United States
Environmental Protection
Agency
Industrial Environmental
Research Laboratory
Research Triangle Park NC 27711
Research and Development
EPA-600/S8-84-016 July 1984
Project Summary
Cost and Performance Models for
Electrostatically Stimulated Fabric
Filtration
Andrew S. Viner and Bruce R. Locke
A survey of the literature on perform-
ance models for pulse-cleaned fabric
filters is presented. Each model is
evaluated for its ability to predict
average pressure drop from pilot plant
data. The best model is chosen and used
in conjunction with pressure drop
reduction data from an electrostatically
stimulated fabric filter (ESFF) pilot
plant to produce a model of ESFF
performance. The accuracy of the
models is limited by their primitive
nature and the size of the pulse-jet
performance data base. Where the
baghouse, dust, and fabric to be
modeled are very similar to the pilot
plant from which the model was
developed, the model should perform
adequately for comparison between
ESFF and non-ESFF baghouses.
Published correlations relating equip-
ment size and cost are used in a model
for predicting the capital and operating
costs of conventional pulse-jet bag-
houses. A comparison between predict-
ed capital costs and independently
obtained esitmates shows that the
baghouse cost model is capable of
±20% accuracy. A prototype design for
ESFF hardware is developed and cost
quotes from vendors are incorporated
into a predictive equation for ESFF
costs. Because there are no pulse-jet
ESFF baghouses, the prototype design
is subject to revision. This lack of
certainty in the hardware design restricts
the accuracy of ESFF cost predictions
to ± 30%. The cost model is best used in
comparing cost estimate of ESFF and
non-ESFF pulse-jet baghouses and in
comparisons of different sizes of
conventional pulse-jet baghouses.
The performance and cost models are
incorporated into a computer program
for two different computers: the Tek-
tronix series 4050 computers and the
TRS-80 Model 1, III, and 4 microcom-
puters. The program requires pulse-jet
design data as input and predicts
average pressure drop, capital cost,
operating cost, and net present value.
Complete program documentation is
also included.
This Project Summary was developed
by EPA's Industrial Environmental
Research Laboratory. Research Triangle
Park, NC. to announce key findings of
the research project that is fully docu-
mented in a separate report of the same
title (see Project Report ordering
information at back).
Introduction
Electrostatic stimulation of fabric filtra-
tion (ESFF) involves applying an electric
field to the surface of a fabric filter to
enhance the collection of particulate
matter. An added benefit of ESFF is a
reduction of the pressure drop across
the filter. This technology has been
successfully demonstrated in the labora-
tory on a pilot scale pulse-jet fabric filter.
Preliminary estimates show that this
technology may also be economically
feasible.
This report summarizes work done for
the U.S. Navy and the U.S. Environmental
Protection Agency to develop a computer-
ized model of pulse-jet ESFF performance
with the ability to predict capital and
operating costs as well.
-------
Performance Model
Design of a pulse-jet fabric filter
requires predicting the maximum and
average pressure drops for the filter and
for the gas flow through the entire
system, and determining the maximum
penentration of particles and agglomerates
through the filter. Predicting the cost of a
fabric filter is usually more sensitive to
the pressure drop calculations since the
collection efficiency is generally greater
than 99 percent. Thus, for the perform-
ance model in this cost analysis, only the
pressure drop is considered.
The state of the theoretical and empiri-
cal models for predicting pressure drops
in conventional pulse-jet fabric filters is
not well developed. The empirical models
are limited to the conditions under which
they were developed. The theoretical, or
quasitheoretical, models have not been
well tested for conditions other than
those used by their authors; furthermore,
they require some data for determining
unknown constants. Thus it is necessary
to rely primarily on full-scale and pilot-
plant experience for designing new
systems.
As with the theory of pulse-jet filter
pressure drops, the theory of the effect of
electric fields on fabric filter pressure
drop is limited to rationalizations of pilot-
plant observations. The laboratory studies
that have been done do not yield sufficient
information to extend the results to full-
scale baghouses. Of the pilot-plant
studies reported in the literature, that
referred to earlier is the most applicable
for this study. Consequently, the models of
baghouse performance will rely on user
experience for the prediction of conven-
tional baghouse pressure drop and on the
referenced operating experience for the
electrostatic enhancement effect. The
data requirements for each model are
described below.
Figure 1 shows how pressure drop
increases as the amount of dust on the
fabric (W) increases for a typical pulse-jet
filter. The pressure drop increases rapidly
just after a bag has been cleaned. After a
dust cake becomes established on the
filter, the pressure increases linearly with
time. The slope of the linear portion of the
curve is called K2, the specific resistance
coefficient. There is no satisfactory way
to predict a priori the nonlinear portion of
the pressure drop curve. The simplest
way to predict the overall performance is
to postulate an effective residual pressure
drop Pe (i.e., the pressure drop just after
cleaning) and assume that the pressure
drop increases linearly for the entire
• max
p,
Figure 1.
t(W0=C,nVtj
Pressure drop versus cycle time for pulse-jet fabric filters.
filtering time. The pressure drop across
the bag can then be written:
P = Pe + K2VW (1)
where V is the filtration velocity. To
develop a pressure drop model independ-
ent of filtration velocity, it is necessary to
define the drag: S = P/V. Substituting into
Equation (1 ):
S = Se + K2W (2)
where Se is the "effective" residual drag
just after cleaning: Se = Pe/V. There is no
satisfactory way to predict Be-
Thus, to design a conventional fabric
filter for a given set of inlet conditions
(including inlet dust concentration,
velocity, and cycle time), either (1) Pe(Se)
and K2, or (2) P, (S,) and the nonlinear
behavior of P (S) with loading must be
measured and used as input to the model.
To extend the conventional pulse-jet
model to the ESFF pulse-jet, the change
in these parameters with the applied field
must be predicted. Since few data are
available to predict these effects, rough
approximations must be made based on
available data.
Experimental data from ESFF units are
reported in terms of the pressure drop
ratio (PDR) defined as:
PDR = (Pmax-Pr)ESFF/(Pmax-Pr)conventional (3)
where Pmax is the maximum pressure
drop (see Figure 1) that is reached in one
cycle of filtration and Pr is the true
residual pressure drop. If PDR is known
as a function of applied voltage and (Pmax-
Pr) conventional is known, then (Pmax-
PrJESFF can be calculated. Tofind(Pmax>ESFF
the (Pr)ESFF must be known.
The residual pressure drop (Pr) has
been found to be significantly affected by
the presence of an applied field; however,
due to limited testing and scatter in
available data, a correlation could not be
developed between applied field strength
and the reduction in P,. An average
reduction in residual pressure drop of
0.42 with a standard deviation of 0.27
was found for a wide range of operating
conditions. The reduction in residual
pressure drop is defined as:
r — (Pr conventional'",- ESFF)/Pr conventional-
(4)
Lacking further information, the model
for this program uses a reduction in
residual pressure drop of 0.42 when
comparing the ESFF to the conventional
pulse jet. Thus given r and (Pr)convent,0nai,
then (Pr)esFF can be predicted from
Equation 4 and thus (Pmaxhspp can be
found from Equation 3.
Using the earlier observed dependence
of PDR on the applied field, the data were
empirically fit to give:
PDR = 0.77 exp(-0.25f) 5 > f > 0.75 (5]
PDR = 1.0-0.63f + 0.21f2-
0.024f3f <0.75 (6]
where PDR = (P - Pr)esFF/
(P - Pr)conventional, and (7]
f is the applied field in kV/cm.
Field data indicate that an optimum PDR
is achieved at field voltages of 2.5 - 3.C
kV/cm. Operation at applied fields
greater than this would not result in
much better performance.
In summary, due to a lack of a well-
developed theory on the performance ol
conventional pulse-jet filters, a simple
-------
model of baghouse pressure drop (Equa-
tions 1 and 2), requiring user input of
critical parameters, is used in the
performance model for predicting bag-
house pressure drop. If the baghouse is to
be operated as an ESFF unit, the user-
supplied value of the residual pressure
drop for the conventional baghouse is
adjusted to account for the presence of
the electric field (Equation 4). The rise of
the pressure drop over the filtration cycle
is also adjusted, using the estimated PDR
(Equations 5-7) to yield an estimate of
pressure drop performance of the ESFF
baghouse. This method is summarized in
Table 1.
Note that the performance model given
here is based on a limited amount of data.
The quality of the predictions obtained
from this model will depend on the quality
of the user-supplied inputs and on the
similarity of the coal and fabric type used
to those employed during the ESFF pilot-
plant evaluation. Laboratory and theoret-
ical studies are in progress to improve the
understanding of the ESFF mechanism
and its effects on pressure drop.
Economic Measures of Merit
Before the construction of a pulse-jet
baghouse can begin, the optimum design
must be determined. Is a small baghouse
with a large pressure drop better than a
large baghouse with a low pressure drop?
Is the added expense of electrostatic
stimulation worthwhile? An objective
criterion is needed to answer these
questions. For this reason several
economic measures of merit have been
established. The three measures of
merit—savings-to-investment ratio, pay-
back period, and net present value—are
the objective criteria needed for such an
evaluation.
A measure of merit describes the
economic feasibility of a project. For the
Navy, the evaluation of a baghouse falls
under the category of a Fundamental
Planning Analysis (FPA). There are two
kinds of FPAs: Types I and II. A project to
retrofit a baghouse with ESFF hardware
to reduce its pressure drop and therefore
lower operating costs requries a Type I
FPA. The construction of a new baghouse,
either with or without ESFF, requires a
Type II FPA. AType I FPA results in a value
of the Savings/Investment Ratio (SIR),
which is similar to the Return on
Investment (ROI) indicator discussed by
other analysts. Another measure of merit
that can be used to describe a Type I FPA
is the payback period. Either measure of
merit (SIR or payback period) can be used
to determine the economic feasibility of a
project under both Navy and EPA guide-
lines.
The SIR gives a direct measure of the
"profitability" of a proposed modification
in current operations. Before the savings
gained by retrofitting a baghouse with
ESFF are calculated, the annual operating
and maintenance (O&M) costs for the
ESFF baghouse must be computed, then
substracted from the current annual
O&M costs to determine annual savings
in current dollars. This difference must be
positive if ESFF is to offer an economic
advantage over the status quo. Before
determining the economic benefit of an
ESFF retrofit over the remaining life of the
baghouse, the cumulative savings must
be computed, using the cumulative
uniform series factor. This factor, which
discounts the value of money that is to be
paid out in the future to the value at the
present time, is calculated as:
h= exp(n-ln(1+R))- 1 8
ln(1+R)-exp(n-ln(1+R»
where
n = the number of years the annual
payments are to be made, and
R = the effective annual discount rate.
Table 1.
Performance Model
Input:
Option 1—PJSJ Pe(S
-------
annual payment. Consideration of these
factors is beyond the scope of this project,
though it is still desired to estimate the
appropriate measures of merit. There-
fore, the reported measures of merit,
described under Computer Program,
below, are based only on the capital,
operation, and maintenance costs. As
such, the reported values are "before-
tax" estimates of the measures of merit.
For those who wish to pursue a more
rigorous evaluation of the measure of
merit, the capital cost and operation and
maintenance costs are also reported.
Pulse-jet Baghouse Cost
Estimation
Each measure of merit discussed above
requires a knowledge of the initial
investment in equipment and the annual
cost of operating the equipment. The
initial investment is the capital cost of the
project, including the cost of purchasing
and installing the baghouse system. The
method for estimating the costs used in
this project is based on obtaining costs for
the major items in the plant (e.g.,
baghouse, ducting, fan), based on the net
cloth area available for flow. Then Lang
factors are used to estimate the installa-
tion charges and the indirect costs as
percentages of the cost of the major plant
items.
The annual operating and maintenance
costs are calculated based on the size of
the equipment (net cloth area) and the
flow rate of gas. The indirect annual costs
are based on the capital cost of the
equipment.
The procedures used for estimation of
capital and operating costs are too
lengthy to be described here.
Correlations for predicting the cost of
equipment were developed from vendor
quotes. One subtask of this contract was
to develop an independent check of the
accuracy of those correlations. Unfortu-
nately, very little data have been published
in the literature on the cost of pulse-jet
baghouses. As a result, a consultant was
hired to obtain new vendor quotes for
pulse-jet baghouses.
Three different sizes of baghouses
were priced: 26, 85, and 165 acms (55,
180, and 350 k acfm). The equipment
includes the baghouse, its insulation,
woven glass bags and cages, inlet and
outlet manifolds and dampers, hopper
heaters, controls, and structural supports.
The baghouses were specified with air-
to-cloth ratios of 0.02 m/s (4 ft/min).
With this information, earlier correlations
were used to develop the costs in Table 2.
A rigorous effort was made to guarantee
that the vendor quotes were directly
comparable with the predicted numbers.
The agreement between the predicted
values and the reported values is surpris-
ing because both are based on vendor
quotes. Typically such price quotes can
vary by a factor of 2 between different
vendors. Note that the cost modeling
equations are based on design standards
that are at least 7 years old, so that
changes since then are not included in
the costs reported in the table. Note also
that this comparison does not include the
price of ducting, ash conveying systems,
ash ponds, or installation. Also, at the
time of this writing, data were not
available to validate the ESFF costs.
In summary, the equations for predic-
ting the cost of the baghouse, insulation,
dampers, and fabric seem to be in good
agreement with expected values. The
cost predictions of ESFF hardware, other
auxiliary equipment, and operating and
maintenance costs have not been vali-
dated. An earlier report, that cost
equations (except the ESFF hardware)
should be accurate to ±20 percent, is
consistent with the results given here.
The reliability of ESFF hardware-cost
predictions is unknown. The earlier pro-
totype system is quite simple, and it is
likely that some items have been over-
looked. In situations such as this, it is
common to specify the accuracy of the
estimate as ±30 percent. Likewise, for
lack of reliable data, it is assumed that
annual operating and maintenance cost
predictions should be accurate to within
±30 percent.
Computer Program
A computer program has been written
that incorporates the performance and
cost models described above. The program
allows the user to predict pulse-jet
baghouse performance, and then use the
air-to-cloth ratio, air flow rate, and
predicted pressure drop to predict the
capital and annual operating costs. The
appropriate measure of merit (NPV for
new baghouses and SIR and payback
period for baghouses retrofit with ESFF
hardware) is calculated from the predicted
values.
The computer program (PULSEJET)
was developed for two different types of
microcomputers: the Tektronix® series
4050 computers and the TRS-80®
Models I, III, and 4. For the PULSEJET
program to run successfully on a Tektron-
ix® 4051, 4052, or 4054, the computer
must have a minimum of 32 kilobytes of
available memory (RAM). Floppy disk
drives are not required for program
operation. The program can provide hard
copies of the program input and output on
a Tektronix® Hard Copy Unit.
Table 2. Comparison Between Predicted Values and Vendor Quotes
Item
Baghouse
Insulation
Bags
Dampers
Total
Instruments and controls (1O%I
Purchased Equipment Cost
Foundation and Supports (4%)
Capital Cost (December 1977)
Purchased Equipment Cost fCE Plant Cost Index)
(mid-1983)
Predicted Unit Cost ($/mz fabric)
Vendor Quote ($/m fabric)
Ratio: Predicted Unit Cost/Vendor Quote
Q = 26 m3/s,
A = 7,275 m2
109.870
37.910
11.380
2.783
161.943
16. 194
178.137
7.125
185.262
281.526
221
258
0.857
Cost ($)
Q = 85 m3/s
A = 4,274 rrf
354,970
115.310
31.192
6.608
508.080
50.808
558.888
22.356
581.244
883.263
207
215
0.963
Q-165 m3/s
A =8. 194 m3
675.690
216.590
57,034
11.810
961.124
96.112
1.057.236
42.289
1.099.525
1.670,847
204
183
1.115
-------
The minimum system required to run
the TRS-80® version of the program
includes a TRS-80® Model I, III, or 4
microcomputer with 32 kilobytes of avail-
able memory and one 5-1/4-in. floppy
disk drive. The program requires either
the TRSDOS® or NEWDOS® Disk Opera-
ting System (DOS) or any other DOS that
can run Microsoft Disk Basic and is file-
compatible with TRSDOS. The pulse-jet
program is also required to list results on
a line printer.
The PULSEJET program is menu-
driven in both operation and data entry;
it includes default values if input values
are unknown. English and metric units
are available and program execution is
very fast. Complete instructions on
program operation and helpful informa-
tion are available for programmers who
wish to modify the code.
Summary
A simple model of pulse-jet baghouse
performance has been adapted for
prediction of ESFF performance. The
model predicts maximum and average
pressure drops across the baghouse
based on user-supplied parameters. Pilot
plant data have been used to estimate the
reduction in pressure drop that results
from applying an electric field to the fabric
surface. This model of an electrostatically
enhanced pulse-jet baghouse has been
incorporated into a computer program,
along with a model for predicting capital
and operating costs. The program pre-
dicts the performance of a conventional
or ESFF baghouse, along with capital and
operating costs. The program also reports
the appropriate measure of merit (depend-
ing on whether the equipment is new or
retrofit). Complete documentation of the
performance model, the cost model, and
the computer program are available.
•&U. S. GOVERNMENT PRINTING OfFICE: 1984/759-102/10629
-------
A. Viner and B. Locke are with Research Triangle Institute, Research Triangle
Park, NC 27709.
William B. Kuykendal is the EPA Project Officer (see below).
The complete report, entitled "Cost and Performance Models for Electrostatically
Stimulated Fabric Filtration." (Order No. PB 84-207 828; Cost: $13.00, 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:
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park. NC 27711
United States Center for Environmental Research
Environmental Protection Information
Agency Cincinnati OH 45268
Official Business
Penalty for Private Use $300
------- |