f/EPA
United States Industrial Environmental Research EPA-600/7-80-076
Environmental Protection Laboratory April 1980
Agency Research Triangle Park NC 27711
High Resistivity Behavior
of Hot-side Electrostatic
Precipitators
Interagency
Energy/Environment
R&D Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Approval does not signify that the contents necessarily reflect
the views and policies of the Government, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/7-80-076
April 1980
High Resistivity Behavior
of Hot-side Electrostatic Precipitators
by
Roy E. Bickelhaupt
Southern Research Institute
2000 Ninth Avenue, South
Birmingham, Alabama 35205
Contract No. 68-02-2610
Task No. 10
Program Element No. EHE624
EPA Project Officer: Leslie E. Sparks
Industrial Environmental Research Laboratory
Office of Environmental Engineering and Technology
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
ABSTRACT
The objective of this research was to obtain evidence to
prove a hypothesis explaining the time dependent degeneration
of hot-side precipitator performance.
Previous research by the author had shown that electrical
volume conduction through a fly ash layer is controlled by an
ionic mechanism depending principally on the migration of sodium
ions to serve as charge carriers. Chemical transference experi-
ments, which are conducted under continuously applied direct
voltage for a long time, revealed a significant change in the
concentration of sodium ions in the fly ash that was contiguous
with the electrodes. At the positive and negative electrodes,
there was, respectively, a depletion and an increase in sodium
ion concentration of the ash.
It was hypothesized that the time dependent, deteriorative,
hot-side precipitator performance was caused by a thin film of
ash adhering to the collection plates even though recommended
rapping procedures were being used. Long exposure of a semiper-
manent, thin layer of ash to a negative corona discharge will
produce a depletion of sodium ions in the ash immediately adja-
cent to the plate. The resistivity of the ash film becomes much
greater as the sodium is depleted. This adherent, high-resistiv-
ity ash layer in series with lower-resistivity, continuously
collected fly ash produces an objectionably high resultant resis-
tivity leading to the degradation of precipitator performance
with time.
A miniature corona discharge device was constructed to run
experiments in the laboratory under thermal and environmental
conditions simulating a hot-side precipitator. Principally,
the data were current density-voltage curves taken with either
positive or negative corona using a hand-placed ash layer 5 mm
thick. Three major experiments were conducted to evaluate the
effect on current density-voltage relationships of: long expo-
sure to negative corona, long exposure to positive corona, and
cyclic exposure to negative and positive corona.
It was observed that exposure of a stationary ash layer
to continuously applied negative or positive corona caused a
degradation of the current density-voltage relationship. That
111
-------
is, the current density resulting from a given applied voltage
decreased with time. Furthermore, it was found that the degraded
relationship could be restored to the initial condition by revers-
ing the polarity of the corona and continuously applying the
reverse polarity for a time electrically equivalent to the initial
polarity. Reversing corona polarity on a 24-hour schedule ap-
parently inhibited degeneration of the current density-voltage
relationships. Since a stationary ash layer was used, the ash
nearest the corona wire was not periodically replaced as in a
precipitator. Because of this, there was ambiguity with respect
to the reason for the degradation of the current density-voltage
relation when positive polarity was used. Degradation could
result from the sodium depletion in the ash nearest the discharge
electrode or from the build-up of these ions in the ash nearest
the "c611ection" electrode.
All evidence obtained suggests that the hypothesis stated
above is correct. Furthermore, the data indicate that some form
of reversible polarity corona could remedy the difficulty. Obvi-
ously, steps taken to improve rapping to eliminate the thin
adherent ash layer or to develop a suitable conditioning agent
with respect to resistivity attenuation or ash rapping and reen-
trainment would be desirable.
This report is submitted in partial fulfillment of Contract
68-02-2610, Task 10 by Southern Research Institute under sponsor-
ship of the U.S. Environmental Protection Agency. This report
covers the period October 1, 1978, to March 10, 1979, when it
was completed.
IV
-------
CONTENTS
Abstract iii
Figures vi
Acknowledgement vii
SECTIONS
1. Introduction 1
2. Summary and Conclusions 2
3. Remedial Considerations 4
4. Background 7
5. Experimental Equipment and Procedures 14
6. Results and Discussion 17
7. References 29
-------
FIGURES
Number Page
1 Dielectric strength versus environmental pressure
at 350°C 8
2 Resistivity at electrical failure versus
pressure 10
3 Attenuation of current density with time of con-
tinuously applied direct current voltage;
parallel plate resistivity test cell 12
4 Schematic and electrical circuit diagram for wire-
plate corona discharge device 15
5 Current density as a function of continuously
applied negative corona, 10 kV 18
6 Distribution of charge-carrying ions in an ash
layer after long-time exposure to an applied
direct current voltage 19
7 Current density as a function of applied negative
voltage 21
8 Current density as a function of applied positive
voltage 23
9 Influence of corona polarity reversal on current
density-voltage curves 24
10 Effect of periodic reversing of corona polarity
for various increments of time 26
11 Current density as a function of applied positive
or negative voltages 27
VI
-------
ACKNOWLEDGEMENT
Mr. Sabert Oglesby, Jr. introduced the author to the subject
problem and vigorously supported this work. The author is grate-
ful for the opportunity to conduct this research and appreciates
the helpful conversations held with various colleagues.
VII
-------
SECTION 1
INTRODUCTION
High electrical resistivity has a significantly adverse
effect on the dry, electrostatic collection of fly ash. At a
typical cold-side precipitator temperature of 150°C, fly ash
produced from low-sulfur (<1%), low-sodium (<1%) coal will typi-
cally have a resistivity value in the range 1 x 1011 to 5 x 1012
ohm cm. Of the several options available with which the elevated
resistivity can be combated, one is the use of hot-side precipi-
tation. The selection of this option has increased in recent
years. Between 1967 and 1977, approximately 150 hot-side preci-
pitators were purchased. These units are associated with 25
percent of the total electrical generating capacity of the coal-
fired boilers purchased with precipitators in this period.1 Hot-
side precipitators normally operating at 350°C are used, because
the resistivity of the above described fly ash is usually in
the range of 1 x 109 to 5 x 1010 ohm cm at this temperature.
The increased usage of hot-side precipitators brought forth
problems, usually associated with poor performance, previously
not observed or, at least, little discussed. A principal ob-
servation was the degradation of precipitator performance as
a function of time. When clean, the precipitator functions well;
however, after several weeks of operation, performance gradually
becomes poor with the advent of back-corona. After the precipi-
tator is thoroughly cleaned, good performance returns. One
of the most intriguing observations was current density-voltage
curves indicative of back-corona occurring concomitantly with
ir\ situ and laboratory resistivity values of about 5 x 109 ohm
cm.
This report describes research conducted to determine the
explanation for the above described behavior and to suggest
possible remedial steps to improve performance.
-------
SECTION 2
SUMMARY AND CONCLUSIONS
It has been observed that the performance of certain hot-
side electrostatic precipitators degrades as a function of time.
Thorough washing of the precipitator restores the desirable per-
formance only to attenuate again with time. This degradation
often is concomitant with evidence of back-corona although in
situ measurements suggest no resistivity problems. Cursory
laboratory experiments and additional field observations showed
that explanations for the above behavior based on the effect
of reduced pressure at high altitude or decreased dielectric
strength at hot-side temperatures were not satisfactory.
It was hypothesized that the observed performance is a mani-
festation of electrode polarization. If the rapping process
is imperfect and a thin layer of ash collected using negative
corona remains on the collection plate, this layer will become
essentially void of mobile, postive charge-carrying ions (sodium).
Since the collection plate is a blocking electrode; that is,
incapable of supplying the charge carrying species, it is termed
polarized. Depending on technical discipline, this can be refer-
red to as sodium depletion or an ash layer space charge effect.
This effect should occur due to imperfect rapping for hot-side
and cold-side precipitators that depend on sodium concentration
for electrical conduction rather than adsorbed sulfuric acid
vapor. When the thin layer of ash tenaciously adhering to the
collection plate in spite of rapping has been exposed to the
electrical field for a sufficiently long time, conduction through
the layer is limited to a mechanism other than sodium migration;
for example, the migration of anions such as oxygen. This thin
layer having very high resistivity develops sufficiently high
field strength near the plate to cause dielectric breakdown.
This hypothesis was based on literature to be cited later that:
a) suggested the possibility of the electrode polarization prob-
lem, b) showed the attenuation of current with time of applied
electrical potential, and c) determined that a sodium concentra-
tion gradient was produced in an ash layer subjected to long-
time direct current electrical potential at both cold-side and
hot-side temperatures.
Since the hypothesis was based on evidence obtained using
parallel plate resistivity cells and temperatures both above
and below normal hot-side precipitator operation, additional
-------
information was desired. The data herein reported were obtained
at 350°C using a small, wire-plate corona discharge device con-
taining a hand-placed layer of fly ash. Three experiments were
conducted in which the effect of: a) long-time exposure to nega-
tive polarity corona, b) long-time exposure to positive polarity
corona, and c) cyclic exposure to negative and positive polarity
corona for various time increments were evaluated by determining
either negative or positive current density-voltage curves at
appropriate times.
None of the data obtained conflict with the above hypothesis.
Current density-voltage curves were made to degrade as a function
of time of applied electrical field for coronas of positive or
negative polarity. Furthermore, the degraded current density-
voltage curves could be restored to the initial condition by
reversing the polarity. The data also illustrated the phenomenon
of storing charge carriers using one polarity to delay the degra-
dation of the current density-voltage relationship for the oppo-
site polarity. For an ash of relatively high inherent resis-
tivity, the data showed that in a short period of time (3 to
6 days), a condition suggesting back corona could be developed.
These observations in which the current density decreases as
a function of time of electrical field application are identical
to those found with parallel plate resistivity cells and are
the result of the ionic conduction mechanism in fly ash and the
use of blocking electrodes.
-------
SECTION 3
REMEDIAL CONSIDERATIONS
Under certain conditions a thin, essentially continuous
layer of fly ash adheres to the collection plates even though
good rapping technique is employed. This layer of fly ash is
consequently subjected to an electrical field for a relatively
long time. Since electrical conduction through the layer at
hot-side temperatures is dependent on a finite concentration
of charge carriers (sodium ions), prolonged exposure to the elec-
trical field causes a significant increase in the resistivity
of the layer due to the migration of the charge carriers out
of this zone.
The situation can be aggravated in several ways. Anything
that can contribute to ineffective rapping is undesirable. The
rapping interval and intensity should be optimized and plate
alignment, linkage and suspension should be in proper working
order to transmit the rapping energy to the ash layer.
Usually the subject problem is associated with hot-side
precipitators designed to operate at 350°C (660°F) and higher.
Whether one is considering the initial condition or after precip-
itator performance has degraded, the conduction process through
the ash layer is thermally activated. Therefore, lower precip-
itator temperatures will produce higher resistivity values.
The character of the fly ash can also affect the precipitator
performance degradation. To circumvent high resistivity under
cold-side conditions, hot-side precipitators are often installed
when design coals will produce insignificant concentrations of
sulfuric acid vapor in the flue gas. When these coals also pro-
duce fly ashes with small concentrations of sodium, resistivity
can be high even at hot-side temperatures. The greater the in-
herent resistivity of the fly ash, the more quickly a situation
of degraded performance can develop and the more severe will
be the problem. Another ash characteristic of importance is
interparticle cohesion and ash-layer density. Fly ash composi-
tion and particle size distribution influence these factors.
Low density particulates of extremely small particle size can
produce collected layers of low density and little cohesion.
These layers are difficult to rap as a sheet into the collection
-------
hoppers. A thin layer of ash is often left on the plates. Fine,
low density ashes high in alkaline-earth elements such as calcium
and low in iron concentration characterize these difficult ashes.
At the present time, only a thorough cleaning of precipitator
electrodes positively relieves the subject difficulty for a short
time. The specific and general approaches cited below toward
the solution of the problem are intended only for the reader's
consideration and investigation. Potential remedial solutions
have yet to be proved.
• Reversible polarity corona - In this report the experi-
ments used to demonstrate the cause of the problem also suggest
a potential solution. The periodic use of positive polarity
corona causes the migration of sodium ions into the ash layer
depleted of these ions during the use of negative polarity corona.
Thus a discharge electrode design with systematic reversible
polarity could balance the migration of sodium near the collec-
tion plate yielding an average performance superior to that of
a completely degraded system using continuously negative corona.
• Continuous positive corona - This system would have the
inherent disadvantages of lower current density for a given volt-
age in comparison to negative corona. It was also shown in this
report that the current density-voltage relationship for positive
corona degrades with time. However, it is not known whether
this degradation is related to a depletion of sodium ions at
the ash surface nearest to the corona wire or whether it is due
to a build-up of sodium ions at the "collection plate." In a
precipitator only the latter occurrence would present a problem
since it is only the ash immediately adjacent to the plate that
is of concern. A difficulty associated with a build-up of sodium
at the ash-collection plate interface would require a sink for
the sodium; for example, perhaps a carbon-cladded collection
plate.
• Continuous negative corona with novel collection plate
design - An innovative collection electrode design incorporating
a source of charge carriers would overcome the depletion of
sodium ions in the adjacent ash layer. The most elementary
approach to this solution would be the deposition of a sodium
bearing compound on the collection plates using a washing tech-
nique.
• Improved rapping technique - Perhaps truly clean collec-
tion plates can be produced by conventional rapping if: a) ma-
terials and surface finish are altered, or b) a "soot" blowing
device is designed to periodically blast the collection plate
surfaces.
-------
Conditioning - Potential coal or flue gas additives that
are thought to cause densification, agglomeration or increased
cohesiveness of the ash to promote better rapping should be given
consideration.
Conditioning to attenuate resistivity should also be effec-
tive. The potential agents are compounds with which sodium or
other charge carrying ions can be added to the fly ash. Sodium
compounds can be introduced dry or as solutions to the flue gas
or to the coal feed. It should be emphasized that this conditioning
technique is sensitive to the degree of intimacy attained between
the agent and the fly ash.
If the compound is added to the flue gas and the agent does
not decompose, one probably must depend on coprecipitation of
fly ash and conditioning compound to develop an ash layer having
lower resistivity. If the coprecipitated material is not homo-
geneous or if it is simply deposited on a high-resistivity,
sodium-depleted layer, one would not expect effective conditioning.
However, if a reasonably homogeneous, coprecipitated layer is
deposited on clean plates, it is possible that satisfactory opera-
tion can be extended to a tolerable length of time. It is con-
ceivable that this type of conditioning would only be required
during start-up after outage for precipitator cleaning.
If the conditioning agent is added to the coal feed and
the compound decomposes or is volatilized, fine sodium-rich par-
ticulates can form or a vapor can condense on the fly ash surface.
In either case one should find the conditioning species intimately
associated with almost all fly ash particles. This mode of con-
ditioning probably would require less additive than in the case
of flue gas injection, and as described above, may be required
only during start-up. It is also conceivable under these con-
ditions that the intrinsically conditioned ash could serve as
the driving force for the chemical diffusion of the sodium toward
the collection plate thereby compensating for the existing sodium-
depleted zone that results from the effects of the electrical
field. If this mechanism is operable, one should observe some
improvement in precipitator performance after operating a unit
for some period of time at temperature but electrically deenergized.
Presently preliminary field experiments are being conducted
for remedial procedures involving polarity reversal and sodium
conditioning.
-------
SECTION 4
BACKGROUND
Initial reports2'3 of the symptoms of poor performance de-
fined above came from installations in the western part of the
United States located at altitudes of 5000 feet. This led to
the rationalization that reduced atmospheric pressure played
a role by permitting dielectric breakdown of the interstitial
gas in the ash layer at lower than normal field strengths.
It was also suggested that the higher temperature associated
with hot-side precipitators caused a decrease in dielectric strength,
If true, the allowable value of resistivity for a desirable
oocurrent density would be reduced from the value normally con-
sidered acceptable for cold-side operation.
Careful consideration suggested that these views be evaluated
even though it was known that the degradation in performance
was time dependent. Subsequently, poor hot-side precipitator
performance having the same symptoms as the units at high altitude
occurred at sea level.
Before conducting the principal research efforts reported
herein, a few cursory experiments were executed at reduced pres-
sures and elevated temperature to examine the effect of pressure
and temperature on resistivity and dielectric failure. The tests
were run at 350°C using apparatus previously described. "* The
environment was a simulated flue gas containing 9 percent water
and no sulfur oxides. The pressure in the environmental chamber
was determined by a manometer and controlled with a vacuum pump
and leak valve. The fly ash used had a resistivity of about
2 x 109 ohm cm and was obtained from a unit experiencing the
subject difficulties.
Figure 1 illustrates the average field strength at electrical
failure as a function of pressure for ash layers approximately
2 and 7 mm thick. The data show that the field strength for
dielectric failure decreases linearly with decreasing pressure.
The decrease in field strength for failure associated with a
pressure change from 760 to 635 mm of mercury is insignificant
with respect to the problem being considered. The single data
point at a pressure of 635 mm of mercury using a 2 mm thick speci-
men was obtained to illustrate that under these conditions of
pressure and temperature, a high dielectric failure stress was
possible. The lower values associated with the thicker ash layer
are typical for laboratory resistivity specimens, while the value
-------
20
u
5
oc
D
<
z
•z.
UJ
DC
co
O
15
10
ASH LAYER
THICKNESS
O f 7 mm
D ** 2 mm
I
I
I
0 200 400 600 800
APPROXIMATE PRESSURE, mm of mercury 3868-7
Figure 1. Dielectric strength versus environmental pressure at 350°C.
8
-------
for the thin specimen is typical for iri situ resistivity speci-
mens. The increase in dielectric strength as the test specimen
thickness decreases is commonly observed for ceramic insulating
mater ials.
Resistivity was also determined as a function of environ-
mental chamber pressure. These data are shown in Figure 2.
Current readings just prior to dielectric breakdown were used
to calculate resistivity. Since the electrical field causing
dielectric failure decreased with decreasing pressure, the cur-
rent measurements with which resistivity was calculated were
made at successively lower electrical field strengths. There-
fore, the increase in resistivity as a function of decreasing
pressure shown in Figure 2 is likely a manifestation of the lower
electrical fields associated with the lower pressure.
In the tests described above, the voltage was always applied
for 1 minute before a current reading was taken or a higher volt-
age was applied. Subsequently, dielectric failure tests were
run at 350°C and a pressure of about 500 mm of mercury using
the above mentioned fly ash and another fly ash having a resis-
tivity of 1 x 1011 ohm cm. The ash layers were about 7 mm thick.
In these tests, each successively higher voltage was applied
for 1 hour instead of 1 minute to determine any tendency for
electrical fatigue. The ash with the higher resistivity failed
immediately at 4000 volts in both tests. The ash with the lower
resistivity failed immediately at 4700 volts using the 1 minute
application and failed at 4000 volts after 6 minutes of the 1
hour application. Although these data do not suggest a strong
tendency toward electrical fatigue, additional tests for long
periods of time are needed to evaluate this point.
The results of the superficial experiments detailed above
do not conflict with the observations that: a) the subject hot-
side precipitator problem can be transiently eliminated by wash-
ing and, b) the problem happens at all altitudes. It was con-
cluded that some phenomenon other than the effect of temperature
and pressure was responsible for the time dependent degradation
of hot-side precipitator performance.
Early in 1978 the writer suggested that the characteristic
observations regarding poorly performing hot-side precipitators
could be caused by electrode polarization. Depending on one's
technical discipline, this phenomenon also can be referred to
as sodium or carrier ion depletion, or as a space charge event.
Whatever name is applied, the phenomenon involves the migration
of charge carrying ions of limited, inherent concentration under
an applied direct current voltage to create a region or zone
depleted of charge carriers which produces an ash layer having
higher resistivity. This phenomenon was described5 previously
in detail with respect to ionic conduction through precipitated
ash layers.
-------
10
11
£
u
E
V)
U)
109
108
_L
I
_L
200 400 600 800
APPROXIMATE PRESSURE, mm of mercury 3868-8
Figure 2. Resistivity at electrical failure versus pressure.
10
-------
Briefly with respect to fly ash and hot-side, negative corona
precipitation, the following event would occur. Sodium ions
in the collected ash layer migrate toward the discharge electrode.
No excessive build-up of sodium would be expected because charged
particles are constantly being added to the collected layer and
ash is being rapped off. However, at the positive collection
plate, a different situation exists. As the sodium ions migrate
away from the anode, the conduction process becomes dependent
on another ionic mechanism that requires more energy to activate.
Therefore, that portion of the ash layer contiguous to the col-
lection plate develops a resistivity much greater than the re-
mainder of the collected layer. The collection plate can be
described as a "blocking electrode" because it is incapable of
supplying charge carriers; in this case, sodium ions.
The normal electrical manifestation of this situation is
an attenuation of current as a function of time of applied voltage.
This was observed when chemical transference experiments5'6'7
were conducted both at 60°C and 600°C. The phenomenon will take
place under cold-side or hot-side conditions if the conduction
process is primarily dependent on the inherent, limited concentra-
tion of charge carrying alkali metal ions in the ash.
Figure 3 illustrates the above described effect. When the
circuit was initially closed, a current density in excess of
100 nanoamps/cmz resulted. After 140 hours of continuously ap-
plied voltage, the current density asymptotically approached
a value of <10 nanoamps/cm2. This time dependent degradation
of current carrying ability is similar to the problem faced with
certain hot-side precipitators. The apparent lower limit for
current density is related to the development of a stable layer
of ash through which conduction no longer depends on the presence
of mobile alkali metal ions.
The data shown in Figure 3 were acquired at 350°C using
an ASME, PTC-28 resistivity cell containing an ash layer 3 mm
thick having a resistivity value of about 1 x 109 ohm cm. After
the test was completed, it was determined that the current through
the ash layer varied exponentially with a decrease in voltage
from the maximum applied and that the experimental activation
energy for volume conduction had increased from 0.7 electron
volts to 1.1 electron volts. It is assumed that the later activa-
tion energy is associated with the rate controlling conduction
mechanism prevailing after the sodium ions have been depleted
in the ash layer adjacent to the anode.
It was hypothesized that the above described phenomenon
can be responsible for the time dependent decay of precipitator
performance. All that is required is that a layer of ash contig-
uous to the positive collection plate is not removed by rapping.
If a film of ash remains on the plates for a long period of time,
11
-------
140
120 O-
100
oc
cc
o
40
20
I
I
I
I
20 40 60 80 100
TIME, hours
120
140 160
3858-9
Figure 3. Attenuation of current density with time of continuously
applied direct current voltage; parallel plate resistivity test cell.
12
-------
the current carrying capacity will degrade. Poor rapping design,
low ash-layer density, low interparticle cohesion, high electro-
static attraction of the ash to the plate, and high adhesion
of the ash to the plate will aggravate the situation.
The initial experiment described above was accomplished
with a resistivity test cell. This report is concerned with
the effort to verify the hypothesis using a small wire-plate
corona discharge device. The objective was to demonstrate the
degradation of current density-voltage curves toward positions
of higher resistivity and symptoms of back corona as a function
of time of applied voltage. This was to be accomplished for
both positive and negative corona. It was also desired to demon-
strate that the effect is reversible.
13
-------
SECTION 5
EXPERIMENTAL EQUIPMENT AND PROCEDURES
The general resistivity test apparatus described in refer-
ence 4 was used in this work. The environmental chamber was
altered to accept the electrical feed-throughs for the wire-plate
corona discharge device, and a reversible-polarity, direct-current
voltage source having a capacity of 50 kV was installed in place
of the more precise, lower capacity source used for resistivity
determinations. In all tests, the environment was a simulated
flue gas containing 9 percent water and no sulfur oxides main-
tained at 350°C. Figure 4 shows a schematic of the wire-plate
corona discharge device. The materials used, critical dimensions
and electrical circuitry are self-explanatory in the figure.
After initial short term experiments were conducted to estab-
lish the optimum test conditions and determine current density-
voltage curves for negative and positive polarity under clean-
plate conditions, three long term experiments were run.
EXPERIMENT 1.
The objectives of this experiment were to demonstrate the
degradation of a current density-voltage curve as a result of
a continuously applied negative corona and to determine if the
application of positive voltage for a period of time would restore
the current carrying capacity of the ash layer. After the ash
was placed in the test cell shown in Figure 4, the apparatus
was thermally equilibrated at 460°C for 16 hours in dried nitrogen.
The apparatus was then thermally equilibrated at 350 C in the
presence of the previously described simulated flue gas environ-
ment. A negative corona was established, and the initial current
density-voltage curve was determined. After data for this curve
were obtained the corona voltage was set to an arbitrarily selected
value of 10 kV (negative) to continuously apply a voltage drop
across the ash layer. The resulting current flow was monitored.
Current density-voltage curves were obtained at least one time
every 24 hours. The voltage was reset to 10 kV after each deter-
mination. After negative corona had been applied for 192 hours,
the final current density-voltage curve was taken, and the polarity
was then reversed to positive. After 90 hours of positive corona
with positive current density-voltage curves being periodically
14
-------
^
/
/•
o o
V-
/
^
J
i
TERMINAL CORONA WIRE
INSULATOR CONNECTOR
^_^. INNER WALL
"*" ENVIRONMENTAL
CHAMBER
Y//s\ STAIN LESS STEEL
b#^
r
f
/
CORONA WIRE
0.018 cm
STAINLESS STEEL
o
O
^»^»i-^
^%^<
!
d
r^F
Is
, K
k^-
/
ELECTRICAL
FEED-THROUGH
INSULATOR
FLY ASH
>
,/
/
9 cm DIA. *4
^^^^
;x^x%
'
X^
1^J
\^
^•/•^'•^^x
^<^
^^.
^
y '
\v
,\
X
x^
^x
i
0.54 cm
T
i
%^y
r^\ .
HIGH
VOLTAGE
SUPPLY
_
3858-10
Figure 4. Schematic and electrical circuit diagram for wire-plate
corona discharge device.
-------
taken, a negative current density-voltage curve was established
after which positive polarity was resumed. When a total of 168
hours of positive corona had been impressed on the specimen,
the final negative corona current density-voltage curve was
obtained.
EXPERIMENT 2.
The objective of the second experiment was to determine
the effect of periodic corona polarity reversal on the current
density. The preliminary steps taken in experiment 2 were identi-
cal to those used in experiment 1.
After an initial current density-voltage curve using nega-
tive polarity was established, the continuously applied negative
voltage was selected to produce 15 nanoamps/cm2 current density.
After 24 hours had passed, another negative current density-
voltage curve was obtained. The polarity of the applied voltage
was then reversed to positive. A current density-voltage curve
was taken, and the continuously applied positive voltage was
selected to give a current density of 15 nanoamps/cm2. After
24 hours had elapsed, both positive and negative current density-
voltage curves were recorded. These data were then compared
with curves established after the first 24 hour cycle of negative
polar ity.
For a period of 320 hours, a sequence of polarity reversals
was maintained. The number of hours that a given polarity was
used was varied. Principally, an attempt was made to show the
effect of shortening the period of positive polarity between
two extended periods of negative polarity.
EXPERIMENT 3.
The objectives of the third experiment were to determine
if the use of positive polarity would lead to a condition of
electrode polarization and what effect this might have on sub-
sequent current density-voltage curves using negative polarity.
As in the case of experiments 1 and 2, a fresh layer of high
resistivity fly ash was placed in the corona discharge device,
and the aforementioned preliminary steps were taken.
A positive voltage to produce a current density of 15
nanoamps/cm2 was maintained continuously for about 120 hours
with current density-voltage curves taken periodically. After
this, the corona voltage was reversed and maintained again at
15 nanoamps/cm2. Negative current density-voltage curves were
obtained periodically until the test was terminated.
16
-------
SECTION 6
RESULTS AND DISCUSSION
When an electrical field is maintained across a layer of
ash for an extended period of time, the current conducted through
the layer will gradually diminish, asymptotically approaching
some minimum value. This was illustrated in Figure 3 for an
ash layer sandwiched between parallel plates in a resistivity
test cell. In Figure 5 similar data are shown for a 5 mm thick
ash layer resting on a "collection plate" exposed to a corona
wire at a negative 10 kV potential. At time zero the current
density is commensurate with the inherent conductivity of a
"freshly collected" layer of ash on a previously clean collection
plate surface. This is shown for two fly ash layers having dif-
ferent electrical conductivities due to variation in sodium con-
centration. These hand-placed, stationary layers of ash are
meant to simulate the situation in which a layer of ash on a
collection plate of an electrostatic precipitator fails to dis-
lodge and fall during rapping. As the time of exposure to the
electrical field increases, the current density decreases.
It has been hypothesized that the attenuation of current
density is a result of the depletion of charge carrying ions,
principally sodium, in a very thin film of ash at the anode
(collection plate) surface and the concomitant build-up of these
ions at the ash layer surface nearest the negative corona wire.
This distribution of charge carriers is schematically illustrated
in Figure 6. In practice this situation results from a thin
film of ash remaining on the plates after rapping.
It can be assumed that the initial concentration of mobile
charge carrying alkali metal ions, Cj, is adequate to produce
a resistivity of 1 x 1010 ohm cm and that an ash layer 5 mm thick
has been collected. Because of the failure to rap cleanly, a
layer of ash depleted of positive charge carriers develops up
to the distance x. If it is assumed that this layer depleted
of positive charge carriers is about 0.05 cm thick and that the
average resistivity for this region is 1 x 1013 ohm cm, then
it can be calculated that the resistivity of the entire 5 mm
thick layer is about 1 x 1012 ohm cm. The localized field strength
in this region can be great, leading to dielectric failure even
though the dielectric strength of this region may have been en-
hanced. This example suggests an explanation for back corona
17
-------
HIGH RESISTIVITY ASH
LOW RESISTIVITY ASH
5 mm ASH LAYER
SIMULATED FLUE GAS
350°C
80 120
TIME, hours
160
200
3868-11
Figure 5. Current density as a function of continuously applied
negative corona, 10 k V.
18
-------
/
-4- - 4 _4-4--4-4-
4-4- — 4- — 4- - 4- - 4- 4
•-•»•- + -+-
4 *
4 4
4 4
ASH
LAYER
NEGATIVE
DISCHARGE
ELECTRODE
INTERRACIAL
CHEMICAL 02t+ 4e~ <- 2O=
REACTIONS
4Na°
DISTANCE
3858-12
Figure 6. Distribution of charge carrying ions in an ash layer after long-time
exposure to an applied direct current voltage.
19
-------
being observed, while in situ and laboratory resistivity deter-
minations made over a relatively short period of time yield
values of about 1 x 1010 ohm cm.
Electrical conduction does not cease but approaches some
minimum value as illustrated in Figure 5. After there is little
decrease in current-density with the additional passage of time,
it can be assumed that a stable film of ash, essentially void
of mobile, positive charge carriers, has developed and that con-
duction is dependent on another mechanism. Glass technology
suggests that under these conditions conduction takes place via
the migration of oxygen ions toward the collection plate result-
ing in the formation of neutral oxygen molecules with the release
of electrons. Because of the low concentration of unassociated
and unbound oxygen atoms and the increased energy requirements,
the resistivity of this thin ash film is great. It can be specu-
lated that "minimum" current densities for the two ashes shown
in Figure 5 will remain separated because the ash with the greater
initial concentration of mobile alkali metal ions will necessarily
have a greater number of mobile oxygen ions. Suitable electro-
chemical expressions are given in Figure 6 to illustrate the
charge exchange at the ash layer boundaries while maintaining
electrostatic balance in the system. In the subsequent discus-
sion an attempt is made to analyze the results of the three pre-
viously described long-time experiments in terms of the above
interpretation regarding electrode polarization or sodium de-
pletion.
The initial experiment involved a layer of high resistivity
fly ash being exposed first to a negative corona for a long time
followed by exposure to a positive corona. The effect was evalu-
ated by periodically obtaining either negative or positive current
density-voltage curves. In Figure 7 the initial negative current
density-voltage curve, A, was obtained after the negative voltage
corona had been on for 6 minutes. Comparing this with the clean-
plate curve indicates that the fly ash had a relatively high
inherent resistivity. After impressing the negative 10 kV corona
on the ash layer for 192 hours, the negative current density-
voltage curve, B, was obtained. The dramatic shift to lower
current density for a given voltage illustrates the increase
in effective resistivity for the ash layer. At this point, the
fly ash contiguous with the "collection electrode" was depleted
of mobile sodium ions and the ash surface nearest the negative
corona wire contained an excess of these ions, as depicted in
Figure 6. Current density for a constant voltage had reached
a "minimum" value with respect to time. Without physically
disturbing the apparatus, the polarity was reversed, and after
90 and 168 hours of continuously impressed positive corona, the
negative current density-voltage curves, C and D, respectively,
were taken. After 90 hours of positive polarity, the negative
20
-------
30
I I i i I i r
ASH LAYER EXPOSURE TO ELECTRICAL FIELD
0.1 HOURS NEGATIVE
192 HOURS NEGATIVE
192 HOURS NEGATIVE PLUS 90 HOURS POSITIVE
192 HOURS NEGATIVE PLUS 168 HOURS POSITIVE
25
20
CLEAN
PLATE
15
o
VI
Q.
-------
current density-voltage curve was partially shifted back to its
original position. The shift to the original position was
complete after 168 hours of positive corona. These observations
are interpretable using the depleted charge carrier idea, and
the events are reversible by changing polarity. The distorted
distribution of charge carrying ions produced by an electrical
potential of one sign apparently can be eliminated by changing
polarity until an approximately equivalent charge is passed in
the opposing direction.
Positive voltage-current density curves were also obtained
during experiment 1. These results are shown in Figure 8. The
distorted carrier ion distribution produced by 192 hours of con-
tinuous negative corona also adversely affected the current carry-
ing capacity under positive polarity. The ash surface nearest
the corona wire contained an excess of alkali metal ions, the
positive charge carriers. The first positive corona current
density-voltage curve, A, in Figure 8, showed an undesirable
situation. As the time of continuously applied positive voltage
was increased, the current density-voltage curves improved to
B and then C. It is believed that continued positive voltage
would ultimately cause curve C to degenerate back toward B as
the carrier ion concentration becomes distorted in the opposite
direction. This point will be demonstrated when experiment 3
is discussed below. Comparison of curve C in Figure 8 with Curve
A in Figure 7 shows the inherently lower current density produced
by positive corona for a given voltage in contrast to negative
corona. The greater current density produced by a given negative
corona voltage is probably due to the presence of free electrons
in addition to negative ions to serve as charge carriers in the
interelectrode space.8 This observation may be exaggerated be-
cause of the rather small distance between discharge and col-
lection electrodes.
The observation that the undesirable degradation of current
density-voltage curves can be arrested by a change in polarity
suggested the second experiment. Again negative and positive
current density-voltage curves were used to evaluate the effect
of periodically reversing the corona polarity. Variation in
the time increment for each polarity was examined. Because of
the lower current densities associated with positive corona,
it was of interest to see if relatively long periods of negative
corona could be used with relatively short periods of positive
corona.
Figure 9 shows the data acquired using a 24 hour schedule
of alternately negative and positive corona. Twenty-four hours
of continuously applied negative corona caused a shift in the
negative current density-voltage curve from A to B. At this
point the voltage was reversed to positive, and the curve labeled
22
-------
20
I I I I I I I T 1
CURVE ASH LAYER EXPOSURE TO ELECTRICAL FIELD
A O 192 HOURS NEGATIVE CORONA AND 0 HOURS POSITIVE
B D 192 HOURS NEGATIVE CORONA AND 90 HOURS POSITIVE
C A 192 HOURS NEGATIVE CORONA AND 168 HOURS POSITIVE
16
(N
u
>
a
ra
O
n
C
Ill
Q
cc
cc
12
I
I
9 10 11
VOLTAGE. kV
12
13
I
14 15
3858-14
Figure 8. Current density as a function of applied positive voltage.
23
-------
CM
U
CA
a
ID
o
c
(0
GO
Z
LLJ
O
cc
-------
C was obtained. After twenty-four hours of continuously applied
positive corona, the positive curve D and the negative curve
E were obtained. Apparently the negative current density-voltage
relationship which had been degraded by a continuous negative
potential producing a current density of 15 nanoamps/cm2 could
be restored by operating under positive corona for a similar
length of time and current density.
Figure 10 shows the results obtained when the time increment
for negative potential was maintained at 20 to 24 hours while
the time increment for positive potential was decreased from
24 to 4, then 2, and then 1 hour. For two cycles while the posi-
tive increment was 4 hours (Figure 10 abscissa, about 100-160
hours), it would appear that the negative current density-voltage
relationship was being reasonably restored. However, when the
positive cycle was reduced to 2 hours and then 1 hour (Figure
10 abscissa, about 160-200 hours), the negative current density
did not return to its initial highest level and, with respect
to time, quickly fell to its lower level. In the later stages
of the experiment, the positive cycle was lengthened again to
24 hours. The current density values for negative corona once
again approached their initial condition.
It would intuitively seem that the quantity of electricity
passed using positive polarity would have to equal that passed
under negative polarity to restore the negative current density-
voltage curve to an initial condition. However, the results
shown in Figure 10 do not completely substantiate this point,
and the issue deserves additional examination. It should also
be considered that in practice the positive corona would not
have to overcome a surface saturated with positive charge carry-
ing ions, because the surface layer of collected ash nearest
the corona wire is probably renewed by the sequence of rapping
and additional collecting.
A third experiment was conducted to demonstrate the effect
of a long-time exposure to positive corona using fresh fly ash.
The experiment was essentially identical to the first experiment
described except that the initial corona was positive. The
results are shown in Figure 11.
The first positive current density-voltage curve, A, was
taken a few minutes after the start of positive corona. After
approximately 120 hours of continuously applied positive polarity
corona at a current density of 15 nanoamps/cm2, the positive
current density-voltage curve, B, was obtained. The degradation
of the current density-voltage relationship shown by curves A
and B was expected due to the distortion of the charge-carrying
ion distribution. In this case, a deficiency of sodium ions
at the surface nearest the corona wire and an excess at the "col-
lection plate" would occur. It is believed that curves C in
25
-------
O NEGATIVE POLARITY
D POSITIVE POLARITY
40
160 200
TIME, hours
Figure 10. Effect of periodic reversing of corona polarity for various
increments of time.
26
-------
60
50
40
a
re
o
c
re
c
t 30
00
Z
UJ
a
UJ
C 20
O
10
TIME
10:00
8:00
8:10
8:05
8:25
POLARITY
POSITIVE
POSITIVE
NEGATIVE
NEGATIVE
NEGATIVE
I
I
10 11 12 13 14
APPLIED VOLTAGE, kV
15
16
17
18
19 20
3868-17
Figure 11. Current density as a function of applied positive or
negative voltages.
27
-------
Figure 8 and D in Figure 9 would also show degradation with addi-
tional time of continuous positive polarity corona.
After 120 hours of positive polarity corona had been used,
a negative current density-voltage curve was taken, curve C of
Figure 11. For the remainder of the experiment, negative polarity
was used and curves were obtained after 24 and 72 hours had elapsed,
D and E, respectively. The distorted concentration profile of
sodium ions in the ash layer due to the initial use of positive
polarity produced an interesting effect when the experiment was
converted to negative polarity. Note that after 24 hours of
continuous negative polarity, the current density-voltage relation-
ship, curve D, had not degraded. This is in contrast to the
degradation illustrated by comparing curves A and B in Figure 9.
After about 72 hours of negative polarity corona, the current
density-voltage curve, E in Figure 11, had degraded to the level
of B in Figure 9. The additional time required to degrade the
curve under negative polarity was due to the excess of sodium
ions available at the "collection plate" as a result of the
previously applied positive polarity corona.
28
-------
SECTION 7
REFERENCES
1. Walker, A.B. Operating Experience with Hot Precipitators
on Western Low Sulfur Coals. Presented at: American Power
Conference, April 18-20, 1977. Palmer House, Chicago,
Illinois.
2. Marchant, G.H., Jr., and Gooch, J.P. Performance and Economic
Evaluation of a Hot-Side Electrostatic Precipitator. EPA-
600/7-78-214, Environmental Protection Agency, Research
Triangle Park, North Carolina, November 1978.
3. Rugg, D.E., and Patten, W. Voltage and Current Relation-
ships in Hot-Side Electrostatic Precipitators. EPA-600/7-
79-044a, pp 421-431, Environmental Protection Agency, Re-
search Triangle Park, North Carolina, February 1979.
4. Bickelhaupt, R.E. Measurement of Fly Ash Resistivity Using
Simulated Flue Gas Environments. EPA 600/7-78-035, Environ-
mental Protection Agency, Research Triangle Park, North
Carolina, March 1978.
5. Bickelhaupt, R.E. Electrical Volume Conduction in Fly Ash.
APCA Journal 24 (3): 251-255, 1974.
6. Bickelhaupt, R.E. Volume Resistivity-Fly Ash Composition
Relationship. Environ. Sci. & Technol. 9 (4): 336-342,
1975.
7. Bickelhaupt, R.E. Surface Resistivity and the Chemical
Composition of Fly Ash. APCA Journal 25 (2): 148-152,
1975.
8. McDonald, J.R., Anderson, M.H., and Mosley, R.B. Charge
Measurements of Particles Exiting Electrostatic Precipi-
tators. EPA-600/7-80-077, Environmental Protection Agency,
Research Triangle Park, North Carolina, April, 1980.
29
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
I. REPORT NO
EPA-600/7-80-076
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
High Resistivity Behavior of Hot-side Electrostatic
Precipitators
5. REPORT DATE
April 1980
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
a. PERFORMING ORGANIZATION REPORT NO.
Roy E. Bickelhaupt
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Southern Research Institute
2000 Ninth Avenue, South
Birmingham, Alabama 35205
10. PROGRAM ELEMENT NO.
EHE624
11. CONTRACT/GRANT NO.
68-02-2610, Task 10
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Task Final; 10/78-10/79
14. SPONSORING AGENCY CODE
EPA/600/13
15. SUPPLEMENTARY NOTES jERL-RTP project officer is Leslie E. Sparks, Mail Drop 61,
919/541-2925.
s. ABSTRACT
repOr|. gives results of experiments to explain the high resistivity
behavior of hot-side electrostatic precipitators (ESPs) collecting fly ash. The wor-
king hypothesis is that the behavior is the result of the buildup of a thin layer of
sodium-ion-depleted fly ash which has a high electrical resistivity near the collector
plate. The hypothesis was tested by experiments in a miniature corona discharge
device under thermal and environmental conditions simulating a hot-side ESP. Cur-
rent density- voltage curves were taken with positive-negative corona using a hand-
placed ash layer 5 mm thick. Three major experiments were conducted to evaluate
the effect on current density -voltage relationships of: long exposure to negative cor-
ona, long exposure to positive corona, and cyclic exposure to negative and positive
corona. Exposure of a stationary ash layer to continuously applied negative or posi-
tive corona degraded the current density -voltage relationship. All evidence suggests
that the examined hypothesis is correct. The data also indicate that some form of
reversible polarity corona could remedy the difficulty. Improved rapping (to elimi-
nate the thin adherent ash layer) or development of a suitable conditioning agent
(with respect to resistivity attenuation or ash rapping and reentrainment) would be
desirable.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution Cleaning
Electrostatic Precipitators
Electrical Resistivity
Fly Ash Treatment
Dust
Electric Corona
Pollution Control
Stationary Sources
Particulate
Conditioning
Rapping
13B
131
20C
2 IB
11G
13H
14B
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
36
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
30
------- |