United States
Environmental Protection
Agency
Air and Energy Engineering
Research Laboratory
Research Triangle Park NC 27711
Research and Development
EPA/600/S7-85/049 Jan. 1986
&ERA Project Summary
Development of Advanced
ESFF Technology
G. E. R. Lamb, R. I. Jones, K. T. Duffy, D. A. Saville, and B. A. Morris
This report summarizes work done to
explore ways to magnify the effects and
broaden the scope of electrical stimula-
tion of fabric filtration (ESFF). The im-
portance of these effects was estab-
lished in earlier work, both in the
laboratory and in pilot plant trials. The
present work covered a number of sep-
arate topics: a) effects of particle
charge levels, b) mechanisms of pres-
sure drop reduction, c) programming
applied electric fields with respect to
time, d) cleaning fabric filters, and
e) printed electrodes.
The results showed that particle
charge has a strong effect on the re-
sponse of filtration performance to
ESFF and that some form of precharg-
ing would be cost-effective in many
cases. Measurements of dust deposi-
tion patterns showed that electrical ef-
fects cause shifts in the dust deposits:
a) toward the entrance of the bag, b) to
the electrodes, and c) toward the sur-
face of the fabric. Results of modeling
studies agree quantitatively with the
observed effects. Studies of pro-
grammed voltages examined the possi-
bility of using sudden changes in elec-
tric field to aid bag cleaning and also
the effects of using ac rather than dc
fields, but neither approach held out
promise of significantly improved per-
formance. When pulse-cleaned bags
were run at higher than conventional
velocities, penetration rose to unac-
ceptable levels; whereas, with bags
cleaned with reverse air, pressure drop
reached large values. The difference ap-
pears to be related to the energy levels
of the two cleaning methods. "Printed"
electrodes (PEs) are stripes of conduct-
ing material replacing metal wire elec-
trodes and having the advantages of
lower cost and the possibility of apply-
ing electrodes of complicated design;
they appear to be as effective as wire
electrodes.
This Project Summary was devel-
oped by EPA's Air and Energy Engineer-
ing Research Laboratory, Research Tri-
angle Park, NC, to announce key
findings of the research project that is
fully documented in a separate report
of the same title (see Project Report
ordering information at back).
Background
The cost of emissions control can be a
substantial fraction of the total capital
and running costs of electric utilities.
Where the chosen means of particulate
removal is a baghouse, some cost re-
duction could be achieved by increasing
face velocities, allowing the use of
fewer bags and a smaller baghouse. It
has already been shown that with elec-
trostatic augmentation, a baghouse can
operate continuously at about double
the conventional face velocity. Reduc-
tion in cost has been estimated at 30
percent for a pulse-jet baghouse if, by
means of ESFF, face velocity is increased
from 2 to 3 cm/s. Similar pilot plant re-
sults have been obtained for reverse-air
baghouses.
The promsing results of the two pilot
plant tests were obtained with a rather
simple modification of conventional
bags. It is reasonable to suppose that
this did not represent the best possible
design and that additional research
might indicate ways of further improv-
ing the response to ESFF. The work dis-
cussed in this report consisted of a
number of studies, outlined in the ab-
stract. Some were experimental; others
-------
sought a better understanding of ESFF
mechanisms through modeling. The
several studies addressed the following
problems:
1. The charge level on the aerosol
has received limited attention in the
past. In the Apitron process, charging is
known to produce great improvement
in performance, but little is known
about the effects of charge when an ex-
ternally applied parallel field is applied
to the fabric. In the pilot plant tests of
ESFF, the charge level was neither con-
trolled nor measured.
2. The reason that ESFF brings about
lower pressure drop (Ap) has usually
been given, qualitatively, as a "more
porous dust cake." In this study, a more
precise quantitative treatment of the
mechanism involving a shift of the dust
cake upstream into the f Iuffy layer of the
filter was undertaken. Other mecha-
nisms were also considered (i.e., dis-
placement of the dust mass in axial and
tangential directions in the bag. This
was experimentally verified.
3. Most ESFF work to date has em-
ployed dc voltages. Part of this study
examined reversals of voltage polarity
as an aid to cleaning and use of ac
voltages, since in some cases this might
bring some cost advantages.
4. The purpose of using ESFF is oper-
ation at higher face velocities. Even with
ESFF, however, an upper limit to veloc-
ity is imposed by uncontrolled rise in
Ap. This limit can be assumed to reflect
the point at which the cleaning removes
less dust than was collected in the cycle.
Thus, improvement in cleaning should
allow greater increases in velocity than
are possible with ESFF alone.
5. To control buildup of static elec-
tricity, a technique for applying conduc-
tive bands or stripes to fabrics has been
developed commercially. Part of this
project examined the possible use of
such "printed" electrodes as substitutes
for wire electrodes.
Results
With the exception of measurements
for the modeling of pressure drop
mechanisms, all experimental work was
carried out with a single-bag laboratory
baghouse. The bag length and diameter
were 122 and 11.4cm, respectively. The
aerosol was redispersed coal fly ash.
1. The initial object of altering aero-
sol charge levels was to explain why,
with the TRI baghouse, the improved
performance due to ESFF would dimin-
ish or "fade" with time. The results of
charge measurements showed that the
redispersed fly ash aerosol carried such
a small charge that it was not measure-
able. It is assumed that some charging
occurred from corona from the bag
electrodes, but this was suppressed in
time by dust buildup. Precharging the
aerosol eliminated fading, but this par-
ticular result is of interest only in ex-
plaining laboratory results, since (in the
field) fly ash from a boiler carries con-
siderable charge and fading is unlikely
to occur. Figure 1, for a bag cleaned
with reverse air, however, shows that
there is a definite dependence of the re-
sponse to ESFF on charge levels, which
were approximately 2.5 and 5 (xC/g at 9
and 15 kV precharger potentials, respec-
tively. Weighing the bag showed that
the mass of dust collected was the same
no matter what the charge level, so that
the reductions in Ap were not due to
loss of dust in the precharger. The same
dependence on particle charge was ob-
served with a pulse-jet cleaned bag.
2. To study the mechanisms respon-
sible for Ap reductions, a modified bag
was made that could easily be opened
to examine the dust deposit. This was
done after operation with various com-
binations of precharger and bag poten-
tials. Visual observation (Figure 2) and
measurement of dust mass distribution
showed that, with precharging, dust
settled closer to the bag entrance and
preferentially on the electrodes. Calcu-
lations of Ap based on measured
skewed dust distributions gave good
agreement with the observed Ap. Calcu-
lations of the trajectories of charged
particles in the electric field generated
by charged electrodes in the bag also
indicated concentrations of dust on the
electrodes, in agreement with the ob-
served patterns. The same calculations
also showed the presence of a cylindri-
cal zone in the center of the bag where
the electric field was essentially zero.
A separate study examined the
change in Ap due to shifting of the dust
deposit upstream to a low solidity sur-
face layer of the filter fabric. Measure-
ments with layered filters showed an in-
creasing shift to the upstream layer as
the electric field increased. Calculations
of the reduction in Ap expected to result
from these shifts were made based on
the Happel cell model. It was found that
the calculated Ap depended on the as-
sumed form of the collected dust; that
is, agreement between theory and ex-
perimental values required an arbitrary
"dendrite fraction" of 0.25; i.e., it was
assumed that one quarter of the dust
1.0
•S 0.8
«J
CC
& 0.6
Q
0)
0.4
02
— 3 cm/?
—1.5 cm/s
Average Field. kV/cm
Figure 1. Dependence of pressure drop
ratio (PDR) on bag field after
extended run for a woven-glass
reverse-air bag. (Top curves: no
orecharger.)
Figure 2. Dust deposition pattern with 4
kV/cm field and -15 kV pre-
charger potential.
settled on the fibers in the form of den-
drites. The actual value of the dendrite
fraction was not determined experi-
mentally.
3. Sudden large electrical potentials
of reversed sign were applied to the
electrodes on the dust-laden bag during
the cleaning cycle. Initially, results
seemed encouraging, since there was a
sharp drop, as much as 30 percent, in
the residual Ap. However, this drop was
quickly reversed when filtration was re-
sumed, leading to the assumption that
-------
only a small area of the dust cake near
the electrodes was affected, the dust
cake in this area quickly becoming re-
stored, so that practically no overall im-
provement was obtained. The drop in
residual Ap was even smaller when a
large potential was applied to the elec-
trodes during the filtration part of the
cycle. When ac was used in either the
bag or the precharging electrodes, the
improvement in performance was
somewhat smaller than with dc poten-
tials equal to the RMS ac. Thus little ad-
vantage can be derived from use of ac
potentials.
4. Bag cleaning studies were made
with a Teflon® felt bag cleaned by
pulse-jet (T-PJ), a woven glass bag
cleaned by pulse-jet (G-PJ), and a Teflon
felt bag cleaned by reverse air with
shaking (T-RA). It was found that with
T-PJ, Ap remained within acceptable
levels (~ 1.0 kPa) even at face velocities
of 7.5 cm/s. However, penetration
slowly rose to the order of 0.1 and did
not return to acceptable levels even
when the velocity was lowered. With
G-PJ, penetration was almost 0.1 even
at low velocity. As with T-PJ, however,
Ap remained low. With T-RA, penetra-
tion remained very low throughout,
never rising above 10 2 with no ESFF
potential or above 10~3 with 4 kV on the
bag electrodes and 15 kV on the
precharger. Even with these potentials,
however, Ap rose out of control at 4.5
and 6 cm/s. The different behavior is ex-
plained by the large difference in the
energy input for the two methods. The
shaker used released about 1 J/m2 of
fabric per shake. The 30 psi (207 kP)
pulse released 350 J/m2. Future re-
search should examine performance at
intermediate energy levels.
Cleaning by bag shearing was ex-
plored. The upper bag support is ro-
tated about the bag axis so that the fab-
ric is sheared. This mechanical action
was found to be as effective as shaking
in lowering Ap when the shear strain
was about 6°. This short program did
not examine the effects on penetration.
5. Measurements with commercially
produced filter felts fitted with printed
electrodes (PEs) showed improvement
in performance very similar to that ob-
tained with wire electrodes (Figure 3).
Measurements with woven fabrics like-
wise gave similar results with PEs or
wires, but a separate problem was the
fragility of PEs when applied to woven
fabrics, which have a high in-plane
shear compliance so that the PEs were
too easily strained to the breaking point.
Electrical continuity was then de-
stroyed. This problem was overcome by
making PEs of mixtures of carbon black
and rubber cement. However, this rub-
ber degrades at utility baghouse tem-
peratures: a temperature-resistant for-
mulation does not exist.
7.00-
01
CL
01234
Average Electrical Field, kV, cm
Figure 3. Performance of polyester felt
bag fitted with printed elec-
trodes
Conclusions
The electric stimulation of fabric fil-
ters is strongly dependent on the level
of charge on the incoming particles. If
the electrode wires can be kept clear of
dust cake, charging the dust by corona
from the electrodes is sufficient to re-
duce pressure drop by an order of mag-
nitude. Failing that, the charge level can
be maintained by a separate charging
device, and significant improvement in
performance can be obtained.
When a precharger is used, the paths
of the dust particles as they approach
the fabric are altered so that dust collec-
tion occurs preferentially on the elec-
trodes and near the bag entrance. Mea-
surements show that the dust mass
redistribution over the surface of a bag
accounts for a large part of the reduc-
tion in Ap. Calculated particle trajecto-
ries are consistent with the observed
pattern of dust deposition on the bag
wall, but also indicate a region with ra-
dius equal to half the bag radius in
which the electric field strength is al-
most zero and in which particles re-
spond only to the flow field.
Besides responding to changes in
dust mass distribution, Ap is also re-
duced because of the higher permeabil-
ity of a dust cake formed in a strong
electric field. The relative importance of
the two effects varies as the dust is
more or less uniformly distributed over
the bag surface. The increased perme-
ability is largely due to a shift upstream
into a region of the fabric having lower
solidity. This has been demonstrated
experimentally, and mathematical mod-
eling shows that the Ap is also sensitive
to whether the dust collects as dendrites
or as compact coatings on the fibers.
Agreement between models and exper-
imental results requires assumption of
an intermediate mode, expressed as the
dendrite fraction, which has not yet
been determined experimentally.
There is no benefit to be derived from
programmed variations of the voltage
applied either to the bag electrodes or
to the precharger. Sudden reversal of
the voltage polarity when the bag is
cleaned gives negligible improvements
in cleaning. Use of ac rather than dc po-
tentials resulted in slightly higher Aps,
higher penetrations, and large electrode
currents. However, although improve-
ments in performance with ac were
smaller than with dc, it is worth noting
that use of ac in ESFF need not be ruled
out.
As face velocity is increased, perform-
ance depends increasingly on cleaning
energy. High cleaning energy, as in
pulse-jet cleaning, leads to large pene-
tration, while Ap remains within a com-
mercially acceptable range. Low clean-
ing energy, as in reverse air with
shaking, leads to large Ap, with penetra-
tion remaining low. Shear cleaning, a
method which uses in-plane shear de-
formation of the fabric to aid reverse air
cleaning, appears to have some poten-
tial as a nontraumatic substitute for
shaking.
"Printed" electrodes (i.e., bands of
conducting material deposited on the
filter fabric) are as effective as wires in
producing ESFF. Their use by utilities
depends on finding a formulation for
the material that will resist the tempera-
ture and chemical environment of bag-
houses.
Recommendations
Designing electrode systems to maxi-
mize field intensities and particle charge
levels would be desirable. Replacing the
lightning rod with an axial wire for
corona charging should achieve both
ends. For pulse-jet baghouses, the
wires would be suspended between the
bags.
-------
Using the highest voltages would
maximize electric effects. This, how-
ever, may lead to problems analgous to
the back corona experienced with elec-
trostatic precipitators. Studying the
possible occurrence of this phe-
nomenon, as well as means for prevent-
ing it, would also be desirable.
Another benefit would be the devel-
opment of fabrics, having an upstream
region of low solidity, that are suitable
for commercial fabrication of bags,
tested to determine long-term perform-
ance.
Formulating and pilot-plant testing of
conducting substances, to be used for
printed electrodes that can withstand
baghouse conditions, would solve a se-
rious problem in this area.
Other problems would be solved by
investigating the potential usefulness of
shear cleaning, including its effects on
penetration and bag life.
G. Lamb, R. Jones, K. Duffy, D. Saville, and B. Morris are with Textile Research
Institute, Princeton, NJ 08540.
Louis S. Hovis is the EPA Project Officer (see below).
The complete report, entitled "Development of Advanced ESFF Technology,"
(Order No. PB 86-122 595/AS; Cost: $11.95, subject to change) will be
available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Air and Energy Engineering Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
U.S.OFRCI^LMAJL'
rUlPOSfAGf !•
Official Business
Penalty for Private Use S300
EPA/600/S7-85/049
0000329 PS
U S ENVIR PROTECTION AGENCY
REGION 5 LIBRARY
230 S DEARBORN STREET
CHICAGO IL 4Q604
------- |