Professional Documents
Culture Documents
ISSN 2229-5518
Abstract: The two-stage electrostatic precipitator used as an approach among others to avoid difficulties in collecting of high
resistivity dust through the separation between the charging and collection processes and accomplish them in two different
stages. Removal of cement dust from air stream was investigated by parallel-plate electrostatic precipitator, first by two-stage,
and second by single-stage types. Comparison between the performance of each one to remove the cement dust was carried
out. A system of many functions was provided and designed for this work, consists of air supply with variable flow rate to alter
the residence time in the electrostatic precipitator, dust feeding and mixing with air, and electrostatic precipitator which operated
in a single-stage or two-stage. A high-voltage power supply was used to apply a varying negative voltage (11-15 kv) to the
corona electrode and the collection electrode was grounded. The experimental results show that the collection efficiencies of
ESP were increased with increasing voltage and with decreasing air velocities. Although the collection efficiency of two-stage
(94.5%) is found to be lower than that of single-stage (97.5%), it gave positive results which can be developed to obtain the
required efficiency. Various mathematical models were examined and found that the semi-theoretical and statistical model gave
a good results compared with experimental results (The average absolute error equal to 1.7% for single-stage ESP and 1.4% for
two-stage ESP).
Key Words: Applied voltage, Electrostatic precipitator (ESP),Corona, Collection efficiency, Migration velocity, Single-stage, Two-
stage.
required to initiate corona is that which will An analysis of this equation by Crawford [8]
produce electron energies sufficient to cause lead to the following result which gives the mean
ionizing collisions in the gas species present. Peek field strength as:
[6] has shown semi-empirically that the onset of
corona for smooth round wires and air of relative
2Eoro (0.8 w ro )Cnv
E
density δ at an electric field strength Eo defined by: 0.8 w[(k 2) / kd p2 0.8 wCnvo]
(9)
EO 3.2 10 6 f ( 0.03 ) (1)
2Eo ro (0.8 w ro )
rw
The voltage that must be applied to the wire
0.64w
to obtain this value of field, Vo, is found by And the collecting field strength as:
integrating the electric field from the wire to the
plate. Vo is given by [7]: Eoro (0.8 w ro )Cnv
EC
VO rwEO ln( rc / rw) (2) 0.8 w[(k 2) / 3kd p2 0.8 wCnvo / 3]
(10)
No current will flow until the voltage reaches this
value, but the amount of current will increase ro Eo
steeply for voltages above this value. 0.8 w
A fundamental step in electrostatic At a very low current, the distortion of the
precipitation is the process of charging the electric field by the presence of space charge is so
particulate matter .The particulate must be small that no interaction between the space charge
charged for the electric field to impart a driving and the electrostatic field needs to be considered.
IJSER © 2013
http://www.ijser.org
International Journal of Scientific & Engineering Research Volume 4, Issue 2, February-2013 3
ISSN 2229-5518
IJSER © 2013
http://www.ijser.org
International Journal of Scientific & Engineering Research Volume 4, Issue 2, February-2013 4
ISSN 2229-5518
q p Ec 1 exp(z 2 / 2)
(20) 1 dz (29)
6a 2 exp(Kxg .exp(z ln g ))
Allander [19],18] evaluated analytically the
Collection Efficiency efficiency expression of equation (29) for the
The collecting efficiency of an electrostatic important case of the log-normal particle size
precipitator is generally calculated by means of the distribution as follows:
formula evolved by Deutsch-Anderson which is
for plate-type precipitators state [5],[17],[18]: exp[ (1 / 2) ln 2 g ]
1 f ( , h) (30)
1 exp( L / wu ) (21) 2 ln g
Equation (2-30) can be modified to yield [17]:
where:
1 exp( SC ) (22) = Kxg exp (-ln2σg) ,
Where Sc (s m-1) representing the specific 1
collecting area, h
SC AC / Q (23) 2 ln g
The residence time, t (sec) which a particle above equation represent overall efficiency which
spends in the active space of the precipitator can be solved by two ways graphical or numerical
defined as L/u (equation 17). Similarly, the particle solutions. It is evident from equation (30) that the
separating time τ (sec) equals w/ω .Consequently, efficiency depend only on the two dimensionless
equation (21) can be reformulated as [14],[17]: quantities Kxg and σg. Where the particle-size
η = 1- exp (-t/τ) (24) parameter xg and σg have been found from test
[2],[20],[18].
White [2],[18]derived equation (25) to The theoretical collection efficiency equations
calculate the efficiency of non-uniform particle size were developed on the basis of uniform conditions
distributions from equation (22) by integration among which, the gas flow, which is not realized
methods. in the practical case. Therefore several approaches
were employed to make the resulting expression
more useful for design purpose [6,21].
η = 1 exp( Kx) ( x)dx
0
(25)
In one approach, the concept of non-
Where, K a parameter defined by: uniform particle size distribution is generalized to
include variations in gas velocity, corona current
Ac o E E c density, and electric field. Computations based on
K= (26)
Q this concept have been carried out by Matts and
And, γ(x) represents the size-frequency particle Ohnfeld [18],[22]which leads to an efficiency
size distribution function. equation of the form of equation (31) which is
The log-normal particle-size distribution is the based on semi-empirical and statistical methods
most important one in practice because it holds for rather than on theory[18],[5].
a wide range of industrial and natural dispersions. η = 1- exp (-ωmSc)m (31)
The size-frequency function for the log-normal Where m is a parameter of the statistical spread of
distribution is [18],[2],[16]: the variable quantities usually 0.4 to 0.6,
depending on the standard deviation of the
1
( z) exp( z 2 / 2) (27) particle size distribution and other dust properties
2 affecting collection efficiency, however, the value
Where: of 0.5 gives satisfactory results.
ln x ln x g Experimental Work
z or The schematic diagram of the experimental
ln g pilot plant are shown in figures (3). The air was
x= x g exp( z ln g ) (28) supplied to the system device by means of air
blower. The powder of cement (specification of
Where xg is the geometric mean size and σg the cement was shown in table (1)) was introduced to
geometric standard deviation. the mixing chamber by means of screw conveyor
The expression for efficiency is obtained by from small conical hopper. The polluted air is
White [2],[18] by substitution of these equations directed to the ESP unit to be cleaned from dust
(27 and 28) in equation (25) yield:
IJSER © 2013
http://www.ijser.org
International Journal of Scientific & Engineering Research Volume 4, Issue 2, February-2013 5
ISSN 2229-5518
COMPOUND COMPOSTION %
CaO 63.6
SiO2 22.1
Al2O3 6.5
Fe2O3 3.8
MgO 1.6
SO3 2.4
IJSER © 2013
http://www.ijser.org
International Journal of Scientific & Engineering Research Volume 4, Issue 2, February-2013 2
ISSN 2229-5518
11
14
6 12
4
13
5
3 7
1
2
8
8
1-blower, 2-orifice meter
3-mixing chamber, 4-screw feeder 9
E 2
5-motor for screw, 6-cement vessel
7-ESP, 8-cascad
impactor or filter compartment, A B
9-rotameter 10
10-vacuume pump, 11-H.V.D.C.
Power supply, 12-H.V. Probe
13-current meter, 14-volte meter
A, B- Valves. E - Earth
One sampling train
IJSER © 2013
http://www.ijser.org
International Journal of Scientific & Engineering Research Volume 4, Issue 2, February-2013 2
ISSN 2229-5518
1350.00 90.00
1200.00
80.00
barbed wire
1050.00
single-stage 70.00
Corona Current (microamper)
50.00
750.00
40.00
600.00
30.00
u= 0.7 m/s
450.00 20.00 u= 1.0 m/s
u= 1.4 m/s
300.00 u= 1.7 m/s
10.00
150.00
5.00
1.00 10.00 100.00
0.00 Particle Diameter (micrometer)
10.00 11.00 12.00 13.00 14.00 15.00 16.00
70.00
size distribution is often described by log-normal
60.00
parameters, the two parameters needed to describe
50.00
it are the mass median (or mean) diameter and the
40.00
geometric standard deviation.
30.00 u= 0.7 m/s
To obtain the particle size-distribution with air
inlet
velocities, the mass on the four discs were obtained 20.00
outlet
IJSER © 2013
http://www.ijser.org
International Journal of Scientific & Engineering Research Volume 4, Issue 2, February-2013 3
ISSN 2229-5518
Efficiency Percent
shown in figure (7), which attributed to only part 98.00
experimental
99.90
99.00
98.00
99.50
Two-Stage
Applied Voltage= 15kv
Efficiency Percent
theoretical
95.00
98.00 experimental
90.00
95.00
Barbed wire
u = 0.7 m/s
single-stage
90.00
80.00 two-stage
70.00 80.00
10.00 11.00 12.00 13.00 14.00 15.00 16.00
IJSER © 2013
http://www.ijser.org
International Journal of Scientific & Engineering Research Volume 4, Issue 2, February-2013 4
ISSN 2229-5518
Efficiency Percent
ηexp = 1- exp (-ωexp Ac/Q) , Which can be 95.00
affected by the particle size, saturation charge, and Figure (10) collection efficiency of single-and
electric field. Table (3) shows the migration two-stage ESP as a function of specific collection
velocity at various applied voltage for both single- area.
and two-stage ESP where the migration velocity
increases with increasing applied voltage.
Comparison between Various Models for
TABLE (3) Theoretical and experimental migration Cement Dust Collection by Electrostatic
velocity for various voltages at Precipitator
a given velocity (0.7 m/s) for single- and two- Cement dust collection by electrostatic
stage ESP. precipitation has many parameters which control
this process; therefore many researchers present
single-stage ESP two-stage ESP theoretical models for predicting the collection
Vw performance of an electrostatic precipitator. The
ire ω ωexp ηth ηe ω ωexp ηth ηe
(m (m eo xp (m (m eo xp
existing theoretical models of ESP were validated
(kv) by comparing its predictions with existing
/s) /s) % % /s) /s) % %
11 0.1 0.0 97 92 0.0 0.0 94 88 experimental data. These models are:
12 01 75 97 95 94 62 .7 .6 a-Deutch-Anderson equation which use the
13 0.1 0.0 .8 96 0.0 0.0 95 91 average particle size of
14 09 84 98 97 96 73 .2 93 the dust (equation 22):
15 0.1 0.0 .5 97 0.0 0.0 95 93 theo 1 exp( A C / Q)
21 90 99 .5 99 79 .6 .8 b- Overall collection efficiency (log-normal
0.1 0.0 99 0.1 0.0 96 94 distribution model), equation (30):
32 98 .3 04 82 .7 .5
0.1 0.1 0.1 0.0 97
43 21 20 87 .3
IJSER © 2013
http://www.ijser.org
International Journal of Scientific & Engineering Research Volume 4, Issue 2, February-2013 5
ISSN 2229-5518
exp.
Deutch
Efficiency Percent
95.00
overall
(31)
ηsemi = 1- exp (-ωm Ac/Q)m
90.00
Where ωm is the equivalent of a performance of
ω which equal to[15]:
1 80.00
1 exp
ωm = ω ln ( )
70.00
0.20 0.40 0.60 0.80 1.00 1.20 1.40 1.60 1.80 2.00
the collection efficiency; the results were plotted in Figure (12) comparison between three theoretical
figure (11) for single-stage ESP and figure (12) for models for calculating efficiency of ESP and
two-stage ESP on a linear-probability scale. experimental results for two-stage ESP.
overall
98.00 gave close result and converge with increasing
velocity (average absolute error equal to 1.4%).
95.00 Deutch model far away from the experimental
results (average absolute error equal to 2.5%)
90.00 especially at low velocity because did not consider
non-ideal parameters.
80.00 Since the air velocity through the electrostatic
precipitator that required in the industry between
70.00
0.20 0.40 0.60 0.80 1.00 1.20 1.40 1.60 1.80 2.00 (1-1.5m/s), therefore the semi-theoretical model is
Velocity (m/s) the best model for ESP design.
Figure (11) comparison between three theoretical
models for calculating efficiency of ESP and
experimental results for single-stage ESP. CONCLUSIONS
- The main and important contribution of this
As can be seen from figure (11) the semi- present research work is to present such a device
theoretical model provides a better prediction of which known as two-stage ESP which can be used
the collection efficiency than that given by the to avoid back corona discharge, when the dust of
others theoretical models, where the average high resistivity (such as cement or fly ash) is to be
absolute error equal to 1.7% for semi-theoretical removed, where the charging and collection
model, while its equal to 3.0% and 5.9% for Deutch processes can be separated and accomplished in
and overall collection efficiency models two different stages.
respectively. The overall model gave good results - Collection efficiency was increased with
at low air velocity (< 1m/s) while it diverge with increasing applied voltage where its equal to (92%
increasing velocity where re-entrainment was and 88.6%) at 11kv for single- and two-stage ESP
increased until it was become far away from the respectively, while its equal to (97.5% and 94.5%)
experimental results compared with others. at 15kv for single- and two-stage ESP respectively.
Deutch model has a large values compared with - Collection efficiency was inversely proportional
the experimental results, because did not consider with air velocity where its equal to (97.5% and
non-ideal parameters. 94.5%) at 0.7 m/s for single-and two-stage
respectively, while its equal to (84% and 80%) at
1.7 m/s for single-and two-stage respectively.
IJSER © 2013
http://www.ijser.org
International Journal of Scientific & Engineering Research Volume 4, Issue 2, February-2013 6
ISSN 2229-5518
- The collection efficiency was increased with [12] Moor, A.D., "Electrostatics and Its
increasing specific collection area of ESP. Applications", John Wiley & Sons, New York, 1973.
- The mean diameter of input particles (6.4 μm) [13] Cooperman, G., "A new Current-Voltage
was larger than that of output particles (2.7 μm) for Relation for Duct Precipitators Valid for Low and
a given velocity (0.7 m/s), and the average particle High Current Density", I.E.E.E. transaction on
size was increased with increasing air velocity. industry application, Vol. 17, 1981.
- The migration velocity would increase with [14] Vincent, H. James, "Electrostatic Precipitator
increasing the electric field and saturation charge Performance from the Point View of Gas
of the particles. Treatment Time", JAPCA, Vol. 30, No. 4, 1980.
- Although the collection efficiency of two-stage .[16] Hinds, C. William, "Aerosol Technology,
ESP relatively was lower than that for single-stage Properties, Behavior, and Measurement of
ESP at a given condition tested, it gave enough Airborne Particles", 2nd edition, John Wiley &
positive results which may develop to accomplish Sons INC., 1999.
the target efficiency through changing some [17] Storch, O., Albrecht, J., Hejma, J., Pojar, K.,
parameters such as the length of charging section. Urban, J., "Industrial Separators for Gas Cleaning",
- Many mathematical models were examined on Elsevier Scientific Pub.Co., 1979.
ESP and compared with experimental results. The [18] White, J. Harry, "Electrostatic Precipitation of
semi-theoretical efficiency model was found to be Fly Ash", APCA Reprint series, 1977.
the best. [19] Allander, C. and Matts, S., "The Effect of
Particle Size Distribution on Efficiency in Electrical
Precipitators", Stub 52:738, 1957.
REFERENCES [20] Hamoodi, A. Kadim, "Removal of Oil Ash
[1] Technical Support Document for HWC MACT from Air by Electrostatic Precipitator", M.Sc.
Standards, Thesis, University of Baghdad, 1994.
[21] Gooch, P. John and Francis, L. Norman, "A
http://www.epa.gov/epaoswer/hazwaste/comb Theoretically Based Mathematical Model for
ust/tech/tsd.pdf, (February 1996). Calculation of Electrostatic Precipitator
[2] White, J. Harry, "Industrial Electrostatic Performance", Fine Particulate Control
Precipitator", Addison-Wesley Pub. , 1963. Technology, APCA reprinted series, 1975.
[3] Raouf, S. Tawfeeq, "Some Aspects of Computer [22] Matts, S. and Ohnfeld, P., "Efficient Gas
Simulation in the Field of Cleaning with SF Electrostatic Precipitators",
Electrostatic Precipitation", M.Sc. Thesis, bulletin of A. B. Svenska Flaktfabriken, Stockholm,
University of Baghdad, 1994. Sweden.
[4] Gary, P. Greiner and McKenna, D. John, [23] Mustafa, Y. Abdul Azez, "Removal of Cement
"Electrostatic Precipitators", Dust from Air by Granular Bed Filter",
http://www.eny.hut.fi/research/combustion- PH.D.Thesis, University of Baghdad, 2004.
waste/publications.htm, 1995.
[5] Lloyd, D.A., "Electrostatic Precipitator Hand
Book", Adam Hilyer, 1988.
[6] Oglesby, S., "Electrostatic Precipitation", Marcel LIST OF SYMBOLS AND
Dekker INC., New York, 1978. ABBREVIATIONS
[7] Particulate Matter Controls, Ac = Surface collection area (m2).
http://www.epa.gov/ttn/catc/dir1/cs6ch3.pdf, a = Radius of particle (m).
EPA/452/B-02-001, (September 1999). Cd = Discharge Coefficient.
[8] Crawford, M., "Air Pollution Control Theory", Cmv = Particle concentration (kg/m3).
McGraw-Hill, New York, 1976. Cnvo= Particle concentration at the entrance of
[9] Turner, H. James, Greiner, P. Gary, Vatavuk, M. ESP (particle/m3).
William, "Electrostatic Precipitators", Journal of Cnv = Particle concentration at the end of ESP
Electrostatics, Vol. 48, 1999. (particle/m3).
[10] Allen, N., "Field Theory", Charles E. Merrill dp = Particle diameter (m).
Books INC., USA, 1966. E = Electric field strength (v/m).
[11] Davidson, J.H., "Recent Trends in Electrostatic Ec = Collecting field strength (v/m).
Precipitation", http://www.Icac.com, Control
E = Mean electric field strength (v/m).
Technology Information, 2000.
Eo = Corona onset field at the wire surface (v/m).
IJSER © 2013
http://www.ijser.org
International Journal of Scientific & Engineering Research Volume 4, Issue 2, February-2013 7
ISSN 2229-5518