N ELECTffOSTATIC PRECIPITATOR SYSTEMS STUDY
'x
puthem Research Institute
Lrmingham, Alabama
D October 1970
NATIONAL TECHNICAL INFORMATION SERVICE
Distributed . . .'to foster, serve and promote the
nation's economic development
and technological advancement.'
•
•
•*
• •
••
-------
AN ELECTROSTATIC PRECIPITATOR SYSTEMS STUDY
Final Report ,To ..
~f-\--'-
THE NATIONAL AIR POLLUTION CONTROL ADMINISTRATION
Cincinnati, Ohio
Contract CPA 22-69-73
SOUTHERN RESEARCH INSTITUTE
2000 9th Avenue S. Birmingham, Alabama 3520S
NATIONATTK&NICAL
INFORMATION SERVICE
AN ELECTROSTATIC PRECIPITATOR SYSTEMS STUDY
Final Report
To
THE NATIONAL AIR POLLUTION CONTROL ADMINISTRATION
Cincinnati, Ohio
Contract CPA 22-69-73
Southern Research Institute
Birmingham, Alabama
October 30, 1970
-------
FOREWORD
This report is a summary of the work performed under Contract
CPA 22-69-73 by Southern Research Institute. The program was adminis-
tered under the direction of the Division of Process Control Engineering,
National Air Pollution Control Administration, with Mr. Timothy W. Devitt
serving as project officer.
SOUTHERN RESEARCH INSTITUTE
TABLE OF CONTENTS
Page No.
INTRODUCTION 1
LITERATURE SURVEY AND PREPARATION OF THE
BIBLIOGRAPHY 2
-REVIEWref THE STATUS OF PRECIPITATOR
TECHNOLOGY 5
Descriptive summary of review of precipitator
fundamentals 5
APPLICATION OF ELECTROSTATIC PRECIPITATORS IN
INDUSTRIAL DUST CONTROL, 25
Cement 26
Pulp and paper 26
Magnesia, 26
Phosphorus.". 28
Lime 28
Gypsum 28
Sulfuric acid 28
Petroleum 28
Iron and steel. 28
Electric power generation. 28
- Nonferrous metals, 29
Incinerators 29
Data sources 29
Descriptive summary of application of electrostatic
precipitators in major application areas 29
RESEARCH AND DEVELOPMENT RECOMMENDATIONS 54
BASIC RESEARCH PLAN 58
REVIEW OF SPECIFICATION AND CONTRACTING
PRACTICE 64
-------
Table No.
1
2
3
LIST OF TABLES
Sources Searched for Preparation of Bibliography
of Precipitator Literature
A Survey of the Use of Electrostatic Precipitators
in Various Industries
Research and Development Plans
Page No.
27
55
-ii-
•OUTHIHN HUIAKCH IHSTITUTI
AN ELECTROSTATIC PRECIPITATOR SYSTEMS STUDY
Introduction
This is the final report under Contract CPA 22-69-73 covering the
systems study of electrostatic precipitators. Work under this contract was
performed during the period from April 15, 1969 to September 30, 1970.
The purposes of this program have been: (1) to identify and assess
the status of electrostatic precipitator technology, (2) to define current
practices in the application of electrostatic precipitators for air pollution
control, (3) to identify new potential areas of application of electrostatic
precipitation, (4) to investigate existing electrostatic precipitator systems,
and (5) to define research and development areas needed to improve the
performance of electrostatic precipitators.
The results of this study are presented in a series of documents
which, in addition to the final report, include:
(1) a comprehensive bibliography of electrostatic precipitator
literature,
(2) a selected bibliography prepared from the comprehensive
bibliography which includes the more pertinent articles in
the literature (this selected bibliography has been indexed
by keyword descriptions for rapid reference to articles
covering a given subject), and
(3) a manual of precipitator technology covering fundamentals
of electrostatic precipitator technology and the application of
precipitators to the control of dust emissions in each of
nine major application areas.
The program has been a cooperative effort between Research-
Cottrell, Inc., Rust Engineering Co., and Southern Research Institute.
Dr. Harry J. White, Head of the Department of Applied Science,
Portland State College, Portland, Oregon, has served as consultant
to SRI throughout the program. Research-Cottrell has provided the
comprehensive bibliography of the literature, the keyword descriptors
for the selected bibliography, and data on cost and application of
-------
electrostatic precipitators for various types of dust control systems.
Rust Engineering has provided information on the application of electrostatic
precipitators in the pulp and paper industry. Dr. White has assisted in the
review of the fundamentals of precipitator operation and in the review of the
data on precipitator performance in the major application areas.
This report is broken down into topics covering the major areas investi-
gated during this program. It covers primarily the methodology of carrying
out the program and a summary of the results which are covered in the
other documents.
Literature Survey and Preparation of the Bibliography
Organization of the literature survey included searching of specific
technical information sources by SRI, followed by a more comprehensive
search by Research-Cottrell.
The literature review was fi-rst prepared covering the major sources
likely to contain articles on precipitators or related subjects. A list of the
sources searched is shown in Table 1. The search covered the period
from the initiation of the abstracting service to June 1969.
From the gross bibliography, a selection was made of articles
containing significant precipitator data as judged by the journal, author,
and title of the paper. From the 300f3ntries in the original bibliography,
about 1000 were selected as being of lasting or timely interest. Each of
the articles in the selected bibliography was provided with descriptors
selected from a thesaurus of about 100 items. The articles, with their
descriptors, were arranged for computer access by keyword.
The bibliography was arranged alphabetically by author and numbered
sequentially. In the back of the bilbiography, all of the descriptors com-
prising the thesaurus were listed with numbers corresponding to the
articles for which the particular descriptor is appropriate. In this way,
the articles covering a particular subject can be retrieved without the
necessity for access to the computer.
SOUTHERN RESEARCH INSTITUTE
-3-
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
Table 1
Sources Searched for Preparation of Bibliography
of Precipitator Literature
Air Pollution Control Association Abstracts.
Electrical and Electronics Abstracts (Science Abstracts, Series B).
Physics Abstracts (Science Abstracts, Series A).
Chemical Abstracts (foreign).
Nuclear Science Abstracts.
Index to Publications of The Iron and Steel Institute.
Fuel Abstracts and Current Titles.
Engineering Index.
Applied Science and Technology Index.
British Technology Index.
Research-Cottrell Technical Information Center.
British Coal Utilization Research Association - Monthly Bulletin
(foreign).
Air Pollution Titles.
Reprint collection of M. Robinson.
Bibliography on Aerosols, 1950-1955, W. J. Sheffy, supplement to
AEC Report SO-10003, 1956.
Bibliography on Aerosols, R. A. Stehlow, AEC Report SO-1003, 1951.
-------
-4-
17. Bibliography of Selected Articles on Electrical Precipitation, Anon.
Cement Mill and Quarry 16, 18 (Feb. 20, 1920).
18. Southern Research Institute Bibliography derived from:
a) Air Pollution Technical Information Center,
b) Defense Documentation Center,
c) Pennsylvania State University Center for Environmental
Studies, and
c) National Aeronautics and Space Administration.
19. Bibliography of Chapter 5, Air Pollution Control, W. Strauss, ed.,
New York, in press.
20. Bibliography of Handbuch der Staubtechnik, W. Koglin, ed.,
Maschinenfabrik Beth, Lubeck, Germany (foreign).
21. Bibliography of Industrial Electrostatic Precipitation, H. J. White,
Addison-Wesley, Reading, Mass., 1963.
22. Bibliography of Electrostatyczne Odpylanie Gazow, J. Lutynski,
WNT, Warsaw, 1965 (ioreign).
23. Bibliography of Ochistka Promyshlennykh Gazov Elektrofiltrami,
V. N. Uzhov, Gaskhimizdat., Moscow, 1967 (foreign).
24. Bibliography of An Introduction to Electrostatic Precipitation in
Theory and Practice, H. E. Rose and A. J. Wood, Constable,
London, 2nd ed., 1966.
25. Bibliography of Sanitary Protection of Atmospheric Air. V. N.
Uzhov, Medgiz, Moscow, 1955 (foreign).
26. Bibliographies of numerous papers on electrostatic precipitation.
-5-
Review of the Status of Precipitator Technology
To accomplish the objectives of assessing the status of engineering
technology on electrostatic precipitators, a review of the literature covering
precipitator fundamentals was made. In addition, visits were made to
discuss the current research in the area of electrostatic precipitation.
These visits included laboratories in the U.S. and in Europe to review
work in some of the areas where the status of technology has not been
clearly defined.
Details of the review of precipitator fundamentals are covered in the
section of the manual covering the various topics of particle charging,
corona generation, dust removal, resistivity, etc. The following is a
descriptive summary extracted from the manual which describes the major
topics reviewed.
Descriptive summary of review of precipitator fundamentals
Electrostatic precipitation utilizes the forces acting on electrically
charged particles in the presence of an electric field to effect the separation
of solid or liquid aerosols from a gas stream. In the precipitation process,
dust suspended in the gas is electrically charged and passed through an
electric field where electrical forces cause the particles to migrate toward
the collection surface. The dust is separated from the gas by retention on
the collection electrode and subsequently removed from the precipitator.
Various physical configurations are used to accomplish these basic
functions of charging, collection, and removal, depending upon the type
of application and properties of the dust and gas.
While particles in a gas stream normally have a small inherent elec-
tric charge, it is orders of magnitude too small for effective electrostatic
collection. Consequently, the precipitation process must provide a means
for particle charging. In all commercial precipitator applications, the
charging is accomplished by means of a high-voltage, direct-current
corona.
Corona generation. Corona, as applied to electrostatic precipitators,
is a gas discharge phenomenon associated with the ionization of gas mole-
cules by electron collision in regions of high electric field strength. The
process of corona generation requires a nonuniform electric field, which
is obtained by the use of a small diameter wire as one electrode and a plate
-------
-6-
or cylinder as the other electrode. The application of a high voltage to
this electrode configuration results in a high electric field near the wire.
The electric field decreases inversely with the radius from the wire
surface.
The corona process is initiated by the presence of electrons in
the high field region near the wire. Electrons for corona initiation are
supplied from natural radiation or other sources and, since they are in
a region of high electric field, they are accelerated to high velocities
and possess sufficient energy so that on impact with gas molecules in
the region, they release orbital electrons from gas molecules. The
additional free electrons are also accelerated and enter into the ionization
process. This avalanche process continues until the electric field
decreases to the point that the electrons released do not acquire sufficient
energy for ionization.
Within the region defined by the corona glow discharge, where ioniza-
tion is taking place, there are free electrons and positive ions resulting
from electron impact ionization. The behavior of these charged particles
depends upon the polarity of the electrodes, and the corona is termed
negative corona if the discharge electrode is negative, or positive corona
if the discharge electrode is positive. Both positive and negative corona
are used in industrial gas cleaning applications; however, the negative
corona is most prevalent within the temperature range of most industrial
applications.
In the case of the negative corona, positive ions generated in the
corona region_as a result of electron impact are attracted toward the
negative wire electrode and electrons toward the positive plate or cylinder
electrode. Beyond the corona glow region, the electric field diminishes
rapidly, and if electronegative gases are present, electrons will be
•captured by the gas molecules on impact. The negative ions thus
generated move toward the collection electrode and serve as the principal
means for charging the dust.
In the corona process, there must be a source of electrons to initiate
and maintain the avalanche process. The electrons are supplied from
naturally occurring ionizing radiation, photoionization due to the presence
of the corona glow, and, in the case of high temperature operations, from
thermal ionization at the electrode surface. For negative corona, elec-
trons are also provided by secondary emission from the impacts between
SOUTHERN RESEARCH INSTITUTE
-7-
the positive ions and the discharge electrode. Cosmic and terrestrial
radiation provide approximately 20 ion-electron pairs per cubic centimeter
of gas per second.
Mobilities of the various charge carriers play an important role in
the corona generation process. Electron mobility in high fields is
approximately 400 times that for ions. In the negative corona case, the
electron mobility is such that sparking would occur when the field required
to initiate corona is reached unless the electrons are attached to gas mole-
cules to form a stabilizing space charge.
In most industrial gas cleaning applications, there are sufficient
quantities of electronegative gases, such as oxygen, so that practically all
of the electrons are attached to gas molecules. Gases such as nitrogen,
helium, argon, etc., do not form negative ions, and hence a stable
negative corona is not possible in these gases.
In positive corona, the electrons generated by the avalanche process
flow toward the collection electrode. Since the positive ions are the charge
carriers, they serve to provide an effective space charge, and the presence
of an electronegative gas is not required for positive corona. Sources of
electrons for initiating and maintaining avalanche in a positive corona are
cosmic radiation and phofoionization due to the corona glow.
Positive and negative corona differ in several important aspects. In
appearance, the positive corona is a rather uniform sheath surrounding the
discharge electrode. In contrast, negative corona appears as localized
discharges from points on a clean wire and as localized tufts along the dust-
coated electrode. The voltage-current characteristics of the negative
corona are superior to those of positive corona at the temperature at which
most precipitators operate. Higher operating voltages and currents can
be reached prior to disruptive sparking. It is postulated that a spark or
arc breakdown in the interelectrode space occurs by formation of a
streamer originating at the positive electrode surface. In the positive
corona, the origin of the streamer would be at the surface of the discharge
wire, and hence in a high field region. In the negative corona, the positive
electrode is the collection plate and the field near this surface is con-
siderably less than at the discharge electrode; hence a higher voltage
would be required for spark propagation.
-------
-8-
Negative corona is accompanied by the generation of ozone, and there-
fore is usually not used for cleaning air in inhabited space. However,
most industrial gas cleaning precipitators utilize negative corona because
of its inherently superior electrical characteristics which lead to increased
efficiency at the temperatures at which they are used.
Geometry of the electrodes, gas composition, and gas conditions have
important influences on corona generation. The diameter of the discharge
wire and the electrode spacing determine the voltage gradient, and hence
the variation in electric field strength. The electric field varies as the
reciprocal of the radius near a small diameter wire. Hence, with a very
small wire, the electric field near the surface can be quite high, often in
the range of 50-100 kV/cm. The avalanche process requires the presence
of a high electric field over a given distance. In general, the smaller
diameter wire requires a higher electric field strength for initiation of
corona. For a given spacing, however, the onset of corona occurs at a
lower voltage for the smaller diameter wire. Also, for a given voltage,
higher currents are obtained with-sraaller diameter discharge electrodes.
Temperature and pressure influence the generation of corona by
changing the gas density. In the avalanche process, the time available
for accelerating an electron between collisions is a function of gas
density. With increased molecular spacing, higher velocities can be
achieved between collisions. Thus, ionizing energy can be achieved with
low electric fields for low gas densities.
A second effect, in the case of the negative corona, is that the
increased molecular spacing results in the penetration of free electrons
further into the interelectrode region before capture to form a negative
ion. This results in an increased average mobility in the interelectrode
space, and hence a higher current. Additionally, at very high tempera-
tures (above about 1500°F) thermionic emission increases, further
increasing the number of free electrons and the effective average charge
mobility. These effects reduce the voltage required for sparkover, so
that at high temperatures, positive corona would perhaps give superior
voltage-cur rent relationships and improved collection efficiency since
electrons move toward the discharge electrode in positive corona.
Corona generation studies of a basic nature are most often made
with clean electrodes under laboratory conditions. These conditions are
highly idealized in comparison to industrial precipitators. In practical
•OUTHERN RESEARCH INSTITUTE
-9-
precipitators, the presence of a dust laden gas has several effects on
corona generation. First, the dust entering the interelectrode space
becomes charged by attachment of negative or positive ions. Because
of the much lower mobility of the charged dust, it constitutes a signifi-
cant space charge. The magnitude of the space charge depends upon the
size and quantity of the dust and the magnitude of its charge. The effect
of the space charge is to reduce the electric field in the vicinity of the
corona glow region and thus it tends to quench the corona and reduce the
current. This effect is particularly significant at the inlet section of a
precipitator where dust concentrations are highest. Special electrode
shapes are often used to minimize this space charge problem at the
inlet section.
A second important consideration of the effects of dust on corona
generation is the deposits formed on both collection and discharge
electrodes. On the collection electrode, dust deposits alter the electric
field and sparking conditions as a result of the voltage drop within the dust
layer. This effect limits the voltage and current at which the precipitator
can operate, and is its chief influence on corona generation.
Dust deposits also form on the discharge electrode in operating
precipitators. This deposit can be quite heavy in the case of some types
of dust. The effect on corona can be considerable depending on the nature
of the deposit and the electrical properties of the dust. If the dust resis-
tivity is reasonably low, the effect of the deposit will be to effectively
increase the diameter of the discharge wire. This results in higher voltage
required for corona initiation or reduced corona current for a given voltage.
If the dust deposit is uneven, an uneven distribution of corona along the
length of the wire may result.
If the dust resistivity is high, the effect generally would be to reduce
corona current for a given voltage. However, if the deposit is somewhat
porous, breakdown of the gas in the interstitial region can occur and the
effect of the deposit may be reduced.
The current in a precipitator is carried by free electrons, ions, and
charged dust particles. The magnitude of the current carried by each of
these is related to the number densities of the carrier, the mobility of the
carrier, and the electric field. Mobilities are related to the various physi-
cal parameters of the charged particle, but primarily to the charge-to-
mass ratio of the carrier. Electrons, with their low mass, have the highest
-------
-10-
mobility of all the current carriers (He~750 cm2/volts sec). Oxygen ions
would have a lower mobility, about 1/400 that of an electron (Ho ~1.9 cm2/
volts sec), and charged dust particles would have the lowest mobility of all
the current carriers (in the range of 0. 02 cm2/volts sec). These mobilities
are for conditions normally encountered in commerical installations. It is
thus apparent that for a given electric field, the current carried by each of
the carriers is in proportion to their mobilities for the same number den-
sities. For a negative corona, the number densities of the charged parti-
cles change in traversing the glow region from the discharge electrode
surface. At the surface, the positive ion density is greatest, since all of
the positive ions from the avalanche process flow to the discharge electrode.
At the boundary of the corona glow region, ion generation ceases; hence
the concentration of positive ions is zero beyond that point.
The number of free electrons at the wire surface is minimal and
ir.creoKes, because of the avalanche process, to a maximum at the bound-
ary of the corona glow region. Moving into the quiescent zone, free elec-
trons are captured by the electronegative gas molecules. The probability
of electron capture is high due to the number of gas molecules present, so
that the electron density rapidly decreases beyond the corona glow boundary.
As the electrons from the corona are captured by the gas molecules
to form negative ions, the negative ion concentration increases. Traversing
further into the interelectrode space, the negative ions attach to the dust
particles present to form charged particles. The number of free ions
present then decreases as they are consumed in the particle charging
process. A considerable fraction of the negative ions present go to
charging the dust particles, although due to the difference in mobilities,
the percentage of the current due to bound ions is small.
The currents carried by these various carriers can be determined by
their number densities, mobilities, and the electric field strength. It
should be remembered, however, that the number densities of the various
carriers, as well as current, are important in analysis of precipitator
- operation.
Particle charging. There are two physical mechanisms by which gas
ions impart charge to dust particles in the precipitator. Particles in an
electric field cause localized distortion of the field so that electric field
lines intersect the particles. Ions present in the field tend to travel in the
direction of maximum voltage gradient, which is along electric field lines.
SOUTHERN RESEARCH INSTITUTE
-11-
Thus, ions will be intercepted by the dust particles, resulting in a net charge
flow to the particle. The ion will be held to the dust particle by an induced
image charge force between the ion and dust particle. As additional ions
collide with and are held to the particle, it becomes charged to a value
sufficient to divert the electric field lines such that they do not intercept
the particle. Under this condition, no ions contact the dust particle and it
receives no further charge. The electrostatic theory of the process shows
that the saturation value of the charge on the particle is related to the
magnitude of the electric field in the region where charging takes place,
the size of the particle, and the dielectric constant of the particle. The
saturation charge is proportional to the square of the particle diameter,
thus larger particles are more easily collected than small ones. This
mechanism of charging is called field-dependent charging.
The time required for a particle to acquire a saturation charge is an
important factor that is often neglected in precipitation theory. Field de-
pendent charging may be described by an asymptotic final value function,
and is dependent upon ion density and mobility. If a single particle is intro-
duced into a high ion density field, charging takes place in milliseconds.
However, charging times by the field-dependent regime can be appreciable
under practical conditions of large dust loading and low currents.
For small particles (diameter <0. 2M), the field-dependent charging
mechanism is less important, and collision between the particles and gas
ions is governed primarily by thermal motion of the ion. Equations de-
scribing the rate of charging can be derived assuming that charging rate
is independent of the magnitude of the electric field. Since ion movement
is so greatly influenced by the electric field, the assumption obviously
generates considerable error. However, the factors influencing charging
rate can be seen under this assumption to be particle diameter, free ion
density, and thermal velocity of the ions.
Since the range of thermal velocities has no upper boundary, there
is no saturation value associated with diffusion charging. However, as the
charge on a particle increases, the probability of impact decreases, so
that there is a decreasing charging rate associated with an increasing par-
ticle charge. This second charging process is called diffusion charging.
Recent work in the area of diffusion charging has been directed toward
obtaining better agreement between experimentally determined values of dif-
fusion charging and those predicted by theory. Studies of the influence of
-------
-12-
electric field have been made on the basis of approximate solutions to the
equations that include the effect of the electric field.
In practical precipitators, field-dependent charging is usually of most
interest, but in some applications particles are present in the range where
diffusion charging is the predominant mode (<0. 2ji diameter) as well as the
area in which both mechanisms are significant. Unified charging equations
have been developed covering the size range where both charging regimes
are important, and results agree reasonably well with experimental labor-
atory values.
Particle charging theory indicates several important factors govern-
ing precipitator performance. Since the magnitude of the particle charge
is dependent upon the magnitude of the electric field in the field-dependent
mode, it is important that field strength be kept as high as practical in
the region where charging takes place.
A second factor of importance is the rate of charging of the particles.
Practical precipitators generally ip+T,duce heavy concentrations of uncharged
dust in the inlet section of the precipitator. Calculations show that the num-
ber of ion» required to charge this dust to its saturation value may be large,
hence the number of free ions present may be substantially !>educ?d and
charging times are not insignificant.
The waveform of the applied voltage is significant, as it influences the
peak value of the electric field and the charging time. The saturation value
of the particle charge is determined by the peak value of the electric field.
However, charging only occurs during the interval of time that the applied
field exceeds the self-field corresponding to the charge on the particle.
Consequently, longer times will be required to reach saturation for a varying
voltage than for a pure d-c voltage. However, the varying voltage is
preferable for spark quenching and permits operation at a higher average
voltage.
The electric field in a precipitator determines the maximum value
of the particle charge due to field-dependent charging and also the force
acting on a charged particle. Since the electrical field enters the collection
efficiency equations effectively as a squared term, it is important that the
magnitude of the field be maintained as high as practical.
•OUTHCRN RESEARCH INSTITUTE
-13-
The electric field strength is determined by the electrostatic compo-
nent, which is related to the precipitator geometry and the applied voltage;
and by the space charge component, which is related to the presence of
charged particles (ions and charged particulate) in the interelectrode space.
The design of the precipitator can be varied to alter the geometry of the
discharge electrode and the electrode spacing. These factors can determine
the magnitude of the electrostatic component. Variations in electrode
geometry can also alter the corona current, which in turn influences the
electric field by changing the space charge contribution. Equations des-
cribing the electric field for clean electrodes have been developed from
classical electrical theory. Utilizing these equations, it is possible tt>
predict the field at any location, assuming clean electrodes.
In a practical precipitator, dust accumulations on the collection elec-
trodes limit the maximum voltage, and hence the electric field strength at
which the precipitator can operate. The voltage drop across the dust
layer is dependent upon the corona current density t the electrical resistivity
of the dust, and the thickness of the dust deposit. For high resistivity dusts,
the voltage drop across the collected dust layer can be in the neighborhood of
10-20 kV, assuming reasonable current densities and dust deposit thicknesses.
Obviously, the electrical energization equipment must be capable of provid-
ing sufficient voltage to accommodate this voltage drop, while maintaining
adequate voltage across the interelectrode spaces. The effect of the resis-
tivity of the dust layer, however, is more severe than its influence on power
supply voltage. The electric field in the dust layer can be quite high for
high resistivity dust. This high field region at the anode surface can lead
to sparkover at lower applied voltages, thus limiting the maximum oper-
ating voltage of the precipitator.
A second condition associated with high dust resistivity can also influ-
ence particle charging and the magnitude of the electric field. Once spark-
ing occurs, a crater is formed in the dust layer, and current densities in
the localized area can result in localized electric fields that are sufficiently
high to initiate a corona emanating from the base of the crater. This co-
rona results in positive ion production and, due to the direction of the elec-
tric field, these ions flow toward the discharge electrode. Collisions with
dust particles tend to charge them with opposite polarity to that required
for collection. Also, collision with negative ions tends to neutralize them,
reducing the ion density in the interelectrode space.
-------
-14-
If dust resistivity is further increased, a diffuse corona glow will
appear over the large areas of the dust surface. Under these conditions
the positive ion production by the reverse corona is sufficient to completely
disrupt the charging process, and effective precipitation is not possible
under these conditions.
Particle collection. The forces acting on a charged particle in a pre-
cipitator are gravitational, inertial, electrical, and aerodynamic. The
latter two are the principal ones of importance in electrostatic precipitation.
If a particle is suspended in a laminar gas flow stream in a pipe and
wire precipitator, a force due to the electric field and particulate charge
will act on the particle in the direction of the collection electrode. This
force is opposed by the viscous drag force of the gas. In sufficient time,
which is short for small particles, the particle would reach a terminal
velocity at which point the electrical and viscous drag forces would be
equal. In precipitator terminology, this is called the migration velocity.
The other force acting on the particle is the aerodynamic force by the gas
stream. The motion of the particle will be along the line defined by the
vector sum of these two forces. Under laminar flow, all particles would
be collected in a given length of the precipitator and the collection efficiency
for shorter lengths would be linearly related to precipitator length.
In practical size precipitators, however, laminar flow is practically
never achieved. Consequently, the turbulent gas flow causes particles to
follow a random path through the precipitator. The magnitude of the forces
due to the turbulent gas flow is large compared to the electrical forces.
However, at the boundary layer the gas flow is laminar,and particles enter-
ing the boundary layer will be collected. The collection efficiency is there-
fore related to the probability of a particle entering the boundary layer.
Studies by Anderson, Deutsch, and White of particle collection in a turbu-
lent gas stream have shown theoretically that collection efficiencies are
exponentially related to the collection surface, the gas volume handled,
and the migration velocity of a particle. The equation, known generally as
the Deutsch-Anderson equation, is of the form 17 = 1 - exp (- _^_w).
vg
The derivations of the efficiency equation are based on the assump-
tion that there is a reasonably constant distribution of the particles in any
cross section of the precipitator due to turbulent mixing of the gas. In
addition to this assumption, there are several basic conditions that apply to
SOUTHERN RESEARCH INSTITUTE
-15-
Utilizing the relation for the theoretical
— generally associated with the efficiency
the derivation of the equation,
2
migration velocity w = —
•/
equation neglects any contribution of diffusion charging. Also, the calcula-
tion is based on a single particle size. The equation also includes no term
to account for reentrainment of collected dusts, uneven distribution of the
gas flow, or other factors inherent in practical precipitator operation.
A principal practical use of the Deutsch-Anderson equation has been in
relating measured collection efficiency to the collecting surface area and gas
volume. In such cases, the term w as calculated from the Deutsch-Anderson
equation is a parameter, rather than the migration velocity given by theoreti-
cal considerations. In this case, it is called effective migration velocity
or precipitation rate parameter. The term is useful in describing the
effectiveness with which a given dust can be collected, and is widely used
in design and analysis of precipitators.
From a theoretical as well as a practical standpoint, the distribution
of particles within the precipitator is important. There is some evidence
to indicate that particle distribution within the precipitator may not be uni-
form and that diffusional forces may also play a role in collection efficiency.
Removal. Once collected, the dust or liquid aerosol must be removed
from the precipitator. This can be accomplished by flowing a liquid down
the collection electrode to wash the collected dust, or by rapping the elec-
trodes to impart an acceleration to dislodge the dust, which falls into a
hopper for subsequent removal. Liquid aerosols normally coalesce and
drain from the plates so that removal is not a problem.
In dry removal systems, rapping of the collection electrode to remove
the dust is normally done on a periodic basis. Successful rapping depends
upon accumulation of sufficient thickness of material on the plate so that it
falls in large agglomerates into the hopper. There is always some reen-
trainment of the dust so that effective rapping must minimize the amount
of material reentrained in the gas stream.
The accelerations required to remove the collected dust vary with the
properties of the dust and gas stream. Forces of cohesion and adhesion
consist of molecular (van der Waals), electrical, and mechanical forces.
Some dusts adhere tenaciously to the collection surface and require sub-
stantial accelerations to dislodge them. Variations in operating tempera-
ture, gas composition, or both, can alter the forces required for success-
ful rapping. Electrical forces, which are related to current density and
-------
-16-
dust resistivity, are also significant in holding the collected material to
the plate, and therefore affect the forces required for rapping. Since
current densities are higher at the discharge electrode than at the collecting
electrode, greater forces are often required to maintain them relatively
free of dust deposits than are required at the collection plates.
Reentrainment of the dust during rapping is evidenced by increased
dust loadings at the precipitator exit following a rap. To minimize this
effect, only small sections of the precipitator are rapped at one time.
Electrical energization. The function of electrical energization is to
provide optimum electrical conditions for particle charging and collection.
The energization equipment consists of a transformer to step up the voltage
from the normal supply line to between 30-100 kV which is required for
precipitation. The particular voltage is a function of electrode (wire-to-
plate) spacing, and the spacing is generally chosen to minimize the influ-
ence of misalignments of the electrodes resulting from faulty construction,
warping, etc. A rectifier, usually of the vacuum tube or silicon type,
converts the alternating voltage to d-c to give unipolar ions. An automatic
control system is usually provided to maintain optimum voltage conditions
for precipitation. The control system can operate from spark rate,
secondary or primary voltage, or other parameters.
The power capacity of the energization equipment is determined by
the precipitator size or quantity of gas flow, design efficiency of the pre-
cipitator, and properties of the dust and gas. The manner in which the
power is applied varies with the application, and the policies of the equip-
ment manufacturer. The principal variable is the number of independently
powered precipitator sections that make up the total power supply.
Increasing the number of sections by using a large number of smaller
power supplies is advantageous if the precipitator is operating in a spark-
limited mode, as a higher voltage can be maintained for a given spark rate.
The size of the power supply also determines the internal impedance of the
supply, hence large power supplies with high current capacities have low
impedances, which are not as effective in spark quenching as the smaller,
high impedance supplies. Smaller power supply sections also tend to
minimize effects of plate misalignment or other localized disturbances.
Finally on large systems, the influence of an outage of a section of the pre-
cipitator is not as pronounced in systems comprised of smaller power
supply sections where many independent sections are used.
SOUTHERN RESEARCH INSTITUTE
-17-
Analysis of the electrical characteristics of a precipitator shows that
it consists primarily of a parallel capacitive and resistive load, the value of
which is determined by the corona current and voltage and precipitator
dimensions. The value of the capacitance is large enough to maintain the
voltage at a high level between cycles, even though the power supply is
unfiltered. The electrical characteristics of the dust deposit also influence
the electrical operation of precipitators. The effect of the filtering charac-
teristics of the dust layer is to smooth out voltage fluctuations so that higher
peak voltages can be applied for a given spark rate if the voltage waveform
has a high rise rate. On the other hand, charging times may be increased
by these voltage waveform alterations, hence for the optimum design, both
effects must be considered.
Systems analysis. Despite its seeming simplicity, the interaction of
the variables makes it difficult to analyze the performance of a precipitator
without considering the system as a whole. The systems analysis approach
permits a review of the various parameters controlling system performance
and permits a rational basis for precipitator design and analysis.
A simplified systems analysis has been developed which relates most
of the important variables and permits a computerized method of precipita-
tor analysis or performance evaluation.
The systems model is programmed to determine the collection effi-
ciency for each discrete particle size range for each increment of length
of the precipitator by calculating the following: the field strength as a func-
tion of radius at each increment of distance through the precipitator, the
saturation charge on the dust, the actual charge on the dust at each incre-
ment from considerations of the free ion densities, and the amount of
material in each size range that is removed at each increment. From this,
the total collection efficiency can be calculated.
Comparisons of efficiencies predicted from the systems model with
those measured in field tests show good agreement, except where unusual
conditions, such as excessive reentrainment, were suspected.
The general philosophy behind the systems model is that it permits a
satisfactory calculation of performance based on known theory, whereas
theoretical calculations based on very elementary or simplified theory do
not. Refinement of the model to include such factors as reentrainment,
voltage-current characteristics, sparkover effects, spatial distribution
of the dust, diffusion charging, and gas distribution, should permit a
-------
-18-
more theoretically correct analysis and should give better agreement
between measured and calculated performance.
Design. The design of precipitators involves the determination of
precipitator size and electrical energization equipment required to give a
desired efficiency, the design of a gas flow system to provide acceptable
gas flow quality, structural design of the precipitator shell and supports,
selection of the rapping equipment, and selection of the electrode config-
uration.
Present design methodology is generally based upon empirical
relations, the values of which have been obtained from experience with
similar applications. There are several approaches to the selection of
the precipitator size. In general, these methods involve the selection of
a precipitation rate parameter and determination of the collection plate
area required from the Deutsch-Anderson equation or from design curves
based upon field experience. The precipitation rate parameter varies for
different applications and often varies considerably within the same appli-
cation area dBe io variations in gas and dust properties.
Selection of the precipitation rate parameter can be made on the
basis of experience with similar installations- or- from experimentally
derived curves relating precipitation rate parameters to dust properties.
For many applications, the range of precipitation rate parameter varia-
tions is small (of the order of ± 10%). In such cases, the uncertainty in
plate area requirements is of the same magnitude. In other instances,
variations can be as high as 400-500%, so that some method for reducing
the uncertainty is highly desirable. In general, some property of the
effluent from the industrial process has been related to precipitation rate
parameter and an empirical relationship is'derived to predict the value of
the precipitation rate parameter. In the case of fly ash precipitators,
sulfur content of the coal and resistivity of the dust are commonly used to
establish this value. Particle size distribution is another significant
variable, and curves relating precipitation rate parameter with any other
variable should be modified to compensate for particle size variations if
sufficient data are available.
Power requirements for a precipitator vary with collection efficiency.
Selection of the power requirements is generally based on curves relating
efficiency with corona power per unit volume of gas flow (watts/cfm). These
curves are experimentally developed for each type of application and vary
SOUTHERN RESEARCH INOTITUTC
-19-
with dust properties. These curves are usually based on total delivered
secondary power, and power supply capacity is selected on the basis of a
power supply efficiency (from 60-75%) and the standard power supply size
that will meet the efficiency requirements.
Sectionalization is also based on empirical information derived from
experience. These curves must be consistent with those based on power
requirements and on relationships involving collection surface area, since
the same efficiencies can be achieved through the use of fewer sections
and greater collection area for installations operating in a spark rate
limited mode.
Design of the gas flow system is generally based on model studies
with large systems, and its importance to good precipitation cannot be
overemphasized.
Selection of the number of rappers, type of electrodes, etc., varies
among manufacturers and with the type of dust being collected. Struc-
tural design is relatively straightforward.
One of the principal difficulties encountered by the users of electro-
static precipitators is the evaluation of bids for specific installations.
Although precipitator bids are based on guaranteed performance, there
are many examples of precipitator installations that fail to meet design
performance by a wide margin. It is not uncommon for bids to vary con-
siderably in the collection surface area, the amount of power and degree
of electrical Sectionalization, and the type of gas flow distribution systems.
The user is therefore faced with the problem of evaluating the adequacy
of each design for meeting his requirements. The ability to interpret bids
requires experience with the particular type of dust or a method of assess-
ing the effect of the design variations on performance.
A technique for evaluating design parameters for given dust and gas
conditions has been developed based on the method of regression analysis.
Based on a group of approximately 75 tests on 20 installations, equations
predicting the precipitation rate parameter as functions of gas and fuel
properties have been developed. An overall correlation coefficient of
about 0. 85 has been obtained, with maximum uncertainties of about 25%
in the precipitation rate parameter for fly ash precipitators. This is
rather good agreement since design and performance parameters may vary
considerably more than for fly ash collection.
-------
-20-
Mechanical components. The two general types of precipitators
currently used are the tubular and plate types. The tubular precipitator
consists of cylindrical collection electrodes with discharge electrodes
located on the axis of the cylinder. Gas to be cleaned passes through the
annular space between electrodes and the dust is collected on the outer
cylinder. This type construction is used in wet electrode systems, in high-
pressure, high-temperature applications, in some types of precipitators
collecting liquid aerosol particles, and for small dry collection installations.
The plate type precipitator consists of parallel collecting plates with
discharge electrodes located between the plates. Collecting electrode plates
are usually 12 to 40 feet high and spaced 6 to 12 inches apart. The length of
the ducts in the direction of gas travel can vary according to the precipitator
design. Total length can be from around 12 to 24 feet or longer for very
large installations.
Mechanical components which make up a precipitator are the shell,
electrodes, hoppers, rappers, support members, and the necessary
electrical feedthrough and support backings.
The shell can be rectangular or cylindrical and can be constructed
of steel, tile or concrete. Thermal insulation is .usually provided in the
case of corrosive gases to maintain the shell above the dew point tempera-
ture to minimize corrosion. Design of the shell generally follows straight-
forward structural engineering practice.
The shell and electrode configurations can be arranged to divide the
gas flow and to provide independent sections that can be separately energized.
When parallel gas flow paths are provided, each path is referred to as a
duct or chamber.
Discharge electrodes in both cylindrical and plate type precipitators
are of a variety of types depending upon the application and the precipitator
manufacturer. The discharge electrodes can be small diameter (~0. 1 inch)
wire, square wire, or fabricated structures of various types. The pri-
mary consideration in discharge electrode selection is to obtain desirable
voltage-current characteristics and to provide mechanical strength for
resistance to corrosion and fatigue. In general, American practice is to
utilize electrode structures supported at the top, with weights at the hopper
end to maintain the electrode taut. Guides are provided to maintain
alignment. European practice is generally to use a mast or frame for
rigid support of the discharge electrodes. Various discharge electrode
SOUTHERN HCKAMCH INBTITUTC
-21-
configurations can be used with either type of mounting.
Collecting electrodes in plate type precipitators are generally flat
plates with various types of stiffeners and baffles. The baffles provide
shielding of the collected dust to reduce reentrainment during rapping
and to reduce scouring of the plates due to gas flow.
Single impact or vibrational rappers are provided for dust removal
in dry type precipitators. Rappers can be electromagnetically, pneu-
matically, or mechanically actuated. The major requirement for
successful rapping is to deliver sufficient impact to the electrodes to
dislodge the dust without causing excessive reentrainment. The
acceleration required to remove the dust varies with the type of dust and
gas composition. Rapping is generally specified in terms of the energy
delivered per rap and the number of rappers per square foot for the col-
lecting surface or per length of wire for the discharge electrode.
Dust removal can be through flat bottom pans with scrapers to move
the collected material to a screen conveyor or through pyramidal hoppers
where it is removed by conveyors or vacuum systems. The latter type is
the most prevalent.
Gas flow. Gas flow in practical precipitators is well within the
turbulent region. When exiting from the process, gas velocity usually is
relatively high and often uneven. Gas velocity must be reduced to a
relatively low level and turbulence controlled before entering the precipi-
tator for good precipitation. Poor quality gas flow can affect precipitator
performance by scouring the plates in localized regions of high gas velocity
and by reducing performance due to the exponential relationship between
efficiency and gas volume flow.
Often, space limitations preclude more conventional methods of
achieving uniform gas flow, and turning vanes, splitters, straighteners,
and diffusion plates must be designed to provide adequate gas flow quality.
The Industrial Gas Cleaning Institute recommends a minimum gas flow
quality such that 85% of local velocities is within 25% of the mean, with
no reading more than ±40% from the mean.
Because of the difficulties in predicting gas flow quality, the use of
physical models is almost universal. Models are usually constructed
of pressed hardwood or acrylic plastic. Smoke streamer patterns and
pilot tube traverses are used to indicate gas flow uniformity. Models
-------
-22-
have been historically constructed — to -|- size with the Reynolds number
held constant. However, with the trend toward larger plants, models con-
structed to — scale have been used. The use of smaller scale models
tends to give less accurate results.
Resistivity. Electrical resistivity of the dust is an important factor
in the performance of electrostatic precipitators. If the resistivity of the
collected dust is higher than about 2 x 10 ohm-cm, excessive sparking or
reverse corona can occur, thereby limiting precipitator performance.
Two distinct types of electrical conduction occur. One type is con-
duction by free electrons within the particles. This type of conduction
depends upon the electron activation energy (a material property) and
temperature. Many industrial dusts are composed of metallic oxides,
sulfates, etc., which have low activation energies so that the electrical
conductivity is low at temperatures in the range of 300-400°F. At higher
temperatures, however, conductivity becomes greater and for most dusts
this is the primary conduction mechanism at temperatures above
450-500°F.
The second kind of conduction is conduction over the particle surfaces
owing to adsorption of moisture or certain chemicals such as sulfuric acid.
Adsorption increases with decreasing temperature and hence particle con-
ductivity also increases with decreasing temperature. Moisture is often
referred to as the primary conditioning agent and other chemical adsorbates
as secondary conditioning agents.
The role of the secondary conditioning agents is not clearly under-
stood in all cases. It is postulated that the secondary conditioning agent
may alter the surface of the dust, thus enhancing the rate of moisture
adsorption and hence the conductivity. The effectiveness of the secondary
conditioning agent varies with the type of dust. There is evidence that an
acid conditioning agent, such as SO3, is more effective in conditioning a
basic dust, whereas a basic conditioning agent, such as NH3, is more
effective in conditioning an acidic dust. There is some evidence that the
conditioning agent and moisture adsorb or condense onto the dust surface
together to form the conductive layer and subsequently may react with the
dust layer to alter the resistivity. Effects of additions of limestone to
particulate emission from sources such as power generator boilers and
sinter machines tend to indicate that reactions between dust and conditioning
agent may detrimentally influence electrical conductivity.
SOUTHERN RESEARCH INSTITUTE
-23-
In some instances of high resistivity dust, additions of a conditioning
agent to the effluent gas have resulted in substantial reductions in resis-
tivity and enhanced collection. Examples are SO, additions to the gas from
power generator boilers and ammonia additions to gas from catalytic
cracking units used in petroleum refining. However, additions of relatively
large quantities of chemical additives have failed to improve performance
in some applications. But since secondary conditioning agents are known
to be highly specific and selective in effectiveness, there is no appropriate
reason for expecting random additives to work. Causes for this condition
are not fully understood.
Measurement of performance. Measurement of precipitator perform-
ance generally includes dust loadings at the precipitator inlet and outlet,
gas velocity distribution, electrical current and voltage input to the precipi-
tator, and gas composition. Dust resistivity is an important parameter that
should be measured, but often is not.
Dust loadings are normally measured by a sampling probe and a
dust collector, such as a thimble. Prior to sampling, a gas velocity dis-
tribution traverse is made so that dust samples can be taken isokinetically
to minimize selective size collection. Sources of error in sampling are
usually attributed to anisokinetic sampling, improper probe handling, dust
collection on the walls of the sampling train or dust collector, and errors
in gas flow measurement. Sampling procedures require careful attention
to detail and the development of skills in the sampling procedures.
Resistivity of the dust sample must be made in-situ to be meaningful.
Laboratory measurements may be high by orders of magnitude and often
fail to correlate with in-situ values.
In-situ resistivity can be measured by determining the current flow
through a given volume of collected dust when a known voltage is impressed
across it. Several types of resistivity probes can be used for this pur-
pose. In some types, the dust is deposited electrostatically utilizing a
point-plane precipitator. Other type probes utilize a cyclone type mechan-
ical collector. The errors involved include those associated with obtain-
ing a representative sample. Particle size and degree of packing can
influence the apparent resistivity of the dust layer.
Troubleshooting and maintenance. Causes for a precipitator to fail
to achieve its design efficiency can be due to inadequate design, electrical
-------
-24-
difficulties, improper gas flow, inadequate rapping, installation problems,
electrode misalignment, poor maintenance, or improper operation.
Indications of electrical difficulties can usually be observed from
the levels of corona power input. Efficiency is generally related to power
input, and if inadequate power densities are indicated, difficulties can
usually be traced to dust resistivity, unusually fine particle size, elec-
trode misalignment, improper control operation, or dust accumulation
on the electrodes. Improper gas flow can be determined by measure-
ment of the gas flow distribution. Poor gas flow can result from improper
design, plugging of distribution plates, build-up of dust on dust walls and
turning vanes, etc.
Improper rapping is usually manifested in excessive dust deposits on
the collection and discharge electrodes. Adequacy of rapping can be
measured by accelerometers mounted on the electrodes. Accelerations
required to dislodge the dust vary with type of dust and operating conditions,
and rapping must be adjusted to maintain the desired thickness of the dust
deposit without excessive reentrainment. Visual observation of the interior
of the outlet flue from the precipitator is often a sensitive and revealing
method for checking rapper operation.
Maintenance schedules should be set up for inspection, servicing,
and repair of critical components. These components include rappers,
feed-through bushings, transformers, electrodes, ash removal equipment,
and electrical controls. Excessive dust deposits, corrosion, broken
wires, etc. are common types of difficulties encountered.
Electrostatic augmentation and unusual designs. Electrostatic
augmentation of fabric filters, packed bed filters, and wet scrubbers
has been studied in some detail. Performance of fabric filters varies
with the type of dust deposit and the thickness of the deposit. Pressure
drop is also generally related to the dust deposit. The filtering action
is generally considered to occur by the trapping of dust on the collected
layer. If a single particle is too large to pass through an interstitial
region, it is trapped on the dust layer and separated from the gas stream.
Electrostatic augmentation involves establishing an electric field
between the fabric and another electrode, precharging the dust particle,
or both. The effect of electrostatic augmentation is that the interstitial
•OUTHCRN KEKARCH INSTITUTE
-25-
openings in the fabric material function as if they were smaller and
hence smaller particles are retained. Its principal advantage has been
in the more rapid build-up of the dust layer and somewhat higher efficiency
for a given pressure drop.
Loose bed filters are notably inefficient in collection of small
particles. However, if charged particles are introduced or if an elec-
trostatic field is impressed across the filter, or both, the dust particles
will deposit on the filter bed. The effect is to provide increased collec-
tion surface area in a two stage precipitator. Greatly increased effi-
ciency can be achieved over that for a conventional loose bed filter. The
potential for increased collection surface would indicate the possibility
of enhanced collection over a two stage precipitator with the same metallic
plate area. The obvious disadvantage is in removal of the collected dust,
which would require liquid backwash or circulation and cleaning of the
filter material.
Augmentation of wet scrubbers is intended to provide better contact
between the particulate and the scrubbing liquid by utilization of the attrac-
tive force between the charged particles. The attractive force between a
charged dust particle and an oppositely charged liquid droplet varies with
the distance of separation. Thus, with separation distances of 10/i, the
attractive force is approximately equal to the gravitational force.
The use of the space charge, developed as a result of the dust charge
or as a result of other charged material, has often been the basis of a pre-
cipitator concept in which no high voltage is applied to the collection plate.
Fields developed as a result of space charge can be quite high pro-
vided the concentration of charged dust is high and there is ample spacing
between electrodes. However, as dust is collected, the space charge
decreases and collection efficiency decreases. Charged water drops have
also been utilized to maintain a higher space charge field. These approaches
are more suited to specialized applications and their practicality for indus-
trial gas cleaning has not been demonstrated.
Application of Electrostatic Precipitators in Industrial Dust Control
The principal use of electrostatic precipitators in the control of
industrial dusts has been in the following major areas.
-------
-26-
1. Electric Power Generation
2. Pulp and Paper
3. Rock Products
4. Iron and Steel
5. Nonferrous Metals
6. Petroleum
7. Chemical Industry
8. Municipal Incinerators
9. Miscellaneous Applications
The use of electrostatic precipitators was reviewed in these areas
to determine the extent of the use of precipitators in each area, the range
of input variables, and the design factors, costs, and problems peculiar
to the use of precipitators in the particular industry. The approach followed
in this study was to: (1) review the process to determine the dust and gas
compositions and properties, (2) determine from the users the extent to
which design, test, and cost data were available, and (3) integrate the data
supplied by Research-Cottrell into an application area report. The reports
from the various industries are compiled into a single volume covering the
major areas of application.
The results of the survey are tabulated in Table 2 by industry.
Cement
In the cement industry, a total of nine companies were contacted.
Of these, five were visited and made data available to the survey.
Pulp and Paper
Data from the pulp and paper industry were supplied by Rust
Engineering Co. Much of the data were taken directly from bids on
installations for which Rust was the engineering contractor. In addition
to this source of data, test results were obtained from nine installations.
The test data were obtained from the mills as a result of personal visits
and telephone contacts.
Magnesia
Data were supplied for all three magnesia plants contacted.
of these were visited and the third furnished the data requested.
Two
SOUTHERN RESEARCH INSTITUTE
-27-
Table 2
A Survey of the Use of Electrostatic Precipitators
in Various Industries
Cement
Pulp and Paper
Magnesia
Phosphorus
Lime
Gypsum
Petroleum
Iron and Steel
Electric Power Generation
Nonferrous Metals
Municipal Incinerators
Total
Contacted
9
19
3
5
5
4
3
11
5
2
1
Number Furnishing Installations on
Data Which Data Rec'd
5
19
3
3
4
1
0
7
5
0
1
5
19
3
3
4
1
0
28
53
0
2
-------
-28-
Phosphorus
Five producers of phosphorus were contacted.
visited and furnished data.
Lirne^
Three plants were
Five companies with electrostatic precipitators installed on lime
kilns were contacted, one was visited and furnished data. Three companies
furnished the data requested on the survey form.
Several companias v/ere contacted, one furnished data and one had
no data in their files.
Sulfuric Acid
Several companies were contacted and one furnished data.
Petroleum
Three companies were contacted. All wanted to work through the
American Petroleum Institute Committee. The request submitted to the
committee was not honored.
Iron and Steel
Eleven companies were contacted for data covering the iron and steel
applications. Of these, seven companies supplied data on at least some
applications. Of those furnishing data, information was supplied on one
precipitator for an iron cupola, four sinter machines, five open hearth
furnaces, 15 blast furnaces, 11 by-product coke ovens, and three EOF
furnaces.
Electric Power Generation
Data were requested from five electric utility companies and were
received on 53 precipitators from all five utilities.
•OUTMUN MEHAMCH INOTTUTB
-29-
Nonferrous Metals
Data were requested from two companies, but no user data were
obtained.
Incinerators
One installation was visited and limited data obtained since the bid
data were incomplete and no test data were available.
Data Sources
In addition to the data received from the survey, published literature
provided data on some specific installations in sufficient detail for use.
The data supplied from the various sources consisted of design and
test data. Design data were generally more readily available since many
installations had not been tested. Where tests have been conducted, the
primary goal has been to determine stack emissions and electrical data
were often not recorded.
Table 3 is a summary of the precipitator applications by year
showing the total number of precipitators installed in each area. These
data are based on records supplied by Research-Cottrell and represent
"lost jobs" as well as those furnished by Research-Cottrell. The following
is a descriptive summary of the section of the manual covering applications
of precipitators in various industries.
Descriptive Summary of Application of Electrostatic
Precipitators in Major Application Areas
Part II of the manual on electrostatic precipitators is a review of
the application of electrostatic precipitators in each of eight major appli-
cation areas. In addition, an analysis is presented of the use of dust
control equipment in a number of manufacturing operations and the factors
influencing the use of electrostatic precipitators in new dust control appli-
cations.
Within each application area, a review of the process is given with
particular emphasis on the dust and gaseous emissions. This is followed
by a tabulation of input and design parameters for precipitators operating
on various types of dust control problems and an analysis of critical design
-------
-30-
parameters and test results. Cost data are also presented to show the
range of FOB and erected precipitator costs for various efficiency levels
and gas volumes.
The information presented has been derived from a review of the
literature pertaining to the particular application, a tabulation of informa-
tion by Research-Cottrell, Inc., of design, cost, and input parameter data,
and, finally, an analysis of design and test data obtained from a survey of
operating installations. The following is a summary of the information
presented in greater detail in the following chapters of the manual.
Electric power generation. Electrostatic precipitators are used in the
electric power generation industry principally for the control of emissions
from coal-fired steam generating plants. Although there is a trend toward
the use of nuclear power generation, the expected increase in the total power
consumption and the upgrading of existing control equipment is expected to
result in the expanded use of electrostatic precipitators in this application
for a number of years.
Fly ash is generated from both pulverized-coal-fired boilers and
cyclone boilers. The character and amount of'fly ash from these two types
of boilers vary with the type and chemistry of the coal being burned and
the particular operating conditions of the boiler. The ash content of the
coals being burned varies from 5-25%, and, together with the ash-fusion
temperature and boiler operation, determines the dust load that must be
handled. Typical dust loads range from 2-7 gr/cu ft of gas.
The particle size distribution of fly ash varies with the type of boiler
and type of coal. For a pulverized fuel boiler, the mass median diameter
of the fly ash is around 10-15 microns. For a cyclone furnace, the mass
median diameter varies over a wider range and is generally smaller
(5-15 microns). It is rather common practice to precede fly ash precipi-
tators with mechanical collectors which remove mainly the large size frac-
tion of the dust. The mass median diameter of the fly ash to a precipitator
preceded by a mechanical collector is around 3 microns.
The resistivity of the collected fly ash is perhaps the most important
property influencing collection by electrostatic precipitators. If the resis-
tivity is high (above about 2 x 1010 ohm-cm), the voltage and current to the
precipitator must be kept low to prevent excessive sparking and back corona.
Under these conditions, the charge acquired by the dust will be low, the
charging time will be long, and the collection field will be low. Conse-
quently, the performance of the precipitator will be impaired.
SOUTHERN RESEARCH INSTITUTE
-31-
At the other extreme, too low a resistivity will permit reentrainment
of the collected dust and result in low efficiency. There is an optimum
resistivity therefore for maximum precipitator performance.
Resistivity of fly ash is determined by the temperature of the flue gas
and the chemistry of the coal. At temperatures of around 450-500°F or
above, volume conductivity predominates and the resistivity is always below
the critical 2 x 1010. As the temperature decreases, resistivity increases.
This trend would continue under bone-dry conditions. However, at tempera-
tures in the range of 300°F, moisture in the flue gas is adsorbed on the fly
ash surface and alters its resistivity by a mechanism called surface con-
duction. The lower the temperature, the greater the rate of adsorption, so
that resistivity continues to decrease with decreasing temperatures.
In addition to temperature, the percentage of sulfur in the coal also
infl".ences fly ash resistivity. Studies of fly ash resistivity indicate that
the SO3 present in the flue gas acts to alter the rate of moisture adsorption
and serves as a secondary conditioning agent. On the average, the amount
of SO3 present in the flue gas is directly related to the sulfur content of the
coal. However, operating conditions in the boiler, and perhaps other
constituents of the fly ash, govern the quantity of sulfur appearing as SO3.
In addition to the temperature and the amount of SO3 present in the
gas, the chemistry of the fly ash appears to influence the adsorption rate.
A basic ash is reported to contribute to a higher rate of adsorption of SO3.
Thus, even though there is a good statistical correlation between sulfur
content of the coal and fly ash resistivity, the variations for a single con-
dition are too great to permit accurate prediction of resistivity based on
sulfur content alone.
Measurement of resistivity of fly ash should be made in-situ,
utilizing the techniques described in Chapter 12 of Part I of the manual.
Fly ash precipitators are generally designed on the basis of the
Deutsch-Anderson equation relating efficiency to gas volume and collecting
surface area. Experience with large numbers of precipitators has shown,
however, that the precipitation rate parameters can vary between around
3 cm/sec to 17 cm/sec. The major problem in the design, therefore, is to
narrow the range of uncertainty in selection of the precipitation rate
parameter.
-------
-32-
The value of the precipitation rate parameter for most fly ash pre-
cipitators is around 10 cm/sec (0.33 ft/sec). Variations from this value
can occur if the fly ash properties are either more favorable or more
adverse than normal.
Low values of precipitation rate parameter are normally associated
with excessive gas velocities, uneven gas flow distribution, high dust
resistivity, low dust resistivity.or improper rapping. Problems of gas
now distribution and gas velocity can be handled by proper design through
the use of models, etc. These problems are perhaps made more severe
in the case of fly ash precipitators by the large gas volumes that must be
handled.
High resistivity problems are usually associated with high ash, low
sulfur coals. Low resistivity problems can occur if the gas temperature
is low and high sulfur coal is being burned.
Problems with high fly ash resistivity can be resolved by several
approaches which include: (1) increasing precipitator size, (2) changing
the flue gas temperature, or (3) adding chemical conditioning agents.
The first of these alternatives is straightforward; however, about
three times the normal collection surface area may be required and the
costs of this alternative may be prohibitive.
Precipitators are normally located downstream of the air heater
and operate at temperatures in the range of 250-350°F. However, they
can be located ahead of the air heater where gas temperatures are in the
vicinity of 700°F. At these temperatures, the dust resistivity is deter-
mined principally by volume conductivity and is independent of the sulfur
content of the coal. Since gas volumes are about 1.5 times those at
300°F, additional precipitator capacity is required. However, this
alternative may be attractive for many applications.
Operation of precipitators following the air heater can utilize the
option of lowering the flue gas temperature to reduce dust resistivity.
Reduction of temperatures from 300°F to 260-270°F can reduce fly ash
resistivities by factors of 10 or more, and can minimize the high resis-
tivity problem. Because resistivities can change so rapidly within this
temperature range, control may become a problem.
•OUTHCKN KUCARCH INOTTUTB
-33-
Use of chemical additives to control dust resistivity is a third
alternative. Small additions of SOj (up to 10-15 ppm) to the flue gas have
altered dust resistivity by factors of 10 or more and resulted in substantial
improvements on precipitators which are limited by sparking due to high
dust resistivity. Other chemical additives, such as H2SO4, are being
evaluated as conditioning agents.
Problems with low dust resistivity have been encountered when
burning high sulfur coal and operating at low flue gas temperatures. The
difficulty has been identified as excessive scouring or reentrainment due
to the reduction in force holding the dust to the collection plate. The prob-
lem is associated with high gas velocity and is apparent when relationships
between gas velocity and precipitation rate parameter are plotted for vari-
ous resistivities. This problem can be resolved by increasing the flue gas
temperature, injection of ammonia, reduction in gas velocity, or a com-
bination of these effects.
In addition to determining the collection surface area, design of a
fly ash precipitator also includes determining the power supply require-
ments. It has been shown empirically that efficiency is related to the
corona power delivered to the precipitator. Curves showing this relation-
ship for fly ash precipitators show a good correlation and are useful in both
design and troubleshooting of fly ash precipitators. _
The number of independently energized bus sections is another design
variable. Since fly ash precipitators are generally designed to handle
large gas volumes, the degree of sectionalization is of greater importance
than in some other applications. If very high collection efficiencies are
desired, a higher degree of sectionalization should be used. The advantages
are: (1) higher operating voltages in the spark-limited mode, (2) lower
internal impedance with better spark quenching, and (3) less percentage of
the precipitator would be disabled by the outage of a single section.
Since a number of factors influence precipitator performance, and
since these vary between installations, a technique of regression analysis
has been developed by several investigators exploring precipitator operation.
A method relating the expected precipitation rate parameter with the sig-
nificant parameters has been developed to serve as a guide to analysis of
precipitator specifications. The technique gives a correlation coefficient of
about 0.85. A second regression analysis performed on data from a group
of installations gave a correlation coefficient of around 0. 9 when using more
fundamental precipitator and dust parameters.
-------
-34-
Pulp and paper. Electrostatic precipitators are used in the pulp and
paper industry to remove particulates carried by the effluent gases from
black liquor recovery boilers.
The kraft or sulfate process constitutes approximately 55% of the total
production in this country. Of these kraft mills, about 65% have electrostatic
precipitators installed on the recovery boiler.
The economics of the kraft process requires the recovery and reuse
of the spent chemicals used in the cooking process and recovery of the heat
content of the concentrated spent liquor. This is accomplished by combustion
in the recovery boiler which releases large volumes of flue gases. During
combustion, a significant fraction of the recoverable chemicals is entrained
as particulates with the flue gas. Recovery of these particulates constitutes
a significant economic factor as well as a means for controlling air pollution.
Particulate emissions from the recovery boiler are extremely
fine hygroscopic particles, composed principally of sodium sulfate and
sodTuiircaTbonate with small quantities of sodium chloride, sulfide, and
sodium sulfite. Because of its hygroscopic nature, sampling to
determine particle size distribution is difficult. A technique for pre-
cipitating a sample onto a copper mesh electron microscope target and
subsequently counting the particles in various size ranges has been
developed. The collected sample must be encapsulated in a protective
atmosphere to prevent moisture pickup during transfer from the duct to
the microscope. The mass median particle size for recovery furnace dust
is around 1. 9 microns. By count, the median size is about 0.4 micron.
Electron photomicrographs of the dust samples from recovery boilers
show a change in the character of the dust depending upon operating
temperature. At temperatures in the range of 350-360°F, the dust is pri-
marily spherical particles. At a temperature of about 280-290°F, the dust
contains large numbers of needle-like particles. These differences in par-
ticulate structure are thought to be the cause of the variations in the trans-
porting and rapping properties of the collected dust.
Resistivity data for recovery furnace dust have not been reported
extensively, primarily because high resistivity has not been identified as a
problem.
SOUTHERN RESEARCH INSTITUTE
-35-
Electrostatic precipitators for recovery boilers are of a variety of
types including vertical and horizontal-flow types, wet and dry bottom types,
and steel, tile or concrete shell types.
Dry bottom precipitators have been used extensively in recovery
boiler precipitators from 1930 to 1948. In the dry bottom precipitators,
dust from the collection plates was collected in pyramidal-type hoppers.
The wet bottom precipitator was introduced to minimize problems associated
with the collection, transporting, and redissolving of the collected salt cake.
In the wet bottom precipitator, a solution of 50% liquor is pumped into the
collection hopper. The collected dust is rapped from the plates and falls
into the liquor solution, where it is dissolved and subsequently removed.
Current emphasis on odor control from kraft mills has resulted in a
trend toward elimination of the contact between the flue gas and black liquor.
Consequently, wet bottom precipitators may not be used on new kraft process
mills. Also, elimination of the direct contact evaporators may influence
the properties of both the gas and particulates to be handled by electrostatic
precipitators.
Precipitators for recovery furnace boilers are designed for dust
concentrations in the range of 1-9 gr/acf, with the bulk of the installations in
the range of 2-5 gr/acf. Inlet temperatures range from 225-375°F, with
the majority of installations operating at temperatures of 275-325T.
Gas velocities range from 2-8 ft/sec, with the largest number of precipi-
tators operating in the 3-6 ft/sec range.
Connected input power for the majority of installations is in the
range of 100-200 watts/1000 cfm. Design field strengths range from
around 7-16 kV/in. , computed as the ratio of average design voltage to
plate spacing.
Design precipitation rate parameters for recovery boiler precipi-
tators range from around 0.2 to 0.35 ft/sec.
Iron and steel industry. The application of electrostatic precipitators
in the iron and steel industry has been in the cleaning of gaseous effluents
from steelmaking furnaces, blast furnaces, foundry cupolas, sinter machines,
and by-product coke ovens.
-------
-36-
Steelmaking processes have undergone a number of changes in pro-
duction methods. One of the earliest steelmaking furnaces was the
Bessemer converter. About 1868, the open hearth process was invented
and was the primary method of steelmaking for a number of years. The
basic oxygen furnace, introduced around 1950, has increased in impor-
tance, and in 1969 about 40% of the total steel made was produced in EOF
furnaces.
By-product coke
Production of metallurgical coke is made by the beehive process and
in by-product coke ovens. The latter accounts for about 98% of all the
metallurgical coke produced in this country.
Precipitators are used in the by-product coking process to remove
the tars and particulate matter from the gases prior to recovery of the
hydrocarbons.
The precipitator most often used for detarring is a cylindrical type
consisting of collection electrodes made from 6-8 in. pipe, 6-9 ft long
suspended from a header plate in a round shell. Discharge electrodes are
suspended axially through the cylinders. The precipitators are main-
tained at a temperature high enough so that the collected tar drains from
the plates and no rapping is required.
Three problems associated with the use of precipitators for cleaning
coke oven gases are as follows: (1) Collection of the fluid tar makes elec-
trical insulation difficult and requires that insulators be kept out of the gas
stream. The insulators are also heated and cleaned on a regular scheduled
basis. (2) The gases being cleaned are combustible when mixed with air,
so that no inleakage of air must occur. Design often includes operation at
positive pressure to insure no inleakage. (3) The gases can be corrosive.
Protection against corrosion is provided by the film of tar on the interior
surface of the collection electrodes. Spraying of tar on the exterior collec-
tion electrode surface can minimize corrosion at that point.
Coke oven gas precipitators are designed to handle gas volumes in the
range of 5-15,000 acfm with efficiencies in the range of 95-98%. Precipi-
tation rate parameters are about 0.2 - 0. 3 ft/sec. Inlet loadings are in
the vicinity of 0. 5 gr/acf. Gas velocities average around 8 ft/sec, which is
higher than for some applications since reentrainment is not a problem.
•OUTHCRN MEBEAHCH INWTUTI
-37-
Power densities are 8-10 watts/ft2.
Sinter plants
Sintering is a process for agglomerating iron-bearing fines to
prevent their loss during reduction in the blast furnace.
The raw material is composed of iron-bearing fines, coke or coal
dust, and a fluxing material such as limestone or dolomite. As the material
moves through the sinter stand, it is ignited by surface burners and com-
bustion is sustained by air drawn through the mixture by fans. The products
of combustion are collected in a group of compartments called windboxes
located beneath the grates of the sinter machine.
The particulate material emitted from the sintering process is a
result of the mechanical handling of the raw material and combustion of the
coal or coke. Under normal conditions, 5-100 Ib of dust is produced per
ton of sinter produced. Gas volumes vary from 100,000-450,000 cfm, with
dust loadings of 0. 5 - 6. 5 gr/scf. About 80-90% of the particulate is greater
than 20 microns. The dust contains fewer fines than metallurgical furnace
operations. However, when preceded by a mechanical collector, the dust
can be relatively small. Typically, the mass median diameter of the par-
ticles following a mechanical collector would be around 6 microns, whereas
without a mechanical collector, the mass median diameter would be around
50 microns.
The ele-U-ical resistivity of sinter machine dust can vary radically
depending upon the type and amount of fluxing material in the burden
makeup. The amounts of fluxing material can vary from around 10-35%,
and this variation can change the resistivity of the dust by several orders
of magnitude. The very high resistivities associated with the higher per-
centages of fluxing materials can result in excessive sparking, lower
operating voltage and power, and generally poorer precipitator performance.
The effect of the addition of the flux has not been fully explored;
however, the condition is similar to that encountered in fly ash precipitators
where limestone is added to the boiler for removal of sulfur oxides. It is
postulated that the limestone reacts preferentially with the SO, present in
the gases. Since the SO3 is a secondary conditioning agent, surface con-
ductivity would be decreased.
-------
-38-
Studies of the mechanisms of conduction and possible corrective
measures have not been explored for sinter machine precipitators to the
same extent that they have for fly ash precipitators.
Design of sinter machine precipitators has conventionally been based
on conditions in which resistivity has not been a problem. Because of the
relatively large particle size of the dust, the range of design precipitation
rate parameters has been from around 0. 25 to 0.4 ft/sec. However, test
precipitation rate parameters as low as 0. 08 ft/sec have been obtained
when high dust resistivity has been encountered.
Other design parameters for sinter machine precipitators are:
gas velocity - 4-5 ft/sec, temperature - 250-300°F, electric field -
8-10 kV/in., average inlet dust loading - 1 gr/acf, and average precipi-
tator power - 70 watts/1000 cfm.
Blast furnace
The effluent gases f^m blast furnaces have heating values of about
100 Btu/scf, which makes them valuable as a fuel for heating. The gas,
however, contains particulate material carried over from the blast furnace
and must be cleaned to prevent clogging of gas burucr-s-and gas mains.
Electrostatic precipitators have typically been used for this service.
Blast furnace gas is usually passed through a dust catcher where the
heavier particles are separated by inertial forces. From the dust catcher,
the gas is passed to a wet scrubber and an electrostatic precipitator.
Precipitators for cleaning blast furnace gas are conventionally of
the vertical-flow type employing cylindrical collection electrodes, although
horizontal-flow, plate-type precipitators are also used.
Since the gas delivered to a blast furnace precipitator is cooled to
saturation by the preceding wet scrubber, a wet type precipitator is used.
Particulate removal is through a slurry hopper. The slurry is normally
piped from the hopper to a dust disposal system.
The use of precipitators for cleaning blast furnace gases has
generally followed the blast furnace production capacity. However, in
recent years, the sales of blast furnace gas precipitators have been low
-39-
compared with the peak periods during the 1940-1960 period.
Inlet gas temperature for blast furnace gas precipitators is low as
compared with other applications. Typical temperatures are in the range
of 70-100°F. Inlet dust loadings are also low, of the order of 0. 05 - 0. 3
grains/scfd for most installations.
Gas velocities are relatively high, around 6-15 ft/sec, since reen-
trainment is not a serious problem with wet precipitators. The average
field strength (average voltage-to-electrode-spacing ratio) varies from
around 9-15 kV/in. Input power ranges from 50-300 watts/1000 cfm.
Design precipitation rate parameters for blast furnace gas precipi-
tators are around 0. 2 - 0.4 ft/sec. Test data vary over a considerably
wider range.
Open hearth
Open hearth steelmaking furnaces consist of a large refractory lined
dish into which metal from the blast furnace, steel scrap, iron ore, and
limestone is charged. Heat is provided by furnace burners which burn
oil, natural gas or tar with combustion air that is heated in regenerative
heat exchangers called checkers. Open hearth furnaces are often
equipped with oxygen lances to facilitate oxidation of the carbon and other
elements to be removed.
The particulate material carried out by the exhaust gases comes from
a variety of sources, including dirt and other fines on the charge material,
oil and grease, and volatile metal oxides from the scrap charge. During
oxygen lancing, large amounts of iron oxide are evolved together with
lesser amounts of nonmetallic oxides from the slag. The quantity of
exhaust gases, particulate loading, and particulate composition varies
widely during the period of the heat. Consequently, the precipitator input
conditions vary depending upon operating conditions of the furnace.
The sizes of the particulate matter from open hearth furnaces range
from less than 0.03 micron to several microns. Composite samples
representing dust evolved during the entire heat indicate that 50% of the
particulate is less than 5 microns. However, during the lime boil, the
dust is considerably finer with as much as 77% less than 5 microns and
20% less than 1 micron. Resistivity of open hearth furnace dust varies
with moisture, temperature, and composition.
-------
-40-
Precipitators for cleaning open hearth gases are generally the
horizontal-flow, duct-type with a steel shell. In large steelmaking shops,
a multiplicity of furnaces are serviced by a common air pollution control
system, which may consist of mechanical collectors, washers, and electro-
static precipitators singly or in combination. Open hearth precipitators
are designed with precipitator rate parameters in the range of 0.15 - 0.3
ft/sec.
The total installed capacity of open hearth precipitators has steadily
increased since 1950. In 1969, the total capacity reached approximately
12 million acfm.
Basic oxygen furnaces
The basic oxygen process of steelmaking utilizes high pressure oxygen
introduced into the mouth of a basic, refractory lined converter to oxidize
carbon and other elements from the charge. There is no external source
of heat in the converter, and molten pig iron from the blast furnace con-
stitutes the major constituent of the furnace charge. After charging, oxygen
is blown into the converter at high pressure through a water-cooled lance.
Gas and-ii'ast emissions from BOF converters vary gjfeaily with the
stage of heat. During charging, the gaseous and particulate emission
levels are low. During the oxygen blow, which may last for 20 minutes,
large amounts of fume and gas are evolved. Gas volumes range from
200, 000 - 1, 200, 000 cfm at temperatures of 3000 - 3200°F and may carry
300 Ib/min or more of dust.
Fume from the basic oxygen furnace is composed primarily of iron
oxide in amounts of 20-60 Ib/ton of steel. Fume concentrations of up to
15 gr/scfm are produced under peak conditions. The fume is finely
divided with most of the particles ranging from 0.1 - 1 micron in size,
although on a weight basis, a considerable portion of the burden exists
as relatively large particles.
Resistivity of BOF dust is a function of moisture content and tempera-
ture. Moisture comes from the evaporation of water in the cooling tower
located between the precipitator and the furnace.
SOUTHERN RESEARCH INSTITUTE
-41-
The cooling tower is an important part of the gas cleaning facility.
It should be large enough to insure complete evaporation of the water to
prevent carryover and subsequent clogging of the precipitator. Since the
gas temperature varies, the rate of heat transfer varies, so that water
quantities to the cooling tower must be controlled.
Precipitators for BOF gas cleaning are generally the horizontal-flow,
duct-type with steel shells. Design precipitation rate parameters vary from
around 0. 15 - 0. 25 ft/sec. The generally smaller particle size distribution
accounts for the lower values of precipitation rate parameter.
Gas temperatures for BOF precipitators range from around 250-350°F
depending upon design philosophy. Design efficiencies are generally 99+%
for newer installations.
One problem associated with BOF precipitators is that the collection
efficiency at the start of the oxygen blow is reduced. Cooling towers are
generally set to operate the water sprays when the gas temperature reaches
around 500°F. In the interim, dust resistivity is high and emissions from
the stack can be relatively high during this period. The condition is called
a lance puff and can be controlled'by introduction of steam to condition the
dust during the period before the cooling tower sprays come on.
Electric arc furnace
The electric arc furnace as used in the steelmaking process consists
of a refractory lined structure with a dish-shaped bottom and a domed roof.
Steel scrap, and perhaps hot metal from a blast furnace, are charged into
the furnace and heated by an electric arc developed between graphite elec-
trodes which are lowered into the furnace charge.
After meltdown, oxygen is introduced to remove the carbon and
other elements. Oxygen sources can be from various sources including
oxygen gas, oxides of alloying elements, iron ore, decomposition of
limestone, etc.
The flue gases from electric furnaces contain large concentrations
of carbon monoxide which must be converted to COj prior to entering the
precipitator to minimize the explosion hazard. This is usually accom-
plished by admitting air to the hot gas stream and allowing combustion to
take place in the high temperature region. After combustion, the gases
-------
-42-
are passed through a cooling tower to reduce the temperature before
entering the gas cleaning equipment.
The participate emissions from electric arc furnaces average
around 5-30 Ib/ton of steel produced. The size and composition of the
dust emitted from electric furnaces vary with the type and cleanliness of
scrap and the metal refining procedure. Nonferrous impurities in the
scrap can lead to significant quantities of oxide fumes. Also, presence of
oil and grease in the scrap can produce large amounts of carbonaceous
particulate matter during early stages of meltdown.
Dust from electric arc furnaces tends to be extremely fine. Data
on particle size distribution vary. Some sources indicated that as much
as 90-95% of the fume is below 0.5 micron. Other data would indicate a
somewhat coarser dust.
Electrostatic precipitators have been used to clean electric furnace
gases in the United States and Europe. However, the small size of the
particles necessitates the use of a large precipitator to achieve high
collection efficiencies, and other methods of cleaning electric furnace gases
have been used to advantage in some instances.
Design precipitation rate parameters for electric arc furnace pre-
cipitators range from around 0. 12 - 0. 16 ft/sec.
Scarfing machines
Scarfing is the operation of removing the skin of the slab in the pro-
duction of steel. In the scarfing machine, the slab passes under cutting
torches which burn through the slab. The process generates iron oxide
particles and fume which constitute an emissions problem. The application
of electrostatic precipitators to this service is limited; the total capacity
installed being around 500, 000 cfm.
Cupolas
The iron cupola is used to provide a source of molten metal for cast
iron foundries. The cupola is a refractory lined, cylindrical furnace which,
when charged with pig iron, scrap, coke, and flue, provides a self-sustain-
ing exothermic reaction to melt the charge and maintain it at the proper
temperature.
SOUTHERN RESEARCH INSTITUTE
-43-
Particulate emissions from iron cupolas consist primarily of oxides
of iron, silicon, calcium, aluminum, magnesium, and manganese. The
dust load in the effluent gases varies widely with ranges from 0. 9 - 6. 5
and 1.3 - 11 grains/scfm reported for cold and hot blast cupolas, respec-
tively.
Particle size distribution of cupola dust varies between wide limits
depending upon melt rate, type and cleanliness of scrap, and furnace
operating conditions. Values range between 10% of the dust less than 10
microns to 40% less than 1 micron.
The gases from the cupola are high in carbon monoxide and must
be burned prior to entering an electrostatic precipitator because of the
potential explosion problem. Burning normally is accomplished by intro-
duction of combustion air at the exit of the cupola. The combustion process
can also burn carbonaceous particulates, and thus can have an influence on
the particle size distribution as well as the composition.
Use of electrostatic precipitators on iron cupolas in this country has
been extremely limited. Poor experience with operation of precipitators on
a few installations has influenced the choice of dust collection equipment to
a considerable extent. Also, the extreme variability of cupola operation
resulting in extremely wide ranges of particle size and composition has
tended to limit use of precipitators for .this application.
Rock products. Electrostatic precipitators are used in the rock pro-
ducts industry in the collection of dust from cement kilns, gypsum kettles,
and from auxiliary grinding, transporting, and handling operations.
Portland cement
Portland cement is produced by a wet or dry process which defines
the conditions under which the ingredients are ground and fed into the
calcining kiln. There are variations of these two production methods which
are often termed semi-wet or semi-dry processes.
In the basic cement production process, the raw materials, con-
sisting of lime, alumina, iron oxide, and a fluxing material are ground
in a mill and introduced into a kiln. The material fed into the kiln is
dried in the initial section, calcined as it passes down the kiln, and
finally fused into a clinker in the final section of the kiln. The clinker
is then removed, cooled, and ground to produce the final product.
The effluent gases from the kiln range from 40, 000-100, 000 cfm
depending upon the type of kiln, method of gas cooling, and the method of
-------
-44-
preheating the raw materials. The gases are composed of nitrogen,
water vapor, carbon dioxide, and small concentrations of oxygen and
sulfur dioxide.
Particulate carried by the kiln gases originates from the abrasion
of the charge as it tumbles through the kiln, the release of particulate
due to the gas release associated with calcination, the ash from the fuel
if the kiln is coal fired, and the fume resulting from the vaporization and
condensation of alkali. The particles resulting from the mechanical
abrasion are generally large in comparison with those produced by alkali
condensation.
Resistivity of cement-kiln dust is dependent upon moisture content and
temperature. In the range of 500°F and above, resistivities are generally
in the range suitable for collection by electrostatic precipitators regardless
of moisture. However, from 300-400°F, resistivities are extremely
dependent on moisture content. In the wet process kiln, moisture is
provided by the evaporation of *be water from the slurry feed to the kiln.
In the dry process, moisture must be provided by evaporation of water
in the cooling tower.
Electrostatic precipitators used on cement kilns are of the horizontal-
flow, duct-type with insulated steel shells. The application of precipitators
to control of cement kiln dust has steadily increased over the past 50 years;
the present installed capacity being around 40 million acfm. The trend in
the application of precipitators is toward a higher collection efficiency;
the average for the past five years being designed for around 99.7%.
Inlet dust loadings for cement kiln precipitators vary from a mini-
mum of around 3 gr/scfd to around 50 gr/scfd. Gas velocities range from
around 3-8 ft/sec. Moisture contents range from 4-14% for dry process
kilns to 13-40% for wet process kilns. Inlet gas temperatures range from
350-650°F on wet process kilns to 500-700°F on dry process kilns. Precipi-
tation rate parameters for cement kilns range from around 0.25-0.45 ft/sec.
Precipitator applications to cement kilns have been particularly
favorable from the standpoint of recovery of cement as well as air
pollution control. The effluent dust from the cement kiln has about the
same composition as the kiln charge. Consequently, its recovery is of
direct economic importance. However, alkali present in the dust cannot
be recycled. A fundamental property of the electrostatic precipitator
is that larger particles tend to be separated first, and the smaller,
alkali-containing fraction of the dust tends to be removed in the last
stages. This fractionating effect has been used to separate the dust that
is relatively free from alkali and to recycle it through the kiln.
SOUTHERN RESEARCH INSTITUTE
-45-
There are several problems associated with the use of electrostatic
precipitators in cement kiln applications. (1) In the dry process kilns,
the moisture content is low, and consequently resistivity would tend to be
high in the temperature range of 300-400°F. There is a trend toward the
use of fabric filters on dry process kilns in this country. (2) Upsets in
kiln operation can create conditions under which combustible gas can be
introduced in the precipitator. Instances of fire in the precipitator have
been reported which limit acceptance of precipitators, although this can
also be a problem in other types of dust control equipment. (3) Operation
at certain temperatures can result in a deposit on the electrodes that
impairs precipitator performance.
Gypsum
Gypsum is a hydrated calcium sulfate which, when heated, looses
water to form plaster. Removal of the water occurs in a process called
cooking. Electrostatic precipitators are used to control emissions from
gypsum kettles.
Particulate emissions from the cooking process result from the cal-
cination process and the mechanical agitation of the charge. The size of
the dust particles is relatively large compared with processes where the
vaporization and condensation of material occurs.
The process of cooking increases moisture of the exit gases. The
wet gas is mixed with ventilating air and flue gas to give a mixture of gas
that ranges from 300-400°F with a moisture content of 30-35%. As a
result of the moisture, resistivity of the dust is not usually a problem.
However, at times, water sprays are used in the flue ahead of the pre-
cipitator.
The use of electrostatic precipitators in the production of gypsum
has steadily increased from around 1935 to the present. The total
installed capacity is approximately 1.8 million acfm. The gas flow
ranges from a minimum of around 3000 acfm to 14 acfm for kettles to
a maximum of around 80, 000 acfm for rotary calciners. Efficiencies
are generally in the range of 95-99+%. Gas velocities range from 1. 5-
8 ft/sec. Inlet dust loadings are from around 4-60 gr/scfd. Inlet gas
temperatures are around 200-350°F for calciners and 125-250°F for
rock dryers. Design precipitation rate parameters for precipitators
used in the gypsum industry are around 0.4 - 0. 5 ft/sec.
Chemical industry. Precipitators are used in the chemical industry
in the production of sulfuric and phosphoric acids.
-------
-46-
Sulfuric acid
Sulfurtc acid is made by the oxidation of sulfur dioxide to sulfur
trioxide and subsequent absorption in a recirculating sulfuric acid solution
to make acid of the desired concentration. About 95% of the sulfuric acid
produced in this country is by the contact process in which SO^ is cataly-
tically oxidized by atmospheric oxygen in the presence of a vanadium
pentoxide catalyst. Sources of SC^ for the process are burning of elemental
sulfur, roaster gases from metallurgical operations, and burning of hydrogen
sulfide and spent acid from petroleum refineries.
Electrostatic precipitators can be used in the manufacture of
sulfuric acid in two ways. If the source of SO2 is smelter gas from non-
ferrous metallurgical operations, the gases contain approximately 3-10%
SOj and are contaminated with dust, which must be removed prior to
being introduced to the converter to prevent fouling of the catalyst. Clean-
ing can be accomplished by electrostatic precipitators or wet scrubbers.
In the converter, SO2 present in the gases is converted to SO3.
Gases from the converter exit at around 450°F and pass to the absorber
where the SO3 combines with water tc pnoduce 98-99% sulfuric acid.
Casts leaving the absorber contain unabsorbed SO3 and H2SO4
mist. The mist is of very small size and will pass through the absorber
without being collected and would be emitted from the process unless
suitable collection equipment is provided.
The type of collection equipment needed to remove the acid mist
depends upon the size of the particles. Wire mesh eliminators have
collection efficiencies of over 90% when most of the particles are
greater than 3 microns, which is the case when 98% acid is being pro-
duced. When oleum is also produced, 85-95% of the particles leaving
the oleum tower is less than 2 microns and wire mesh pads may not
effectively remove these fine particles. Electrostatic precipitators have
been extensively used for acid mist removal and are effective for the smaller
sizes.
Precipitators for acid mist collection have historically been of the
vertical up-flow type with cylindrical shells constructed of sheet lead
supported by steel banding. Discharge electrodes are of lead with a star
cross section. Acid mist precipitators with all-steel construction have
been installed in recent years without apparent difficulty.
Between 1945-1969, about 120 sulfuric acid mist precipitators
were installed with a total capacity of 2, 230, 000 cfm. The average
SOUTHERN RESEARCH INSTITUTE
-47-
precipitator size has increased from 10,000 acfm in 1945 to around
30,000 acfm in the 1964-1968 period. Design efficiencies are in the range
of 97. 5 - 98.5%.
Gas velocities for acid mist precipitators range from 2-8 ft/sec,
with the majority falling in the 3-5 ft/sec range. Inlet loadings range
from about 0. 2 - 2. 5 gr/scf. Operating temperatures range from 80-
180°F. Input power varies from around 150-700 watts/1000 cfm for the
majority of installations. Design field strengths are from 8. 5 - 13 kV/in.
Design precipitation rate parameters are around 0. 2 - 0. 3 ft/sec.
Phosphorus
Precipitators are used in the production of phosphorus and phosphoric
acid. In the production of phosphorus, phosphate rock, silica and coke are
charged in an electric furnace and heated to around 2300-2700°F to liberate
PjOj, which is reduced to elemental phosphorus by the carbon. The effluent
from the furnace is cleaned by an electrostatic precipitator operating above
the condensation temperature of the phosphorus (525Hg20*F) to prevent dust
contamination of the phosphorus as it is condensed. Precipitators typically
remove 90-99% of the dust.
Precipitators for use in cleaning phosphorus furnace gas are
typically vertical gas flow, single-stage types with cylindrical collec-
tion electrodes. The collection electrodes are heated to avoid phos-
phorus condensation and special rappers are often used to avoid damage
during rapping.
Special types of hoppers are used for dust removal since the
collected dust contains some absorbed elemental phosphorus which
ignites on exposure to air.
During 1938-1969, about 20 precipitators have been installed for
hot phosphorus applications. The total gas volume handled is around
165, 000 acfm. Gas velocities for phosphorus precipitators range from
1-6 ft/sec, with the average around 2-3 ft/sec. Average gas tempera-
tures range from 500-600°F, with some as high as 800°F. Dust loadings
range from 4-15 gr/scf, with the majority in the range of 12-14 gr/scf.
-------
-50-
of the gases to the precipitator to around 600-700°P.
The inlet dust concentration varies with the type of mechanical
collector. A typical range is 0. 2 - 1.0 gr/scfd. Mass median particle
size is around 10-12 microns. Resistivity varies with temperature and
moisture content. Conditioning of the catalyst dust by additions of
ammonia to the effluent has been reported to reduce resistivity and
improve performance in instances where high resistivity dust lias been a
problem.
Detarring
Precipitators used for detarring of gases are commonly of the
single-stage, vertical wire and pipe type, with collection electrodes
suspended from a top heater. The collected oils and tars are usually
free flowing and no rapping is required for removal.
During the period 1940-1963, approximately 55 precipitators were
installed for detarring of carburetted water gas, 3 for detarring oil gas,
3 for reformed gas, '2 tor shale oil, and 1 for acetylene. The total gas
volume for all of these applications is around 400, 000 acfm. Gas temp-
eratures for detarring precipitators range from around 70-120°F and
collection efficiencies are around 95%.
The total market for detarring precipitators is small since very
little manufactured gas is sold in this country and the requirements for
shale oil processing and acetylene manufacturing are relatively small.
Nonferrous metals. The commercial use of electrostatic precipi-
tators has been standard practice by copper, lead, and zinc smelters in
cleaning the off-gases from the extraction process. Precipitators are
also used in cleaning gases from electrolytic cells in the reduction of
bauxite to produce aluminum.
Extraction of nonferrous metals from their ores is carried out
in a number of types of processing equipment, many of which are
common to all nonferrous operations.
Sinter machines are used to convert metallic ores, fines, and
plant process dust into larger material that can be handled in the reduc-
tion process. Pellets to be sintered are spread on grates that move the
•OUTHUN MUKAMCH INmTUTC
-48-
Predictions for phosphorus furnace applications are for continued
growth, although the total volume is small in comparison with other areas.
Precipitators for use in the production of phosphoric acid are for
mist elimination. During 1930-1960, the cumulative installed capacity
was around 150, 000 acfm.
The corrosive nature of the gas creates some problems in the
choice of materials. Stainless steel pipes of 5-15 inch diameter are used
as collection electrodes. The collected mist drains from the plates so
that no rapping is required.
Inlet gas velocities for phosphoric acid mist precipitators range
from 2-8 ft/sec. Inlet gas temperatures vary from 150-300°F, and
inlet concentrations range from around 5-35 gr/scfd.
Carbon black
Precipitators used in the collection of carbon black are for the
purpose of agglomeration of the particles so they can be collected in a
mechanical collector. About 18% of the carbon black is collected in the
precipitators and 72% in the mechanical collector. Bag filters are often
used following the mechanical collectors to reduce the emissions to the
atmosphere.
Carbon black particles are extremely fine, ranging from 0. 02 -
0.4 micron. The conductivity is also very high, so that the particles
would tend to be discharged upon contact with the collection electrode.
However, the carbon black particles cling tenaciously to both the dis-
charge and collection electrodes and cleaning is a problem.
It is estimated that about one hundred sets of electrostatic preci-
pitator - mechanical collector units have been built since 1926. No
information was found on installations made since 1958.
Municipal incinerators. Use of electrostatic precipitators for
control of municipal incinerator emissions is a relatively new application
in this country, although the practice is rather widespread in Europe.
There are two principal types of furnaces in general use for in-
cineration of municipal wastes—these are water-cooled and refractory
-------
-49-
lined furnaces. The type of refuse handled is highly variable between
countries and between different sections of the same country. This leads
to large variations in the properties and composition of the particulate.
Particle size of the fly ash from municipal incinerators ranges
from a mass median diameter of 15 to 30n. Resistivity of the ash varies
with temperature and moisture content and with particle size.
Gases from municipal incinerators are at temperatures in the
neighborhood of 1200-1800°F and must be cooled before entering the pre-
cipitator. Cooling can be provided by evaporation of water in a water
cooling tower or by heat exchangers in a system utilizing heat recovery.
Design precipitation rate parameters for a group of precipitators
installed on European incinerators range from around 0. 2 - 0.4 ft/sec.
Gas velocities range from 2-4 ft/sec. Power densities range from 50-200
watts/1000 cfm.
Precipitation rate parameters for European precipitators vary with
precipitator inlet gas temperature. Data from the U. S. installation, for
which information was available, agree well with those data from the
European installations.
No test data are available on the municipal incinerators in this
country since they have been installed so recently.
Petroleum industry. The principal uses of electrostatic precipitators
in the petroleum industry are in the collection of particulate emissions
from fluidized bed catalytic cracking units and the removal of tar from
various gas streams, such as fuel gases, acetylene, and shale oil dis-
tillation gases. The first of these areas, recovery of catalyst dust, origi-
nated with the production of high octane gasoline during World War II.
Electrostatic precipitators were used to recover catalyst from the dis-
charge stream of the catalyst regenerators as a part of the process.
Improvements in mechanical collectors inside the regenerators have
eliminated the process requirements. However, precipitators are
presently used for control of dust emissions from the process.
Gases from the catalyst regenerator are exhausted from the top
through a series of mechanical collectors which return all but the fine
particles to the process. The regenerator may be followed by waste
heat boilers that recover some of the energy and reduce the temperature
SOUTHERN RESEARCH INSTITUTE
-51-
material through the sinter machine. Gas-fired burners ignite the
material and air is supplied by fans to maintain combustion. Exit gases
are collected in windboxes and passed to the electrostatic precipitator
for cleaning.
Ore roasting can be accomplished in a variety of types of fur-
naces including multiple hearth, flash roasters, and fluid bed roasters.
Smelting and refining operations are carried out in reverberatory fur-
naces and blast furnaces. Converters are used to convert matte to
metallic copper. Cupolas are used in the nonferrous metals industry
to melt and reduce copper brasses, bronzes, and lead.
Aluminum is produced by the electrolytic reduction of alumina .
(A12O3) dissolved in a molten cryolite bath. During reduction, electro-
lytic, thermal, and chemical action in the cell results in the evolution
of carbon and alumina dust, other particulates, and gaseous fluorides.
Hoods over and around the cells collect the effluent which is sent to
mechanical collectors followed by an electrostatic precipitator and then
to a scrubber for removal of the remaining gaseous fluorides.
In copper production, the copper-bear ing ores are roasted to-
eliminate some of the sulfur from the concentrate and to volatilize
zinc, arsenic, and antimony present in the ores. The roasted ores are
then smelted to produce a molten sulfide of iron and copper. The copper
matte from the smelter is then converted to copper by blowing it with air
in a converter. The blister copper from the converter is then refined by
either fire or electrolytic methods.
Lead production begins with either oxide or sulfide ores. Oxide
ores can be directly reduced in a blast furnace, whereas sulfide ores
must be first converted to oxides. This is accomplished by roasting
or sintering in an oxidizing atmosphere. After converting to the oxide,
the lead ore is reduced in a blast furnace by reaction with carbon
supplied by coke charged in the furnace. Refining of the lead is
accomplished by electrolytic refining or by a kettle or reverberatory
furnace.
-------
Zinc
-52-
is generally extracted from ores containing both zinc and lead
suifides, although some zinc and copper-zinc ores are processed. The
zinc ores are concentrated by flotation and processed in a roaster to
convert the zinc suifides to zinc oxide. Metallic zinc is produced from
the roasted ore by retorting, electrolysis or fractional distillation.
Most of the electrostatic precipitators used in the nonferrous
metals production have been designed by the large western smelters,
and hence information on the applications in the nonferrous metals industry
is limited. Information that is available indicates that design precipitation
rate parameters vary from 0. 05 - 0. 07 ft/sec for precipitators operating
on converter gas, 0.12 - 0.14 ft/sec for precipitators used on copper
roasters and reverberatory furnaces, and around 0.25 ft/sec for sinter
machine precipitators.
High temperature, high pressure gas cleaning. A unique application
of precipitators would be in the cleaning of high temperature, high pressure
gases. One of the early applications of precipitators was in the removal
of ash from the products of high temperature gasification of coal. This
work was a pilot-scale research operation conductedMn conjunction with
the development of the coal-fired, gas-turbine locomotive. Temperatures
of 1500°F and pressures up to 600 Ib/in. were used in this study.
Current interest in high temperature, high pressure gas cleaning
is for the use of the gases produced from waste incineration to produce
electric power from gas-turbine-driven generators.
Research has been conducted on corona generation at temperatures
up to 1700°F and pressures of 100 psig. Precipitation rate parameters of
0.23 - 0.26 ft/sec were obtained on pilot-scale units and extrapolations
made to full-scale precipitators.
New application areas. New applications for electrostatic precipi-
tators can take the form of replacement of other types of dust control
systems, or the control of participate emissions from sources where no
particulate control devices are now used.
•OUTMBItN NnCAMCH INSTITUTE
-53-
The major advantages of electrostatic precipitators are that high
collection efficiencies can be achieved even with small particles and the
pressure drop across the precipitator is low. The latter characteristic
makes precipitators especially attractive when large gas volumes are
to be handled.
The high initial cost of electrostatic precipitators is a disadvantage.
However, when determining air pollution control costs, total costs over
a period of years should be determined.
Two factors that limit application of electrostatic precipitators are:
(1) high resistivity dust results in limitation of the operating voltage and
current so that the resulting precipitation rate parameter is low. This
necessitates the use of an excessively large precipitator, alterations in
inlet gas temperature, or additions of chemical conditioning agents to
alter dust resistivity. These constitute additional costs and can make
other dust control methods more attractive. (2) Very fine particles do
not acquire a charge sufficient for good precipitation. This again results
in low precipitation rate parameters requiring larger precipitators with
higher costs.
Two primary areas for increased precipitator applications are in
the control of emissions from municipal incinerators and in control of
foundry cupola emissions. The use of precipitators on municipal incin-
erators is relatively common practice in Europe. Within the past year,
several installations have been made in this country.
The control of cupola emissions by electrostatic precipitators is
potentially a promising area. However, details of accommodating the
highly variable emission rates and character of the emissions must be
resolved.
A summary of the use of the various types of dust control equipment
in each of the areas identified by SIC classification for the 1966-1967
period is given in Part II of the manual. These represent potential use
areas for electrostatic precipitators.
-------
-54-
Research and Development Recommendations
Based on the precipitator systems study, a research and development
program has been prepared to provide additional fundamental information
on precipitator technology and applied research aimed at resolving some of
the problems found that limit precipitator performance and hence their
application to various dust control problems.
The program is divided into two broad areas covering a plan for
basic research and a plan for applied research. The applied research
program is further divided into research directed toward collection of
dusts with high electrical resistivity, collection of fine particles or fume,
and general application problems.
In addition to these defined areas, provision was made in the program
for investigation of new precipitator concepts and for studies of electro-
static augmentation of other types of dust control equipment.
In addition to the technical research areas, provision was made for
a critical review and analysis of the methods of specifying precipitators
for specific applications and for the most appropriate type of contract for
the purchase of precipitators. The latter should encompass the negotiated
type of contract, which has been used by some electric utility companies
and the penalty-bonus contracts that have recently been employed. Finally,
performance guarantees should be reviewed, especially in regards to
the methods of acceptance testing and the desirability of delayed acceptance
tests.
Table 3 lists the suggested research plan. The research topics
are listed for a five year research effort. Man hours are shown for a
program of approximately five million dollars over the five year period.
The contract calls for two research plans, one for five million and the
other for a seven million dollar expenditure over the five year period.
For the seven million dollar program, it is recommended that the
additional funds be used during the fourth and fifth years for demonstrations
of the effectiveness of the research programs in full size plant operations.
This may require design and building of precipitators or at least sub-
stantial modifications, so that funds will be required for purchase of
components and for construction and modification.
SOUTHERN RESEARCH INSTITUTE
Ja
I]
I!
?a? -gasp
lil !§ff
ih l!
if
*• "^i •=£
?: s"l !is
riH'11
m II? f£f
!?5 gal £;l
c=-a "on trE
|la II
S E3 =!
fl^ I
afl!f
ifft
pi!
35s
s?5
fH
3 is
si
III
S H:
ID
OS
3.
2
B*
3
Ml
i!
5 3 S 2
2.1- a
-------
Table 3 (continued)
Year 2
Study effect! of odor
control method* of design
and operation of preclpl-
tatora for pulp and paper
recovery boiler*
Application Problem*
Investigate method* of com*
batting "snowing" fron
recovery boiler preclpltators
Man-Jfears.
Study of hifb reiiatlvity
problems In alnter machine
precipttators
Review and analysis of the 9
procedures for spacifLcutlon
of electrostatic prcctpltator*.
Review of rontractlng proce-
dures and of methode of
establishing performance
guarantees - applicability of
penalty-bonus clause
Study economics of du*t control
on foundry cupolu and review
potential for electroetatlc pre-
ctpltatore in this application
Analysis of method* of com-
batting the hlfh reetottvlty
probleme aeaoclated with dry
proceea Ulna
Review and analyaU of methodK
of bid analyite
Analysis of the Impact of
control of gaseous pollu-
tants on application of
electrostatic preciplUton
Review method* for
improved prec ipltator
operation on Intermittent
metallurgical processes.
Potential Tor rapping during
periods of heavy dust loading
Analysis of fire and explosion 2
hatarda associated with elec-
trostatic preclpitator* opera-
ting on cement kilns and method*
of combatting these problem*
Comparison of the 4
the ecooonitca of
combination low
energy «crubber*
and «l*ctroeUtic
prec ipltator* with
high energy to rubber*
for both gae*oua and
parttculate control
Review of Bulfate
buildup on cement
kiln precipttator*
Review of problem• 3
of ash handling from
hopper*
Table 3 (continued)
APPLIED RESEARCH PLAN
Collection of Ht^h Realattvlty Oust
Man- Y«i
Inv*>t<(ate potential for corroelon
and fouling of air heater at low flu*
ga* temperatures
Year 2
Review of high temperature
fly a*h piwcipltator opera-
tion.- Identify problem area*
with extttlng plant*
Review economic* of low
temperature operation
Including increased boiler
efficiency and Increaaed
maintenance coat*. If any
Analyai* of operation of
fly aah conditioning plant*
Year!
Investigate potential problem*
with high temperature fly aeh
precipltator* - potential for
fly aah suit*ring, effect* of
ga* velocity on reentratninent
at high temperature*, ga*
vl*co*lty - change* In V-l
characteristic* of ga*, etc.
Review of *iperlence* with
low temperature operation
of precipltator*
Review of potential for con-
ditioning of high resistivity
dust* other than fly aab
Study prec Ipltator design
for handling high resistivity
dust*
lection of Small particle* and Fume*
Review economic* of
high temperature fly
a*h prtcipltator operation
Pilot scale • turtle* with
low temperature precipl-
tators
Pilot and full
scale testa of
condition ing plants
Pilot studies pf mod if Led
preeip tutor
Analysis of method* for
increasing charge on small
particles and fumes, or for
agglomerating particle*
Pilot plant studies of pre- 4
eipltator designed for fine
particle collection
Field trial* of
•mall particle
-------
-58-
The following discussion is a further amplification of the suggested
research topics listed in Table 3.
Basic Research Plan
The basic research plan is intended to define limitations of precipi-
tator performance and to obtain information that can provide a more
rational basis for design. The following topics are suggested.
Refine system model. A simplified system model was developed
under Contract CPA 70-166. This model was a first step in showing
interrelationships between the variables that influence precipitator per-
formance. However, several important relationships were not included.
These include (1) diffusion charging, (2) effects of gas flow distribution,
(3) limitations due to sparking which defines current-voltage relationships,
and (4) reentrainment effect. The program is designed to include the
effect of these variables in the overall model. The second year's effort
would Include verification of the model on pilot and full scale units. The
third year's effort would include upgrading of the model to more accurately
predict performance of field precipitators. The data for upgrading would
be provided from carefully conducted field tests and from quantitative
relationships developed under other research programs. The fourth
year's effort would involve optimization of precipitator design based on
studies with the mathematical model. Optimization might include vari-
ations in current-electric field, resistivity, and gas velocity to give the
best efficiency of collection. The fifth year's effort would be further
field tests to verify the optimization conditions.
The objectives of the model study would be (1) to determine whether
a better basis for design can be developed based on theory as opposed to
the empirical methods currently used, (2) to provide a method for ana-
lyzing the problems of a particular installation where poor performance
is being experienced, and (3) to provide a convenient method for ana
lyzing the interaction of precipitator, dust, and gas variables, and to
permit optimization of these variables in a systematic approach.
Role of turbulence and electric wind. Controversy exists as to the
effect of turbulence and the electric wind on precipitator performance.
Since, in a turbulent flow condition, collection occurs only within the
boundary zone, the transport of the dust to the boundary zone must be by
SOUTHERN RESEARCH INSTITUTE
-59-
means of gas turbulence or particle diffusion. Collection efficiency is
related to the ratio of the dust concentration within the boundary layer
and that in the inter electrode space. Thus, if there is a variation in the
dust concentration brought about by gas turbulence, or other factors, the
efficiency equations would be altered.
The magnitude of the electric wind has been determined by a point
plane apparatus with large current flows. Also, studies have been made
of the potential of the corona process to serve as a means for moving air.
However, there is still considerable uncertainty regarding the magnitude
of the electric wind in commercial precipitators and its influence on
collection efficiency.
Fundamental studies of spark propagation. The voltage and current
relationships in a practical precipitator are limited by sparkover. The
thickness of the dust layer, the resistivity of the dust layer, and the
magnitude of the electric field at the dust surface influence the voltage at
which sparking occurs. In analysis of precipitator performance, quanti-
tative relationships are needed to establish the voltage and current at
which a precipitator will operate for a given set of dust conditions. Fur-
ther studies are also needed to quantitatively establish conditions for
spark quenching. Effects of power supply size, sectionalization, etc.,
are needed to give a firm basis for design of energization equipment.
Equivalent circuits of precipitators and power supplies should be
developed to permit rapid evaluation of the effects of charging parameters
on sparking and spark quenching. Finally, studies are proposed to show
the effects of the rate of voltage rise on the peak and average voltages
that can be attained for a given spark rate. The latter is significant in
terms of the interest in pulsed power supply operation.
Reentrainment. Studies of rapping requirements have so far been
limited to various types of dust with power on and power off. Since the
force holding dust on the collection plate is so much a function of the
electric field strength, rapping requirements should be related to the
basic holding force. Too intense rapping can result in significant reen-
trainment, and since the holding force is a function of resistivity, there
would be a tendency for low resistivity dust to be easily reentrained by
too severe rapping.
Also, little work has been done on relating rapping requirements to
the condition of the dust. Dusts can vary from light fluffy conditions to
-------
-60-
wet sticky conditions as a result of gas temperature and composition.
Wet sticky dust can be difficult to remove by normal rapping techniques.
Little attention has been given so far to the collection of medium-
to-low resistivity dusts as a function of gas velocity. There is considerable
evidence to indicate that low dust resistivity can result in excessive reen-
trainment of the dust layer, as well as higher losses during rapping.
Quantitative data are needed to relate reentrainment to dust resistivity so
that conditions of optimum resistivity and gas velocity can be determined.
Effect of dust layer on corona generation. The presence of deposits
on the discharge wires appears to have varying degrees of influence on
precipitator performance. In some instances, corona current suppression
occurs, whereas in other instances, normal precipitation appears to occur
even with reasonably large buildup on the discharge wire. The effect
appears to be related to the tightness of the deposit, resistivity of the
deposit, and evenness of the deposit.
High temperature corona. Use of electrostatic precipitators for
cleaning of gases at temperatures above around 100° F is a unique appli-
oiti-3B. Because of increaset!-thermionic emission from the discharge
electrode at elevated temperatures, sparking occurs in negative corona
at reduced voltage. Further studies are needed in the definition of the
most desirable conditions for corona generation at temperatures above
around 1500°F and pressures around 100-200 psi.
Collection of small particles. Research should be directed toward
the improved collection of small particles by electrostatic precipitators.
Because of the inherently lower charge that can be applied to a small
particle and the resulting lower precipitation rate parameter, techniques
for enhancing the charge on small particles should be investigated.
The small particle problem is especially important in the collection
of metallurgical fume. Charging particles of less than around 0. 3M diameter
occur primarily by diffusion charging. Methods of enhancing the charge
by this mechanism should be studied. It has been shown that the effect of
electric field on diffusion charging is to increase the charging rate. Also,
increased ion densities would increase the charging rate. Studies should
include methods for minimizing space charge suppression of the corona
and for agglomerating the particles.
High resistivity problems. High dust resistivity limits the current
and voltage at which precipitators can operate and reduces the efficiency
of collection. Since many industrial dusts are in the high resistivity
SOUTHERN RESEARCH INSTITUTE
-61-
region, techniques must be developed for handling these dusts. These
techniques include modifications to the basic precipitator, use of larger
collection surface areas or modification of the dust resistivity. Dust
resistivity can be modified by increased gas temperature to enhance
volume conductivity, reduced gas temperature to enhance surface con-
duction or by increased moisture or secondary conditioning agents to
enhance surface conduction.
Low temperature operation. Operation of electrostatic precipitators
at gas temperatures below around 260-270°F is an alternative method of
reducing resistivity. Experience has shown that efficiencies can be sub-
stantially improved by reducing the flue gas temperature when a precipi-
tator is limited by high dust resistivity. High dust resistivity is normally
associated with low sulfur coals in the case of fly ash precipitators.
There is concern that low gas temperatures can result in excessive
corrosion of the air heaters. However, many installations are functioning
at reduced gas temperatures even with high sulfur coal.
Further research is needed to define the conditions under which
corrosion becomes excessive, especially when burning low sulfur coal.
Problems of control of the temperature to minimize excessive tempera-
ture excursion should be identified.
High temperature precipitation. Operation of fly ash precipitators
ahead of the air heater is a potential method of reducing fly ash resistivity.
Experience with electrostatic precipitators in other applications in which
gas temperatures are in the range of 500-600°F would indicate that the
basic precipitator can function without difficulty in this temperature range.
Studies of the economics of high temperature operation should be made to
determine the comparative costs of this alternative to the high resistivity
problem. Analysis of operating problems with high temperature precipi-
tators should be made to determine the possibility of fusing of the fly ash
or other factors that can influence the choice of this method of dust control.
Dust conditioning. The role of conditioning agents in influencing
resistivity of the dust needs further study. Where secondary conditioning
agents are effective, as in combustion processes, the relationships
between fuel properties and the amount of conditioning agents should be
resolved. Also, the mechanism by which the conditioning agents act to
influence resistivity should be explored. The influence of dust composition
on the action of the conditioning agent should be investigated.
Studies should also be made of the influence of artificial conditioning
agents on the resistivity of various types of dusts. The majority of work
-------
-62-
on conditioning has been on fly ash from electric power boilers and petroleum
catalytic crackers. The application to other types of dusts should be
explored.
Precipitator modifications to handle high resistivity dusts. Studies
should be made of the voltage-current requirements in a precipitator to
determine the optimum conditions for handling high resistivity dusts.
These studies should relate charging time and electric field requirements
for optimum performance.
Application problems. Several problems have been noted which relate
to the application of electrostatic precipitators in specific areas and to the
overall acceptance of precipitators. Suggested research on these problems
are grouped together in the suggested research plan.
Pulp and paper industry
Application of precipitators in the pulp and paper industry has been
in the control of emission from black liquor recovery boilers. Recent
interest in the control of odor emission from paper mills has changed
the recovery process to eliminate the contact between black liquor and
flue gas. Effects of these changes on precipitator operation is an area
that needs further study and aeTBrttton.
A problem with recovery boiler precipitators has been the periodic
release of light fluffy particles from the precipitator. The effect is
termed "snowing" and creates problems with corrosion in the area
immediately surrounding the plant. Methods are needed to prevent the
occurrence of this problem.
Electric power generation
Collection of fly ash from electric power generation plants is the
largest single application of electrostatic precipitators. Current activity
in the use of wet scrubbers to control SO2 emissions, methods of removing
sulfur from the fuels, and control of NOX emissions can influence the
application of precipitators in this important area. Since scrubbers are
currently being considered as a means of SO2 control, the question of the
effectiveness of using high energy scrubbers for removal of particulate
as opposed to the use of precipitators with low energy scrubbers should
be resolved.
SOUTHERN RESEARCH INSTITUTE
-63-
Metallurgical industry
Electrostatic precipitators have historically been used for collection
of dust emissions for sinter machines used in the processing of fines from
the iron and steelmaking operation. However, use of high percentages of
lime has resulted in high resistivity dust which has greatly reduced the
effectiveness of precipitators for this application. Further research is
suggested to find means for conditioning the dust or other ways of
combatting the problem.
Use of precipitators for collection of foundry cupola dusts has been
a promising new area of application. Review of the cost factors involved
in the collection of foundry cupola dusts with various dust control methods
should be made. Potential for use of precipitators should be explored,
especially if methods of overcoming the small particle and high resistivity
problems can be developed.
Some metallurgical problems are intermittent, which provides the
opportunity for more intense rapping during the off periods. This offers
trie'promise of starting with clean plates with perhaps no plate rapping
during peak dust load periods. The technique may improve overall
efficiency of operation.
Cement industry
Use of electrostatic precipitators for control of emissions from
cement kilns is limited by (1) the high resistivity of the dust from dry
process kilns, (2) history of fires and explosions in cement kiln precipi-
tators, and (3) buildup of sulfate on the electrodes at certain operating
conditions. The high resistivity problem has been a factor in the decline
in use of precipitators or dry process kilns in this country. Successful
methods of combatting this problem would enhance the use of electro-
static precipitators in this application.
Fires occur in cement kilns due to upsets in the kiln which result
in combustible gases being released from the kiln. The condition can
occur during start-up or during operation when a combustible mixture
is formed by introduction of air following the kiln.
The problem with fire in the dust collection equipment is not
limited to precipitators, but is made more severe by the presence of
-------
-64-
ignition sources resulting from sparking within the precipitator. -
The problems with fire in the precipitator are largely due to operating
procedures, but their occurrence limits precipitator application in this
area.
The buildup of sulfate deposits on the discharge wire and collection
electrodes of cement kiln precipitators is apparently the result of operation
within certain critical temperature regions. However, difficulty has been
noted with precipitator operation due to the sulfate buildup. Resolution of
the problem would enhance operation of cement kiln precipitators and
increase their use in this area.
Many precipitator operating problems are due to improper ash
removal. Clogging of the ash removal system can cause the precipitator
to be shut down for short periods. Failure of the ash removal system can
also result in excessive sparking or arcing of the precipitator. Formation
of a fused fly ash mass can result which causes further ash removal
problems. A condition called "concreting" of the ash can result if a damp
ash is encountered.
A better definition of the ash handling problems, relationships for
sizing ash handling systems, and operating conditions for optimum per-
formance are required.
Review of Specification and Contracting Practice
The study of the application of precipitators indicates the desirability
of a review of procedures for specifying, bid evaluation, and contracting
procedures. Such a review would be of benefit to both precipitator manu-
facturer and user. Specifications are often too loose so that there is no
firm basis for fair competitive bidding. When low bids are accepted, the
precipitator design is often too marginal to perform adequately except
under ideal conditions.
Current air pollution legislation is imposing more strict limitations
on particulate loadings. This has resulted in efforts on the part of the
users to impose more severe penalties for failure to meet design specifi-
cations. The penalty-bonus type of contract has been used in an effort
to insure adequate performance. However, experience with this type of
SOUTHERN RESEARCH INSTITUTE
-65-
contract is limited, and the desirability of this type contract has not been
established.
The main goal of this proposed study is to arrive at the most
desirable method of specifying and contracting that will give adequate
assurance of performance and minimize the tendency to provide a pre-
cipitator whose performance is marginal.
Performance testing practice to determine whether a particular
installation meets contract guarantees also needs review. Desirability of
two-step testing, one following completion of the installation and a delayed
test, should be investigated. This practice is followed in other countries
and appears to have merit in insuring long-term reliability.
Submitted by:
"Albert Oglesby,-'Jr., pfrector
Engineering Research
A378-2291-XDC
(75: lr: 1: 30) mat
------- |