYESTER,
93 DACRON (_40% j
* I20 DRALON T C30%J 19 POLYESTER a PVA ^45%
9 50/50 DA/OR Q35% J
77 GLASS C77%I! * 2 NOMEX
118 POLYESTER Q70%I]
7 ACRYLIC, Z
107 DACRON, NAP C60%D B7 DRALON T C85%J* 41 ACRYLIC, 7.
* 12 ORLON C30%H 42 QRLON [>0%D
10 75/25 DA/OR f>o%D . 3 DRALON T
16 DACRON SI £30% J
83 POLYPROPYLENE
4-- DARLAN S546
so GORE CNOMEX BASED LTS
-6
37 TEFLON
65 KEVLAR
APPROXIMATE LOCATIONS
FROM TRIBOELECTRICIFICATION DATA BY PROPORTIONAL CALCULATIONS
HI = RELATIVE DISCHARGE RATE Cl-OSS C%D IN 2 MINUTES^ AT 50% RH
NOTE: NUMBER PRECEDING EACH FABRIC IS AN ARBITRARY FABRIC
REFERENCE NUMBER CSEE TABLE O
14
-------
8) Turn on the motor switch momentarily (1 to 2 sec) and allow
the reference wheel to stop. (Within the limits of about 100
to 500 rpm, the number of revolutions of the reference wheel
is not critical).
9) Remove the weight attached to the test fabric support frame
by lifting it over the frame to a resting position on the far
left of the test unit (as in Figure 3B). This will remove
the test fabric from contact with the reference fabric.
10) With the probe support in the far right position, move the
motor base support to the rear until the probe holder reaches
the center of the test fabric.
11) Slide the probe support to the far left position so that the
probe contacts the rubbed test fabric.
12) Read the meter.
13) Slide the probe support to the far right position.
14) Slide the motor base support to the rear position.
15) Ground off the charges from the test fabric by rubbing ground
wire over the entire rubbed area and beyond.
16) Repeat the above measurement at least four times, recording
the fourth and fifth measurements as the true charge intensity.
Triboelectrification and Discharge Rate Determination.
1) Repeat the operations indicated above through operation 9.
2) Allow the test sample to remain in this condition for a per-
iod of two min after the rubbing operation.
3) Slide the probe support to the far left position so that the
probe contacts the rubbed but discharging test fabric.
4) Read the meter.
5) Two or three tests of this type are usually adequate for pro-
viding data to be used with that obtained in operation 12
for indicating the rate of charge dissipation.
6) The charge dissipation (percent) after two min equals 100 x
[(data from operation 12) minus (data from operation 4 of
this section)] divided by (data from operation 12).
Triboelectric Data for Some Filter Fabrics
The direct influence of a fabric's electrical resistivity on
electrostatic charging and the rate at which the charge decays is
15
-------
TABLE 3. ELECTRICAL VARIATIONS IN SOME FABRIC FILTER MEDIA
No.
Fabric Type
Resistivity, n/Q
i Room @ Room
Conditions Conditions
(80°F/39% RH) (after 150°F/16 h)
Relative Electrostatic Data [p.l2f
vs. vs. 2-min
nylon, darlan, loss,
volts volts %
2
2
30
30
18
18
41
41
Aramid, fil. (D.F.TW.) as received
Aramid, £11. (D.F.TW.)
wash-nonionic, rinsed well
Aramid, fil. (3 x 1 TW) as received
Aramid, fil. (3 x 1 TW)
wash-nonionic, rinsed well
Polyester, sp. (2x2 TW) napped
as received
Polyester, sp. (2 x 2 TW) napped
wash-nonionic, rinsed well
Acrylic, sp. (3x2 TW) as received
Acrylic, sp. (3 x 2 TW)
wash-nonionic, rinsed well
7.0 x 1010 - 0.5
5.0 x 1013 -- -10.0
2.0 x 1010 -- - 0.5
2.0 x 1013 - 6.6
8.3 x 109 7.4 x 1013 - 1.3
2.0 x 1012 5.0 x 10II+ -10.0
2.0 x 10n - 1.0
(bone
1.9 x 1013 2.7 x 1014 dry) -10.0
(from oven) to
7.0 x 10 13 (conditioned)
+0.5 100
+3.0 30
+ 1.0 100
+ 3.0 20
+ 3.9 80
+11.6 25
+ 1.0 100
+ 5.8 25
AATCC Method 76-1972 (ANSI L14.112-1973-2/15/73)
70°F/50% RH
-------
shown in Table 3. The data reported here are for only four com-
mercial fabrics and indicate that just a modest reduction in sur-
face resistivity (i.e. to 1010 n/Q) can affect chargeability to
a significant extent; thus confirming the relationship given in
Table 1. The effect shown on resistivity and, thereby, on electro-
static charging of a wash-removed antistatic finish is evident.
It can also be seen that the finish is not active when desiccated
(note the change due to heating fabric #18) and is fugitive. Data
such as these have been obtained for an exceptionally large number
of commercial filter fabrics. If, therefore, electrostatic charges
are as important in the filtration process as this report will
attempt to show, the presence of such finishes must be known and,
for most applications, removed prior to use. The easy-to-carry
out resistivity test, therefore, would seem to be a necessary part
of any filter media control program.
The results of rubbing tests conducted on some of the fabrics
available for filtration evaluation, according to the procedure
noted above, are included in Table 4. The relatively low resist-
ivities of some as-received fabrics, most of which develop much
higher values after laundering, clearly indicate the presence of
a removable antistatic finish, most likely the fiber producer's
treatment applied to enhance processing. That this finish is re-
movable by laundering [or simply a hot (140°F) water rinse] is
evident from the change in both resistivity (from 1010 to 1013
or higher values) and the triboelectrification data.
The triboelectric series provided in Table 2 is offered to
illustrate the usefulness of the rubbing tests for establishing
such a series. Because a proportional kind of analysis of these
rubbing data was used to develop the Table, the positions should
not be considered to be absolute. The real locations can be
specified accurately only by using all of the samples as both the
rubbed and rubbing materials.
While this restriction on the use of the series as given is
not to be overlooked, it is interesting to speculate on some of
the apparent similarities and anomalies shown. For example, the
four acrylics, despite their obvious chemical difference, are
located close together in the -0.2 to -1.5 region of the arbitrary
scale. Dacron, on the other hand, shows a spread of from about
+4.8 to -2.5 on this scale. Such wide differences in the tribo-
ele"ctric data for Dacron were shown earlier. : Also significant
and undeniable, is the very electropositive character of the wool/
nylon fabric and the extremely negative properties of Teflon and
Kevlar. Actually, Kevlar appears to be the most electronegative
fabric and it is also interesting because of its tendency to lose
its charge faster than Teflon.
Frederick, E. R.3 Chem. Eng. 68:107 107-114 (June 1961).
17
-------
TABLE 4. SOME PROPERTIES OF VARIOUS FILTER MEDIA
Fabric
Ho.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Hfr.
JPS
JPS
JPS
JPS
JPS
JPS
JPS
AF
AF
AF
Eld
AF
AF
AF
AF
JPS
JPS
JPS
(continued)
Style No.
D-8500-1
04-D8301
55-90077
D-8328
90074-57
(51b)
04-D-8453
0» Bxp-)
90074-57
(6U)
S-574
2355
Ex
2430
S-575
HF-DB
245S
CS-1316
04-39703/5
04-33141/1
80-D-8339
Fiber
Generic
olefln
amid
acrylic
aramld
acrylic
polyester
acrylic
acrylic
polyester
acrylic
polyester
acrylic
polyester
spun bonded]
acrylic
wool
acrylic
polyester
polyester
olefln
polyester
Trade
polyprop
Mo«x
Dralon T.
NOMJC
Dralon T
Allied 1-270
Kodel
Zefran
Orion
SOXDacron
TO Orion
75/Dacron
-75 Orion
Reemay
Orion
wool
llcrotaln
Racron
Dacron (55)
Thlofcol p.p.
unbranded
Type
fll.
fll.
at.
fll. &
Taa.
t.-2V
at. -IV
St. -2"
at.
t.
at.
St.
St.
St.
St.
fll.i at.
fll.
st.-lH"
et.-lV
Tarn
fll.
fll.W
sp.F
ap.
fll.W
tex.F
ap.
sp.
ap.
ap.
sp.
ep.
sp.
sp.
fll.W
ep.F
fll.
sp.
ap.
Construction
Weave
For
plain
d.f. twill
3x1 twill
d.f. twill
t nap
2x2 twill
1x3 twill
W/P 1-270
rodel
2x2 twill
(bonded)
felt
twill
twill
plain
2x2 twill
nap
(one aide)
Per*.
cfn/ft2 9
0.5" w.c.
7
S3
19
45
30
16.5
19
119
176
176
416
66
22
61
39
36
61
38
Electrical Propertlea
Resistivity (II /a)
as received | na waulied
t /0*-80*F/40-50IRII
> 2 x 101"
7 x 10"
2 x 1013
4 x 1013
2.4 x 10*
> 2 x 10'"
2 x 1010
> 2 x 10'*
> 2 x 10"1
> 2 x lO1*1
> 2 x 101*
> 2 x 101"
5 x 10"
> 2 x 101"
> 2 x 10l<1
> 2 x 10'"
2 x 1010
1.6 x 1010
~
> 10"-
> 101"
4 x 1013
4 x 10"
> 10'"
~
~
1.4 x 10'3
(either aide)
> 2 x 10'"
> 2 x 10'1*
> 10'*
> 10'*
Rel. Trlboelectric
Foa. & (decay rate]
70*P/50XRH
+ 0.6 (60]
- 1.7 130]
4 0.7
4 1.9 (20)
- 0.2 [40]
- 0.5 (30]
4 0.7 (35)
- 1.3 (40)
4 1.7
- 0.7 (30]
v 4 7
4 4.8 (45)
- 2.5 (30]
+ 1.4 (25)
(nap side)
Rub Voltages
vs nylon/vs Dorian
-7/+6
-10/+3
-3.4/+3
-7.6/+11
-8.3/45
-7.5/+A
-7.3/+6.S
-10.8/H
-2.7/+3.6
-6.6/+3.Z
+0.9/+6.5
-1. 3/49. 5
-7.5/+1.3
-10/411.6
00
-------
TABLE 4 (continued)
Fabric
Mo.
19
20
21
22
23
24
23
26
27
28
29
30
31
32
33
34
35
36
37
(con
ft.
IPS
AF
AF
Eld
AF
AF
AF
AF
Eld
JPS
JPS
JPS
JPS
JPS
JPS
JPS
AF
JPS
JPS
tlnued
Style No.
80-90067/71
S-580
S-1414
F4xR-stet.
B-Bu
S-2353B
S-1240
S-610 (826)
2470
S-632/54
S-401/38
4-38325/2
F-4143-00
(1402)
04-49832/15
04-90055/3
80-D-8339/1
S-1152
D-8514
4N-22B1/1-2
)
Fiber
Generic
polyester
polyester
wool/nylon
arnald
wool /cot ton
polyester
olefln
acrylic
polyester
(spun bonded]
fiberglass
fiberglass
a ran Id
araald
polyester
polyester
polyester
polyester
p.t . f ,e.
Trade
unbrandcd &
PVA
Dacron
TSxwool
'TS nylon
99X Nostex
/1 S. Steel
95.x polyester
J epltroplc
polyprop
Orion
Re cits y
Trltenp
Trltenp
Noaeii
Nonex
Dacron
Dacron T-55
T-54
unhranded
Darron
Teflon
Type
HI. -IS"
St.
St.
St.
St.
St.
at.
St.
St.
fll.
f 11.
fll.
St.
fll.
HI. /at.
st. -IS"
fll.
-
Ml.
Construction
Ysrn
sp.
sp.
sp.
sp.
ap.
sp.
sp.
ap.
fll.
Ml.
fll.
Ml.
ML.W
sp.F
sp.
~
f 11.
Weave
Form
3x1 twill
plain
plain
plain
non-woven
(bonded)
2x2 twill
3x1 twill
3x1 twill
felt
(needled)
3x1 twill
3x1 twill
2x2 twill
(singed)
3x1 twill
Pen.
cfsi/ft2 9
0.5" w.c.
24
68
35
120
75
46
26
75
101
48
18
31
29
17
24
~
76
21
Electrics! Properties
Resistivity (()£))
as received as washed
70'-80*F/40-50ZRH
3 x 10"
8 x 10»2
2.8 x 1012
1 x 108
1 x 101J
5 x 107
1 K 1010
> 2 x 101"
> 2 x 10'"
1 x 10"
5 x 10"
3 x 101"
5 x 10"
> 2 x 10'1"
> 2 x 10'1*
5 x 10"
> 2 x 101"
v 6 > 10"
> 2 x 10'1*
10'"
> 10'"
3 x 1013
--
--
> 10"-
~
~
Rel. Trlboeloctrlc
Pos. 6 [decay rate]
70*F/50IRH
4 0.4 (401
- 1.7 115]
< + 8 [20]
n.d. [100]
>. + 7 [95]
~
>. + 5
~
- 0.9 [201
~
--
v - 6 [0]
Rub Voltage
vs nylon/vs
-7. !/+'). ft
-11.4/+3.'
+1/+10
--
-f>.6/43
-
--
-
-
-H/-3.5
-------
TABLE 4 (continued)
Fabric
Ho.
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
Mfr.
JPS
JPS
JPS
JPS
JPS
JPS
JPS
JPS
JPS
Eld
AT
AF
JPS
JPS
JPS
D(A)
AF
AF
AF
Style No.
04-D-8328/1
04-90051-5
55-90078
55-90078/10
55-90062/50
80-90059/1
4-33106/1
04-49815/77
04-90108/1
T1056-95
(+ Zepel)
820 B
869C-15
80D-8339/1
04-90051/5
80-90059
659 NA
BB-BS
40/601
136B
Fiber
Generic
araild
raid
cry lie
acrylic
acrylic
mid
olefln
acrylic
acrylic
araald
acrylic
polyeater
polyester
aruld
araild
wool
acrylic
polyeater
Trade
NoMX-Taa
No«ex-T450
Dralon T
Zcfran
Orion, T-75
Ncnax-T450
polyprop
(Brt. Alamo)
Dralon T
Dralon T
Nomox
(+ Zepel)
Orion
Dacron
Dacron
Noaex
NOMX
wool
Darlan
Dacron
Type
fll./at.
at. -2V
t.-2"
at. -2"
at. -2"
fll.
fll.
fll. /St.
at.
at.
at.
St.
at.
Construction
Tarn
fil.U
ep.r
p.
ap.
ap.
sp.
fll.
fll.
fll.W
ep.P
P.
Weave
For*
twill
3x2 twill
3x2 twill
2x2 twill
3x2 twill
3x1 twill
2x2 twill
satin
felt
(needled)
twill
twill
felt
(pressed)
plain
felt
(needled)
felt
(needled)
Pena.
CfB/fl2 0
0.5" w.c.
59
25
35
28
42
36
20
28
11
40
61
94
86
35
43
49
32
44
Electrical Propertlea
Realatlvlty (l)£j)
an received ' aa waahed
70*-80*F/40-50XRH
3 x 1012
> 2 x 101"
2 x 1011
2 X 1011
3 x 1013
> 2 » 10U
> 2 x 10U
> 8 x 1011
> 2 x 101*
> 2 x 10I1(
» x 10l°
4 x 1010
2 x 10IN
> 101"
2 x 1013
> 101"
> 10»*
> 10" *
5 x 1010
Rel. Trlboelectrlc
Poa. & [decay rate]
70'F/SOZRH
+ 2.0 |20)
- 0.3 |25]
- 0.3 [60)
- 2.7 (15)
- 0.3 (50)
+ 0.4 (30)
- 0.8 [25]
- 3.7 [30]
Rub Voltage
va nylon/va Darlan
-8.3/+12.5
-10/+5.8
-8.4/-M.9
~
-6.5/+1
-7. SAM. 5
-9.5/+7.S
~
~
-12.5/+6
~
-14.7/+1
to
O
(continued)
-------
TABLE 4 (continued)
Fabric
No.
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
Mfr.
AP
AF
AP
AF
AF
AF
AF
PC
tit
AF
AF
AF
AF
AF
AF
AF
AF
AF
Style Mo.
-D-282
(H-70079A)
-D-1S4
40/894
40/306
40/804
E40/455
(antl atat)
E40/804
(antl etat)
Pot.
181 III
S-237
136B
B.W.
GIU
2005
8-225
-D-262
(H-82936A)
40/300
-D-602
40/449
Fiber
Generic
acrylic
wool
acrylic
acrylic
acrylic
polyeater
polyester
acrylic
polyester
amid
wool
polyeater
wool
wool
wool
wool
acrylic
Trade
Orion 42
wool
Orion
Orion
75x Orion
'TS Dae r on
Dacron
(4 a. a. fin.)
75^ Orion
''25 Dacron
(& a. a. fin.)
Kevlar
wool
Dacron
wool
wool
wool
wool
Orion
Type
St.
at.
at.
at.
St.
at.
nt .
--
f 11.
at.
at.
at.
at.
at.
at.
nt.
at.
at.
Construction
Tarn
--
fll.
HP.
ep.
ap.
ap.
ap.
Weave
For»
felt
(needled)
felt
felt
(needled)
felt
(needled)
felt
(needled)
felt
(needled)
felt
(needled)
singed
3x1 twill
felt
(needled)
plain
plain
plain
felt
(needled)
felt
(needled)
felt
(needled)
Pern.
cfa)/ft2 9
0.5" w.c.
28
30
23
47
29
52
19
53
31
46
44
54
46
Electrical Properties
Realatlvlty (a/Ot
aa received ' aa waahed
70*-80*F/40-50*RH
2 x 1010
4 x 1012
2 « 10IJ
8 x 1011
> 2 x 101"
v 6 x 10e
v 6 x 108
> 2 X 10"
10'"
10"
Eel. Trlboelectrlc
Poa. & [decay rate]
70*F/5OTRH
_.
antistatic |100|
antistatic [100]
% - 8 |45I
4 5.6 (75)
Rub Voltage
va nylon/va Darlan
--
-7.3/-10.7
-0.3/+8
N)
(continued)
-------
TABLE 4 (continued)
Fabric
Ho.
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
Hfr.
HSA-1
MSA-2
AT
AP
AT
AF
AF
AF
AF
Hb
Hb
R-t
R-T
AF
Core
Gore
lore
AF
Style No.
Reap. unt.
Reap. & ZnR
FG-fil
8-1*14
2348
868-B-6
960
S-1441H
S-1290D
H06309
H06602
R & H
Imp. nut.
R & H
lap. -glazed
2339
L10564
L1056S
L10566
0/601
Fiber
Generic
wool/acrylic
wool /acrylic
glaaa
7S, vool
'75 nylon
polyeater
polyester
cotton
91/-polyeatei
/5 a. ateel
polyprop
9S, polyeater
^f a. ateel
95>- polyeater
'S a. steel
acrylic
(f H-a.a.)
acrylic
(+ H-a.a.)
acrylic
polyester
(& Tef. Ian.)
araald
(& Tef. Ian.)
p.t.f.e.
(& Tef. Ian.)
acrylic
(dlnitrlle)
Trade
wool/acrylic
wool/acrylic
glaea
wool/nylon
Dacron
Dacron
cotton
9V Dacron
X5 a. ateel
9 5x polyester
'S a. ateel
9j>acryllc
^5 a. ateel
Draloo T
(& H-a.a.)
Dralon T
(4 H-a.e.)
Orion
polyeater
(& Tef. Ian.)
Nomex
(& Tef. laau)
Teflon lasi.
Darlan
Type
at.
at.
at.
et.
at.
at.
at.
at.
at.
at.
at.
at.
at.
at.
at.
at.
at.
Construction
Tarn
~
ap.
p.
ap.
ap.
ap.
ap.
ap.
ap.
Weave
Fora
felt
felt
plain
plain
(napped)
twill
satin
felt
(needled)
twill
(napped)
felt
(needled)
felt
(needled)
felt
(needled)
felt
(needled)
felt & 1««.
felt 4 Ian.
felt 4 IBM.
felt
(needled)
Para.
cf«/ft2 e
0.5" w.c.
103
67
30
33
79
33
16
39
73
62
26
21
22
28
~
35
Electrical Properties
Resistivity (n/b)
as received ' aa washed
70*-80'F/40-50XRH
10"
3.8 x 1011
3.6 x 10n
9 x 1012
6.6 x 1013
5 K 101*
6.7 x 108
2.1 x 10'
2 x 107
2 x 107
10»
109
1.3 x 101 10>*
> 10>"
, 10'"
Ral. Trlboelectrlc
Poa. & (decay rate]
70'F/SOXRH
+
very positive
* + 7 [80]
n.d. (100]
n.d. (100]
n.d. [100]
antlatatlc [100]
antlatatlc (100)
- 3.6 (55)
> - 5 (75J
Rub Voltage
va nylon/va Darlan
+1.3/+10
--
-1.6/+0.7
-0.8/-HJ.5
-14/+2.8
-2.3/-1
to
NJ
(continued)
-------
TABLR It (continued)
U>
Fabric
No.
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
Mfr.
AF
Eld
.IPS
AF
AF
AF
AF
AF
AF
Horn
Horn
Horn
AF
U-J
AF
P & S
P & S
AF
Style No.
136B
(Here.)
S.N. 529-75
(XT0954)
Ram.
S-1414X
S-1414P
S02B
40/300
40/894
60306 nat.
908 nat.
970 nat.
812
AW155V0335
S-197
U-.I
U-.l
802BC
Fiber
Generic
polyester
p. t.f .e.
polyester
16 oz.
15s wool
-^5 nylon
60/ wool
^40 nylon
nylon
acrylic
wool
wool
wool
wool
acrylic
polyester
polyester
polyester
polyester
nylon
Trade
Dacron
Teflon
75^ wool
"T5 nylon
60> wool
"40 nylon
nylon
Orion
wool (91)
wool (06)
wool (08)
Microtain
Uacron
Dacron
Dacron
Dacron
nylon
Type
St.
fll.
St.
St.
St.
St.
St.
St.
St.
St.
Ml.
fil.
St.
St.
St.
St.
Construct Ion
Yarn
fil.
sp.
sp.
sp.
sp.
sp.
Bp.
fil.
fil.
sp.
sp.
Weave
Form
felt
(needled)
felt
(needled)
plain
plain
plain
felt
(needled)
3x1 twill
plain
(napped)
felt
(singed
b.s.)
felt
(singed
b.s.)
twill
(napped)
Perm.
cfm/ft2 1?
0.5" w.c.
52
32
102
76
42
48
125
118
31
31
56
66
56
32
Electrical Properties
Resistivity (a/a)
as received as washed
70°-RO°F/40-50ZR11
> 10'"
> 10'"
1.4 x 1012
3 x 1013
3.7 x 10
5 x 1013
2 x 101"
2 x 101"
1.4 x 1012
3 x 1011
3.6 x 1012
9 x 10"
(used)
3 x 10"
(used)
1 x 10"
2.9 x 1013
> 101'1
Pel. Triboelectrlc
Pos. & [decay rate)
70°F/50ZRII
+ 1.5 [40]
+ 0.9 [95]
+ 5.7 [90|
v. + 8 [80]
+ 4.7
~
+ 5.5 [85]
+ 3.9 [80]
v. + 8 [80]
+ 0.4 [90]
pi. -0.2 [50]
nap -0.4 [60]
pi. -2.2 [25]
singed -2.8 (25)
pi. +3.7 [75]
nap +3.1 [75]
Rub Vnl lage
vs nylon/vs Marian
-12/+I4.3
-5.2/+5.1
-0.3/+11.9
+3.8/+I4.1
-2.3/+14.S
-0.9/+17.5
-3/+11.5
+2.5/+11.5
-1.9/+1.5
-8.S/+6.3
-15/+8.3
-13.5/+2.8
-15.5/+2.1
-3.2/+1I
-5/-U2.5
(continued)
-------
TABLE 4 (ciiTit I nurd)
ro
Fabric
No.
Ill
112
113
II'.
US
116
11?
ne
119
120
121
122
123
124
125
126
127
Mfr.
AF
AF
AF
AF
AF
Cnrb
Carb
Cnrh
AF
AF
JPS
Curb
F.Id
Eld
AF
ll-M
AF
Stylo Nn.
961
C868B
S-2283NRHH
S-2473 NBR
XD3766N
Cn-STOl-8
Ca-SA02-8
CH-ST02-8
B10AI
810LC
90112
Cn-SAOX
4057-18-2
4057-18-10
S-'>47
USS-H
C-H171.CF.
Fiber
Ci-nrrlc
ct»t ton
polyester
araald/glaas
polyester
glass
wool/acrylic
polyester
acrylic
polyester
acrylic
acrylic
acrylic
acrylic
nrnmld
aranld
acrylic
polyester
acrvl Ic
Trade
cotton
Dacron
tl, NOMX
'T3 glass
Vis polyester
>ll glass
5ft,» wool
TO acrylic
unbranded
ZeA507
unbranded
Hlcrociln
Hlcrotaln
ZeA507
Oralon T
(dyed 7)
Noncx
Noaex
Darlan
Hlcrotaln
Typo
St.
at.
at.
It.
at.
St.
at.
St.
at.
at.
at.
St.
Rt.
St.
at.
ft.
Construction
Yarn
sp.
sp.
sp.
sp.
sp.
sp.
sp.
sp.
sp.
up.
up.
sp.
Kp.
sp.
up.
Heave
For*
sateen
(napped)
twill
frit
(needled)
singed o.s.
felt
(needled)
cal. o.s.
Celt
(needled)
knit
(aeaaleas)
knit
(seaailess)
knit
(seamless)
twill
twill
(nnpped)
knit
(seamless)
napped
felt
(scr Iraless)
folt
(nnpped)
Pern.
cf«/ft2 9
0.5" w.c.
18
3ft
25
19
130
54
99
50
46
50
44
45
47
31
--
J7
'i7
Electrical Properties
Resistivity (0/0)
as received aa washed
70*-80*F/40-50ZRH
6.7 x 108
2.7 x 10"
1.9 x 1010
2.4 n 1011
5.9 x 10'°
3 x 1013
1.4 x 1013
1 x 10"
1.3 x 1012
3 x 10l)
--
1.2 x 10'°
* 101U
> in1'1
7 x 101"
2.H « in1'1
~
--
6.7 x 1013
~
~
* 10'"
* 10' *
> 10>*
~
Rel. Trlboelectrlc
Pos. & (decay rate)
70*F/50«H
n.d. (100)
- 0.7 |65|
pi. H .5 (40)
United +3.9 (40|
col. +1.9 (9U|
pl. «.4 [9fl|
-
+ 1.9 (75)
f 2.3 |90)
- 1.1 175]
nap +5.1 |90|
pl. +4.8
Rub VnltnRc
VB nylon/vn llnrlan
-S.6/+4.3
-10.1/+13
-7.5/+2H
-1 .'./+J.2
-2. <*/);./
-
-'>.1/H7.3
-
-4.1/»7.8
--
-0.4/+4. \
-0. 57+3.8
~
"
(continued)
-------
TABLE 4 (continued)
Fabric
No.
128
121
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
Hfr.
USF
USF
USF
USF
USF
USF
USF
USF
USF
Troy
W-Ac
W-Ac
W-Ac
AK
W-Ac
Gore
AF
Style No.
25-830-030
4A25-831
4B25-838
2A16-140
2B16-143
3A25-591
3B25-644
1A15-011
1A15-014
4582903-16
072-22p
F1L-11
Kernu-l
F1M)
MC-2-811L
SR8166
S-1414X
Fiber
Generic
aranld
aranld
aranld
acrylic
acrylic
polyester
polyester
polyester
polyester
polyester
--
acrylic
p . t . f . e . 1 am
on polyester
wool/nylon
Trade
Nooex
(4 Tef. fin.)
NOMX
Nonex
(& Tef. fin.)
HOMO acrylic
HOBO acrylic
(& Tef. fin.)
unbranded
unbranded
(4 Tef. fin.)
unbranded
(4 all.)
unbranded
(6 Tef. fin.)
Dacron
Mlcrota in
10'"
> 10'"
> 10'"
10'°
> 10'"
10"
> 10>"
10"
10' s
6.9 x I0'<>
6.9 x 1010
4 x 10'"
1.5 x 10n Tef.
10 '7 fell
2 x 10"
10'"
«
__
__
> 10'"
> 10'"
2 x 10"
3.6 X 10''
__
~
Rel. Trlboelectric
Poa. & (decay rate)
70*F/50XRB
_
._
_..
h. cover -2.4 |75]
plain -1.1 |80|
~
__
--
Rub Voltage
vs nylon/va Darlin
--
--
--
--
~
N>
Ul
(continued)
-------
TABLE 4 (continued)
Fabric
Pibsr
Construction
Electrical Properties
Ho.
Hfr. Style Ho,
Generic
Trade
Type
Tarn
Weave
Pom
Pera.
cfm/ft2 9
0.5" w.c.
Resistivity (n/b)
as received as washed
70*-80'P/40-50ZRH
Rel. Trlboelectrlc
Pos. 4 [decay rate)
70*P/50XRH
Rub Voltage
va nylon/vs Darlan
0\
144c
US
146
U7
148
149
ISO
151
1S2
153
AP
C-D
AF
UF
UP
UP
UP
Core
API
AFI
S-1414XG
S1472L
5-19-77
5-19-77-K!
R-C
-------
Particulate Resistivity
The apparatus used to determine the resistivity of particul-
ates has been described in detail by G. W. Penney.2 The dust
sample is equilibrated in the apparatus to an appropriate condi-
tion of dry and wet bulb temperature. The current obtained by
the impressed voltage is determined by means of an electrometer
and resistivity is calculated from the ohm's law relationship.
The pressure on the samples is approximately 4.75 g/cm2- Under
these conditions, the determined resistivity (ft-cm) is a function
of the applied humidity, temperature, packing and voltage. (Table 5)
Electrostatic Properties of Particulates
Particles become charged, in fact White3 states that it is
almost impossible not to charge particles in the course of normal
handling and processing. Whether the accumulated charge is posi-
tive or negative and will be retained for a significant time de-
pends upon many conditions - inherent properties, the nature of
the contacted material, the environment, etc. High resistivity
insures charge retention but an easy-to-conduct and reproducible
method for determining triboelectric (T.E.) locations is still
needed. Experimentally, Penney has employed an impingement tech-
nique in which silica dust was evaluated for charge polarity and
magnitude after being "bounced-off" eight different fabrics that
had already been located in the T.E. series. These results (Table
14) indicated that this silica is located at about a -3 position
in the scale of Table 2. The method offers significant promise
and deserves further consideration.
PARTICULATE AGGLOMERATION OR POROUS CAKE FORMING TENDENCIES
For its potential value as a change that may occur during
filtration, it is desirable to determine the relative agglomera-
ting or porous cake forming tendencies as well as density chang-
ing characteristics of particulates. Clues to such changes may be
obtained by a simple rolling test. In practice, the first stage
of the test employs an apparent density type of evaluation, con-
ducted by introducing a weighed and measured (usually 100 ml)
volume of dust into a liter, wide mouth polyethylene bottle. The
usual apparent density determination practice is used whereby the
graduate is gently and uniformly tapped on a resilient base at
each stage when the 25 ml, 50 ml, 75 ml, and 100 ml levels are
reached. This procedure is applied to the dust before and after
the rolling operation. The rolling may extend for any reasonable
time but a constant period of about 30 min (or less) at a moder-
ately slow speed (30 rpm) is preferred. Frequently, an increase
in density occurs but, most important, the tendency for agglomera-
2Penney, G. W. , AIEE Transactions. 7£ Section 1-201:1-3 (1951).
3White, H. J., Industrial Electrostatic Precipitation, Reading, MA,
Addison-Wesley (1963).
27
-------
TABLE 5. ELECTRICAL RESISTIVITY OF SOME PARTICULATES
CO
Particulate
Electric Furnace, U-D
Electric Furnace, U-D
Electric Furnace, U-D
Electric Furnace, U-D
Electric Furnace, U-D
Electric Furnace, U-D
Electric Furnace, U-D
Elec. Furn. , s.s., U-J
Elec. Furn., s.s., U-J
Elec. Furn., s.s., U-J
Elec. Furn., s.s., U-J
Fly Ash, WP-S
Ferromoly b-p . , C-L
Moly-Met, R-S
P-6140 resin
Steel Grinding Dust, R-C
Steel Burning Dust, R-C
Steel Burning Dust, R-C
(Lot)
(A)
(A)
(A)
(A)
(A)
(A)
(A)
(A)
(B)
(C)
(D)
(A)
(A)
(A)
(A)
(A)
(A)
(B)
Temp.,°F
70
240
200
150
125
100
85
77
77
73
73
73
73
73
73
78
78
78
RH, %
35
v. low
v. low
v. low
v. low
v. low
v. low
42
42
33
33
33
33
33
33
55
55
55
Applied
Voltage, V
1665
1665
1470
1470
72
1490
1480
1480
1000
1750
Amperage
I, amp
0.78 x 10"1*
0.65 x 10~"
2.30 x 10~5
2.20 x 10~5
1.35 x 10~2
0.59 x 10"1*
0.52 x 10"1*
< 10~12
4.20 x 10~7
10~6
Resistance
R, ohm
2.1 x 107
2.5 x 107
6.4 x 107
6.9 x 107
5.3 x 103
2.5 x 107
2.8 x 107
> 1015
2.0 x 107
2.4 x 109
1.8 x 109
Resistivity
p = 15.71 R
ohm cm
6.5 x 108
2.8 x 1011
9.3 x 1012
1.2 x 1013
5.0 x 1012
2.5 x 1012
1.7 x 1012
3.3 x 108
4.0 x 108
1.0 x 109
1.1 x 109
8.4 x 101*
4.0 x 108
4.4 x 108
> 1015
3.1 x 108
3.8 x 1010
2.8 x 1010
-------
tion implies that a similar change may be expected to occur during
filtration if a fabric of preferred electrostatic properties were
used. (Table 6)
VISUAL EXAMINATIONS OF PARTICULATES (IN THE COURSE OF FILTER TESTS)
Visible examinations of the particles are made as they appear
in the aerosol and as they appear after having been collected.
The condition of the particulate at both locations is recorded
photomicrographically (Figure 6). A sample of the incoming parti-
culate suspension is obtained by drawing it from the mainstream
onto the microscope slide in the settling chamber (F) of Figure 1.
In practice, the air flow to this chamber is stopped when the
microscope slide is inserted over the chamber port, the suspended
dust in the cloud over the slide is then allowed to settle, thus
providing a sample of the particles as they occur in the incoming
aerosol. At the header, the collected particulate from the filter
bag is also allowed to deposit on a microscope slide. Photomicro-
graphs of the two samples provide a record of the material at the
two locations and, by including a reference wire of known size in
the picture, an estimate of particle and/or aggregate size is
allowed.
REPRESENTATIVE AND EXPERIMENTAL FILTER FABRICS
Through the cooperation of fiber and/or fabric manufacturers,
a broad spectrum of filter media have been obtained for the exper-
imental studies (Table 4). This group of samples includes most of
today's practical filter media. Identically sized test bags were
prepared for use in the experimental collector. Separate swatches
have been cut (3 in. x 12 in.) for triboelectrification tests and
(6 in. x 8 in.) for permeability measurements. Resistivity data,
verified by triboelectrification studies, have shown that many
materials as-supplied had an antistatic finish. Several of the
fabrics, although made from different fibers, also carried the
same style number. Different kinds of acrylic fiber, for example,
seem to be used interchangeably without differentiation.
Variations such as these in fabrics offered to bag makers
and/or users, can have an adverse influence on filtration perfor-
mance. Accordingly, one of the purposes of this report is to
encourage fabric producers to add resistivity to the list of para-
meters included in filter fabric specifications and, of course,
to indicate clearly the kind of fiber(s) used in the fabric. These
characteristics of filter media are just as important as permeabil-
ity, weight, etc. and need to be included in the information that
relates fabric properties to performance.
EXPERIMENTAL FILTRATION STUDIES
In the course of shake-down trials conducted on the bench-
scale filter unit, a number of commercial particulates, including
29
-------
TABLE 6. AGGLOMERATION AND APPARENT DENSITY CHANGES OF SOME PARTICIPATES BY ROLLING
Particulate
P'K-3085 resin
P-6140 resin
Ferromoly. b.p., CM-L*
Moly-met, R-S*
Elec. Furn. , s.s., U-J*
S.S. Burn., U-H*
Steel Grinder, R-C*
Steel Burn. , R-C*
£ Steel Burn. , R-C*
Steel Burn. , R-C*
Silica, as rec'd., F-HH*
Silica, as rec'd., F-HH*
Silica, as rec'd., F-HH*
Silica, used, F-HH*
Silica, used, F-HH*
Silica, used, F-HH*
Silica, used, F-HH*
Silica, as rec'd., PG-C*
Silica, as rec'd., PG-C*
Silica, as rec'd., PG-C*
Silica, as rec'd., PG-C*
Tendency to
Agglomerate Weight
(Lot) Further g
(A)
(A)
(A)
(A)
(A)
(A)
(A)
(A)
(B)
(Oa
(A)
(B)
(C)
(A)
(B)
(C)
(D)
(A)
(B)
(C)
(D)
+
+
+
+
+
+
+
+
+
+
_b
_b
_b
_b
_b
_b
_b
_b
_b
_b
_b
38.7
46.7
131.8
90.5
24.6
106.6
170.0
87.5
83.5
80.0
120.4
124.2
118.9
119.4
136.9
132.3
121.1
115.1
116.2
119.7
112.0
Initial
Vol.
ml
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
Initial
App . Den .
g/ml
0.39
0.47
1.32
0.91
0.25
1.07
1.70
0.88
0.84
0.80
1.20
1.24
1.19
1.19
1.37
1.32
1.21
1.15
1.16
1.20
1.12
Final
Vol.
ml
88
92
80
73
92
86
94
71
78
60
106
108
102
110
108
105
105
112
107
107
105
Final
App . Den .
g/ml
0.44
0.51
1.66
1.24
0.27
1.24
1.82
1.22
1.07
1.33
1.14
1.16
1.16
1.09
1.27
1.26
1.15
1.03
1.09
1.12
1.07
^ App. Density
Change
%
+13
+ 9
+26
+37
+ 7
+16
+ 7
+39
+28
+66
- 6
- 7
_ 2
- 9
- 7
- 5
- 5
-11
- 7
- 7
- 5
tolling carried out in 1000 ml glass bottle
Tendency to deagglomerate evident
Source code
-------
ticulatt- in i
HO magnifications
RELATIVE SIZES OF PARTICULATES
i A molvMcnum b.ist-d lust-
part
Figure 6 Particulates Before and After Fabric Contact
31
-------
cement, flyash, shredded paper, and potato granules were examined
to determine the performance characteristics and reliability of
the equipment as well as to obtain information on the collectabi-
lity of the dusts. An industrial problem manifested especially
by poor fabric cleanability because of the substantial adhesion
between the paper particles and the collecting fabric was verified
experimentally and eliminated by making the particles, not the
medium, antistatic. A cationic treatment of the shredded paper,
an operation that might be carried out in the course of the manu-
facturing process, was shown to minimize the filter fabric clean-
ing problem.
Another serious problem experienced commercially in the fil-
tration of potato particles was reproduced experimentally and
found to be resolvable by heating the potato-dust aerosol to a
temperature at which its relative humidity was below 15 percent.
This particulate was shown clearly to be moisture sensitive. It
had to be handled and filtered under relatively dry conditions
in order to avoid particulate solvation, swelling, and tackiness.
The Collection of Flyash, WPP-S
The collectability of flyash (WPP-S) was found to vary con-
siderably depending upon the nature of the different test filter
media. Electrostatics, as well as construction features, provide
an explanation. The studies were carried out at essentially con-
stant conditions of dust loading, A/C ratio (5.4) and limiting
pressure drop (6 in. w.c.). Significant differences were found
among the eight tested fabrics in regard to collected solids,
plug weight, plugged cloth pressure drop, and leakage (Table 7).
Variations in performance as a result of electrostatic effects
are detectable when the influence of the fabrics' charge polarity
and charge intensity are recognized. In making the analysis, the
highly electropositive features of the wool/nylon fabric #21, for
instance, must be noted. Also noteworthy are the moderately high
charge but more electronegative properties of fabrics #40, #41,
#44, and #42, and the very electronegative characteristics of the
Teflon fabric (#37).
The fabrics (#40, #41, #44, and #42) in the middle region
(-2.7 to +2.0) of the triboelectric series and especially the
three that develop a high charge (#40, #41, and #42), collect the
largest amounts of the flyash before reaching the limiting pres-
sure drop of 6 in. w.c. Some of the high collectability indicated
by fabric #44, a filament polypropylene, however, must be attributed
to the fact that it leaked badly, never really forming an effective
filter cake.
It is interesting to speculate further concerning the effect
of electrostatic differences on the plugged cloth data shown for
fabrics #40, #41, and #42. These fabrics, each made from a dif-
32
-------
TABLE 7. EXPERIMENTAL FILTRATION OF FLY ASH, WPP-S
A/C - 5.4, APc limit = 6 In. w.c.. 75 - 85°F (^ 60% RH) shake cleaning
Fabric
No. Mfr.
40 JPS
21 AF
41 JPS
u
U)
15 AF
44 JPS
42 JPS
37 JPS
28 JPS
Style Yarn Perm.
90078 sp. 35
S-1414 sp. 35
90078/10 sp. 28
CS-1316 fil. 39
33106/1 fil. 20
90062/50 sp. 42
4N-2281/1 fil. 21
S-632/54 tex. 48
Rel. T.
E. Pos. Fiber
Total rub Volt. Type
-1-2.
20.
+8
11
-0.
15.
+4.
10.
-2.
7.
-0.
13.
*\/ -6
11.
^ +5
0
8
3
8
8
8
7
6
3
3
5
acrylic
Dralon T
75/> wool
^25 nylon
acrylic
Zefran
polyester
Dacron
Olefin
polyprop.
acrylic
Orion
Teflon
fiberglass
Filtration Parameters
Collected
solids, g
240
120
200
130
230
230
195
175
Plug Wt. Plug AP Relative
g in. w.c. Leakage
8 0.5 v. low
48 2.6 v. low
26 1.0 v. low
20 1.8 mod. high
17 1.0 v. high
16 0.8 v. low
16 1.4 mod. high
31 1.2 low
fil.
-------
ferent acrylic and quite similar in construction, provide distinctly
different plug weight-plugged cloth-pressure drop data that imply
a preference for a moderately electropositive media, like #40 at
+2.0 rather than the slightly electronegative media like fabrics
#41 and #42, both of which are at a -0.3 triboelectric position.
Most striking are the very significant and real differences in the
plugged cloth features of these three filter fabrics. The bag
made from fabric #40 retains about 8 g of flyash and shows a plug-
ged cloth pressure drop of only 0.5 in. w.c. compared to the much
higher values found for the other two acrylics. The comparatively
low collectability provided by the highly positive wool/nylon #21
and Dacron #15 and even the glass #28 fabrics that also show a
moderately high plugged cloth pressure drop and high plug accumu-
lation, also suggest adverse effects produced by unfavorable elec-
trostatic properties.
Following the preliminary trials, investigations were carried
out to determine the collection properties of several selected but
very different fabrics with a variety of polymeric and metallurgi-
cal dusts.
The Collection of Some Metallurgical Dusts
Particulate Emissions from Electric Arc Furnaces
The large electric furnace baghouses that control emissions
through an entire building evacuation system are quite prevalent
and, too often, present energy related if not serious operational
problems. The bags in these massive installations are large
(mostly about 1 ft in diameter and 34 or 35 ft long), their num-
bers run into many thousands (sometimes more than 6000) and they
handle as much as 1 3/4 million or more cubic feet of aerosol per
minute. Most often, cake removal from these bags is accomplished
by reverse air flow, a method that requires the use of an easy to
clean fabric, usually considered to be a filament-like medium.
A number of these electric furnace baghouses, large and small,
have experienced problems, some more serious than others. Exces-
sive pressure drop, poor cleanability, and abnormally high energy
demands have been reported.
The Collection of Electric Furnace Dust, U-DAt the very
large plant of U-D, 5472 - 11 3/4 in. diameter by 34 ft long fila-
ment bags collect the particulate from 1 3/4 million ft3/min
emissions of an electric furnace shop at a pressure drop of 8 to 9
in. w.c. This level of flow resistance is considered to be exces-
sive by ordinary filtration standards; a threat to high collection
efficiency, too energy intensive, and excessively costly to oper-
ate at the reported $3000/day. Bag life is very good, but this
excessive pressure drop and the attendant abnormally high energy
demands, suggested that an experimental filtration study should
be conducted to determine how other fabrics might perform and,
possibly, how electrostatics might influence the process. Three
34
-------
polyester fabrics of different construction were selected for the
first trials. One was made of filament type yarns, probably not
too unlike that used in the bags in commercial service. The other
two were made of staple yarns, one of which was napped. The re-
sults of the abbreviated study are summarized in Figure 7, and
describe the pressure vs collected solid features of the three
kinds of fabric through several cycles of operation at an A/C of
5.4 and to a limiting pressure drop of 6 in. w.c. Additional runs
were made and the results are considered later, but the data pre-
sented here show clearly that the napped polyester fabric #18 far
surpasses the performance of the filament yarn #15 and the un-
napped staple yarn #16 fabrics. Only the relatively high plug
weight (26 g) detracts from the outstanding collection properties
of this napped polyester filter medium. (The potential danger of
blinding problems in service must be considered.) It is also im-
portant to note that the fabrics of higher plug weight display
lower plug pressure drop values.
The superior overall performance of the napped staple fiber,
spun yarn fabric #18 and the better performance of the combination
filament warp/spun filling fabric #15 compared to the all filament
fabric #16 is noteworthy. While all three fabrics are polyesters
and of about equivalent permeability (between 36 and 39), the
significant differences in relative electrostatic chargeability
among these fabrics would imply that different polyesters make up
the three fabrics and, accordingly, that electrostatic polarity
and intensity may offer a clue to filter fabric variability and
therefore, to media specification. For example, on a triboelectric
scale of 10 with nylon located at +6 and Darlan at -4, fabric #15
is located at +4.8, fabric #16 at -2.5, and fabric #18 at +1.4.
Also important, perhaps even critical in this instance, is the
intensity of the charges that develop on the different media.
For example, fabric #18 charges triboelectrically to a total of
21.6 V (-10 V vs nylon and +11.6 V vs Darlan), fabric #15 to a
total of 10.8 V and fabric #16 to a total of 8.8 V.
Additional runs were carried out on fabric #18 (through 26
cycles), the same fabric (#18) with 0.69% add-on of conductive
graphite, fabric #20 (polyester, sp. 68 perm), fabric #31 (Nomex
felt, 29 perm), fabric #40 (Dralon T, sp. 35 perm), fabric #41
(Zefran, sp. 28 perm), fabric #40 and #41 very lightly hand napped,
and fabric #4 (Nomex fil. Taslan, napped 45 perm). Although the
graphite finished fabric #18 was lowered in electrical resistivity
to 107 n/f-i (antistatic), only slight (^ 15%) improvement in plug
reduction was achieved by this treatment. Hand napping was too
ineffective to alter significantly the surface characteristics
and performance of fabrics #40 and #41.
That the two acrylics, Dralon T and Zefran fabrics (#40 and
#41) perform differently despite similarities in construction is
evident from Figure 8. Although additional data for fabric #41
were obtained, these are not plotted in order to avoid crowding
of the curves. For this fabric, a plug weight of 16.5 g was found
35
-------
CONDITIONS - A/C = 5.4, APc limit = 6 in. w.c., 150°F, shake (moderate) cleaning
Run
Fabric
u>
4 -
0
LEGEND: No. No.
....... 4° 15
42 18
IIIIIIIIIIIMIIIIIIIIII
11 AH
Type Fiber Perm. g Total Rub Voltage
fil.W/fil.F Da 55/Da 58 & Si. 36 3.6 -2.5/8.8
fil.W/sp.F Da 39 9.7 +4.8/10.8
sp.W/sp.F p.e. & nap. 38 26.9 +1.4/21.6
]A i
m : /
Jl /
-»f - ^
M/
/" (
fx'*
i i
10 20
!3|/
;
\
%
/
/
/
/
»*
/
^%*
/
s
J
*
/
/
%
/
/
/
/
/
p% 1
1
/-
^f
*/£ N
^6.9j)
i i i i i i i ^
30 40 50 60 70 80 90 100
TOTAL COLLECTED PARTICIPATE, g
FIGURE 7. EXPERIMENTAL FILTRATION OF ELECTRIC FURNACE DUST, U-D (I)
-------
CONDITIONS - A/C = 5.4, APc limit = 6 in. w.c., 150°F, shake (moderate) cleaning
LEGEND:
Run
No.
45
43
48
47
No.
40
20
4
41
Fabric
Type
sp.W/sp.F
sp.W/sp.F
fil.W/tex F
sp.W/sp.F
Fiber
Dralon T
Dacron
Nomex
Zefran
Permeability 1
cfm/ft^ @ 0.5 in. w.c.
35
68
45
28
Plug W'
.
6.7
38.5
55.6
9.6
t. Rel. l.t. rob
Total Rub V
+2/20.8
-1.7/14.9
+0.7/6.4
-0.3/15.8
Ul
4
< 2
%
CO
0
10
20
FIGURE 8.
30
80
90
40 50 60 70
TOTAL COLLECTED PARTICIPATE, g
EXPERIMENTAL FILTRATION OF ELECTRIC FURNACE DUST, U-D (II)
-------
after 10 cycles when the collected solids reached only 78.2 g.
Thus, Zefran performed much less favorably than Dralon T and of
course, neither fabric #20 (sp. Da) nor fabric #4 (fil. Tas Nomex)
performed better than the Dralon T that showed a relatively low
plug weight (6.7 g) and reasonably good collectability.
It would appear that if a napped fabric like #18 or possibly
a napped variation of #40 would not become blinded, or if mechan-
ical cleaning could be applied to limit plug concentration in com-
mercial service with this electric furnace dust, an appreciable
lowering of the bag pressure drop could be achieved to realize
very significant savings in fan energy and thereby operating costs.
According to earlier speculations, much more reliability in
selecting the preferred fabric to collect electric furnace dust
should be allowed if the triboelectric location of the particulate
were known. If, as suggested before, agglomeration at the col-
lecting fabric surface is encouraged more and more as the charge
polarity difference and intensity increases between the two mat-
erials, then knowing the T.E. series location of the dust as well
as that of the fabric should permit considerable accuracy in spec-
ifying the preferred fabric. With this added information, it
would be relatively easy to determine whether increased charge
intensity was needed. Construction changes for more or less bulk
to alter charge intensity in an appropriate way, or other changes
to influence charge bleed-off as needed, could then be determined
rapidly with certainly.
The Collection of Electric Furnace Dust, C-TFrom a limited
experimental filtration study, it becomes apparent that the fila-
ment like polyester fabric used at the C-T steel plant collected
electric furnace dust in an inferior manner compared with several
other fabrics (refer to Table 8). The cake that forms on this
fabric, made of fine staple polyester fiber on the cotton system,
tends to be highly resistant to air passage. Such a deposit leads
to a rapid rise in pressure drop and to short cycling. On the
favorable side, the deposited dust is removed quite readily from
the relatively smooth-surfaced, low charge fabric, suggesting that
reverse air cleaning should be effective. This feature is indi-
cated in the data but again another very important observation
regarding the plugged cloth pressure drop should be noted. Impor-
tant variations occur between the plug AP values of the filament-
like plant-used polyester medium and the other, especially the
acrylic, media. Even though the quantity of plug (the dust re-
tained on the plant-used polyester fabric after reverse air clean-
ing) is low, the resistance of this residue to air transfer is
greater than that recorded for the very highly plugged (weightwise)
acrylic fabrics. Whereas only four grams of plugged dust in the
plant-used medium produces a pressure drop of 3.5 in. w.c., as
much as 84 g of plugged dust in the acrylic fabric produce only a
2 in. w.c. resistance. These differences in plug as well as the
variations in the collected dust in and on the two types of fabric
indicate that the deposits differ substantially in porosity. It
38
-------
TABLE 8. EXPERIMENTAL FILTRATION OF ELECTRIC FURHACE DUST, C-T
6. APc llait - 6 in. v.c., 130*-140'F. simulated reverse air cleaning
Fabric
No.
145
18d
120
3c
3f
146b
146a
Hfr.
JPS
AF
JPS
JPS
AF
AP
Style
(C-T)
80-D8339
810LC
55-90077
55-90077
t C
S-14J2L
8-1472L
Fiber
p.e.
p.e.
acrylic
cry lie
acrylic
& cond.
graph.
acrylic
acrylic
Yarn
fil.tt
ap.F
ap.
ap.
ap.
ap.
fll.W
sp.r
(11. U
ep.F
Weave
3x1 tu
(c.e.)
2x2 tvt
napped
2x2 tu
<«.».)*
napped
3x1 tu
(o a)1
Jxl tu
(«. a.)1
3x1 tu
(c. a)1
3x1 tu
Permeability
cfn/ft2 g 0.5"
15
38
50
19
19
40
40
Resist.
n/a
> 10'"
3x10' J
> 10'"
2*10"
2xl07
> 10'"
> 10'"
Rel.
Triboelectrlc
Position
*
+ 1.4
- 1.1
- 1.7
- 1.7
(antlatat)
- 1.0
-1.0
Filtration Parameters
20 aec rev-air cleaning
Cake
Wt., g
0.5
2.1
5.4
4-5
4-5
3.6
4.5
Plug
Wt., g
40
high
84
38
25
54
AP in uc.
3.5
2.1
2.0
2.5
1.9
1.5
36 1 1.3
1
40 aec rev-air cleaning
Cake
Wt., g
2.8
5.5
4.5
4.8
aide to
5.0
it aide
Plug
HE., g
59.5
89
44
46
1 ,.
AP in. w.c.
1.5
1.5
l.H
1.0
1.2
Relative
Leakage
mod. high
mod. low
low
low
low
mod. low
mod. lav
1
VO
'cotton system
*voolen aye tee>
-------
is suggested that favorable electrostatic properties (polarity
and intensity) contribute to the porous deposits (plug and cake)
in and/or on the acrylic fabrics and that the absence of these
preferred electrostatic features lead to the compact, highly re-
sistant deposits in and on the plant-used, filament-like polyester
medium.
If, after collectability, the plug pressure drop, not the
plug weight, were the only other parameter of concern, it is evid-
ent that any of the acrylics, particularly fabric #146af might be
preferred for optimum performance. However, without information to
indicate whether or not the residue might change phase or otherwise
develop into a blinding type of plug to close-up the fabric, it
would seem advisable to avoid excessively high levels of plug. Ac-
cordingly, more vigorous cleaning action (than reverse air) would
appear to be the safest practice to adopt if a change to one of the
high bulk acrylic fabrics were made for the collection of the C-T
electric furnace dust. Combining shaker-type cleaning with a medium
such as 3c, for example, would be a most reasonable recommendation
for collecting this particulate commercially in an effective way.
The Collection of Electric Furnace (Stainless Steel) Dust/ U-J-
Prompted by reports of collection problems with a stainless steel
electric furnace (ssef) dust possibly related to electrostatics
and reflected specifically by abnormally high pressure drop in an
industrial, shaker-type baghouse, a filtration study was carried
out. An evaluation of at least twenty (20) different media was
carried out. Initially, comparisons were made of the more popular
and relatively standard filter fabrics, but later other woven fab-
rics were examined together with knitted materials. The results
indicated that the moderately electronegative and high charge in-
tensity media, such as napped acrylic, napped cotton and napped
polypropylene (in this order for increasing effectiveness) per-
formed better than a variety of relatively electropositive poly-
ester, nylon, and wool fabrics. Better performance was judged
largely by better collectability as the test bags reached a pres-
sure drop limit of 6 in. w.c. under equilibrium conditions at an
A/C of 5.4.
Among the fabrics that performed best, the acrylics (Zefran
or Dralon T) were considered to be the preferred practical choice
for industrial service, inasmuch as problems could develop with
the otherwise effective but more heat sensitive polypropylene and
cotton media. However, when a 7.25 oz twill weave acrylic medium
was recommended for use in the stainless steel electric furnace
baghouse, the equipment vendor insisted that a knitted version be
considered and samples were submitted for study. A Zefran knitted
medium as well as a polyester knitted filter fabric were supplied
and evaluated, filtrationwise, with the ssef dust. The knitted
acrylic collected quite well and was found to surpass the perfor-
mance of a polyester knit fabric. However, it was not acceptable
because of excess leakage. Accordingly, a new, lower permeability
(< 30) knit fabric was requested. Such a fabric was received as
40
-------
a 10+ ounce knit material. This bluish-grey knitted fabric dis-
played extremely high electropositive properties, essentially as
positive as nylon and definitely not like Dralon T or Zefran in
the triboelectric series. After proving in filtration tests that
this was not a favorable medium for collecting ssef dust, it was
rejected for use in the plant baghouse in deference to an acrylic
medium, specified preferably as Dralon T for optimum stability,
with the following other properties:
weight -v 7+ oz/yd2
fiber staple
yarn spun
weave 3x1 twill
finish napped one side (dust side)
permeability . . . 15-25
The U-J baghouse collecting the emissions from this stainless
steel electric furnace was reclothed with the recommended medium
in May 1977 and has been functioning quite satisfactorily.
While the above' specifications indicate that the preferred
medium is one made on the cotton system, it should be apparent
that a similar but heavier material made on the woolen system may
be used, especially if durability and efficiency problems should
develop in the shaker-type industrial baghouse.
The Collection of Stainless Steel Burning Dust, U-HThe
emissions from a stainless steel burning (ssb) operation at the
U-H Steel plant are evacuated to a baghouse. Serious problems in
regard to both dust leakage to the atmosphere and excessively high
pressure drop in the collector brought the issue to the attention
of the writer.
In the course of the batch-like operations, the vendor-installed
polyester bags showed a clean cloth pressure drop of about 2 in. w.c.
at start up in a Torit collector. Soon (1 to 2 min) after the emis-
sions entered the plant baghouse, the pressure across the bags in-
creased to 10 - 12 in. w.c. and then progressively but more slowly
during a subsequent 20 min period to 16 - 18 in. w.c.
The normal operating temperature (^ 100°-120°F) is not exces-
sive for a number of available filter media but a filament-like,
light weight, combination yarn, polyester fabric with the filament
side to the dust was supplied by the equipment vendor. The ssb
dust is fine, relatively low in apparent density and magnetic, but
visual examination and the rolling test indicated that it does
agglomerate.
41
-------
After the intermittent collection cycles of the baghouse, bag
cleaning is applied by means of a 90 psi air jet employing six
pulses. No information was provided concerning the A/C or dust
loading conditions. The results of the initial filtration study
carried out at A/C of 5.4 to a limiting pressure drop of 6 in. w.c.
are summarized in Table 9. These data indicate that most of the
tested media, regardless whether of high, low, or medium permeabil-
ity, operated for only 30 to 40 sec, compared to the 20 sec cycle
experienced with the vendor-supplied medium. Only a Teflon web
(micro porous) laminated fabric (Gore-Tex L10564) performed sig-
nificantly better. The filtration time for this fabric, all other
parameters essentially the same, increased to 115 sec for an im-
provement of almost 500 percent.
Recognizing that even this extent of improvement might not
be adequate, further examination of the fabrics was carried out
at an A/C of 1.8 with the other parameters the same as before and
reasonably constant. As expected, significantly longer filtration
times were recorded to reach the same (6 in.) pressure drop with
the filtration cycles extending from 20 to 25 min for the vendor-
supplied and the usual media, and over 30 min for the microporous
Teflon surfaced fabric.
On the basis of these limited data, the following recommenda-
tions were made at that stage of the study:
lower the A/C ratio (increase capacity) substantially to 2
or less,
provide an extra increase in capacity to have each compartment
off-line for cleaning during the filtration period, and
install the microporous Teflon surfaced (Gore) bags.
An examination of the data raises some interesting questions.
For example, why does a low permeability fabric like #25a pass the
stainless steel burning dust so seriously that the filtration test
could not be continued while a relatively open (101 perm.) spun
bonded fabric retained this dust and, in fact, collected it at
least as well as the vendor-supplied, filament/spun yarn medium?
Why too did the wool/nylon fabric, a relatively bulky material and
very electropositive material with considerably fiber cover also
leak badly? Why did the Teflon laminate-surfaced fabric collect
500 percent better than the vendor's fabric at the 5.4 A/C, but
only 50 percent better at 1.8 A/C?
Obviously, the usually considered physical variations of the
test fabrics provide no answers. If, however, the high intensity,
very electronegative electrostatic charge features of Gore-Tex
preferentially contribute to the formation of a porous cake; and
if aerosol velocity and charge generation are proportional, then
one reasonable explanation is available.
42
-------
TABLE 9. EXPERIMENTAL FILTRATION OF STA1NLF.SS STEF.I. RURNINC DUST, U-H
(A/C S.t, AP limit " 6" v.c. - roon temperature (i 70'F). duet loading x 20 uratna/ft3 - puUe-llkc cleaning)
Run
No.
141
142
144
14*
140
Fabric
No.
12«,
97
2Sa
27b
12?
lllb
Hfgr. style No. Ynrn
I 1
Vendor (T»rlt) tnat.
7 Hupp lied hy U-ll
tcel
1
AF
rid
AF
AF
S-IM4P
C-ni7LCK
9M
P.
f 1 1
«P.
«r.
WC.WC
fll. W.p.F
1x1 cwlll
c...'
plain
V.,.'
C . « , '
txi twin
nnpped
w...1
not tn
nnpped
1* 1 rul 1 1
b(i»c w.«. '
Rrnlflt . '
n/o
7x1(1 "
1X10"
2x10'''
2.8x10'''
*.7«10"
Prrmeahlllty
cfi»/ft2 » 0.5" v.c.
17
7f>
26
ini
47
U.5
jq 3
Fiber
Generic
polye»ter
7> v°°»-
T5 Nylon
pol yoRtpr
nc rvl Ic
ml ton
Trade
j
Wool /Nylon
Reemay
Htcrotatn
rot ton
For*
Ml. t. St.
(fll, to duat)
t.
at .
at
(nnp to rtttdl )
Filtration Dat.n
A/C -5.4
Filter
Cycle-Sec.
111-20
.«
sn
30
]n
Re).
Leakage
Hod.
tow
I^W
1 nw
A/C 1.8
Filter
Cycle-Mtn.
20-25
22-24
22-24
__
Bel.
MnrJ. Low
Low
tow
_-
LJ
'/it 50« Ml. 70'F
'cotton »y»tr-ni 'wool
-------
Further studies were carried out to compare the vendor-
supplied filament-like polyester fabric and the microporous Teflon
surfaced fabric (Gore-Tex L10564) at different dust loadings at
both the 5.4 and 1.8 A/C ratios, mostly to a pressure drop limit
of 6 in. w.c., but in some trials to a 3 in. w.c. limit. The
aerosol temperature was also increased from the previous 65°F to
180 °F in these later studies, in an effort to reproduce more ex-
actly plant conditions.
The results are summarized as follows:
at an A/C = 6 and particulate loading of ^ 0.5 to 0.6 gr/ft3,
the Gore-Tex collected two to three times more dust than the
vendor's polyester up to the 6 in. AP limit.
at an A/C = 6 and particulate loading of ^ 1 gr/ft3, the
Gore-Tex collected about six times more dust than the vendor's
polyester up to a 6 in. AP limit.
at an A/C = 6 and particulate loading of * 2 to 3 gr/ft3, the
Gore-Tex collected about ten to eleven times more dust than
the vendor's polyester up to a 6 in. AP limit.
at an A/C = 6 and particulate loading of ^ 7 to 8 gr/ft3, the
Gore-Tex collected about 30 to 40 percent more dust than the
vendor's polyester up to a 6 in. AP limit.
,
at an A/C = 1.8 and particulate loading of * 6 to 7.6 gr/ft3
the Gore-Tex and the vendor's polyester collected about the
same quantity of solids up to a 6 in. AP limit.
at an A/C = 1.8 and particulate loading of * 1.5 gr/ft3, the
Gore-Tex collected over 80 percent more dust than the vendor's
polyester up to a 3 in. AP limit.
The plugged cloth pressure drop of the Gore-Tex fabric re-
mained lower, consistently, than that of the vendor's polyester
and ranged from 0.4 to 0.6 in. w.c. (APp) , compared to the 1.4 to
2.5 in. w.c. (APp) value found with the vendor's fabric when used
at the 6/1 (A/C ratio) velocity. At the 1.8 A/C flow rate, the
APp values were 0.1 to 0.2 in. w.c. for the Gore-Tex and 0.2 to
0.8 in. w.c. for the vendor's filament-like polyester filter.
These data substantiate but extend the basic conclusions
reached earlier regarding the relative effectiveness of the two
fabrics. Only at the low A/C flow conditions with moderate dust
loadings do the advantages fall off for the Teflon microporous
membrane-coated fabric in collecting this stainless steel burning
dust. There is also an indication that an optimum level of (ssb)
dust loading exists for this unusual fabric in the region of 3
gr/ft3 at the 6/1 A/C ratio flow condition.
44
-------
It is desirable, if not important, to attempt an explanation
for the superior performance of Gore-Tex as a filter medium for
the ssb dust. Acceptance of the screening-type action indicated
for this fabric in the EPA 600/2-76-168c (December 1976) report is
certainly plausible, but this does not go far enough. If only
screening were the filtration mechanism by which the microporous
Teflon membrane fabric functioned, such fabrics would become univ-
ersal, all-purpose filter media. All, or at least most dusts,
therefore, should be collected in an effective and efficient man-
ner by Gore-Tex. That this is not the case clearly suggests that
other forces are at work.
Assuming the screening action to be predominant in Gore-Tex
filtrations and that particulate agglomeration is also a common
factor in the process (although it is variable and dependent upon
the agglomerating tendencies of each aerosol), it seems likely that
the electronegative and relatively high charge (-3.6/16.8 V) and
moderately high discharge rate (^ 55%) features of this Gore-Tex
fabric can play a significant and possibly controlling role in the
filtration process.
According to my observations, the easy-to-agglomerate dusts,
especially those located away from Gore-Tex in the triboelectric
series, should respond to the electronegative Teflon surface by
forming aggregates. Obviously, these larger agglomerated particles
would be easily sieved out of the air stream, but because the rate
of charge loss from the fabric is high, the collected solids would
not be retained on the slick surface. As a result, some particul-
ates are more effectively filtered than others by Gore-Tex. (A
relatively clear indication of such differences is provided in the
filtration results provided from the study of the two polymeric (P)
dusts, referred to later in this report.)
The variations in filtration performance of Gore-Tex occurring
as a result of different ssb dust loadings and different velocities
are not so readily explained. Why, for instance, should Gore-Tex
be so superior to the vendor's polyester at one level of dust load-
ing and not at lower or higher loadings? Do the screening, charge
generating and charge transfer features contribute significantly
here too? A case could be made for velocity related frictional
changes with resultant differences in charging and for particle-
to-particle contributions as a function of concentrations and dust
loading. Obviously, more study is needed in order to arrive at an
acce'ptable explanation.
The Collection of a Ferromolybdenum By-Product Dust, CM-L
The filtration study of this (fmbp) dust was carried out in two
phases. The results are considered here essentially as they were
reported at the end of each stage to demonstrate that in the course
of any such evaluation a representative variety of media should be
examined and also that more than the common physical properties of
the media and particulates must be considered. Reliable interpre-
tations of performance differences cannot be accomplished with just
45
-------
the usual information on fiber type, fabric construction, perme-
ability and weight together with the usual particulate parameters
of size and chemistry.
For the first five fabrics evaluated for collectability with
this dust, as reported in Table 10, the major differences in per-
formance can be attributed as much to construction variations
(i.e. filament vs spun yarns) as to differences in electrical
properties. The three spun yarn fabrics #102, #120, and #141,
whether of 50 or 21 permeability, provide high collection effi-
ciency mostly because they retain a high level of dust (after
cleaning) that serves as a filter-aid. The filament yarn fabrics
#37 and #44, on the other hand, retain less dust, are inherently
less bulky, and leaked the dust quite seriously throughout the
entire filtration cycle, especially at start-up. Observations
such as this are not uncommon and the explanation is usually
credited to the obvious physical differences. But the porosity
differences in the non-removable (by the adopted cleaning method)
dust (plug) in and/or on the two types of fabric must contribute
to the differences in flow resistance. That these differences
are significant is quite evident. Although the filament type
fabrics retain one tenth or less as much plug as the spun yarn
fabrics, they exhibit about the same air flow resistance as shown
by the plugged cloth pressure drop. Obviously, the dust residues
in the two types of media differ in porosity.
In subsequent filtration studies of the ferromolybdenum by-
product dust (fmbp), five other fabrics were evaluated. The re-
sults are summarized in Table 11. From the results of these
trials as well as from the rolling tests carried out on the dust,
it became evident that this fmbp particulate agglomerated quite
easily, especially in contact with the more electropositive, high
charge intensity fabrics. As a result, the media that perform
best are those that provide these properties. The electropositive
fabrics #102, #144, and #111 perform well despite high electro-
static discharge rates, whether achieved by means of a conductive
(graphite) finish or provided naturally (cotton). Quite possibly,
the highly electropositive media are more effective in promoting
agglomeration of this dust than those that are electronegative.
This would suggest that this relatively easy-to-agglomerate ferro-
molybdenum by-product dust should be located about midway or at
an electronegative position in the triboelectric series. Apparently,
so long as the fabric used to collect this dust is quite electro-
positive, agglomeration proceeds well and collectability is quite
good.
The addition of a high discharge rate finish [conductive
graphite] does not detract from the performance of an otherwise
favorable (electropositive) fabric, #144. On the other hand,
neither the slightly negative material (acrylic as in #120) nor
the very electronegative fabric (Teflon #37) performs well, despite
favorable construction features. However, when the electropositive
46
-------
Run
No.
154
155
156
157
158
150-
TABLE 1C. KXPEKI MENTAL FILTRATION OF A FERROMOLYHDENUM BY-PRODUCT DUST, CM-L (I)
(A/C = 6, Al'c: Jim it = 6 In. w.c.. nO"-l40°F, 3-5 sr/ft3 loading, shake- cleanlnu)
1'abric Filtration Parameters
No. _
120 r
12U
h
Mir.
AF
AK
AF
.IPS
.IPS
Hum
norrnnl cover stirt
Style
8 1 Ol.C
8IOI.C
MC2-8I 1L
4-33106/1
4N-2281
ftOlOb nut
bna
J>EH._
wov. , sp.
wov. , sp.
napped
wuv . t sp.
TFF. fin.
wov., fll.
wov. , fil.
wov. , sp.
Tiber
aery 1 ic
(mlcroLain)
aery lie
(roicrotaln)
acryl ic
(microtain)
polyprop
Teflon
wool
Permeability
dm/ ft2 @ 0.5"
50
50
21
20
21
50.5
K>;1. Trlbiielertrlc
Position
- 1.1
- 1. I
- 2.4
- 2.7
»- a.o
t- 5.5
Rute of
Cbargu Loss
Z (2 uln)
75
75
75
10
0
85
Collected
Participates, g
x- 16
t 16
< 12
i- 23
> 12
>. 16
Plu
AP
in. w. c .
0.8
1. 1
1.4
0.7
0.9
0.6
a
Ml.
£
.,. 45
v 76
>. 56
4.5
4.6
>. 45
Relative
Leakage
low
low
low
v. high
v. high
low
TAI1I.L 11. LXI'l.KIMIiNTAl. FILTRATION OK A FERKOMIU.YBDKNUM BY- PKOIHICT DUST, CM-I. (II)
. J.A./1: " 6 t" Al'c limit of t .ijicbes w.;:., in. l.'»(.r.vl*!i''ii..J-.5 Bl'/J'.! J .l^S'lJUlii t^.*. dv:!!!
Run
No.
I5S
160
161
h,2
161
No.
102 a
I2U h
120 1
I44f
Mir.
Horn
AF
AF
AF
AF
Sl_^ 1 c.
t.inou ii. a .
HMII.C (,
a.s. (ll)
81 Ol.C &
:,.,. (.;)
S-I4I4 &
a.s. (C)
S- 141*
wovei^ s tail If
woveiv slap I f
t. Downy
woviiu, staple
& conJ.
woven, staple
& cond.
,;ra|.l,ite
wovciv ula|>le
napped
Fabr ic
Fil.el-
uool
iii-ryl If
.i.iryl Ic
75, wool
/L"i nylon
75, woo|
^25 nylon
Pel liu-ali 1 lily
liii/fl1 (J 0.5"
50.5
bl)
50
50
50
1 7 5
Rfl. Tr
Position
1 5.5
i. - 1
- - 1
very 1-
very +
IboeK-i-l r Ic
'll.ill: of
i:li.-||-Bi- Loss
Z (2 mill)
85
v. hlKl,
V. l,lKli
V. hiu.1,
80
Fill r.i
Part i < ii 1 al r:; , £
- Id
-.. i,
< II
... 2,
19
Ion Pa I', nil
I'll
In. w.c.
(I.I,
2 . '
1.0
0.5
O.'J
e 1 e r s
\ ..
Ml .
B
- 45
. HO
60
>. 40
. 65
lu-lal Ivi
l.i'ak.i^e
low
low
low
low
lou
-------
characteristics are satisfied, then suitable construction and high
electrostatic discharge rate features seem to further improve per-
formance. [It should be apparent that these results substantiate
other evidence indicating how antistatic qualities (i.e. conduct-
ivity through graphite finishing) do not eliminate charging but
only cause the charge to bleed-off rapidly. The Downy (antistatic)
finish would seem to impart tackiness that detracts from any favor-
able charge bleed-off feature that it might convey. Of course, it
could not provide such properties at the elevated filtration temp-
erature. ]
The direct comparison provided by runs 161 and 162 (Table 11),
indicating 100% better collectability by fabric #144c compared to
fabric #120i, with both fabrics at the same permeability, tends to
make a strong case for the influence of charge polarity and inten-
sity, especially since, on a basis of weaves, the plain weave fab-
ric (#144) should not be quite as effective as the twill weave
fabric (#120).
The #44b fabric, a light-weight filament polypropylene (refer
to Table 10), provides a high level of collectability but leaks
seriously. The performance differences between this fabric and a
negative (much more negative at *> -8) Teflon fabric, also in the
filament construction, is not so easily explained unless the dust
is located at a position in the triboelectric series quite near
Teflon. Also of possible but lesser importance is the very low
rate of charge decay of this fluorocarbon fabric.
The questions arising from these studies tend to be more
numerous than the answers. It seems clear, however, that without
such evaluations, no fair indication of filter media preferences
can be given and, further, that construction parameters are not
the only criteria upon which filter media must be specified.
The Collection of a Steel Grinding Dust, R-CAn effort was
made, actually without even nearly adequate success, to collect
the particulate products from a billet grinding operation by means
of a pocket type filter. At least three different media were
tried in the commerical collector designed for an A/C ratio of 3.9.
All failed because they could not be cleaned of the tenaciously
adhering dust by the relatively mild (^ 3 psi) reverse air. After
a conventional polyester felt (^ 14 oz) did not respond to the
reverse cleaning, a similar polyester felt with an exceptionally
smooth (glazed) surface (permeability = 16 cfm/ft2 at 0.5 in. w.c.)
was tried. Even this extremely smooth surfaced fabric did not
release the grinding dust and, as with the conventional felt fil-
ter media, the pressure drop in the baghouse increased so exces-
sively (» 10 in. w.c.) that the grinding emissions could not be
collected and had to be vented. In the course of the field trials
a relatively light weight (^ 6 oz) woven acrylic fabric was tried.
Although somewhat better performance may have been realized, even
though moderate at best, this fabric was also discarded after only
48
-------
limited service because of excessive fabric wear and inadequate
cleanability.
Experimental filtration trials of the commercially tried -
and several other - media, reaffirmed the poor cleanability of
the grinding dust (refer to Table 12). Under the test conditions,
the ultrasmooth polyester felt that failed in the commercial trial
collected only 3.6 g of the grinding particulate during the very
short (20 sec) filter test to the 6 in. w.c. pressure limit. All
of the other trial media performed better and the more practical
fabrics collected 8 to 10 times as much dust before reaching this
limiting pressure drop.
While the better performance of bag #143 (Gore-Tex) might be
attributed to the slick microporous Teflon surface, it should also
be noted that this fabric is very electronegative and also develops
a high level of charge in the triboelectric tests. Furthermore,
if surface slickness were the all-controlling factor, the poor
behavior of bag #148 and better performance of fabric #113e (singed
87/13 Nomex/e fiberglass felt) would be difficult to explain.
Actually, singing of this fiber blend burns out the Nomex cover
fibers to leave a high concentration of glass on the surface.
This causes the electrostatic charge intensity to increase by 50
percent to an extremely high value and, at the same time, raises
its position in the triboelectric series (refer to Table 2) from
^ 1.5 to ^ 3.9, reflecting the presence of a high concentration of
the electropositive glass fibers. The relatively good behavior,
collectionwise and in low plugged cloth pressure drop, of the
highly electropositive wool fabrics (bags #58 and #75) also sug-
gests that electrostatics influence the collection of this magnetic,
moderately conductive (resistivity = 3 x 108 fi-cm)dust. Incident-
ally, despite this noted conductivity by the bulk method referred
to on page 11, a substantial layer of the collected particulate
on the filter fabric provided only a very high (> 10* ** n'f~)) measure
of resistivity by the square method (page 10).
Although it was conceded that more study was needed, an immed-
iate decision had to be made regarding filter medium requirements
and even more basically, whether the pocket type baghouse was
indeed a viable system for collecting this and a burning dust too.
Accordingly, the pocket filter was fitted with the #113 Nomex/glass
fabric, turning the singed surface to the dust side. After pre-
treating the filter with limestone dust, according to the user's
specification, the system was placed into service. Start-up was
uneventful and performance was quite adequate at least to an A/C
ratio of 2.5. As the flow increased to the design A/C ratio of
3.9, however, the pressure drop increased to an excessive level.
Again, as in previous trials, the dust caked onto the filter sur-
face and could not be discharged by the mild reverse air cleaning
pulses. As a result, the decision to change to another, more
vigorous cleaning type collector was made and except for one or
two remaining trials, the experimental filtration program was ter-
minated.
49
-------
U1
o
TA11LE 12. EXI'liRlMENTAL FILTRATION nF STKF.I. GRINDER DUST, H-C
(A/C 6. APc limit - 6 in. w.r., 130*-140*F, 10 pop i-UsmIng oporntionK)
Run
No.
171
172
173
174
175
178
176
177
179
180°
Fabric
No.
148
147
143
141
U3e
U3C
148
58
75
113d
Mfr.
U.F.
U.F.
Gore
AF
AF
AF
U.F.
AF
HSA
AF
Style
? p.c. wov. sp.
? p.c, wov. sp. &
0.6Z cond. graph.
Gore-Tex
TFR on p.e. felt
MC2-8UL - TFE
fin. on acrylic
S22B3NRMM
87XNomex
>13 C glass,
singed
same as 113c &
1.5Z cond. graph.
? p.e. felt
Rlazcd
(D1R4) felt
wool
Rcsp.-unt.
wool/acrylic
same- as 113e
Permeability
cfm/ft3 9 0.5"
14
14
6
21
42
42
16
30
103
42
Resistivity
n/n
3 x 10U
107
1 x 1013
1.6 x 1010
6.6 x 109
107
5 x 1012
to11
6.6 x 101
Filtration P. if. i miners
Hartlculnte
Loading
g/min
11
18
20
16
16
15
11
16
15
13
Col looted
Pnrt Iculnte
E
20.2
12
35
21
28a
36a
34
3.6
33
39b
28
Cnko Wt.
K
10
10
20
16
3'.
27
3.5
28
30
25
l-'l Her
< yr I o
ni i n : HOC
1:05
0:40
1:45
1:20
1:4S
2: IV"1
2:15
0:20
2:05
2:35b
2:10
Tlu
wt._, g
44
53
15
79
31
42
15
116
222
33
8
APp
3.2
3.2
1.6
2.5
1
1
1.4
1.1
0.8
I
Relative
Leakatc
Moderate
Low
V. I.DW
High
V. Low
V. Low
V. Low
V. Low
V. Low
V. HiRh
Low
"Additional 10 shakes applied
S instead of 10 pop cleaning operations
'"Run made after bag precoated with limestone dusts
-------
In the course of other tests, some extremely interesting and
potentially significant observations were made. For example, in
the rolling tests, both the grinding and burning dusts deposited
a tenacious, brown resin-like film on the contact surface. The
dust had a distinct odor resembling a burned organic material.
Furthermore, the grinder dust was shown by scanning electron micro-
scopy to be made up of small but exactly spherical particles, and
none were rough. Also, on the inside, clean air side of the pocket
filter fabric, a dark brown, narrow line of about 1/2 in. in length
occurred at each point of contact with the metal spring used to
keep the pocket from collapsing. These marks were first thought
to be burns, but closer examination showed no fused fibers and
treatment with a petroleum solvent caused them to be removed al-
most completely.
These findings need verification and they raise other ques-
tions that should be answered. Nevertheless, some speculation is
warranted even on the basis of the very limited information. For
example, it seems that a component, possibly organic or a form of
sulfur, and developing during both the processes of burning and
the grinding of the billets, is common to both dusts, producing
a very adhesive binder. This ingredient, assumed to be a material
of high electrical resistivity with strong electrostatic activity,
could migrate preferentially (because of the charge) or otherwise
through the fabric to the grounded spreader wires inside the filter
pockets producing the observed brown line marks. Certainly, the
smooth spherical particles as shown in the SEM photo (Figure 9)
of the grinder particles would have little tendency to bind mech-
anically to the fabric and account for poor cleanability. Even
static adhesion between the particulate and fabric, especially
when using a conductive medium, would not be as tenacious as shown
in the filter trials. Some obvious questions arise from these
real and not so real findings. Where, for example, would a common
contaminant originate? Could it be that a protective coating is
applied to the billets and that this, in course of heating, is
transformed into a super binder? It is known that a flux is used
in the burning operation and that the grinder wheels contain a
phenolic binder. Certainly, some of both materials or their pyro-
lytic products get into the dust, but are these products so similar
that they form the same kind of reddish brown film in the rolling
tests and impart the adhesive properties? At this writing, no
better data have been obtained and no further explanations have
been made available. The questions remain unanswered.
The Collection of Polymeric Dusts
The Collection of Polymeric Dusts P-K-3085 and P-6140
Two polymeric dusts, P-K-3085 , an aromatic copolymer, and
P-6140, an alicyclic polymer mixture, were evaluated filtration-
wise with several selected filter fabrics. The Teflon fabric used
in the plant shaker-type collector served as the reference medium.
It is not clear why a Teflon fabric is used in the plant collector
51
-------
FIGURE 9. STEEL GRINDING DUST, X 500
by
SCANNING ELECTRON MICROSCOPE
52
-------
that operates at ambient temperature and humidity, unless one or
both dusts exhibit critically serious blinding properties or be-
cause the dusts are flammable.
Both dusts are flammable, thermoplastic and develop high
levels of electrostatic charge. Resistivity measurements at 70°F/
40%RH also indicate that both dusts have values greater than 1014
n/t> In the rollin9 test, the apparent density of P-K-3085 in-
creased by about 13.7%, and that of P-6140 increased by 9% while
both showed evidence of aggregation.
Serious leakage, especially during the cleaning operation,
is the most critical problem experienced during the commerical
filtration of these dusts. The resulting dust seepage leads to
difficulties with plugged chambers, clogged pressure taps, and to
the malfunctioning of shaker controls and the dust removal system.
The equipment specifications and operating conditions were
given as follows:
shaker-type collector designed for A/C = 5.3 and operated
at 3.3,
the maximum AP is 10 in. w.c. but operates at 2 to 4 in. w.c.,
cleaning is accomplished by eccentric, horizontal mechanical
shaking of 1-2 min,
the filter cycle is 8 hr (to the 2 to 4 in. w.c.),
the clean cloth AP is 2 in. w.c.,
dust loading ranges from 0.9 to 1.4 gr/scfm.
The filter test results, although limited to an evaluation
of seven media - three very electropositive fabrics, three very
electronegative fabrics and a moderately positive fabric (refer
to Table 13) permit some tentative conclusions of practical sig-
nificance. While both of these polymeric dusts show the same high
level of electrical resistivity (> 101£* fi/£)) , they differ appreci-
ably in their aggregating and filtering properties. Of the two
dusts, the P-6140 resin is the most critical, in regard to its
ability to agglomerate and with respect to its overall collect-
ability. With the P-K-3085 dust, all but the plant-used fabric
performed well without leakage. The more critical-to-collect
P-6140 dust is filtered best without leakage only by the high
charge, electropositive fabrics.
It is especially interesting to note that the very electro-
negative Kevlar fabric #65 performs only slightly better than the
other electronegative filters, including the plant-used Teflon
#94. Actually, the Kevlar filter performed more like the Teflon
53
-------
U1
TABLE 13. EXPERIMENTAL FILTRATION OF POLYMERIC DUSTS P-K-3085 AND P-6140
A/C - 6, APc limit 6 In. w.c., 70-75'F. shake cleaning
Run
No.
Ill
109
106
110
105
108
107
112
113
116
115
117
Fabric
No.
71
94
89
96
78
78a
18
65
Hfr.
AF
Eld
Gore
AF
AF
AF
JPS
Eld
Style
S-225
XT0954
L10564
S-1414
S-1414
S-1414
D-8339
181 lit
Flber(s)
wool
Teflon
Coretex
Teflon lam.
on p.p.
7V wool
'25 nylon
75, wool
"55 nylon
T^- wool
'25 nylon
0.9Z Hyamine 3500
p.e. ep. &
napped
Kcvlar fll.
Perme.iblH ty
cfm/ft? C 0.5"
46
37
19.5
102
33
33
38
31
Resist.
n/a
10"
.»
.»
2.9xl013
4 x 101?
10n
) 10"'
10'"
Rel. Tril.o.
Position
Total Rub
Voltage
+5.6
8.3
v -6
11.5
-3.6
16.8
+5.7
12.2
x. +7
11.3
antistatic
+1.4
21.6
" -8
18
Fll trot Inn P.ir.iroetprs
Collected
Partlculntp, g
K-3085 6140
171
148
71
207
88
190
13(1
193
113
102
230
124
96
Plug
Weight, g
K-3085
15.6
1.4
9.3
10
9.3
12.6
6140
6.2
8.4
33
12.4
13
15
5.4
AP In. w.c.
K-3085
0.15
0.7
0.7
0.1
0.2
,..
0.1
6140
1.2
0.7
0.1,
0.4
1
o.s
0.2
O.R
low
mod. htgli
v. high
1 I»W
low
slight
s light
(cleaning)
low
low
low
low
lew
hlKh
-------
and Gore #89 fabrics in regard to collected solids, plug retention,
and plugged cloth resistance than the electropositive fabrics. In
collecting the P-6140 dust, therefore, the more electropositive,
high charging media perform far better than the electronegative
materials. The outstanding performance of the high charge, only
moderately high electropositive polarity, low dissipation rate
polyester fabric #18 with this dust is noteworthy. For collecting
the P-6140 dust without leakage, these results seem to suggest that
the preferred medium is electropositive but not necessarily at the
top level in the triboelectric series, capable of generating a
high intensity of charge and with a relatively low rate of charge
bleed-off. These features, presumably, are effective in encourag-
ing suitable aggregation of the P*6140 dust on the collecting fab-
ric. The same fabric should also be effective in collecting the
P-K-3085 resin even though it seems to respond favorably to lesser
extremes in charging.
The Collection of a Fine Polymeric Dust, RH-P
A fine, high resistivity, very electrostatically active acrylic
polymer dust that posed serious commercial collection problems was
shown to behave similarly in experimental filtration studies. Only
two among the first 14 tested media made of different fibers in a
variety of constructions showed plug weights below the weight of
the collected dust. Dust release from all of the tested bags was
shown to be limited whether cleaning was attempted by shaking or
by popping, the latter action being similar to that provided by
pulse jet. The dust-fabric adhesion was so intense that plug
weight/cake weight ratios ranged from 80-90/20-10 for all but the
two test media noted above. The plug weight/cake weight ratio was
substantially lower for one of these - a filament Teflon filter
fabric, but dust leakage was critically serious throughout each
filtration cycle. In these trials, only a Dralon T woven filter
fabric performed at a relatively low plug weight/cake weight ratio
without leakage. It is important to note that an almost identical
filter fabric (with respect to weave and permeability), made of
another acrylic (Zefran) at a different location in the T.E. series,
held about 50 percent more plug and provided a 70 percent higher
plug pressure drop with a high plug weight/cake weight ratio at
85/15. These performance differences are significant since the
tests were conducted at the same 5.4 A/C ratio to a pressure drop
limit of 6 in. w.c. under conditions of relatively constant dust
loading and temperature (140 - 150°F).
" Additional trials were carried out using 16 variations of
needled (felt) fabrics, some of these were prepared with and with-
out glazed filter surfaces and with and without organic-type anti-
static treatments or included conductive fibers. The following
conclusions were reached in the course of this phase of the study:
glazing of the felt surface improved collectability. Both
plug weight and plug pressure drop values were lowered when
a glazed filter surface was used.
55
-------
a cationic antistatic finish (Hyamine 3500), whether used on
the glazed or on the normal felt surface, improved collect-
ability and reduced the plug weight in each case.
A single filter trial with the Dralon T fabric 13 from the
first test series was carried out using the acrylic dust after
this dust had been treated with the cationic Hyamine 3500 (anti-
static) surfactant to a 1 percent add-on. The resistivity of the
original test dust was reduced from > 1013 n-cm to 109 n-cm by
the treatment, but this change had no beneficial effect on the
collection process.
As a matter of interest, with a potential for explaining
performance differences, is the observation that this acrylic
dust with the antistatic treatment showed little tendency to dis-
perse easily into a cloud. It was also shown to densify (by 20
percent) and become highly aggregated in the rolling test. The
untreated dust, on the other hand, remained dusty or prone to dis-
persion, did not aggregate significantly and did not increase in
density in the rolling test.
The overall results from the experimental studies may be
summarized as follows:
an acrylic fiber content filter performed better than fabrics
made from other fibers,
one acrylic, Dralon T, collected the polymeric acrylic dust
better than another,
a Dralon T felt functioned better than a woven Dralon T filter
material,
glazing this felt surface offered further improvement, and
an antistatic finish applied to the glazed Dralon T felt
provided additional advantages.
Subsequently, information obtained from the processor of the
acrylic polymer dust indicated that the filter bags that performed
best in commercial service were made from a European manufactured
felt constructed with an acrylic fiber but treated by the user
with a cationic (antistatic) finish. After relatively short per-
iods (about one month) of plant use with different products, the
bags were laundered and retreated with this finish before reuse.
Without knowing the location of this acrylic dust in the
triboelectric series, it is not too safe to speculate regarding
electrostatic effects. Nevertheless, the rolling test indicated
that the dust did not aggregate or at least did not agglomerate
easily. With its high level of electrostatic activity and this
restriction on aggregation, especially in contact with a fabric
56
-------
of suitable charge and/or charge polarity, it seems reasonable to
suggest that construction (felt for depth filtration and a glazed
surface for enhanced cleanability) must be relied upon for the
basic filtration needs, but that the preferred filter medium might
be that having a location in the triboelectric series near the
position of the dust. Such close proximity would be expected to
limit electrostatic attraction or the adhesion between dust and
fabric allowing, thereby, improved cleanability. In addition, the
presence of the antistatic finish would also tend to enhance clean-
ability through its ability to promote rapid charge bleed-off.
The evidence, albeit circumstantial, implies that collectabil-
ity of even those fine, electrostatically active dusts that resist
aggregation may be achieved best by employing media with preferred
electrostatic-as well as constructional-properties.
The Collection of Silica PG-C
Only a few particulates seem to resist aggregation and fail
to respond to electrostatic charge induced agglomeration on a
suitable filter medium. While silica does agglomerate, it seems
to reach a state of limited aggregate stability. In fact, in the
rolling test (Table 6), arbitrarily adopted as a guide to whether
dusts will agglomerate, silica (at least the two varieties exam-
ined) actually seems to disperse somewhat into more "feathery"
aggregates and consistently showed a reduction rather than an
increase in density. In the collection of silica by fabric fil-
tration, therefore, further agglomeration beyond a frail aggregate
structure cannot be relied upon for optimizing filter performance.
Fabric selection, in order to realize the best performance, seems
possible on the basis of other operating parameters.
Since extended processing may tend to further disperse silica
and almost any commercial method of handling this material can be
expected to induce such a change, selection of the preferred medium
will most likely have to be directed more to attain and maintain a
high level of efficiency than to realize exceptionally low operat-
ing pressure drop. In other words, fabric construction would seem
to have to be relied upon primarily to achieve a good pressure
drop-collection relationship. But the electrostatic properties of
the medium might be used to hold an effective filter-aid like cake
on its surface at all times and, thereby, to enhance efficiency.
This suggests that the polarity difference between the dust and
the filter medium should be high in order to offer substantial
attra'ction between the two materials and, further, that the dis-
charge rate of the fabric should be sufficiently low to maintain
the attractive forces for an adequate period of time to keep some
of the dust on the surface of the fabric, even after the cleaning
operation. Obviously, cleaning effectiveness may also have to be
controlled at less than a maximum level.
57
-------
According to the data reported earlier in Part A of this
report by Professor Penney, some of which is reproduced in Table 14,
silica should be located at about -3, between -2.5 and -3.6 on the
arbitrary scale of our triboelectric series. In this location, a
considerable electrostatic influence should be exerted on silica
dust by a highly positive fabric like #97 (wool/nylon) or by a
highly electronegative fabric like #65 (Kevlar), both of which are
widely separated from the silica in the T.E. series. In a favor-
able construction, the wool/nylon and a Kevlar fabric might be
expected to perform similarly, filtrationwise (refer to Table 14).
The Kevlar fabric, however, displays a higher rate of charge dis-
sipation (45% in 2 min) compared to the 20% value shown for the
wool/nylon material. More serious, though, is the significant
difference in the bulkiness of the two materials with the filament
Kevlar fabric offering little more than screen-like properties
compared to the high bulk, good cover, high charge features of the
wool content medium. On all counts, then, the woven wool/nylon
#97 should perform at high collection efficiency in shaker-type
cleaning systems, and a felted wool should be best in a pulse-
or reverse air-jet type system when collecting either of the two
silicas.
Unfortunately, without a spun yarn Kevlar fabric more like
the wool/nylon material in construction, it is not possible to
prove without question the hypothetical relationship suggested
above. Nevertheless, there are some important observations that
appear to provide a measure of support. For example, although the
wool/nylon #97 fabric leaked the silica dust quite seriously at
first, the loss of silica through this filter bag became less and
less with progressive runs, despite the vigorous shaking type
cleaning operations. Even with such energetic cleaning (ten hori-
zontal strokes through 1.5 in. displacement at the top cap), the
plug pressure drop (APp) increased from 0.2 in. to 0.6 in. w.c.
after about eight (8) cycles and remained at that level until the
test was terminated after eleven (11) cycles. Similarly, the
plug weight increased from 19.3 g to a high and essentially con-
stant value of -^ 50 g. Although the Kevlar filter bag could not
be carried through as many runs because of excessive leakage, the
initial AP of 0.6 in. w.g. and, especially, the relatively high
initial plug weight of 21 g (6 cycles) (high for a filament yarn
fabric) would seem to reflect the attractive features of a fabric
with a triboelectric charge polarity substantially opposite to
that of the silica dust.
The filtration data (refer to Table 15) obtained in the col-
lection of the same silica dust with a needled polyester felt,
when compared with the above results, seems to substantiate the
hypothesis. Certainly, compared to the filament yarn fabric or
even the woven wool/nylon fabric, the data recorded for fabric
#109 are most striking. Despite the fact that bags made of felt-
like materials do not shake as well as the more flexible woven
fabrics, the plugged cloth pressure drop value (APp) for the poly-
58
-------
TABLE 14. TRIBOELECTRIC POSITION OF SILICA, PG-C
(RELATIVE TO SOME FABRICS)3
Fabric
Dust, Charge (Current, 1 x 10 9 a)
Identification
Rel. T.E. Position
Relative to Fabric
Ul
// 97 75 wool/25 nylon
// 15 Dacron
# 9 50 Dacron/50 Orion
Dacron (Type 55) & Silicone
//89 Gore-Tex Teflon on p.e.
#55 Darvan (felt)
//37 Teflon
#65 Kevlar
-v +8.0
+4.8
+0.7
-2.5
^ -3 Silica PG-C
-3.6
-3.7
-v- -6.0
* -8.0
-4.5
-1.4
-1.1
+0.9
+3.8
+4.2
Penney, G.W. - Progress Report of October 15-November 15, 1976 to EPA Grant No. R803020
-------
TABLE 15. EXPERIMENTAL FILTRATION OF SILICA, PC-C
(A/C - 6. ftPc limit - 6 in. w.c.. 130"-140'P. shake cleaning)
Run
Ho.
147
148
149
150
151
Fabric
No.
65
97
109
16c
55
Hfr.
Eld
AF
P&S
JPS
AF
Style No.
181 III
S-1414P
U-J (used
ft washed)
04-39703/5
40/601
Fiber
Kevlar
aramld
60.x wool
^Co nylon
Dae r on
p.e.
Dacron
|).C.
Darlan
acrylic
Type
wov. fil.
wov. sp.
felt
wov . sp .
felt
Permeability
cf«/ft2 9 0.5"
31
76
56
26
32
Rel. T.E.
Position
v. -8
-v +8
-2.S
-2.5
-3.7
Rate of
Charge Loss
2 (2 oln)
45
80
30
30
30
Filtration Parameter;)
Av. Collected
Participates
K
130
125
105
22
40
_P1<
«t., g
21
50
45
5
45
>K
app
n.e
0.6
0.3
0.9
0.3
Relative
Leak.ige
hif,h
mod- low
mod -low
high
low
-------
ester felt increased from an initial value of 0.1 in. to only
0.3 in. w.c. and the plug weight, again despite the fabric's com-
paratively very high bulk, increased from about 30 to only 45 g.
The plug weight, although seemingly somewhat elevated, is really
relatively low for a felt-like filter bag and yet it is below
that found in the far more open (higher permeability), less bulky,
wool/nylon bag of exactly the same filter area. In addition, com-
pared to this polyester needled bag, the wool/nylon fabric collected
about 15% more dust before reaching the same limiting pressure drop
(APc) of 6 in. w.c. Since some of this better collectability might
be attributed to permeability difference, major emphasis must be
placed on the plug weight and plugged cloth pressure drop data to
provide a measure of evidence for the effectiveness of electro-
statics in producing and retaining the attractive forces needed
to form and hold a filter aid type of surface of silica on the
fabric during each filtration cycle.
Also as given in Table 14 data, the #16 Dacron with a Silicone
finish (this finish does not contribute significantly to the tribo-
electric properties) is located somewhat closer to silica in the
T.E. series. As a result, the forces of attraction between the
fabric and dust should be relatively low. Such appears to be
verified by the filtration data that shows a low plug weight (^ 5g) ,
high APp and low collectability for this woven filter fabric .
Based on these results, it would appear that the efficient
collection of silica dust in a shaker-type system may be accom-
plished best with a moderately heavy (^ 8 oz/yd2), high cover,
very electropositive, woven woolen fabric of about 30 to 40 perm-
eability. If a pulse-jet type collector were to be used and
similarly high efficiency were required, a wool felt of about
12 oz/yd* at a permeability of about 50 to 60 permeability might
be recommended.
61
-------
SUMMARY
Admittedly, many of the conclusions reached in the foregoing
discussions relate electrostatic properties of filter fabrics with
those of the particulates to explain collectability, cleanability
or efficiency are developed largely upon circumstantial evidence.
But any other fabric or particulate property fails completely to
offer a more reasonable explanation. The basic premise of the
electrostatic involvement hypothesis deals with particulate aggre-
gation. Whereas this electrostatic charge-agglomeration reaction
was first merely hypothesized, Professor Penney's work now removes
some of the guesswork. His studies have demonstrated, for example,
that "impact" charged particles form a "chain-like," porous or
agglomerated deposit on a fabric without the use of high voltage
either on the particulate or on the collecting surface. This ob-
servation serves to confirm the premise that natural charging can
produce aggregates just as artificial charging (i.e. in an electro-
static precipitator) often leads to such a change in particulate
qualities.
In another of Penney's tests, again without an external pot-
ential being impressed on the filter fabric, corona-charged part-
icles (electronegative) also became deposited in a "chain-like"
aggregated manner on just the electropositive fiber of a composite,
two fiber filter. This filter is pictured in the photomicrograph
(Figure 10)3 showing a blend of two different (wool and acrylic)
fibers, each of about 3 ym in diameter. The fact that one fiber
remains clean while the adjacent fibers and only these electro-
positive fibers, collect the negatively charged particles as a
porous aggregate is viewed as supporting evidence for the original
hypothesis that stresses the effectiveness of high, often opposite
charges in promoting the agglomeration of difficult-to-agglomerate
particulates.
The other electrostatic-filter fabric relationships are more
readily appreciated and justified. For example, that opposites
attract and that as materials become more opposed their attraction
increases is generally accepted. Less clear is the relative im-
portance of charge intensity and triboelectric position in the
3Penney, G. W., J. Air Poll. Control Assn., 26:58 (1976)
62
-------
FIGURE 10. AGGREGATED DEPOSIT OF PRECHARGED
PARTICULATE ON A FILTER FIBER
63
-------
aggregation process. But these and other charge-filtration rela-
tionships seem to provide the best explanation for many observed
anomalies.
Rapid charge dissipation from a fabric or from a dust, too,
may be realized if their entire surfaces are treated with a con-
ductive finish. Much less effective, if at all effective in a
fabric, is the use of conductive fiber elements. These fibers
produce essentially no continuity over the fiber surface on the
micro scale and, thereby, provide little benefit and certainly
not the kind of overall charge dissipation offered by continuous
conductive (cationic or graphite) finishes. Nevertheless, using
the fugitive (organic based antistatic agents) in filtration
processes that depend on such finishes for charge bleed-off, must
be done with the knowledge that these agents, whether they depend
upon the ionization of quaternary or hydroxy compounds for charge
conductivity, must have moisture present for this function. Fur-
thermore, both chemicals are temperature sensitive and the quat-
ernary compounds are believed to form as intermediate product, an
electropositive amine, in the "burning-off" process.
The value of antistatic treatments, sometimes even more
effective on the particulate than on the filter fabric, seems to
have been reasonably well demonstrated in some of the tests, and
these revelations can have important implications. In considering
antistatic finishes, it is important to note that such finishes
do not eliminate electrostatic charging, but rather, they only
provide a means for the generated charge to bleed-off very rapidly.
In the course of the experimental studies reported here, the
considered effects of electrostatics on the filtration processes
have been utilized to explain performance characteristics and to
identify preferred media. Whenever particulates were capable of
agglomerating, fabrics of high polarity differences and/or high
charge intensities were employed to realize this effect. In at
least one instance, a fabric capable of developing a high charge
but at only a medium triboelectric position, produced the needed
aggregation. In collecting another dust that did not seem to
agglomerate to a stable aggregate, optimum filtration properties
were achieved with a filter medium that was separated significantly
from the particulate in the triboelectric series. Fortunately,
the location of this dust in the triboelectric series had been
established from the Penney studies. For the first time, the
reliability of the dust-filter fabric relationship in the tribo-
electric series had been established. Accordingly, the analysis
that prescribed such a difference for optimum performance would
seem to be based on reliable evidence.
In at least one instance, the extent of improvement in fil-
tration performance suggested for a supposedly better medium based
on electrostatic properties, was inadequate. Further analysis of
the overall problem now suggests that other factors may have had
64
-------
a completely overriding influence and critical fabric parameters,
including the electrostatic, could never be altered favorably
until these limiting conditions were eliminated.
Hopefully, this report of some fundamental and practically
oriented studies, directed especially to considerations of elec-
trostatic effects in fabric filtration, provides a contribution
to the advancement of technology that allows otherwise unexplained
filtration events to be solved better and and faster, even though
not necessarily diagnosed with complete accuracy. Hopefully, too,
other investigators will find this report sufficiently motivating
to proceed further into this important area of fabric filtration.
TABLE 16. METRICATION OF SOME FILTER PARAMETERS
PROPERTY
Width of cloth
Diameter of bags
Length of bags
Thread count
Thickness
Unit mass of cloth
Density of material
Permeability*
Temperature
Pressure drop
Dust loading
Dust mass
Particle size
*In metric practice, 20 mm is usual AP, but 200 Pa is being suggested. (Since
permeability is not strictly proportional to AP, the factor of 7 is only approx-
imate and must be established experimentally).
METRIC UNIT
nun
cm
m
per cm
inn
g/m2
g/m3
1/s/m2 @ 200 Pa
8C
mm w.c .
g/m3
g
urn
BRITISH UNIT
in.
in.
ft
per in.
in.
oz/yd2
lb/ft3
cfm/ft2 @ 0.5" w.c.
°F
in. w.c.
gr/ft3
gr
in.
CONVERSION
FACTOR
(TO METRIC)
25.4
2.54
0.3048
0.394
25400
33.9
16018
7*
(F-32) 5/9
25.4
2.288
0.0648
25400
65
-------
INDEX
Abbreviations, XII
Acrylic Dust, 55
Adhesive Forces, IV
Aerosols, IV, V, 27
Charged, IV, V, 27
Resistivity, 27, 28
Agglomerates, IV, V, 27, 30
31, 63
Aggregates, (see agglomerates)
Antistatic, 12
Burning Dust, 41
Cake, V, VI, 11, 27
Porous, V, VI 27
Charge, V, 13, 27, 57
Aerosol, V, 27, 57
Intensity, 13
Charged Particles, V, 27, 57
Charging, 8, 9, 12
Corona, 62
Frictional, V, 8, 9, 12,
27, 57
Particulate, V, 27, 57
Cleaning, 3
Contents, Table of, VII
Conclusions, 1
Deposit, V, VI, 27, 62
Porous, V, VI, 27, 62
Dust, V, VI
Agglomeration, V, VI, 27, 30
31, 62, 63
Electric Furnace, 34, 36, 37,
38, 39, 40
Electrostatic Properties, 27,
28
Metallurgical, 34, 38, 39, 40
Polymeric, 51, 52, 54, 55
Resistivity, 27, 28
Silica, 57, 59, 60
Def initions, XI
Efficiency, V
Electric Furnace Dust, 34,
36, 37, 38,
39, 40
Electrical, IV, 6, 9, 12, 16,
18, 19, 20, 21,
22, 23, 24, 25,
26, 28
Augmentation, IV
Properties, IV, 6, 9, 16,
18, 19, 20,
21, 22, 23,
24, 25, 26,
28
Resistivity, 6, 9, 12, 16,
18, 19, 20,
21, 22, 23,
24, 25, 26,
28
Electrostatic, II, IV
Augmentation, V
Charge, V
Effects, V
Evaluator, 8
Fabric (properties), 14, 16
Filter, V
Forces, V
Generator, 8
Particulate (properties), 27
Ferromolybdenum By-Product
Dust, 45
Figures (List of), VIII
Filter, XI
Cake, XI
Efficiency, XI
Electrostatic, IV, 62
Fabric, 18, 19, 20, 21, 22,
23, 24, 25, 26, 29
Performance, V
Filtration, VI, 1, 3, 4, 5,
6, 7, 8, 9, 11, 29
Equipment, 4, 5, 6, 7, 8,
9, 11
Experimental, 3, 29
66
-------
Facilities, 3
Procedure, 5
Studies, VI, 1, 3
Flyash, 32, 33
Grinding Dust (Steel), 48
Humidity, 12
Instrumentation, 6
Equipment, 4, 5, 6, 7, 8, 9,
11
Metallurgical Dusts, 34, 38,
39, 40
Filtration, 34, 38, 39, 40
Metrication (Filter
Parameters), 65
Preface, IV
Particles, see particulate
Particulate, V, IV, 27, 28, 29,
30, 31, 63
Agglomeration, V, IV, 27, 30,
31, 63
Electrostatic (Properties),
IV, 27, 28, 62
Resistivity, 27, 28
Triboelectrificaton, 27
Permeometer, 7
Plug, 12
Polymeric Dusts, 51, 52, 54, 55
Fabric, 11, 12, 16
Particulate, 27, 28
Surface, 11, 12
VS Triboelectrification, 16
Volume, 12
Recommendations, 2
Silica, 57, 59
Steel Dusts, 34, 36, 37, 38,
39, 40, 41, 48
Burning, 41
Electric Furnace, 34, 36, 37,
38, 39, 40
Grinding, 48
Summary, 62
Tables (List of), IX
Triboelectric, v, VI, 8, 12, 14,
16, 18, 19, 20,
21, 22, 23, 24,
25, 26, 27, 28,
59, 62, 63, 64
Data, 14, 16, 18, 19, 20,
21, 22, 23, 24, 25,
26,
Determination, 8, 12
Effects, V, VI, 62, 63, 64
Fabrics (properties of),
14, 16, 18, 19,
20, 21, 22, 23,
24, 25, 26
Particulates (properties of),
27, 28
Properties, 14, 18
Series, 14
Resistivity, 6, 9, 12, 16, 18,
27, 28
Dust, 27, 28
Electrical, 6, 9, 12, 16, 18,
19, 20, 21, 22, 23,
24, 25, 26, 27, 28
67
-------
TECHNICAL REPORT DATA
/Please read laaructions on the reverse before completing}
1. REPORT NO.
EPA-600/7-78-142b
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Electrostatic Effects in Fabric Filtration: Volume n.
Triboelectric Measurements and Bag Performance
(Annotated Data)
5. REPORT DATE
July 1978
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
E.R. Frederick
9. PERFORMING ORGANIZATION NAME ANO ADDRESS
Carnegie-Mellon University
Schenley Park
Pittsburgh, Pennsylvania 15213
10. PROGRAM ELEMENT NO.
E HE 624
11. CONTRACT/GRANT NO.
Grant R803020
12. SPONSORING AGENCY NAME AND ADDRESS
13. TYPE OF REPORT AND PERIOD COVERED
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND
Final; 9/73-5/78
14. SPONSORING AGENCY CODE
EPA/600/13
is.SUPPLEMENTARY NOTES TERL-RTP project officer is James H. Turner, Mail Drop 61,
919/541-2925.
is.ABSTRACT ijine repOrt describes the construction and application of a bench-scale,
single-bag, experimental filter. It also describes several complementary evaluation
procedures and their data. Especially significant are the methods for, and results
of, electrical determinations that are not normally applied to filter media and parti-
culates. The effect of these electrical parameters on the collection process is used
to explain performance variations. Results of several filtration studies on several
industrial particulates (e.g., from a power plant, and from metallurgical and chem-
ical processes) are reviewed in detail and explained on the basis of electrostatic
properties. Flyash collection, for example, was favored by the use of mid-triboelec-
tric position media and not by the highly electropositive or electronegative fabrics
that are used for their high temperature properties. Three different electric furnace
dusts tended to respond best (filtrationwise) with mid-triboelectric position fabrics,
modified for different cleaning practices. Steel grinding and burning dusts offered
very critical filtration characteristics that demanded control of aerosol flow and
particulate loading, as well as special care in filter media selection. A ferromoly-
bdenum by-product dust was collected best by very electropositive fabrics.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Air Pollution
Dust
Aerosols
Fly Ash
Filtration
Fabrics
Electrostatics
Measurement
Electric Power Plants
Metallurgy
Chemical Industry
Electric Furnaces
Air Pollution Control
Stationary Sources
Particulate
Fabric Filtration
Triboelectric Measure-
ment
13B
11G
07D
21B
11E
20C
14B
10B
11F
07A
13A
18. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
78
20. SECURITY CLASS (This page I
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
68
------- |