United States
Environmental Protection
Agency
Air and Energy Engineering
Research Laboratory
Research Triangle Park NC 27711
Research and Development
EPA/600/S8-86/030 Jan. 1987
Project Summary
An Interactive Computer Model
for Calculating V-l Curves in
ESPs Version 1.0
Phil A. Lawless and Leslie E. Sparks
The manual describes two microcom-
puter programs written to estimate the
performance of electrostatic precipitators
(ESPs): the first, to estimate the electrical
conditions for round discharge electrodes
in the ESP; and the second, a modification
of the EPA/SRI ESP model, to estimate the
particle collection efficiency of the ESP
operating with the electrical conditions
predicted by the first program. Both pro-
grams, written in the BASIC computer lan-
guage, are designed for the IBM-PC and
compatible computers and the Tandy
2000 computer. The programs require
Advanced BASIC and a color graphics
adapter.
The V-l program allows prediction of
electrical conditions for both positive and
negative corona and for arbitrary dis-
charge electrode diameters and wire-to-
wire spacing. This means that most com-
mercial discharge electrode geometries
can be simulated by proper selection of
wire diameters and spacing.
Both the V-l and performance models
are completely documented in the report,
including a discussion of the theory on
which the models are based. Most of the
report is devoted to instructions, with ex-
amples, for using both models. Guidance
for adapting the programs to other com-
puters is also presented.
The models are quite useful and allow
rapid assessment of expected ESP perfor-
mance under a wide variety of conditions.
This Project Summary was developed
by EPA's Air and Energy Engineering
Research Laboratory, Research Triangle
Park, NC, to announce key findings of the
research project that is fully documented
in a separate report of the same title (see
Project Report ordering information at
back).
Introduction
Electrostatic precipitators (ESPs) are
used to collect particulate pollutants from
a wide range of sources. Considerable
effort has been spent developing com-
puter models for predicting the perfor-
mance of ESPs. The most successful of
these models is the EPA/SRI ESP model.
The predictions of the EPA/SRI ESP model
have been verified for several sources —
especially for coal-fired power plants.
Before the EPA/SRI ESP model can be
used, the electrical operating conditions in
the ESP must be estimated by the user be-
cause the model cannot calculate elec-
trical conditions.
Because electrical conditions in the ESP
are so important, the V-l predictor model
described in this report was written. In the
initial development of the model, it be-
came apparent that rapid mathematical
computation would be necessary. Thus, it
was decided to use analytic approxima-
tions as much as possible and to avoid use
of numerical solutions of partical differen-
tial equations. This led to the development
of several new approximations for the
electrical conditions in an ESP. These ap-
proximations have been checked against
experimental data and the detailed numer-
ical solutions. In all cases, the answers
provided by the approximations were in
excellent agreement.
Model Version 1.0 enables the prediction
of the electrical conditions for round-wire
discharge electrodes. The wires in a given
section can have varying diameters and be
at various spacings. The effects of an im-
mobile, constant space charge can be in-
cluded, and this component of space
charge can decrease by a certain fraction
for each wire, thereby simulating the ef-
fects of real particulate space charge. This
-------
version has been verified experimentally
for clean wires operating in positive or
negative glow corona. Extensions of the
model (including negative tuft corona, pos-
itive streamer corona, dirty wires, and par-
ticulate space charge) can be made but
currently are supported only by theory.
The model can be used to simulate the
electrical operating characteristics of
various commercial discharge electrode
geometries by careful selection of wire
diameters and wire-to-wire spacing.
A version of the EPA/SRI model is cou-
pled to the V-l predictor model so that par-
ticle collection efficiency can be esti-
mated. The entire set of models have been
implemented in the BASIC computer lan-
guage for the IBM-PC and compatibles and
the Tandy 2000 microcomputer.
Prediction of V-l Curves
The first step in predicting the electrical
conditions (current density and electric
field) of an ESP is to predict the V-l curve
for the ESP. The model uses the analytical
solutions for the electric field around a
round wire as the starting point. The
model considers the perturbations of the
field around one wire due to the nearby
wires at the same potential and then
determines the corona onset voltage for
each wire electrode.
The next step is evaluating the current
density on the plate directly opposite the
wire. Both experimental evidence and the-
oretical considerations have shown that
this current density can be described by
the same relations that predict the current
density in wire-cylinder corona devices.
The third step is determining the current
density on the plate at points between the
wires. Several useful approximate relations
for the angular dependence of current den-
sity and electric field have been discovered
and verified by numerical computations.
Experimental evidence also indicates that
the currents emanating from adjacent
wires become truncated at their common
boundary. These conditions make it pos-
sible to evaluate the voltage dependence
of the current from each wire. This
amounts to the prediction of the V-l curve
for each wire.
The last step is evaluating the V-l curve
characteristics of the whole electrical sec-
tion. This depends on interpolating the V-l
curves of each wire at a common voltage
and then averaging the current over the
whole plate area. Thus, the composite V-l
curve extends only from the lowest corona
starting voltage to the maximum wire volt-
age on the individual V-l curves.
A modified version of the EPA/SRI ESP
performance model was written to inter-
face with the V-l predictor model. This
modified EPA/SRI model uses the electrical
conditions calculated by the V-l prediction
model as input data for calculating the per-
formance of an ESP. The performance
model uses the same input data format as
does the V-l model.
The Models
The entire operation of the program is
governed by menus. A menu is displayed
on the screen of the computer with the
user selecting actions to be taken. The
cursor positioning keys (arrow keys) move
the flashing cursor to the desired location
on the screen. The action that follows
depends on the type of menu involved.
There are three types of menus: operat-
ing menus, single-parameter data entry
menus (or forms), and multi-parameter
data entry menus (or forms). In an operat-
ing menu (once the cursor is positioned),
the only possible actions are to press the
key to activate the operation,
the key to abort the operation, or
the cursor keys to move to anothc
location.
In a single-parameter data entry mem
either numbers or letters may be entere
at the cursor location. The entry term
nates by pressing the key. Th
value entered then replaces the currer
value (displayed to the left of the screei
when the replacement operation is ind
cated specifically.
In a multiple-parameter data entry men
the operation proves somewhat mo:
complex. Here, several parameters of tr
same type are grouped together. The ail
of this menu is to allow for quick revie
and modification of all these values. A
though only one set may be displayed c
the screen at a time, cursor operatior
allow other sets to be displayed easil
New values may be entered in the sarr
way as in a single-parameter data men
When values are changed, however, th<
take effect immediately.
Examples of the menus for the V-l pr
dictor model are shown in Figures 1, 2, ar
3. Examples of the menus for the ESP p«
formance model are shown in Figures
and 5.
V-l Model Operating Options
Glow Corona
(Main Menu)
-) Input Options
Calculation Options
Display Options
Calculation Output
Program Configuration Options
Run EPA/SRI ESP Performance Model
Return to MS-DOS
Use the up and down arrows to position the
cursor at the selection desired.
Press (Enter) to choose the selection.
\
Figure 1'. Opening menu of the model
program.
-------
The current file name is: default.esp
New name ?
<Łnter> uses the current name.
+<Łnter> wilt abort the operation.
Figure 2. Example of single parameter
menu.
Program outputs from the V-l model
range from the simple output of the oper-
ating voltage and current density for use
by the performance model to detailed
graphical output for the current density
and electric fields in the electrical section.
The outputs include graphical output of
the voltage current curves and numerical
printout of the curves. Examples of the
output are shown in Figures 6, 7, and 8.
Outputs of the EPA/SRI performance
model include summary performance as
well as detailed performance as a function
of particle diameter. Examples of the out-
put from the performance model are
shown in Figures 9 and 10.
System Requirements
The programs, developed on a Tandy
2000 microcomputer, make full use of all
its graphic capabilities. Because of the
large number of IBM-PCs in use, the pro-
gram was written to include the capability
to run on those machines without modifi-
cation. However, some sacrifices in graph-
ics are noticed when the models are run
in the IBM mode.
Tables 1 and 2 list the minimum hard-
ware and software requirements for run-
ning the program and suggest options that
will enhance the use of the programs.
The models could be rewritten to run on
other computers. However, the modifica-
tions required to adapt the data entry
portions of the program to computers that
do not use Microsoft BASIC would be
extensive.
Plate-plate Spacing (in.)
Temperature (deg F)
Pressure fatm.)
Ion Mobility (u2/Vs)
Polarity
Space Charge (uC/u3)
Spc Chg Decay Rate (/wire)
Environment Inputs
Present Value
9.00
70
1.00
1.79E-04
/Positive)
10.0
0.10
New Value
Enter values (F1)
Return to prior menu (Esc)
Use the up and down arrows to position the cursor at the selection desired.
Use "Enter values" or (F1) to enter all the changed values.
Press (Esc) to return without making any changes.
Figure 3. Data input menu for environmental inputs.
ESP Model Data Input. Main Menu. Select Item to Enter
-) Change type of computer and save new file now 'TANDY
Enter general data eg overall SCA, gas temp
Enter particle size data
Get VI data from VI model file
Enter sectional data
Enter non-ideal data
Store data in file
Read data from file
Run program.
Run VI Calculation (Note be sure to save info first)
Quit and return to MS-DOS
Configure a new printer.
Arrows move cursor up and down. Press (ENTER) to perform function.
Figure 4. Main menu.
-------
Data entry form for general ESP data.
Title EXAMPLE 12 FROM REV 3(10 FROM REV 1)
-)
Inlet dust loading
Particle density
Dielectric constant
Ion mobility
Total SCA
Total length of ESP
Gas flow rate
Gas velocity in ESP
Gas temperature
Gas pressure
Gas viscosity
(C)old side or (H)ot side
Number of electrical sections
Estimated efficiency
Number of iterations
Resistivity
No. of non ideal conditions
RETURN TO MAIN PROGRAM (Esc)
3.00 gr/acf
2.56gm/cm3
100
2.20E-O.4 m2/Vs
360.0 ft2/kacfm
36.0 ft
500,000 acfm
4.00 ft/s
300 deg F
1.00 atm
2.005-04 cp
C
4
99.00%
2
2.00E+10 ohs-cm
2
ARROW KEYS ARE ACTIVE. PRESS (ESC) TO RETURN. ENTER ITEM.
Figure 5. Data entry form for general ESP data.
EPA/ SRI Model Section Output Data
Voltage
Section (V)
/ 55200
What is the section number?
CD
(nA/cm2)
106.7
E Plate
(kV/cm)
3.07
File
default.esp
A section number ofO will abort any update.
Figure 6. Example output from V-l program.
-------
h
r
»
K
6
v ^
r~
nA/cm2
70
60
50
40
30
20
10
0
4
A
Section Voltage-Current Density Relationship
^
X
X
x
X
"
^
y
/
'
0 50 60 70 1
Kilovo/ts i
>(\
^ y
Figure 7. Graphic display of total section current density curve.
CURRENT DENSITY DISTRIBUTION AT 65420 VOLTS
The distributions are symmetrical about the center axis.
but are shown on one side only for clarity. Press (EscJ to abort.
^>
5 10 20 50 100 200 nA/cm2
^
figure 8. Graphic display of current density distribution when Displays is B& W.
-------
Summary output data for non-ideal correction set No. 1
Stated efficiency - 99.2256% Ideal calculation = 99.2127%
Inlet MMO =15.0 microns with sigma -3.6
Gas velocity sigma = 0.00 with 0.00 % sneakage over 4. stages
Cold site ESP
SCA = 360. ft2/1000 acfm
Rapping size distribution - and 6.0 microns with sigma 2.5
Adjusted no-rap efficiency = 99.7595%
No rap emissions 0.00722 SR/ACF OR 0.0188 LB/MMBTU
No-rap outlet MMD =1.5 microns
No-rap outlet sigma =1.9
Quality of fit = 0.9962
Efficiency with rapping = 99.6740%
Emissions with rapping 0.010 GR/ACF OR 0.0255 LB/MMBTU
Corrected outlet MMD =1.9 microns
Corrected outlet sigma = 2.3
Quality of fit = 0.9965
PRESS ANY KEY TO CONTINUE
Figure 9. Results of the model calculations for the first non-ideal data set.
Particle size results for non-ideal set 1
IDEAL
NO-RAP with SNEAKAGE
with RAPPING
Diam
0.3000
0.5500
0.8500
1.2500
1.7500
2.5000
3.5000
4.5000
6.0000
8.5000
12.5000
20.0000
27.5000
65.0000
effic
0.937449
0.927252
0.942833
O.963790
0.980908
0.993077
0.998234
O.999550
0.999944
0.999998
0.999999
0.999999
0.999999
0.999999
pt
0.062551
0.072748
0.057167
0.036210
0.019092
0.006923
0.001766
0.000450
0.000056
0.000002
0.000001
0.000001
0.000001
0.000001
effic
0.997994
0.994156
0.993540
0.994794
0.996516
0.998152
0.999125
0.999550
0.999944
0.000008
0.999999
0.999999
0.999999
0.999999
Pt
0.002006
0.005844
0.006460
0.005206
0.003484
0.001848
0.000875
0.000450
0.000056
0.000002
0.000001
0.000001
0.000001
0.000001
effic
0.99179
0.98682
0.98346
0.98343
0.98898
0.99480
0.99604
0.99560
0.99564
0.99761
0.99911
0.99980
0.99993
0.99999
Pt
0.00821
0.01318
0.01654
0.01657
0.01102
0.00520
0.00396
0.00440
0.00436
0.00239
0.00089
0.00020
0.00007
0.00001
Figure 10. Example output from performance model.
-------
Table 1. Tandy 2000 Hardware Requirements
System Minimum Beneficial
RAM 256 >256
Disks 1 floppy, 2 floppies or
double sided 1 floppy/1 hard
Graphics High resolution
Color Optional Standard or Tandy
CGP-220 for graphics3
Software GW-BASIC
DOS 2.11 or higher
alf high resolution color is not available, the Tandy 2000 can be run as an
IBM configured computer.
Table 2. IBM/PC/PC-XT/PC-AT Requirements
System Minimum
RAM
Disks
Graphics
Color
Printer
Software
DOS
128K
1 floppy,
double sided
Optional
Optional
Optional
BASICA (Advanced BASIC)
2.0 or higher
Beneficial
>256K
2 floppies or
1 floppy/1 hard
High resolution, B&W
Color monitor
Standard
P. A. Lawless is with Research Triangle Institute. Research Triangle Park, NC
27709; the EPA author, Leslie E. Sparks (also the EPA Project Officer, see
below), is with the Air and Energy Engineering Research Laboratory, Research
Triangle Park. NC 27711.
The complete report consists of two parts, entitled "An Interactive Computer
Model for Calculating V-l Curves in ESPs, Version 1.0,"
Paper copy (Order No. PB 87-100 046/AS; Cost: $24.95)
Computer diskettes (2) (Order No. PB 87-126 025; Cost: $90.00)
The above items will be available only from: (costs subject to change)
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
Official Business
Penalty for Private Use $300
EPA/600/S8-86/030
Center for Environmental Research
Information
Cincinnati OH 45268
0000329
<0<0t
------- |