Journal of Electrostatics, 10 (1981) 73--80
73
Elsevier Scientific Publishing C o m p a n y , A m s t e r d a m -- Printed in T h e Netherlands
SENSING OF BACK DISCHARGE AND BIPOLAR IONIC CURRENT
S. MASUDA and Y. NONOGAKI Department
of Electrical
Engineering,
University
of Tokyo,
Tokyo
(JAPAN)
ABSTRACT A novel type of ionic current probe is developed in order to diagnose back discharge
inside an electrostatic
severity.
It is called as a "bipolar current probe",
measurement
precipitator
of positive and negative
and to identify aualitatively
the separate
ionic current density in the bipolar ionic
field. The probe can be used also under monopolar indicated
and it enables
that the portion of positive
condition.
The probe diagnosis
ionic current density from back discharge
amounts as high as 30-40 % of the normal negative
ionic current density from discharge
electrodes when an intense back discharge is occuring at dust resistivity ohm-cm.
its
The theory of the probe and the results of its measurement
of i012-10 lh
are presented.
i. Introduction The performance electric:
of an electrostatic
mine the particle
charge and its charging rate. ~ e n
dust, an abnormal phenomenon positive
precipitator
(ESP) depends on the magnitude
field strength and ionic current density in collection
called back discharge
of
fields, which deter-
collecting high resistivity (ref. i) takes place, providing
ions to the field to result in a reduction of negative particle charge and
the field intensity as well.
In this case, the degree of decrease in particle charge
is determined by the ratio of positive already reported by Pauthenier The authors developed
disturbance
one, as
(ref. 2).
a "bipolar current probe", which enables the on-line measure-
ment of positive and negative a negligible
ionic current density to negative
ionic current density at any point inside an ESP, with
to the measuring
field.
In this PaDer are reported the theory
of the bipolar current probe and the results of its experiments
thus far obtained.
2. Theory Figure i shows the schematic of the bipolar current probes of both spherical and cylindrical
types, each having three electrodes
A, B and C. The probes are inserted
into a bipolar ionic field so that the central plane D becomes perpendicular local electric
field, as illustrated
in figure 2. The measuring electrodes
0 3 0 4 - 3 8 8 6 / 8 1 / 0 0 0 0 - - 0 0 0 0 / $ 0 2 . 5 0 © 1981 Elsevier Scientific Publishing C o m p a n y
to the
A and B,
74 A : ELECTRODE A B : ELECTRC,DE B C : CENTERELECTRODE 6 : 6UARDELECTRODE
I Bipolarcurrentprobe
(a) S p h e r i c a l Type 2a=4.6ram, d=l. 2rran
(b) C y l i n d r i c a l Type 2a=3.0ram, d=l. O m m
i=i0.Onun Fig. 1
Fig. 2
Bipolar current probe,
Field configulation around probe at balanced conditloo.
and the detecting electrode C are connected through a current meter and wires to a variable dc high voltage source to be applied with the same voltage, Vo (figure 3). When the magnitude of V o is adjusted to become equal to the original local potential of the probe center point,
the electric lines of force in its surrounding takes
a symmetric configulation respect to the central plane D (figure 2). Then, measuring electrodes, ionic current,
the
A and B, accept respectively only the negative and positive
I- and I+, which can be easily calculated as
(a) Spherical probe:
I_ =-3~a2i-cos2(d/2a),
I+ = 3~a2i+cos2(d/2a)
(i)
(b) cylindrical probe:
I- =-4al i_cos(d/2a),
I+ = 4al i+cos(d/2a)
(2)
where a = radius of probe, d = width of electrode C, 1 = length of C in a cylindrical probe, and i+, i- = positive and negative ionic current density this "balanced condition",
(see appendix),
Under
the number of electric lines of force entering into the
electrode C is equal to that leaving from it. But, a slight current is detected by C owing to the difference in the mobility and space concentration of positive and negative ions. It is evident that this probe can also be used in the usual monopolar ionic field.
Vm
Needle electr°deV ~ Bipolarcurrentprobe T / (sphericaltype)
[
Ic
, ......... I
;
Figures 4 and 5 shows, for the monopolar '
and bipolar ionic field respectively, 0
' Fig. 3
condition can be derived by the following methods: i) Graphical method:
-i ~
The values of I_ and I+ at the balanced
~ VariableH.;.so~urcel Diagnosis of back discharge in a point-to-plane electrode.
the
values of I+, I_, and I c as the functions of V o. Theory predicts that each curve in
75 [~A] -IOC
~ ~
J
Vm=3OkV
Vm=3OkV
•
L
*~-5C
~ -5C
~jN ~I ~2 P -~ Probe P o t e n t i a l
Fig. 4
•
-15 Vo
8
,
o
_;'; .,?
c -2C
[kV]
Probe P o t e n t i a l
Fig. 5
Probe currents vs. probe potential in monopolar ionic field (spherical probe).
Vo
[kV]
Probe currents vs. probe potential in bipolar ionic field (spherical probe).
both monopolar and bipolar cases consists of three parts, two straight lines with different gradients (L and N) and a curved line connecting them (M). In the monopolar case the second straight branches coincide with the horizontal axis. Theory further tells that the extrapolation of these two straight line branches provides a crossing point p on the horizontal axis, which corresponds to the balanced potential (see appendix). This method requires a tedious procedure, and the ionic field may subject to disturbances in the mean time. ii) Use of correction factors at "quasi-balanced potential" It is much easier in the bipolar ionic field to determine the probe potential at Ic = 0. This potential, called "quasi-balanced potential", slightly differs from the true balanced potential owing to the difference in positive and negative ionic current density. But its deviation as well as the deviations of I+ and I_ can be calculated theoretically.
Thus, the values of I+ and I_ at the true balanced potential to be used
in equation (i) and (2) can be obtained by dividing their values at the quasi-balanced potential by the correction factors k+ and k- indicated in figure 6. These correction factors are calculated for different values of the probe parameters, d/2a, and depicted as the functions of
"~ ~ o
key [ d/2a ba ,I0.333
I"'=
c
1.4'\
d
Type Spherical
0.333--
0.200 Cylindrical
~ k ~
~.6
--
1.3 ~ \ , ~
/a bu
i~(true) =
=~*i*(m~asured)~7
,~ ~
).5
1.1 ~
O ~ f~
I+/I_ at I c = 0. In the monopolar field
~
the method as described above
~
fails because the exact
~
determination of the quasi-
~.8
~
balanced potential (Ic = 0)
19
~,~
becomes extremely difficult
o~
(see point q in figure 4).
,o'
02
o!4
o.'s
0.8
'1o
Measured value of -I~/I~ Fig. 6
Correction factors k+ and k- vs. I+/I_.
76
However,
theory indicates that, at the balanced potential in this case, the ratio of
Ic/I- takes a definite value depending upon the type and geometry of the probe:
(a)
Spherical probe:
Ic/I_ = tan2(d/2a)
(3)
(b)
Cylindrical probe:
Ic/I_ = sec(d/2a) - 1
(4)
In the use of this probe in a back discharge field the negative ionic current from the discharge electrode is partly obstructed by the probe to produce its shadow on the dust deposit on the collecting electrode. As a result no back discharge occurs (figure 7). Then, the positive ionic current detected by the lower electrode becomes extremely low or almost zero. This error can be avoided by providing a gas flow perpendicular to the field lines at a low velocity,
the magni-
tude of which must be determined by the probe dimension and its distance from the collecting electrode.
~ischarge
Shadow
hadow
(a) without airstream (b) with Fig. 7
airstream
Formation of negative ion shadow and its movement by air flow.
3. Verification of theory The saturation charge of a spherical particle by field charging under bipolar condition is derived by Pauthenier as I - ~ Qbi
1 where
Qmono represents
field intensity.
+
~+/i_
(5)
Qmono
the saturation charge under monopolar condition at the same
The authors verified experimentally
the probe theory in the preceding
section based on this equation. Figure 8 represents the experimental apparatus used, which produces a bipolar ionic field with a desired field intensity and current density.
The saturation charge of a
steel ball with 5 mm in diameter is measured and compared with its theoretical value from equation
(5), where i+/i_ measured by the probe is used. In order to supply
positive and negative ions, a pair of planer ion sources are located in parallel to each other. Each one of the ion sources comprizes a series of parallel,
strip-shaped
discharge electrodes attached on a glass plate. On the back side of each plate is attached a metal plane serving as an exciting electrode. When an ac exciting voltage is applied between the discharge and exciting electrodes,
alternating surface dis-
77 charge occurs along the whole edges of the discharge electrodes over the entire inter-strip
surfaces
so that plasma appears
to form a planer ion source.
The monopolar
ions
of the same polarity as each ion source is supplied from the plasma to the charging field when a main dc voltage and negative
is applied between the two ion sources.
The positive
ionic current density can be controlled by changing the magnitudes
exciting voltages,
Vex + and Vex-, where
the main field intensity
of the
is kept constant at
E = 2 kV/cm. The steel ball is hanged by a nylon string in the charging field for a sufficient time to acquire the saturation
charge. After this, the supply of ions are stopped.
This must be made rapidly and simultaneously
at both sides. Otherwise
the charge on
the steel ball may be altered by ions remaining in the field. Then, the steel ball is brought into inside of a Farada
~ •
saturation charge measured. s~ plate
so~ _
i
age and its
The total current
density at the measuring position,
i = i+ + i-
is kept small at 0.3 mA/m 2 so that the effect
Discharge electrode Excitinq
of ion recombination
can be neglected.
The saturation charge at a different position, h, in the field is measured and plotted in figure 9, for various exciting voltages, Vex+ and Vex_.
It can be seen that the mean
value of the saturation charge remains almost the same at different
~1 ]r~ Faradaycaqe
]kHz FiR. 8
~c]
~
positions.
Experimental verification
apparatus for the of probe theory.
{O [Vex+ = 0.0 kV] Vex2.7 I"01
3
measured by • Spherical probe o cylindrical probe
[Vex+ = 1.3 kV] Vex_ 2.7
2
05
IN?
[Vex + = 1.9 kV]
Theoretical
O -H
5
75
I
I
10 "
h [cm]
I
0 o
q
Saturation charge of s steel ball in bipolar ionic field (Vm = 30 kV, i = 0.3 mA/m2).
~i
1.0, ,~-~L~
Current density ratio i+/i_
[Vex+ = 2.7 kV] Vex_ 2.2 Fig.
05
Fig. i0
Normalized saturation charge vs, current density ratio.
78
After this, the bipolar current probe is inserted instead of the steel ball, and the values of i+ and i_ are measured at h = 7.5 cm under the same conditions of the charge measurement. measured,
The relationship between the saturation charge and i+/i_, thus
is shown in figure i0 together with the theoretical curve of equation (5),
where the saturation charge is indicated in a normalized form, Qbi/Qmono. ment between the theory and the measured values is ~ery satisfactory, the validity of the probe theory described.
The agree-
supporting
The deviation of the measured values may
be caused by the instability and spacial non-uniformity of the ion sources. 4. Diagnosis of back discharge
(refs.
3 and 4)
The magnitude of both positive and negative ionic current density,
i+ and i_, are
measured using the present probe in a point-to-plane electrode system (figure 3), under back discharge condition of different modes. with dust layer (CaCO B powder or fly-ash,
The plane electrode is covered
thickness t = 1.5 mm), and the dust resist-
ivity is altered by changing the ambient air temperature
from i0 to 80°C. In this
case the probe is set at the quasi-balanced potential and the correction factors (figure 6) is used. The current density ratio, i+/i-, a parameter to represent the back discharge severity,
is plotted against the main voltage, Vm, for various levels of dust resist-
ivity, Pd. Figure ii shows the results for CaCO 3 sample, and figure 12 those for fly-ash. In the case of CaCO 3 powder,
the ratio, i+/i_, measured by the spherical probe
amounts to 25-35 %, suggesting the saturation charge, Qbi, to drop to about 30 % of the ideal value, Qmono"
In this case the back discharge severity tends to decrease
with the increase in dust resistivity,
although the back discharge starts at a lower
voltage. In the case of the fly-ash,
the ratio amounts to 30 % at the resistivity Pd =
8.0 × i0 II ~.cm. At a lower resistivity Pd = 1.4 × I0 II ~-cm, a very weak back discharge occurs only at a high voltage just below the flashover threshold. back discharge can also be clearly detected by the present probe.
50
40
x
The results obtained Flashover
lO I I S p h e r i c a l
o 1.4 x I0 II Cylindrical
, 3(
i
.~
8.0 x i0 ~ Spherical A 8.0 x I0 II Cylindrical •
i 1,4
This weak
(~io cm) TM x iol~
20
x
x i o 12 x io II i
0
10 20 30 40 50 60 Main Voltage Vm
[kV]
J
,/,J
I~0 20
Main
Fig. ii i+/i_ vs. main voltage V m (CaCO 3 powder, measured by the spherical probe).
30
40
Voltage
50
[kV]
Vm
Fig. 12 i+/i_ vs. main voltage Vm (fly-ash).
?9 by the spherical and cylindrical probes show a good coincidence with each other. It should be noted that, when the back discharge takes the streamer-mode
(ref. i), and
extends to reach the probe, the measurement with the present probe becomes difficult.
5. Conclusions A bipolar current probe has been developed to measure separately positive and negative ionic current density in the bipolar ionic field, and the following results were obtained: l)
The theory of the bipolar current probe was presented, and its validity tested by experiments based on Pauthenier's theory.
2)
The bipolar current probe proved to be applicable in diagnosis of the back discharge severity in a corona field.
3)
The positive ionic current density can amount to as high as 30-40 % of the negative one under a severe back discharge condition at Pd = 1012-10 I~ ~'cm. This lowers the particle charge to 30-20 % of its ideal value under normal monopolar condition.
APPENDIX Characteristic of probe currents I+, Ic and I_ differ in the regions (I)-(V): (i) Spherical probe region region region region region
I : V II : -3aE < V III: -3aEsin~ < V IV : 3aEsin~ < V V : 3aE < V
< < < <
(ii) Cylindrical probe
-3aE -3aEsin~ 3aEsin~ 3aE
-4~acoE -4~acoEsin~ 4~aeoEsin~ 4~a~oE
< -4~aeoE < ~ < -4~aeoEsin~ < ~ < 4~aeoEsin~ < ~ < 4~aeoE <
(i) Spherical probe I+ = i+[3va2cos2~
2~a(l-sin~) V] E
= i+[3n(a - 3 ~ V)2] - i-[3~(asin6 - 3 ~ V)2] V] = -i-[-3~a2c°s2~ + 2~a(l-sin6) E 4~asin~
: i+[3~(asin6 - 3 ~ v) 2] - i-[3~(asin6 + 3 ~ V)2] -i_
4~asin6 - V E
(region V)
(region III) (regions IV,V)
-
I_ = i+[-3~a2eos2~
(region IV)
(regions I,II)
I c = -i+----f---v
=
(regions I,II,III)
2~a(l-sin~) V] E
= i+[3~(asin6 + 3 ~ V) 21 - i_[3~(a + ~ E V) 21
(region I) (region II)
80
I_ = -i_[3~a2cos26
where
+ 2va(l-sin~) E
V]
(regions
Ill,IV,V)
I+, I c and I_ = probe current flowing into electrodes A, C and B, a = radius of probe, = d/2a, d = width of electrode C, Co= dielectric constant of vacuum, E = electric field intensity, i+ and i_ = positive and negative ionic current density, and V = deviation of probe potential from the balanced potential.
(ii) Cylindrical
probe ~-26 - 2%Eo E Q]
I+/l = i+[4acos~
(regions
~-2D = i+[4acosD - 2~Co E Q] - i-[4a(cosD = -i_[-4acos& 26 Ic/l = -i+ ~Eo-~
~-2~ + ~
D-6 -cos~) + c - - ~ Q]
I,II,IIl)
(region IV)
Q]
(region V)
Q
(regions
~-D O] - i_[4a(cosD - cosg) + ~eo~ ~+D O](region = i+[4a(cosD - cos6) - --~oE 26 = -i_ ~o---E Q
I_/l = i+[-4acos6
(regions ~-26 - 2~¢o~-~ Q]
-~-D = i+[4a(cosD - cosS) - ~ 0] - i_[4acosD = -i_[4acos6
where
~-2~ + ~
I,II) III) IV,V)
(region 7) ~+2D
+ 2~-oE Q]
Q]
I = length of electrode C, Q = charge induced on the probe per unit length, which D = sin-l(Q/4~aeoE).
(region IT) (regions
III,IV,V}
is proportional
to V, and
REYERENCES (I) S. Masuda and A. Mizuno, J. Electrostatics, voi.3(1977), p. 43. (2) M. Pauthenier, La Physique des Forces Electrostatiques et Leurs Applications, CNRS(1961), p. 279. (3) S. Masuda and Y. Nonogaki, Conf. Record of IEEE/IAS Annual M e e t i n g 1980, p. 912. (4) S. Masuda, Y. Nonogaki, H. Nakatani and T. Oda, Proc. US-Japan Seminar on Measurement and Control of Particulates Generated from Human Activities, Kyoto (1980), p. 66.